Chapter 1
Overview of Brachytherapy Physics Ravinder Nath, Ph.D. Department of Therapeutic Radiology Yale University School of Medicine New Haven, Connecticut Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Different Types of Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Intracavitary Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Interstitial Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Intraluminal Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Surface Molds and Plaques for Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Different Types of Brachytherapy Loading Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Manual “Hot” Loading Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Manual Afterloading Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Remote Afterloading Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Image-Guided Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Different Types of Brachytherapy Based on Dose Rate and Duration of Brachytherapy . . . . . . 4 Different Durations of Brachytherapy Typically Used . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Permanent Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Temporary Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Dose Rates Used in Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Introduction Brachytherapy is a special procedure in therapeutic radiology that utilizes the irradiation of a target with radioactive sources placed at short distances from the target. Typically, the sources are implanted in the target tissue directly (interstitial brachytherapy) or are placed at distances of the order of a few millimeters from the target tissue, in body cavities such as the uterus, mouth, etc. (intracavitary brachytherapy), or externally on structures such as the eye, skin, etc., (surface plaques or molds). Brachytherapy generates highly conformal dose distributions in a target volume because radioactive seeds are implanted directly within or in the vicinity of the target tissue. For example, in a typical interstitial brachytherapy implant, 50 to 100 radioactive seeds, each about the size of a rice grain, are implanted in the tumor using image-guided implantation techniques such as ultrasound, computed tomography (CT), or fluoroscopy, which allow the physician to place radioactive seeds precisely at desired locations with minimal invasiveness. For these applications, low-energy, photon-emitting radionuclides such as 125I (27 keV) and 103Pd (21 keV) are preferred because these sources provide adequate coverage of tumor when used in a grid of about 1 cm spacing and produce minimal exposures to distant organs in the patient, to the hospital personnel performing the procedure, and to the family members and friends of the patient after he/she is released from the hospital with the radioactive seeds in place. Most brachytherapy procedures today are performed in one-day surgery suites without the need for hospitalization. These factors and the depth dose characteristics make brachytherapy a very cost effective and patient friendly procedure compared to 3-D conformal radiotherapy (3DCRT) or intensity-modulated radiation therapy (IMRT), which also produce highly conformal dose distributions. A key advantage of 3DCRT or IMRT over brachytherapy is that it is noninvasive. However, both 3DCRT and IMRT are very sensitive to patient localization and setup errors because of high dose gradients at the periphery of the target volume. Therefore, the target must be placed at the right position with a precision of about a millimeter relative to the linear accelerator (linac) daily over a course of 5 to 6 weeks of 3DCRT and IMRT, and the clinical target volume
1
2
Ravinder Nath
expanded by 5 to 10 mm for intrafraction movement. In contrast, brachytherapy requires a single visit to a one-day surgery clinic or 3 to 5 visits in an outpatient clinic for high dose-rate (HDR) intracavitary brachytherapy. Unlike 3DCRT and IMRT, brachytherapy is far more forgiving of localization and target motion errors because the implanted sources of radiation in an interstitial implant move with the target. Thus, brachytherapy solves a critical problem of 3DCRT and IMRT, which is that they have a potential to “miss the target very precisely” unless they are implemented with extreme precision. This high-precision requirement makes 3DCRT and IMRT very expensive, labor intensive, and technically complex. For these practical reasons and other important radiobiological reasons related to continuous low doserate (LDR) irradiation, brachytherapy remains a valuable treatment modality for selected cancers despite the current trend of widespread adoption of 3DCRT and IMRT in radiotherapy. In this chapter, an overview of the rich spectrum of various brachytherapy procedures is presented so that the students have an appreciation of the overall field. Most brachytherapy uses gamma-emitting radionuclides. Many use radionuclides such as 137Cs and 192 Ir, which emit high-energy gammas, that penetrate deeply and also require heavy shielding for radiation protection of the personnel and the patient’s family. Whenever possible, brachytherapy with low-energy gamma emitters, like 125I and 103Pd is preferred because it requires minimal shielding for radiation protection. On the other hand, beta-emitting sources are commonly used as unsealed sources for systemic brachytherapy similar to nuclear medicine diagnostic procedures. Beta particles are absorbed within a few millimeters in tissues. Hence, beta emitters require minimal radiation shielding unless there is spillage of radioactive material in a liquid form and subsequent contamination of work surfaces. Beta emitters commonly used for brachytherapy include 32P, 106Ru, 90Sr, and 90Y. In the last decade, the betaemitting sources also generated a great deal of interest in intravascular brachytherapy for prevention of restenosis after angioplasty because the distances of interest for treatment are about 2 mm, typical wall thickness of a coronary blood vessel. In this application, the sealed sources of 32P and 90Sr were widely used and 90Sr is still being used today. The only neutron emitter, which has been used clinically, is 252Cf. The advantage of neutron emitters is that their interactions with tissues produce heavy charged particles that have much higher linear energy transfer (LET) than the secondary electrons produced by gammas or betas and because of the higher LET, neutrons are theoretically more effective against hypoxic tumors. Neutrons are, however, notoriously difficult to shield and present a far greater challenge in reducing the radiation hazards to personnel. For these reasons, neutrons are rarely used except in a few major medical research centers.
Different Types of Brachytherapy Intracavitary Brachytherapy Intracavitary techniques involve placing radioactive sources into custom-designed applicators, which are placed in body cavities. This is commonly used for the treatment of gynecological tumors where the radioactive material can be placed in the uterine cavity and vagina. This is also the most common brachytherapy procedure done worldwide.
Interstitial Brachytherapy Interstitial brachytherapy involves placing the sealed radioactive sources within tissues. Thus, most nongynecological implants are performed using interstitial techniques. This is the most common brachytherapy procedure performed in the United States. The most common application is permanent implantation for treatment of prostate cancer. Less common are interstitial implants such as the Syed implants for gynecological tumors, which are more challenging to execute because large variations in
1–Overview of Brachytherapy Physics
3
size, shape, and location prevent them from being visualized and implanted with a uniformly applicable algorithm, like prostates.
Intraluminal Brachytherapy This is really a subclass of intracavitary brachytherapy in which the radioactive sources are inserted in the lumen of a vessel such as the blood vessel, bronchus, esophagus, or bile duct.
Surface Molds and Plaques for Brachytherapy In this technique, the radioactive sources are placed in custom-designed molds or plaques, which are then placed on the surface of the target tissue rather than being placed inside the target tissue. These techniques are not very commonly used.
Different Types of Brachytherapy Loading Systems Over the years, many techniques have been developed for handling and loading radioactive materials in order to reduce the radiation exposure of the personnel. These techniques are based upon the three cardinal principles of radiation protection, reducing time of exposure, increasing the distance from the radioactive material, and inserting shielding between the user and the radioactive material. Most of these techniques fall in one of the following categories.
Manual “Hot” Loading Techniques When brachytherapy was first introduced over a century ago, the radiotherapy sources were manually introduced into the tumor, hence subjecting the physicians and other operating room personnel to unwanted radiation exposure and the subsequent adverse effects of radiation. Direct “hot” loading is almost never used today because of its associated radiation hazards that can now be reduced by the adoption of alternative techniques.
Manual Afterloading Techniques Since 1950s or so, most brachytherapy procedures have been performed using afterloading techniques whereby hollow needles, catheters, or applicators are first inserted into the target volume. The applicators are usually inserted in the operating room. Once the position of the applicators is confirmed, the radioactive material is introduced manually into the applicator. This procedure is usually performed in the patient’s room rather than the operating room. This afterloading procedure improves placement accuracy because the clinicians can take their time in optimizing the positioning of the applicator without incurring excessive radiation exposure.
Remote Afterloading Techniques Although the manual afterloading techniques mentioned above reduce radiation exposure for the physician and the operating room staff, the nursing staff and other caregivers as well as visitors continue to be exposed to a small amount of radiation exposure. Even this small exposure can be virtually eliminated by the use of remote-controlled afterloading, in which the radioactive material is loaded into and out of the applicator by microprocessor controls positioned and operated remotely by the caregivers in an adja-
4
Ravinder Nath
cent room. Advances in the developments of remote afterloading techniques, especially remote HDR brachytherapy, have changed profoundly the clinical practice of brachytherapy, so that HDR has become a very popular form of brachytherapy, especially in developed countries.
Image-Guided Techniques Before the use of computerized treatment planning systems became popular, clinicians relied on generalized systems of rules (i.e., the Paterson-Parker System, the Quimby system, the Memorial system, the Paris system, the Manchester system, etc.) for pre-implant or intraoperative planning. All of these early systems depended on tables and nomograms to select number, location, and activities of radioactive sources in an implant to achieve desired dose coverage of an estimated target volume. While traditional systems are the culmination of years of clinical experience and still provide a good point of comparison with modern computerized treatment planning, the description in traditional systems of the actual dose distribution and the underlying anatomy is at best approximate and qualitative. In the last decade, two major advances have been made in brachytherapy treatment planning. One is the increased use of imagebased treatment planning and dose evaluation and the other is the development of computer-assisted dose optimization algorithms for brachytherapy. Developments in three-dimensional tomographic image-based source localization techniques have made it possible to calculate and display dose distributions directly on top of the tomographic images of involved anatomy. This enables clinical decisions to be made based on the visualization of dose distributions with respect to patient anatomy instead of dose tables. It has almost become routine in the last several years that the target volume and normal organs be localized from CT, magnetic resonance imaging (MRI), or ultrasound (US) images. Pre-implant planning can now be carried out by inspecting the instantaneous dose distribution changes resulting from different source placements. In addition, some modern analytic tools including dose volume histograms, normal tissue complication probabilities, and tumor control probabilities can be used to estimate the quality of an implant more quantitatively.
Different Types of Brachytherapy Based on Dose Rate and Duration of Brachytherapy Different Durations of Brachytherapy Typically Used Permanent Implants In a permanent implant, the radioactive sources are permanently implanted into the tumor, the patient is released from the hospital with radioactive material in him or her, and radioactivity is allowed to decay within the patient. Hence, the dose, dose rate, or the dose distribution cannot be changed after the initial insertion. Many of these procedures are relatively simple, and can be performed in an outpatient clinic and one-day surgery units. Other clinical advantages of the permanent implant are that, in deep-seated tumors, it may be safer because of the lower risk of infection and that a second operation for its removal is not required. Permanent implants are performed with relatively short half-life radioisotopes like 125I, 103 Pd, or 198Au so that the radioactivity decays to a safe level within a few weeks or months and does not present the risk of radiation-induced carcinogenesis due to long-term chronic radiation exposure. Temporary Implants In a removable implant, the radioactive material is temporarily implanted into or close to the tumor and is removed once the desired radiation dose has been delivered. When the treatment time is short (less than 20 minutes), there is better control of the total dose and the dose distribution. Removable implants with
1–Overview of Brachytherapy Physics
5
short treatment times require more precision compared to permanent implants. On the other hand, lower precision and accuracy can be tolerated in LDR temporary implants, because in such implants dwell locations and times may be adjusted to correct for poor needle placement. In intracavitary implants, if the dosimetry is inadequate, the dose for that fraction may be reduced and positioning or packing adjusted for subsequent fractions. If radiobiologically equivalent doses for each modality have been adopted, the decision to use either temporary or permanent implantation depends upon the judgment of the radiation oncologist and his skill and experience with a given system.
Dose Rates Used in Brachytherapy Brachytherapy can be delivered at enormously different dose rates over a very wide range of treatment times varying from minutes to months. Brachytherapy dose rates have been divided into low, medium, and high dose rates by International Commission for Radiation Units and Measurements (ICRU) Report No. 38 as follows: 1. Low Dose Rate (LDR): 0.4 to 2.0 Gy per Hour This is the traditional dose rate for permanent and manually afterloaded brachytherapy. Temporary implant LDR techniques result in typical treatment times of 3 to 5 days. This requires hospitalization of the patient. These implants are generally manually afterloaded, although LDR remote afterloaders are also available, but not commonly used. Most of the long-term clinical experience with brachytherapy is with LDR. Many of the advantages of brachytherapy are attributed to the radiobiology of continuous LDR irradiation in LDR brachytherapy. 2. Medium Dose Rate (MDR): 2 to 12 Gy per Hour This dose rate, also called “intermediate dose rate,” is rarely used since it gives excessive exposure if such an implant is manually loaded, and this dose rate does not have the advantages of outpatient brachytherapy afforded by the HDR technique. Pulsed dose rate (PDR) brachytherapy afterloaders were developed in this dose rate realm to replicate the LDR experience in terms of total treatment duration but with the source exposed in pulses for only 5 to 10 minutes per hour. 3. High-Dose-Rate (HDR): More than 12 Gy per Hour HDR brachytherapy utilizes very high activity sources, typically a 10 Ci 192Ir source, which produces a very intense radiation field around the source. Since HDR brachytherapy is associated with high radiation exposure rates, it is only used in well-shielded bunkers. Depending on distance and usage factors, 1 to 2 feet of concrete shielding or its equivalent in other materials is required. Treatment is delivered by remote-control techniques rather than manual loading. The usual dose rate in the commercially available HDR brachytherapy systems is about 100 to 300 Gy per hour, allowing the treatments to be given in only a few minutes on an outpatient basis. The introduction of HDR remote brachytherapy, with its advantages of thorough radiation protection and outpatient treatments, has led to a resurgence of interest in brachytherapy.
Chapter 2
Radiobiology: A Briefing for the Brachytherapist Marco Zaider, Ph.D. Department of Medical Physics Memorial Sloan-Kettering Cancer Center New York, New York Who Needs It? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 The Nuts and Bolts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Kinds of Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Dose-Response Curves and Their Meaning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Hits: What Are They? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Temporal Aspects (I) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Enhancement Ratios: RBE and OER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Cellular Proliferation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Mechanisms of Radiation Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Microdosimetry in a Nutshell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 A Microdosimetric Account of the Linear-Quadratic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Temporal Aspects (II) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Equivalent Treatments (Painless Version) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Equivalent Treatments (II) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Radiation Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 The Real World . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Further Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Who Needs It? The present function of radiation biology in treatment planning is threefold: (1) to provide information on biologically equivalent temporal patterns of dose delivery; (2) to quantify the effect of radiation quality; and (3) to guide the implementation of biologically based optimization algorithms. Of these three functions the first one is most often invoked, particularly in brachytherapy where low and high dose rate (LDR and HDR) regimens are commonly employed. The issue of radiation quality (to be defined more precisely below) is mostly relevant to brachytherapy where low-energy electrons (“low” compared to standard megavoltage radiation) are known to have enhanced biological effectiveness. The topic of biological optimization, although relatively new, is expected to dominate treatment planning in the years to come and, in fact, to replace the current dose-prescription modus operandi. The goal of this chapter is to familiarize the medical physicist with the basic language and some of the fundamental theories of quantitative radiobiology and to offer the brachytherapy practitioner the tools necessary for its implementation in daily activities. The first section explains the basic dose-response relationships, relative biological effectiveness (RBE), and enhancement factors, in particular oxygen enhancement ratio (OER). Factors that impinge on the dose-survival response, e.g., the effects of cell proliferation or radiation quality are briefly introduced. The key tools for understanding the mechanisms of radiation action and their connection to the physical description of the radiation field are provided by the discipline of microdosimetry. Elements of microdosimetry are introduced with the purpose of explaining the well-accepted linear-quadratic (LQ) formalism. An account of elementary [the effect of dose rate, biological equivalent dose (BED)] and advanced [tumor control probability (TCP), normal-tissue complication probability (NTCP)] applications concludes this chapter.
7
8
Marco Zaider
In line with the adage “teach a man to fish and he will never go hungry,” the author could think of no better introduction to the application of these topics than letting the reader deduce a number of familiar results as applied to brachytherapy. Many important topics, as well as specific references, are absent. Because of limitations of space and time a choice was necessary and I hope that references given under “Further Reading” will stimulate the reader to complete his knowledge.
The Nuts and Bolts Kinds of Effects It is useful to recognize that radiation effects may be classified into two broad classes: stochastic and non-stochastic (or deterministic). Stochastic effects are of the all-or-none variety; for instance, cell killing. Quite obviously, radiation— or any other agent—has no effect on the nature of this sort of effect. What radiation does instead is to change the probability of producing the effect. For instance, a dose-response relationship connects probability of effect to absorbed dose. Tumor control probability (TCP), which measures the likelihood that a particular treatment dose leaves all cancer cells sterilized, illustrates this state of affairs. In contrast, deterministic effects are such that their severity (or magnitude) can be changed as a function of dosage. Cataract-induction by ionizing radiation is one such example. Effects of radiation on normal tissues are mostly of the deterministic kind—as the ability of that particular organ to function is progressively altered with increasing dose. However, one can always transform a non-stochastic effect into a stochastic one by setting up a threshold that divides the deterministic effect in question into, say, acceptable and unacceptable domains. Normal tissue complication probability (NTCP) is the result of dichotomizing the effect as described. In the following we shall denote probability of effect by E(D). The absence of effect (colloquially known as survival probability) is S(D) = 1-E(D).
Dose-Response Curves and Their Meaning Progress in molecular biology notwithstanding, the dose-response relationship remains the principal tool available to the radiation biologist for testing assumptions concerning the mechanisms responsible for radiation action. Examples of questions commonly asked are: (a) Is the effect a result of single-lesion action or a consequence of the accumulation of multiple sublesions? (b) Does the temporal pattern of dose delivery matter? (c) Does the response depend on the position of the cell in the cell cycle? (d) Is radiation-induced damage repairable? Consider the two dose-response curves shown in Figure 1 (note the semilogarithmic representation). The curve in Figure 1a is a purely exponential function, exp(−αD) where α > 0; it describes an outcome where equal increments in dose result in equal fractions of the exposed population acquiring the effect. The interpretation of this kind of response is that surviving cells1 have no memory of previous exposure
1
For convenience we shall refer throughout this chapter to “cells” as the biological units under observation. However, the statement applies equally well to any kind of biological object, from DNA to a human being.
2–Radiobiology: A Briefing for the Brachytherapist
9
Figure 1a. An exponential dose response curve indicates a single-hit mechanism of radiation action.
Figure 1b. A concave downward dose effect curve shows an accumulation of sublesions and represents a multihit mode of radiation action.
to radiation. Alternatively, one speaks of single-hit mechanism in the sense that, when exposed to radiation, the cell will either show the effect or remain unaffected. The curve in Figure 1b is concave downward and depicts the LQ equation: S ( D ) = e− α D − β D , α , β > 0 . 2
(1)
10
Marco Zaider
In this case sequential exposures to equal increments in dose result in an increasing fraction of affected cells. The obvious interpretation of this type of response is that cells exposed to radiation carry (and accumulate) damage not sufficient to produce the effect on its own but enough to sensitize the cell to further radiation exposure. The term sublesion will be used for this kind of damage, and lesion will denote the biological alteration responsible for the end-point under investigation. The radiation response illustrated in Figure 1b is known as multi-hit action. Finally, if—as a result of exposure to radiation—cells become less sensitive, one expects a doseresponse curve, which is concave upwards. Mathematically, the function that determines the behavior of the curves shown in Figure 1 is2:
h( D) = −
d log ( S ) dD
constant ( curve A ) = monotonicallly increasing ( curve B) monotonically decreasing .
(2)
Hits: What are They? We have spoken of single-hit and multi-hit action and at this time we need a precise definition of the term “hit.” Loosely speaking, a hit (microdosimetrists prefer the name event) indicates the traversal of the radiation-sensitive region of the system under observation by an ionizing particle (Figure 2). More precisely, a hit (or event) is the production of statistically correlated alterations in the sensitive site. Essentially, alterations result from local energy deposition via ionizations or excitations. The crucial feature here is that events are statistically independent.
Temporal Aspects (I) If cells respond to radiation according to the single-hit mechanism, the temporal distribution of hits is clearly a matter of indifference. However, in multi-hit mode and in the presence of repair mechanisms sublesions may be eliminated before the next hit arrives and thus dose rate (proportional to the number of hits per unit time) becomes relevant. As the dose rate is progressively lowered, the term responsible for multi-hit behavior [βD2 in equation (1)] must grow increasingly smaller; at the limit of very low dose rates the dose response curve becomes exponential (Figure 3). (A more precise definition of “low” dose rate will be given below; however, it should be apparent that this quantity depends on the characteristic sublesion repair time, which is of the order of one hour.)
Enhancement Ratios: RBE and OER The shape of the dose-response curve depends on many factors (radiation quality, oxygen concentration, position in the cell cycle, etc.). The enhancement ratio (ER) quantifies such differences as follows: let DA and DB be isoeffective doses corresponding, respectively, to dose-response curves A and B. Then: ER =
2
In epidemiology h is known as the hazard rate.
DA DB
.
(3)
2–Radiobiology: A Briefing for the Brachytherapist
11
Figure 2. The dots indicate the locations of energy deposition (ionizations or excitations) by a hypothetical charged-particle track in tissue. Also shown are two cells—one traversed by the main track and the other by a secondary electron (delta ray). The ensemble of energy deposition sites within the radiosensitive volume, which are associated with one track, defines a (microdosimetric) event.
Figure 3. As the dose rate decreases the quadratic term (βD2) becomes increasingly smaller. At very low dose rates only the linear term, αD, remains (dashed curve). Lower solid curve: acute exposure; upper dashed curve: protracted exposure.
12
Marco Zaider
The following are two important examples: Relative biological effectiveness (RBE) is a quantity used to compare the biological effects of two different radiations. If X denotes the reference radiation, the RBE of A is given by: RBE =
DX DA
.
(4)
Typically, high-energy electrons (or, equivalently, high-energy photons) are taken as reference radiation. Anoxic cells (poor in oxygen) are more resistant to radiation than aerobic (oxic) cells. This phenomenon is known as oxygen effect. To quantify the oxygen effect, one makes use of the oxygen enhancement ratio (OER) defined as: OER =
Danoxic Doxic
.
(5)
In general, the ER depends on the dose level [e.g.. DA in equation (4)]. An agent for which the ER does not depend on dose is known as a dose-modifying factor. (The oxygen is believed to be a dosemodifying factor.)
Exercises 1. Find an expression for RBE for two radiations characterized by (αA, β) and (αX, β): see equation (1). 2. Plot RBE as a function of DA for the result obtained above. 3. Find the limiting value of RBE when DA → ∞, and when DA → 0. 4. Show that when β = 0, RBE does not depend on dose. 5. Find a linear-quadratic expression for an agent, which is a dose-modifying factor.
Cellular Proliferation Stem cells (responsible for recovery of self-renewing tissues from radiation injury) and malignant cells (which account for tumor growth) have the capability to proliferate. Knowledge of the growth curve of these cell populations is of obvious importance in treatment planning. Additionally, one must be aware that radiation response depends drastically on the cell position in the cell cycle. We shall start by a brief review of this latter aspect. The cell cycle of actively dividing cells is conventionally divided into four stages: G1, S, G2, and M. G1 and G2 (known as first and second gaps) bracket the S phase (where DNA synthesis takes place) and the M phase—where the actual cell division occurs. Empirical evidence shows that (1) cells are most radiosensitive in M phase and most radioresistant in the latter part of S; (2) exposure to radiation results in a lengthening of the G2 phase (this is known as the G2 block); and (3) exposure to radiation decreases the division probability.
2–Radiobiology: A Briefing for the Brachytherapist
13
The kinetics of cell growth can be described in terms of three quantities: (1) the average duration of the cell cycle, Tc; (2) the growth fraction, GF, which is the fraction of cells actively proliferating; and (3) the cell loss factor, φ, which represents the rate of cell loss divided by the rate of cell production. With this, the cell number doubling time, Td, is given by: Td = Tc
1
ln 2
1 − ϕ ln (1 + GF )
.
(6)
In particular, if no cell loss occurs (φ=0), the doubling time is known as Tpot, which stands for potential doubling time. Tpot = Tc
ln 2 ln (1 + GF )
(7)
Another useful relation is:
ϕ = 1−
Tpot Td
.
(8)
For a typical human tumor Tc ≈ 1−2 d, φ ≈ 0.8−0.9, and GF ≈ 0.4−0.5. The expressions, equations (6 to 8), will be useful for calculating TCP.
Mechanisms of Radiation Action Microdosimetry in a Nutshell The energy deposited by ionizing radiation in any given site is a stochastic quantity and absorbed dose— the quantity commonly used to measure the “amount of radiation”—is in fact only the average value of its distribution. The response of an organism (cell, virus, etc.) to radiation depends on the energy locally deposited and, understandably, bears no relation to its expected value over an ensemble of such individuals. The belief that the average effect will depend on dose (average energy per unit mass) is not borne out by experiment either; it is well established that the same dose of radiation may result in vastly different effects, when delivered by different types of radiation. The reason for this state of affairs can be found in the observation that generally the biological effects of radiation depend on the (highly non-uniform) macroscopic pattern of energy deposition. Microdosimetry is the systematic study and quantification of the spatial and temporal distribution of energy in irradiated matter. The principal microdosimetric quantity is specific energy, z, defined as the quotient of ε by m, where ε is the energy imparted by ionizing radiation to matter of mass m: z=
ε
.
(9)
m
The unit of specific energy is the gray (Gy), equal to 1 J/kg. Specific energy z can be imparted to matter in single events or in multiple events; the respective distributions are labeled f1(z) and f(z). Note that f(z)dz—i.e., the probability that specific energy is in the interval dz about z—depends on the geometry and physical composition of the site in which z is determined. In contrast, absorbed dose, D, is defined
14
Marco Zaider
at a point in matter through the following limiting process. D = lim z
(10)
m→ 0 ρ = const
where is the first moment of f(z) and the limit is taken at constant mass density, ρ. When the site is sufficiently small (almost invariantly the sizes of interest are of the order of micrometer) D = z . Two averages of the single-event distribution, zF (the frequency average) and zD (the dose average) are defined as follows: ∞
zF = ∫ zf1dz ;
1
zD =
zF
0
∞
∫z
2
f1dz .
(11)
0
For a given dose, D, the average number of events traversing the site is n=
D
.
(12)
zF
Alternatively: ∞
n = ∑ k P ( k,; n )
(13)
k=0
where P(k;n)—the probability of exactly k events—is the Poisson distribution: P ( k; n ) = e n
nk . k!
(14)
Let fk(z) denote the distribution of z in exactly k events. It follows that: f0 ( z ) = δ ( z )
f1 ( z ) ≡ f1 ( z ) f2 ( z ) =
∞
∫ f ( z ') f ( z − z ') dz ' 1
1
0
(15)
fk ( z ) =
∞
∫ f ( z ') f ( z − z ') dz ' 1
k −1
0
and thus: ∞
f ( z ) = ∑ p ( k; n ) k=0
fk ( z ) .
(16)
2–Radiobiology: A Briefing for the Brachytherapist
15
The expression, equation (16), shows that—given the single-event distribution f1(z)—one can calculate f(z) for any dose, D (or, equivalently, n).
Exercises 1. Let Φ(ω,n) and Φ1(ω) be the Fourier transforms of f(z) and f1(z), respectively. Prove, using the convolution theorem that Φ (ω , n ) = exp ( − n ) exp [ nΦ1 (ω )] . 2. Let mk and Mk represent, respectively, the k-th moments of f1(z) and f(y). Show that: k −1
mk + 1 = M k + 1 / n − ∑ C jk M k − j m j + 1 . j=0
From the result obtained at Exercise 2, it follows that: M 1 = z = nzF M 2 = z 2 = z ( z + zD ) ,
(17)
where [from equation (11)] zF = m1 and zD = m2/m1. If D = , equation (17) becomes: z 2 = zD D + D 2 .
(18)
A Microdosimetric Account of the Linear-Quadratic Model The LQ expression for the survival probability, equation (1), is understood to stand for the probability that following exposure to dose D, no lesions were produced in the organism. If the actual number of lesions (ε, a stochastic quantity) is Poisson-distributed3, it follows that
ε ( D) = α D + β D2
(19)
is the average number of lesions. Moreover, the LQ dependency of the yield of lesions on dose is compatible with the assumption that each lesion is the product of two sublesions. To make this clear, consider the “prototype” lesion responsible for preventing cell proliferation4, namely the dicentric aberration (Figure 4). A dicentric is the result of two (unrepaired) chromosome breaks misjoining their split ends. A cell that has a dicentric in it at the time of mitosis will not be able to go through the normal process of division. The two single-chromosome breaks are the sublesions and the dicentric is the lesion. Each sublesion is the result of one particle (event, or hit) traversing a chromosome. The pair of sublesions needed to produce a dicentric can be generated in two different ways: (1) by a single particle traversing both chromosomes, or (2) by two (independent) particles. The yield of the former mechanism is proportional
3
The distribution of lesions is not Poissonian; however, for low-LET (linear energy transfer) radiation (energetic electrons or photons) the Poisson distribution is a fair approximation. 4 Radiation biologists refer to this end point as “cell killing” although the cell is not “killed” in any sense of the word.
16
Marco Zaider
Figure 4. A dicentric chromosomal aberration results when two adjacent single-chromosome breaks (sublesions) rejoin to produce a chromosome with two centromeres (a lethal lesion). In what is termed “intra-track” event, a single particle produces both breaks and the yield of lesions thus induced is proportional to the number of particles, or equivalently to dose (upper panel). In the other possible mechanism (inter-track event) two independent tracks produce each one chromosome break. The yield of dicentrics thus produced is proportional to D2 (lower panel).
to the number of particle traversals, that is to dose, D; while the latter will depend—for similar reasons— quadratically on dose. The result is equation (19). Microdosimetry offers a physical explanation of the constants α and β in equation (19). In line with the interpretation given above, assume that the yield of lesions (alterations responsible for the end point observed) depends quadratically on specific energy: E ( z) = β z2.
(20)
On average [see equation (18)]:
(
)
E ( D ) = β z 2 = β zD D + D 2 .
(21)
Thus: zD =
α β
.
(22)
To summarize: β is proportional to the yield of sublesions (e.g., single-chromosome breaks) per unit specific energy; and the ratio α/β, a purely physical quantity depends on radiation quality and on the distribution of radiosensitive matter in the cell.
2–Radiobiology: A Briefing for the Brachytherapist
17
Temporal Aspects (II) To account for sublesion repair—a factor that affects the quadratic term, βD2—the LQ expression is modified as follows:
ε ( D ) = α D + q ( t ) β D 2.
(23)
The dose-rate function, q(t), quantifies sublesion damage repair that occurs in-between events. It is calculated as follows: ∞
∫ τ (t ) h (t ) dt ,
q=
(24)
0
where τ(t) = exp(−t/t0) gives the rate of sublesion damage elimination as a function of time, t0 is the sublethal repair constant, and h(t)dt is the distribution of time intervals t between consecutive events (hits). Thus: h (t ) =
2 D
2
∞
∫ I ( s ) h ( s + t ) ds .
(25)
0
In this expression I(s) is the dose rate as a function of time. As expected, for acute (HDR) exposures q → 1; for protracted exposures q → 0. For instance: (a) For irradiations at a constant dose rate: q(r ) =
2 r
−
2 f2
(1 − e ) −r
(26)
where r = t/t0. In particular, when t >> t0 : q≅
2t 0
.
(27)
t
(b) For f well-separated fractions (complete sublesion repair in between fractions): q=
1 f
.
(28)
In a typical LDR treatment in brachytherapy the total irradiation time is of the order of several days and therefore q ª 0. It follows that the probability of cell survival, S(D), is quasi-exponential and the RBE is determined by the linear (α) coefficient—or in microdosimetric terms, by zD. In HDR brachytherapy the dose rate is of the order of 1.5 Gy/min; for a prescription dose of, say, 6 Gy and taking t0 = 60 min one obtains q = 0.978 and thus both terms, linear (αD) and quadratic (βD2), contribute significantly.
18
Marco Zaider
Exercises 1. Demonstrate equation (26). 2. Demonstrate equation (27). 3. Demonstrate equation (28).
Applications In this section we discuss some simple applications of the items described above. For numerical estimations the following may come in handy: Typical initial dose rates in brachytherapy are 0.07 Gy/h (125I) and 0.24 Gy/h (103Pd) for permanent implants; 0.5 Gy/h for interstitial implants; 50 to 100 Gy/h for socalled “HDR” treatments. It is common to take α/β = 10 Gy for early responding tissues (tumors and tissues with fast turnover) and α/β = 3 Gy for late responding tissues. For the sublethal repair constant t0 of equation (24), a value of 0.5–1 h would be representative for most cell lines. The simplest (but not necessarily accurate) way to account for proliferation during time t is through a term exp(t/Tpot).5 Since only the quadratic term in dose (qβD2) is affected by temporal effects, the terminology “low” or “high” dose rate must be understood in terms of the relative contribution of this term to the total yield of lesions. Thus, if the dose rate is such that αD >> qβD2, one is in low dose rate regime. Note that the lower the dose, the larger the range of dose rates that classify as “low.”
Exercise Using α = 0.1 Gy−1, β = 0.01 Gy−2, and t0 = 1 h, find—as a function of dose—the range of dose rates that qualify as “low” if one defines low dose rate by the condition 0.1 αD = qβD2.
Equivalent Treatments (Painless Version) Two treatments [different with respect to either one or any combination of the following: radiation quality and type of tissue (α, β), dose rate, Tpot, sublesion repair constant (t0)] can be made equivalent by changing the total dose, the dose rate, and/or the treatment time. It is understood that the temporal pattern of dose delivery is given by prescription. However, one must ask first “equivalent in terms of what?” The two choices are equivalent TCP or equivalent NTCP. It is generally assumed (incorrectly, in fact) that for both TCP and NTCP treatments that result in equal survival probabilities for the cells that make up the respective tissues cause equivalent biological outcome. For instance, in a tumor containing n malignant cells treated uniformly to dose D, the average number of surviving cells is nS(D). If the number of surviving cells is Poisson-distributed (this is true only under certain conditions), the probability that no cell will survive the treatment is: TCP = e− nS ( D ) ,
(29)
and in this case iso-survival probability is indeed equivalent to iso-TCP. The case for NTCP is less clear and the assumption mentioned above (equal survival means equal normal tissue damage) is by and large a matter of faith. 5
See, however, the section herein Equivalent Treatments (II).
2–Radiobiology: A Briefing for the Brachytherapist
19
Under these conditions to find biologically equivalent treatments, one would use the following expression:
α D + β qD 2 −
t Tpot
= α 1 D1 + β1q1 D12 −
t1 Tpot ,1
.
(30)
This equation must be applied separately to early- (e.g., tumor) and late-responding tissues; and if the dose is designed to match, say, the effect on the tumor, one must evaluate if the effects on normal tissues will increase or decrease. A worked example [In the following we shall use α/βearly = 10 Gy, α/βlate = 3 Gy, Tpot = 2 d (early) or 60 d (late), β = 0.01 Gy−2, t0 = 1 h). 1. A tissue is known to tolerate N = 30 daily fractions of 1.8 Gy each (total dose, D). What is the equivalent total dose, D1, required if the treatment is delivered in one single fraction over 7 days? (ignore cell proliferation) Solution:
2t 1 2 D = α D1 + β 0 D12 t N
αD + β
α 2t 0 2 α 1 2 D1 + D1 − D + D = 0 N β t β 2
2t α α α 1 − + + 4 0 D + D2 N β t β β D1 = 4t0 t For early responding tissues D1 = 59.5 Gy. For late-responding tissues D1 = 68.0 Gy. 2. If the dose is designed to match the tumor tissue, will the effect on late-responding tissues increase or decrease? 3. Repeat the calculation above by including cell proliferation. Now (with obvious notations and ∆t = 1 day):
20
Marco Zaider
S( D) = e
2
− α D − β qD + t / T pot
2t t 1 2 t D − = α D1 + β 0 D12 − 1 N t2 T T1
αD + β
2t 0 2 α α 1 2 N ∆t t1 t D1 + β D1 − β D + N D − T β + T β = 0 1 1 2
− D1 =
N ∆t t 2t α α α 1 + 1 + + 4 0 D + D2 − N β T β T1 B t1 β β
4t0 t 1
Answer again question 2, but for this case.
Equivalent Treatments (II) For any temporal protocol of dose delivery, a general formulation that describes the probability of tumor control—defined here as the probability that at time t there are no clonogens alive—is:
S ( t ) e(b − d ) t TCP ( t n, S ) = 1 − t dt ' (b − d )t 1 + bS t e ( ) ∫ S ( t ' ) e(b − d ) t ' 0
n
.
(31)
In equation (31) S(t) is the survival probability (proliferation processes excluded) at time t of the n clonogenic tumor cells present at the time treatment started (t = 0), and b and d are, respectively, the birth and (radiation-independent) death rates of these cells. Equivalently, b = 0.693/Tpot and d/b is the cell loss factor, φ, of the tumor. In this expression t refers to any time during or after the treatment; typically, one would take for t the end of the treatment period or (better) the expected remaining life span of the patient. [Unlike equation (29) the expression, equation (31), is valid for both temporary and permanent implants.] The quantity: Seffective =
S ( t ) e(b − d ) t t
1 + bS ( t ) e(b − d ) t ∫ 0
dt '
(32)
S ( t ') e
(b − d )t '
in equation (31) represents the effective survival probability (proliferation included) of a cell present in the system at time t. In the same way, one may define a biologically equivalent dose (BED) by observing that
2–Radiobiology: A Briefing for the Brachytherapist
BED = −
1
α
21
S ( t ) e(b − d ) t
log
1 + bS ( t ) e
(b − d )t
(33)
t
dt '
∫ S (t ') e(
b − d )t '
0
has a dimension of dose. In the absence of cell proliferation (b = d = 0) BED takes the more familiar form:
β BED = D 1 + q ( t ) D . α
(34)
For a fractionated treatment (f fractions separated by time ∆1, S0 = survival probability after a single fraction), the probability of cure—defined as the probability that as time post-treatment tends to infinity there will be no clonogenic tumor cells left—is given by the following expression:
f λ∆ S0 e ) ( TCP = 1 − λ∆ λ ∆ ( f − 1) − 1) b eλ ∆ 1 + 1 − e− λ ∆ S0 e ( S0 e ) ( λ1 S0 e λ ∆ − 1 1
1
1
1
1
1
n
1
1
(35)
1
1
1
1
(λ1 = b-d) or, within a very good approximation:
f λ∆ S0 e ) ( n TCP ≈ exp − . λ∆ λ ∆ ( f − 1) − 1) b eλ1 ∆1 1 + 1 − e− λ ∆ S0 e ( S0 e ) ( λ1 S0 e λ ∆ − 1 1
1
1
1
1
1
(36)
1
1
1
1
Radiation Quality The photon or electron energies used in brachytherapy are generally low. At low doses or low dose rates the RBE of 100 keV photons relative to, say, 1-MeV photons may exceed 2. For instance, the RBE for 125 I has been extensively studied and results of 1.2 to 2 have been reported for the dose-rate range of 0.03 to 9 Gy/h. Similarly, for 103Pd a study performed at 0.07 to 0.8 Gy/h reported an RBE value of 1.9. While considerably lower RBE values apply to the much higher doses or dose rates usually employed in external radiotherapy, differences in biological effectiveness of the order of 10% to 15% remain. In a typical LDR treatment in brachytherapy the total irradiation time is of the order of several days and therefore q ª 0. It follows that the probability of cell survival, S(D), is quasi-exponential and the RBE is determined by the linear (α) coefficient—or in microdosimetric terms, by zD. The expression, equation (22), links the ratio α/β to the microdosimetric quantity zD and because β changes very little with electron energy one can assume that—on a relative scale—the coefficient α changes in the same manner. The table below gives numerical examples of ratios RBE = zD/zD(60Co) for radiations of interest in brachytherapy.
22
Marco Zaider
zD/zD(60Co)
Radionuclide 103
Pd
2.3
125
I
2.1
Am
2.1
241
192
Ir
1.3
60
Co
1.0
The Real World The treatment presented here disregards many important factors present in real-life clinical situations and the unwary should not be carried away by mathematical enthusiasm. Here are four examples: 1. It is rarely the case, and it never happens in brachytherapy, that any given treatment area (tumor or healthy organ) is uniformly exposed to the same dose. To the extent that individual cells (or functionally linked groups of cells) respond independently to radiation the device known as dosevolume histogram (DVH) can be used to take this into account. For instance, the TCP expression, equation (31) becomes: N
TCP ( t ) = ∏ TCP ( t ni , Si ) ,
(37)
i =1
where ni tumor cells have the same (dose-dependent) survival probability, Si. 2. Tissues (malignant or not) consist of cells that have a wide spectrum of α, β, Tpot, and φ values. Theoretical considerations indicate that—unless one particular cell type numerically dominates the tissue in question—tumor response and normal-tissue toxicity is primarily determined by a (possibly) small number of cells—radioresistant, fast-proliferating in the case of TCP, or radiosensitive for NTCP. Empirical evidence for this conjecture comes from the fact that, quite invariably, analyses of TCP dose-response data indicate that the number of relevant cells [n in equation (31)] is remarkably small. 3. As actively proliferating cells progress through the mitotic cycle, they change radiosensitivity parameters by as many as one to two orders of magnitude. 4. Cells do not generally exist in isolation and their response to radiation may depend on many extracellular factors and fluctuating environments. Nevertheless, one should remember that crude though such mathematical models may be in terms of biological realism, their limitations fade into insignificance when compared with the clinical rewards that can result from increased knowledge of the mechanisms responsible for radiation action.
Further Reading Brahme, A., J. Nilsson, and D. Belkic. (2001). “Biologically optimized radiation therapy.” Acta Oncol 40:725–734. Dale, R. G. (1985). “The application of the linear-quadratic dose-effect equation to fractionated and protracted radiotherapy.” Br J Radiol 58:515–528. Fowler, J. F. (1989). “The linear-quadratic formula and progress in fractionated radiotherapy.” Br J Radiol 62:679–694.
2–Radiobiology: A Briefing for the Brachytherapist
23
Hall, E. J. Radiobiology for the Radiologist. New York: Harper and Row Publishers, 1994. Kellerer, A. M., and H. H. Rossi. (1978). “Generalized formulation of dual radiation action.” Radiat Res 75:471–488. Kiefer, J. Biological Radiation Effects. New York: Springer-Verlag, 1990. Rossi, H. H., and M. Zaider. Microdosimetry and Its Applications. Springer-Verlag Telos, 1996. Steel, G. Gordon. Growth Kinetics of Tumors. Gloucestershire, UK: Clarendon Press, 1977. Steel, G. Gordon (ed.). Basic Clinical Radiobiology. Oxford: Oxford University Press, 1997. Thames, H. D., and J. H. Hendry. Fractionation in Radiotherapy. Bristol, PA: Taylor and Francis, 1987. Zaider, M., and G. N. Minerbo. (2000). “Tumor control probability: A formulation applicable to any temporal protocol of dose delivery.” Phys Med Biol 45:279–293. Zaider, M., and H. H. Rossi. Radiation Science for Physicians and Public Health Workers. Norwell, MA: Kluwer Academic Publishers, 2001.
Chapter 3
Sources and Delivery Systems I: Radionuclides Ravinder Nath, Ph.D. Department of Therapeutic Radiology Yale University School of Medicine New Haven, Connecticut Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 Photon-Emitting Radionuclides Used in Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 226 Ra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 137 Cs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 192 Ir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 198 Au . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 125 I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 103 Pd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Beta-Emitting Radionuclides Used in Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 90 Sr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 32 P . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 90 Y . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Neutron-Emitting Radionuclides Used in Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 252 Cf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Other Radionuclides with Potential Applications in Brachytherapy . . . . . . . . . . . . . . . . . . . . . . 29 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Further Reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Introduction In the early applications of brachytherapy, naturally occurring radionuclides of radium and radon were used extensively. As the artificially produced radionuclides became available from nuclear reactors and particle accelerators, many new radionuclides entered the brachytherapy field. Today in medicine, there is an extensive menu of radionuclides emitting gammas, betas, or neutrons with wide spectra of energies and half-lives. For brachytherapy, these radionuclides are encapsulated in sealed, biocompatible capsules made of materials such as titanium, stainless steel, etc. The physical characteristics of some of the commonly used radionuclides in brachytherapy are described in this chapter. Details of the fabrication and composition of actual brachytherapy sources in use are described in the next chapter.
Photon-Emitting Radionuclides Used in Brachytherapy 226
Ra
226
Ra is the sixth member of the uranium decay series that starts with 238U and ends with stable 206Pb. Radium decays into radon, with a half-life of about 1622 years; radon is a heavy, inert gas that in turn disintegrates into its daughter products. As a result of the decaying processes from 226Ra to 206Pb, at least 49 photons are emitted, with energies ranging from 0.184 to 2.45 MeV. The average energy of the gamma rays from radium in equilibrium with its daughter products and filtered by 0.5 mm of platinum is 0.83 MeV. There are also some high-energy beta particles and alpha particles emitted by this radionuclide that are absorbed by the encapsulation material. The half-value layer (HVL) of the gamma rays from an encapsulated source containing this radionuclide is about 14 mm of lead. 226 Ra and 222Rn are virtually unused today, primarily because of the hazards of chemical and radioactive toxicity of radium and its by-products and because of the high energy of photons emitted by these
25
26
Ravinder Nath
sources. The high energy of the photons makes it difficult to shield health professionals and others from unwanted radiation exposure. For these reasons, several other radionuclides (as described below), which emit lower-energy photons, have been introduced. In the early radium sources, the radioactive material was supplied mostly in the chemical form of radium sulfate or radium chloride mixed with an inert filler and loaded into cells about 1 cm long and 1 mm in diameter. A typical radium source might contain 1 to 3 cells, depending on the source length. Radium sources were manufactured as needles or tubes in a variety of lengths and activities. 137
Cs
137
Cs is a by-product of nuclear fission and is generated in nuclear reactors. It has a half-life of about 30 years and decays through beta emission (93.5%) to the metastable state of 137Ba (half-life 2.5 minutes) and then through gamma-ray emissions to its ground state (Figure 1). The gamma rays emitted have a photon energy of 662 keV. The HVL for the gamma rays from this radionuclide is about 5.5 mm of lead. For 137Cs sources, the radioactive material is supplied in the form of insoluble powders or ceramic microspheres labeled with cesium and doubly encapsulated in stainless steel for both needles and tubes. The beta particles and low-energy characteristic x-rays are totally absorbed by the stainless steel encapsulation, making the clinical source a pure gamma emitter. The active length of the tube sources typically used is 10 to 50 mm and diameters are approximately 1.5 to 2.0 mm. 137 Cs sources are less hazardous and require less shielding compared to that required for 226Ra sources. Although the half-life of 137Cs is much less than that of 226Ra, some of the cesium sources may have to be replaced after about 7 years. In clinical use, the activity of the sources must be adjusted for the decay of 137 Cs over time. 137Cs sources have been used in both interstitial and intracavitary brachytherapy. 192
Ir
192
Ir is produced in a nuclear reactor via neutron capture by stable 191Ir. 192Ir has a half-life of 73.83 days and decays primarily by beta emission and electron capture to excited states of 192Pt and 192Os. Subsequently, the daughters decay to the ground states by gamma-ray emission. 192Ir has a very complicated gamma-ray spectrum. The average energy of gamma rays is about 0.38 MeV. Because the gamma-ray energy is lower than 226Ra or 137Cs7, 192Ir sources require less shielding. The HVL thickness for this radionuclide is about 2.5 mm. 192 Ir sources are available in the form of small sources (about 3 mm long with a diameter of about 0.5 mm) placed in nylon ribbons for safety purposes. Two different source designs are commercially available in the United States. One of them has an inner core alloy composed of 30% iridium and 70% platinum, encapsulated in stainless steel. The other has an inner core alloy composed of 10% iridium and 90% platinum surrounded by a 0.1-mm thick cladding of platinum. Typically, these sources are press-fitted,
Figure 1. Decay scheme for 137Cs.
3–Sources and Delivery Systems I: Radionuclides
27
normally at 1-cm intervals, into nylon ribbons with an outer diameter of about 0.8 mm. 192Ir sources are also available as wires composed of an alloy of 25% iridium and 75% platinum enveloped in pure platinum. The wires are available in two sizes with outer diameters of 0.3 mm and 0.5 mm. These are more popular in Europe than in the United States. 192Ir wires and sources in ribbons are particularly suitable for afterloading technique. These sources are used for temporary brachytherapy implants only, because their high average photon energy makes them unsuitable for permanent implants due to excessive public radiation exposure. High-intensity 192Ir sources are also available for the high dose rate (HDR) remote afterloader units. Very high activity sources (typically 10 Ci) of 192Ir can be fabricated in a very small volume source because of the high specific activity of this radionuclide. 192 Ir is the only gamma isotope that has been clinically used for intravascular brachytherapy (IVB) thus far. In one of the IVB systems, 192Ir sources were used as trains of sources in nylon ribbons containing 6, 10, 14, 18, or 22 seeds with 1 mm seed spacing (i.e., 4 mm center to center seed spacing). Seed activities were typically 33 mCi per seed. Treatment times for IVB were typically 15 to 40 minutes to deliver 15 Gy at a distance of 2 mm from the source center. A second 192Ir system for IVB consisted of a wire source (192Ir source inside a 0.4 mm diameter nitinol wire) rather than seeds. 198
Au
198
Au is produced in a reactor by bombarding a gold target with neutrons. 198Au has a half-life of 2.7 days and 99.9% of its gamma rays are emitted with an energy of 0.412 MeV; its primary beta emission has a maximum energy of 0.96 MeV. 198Au has three gammas (412, 676, and 1088 keV) and three betas. The nuclide may also undergo an isomeric transition (five prominent gammas from 100 to 334 keV) prior to beta decay. A typical gold seed, also known as a gold “grain,” is encapsulated in 0.1 mm of platinum, which is sufficient to absorb the beta rays emitted by 198Au. The outside dimensions of a 198Au source are 2.5 mm long and 0.8 mm in diameter. Because of their short half-life, 198Au seeds are used in permanent implants only. Although commonly used in Canada, the 198Au sources have been used only sparsely in the United States. 125
I
125
I has a half-life of 59.6 days decaying exclusively by electron capture process to an excited state of 125Te followed by spontaneous decay to the ground state with the emission of 35.5 keV gamma rays. Characteristic x-rays in the range of 27 to 32 keV are also emitted as a consequence of the electron capture and internal conversion processes. One of the advantages of this radionuclide over radon and gold is the lower energy of the photons that it emits, which makes reducing radiation exposures around the patient much easier. The HVL thickness for the photons emitted by encapsulated sources containing this radionuclide is about 0.025 mm of lead. Two models of the 125I seeds were originally introduced in the 1960s. In one of the models, the 125I was adsorbed onto a 3-mm long silver rod that is the central core of the source. The rod was encapsulated in 0.5 mm of titanium; the resultant seed is 4.5 mm long and 0.8 mm in diameter. In the other original model, the seeds consisted of 2 to 4 ion exchange resin spheres containing 125I inside a titanium tube. Titanium encapsulation served to absorb low-energy electrons and x-rays. The ion exchange resin beads were impregnated with 125I in the form of iodide ion. This technique offered an efficient absorption of 125I, which allows fabrication of higher activity sources than is possible with the other model. Except for their use in 125 I sources in eye plaques for temporary brachytherapy, 125I seeds are used principally for permanent implants. Seeds are available in strengths up to 50 U. In the last decade, numerous other models of 125I sources have become commercially available, as described further in the next chapter.
28
Ravinder Nath
103
Pd
103
Pd can be produced in a nuclear reactor when stable 102Pd captures a thermal neutron. It can also be produced in a cyclotron by bombarding protons into a rhodium target. With a half-life of 17 days, it decays via electron capture process with the emission of characteristic x-rays in the energy range of 20 to 23 keV and Auger electrons. The weighted mean photon energy is 20.7 keV, and the HVL of photons emitted from encapsulated sources of 103Pd is about 0.004 mm of lead. Original 103Pd sources consisted of a cylindrical titanium tube sealed at both ends with laser welded titanium end cups. Enclosed in the tube were two 103Pd graphite cylinders and a lead rod x-ray marker for radiographic identification. The cylindrical tube was about 5 mm long and 0.8 mm in diameter. In the last decade, numerous other models of 103Pd sources have become commercially available, as described further in the next chapter. 103 Pd seeds are primarily used for permanent implants. However, high-activity 103Pd sources have been fabricated which make it possible to use 103Pd for temporary brachytherapy applications.
Beta-Emitting Radionuclides Used in Brachytherapy 90
Sr
90
Sr decays with a half-life of 28.9 years to 90Y and yields only beta rays with a maximum energy of 0.5 MeV. The daughter isotope 90Y is a nearly pure beta emitter (99.99% of decay events) with a 64-hour halflife, and emits betas with a maximum energy of 2.27 MeV. Like 90Sr, 90Y decays solely by beta emission, but a tiny fraction of decays (0.01%) result in one of two excited levels of the daughter isotope, which then relaxes by emission of low-intensity gamma rays. It is primarily the 90Y betas that are useful for therapy. This radionuclide is a beta-ray emitter suitable for treatment of superficial lesions, e.g., a 90Sr ophthalmic applicator is used for treatment of lesions in the eye where the depth of penetration needed is a few millimeters. The second most common application for 90Sr is IVB for prevention of restenosis in coronary or peripheral arteries. One of the IVB systems consists of a string (“train”) of 90Sr seeds. Each seed is encapsulated in stainless steel, and the activity is contained in a ceramic matrix. The seeds are stored in a hand-held delivery device and are advanced by a closed loop hydraulic system, which uses sterile saline to advance (and then retract) the seeds into and out of a non-centered catheter. 32
P
32
P is a pure beta emitter with 14-day half-life, and betas with a maximum energy of 1.71 MeV. Recently, it has been used in a commercial IVB system, which consisted of a 32P wire source. In one version, the delivery system consisted of a sophisticated computer-controlled stepping motor, similar to a conventional HDR unit, and a spiral-shaped perfusion-centering catheter. The source itself contained 32P hermetically sealed inside a 0.5 mm diameter nitinol wire. Another IVB system utilizing 32P consisted of an angioplasty balloon impregnated with 32P. 90
Y
90
Y is the daughter product of the decay of 90Sr and is a beta emitter with a 64-hour half-life, and 2.27 MeV maximum beta energy. 90Y has been tested as a stand-alone source (i.e., not in equilibrium with 90Sr) in clinical trial for IVB. In this trial the 90Y source consisted of a 0.1 mm diameter 90Y wire coated with
3–Sources and Delivery Systems I: Radionuclides
29
titanium with a computer-controlled HDR-like delivery device and a non-perfusion, segmented balloon centering system.
Neutron-Emitting Radionuclides Used in Brachytherapy 252
Cf
252
Cf has a half-life of 2.65 years decaying via alpha emission (97%) and nuclear fission (3%). The fission decay path emits neutrons with a spectrum of neutron energies similar to that of a fission reactor and with an average energy of 2.15 MeV. A significant number of gamma rays with an average energy of 0.7 to 0.9 MeV are also emitted from the fission events and from the decay of fission products. One of the clinical 252Cf sources has a central core consisting of a ceramic metal mixture of 252Cf oxide and palladium. This core is encapsulated with platinum-iridium alloy (10% iridium) to limit the combined gamma-ray and beta-ray dose to one-third the effective neutron dose. The seed is about 6 mm long and 0.8 mm in diameter. 252Cf sources are intended for use in temporary implants only because of the high risk of radiation-induced carcinogenesis from chronic low-dose irradiation by neutrons.
Other Radionuclides with Potential Applications in Brachytherapy Other radionuclides that have been used in brachytherapy include 60Co, 182Ta, 106Ru, 106Rh, 241Am, 145Sm, and 169Yb. The long half-life of 241Am and its intermediate photon energy of about 60 keV indicate that this radionuclide may be well suited for intracavitary use in which custom-designed shielding may be very effective. 169Yb has similar average photon energy but a much shorter half-life, which would make it commercially attractive as an HDR source. 145Sm has a half-life of 340 days and photon energies in the range of 38 to 45 keV. With its low energy and relatively long half-life, 145Sm is being considered as a substitute isotope for removable interstitial implants. Several other radionuclides have also seen limited clinical testing, but do not appear to be likely candidates for immediate commercial development. One of these radionuclides is 188W (69.4-day half-life, 349 keV maximum beta energy) in equilibrium with its short-lived daughter 188Re (17-hour half-life, 2.12 MeV maximum beta energy). The 188Re beta decay is accompanied by 155 keV gammas (10%) plus another 2% of gammas with energies up to 932 keV. In a version of the IVB system, 50 to 100 mCi of liquid 188Re were “milked” from the generator for use in the radioactive liquid balloon delivery system for IVB. Another version of this system, 188Re in a sealed, catheter-based 188W/188Re source has been investigated for IVB. In this case, the source comprised a neutron-activated tungsten coil mounted on a flexible nitinol wire. One can only surmise that many new brachytherapy applications will emerge in future as a vast number of radionuclides are either present in nature or can be artificially produced and only a small fraction of these radionuclides have been clinically exploited so far. Choice of a radionuclide for a particular type of brachytherapy is a complex decision. The selected radionuclide must have dosimetric characteristics (i.e., energy and half-life) suitable for the specific clinical need. In addition, other critical factors are easy availability, reasonable cost, high-activation cross sections, low levels of undesirable contaminants, and high specific activity. The search for the ideal radionuclide for specific clinical applications continues to engage research physicists. In summary, the basic physical characteristics of commonly used radionuclides in brachytherapy are presented in Table 1.
30
Ravinder Nath Table 1. Physical Characteristics of Various Radionuclides Used for Brachytherapy
Radionuclide
Half-life
Principal or mean energies from encapsulated sources, MeV* Photon
Radium-226 Cesium-137 Iridium-192 Gold-190 Iodine-125 Palladium-103 Strontium-90 Ytterbium-90 Phosphorus-32
1622 y 30 y 74 d 2.7 d 60 d 17 d 29 y 64 h 14 d
Californium-252
2.65 y
Beta
Neutron
0.830 0.662 0.380 0.412 0.028 0.021 0.50 2.27 1.71 2.15
*These are nominal values assuming typical encapsulation for sources.
Acknowledgments Some of the information presented here has been taken from an article that was coauthored by the present author. More details and many other items are presented in the original article entitled “Basic Physics of Brachytherapy” by Ali Meigooni, Cheng B. Saw, and Ravinder Nath in the textbook entitled Principles and Practice of Brachytherapy edited by Subir Nag, Futura Publishing Company, Inc., Armonk, NY, 1997.
Further Reading Meigooni, A., C. B. Saw, and R. Nath. “Basic Physics of Brachytherapy” in Principles and Practice of Brachytherapy. Subir Nag (ed.). Armonk, NY: Futura Publishing Company, Inc., 1997. Nath, R. “Physical Characteristics and Clinical Uses of Brachytherapy Radionuclides” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath. Madison, WI: Medical Physics Publishing, pp. 7–37, 1995.
Chapter 4
LDR Sources: Design and Delivery Systems Robert E. Wallace, Ph.D. UCLA Clinical Professor at Radiation Oncology Cedars-Sinai Outpatient Cancer Center at The Samuel Oschin Comprehensive Cancer Institute Los Angeles, California Low Dose Rate Brachytherapy: Definition and Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 History of LDR Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Overview of Manchester, Quimby, Paris Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Patterson-Parker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Fletcher and Other Radium and Cesium Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Requirements on Source Design Deriving from Implanting Systems and Methods . . . . . . . . . . . 33 Encapsulation, Embodiments, and Desirable Features of Sources . . . . . . . . . . . . . . . . . . . . . . . 33 Quick History of Dosimetry and Strength Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Source Archetypes, 125I: Amersham and 103Pd Theragenics Seeds . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 History of Source Designs for 125I and 103Pd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Conclusions and Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Low Dose Rate Brachytherapy: Definition and Rationale The use of isotopes in treating tumors was first suggested soon after the discovery and separation of radium (Bell 1903). Bell observed one of the tenets of brachytherapy in that, by inserting (sealed) radioactive material directly into tumors, tumor dosage can be improved while the effects on healthy tissues can be reduced when compared to external x-rays (i.e., Crookes tube roentgen rays). That is, dose is not deposited in healthy tissues due to the traversal of x-rays from an external source to an internal target. Dose to healthy surrounding tissues in brachytherapy is a function of the placement of the radioactive sources and the type and energy of the radiation, the ability of the radiation to penetrate distances greater than the dimensions of the tumor, and the physical design of the source. There are radiobiological advantages to the dose distribution, the (usually short) time of treatment, and the use of low dose rate (Brenner 1997). Brenner observes that an array of implanted sources offers an “optimal conformal” dose distribution relative to that achievable using external beam systems. An appropriately designed implant can deliver high doses to the tumor while “minimizing radiobiological damage to normal adjacent tissues” (Brenner 1997). Brenner further observes that “good dose distributions spare: (1) early responding normal tissues, which, in external beam radiotherapy, typically produce the complications that force treatments to be prolonged over more than one month; and (2) late-responding normal tissues, which, in external beam radiotherapy, often represent the dose-limiting endpoint.” Shorter treatment time in brachytherapy generally overcomes accelerated tumor repopulation (Withers, Taylor, and Maciejewski 1988; Brenner 1993). Since in an implant, the dose to adjacent normal tissue is typically reduced, accelerated treatment has less effect on normal tissue. Low dose rate (LDR) irradiation generally reduces late effects in normal tissue more than it reduces tumor control (Coutard 1932; Lea and Catcheside 1942). From these observations, general aspects of brachytherapy source design are indicated: materials sealed in capsules, energy and radiation type selected for local penetration, activity and half-life selected for low dose rates, encapsulation, and internal isotope distribution designed to support conformal tumor dose distribution when an ensemble of sources are used.
31
32
Robert E. Wallace
History of LDR Brachytherapy In the early twentieth century, rigid and large brachytherapy sources contained radium and were sealed to contain radon gas and other toxic decay products and to filter α and β ray emission (Dominici 1907; Regaud 1924, 1930). By midcentury, man-made isotopes, tantalum-182 (182Ta) and iridium-192 (192Ir), in the form of wires made implantation more feasible (Van Miert and Fowler 1956; Hall, Oliver, and Shepstone 1966; Pierquin et al., 1978). Other such isotopes including cesium-137 (137Cs), gold-198 (198Au), cobalt-60 (60Co), iodine-125 (125I), ytterbium-169 (169Yb), phosphorus-32 (32P), and palladium-103 (103Pd) have been used in discrete, encapsulated sources of different sizes and applications (Hendee and Ibbott 1996). Fission spectrum neutron sources encapsulating californium-252 (252Cf) have been used to exploit the differences in tissue/tumor repair capacity to damage from high LET (linear energy transfer) radiations (Oliver and Wright 1969; Krishnaswamy 1971; Anderson 1974; Smith, Almond, and Delclos 1974; Rivard 2000). The development of LDR sources in the past 10 years has largely focused on sources containing 125I and 103Pd owing to the increase in their use in treating prostate cancer (Russell and Blasko 1993). Development in the field of brachytherapy sources includes designs encapsulating cesium-131 (131Cs) (Murphy et al., 2004), first suggested earlier (Henschke and Lawrence 1965). The use of sources in treating cancer quickly led to the development of systems, tables, and nomograms to facilitate and expedite clinical applications. The assumptions and the dosimetry of these systems indicate some of the desired features and some requirements for source design. Overview of Manchester, Quimby, Paris Systems Prior to the advent of robust computer treatment-planning systems for brachytherapy, several different systems of distribution (i.e., placement) rules, dose tables, and dose specification criteria were developed to optimize particular implant geometries. Each system had assumptions about what determined a “good dose distribution” and each was constrained by the available sources. The Manchester system (Paterson and Parker 1934, 1938, 1995; Meredith 1967) was designed to maximize dose homogeneity within a treated plane or volume using planar or volume arrays of radium sources. The loading rules dictate a variable loading in the implant, some regions having more or stronger sources than others. As such, implementation of the system requires a large variety of radium needles of different active lengths and activities. Using wires or seeds having uniform activity, one can achieve the differential loading of the Manchester system by altering the geometric loading rules and source-spacing requirements of the system, originally fixed to 1-cm spacing. In one approach, the differential loading is accomplished by using sources of differing lengths and strengths in a uniform geometry. In another, identical sources are placed in a variable geometry to achieve differential loading. The Quimby system (Quimby 1944; Quimby and Castro 1953) for implants is distinguished by its use of uniform spatial distribution of uniform source activity within a plane or a volume. Spacing is uniform but varies in dimension between 1 to 2 cm between sources. The Paris system (Pierquin, et al. 1978; Dutreix and Marinello 1987) was designed for uniformly spaced, uniform linear density 192I wire sources in one or two planes for planar and volume implants. Spacing between wires is 1/2 to 2 cm. In two-plane implants covering a volume, wires are distributed in equilateral triangle or square geometry. The system applies also to strands of 192Ir seeds, spaced uniformly (Pierquin and Marinello 1997). Paterson-Parker In the development of the Manchester system for radium implants, Parker considered the non-uniform distribution of activity along linear sources that are placed on a uniform grid (Meredith 1967). The calculations of this model indicate a ratio of activity of 3:1 between the rind (periphery) and the core (contained
4–LDR Sources: Design and Delivery Systems
33
volume). That is, approximately 75% of the activity is placed on the periphery, the balance uniformly distributed in the enclosed volume. Modified Loading for Modern Prostate Implants. In planning a permanent implant of the prostate using low-energy sources of 125I or 103Pd, a modified Paterson-Parker-like loading can limit the urethra dose to just above the prescribed prostate dose. In studies of possible loading schemes for prostate therapy, a modified uniform loading placing sources toward the periphery while removing sources proximal to the urethra can improve dose coverage while reducing urethral dose (Butler et al., 2000). The recommendations of the AAPM Task Group 56 (TG-56) are that “the treatment plan is designed to place seeds peripherally to improve dose homogeneity and to avoid unnecessary radiation damage to the urethra” (Nath et al., 1997). Whether the ratio of peripheral to core activity is 2:1 or 3:1, modern prostate implants reflect the earlier observations of Parker (Meredith 1967) regarding the achievement of uniform volume dose by using non-uniform distribution of activity, or a non-uniform distribution of a number of sources of equal activity. Fletcher and Other Radium and Cesium Methods Intracavitary applicators for the treatment of cervical cancers at Manchester were first described for use with radium in 1938 (Tod and Meredith 1953; Meredith 1967a). The applicators were similar to later devices in geometry, having an intrauterine linear array of sources with paired vaginal sources possibly with some shielding for the bladder and rectum (Fletcher 1953; Haas, Dean, and Mansfield 1985). Sealed sources of radium and later 137Cs are used with no specific requirement on source design other than consistency with the apertures in the rigid applicator bodies. Modern cesium sources from any manufacturer have utilized one of the few designs that existed when the applicators themselves were designed. This reduces cost because patents no longer protect those designs and using an established design speeds regulatory approval. Requirements on Source Design Deriving from Implanting Systems and Methods The requirements on modern sources are fairly obvious: sealed and durable encapsulation; availability (in quantity) in different strengths; existence of a standard for source strength; the existence of standardized single-source dosimetric parameters in a form usable in treatment planning; visibility using radiographic, ultrasound, or magnetic resonance imaging; and compatibility with existing source designs and source delivery systems (e.g., applicators, needles, catheters). The use of discrete sources in ultrasound-guided prostate implants leads to some desirable features in x-ray source dosimetry: high dose per contained activity (may reduce cost), relatively isotropic distribution around the source (since source orientation is possibly indeterminate), and limited penetration (i.e., low energy) to spare adjacent tissues. Sources in permanent implants for prostate brachytherapy should be relatively small, given that the prostate is not a large target.
Encapsulation, Embodiments, and Desirable Features of Sources Bell had suggested encapsulating radium salts in glass tubes (Bell 1903), while the benefits of filtering and sealing radium sources in a platinum-iridium needle were observed by Dominici (1907). Radium needles were manufactured in different active lengths with a variety of internal distributions of activity: uniform, dumbbell (higher activity at needle ends), and club (higher activity at one end). Cesium-137 is a substitute for radium and is available in needles and in tubes. The advent of afterloading allowed the replacement of needle-based brachytherapy with tube sources that would be placed within implanted catheters, hollow needles, or intracavitary applicators. Typical dimensions for stainless steel, encapsulated, ceramic 137Cs sources in current production are approximately 3 mm in diameter by 10 to 20 mm physical length with the tube ends plugged and welded (Devine 2004). Recent evaluation
34
Robert E. Wallace
of dosimetry for a variety of 137Cs source designs allows for treatment planning using the current AAPM TG-43 formalism (Liu, Prasad, and Bassano 2004; Rivard et al., 2004) Outdated in use, 182Ta sources had been available for temporary implants in the form of flexible wires, 0.2 mm diameter, encased in a 0.1 mm thick platinum sheath (Cohen 1955). Wires of 192Ir are constructed with a core of 25% iridium, 75% platinum, 0.1 to 0.3 mm in diameter, encased in a 0.1 mm thick platinum sheath (Hendee and Ibbott 1996). Iridium seed sources are available in two designs, the first has a 0.3 mm diameter core 10% 192Ir/90% platinum clad in a 0.1 mm thick platinum sheath (Best Industries, Springfield, VA). The second design has a 0.1 mm diameter core of 30% iridium/70% platinum clad in a 0.2 mm thick stainless steel sheath (Alpha-Omega Services, Bellflower, CA). Iridium seeds are typically delivered in nylon strands at uniform 1 cm center-to-center spacing. As a substitute for gold-encapsulated pellets containing radon gas, “seed” type small sources containing 198Au are used in some applications (Slanina et al. 1982; Hendee and Ibbott 1996). Present manufacture provides these sources with the active gold in a 2.2 mm by 0.5 mm diameter pellet encapsulated in a 0.15mm thick platinum sheath, resulting in a seed that is 2.5 mm long by 0.8 mm in diameter (Best Industries, Springfield, VA). Seed sources containing 125I or 103Pd are available in a number of designs, differing principally in internal configuration. At one point the market in the United States had nearly 20 different manufacturers of 125 I seeds and 5 manufacturers of 103Pd seeds. The large number of manufacturers and differences in source design derive from the increased demand due to increased popularity in the late 1990s of their use in permanent-implant prostate brachytherapy, and the existence of active, enforceable patents on existing source designs. The market competition has led to novel designs that variously improve single-source dosimetry, improve visibility, improve availability, and reduce manufacturing costs.
Quick History of Dosimetry and Strength Standards The dosimetry of brachytherapy sources has developed through a series of reports over the past 18 years. The reports covered source strength standards, dosimetry, and use of brachytherapy sources. Reports of exposure standards were published by the National Bureau of Standards [NBS, which is now the National Institute of Science and Technology (NIST)] for 137Cs (Loftus 1970), 192Ir (Loftus 1980), and 125I (Loftus 1984). Kubo observed that 125I exposure measured in the NBS 1984 standard using the Ritz chamber (Ritz 1960) included the effects of fluorescent x-rays produced in the titanium encapsulation of the 125I sources. These titanium soft (approx. 5 keV) x-rays have little practical significance in phantom dose measurements (Kubo 1985; Williamson 1988). As long as a consistent standard and individual source dose-rate constant data were used, under the 1984 standard, patient administered dose was not significantly affected. In 1987, the AAPM TG-32 published, in AAPM Report No. 21 (AAPM 1987), the recommendation that brachytherapy source strength should be specified in terms of air kerma strength and that such should be adopted by NBS/NIST and by the medical physics community for clinical use. NBS responded with a standard (Loftus 1988), a subsequent paper provided direction for clinical implementation (Williamson and Nath 1991), and the use of air kerma strength was endorsed by the American Endocurietherapy Society (Williamson et al., 1993). Also, in approximately 1988, a manufacturer implemented an ad hoc 103Pd source activity standard that was based on intercomparison with a 109Cd source having NIST-traceable activity calibration (Williamson et al., 2000). In 1999, NIST reported an air kerma strength standard (Seltzer et al., 1999; Seltzer and Lamperti 1999; Seltzer et al., 2003) for both 125I and 103 Pd using a new device (Loevinger 1993) that addressed the effect observed by Kubo. The 1999 NIST standard for 125I and 103Pd air kerma strength had direct bearing on revision of dosimetric parameters. Because air kerma strength under the NIST 1999 standard for 103Pd was significantly different from that derived under the existing ad hoc standard, patient dose prescription when using 103Pd was also affected. Suggestions for revision of the prescribed dose were enunciated by the AAPM (Williamson et al., 2000).
4–LDR Sources: Design and Delivery Systems
35
The 1995 AAPM TG-43 report unified dosimetry for 192Ir, 125I, and 103Pd brachytherapy sources (Nath et al., 1995). Subsequent papers described clinical implementation of the recommendations (Luse, Blasko, and Grimm 1997; Bice et al., 1998). In anticipation of the 1999 NIST standard for 125I, the report of the AAPM ad hoc committee on 125I sealed-source dosimetry addressed the implementation of TG-43 dosimetry in the clinic, dose prescription changes for 125I using the published TG-43 parameters, and clinical incorporation of the proposed air kerma strength standard from NIST (Kubo et al., 1998). That report addressed the 125I sources available in September 1997, models 6702 and 6711 (Nycomed-Amersham, Arlington Heights, IL), and provided a formulary for implementing the 1999 NIST standard. The arrival of a new source design in the market in early 1998 (Wallace and Fan 1998) and the prospect of further designs (Wallace and Fan 1999) prompted the delineation (Williamson et al., 1998) of requisite procedures to characterize any new source not considered in TG-43. These recommendations by the AAPM ad hoc subcommittee on low-energy brachytherapy source dosimetry were followed by specific instructions to users (Williamson et al., 1999) regarding changes in the NIST standard for 125I sources implemented early in 1999. The arrival in January 1999 of new design 103Pd sources (Wallace and Fan 1999, 1999a; Li, Palta, and Fan 2000; Nath et al., 2000; Williamson 2000) and the contemporaneous enunciation of the 1999 NIST standard for 103Pd air kerma strength raised two issues. The first issue concerned the appropriate therapeutic dose prescription for 103Pd under the new 1999 NIST standard and how to convert a decade of clinical experience using a source having an activity intercomparison standard. This was addressed in an exhaustive report of the ad hoc subcommittee (Williamson et al. 2000). This report also provides instructions for clinical implementation of the new 103Pd source strength standard and new dosimetric evaluations. The second issue was the continued use of apparent activity both by manufacturers and in clinical application. In their communication, the ad hoc subcommittee reaffirmed earlier recommendations and provided direction in the interim use of apparent activity during the transition to TG-43 based dosimetry (Williamson et al., 1999a). NIST and the manufacturing community recognized anomalous results in air kerma strength standardization measurements made under the primary standard for late 1998 through 1999. Discrepancies of 2% to 7% were identified, but users of the affected 125I and 103Pd source designs were advised to make no changes to any dosimetry parameters until a detailed formal analysis was completed (Williamson et al., 2000a). An update of the TG-43 report was subsequently published that reviewed dosimetric formalism, methods for measurement and analysis of data for new sources, and that compiled consensus data for selected sources (Rivard et al. 2004, 2004a). All measured dose-rate constant data in the updated report reflect corrections for the 1999 anomaly and should be applied to the affected sources in the clinic.
Source Archetypes, 125I: Amersham and 103Pd Theragenics Seeds In the United States, low-energy 125I sources were first manufactured four decades ago by the Lawrence Soft X-ray Company. The 3M Company took over the manufacturing and then sold that part of their business to what are now the Amersham Health/MediPhysics (Arlington Heights, IL) and Oncura (Plymouth Meeting, PA) companies. These sources, the models 6701, 6702, 6711, and the more recent 6733 demonstrate most of the design issues in 125I sources. In 1988–1989, Theragenics Corporation (Norcross, GA) released one design of a 103Pd source, the model 200, that has undergone small, incremental manufacturing variations to the present day. A 103Pd source, model MED 3933, was released in 1999 by North American Scientific, Inc. (NAS, Chatsworth, CA) and a new source from Mills Biopharmaceuticals, Inc. (MBI, Oklahoma City, OK) recently was reported (Rivard, Melhus, and Kirk 2004). These three designs provide archetypes for 103Pd sources. A unique 103Pd source design, RadioCoil™, (RadioMed Corporation, Tyngsboro, MA) has been reported where the source is a coiled wire of activated rhodium (Meigooni et al., 2004).
36
Robert E. Wallace
History of Source Designs for 125I and 103Pd The model 67xx source designs are depicted in Figure 1. The first source introduced in this series, the 6701, was a titanium cylinder, 4.5 mm nominal length, and 0.8 mm outer diameter, end plugged and welded, depicted in Figure 1a. This design contained two active resin spheres (0.5 mm diameter) that flank a central radiopaque marker. 125I is incorporated into the resin via an ion-exchange mechanism when the resin spheres are soaked in a concentrated solution of an ionic salt containing 125I. This is one of the more common mechanisms of incorporating radioactive iodine in sources. The use of only three spheres appears to have been dictated by the process of welding plugs at the source container ends where heat dissipation may have been an issue. Modern laser welding techniques are designed to minimize this problem and typically remain as industrial secrets. The 6701 is visible in a radiograph while the 6702 design is not. In the 6702, the central marker is replaced with a third active sphere, apparently to increase the activity range of the design. Both designs are presently unavailable for routine use. In Figure 1b, the model 6711 design is shown. In this design, the 125I is adsorbed (i.e., plated) onto a 3 mm long, 0.5 mm diameter silver rod using proprietary techniques. This design provides improved radiographic visibility with the possibility of determining both location and orientation. The incorporation of the silver substrate adds characteristic x-rays from silver to the 125I spectrum. This had a significant effect on the calibration of 125I source strengths at NIST, as outlined above. The use of ultrasound localization in prostate brachytherapy generated a desire to improve acoustic visibility of sources. The model 6733 source (Figure 1c) represents an approach to improved visibility of individual sources under ultrasound. The principal difference between this design and the 6711 is the encapsulation. The corrugation of the cylindrical titanium tube portion of the capsule provides a wide-angle dispersion of reflected sound back to the ultrasound transducer. Since reflection is specular, a flat surface will reflect sound at an angle that mirrors the incident angle, usually away from the transducer. Corrugation provides several different reflection angles, thereby improving the likelihood that some reflected sound will be detected at the transducer. Also, in the 6733, the end-welds are less massive than in the 670x and 6711 designs. The design of the model 200 103Pd source is shown in Figure 2. In the model 200 design, the active elements have a graphite substrate and the marker is lead. The inverted end-cups sealing the ends of the titanium cylinder allow welding to be distant from the internal components while presenting lower attenuation than a 6701 type end plug. The concavity and extent of the end-cups serves to limit the motion of internal elements of the source. There are many other designs for sources that are available for use. The AAPM updated TG-43 report (Rivard et al., 2004) provides comprehensive analysis of the dosimetry of sources, while a review of how source design affects dosimetry of 125I sources is also available (Heintz, Wallace, and Hevezi 2001). In terms of dosimetry, the model 670x sources differ from model 6711/33 sources. In these latter source designs, the inclusion of characteristic x-ray from the interaction of the 125I x-rays in the silver substrate lowers the average energy of the emitted spectrum, decreases the penetration in water, and decreases the reference dose per unit source strength (i.e., the dose-rate constant). Compared to the 670x sources, the dose-rate constants for the 6711/6733 sources are approximately 6% to 7% lower than those of the 670x sources (Heintz, Wallace, and Hevezi 2001; Rivard et al. 2004, 2004a). The Mills 125I and 103Pd sources share the same design, Figure 3. These sources have a standard size, 4.5 mm length by 0.8 mm diameter, titanium capsule with laser-welded ends. The end welds of these sources are somewhat less massive than those found in the 670x or 6711 sources. The isotope bearing elements are five silver spheres, plated with either 125I or 103Pd. These active and radiopaque elements move within the source, with random orientations shown in Figure 3a that compare to an idealized configuration, Figure 3b, used for dosimetry modeling studies (Li 2002). The net effect of internal motion is to introduce a small (+0.3/–0.2) uncertainty in the dose-rate constant for the 103Pd source that is about 3% lower than for other available sources (Rivard, Melhus, and Kirk 2004). This is not clinically untenable considering that the spread in the dose-rate constant among the various 125I source designs is on the order
4–LDR Sources: Design and Delivery Systems
37
Figure 1a. 6701, 6702 sources. Model 6702 is shown. In the 6701, the central resin sphere is replaced with a radiopaque marker. [Reprinted from Heintz, B. H., R. E. Wallace, and J. M. Hevezi, “Comparison of I-125 sources used for permanent interstitial implants,” Med Phys 28:671–682. © 2001, with permission from AAPM.]
Figure 1b. Model 6711. [Reprinted from Heintz, B. H., R. E. Wallace, and J. M. Hevezi, “Comparison of I-125 sources used for permanent interstitial implants,” Med Phys 28:671–682. © 2001, with permission from AAPM.]
Figure 1c. Model 6733. [Reprinted from Meigooni, A. S., S. A. Dini, K. Sowards, J. L. Hayes, and A. Al-Otoom, “Experimental determination of the TG-43 dosimetric characteristics of EchoSeed™ model 6733 125I brachytherapy source,” Med Phys 29:939–942. © 2002, with permission from AAPM.]
38
Robert E. Wallace
Figure 2. Model 200. [Reprinted from Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson, “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations,” Med Phys 31:633–674. © 2004, with permission from AAPM.]
Figure 3a. MBI 125I and 103Pd sources, radiographs. [Reprinted from Rivard, M. J., C. S. Melhus, and B. L. Kirk, “Brachytherapy dosimetry parameters calculated for a new 103Pd source,” Med Phys 31:2466–2470. © 2004, with permission from AAPM.]
Figure 3b. MBI 125I and 103Pd sources containing five 0.5-mm diameter silver spheres plated with 125I or 103Pd. [Reprinted from Li, Z., “Monte Carlo calculations dosimetry parameters of the Urocor Prostaseed 125I source,” Med Phys 29:1029–1034. © 2002, with permission from AAPM.]
of 5% (Heintz, Wallace, and Hevezi 2001; Rivard et al., 2004, 2004a). These designs are similar to the 6711 and 6733 in that the isotope is distributed on the surface of a silver substrate. The model IS-125-01 125 I source (Imagyn Medical Technologies, Denton, TX) is similar in design to the Mills sources where the five silver spheres are 0.64 mm in diameter. The model 1A1-125A (IsoAid, Port Richey, FL) resembles the 6711 design, yet with thinner end-welds. The model 130.002 125I (Nucletron Corporation, Columbia, MD) also resembles the 6711 source with a longer marker, 3.4 mm versus 3 mm (Karaiskos et al., 2001). Consequently, the Imagyn, IsoAid, Mills, and Nucletron 125I sources have an emitted spectrum that includes characteristic x-rays from silver. The palladium substrate of the model BT-125I 125I source (no longer available; Syncor Corporation, Golden, CO) provides characteristic x-rays comparable to those from silver. The use of silver or palladium substrate results in a lowered dose-rate constant that is similar to that for the models 6711/6733 (Heintz, Wallace, and Hevezi 2001; Karaiskos et al., 2001;
4–LDR Sources: Design and Delivery Systems
39
Rivard et al. 2004, 2004a). The use of a tungsten substrate (and marker rod) in the model 2301 (Best Medical Industries, Springfield, VA) adds characteristic x-rays that do not significantly affect the otherwise pure 125I emitted spectrum (Heintz, Wallace, and Hevezi 2001). The double-walled container of the model 2301 presents very thin end walls (about 0.04 mm) and 0.08 mm thick body walls. (Most other sources have body wall thickness on the 0.04 to 0.05 mm.) The model 3500 125I source (Implant Sciences Corporation, Wakefield, MA) contains a silver marker of smaller diameter than used in the 6711 but with pointed ends to enhance radiographic visibility. Due to the reduced mass of the silver marker as compared to the model 6711, the dose-rate constant for this source is between that of the 6702 and 6711 sources (Rivard 2002). In the model 3500, 125I is incorporated into a quartz tube via the ion implantation of 124Xe, subsequently neutron-activated to 125Xe (17 hour half-life) and 125I. After decay of the 125Xe, the active quartz tubes are processed into brachytherapy sources (Munro 2004). The model I125.S06 (BeBig/Theragenics, Norcross, GA), and the model STM125I (Source Tech Medical, Carol Stream, IL) each contain linear gold markers in novel configurations that place the 125I in either a ceramic matrix, in the I125.S06, or plated on a copper sheath covering an aluminum tube that itself contains the gold marker rod. As the case for the tungsten marker in the model 2301, the characteristic x-rays from the gold, copper, and aluminum components do not significantly affect the dosimetry of these sources (Heintz, Wallace, and Hevezi 2001). Of the sources discussed thus far, all but the models 6701 125I and 200 103Pd sources are essentially linear sources (or an uninterrupted linear array of active elements—spheres). In radionuclide distribution, these sources resemble the uniform radium needles. The model 200 103Pd source, the NAS model MED-3631-A/M 125I , MED3633 103Pd sources, the Best Industries model 2335 103Pd source, and the model LS-1 125I source (Draximage, Inc., Kirkland, Quebec, Canada) each place the active elements on organic materials separated by a marker. This echoes the 6701 design. These designs also reflect the “dumbbell” designs of radium needles, and for the same reasons. The dose distributions around these sources should be more isotropic than a line source (Heintz, Wallace, and Hevezi 2001; Rivard et al., 2004, 2004a), particularly when the end-walls/-welds are less massive than those of the 670x designs. The NAS sources are depicted in Figure 4, and contain four active elements (resin beads suffused with the isotope) separated by two gold/copper alloy markers. In the Best model 2335 source, six polymer beads containing the 103Pd are separated by a short tungsten marker within a double-wall container (Meigooni et al., 2001). The Draximage model LS-1 source is likely the most complicated design that is currently available, shown in Figure 5 (Heintz, Wallace and Hevezi 2001). In the source, two glass beads carrying the 125I are separated by a Pt/Ir(10%) marker with centralized extra wall thickness using internal titanium spacers that also fix the locations of the glass beads. The weld of this source is also centralized to allow minimal end-wall thickness. In this source, the activity contained
Figure 4. NAS Model 3631-A/M 125I and Model 3633 103Pd sources. [Reprinted from Heintz, B. H., R. E. Wallace, and J. M. Hevezi, “Comparison of I-125 sources used for permanent interstitial implants,” Med Phys 28:671–682. © 2001, with permission from AAPM.]
40
Robert E. Wallace
Figure 5. Draximage model LS-1 125I source. [Reprinted from Heintz, B. H., R. E. Wallace, and J. M. Hevezi, “Comparison of I-125 sources used for permanent interstitial implants,” Med Phys 28:671–682. © 2001, with permission from AAPM.]
in/on the glass beads should be identical in order to obtain an ideally isotropic dose distribution. The same is true for the model 200 source, Figure 2. The use of multiple active elements in the model MED3631A/M, MED 3633, and 2335 sources allows some variation in the activity carried by each bead so long as the net activity at each source end is nearly identical. The manufacturing issue reduces to efficient sorting in these sources rather than exact activation in the models 200 and LS-1. The model 1251L (International Brachytherapy, Norcross, GA) 125I source, Figure 6, is a concentric tubular design that allows a suture to be run through the source to form strands of sources. Dumbbellstyle end-loading and relatively thin encapsulation provide for a relatively high dose-rate constant and better isotropy than a 6711 type source (Heintz, Wallace and Hevezi 2001). A novel design 103Pd source, the RadioCoil, is outlined in Figure 7. This source, available in various lengths, combines the flexibility of the Paris system wires with the low-energy of 103Pd. The source is fabricated of a coiled flat-wire of high-purity rhodium that is neutron-activated to 103Pd. The 5 mm length source, a size comparable to other 103Pd seeds, also demonstrates equivalent dosimetry: dose-rate constant, penetration in tissue, and dose isotropy (Meigooni et al., 2004). A new 131Cs source, model CS-1 (IsoRay, Inc., Richland, WA), has been reported that resembles the model I125.S06, 3500, and STM-125I sources. In Figure 8, the source is shown with an axial gold marker rod, a glass/ceramic tube bearing the 131Cs, with ring welded end-caps. Relative dosimetry appears to be comparable to 125I sources in penetration (due to similar average energy, 29 to 34 keV) and dose anisotropy. The initial dose rate is higher than 103Pd due to the 9.7-day half-life and the dose-rate constant is nearer that for 125I sources than 103Pd sources (Murphy et al., 2004). As of this writing, no formal air kerma strength standard exists for 131Cs brachytherapy sources such as are available for either 103Pd or 125I sources.
Conclusions and Remarks There has been considerable effort in improving the design and manufacture of brachytherapy sources in the past decade, focusing primarily on seeds containing 103Pd and 125I. Legacy designs for 198Au, 192Ir, and 137Cs remain available for use. A new seed for 131Cs has been reported for clinical use, possibly in place of 125I, but an air kerma strength standard needs to be enunciated. The principles and desirable features of source design were laid out in the early twentieth century, notably by Paterson and Parker in 1934, and endure in current designs. Not specifically mentioned in this chapter are significant advances in packaging: seed cartridges, preloaded needles and catheters, uniform and customized non-uniform spaced strands of sources, or automated needle loading systems.
4–LDR Sources: Design and Delivery Systems
41
Figure 6. IBT model 1251l. [Reprinted from Heintz, B. H., R. E. Wallace, and J. M. Hevezi, “Comparison of I-125 sources used for permanent interstitial implants,” Med Phys 28:671–682. © 2001, with permission from AAPM.]
Figure 7. RadioCoil™ 103Pd activated rhodium source 6733. [Reprinted from Meigooni, A. S., H. Zhang, J. R. Clark, V. Rachabatthula, and R. A. Koona, “Dosimetric characteristics of the new RadioCoil™ 103Pd wire line source for use in permanent brachytherapy implants,” Med Phys 31:3095–3105. © 2004, with permission from AAPM.]
Figure 8. IsoRay model CS-1 131Cs brachytherapy source. [Reprinted from Murphy, M. K., R. K. Piper, L. R. Greenwood, M. G. Mitch, P. J. Lamperti, S. M. Seltzer, M. J. Bales, and M. H. Philips, “Evaluation of the new cesium-131 seed for use in low-energy x-ray brachytherapy,” Med Phys 6:1529–1538. © 2004, with permission from AAPM.]
42
Robert E. Wallace
References American Association of Physicists in Medicine (AAPM). Report No. 21. “Specification of Brachytherapy Source Strength.” New York: American Institute of Physics, New York, 1987. Anderson, L. L. (1974). “Characteristics of 252Cf neutron sources used for radiotherapy.” Eur J Cancer 10:203–205. Bell, A. G. (1903). “The uses of radium.” Ann Med 6:261. Bice, W. S., B. R. Prestidge, J. J. Prete, and D. F. DuBois. (1998). “Clinical impact of implementing the recommendations of the AAPM Task Group 43 on permanent prostate brachytherapy using 125I.” Int J Radiat Oncol Biol Phys 40:1237–1241. Brenner, D. J. (1997). “Radiation biology in brachytherapy.” J Surg Oncol 65:66–70. Brenner, D. J. (1993). “Accelerated repopulation during radiotherapy–Evidence for delayed onset.” Radiat Oncol Invest 1:167–172. Butler, W. M., G. S. Merrick, J. H. Lief, and A. T. Dorsey. (2000). “Comparison of seed loading approaches in prostate brachytherapy.” Med Phys 27:381–392. Cohen, L. (1955). “Protracted interstitial irradiation of tumors using 182Ta.” Br J Radiol 28:338. Coutard, H. (1932). “Roentgentherapy of epitheliomas of the tonsillar region, hypopharynx, and larynx, from 1920 to 1926.” Am J Roentgenol 28:313–331, 343–348. Devine, M. (2004). Private communication. Isotope Products Laboratories, Valencia, CA. Dominici, H. (1907). “Modifications histologiques déterminées par le rayonnement du radium.” Arch D’elect Méd 15:835. Dutreix, A., and G. Marinello. “The Paris System.” Chapter 5 in Modern Brachytherapy. Chicago, IL: Year Book Medical Publishers, 1987. Fletcher, G. H. (1953). “Cervical radium applicators with screening in the direction of bladder and rectum.” Radiology 60:77–84. Haas, J. S., R. D. Dean, and C. M. Mansfield. (1985). “Dosimetric comparison of the Fletcher family of gynecologic colpostats 1950–1980.” Int J Radiat Oncol Biol Phys 11:1317–1321. Hall, E. J., R. Oliver, and B. J. Shepstone. (1966). “Routine dosimetry with tantalum-182 and iridium-192 wires.” Acta Radiol 4:155. Heintz, B. H., R. E. Wallace, and J. M. Hevezi. (2001). “Comparison of I-125 sources used for permanent interstitial implants.” Med Phys 28:671–682. Hendee, W. R., and G. S. Ibbott. Radiation Therapy Physics, Second Edition. St. Louis, MO: Mosby, 1996. Henschke, U. K., and D. C. Lawrence. (1965). “Cesium-131 seeds for permanent implants.” Radiology 85:1117–1119. Karaiskos, P., P. Papagiannis, L. Sakelliou, G. Anagnostopoulos, and D. Baltas. (2001). “Monte Carlo dosimetry of the selectSeed 125I interstitial brachytherapy seed.” Med Phys 28:1753–1760. Krishnaswamy, V. (1971). “252Cf dosimetry.” Br J Radiol 527:903. Kubo, H. (1985). “Exposure contribution from Ti K x rays produces in the titanium capsule of the clinical I-125 seed.” Med Phys 12:215–220. Kubo, H. D., B. M. Coursey, W. F. Hanson, R. W. Kline, S. M. Seltzer, R. E. Shuping, and J. F. Williamson. (1998). “Report of the ad hoc committee of the AAPM Radiation Therapy Committee on 125I sealed source dosimetry.” Int J Radiat Oncol Biol Phys 40:697–702. Lea, D. E., and D. G. Catcheside. (1942). “The mechanism of the induction by radiation of chromosome aberrations in Tradescantia.” J Genet 44:216–245. Li, Z. (2002). “Monte Carlo calculations dosimetry parameters of the Urocor Prostaseed 125I source.” Med Phys 29:1029–1034. Li, Z., J. Palta, and J. Fan. (2000). “Monte Carlo calculations and experimental measurements of dosimetry parameters of a new Pd-103 source.” Med Phys 27:1108–1112. Liu, L., S. C. Prasad, and D. A. Bassano. (2004). “Determination of 137Cs dosimetry parameters according to the AAPM TG-43 formalism.” Med Phys 31:477–483. Loevinger, R. (1993). “Wide-angle free-air chamber for calibration of low-energy brachytherapy sources (abstract).” Med Phys 20:907.
4–LDR Sources: Design and Delivery Systems
43
Loftus, T. P. (1970). “Standardization of 137Cs gamma-ray sources in terms of exposure units (roentgens).” J Res Nat Bur Stand 74A:1–6. Loftus, T. P. (1980). “Standardization of 192Ir gamma-ray sources in terms of exposure.” J Res Nat Bur Stand 85:19–25. Loftus, T. P. (1984). “Exposure standardization of 125I seeds used for brachytherapy.” J Res Nat Bur Stand 89:295–303. Loftus, T. P. (1988). “Calibration of gamma-ray emitting brachytherapy sources.” Nat Bur Stand Spec Pub 89:2250. Luse, R. W., J. Blasko, and P. Grimm. (1997). “A method for implementing the American Association of Physicists in Medicine Task Group-43 dosimetry recommendations for 125I transperineal prostate seed implants.” Int J Radiat Oncol Biol Phys 37:737–741. Meigooni, A. S., Z. Bharucha, M. Yoe-Sein, and K. Sowards. (2001). “Dosimetric characteristics of the Best® doublewall 103Pd brachytherapy source.” Med Phys 28:2568–2575. Meigooni, A. S., S. A. Dini, K. Sowards, J. L. Hayes, and A. Al-Otoom. (2002). “Experimental determination of the TG-43 dosimetric characteristics of EchoSeed™ model 6733 125I brachytherapy source.” Med Phys 29:939–942. Meigooni, A. S., H. Zhang, J. R. Clark, V. Rachabatthula, and R. A. Koona. (2004). “Dosimetric characteristics of the new RadioCoil™ 103Pd wire line source for use in permanent brachytherapy implants.” Med Phys 31:3095–3105. Meredith, W. J. “Physical Aspects of the Interstitial Treatment System.” Chapter 10 in Radium Dosage: The Manchester System, ed. 2. Edinburgh: E. & S. Livingstone, pp. 83–107, 1967. Meredith, W. J. “Dosage for Cancer of the Cervix Uteri.” Chapter 6 in Radium Dosage: The Manchester System, ed. 2. Edinburgh: E. & S. Livingstone, pp. 42–50, 1967a. Munro, J. J. (2004). Private communication, Implant Sciences Corporation, Wakefield, MA. Murphy, M. K., R. K. Piper, L. R. Greenwood, M. G. Mitch, P. J. Lamperti, S. M. Seltzer, M. J. Bales, and M. H. Philips. (2004). “Evaluation of the new cesium-131 seed for use in low-energy x-ray brachytherapy.” Med Phys 6:1529–1538. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni, (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43.” Med Phys 22:209–234. Also available as AAPM Report No. 51. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56. American Association of Physicists in Medicine.” Med Phys 24:1557–1598. Also available as AAPM Report No. 59. Nath, R., N. Yue, K. Shahnazi, and P. J. Bongiorni. (2000). “Measurement of dose-rate constant for 103Pd seeds with air-kerma strength calibration based upon a primary national standard.” Med Phys 27:655–658. Oliver, G., and C. Wright. (1969). “Dosimetry of an implantable 252Cf source.” Radiology 92:143. Paterson, R., and H. M. Parker. (1934). “A dosage system for gamma-ray therapy.” Br J Radiol 7:592–632. Paterson, R., and H. M. Parker. (1938). “A dosage system for interstitial radium therapy.” Br J Radiol 11:252–266. Paterson, R., and H. M. Parker. (1995). “A dosage system for gamma ray therapy. 1934.” Br J Radiol 68:H60–100. Pierquin, B., and G. Marinello. A Practical Manual of Brachytherapy. Translated by F. Wilson, B. Erickson, and J. Cunningham. Madison, WI: Medical Physics Publishing, 1997. Pierquin, B., A. Dutreix, C. H. Paine, D. Chassagne, G. Marinello, and D. Ash. (1978). “The Paris system in interstitial radiation therapy.” Acta Radiol Oncol 17:33–48. Quimby, E. H. (1944). “Dosage table for linear radium sources.” Radiology 43:572. Quimby, E. H., and V. Castro. (1953). “The calculation of dose in interstitial radium therapy.” Am J Roentgenol 70:739–749. Regaud, C. (1924). “Some biological aspects of radiation therapy of cancer.” Am J Roentgenol 10:97. Regaud, C. (1930). “Sur les principes radiophysiologiques de la radiothérapie des cancers.” Acta Radiol 11:455. Ritz, V. H. (1960). “Standard free-air chamber for the measurement of low-energy x-rays (20 to 100k – constant potential).” J Res Nat Bur Stand 64:49. Rivard, M. J. (2000). “Neutron dosimetry for a general 252Cf brachytherapy source.” Med Phys 27:2803–2815. Rivard, M. J. (2002). “Comprehensive Monte Carlo calculations of AAPM Task Group Report No. 43 dosimetry parameters for the Model 3500 I-Plant 125I brachytherapy source.” Appl Radiat Isotop 57:381–389.
44
Robert E. Wallace
Rivard, M. J., C. S. Melhus, and B. L. Kirk. (2004). “Brachytherapy dosimetry parameters calculated for a new 103Pd source.” Med Phys 31:2466–2470. Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. (2004). “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Also available as AAPM Report No. 84. Rivard, M. J., W. M. Butler, L. A. DeWerd, M. S. Huq, G. S. Ibbott, Z. Li, M. G. Mitch, R. Nath, and J. F. Williamson. (2004a). “Erratum: ‘Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.’” [Med Phys 31:633–674 (2004)].” Med Phys 31:3532–3533. Russell, K. J., and J. C. Blasko. (1993). “Recent advances in interstitial brachytherapy for localized prostate cancer.” Probl Urol 7:260–278. Seltzer, S. M., and P. J. Lamperti. (1999). Status of NIST Primary Standards for 125I and 103Pd Therapy Seeds Based on the Wide-Angle Free-Air Chamber (WAFAC): Proceedings of a CIRMS Workshop, National Institute of Standards and Technology. Seltzer, S. M., P. J. Lamperti, R. Loevinger, C. G. Soares, and J. T. Weaver. (1999). “New NIST air-kerma strength standards for 125I and 103Pd brachytherapy seeds (abstr).” Med Phys 25:170. Seltzer, S. M., P. J. Lamperti, R. Loevinger, M. G. Mitch, J. T. Weaver, and B. M. Coursey. (2003). “New national air-kerma-strength standards for 125I and 103Pd brachytherapy seeds.” J Res Nat Bur Stand 108:337–358. Slanina, J., M. Wannenmacher, K. Kuphal, H. Knufermann, C. Beck, and W. Schilli. (1982). “Interstitial radiotherapy with 198Au seeds in the primary management of carcinoma of the oral tongue: Results in Freiburg/Breisgau from January 1964 to July 1980.” Int J Radiat Oncol Biol Phys 8:1683–1689. Smith, A., P. Almond, and L. Delclos. (1974). “Evaluation of 252Cf neutron emitter for interstitial and intracavitary radiation therapy.” Eur J Cancer 10:369. Tod, M., and W. J. Meredith. (1953). “Treatment of cancer of the cervix uteri, a revised Manchester method.” Br J Radiol 26:252–257. Van Miert, P. J., and J. F. Fowler. (1956). “The use of tantalum-182 in the treatment of early bladder cancer.” Br J Radiol 29:508. Wallace, R. E., and J. J. Fan. (1998). “Evaluation of a new brachytherapy iodine-125 source by AAPM TG43 formalism.” Med Phys 25:2190–2196. Wallace, R. E., and J. J. Fan. (1999). “Report on the dosimetry of a new design 125iodine brachytherapy source.” Med Phys 26:1925–1931. Wallace, R. E., and J. J. Fan. (1999a). “Dosimetric characterization of a new design 103palladium brachytherapy source.” Med Phys 26:2465–2470. Williamson, J. F. (1988). “Monte Carlo evaluation of specific dose constants in water for 125I seeds.” Med Phys 15:686–694. Williamson, J. F. (2000). “Monte Carlo modeling of the transverse-axis dose distribution of the model 200 103Pd interstitial brachytherapy source.” Med Phys 27:643–654. Williamson, J. F., and R. Nath. (1991). “Clinical implementation of AAPM Task Group 32 recommendations on brachytherapy source strength specification.” Med Phys 18:439–448. Williamson, J. F., L. L. Anderson, D. W. Grigsby, A. Martinez, R. Nath, D. Neblett, A. Olch, and K. Weaver. (1993). “American Endocurietherapy Society recommendations for specification of brachytherapy source strength.” Endocuriether/Hypertherm Oncol 9:1–7. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, and R. Nath. (1998). “Dosimetric prerequisites for clinical use of new low energy photon interstitial brachytherapy sources.” Med Phys 25:2269–2270. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, R. Nath, and G. Ibbott. (1999). “Guidance to users of Nycomed Amersham and North American Scientific, Inc. I-125 interstitial sources: Recommendations of the AAPM Radiation Therapy Committee ad hoc Subcommittee on low-energy seed dosimetry.” Med Phys 26:570–573. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, R. Nath, M. J. Rivard, and G. Ibbott. (1999a). “On the use of apparent activity (Aapp) for treatment planning of 125I and 103Pd interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Subcommittee on low-energy brachytherapy source dosimetry.” Med Phys 26:2529–2530.
4–LDR Sources: Design and Delivery Systems
45
Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, R. Nath, M. J. Rivard, and G. Ibbott. (2000). “Recommendations of the American Association of Physicists in Medicine on 103Pd interstitial source calibration and dosimetry: Implications for dose specification and prescription.” Med Phys 27:634–642. Also available as AAPM Report No. 69. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, G. Ibbott, R. Nath, and M. J. Rivard. (2000a). “Important notice for radiation therapy physicists using 125I or 103Pd brachytherapy sources (Letter).” Online: http://rpc.mdanderson.org/rpc/htm/Home_htm/Low-energy%20documents/ImpNotice_EndUsers_v12.pdf. Withers, H. R., J. M. G. Taylor, and B. Maciejewski. (1988). “The hazard of accelerated tumor clonogen repopulation during radiotherapy.” Acta Oncol 27:131–146.
Chapter 5
Quality Management of Low Dose Rate Brachytherapy Sources Gary A. Ezzell, Ph.D. Mayo Clinic Scottsdale Scottsdale, Arizona Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Acceptance Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Preparation for Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Long-lived Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Short-lived Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Periodic Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Per-patient Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 Form 1: Radioactive Materials Requisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Form 2: Cesium In-Use Inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Form 3: Iridium In-Use Inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Introduction The purpose of this chapter is to review one particular part of the quality management (QM) program for low dose rate brachytherapy: that associated with the sources themselves. It is the first link in the quality chain, since a problem with the sources will certainly affect the quality of the brachytherapy treatment. The QM program has three main components: acceptance testing of new sources, periodic testing, and per-patient testing. This discussion is organized in that fashion. The requirements and procedures for each component depend somewhat on the characteristics of the sources being used. Useful divisions are between sources used for temporary or permanent implants (sources which emit primarily photons with an average energy >50 kV and other sources, respectively), long- or short-lived isotopes, and small or extended sources (e.g., 125I seeds or 137Cs tubes.) The discussion for each section reflects these major divisions. Major references guiding this discussion are, of course, the AAPM TG-40 report on comprehensive quality assurance in radiation oncology (Kutcher et al. 1994) and the AAPM TG-56 report on the brachytherapy code of practice (Nath et al. 1997).
Acceptance Testing Preparation for Calibration The local physicist has the responsibility of assuring the accuracy of the calibration of the sources used for brachytherapy. TG-56 states “Every institution practicing brachytherapy shall have a system for measuring source strength with secondary traceability for all source types used in its practice.” Secondary traceability means that the source strength is measured either with an instrument that has been itself calibrated for that source type either at the National Institute of Standards and Technology (NIST) or an Accredited Dosimetry Calibration Laboratory (ADCL) or by comparison to a source of the same type that has been calibrated at NIST or an ADCL. In practice, such measurements are best done with a well-type chamber. So the first step in the process is to obtain such a chamber/electrometer system and arrange with
47
48
Gary A. Ezzell
an ADCL to calibrate either that system or an appropriate source. (The details of the calibration process and its associated QM program are covered in another chapter.) Long-lived Sources For long-lived sources such as 137Cs tubes, TG-56 recommends the following steps: 1. Document the physical/chemical form and encapsulation based on the manufacturer’s specification in order to support dosimetry calculations. 2. Document the initial leak test based on the manufacturer’s certification, and, optionally, repeat with local equipment (see below). 3. Verify the uniformity of the activity distribution within each source (most likely with an autoradiograph; no tolerance limit is given). 4. Verify the location of the activity distribution within each source relative to the exterior dimensions to a tolerance of 1 mm (most likely by combining an autoradiograph with a transmission radiograph). 5. Verify the identification of each source (serial numbers, color coding, etc.). 6. Calibrate each source. If the measured source strength agrees with the manufacturer’s specification to within 3%, then either may be used for dose calculations. Differences larger than 3% should be investigated, and differences larger than 5% should be reported to the manufacturer. TG-56 asserts “It is always advisable to ask the manufacturer to review its calibration of the sources to help resolve these discrepancies.” When sources are batched for dosimetry purposes, then the 3% tolerance applies to the mean of the batch, and the range of source strengths within the batch should not exceed 5%. Long-lived sources need to be leak tested at intervals specified by the radioactive materials license, typically 6 months or 3 years, depending on the source type. Such sources should have certified leak tests before being shipped, and that documentation can show initial compliance. Users of sources such as 137Cs tubes may decide to repeat the leak test locally during acceptance; certainly that should be done if there is any suggestion of damage. Properly leak testing sealed sources requires sensitive instrumentation, such as a NaI well counter, a means of determining the counting efficiency of the system for the energies being analyzed, and careful consideration of counting times in order to determine the minimum detectable activity (MDA). In the United States, the MDA must be less than 0.005 µCi. A comprehensive discussion of leak test considerations is given in Thomadsen (2000, pp 21–28). Figure 1 demonstrates localizing the activity distribution within 137Cs tubes. This was accomplished using Gafchromic RTQA film (International Specialty Products, Wayne, NJ). First, the position of each source was drawn on the film surface with lines showing the axis and endpoints of each tube assembly, along with the serial numbers. Then, the sources were laid on the film. After approximately 10 minutes of exposure for these nominal 20 mgRaeq sources, the self-developed autoradiograph shows the location of the distribution relative to the physical source. Short-lived Sources For short-lived sources such as 125I seeds, TG-56 recommends the following steps: 1. Document the physical/chemical form and encapsulation based on the manufacturer’s specification in order to support dosimetry calculations.
5–Quality Management of Low Dose Rate Brachytherapy Sources
49
2. Verify the uniformity of the activity distribution within each source, where applicable, or the distribution of seeds along an extended ribbon. 3. Calibrate the sources, either individually or as a batch. For large batches of loose seeds, TG-56 allows for “secondary traceability by statistical inference” in which a “suitable random sample” is calibrated with secondary traceability. TG-40 calls for that sample to comprise at least 10% of the total batch. Some institutions have developed procedures for measuring all the seeds in a batch together, applying a correction factor that depends on the batch size (Lee et al. 1999; Mellenberg and Pennington 1999). A disadvantage of batch measurements is that one or more significant outlier sources may be lost in the background. For large numbers of seeds in ribbons, TG-40 calls for measuring a minimum of 10% of the seeds or two ribbons, whichever is larger. It is possible to obtain inserts for well chambers that allow ribbons to be coiled into reproducible positions, but it may be necessary to develop correction factors for ribbons of different lengths (Thomadsen et al. 1999) The same as with long-lived sources, the expectation is that the local calibration should agree with that provided by the manufacturer within 3%. The range of source strengths should not exceed 5%, but TG-56 acknowledges that verifying this for a large batch of seeds may not be practical. For sources purchased in a sterile assembly, TG-40 recommends, “purchasing and calibrating a single (non-sterile) seed for each designated-strength grouping.” This problem of sterile assemblies deserves additional discussion. Calibrating a loose seed as a surrogate for the actual seeds implanted assumes that there is some meaningful connection between the manufacturing process that produced the loose seed and that for the assembly. For example, if the loose seed comes from the same manufacturing lot as those used to create the assembly, then one might consider it a random sample of the seeds used to treat the patient. If the loose seed comes from a different lot, but the manufacturer declares that it has passed through the same
Figure 1. Autoradiograph of 137Cs tube sources using Gafchromic RTQA film. The exposure time was approximately 10 minutes for these nominal 20 mgRaeq sources. Left: lines drawn to show the axis and endpoints of the sources, along with the serial numbers. Middle: sources placed on the prepared lines. Right: self-developed densities indicating the radioactive material distribution.
50
Gary A. Ezzell
calibration process as those used to create the assembly, then the connection is weakened, but may still be sufficient. There are several ways to deal with the problem of assuring the calibration of seeds that come in sterile assemblies. 1. Purchase and calibrate some loose seeds of the same type and activity. 2. Purchase additional assemblies, which are then calibrated and discarded. 3. Break sterility on some assemblies, which are then calibrated and re-sterilized. 4. In the operating room (OR), calibrate some of the assemblies under sterile conditions (Feygelman et al. 1996; Butler et al. 1998). 5. In the OR, remove some of the seeds from the assemblies under sterile conditions, and then go to a non-sterile area to calibrate and subsequently discard them. 6. Purchase the seeds from the manufacturer and have them sent to a third party that can provide an independent calibration and sterilize the assemblies. The third-party supplier of pre-loaded needles or stranded seeds may be asked to hold aside a few loose seeds from the manufacturer’s shipment for end-user calibration, sterilization, and possible implantation. None of these options is clearly best. The loose seeds in option (1) above may not be representative of the seeds implanted. Purchasing material for calibration only as in (2) or (5) adds expense, and (5) treats less of the patient because the assemblies are shortened. Resterilization (3) may not be possible or permitted by the package insert. Sterile calibration (4) is probably the best solution but has some logistical challenges. The last option listed, (6), provides an independent calibration that is not under the control of the local institution and with no assurance of the third party’s qualifications of traceability. Whether or not that is acceptable is somewhat debatable and is presently under review by an AAPM working group. Note that leak testing short-lived sources is not usually necessary, since they are typically not kept in active inventory for longer than 6 months. TG-56 does recommend leak-testing 125I seeds before using them for a second patient or after extensive handling, since the titanium encapsulation is relatively fragile and the radioactive material highly volatile.
Periodic Testing For sources that are maintained in inventory, the quality management program should include some periodic activities. TG-56 recommends the following steps: 1. A formal inventory check needs to be completed quarterly; this is a Nuclear Regulatory Commission (NRC) requirement that applies to federal and state licensees. Note that this could apply to 125 I seeds if an inventory is maintained. 2. The calibration of each source should be checked annually and the identification system (e.g., color-coding) renewed, if applicable. Deviations from the apparent half-life may indicate contamination with a different isotope. Should any such deviation be found and confirmed, the dose distribution around the source may differ from standard tables. Assessing that dose distribution is beyond the capabilities of most clinics, so any such sources would typically be returned to the manufacturer. 3. Leak testing should be performed at the intervals required by the radioactive materials license. Note that this could also apply to 125I seeds kept in service for longer than 6 months (an unlikely but not impossible situation.)
5–Quality Management of Low Dose Rate Brachytherapy Sources
51
Rechecking the uniformity of the activity within the source, while not specifically recommended by the AAPM, is another possible element of the QM program.
Per-Patient Testing For long-lived sources maintained in an inventory, such as 137Cs tubes, each time the sources are used there should be a mechanism for checking the proper selection and arrangement of the sources. Typically, this is accomplished by using a color-coding system. Where staff resources permit, one can have one person build the source assembly and another check it. It is a regulatory requirement that an inventory system be established that records the use of the sources. The system should clearly show which sources are out of storage and where they are. The importance of inventory control cannot be overstated. Dealing with a lost source is best read about, not experienced firsthand. The appendix shows worksheets developed at Wheeling Hospital, Wheeling, WV, to aid in specifying the sources to be used in an implant and in documenting the inventory. For short-lived sources that are obtained for a specific patient’s treatment, the tests listed under “Acceptance Testing” are both necessary and sufficient.
Conclusion The topic for this chapter has been the quality management system for the sources themselves. The table below summarizes the recommendations. It should be apparent to any physicist with some clinical experience that this element of the overall quality management program is necessary, since an error here can affect a large number of patients. Most errors that do occur, however, relate to how the sources are used. Thus, quality management of the sources is where vigilance in brachytherapy begins.
52
Gary A. Ezzell Table 1. Quality Control Tests for Low Dose Rate Brachytherapy Sources Long-lived sources, maintained in an inventory Test
Physical/chemical form; encapsulation Check integrity with leak test
Activity location and uniformity of distribution Verify identification system Calibration Inventory
Method
Frequency
Shipment documentation
Acceptance
Leak test
Acceptance; then every 6 months unless otherwise specified by license Acceptance; optionally at annual calibration Acceptance; annually; each use Acceptance; annually Acceptance; quarterly; each use
Radiograph/autoradiograph Check color-coding or equivalent Well-chamber Inspection
Short-lived sources, ordered for specific patients Physical/chemical form; encapsulation Activity location and uniformity of distribution (where applicable, such as wires or source trains) Calibration
Shipment documentation
Acceptance
Inspection or autoradiograph
Acceptance
Well-chamber
Acceptance
(Adapted from Kutcher et al. 1994; Nath et al. 1997; and Thomadsen 2000).
Appendix Worksheets for specifying the sources to be used in an implant and in documenting the inventory. (Courtesy of Dr. Wayne Butler, Wheeling Hospital, Wheeling, WV.)
5–Quality Management of Low Dose Rate Brachytherapy Sources
53
Form 1: Radioactive Materials Requistion
Radioactive Materials Requisition Patient: ____________________________________
Room: ____________________
( ) Cesium - 137 Tandem Loading (mg Ra eq.)
Upper, inner end ________
________
________
________
________
________
________
________
Outer end ________
________
Color
————— Spacers —————
10 = green
—————
15 = yellow
—————
20 = orange
————— Color
Ovoid Loading Right _________ ________
(mg Ra eq.)
Left __________ ________
( ) Iridium - 192 Number of Ribbons
No. of Seeds per Ribbon
Seed Spacing (mm)
Activity per seed (mg Ra eq) (mCi)
Ribbon Color
Date and Time needed: ________________________________ Radiation Oncologist: ___________________________________ Date: _______________________ Requisition filled by: ___________________________________ Date: _______________________ Checked by: __________________________________________ Date: _______________________
54
Gary A. Ezzell Form 1: Radioactive Materials Requistion (continued)
Radioactive Materials Requisition Iridium - 192 (>4 ribbon types) Patient: ____________________________________ Number of
No. of Seeds
Seed Spacing
Ribbons
per Ribbon
mm
Room: ______________________ Activity per seed
mg Ra eq
mCi
Ribbon Color
Date and Time needed: ________________________________ Radiation Oncologist: ___________________________________ Date: _______________________ Requisition filled by: ___________________________________ Date: _______________________ Checked by: __________________________________________ Date: _______________________
5–Quality Management of Low Dose Rate Brachytherapy Sources Form 2: Cesium In-Use Inventory
Cesium In-Use Inventory
Patient: ________________________________________________
Room: _____________
Sources removed: ________________________________________
Date: _____________
________________________________________
Time: ____________
Cesium in Hot-Cell Storage Safe Quantity After Removal
Quantity After Return
Color Code
Quantity Stocked
Identity
Green
6
10 mg Radium equivalent
Yellow
6
15 mg Radium equivalent
Orange
4
20 mg Radium equivalent
←
Authorized Handler Initials
→
Sources returned: _________________________________________
Date: _____________
_________________________________________
Time: _____________
*********************** Authorized Handlers:
(List here)
55
56
Gary A. Ezzell Form 3: Iridium In-Use Inventory Iridium-192 Inventory Log Receipt Manufacturer: _____________________ Received by: _________________ Survey instrument check
Package Damaged?
Package Wipe Test <6.6kdpm/300cm2
Date: _____________
Transport Index
Time: ___________
Exp. Rate @ 1m mR/hr
Max. Surface mR/hr
Seed Calibration Lot Number / Ribbon color
Calibration Date
Calculated Activity (t _ = 73.83 d)
Activity per seed (mg Ra eq)
Upon Receipt days / D.F.
At start of treatment:
mg Ra eq
mCi
days / D.F.
mCi
Inventory for Implant Ribbon Color
Number of Ribbons
No. of Seeds per Ribbon
Act. / Ribbon (mCi)
Dose Calibrator Check (mCi)
Total Number of Seeds
Total Activity (mCi)
Totals
Storage and Use Inventory In Storage Date
Time
No.
mCi
Removed No.
mCi
Returned No.
mCi
Patient/Room
Init
Seed Return Shipper:
Date:
Total number of seeds
Total Activity:
mCi =
Exposure @ 1 meter:
mR/hr
Max. Surface Activity:
mR/hr
Act./Seed:
Transport Index:
*********************** Authorized Handlers: (list here)
mCi GBq
5–Quality Management of Low Dose Rate Brachytherapy Sources
57
References Butler, W. M., A. T. Dorsey, K. R. Nelson, and G. S. Merrick. (1998). “Quality assurance calibration of 125I Rapid Strand in a sterile environment.” Int J Radiat Oncol Biol Phys 41:217–222. Feygelman. V., B. K. Noriega, R. M. Sanders, and J. L. Friedland. (1996). “A simple method for verifying activity of iodine-125 seeds in rigid absorbable suture.” Med Dosim 21: 261–262. Kutcher, G. J., L. Coia, M. Gillin, W. F. Hanson, S. Leibel, R. J. Morton, J. R. Palta, J. A. Purdy, L. R. Reinstein, G. K. Svensson, M. Weller, and L. Wingfield. (1994). “Comprehensive QA for radiation oncology: Report of the AAPM Radiation Therapy Committee Task Group No. 40.” Med Phys 21:581–615. Also available as AAPM Report No. 46. Lee, P. C., S. J. Starr, K. Zuhlke, and B. J. Moran. (1999). “Comparisons of a proposed five-seed assay method with the single-seed and batch assay methods for I-125 seeds in ultrasound-guided prostate implants.” Radiat Oncol Invest 7:374–381. Mellenberg, D. E., and E. C. Pennington. (1999). “103Pd loaded cartridge air kerma strength verification.” Med Dosim 24:73–75. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24:1557–1598. Also available as AAPM Report No. 59. Thomadsen, B. R. Achieving Quality in Brachytherapy. Philadelphia: Institute of Physics Publishing, 2000. Thomadsen, B., L. DeWerd, T. McNutt, S. DeWerd, and D. Schmidt. (1999). “Assessment of the strength of individual 192Ir seeds in ribbons.” Med Phys 26:2471–2475.
Chapter 6
High Dose Rate Sources and Delivery Systems Rupak K. Das, Ph.D., and Bruce R. Thomadsen, Ph.D. Department of Human Oncology University of Wisconsin, Madison, WI Introduction Features of Remote Afterloaders Currently Available . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Components of a High Dose Rate Remote Afterloader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Shielded Safe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Radioactive Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Source Drive Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Indexer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Transfer Tubes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Treatment Control Station . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 Treatment Control Panel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Safety Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Emergency Switches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Emergency Crank . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Door Interlock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Audio/Visual System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Radiation Monitor and TREATMENT ON Indicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Emergency Service Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Backup Battery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Shielding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Quality Assurance (QA) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Verification of Dose Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Verification of Position Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Verification of Proper Operation of Safety Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Advantages and Disadvantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Advantages of HDR Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Dose Reduction to Normal Tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Application Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Outpatient Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Disadvantages of HDR Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Investment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Complexity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Compressed Time Frame . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Radiobiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Introduction High dose rate (HDR) brachytherapy devices have become common with the advent of methods to deliver the dose at a much higher dose rates, up to 700 cGy/min at 1 cm from the source. All HDR treatments are temporary and treatments are administered using discrete fractions. Treatment is delivered by a remote afterloader (RAL), which is a computer-driven system that transports the radioactive source from a
59
60
Rupak K. Das and Bruce R. Thomadsen
shielded safe into the applicator placed in the patient. Upon termination or interruption of the treatment the source is driven back to its safe. The device may move the source by one of several methods, most commonly pneumatic pressure or cable drives. The currently available HDR RALs use stepping-source technology, which is a particular design of the treatment unit that consists of a single source at the end of a cable that moves the source through applicators placed in the treated volume. The treatment unit can treat implants consisting of many needles or catheters in the patient. Multiple catheters are often required to cover the target with adequate radiation doses. Each catheter or part of an applicator is connected to the RAL through a channel. The computer drives the cable so that the source moves from the safe through a given channel to the programmed position in the applicator (dwell position) for a specific amount of time (dwell time). In any applicator, there may be many dwell positions. After treating all the positions in a given catheter (channel), the source is retracted to its safe and then driven to the next channel. The dwell positions and the dwell time in each channel are independently programmable, thereby giving a high level of flexibility of dose delivery.
Features of Remote Afterloaders Currently Available Currently, there are three HDR RALs available in the market: The GammaMed (Figure 1) and VariSource™ (Figure 2), both marketed by Varian Associates (Palo Alto, CA) and microSelectron (Figure 3), marketed by Nucletron (Veenendaal, The Netherlands). Table 1 summarizes the specific features of the three currently marketed HDR RALs.
Components of a High Dose Rate Remote Afterloader All available HDR RALs consist of the same general components even though they differ somewhat in their technological details as shown in Table 1. Figure 4 gives a general overview of the systems, with the major parts described below.
Shielded Safe A stepping-source remote afterloader uses a 192Ir source of 5 to 10 Ci to provide a dose rate up to 700 cGy/min at 1 cm from the source. To house this highly radioactive source, a shielded safe made of tungsten or depleted uranium of sufficient thickness to provide enough radiation shielding is an integral part Table 1. Specific Features of Three HDR RALs
Vendor
MicroSelectron V2
GammaMed+
VariSource 200/200t
Nucletron
Varian
Varian
192
192
10 Ci of 192Ir
Sources
10 Ci of
Source Dimension
3.5 mm L, 1.1 mm OD
4.52 mm L, 0.9 mm OD
5 mm L, 0.59 mm OD
Source Cycle
25,000 transfer
5000 transfer
5000 transfer
Channels
18
2, 3, or 24
20
Source Extension
1500 mm
1300 mm
1500 mm
Channel Length
Variable
Fixed
Variable
Source Movement
Stepping forward
Stepping backward
Stepping backward
Step sizes
2.5, 5, or 10 mm
1–10 mm, 1 mm steps
2–99 mm, 1 mm steps
Dwells/channel
48
60
20
Ir
10 Ci of
Ir
06–High Dose Rate Sources and Delivery Systems
Figure 1. The Varian VariSource™ remote afterloader.
Figure 2. The Nucletron MicroSelectron V2 high dose rate remote afterloader.
61
62
Rupak K. Das and Bruce R. Thomadsen
Figure 3. The Varian GammaMed remote afterloader.
Figure 4. Schematic diagram of a single stepping-source remote afterloading device. (Courtesy of Nucletron Corporation, Columbia, MD.)
06–High Dose Rate Sources and Delivery Systems
63
of the treatment unit. When the source is not in treatment mode, the shielding reduces the air kerma rate to 1 to 4 µGy/h on the surface of the safe. Once in treatment mode, the source is driven out of the safe while it follows the program through the dwell positions. In the event of an interruption or termination of the treatment, the source is driven back to the shielded safe.
Radioactive Source While delivering a high dose rate requires an intense source, passing the source through needles placed through a tumor requires a source of a small size. The radionuclide used for all HDR RALs now marketed is 192Ir, although early versions of HDR RAL used 60Co. Since 192Ir has a high specific activity of about 450 Ci/g and effective gamma-ray energy of 0.38 MeV, a 10 Ci sources made out of this radioactive material can be smaller and easier to shield compared to 60Co or 137Cs. The radioactive source in an HDR RAL is usually 3 to 10 mm in length and less than 1 mm in diameter, fixed at the end of a steel cable (Figure 5). The Nucletron source is placed in a stainless steel capsule and welded to the cable, while the Varian source is placed in a hole drilled into the cable and closed by welding. Figure 5 shows an older version of the VariSource with a length of 10 mm. Currently, Varian markets a source with a length of 5 mm. Since 192 Ir has a half-life of only 74 days, these sources need to be changed every 3 months to keep the treatment in the HDR range. As required by national or state regulations, physicists need to calibrate the source after each installation using a well-type chamber and an electrometer, both of which have been calibrated by Accredited Dosimetry Calibration Laboratories (ADCLs). The resulting source calibration should be verified against the vendor’s calibration and the two values should be within 3% of each other (AAPM 1994). For clinical implementation, the user can then either use the manufacturer’s or the institution’s source strength.
Source Drive Mechanism When the machine receives a command to start a treatment, the check cable (an exact duplicate of the radioactive source along with its cable, except not radioactive) stepper motor drives the check cable to the programmed length plus a couple of millimeters to verify the integrity of the system. This consists of checking the proper attachment of the transfer tube to the indexer ring and also the attachment of the transfer tube to the applicator or catheters used for the treatment. A noneventful run of the check cable initiates the source cable stepper motor connected to the reel containing the source drive cable to turn. This causes the source cable to advance from the shielded safe along a path constrained by transfer tubes to the first
Figure 5. Schematics of the two types of sources used in stepping-source remote afterloaders.
64
Rupak K. Das and Bruce R. Thomadsen
treated dwell position in the applicator attached to the first channel. The source dwells at that position for a predetermined duration (dwell time) as calculated by the treatment-planning system. After completing that dwell it goes on to the subsequent dwell positions. Some units step as the source drives out (microSelectron), stopping first at the dwell position most proximal to the afterloader, while in the others (VariSource and GammaMed) the source travels first to the most distal dwell (towards the tip of the applicator), and a bit farther, and then steps as the source returns towards the safe. Stepping on the outward drive obviates any concern about the effect of slack in the drive mechanism affecting the accuracy of the source position. The unit that steps on the way back into the unit includes correction for slack in the calibration of the source location. Upon completion of the treatment for the first channel, the source is retracted into the safe, and redirected to travel to the second channel after a safety check run by the check cable. The process is repeated for all the subsequent treatment channels. The programmed movement of the source is verified by means of an optical encoder or other devices that compare the angular rotation of the stepper motor or cable length ejected or retracted with the number of pulses sent to the drive motor. This system is capable of detecting catheter obstruction or constriction as increased friction in the cable movement. Under certain fault conditions, such as if the stepper motor fails to retract the source, a hightorque, direct-current (DC) emergency motor will retract the source. The confirmation of the source exit from and return to the safe is carried out by an “opto-pair,” consisting of a light-sensitive detector and an infrared light source, which detects the cable when its tip obstructs the light path.
Indexer The indexer (Figures 6 and 7) is the part of the RAL that directs the source cable from the exit of the safe to one of the exit ports from the unit (channels). This mechanism gives the option of connecting a number of catheters (N = 2, 3, 18, 20, or 24) to the RAL. Items 10 to 14 of Figure 4 show the details of the Nucletron HDR RAL indexer. The various catheters or applicator parts connect to these channels, usually through connecting guides called “transfer tubes.” Different units have between 2 and 24 channels available for connection. If a patient’s treatment requires more than the number of channels on a given treatment unit, the treatment must be broken into sessions, where the catheters are connected up to the number of channels available and treated. Then the transfer tubes are disconnected from the catheters just treated and reconnected to the next set of catheters for continuation of the treatment. If a treatment plan contains more catheters than the maximum channels the machine can handle, the computer treatment-planning system breaks the plan into two or three sessions as needed.
Transfer Tubes As the name suggests, transfer tubes are long tubes that act as a conduit to transfer the source from the RAL to the applicators or catheters for treatment. One end of the transfer tube is attached to the indexer of the RAL (Figure 8), while the other end is attached to the interstitial, intracavitary or transluminal applicators (Figure 9). Often, the applicator end of the transfer tube contains spring-loaded ball bearings that block the path through the tube if no applicator is attached. When an applicator is inserted, it pushes aside the ball bearings, opening the path for the source cable. When the check cable makes its test run, if no applicator is attached to the transfer tube, the check cable hits the obstacle of the ball bearings, and prevents ejection of the source. Each type of applicator has its own type of transfer tube.
Treatment Control Station The treatment control station (Figure 10) allows the user to select the dwell positions and dwell times to be used in each channel. There are three ways of entering this information: (1) manually at the control
06–High Dose Rate Sources and Delivery Systems
65
station by the keyboard/mouse; (2) recalling a standard plan from the computer which has been saved before and then editing the data without affecting the standard plan from which it originated; or (3) by importing the data from a treatment planning system via transfer medium [floppy disk or compact disk (CD)] or a local area network (LAN) connection to the treatment control station.
Treatment Control Panel The treatment control station transfers the data to the treatment control panel. A hard or soft START button initiates the execution of the treatment according to the program. In addition, there is an INTERRUPT button, which when pressed retracts the source and stops the timer, allowing the user to enter the treatment room without receiving radiation exposure. A RESUME or START button resumes the treatment from the time and the dwell position where it was interrupted. A master EMERGENCY OFF button initiates the high-torque DC emergency motor to retract the source. In the normal course of a successful termination of the treatment, the timer runs to zero and the machine automatically retracts the source. Figure 11 shows an example of the treatment control panel.
Figure 6. Nucletron’s MicroSelectron high dose rate remote afterloader, consisting of 18 channels.
Figure 7. Varian’s VariSource™ indexer with 20 channels.
66
Rupak K. Das and Bruce R. Thomadsen
Figure 8. A view of two types of transfer tubes hooked up to the indexer of a remote afterloader.
Figure 9. View of the transfer tubes connected to a gynecological applicator. Ball bearings beneath the grey polymer coating allow verification of the proper connection of the transfer tubes to the applicator. The numbers 1 and 2 represent that these transfer tubes must connect the channels 1 and 2 of the indexer ring and the similarly numbered parts of the applicator.
Figure 10. A view of the monitor of the treatment control station.
06–High Dose Rate Sources and Delivery Systems
67
Safety Features HDR RALs are complicated devices containing very high activity radioactive sources. Serious accidents can happen quickly. All such units have many safety features and operational interlocks to prevent errant source movement or to facilitate rapid operator response in the event of a system failure.
Emergency Switches Numerous EMERGENCY OFF switches are located at convenient places and easily accessible, in case a situation arises. One EMERGENCY OFF switch is located on the control panel. Another EMERGENCY OFF button is located on the top of the RAL treatment head. Vendors usually install one or two emergency switches in the walls of the treatment room. In the event a treatment is initiated with someone other than the patient in the treatment room, that person can stop the treatment and retract the source by pressing the EMERGENCY OFF button. Figure 11 shows the EMERGENCY OFF switch on the treatment control panel.
Emergency Crank All HDR RALs have emergency cranks to retract the source cable manually if the source fails to retract normally and the emergency motor also fails to reel in the source. Figure 12 shows such a crank for the microSelectron and Figure 13 for the VariSource. Using the crank requires the operator to enter the room with the source unshielded. Exposure rates for this situation are considered below.
Door Interlock Interlock switches prevent initiation of a treatment with the door open. When a treatment is in progress, opening the door interrupts the treatment. This safety feature protects the medical personnel from radiation exposure in the event somebody enters the treatment room without the knowledge of the operator. If a door is inadvertently opened during the treatment, the treatment is interrupted and the source returns to the safe. The treatment can be resumed at the same point where it was interrupted by closing the door and pressing the START or the RESUME button at the control panel.
Figure 11. The Treatment Control Panel of the Nucletron, MicroSelectron afterloader. The START button is the button on the right side of the panel, inside the frame, while the EMERGENCY OFF button is the large button on the extreme left side of the panel.
68
Rupak K. Das and Bruce R. Thomadsen
Figure 12. A view of the access panel of the microSelectron treatment unit. The EMERGENCY OFF button (in the center of the box at the bottom of the picture) on the unit is to withdraw the source into the shielded safe. Also shown are the manual retraction cranks for the radioactive source cable on either side of the keyhole in the middle, for the source on the left and the check cable on the right.
Figure 13. The back panel of the VariSource™ showing the crank for manual source retraction in an emergency.
06–High Dose Rate Sources and Delivery Systems
69
Audio/Visual System All HDR suites are equipped with a closed-circuit television system (CCTV) or shielded windows and/or mirrors for observing the patient, and a two-way audio system to communicate with the patient during treatment.
Radiation Monitor and TREATMENT ON Indicator Three separate independent systems alert personnel when the source is not shielded. One radiation detector is part of the treatment unit and indicates on the control panel when it detects radiation. An independent unit, usually mounted on the treatment room wall with displays both inside and outside the room, also alerts the operator and other personnel when the radioactive source is out of the safe. A TREATMENT ON indicator outside the room, activated when the source passes the reference optical pair discussed above and shown in Figure 4, also indicates that a treatment is in progress.
Emergency Service Instruments In the event the radioactive source fails to retract after termination, interruption, pushing the EMERGENCY switch or cranking the stepper motor manually, the immediate priority is to remove the source from the patient. Table 2 gives the exposure rates at various distances from a 10 Ci 192Ir source. The table shows that the dose to the patient, with the source in contact, can cause injury in a very short time. On the other hand, the operator, working at a greater distance, is unlikely to receive a dose exceeding regulatory limits for a year, let alone one that would cause health problems. Once the source is removed from the patient and moved to distance of even a meter, the exposure rate is quite low, and whatever actions need be taken to remove the patient from the room can be performed safely. Most institutions set the effective annual limit to the body at 10 times less than the U.S. Nuclear Regulatory Commission (NRC) limit of 0.05 Sv in keeping with the principle to keep exposures as low as reasonably achievable. Ideally, such a dose should not be received in one, short exposure. The allowed exposure to the hands is 10 times that to the body. The preferred approach to a source that cannot be retracted by any method is to remove the applicator from the patient as quickly as possible and to place the applicator containing the source in a shielded container (Figure 14). If it is clear that the cable is caught in the transfer tube and not in the applicator OFF
Table 2. Exposure Rates from an Exposed 10 Ci 192Ir Source
Typical situation
Distance [m]
Dose equivalent rates [Sv/h]
Time to receive 10 Sv
0.05 Sv
(Likely injury)
(Annual body limit) 0.007 minutes 0.4 seconds 0.67 minutes
In Patient
0.01
4.6R/(mCi h) × 0.966 rem/R = 444 Sv/h
1.35 minutes
Handling with Kelly Clamps
0.1
4.44
2.3 hours
(6.8 minutes for hand limit)
Handling with Kelly Clamps
0.3
0.5
20 hours
0.10 hour 6 minutes
Standing near
1
0.044
9.5 days
1.1 hours
Standing far
2
0.011
37.5 days
4.5 hours
70
Rupak K. Das and Bruce R. Thomadsen
Figure 14. The shielded container for emergency placement of an unretracted source.
itself, the applicator or catheter may be disconnected from the transfer tube and the source pulled from the applicator. In some cases, this will be faster than removing the applicator. The reason to avoid disconnecting the applicator from the transfer tube is that a source may stay in the applicator if the source capsule shatters. In that case, removing the applicator attached to the transfer tube keeps the system closed, while disconnecting the two opens a path for parts of a broken source to fall from the applicator into body cavities or crevices, or roll onto the floor. A situation may arise when the source needs to be detached manually from the treatment unit. One (still unlikely) scenario would be if the source were stuck out of the treatment unit, the sources or the closed applicator had been removed from the patient, some person were pinned very close to the source so neither they, nor the treatment unit, could be moved, and the source on the cable could not reach the shielded container. In this special situation, the source cable should be cut from the unit and the source placed in the shielded container always present in the room. In cutting the source cable, it must be clear that the cut is not through the source capsule. For units with the capsule welded on the cable, the cut must be through the braided cable as opposed to the smooth steel capsule (Figure 15). For sources imbedded in the cable, a sufficient length of the cable must be seen to assure the cut occurs behind the source. Thus, emergency tools that must be present in the treatment room and always readily accessible include a wire cutter, a pair of forceps, and a shielded service container.
Backup Battery In case of a power failure during the treatment, the machine is equipped with a backup battery to provide retraction of the source to its safe. The batteries should be tested with each source change.
06–High Dose Rate Sources and Delivery Systems
71
Figure 15. Cutting the source cable from a treatment unit. This procedure should only be performed in very special, rare situations as described in the text. Great care must be taken to ensure the cut is through the cable and not the source capsule.
Shielding The radioactive source in the HDR machine starts about 10 Ci with an exposure rate at a distance of 1 m from the source of about 46 mSv/h. According to the rules and regulations of the U. S. NRC (USNRC 1993), the annual limit for radiation exposure to the public is 1 mSv, and the annual occupational limit is 50 mSv. Following the principle of maintaining exposures as low as reasonably achievable, the NRC usually expects licensees to create warning alert levels effectively to limit occupational personnel exposures to 10% of the NRC limit, which becomes 5 mSv for whole body exposure. In addition to the annual limit, the NRC requires that in an unrestricted area, the dose equivalent rate may not exceed 0.02 mSv in any hour. Thus, the HDR machine needs to be housed in an adequately shielded room. To meet these requirements in an HDR suite, where the walls and the ceiling are at least 5 feet from the machine head, concrete walls of about 43 to 50 cm (or 4 to 5 cm of lead) are needed. For larger rooms the concrete wall thickness will be lower since the exposure rate is inversely proportional to the square of the distance from the radioactive source. Details on the procedures for calculating the thickness of barriers for a particular facility are elaborated in Cember (1996) and McGinley (2002).
Quality Assurance (QA) In order to maintain the quality of patient care, a quality management program is required in every facility that provides HDR treatment. Such a program generally follows standards set by professional organizations and intends to minimize untoward events caused by the malfunction of the machine or human error. Such programs become exceedingly important in HDR brachytherapy because the planning and the treatments tend to happen very quickly, increasing the likelihood of accidents and mistakes. QA tests measure some performance aspect of the treatment unit and compare the results with expectations in order to demonstrate proper operation. QA is performed at various intervals: some for each patient, some once each treatment day, and others with each source change. Moreover, for HDR machines, the USNRC (2003) mandates that users meet certain standards, including education and training on operating the machine, emergency procedures, radiation monitoring, pretreatment safety checks, safe and accurate delivery of the treatment, and monthly/initial calibration of the source. The details of QA are outside the scope of this chapter and are covered in the next. Interested readers can refer to the reports of Task Groups 59 (AAPM
72
Rupak K. Das and Bruce R. Thomadsen
1998) and 56 (AAPM 1997) of the American Association of Physicists in Medicine (AAPM) and relevant texts (Thomadsen 1999). In general, the problem of QA is ensuring that the treatment will deliver the correct dose, to the correct location, safely. Thus, the QA tests generally follow the outline below: 1. Verification of dose variables a. Checking the strength of the source compared with that projected from the initial calibration based on radioactive decay. b. Checking the proper operation of the controlling timer. 2. Verification of position control a. Checking that the source goes to the location programmed. b. Checking coincidence between the programmed positions and the respective positions indicated by imaging markers. c. Checking consistent movement of the source. 3. Verification of proper operation of safety features a. Checking operation of the door interlocks. b. Checking the operation of a handheld radiation detector. c. Checking the operation of the on-board and on-wall radiation detectors. d. Checking the operation of the check cable runs and interlocks. e. Checking the operation of the EMERGENCY OFF and TREATMENT INTERRUPT buttons.
Costs Currently two vendors (Nucletron Corporation and Varian Medical Systems) market their RAL treatment unit in the United States. Both the devices require a capital expenditure of about $0.5M to $1.0M, which includes the treatment unit and a variety of transfer tubes, along with the software and hardware for the treatment-planning system. Applicators that are needed for placement in the tumor cost extra. The costs of preparing a shielded room along with ancillary equipment for x-ray imaging and operating room procedures can be another $0.5M to $0.75M. Hence the total cost can run in between $1M and $2M.
Advantages and Disadvantages HDR brachytherapy compared with low dose rate (LDR) brachytherapy offers several advantages and disadvantages. Being aware of these permits safe and effective operation and application of HDR brachytherapy.
Advantages of HDR Brachytherapy Safety One of the major advantages of an RAL is the reduction or elimination of radiation exposure to the radiotherapy staff. In conventional LDR manual afterloading, the radiotherapy staff receives radiation exposure while loading the applicators with the radioactive sources, and the nursing personnel are exposed during
06–High Dose Rate Sources and Delivery Systems
73
patient care through the duration of the treatment (1 to 4 days). With either HDR or LDR remote afterloading, the radiotherapy personnel are outside the shielded room during the treatment and, hence, are exposed to minimal radiation. Optimization The design of the HDR RAL with the stepping source allows greater flexibility and control over dose distribution. The stepping-source allows optimization of the dose distribution by adjustment of the dwell times for each dwell position in each channel. The dwell times can be varied almost infinitely, permitting very fine control of the dose distribution. In LDR brachytherapy, either manual or using an RAL, the finite number of activities available (usually four at most) and the larger sources used with manual applications impose a restriction on the ability to conform the dose distribution to the target. Stability Because HDR intracavitary treatments take so little time (about an hour), applicators can be fixed in place much more stably than for the multi-day treatments using LDR brachytherapy. Dose Reduction to Normal Tissue As with stability, the short duration of HDR intracavitary treatments allows displacement of normal tissue structure (i.e., pushing them away from the source paths) to a greater extent than with LDR treatment. Applicator Size The small size of the HDR source permits the use of smaller applicators than those required for the LDR applications, increasing the comfort to the patient. Outpatient Treatment Almost all HDR patients are treated on an outpatient basis compared to LDR patients, who usually are treated as inpatients. Outpatient treatment is more convenient for the patients and sometimes results in lower overall costs.
Disadvantages of HDR Brachytherapy Investment The initial expense of an HDR RAL is very high. Machines and site preparation costs can be anywhere between $1M and $2M. Complexity The technological complexity of HDR RALs increases the probability of errors, and leads to increased regulatory scrutiny. Compressed Time Frame As mentioned above, the rapidity with which procedures progress in HDR brachytherapy increases the probability of executing errors. Radiobiology As the dose rate increases, the radiosensitivity (damage per unit dose) increases for both normal tissues and tumors. Unfortunately, the radiosensitivity for the normal tissue increases faster than that for tumors,
74
Rupak K. Das and Bruce R. Thomadsen
increasing the likelihood of injuring the patient while controlling the tumor. Overcoming this radiobiological handicap requires the use of the advantages of optimization, geometry, stability, and dose reduction to normal tissues, in addition to fractionation. As with external beam radiotherapy delivered using a linear accelerator, which also operates at high dose rates, spreading the treatments over many smaller fractions delivered over several days reduces the difference in radiosensitivities between the tumor and the normal tissues.
References American Association of Physicists in Medicine (AAPM). (1994). “Comprehensive QA for radiation oncology. Report of AAPM Radiation Therapy Committee Task Group 40.” Med Phys 21:581–618. Also available as AAPM Report No. 46. American Association of Physicists in Medicine (AAPM). (1997). “Code of practice for brachytherapy physics: AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24:1557–1598. Also available as AAPM Report No. 60. American Association of Physicists in Medicine (AAPM). (1998). “High dose-rate brachytherapy treatment delivery: AAPM Radiation Therapy Committee Task Group No. 59.” Med Phys 25:375–403. Also available as AAPM Report No. 61. Cember. H. Introduction to Health Physics, 3rd ed. New York: McGraw-Hill, 1996. McGinley, P. Shielding Techniques for Radiation Oncology Facilities, 2nd ed. Madison, WI: Medical Physics Publishing, 2002. Thomadsen, B. R. Achieving Quality in Brachytherapy. Bristol: Institute of Physics Publishing, 1999. U.S. Nuclear Regulatory Commission. (2003). Code of Federal Regulations, title 10, chapter 1: Energy, Part 35: Medical Use of By-product Material. Washington D.C.: Nuclear Regulatory Commission. U.S. Nuclear Regulatory Commission. (1993). Code of Federal Regulations, title 10, chapter 1: Energy, Part 20: Standards for Protection Against Radiation. Washington D.C.: Nuclear Regulatory Commission.
Chapter 7
Quality Assurance for High Dose Rate Remote Afterloaders: Safety Standards for Clinical Implementation of High-Activity Brachytherapy Sources Mark J. Rivard, Ph.D., and Christopher S. Melhus, M.S. Department of Radiation Oncology, Tufts University School of Medicine Boston, Massachusetts Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Rationale for HDR QA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Ethics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Recommendations of AAPM TG-56 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Recommendations of AAPM TG-59 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Modern Regulatory Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 IEC 60601 and International Standards and Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 A Comprehensive Quality Assurance Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 QA for a New HDR Brachytherapy Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 QA for the Treatment Device and Related Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 New Source QA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Monthly QA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Daily QA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Treatment-related QA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 QA for Imaging-related Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Treatment-Planning System and Related Equipment QA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 Emergency Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 System Failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Physical Plant Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Minor Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Major and Medical Emergencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Staffing, Credentialing, Assignments, and Training Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Radiation Safety Officer (RSO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Physician . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Radiation Therapy Technologist (RTT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Medical Physicist . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Case Reports from NRC Public Postings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Example 1: NRC Information Notice 00-05 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Example 2: NRC Significant Enforcement Action EA-04-093 . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Example 3: NRC Significant Enforcement Action EA-04-087 . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Example 4: NRC Significant Enforcement Action EA-99-257 . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Example 5: NRC Significant Enforcement Action EA-97-284 . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Example 6: NRC Significant Enforcement Action EA-97-245 . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
Introduction History High dose rate (HDR) remote afterloaders (RALs) for brachytherapy administration have been used since the 1960s with a variety of radionuclides including 60Co, 137Cs, 252Cf, and, more recently, 192Ir. These clin-
75
76
Mark J. Rivard and Christopher S. Melhus
ical applications were largely initiated in Europe, and later adopted in the United States. Over the past four decades, efforts have been made to make the HDR sources smaller in both length and outer diameter, to make the RAL more robust, and to provide a greater assortment of clinically relevant features. Furthermore, substantial research has been performed to assess the safety and efficacy of HDR brachytherapy in comparison to brachytherapy using low dose rate (LDR) sources. To date, evidence indicates that clinical outcomes using HDR brachytherapy are at least equal if not better than when using LDR brachytherapy. While additional research is underway to develop new HDR RALs using the 169Yb (32 day half-life and slightly lower average photon energy than 192Ir) and 252Cf (2.645 year half-life and neutronemitting) radionuclides, the lion’s share of RALs in both the United States and worldwide employ 192Ir as the radioactive source. With new medical technology comes the need for new quality assurance (QA) standards and recommendations for clinical practice as first suggested in the AAPM Task Group No. 41 (TG-41) report (Glasgow et al., 1993). Therefore, HDR 192Ir RALs and associated QA will be presented in this chapter.
Rationale for HDR QA Ethics In comparison to LDR brachytherapy, treatments using HDR brachytherapy afford the following advantages: (a) ability to provide treatments quickly on an out-patient basis without the need for the patient to have an overnight hospital stay, (b) brachytherapy source positioning controlled via a computer for intensity modulation of source dwell positions and inverse treatment-planning optimization, (c) clinical outcomes at least equal if not better than LDR brachytherapy, (d) decrease in radiation exposure to hospital personnel, (e) electronic and/or paper record of HDR brachytherapy therapy delivery via the control console whereas LDR brachytherapy is generally performed manually, and (f) financial reimbursement for HDR brachytherapy is much more lucrative than for LDR brachytherapy. This reality offsets the substantial startup costs (to be discussed later) of an HDR brachytherapy program. With these and other advantages, HDR brachytherapy is often favored over LDR brachytherapy. However, there are also several disadvantages of HDR brachytherapy. Some of these are listed below: (a) reduction of the radiobiological advantage in comparison to LDR brachytherapy, (b) increase in complexity of the HDR treatment planning and RAL systems, which increases the time needed for training, and (c) compression of the time frame for dose delivery, which further reduces the margin of error. The AAPM TG-59 report (Kubo et al., 1998) includes a thorough discussion of these and other characteristics of HDR brachytherapy. Further ethical concerns focus on the need for the medical physicist to develop a quality management program (QMP) and to perform requisite QA to assure that this high technology is put to good use. Fortunately, many recommendations are already in place that outline the standards of care for HDR QA and treatment delivery.
7–Quality Assurance for High Dose Rate Remote Afterloaders
77
Recommendations of AAPM TG-56 In 1997, the American Association of Physicists in Medicine (AAPM) published the “Code of Practice for Brachytherapy Physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56” (Nath et al., 1997). Part I of the TG-56 report is addressed to radiation oncology administrators and is intended to generally describe the complex and specific requirements of a brachytherapy program while delineating the critical role of the medical physicist in establishing and maintaining such a program. For this chapter, Parts II and VI is the meat of the TG-56 report, outlining specifics on the scope and tolerances for general brachytherapy implementation. Specifically, seven tables are included which delineate various procedures, endpoints, their frequency of performance, and the test methodology as applicable to brachytherapy, with Tables V and VI specific to remote afterloading. Many of these tasks are similar in spirit to the QA oversight one would implement for a teletherapy QA program, and the reader is prompted to carefully review these AAPM-approved recommendations. Practical implementation of these and other recommendations are made in the enclosed QA forms for HDR brachytherapy. Recommendations of AAPM TG-59 In comparison to TG-56 which focused on the code of practice for HDR and LDR brachytherapy, the TG-59 report “High Dose-Rate Brachytherapy Treatment Delivery: Report of the AAPM Radiation Therapy Committee Task Group No. 59” was prepared to aid the medical physicist in ensuring safe delivery of HDR treatments (Kubo et al., 1998). Towards that goal, the TG-59 report describes the principles of an HDR brachytherapy program; the staff required to carry out an HDR brachytherapy program and their related training; recommendations for treatment-specific QA; and emergency procedures for HDR RALs. These topics are expanded upon in the sections below through example forms incorporating these recommendations. Modern Regulatory Requirements Today, computers play an even more important role in HDR brachytherapy delivery, there are more features available than when the aforementioned AAPM task group reports were released, and there are additional requirements included in a number of U. S. Nuclear Regulatory Commission (NRC) and state regulations. Thus, the medical physicist is ethically bound to consider including QA oversight of these new features into the current QMP paradigm. Examples of these features largely unused in HDR brachytherapy during the late 1990s were: image fusion capabilities and implant localization in conjunction with teletherapy (Lerma and Williamson 2002). Until new, national recommendations are made which specifically include these features and related functionality, it is recommended that each physicist propose local QA criteria for assuring sophisticated HDR brachytherapy using these features is delivered safely. IEC 60601 and International Standards and Requirements Though not strictly of interest to the medical physicist, it is worth noting that the International Electrotechnical Commission (IEC) has established a worldwide set of standards to promote consistency of electrical equipment testing and design conformity (IEC). Medical device manufacturers follow these standards for the safe design and use of their equipment. Such “upstream” efforts in equipment design obviously impact the results of clinical implementation and machine QA. Specific to this chapter, IEC report 60601-2-17 Amendment 2 covers: Medical Electrical Equipment-Part 2: Particular Requirements for the Safety of Remote-Controlled Automatically Driven Gamma-Ray After-Loading Equipment (IEC 60601-2-17). In the United States this report has been adopted as UL 60601, with slight deviations specific to the U.S. market and infrastructure (UL 60601).
78
Mark J. Rivard and Christopher S. Melhus
A Comprehensive Quality Assurance Program Having a comprehensive QA program demands that details resultant from a QMP be carefully thought through, and that these QA tests be coordinated such that complete coverage from all possible perspectives is provided, yet overlap is minimized so as to prevent wasting limited resources like the physicist’s time. Toward coordinating these efforts, one may divide the scope of HDR brachytherapy into various categories as presented below.
QA for a New HDR Brachytherapy Program Unlike the extensive tests used to acceptance test and commission a new linear accelerator (linac), the efforts needed to start a new HDR brachytherapy program are less involved. This is largely due to the output of the source being more constrained than that of a linac. Specifically, the physicist primarily measures the source strength and the relative three-dimensional (3-D) dose distribution of the source. In contrast to linac-based radiotherapy, the 3-D dose distribution in HDR brachytherapy is fixed due to radiation emissions being radionuclide based. For a linac, there are many more beam-modifying features, and radiation output is much more variable among manufacturers. Starting a new HDR brachytherapy program, initially one will need to address room shielding design, licensing, equipment storage, maintenance, contractual agreements, patient/staff ergonomics and data management processes, and departmental policies, etc. These efforts are quite specific to the individual institution. Advice on room shielding design can be found in chapter 10, and supplemental recommendations can be found in the AAPM TG-41 (Glasgow et al.. 1993), TG-56 (Nath et al., 1997), and TG-59 (Kubo et al., 1998) reports. For the established HDR brachytherapy program in steady-state, the following are QA guidelines and practical examples.
QA for the Treatment Device and Related Equipment New Source QA Few if any users in the United States are shipped a high-activity 192Ir source for source exchange by the medical physicist. Manufacturers typically contact the physicist preceding source shipment to coordinate the source exchange. The new source arrives, is surveyed on the institution loading dock, the package is wipe tested, moved to a secured location (e.g., radionuclide storage room), added to the institution’s radioactive inventory, and typically awaits arrival of the manufacturer field service engineer a couple of days later. This service engineer will swap out the ~3 month old (i.e., 4 Ci) HDR 192Ir source into a transport pig for return back to the manufacturer, then perform planned maintenance and inspection of the RAL and control console. Sometimes emergency procedure training is provided during this service phase. Afterwards, the new HDR 192Ir source (i.e., 10 Ci) will be transferred into the RAL, and the service engineer will ask the physicist to review the work performed, before leaving the institution. It is then the physicist’s responsibility to arrange for shipment of the weaker HDR 192Ir source back to the manufacturer for decay/disposal, and to oversee a radiation survey of the environment with the new device in place (Figure 1). With these tasks completed, the physicist must perform a set of complex QA tasks needed to characterize the output of the source and behavior of the RAL system and control console. Instead of including a lengthy description of these tasks, a New Source QA form is included (Figure 2), which largely includes the tasks recommended by the aforementioned AAPM task groups and some specifics to the Tufts-New England Medical Center (Tufts-NEMC) radioactive materials license (RML) for HDR 192Ir. Note the offset margins which permit double-sided printing. Not apparent in this printed form are the hyperlinks that facilitate mathematical analysis of the input data in light of the reported tolerances. The practice of keeping
7–Quality Assurance for High Dose Rate Remote Afterloaders
79
Tufts-New England Medical Center Department of Radiation Oncology New HDR 192Ir Source QA Survey Form
192
Ir Source Serial # ___________________ Mallinckrodt Activity _______________ U Assay Date _____________ Exposure Rates Near Treatment Unit (mR/h) (orientation facing indexer, source retracted in safe) @ surface
@ 1 meter
Right________________
Right ______________
Left ________________
Left _______________
Front _______________
Front ______________
Rear ________________
Rear ______________
Above ______________
Above _____________
Below ______________
Below _____________
Exposure Rates @ Adjacent Areas (mR/h) (source extended, ask Therapist or Medical Physicist to operate unit) Entrance Door ___________________________________ HDR Control Console _____________________________ Main Hallway ___________________________________ Back Hallway ____________________________________ Simulator Control Room ___________________________ Proger 1 Room___________________________________
Survey Meter (make / model / Serial #) ____________________________________ Calibration Date ____________ Health Physicist name________________________Signature _____________________ Date __________________
Figure 1. QA New Source survey form.doc.
80
Mark J. Rivard and Christopher S. Melhus
Tufts-New England Medical Center, Department of Radiation Oncology New HDR 192Ir Source QA Autoradiograph Tape clear 6F endobronchial catheters to readypack V-film with two GYN dummy #1 inserted. Place on simulator table at 100 cm SFD with 4 x 25 cm2 field size (no wires present) and take 3 exposures (small focus, 75 kVp, 320 mAs). Remove dummy markers and bring to HDR unit. Program HDR control console with step=5.0 mm; L=995 mm (separately); and t=0.2 second for positions below (tIRR = 3.0 s total). Autoradiograph film without bolus to minimize scatter. Process film, and analyze shifts within 0.5 mm.
Position1 1000 mm 6F Shift [mm] 1
3
1
5
7
9
3
11
5
9
11
13
15
17
13 15 17 19 21 23 25 27 29 31
19
23
25
27
29
31
33
33 35 37 39 41 43 45
ŶŶŷŶŷŶŷŶŷŶŷŶŶŷŶŷŶŷŶŷŶŷŶŷŶŷŶŷŶŷŶŷŶŶŷŶŷŶŷŶŷŶŷŶŷŶŷŶŷŷŷŷŷ Perform two autoradiographs using tdwell = 3 s and 30 s to evaluate source homogeneity.
Source homogeneity acceptable (visual assessment) ? ___________ Average measured offset _______________ mm (range 0.0 ± 1.0 mm) If offset is within range, write “NA” below. Otherwise enter the correct L (≠ 995 mm) and complete the following 3 tasks: Update daily check program with new length L Update program cards for any ongoing patients with new length L Post new length L in planning room
___________ ___________ ___________
Check of Guide Tubes GYN #1: length = __________ mm Acceptable [748.5 mm +/- 1.0 mm]? GYN #2: length = __________ mm Acceptable [748.5 mm +/- 1.0 mm]? GYN #3: length = __________ mm Acceptable [748.5 mm +/- 1.0 mm]? Is mechanical integrity of all applicators, connectors, and guide tubes acceptable based on visual inspection?
___________ ___________ ___________ ___________
Timer Offset, Linearity, and Accuracy Insert a clear, 6F catheter into HDR-1000. Setup control console with L=995, step 2.5 mm, position=18, and variable dwell time (tdwell). Use electrometer without filter, fit to straight-line, and determine toffset and time linearity (R2). Use stopwatch to independently measure tdwell(120) accuracy. If toffset, R2, or tdwell(120) are not within specifications, have second physicist review results. If they are still outside specification, contact Nucletron.
tdwell [seconds] 0.1 1.0 10.0 0.5 5.0 0.2 2.0 20.0
measured charge (2 µC scale)
what is the value of toffset ?
what is the value of R2 ?
what is the value of tdwell(120)/120 ?
is toffset –0.55 ± 0.03 s ?
is R2 > 0.99 ?
is tdwell(120)/120 1.000 ± 0.010 ?
Source Strength Calibration 1.
Setup control console with L=995 for 5F/6F caths in HDR 1000+ and 10 seconds each for positions 14-22. Use 200 nA scale on the Keithley 6517A and the AVG(10) filter function.
Position 22 21 20 19 18 17 16 15 14 Distance [mm] 942.5 945.0 947.5 950.0 952.5 955.0 957.5 960.0 962.5 Rdg Determine maximum reading (MAXRdg) ____________________ at time: ________________ 2. Calculate in-house measure of air kerma strength (NEMCSK); 192Ir t½ = 73.83 days with Ȝ = 0.009388 days–1: Decay factor (NEMCF) to 0:00 for the current day _____________> 1, typically ~ 1.0071. Measure temperature _________ [ºC] and pressure ___________ [mbar] and calculate CTP ___________ Standard Imaging HDR-1000 well chamber S#A940704 calibrated 2/27/2003 CWELL = 0.5035 mGy m2 h–1 nA–1 Keithley 6517A electrometer S#0857926 calibrated 9/15/2003 CEL = 0.999 A Rdg–1 on 200 nA scale or Keithley MK 602 (CNMC) electrometer S#34635A calibrated 1/25/2002 CEL = 0.0981 nA Rdg–1 10–8 A scale
SK = MAXRdg x NEMCF x CTP x CWELL x CEL = _____________ [mGy m2 h–1] For air kerma rate @ 1 meter (as required by license) divide above number by 10 _____________ [cGy/h] NEMC
Using a conversion factor of 2.45 Ci-h/cGy, enter Aapp [Ci] on the yellow placard on the RAL and console ________ if not already done by the Nucletron Service Engineer.
Figure 2. New Source QA.doc. page 1
7–Quality Assurance for High Dose Rate Remote Afterloaders
81
Tufts-New England Medical Center, Department of Radiation Oncology New HDR 192Ir Source QA 3. Comparison of in-house determination of air kerma strength with source strength reported by Mallinckrodt. The microSelectron remote afterloader (RAL), model # 080.000 serial # 9511, is made by Nucletron BV. The source is manufactured by Mallinckrodt Medical BV, model CI L BV. Source ID # ____________________ Calibration date __________________________ Mallinckrodt air kerma strength (VENDORSK) calibration [mGy m2 h–1] ÷ 10 ________________ [cGy m2 h–1] Decay factor (VENDORF) to Mallinckrodt calibration (X days +5 hours due to timezone differences) ____________< 1
Calibration ratio
=
NEMC VENDOR
SK
F ×VENDOR S K
=
_____________ Is this ratio 1.00 ± 0.05? ___________
If yes, 1. update source strength (NEMCSK) in the microSelectron treatment console (submode 11) compare predicted treatment time for a standardized VagCyl treatment plan with catalog 2. update source strength (NEMCSK) in the Plato v.14.1.2 planning system (passwd=hdr) 3. update source strength (NEMCSK) in the Pinnacle planning system (passwd=physics) 4. update source strength (NEMCSK) in the MammoSite Excel spreadsheet 5. make 2 new decay tables and post at the planning workstation and in HDR console logbook
___________ ___________ ___________ ___________ ___________
If ratio is not within 5% of unity, contact another Physicist or the RSO and consider cross-calibration using the IVB-1000 well chamber.
Safety Checks
(Y or N)
1. HDR 192Ir source retraction following AC power failure simulation?
___________
2. Door light and Primalert function with AC power off using Cs micrad check source? (Unplug “Connector In” from PRIMAPAK)
___________
3. Emergency STOP (on wall near console) functions with printout of irradiation time?
___________
4. Door interlock functions with printout of irradiation time?
___________
137
5. Interrupt button on the console functions with printout of irradiation time?
___________
6. Survey meter battery check and constancy check OK?
___________
7. Was survey meter calibrated within 365 days ? (If not, notify Health Physics)
___________
8. Are Emergency Procedures in HDR Log Book?
___________
9. Is error 129 detected when indexer ring is not latched?
___________
10. Is error 129 detected when applicator is not inserted?
___________
11. Is error 129 detected when interstitial interconnect tube is installed without catheter present?
___________
12. Review of daily checks since last monthly/new source QA?
___________
13. Daily checks performed? (attach and complete QA-Daily.doc)
___________
Additional Notes (if needed): Authorized Physicist Name _____________________________ Signature/Date _____________________________
Figure 2. New Source QA.doc. page 2
82
Mark J. Rivard and Christopher S. Melhus
printed copies of these forms, and hardcopies of the irradiated films, will soon be replaced with an electronic spreadsheet-based approach that is paperless (see Figure 3). Based on past inspections by the Massachusetts Radiation Control Program, sole use of electronic records is an acceptable practice. While somewhat quite painful, it is recommended by AAPM TG-56 that QA of device-related equipment (catheters, applicators, tubes, dummy wires, etc.) be performed once every 3 months. In addition to confirming device integrity and inventory, as these items occasionally get moved or are lost just when they are needed for that rare and complex procedure, one should physically determine that the device performs according to expectations. For instances, the metal nibs on a dummy marker can become uncrimped and slide along the wire. When used in vivo, misleading distances may be obtained and the whole HDR 192Ir brachytherapy procedure could be aborted, or worse yet, could proceed incorrectly due to erroneous information. Similarly, the parts for an HDR vaginal cylinder kit may get mixed up with an HDR tandem-and-ovoid kit. In addition to being misplaced, possibly in the operating room (OR) storage room, parts could become damaged if subjected to incompatible sterilization procedures (e.g., flash versus gas). Monthly QA While no longer the accepted practice in the U.S. with the advent of the revised Code of Federal Regulations 10 CFR part 35, our license demands that patients not be treated if monthly or new source QA was performed more than 31 days ago. When patients are not scheduled for treatment, there are occasions when we will actively decide to let the date for QA creep into the next month so as to exert resources elsewhere. This approach is feasible at our institution since we can generally complete the QA tasks outlined below in the Monthly QA form (Figure 4) within a matter of 4 hours since the RAL is in a vault with a seldom-used linac. Toward resources management, we are working with the Massachusetts Radiation Control Program to resolve the monthly versus quarterly QA requirements, and integration of NRC regulations into our agreement state regulations. Daily QA Based upon the level of staffing at a given facility, either the medical physicist or the radiation therapy technologist (RTT) performs daily QA tests preceding patient treatments. At some facilities, a trained radiotherapist will perform the daily QA tests and the physicist will occasionally assist the RTT team by performing the daily QA toward hastening patient treatments and coordinating departmental activities. The TG-56 report provides guidance on core daily QA tests recommended for HDR brachytherapy facilities. An example of a Daily QA checklist used at Tufts-NEMC is provided in Figure 5. Treatment-related QA A form, Treatment QA (Figure 6), has been in use for many years at Tufts-NEMC to coordinate responsibilities of all parties preceding, during, and immediately after HDR brachytherapy. This checklist approach ensures that tasks are appropriately delegated and consistently performed. We keep this completed form in the patient chart, along with a printed record of the treatment that has been signed by the physician, treatment plan, and written directive. QA for Imaging-related Equipment Simple QA tasks of image quality, spatial accuracy and orientation, and accurate data depiction can and should be performed during the patient simulation phase. Sometimes a therapist will mistakenly label a film as right anterior oblique instead of left anterior oblique, or record the wrong source-to-film distance value. These mistakes can significantly delay the treatment-planning process for the physicist, and caution versus cynicism is recommended. However, these simple tasks must supplement a rigorous, regularly performed QA program for the imaging component of the brachytherapy simulation process.
7–Quality Assurance for High Dose Rate Remote Afterloaders
83
HDR brachytherapy is a complex endeavor that relies on a multitude of imaging equipment to properly deliver the therapy. While somewhat outside the scope of this chapter, it goes without saying that “garbage-in is garbage-out” regarding the quality of imaging data used to prepare an HDR brachytherapy treatment plan. QA of the simulator unit is typically performed monthly, sometimes only quarterly. However, more sophisticated plans rely on 3-D computed tomography (CT)- and magnetic resonance imaging (MRI)-based image datasets. With physicist specialization, it is often the case that the physicist primarily involved with brachytherapy planning and delivery will not be the one performing QA of the imaging equipment. Remember that brachytherapy is limited in that it does not have the in vivo delivery confirmation feature that teletherapy does through port-films, electronic portal imaging devices, or kVpbased image-guided radiation therapy. Use of a diode integrating output or surface dose at a point is not a substitute. Treatment-Planning System and Related Equipment QA After the patient has been simulated, the clock is often ticking and the physicist is under pressure to complete the radiation treatment plan (RTP). Everyone, including the sick patient, is waiting to start the treatment. This is not the time to feel uncertain about your HDR QA program and possible inadequacies of RTP QA. At a minimum, consistency of RTP performance regarding source decay and calculation of point doses results. At our institution and many others, an atlas of preplanned HDR treatment plans is available for the purposes of QA and for facilitating quick implementation of relatively simple multifractionated brachytherapy implants. For our vaginal cylinder atlas, plans are available for a wide range of doses, cylinder diameters, and prescription points. We use air kerma strength at Tufts-NEMC, and simply taking the ratios of the source strengths permits a physicist opportunity to check the corrected treatment times (both total time and individual dwell times) with the times calculated by the HDR control console (in computer memory) and HDR RTP system. Of course, it is senseless to perform all these tests, and not document your efforts. Strive for an electronic document (backed up occasionally) that permits easy data entry and record keeping. In addition to tests of the RTP system, QA tests must be performed of the associated peripherals such as the digitizer, printer, treatment card, or export function, etc. One must have confidence during the treatment-planning process that printed and exported results are accurate and as expected. An example of comprehensive QA tests of modern RAL equipment, albeit for LDR 125I, is provided by Rivard, Evans, and Kay (2005) where suitability of various AAPM task groups is considered.
Emergency Procedures The AAPM TG-59 report provides guidance toward the development of adequate HDR brachytherapy emergency procedures. Emergency procedures should be able to direct treatment personnel through the worst-case scenario accident, e.g., the HDR 192Ir source becoming detached from the source guide wire and remaining inside the catheter or applicator. In some extreme instances, source retrieval from the patient could include surgical intervention. However, it is more likely for minor emergencies to occur, which include, but are not limited to, loss of power and abnormal performance of the RAL. Because of the severe consequences that could occur after only a short period of time, the NRC directs the licensee to provide appropriate staff and equipment during HDR brachytherapy. Hence, the presence of an authorized medical physicist and authorized user is often required during HDR brachytherapy treatment.
2.399 2.376 2.354 2.332 2.310 2.289 2.267 2.246 2.225 2.204 2.184 2.163 2.143 2.123 2.103 2.084 2.064 2.045 2.026 2.007 1.988 1.969 1.951 1.933 1.915 1.897 1.879 1.862 1.844 1.827 1.810 1.793
2.387 2.365 2.343 2.321 2.299 2.278 2.257 2.236 2.215 2.194 2.173 2.153 2.133 2.113 2.093 2.074 2.054 2.035 2.016 1.997 1.979 1.960 1.942 1.924 1.906 1.888 1.870 1.853 1.835 1.818 1.801 1.785
7/5/05 7/6/05 7/7/05 7/8/05 7/9/05 7/10/05 7/11/05 7/12/05 7/13/05 7/14/05 7/15/05 7/16/05 7/17/05 7/18/05 7/19/05 7/20/05 7/21/05 7/22/05 7/23/05 7/24/05 7/25/05 7/26/05 7/27/05 7/28/05 7/29/05 7/30/05 7/31/05 8/1/05 8/2/05 8/3/05 8/4/05 8/5/05
1.472 1.458 1.445 1.431 1.418 1.405 1.391 1.378 1.366 1.353 1.340 1.328 1.315 1.303 1.291 1.279 1.267 1.255 1.243 1.232 1.220 1.209 1.197 1.186 1.175 1.164 1.153 1.142 1.132 1.121 1.111 1.100
1.465 1.452 1.438 1.425 1.411 1.398 1.385 1.372 1.359 1.346 1.334 1.321 1.309 1.297 1.285 1.273 1.261 1.249 1.237 1.226 1.214 1.203 1.192 1.181 1.170 1.159 1.148 1.137 1.126 1.116 1.106 1.095
Figure 3. Decay table for an HDR 192Ir brachytherapy source (1266.xls). continued on next page
5/14/05 5/15/05 5/16/05 5/17/05 5/18/05 5/19/05 5/20/05 5/21/05 5/22/05 5/23/05 5/24/05 5/25/05 5/26/05 5/27/05 5/28/05 5/29/05 5/30/05 5/31/05 6/1/05 6/2/05 6/3/05 6/4/05 6/5/05 6/6/05 6/7/05 6/8/05 6/9/05 6/10/05 6/11/05 6/12/05 6/13/05 6/14/05
3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05
3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082
3.890 3.854 3.818 3.782 3.747 3.712 3.677 3.642 3.608 3.575 3.541 3.508 3.475 3.443 3.411 3.379 3.347 3.316 3.285 3.254 3.224 3.194 3.164 3.134 3.105 3.076 3.047 3.019 2.991 2.963 2.935 2.908
3/23/05 3/24/05 3/25/05 3/26/05 3/27/05 3/28/05 3/29/05 3/30/05 3/31/05 4/1/05 4/2/05 4/3/05 4/4/05 4/5/05 4/6/05 4/7/05 4/8/05 4/9/05 4/10/05 4/11/05 4/12/05 4/13/05 4/14/05 4/15/05 4/16/05 4/17/05 4/18/05 4/19/05 4/20/05 4/21/05 4/22/05 4/23/05
3.908 3.872 3.836 3.800 3.764 3.729 3.694 3.660 3.625 3.592 3.558 3.525 3.492 3.459 3.427 3.395 3.363 3.332 3.301 3.270 3.239 3.209 3.179 3.149 3.120 3.091 3.062 3.033 3.005 2.977 2.949 2.921
DATA ENTRY Date of Strength Calibration
HDR 192Ir source #D35A-1266 = 3.908 cGy m2/h at 00:00:00 3/23/2005 by CSM length 995 mm 73.83 day HL Strength (cGy m2/h) Strength (cGy m2/h) Strength (cGy m2/h) date MIDNIGHT NOON date MIDNIGHT NOON date MIDNIGHT NOON
84 Mark J. Rivard and Christopher S. Melhus
1.776 1.760 1.743 1.727 1.711 1.695 1.679 1.663 1.648 1.632 1.617 1.602 1.587 1.572 1.557 1.543 1.528 1.514 1.500 1.486
1.768 1.751 1.735 1.719 1.703 1.687 1.671 1.655 1.640 1.625 1.609 1.594 1.579 1.565 1.550 1.536 1.521 1.507 1.493 1.479
8/6/05 8/7/05 8/8/05 8/9/05 8/10/05 8/11/05 8/12/05 8/13/05 8/14/05 8/15/05 8/16/05 8/17/05 8/18/05 8/19/05 8/20/05 8/21/05 8/22/05 8/23/05 8/24/05 8/25/05
1.090 1.080 1.070 1.060 1.050 1.040 1.030 1.021 1.011 1.002 0.992 0.983 0.974 0.965 0.956 0.947 0.938 0.929 0.921 0.912
1.085 1.075 1.065 1.055 1.045 1.035 1.026 1.016 1.006 0.997 0.988 0.979 0.969 0.960 0.951 0.942 0.934 0.925 0.916 0.908
Figure 3. Decay table for an HDR 192Ir brachytherapy source (1266.xls). continued from previous page
6/15/05 6/16/05 6/17/05 6/18/05 6/19/05 6/20/05 6/21/05 6/22/05 6/23/05 6/24/05 6/25/05 6/26/05 6/27/05 6/28/05 6/29/05 6/30/05 7/1/05 7/2/05 7/3/05 7/4/05
3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05 3/23/05
3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082 3.9082
2.880 2.854 2.827 2.800 2.774 2.748 2.723 2.697 2.672 2.647 2.622 2.598 2.574 2.550 2.526 2.502 2.479 2.456 2.433 2.410
4/24/05 4/25/05 4/26/05 4/27/05 4/28/05 4/29/05 4/30/05 5/1/05 5/2/05 5/3/05 5/4/05 5/5/05 5/6/05 5/7/05 5/8/05 5/9/05 5/10/05 5/11/05 5/12/05 5/13/05
2.894 2.867 2.840 2.814 2.787 2.761 2.736 2.710 2.685 2.660 2.635 2.610 2.586 2.562 2.538 2.514 2.490 2.467 2.444 2.421
DATA ENTRY Date of Strength Calibration
HDR 192Ir source #D35A-1266 = 3.908 cGy m2/h at 00:00:00 3/23/2005 by CSM length 995 mm 73.83 day HL Strength (cGy m2/h) Strength (cGy m2/h) Strength (cGy m2/h) date MIDNIGHT NOON date MIDNIGHT NOON date MIDNIGHT NOON
7–Quality Assurance for High Dose Rate Remote Afterloaders 85
86
Mark J. Rivard and Christopher S. Melhus
Tufts-New England Medical Center, Department of Radiation Oncology Monthly HDR 192Ir Source QA Autoradiograph Tape clear 6F endobronchial catheters to readypack V-film with two GYN dummy #1 inserted. Place on simulator table at 100 cm SFD with 4 x 25 cm2 field size (no wires present) and take 3 exposures (small focus, 75 kVp, 320 mAs). Remove dummy markers and bring to HDR unit. Program HDR control console with step=5.0 mm; L=995 mm (separately); and t=0.2 second for positions below (tIRR = 3.0 s total). Autoradiograph film without bolus to minimize scatter. Process film, and analyze shifts within 0.5 mm.
Position1 1000 mm 6F Shift [mm] 1
3
5
1
7
9
3
11
5
9
11
13
15
17
13 15 17 19 21 23 25 27 29 31
19
23
25
27
29
31
33
33 35 37 39 41 43 45
ŶŶŷŶŷŶŷŶŷŶŷŶŶŷŶŷŶŷŶŷŶŷŶŷŶŷŶŷŶŷŶŷŶŶŷŶŷŶŷŶŷŶŷŶŷŶŷŶŷŷŷŷŷ Perform two autoradiographs using tdwell = 3 s and 30 s to evaluate source homogeneity.
Source homogeneity acceptable (visual assessment) ? ___________ Average measured offset _______________ mm (range 0.0 ± 1.0 mm) If offset is within range, write “NA” below. Otherwise enter the correct L (z 995 mm) and complete the following 3 tasks: Update daily check program with new length L Update program cards for any ongoing patients with new length L Post new length L in planning room
___________ ___________ ___________
Timer Offset, Linearity, and Accuracy Insert a clear, 6F catheter into HDR-1000. Setup control console with L=995, step 2.5 mm, position=18, and variable dwell time (tdwell). Use electrometer without filter, fit to straight-line, and determine toffset and time linearity (R2). Use stopwatch to independently measure tdwell(120) accuracy. If toffset, R2, or tdwell(120) are not within specifications, have second physicist review results. If they are still outside specification, contact Nucletron.
tdwell [seconds] 0.1 1.0 10.0 0.5 5.0 0.2 2.0 20.0
measured charge (2 PC scale)
what is the value of toffset ?
what is the value of R2 ?
what is the value of tdwell(120)/120 ?
is toffset –0.55 ± 0.03 s ?
is R2 > 0.99 ?
is tdwell(120)/120 1.000 ± 0.010 ?
Source Strength Calibration 1. Setup control console with L=995 for 5F/6F caths in HDR 1000+ and 10 seconds each for positions 14-22. Use 200 nA scale on the Keithley 6517A and the AVG(10) filter function. Position 22 21 20 19 18 17 16 15 14 1000 mm 6F Determine maximum reading (MAXRdg) ____________________ at time: ________________ 2. Calculate Monthly QA air kerma strength (SK); 192Ir t½ = 73.83 days with Ȝ = 0.009388 days–1: Decay factor (NEMCF) to 0:00 for the current day _____________> 1, typically ~ 1.0071. Measure temperature _________ [ºC] and pressure ___________ [mbar] and calculate CTP ___________ Std Imaging HDR-1000 well chamber S#A940704 good until 2/17/2006 CWELL = 0.5029 mGy m2 h–1 nA–1 Keithley 6517A electrometer S#0857926 good until 9/15/2005 CEL = 0.999 nA Rdg–1 on 200 nA scale or Keithley MK 602 (CNMC) electrometer S#34635A good until 2/9/2006 CEL = 0.0998 nA Rdg–1 10–8 A scale Monthly QA SK = MAXRdg x NEMCF x CTP x CWELL x CEL = _____________ [mGy m2 h–1] For air kerma rate @ 1 meter (as required by license) divide above number by 10 _____________ [cGy/h] 3. Comparison of Monthly QA determination of SK with tabulated value from decay chart. Source ID # ____________________ Calibration date __________________________ Tabulated (Midnight) SK value from decay table _________________ Ratio- divide Tabulated by air kerma rate @ 1 meter __________ Acceptable (Y/N)? _____ (1.000 r 0.020)
Figure 4. Monthly QA.doc. page 1
7–Quality Assurance for High Dose Rate Remote Afterloaders
87
Tufts-New England Medical Center, Department of Radiation Oncology Monthly HDR 192Ir Source QA Check of Guide Tubes GYN #1: length = __________ mm Acceptable [748.5 mm +/- 1.0 mm]? GYN #2: length = __________ mm Acceptable [748.5 mm +/- 1.0 mm]? GYN #3: length = __________ mm Acceptable [748.5 mm +/- 1.0 mm]? Is the mechanical integrity of all applicators, connectors, and guide tubes acceptable based on visual inspection? Safety Checks
___________ ___________ ___________ ___________ (Y or N)
1. HDR 192Ir source retraction following AC power failure simulation?
___________
2. Door light and Primalert function with AC power off using Cs micrad check source? (Unplug “Connector In” from PRIMAPAK)
___________
3. Emergency STOP (on wall near console) functions with printout of irradiation time?
___________
4. Door interlock functions with printout of irradiation time?
___________
5. Interrupt button on the console functions with printout of irradiation time?
___________
6. Survey meter battery check and constancy check OK?
___________
7. Was survey meter calibrated within 365 days ? (If not, notify Health Physics)
___________
8. Are Emergency Procedures in HDR Log Book?
___________
9. Is error 129 detected when indexer ring is not latched?
___________
10. Is error 129 detected when applicator is not inserted?
___________
11. Is error 129 detected when interstitial interconnect tube is installed without catheter present?
___________
12. Review of daily checks since last monthly/new source QA?
___________
13. Daily checks performed? (attach and complete QA-Daily.doc)
___________
137
Additional Notes (if needed):
Authorized Physicist Name _____________________________ Signature/Date _____________________________
Figure 4. Monthly QA.doc. page 2
88
Mark J. Rivard and Christopher S. Melhus
Tufts-New England Medical Center Department of Radiation Oncology Daily HDR 192Ir Source QA The microSelectron remote afterloader, model # 080.000 serial # 9511, is made by Nucletron BV. The source is manufactured by Mallinckrodt Medical BV, model CI L BV. Name: ____________________________ Signature: _________________________
Date: _______
1. Are the TV monitors at the control console operational? 2. Is the intercom operational? (CL-4 must be powered) 3. Does the radiation area monitor function? 4. Does the radiation survey meter function? 5. Is the emergency basket stocked and the emergency storage container in place?
_______ _______ _______ _______ _______
[forceps/tongs, cutters/scissors, suture removal kit, and “CAUTION RADIATION AREA” tape]
6. Are the source guide tubes free of kinks, etc? 7. Did control console pass self-diagnostic checks when key switch is first turned? 8. Is the emergency STOP button in the treatment room operational?
_______ _______ _______
[verify illumination of “Activate” light on Emergency Stop box near console]
9. Is the emergency STOP button on the HDR unit operational?
_______
[verify illumination of “Activate” light on Emergency Stop box near console]
Attach source position check ruler to channel 2 (incorrect channel). Turn on ceiling light and adjust TV camera so that check ruler is clearly visible. Run program card #1. This will program: Positions 4, 8, 12, 16, 20 Length 995 Step size 2.5 mm Dwell times at least 3 sec for positions 8,12,16,20 at least 20 sec for position 4 10. Confirm program and dwell times, press START to initiate treatment, and verify that Error Code 0129 (or 014) appears on printout and that the source is not exposed.
_______
Attach the source position check ruler to channel 1 (correct channel). Press START to initiate treatment. 11. Do the ends of the live and dummy sources align with the marked distance indicators to within 1 mm of 950, 960, 970, 980, and 990?
_______
While the source is dwelling at position 4, check the door light, interrupt key, door interlock, and emergency stop. The source should return to position 4 (end at 990) after each restart. 12. Is the “HDR IN USE” door light on? 13. Is the door interlock operational? 14. Does the interrupt button on the console function properly? 15. Is the emergency STOP button near the HDR console operational?
_______ _______ _______ _______
On the printout: 16. Is the date correct? 17. Is the time correct? 18. Is the source strength on the printout within 0.02 cGy m2/h of the posted activity? 19. Is the paper supply in the console printer sufficient for the day's treatment?
_______ _______ _______ _______
Figure 5. Daily QA.doc.
7–Quality Assurance for High Dose Rate Remote Afterloaders
Figure 6. Treatment QA.doc.
89
90
Mark J. Rivard and Christopher S. Melhus
System Failures Physical Plant Emergencies For natural disasters such as fires, earthquakes, tornadoes, or floods, the brachytherapy treatment team should generally stop HDR treatment and follow hospital or clinic-specific emergency procedures. Even if electrical power is interrupted, RAL units are designed with sufficient redundancy (battery backup) to retract the HDR 192Ir source and to record the amount of treatment completed using a backup battery or power pack. The medical physicist should include regular tests of backup power systems through QA tests. AAPM TG-56 recommends quarterly testing of backup power systems. Minor Emergencies Minor emergencies arise through abnormal operations of the RAL unit. These may include kinked treatment catheters, catheter obstruction, loose guide tube connections, operation of treatment vault doors during treatment, patient motion or discomfort during treatment, or other similar events. Manufacturers of HDR brachytherapy devices incorporate displayed messages and error codes for the operator to interpret the root cause of a minor emergency. Training should prepare hospital brachytherapy personnel to recognize common error codes and/or enable them to interpret appropriate reference manuals to determine the meaning of a given error code. In addition, operators must be able to determine whether treatment should continue in the wake of a minor emergency. Typically, minor emergencies do not result in misadministrations to the patient, and treatment may resume after correcting the problem. Major and Medical Emergencies A major emergency is one that can only be resolved through physical intervention of a member of the treatment team. For example, source retraction failure, controller (computer) failure, or patient medical emergency would qualify as a major emergency. As this type of emergency requires entry into the treatment vault while the source is out of the safe, the potential exists for very high personnel and patient doses. The dose rate can be over 7 Gy per minute at a distance of 1 cm from a 10 Ci 192Ir source. Therefore, individuals should be trained to work quickly and efficiently during the resolution of a major emergency. RAL manufacturers often prepare a single-page instruction sheet on how to retract the source manually in the event of a major emergency. All treatment personnel should be trained to perform this task and retraining should be offered at least annually. Ideally, source retraction/retrieval would occur within 1 to 2 minutes of a major emergency to limit detrimental outcomes to the treatment personnel and to the patient. The medical physicist should maintain an emergency kit near the HDR treatment vault. At a minimum, this kit should include tools for transferring a loose HDR source to a backup shielded container, a suture removal kit, and appropriate signage to secure a room after a major emergency. The contents of an emergency kit should be checked before every treatment. An example of an Emergency Procedure form is included in Figure 7. When printed in color and posted in the clinic, the form facilitates rapid understanding of the proper actions to manually retract the HDR 192Ir source for personnel responding to an emergency.
Staffing, Credentialing, Assignments, Equipment, and Training Frequency Radiation Safety Officer (RSO) While the RSO generally does not directly operate the HDR brachytherapy device, the RSO is directly responsible for the safe use of radiation administration at an institution. In addition, the RSO communicates with regulatory agencies regarding the administration of a radioactive materials program. As a result, the RSO shall review and approve procedures and emergency plans. However, minor changes to forms for improvements to workflow, clarity, or practicality need not go through a laborious review process.
7–Quality Assurance for High Dose Rate Remote Afterloaders
Figure 7. Emergency Procedure.doc.
91
92
Mark J. Rivard and Christopher S. Melhus
Physician The radiation oncologist should be board-certified, have substantial expertise and training in brachytherapy, and be licensed (authorized) to operate an HDR 192Ir unit. While the physician is most knowledgeable about clinical aspects such as diagnoses, prescription doses, and short-/long-term effects, the radiation oncologist should have annual emergency procedure training and QMP review, and be shown the practical operation techniques of the RAL at least annually. Without a high frequency and running the risk of “crying wolf,” it is often humbling to call for surprise emergencies to glean how the physicians would address a medical emergency and perform their responsibilities of the emergency procedures. Radiation Therapy Technologist (RTT) The treatment unit operator is typically a radiation therapy technologist (RTT), and should have vendorsupplied training on the HDR unit or be trained by an in-house expert using a documented and RSO-approved training program. Like the physician and physicist, the RTT should have annual emergency procedure training and should review the QMP for a general sense of the program goals. Furthermore, the treatment-unit operator should have the opportunity to practice operating the machine before administering patient treatment. This is why it is not suggested that the medical physicist perform the daily QA as a favor since this gesture could inadvertently result in a misadministration. Medical Physicist AAPM TG-59 recommends that one full-time equivalent medical physicist be allocated for an average of 10 HDR fractions per week, including treatment planning, treatment supervision, and periodic QA and training. The medical physicist should be board-certified in therapeutic radiological physics and should attend vendor-supplied training for both the treatment unit and the treatment-planning system. If possible, the medical physicist should spend 1 to 3 weeks at an established HDR program to observe all aspects of HDR brachytherapy treatment and QA. Since physicists typically coordinate the training and technical aspects of the HDR QA program, sometimes the continuing training of the physicist is overlooked. In addition to the annual emergency procedure training and QMP review, the physicist should re-examine the credentialing program to assure that it is up-to-date, and that s/he would still pass the competency test. A sample test (Figure 8) for physicians, therapists, and physicists is included (with a key).
Case Reports from NRC Public Postings Towards learning from the mistakes of others, it is instructive to review the problems with HDR brachytherapy experienced at other institutions. Follows are six examples which document treatment mishaps that could have been prevented had high QA standards been enforced. Example 1: NRC Information Notice 00-05 In NRC Information Notice 00-05 titled “Recent Medical Misadministrations Resulting From Inattention To Detail,” a series of eight preventable misadventures occurring between January 1999 and March 2000 were listed, including the following two in HDR brachytherapy. A patient received an underdose to an intended site and a dose to an unintended site during a high dose rate afterloader (HDR) treatment. An error, made by a technologist when programming the parameters into the HDR, was not detected by the physicist required to verify the information keyed into the console. The technologist responsible for setting up the HDR was unfamiliar with the skip process and did not program the skip into the HDR, even though it was clearly marked on the written directive. The skip process, in this particular remote afterloader, involves the sending of the source to the end of the catheter as a starting point and then moving the source back to the designated first radiation point. Because the
7–Quality Assurance for High Dose Rate Remote Afterloaders
93
skip was not programmed, the source started the radiation treatment from the end of the catheter, resulting in a dose to an unintended site. Ineffective training and inattention to detail were considered the root causes of this misadministration. A patient received an underdose to the intended site and a dose to an unintended site during an HDR treatment. After several unsuccessful attempts to electronically transfer a patient’s treatment plan to the HDR treatment system, the licensee manually entered the treatment plan directly into the treatment system’s control station. While manually entering changes to the source dwell times, an unintended change to the source step size occurred. The licensee did not notice this change and the patient was treated using the incorrect step size resulting in the misadministration. The licensee attributed this unintended and unnoticed change in step size to a software problem. However, had the licensee properly reviewed the complete final treatment plan, which clearly showed the changed step size, this misadministration could have been prevented.
More information regarding the latter incident is available in NCR Information Notice 99-09. Example 2: NRC Significant Enforcement Action EA-04-093 In the instance described in EA-04-093, the NRC determined that a Severity Level III violation occurred when the licensee failed to ensure that the HDR source transfer tube of the correct length was attached to the treatment applicator. In summary, [the medical center] did not develop written procedures to assure the source transfer tube remained secured to the vaginal cylinder during an HDR brachytherapy treatment and thus ensure that NRC-licensed material was administered in accordance with the written directive from an authorized user physician, as required by 10 CFR 35.41, “Procedures for Administrations Requiring a Written Directive.” An authorized user physician prepared a written directive prescribing three HDR remote afterloading brachytherapy treatments to a patient after the patient received treatment from external beam radiation. Each HDR treatment was to consist of 700 rads from an Iridium-192 source of approximately 3.5 curies by vaginal cylinder. Erythema was noticed on the inner portion of each of the patient’s thighs, subsequent to the third treatment on March 11, 2004, indicating the radiation source did not transverse the entire length of the transfer tube during the HDR treatment. The authorized user physician did not expect any adverse medical effects to the patient as a result of this medical event. Corrective actions included: (1) developing written procedures to determine if the transfer tube is in place before and after treatment; (2) marking the transfer tube as part of a visual check to determine whether the transfer tube moves during treatment; and (3) training the medical physics and nursing staffs on the procedures.
Example 3: NRC Significant Enforcement Action EA-04-087 The misadventure in this example occurred when one medical physicist did not correctly implement the prescription depth requested by the authorized user in the written directive, and a second medical physicist did not review the work until two of five fractions were completed. Thus, there was a dose error of +60% for the first two fractions. This event reinforces the value of communication between the treatment planner (medical physicist or dosimetrist) and the authorized user, and the importance of treatment plan checking prior to radionuclide administration. An authorized user physician prepared a written directive prescribing five planned HDR remote afterloading brachytherapy treatments to a patient. Each treatment was to consist of 500 rads at the surface of a 25 millimeter (mm) vaginal cylinder for a total dose of 2500 rads, according to the directive prepared by the physician. A medical physicist, not realizing the physician’s written directive specified the intended dose of 500 rads per fraction was to be delivered at the surface of the cylinder, used
94
Mark J. Rivard and Christopher S. Melhus
KEY HDR 192Ir BRACHYTHERAPY EXAMINATION FOR PHYSICIAN AUTHORIZED USERS AND DEVICE OPERATORS 1. For HDR 192Ir, what requirements need to be satisfied to be a physician authorized user? 5 treatments, emergency procedure training, board certification or RSC & State Credentialing, 75% > this exam 2. For HDR 192Ir, what requirements need to be satisfied to be an authorized medical physicist? 5 treatments, emergency procedure training, board certification or RSC & State Credentialing, 75% > this exam 3. For HDR 192Ir, what requirements need to be satisfied to be an authorized radiation therapist? 5 treatments, emergency procedure training, registered RTT, 75% > this exam 4. Who may have access to keys to the HDR unit? Only authorized physician, physicist, and therapist 5. What is the required action should the door interlock malfunction? Guard the door entry during treatment, and contact Health Physics (x6-6168) 6. How may treatment planning data be entered into the HDR 192Ir control unit? plastic data card, standard programs in console memory, manual entry 7. According to our radioactive materials license, who must be physically present or within the audible range of normal human speech at the console area during all HDR treatments? Authorized physician, physicist, and therapist 8. What are the radiation survey requirements for HDR 192Ir procedures? Before and after treatment, survey the room and patient
Figure 8. HDR Exam3_key.doc. page 1
7–Quality Assurance for High Dose Rate Remote Afterloaders
95
KEY 9. According to our license, should the following checks be performed daily, monthly, or quarterly? a. b. c. d. e. f. g.
Check of the permanent radiation monitor Check of the TV monitor and intercom Check of treatment room door interlock Timer accuracy and linearity Check of backup battery Calibration of source Check of source homogeneity
D D D M M Q M
10. If the permanent radiation monitor is found to be inoperable, under what condition may a patient be treated? Guard the door during treatment (preferably Health Physics) 11. What is the maximum exposure level 1 meter from the surface of the HDR unit when the source is in the shielded position (in the tungsten safe)? (a) 0.25 mR/h
(b) 2.5 mR/h
(c) 25 mR/h
(d) 250 mR/h
12. What is the typical activity range of the HDR 192Ir source? (a) 4-10mCi
(b) 40-100 mCi
(c) 4-10 Ci
(d) 40-100 Ci
13. What is the typical source strength (U = 1 cGy/h/cm2) of the HDR 192Ir source? (a) 10-40 U
(b) 100-400 U
(c) 1,000-4,000 U
(d) 10,000-40,000 U
(c) 192Ir
(d) 252Cf
14. What radionuclide is used in the HDR unit? (a) 137Cs
(b) 60Co
15. What is meant by the term dwell time? Time HDR 192Ir source is in treatment positions 16. According to Nucletron’s definition, what is meant by the term secondary time? Time HDR 192Ir source is out of RAL safe 17. In the event of an emergency, what actions must be undertaken, in what order, and who is responsible for carrying them out? Therapist Physicist Physician Anyone
try all emergency off buttons enter room, crank source, survey patient remove applicator, place source in bailout pig contact Health Physics division (x6-6168)
18. Why should you not use a connecting tube for endobronchial or esophageal implants? Adding the connecting tube will not allow the source to reach the intended treatment site
Figure 8. HDR Exam3_key.doc. page 2
96
Mark J. Rivard and Christopher S. Melhus the standard treatment parameter of 5 mm beyond the surface of the cylinder in developing the treatment plan. The patient was then treated on March 15, 22, and 29, 2004, according to the treatment plan prepared by the medical physicist. Prior to the fourth treatment, a second medical physicist independently reviewed the previously approved treatment plan and discovered the 5 mm error in that plan. As a result of the 5 mm error, an additional dose of 318 rads was delivered at the cylinder surface during each fractionated treatment for a total of 818 rads administered to the patient per treatment instead of the prescribed 500 rads. Corrective actions included: (1) requiring your staff to independently check treatment plans prior to each treatment; (2) requiring a second medical physicist to review the treatment plan and verify the treatment parameters are correct; (3) requiring the medical physicist to manually calculate the treatment dose and compare it to the dose developed in the treatment plan and written directive prior to administering the treatment to the patient*; and (4) disseminating the results of the investigation, including root cause and corrective actions, to all members of the medical physics staff and authorized user physicians involved in HDR treatments.”
Example 4: NRC Significant Enforcement Action EA-99-257 Although there were no adverse effects as a result of the incident described, in the following is an example of failure to verify the vendor-supplied source strength after HDR source exchange prior to patient treatment. [This action] involved the treatment of patients with your High Dose Rate (HDR) afterloader after the source was replaced, without first checking the source to ensure that the source strength was consistent with the value provided by the manufacturer. Although some quality assurance checks were performed on the HDR unit prior to use, your medical physicist was unable to verify the manufacturer’s source strength certification because the dosimetry system used to verify (calibrate) the source unit was not available. This was a violation of your facility’s procedures that require a check of the source strength after each installation. Although there were no actual consequences in this case, since the source strength was subsequently checked and was consistent with the manufacturer’s values, the failure to check the source prior to use on patients could have resulted in administration of an incorrect dose to the patient, possibly much higher than the prescribed dose for treatment. Credit for corrective actions is warranted because your corrective actions were considered prompt and comprehensive, once the violation was identified by the NRC. These actions include, but are not limited to: (1) a complete review of the HDR program by the Medical Physicist and the Radiation Oncology Manager; (2) a mandatory inservice for all staff associated with the HDR program on the requirements of the license; (3) the development of a “job specifics checklist” for newly hired medical physicists that includes instruction on all requirements of the NRC license; and (4) revision of HDR quality assurance forms highlighting items that must be completed prior to treating patients with the HDR unit.
Example 5: NRC Significant Enforcement Action EA-97-284 In this Enforcement Notice, the licensee was cited for administering radiotherapy treatments that were not explicitly allowed by their RML.
*
Note that item (3) was corrective action adopted by the particular medical center as a result of the misadventure. Manual calculation of HDR treatment doses is recommended by TG-59. However, there is significantly more reliance on computer-based treatment planning systems now than in 1998, when TG-59 was published. Thus, it is more likely for a physicist to carefully review the input parameters to the treatment plan, as opposed to manually calculating the dose at a select number of reference points.
7–Quality Assurance for High Dose Rate Remote Afterloaders
97
The violation involves the use of iridium-192 in a HDR afterloader for surface treatment of skin cancer without your NRC license authorizing such use. Your license states that the HDR unit may be used for the treatment of humans only for interstitial, intracavitary, or bronchial therapy. The license does not authorize the use of the HDR for topical skin treatments.
While this example is not useful for most day-to-day application of HDR brachytherapy, it is important for the medical physicist to be aware of the stipulations of their radioactive materials license. This applies to both LDR and HDR brachytherapy and the range of radionuclides and treatment sites/methods that may be employed at a given facility. Example 6: NRC Significant Enforcement Action EA-97-245 As in Example 5 above, this licensee did not ensure that patient treatments were conducted in accordance with their RML. In this instance, the medical physicist performing tasks associated with HDR brachytherapy treatment was not listed on the license, and the medical physicist did not follow protocols outlined in the RML. The violation involves the failure to follow your Quality Management Program (QMP) for your High Dose Rate Afterloading (HDR) treatment program in that certain activities required to be performed by your authorized Medical Physicist listed on the license were, at times, being performed by an individual other than the Medical Physicist listed on the NRC license. Specifically, from January 1997 to May 1997, for several HDR patients’ treatments, the calculations of the required treatment times and verification of the accuracy of the treatment input parameters were performed by an individual other than the Medical Physicist listed on the license. In addition to the violation, one of the requirements of NRC Bulletin 93-01, issued in 1993 following a serious HDR event at a facility in Pennsylvania, is that an authorized user and the RSO or the Medical Physicist must be physically present during each HDR treatment. There is only one individual who is named both the RSO and also the Medical Physicist on your NRC license. During the inspection, you stated that since January 1995, you used physicists other than the Medical Physicist named on the license, to be present during HDR treatments to meet the NRC Bulletin. Also, you indicated that since January 1997, some of the treatment plans were performed by an individual who is not authorized by the NRC license as a medical physicist. You further indicated that at times, the individuals, while in the general vicinity of the treatment areas, were not physically present for the treatment. At the conference, you indicated that since the inspection, the Medical Physicist named on the license has been present for all HDR treatments.
References Glasgow, G. P., J. D. Bourland, P. W. Grisby, J. A. Meli, and K. A. Weaver. “Remote Afterloading Technology.” AAPM Report No. 41. New York: American Institute of Physics, 1993. International Electrotechnical Commission (IEC), Geneva, Switzerland, http://www.iecee.org/cbscheme/ Standard/med.htm last accessed May 1, 2005. IEC 60601-2-17 http://www.astro.org/publications/astronews/2005/Jan/IECReport.htm last accessed May 1, 2005. Kubo, H. D., G. P. Glasgow, T. D. Pethel, B. R. Thomadsen, and J. F. Williamson. (1998). “High dose-rate brachytherapy treatment delivery: Report of the AAPM Radiation Therapy Committee Task Group No. 59.” Med Phys 25(4):375–403. Also available as AAPM Report No. 61. Lerma, F. A., and J. F. Williamson. (2002). “Accurate localization of intracavitary brachytherapy applicators from 3D CT imaging studies.” Med Phys 28:325–333.
98
Mark J. Rivard and Christopher S. Melhus
Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24(10):1557–1598. Also available as AAPM Report No. 59. Rivard, M. J., D-A. R. Evans, and I. A. Kay. (2005). “A technical evaluation of the Nucletron FIRST® system: Conformance of a remote afterloading brachytherapy seed implantation system to manufacturer specifications and AAPM Task Group report recommendations.” J Appl Clin Med Phys 6:22–50. UL 60601 http://www.devicelink.com/mddi/archive/04/02/001.html last accessed May 1, 2005. U.S. Nuclear Regulatory Commission Enforcement Action 99-257: “Clara Maass Medical Center – Notice of Violation.” Washington, D.C., 2000. http://www.nrc.gov/reading-rm/doc-collections/enforcement/actions/materials/ ea99257.html. U.S. Nuclear Regulatory Commission Enforcement Action 04-093: “Missouri Baptist Medical Center – Notice of Violation.” Washington, D.C., 2004. http://www.nrc.gov/reading-rm/doc-collections/enforcement/actions/materials/ea04093.html. U.S. Nuclear Regulatory Commission Informational Notice 00-05: “Recent medical misadministrations resulting from inattention to detail.” Office of Nuclear Material Safety and Safeguards, Washington, D.C., 2000. http://www.nrc.gov/reading-rm/doc-collections/enforcement/actions/materials/ea99257.html. U.S. Nuclear Regulatory Commission Enforcement Action 97-284: “Centre Community Hospital – Notice of Violation.” Washington, D.C., 1997. http://www.nrc.gov/reading-rm/doc-collections/enforcement/actions/materials/ ea97284.html. U.S. Nuclear Regulatory Commission Enforcement Action 04-087: “St. Vincent Hospital & Health Care Center – Notice of Violation.” Washington, D.C., 2004. http://www.nrc.gov/reading-rm/doc-collections/enforcement/ actions/materials/ea04087.html. U.S. Nuclear Regulatory Commission Enforcement Action 97-245: “Mountainside Hospital – Notice of Violation.” Washington, D.C., 1997. http://www.nrc.gov/reading-rm/doc-collections/enforcement/actions/materials/ ea97245.html.
Chapter 8
Continuous Low Dose Rate and Pulsed Dose Rate Remote Afterloader Units John L. Horton, Ph.D., Ann Lawyer, M.S., and Firas Mourtada, Ph.D. University of Texas M. D. Anderson Cancer Center Radiation Physics Department Houston, Texas Continuous Low Dose Rate Remote Afterloader Units (LDR RAUs) . . . . . . . . . . . . . . . . . . . . . 99 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Selectron LDR RAU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Advantages of LDR RAUs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Disadvantages of LDR RAUs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Facility Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Licensing Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 License Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Record Keeping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Nucletron Pulsed Dose Rate Remote Afterloader Unit (PDR RAU) . . . . . . . . . . . . . . . . . . . . . 104 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Advantages of PDR RAUs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Disadvantages of the PDR RAUs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Facility Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Licensing Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Commissioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Quality Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Continuous Low Dose Rate Remote Afterloader Units (LDR RAUs) Introduction University of Texas M. D. Anderson Cancer Center (UTMDACC) has a long history of continuous low dose-rate (LDR) brachytherapy treatment for gynecological cancers. The Fletcher Suit Delclos (FSD) applicators and their variants are probably the most widely used applicators in the United States. Approximately 18 years ago UTMDACC adopted clinical use of the Selectron low dose-rate remote afterloader unit (LDR RAU), shown in Figure 1, as a replacement for manually afterloaded 226Ra treatments. The choice of the LDR RAU was predicated on the desire to maintain LDR treatments. Currently, approximately 90% of our gynecological tandem and ovoid and tandem and cylinder treatments are performed with the Selectron LDR RAU. The remaining 10% are performed with manually loaded 137Cs with an occasional patient treated with the Nucletron high dose-rate remote afterloader (HDR RAU). Treatments average about one patient per month with the HDR RAU, compared with about three per week with LDR treatments. The HDR RAU patients are typically those with medical problems that prevent them from spending 48 hours in a hospital bed. Our principal use of HDR RAU for gynecological patients is limited to those treated for vaginal cuff disease with dome and cylinders. Treatments average about three patients per week with this modality.
99
100
John L. Horton et al.
Figure 1. Selectron Low Dose Rate Remote Afterloader Unit.
Selectron LDR RAU The Selectron LDR RAU was chosen as a replacement for 226Ra because an “equivalent” dose rate and dose distribution could be achieved, allowing us to build on several decades of prior clinical experience. The scattering and attenuation characteristics of 137Cs gamma rays in tissue are similar to 226Ra gamma rays, providing very nearly the same relative dose distribution. By proper choice of activity of the 137Cs pellets, the 226Ra dose rate could be approximated. The UTMDACC Selectron LDR RAUs have four channels, one for the tandem, one for each ovoid, and one for backup. Combinations of 48 active and inactive pellets are placed in each channel to simulate manually loaded intracavitary brachytherapy systems. The active pellets have a nominal activity of 5 mgRaeq (36.135 U) each. All pellets are 2.5 mm in diameter. The inactive pellets are ferromagnetic and the active pellets are nonmagnetic. The RAU is controlled by a microprocessor. The positions of the active pellets in each channel are programmed into the microprocessor. The pellets are transferred from the main safe to an intermediate safe. The pellets are sorted by a magnetic sorting technique. The pellets are stored in their programmed order in the intermediate safe. When treatment is initiated, the pellets are transferred from the intermediate safe in the afterloader to the brachytherapy applicators. When treatment is interrupted for nursing care or visitors, the pellets are transferred back to the intermediate safe. At the end of treatment, the pellets are transferred to the main safe. All pellet transfers are performed with compressed air. Prior to clinical implementation of the Selectron LDR RAU, extensive treatment planning comparisons were performed between historical 226Ra treatments and treatments to be performed with the Selectron LDR RAU. To ease the transition, a table was developed that translated 226Ra treatments to positions of active sources for the Selectron LDR RAU treatments.
8 – Continuous Low Dose Rate and Pulsed Dose Rate Remote Afterloader Units
101
Advantages of LDR RAUs The principal advantage of all RAUs is the reduction of personnel exposure to all staff members, including the radiation oncologist, physicist, dosimetrist and nursing staff, because the sources remain in the shielded radiation safe when personnel are in the room caring for the patient. In the European Union nations with stricter limits on personnel exposure, this is more of an issue than in the United States. Usually, in the United States, the reduced personnel exposure is not a sufficient advantage to offset the increased costs in the minds of most hospital administrators. Manually afterloaded brachytherapy implants can be performed while adhering to personnel exposure limitations by following a well conceived and executed plan limiting the dose to radiation workers. A second advantage of the LDR RAU is the reduced dose to the rectum. This reduced dose results from the LDR RAU ovoids being 5 mm longer than the ovoids that were used for 226Ra treatments and from the source activity loaded within the ovoid being in a more anterior position. The standard LDR RAU ovoids in use at UTMDACC are 33 mm long rather than the 28 mm of the manually loaded FSD ovoid. Also, the activity within the 226Ra sources and within the 3M 137Cs sources is asymmetrically placed within the encapsulation, because of the eyelet in these sources. This arrangement places the activity to the posterior pole of the ovoid rather than being centered in the ovoid. The newer J sources from Amersham center the activity in the encapsulation, but the means of orienting the sources within the ovoid places the center of the activity posterior to the center of the ovoid. With the Selectron RAU, at UTMDACC the ovoid activity is centered within the ovoid. A third advantage is the higher degree of “optimization” of the dose distribution that can be achieved with the LDR RAU because of the flexibility of placing the sources within the applicator. During the time of 226Ra treatments, activity loadings were specified in terms of “sources” or “inches,” depending on the length of the tandem relative to its position with the external os and the ovoids. The 226Ra sources were 22 mm long. If the tandem sources were abutted, the loading was prescribed as a “sources” loading. However, if the physician desired to spread the sources out over a greater length, 3-mm spacers were used between the 226Ra sources; this prescription was considered an “inches” loading. As the change was made to manually afterloaded 137Cs sources, this schema continued. However, as the 137Cs sources were only 20 mm in length, rather than the 22 mm of the 226Ra sources, 2 mm-spacers were used for “sources” loading and 5-mm spacers for “inches” loading. “Short sources” loading was also used for 137Cs. For the short sources prescriptions for 137Cs, the sources were abutted. This nomenclature continues today at UTMDACC with the LDR RAU treatments, and is detailed in two sets of standard loading tables. Table 1 was developed for “sources” loading in the tandem, with the numbers referring to the position of the active pellets. Table 2 is similar except it is for “inches” loading in the tandem. Other tables are available for different mg Ra eq loadings in the ovoids. The “custom” loadings for tandem and ovoid patients usually affect only the more inferior sources in the tandem. This customization gives the radiation oncologist more flexibility in dealing with relative geometry between the tip of the tandem and the flange. In recent years the degree of custom loading of UTMDACC cases has increased over the examples seen in Tables 1 and 2. There are tables for “short sources” and tables for tandems with various length spacers at the tip. UTMDACC dosimetrists frequently customize the treatments for patients beyond these loadings, which the many tables address. The decisions on these custom loadings may be made on the physician’s experience, based on the extent of the disease, geometry of the implant, and external beam dose, or the physician may want to see a treatment plan before beginning treatment. Tables 1 and 2 are given here only for illustration and should not be used in other clinics without first verifying that they are appropriate for that practice. More importantly, the flexibility in source positioning provided by the LDR RAU is a great asset for tandem and cylinder treatments. The high degree of customization with the LDR RAU allows the
102
John L. Horton et al. Table 1. “Sources” Loading Numbers refer to active source positions.
Superior source
Inferior source
5 mgRaeq
3
11
20
29
38
10 mgRaeq
1,3
10,12
19,21
29,31
37,39
15 mgRaeq
1,3,5
10,12,14
19,21,23
28,30,32
36,38,40
20 mgRaeq
1,2,3,4
10,11,12,13
19,20,21,22
28,29,30,31
37,38,39,40
Table 2. “Inches” Loading Numbers refer to active source positions.
Superior source
Inferior source
5 mgRaeq
3
13
23
33
44
10 mgRaeq
1,3
12,14
22,24
32,34
43,45
15 mgRaeq
1,3,5
11,13,15
21,23,25
31,33,35
42,44,46
20 mgRaeq
1,2,3,4
11,12,13,14
22,23,24,25
32,33,34,35
42,43,44,45
dosimetrist to design a plan that fulfills the prescription of the radiation oncologist much more easily than for the case with manually afterloaded patients. As can be seen from this discussion, intracavitary brachytherapy treatments as practiced at UTMDACC are almost as much art as science with the patient treatment based on semi-empirical rules articulated by Fletcher (Fletcher 1980) and Katz and Eifel (Katz and Eifel 2000).
Disadvantages of LDR RAUs The principal disadvantage of the LDR RAU is the possibility of failure during treatment. The great majority of these failures occur when the pellets are being transported from the radiation safe to the patient. Failures typically occur because of low air pressure, a power failure because of weather conditions, or a crimped transfer tube. The nursing staff can handle the majority of these failures, simply by restarting the transit. Occasionally a dosimetrist or medical physicist may be required to reprogram a unit. Maintenance may be indicated when the failures become too frequent. Maintenance issues typically arise when parts in the afterloader that are involved in the transfer of the pellets become worn and “out-of-round.” The pellets also can become out-of-round, leading to failures in their transport. To minimize the problems with pellets, Nucletron limits the number of pellet transfers to 7000. The second disadvantage of the LDR RAU is the desire of the patient to have visitors during treatment. Normally one would think that this would be an advantage of the LDR RAU. For patients being treated with manually loaded sources, visitors are limited to 30 minutes a day to prevent exceeding the visitor’s regulatory dose/exposure limits, while the patient treated with the LDR RAU can receive visitors for extended periods of time, since the visitor will not be irradiated. The sources will be within the radiation safe while a visitor is present. However, when the sources are within the safe, the patient is not being treated and the treatment time is extended. The physician counsels patients about the issue of extending the treatment time by having visitors. Given the choice, patients tend to limit their visitors in order to leave the hospital in a timely manner.
8 – Continuous Low Dose Rate and Pulsed Dose Rate Remote Afterloader Units
103
Facility Considerations Planning for the facility to house an LDR RAU involves shielding, location, ancillary services, and audiovisual communication issues. To design appropriate shielding, one must decide what treatments will occur in the room and what will be the maximum air kerma strength per hour and maximum total reference air kerma (TRAK) per week for these treatments. The room shielding must limit the dose equivalent in uncontrolled areas to less than 1.0 mSv per year, with a maximum dose equivalent of 0.02 mSv in any one hour (NCRP 1987). Radiation workers may receive up to 0.05 Sv per year, but based on ALARA (as low as reasonably achievable), rooms should generally be designed to limit exposure of radiation workers to 0.1 mSv per week in controlled areas (NCRP 1987). Some states also limit the instantaneous dose equivalent rate received for both the general public and radiation workers. The instantaneous dose equivalent rate limit is usually not an issue for LDR RAU. As always, a facility and an individual must comply with the applicable regulations in their state. Although a bedside shield may be used to reduce the dose equivalent levels to comply with the regulations, adequately shielded walls are the preferable alternative. This situation does not rely on someone to place the shield in the appropriate position for each case and provides more usable floor space in the room. LDR RAU rooms are typically hospital rooms that should be located close to the nurses’ station where the nurse can keep the patient and the room under surveillance. If there are multiple shielded rooms, they should be adjacent to each other and remote from unshielded hospital rooms. The nurses need an intercom and a closed circuit television system to observe and to communicate with the patient during treatment. The LDR RAU treatment control console is placed immediately outside the patient room, with a remote at the nurses’ station. The power for the LDR RAU should be on the emergency power circuit. Compressed air is also required for the Selectron LDR RAU. The room door should be interlocked. An area radiation monitor should be in the room with a remote outside the room to indicate when the sources are exposed. As with any brachytherapy room, there should be emergency equipment and an emergency container (bail out pig) in the room. The door should have a plaque for mounting radiation signs and instructions when patients are being treated.
Licensing Issues RAUs and the sources they contain must be licensed by the U.S. Nuclear Regulatory Commission in nonAgreement States and federal facilities or the appropriate state or local agency in Agreement States. Before a source or device can be licensed, it must be on the Registry of Sealed Sources and Devices. Typically, the process of entering a source or device on this registry requires many months and significant paperwork from the vendors. The vendor should always supply a copy of its registration before you consider the purchase of the source or device. License Application Whatever you state you will do in your license, you must perform. You can always increase your tasks from what is stated in your license, but you can never decrease what you indicate in your license without a time-consuming license amendment and thorough justification. Always promise less than you plan to deliver in these situations. Place only the minimum required information in your license application. The license application requires a description of the source and its registry number, the manufacturer and model number of the afterloader, the authorized users and their qualifications, the planned use of the device, the location of this planned use, and suitability of the room shielding. You should also include the radiation detection devices to be used, the radiation warning devices, the audiovisual equipment to view and communicate with the patient, and a description of the security of the area and for the sources. Details of your quality control program must be discussed, including calibration techniques and frequency, routine
104
John L. Horton et al.
quality control procedures and frequency, and leak test procedures and frequency. The qualifications of the individuals who will perform these tests and procedures, training of operators, emergency procedures and where they are posted, and disposal of decayed sources are required items. You also need to specify the maximum activity of the specified isotope that will be at your facility at any one time. Remember that when you have a source exchange, you have not only the new source but also the old source in your facility. It is usually acceptable to specify twice the activity of the new source to account for this. Of course, if you have other sources of the specified isotope in addition to what is in the afterloader, you also must ensure that these sources are accounted for in your specification of the maximum activity. Record Keeping Finally, the regulatory agency has requirements for record keeping of calibrations, quality control tests, room surveys, and treatment head surveys. These regulations include the length of time these records must be maintained. Be sure you adhere to all these requirements.
Training In addition to the standard radiation protection training, personnel require training specific for the RAU. Medical physicists and dosimetrists should be trained in the operation, programming, and emergency response procedures for the LDR RAU. The nursing staff needs to know how to start and stop the unit and be familiar with emergency response procedures. Physicians also need to be familiar with emergency response procedures.
Nucletron Pulsed Dose Rate Remote Afterloader Unit Introduction The Nucletron Pulsed Dose Rate Remote Afterloader Unit (PDR RAU), shown in Figure 2, is based on the Nucletron HDR afterloader described in a previous chapter. The physical construction of the PDR source is identical to the Nucletron HDR source (mHDR v2). The only difference is the lower activity of the PDR source (between 37 and 74 GBq at installation, depending on the institution’s needs and room shielding) compared to the HDR source (approximately 370 GBq at installation). The PDR RAU simulates continuous LDR treatments by delivering the same total dose in the same total time. With the higher activity source, this is accomplished by exposing the source from the afterloader for only a fraction of the time for each hour. Consider this example: the desired dose rate is 50 cGy/hr at the point of interest for a total treatment time of 48 hours. The PDR source delivers an instantaneous dose rate of 500 cGy/hr at that point. In this instance, the source should be exposed for a tenth of an hour, or 6 minutes, each hour for the same total treatment time of 48 hours. This schema delivers the same total dose in the same total time yielding the same average dose rate.
Advantages of PDR RAUs An essential component of the UTMDACC brachytherapy program is to continue the LDR program or a modality that provides a radiobiologically equivalent treatment. Radiobiological models and measurements indicate that PDR provides this capability (Brenner and Hall 1991; Fowler and Mount 1992; Millar, Hendry, and Canney 1996; Brenner et al. 1996; Armour et al. 1997; Sminia et al. 1998). The PDR RAU provides the same personnel protection as the LDR RAU, because no personnel are in the room when the sources are loaded into the applicators. However, the PDR provides the additional advantage in that nursing care can occur during the time between treatment pulses. For all patients, but
8 – Continuous Low Dose Rate and Pulsed Dose Rate Remote Afterloader Units
105
Figure 2. Nucletron Pulsed Dose Rate Remote Afterloader Unit.
especially for patients who require extra nursing care, this means that the extra care can be given without extending the overall treatment time. This also permits the patient to have visitors without extending the treatment time. These two advantages should reduce the stress of the treatment for the patient. Another advantage of the PDR RAU is the ability to achieve greater conformation of the dose to the treatment volume than is possible with the LDR RAU. This advantage results from being able to perform computer optimization of individual dwell times. The PDR RAU provides the ability to vary the dwell times at each source position during each treatment pulse. As the LDR RAU sources have a nominal activity of 5 mgRaeq, two sources are required to simulate a 10 mgRaeq source and three sources for a 15-mgRaeq source, etc. The LDR RAU provides limited flexibility in plan optimization in that it may simulate a 12.5-mgRaeq source by programming the unit to deliver half the treatment with three sources and half with two sources. However, to accomplish this requires a dosimetrist or medical physicist to reprogram the unit halfway through the treatment. In reality, for true optimization, a critical aspect of the process is the ability to image the implant with computed tomography or magnetic resonance. Analogous to external beam treatment, the physician must be able to visualize and to contour the treatment volume and all critical structures to make meaningful decisions about the optimization process.
Disadvantages of PDR RAUs The most serious disadvantage of the PDR RAU is the potential for afterloader failures with the source in the exposed position. A very large dose could result if the high activity source were to stick at one dwell position during treatment. As a first order approximation, consider that a typical tandem and ovoid treatment may consist of 15 dwell positions, which are treated for a total time of 10 minutes each hour. This implies the average dwell time at each position is approximately 10 minutes divided by the 15 positions, or 40 seconds. Clearly, if the source remains at one dwell position, an excessive dose will be
106
John L. Horton et al.
delivered in a very short time. If it takes 10 minutes to respond to the situation and retract the source, the overdose at that dwell position will be a factor of 15 too high. This situation requires an immediate response by a well-trained staff. Since the patient is in a hospital room for up to 72 hours, the medical physicist cannot be on site during the entire treatment. In a large hospital he/she may be remote from the treatment area and it may take 10 minutes to respond even if the medical physicist is on site. Preparation for the worst-case scenario requires that the nursing staff be able to respond to the emergency situation in a very timely manner. Fortunately, the probability of the source not being retracted is very low, based on data from the Academic Medical Center-Amsterdam (Blank 2004). They have reported their experience with PDR RAU for 417 treatments consisting of 5117 pulses. Of the 5117 planned pulses, 272 pulses (5.3%) encountered an obstruction during the check cable run. There were only five cases (0.1%) of an obstruction during an active sources run, all of which occurred when the active sources was exiting the PDR RAU. In these cases the machine retracted the active source without incident or the need for human intervention. Of the 417 treatments analyzed, five cases (1.2%) required an adjustment of the implant as a result of the catheters being kinked. Another five cases had their treatment discontinued for medical reasons. The Academic Medical Center-Amsterdam concluded based on their experience that 1. Errors can occur with the PDR RAU and on rare occasion require a treatment adjustment. 2. Errors do not always occur during the first pulse; hence an error-free start does not assure an errorfree treatment. 3. The check cable run has proven an effective system to prevent accidents with the active source. Academic Medical Center-Amsterdam has two physicists and four dosimetrists who are on call to address issues during the PDR treatments. They state that 95% of the issues can be solved by telephone conversation with the nurse.
Facility Considerations The facility requirements for a PDR RAU will be the same as for the Selectron LDR RAU, except for the compressed air. If the room has been shielded for LDR treatments, the shielding should be adequate for the equivalent PDR treatments. The instantaneous dose equivalent rate will be higher with the PDR RAU than for LDR treatments, but the dose equivalent rate averaged over an hour should be the same as for the LDR treatment. However, if local regulations include an instantaneous dose rate limit, it must be verified that these limits will not be violated. From a shielding standpoint, one advantage PDR has over LDR is that the half-value layer (HVL) in concrete for 192Ir is less than the concrete HVL for 137Cs (NCRP 1972), because of the lower average energy of the 192Ir. One caveat though is that if the new PDR treatments are planned that were not performed previously with LDR, these treatments may result in a higher average dose equivalent rate or higher total dose equivalent. For instance, large interstitial implants may result in higher average dose equivalent rate and higher total dose equivalent than the gynecological treatments. In this case, the adequacy of the shields must be verified.
Licensing Issues The issues involved with licensing this unit are essentially the same as discussed above for the LDR RAU and are common to all RAUs.
8 – Continuous Low Dose Rate and Pulsed Dose Rate Remote Afterloader Units
107
Training As with the LDR RAU, medical physicists and dosimetrists should be trained in the operation, programming, and emergency response procedures for the PDR RAU. The nursing staff needs to know how to start and stop the unit and be very familiar with the emergency response procedures. Physicians also need to be familiar with emergency response procedures. Because of the much higher activity of the source, everyone’s knowledge of emergency response procedures is much more important for operating the PDR RAU than for the LDR RAU.
Commissioning Because the Nucletron PDR RAU is basically the same machine with different software as the Nucletron HDR RAU, commissioning of the PDR RAU is similar to the HDR RAU commissioning previously discussed in other chapters. However, for the PDR RAU the determination of the transit dose from the source is important because of the large number of source transits during a treatment. Consider a large interstitial implant of 18 catheters treated for 5 days. As the source will be exposed once each hour, the number of transits is 120 transits per catheter × 18 catheters = 2160 total transits. The treatment control station software accounts for much of the transit time between dwell positions by subtracting at least a 0.1 s from each dwell time. Most of the transit dose to critical structures near the implant results from the return of the source at the completion of a treatment session. Transit dose may also be delivered to normal tissue external to the implant as the source enters and exits the treatment volume. The transit dose may be estimated by simulating a large implant with a multicatheter applicator and measuring the transit dose with either high-speed radiographic film or another sensitive dosimeter, such as, a MOSFET (metal oxide semiconductor-field effect transistor) or diode. Another important aspect of a PDR unit is the ability to transfer the source to and retract the source from the applicator multiple times without failure. Nucletron warrants the PDR unit to perform 25,000 transfers without failure. Tests should also be performed with varying angles of flexion of the transfer tubes to determine the angle of flexion that will cause a failure. At UTMDACC the foremost consideration for gynecological treatments, after verifying machine reliability, establishing adequacy of the room shielding, developing procedures for emergency response, and training all personnel in emergency response and machine operation, is the difference in the gynecological applicator designs. Comparison of Monte Carlo calculations and radiochromic film measurements in phantom around the PDR RAU Fletcher Williamson applicators to similar calculations and measurements for the FSD applicators used with the LDR RAU is currently underway (Gifford et al. 2005). Also being compared is a selected group of patients for whom there are CT scans of their FSD treatments with the 137Cs LDR RAU available. Gifford (2004) performed a Monte Carlo analysis including the effects of the shields on the doses these patients received. These results will be compared to Monte Carlo calculations assuming the same source positions in the applicators, but for 192Ir and the PDR applicators. A research program to study other applicators and perhaps to develop new applicators that can provide an improved dose distribution with 192Ir has begun.
Quality Control Because the PDR RAU is basically the same as the HDR RAU discussed in an earlier chapter, the quality control should address the same issues of source position accuracy, room and head radiation surveys, interlock checks, audiovisual device checks, and source calibrations. The difference with the PDR RAU is the importance of the nursing staff training, as discussed above. Because the PDR RAU provides an order of magnitude lower dose rate than the HDR RAU, medical physicists are not required to be present
108
John L. Horton et al.
throughout the treatment. However, because the PDR RAU has an order of magnitude higher dose rate than the LDR RAU, the nursing staff must be able to respond to an emergency situation in a very timely manner. The nursing staff must be the first line of defense. This means that the nursing staff must be well trained before beginning a PDR RAU program and frequent in-services to reinforce this training are a must. A well-planned and documented program of appropriate actions based on the errors codes from the computer is also essential.
Conclusions The PDR RAU offers several advantages over either manual afterloading or remote LDR afterloading. The PDR RAU provides the same average dose rate and radiobiologically equivalent treatment as continuous LDR treatment, allowing us to build on our many years of experience with continuous LDR treatments. PDR RAU provides personnel protection for both hospital staff and visitors. The nursing care and visiting time can occur between treatment pulses without extending the total treatment time. The PDR RAU provides the means to more tightly conform the radiation dose to the treatment volume than the manually afterloaded or remotely afterloaded LDR implants. However, a careful comparison of the applicators used and dose distributions delivered via the LDR RAU with the PDR RAU is essential before changing from a well-established technique. Finally, the nursing staff must be well trained and accept their new responsibilities before embarking on a PDR program.
References Armour, E. P., J. R. White, A. Armin, P. M. Corry, M. Coffey, C. DeWitt, and A. Martinez. (1997). “Pulsed low dose rate brachytherapy in a rat model: Dependence of late rectal injury on radiation pulse size.” Int J Radiat Oncol Biol Phys 38:825–834. Blank, L. Overview of LDR-PDR Differentiators. Presented at Nucletron Workshop: From Traditional to Modern Brachytherapy. Transferring from LDR to PDR. Amsterdam, The Netherlands, 2004. Brenner, D. J., and E. J. Hall. (1991). “Conditions for the equivalence of continuous to pulsed low dose rate brachytherapy.” Int J Radiat Oncol Biol Phys 20:181–190. Brenner, D. J., E. J. Hall, G. Randers-Pehrson, Y. Huang, G. W. Johnson, R. W. Miller, B. Wu, M. E. Vazquez, C. Medvedosky, and B. V. Worgul. (1996). “Quantitative comparisons of continuous and pulsed low dose rate regimens in a model late-effect system.” Int J Radiat Oncol Biol Phys 34:905–910. Fletcher, G. H. Textbook of Radiotherapy. Third Edition. Philadelphia, PA: Lea & Febiger, 1980. Fowler, J., and M. Mount. (1992). “Pulsed brachytherapy: The conditions for no significant loss of therapeutic ratio compared with traditional low dose rate brachytherapy.” Int J Radiat Oncol Biol Phys 23:661–669. Gifford, K. A 3-D CT Assisted Monte Carlo Evaluation of Intracavitary Brachytherapy Implants. Ph.D. dissertation. The University of Texas Graduate School of Biomedical Sciences. Houston, TX, 2004. Gifford, K., J. Horton, E. Jackson, T. Steger, M. Heard, F. Mourtada, A. Lawyer, and G. Ibbott. (2005).”Verification of Monte Carlo calculations around a Fletcher Suit Delclos ovoid with radiochromic film and normoxic polymer gel dosimetry.” Med Phys (Accepted for publication). Katz, A., and P. J. Eifel. (2000). “Quantification of intracavitary brachytherapy parameters and correlation with outcome in patients with carcinoma of the cervix.” Int J Radiat Oncol Biol Phys 48:1417–1425. Millar, W. T., J. H. Hendry, and P. A. Canney. (1996). “The influence of the number of fractions and biexponential repair kinetics on biological equivalence in pulsed brachytherapy.” Brit J Radiol 69:457–468. National Council on Radiation Protection and Measurements (NCRP). NCRP Report No. 40. Protection Against Radiation from Brachytherapy Sources. Bethesda, MD: NCRP, 1972. National Council on Radiation Protection and Measurements (NCRP). NCRP Report No. 91. Recommendations on Limits for Exposure to Ionizing Radiation. Bethesda, MD: NCRP, 1987. Sminia, P., C. J. Schneider, K. Koedooder, G. van Tienhoven, L. E. Blank, and D. G. Gonzalez. (1998). “Pulse frequency in pulsed brachytherapy based on tissue repair kinetics.” Int J Radiat Oncol Biol Phys 41:139–150.
Chapter 9
An Aperçu of Codes, Directives, Guidances, Notices, and Regulations in Brachytherapy Glenn P. Glasgow, M.S., Ph.D., F.A.A.P.M., F.A.C.R. Department of Radiation Oncology Loyola University Chicago Stritch School of Medicine Maywood, Illinois Brachytherapy Regulatory Overview (What’s New Since 1994 Summer School?) . . . . . . . . . 109 Is the regulatory climate hot (for more regulations) or cold (quiescent)? . . . . . . . . . . . . . . . . . . . . . 109 There is no excuse not to know the regulations: They are on the web! . . . . . . . . . . . . . . . . . . . . . . 110 What’s hot and what’s not in brachytherapy procedures? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Do you know where your sources are? Who else knows? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 You still haven’t read the “new” (4-24-02) USNRC 10 CFR 20, 32, and 35 regulations? . . . . . . . . 111 I’m an agreement state licensee! Why should I be concerned about changes in federal codes? . . . 112 Brachytherapy Regulatory Overview (What’s Unchanged Since 1994 Summer School?) . . . 112 10 CFR 19 (Notices, Instructions, and Reports to Workers; Inspections) . . . . . . . . . . . . . . . . . . . . 112 10 CFR 20 (Standards for Protection Against Radiation) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Regulatory Review: What’s Changed in 10 CFR 20 (Standards…Radiation)? . . . . . . . . . . . . . 112 10 CFR 20.1002/Scope; –.1003/Definitions, and –.1301/Dose Limits for Individual Members of the Public (USNRC 2002) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Regulatory Review: What’s Changed in 10 CFR 32 (Specific…Material)? . . . . . . . . . . . . . . . 113 Regulatory Review: What’s “New” in 126 Sections of 10 CFR 35 (Medical Use…Material)? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Components of 10 CFR 35 Applicable to All Forms of Brachytherapy (Tables 4A,B) . . . . . . . . . . 114 Some Components of 10 CFR 35 (F) Applicable to Manual Brachytherapy (Table 4C) . . . . . . . . . 115 Some Components of 10 CFR 35 (H) for Photon-emitting Remote Afterloaders (Tables 4D,E) . . . 115 Some Components of 10 CFR 35 (L) (Record Retentions) (Table 5) . . . . . . . . . . . . . . . . . . . . . . . . 120 Some Components of 10 CFR 35 (M) (Reports…Medical Events…Sources) (Table 6) . . . . . . . . . 120 Bulletins, Directives, Guidances, Information Notices, Newsletters, and Regulatory Summaries, for Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Bulletins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Directives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Guidances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Information Notices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Newsletters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Regulatory Summaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Recent NRC Activities—Specialty Boards and Training Programs . . . . . . . . . . . . . . . . . . . . . . . . . 122 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Brachytherapy Regulatory Overview (What’s New Since 1994 Summer School?) Is the regulatory climate hot (for more regulations) or cold (quiescent)? Regulations are like taxes: Everyone thinks they should be reduced! However, it rarely occurs. And, like taxes, when the “Code” is revised, taxes or regulations that appear to have been eliminated in one section often appear in other sections, often at higher rates! Occasionally, a lonely voice cries out for reform, but usually the cry goes unanswered. Mossman proposes that regulatory effectiveness and efficiency would be improved by three changes: (1) adoption of a dose-based rather than the current risk-based system;
109
110
Glenn P. Glasgow
(2) adoption of the International System of Units; and (3) establishing a single, independent office to coordinate nuclear regulations established by U.S. federal agencies and departments (Mossman 2003). However, the pace of regulatory change is glacial—usually advancing, not retracting! So, let us review the significant changes in the past 10 years in brachytherapy and the corresponding changes in regulations. The significant changes were: (1) Increased use of the World Wide Web Internet to disseminate regulations; (2) changes in types of brachytherapy procedures; (3) concerns about byproduct material security (or lack thereof) in medical facilities; (4) the adoption of new federal codes, and (5) implementation of these codes by Agreement States. The opinions expressed are exclusively those of the author, based on 30 years of experience working directly, and concurrently, with regulators under both an NRC license (Edward Hines VA Hospital) and Agreement State (Illinois) license (Loyola University Chicago Medical Center).
There is no excuse not to know the regulations: They are on the web! A comprehensive overview (Deye 1988), now dated, still serves as an excellent tutorial for those unfamiliar with organizations, such as the International Atomic Energy Agency (IAEA), International Commission on Radiological Protection (ICRP), and International Commission on Radiation Units and Measurements (ICRU), and codes and regulations. Table 1 presents useful, current web addresses for international organizations that often make recommendations that become the basis of subsequent federal regulations made by federal agencies. The National Council on Radiation Protection and Measurements (NCRP) makes, via statements and publications, recommendations on matters of radiation usage and, subsequently, safety issues. Through the Environmental Protection Agency (EPA), the federal agencies, Departments of Energy (DOE) and Transportation (DOT), Food and Drug Administration (FDA), currently coordinate their regulatory efforts. The U.S. Nuclear Regulatory Commission (NRC) is the lead agency for promulgating regulations related to brachytherapy sources and their use. Their regulations apply in 17 non-Agreement states and in federal facilities directly holding federal licenses. Agreement States, currently 33 in number, agree, under certain terms, to apply those regulations adopted by the NRC. The Conference of Radiation Control Program Directors (CRCPD) coordinates regulatory issues among its member states and proposes model state licensure and regulatory language. Their web site has electronic links to all state regulatory agencies. The Radiation Safety Officers Toolbox, via Idaho State University (http://www.physics.isu.edu/radinf/rsotool box.htm), provides direct links to the Code of Federal Regulations (CFR) discussed later in this chapter. Other organizations, such as the American Association of Physicists in Medicine (AAPM), Health Physics Society (HPS), Society of Nuclear Medicine (SNM), often issue position statements regarding proposed regulations. Higson (2001) offers a useful international view of the regulation of medical devices for public health and safety. Benedetto (1995) reviews how regulations arise in the United States.
What’s hot and what’s not in brachytherapy procedures? Brachytherapy prospers in the United States. Interstitial prostate implants with 125I and 103 Pd seeds are popular and show no signs of decaying! Remote high dose rate (HDR) afterloading procedures are increasing; mobile units serve multiple hospitals in heavily populated urban areas. Traditional single-session procedures (gynecology treatments with manually loaded 137Cs sources, interstitial sarcoma implants, etc.) are being replaced with multiple fractional HDR twice-a-day treatment regimens. Our original (7/88) HDR program at Hines VA Hospital was a Gamma Med IIi located in a 60Co teletherapy vault. Treating lung, esophagus, and the vagina, we performed about 49 procedures yearly. Since relocating (8/03) the program to a new facility at Loyola with a dedicated HDR vault, and opening a full gynecologic service and interstitial sarcoma and head and neck implant service, we perform 200+ cases yearly. Treatment of a new site (breast) with MammoSite® applicator additionally increased the total number of annual procedures.
9–An Aperçu of Codes, Directives, Guidances, Notices, and Regulations in Brachytherapy 111 Table 1. Useful Web Sites with Information about Radiation, Regulations, and Regulatory Issues Agency Conference Radiation Control Program Directors (CRCPD) Department of Energy (DOE) Department of Transportation (DOT) Environmental Protection Agency (EPA) Food and Drug Administration (FDA) Health Physics Society (HPS) International Atomic Energy Agency (IAEA) International Commission on Radiological Protection (ICRP) International Commission Radiation Units and Measurements (ICRU) National Council on Radiation Protection and Measurements (NCRP) Nuclear Regulatory Commission (NRC) Idaho State University
Internet Address (“http://www.) crcpd.org energy.gov dot.gov
Electronic Mail Address Not given on web page Not given on web page
[email protected]
epa.gov
Not given on web page
fda.gov
Not given on web page
hps.org iaea.org
[email protected] [email protected]
icrp.org
[email protected]
icru.org
[email protected]
ncrp.com
Not given on web page
nrc.gov
Not given on web page
physics.isu.edu/radinf/rsotoolbox
Not given on web page
Regulatory hurdles to treating with pulsed dose rate (PDR) remote afterloading devices have been lowered. Intravascular brachytherapy (IVB), popular for a short time, unfortunately exhibited a reasonably short half-life and was rapidly abandoned at many facilities.
Do you know where your sources are? Who else knows? A new international and national concern is the security of byproduct sources in medical facilities. Most medical licensees have small (multiple millicurie) quantities of long-lived byproduct materials (137Cs, 60Co, etc.), ideal components for a dispersal “dirty bomb.” The IAEA has developed an action plan to combat nuclear terrorism (Health Physics News and Notices 2002, 2003). These international efforts likely will lead to new national and state regulations requiring greater security for radioactive sources.
You still haven’t read the “new” (4-24-02) USNRC 10 CFR 20, 32, and 35 regulations? Shame on you! You will not be spared! This limited presentation (NB: think of it as “Regulations Lite”) reviews the cogent details of regulations that did not change as well as those that did.
112
Glenn P. Glasgow
I’m an agreement state licensee! Why should I be concerned about changes in federal codes? Agreement States have certain periods (5 years or more) within which state regulations must become compliant, at certain levels of compliance, with NRC regulations. A licensee may be in “regulatory transition,” a dangerous period of change! For example, at this time (2005) Illinois is enforcing state regulations based on NRC regulations in force prior to the 2002 regulatory changes, as they prepare new state regulations compliant with the recent changes in federal codes. Other states are in similar regulatory transitions.
Brachytherapy Regulatory Overview (What’s Unchanged Since 1994 Summer School?) 10 CFR 19 (Notices, Instructions, and Reports to Workers; Inspections) This long-standing regulation (USNRC 1981), with 14 sections, issued 12/18/1981, un-revised, remains in force. Table 2 lists seven important sections with brief comments about their content. Staying compliant with “instructions” to a constantly changing workforce is a regulatory challenge!
10 CFR 20 (Standards for Protection Against Radiation) These standards, consisting of 69 sections, are (with one exception discussed later) mostly unchanged from the 5/21/1991 release (USNRC 1991). Tables 3A and 3B list 10 key headings with brief comments about their content. Note that security of radioactive materials (RAM) is addressed in §20.1801.
Regulatory Review: What’s Changed in 10 CFR 20 (Standards…Radiation)? 10 CFR 20.1002/Scope; –.1003/Definitions, and –.1301/Dose Limits for Individual Members of the Public (USNRC 2002) 20.1002/Scope now states that “…limits in this part do not apply…to exposures from individuals administered RAM and released under §35.75….” 20.1003/Definitions adds “Occupational dose does not include…dose…from individuals administered RAM and released under §35.75….” “Public dose does not include…dose…from individuals administered RAM and released under §35.75….” 20.1301/ Dose Limits for Individual Members of the Public now adds to the exclusion of dose from RAM in sanitary sewers, the following: “…does not exceed 0.1 rem (1mSv) in a year exclusive of the dose contributions from background radiation, from any medical administration to the individual, from individuals administered RAM and released under §35.75, from voluntary participation in medical research programs….” Also, added: “…a licensee may permit visitors to an individual…to receive a radiation dose greater than 0.1 rem (1 mSv) if (1) the radiation dose…does not exceed 0.5 rem (5 mSv) and (2) the authorized user has determined before the visit that it is appropriate.”
9–An Aperçu of Codes, Directives, Guidances, Notices, and Regulations in Brachytherapy 113 Table 2. 10 CFR 19 (Notices, Instructions, & Reports to Workers; Inspections) (Partial Contents) Section
Major contents of section
.3/Definitions
Workers, licenses, restricted areas defined.
.11/Postings notices to workers
a) Post regulations, (i) license & its conditions; (ii) operating procedures; (iii) violations; b) Documents, forms must be conspicuous.
.12/Instructions to workers
Inform about: a) storage, use RAM; b) health protection problems; c) procedures to reduce exposures; d) regulations; e) report conditions, violations; f) response to warnings; g) their exposures.
.13/Notification & reports to individuals
a) Written exposure reports; b) annual exposure reports per workers request; c) other provisions not stated here.
.14/Presence of licensee’s & workers representatives during inspections;
a) Licensee to allow inspections; b) inspectors may meet workers; c) representatives may accompany inspectors during inspections; d) other provisions not stated here.
.15/Consultations with workers during inspections
a) Inspectors may consult privately with workers; b) workers may consult privately with inspectors.
.16/Requests by workers for inspections
Workers may request, without retribution, inspections.
Regulatory Review: What’s Changed in 10 CFR 32 (Specific…Material)? The “new” (4-24-2002) 10 CFR 32 (Specific Domestic Licenses to Manufacture or Transfer Certain Items Containing Byproduct Material) changes are only notational bookkeeping, changing the paragraphs numbers and sections in Part 32 to correspond with the corresponding sections of the new 10 CFR 35.
Regulatory Review: What’s “New” in 126 Sections of 10 CFR 35 (Medical Use…Material)? With 126 sections, we focus only on those of direct interest or applicability to brachytherapy. Tables 4A through 4E summarize, using some shorthand notations, the major contents of the important sections. The bulk of regulatory changes relative to brachytherapy occur in these sections.
114
Glenn P. Glasgow Table 3A. Unchanged Components of 10 CFR 20 (Standards for Protection Against Radiation) Section
.1101/Radiation Protection Program (RPP)
Major contents of section a) RPP must be developed, documented, implemented, commensurate with extent & scope of licensed activities; b) ALARA for occupational & public doses; c) Annually review RPP content & implementation.
.120/Occupational Dose Limits
a) Annual TEDE 0.05 Sv; sum of DDE & CDE of organs 0.5 Sv; eye DE 0.15 Sv; shallow skin or extremity DE 0.5 Sv; b) Excess DEs must be planned; c) Other provisions not stated here.
.1208/Dose to an Embryo/Fetus
a) 5 mSv dose to embryo/fetus, entire pregnancy, occupational exposure of mother; b) Avoid variations in uniform monthly doses; c) Dose is sum of DDE of mother & radionuclides in mother & embryo/fetus; d) Other provisions not stated here.
.1502/Individual Monitoring of External/Internal Occupational Doses
a) Those likely DE 10% of limits; b) Those in high & very high radiation areas; c) Those likely to receive CEDE of 10% from radionuclides; d) Other provisions not stated here.
.1801/Security of radioactive materials
a) Secure from unauthorized removal or access licensed material stored in controlled or unrestricted areas; b) Licensed material not in storage shall have control and constant surveillance.
ALARA: as low as reasonably achievable CDE: committed dose equivalent CEDE: committed effective dose equivalent
DDE: deep-dose equivalent DE: dose equivalent TEDE: total effective dose equivalent
Components of 10 CFR 35 Applicable to All Forms of Brachytherapy (Tables 4A,B) A new term, “Authorized Medical Physicist (AMP)”, and the training thereof, is defined, as well as types [LDR (low dose rate), PDR, HDR] of remote afterloading units (RAU), including medium dose rate (MDR). Mobile services and medical events are new additions. Roles of management, the radiation safety officer (RSO), and authorized users (AU) supervision of individuals are explained. Dose prescriptions, or written directives (WD) details and procedures are enumerated. Table 4B notes source inventories are now at 6-month intervals. Section 35.75 explains new release criteria for patients (NB: See Glasgow (2002b) for a discussion relative to 192Ir seed patients) (USNRC 1997). Some requirements for mobile medical services are in this section, as well as rules for decay-instorage of RAM.
9–An Aperçu of Codes, Directives, Guidances, Notices, and Regulations in Brachytherapy 115 Table 3B. Unchanged Components of 10 CFR 20 (Standards…Protection…Radiation) Section
Major contents of section
.1901/Caution Signs
Radiation symbol (trefoil) color schema (magenta, purple, black) on yellow & design defined.
.1904/Labeling Containers Radioactive Materials
a) Containers of RAM must be marked either “CAUTION” or “DANGER”, RADIOACTIVE MATERIAL; b) Label must identify quantity, date, radiation levels, kind of material; c) Remove/deface labels on empty containers.
.1906/Receiving/Opening Packages
a) Package receipt & monitoring procedures; b) Carrier notified if wipe test or radiation levels exceed limits; c) Package opening procedures; d) Other provisions not stated here.
.1501/Surveys and Monitoring
a) Make necessary surveys; b) Equipment used for surveys calibrated; c) Excluding direct/indirect pocket dosimeters, NVLAP accreditation for badge processor.
.2001/Waste Disposal
a) By transfer to authorized recipient; b) By decay in storage; c) By effluent release within limits; d) Others provisions not stated here.
NVLAP: National Voluntary Laboratory Accreditation Program
Some Components of 10 CFR 35 (F) Applicable to Manual Brachytherapy (Table 4C) One major change is a requirement to decay output or source activities in 1% intervals. Another section adopts AAPM good practices, per various protocols, for quality assurance of therapy planning systems, as a regulation! (NB: To present and future task group members: Be careful what you write lest it become a required regulation!)
Some Components of 10 CFR 35 (H) for Photon-emitting Remote Afterloaders (Tables 4D,E) In the nine sections, the most significant change is the requirements for MDR and PDR units. Physicians other than AUs, trained in MDR and PDR operation, emergency procedures, and source removal, may work under the supervision of an AU [NB: I will denote them as “substitute authorized users (SAU)”]. For the initial treatment, the AMP and AU or SAU must be present; during subsequent (continuation) treatments, the AMP, AU, or SAU must be immediately available. (NB: In the medical world, that’s by pager!) These requirements are less onerous than the prior requirements of the AU always being present during all treatments. These changes may (or may not!) allow PDR to develop in the United States. Another section adopts AAPM good practices, per various protocols, for quality assurance of RAU therapy planning systems, as a regulation!
116
Glenn P. Glasgow Table 4A. Components of 10 CFR 35 (A, B) Applicable to All Forms of Brachytherapy Section
.2/Definitions
Major contents of section a) Authorized medical physicist defined; b) LDR, MDR, HDR, PDR defined; c) Mobile medical service defined; d) Medical event (no more misadministrations!) explained; e) Manual prescribed dose (total sources strength & time, or dose per WD) given; f) Remote prescribed dose (total dose & dose per fraction per WD) given.
.24/Authority Radiation Protection Program
a) Defines a stronger management role; b) Defines & strengths RSO role.
.27/Supervision
Explains role of authorized user (AU) & supervised individuals with respect to process & procedures with RAM.
.40/Written Directives (WD)
a) Written directives required or oral directives with 48 h for written; b) HDR: radionuclide; site, fx dose, # fxs, total dose; c) Others; before tmt: radionuclide; site, dose; before finish: # sources, total source strength & time (or total dose); revisions allowed during treatment.
.41/Procedures…written directives
a) ID patient; b) Administration per WD; c) Check manual, computer dose calculations; d) confirm console data.
.51/Training authorized medical physicist
a) Board certifications; b) Degrees + 1 y training + 1 y experience; c) Preceptor’s written statement regarding training.
Requirements for dosimetry systems (DS), full calibrations (FC), and spot-checks (SC) are described, including those for mobile services.
9–An Aperçu of Codes, Directives, Guidances, Notices, and Regulations in Brachytherapy 117 Table 4B. Some Components of 10 CFR 35 (C) Applicable to All Forms of Brachytherapy Section .67/Requirements for possession
Major components of section a) Leak tests (<5 nCi) before 1st use, 6 mos.; b) Exempt 192Ir seeds in ribbons & unused sources; c) 6-months inventory.
.75/Release…patients containing…RAM
a) OK if others TEDE < 5 mSv/y; b) Instruction if others TEDE > 1 mSv/y.
.80/Mobile medical services
a) Facility agreement letters; b) On-site, before use survey meter checks; c) Post-treatment surveys; d) Possession licenses required for all sites.
.92/Decay in storage
a) T1/2 < 120 d; decay to background level; b) Remove labels; keep records.
Table 4C. Some Components of 10 CFR 35 (F) Applicable to Manual Brachytherapy Section .404/Surveys after… implant & removal
Major contents of section a) After implant; source accountability; b) After source removal; keep records.
.406/Source accountability
a)…at all times…in storage& use; record.
.410/Safety instructions
a) Initially, annually…to caregivers; b) Size, type, handling, shielding, visitor.
.415/Safety precautions
a) No room sharing with regular patients; b) Post room (RAM) & visitor limits; c) Emergency equipment for source retrieval from or in patient.
.432/Source calibrations (post 10/24/04)
a) Determine output or activity; b) Positioning in applicators per “protocols”; c) Decay outputs/activities at 1% intervals; keep records.
.433/Decay 90Sr sources
Only AMP shall calculate decayed activity & keep records.
.457/Therapy-related computer systems
a) Acceptance testing per “protocols”; b) Source input parameters; c) accuracy of dose/time at points; isodose & graphics plots; d) localization image accuracy.
118
Glenn P. Glasgow Table 4D. Some Components of 10 CFR 35 (H) for Photon…Remote Afterloaders Section
Major components of section
.604/Surveys of patients
Before releasing patient…survey patient & RAU to confirm…returned to safe.
.605/Installation,…,repair
a) Certain source work, i.e., install, adjust, etc., by licensed person; b) For LDR RAU, licensed person or AMP can do certain source work; record.
.610/Safety procedures
a) Secure unattended RAU; b) Only approved individuals present in room; c) No dual operations; d) Written procedures for abnormal situations; posted copies; initial/annual instructions with drills; records.
.615/Safety precautions
a) Control access with interlock; b) Area monitors; c) CCTV/audio for all except LDR RAU; d) For MPD/PDR an AMP & AU or operatoremergency response MD present at initiation & immediately available during treatments; e) For HDR an AU and AMP physically present at initiation, but, during continuation, AMP & AU or operator- emergency response MD; f) Emergency equipment for unshielded source or source in patient.
.657/Therapy-related computer system
a) Acceptance testing per “protocols”; b) Source input parameters; c) Accuracy of dose/time at points; isodose & graphics plots; d) Localization image accuracy; e) Electronic transfer to RAU accuracy.
9–An Aperçu of Codes, Directives, Guidances, Notices, and Regulations in Brachytherapy 119 Table 4E. Some Components of 10 CFR 35 (H) for Photon…Remote Afterloaders Section .630/Dosimerty system (DS) equipment
Major contents of section a) Except for LDR RAUs, NIST/ ADCL calibrated DS; b) 2 y & after service; or, c) 4 y, if intercompared with calibrated DS within 18 to 30 mo.& < 2% change.
.633/Full calibrations (FC) of RAUs
a) Before 1st use; at source exchanges &/or repairs to exposure assembly; b) For T1/2 >75 d, excluding LDR RAUs, quarterly; c) LDR RAUs yearly; d) FC: 5% output/1 mm positions, source retraction, timer accuracy/linearity; e) Tube lengths & functions; f) Quarterly autoradiographs of LDR RAU sources; g) Decay outputs/activities at 1% intervals; h) FC & decay by AMP; keep records; for LDR RAU can use manufacturer’s data for FC.
.643/Periodic spot-checks (SC) of RAUs
a) For LDR RAUs, before1st treatment; for other RAUs 1st use daily; b) Per WP by AMP; c) AMP review by 15 d; d) SC includes: interlocks, status lights, audio & CCTV, emergency equipment, source position monitors, timer, clocks, decayed source activity.
.647/Additional requirements…mobile RAUs
a) Survey meter checks; b) Source inventory; c) All .643 checks; d) Interlocks, status lights, radiation monitors, source positioning, before 1st use, simulated treatment at each address.
ADCL: Accredited Dosimetry Calibration Laboratory NIST: National Institute of Standards and Technology
120
Glenn P. Glasgow Table 5. Some Components of 10 CFR 35 (L) (Record Retentions) Record Retention Requirement
Section
Duration of license
.2024/RPP (b) RSO authority.
Duration of program (device)
.2610/Safety procedures for device.
Five years
.2041/Procedures for WP; .2026/RPP changes.
Three years
.2040/WDs; .2061/Meter calibrations; .2067/Leak tests & inventories; .2070/Surveys; .2075/Patient release; .2080/Mobile services; .2092/Decay in storage; .2310/Safety instructions; .2404/Implants & source removals; .2406/Source accountability; .2432/Source calibrations; .2433/Sr-90 decays; .2605/RAU installation, repairs; .2632/Full calibrations; .2643/Spot checks; .2647/Additional mobile records.
Some Components of 10 CFR 35 (L) (Record Retentions) (Table 5) Table 5 summarizes the duration (for license, for program, and for 5 and 3 years) requirements for the retention of records.
Some Components of 10 CFR 35 (M) (Reports…Medical Events…Sources) (Table 6) Misadministration is no more! We now have medical events (ME)! Very careful reading is required for this section, as the ME depends, in some cases, on the difference (presumably lower or higher) in delivered dose and prescribed dose (PD), and in other cases, in exceeding the PD. Moreover, the definitions are not in medical physics terms of absorbed dose in gray (Gy); rather, they are in health physics terms of effective dose equivalent (EDE), and shallow dose equivalent (SDE), in sievert (Sv). Recall that in partial organ irradiation in health physics, organ or tissue weighting factors apply in calculating EDE. As a brachytherapy ME will likely involve adjacent organs, some judgment may be required in deciding on the correct EDE in an ME. Table 6 summarizes the reporting of medical events; reporting requirements are similar to pre-2002 regulations.
9–An Aperçu of Codes, Directives, Guidances, Notices, and Regulations in Brachytherapy 121 Table 6. Some Components of 10 CFR 35 (M) (Reports…Medical Events…Sources) Section
Major contents of section
.3045/Report/notification medical event (excluding patient intervention) (1)
Dose differs from PD more than 0.05 Sv EDE, 0.5 Sv organ/tissue & SDE skin, and, TD, and, TD delivered differs from PD by +20% or falls outside PD range; or single fraction delivered dose differs from single fraction PD +50%.
.3045/Report/notification medical event (excluding patient intervention) (2)
Dose exceeds 0.05 Sv EDE, 0.5 Sv organ/tissue & SDE skin, and, TD from wrong: a) byproduct material; b) administration route; c) person; d) treatment mode; e) leaking source.
.3045/Report/notification medical event (excluding patient intervention) (3)
Excluding migrating permanent implant seeds, dose to skin/organ/tissue other than treatment site that exceeds 0.5 Sv organ/tissue and +50% dose expected from WD.
.3045/Report/notification medical event (excluding patient intervention) (3) (b)
Report any patient interventions producing permanent/physiological damage.
.3045/Report/notification medical event (excluding patient intervention) (3) (c, d)
Notify NRC next calendar day after ME with written report in 15 days; notify referring MD & patient unless referring MD chooses not to for medical reasons; details of reports omitted here.
.3067/Report leaking source.
Report >5 nCi removal contamination within 5 days.
EDE: effective dose equivalent SDE: superficial dose equivalent
PD: prescribed dose TD: total dose
Bulletins, Directives, Guidances, Information Notices, Newsletters, and Regulatory Summaries, for Brachytherapy The USNRC issues to licensees bulletins, directives, guidances, information notices, newsletters, and regulatory summaries as new issues not covered in regulations arise and must be addressed. In some cases, these documents endure for many years, and may actually be incorporated by Agreement States into their regulatory statutes.
Bulletins Apparently there are no recent bulletins pertaining to brachytherapy; the last one was “Release of Patients after Brachytherapy with Remote Afterloading Devices” (USNRC 1993).
Directives Directives appear in several forms. FC86-4, Revision 1–Information Required for Licensing Remote Afterloading Devices, a long-standing (1986) policy and guidance directive, explained the contents for NRC license applications for RAUs (USNRC 1986). While it is not currently on the NRC web site, Illinois (and, I imagine, other states) adopted it, with some changes, into their licensing process for RAUs. FC83-20, Revision 2–Facility Interlocks and Safety Devices for High, Medium, and Pulsed Dose-Rate Afterloading Units, is not on the NRC web site. As the title implies, this release clarified the requirements for
122
Glenn P. Glasgow
interlocks and safety devices (USNRC 1983). It appears that issues raised are addressed in the 2002 10 CFR 35 revisions.
Guidances Guidances often discuss evolving technologies. For example, as intravascular brachytherapy developed, the NRC issued several guidance documents (Glasgow 2002a, USNRC 2004a). These were necessary as the “new” 10 CFR 35 applies only to photon-emitting RAUs; beta-emitting RAUs fall into the “emergent technology” category evaluated on a “case-by-case” basis.
Information Notices Information Notices advise licenses of recent concerns usually arising from medical events reported to the NRC. A recent notice discussed failures of HDR RAUs (USNRC 2003).
Newsletters Newsletters, notably “Nuclear Materials Safety and Safeguards” (NMSS), announce medical events and enforcement actions against those who violate regulations. A recent one reported on a hospital’s failure “…to secure…licensed material…” (USNRC 2004b).
Regulatory Summaries Regulatory Summaries often clarify issues about the interpretation of regulations, such as the calibration measurements for brachytherapy sources (USNRC 2002b).
Recent NRC Activities—Specialty Boards and Training Requirements The 2002 revisions in 10 CFR 35 did not address personnel training. On March 30, 2005, the NRC published the final rule (USNRC 2005) regarding specialty boards and personnel training. The rule identifies (on the NRC web-site, not in the published rule!) various approved specialty boards and describes pathways for approval of RSOs, AMPs, authorized nuclear pharmacists (ANPs), and physicians using many forms of by-product materials. This flexible rule offers multiple pathways by which individuals may achieve authorization to perform various tasks or assume authorized titles, e.g., RSO, AMP, ANP, or physician authorized user, while maintaining the integrity of the approval process. One major pathway is the educational degree → experience → specialty examination → certification path. Another major pathway is the supervised experience → preceptor statement path. Table 7 (replete with necessary acronyms) shows five ways an individual, depending on education, experience, and certification status, can achieve authorization as a RSO. For non-physicians, the education requirements are either (a) a bachelor’s degree or graduate degree in physical science, or, engineering or biologic science with 20 college credits in physical science, or, (b) a master’s degree or doctorate in physics, medical physics, or physical science, engineering, or applied mathematics. Experience requirements vary from 1 year to 5 years depending on the authorization, and are shorter for those with higher degrees. Generally, experience must be gained under a certified medical physicist (CMP) or authorized individual, and documented. Preceptors must document the successful completion of any structured training programs and attest to the individual’s competencies and abilities to perform learned tasks independently. In some instances, structured didactic training programs include classroom and laboratory training in topical areas. For example, topical areas for an RSO include: radiation physics and instrumentation, radiation protection, mathematics pertaining to use and measurement
9–An Aperçu of Codes, Directives, Guidances, Notices, and Regulations in Brachytherapy 123 of radioactivity, radiation biology, and radiation dosimetry are allowed. Table 8 shows the requirements for approval as an AMP or ANP. [NB. No single-page synopsis with two tables can adequately describe the seven pages of new federal regulations on specialty boards and training requirements; interested readers are advised to study the new regulations in their entirety to fully understand them! (USNRC 2005)] Similar tables (not presented here) describe approval processes for physician authorized users for use of by-product materials. Table 7. Some Components of 10 CFR 35 (J) Requirements for Radiation Safety Officers Person
Degree or Certification
Experience
(1) Radiation Safety Officer
B or GD in PS; or, E or BS with 20 cc in PS;
and 5 or more and Passes yrs in HP Exam including 3 yrs in AHP
Or, (2) Radiation Safety Officer
M or PhD in P, MP, or PS, E, AM
and 2 yrs full-time training in MP under supervision by CMP, or, in CNM, by physician AU
or, (3) Radiation Safety Officer
1 yr full-time RS under supervision by RSO
Certification Examination
Classroom Laboratory Training
Preceptor Statement
Special Training
and Passes Exam
and 200 h in topical areas
or, (4) Radiation Safety Officer
CMP
and applicable experience
and has written attestation by preceptor
and training in RS, regulatory issues, & emergency procedures
or, (5) Radiation Safety Officer
AU, AMP, or ANP on license
and applicable experience
and has written attestation by preceptor
and training in RS, regulatory issues, & emergency procedures
ANP = Authorized Nuclear Pharmacist B = Bachelor’s Degree BS = Biological Science CC = College Credits E = Engineering GD = Graduate Degree M = Master’s Degree PhD = Doctoral Degree AHP = Applied Health Physics
PS = Physical Science CMP = Certified Medical Physicist RS = Radiation Safety AU = Authorized User CNM = Clinical Nuclear Medicine MP = Medical Physicist or Physics RSO = Radiation Safety Officer AMP = Authorized Medical Physicist HP = Health Physics
124
Glenn P. Glasgow Table 8. Some Components of 10 CFR 35(J) Requirements for AMPs and ANPs
Person
Degree or Certification
Experience
Certification Examination
(1) Authorized Medical Physicist
M or PhD in P, MP, or PS, E, AM
and 2 yrs under supervision by CMP
and Passes Exam
or, (2) Authorized Medical Physicist
M or PhD in P, MP, or PS, E, AM
and 2 yrs in CRF under supervision by AU eligible physician
and Passes
or, (3) Authorized Medical Physicist
M or PhD in P, MP, or PS, E, AM
and 1 yr fulltime training in MP and 1 yr full-time experience by AMP eligible MP
(1) Authorized Nuclear Pharmacist
Pharmacy; or, passed FPGEC exam
4000 h in nuclear pharmacy
Classroom Laboratory Training
Preceptor Statement
Special Training
and has written attestation of “competency & independency” by MP preceptor
and training in device operation, clinical use, and treatment planning systems
and Passes Exam
or, (2) Authorized Nuclear Pharmacist
ANP = Authorized Nuclear Pharmacist CMP = Certified Medical Physicist RS = Radiation Safety CRF = Clinical Radiation Facility P = Physics AM = Applied Mathematics PhD = Doctoral Degree FPGEC = Foreign Pharmacy Grad Exam Committee
Current, active license
700 h in structured program with 200 h in topical areas
and has written attestation of “competency & independency” by preceptor ANP
PS = Physical Science AMP = Authorized Medical Physicist AU = Authorized User E = Engineering MP = Medical Physicist or Physics MD = Master’s Degree
Conclusions Understanding codes, regulations, and license conditions has to be the least exciting part of a medical physicist’s job! The federal codes are the basis for state codes, but state codes are not identical to federal codes, even in Agreement States. There is no joy being involved in a medical event or discovering a license
9–An Aperçu of Codes, Directives, Guidances, Notices, and Regulations in Brachytherapy 125 violation during an inspection! Compliance with myriad regulations and license conditions is a challenge. However, by knowing the codes and regulations, one can write a better license with which it is easier for one to comply! To be forewarned is to be forearmed! May you always be in compliance with your federal or state license!
References Benedetto, A. R. (1995). “The Brachytherapy Regulatory Environment: Organization of a Radiation Safety Program” and its appendix “Compilation of Selected Rules for Brachytherapy Use of USNRC Licensed Materials” by J. O. Eichling in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (Eds.). AAPM 1994 Summer School Proceedings. Madison, WI: Medical Physics Publishing, pp. 163–173 and 174–183, 1995. Deye, James A. “Codes and Regulations, Radiation” in Encyclopedia of Medical Devices and Instruments. John G. Webster (Ed.). New York: John Wiley & Sons, 1988. Glasgow, G. P. (2002a). “Nuclear Regulatory Commission regulatory status of approved intravascular brachytherapy systems.” Cardiovasc Radiat Med 3:1–11. Glasgow, G. P. (2002b). “Is an Ir-192 permanent seed implant feasible for prostate brachytherapy?” Brachytherapy 1:195–203. Health Physics News and Notices (2002). “IAEA action plan to combat nuclear terrorism.” Health Physics 82:908–909. Health Physics News and Notices (2003). “IAEA and UPU join forces to protect mail.” Health Physics 84:129–130. Higson, G. R. Medical Device Safety: The Regulation of Medical Devices for Public Health and Safety. Bristol, England: Institute of Physics Publishing, 2001. Mossman, K. L. (2003). “Restructuring nuclear regulations.” Environ Health Perspect 111:13–17. U.S. Nuclear Regulatory Commission (1981). Code of Federal Regulations, title 10, chapter 1, part 19 (Notices, Instructions, and Reports to Workers; Inspections). [Online] http://www.nrc.gov/reading-rm/doccollections/cfr/part019. [December 4, 2004]. U.S. Nuclear Regulatory Commission (1983). FC83-20, Revision 2–Facility Interlocks and Safety Devices for High, Medium, and Pulsed Dose-Rate Afterloading Units. Washington, DC. U.S. Nuclear Regulatory Commission (1986). FC86-4, Revision 1–Information Required for Licensing Remote Afterloading Devices. Washington, DC. U.S. Nuclear Regulatory Commission (1991). Code of Federal Regulations, title 10, chapter 1, part 20 (Standards for Protection Against Radiation; Final Rule) [Online] http://www.nrc.gov/reading-rm/doccollections/cfr/part020. [December 4, 2004]. U.S. Nuclear Regulatory Commission (1993). Bulletin 93-01: Release of Patients after Brachytherapy with Remote Afterloading Devices. [Online] http://www.nrc.gov/reading-rm/doc-collections/gen-comm/bulletins/1993. [December 4, 2004]. U.S. Nuclear Regulatory Commission (1997). Regulatory Guide 8.39: Release of Patients Administrated Radioactive Materials. [Online] http://www.nrc.gov/reading-rm/doc-collections/reg-guides/occupational-health/active/ index.html. [December 4, 2004]. U.S. Nuclear Regulatory Commission (2002a). Code of Federal Regulations, title 10, parts 20, 32, and 35 (Medical Use of Byproduct Material: Final Rule). Federal Register, vol. 67, no. 79 (April 24):20250–20397. [Online]. Washington, DC: Nuclear Regulatory Commission. http://www.nrc.gov/reading-rm/doc-collections/ cfr/part020/part032/035. [December 4, 2004]. U.S. Nuclear Regulatory Commission (2002b). NRC Regulatory Issue Summary 2002-20, “Clarification of Requirements under 10 CFR 35.432, “Calibration Measurements of Brachytherapy Sources” [Online] http://www.nrc.gov/materials/miau/med-use-toolkit/reg-issues-sum.html. [December 4, 2004]. U.S. Nuclear Regulatory Commission (2003). NRC Information Notice 2003-21: High-dose rate remote afterloader equipment failure. November 24, 2003. [Online] http://www.nrc.gov/materials/miau/med-use-toolkit/infonotices.html. [December 4, 2004].
126
Glenn P. Glasgow
U.S. Nuclear Regulatory Commission (2004a). Licensing guidance Novoste and Guidant Intravascular Brachytherapy (IVB) systems. May 21, 2004. [Online] http://www.nrc.gov/materials/miau/med-use-toolkit.htmlguidance. [December 4, 2004]. U.S. Nuclear Regulatory Commission (2004b). Newsletter NUREG/BR-0117/04-2: Nuclear Material Safety and Safeguards [Online] http://www.nrc.gov/reading-rm/doc-collections/nureg/brochures/br0117/04-2.pdf. [December 4, 2004]. U.S. Nuclear Regulatory Commission (2005). USCFR title 10, part 35 (Medical Use of Byproduct Material— Recognition of Specialty Boards; Final Rule). Federal Register, vol.70, no. 60 (March 30):16366–16367. [Online]. Washington DC: Nuclear Regulatory Commission. http://www.nrc.gov/reading-rm/doccollections/cfr/part35. [2005, April 19].
Chapter 10
Brachytherapy Facility Design Glenn P. Glasgow, M.S., Ph.D., F.A.A.P.M., F.A.C.R. Department of Radiation Oncology Loyola University Chicago Stritch School of Medicine Maywood, Illinois Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Factors in Program Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Radiation Control and Shielding Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 Manual Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 LDR and Remote PDR Afterloading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 Remote HDR afterloading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Contained Liquid Radionuclide Therapies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Manual Procedure Source Room Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Receipt of Therapeutic Radioactive Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Source Calibration and Quality Assurance Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 Security of TRAM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 Holding TRAM for Decay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Imaging (Source Localization) Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Conventional Imaging Rooms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Portable Imaging Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Computer Tomography and CT/Simulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Isodose/Time Computation Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 Conventional Treatment Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 Pulsed Remote Afterloading Fixed Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 PDR Room Shielding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 PDR Facility Issues: Sources, Security, Calibration, and Quality Control Procedures . . . . . . . . . . 139 Features of a PDR Treatment Room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Permanently Installed HDR RAU Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 HDR Facility Issues: Sources, Security, Calibration, and Quality Control Procedures . . . . . . . . . . 142 Features of HDR Treatment Rooms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 Features of HDR RAUs in Operating Rooms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 HDR Room Designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 Mobile HDR Facilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 Intrafacility Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 Interfacility Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Introduction What’s new in brachytherapy? How do recent changes in the popularity of brachytherapy procedures affect the design of a brachytherapy program or facility? If, as a young medical physicist hired for a new facility, you are asked to make recommendations about brachytherapy programs or services, plan the facility, and implement your plan, what factors would you consider in these processes? Where would you find shielding data? Where would you find information to help you with these tasks? This chapter will provide some, but not all, answers to these and related questions. Beginning with the last question, recent literature searches for articles on brachytherapy facility design yields few new articles (Glasgow 1999, 2002)
127
128
Glenn P. Glasgow
or abstracts (St. Germain 2002). A popular text on shielding techniques (McGinley 2002) offers new information on maze design for remotely afterloaded high dose rate (HDR) brachytherapy facilities. The National Council on Radiation Protection and Measurement (NCRP) has an active workgroup with draft material (NCRP 2004) that includes a section on facility design. Fortunately, older material, discussed later, about facility design (Broadbent 1984; Gitterman and Webster 1984; Glasgow and Corrigan 1995; Glasgow 1995; Glasgow et al., 1993; Houdek et al., 1994; McKenzie et al., 1986; Stedeford, Morgan, and Mayles 1997; Thomadsen et al., 1983) remains relevant even thought some of the material is not so readily available. This is not a shielding tome; we discuss shielding only in the context of facility design. Classical shielding calculation methods still apply (NCRP 1972, 1976; McGinley 2002). So, lacking definitive new literature for guidance, let’s begin anew!
Factors in Program Design Brachytherapy is widely used. From 5% to 15% of radiotherapy patients may be candidates for brachytherapy. Eighteen or more anatomic sites have been treated with some form of brachytherapy. Figure 1, a 3-year review of 45 major journal articles, shows the percentage of the articles reporting results for 10 anatomic sites. Cervix is the most commonly treated; about 16% of cervical patients receive HDR brachytherapy (Eifel et al., 2004). Interstitial prostate implants with 125I and 103Pd seeds are well established (NB: a new 131 Cs source is in clinical test trials (IsoRay 2004). Breast treatment with the MammoSite® balloon applicator is sufficiently popular that a dedicated single-channel HDR remote afterloader unit (RAU) (Varian 2004) is marketed for that specific application. Remote HDR procedures are common for some sites; special three-channel HDR RAUs (Nucletron 2004; Varian 2004) for gynecologic treatments are available. In large urban areas, hospitals in close proximity use shared mobile HDR units to serve multiple locations. Not all brachytherapy succeeds. Intravascular brachytherapy (IVBT), popular for a time, is being abandoned at many facilities. Those designing a brachytherapy program and facility must recognize: (a) treatment programs are highly dependent on specific physicians and their interests and skills; (b) biologically and clinically, the major goal of HDR brachytherapy remains to achieve, with acceptable complication rates, cure rates equivalent to those achieved with LDR brachytherapy; (c) business plan projections (conjectures) about potential patient numbers for specific programs usually are optimistically high by a factor of two to three (NB: often ignoring similar programs in adjacent facilities!); (d) brachytherapy is, for all staff involved, more labor intensive per case than external beam procedures; and (e) a facility will likely last longer than either the
Figure 1. Of 45 articles reporting results for 10 anatomic sites, the percentage of articles by site.
10–Brachytherapy Facility Design
129
physicist designing it or the radiation oncologist using it. Often, particularly in smaller facilities, programs develop because radiation oncologists new to the staff have skills in particular types of brachytherapy and want to do what they know! Conversely, radiation oncologists without prior brachytherapy training are unlikely to lead the charge in developing new brachytherapy programs! Hence, selection of particular brachytherapy technologies depends on realistic estimates of the numbers of patient candidates for treatments; the stability of the radiation oncology staff and their interest and prior experience with the proposed technologies; and realistic estimates of the capital costs, annual costs, and support staff (dosimetrists and physicists) available to support the programs. Brachytherapy is a multifaceted discipline; no one brachytherapy technology is best suited for treating all cancer types and anatomic sites. My on-line literature search of “high dose rate brachytherapy” returned 1669 articles. A quick review indicates some forms of brachytherapy, such as HDR 192Ir brachytherapy, may yield acceptable results in one anatomic site but induce too many complications or inferior cure rates in other sites. Pulsed dose rate (PDR), which presents the technical and radiation protection advantages of HDR while retaining the radiobiological advantages of LDR, is unused in the United States; radiation oncologists are reluctant to adopt it even if it appears to offer advantages over conventional technologies. Multiple brachytherapy technologies likely are required; some necessary but specific programs may treat very few patients yearly. The design of the facility, support space, and selection of quality assurance (QA) equipment are, in some ways, the easiest parts of implementing a program.
Radiation Control and Shielding Calculations Each form of brachytherapy offers the potential of unexpected, unnecessary, and potentially harmful excess radiation exposure to both patients and supporting medical personnel. Regulatory (federal or state) controls apply. Proper facility design can enhance radiation control and reduce personnel dose equivalents (DE). Least anyone not know the applicable DE for occupational personnel and members of the public, they are summarized in Table 1, along with the comments regarding their role in shielding calculations. Shielding calculations, using older concepts and terminologies (exposure rate, X, [R/h], activity, A, [Ci], specific gamma-ray constants at 1 m, Γ, [R m2/h/Ci], inverse reduction of radiation exposure rate with the square of the distance, d, in m, from the source; X=A Γ/d2) are straightforward (NB: they can, of course, be performed in more current medical physics notation; surprisingly, one gets the same shielding thickness!). Patient attenuation, discussed later, is commonly neglected.
Manual Procedures Traditional single-session, in-patient gynecology treatment with manually loaded 137Cs sources is the most common method of manual brachytherapy (Eifel et al., 2004). The 137Cs source inventories established in the 1970s and 1980s have decayed significantly. There is sufficient interest in manual afterloading to warrant production of new 137Cs sources (Radiation Product Design 2004) to replace aging inventories. Interstitial prostate implants with 125I and 103Pd seeds are the most common operative procedure, mostly performed on outpatients. Full-service (calibrated, autoradiographed seeds in sterile packages) purchasing eliminates certain physical plant requirements (calibration equipment, etc.) normally required to support these services.
LDR and Remote PDR Afterloading Remote low dose rate (LDR) and PDR afterloading share a common feature; both require an in-patient hospital room especially designed to accommodate the RAU. LDR never achieved, in the United States, the prominence expected. Fewer than 100 LDR RAUs were sold in the United States and Canada. The
130
Glenn P. Glasgow Table 1. Dose Equivalents (DE) and Area Designations for Protection against Radiationa Member Designation
Commentb
Dose Equivalent
General Public
1 mSv Annually
5 mSv allowed via prior regulatory authorization with demonstrated need; visitors per authorized user approval
General Public
0.02 mSv in any one hour
In unrestricted area; not 0.02 mSv per hour
Embryo/Fetus
5 mSv over gestation period
Avoid substantial variations above uniform monthly exposure rate
Occupational Adults (External Whole Body)
50 mSv Annually
Occupational Adult (Eye Lens)
150 mSv Annually
Deep DE (DDE) at 1 cm
Occupational Adult (Skin & Extremities)
500 mSv Annually
Shallow DE at 7 µm over 1 cm2; knees & below; elbows & below
Occupational Minors
0.1 DE allowed adults
Deep DE (DDE) at 1 cm; excludes planned specials
////////////////////////////////////////////
////////////////////////////////////////////
///////////////////////////////////////////
Area Designation
Limit
Commentb
Unrestricted area
No restriction to access
Radiation Area
0.05 mSv in one hour at 30 cm from radiation source or surface radiation penetrates
Accessible to individuals; “Caution” signs may be used
High Radiation Area
1 mSv in one hour at 30 cm from radiation source or surface radiation penetrates
Controlled locked access; alarmed area; “Caution” or “Danger” signs may be used
Very High Radiation Area
5 Gy (5 Sv) in one hour at 100 cm from radiation source or surface radiation penetrates
Controlled locked access; alarmed area; deliberate entry; “Grave Danger” sign used
a
See 10 CFR 20 “Standards for Protection Against Radiation” for complete definitions. Limited comments only; for a full description see 10 CFR 20.
b
manufacturer, Nucletron Corporation (Columbia, MD), supports existing units, but will phase out LDR technology by 2010. After 14 years, and 188 treatments, we recently closed our LDR program at Loyola. Our radiation oncologist, however, notes that a few gynecology patients yearly, whose fragile anatomy cannot tolerate multiple fraction HDR applicator insertions, are candidates for single-fraction remote LDR treatments. We treat these patients with conventional single-fraction interstitial 192Ir template therapy rather than multiple-fraction HDR therapy. PDR brachytherapy had limited use (mostly in California and Arizona) in the United States before restrictive federal regulations hampered its growth. Unrealistic federal regulatory requirements for a radiation oncologist to be present every time the source came out of its safe have been removed in the recently revised federal code (USNRC 2002). New, less strident regulations on radiation oncologists’ presence may (or may not) allow PDR to be used in the United States (NB: this topic is discussed fully in chapter 9 on regulatory aspects).
10–Brachytherapy Facility Design
131
Remote HDR Afterloading Remote HDR RAUs are now common is the United States; there are over 500 HDR RAUs in use. Most are permanently installed; we will focus on programs and room designs for these devices. Increasingly, mobile HDR RAUs are used as hospitals share a single mobile HDR RAU moving among multiple locations. However, the increased popularly of daily B.I.D. treatments for certain anatomic sites poses scheduling challenges. We will review program features required for these devices.
Contained Liquid Radionuclide Therapies Some medical physicists are involved in cross-department (radiation oncology and nuclear medicine) programs in therapies involving contained liquid radionuclides designed for specific sites, e.g., 125I in intracranial balloons. Technically, these are brachytherapy (in close proximity to tissue) sources. The types and numbers of these therapies are increasing. However, as they are usually handled and prepared in nuclear medicine “hot” labs, we have omitted them from our discussion, as we do targeted radioisotope therapies (166Ho for liver tumors, 89Sr for bone pain, etc.).
Manual Procedure Source Room Facilities There is a major disconnect between the manual source rooms we actually have in our existing facilities and those we would design for a new facility. Generally, the radioactive materials source rooms in existing facilities are rooms of last resort, and are usually small, cramped, and often far removed from patient treatment rooms or areas. The radioactive materials source room at Loyola is a converted stairwell in the sub-basement without air ventilation or water. Current source room activities are little changed from those of years past as described in prior conference (Broadbent 1984) and summer school proceedings (Glasgow 1992). However, let us focus on the activities we perform in these rooms and how we would design the rooms to better perform these activities if we controlled the capital budget committee!
Receipt of Therapeutic Radioactive Materials It is generally recommended that facilities receive therapeutic radioactive materials (TRAM) at a common location, such as the Radiation Safety (Control) Office (RSO), so that, for all departments (radiotherapy, nuclear medicine, etc.), all records (package receipts, required wipe tests, etc.) are kept by RSO personnel at the central receipt location. Following receipt, RSO health physics personnel can distribute packages to departmental source rooms or hot labs. If a central receipt area does not exist, some receipt functions must occur in the source room. A well-designed room, Table 2, has many components. The room should be readily accessible but reasonably isolated away from public and employee work areas. A large room with solid concrete walls and a strong, locked, wide (to accommodate transportable HDR RAU) door with an automatic closer is ideal. Door postings (warning signs and emergency contact numbers) should be permanently mounted. The interior should be well lit. If applicators are cleaned, an appropriate sink with protected (mesh screening) drain and cleaning supply storage is needed. Rooms usually have a well-lit laboratory bench, appropriate electrical outlets, cabinets with drawers, wall pegboard for tool storage, conventional shop tools, a well-lit work area, storage or book shelves for the ever-present 3-ring binders for receipt/shipping papers, a computer terminal for inventories, a telephone for immediate contact with vendors, an area radiation monitor with alarm levels set to respond to low levels of radiation generated by a single misplaced, unshielded source, high-quality flashlights for locating seeds dropped on the floor, and a storage area for shipping cartons with defaced RAM (radioactive material) labels. Table 2 contains a more complete list of features.
132
Glenn P. Glasgow Table 2. Features of a Well-Designed Source Room
Physical Features
Accessible but reasonably isolated (away from public and employee work areas) large room with solid concrete walls; strong, locked, wide door (to accommodate transportable HDR RAU) (with warnings and emergency contact number postings) with automatic closer
Communications
Telephone (outside line for immediate vendor contact); computer (and printer); computer jacks; bulletin boards for regulatory postings
Interior Features
Well lit; water and cleaning supplies, sink with protected drain (mesh screening); workbenches with cabinet drawer storage; adequate electrical outlets above the workbenches; wall pegboards with hooks, bulletin boards; adjustable shelves; seamless tile floor; “Radioactive materials” waste can; film storage in lead bins
Tools
Conventional shop tools; flashlights, source handling tools, tweezers, forceps, etc.
Source Storage
Locked lead safe(s) or lead well; lead “L” blocks, lighted magnification lamp; temporary lead “pigs” for source sorting during preparation
QA Equipment
Area radiation monitor; radiation detection (NaI (Tl), GM) counters, survey meters; wipe test equipment (if needed); well-ionization chamber and electrometer; autoradiography equipment; barometer, thermometer
Transport Equipment
Shielded source transport containers; carts
Empty Container Storage
Cabinet, shelves to storage reusable or disposal containers, empty lead “pigs”
Work Area
Computer area; binders, blank forms, office equipment and supplies, calculators
Applicator Preparation Applicators and accessories (by type) stored in large labeled clear plastic boxes and Storage in cabinets or on shelves Everything Else
All those favorite “gadgets” that you use in your source room!
Source Calibration and Quality Assurance Procedures QA procedures, Table 2, such as package surveys, wipe tests, source calibrations, and source autoradiographs, each require their own instrumentation and radiation protection equipment (source storage safes, long-handled forceps, “L” blocks with viewing windows, well-calibration chambers, survey meters, etc.). Physicists are likely to receive most of their radiation exposures during QA procedures so attention to shields without voids or cracks and processes to reduce exposure are important. One design (Chavaudra 1997) features a mobile “L” shield that slides along the front of a laboratory bench to provide protection at each workstation on the bench. Each QA function requires its own equipment setup and local area shielding so that required tasks can be efficiently and quickly performed in compliance with the ALARA (as low as reasonably achievable) principle.
Security of TRAM Interim storage of TRAM is required. The outer door (with automatic door closer) will be secured (computer locked, dead-bolt lock, etc.); preferably the TRAM in the room should be locked in lead shields or safes. Devices (carriers, carts, etc.) for transport of TRAM and application tools to patient rooms or treatment areas are essential. Often applicators, accessories, and sterilized trays ready and prepared for operating room procedures can be stored in the source room.
10–Brachytherapy Facility Design
133
Holding TRAM for Decay Adequate space and shielding for holding TRAM for decay and subsequent waste disposal is required. In some cases the source room will suffice. At universities, large on-site waste disposal facilities are common.
Imaging (Source Localization) Facilities Source localization for isodose computations is common for interstitial operative procedures and manual and remote afterloading procedures. In some instances, if sources or applicators are inserted in an operating theater, source localization can occur there. However, radiographs with portable x-rays devices often exhibit poor image quality and may fail to achieve radiograph geometries (orthogonal pair, stereo shift, multiple radiographs for random seed entry, etc.) required for isodose computations. Generally, greater accuracy is achieved if source localization occurs in an imaging area designed for that purpose. Whatever imaging facility is used, it is imperative to obtain high-quality accurate images required for accurate isodose computations. We omit discussion of magnetic resonance imaging (MRI) and ultrasound imaging, evolving modalities not widely used for isodose planning.
Conventional Imaging Rooms With planning, conventional x-ray rooms in a diagnostic department can be used for controlled geometry (orthogonal, stereo, etc.) radiographs. Conventional x-ray simulators in radiation therapy facilities exhibit excellent isocentricity and, if available, generally provide excellent localization images. However, both modalities require patient transport and setup that potentially can cause undesired applicator movement between the time of imaging and time of treatment.
Portable Imaging Devices Portable x-ray “C-arms” can be used in the radiation oncology department inside a dedicated treatment room (e.g., an HDR vault). They offer the advantage of imaging the patient in the treatment position without moving the patient. However, without proper planning, they exhibit the same problems that occur in C-arms used in operating rooms. Geometrical reconstruction devices can alleviate some of these problems (Glasgow 1998). Liu et al. (2003) report an improved method of C-arm fluoroscopy for isodose computations.
Computed Tomography and CT/Simulators Computed tomography (CT) and CT/simulators in radiation oncology departments increasingly are used for source localization (if CT-compatible applicators are used) as they provide complete anatomic data relative to the source distribution. It is important to image the patient in the treatment position to keep applicators in their correct position relative to the anatomy. Again, use of a CT/simulator requires patient transport and setup that potentially can cause undesired applicator movement between the time of imaging and the time of treatment. Hence, in facility design, any CT/simulator should be physically close to any dedicated brachytherapy treatment room in the department. Indeed, integration of HDR and CT/simulation in one room offers several advantages but could present scheduling difficulties in a truly busy department.
134
Glenn P. Glasgow
Isodose/Time Computation Devices Software required for isodose and treatment duration calculations generally runs on a dedicated computer platform or runs as a special program on an isodose computation computer that includes software for external beam isodose computations. In either case, the hardware (computer, digitizer, light boxes, etc.) requires allocated space, electrical outlets, and a proper work environment for bioware (personnel; hopefully your dosimetrist!) to operate software on hardware. With the exception of interactive “real-time” prostate seed planning programs used in operating rooms during prostate implants, and as remote afterloaders can receive their treatment instructions via electronic data transfer, there is no physical requirement to locate the remote HDR adjacent to the HDR RAU. Brachytherapy isodose computation centers should conform to the generally good design principles of any well-designed computer workspace. Even high-end portable “laptops” now have the capacity for brachytherapy isodose planning.
Conventional Treatment Facilities The conventional treatment facility is a common hospital room without special shielding for brachytherapy. Such rooms normally house gynecology patients treated with manually afterloaded 137Cs sources, patients treated with 192Ir seeds in nylon ribbons, and permanent implant patients treated with 125I or 103Pd who require overnight observation post-implant. Larger rooms to accommodate afterloading carts, portable bedside shields, and positioning visitors’ chairs far from the patient are ideal. Usually, by inverse-square law considerations, adjacent hallway exposure rates around large rooms are less than around small rooms. A room adjacent to low occupancy areas (stairwells, elevator shafts, soiled linen storage, etc.) is ideal. Carefully consider the thickness of the floor and ceilings and the nature of the occupied areas directly below and above to determine if supplemental shielding is required. Supplemental overhead lighting may be required as some hospital room lights have inadequate local area illumination needed during manual afterloading procedures. Dedicated (partially or totally shielded) rooms afford the best control of radiation levels to adjacent areas. Older reports describing features of shielded rooms are still relevant and offer excellent discussions of the myriad issues of shielding brachytherapy rooms (Thomadsen et al. 1983; Gitterman and Webster 1984; McKenzie et al., 1986; Stedeford, Morgan, and Mayles 1997; McGinley 2002). Obviously, the thickness of wall shielding material needed is facility dependent, and depends on room size, adjacent area use, and number of patients treated. Lead thicknesses from 0.5 cm to 3.5 cm (to shield an adjacent bed in the opposing room) are typical. Shielded, direct entry doors are sufficiently heavy to require an automatic door closer. For those considering shielding a new room or retroshielding an existing room, older material shielding data, Table 3, for brachytherapy radionuclides are applicable. These data are reasonably consistent for all radioisotopes except 192Ir where some variation (1.2 to 2.0 cm) in tenth-value-layer (TVL) is noted. The newest shielding data (Rivard, Waid, and Wierzbicki 1999) are mass attenuation coefficients for Clear-Pb™ for several radionuclides. These data may be useful for those designing facilities with either direct (window) or indirect (window plus parabolic or convex mirrors) for patient observations. While it is common to neglect patient self-attenuation in shielding calculations the degree of safety gained, Table 4 has been described for gynecology patients treated with manually afterloaded 137Cs sources and 192Ir (Glasgow, Walker, and Williams 1985; Glasgow 2002b). For 137Cs a gynecology patient with a 40-cm lateral pelvic width provides about 50% self-attenuation. For either type of room, radiation levels inside the room can be reduced with localized portable shields. Institute of Physics and Engineering in Medicine (IPEM) Report 75 (Stedeford, Morgan, and Mayles 1997) provides a useful discussion of bedside shields. Figure 2(a) shows a homemade (used with some success), fixed-height, slant shield with a step to accommodate both short and tall personnel; Figure 2(b) shows a conventional adjustable-height radiation shield.
10–Brachytherapy Facility Design
135
Table 3. Comparative Selected Broad-Beam TVL (cm) Shielding Data Radioisotope
Concrete
Steel
Lead
Cobalt-60 NCRP 40 & 49a IPEM 75b IPSM 46c Boutroux-Jaffré
20.6 20.6 No data given 22
6.9 No data given No data given 6.7
4.0 4.0 4.6 4.2
Cesium-137 NCRP 40 & 49a IPEM 75b IPSM 46c Boutroux-Jaffré
15.7 15.7 No data given 17.5
5.3 No data given No data given 5
2.1 2.1 2.2 2.2
Ir-192 NCRP 40 & 49a IPEM 75b IPSM 46c Boutroux-Jaffré
14.7 11.3 No data given 14.7
4.3 No data given No data given 4.3
2.0 1.5 1.2 1.6
Au-198 NCRP 40 & 49a IPEM 75b IPSM 46c Boutroux-Jaffré
13.5 13.5 No data given No data given
No data given No data given No data given No data given
1.1 1.1 1.0 No data given
a
Approximate values obtained with large attenuation. No explicit statement that data is broad beam data. c McKenzie et al. (1986). b
Table 4. Effective Transmission Factors for192Ir and 137Cs versus Patient Lateral Half-Pelvic Widths Patient lateral half-pelvic (r) (cm)
Effective transmission factor K(r) for 192Ir (Glasgow 2002b)
Effective transmission factor K(r) for 137Cs (Glasgow 1985)
10
0.72
0.71
12
0.66
0.66
14
0.64
0.62
16
0.55
0.59
18
0.50
0.55
20
0.44
0.50
22
0.40
0.45
24
0.36
0.41
25
0.34
0.38
Generally, properly designed shields reduce bedside radiation exposure levels by more than one-tenth, and proportionally increase bedside work times, Table 5. However, gaining personnel acceptance for use is difficult and requires training; there often are issues associated with shield storage.
136
Glenn P. Glasgow Table 5. Radiation Survey and Working Times Per Unit Dose Equivalent for 192Ir
Location Description
•
X = exposure rate in (mR/h); t = time (minutes) •
Behind 1 in. bedside shield, 6 in. from bed
X (mR/h) t mins
6 in. from bedside with shield removed
X (mR/h) t mins
•
Working Time Per Unit Dose Equivalent Iridium Activity in mgRaeg 20
40
1.4 43
2.9 21
29 2
60 1
The work times per unit dose equivalent are the times in which personnel would receive 1 mrem whole body exposure from the exposure rates at each location. The work times are guides to aid in planning your duties so you may minimize your radiation exposure during your shift. For example, for 20 mg Ra, working 6 in. from bedside with the shield removed for 2 minutes and working behind the shield for 43 minutes would result in 2 mrem exposure.
(a)
(b)
Figure 2. (a) A homemade portable bedside shield featuring a slant design and an adjustable step. (b) A commercial radiation shield with adjustable height.
Pulsed Remote Afterloading Fixed Facilities PDR Room Shielding As PDR is currently little used in the United States, there are few facility descriptions (Speiser and Hicks 1994). A remote PDR RAU room, Figure 3, is similar to a LDR RAU room, except types and quantities of radionuclides are different. Generally, a PDR source would be ~1 Ci of 192Ir while an LDR RAU would be about 0.35 Ci of 137Cs (Glasgow 1995). As the lead TVL for 137Cs (2.1 cm) and 192Ir (2.0 cm) are nearly identical, shielding comparisons are easily made (NCRP 1972). St. Germain notes that, in principle, that PDR brachytherapy requires no additional shielding above that required for corresponding continuous
10–Brachytherapy Facility Design
137
Figure 3. Floor plan of a pulsed low dose rate (P-LDR) microSelectron-PDR remote afterloading facility [Reprinted from “Safety Programmes for Remote Afterloading Brachytherapy: High Dose Rate and Pulsed Low Dose Rate” by B. L. Speiser and J. A. Hicks in Brachytherapy from Radium to Optimization, R. F. Mould, J. J. Batterman, A. A. Martinez, and B. L. Speiser (eds.). Veenendaal, The Netherlands: Nucletron B. V. © 1994, with permission from Drukkers. Courtesy B. L. Speiser.]
LDR implant if the PDR average hourly exposure and total treatment dose are equivalent to those the continuous LDR treatment (NCRP 2004). Williamson offers a similar discussion, but concluded that, as a practical application, additional wall shielding (1.9 cm lead) was required to satisfy regulatory requirements on allowed exposures in adjacent areas (Williamson, Grigsby, and Meigooni 1995). Figure 4 shows one design of an LDR room (Wilson et al. 1986; Glasgow et al., 1993) and Figures 5(a) and 5(b) show our LDR room at Loyola (Glasgow and Corrigan 1995). Could our shielded LDR room, now closed, be
138
Glenn P. Glasgow
Figure 4. A dedicated LDR remote afterloading room with a small maze and viewing window. (Modified from B.M. Wilson et al., Med Phys 13: 608. © 1986, with permission from AAPM. Courtesy J. D. Bourland.]
used as a PDR room? No, the 1 Ci of 192Ir (about 558 mgRaeq) is about four times the activity of 0.35 Ci of 137Cs (about 140 mgRaeq). For gynecology treatments, 70 mgRaeq is a typical loading. For 192Ir, for an equivalent absorbed dose prescription, treatments times would presumably be one-eighth those of 137Cs, e.g.. 1/8 hour. However, during this interval, the instantaneous hallway exposure rate would approach the 0.02 mSv in one-hour limit. Considering a 2.5 m distance from patient to hallway (0.16 inverse square reduction) and 0.5 in. lead (about 0.65 TVL; 0.22 reduction) in the wall, the total reduction is about 0.035. For 558 mgRaeq the exposure rate at 1 m would be 460 mR/h. Applying the reduction factor gives an instantaneous hallway exposure rate of about 16 mR/h. For 1/8 hour, this yields the limit of 2 mR or 0.02 mSv, with no margin of error. While some factors, such as patient attenuation in gynecology patients could reduce the exposure rate, other anatomic sites offer no attenuation and could receive higher absorbed dose prescriptions. Hence, this superficial analysis indicates that additional retroshielding would likely be needed to convert this LDR room to a PDR room. Careful analysis of proposed PDR treatment regimes, duty factors, workloads, and average hourly and weekly exposures must be done in any PDR facility shielding analysis.
10–Brachytherapy Facility Design
139
Figure 5a. A small second floor hospital room renovated to house an LDR remote afterloading device for gynecologic treatments. This room features an internal storage closet in which the LDR unit is stored when not in use. A 1.3 cm (1/2 in.) thick lead projection shield beneath the bed shields the area below; a 0.6 cm (1/4 in.) thick lead shield (not shown) is suspended from the ceiling above. Note the compressed air supply, dedicated electric outlet, radiation monitor, remote control and telephone, power-assisted door opener, 1.3 cm (1/2 in.) lead walls shields, and supplemental lights overhead.
PDR Facility Issues: Sources, Security, Calibration, and Quality Control Procedures Per licensure, PDR RAU source security, calibration, and QA procedures are essentially those required for HDR RAUs. The major issue is that the HDR facility likely is in the radiation oncology department, while the PDR room is in the hospital. This presents some issues with respect to required quality control (QC) procedures. If the PDR hospital room is in new construction, then a small adjacent locked room or vestibule could be designed to accommodate new and spent source storage, calibration equipment, and the other QC equipment. However, it is far more likely that a conventional hospital room may be modified to accommodate a PDR RAU and that adjacent support facilities are not available. Hence, sources and calibration and QC equipment must be transported to the PDR room from some remote location. The best one can do is to recognize the procedural issues when the room is being modified, and try to provide
140
Glenn P. Glasgow
Figure 5b. Adjacent closet houses the Selectron-LDR, a low dose rate remote afterloading device for gynecologic treatments. Note the emergency source retrieval container.
as much space in the room as possible during modification. Obviously, some plan must be made to store and secure the PDR sources as well as the necessary calibration and QC equipment.
Features of a PDR Treatment Room Features of a PDR treatment room (electrical source-retraction door interlocks, radiation area monitor, brachytherapy bed, emergency equipment, nurse communication system, visual/audible status indicators at nurse station, PDR locked storage, supplemental lighting) are similar to those reported for a remote LDR room (Glasgow and Corrigan 1995). Note that 10 CFR 35.615 (USNRC 2002) requires “…viewing and intercom systems to permit continuous observation…” for a remote PDR room, but does not require the same for remote LDR rooms.
10–Brachytherapy Facility Design
141
Permanently Installed HDR RAU Facilities The major topics associated with permanently installed HDR RAUs are: (a) the interim storage (and security) of the HDR source prior/post installation; (b) source calibration and QA procedures (discussed in the context of facility design and space for the equipment); and (c) whether the HDR facility is in a room shared with other equipment (simulator, linac), is in a dedicated HDR treatment room, or is in an operating room for intraoperative procedures. Intrafacility mobility is discussed in the last section on mobile HDR RAUs. Many factors determine which program or programs one develops. Table 6 lists features of a dedicated HDR minor procedures suite. Table 7 describes estimates of facility cost ranges for different programs. Table 8 describes some HDR program options and estimates of their costs.
Table 6. Equipment List: Dedicated HDR Remote Afterloading/Minor Procedures Suite Remote afterloader storage and treatment location; source guide tubes; in-room radiation detector; check source
Anesthesia area: Medical gases and vacuum; designated location and electrical power for patient monitoring equipment; remote displays
Remote afterloader control console
Operating room/procedures light
Visual communications: 2 systems
Applicator storage and medical and nursing supplies
Audio communications
Sink/scrub/gown/glove area
Secure locked door; door interlock
Treatment applicators
Multiple position patient procedure table; x-ray compatible
Applicator positioning clamp: Integral with procedures table
Overhead track-mounted radiography with X-ray generator and X-ray control console; or,
Emergency off buttons at console, in maze, and in room
Mobile C-arm x-ray unit with fluoroscopy; fiducial marker device
Emergency lighting; wall or ceiling mounted
Emergency recovery equipment and container; “Open source” sign, etc.
Emergency power for selected equipment: Audio, video, anesthesia patient monitoring, lighting, radiation detectors and indicators; remote afterloader
Survey meters
Patient monitors
Optional ultrasound imaging system
Optional treatment planning workstation
Table 7. Facility Construction Costsa Program
a
Square Foot Costs
New construction costs for medical building without shielded areas, i.e., offices, etc.
$100
For an HDR suite for minor procedures, i.e., Class B operating room, add $400 per sq ft
$500
$300,000
For an HDR suite for major procedures, i.e., Class C operating room, add $100 per sq. ft.
$600
$360,000
Calendar Year 2004. Assume 400 sq ft for treatment area and 200 sq ft console and support areas.
b
Facility Costs for 600 sq ftb
142
Glenn P. Glasgow Table 8. Program Cost Estimatesa Parameter
Mobile
Low End
High End
HDR Device
$150K–$300K
$250K–$350K
$250K–$350K
Applicators
$100K–Dupl
$50K–Basic
$100K–Many
Table
$35K
$35K
$75K
Van
$25K
Physics QA
$25K
$25K
$50K
Image System
Use Existing
$120K C-Arm
$600K IBU
Vault
Use Existing
$175K–Sm
$235K–Lg
Totals
$310K–$460K
$655K–$815K
$1325K–$1525K
HDR Facility Issues: Sources, Security, Calibration, and Quality Control Procedures HDR interim source storage and security parallels the PDR discussion; in a dedicated facility it may be reasonable to have a secure HDR source storage area adjacent or inside the dedicated room, as this is where the source transfers occur. Prior comments about source room receipt records and inventory records apply. Similarly, as HDR source calibration and QA procedures must occur inside the room where the HDR RAU is located, it is reasonable to provide either space inside the room for calibration and QA equipment, or, a special mobile cart holding all calibration and QA equipment that can be moved to the room.
Features of HDR Treatment Rooms Certain safety features, Table 6, of HDR treatment room (area monitor, door interlocks, emergency equipment, etc.) are likely to be required for licensure in all types of facilities (Glasgow 1996). However, as noted later, for HDR rooms accommodating mobile HDR RAUs it may be desirable to implement these features in a manner different than those with a permanently installed HDR RAU. Note that 10 CFR 35.615 (USNRC 2002) requires “…viewing and intercom systems to permit continuous observation…” for remote HDR rooms.
Features of HDR RAUs in Operating Rooms In addition to the safety features required for permanent installations, operating rooms (ORs) present two unique challenges: shielding staff and maintaining sterile conditions. Two options exist: (1) Locate the OR in the radiation therapy facility, Figures 6 and 7; or (2) locate the HDR RAU in an existing OR. If the former option is used, it may be possible to combine into the design, if desired, both linac intraoperative procedures and HDR intraoperative procedures, as shown in Figure 6. Intraoperative radiation therapy (IORT), which describes external beam IORT, does not describe brachytherapy IORT; however, Dobelbower and Abe (1989) address the general issues of the use of complex equipment in the OR. Local shielding for HDR in ORs can be used effectively (Sephton et al., 1999). St. Germain has addressed the issue of maintaining operating room staff sterile conditions while staff step outside the inner OR while the source is out treating the patient (St. Germain 2002). She suggests an inner/outer room design. During the portion of the procedure when the source is out, the attending staff in sterile gear steps out of the inner operating room into the outer room when sterile conditions are still maintained. She also has designed an adjacent “bail-out” room for storage of the HDR RAU in the event the source sticks in the device and it must be removed from the OR so other procedures can be performed in the OR.
10–Brachytherapy Facility Design
143
Figure 6. Dedicated intraoperative radiotherapy facility at the Regional Cancer Center at Baptist Hospital of Miami. The facility is also used as an HDR suite. [Reprinted from “Design and Implementation of a Program for High Dose Rate Brachytherapy” by P. V. Houdek, G. P. Glasgow, J. G. Schwade, and A. A. Abitbol in High Dose Rate Brachytherapy: A Textbook. S. Nag (ed.). Armonk, NY: Futura Publishing Co., © 1994, with permission from Blackwell Publishing.]
Figure 7. Dedicated HDR brachytherapy suite and OR at the U. of Miami’s Sylvester Comprehensive Cancer Center. The suite can be used for intraoperative procedures. [Reprinted from “Design and Implementation of a Program for High Dose Rate Brachytherapy” by P. V. Houdek, G. P. Glasgow, J. G. Schwade, and A. A. Abitbol in High Dose Rate Brachytherapy: A Textbook. S. Nag (ed.). Armonk, NY: Futura Publishing Co., © 1994, with permission from Blackwell Publishing.]
144
Glenn P. Glasgow
HDR Room Designs HDR room designs obviously take many forms depending on the program designs and the facility sites. The major difference is between those facilities renovating rooms for HDR RAUs and those building new facilities. We offer a few examples from published literature. Often existing unused or underused vaults (orthovoltage, 60Co, etc.) can be modified, at substantial cost savings relative to new construction costs, as an HDR vault. Usually, existing shielding is adequate and the only costs are those of modifying the vault to accept the HDR RAU and its associated electrical interlocks and communication systems. Figure 8 shows an HDR vault in a modified orthovoltage room (Stedeford, Morgan, and Mayles 1997). In this figure note the additional of a partition in the room to create a short-length entry maze to better shield the door. In contrast, Figure 9 shows a dedicated HDR RAU vault with a long maze and minimum entry door shielding (Stedeford, Morgan, and Mayles 1997). Note in this figure a small storage area for the microSelectron™ HDR and the shielded area inside the vault, which is described as a shielded operating theater. McGinley discusses HDR maze exposure distributions, noting that, in the maze described, exposure rates in the direction from the inside of the maze to the entry door “…fall off slower than indicated by inverse square of exponential behavior” (McGinley 2002).
Figure 8. HDR treatment room in a modified orthovoltage room. [Reprinted from Stedeford, B., H. M. Morgan, and W. P. Mayles, “Brachytherapy Room Design” in The Design of Radiotherapy Treatment Room Facilities, Institute of Physics and Engineering in Medicine. © 1997. Reproduced with permission.]
10–Brachytherapy Facility Design
145
Figure 9. HDR treatment room (shielded operating theatre). [Reprinted from Stedeford, B., H. M. Morgan, and W. P. Mayles, “Brachytherapy Room Design” in The Design of Radiotherapy Treatment Room Facilities, Institute of Physics and Engineering in Medicine. © 1997. Reproduced with permission.]
At Loyola, our dedicated HDR facility (Figure 10) was planned as a Class C (full feature OR), but became a Class B limited procedure room. The room accommodates a dedicated C-arm (General Electric Model 9800 Digital Mobile Imaging System) and an ultrasound unit (B & K) for prostate volume studies. The room lacks anesthesia equipment. Patients requiring anesthesia have applicators placed in second floor ORs in the facility; they are then transported to the HDR facility in the basement. Biodex Medical Brachytherapy Table makes an excellent treatment couch that, thus far, has accommodated gynecologic, breast, and head and neck procedures. The room has outer dimensions of 22 ft width by 17 ft depth, inner dimensions of 17.5 ft width by 13 ft depth, with 15 in. thick ordinary density concrete wall and 1.75 in. lead in the power-assisted door. As we are in the basement, the distance from couch to the floor above was about 12 ft with slightly more than 1 ft concrete. The workload was estimated as 50 patients yearly, each an average of 4 fractions for 200 procedures per year. Assuming a constant 10 Ci source (Γ=0.46 R/h per Ci at 1 m) used 2 hours weekly, the exposure rate at 1 m (neglecting patient attenuation) is about 9.2 R/week or 460 R/year. The distance from patient to the door and console area is about 3.75 m, for an inverse square reduction of about 0.07, yielding about 0.64 R/week or about 32.2 R/year. Limiting occupational DE to 5 mSv (0.5 rem) yearly, ALARA, 0.1 mSv/week, requires a wall reduction of 0.1 mSv/week/0.64 R/week, or 0.15, somewhat more than 1 TVL. As the concrete TVL is about 15 cm (about 6 in.), our 15 in. walls (about 2.5 TVLs) provide an adequate margin of safety. The 1.75 in. lead door (about 4.5 cm or 2.25 TVL) provides the same margin of safety. These data are consistent with McGinley’s comment that “Typical room shielding consists of 0.35 m (about 14″) to 0.61 m (about 24″) concrete or 4 to 5 cm lead.” (McGinley 2002). Because adjacent areas were all occupational, the 0.02 mSv in one-hour limit was not a concern around our facility. However, we note, from experience, only one patient can be treated in any one hour.
146
Glenn P. Glasgow
Figure 10. A dedicated “Class B” (minor procedures) HDR suite. (Loyola University Medical Center Outpatient Clinic, 2003)
Overhead, the reduction for 3.65 m (about 0.075) and 30.5 cm (about 2.1 TVL; 0.008) was 0.00067. For four procedures weekly, conservatively assuming one full-hour treatment of 2.3 R at 1 m, in one hour the exposure above is 2.3 R (0.0006) or 0.00014 R in one hour, which is 0.14 mR or 0.014 mSv in one hour, which satisfies the 0.02 mSv in one-hour requirement. This yields a safety margin of about 42%. Preferably, one would like a 100% safety margin. Practically, our 42% is acceptable. Due to many assumptions, the actual radiation survey is a more reliable measure than calculation results. If your facility has space for HDR, but no vault, NELCO offers a pre-fabricated vault, Figure 11, that can be installed in a space about 100 sq. ft. (7 ft deep by 10.5 ft wide by 8.5 ft high) (NELCO 2004). The 50,000-pound unit, designed for 2-hour weekly source exposure time, has all necessary features of an HDR vault and can be installed in 3 weeks. Nucletron features an integrated brachytherapy unit (IBU), Figure 12, designed to perform patient preparation, applicator insertion, imaging, isodose computation, delivery, and verification in the unit (Nucletron 2004). The imaging equipment allows fluoroscopy from all directions. The IBU is designed for those centers with large brachytherapy services and is relatively expensive.
10–Brachytherapy Facility Design
Figure 11. A pre-fab HDR suite. [Courtesy Rick LeBlanc, NELCO, Woburn, MA.]
Figure 12. An integrated brachytherapy unit. [Courtesy E. van ‘t Hooft, Nucletron International B.V.]
147
148
Glenn P. Glasgow
Mobile HDR Facilities A mobile HRD RAU could be an intrafacility device (used at multiple locations within the same facility) or interfacility device [used at multiple (client) locations outside a main (central) facility], Figure 13.
Intrafacility Programs Intrafacility use under a single license is reasonably straightforward. Protocols for the same transport of the HDR RAU through the hospital are required. Each location in the facility where the HDR unit is used must satisfy the safety features of permanent locations previously discussed.
Interfacility Programs For interfacility use, two types of vans are available: The “mobile vault” full-service van that contains not only the HDR RAU and ancillary equipment but also the fully shielded treatment vault, and the van that transports only the HDR RAU and limited ancillary equipment. We focus on the latter as it is more commonly used. For interfacility use both the main user and the client facilities each must hold a license. Those licenses may differ in content but together they cover all aspects of use of the mobile HDR RAU. Federal technical requirements for a mobile service (USNRC 2002) are described in 10 CFR 35.647. Applicable transport regulations must be rigorously adhered to and those transporting must have received hazard materials (HAZMAT) training. Docking (unloading) facilities (tailgate lifts, dock lifts, etc.) may be necessary for the safe transfer of the HDR RAU to and from the van at each treatment facility. The features of the treatment sites may vary depending on the program. Some mobile HDR services transport survey meters, radiation area monitors, and door interlock switches that can be quickly installed and used at each site. They also transport laptop planning computers, small film digitizers, and other necessary QA equipment that can be used at each site. Others design treatment site facilities with all necessary physical plant accommodations that we previously described at permanent facilities. Security of the mobile
Figure 13. “On the road, again!” A mobile HDR van. [Courtesy Plato Lee.]
10–Brachytherapy Facility Design
149
HDR RAU at each site and during transport is of great concern. If stored at a client facility for short times, it must be kept in a secure, locked area, such as a source room. The van must have antitheft alarms; the HDR RAU cannot be left in the van overnight. The van must satisfy either applicable state or federal Department of Transportation requirements for the transport of radioactive materials.
Conclusions All brachytherapy programs, manually afterloaded sources, interstitial permanent implants, remote LDR, MDR, HDR, and PDR RAUs have many similar features and the design of the facilities for each program are reasonably straightforward. Key in facility design is a full understanding of how the brachytherapy sources and services are to be used. I trust that this chapter offers those unfamiliar with facility design a reasonable overview of the topic. May you never lose a source!
Acknowledgments I dedicate this chapter to two prominent medical physicists, deceased, with whom I have co-authored articles. Pavel Houdek, Ph.D., exhibited excellent design skills as exhibited in the room layouts I have included in this presentation. Dale Kubo, Ph.D., had many interests, one of which was ensuring well-designed procedures and processes were used in remote afterloading programs. I regret they are not with us at this 2005 AAPM Summer School; each would have contributed excellently to our facility and design discussions. My thanks to Jean St. Germain for sharing draft NCRP materials, Patton McGinley for his material on HDR mazes, and David Keys and Plato Lee for discussions about their mobile HDR programs.
References Boutroux-Jaffré, F. “Photon Emitting Sources,” Chapter 1 in A Practical Manual of Brachytherapy. B. Pierquin and G. Marinello (eds.). Madison, WI: Medical Physics Publishing, pp. 3–21, 1997. Broadbent, M. V. “Brachytherapy Source Storage, Room Design, and Shielding” in Radiotherapy Safety. B. Thomadsen (ed.). New York: American Institute of Physics, pp. 99–116, 1984. Chavaudra, J. (1997). Chapter 6: “Radiation Protection,” Chapter 6 in A Practical Manual of Brachytherapy. B. Pierquin and G. Marinello (eds.). Madison, WI: Medical Physics Publishing, pp. 83–106, 1997. Dobelbower, R. R., and M. Abe (eds.). Intraoperative Radiation Therapy. Boca Raton, FL: CRC Press, Inc., 1989. Eifel, P. J., J. Moughan, B. Erickson, T. Iarocci, D. Grant, and J. Owen. (2004). “Patterns of radiotherapy practice for patients with carcinoma of the uterine cervix: A pattern of care study”. Int J Radiat Oncol Biol Phys 60:1144–1153. Gitterman, M., and E. W. Webster. (1984). “Shielding hospital rooms for brachytherapy patients: Design, regulatory, and cost/benefit factors.” Health Physics 46:617–625. Glasgow, G. P. “Radiation Safety Program for Radiation Oncology” in Advances in Radiation Oncology Physics. J.A. Purdy (Ed.). 1990 AAPM Summer School. AAPM Monograph No. 19. New York: American Institute of Physics, pp. 596–622, 1992. Glasgow, G. P. “Principles of Remote Afterloading Devices” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, R. Nath (eds.). Proceedings of the AAPM 1994 Summer School. Madison, WI: Medical Physics Publishing, pp. 149–161, 1995. Glasgow, G. P. (1996). “Radiation control, personnel training, and emergency procedures for remote afterloading unit.” Endocuriether/Hypertherm Oncol 12:67–79. Glasgow, G. P. “Isodose Planning: Brachytherapy” in Treatment Planning in Radiation Oncology. F. M. Khan, and R. A. Potish (eds.). Philadelphia, PA: Williams & Wilkins, 1998. Glasgow, G. P. “Brachytherapy” in The Modern Technology of Radiation Oncology. J. Van Dyk (ed.). Madison, WI: Medical Physics Publishing, pp. 695–752, 1999.
150
Glenn P. Glasgow
Glasgow, G. P. “Equipment Selection and Facility Design” in Proceedings of the American College of Medical Physics 19th Annual Meeting and Workshops. American College of Medical Physics, June 3–4, 2002, Jackson Hole, WY, pp. 263–279, 2002a. Glasgow, G. P. (2002b). “Is an Ir-192 permanent seed implant feasible for prostate brachytherapy?” Brachytherapy 1:195–203. Glasgow, G. P., and K. W. Corrigan. (1995). “Radiation design and control features of a hospital room for a low dose rate remote afterloading device.” Health Physics 69:415–419. Glasgow, G. P., S. Walker, and J. D. Williams. (1985). “Radiation exposure rates near brachytherapy patients containing 137Cs sources.” Health Physics 48:97–104. Glasgow, G. P., J. Daniel Bourland, P. W. Grigsby, J. A. Meli, and K. A. Weaver. “Remote Afterloading Technology.” AAPM Report No. 41. New York: American Institute of Physics, 1993. Houdek, P. V., G. P. Glasgow, J. G. Schwade, and A. A. Abitbol. “Design and Implementation of a Program for High Dose Rate Brachytherapy” in High Dose Rate Brachytherapy: A Textbook. S. Nag (ed.). Armonk, NY: Futura Publishing Co., 1994. IsoRay (2004). Cs-131 Seed. [Online] IsoRay. http://www.isoray.com. [2005, March 30]. Liu, L., D. A. Bassano, S. C. Prasad, B. L. Keshler, and S. S. Hahn. (2003). “On the use of C-arm fluoroscopy for treatment planning in high-dose-rate brachytherapy.” Med Phys 30:2297–2302. McGinley, P. H. Shielding Techniques for Radiation Oncology Facilities, 2nd ed. Madison, WI: Medical Physics Publishing, 2002. McKenzie, A. L., J. E. Shaw, S. K. Stephenson, and P. C. R. Turner (eds.). “IPSM Report 46: Radiation Protection in Radiotherapy.” London, Institute of Physical Sciences in Medicine, 1986. National Council on Radiation Protection and Measurements (NCRP). NCRP Report No. 40. Protection Against Radiation from Brachytherapy Sources. Bethesda, MD: NCRP, 1972. National Council on Radiation Protection and Measurements (NCRP). NCRP Report No. 49. Structural Shielding Design and Evaluation for Medical Use of X-rays and Gamma Rays of Energies up to 10 MeV. Bethesda, MD: NCRP, 1976. National Council on Radiation Protection and Measurements (NCRP). NCRP Report SC 91-1. Draft Report; Untitled. Bethesda, MD: NCRP, 2004. (Courtesy J. St. Germain; Private Communication, Dec. 10, 2004). NELCO (2004). (No Date). HDR Shielded Booth. [Online] NELCO. http://www.nelco-usa.com. [2005, March 30]. Nucletron B. V. (2004). (July 7, 2004). Remote Afterloaders. [Online]. Nucletron B.V. http://www.nucletron.com/ content. [2005, March 30]. Radiation Product Design (2004). 137Cs Sources. [Online]. Radiation Product Design.http://www.rpdinc.com. [2005, March 30]. Rivard, M. J., D. S. Waid, and J. G. Wierzbicki. (1999). “ Mass attenuation coefficients of Clear-Pb for photons from I-125, Pd-103, Tc-99m, Ir-192, Cs-137, and Co-60.” Health Physics 77:571–578. Sephton, R., K. R. Das, J. Coles, W. Toye, and P. Pinder. (1999). “Local shielding of high dose rate brachytherapy in an operating theatre.” Australas Phys Eng Sci Med 22:113–117. Speiser, B. L., and J. A. Hicks. “Safety Programmes for Remote Afterloading Brachytherapy: High Dose Rate and Pulsed Low Dose Rate” in Brachytherapy from Radium to Optimization. R. F. Mould, J. J. Batterman, A. A. Martinez, and B. L. Speiser (eds.). Veenendaal, The Netherlands: Drukkers, 1994. Stedeford, B., H. M. Morgan, and W. P. Mayles. “Brachytherapy Room Design” in The Design of Radiotherapy Treatment Room Facilities. York, England: Institute of Physics and Engineering in Medicine (IPEM), pp. 77–89, 1997. St. Germain, J. (2002). “Shielding of HDR, IVB, and PET/CT facilities.” Health Physics 82:S183 (Abstract). Thomadsen, B., J. van de Geijn, D. Buchler, and B. Paliwal. (1983). “ Fortification of existing rooms used for brachytherapy patients.” Health Physics 45:607–615. U.S. Nuclear Regulatory Commission (2002a). Code of Federal Regulations, title 10, parts 20, 32, and 35 (Medical Use of Byproduct Material: Final Rule). Federal Register, vol. 67, no. 79 (April 24):20250–20397. [Online]. Washington, DC: Nuclear Regulatory Commission. http://www.nrc.gov/reading-rm/doc-collections/ cfr/part020/part032/035. [December 4, 2004]. Varian Medical Systems (2004). (July 7, 2004). News and Events. [Online]. Varian Medical Systems. http://www. varian.com/comp/040707.html. [2005, March 30].
10–Brachytherapy Facility Design
151
Williamson, J. F., P.W. Grigsby, and A.S. Meigooni. “Clinical Physics of Pulsed-Dose-Rate (PDR) Remotely Afterloaded Brachytherapy” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.). Madison, WI: Medical Physics Publishing, pp. 577–616, 1995. Wilson, B. M., et al. (1986). Presentation at the AAPM 28th Annual Meeting. Abstract. Med Phys 13:608.
Further Reading Gifford, D., and T. J. Gooden. (1990). “ An analysis of personnel dose records which justifies the application of costbenefit analysis techniques in the design of an afterloading facility and the use of controlled areas and systems of work within suite to control occupational exposure.” Br J Radiol 63:214–218. Klein, E. E., P. W. Grigsby, J. F. Williamson, A. S. Meigooni. (1993). “Pre-installation empirical testing of room shielding for high does rate remote afterloaders.” Int J Radiat Oncol Biol Phys 27:927–931. Kubo, H. D., G. P. Glasgow, T. D. Pethel, B. R. Thomadsen, and J. F. Williamson. (1998). “High dose-rate brachytherapy treatment delivery: Report of the AAPM Radiation Therapy Committee Task Group No. 59.” Med Phys 25(4):375–403. Also available as AAPM Report No. 61. Nag, S., R. Dobelbower, G. P. Glasgow, G. Gustafson, N. Syed, B. Thomadsen, J. F. Williamson. (2003). “Intersociety standards for the performance of brachytherapy: A joint report from ABS, ACMP, and ACRO.” Crit Rev Oncol/Hematol 48:1–17.
Chapter 11
Calibration for Brachytherapy Sources Larry A. DeWerd, Ph.D., F.A.A.P.M. University of Wisconsin Madison, Wisconsin Introduction to the Calibration of Sources and Measurement Devices . . . . . . . . . . . . . . . . . . . 153 Brachytherapy Source Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Importance of Calibration and Standards in the Clinical Situation . . . . . . . . . . . . . . . . . . . . . . . . . . 154 Calibration Concepts and Quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Calibration Quantities and Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Photon Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Beta Particle Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Primary Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Photon Sources for Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Beta Sources for Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 ADCL Standards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 ADCL Source Calibrations and Ophthalmic Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162 Characteristics of Well Chambers for Brachytherapy Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Well-Chamber Characteristics for Photon Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Well-Chamber Characteristics for Beta Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 Clinical Calibration Using Well Ionization Chambers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Clinical Calibration of Photon Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168 Clinical Calibration of Beta Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
Introduction to the Calibration of Sources and Measurement Devices Brachytherapy uses encapsulated radioactive sources to deliver radiation dose to tissues near the source. These sources need calibration before clinical use; this activity is usually accomplished with a well ionization chamber. This chapter discusses source calibration as determined by primary standards in use, the procedures for measuring air kerma strength of brachytherapy sources for both low dose rate (LDR) and high dose rate (HDR) sources, and the absorbed-dose-to-water at 2 mm for beta sources. The use of well chambers for routine clinical source calibration is also discussed. Characterization of brachytherapy sources requires good initial measurements so that the patient dose is known with confidence.
Brachytherapy Source Standards Radiation emissions from brachytherapy sources involve either γ, β, or x-rays, or a combination of these radiations. The majority of sources in use emit γ rays. The type of radiation and energy has an effect on the calibration methodology. The calibration of the source can greatly enhance the precision of the treatment. This is a rapidly changing field, especially concerning intravascular brachytherapy where the use of gamma-emitting 192Ir sources and the 32P beta-emitting source from Guidant used for intravascular brachytherapy applications have come and gone in the past few years. Most of the instrumentation standards for brachytherapy sources are at the U.S. National Institute for Standards and Technology (NIST) or are directly traceable to NIST standards. Brachytherapy sources can be divided into three classes. The first is low energy—LDR sources generally used for prostate implants. The second is high energy—both LDR and HDR sources used for a number
153
154
Larry A. DeWerd
of treatment modalities. The third is beta sources—used for ophthalmic applicators and for intravascular brachytherapy. Depending on the energy of the source, different reference points are used. For example, for low-energy LDR sources, the reference distance is 1 cm; for beta sources, it is generally 2 mm or 5 mm. A short distance from the source may lead to greater uncertainties in the calibration because of the high-dose gradients of these sources.
Importance of Calibration and Standards in the Clinical Situation A common clinical brachytherapy dosimetry practice of the past was to accept the manufacturer-specified source strength without verification as a basis of treatment planning. Acceptance of the manufacturer’s stated source strength without measurement by the clinical physicist is an unwise approach as discussed by AAPM Task Group 40 (TG-40) (Kutcher et al., 1994). To quote from this report, “Each institution planning to provide brachytherapy should have the ability to independently verify the source strength provided by the manufacturer.” It is also recommended by TG-40 that a minimum of 10% of the LDR seeds be calibrated. As has become evident, it is essential that each new HDR 192Ir source be calibrated and that it have a periodic quality check at least once in its clinical lifetime (Glasgow et al., 1993). The need for individual source calibrations is further underscored when the differences in source construction for the same radionuclide, e.g., 125I, are considered. These differences greatly affect the response per unit source strength of calibration instruments. The radiological characteristics of a brachytherapy source are strongly dependent on the source encapsulation and the radioactivity distribution within the source, as discussed in Rivard, Kirk, and Leal (2005). The radioactive material may be continuously distributed or divided in sections within the encapsulation. Figure 1 shows a contact autoradiograph of two 125I sources; Figure 1a shows the 125I on a solid wire and Figure 1b shows the 125I deposited on two balls. Note the difference in the “isodose lines” of Figure 1. The uniformity and precision of the radiation from the brachytherapy source can be greatly affected by the distribution and encapsulation so that individual sources need to be measured. The TG-40 report (Kutcher et al., 1994) elaborates on these ideas. In addition, it has been recommended (Collé 1999) that the uniformity of seed and line sources be evaluated in terms of absorbed-dose-rate-towater at a distance of 2 mm from the source center, both longitudinally and perpendicular to the source axis (equatorial) in a tissue equivalent medium. In LDR applications of 125I and 103Pd seeds, there has been a proliferation of new seeds. To maintain consistency in seed dosimetry, procedures have been published (DeWerd et al., 2004) to assure the calibration and dosimetry values. Included in this publication are quality assurance (QA) procedures for maintaining the air-kerma strength standards for a given model seed. On each Accredited Dosimetry Calibration Laboratory (ADCL) calibration report there is a date to which the given seed is traceable to NIST. This assures that the correct air kerma strength is associated with the correct published dose-rate constant. The American Association of Physicists in Medicine (AAPM), American Brachytherapy Society (ABS), and U. S. Food and Drug Administration (FDA) recognize the need for accurate calibration in terms of well-defined physical quantities. The report of AAPM TG-56 states that the medical physicist should independently measure 10% of the sources to be implanted in the patient (Nath et al., 1997). Recently, other third parties have offered services to measure seeds. Sterilization procedures and insertion of the seeds into sterilized needles have given incentive to this activity. The medical physicist is still responsible for the dose given to the patient so the medical physicist should independently validate the output of seeds. Manufacturers assign uncertainties for their sources; however, many times these uncertainties are not met or are larger than desired for clinical dosimetry. Details on these measurements were given in the 1994 AAPM summer school publication (DeWerd and Thomadsen 1995) and in a course at the Radiological Society of North America (RSNA) (DeWerd 1997) where it is demonstrated that some calibrations actually fell outside ±10%. There are a number of anecdotal examples of the manufacturer stated activity being discrepant by 7% or greater for all source types for both LDR and HDR measurements. While
11–Calibration for Brachytherapy Sources
155
Figure 1a. Surface contour plot of an Amersham model 6711 LDR 125I source made with an autoradiograph.
Figure 1b. Surface contour plot of a North American Scientific model MED3631-A/M LDR 125I source made with an autoradiograph.
156
Larry A. DeWerd
the ±10% limit is workable for shipping or diagnostic purposes, each clinic should realize the importance of calibrating each source, since a 10% error can have deleterious and unacceptable effects on patients. Just as with 60Co teletherapy sources, individual users should calibrate each brachytherapy source as recommended by AAPM TG-41 (Glasgow et al., 1993). The calculation of dose to a patient depends on the accuracy of the well-type chamber or source calibration. Therefore, the accepted practice is to obtain a well chamber calibration from an ADCL. The ADCL is directly traceable to NIST and each maintains a high degree of precision in the transfer of calibration values to the clinical field instrument. The Calibration Coefficient provided by the ADCL is used to correct an instrument’s reading to obtain the reference value, which is the value of a quantity that has a known accuracy related to the primary value from a national laboratory. Calibration factors should be obtained for all instruments used to measure source outputs, e.g., for the ionization chamber and the electrometer used to read the charge accumulated during the measurement by the chamber. AAPM TG-40 (Kutcher et al., 1994) recommends that brachytherapy sources (or well chambers) have calibrations with direct or secondary traceability to national standards.
Calibration Concepts and Quantities Calibration Quantities and Units Source description quantities fall into one of two categories: measures of source output as emitted through the encapsulation and measures of radioactivity actually contained in the encapsulation. The source output is the quantity that is generally measured. When a given source is measured, it then reflects the radiation emitted which accounts for the encapsulation thickness variation and other parameters. This output is given in different units but the endpoint for dosimetry is absorbed-dose-in-water, with the unit being Gy [J/kg]. Apparent activity has been used in the past but it is not a quantity with a NIST-traceable standard. The quantity includes self-absorption or reduction of source output by its encapsulation and is obtained by dividing the measured source output by a known output value for the “bare” output for the radionuclide. This quantity should not be used for medical dosimetry as explained in appendix E of the update of the AAPM TG-43 report (Rivard et al., 2004). Photon Sources The quantity used to measure the output for photon-emitting sources is air kerma strength. The quantity kerma describes the first step in energy dissipation by indirectly ionizing radiation, transferring energy through one of the interactions to charged particles. Thus, the quantity kerma is defined as the energy transferred to charged particles per unit mass in a volume, as the volume limits to a point. Note that the medium (e.g., air, water, or tissue) must be specified. In modern brachytherapy dosimetry, air kerma strength is used, which includes the distance, and the TG-43 dosimetry equation is used to calculate absorbed dose. Air kerma strength is the quantity endorsed by the AAPM (AAPM 1987; Williamson and Nath 1991) and is similar to reference air kerma rate, which is endorsed by many European and international advisory bodies (IAEA 1999; ICRU 1997). Air kerma strength (AAPM 1987), SK, is defined as the product of airkerma rate (ICRU 1980) in free space, K" (d )[Gy / h ] , corrected for air attenuation and scatter, measured along the transverse bisector of the source, and the square of the measurement distance, d [cm],
SK = K" ( d ) ⋅ d .2
(1)
11–Calibration for Brachytherapy Sources
157
Generally, the distance of measurement is corrected to a “reference distance” of 1 m. As recommended by the AAPM TG-32 report (AAPM 1987), the units for air kerma strength, represented by U, are µGy • m2 • h–1 or cGy • cm2 • h–1. Notice that in any of the systems of specification, the reference distance, d0, need not be equal to the measurement distance, d. The ADCL provides calibrations of sources and well chambers in terms of air kerma strength. Beta Particle Sources Beta sources are absorbed in much smaller distances and thus need calibration in closer proximity to where they are used. Thus, the standard distance of 2 mm in water has been chosen as the calibration and delivery distance. This is the distance as chosen by AAPM TG-60 (Nath et al., 1999). In addition, for beta sources a calibration of absorbed-dose-to-water at a reference distance is the quantity of choice. This is accomplished by calibration after passing through a thickness of solid material that approximates water to have similar attenuation properties. NIST has chosen A150 plastic for this purpose (Soares, Halpern, and Wang 1998). In the application of these sources to intravascular situations, lesions in the coronary arteries are on the order of 2 to 4 cm in length in arteries with diameters of 3 mm to 5 mm. Therefore, radioactive line sources of a very narrow diameter, less than 1 mm, are used to deliver dose. Typical geometries include encapsulated seed trains or equivalent type line sources. The physical length of these sources vary but are generally between 3 and 6 cm to adequately cover the lesions undergoing treatment. The uncertainty involved in measurements of reference absorbed dose rate for betas at NIST is large, being given at 15% for k=2 or the 95% confidence level. However, recent communications from NIST have stated that this value has been reduced to 8% because of improvements in geometry and other methodology. However, since it was so large, an attempt was made to quantify the contained activity for one 90Sr source model. The importance of contained activity as a metric for source strength is to provide comparison between model predictions and dosimetry measurements. Monte Carlo calculations predict dose per history, where a history represents the interactions undergone by a single emitted photon or electron. The number of histories can be related to contained radioactivity by disintegration probabilities and branching ratios for complicated decay structures. Thus, Monte Carlo models predict dose rate per unit contained activity. Contained activity for beta-emitting sources can best be determined from a destructive measurement, which involves dissolving a source in a liquid medium that captures all of the contained activity into an aqueous solution (Collé 1999). By a suitable dilution of this solution, contained activity can be determined with a high degree of accuracy by the liquid scintillation technique. A contained activity calibration of a seed or wire beta-emitting source can be transferred to a well-type ionization chamber. This transfer then results in a method to specify a source in terms of contained activity rather than reference absorbed dose rate. The preferred quantity is absorbed dose to water at 2 mm (Nath et al., 1999) and thus, the quantity of contained activity needs conversion to this reference absorbed dose. Note that the problem that could be associated with the contained activity for routine clinical use is the dependence on the generalization of this quantity from the few sources that have been dissolved. Thus, variations in the cladding of the seed sources or other differences between dissolved sources and the source being measured are ignored. The recommended and most generalized quantity for beta source strength is the reference absorbed-dose-ratein-water. Ophthalmic sources also are calibrated to absorbed-dose-to-water; however, these sources are calibrated in terms of surface absorbed dose rate to water (Soares 1991). Since these sources have a distributed dose rate across their surface, the quantity is specified for the central 4 mm diameter, and the variation over the surface of the applicator becomes important.
158
Larry A. DeWerd
Primary Standards The primary standard at NIST is obtained either from calorimeters, an extrapolation chamber, standard chambers with precisely known volumes for high energies, or from free-air chambers (FACs) at low energies. Each of these would be considered a primary measurement device. For example, a FAC measures parameters that are directly in the definition of air kerma, and thus, it falls into the class of an absolute or primary dosimeter. These primary measurements are then transferred to the ADCL, which then transfers the calibration to the clinic through use of a reference class well chamber. The uncertainty in value is increased at each step, but the physicist at the clinic can expect that the uncertainty in the value for brachytherapy measurements from the ADCL is within 2% to 5%, depending on the radionuclide. Photon Sources for Brachytherapy NIST has various means to calibrate LDR photon sources, either with the Wide Angle Free Air Chamber (WAFAC) for low-energy sources or with an ionization chamber for high-energy sources. Currently, there is no direct NIST calibration for HDR 192Ir sources. However, there are primary calibrations on thimble or spherical chambers with external beams, which then can be used for a “primary” calibration on HDR sources. This technique is explained further below. LDR Sources. NIST developed a standard for 192Ir sources (Loftus 1980) using graphite cavity ionization chambers. The procedure has also been described in the literature (Verhaegen et al., 1992). The NIST procedure was to use a large number of LDR sources to provide adequate signal for the relatively insensitive cavity chamber used. The calibrated seeds were then measured individually in a 3.44-liter aluminum re-entrant ionization chamber to transfer the calibration in the future. Thereafter, the re-entrant chamber was used (Weaver, Loftus, and Loevinger 1988). This calibration was only made for two types of LDR 192 Ir seeds; additional types of seeds would require a redundant calibration of the type described by Loftus. The uncertainties at k=2 (95%) in these measurements at NIST is 2.7% for LDR 192Ir sources. The aluminum spherical reentrant chamber has a consistency of 0.2% to 0.3%. NIST uses the WAFAC to collect the radiation from low-energy, photon-emitting brachytherapy sources, such as 125I and 103Pd (Seltzer et al., 2003). For short-lived LDR brachytherapy sources, such as 125 I, a well-type chamber is used that was standardized by using sources calibrated against NIST standards, such as the WAFAC. This well chamber is used for consistency measurements. Figure 2 is a schematic
Figure 2. Schematic of the NIST WAFAC. [Modified from Seltzer et al. (2003).]
11–Calibration for Brachytherapy Sources
159
of the WAFAC, modified from Seltzer et al. (2003). The front window of the WAFAC is located 30 cm from the source, which is rotated. There are two volumes used to determine the source strength for a solid angle of acceptance. The aluminum filter between the seed and the entrance window of the WAFAC is generally used to filter out titanium K-edge characteristic x-rays at 4 keV. With the use of the WAFAC, the standard for these low-energy photon sources was established in January 1999. The history of the standard for low-energy photon brachytherapy sources is outlined below for clarity. In 1984 Loftus calibrated 125I sources using the Ritz Free air chamber (Seltzer et al., 2003). This calibration included contributions from titanium K-edge x-rays. In 1993, Loevinger introduced the WAFAC and added the aluminum filter to eliminate the titanium x-rays. NIST completed testing of the WAFAC in 1998 and developed a plan with the ADCLs and AAPM for implementation of the new standard. The Amersham model 6711 125I source was used for this testing as its calibration has been well characterized. On January 1, 1999, NIST introduced the new standard; Amersham models 6702 and 6711 were based on measurements performed in 1998. Through the model 6711 source, it was found in 2000 that measurements made in 1999 on various seed models had a 3% to 6% error for all 125I and 103Pd sources. For example, the difference in calibration coefficients for the seeds measured in 2000 versus the standard given in 1999 for the University of Wisconsin ADCL is given for 125I in Table 1 and for 103Pd in Table 2. This calibration change in the WAFAC standard resulted in calibration reports having dates of source calibration in 2000 or later as explained further in DeWerd et al. (2004). All the seed models are now based on a calibration value in 2000 or later. Since the dose rate constant for the source was determined from the NIST WAFAC air kerma strength, the old value of the dose rate constants have been changed accordingly to reflect the value from 2000 or after. Finally, the well chambers can be calibrated for LDR sources by the ADCL using a NIST-calibrated source. HDR Sources. As mentioned previously, NIST does not calibrate HDR 192Ir sources. The traceability of these sources is through NIST-calibrated ionization chambers at cesium and x-ray beam, M250, energies. These two beams are presently used to determine an interpolated value at 397 keV for calibration as given in Goetsch et al. (1991). Then a calibration method at multiple distances or with a shadow
Table 1. UW ADCL Observed Change in 125I Calibration for 2000 from 1999 NIST Standard Manufacturer and Model
Percent change in 2000 calibration to 1999 standard
North American Scientific MED3631A/M
–2.4%
Best model 2301
–3.3%
Mills 125SH/125SL
–8.6%
Syncor Pharmacy model
–4.4%
Draximage model LS-1
–9.2%
Table 2. UW ADCL Observed Change in 103Pd Calibration for 2000 from 1999 NIST Standard Manufacturer and Model
Percent change in 2000 calibration to 1999 standard
North American Scientific MED3633
–4.5%
Theragenics TheraSeed model 200
–4.2%
Int Brachy 1031L
–6.1%
Best model 2335
–6.5%
160
Larry A. DeWerd
shield is used with the above-interpolated value. A recent paper has expanded this concept to using a calibration coefficient for each energy line of the HDR 192Ir spectra, but the results agree within 1% (van Dijk, Kolkman-Deurloo and Damen 2004). David Rogers has found (private communication) that the differences are within 0.5%. The standard Goetsch technique has been improved and a comparison made between HDR manufacturers and sources (Stump et al., 2002). This work found that in the calibration of all sources, both the old and new Varian and Nucletron sources in various catheters agree in their calibration within 0.7%. In addition, an intercomparison was made with the Henri Becquerel Laboratory in Saclay, France, and agreement was found to less than 1% (Douysset et al., 2005). The uncertainty analysis for all of these calibrations is well outlined in Stump et al. (2002) to be ±2.1% at the coverage factor of k=2 or the 95% confidence level. The air kerma strength determination requires a calibration of the ionization chamber for the energy spectrum of 192Ir. The spectrum of an HDR 192Ir encapsulated source can be exposure-weighted to obtain an effective energy of 397 keV. This energy falls within 2% of the halfway point between 137Cs (662 keV) and the effective energy (146 keV) of a 250 kVcp medium filtration x-ray beam (HVL = 3.2 mm Cu). NIST provides air kerma calibration coefficients for both of these beam qualities. The ionization chamber wall must be thick enough to block all electrons emanating from the source or capsule, and to provide charged particle equilibrium (CPE) for the highest energy secondary electrons present in the 137Cs beam. Thinner-walled chambers require a buildup cap of at least 0.3 g/cm2; in the absence of a 137Cs buildup cap, a 60Co cap suffices. To establish the response of the chamber, the attenuation of the cap (Ax) must be eliminated. The response of the chamber alone, Nch, to a beam, is given by Nch = Nx Ax ,
(2)
where Nx is the NIST-traceable calibration coefficient for the chamber and cap. The calibration coefficient for 192Ir can then be obtained by interpolation (Stump et al., 2002) between the correction factors for the two bracketing energies from the following equation. (AwNx)Ir = [(AwNx)x-ray + (AwNx)Cs]/2 ,
(3)
where (Nx)Ir, (Nx)x-ray, and (Nx)Cs are the exposure calibration coefficients for 192Ir, 250-kVcp x-rays, and Cs respectively, and (Aw)Ir, (Aw)x-ray, and (Aw)Cs are the corresponding wall effects. From equation (3) the calibration coefficient (Nx)Ir can be determined. The air kerma strength can then be measured in free space along the transverse bisector of the source at a large distance relative to the dimensions of the source and detector. Specifying “free space” requires corrections for absorption and scattering in air, ionization chamber walls, or the room physical environment present in the calibration geometry. This methodology has been explained in previous publications (DeWerd and Thomadsen 1995; Goetsch et al., 1991; Stump et al., 2002). An improved jig giving less scatter and a better setup is described by Stump et al. (2002). Multiple distances are used to account for scatter and error in the setup distance. A comparison of measurements done at one distance versus those done at multiple distances show that the precision for the single distance measurements is ±3%, whereas the precision for multiple distances is ±1.5%. At typical brachytherapy treatment distances, ranging from a few millimeters to a few centimeters, conventional ionization chambers cannot be treated as point-like detectors. In addition, at these short distances, air kerma measurements are extremely sensitive to positional uncertainties. Both of these issues contribute a major part to the overall calibration uncertainty. Corrections for the finite size of idealized chambers (Dove 1959; Kondo and Randolph 1960) become large 137
11–Calibration for Brachytherapy Sources
161
for distances less than 5 cm. A method used to avoid distance uncertainties is to shield the chamber with a lead block to measure the scatter. Of course at the larger distance used, the scatter is a larger percentage of the signal. The shielded readings are subtracted from the unshielded reading to yield the net readings. This method results in very small readings on the ionization chambers and leakage must be accounted for or a larger volume chamber must be used. The method using the series of distances with precisely measured separations can be used to account for scatter and distance setup error. The actual distance for a reading is d′ = d + c ,
(4)
where d′ is the true center to center source chamber distance, d is the apparent center-to-center source chamber distance, and c is the error is setup distance (c can be positive or negative). The contribution of the room scatter radiation, K" S , is included in the measured air kerma rate, K" (d ) . Therefore the air kerma value from the primary radiation only, K" p (d ), is given by K" p (d ) = K" ( d ) − K" s .
(5)
The independence of the room-scatter air kerma rate contribution with respect to distance has been shown to be the case over distances of 10 cm to 40 cm (Goetsch et al., 1991). The air-kerma strength, SK, can then be written based on the inverse square law relationship, as
[
SK = K" p ( d ) i d ' = K" ( d ) − K" s 2
]( d + c )
2
(6)
for any distance. Three measurement groups at three distances can be used to determine the three unknowns, SK, K" S , and c. In actuality, the University of Wisconsin (UW) ADCL uses this technique with seven distances of 10, 15, 20, 25, 30, 35, and 40 cm. The seven distances above redundantly determine the scatter and error in distance since there are 3 unknowns with 35 solutions (DeWerd, DeWerd, and Attix 1993). A computer-generated solution then can be used to average the solutions. Thus, finally, the value of SK is determined without scatter and without distance error. This value then can be used by the ADCL to calibrate well chambers for clinical use. Beta Sources for Brachytherapy For beta sources, NIST uses an extrapolation chamber to measure either a single source or the ophthalmic applicator. In both cases, the chamber volume is extrapolated to a theoretical zero gap to measure the dose at the window surface. For intravascular sources, the absorbed-dose-to-water is specified at 2 mm. For ophthalmic applicators, it is specified as a surface dose. The extrapolation chamber is a primary standard for the determination of absorbed-dose-to-water at 2 mm for beta-emitting sources. Constructional details and operational performance of extrapolation chambers are given in Soares et al. (1998). The extrapolation chamber is basically an air-filled, plane-parallel chamber where the distance between the high-voltage and collecting electrodes (air gap) can be varied. The absorbed dose rate is determined from current measurements at a number of precise air gaps. The current values as a function of air gap are used to determine the slope of the data at the limit of zero air gap. The absorbed-dose-rate-in-water is then given by the Bragg-Gray relationship
162
Larry A. DeWerd
Dw =
water (W / e) ⋅ Sair ( ∆I / ∆$)$→0 kback , ρ0 a
(7)
where (W/e) is the average energy in joules needed to produce one coulomb of charge of either sign in water is the ratio of the mean mass collision stopping power of water to that dry air (33.97 ± 0.05 JC–1), Sair of air, ρ0 is the density of air at the reference temperature and pressure (T0, p0), a is the area of the collecting electrode, ( ∆I / ∆$)$→0 is the rate of change of corrected current (normalized to a reference temperature and pressure) with extrapolation chamber air-gap thickness as the thickness approaches zero, and kback is a correction factor that accounts for the difference in backscatter from the collecting electrode compared to that of water. The area, a [cm2], of the collecting electrode used is critical. The dose rate is determined from the measured currents within this area. The measured dose rate is averaged over this area. The area of the collecting electrode must be smaller than the radiation field being measured, so that the collecting electrode determines the measurement area rather than the radiation field area. The extrapolation chamber has been used at NIST to determine reference absorbed dose rate from a beta-emitting seed or a line of seeds (Soares, Halpern, and Wang 1998). For these measurements the source is inserted in a hole in a tissue-equivalent plastic block with the center of the source at a distance of 2 mm from the surface. At this depth, the radiation field from a seed or line source is such that a collecting electrode diameter of 1 mm can be used to measure absorbed-dose rate. There are problems with this method, mainly due to uncertainties in the effective collecting area of the extrapolation chamber and the divergence effect of the small source/collector geometry. This results in an uncertainty of ±15% (k=2). Taylor and Kuyatt (1994) explain evaluating uncertainties and coverage factors. When corrections are made for the geometric effect and smaller gaps are used, the calibration uncertainty has been decreased to ±8% (k=2).
ADCL Standards The primary air kerma strength calibrations for all brachytherapy sources has been transferred from NIST to the ADCLs. The HDR 192Ir source calibration is determined by the ADCLs as mentioned previously in the section HDR Sources, and is traceable to NIST through external beam calibrations of thimble or spherical chambers. The absorbed-dose-to-water calibrations performed by NIST via the extrapolation chamber have also been transferred to the ADCLs. The ADCLs maintain all these standards to a very high precision. These primary measurements are then transferred to the clinic via the ADCL through use of a reference class well chamber. The calibration precision is demonstrated by proficiency tests with NIST or round robins with each other. For all photon brachytherapy sources, all ADCLs are within ±0.6% of NIST during proficiency tests. For beta sources, agreement among the ADCLs and NIST is ±2%. The UW ADCL has done proficiency tests with NIST for ophthalmic applicators and was within 5%. In all cases, these proficiency tests fall within the stated criterion. In addition to the standards maintained for brachytherapy sources, which are then transferred to well chambers, the ADCLs calibrate electrometers also. Intercomparisons between ADCLs for electrometers are always within ±0.2%.
ADCL Source Calibrations and Ophthalmic Applicators The ADCLs will calibrate sources in addition to well chambers. This involves having a NIST-traceable calibration coefficient on a reference well chamber, since many sources have a relatively short half-life. The source to be calibrated is then inserted in the well chamber, the signal measured, and the air kerma strength determined. Because of the short half-life of many sources, it is generally more useful for the clinic to have a calibration on their well chamber instead of having sources calibrated.
11–Calibration for Brachytherapy Sources
163
Ophthalmic applicators generally use 90Sr, which has a half-life of 28 years. The UW ADCL has a NIST-calibrated source, which is then used with the source to be calibrated. Both sources are placed on radiochromic film at the same time for various exposures. This allows the surface dose on the film to be registered. Net optical density versus dose can then be plotted for the NIST-calibrated source, and the dose in the central 4 mm of the applicator can be determined. In addition, this allows an isodose plot to be made from the contact autoradiograph. The calibration of these sources is about 30% or more different from the prior calibrations done before 1991 (Soares 1991). In addition, the exposure surface is not uniform. For example, Figure 3 shows the autoradiograph of a uniform source. A dose profile in the two orthogonal directions also illustrates its uniformity. Many applicators are not as uniform as shown in Figure 3. For example, Figure 4 shows the autoradiograph of a very non-uniform source. The dose profile in the two orthogonal directions shows the dip in the center, which is shorter in the vertical direction than in the horizontal direction. Note the offset or shift of the dose-weighted isocenter from the physical source center and the non-uniform dose distribution. In some cases, ophthalmic applicators have to be rejected from calibration because of extreme dose profiles or damage of the sources.
Characteristics of Well Chambers for Brachytherapy Sources Well ionization chambers are a common piece of equipment for measurement of brachytherapy source strength. It is highly recommended that the well chamber be dedicated to brachytherapy uses and not be shared with nuclear medicine. The major reasons for this are twofold. First, when the chamber is not under control of the therapy medical physicist, “situations” can develop such as contamination from an unsealed liquid nuclear medicine source. Second, the brachytherapy source calibration is highly dependent on the insert used and subsequent geometry. Calibration at an ADCL is done with a specific insert and it is part
Figure 3. Contour plot of a uniform 90Sr ophthalmic applicator.
164
Larry A. DeWerd
Figure 4. Isodose contour plot of a non-uniform 90Sr ophthalmic applicator.
of the calibration coefficient. The general characteristics for well chambers are given below. The medical physicist should be aware of the equipment used and characteristics that need to be kept in mind. Many well chambers have a “sweet spot” where the output is highest for a given source. Figure 5 shows a typical example for a photon source, in this case an HDR 192Ir source. From this figure it can be noted that a longer source, e.g., a 60-mm train, would result in a 6% decrease in signal at the ends. This falloff is the problem with measurement of source trains because of the axial geometry limitations of the sweet spot of well chambers. Therefore, many manufacturers have developed new chambers or modified existing ones to cover longer sources. These chambers have an extended region of uniform axial response, or a sweet length. An example of such a chamber is shown in Figure 6. As part of the calibration, the ADCL provides a determination of the ion collection efficiency for well chambers. Ion collection efficiency, Aion, is a measurement of the percentage of charge measured by the chamber versus the total charge, which is a measure of the ionization recombination of the chamber. To know the dose absorbed by the air in the ionization chamber requires collecting all the ions formed by the radiation. For many well chambers for these sources, Aion is equal to 1.000 and so can be ignored. However, Aion may not be unity for HDR sources. Well chambers can be open to the atmosphere or pressurized. If they are open to the atmosphere, correction for air density should be made. If they are pressurized, a quality control measurement with time (quarterly) should be made to check for loss of pressure caused by leakage. Air density corrections (temperature and pressure) are calculated according to equation (8). , kTp =
(273.15 + T ) 101.3 (273.15 + T ) 760 • • = (273.15 + T ) p (273.15 + T ) P 0 0
(8)
11–Calibration for Brachytherapy Sources
165
where T is the temperature in °C, p or P is the pressure in kPa or mm Hg, respectively, and T0 is the reference temperature at time of calibration (22.0 °C in the United States). As with all ionization chambers, periodic consistency tests with a long-lived source (137Cs for example, or an external 60Co beam) should be part of the routine QA program. An acceptable method for these checks in addition to those mentioned above is a redundancy check or intercomparison with one or two other chambers on the same source in the same time period. In particular, having two or three chambers that have been calibrated at an ADCL, which then would be used to measure the same source within a short time period is acceptable. If necessary, account should be made for decay of the source. Either the same electrometer for each chamber or independent dosimetry systems can be used for this exercise
Figure 5. Normalized response of a well chamber along its axial length.
Figure 6. An example of a well-chamber response from the bottom of the insert showing the axial sweet length for 192Ir.
166
Larry A. DeWerd
applying the calibration coefficients for each. The chambers or the systems, whether at a given center or from a neighboring center, should be used at the same time and a ratio of results kept. This method relies upon the stability of the systems with respect to each other. This method is suggested in the AAPM TG-56 document (Nath et al., 1997). In addition, the concept of redundant measurements has been proven for external beam chambers in the paper by Rozenfeld and Jette (1984). Such tests not only verify the calibration of the chamber, but also check that sealed chambers maintain their pressure. The response of the chamber should remain constant to within ±0.5% (Goetsch et al., 1992). The ADCL also calibrates electrometers for use with the well chambers. Both the chamber and electrometer correction coefficients must be used to achieve correct results. The electrometer and the well chamber then operate as a system. The scale of the electrometer should be low enough so that good precision and an adequate signal-to-noise are achievable. This is especially crucial for LDR sources where the noise can be a significant amount of the signal for some electrometers. For this reason many manufacturers have made electrometers with pA scales.
Well-Chamber Characteristics for Photon Sources Note in each case, the appropriate source holder must be used since it is part of the calibration. For HDR sources, the ion recombination should be checked, since not all chambers have full collection efficiency. Recombination correction, krecom, may be determined using the two-voltage technique for gamma-emitting sources. If the ratio of the voltage used in this technique is exactly 2 (e.g., if 150V and 300V are used as is often the case with well type chambers) then the recombination correction can be determined from (Attix 1986) 1 krecom
=
4 1 Q300 . − 3 3 Q150
(9)
The ADCL has determined Aion previously upon calibration; the user will have to determine for the source on site. Good-quality chambers generally exhibit negligible recombination effect for gamma-emitting brachytherapy sources. Many well-type chambers have proven their stability with time. Evaluations of commercially available well chambers show a long-term reproducibility of 0.5% or better. If during the constancy checks of the well type chamber it is observed that its response changes more than ±1%, a recalibration is recommended.
Well-Chamber Characteristics for Beta Sources The beta sources presently used for intravascular brachytherapy are 90Sr-90Y sources. Much of the well chamber characteristics were reported on in the previous summer school of 2002 (DeWerd 2002). The source trains are sent to the end of the catheter by pressure from de-ionized water that runs along the inner lumen, past the pellets, and back through the outer lumen. It is important that the catheter be purged with water prior to each measurement. Void regions in the catheter can produce significant variation. Measurements of each source should be done at 90° rotations in the well chamber to average the positioning of the source and the lumen. In this case, the insert used in the chamber is a major part of the calibration, since it has a direct effect on the chamber system. For calibration of long intravascular brachytherapy source trains, the response curve or sweet length (variation of chamber output with position within the well chamber) must be of sufficient length to accommodate the source. DeWerd (2002) elaborates on the measurements made for various chambers. The sweet lengths of chambers for these source trains extend from 50 mm to over 100 mm.
11–Calibration for Brachytherapy Sources
167
For a well chamber to be useful in measuring a beta source, it is necessary to have accurate calibration coefficients by which the dose rate of the source can be determined from the signal generated in the well chamber. Since this source is beta emitting, a number of questions were involved in the calibration of these chambers in addition to the axial response or sweet length. DeWerd (2002) addressed four areas of concern, the first being the axial response. In addition, the type of radiation best measured, e.g., create bremsstrahlung radiation or measure beta particles; catheter uniformity for calibration; and calibrations for the various lengths of seed trains were also addressed. For the type of radiation measured it was found that the best compromise with respect to signal-to-noise ratio, rotational consistency and accuracy were obtained by filtering out low-energy beta particles without the creation of additional bremsstrahlung. For these reasons a 2-mm acrylic holder was chosen for use with the well chambers. A study was also done for any variations among catheters. The results showed that the calibration coefficient for various catheters was the same within ±0.1%. While it is imperative to match the clinically relevant catheter to the specific source train length, there appears to be no concern regarding use of multiple catheters for well chamber measurement of Novoste source trains. Finally, an investigation was conducted for calibrations for the various lengths of seed trains. Each Novoste train was inserted within the sweet length of the chamber and measured through a rotation of 360°, in 90° steps, for a total of five measurements of each train. The relationships between the 30 mm, 40 mm, and 60 mm trains were found to be in proportion to their length or number of seeds. However, some chambers tested were up to 4% different from the expected ratio. DeWerd (2002) showed that the various train calibration coefficients are related in proportion to the length of the train. The ADCL will thus calibrate one seed train length for a chamber that follows the relationship as given in equations (10). D40 = 0.75D30 (10) D60 = 0.50D30 where D30 is the 30 mm calibration coefficient, D40 is the 40 mm absorbed dose-to-water calibration coefficient, and D60 is the 60 mm absorbed dose-to-water calibration coefficient.
Clinical Calibration Using Well Ionization Chambers Well-type chambers provide a reliable method for calibrating brachytherapy sources before clinical use and are recommended for use for both brachytherapy photon sources, and beta sources. The preferred method for traceability in the source calibrations is to have the well-type chamber calibrated against the primary standard at an ADCL. This calibration should be carried out for each source model to be used with the appropriate source holder made by the manufacturer for that source. The source holder is part of the calibration. The accuracy of the source calibration limits the final accuracy of the dose calculation for the patient in the clinic. Therefore, the clinic needs to determine the air kerma strength, SK, for photon sources or the absorbed-dose-to-water for beta sources. In the clinical situation this calibration is accomplished with a calibrated well chamber. The user’s well chamber should come to equilibrium with its surroundings before beginning clinical calibration. The minimal time necessary for this is generally 30 minutes. Care should be taken that the temperature measured is that of the chamber volume and not the room temperature. A minimum of three
168
Larry A. DeWerd
significant digits should always be obtained for charge accumulated or current measurements. Thus charge should be accumulated for a set time depending on the activity of the source. A minimum of three measurements for each source insertion that are neither monotonically increasing nor decreasing should be obtained.
Clinical Calibration of Photon Sources When using chambers for calibrations, the chamber should be located in a “scatter free” location (Podgorsak et al., 1992) .The calibration coefficient of the chamber assumes that the source will occupy the same position as during calibration by the ADCL. This position is usually that of peak response or the center of the uniform response plateau. The user should verify the position before proceeding. See the section, Well-Chamber Characteristics for Beta Sources, and Figure 5 for further information on the sweet spot. The calibration coefficient for the chamber gives air kerma strength per unit reading. Measurements for the HDR source generally use the current mode whereas for LDR sources, it may be either charge or current. For the current measurements, SK = I • Ne • kTP • NSK • Aion • Pion ,
(11)
where: SK is the source air kerma strength, I is the current measured from the well chamber, Ne is the correction coefficient for the electrometer scale, kTP is the correction for temperature and pressure [see equation (8)]; this factor is not used for pressurized chambers, NSK is the ADCL provided air kerma strength calibration coefficient for the well chamber, Aion is the correction for collection efficiency at the time of calibration, from the ADCL calibration report, and Pion is the correction for collection efficiency at the time of measurement. Note that equation (11) applies to LDR photon-emitting sources as well, with two differences. The first involves Aion and Pion, which are generally equal to 1.000 and thus ignored. The second is kTP, which is given by equation (8). Recently, deviations from the standard air density correction have been found for low-energy, LDR photon sources (Griffin et al., 2005; Bohm et al. 2005). These deviations are found at various pressures that enter at different altitudes. There is an additional correction, PA that can be applied after applying the standard kTP given by the relation: PA = k1 ( P ) 2, k
(12)
where P is the pressure and the constants, k1 and k2, relate to the type of source, 103Pd, 125I with a silver base or without a silver base, and the type of well chamber. These constants are given in Griffin et al. (2005) and reproduced in Table 3. In addition, for LDR sources, the signal-to-noise and the accompanying selection of an electrometer may be crucial; see Characteristics of Well Chambers for Brachytherapy Sources.
11–Calibration for Brachytherapy Sources
169
Table 3. Constants k1 and k2 for three chambers and three types of seeds Type of LDR source
HDR 1000 and IVB 1000
PRM WC 2 chamber
k1
k2
k1
k2
Pd
0.0241
0.562
0.171
0.266
I with silver base
0.0490
0.455
0.082
0.378
I without silver base
0.0573
0.431
0.069
0.403
103 125 125
Clinical Calibration of Beta Sources For beta-emitting sources, the calibration procedure is much the same as above for photon sources; however, it is of much greater importance that close attention is paid to the source holder and setup of the chamber. The sweet length and source insert are very important as described previously in the section, Well-Chamber Characteristics for Beta Sources. When performing clinical measurements of the beta sources, make sure the correct settings are used on the insert for the appropriate length source. Measurements should be made at various orientations of the source about its cylindrical axis and the results averaged. Multiple insertions of the sources should also be made for loose seed trains delivered in a catheter. Since there is a polarity effect for beta particles, the calibration coefficient is only valid for the polarity stated in the calibration certificate. The same polarity as used in the ADCL calibration should be used in the clinic. The ion collection efficiency measurements are ignored for beta-emitting source measurements. The correction factor below is for a given train length, e.g., 30 mm. D2mm = Mu • Ne • N2mm • kTp ,
(13)
where the quantities are the same as equation (11) except that D2mm is the absorbed-dose-in-water at 2 mm and N2mm is the calibration coefficient for the beta source provided by the ADCL. Again, if a pressurized chamber is used here, kTP is not used. However, there is a 1% effect at lower air density for beta sources for pressurized chambers (Griffin et al., 2005).
Conclusion The use of brachytherapy sources for brachytherapy is increasing. It is important that the patient receive the maximum benefit from this technique by good physics practices. Critical to this application is the calibration of the source. The present procedures have increased the dosimetric accuracy greatly for the clinic. Well chambers have simplified the measurement problems at the clinical level. This chapter has demonstrated that calibration of sources traceable to national standards, is important.
References American Association of Physicists in Medicine (AAPM). Report No. 21. “Specification of Brachytherapy Source Strength.” New York: American Institute of Physics, New York, 1987. Attix, F. H. Introduction to Radiological Physics and Radiation Dosimetry. New York: John Wiley and Sons, 1986. Bohm, T. D., S. L Griffin, P. M. DeLuca, and L. A. DeWerd. (2005). “The effect of ambient pressure on well chamber response: Monte Carlo calculated results for the HDR 1000 Plus.” Med Phys 32:1103–1114.
170
Larry A. DeWerd
Collé, R. (1999). “Chemical digestion and radionuclide assay of TiNi-encapsulated 32P intravascular brachytherapy sources.” Appl Radiat Isot 50:811–833. DeWerd, L. A. Brachytherapy Dosimetric Assessment: Source Calibration, Radiological Society of North America, Categorical Course in Brachytherapy Physics. B. Thomadsen (ed.). Oak Brook, IL: RSNA, 1997. DeWerd, L. A. “Source Standardization and Calibration for Intravascular Brachytherapy” in Intravascular Brachytherapy/Fluoroscopically Guided Interventions. S. Balter, R. C. Chan, and T. B. Shope, Jr. (eds.). American Association of Physicists in Medicine 2002 Summer School proceedings. Medical Physics Monograph 28. Madison, WI: Medical Physics Publishing, pp. 423–443, 2002. DeWerd, L. A., and B. R. Thomadsen. “Source Strength Standards and Calibration of HDR/PDR Sources” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.). American Association of Physicists in Medicine (AAPM) 1994 Summer School. Madison, WI: Medical Physics Publishing, 1995. DeWerd, L. A., S. M. DeWerd, and F. H. Attix. (1993). Solution to Inverse Square Equations Involving Distance Error and Scatter Correction. Wisconsin Medical Physics Report #197 Madison; WI. (Available from the University of Wisconsin.) DeWerd, L. A., M. S. Huq, I. J. Das, G. S. Ibbott, W. F. Hanson, T. W. Slowey, J. F. Williamson, and B. M. Coursey. (2004). “Procedures for establishing and maintaining consistent air-kerma strength standards for low-energy, photon-emitting brachytherapy sources: Recommendations of the Calibration Laboratory Accreditation Subcommittee of the American Association of Physicists in Medicine.” Med Phys 31:675–681. Douysset, G., J. Gouriou, F. Delaunay, L. DeWerd, K. Stump, and J. Micka. (2005). “Comparison of dosimetric standards of USA and France for HDR brachytherapy.” Phys Med Biol 50:1961–1978. Dove, D. B. (1959). “Effect of dosemeter size on measurements close to a radioactive source.” Br J Radiol 62:202–204. Glasgow, G. P., J. D. Bourland, P. W. Grisby, J. A. Meli, and K. A. Weaver. “Remote Afterloading Technology.” AAPM Report No. 41. New York: American Institute of Physics, 1993. Goetsch, S. J., F. H. Attix, D. W. Pearson, and B. R. Thomadsen. (1991). “Calibration of 192Ir high dose-rate afterloading systems.” Med Phys 18:462–467. Goetsch, S. J., F. H. Attix, L. A. DeWerd, and B. R. Thomadsen. (1992). “A new well ionization chamber for the calibration of iridium-192 high dose rate sources.” Int J Radiat Oncol Biol Phys 24:167–170, Griffin, S. L., L. A. DeWerd, J. A. Micka, and T. D. Bohm. (2005). “The effect of ambient pressure on well chamber response: Experimental results with empirical correction factors.” Med Phys 32:700–709. International Atomic Energy Agency (IAEA) IAEA-TECDOC-1079. “Calibration of Brachytherapy Sources.” Vienna, Austria: IAEA, 1999. International Commission on Radiation Units and Measurements (ICRU). Report No. 33. “Radiation Quantities and Units.” Bethesda, MD: ICRU, 1980. International Commission on Radiation Units and Measurements (ICRU). Report No. 58. “Dose and Volume Specification for Reporting Interstitial Therapy.” Bethesda, MD: ICRU, 1997. Kondo, V. S., and M. L. Randolph. (1960). “Effect of finite size of ionization chambers on measurements of small photon sources.” Radiat Res 13:37–60, American Association of Physicists in Medicine (AAPM). (1994). “Comprehensive QA for radiation oncology. Report of AAPM Radiation Therapy Committee Task Group 40.” Med Phys 21:581–618. Also available as AAPM Report No. 46. Loftus, T. P. (1980). “Standardization of 192Ir gamma-ray sources in terms of exposure.” J Res Nat Bur Stand 85:19–25. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24(10):1557–1598. Also available as AAPM Report No. 59. Nath, R., H. Amols, C. Coffey, D. Duggan, S. Jani, Z. Li, M. Schell, C. Soares, J. Whiting, P. E. Cole, I Crocker, and R. Schwartz. (1999). “Intravascular brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 60.” Med Phys 26:119–152. Also available as AAPM Report No. 66. Podgorsak, M. B., L. A. DeWerd, B. R. Thomadsen, and B. R. Paliwal. (1992). “Thermal and scatter effects on the radiation sensitivity of well chambers used for high dose rate Ir-192 calibrations.” Med Phys 19:1311–1314.
11–Calibration for Brachytherapy Sources
171
Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. (2004). “Update of AAPM Task Group No. 43 report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Rivard M. J., B. L. Kirk, and L. C. Leal. (2005). “Impact of radionuclide physical distribution on brachytherapy dosimetry parameters.” Nucl Sci Engr 149:101–106. Rozenfeld, M., and D. Jette. (1984). “Quality assurance of radiation dosage: Usefulness of redundancy.” Radiology 150:241–244. Seltzer, S. M., P. J. Lamperti, R. Loevinger, M. G. Mitch, J. T. Weaver, and B. M. Coursey. (2003). “New national air-kerma-strength standards for 125I and 103Pd brachytherapy seeds.” J Res Nat. Instit Stand Tech 108:337–358. Soares, C. G. (1991). “Calibration of ophthalmic applicators at NIST: A revised approach.” Med Phys 18:787–793. Soares, C. G., D. G. Halpern, and C.K. Wang. (1998). “Calibration and characterization of beta-particle sources for intravascular brachytherapy.” Med Phys 25:339–346. Stump, K. E., L. A. DeWerd, J. A. Micka, and D. R. Anderson. (2002). “Calibration of new high dose rate 192Ir sources.” Med Phys 29:1483–1488. Taylor, B. N., and C. E. Kuyatt. (1994). “Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results.” NIST Technical Note 1297. Washington, DC: U.S. Dept. of Commerce, 1994. van Dijk, E., I. K. K. Kolkman-Deurloo, and P. M. G. Damen. (2004). “Determination of the reference air kerma rate for 192Ir brachytherapy sources and the related uncertainty.” Med Phys 31:2826–2833. Verhaegen, F., E. van Dijk, H. Thierens, A. Aalbers, and J. Seuntjens. (1992). “Calibration of low activity 192Ir brachytherapy sources in terms of reference air kerma rate with large volume spherical ionization chambers.” Phys Med Biol 37:2071–2082. Weaver, J. T., T. P. Loftus, and R. Loevinger. “NBS Measurement Services: Calibration of Gamma-Ray-Emitting Brachytherapy Sources.” National Bureau of Standards (NIST) Special Publication 250-19, Washington, DC: NIST, 1988. Williamson, J. F., and R. Nath. (1991). “Clinical implementation of AAPM Task Group 32 recommendations on brachytherapy source strength specifications.” Med Phys 18:439–448.
Chapter 12
Localization, Part I: Radiographic Methods and Accuracy Eugene P. Lief, Ph.D. Maimonides Comprehensive Cancer Center Brooklyn, New York Introduction: Requirements for Localization Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Use of Various Imaging Modalities for the Source Localization . . . . . . . . . . . . . . . . . . . . . . . . . 174 Conventional Radiographs and Autoradiographs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Orthogonal Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Stereo-Shift Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Automatic Reconstruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 HDR Dwell Positions Localization Using Autoradiographs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 CT Scanning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Catheters and Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Prostate Seed Post-Implant Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Automatic CT-based prostate Seed Detection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Electronic Portal Imaging Devices (EPIDs), Fluoroscopy, and Cone-Beam CT . . . . . . . . . . . . . . . 181 Source Localization for Various Brachytherapy Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 LDR Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 HDR Brachytherapy Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Intravascular Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Other Methods of the Source Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Introduction: Requirements for Localization Accuracy By definition, “brachytherapy” (“brachy” means “close” in Greek) is radiotherapy at small distances. Such distances (from zero to a few centimeters from the source of radiation) at present can be achieved utilizing small, sealed radioactive isotopes. Due to the inverse square law, which is valid for radiation from small (point) sources, a small inaccuracy in the distance measurements can result in significant difference in dose (Smith, Meigooni, and Chiu-Tsao 1990). For example, if we want to determine radiation dose at a distance of 1 cm from a small radioactive source, inaccuracy of 1 mm would result in a factor of (1.1 cm/1 cm)2 = 1.21 in dose calculation. That means more than 20% error in dose administration, which is a significant discrepancy for radiotherapy. When the dimensions of the sources are comparable with the distance to the tumor or the source radiation is not isotropic, there is an additional consideration of the source orientation, which could be established using plain radiographs (Lief, Davis, and Wallner 1998; Davis et al., 2003). Due to these considerations, source localization was one of the most important parts of brachytherapy. It can be achieved by using x-rays, ultrasound, magnetic resonance imaging (MRI), or other nonradiographic technique. In this chapter we will focus on radiographic localization, just briefly mentioning alternative approaches. Radiographic techniques can be further subdivided into plain radiographs, fluoroscopy, computed tomography (CT), and autoradiographs. By “plain radiographs” we mean taking radiographic films (or digital images obtained by electronic portal imaging devices (EPIDs) using external diagnostic x-ray source. Fluoroscopy utilizes continuous x-ray irradiation that is detected electronically, with the image being continuously updated on the screen. CT methods are based on performing a computer tomography scan and subsequently localizing (manually or automatically) radioactive or “dummy” sources as well as source applicators. One of the latest developments is utilization of a cone-beam CT feature of the Acuity simulator (Varian Medical Systems, Palo Alto, CA) for catheter localization in high dose rate
173
174
Eugene P. Lief
(HDR) treatment of prostate (Prestidge 2004). Autoradiographs are performed by registering the source location on attached radiographic film, using radiation from the radioactive isotope instead of external source. Historically, plain radiographs with external x-ray source and autoradiographs were more commonly used when the CT was less available. Currently, the CT has become the most efficient imaging modality for prostate seeds post-implant localization and also for localization of HDR 192Ir applicators made of special CT-compatible materials. Autoradiographs are predominantly used for HDR 192Ir and intravascular source localization.
Use of Various Imaging Modalities for the Source Localization Conventional Radiographs and Autoradiographs Two or more plain radiographs provide the most accurate way of determining the source position as well as orientation. The most important tasks in the radiograph-based localization are matching the images of the same sources on all the films and correctly determining magnification factors of the films. Another challenge is to find the source location with respect to the tumor. Usually, the treatment target visualization can be performed using the CT easier than with the radiographs; however the CT is generally less convenient for the source localization. Orthogonal Techniques The simplest two-film technique is the one with the film planes orthogonal to each other. Typically, these are anterior-posterior (AP) and lateral films or two films taken at any directions orthogonal to each other. Sometimes these directions have to be different than AP/Lateral to avoid prostheses or other objects which are difficult to image. In such cases, the imaging could be done using paired left anterior oblique (LAO) and right anterior oblique (RAO) radiographs. To correctly determine the magnification factor, a magnification ring (or an opaque circle, e.g., a quarter coin) may be placed as close to the implant area as possible without occluding the view. The magnification factors for both films 1 and 2 in the vicinity of the ring M1(r) and M2(r) can be calculated then as a maximum distance between any two points of the ring on the images 1 and 2, respectively, divided by the actual ring diameter. Usually it is reasonable to assign the same magnification factor to all the seeds on the image, as the size of the implanted domain is usually much smaller than the distance from the radioactive sources to the x-ray tube. Nevertheless, since the location of the ring is typically different from the implanted domain, film magnification factors for the seeds are generally different from M1(r) and M2(r) and could be found using the following formulas (Smith, Meigooni and Chiu-Tsao 1990):
M1 =
(F F + x 1 2
(r )
1
)
F1 F2 + x1 z2
(
(1)
)
z2 / M 1( r ) + F2 z2( r ) − z2 / M 2( r )
and similar for M2 (see Figure 1): M2 =
F1 F2 + x1 z2
(F F + x z ) / M (r )
1 2
1 2
(r ) 2
(
)
− F1 x1( r ) − x1 / M 1( r )
,
(2)
where F1 and F2 are source to film distances for beams 1 and 2, respectively, and x1 and z2 are projections of a point deviation from the isocenter on the planes of the films. F1 and F2 are negative if the
12–Localization, Part I: Radiographic Methods and Accuracy
175
Figure 1. Radioactive source localization using two orthogonal radiographs.
corresponding beam is directed opposite the coordinate axis. In equations (1) and (2), x1(r) and z2(r) refer to x- and z-coordinates of the magnification ring. Using M1 and M2, one can find the actual spatial coordinates of each source. The advantage of the orthogonal method is its simplicity and accuracy. Film technique offers sharp images of the seeds implanted. This is especially noticeable for AP films with better contrast than lateral films. Using films, each coordinate of nonoverlapping seeds can be typically digitized with the accuracy of 0.5 mm. The disadvantage of the method is increasing difficulty of the seed matching with the increase of their quantity. The important factor in seed matching is identity of y-coordinates of the same seed on both projections. Other difficulties include: 1. Possible patient movement between the exposures: if the patient moves between the exposures, the source matching would be much more difficult and much less accurate, and 2. Inability to distinguish different possible coplanar seed locations: an example of the uncertainty of the seed locations would be the registration of two nonexisting “sources,” shown as open circles in Figure 1, instead of two actual seeds (S and I) shown in solid. Stereo-Shift Method As mentioned above, the orthogonal technique becomes inconvenient if there is a large number of sources implanted and especially if many of them lie in the same plane. Another technique, called “stereo-shift,” could be utilized in this case (Smith, Meigooni, and Chiu-Tsao 1990). The idea of this method is to take two images from slightly different positions. It is usually achieved by moving the x-ray tube with respect
176
Eugene P. Lief
to the patient between the two exposures and using a fiducial marker attached to the patient at a predefined location near the implant. The magnitude of the shift should be small enough with respect to the source-to-seeds distance, so the radiographs would look similar and make seed matching easier. On the other hand, the distance of the shift should be significantly larger than the difference in heights of the seeds and the fiducial marker to make the method more accurate. A typical shift is about 20 cm. Figure 2 shows the scheme of the z-coordinate determination for a seed moving with the patient from location S1 to S2, with respect to the x-ray tube. The idea of the stereoshift method is utilizing the fact that as a small object (a radioactive source S or a fiducial marker F) is moved by a fixed distance s with respect to the x-ray source, its “shadow” on the film will move by a distance x which depends upon the object’s height z. The shift s, applied to the source, moves it from position S1 to S2, while the fiducial marker after the same shift moves from F1 to F2. These identical shifts change the distance between x-ray images of the objects S and F on the film from x1 to x2 (Figure 2). Using the magnitude s of the stereoshift and a height of the fiducial marker zf, one can arrive at the following expression for the unknown height of the seed zs (Smith, Meigooni, and Chiu-Tsao 1990):
zs = SFD ⋅
( SFD − z ) ⋅ ( x − x ) − s ⋅ z , ( SFD − z ) ⋅ ( x − x ) − s ⋅ SFD 2
f
f
1
2
f
(3)
1
where SFD is the source-to-film distance. The magnification factor M for each seed can be found, using a formula:
M =
SFD − zs SFD
.
(4)
The advantage of the stereo-shift method is ease of seed identification. However, the accuracy of the stereo-shift method is lower than for the orthogonal films, since a small error in the source coordinate measured from a film results in a parallax with a much larger error in the seed height. For example, if the magnitude of the stereo-shift s = 20 cm, SFD = 140 cm, and zs = 40 cm, then an error ∆x = 0.5 mm in seed digitization translates into a larger error ∆zs in the seed z-coordinate calculation: ∆ zs ≈
∆x ⋅ ( SFD − zs )
s SFD ⋅ tan SFD − zs
≈ 1.8 mm .
(5)
In this formula, the value of the angle in tangent function is expressed in radians. The error ∆y in ycoordinate found from seed digitization is not magnified because of the parallax but simply scaled by the magnification factor. Numbers chosen for this example can vary depending on the setup. Automatic Reconstruction There are a number of various approaches to automatic reconstruction using plain radiographs (Amols and Rosen 1981; Altschuler, Findlay, and Epperson 1983; Biggs and Kelley 1983; Rosenthal and Nath 1983; Stock et al., 1995; Todor et al., 2002). Using the same idea of imaging the implant from two or more different angles, one can use the couch translation or the gantry rotation to obtain those images.
12–Localization, Part I: Radiographic Methods and Accuracy
177
Figure 2. Scheme of the source S localization using stereoshift s and fiducial m ark F.
By matching the source images on the radiographs, measuring their coordinates with respect to the center of the images, and entering the coordinates in the computer, one can reconstruct the spatial distribution of the sources. One of the methods is based on two conventional radiographs; one is typically in AP direction, and the other, at a specific angle (usually, 15° to 45°) with respect to the first one. The two-film method requires identifying seeds on the first film and finding corresponding seeds on the second film. The problem becomes more difficult with an increasing number of seeds. It also becomes more tedious to correlate seed images on both films for larger angle between the gantry positions at which the films were taken. On the other hand, the reduction of this angle significantly reduces the accuracy of spatial seed localization in the AP direction. The compromise is usually achieved at an angle of about 20°. For the two-film method, as for the stereo-shift method above, the largest uncertainty in coordinate determination is in the z-direction. Similar to equation (5), one can find that an error ∆x = 0.5 mm in the seed digitization translates into a larger error ∆zs in the seed z-coordinate calculation:
∆ zs ≈
∆x ⋅ ( SFD − zs ) SFD ⋅ sin ( 20° )
≈ 1.0 mm .
(6)
178
Eugene P. Lief
In this formula, the value of the angle in sine function is expressed in degrees. As in equation (5), the numbers are chosen just for the illustration. They will vary depending on the geometry of each particular case. As mentioned in the orthogonal method description, two-film methods can result in registration of nonexisting seeds instead of the real ones. One way to eliminate this is to use the third radiograph (Amols and Rosen 1981; Rosenthal and Nath 1983). In this approach, there is an AP film and two other films symmetrically taken at the gantry angles ±α, where the angle α could range from 8° to 45°. The 45° approach allows more accurate determination of the z-coordinates of the sources. To make matching easier, one could compare y-coordinates of the seeds. For the same seed, these coordinates should be identical on all three images. If the accuracy of the seed digitization in the three-film method is limited, two seeds with close z-coordinates could be interchanged. This and other difficulties in seed matching can leave a certain number of sources on both films unmatched. In this case, the missing activity can be reassigned among the matching sources. It has been suggested that up to 5% of seeds could be mismatched for the dose calculations to be performed with the acceptable accuracy (Rosen et al., 1982). HDR Dwell Positions Localization using Autoradiographs An important example of brachytherapy source localization is determination of actual coordinates of HDR Ir source dwell positions. Usually, the consistency of the dwell positions is checked using a plain endobronchial catheter of a known length. This can be done by attaching a 15 to 30 cm length of catheter at the closed end to a sheet of a plain ready-pack film, using masking tape. Ready-pack “XV” or “EDR”type films (Eastman Kodak Company, Rochester, NY), which contains a single sheet of film in a lightproof paper envelope, can be used for this purpose. Some other types of film with higher speed, “XTL” or “PPL,” are too fast for the HDR exposures, and the autoradiograph would have excessive blurring from the source motion. If there is a noticeable motion of the film sheet within the paper envelope, the film can be stapled through the envelope at two corners 5 cm or more from the catheter. First, a metal wire with non-active x-ray opaque, nonradioactive markers 10 mm apart (“dummy wire”) is inserted in the empty catheter. The markers are indicating intended dwell positions of the HDR source. The wire has to go all the way into the catheter and be securely fastened by a special cup at the open end. The film with the wire located in the taped empty catheter is exposed to 80 to 120 kVp diagnostic x-rays. The side of the envelope with the catheter should be closer to the x-ray tube. This exposure leaves on the film (after processing) sharp light spots at the markers’ locations. After the first exposure, the catheter remains taped to the film envelope, while the “dummy wire” is removed. The empty catheter is connected to the HDR, and the source is preprogrammed to stop at the positions corresponding to the locations of the “dummy” markers. The dwell time at each position depends on the source strength and the film speed. During the HDR run, the film receives high doses from the source at the dwell locations. After the HDR exposure, the film is processed. The developed film exhibits sharp light spots from the markers overlaid with the dark blurry spots from the HDR source. For the analysis, one should find centers of the dark spots and mark them on the film. Centers of light and dark spots should ideally coincide. The distance between the centers shows the magnitude of the misalignment. Another method, based on autoradiography, allows checking of the HDR dwell positions using only one HDR exposure (Anderson et al., 1995). The idea is to compare the actual dwell positions of the HDR source with predefined stationary markers associated with the empty catheter. A special plastic phantom in the form of a slab has an endobronchial catheter permanently attached to its flat external surface. Several thin metal wires are embedded in the plastic, perpendicular to the catheter, at a small distance from it. The wires are located at regular 1-cm intervals, at the expected HDR dwell positions. The phantom has a thin slot for a film parallel to the catheter and located at a small distance from it. During the HDR exposure the dwell positions are preprogrammed to coincide with the wire locations. The HDR source irradiates the film and the wires simultaneously. Direct film irradiation results in dark
192
12–Localization, Part I: Radiographic Methods and Accuracy
179
blurry spots on the film, like in the previous method. Irradiation of wires generates a stream of photoelectrons, which leaves sharp dark marks on the film. After the HDR run is finished, the film is pulled out of the phantom and processed. During the analysis, one finds centers of each dark spot indicating the center of the HDR source at the dwell positions. In this method, the center of the dark spot should coincide with the dark mark from the wire. If the HDR alignment tests are correctly performed, one could expect an accuracy of 1 mm in the source dwell positions determination. Autoradiographs can be also used for the HDR source localization in some other catheters and applicators (Jones 1989). For Nucletron applicators (Nucletron B.V., Veenendaal, The Netherlands), ready-pack X-OMAT “V” film from Kodak can be used with lead foil as a marker. Radiation-sensitive yellow Detex paper (Detex Corporation, New Braunfels, TX) instead of ready-pack film (Jones 1989). For “V”-type film used with a 370 GBq (10 Ci) 192Ir source, it is recommended to use the following dwell times: bronchus catheter and implant needle = 0.1 s, uterine tube = 0.3 s, esophageal applicator = 2 s. Autoradiographs can be also performed using GAFchromic film (International Specialty Products, Wayne, NJ) (Chiu-Tsao et al., 2004a,b). This film has several advantages: it is not light-sensitive, does not require processing, and has lower speed, which enables longer dwell times and reduces the blurring of the autoradiograph due to the exposure from the moving source.
CT Scanning Catheters and Applicators CT becomes a major imaging modality not only in teletherapy, but also in brachytherapy. Its main advantage is the ability to obtain in a matter of minutes all necessary information about the locations, form, and orientations of the “dummy” sources, catheters, and applicators. The disadvantages of the CT are: lower spatial resolution of small sources and difficulty in reconstruction in the presence of large metal objects, like applicators or even clusters of seeds. Large seed clusters, which sometimes appear after permanent prostate implants, and regular metal applicators can create star artifacts in the form of multiple straight lines intersecting at the metal object location. Such CT-image distortions make it difficult to localize the applicator and invalidate the Hounsfield numbers and dose computation in the vicinity of the metal objects. To use the CT as an imaging modality in the vicinity of small metal sources or applicators, special reconstruction algorithms should be employed to minimize the artifacts. These algorithms make it possible to image small metal seeds used for the permanent prostate implant, but bulky metal objects would still give significant artifacts. HDR 192Ir vendors had to provide special CT-compatible applicators made of solid materials with low diagnostic x-ray attenuation. Modern brachytherapy treatment-planning systems are capable of catheter reconstruction and threedimensional (3-D) dose calculation based on the CT images. The CT image set is imported into the treatment-planning system and associated with the patient. Manually digitizing “dummy” seed markers with a computer mouse, one can reconstruct the form, location, and orientation of a catheter or an applicator (Figure 3). Alternatively, some brachytherapy planning systems can reconstruct the sources, the catheters, and the applicators automatically, based on predefined Hounsfield numbers. Prostate Seed Post-Implant Localization Seed localization after the permanent prostate implant is an important application of CT-scanning in brachytherapy (Roy et al., 1993). It allows for quickly collecting all the information of the seed locations and orientations. Seeds appear as distinct white spots on the CT-images (Figure 4); however, there are two major challenges in the CT-based localization. First, some seeds in the same slice are located so close to each other that they merge into a single white spot in the image. Sometimes, it is possible to distinguish visually the “double seed” from a “single” one judging by the size of the spot; however, it is not always
180
Eugene P. Lief
Figure 3. Reconstruction of HDR catheters and patient’s anatomy by VariSource HDR planning software package from Varian Medical Systems, Inc. (Palo Alto, CA). [Courtesy of Anil Sharma, Ph.D., Long Beach Memorial Medical Center.]
Figure 4. Automatic seed finder module for a commercial implant planning software package “VariSeed v.7.1” from Varian Medical Systems, Inc. (Palo Alto, CA). Different sizes of light spots, representing seeds, make it difficult to distinguish visually between a single source and several sources clustered together, as well as sources that fully or partially lie in a given CT-slice. Note also small reconstruction artifacts in the form of lines connecting different sources. [Courtesy of Ted Jackson, Ph. D., Varian Medical Systems, Inc.]
12–Localization, Part I: Radiographic Methods and Accuracy
181
obvious. There is also a possibility of getting three or more seeds together, but it is much less probable. One of the ways of handling such clusterization is to assign unaccounted seeds to larger spots, which are likely to contain more than one. The second problem is that some seeds appear on both adjacent contiguous slices as two full-size spots. It happens because a part of such a seed lies in one slice, and the other part in the other. For 4.5-mm sources this is especially noticeable if the slice thickness is 3.75 mm or smaller. For instance, if 3-mm helical contiguous scanning is used, the double-registration could be as high as 40% (Lief, Davis, and Wallner 1998). This number could be slightly reduced by eliminating white spots of obviously smaller size that result from the small protruding of a seed into a CT-slice and partial volume effect. Another way to recognize the double-registration is to correlate the locations of the white spots in adjacent images and eliminating one of those that coincide. Unfortunately, this does not work all the time either because most seeds implanted into the prostate are tilted by a certain angle with respect to the CT-axis. In this case, an oblique seed protruding into both adjacent slices can register at two different locations there. One more way to reduce the double-registration is to use the appropriate CT-slice thickness and spacing (Lief, Davis, and Wallner 1998). With all these measures taken, the CT-scanning becomes a powerful tool for the post-procedure localization in permanent prostate implants. CT is the imaging modality recommended by the American Brachytherapy Society (ABS) (Nag et al., 2000) for post-implant dosimetric analysis. Automatic CT-based Prostate Seed Detection There have been many approaches to automatic CT-based seed localization (Brinkmann and Kline 1998; Li et al., 2001; Liu et al., 2003; Holupka et al., 2004). The main purposes of automatic localization are: to save time, to determine seed location more accurately than using the manual methods, and exclude interand intra-observer variability. The idea of the threshold method (Brinkmann and Kline 1998) was to search for 3-D connected objects in the thresholded CT-volume by rastering through the volume voxel by voxel. The method was tested on a special phantom with the seeds implanted according to a typical plan. Using the software written by the authors, about half of all seeds were detected with 0.5 mm accuracy, while for the other half the error was within 2 mm. In a multiple threshold approach (Li et al., 2001), the computer program enabled a graphic user interface allowing varying multiple thresholds for better performance. The purpose of the parameter variation was to localize single seeds and also to distinguish individual seeds in a cluster. The method was tested on two phantoms: one with 20 seeds at discrete locations, and the other with 100 seeds located close to each other. The algorithm was able to identify the seeds within 1 mm of their physical locations for discreet seed localization. It was further able, without the operator intervention, to separate seeds at close proximity to each other with an average localization error of less than 2 mm. Another approach to the automatic seed localization was based on the Hough transform (Holupka et al., 2004). The computer program based on this approach works in fully automatic regime without the operator input. The method has been tested on phantoms and clinically used for post-implant dosimetric assessment on 1000 patients. Some modern treatment-planning systems have a capability of automatic seed detection. Figure 4 shows a Seed Finder™ module in software package Variseed 7.1 (Varian Medical Systems, Inc., Palo Alto, CA).
Electronic Portal Imaging Devices (EPIDs), Fluoroscopy, and Cone-Beam CT Fluoroscopic images, available during the permanent prostate implant procedure, can be used for periodic checking of the seed location during the implant delivery. Several groups (Tubic et al., 2001; Narayanan, Cho, and Marks 2002; Todor et al., 2003; Su et al., 2004) explored fluoroscopy for the seed localization and use of multiple projections for eliminating the source overlap problem. Intra-operative
182
Eugene P. Lief
localization and dynamic dosimetric analysis provide an opportunity to compensate (during the implant delivery) for possible differences between the actual and planned seed positions. EPIDs became a much more used imaging modality in recent years, especially with the introduction of new amorphous silicon detectors. One of the capabilities of these detectors is to enable cone-beam imaging (Cho, Johnson, and Griffin 1995). This capabilities were used for intra-operative CT-scanning (Prestidge 2004), using cone-beam CT with an amorphous silicon detector in the Acuity simulator. Potential of this imaging modality includes intra-operative localization, reconstruction, and real-time dosimetry.
Source Localization for Various Brachytherapy Procedures LDR Implants Low dose rate (LDR) temporary interstitial implants are usually performed using catheters that are surgically inserted in the tumor. The treatment is performed using strings of low-dose sources separated by regular intervals. In order to plan the procedure and order sources of appropriate activities, “dummy wires” are inserted in the catheters exactly in the same positions as active wires. For the localization, two radiographs are taken at two angles, about 20° apart. Special care should be taken to obtain usable images with a clear view of sources and no overlap. Sometimes, a few gantry angles have to be checked, using fluoroscopy, before a suitable projection is chosen. After the films are taken and processed, the dummy markers on the films have to be identified and matched. After digitizing two-dimensional (2-D) coordinates of dummy sources, the 3-D reconstruction takes place, and all coordinates of the sources, which coincide with coordinates of the markers, are obtained. This information is used for planning brachytherapy treatment and choosing activities of the real sources that are located at the spots indicated by the dummy markers. Historically, many permanent prostate implant procedures were performed using fluoroscopy for visualization. These days, permanent prostate implants are usually done with ultrasound guidance that better shows the prostate borders than fluoroscopy. Although, some work has been performed to enhance the seed visibility in ultrasound imaging (Davis et al., 2003), still the accurate seed digitization requires xrays. If the imaging takes place in the operating room, it is typically performed by a C-arm fluoroscopy, which allows obtaining digital images at various angles. The images obtained can be used to reconstruct the implanted seed positions before all the planned seeds have been used. Dosimetric analysis performed in the middle of the procedure provides the dose distribution based on the actual, rather than planned, seed locations. In the case of possible deviations from the original plan, the intended positions of the remaining seeds can be recalculated before the implant delivery is resumed (Todor et al., 2003). After the implant delivery is finished, the patient undergoes a CT scan. Timing of CT varies from institution to institution. Some groups suggest that possible resolution of edema and tumor shrinkage makes dosimetry more realistic if the CT scanning is performed a few weeks after the implant. In addition to the CT, some institutions take a few more radiographs (typically, AP, lateral, and 20° from vertical for the supine setup) for the dose reconstruction in AP and lateral views (Roy et al., 1993). At many institutions the chest radiograph is also taken after the implant. The purpose of this is to check if there are any seeds that were carried away from the prostate with the bloodstream and ended up in the lungs.
HDR Brachytherapy Procedures For HDR treatments of prostate, catheter reconstruction has to be performed using a CT-set (Figure 3). Thin needles used for the catheter insertion can be imaged by either conventional or cone-beam CT (Prestidge 2004). Using dummy markers, one can locate HDR dwell positions that can be used for planning. Although the catheter positions could also be reconstructed using conventional radiographs,
12–Localization, Part I: Radiographic Methods and Accuracy
183
these radiographs would not show the prostate location. Therefore, CT use is necessary to visualize the target. The MammoSite® applicator (Proxima Therapeutics, Alpharetta, GA) is used for breast cancer treatment. The applicator consists of a spherical balloon with the diameter of 4 to 5 cm and a straight HDR catheter passing through its center. The empty balloon is inserted into a cavity in breast and filled with a diluted x-ray contrast agent (3%) and saline. There is only one HDR dwell position in the geometrical center of the balloon. Besides determination of the center position with respect to the patient’s anatomy, it is necessary to make sure that the balloon has a spherical form and is not flattened or squeezed to the shape of a football. The balloon deformation would lead to the distortion of the intended dose distribution. All these tasks can be performed by a CT-scan or a set of plain radiographs. Localization of the conventional tandem and ovoids applicator can be performed in the operation room using a fluoroscopy C-arm. “Dummy wires” are inserted in the applicator, and AP and lateral images are taken. Digital images taken with the use of the C-arm can be copied on a floppy disk or transferred over the computer network to the brachytherapy treatment-planning computer, where the positions of the dummy markers are digitized for the applicator reconstruction. Alternatively, if the applicators are CTcompatible, the CT scan with the applicator in place can be performed, so the applicator with the catheters are automatically reconstructed by the brachytherapy treatment-planning system. Instead of conventional CT, a cone-beam CT can also be used (Prestidge 2004).
Intravascular Brachytherapy Intravascular brachytherapy is used for treatment of benign disorders of coronary arteries. One of the radionuclides used for this purpose is a beta-emitter 90Sr. A small radioactive source is normally located inside a well-shielded container. If a special flexible catheter is attached to the source and a hydraulic pressure is applied using a special syringe, the source can leave the container and move along the catheter to the position. The dwell time is typically about 3 to 4 minutes, depending on the source strength and vessel diameter. After the dwell time elapses, a special manual switch is flipped, and the source returns back into the container driven by the water pressure created by the same syringe. Localization of the dwell position within a patient’s heart is critically important for the success of the procedure. The last U.S. vendor providing intravascular applicators (Novoste Corporation, Norcross, GA) suggested two methods of the source localization. The first method, which was mostly used with the catheter 5-French, involves a special dummy source in a separate container similar to the one containing the active source. After the flexible catheter is inserted to the proper depth, the “dummy” container is attached to it, and the non-active source is sent into position. The interventional cardiologist, using fluoroscopy, performs the test run for two purposes. First is to make sure that the non-active source can negotiate all the turns inside the catheter, and second, that it dwells in the right location. In 2003, Novoste introduced new types of sources used with new catheters 3.5 French. These catheters are coming with a dummy wire with a marker at the end, precisely in the same position to which the active source is supposed to go. When the catheter was initially inserted, the cardiologist would check the position of the dummy marker. If it was acceptable, the wire would be removed, and the catheter would be connected to the container with the active source for the treatment.
Other Methods of the Source Localization: In addition to the methods described above, there are other less common ways to localize radioactive sources. An electromagnetic tracking device was utilized for the intra-operative source localization and real-time treatment planning (Watanabe and Anderson 1995). The idea was to substitute x-rays for localization because of large, metal surgical retractors, which precluded radiographic exams. The
184
Eugene P. Lief
nonradiographic localization suggested was performed by an electromagnetic tracking device, which consisted of a transmitter coil, a receiver coil, and a signal-processing unit. There was an RMS (root mean square) error of only 0.7 mm in determining the distances between points 2 cm apart, thereby demonstrating the feasibility of the method. Another imaging modality used for the prostate seed localization is MRI (Dubois et al., 1997; Dubois, Bice, and Prestidge 2001). The idea behind the method is to utilize an imaging modality that can adequately show both the sources of radiation and its target, i.e., the prostate gland. Generally, MRI gives the best picture of the prostate, allowing definition of its borders. Therefore, MRI was evaluated as a tool for localizing the seed locations. It was found that MRI yields an average localization error of (1.67±0.76) mm, about the same as the CT, which had (1.59±0.64) mm. Furthermore, the difference in isodose volumes for all calculated isodose levels was not statistically significant at 95% confidence levels (Dubois et al., 1997; Dubois, Bice, and Prestidge 2001). This observation was made for both 103Pd and 125I permanent prostate implants. For the HDR source localization, besides radiographic, one can utilize mechanical and optical methods. Thus for the microSelectron HDR 192Ir unit from Nucletron, there is a mechanical fixture which allows direct measurement of the source location (Jones 1989). The device allows the 192Ir source to pass into a transparent channel and to displace a plastic marker along a calibrated scale. This device, which is easy to use, provides direct, accurate information about the machine function. Another type of the device by Varian is utilized for the same purpose. To dertermine the active cable length, the source cable is sent into a special channel with a precise ruler at the end. When the source reaches its end position in this channel, a special digital camera takes a picture of the source side-by-side with the ruler. From the picture it is easy to determine the last dwell position of the active source and to calculate the position correction, if necessary.
References Amols, H. I., and I. I. Rosen. (1981). “A three-film technique for reconstruction of radioactive seed implants.” Med Phys 8:210. Anderson, L. L., F. W. Mick, K. Zabrouski, and Y. Watanabe. (1995). “Photoelectrons facilitate autoradiography for 192 Ir remote afterloaders.” Med Phys 22(11):1759–1761. Altschuler, M. D., P. A. Findlay, and R. D. Epperson. (1983). “Rapid, accurate 3D location of multiple seeds in implant radiation treatment planning.” Phys Med Biol 28:1305–1318. Biggs, P. J., and D. M. Kelley. (1983). “Geometric reconstruction of seed implants using a three-film technique.” Med Phys 10:701–704. Brinkmann, D., and R. W. Kline. (1998). “Automated seed localization from CT datasets of the prostate.” Med Phys 25(9):1667–1672. Chiu-Tsao, S.-T., T. L. Duckworth, N. S. Patel, J. Pisch, and L. B. Harrison. (2004a). “Verification of Ir-192 near source dosimetry using GAFCHROMIC film.” Med Phys 31(2):201–207. Chiu-Tsao, S.-T., T. Duckworth, C. Zhang, N. S. Patel, C. Y. Hsiung, L. Wang, J. A. Shih, and L. B. Harrison. (2004b). “Dose response characteristics of new models of GAFCHROMIC films: dependence on densitometer light source and radiation energy.” Med Phys 31(9):2501–2508. Cho, P. S., R. H. Johnson, and T. W. Griffin. (1995). “Cone-beam CT for radiotherapy applications.” Phys Med Biol 40(11):1863–1883. Davis, B. J., R. R. Kinnick, M. Fatemi, E. P. Lief, R. A. Robb, and J. F. Greenleaf. (2003). “Measurement of the ultrasound backscatter signal from three seed types as a function of incidence angle: application to permanent prostate brachytherapy.” Int J Radiat Oncol Biol Phys 57(4):1174–1182. Dubois, D. F., W. S. Bice, and B. R. Prestidge. (2001). “CT and MRI derived source localization error in a custom prostate phantom using automated image coregistration.” Med Phys 28(11):2280–2284.
12–Localization, Part I: Radiographic Methods and Accuracy
185
Dubois, D. F., B. R. Prestidge, L. A. Hotchkiss, W. S. Bice, and J. J. Prete. (1997). “Source localization following permanent transperineal prostate interstitial brachytherapy using magnetic resonance imaging.” Int J Radiat Oncol Biol Phys 39(5):1037–1041. Holupka, E. J., P. M. Meskell, E. C. Burdette, and I. D. Kaplan. (2004). “An automatic seed finder for brachytherapy CT postplans based on the Hough transform.” Med Phys 31(9):2672–2679. Jones, C. H. “MicroSelectron HDR Source Localisation Techniques.” Proceedings of the 5th International SELECTRON Users’ Meeting 1988. The Hague – The Netherlands. R. F. Mould (ed.). Leersum, The Netherlands: Nucletron International B. V., 1989. Khan, F. The Physics of Radiation Therapy. Baltimore, MD: Williams and Wilkins, pp. 380–381, 385–387, 2003. Li, Z., I. A. Nalcacioglu, S. Ranka, S. K. Sahni, J. R. Palta, W. Tome, and S. Kim. (2001). “An algorithm for automatic, computed-tomography-based source localization after prostate implant.” Med Phys 28(7):1410–1415. Lief, E. P., B. J. Davis, and K. E. Wallner. (1998). “CT scanning for post-operation seed localization in permanent prostate implants.” Med Phys 25:A172 (Abstract). Liu, H., G. Cheng, Y. Yu, R. Brasacchio, D. Rubens, J. Strang, L. Liao, and E. Messing. (2003). “Automatic localization of implanted seeds from postimplant CT images.” Phys Med Biol 48:1191–1203. Nag, S., W. Bice, K. DeWyngaert, B. Prestidge, R. Stock, and Y. Yu. (2000). “The American Brachytherapy Society recommendations for permanent prostate brachytherapy postimplant dosimetric analysis.” Int J Radiation Oncol Biol Phys 46:221–230. Narayanan, S., P. S. Cho, and R. J. Marks II. (2002). “Fast cross-projectional algorithm for reconstruction of seeds in prostate brachytherapy.” Med Phys 29:1572–1579. Prestidge, B. P. “Early Experience with Cone-Beam CT for Brachytherapy.” Materials of 8th Annual International Conference and Workshop “New And Future Developments In Radiotherapy.” Sponsored by Wayne State University, School of Medicine, Detroit, MI. November 12–14, 2004, San Diego, CA, 2004. Rosen I. I., K. M. Khan, R. G. Lane, and C. A. Kelsey. (1982). “The effect of geometric errors in the reconstruction of Iridium-92 seed implants.” Med Phys 9:220. Rosenthal, M. S., and R. Nath. (1983). “An automatic seed identification technique for interstitial implants using three isocentric radiographs.” Med Phys 10(4):475–479. Roy, J. N., K. E. Wallner, P. J. Harrington, C. C. Ling, and L. L. Anderson. (1993). “A CT-based evaluation method for permanent implants: application to prostate.” Int J Radiat Oncol Biol Phys 26(1):163–169. Smith, V., A. S. Meigooni, and S.-T. Chiu-Tsao. “Orthogonal Techniques,” “Stereo-Shift Technique and Automatic Reconstruction Techniques,” and “Specialized Techniques” in Interstitial Brachytherapy: Physical, Biological, and Clinical Considerations. Interstitial Collaborative Working Group. New York: Raven Press, Ltd., pp. 239–252, 1990. Stock, R. G., N. N. Stone, M. F. Wesson, and J. K. DeWyngaert. (1995). “A modified technique allowing interactive ultrasound-guided three-dimensional transperineal prostate implantation.” Int J Radiat Oncol Biol Phys 32:219–225. Su, Y., B. J. Davis, M. G. Herman, and R. A. Robb. (2004). “Prostate brachytherapy seed localization by analysis of multiple projections: Identifying and addressing the seed overlap problem.” Med Phys 31(5):1277–1287. Todor, D. A., G. N. Cohen, H. I. Amols, and M. Zaider. (2002). “Operator-free, film-based 3D seed reconstruction in brachytherapy.” Phys Med Biol 47:2031–2048. Todor, D. A., M. Zaider, G. N. Cohen, M. F. Worman, and M. J. Zelefsky. (2003). “Intraoperative dynamic dosimetry for prostate implant.” Phys Med Biol 48:1153–1171. Tubic, D., A. Zaccarin, J. Pouliot, and L. Beaulieu. (2001a). “Automated seed detection and three-dimensional reconstruction. I. Seed localization from fluoroscopic images or radiographs.” Med Phys 28:2265–2271. Tubic, D., A. Zaccarin, J. Pouliot, and L. Beaulieu. (2001b). “Automated seed detection and three-dimensional reconstruction. II. Reconstruction of permanent prostate implants using simulated annealing.” Med Phys 28:2272–2279. Watanabe,Y., and L. L. Anderson. (1997). “A system for nonradiographic source localization and real-time planning of intraoperative high dose rate brachytherapy.” Med Phys 24(12):2014–2023.
Chapter 13
Localization II: Volume Imaging Techniques and Accuracy for Brachytherapy Dosimetry Jason Rownd, M.S. Medical College of Wisconsin Milwaukee, Wisconsin Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 Volume Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 Volume-based Simulation and Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 Volume Dosimetry and Plan Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 Quality Assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
Introduction Brachytherapy is inherently a three-dimensional treatment for a three-dimensional disease, and one should use as much information as available to correctly identify, localize, plan, and treat the patient. Concepts and definitions will be presented to provide a foundation for detailed discussions on volume-based brachytherapy, with specific case examples from the Medical College of Wisconsin.
Volume Imaging In two-dimensional (2-D) imaging modalities, the smallest image element is the pixel with a length and a width. In three dimensions, this concept expands to become a voxel, or volume element. Inherently, one can think of a three-dimensional (3-D) image as an array of 2-D planes collected together to describe the original object. Thus the three dimensions of a voxel can be thought of as the two dimensions of a pixel in the plane with the slice thickness, or slice separation, of adjacent planes as the third dimension. There are physical lengths that correspond to each side of an imaged voxel. The relationship between the physical length and the voxel unit can be described by its resolution, e.g., how many mm per voxel side. This is usually separated into the area (length and width) within a single imaged plane and the slice thickness, assuming no gaps in the imaged data between slices. In CT scans, each 2-D slice of the volume is mapped onto an X × Y pixel grid. This relationship is the same for all CT scanners, but the value of the resulting in-plane resolution is dependent on the field of view and the number of pixels used to image each plane. For example, with a 512 × 512 pixel array and a 50-cm diameter field of view, each imaged pixel corresponds to 0.98 mm on a side, or roughly 1 mm2/pixel resolution in the plane of the CT scan. The number of pixels and the field of view are machinedependent parameters that can and should be adjusted for the best results. The in-plane resolution places idealized limits on the ability to identify any object within the plane, specifically 1 pixel. A mistake in imaging of 1 pixel would correspond to a localization mistake in the patient of 1 mm. In addition to the in-plane resolution, there is slice separation and slice thickness. The slice thickness is a measure of how much of the real object volume is described by the imaged plane. The slice separation is a measure of how far apart each imaged plane is from an adjacent slice. Ideally, there should be no loss of information because of imaging gaps. The slice thickness should exactly correspond to the center-to-center slice separation of the 2-D views. Typical slice separations can be as small as 1 mm or as large as 1 cm. The idealized limits on voxel resolution apply when there are no intrinsic faults in the imaging system. Any number of mechanical and electronic faults can contribute to decreased resolution. The AAPM Task
187
188
Jason Rownd
Group 66 Report No. 83 (Mutic et al., 2003), AAPM Report No. 1 (Judy et al., 1977), and AAPM Report No. 39 (Lin et al., 1993) detail a list of routine quality assurance (QA) tests and phantoms that can be used to identify problems with a CT-scanning system. In addition to machine-based concerns, the patient themselves can present problems for accurate volume reconstructions. High-Z materials, such as prosthesis or even brachytherapy applicators, can generate imaging artifacts. These artifacts can all but eliminate any useful volume information (see Figure 1). If the artifacts are sufficiently severe, then it can be impossible to achieve the ideal resolution of 1 pixel or voxel. One constraint on collecting all CT scans with the best resolution possible is the amount of memory each scan uses to store this information. The more information you require from the scanned volume, the larger the resulting data file. Fortunately, with modern computing systems, this is rarely a problem. The biggest problem seems to be a limit on the number of CT slices useable by a brachytherapy treatmentplanning system, either because of software-coded limits or because of hardware-constrained planning speed. In most cases, the more images manipulated by the planning system, the slower any updates are handled. There may not be a physical limit, but there could be a practical limit if the system slowdown is too severe. Ideally, the planner should acquire only as many slices as necessary to describe the volumes of interest with a resolution suitable to identify accurately structures of interest with the volume. When magnetic resonance imaging (MRI) studies are discussed, there are additional concerns. There are still the concepts of pixel and voxel resolution for MR scans. There are also concerns about imaging artifacts, but more specifically, imaging distortions. With CT scans, the resulting images are assumed to have nearly ideal reconstruction, where artifacts merely mask the original objects with streaks of highcontrast pixels. The problems with MR scans are more subtle and geometrical in nature. The AAPM Task Group 1 report (Price et al., 1990) gives some detailed QA procedures that can reduce machine-specific problems. The basic theory behind MRI is that mobile protons inside objects can be excited by the application of magnetic fields and ultimately detected. The magnitude of the excitation and the various associated relaxation times are partially dependent on the detailed composition of the object and the applied magnetic fields. The specific variations of tissues within the human body can produce imaged responses that vary
Figure 1. Axial view of a metallic tandem and ring applicator showing common streak artifacts in the CT scan.
13–Localization II: Volume Imaging Techniques and Accuracy for Brachytherapy
189
by orders of magnitude. Thus MR images are very good at resolving tissue-type differences, one of its key reasons for being used in imaging studies. This high sensitivity can also result in imaging artifacts when strong gradients in the magnetic susceptibility are present within an imaged object, specifically at structure interfaces, i.e., bone-air, tissue-metal. Instead of the streaking artifacts one might see in CT scans, one could see distorted geometry in the imaged object. The size and shape of the imaged object can vary by several millimeters from its actual dimensions. Reducing the possibility of such artifacts and distortions involves the proper choice of the field of view to exclude such interfaces whenever possible. In brachytherapy, this can also mean choosing applicators that are closer to tissue-equivalent with respect to their MRI responses, specifically nonmetallic applicators. Another key advantage of MR imaging studies is the variability in reconstructed views. The imaged plane in MR is controlled by the application of three orthogonal magnetic fields. With proper control of the various fields, it is possible to image specifically the axial, sagittal, coronal, or even oblique planes inside the imaged object. With this advantage, MRI can be controlled to image the specific planes that best describe the particular tissue being imaged. This better visualization can lead to better localization of the necessary structures, but may not improve applicator localization. In many treatment-planning systems, only axial views may be used in the planning process. Oblique images, while useful to the physician for imaging the tumor, are often ignored by the planning system. This limitation can be partially offset by comparison of axial planning views with oblique diagnostic views at the time of planning.
Volume-based Simulation and Planning Historically, brachytherapy planning was done based upon standardized rules and tables. Before the advent of modern imaging modalities, brachytherapy dosimetry was limited to standardized plans and calculated point doses. Even now, much of brachytherapy planning is done with radiographic films and planar, 2-D dosimetry. Normal tissue point doses are based upon radiographic films with contrast injected to accentuate the approximate location of normal tissues relative to the brachytherapy applicators (see Figures 2 and 3). Applicator reconstruction is based upon x-ray markers visible on the radiographs. Current planning for external beam radiotherapy treatments relies on the volume information provided by CT and MR scans. Brachytherapy treatment planning can take advantage of the same information these volume scans provide. Accurate brachytherapy treatment planning requires correct reconstruction of the patient anatomy, applicators, and source positions. In most cases, this is a matter of effective scan techniques with the recognition of the image limitations, appropriate applicator choices for the imaging modality, and estimates of the source positions within these applicators. The basic volume concern for any imaging modality is that it correctly describes the scanned object, in this case the patient anatomy with applicators in place. The choice of the imaging modality will affect the correct identification of the target structures. CT scans can be used to identify target structures when the target tissue is sufficiently distinct or when other methods are used to identify the target, such as surgical clips or even the applicator itself. MR scans are often better at imaging gynecological tumors (see Figures 4 and 5). In either modality, reducing the imaging artifacts is necessary for proper visualization of the tissues as well as the applicators. A single, large volume image study can contain hundreds of megabytes of computer data. The number of imaged slices, and the corresponding required data space, are functions of the desired scan volume and slice separation. Imaging only the volume that is needed for treatment planning reduces the burden on the computerized planning system as well as limiting patient exposure during the CT scan. As mentioned before, the imaging resolution for CT/MR scans within an imaged cross section is approximately 1 mm per pixel, based upon the field of view and the scan dimensions. The field of view must contain the entire applicator, tumor volume, and relevant organs at risk. Because CT images may not have sufficient soft tissue contrast, additional injected contrast may be necessary to differentiate between tumor and normal
190
Jason Rownd
Figure 2. AP radiograph of a tandem and ring applicator with x-ray markers and injected contrast.
Figure 3. Lateral radiograph of the same applicator and contrast enhanced anatomy.
13–Localization II: Volume Imaging Techniques and Accuracy for Brachytherapy
191
Figure 4. Axial view of a CT-/MR-compatible tandem and ring insertion imaged by CT with clearly identified applicator and contrast enhanced tissues.
Figure 5. Axial view of the implant imaged by MR with the enhanced image of the tumor location now visible.
192
Jason Rownd
tissues, i.e., rectum and bladder (see Figure 6). Contrast injected only into the Foley balloon is not sufficient to identify bladder proximity to the brachytherapy applicators (see Figure 7). MR scans offer better tissue differentiation without the aid of additional contrast media, but are not as commonly available for brachytherapy planning. In both MR and CT scans, the applicator choice can influence the image quality. Stainless steel applicators, suitable for radiographic film-based treatment planning are poor choices for volume studies. In CT scans, these applicators can cause enough streaking and artifacts to obscure completely the applicator location as well as limit the physician’s ability to visualize patient anatomy (Figure 1). Even in MR scans, nonferromagnetic steel applicators can still distort the shape and location of tissue relative to the applicator because the magnetic susceptibility gradient is so large between the metal applicator and the soft tissue of the patient. Carbon-fiber-based or plastic applicators can safely be used in both imaging modalities to reduce imaging artifacts and distortions. The slice separation and orientation of the applicators affects the accuracy of the reconstruction of the applicator within the scanned volume. Large slice separation values mean more interpolation of applicator geometry, and in some cases, unreasonable extrapolation. When the end of an applicator must be correctly identified, slice separations of 2.5 mm means an uncertainty of 1.25 mm in applicator tip position, and 5 mm of uncertainty if the slice spacing is 1 cm. As mentioned before, artifacts can completely obscure the applicator in cross-sectional views. Additional uncertainties in the accuracy of the reconstructed applicator can also occur when the applicator insertion is oblique to the cross-sectional imaging axis. If only cross-sectional views are used to identify an otherwise continuous applicator, the reconstructed applicator may not resemble the actual applicator. Whenever possible, the best imaging planes should be used to identify the applicator sections. Figure 8 shows the result of a poor reconstruction of the ring section of a tandem and ring applicator, while Figure 9 shows a better reconstruction of the true shape of the applicator when the ring section was reconstructed
Figure 6. Axial view of a tandem and ring insertion imaged by CT with bladder and rectal/sigmoid tissues enhanced by inject contrast.
13–Localization II: Volume Imaging Techniques and Accuracy for Brachytherapy
193
Figure 7. Axial view of the Foley bulb imaged by CT. The isodose lines going through the Foley bulb are clearly less than those going through other sections of the bladder in the previous figure.
Figure 8. Axial view of the ring section of a tandem and ring applicator taken with CT next to the poorly reconstructed result.
194
Jason Rownd
Figure 9. Obliquely reconstructed view of the ring section of a tandem and ring applicator taken with CT and the resulting reconstructed applicator.
on an oblique imaged slice. In some cases, the orientation of the applicator is nearly orthogonal to the imaging access, such as the tandem section of a tandem and ovoid/ring applicator or the needles in a perineal implant. In these cases, the cross-sectional axis is a good choice for reconstruction. Adapting the scan views to visualize the applicator in the imaging studies is crucial to getting the best reconstruction of the applicator. The last part of the reconstruction is the trickiest, correctly identifying where the source positions fall within the actual applicator. In radiographic films, x-ray marker dummies spaced uniformly 1 cm apart can be used to identify a subset of source positions. In CT images, these same markers can be effectively used. The amount of distortion and artifacts produced by these markers is minimal and provide sufficient identification of the necessary source positions. Figure 10 illustrates this feature with respect to the tandem and ring applicator. The uncertainty in the source positions depends on the intrinsic pixel dimensions of the scan and the small amount of artifacts produced by the markers, as well as the proper placement of the markers. Unfortunately, the excellent tumor recognition of MR scans is offset by the lack of a suitable marker to identify possible source positions explicitly. In the case of MR-based treatment planning, other methods for identifying source positions are required. In some cases, it is possible to estimate the source positions based upon knowledge of the applicator geometry. In other cases, some other method must be used to reconstruct source positions. Image fusion of the CT scans and the MR scans is often used in external beam treatment planning and can also be used in brachytherapy treatment planning. CT scans tend to be better for identification of the applicator and source positions for reasons mentioned above, while MR scans can be better for imaging tumors relative to organs at risk. In the best of both worlds, contouring of the necessary targets can be done on the MR scans and fused onto the CT scans with the resulting treatment planning completed on the CT scans. 3-D applicators should be reconstructed by 3-D imaging studies. In single-catheter applications, such as endobronchial treatments, an orthogonal pair of radiographic films can suitably reconstruct the necessary
13–Localization II: Volume Imaging Techniques and Accuracy for Brachytherapy
195
Figure 10. Obliquely reconstructed views of a CT scan showing x-ray marker dummies inside the tandem and ring applicator used to identify possible source positions.
information for treatment planning. Vaginal cylinder applicators, with their simpler design and treatment goals, are also easily reconstructed from radiographic films with sufficient information about the organs at risk (bladder and rectum) being entered based upon injected contrast. Even in the single-catheter applicator treatments for breast brachytherapy, MammoSite® (Proxima Therapeutics, Alpharetta, GA), one could argue that all the planning information, dwell positions, and prescriptions could be obtained from films (see Figure 11). As the brachytherapy applicators get more complex and recognition of tumor and organs at risk become more necessary, accurate volume imaging and reconstruction becomes more necessary. Brachytherapy applicators can be broken up into two varieties: rigid and flexible catheters. Rigid applicators use metal, plastics, or carbon-fiber materials to maintain their shape during the procedure. The simplest rigid applicator is the hollow steel needle. Because of their metallic nature, these applicators can create imaging artifacts in volume studies. They are typically inserted into the patient with the help of a template and held in place by this template for the entire treatment. A 3-D imaging study of needle-based treatments has some benefits that overcome the inherent artifacts produced by steel needles. The main benefit is that the needles tend to run normal-to-axial image views and are easily reconstructed, providing a very accurate picture of the implanted volume. The main drawback is that the metallic needles produce image streaking that makes tissue delineation difficult (see Figure 12). Removing the additional metal of the needle stylets or obturators before taking the CT scan can minimize these artifacts. Additional contrast agents can also be used to identify the nearby rectum and bladder, while the treated volume is identified by needle locations. Another rigid applicator is the tandem and ovoids applicator used in gynecological treatments. This applicator consists of one catheter in the tandem, used to treat the cervix into the uterus and two catheters on either side in the ovoids, used to treat the vaginal surface and nearby tissues. The stainless steel variations can be imaged easily with radiographs, and CT-/MR-compatible versions exist for volume imaging. Similar to the tandem and ovoid applicator is the tandem and ring applicator. The pair of ovoids is replaced by a single ring section that rests up against the cervical os during treatment. The imaging
196
Jason Rownd
Figure 11. En face radiograph of a balloon-based treatment for breast brachytherapy showing x-ray marker dummies and contrast-filled balloon.
Figure 12. Axial view of a cervix implant using steel needles and the resulting artifacts in the CT image.
13–Localization II: Volume Imaging Techniques and Accuracy for Brachytherapy
197
concerns are nearly identical between the tandem and ovoids applicator and the tandem and ring applicator. The straight line of the tandem is easily reconstructed in both. As mentioned before, the shape and placement of the ring section of the applicator make axial views insufficient to correctly reconstruct the applicator. To a lesser extent, the ovoid angle relative to the axial view can also make it difficult to reconstruct the applicator. A simpler applicator is the tandem and cylinder, which is really just a single catheter applicator, where the center catheter of the cylinder is extended by the tandem into the uterus. The relative simplicity of this applicator makes reconstruction trivial and the imaged volumes contain minimal artifacts. Flexible applicators can also be as simple as plastic needles instead of steel needles. The plastic tips of these needles can make their insertion more difficult and the plastic needles may not stay as rigid during the treatment as the steel needles. However, the amount of artifacts produced in scans is greatly diminished. If the template used to hold the needles in places is also MR compatible, then the entire applicator can be imaged as necessary using either MR scans or CT scans. Prostate (Figure 13) and cervix implants can be done this way. Identifying source positions requires knowing the dimensions of the plastic needles and the necessary offsets to be applied to the imaged catheters. More complex implants can be handled using flexible catheters inserted with the aid of hollow needles. The hollow steel needles are removed, leaving the flexible plastic catheters behind in the tissue held in place by sutures or adjustable guides pulled against the skin. Post-resection soft tissue sarcoma beds can be irradiated using a single plane of these flexible catheters, and probably require little if any volume imaging. Orthogonal radiographs and x-ray marker dummies can be used to identify the implanted catheters and treatment region sufficiently.
Figure 13. Axial view of a prostate implant using plastic needles and the negligible artifacts in the CT image.
198
Jason Rownd
Post-lumpectomy volume implants of the breast usually require more than a single plane of catheters. The multi-planar nature of the implant, the irregular nature of the target volume, and the proximity of organs at risk, such as the skin, make imaging crucial for proper coverage and treatment.
Volume Dosimetry and Plan Evaluation Once the brachytherapy applicator, of whichever type, has been implanted into the patient and imaged for treatment planning, a suitable prescription must be developed. Many different methods of prescription are available for brachytherapy treatments. The prescription goal may simply be the dose to an individual point or group of points. It may also be based upon careful selection of the target volume and the percentage of that target covered by the prescription dose. Brachytherapy treatments of the prostate use various scans (ultrasound/MR/CT) to determine target volumes with prescriptions based upon dose volume histogram (DVH) analysis to get the prescribed dose to a specified percent of the target volume. There are two common methods for partial breast irradiation using brachytherapy: the MammoSite balloon and multicatheter volume implants. The prescription goal of the balloon implant is to deliver a desired dose to a distance of 1 cm from the balloon surface. Imaging the balloon within the breast after surgery is done to obtain the balloon diameter and tissue conformance around the balloon (see Figure 14). When using multiple catheters to implant the breast for brachytherapy treatment, CT scans are necessary to identify both the catheter positions and the target volumes for planning (see Figure 15). The prescrip-
Figure 14. Axial view of a breast brachytherapy treatment using the MammoSite® balloon imaged by CT showing exact size and tissue conformance.
13–Localization II: Volume Imaging Techniques and Accuracy for Brachytherapy
199
Figure 15. Axial view of a breast brachytherapy treatment using multiple plastic catheters imaged by CT showing catheters, lumpectomy cavity, and target volume for treatment planning.
tion goals for this type of implant can be more complex. Maximizing target coverage and minimizing normal tissue doses require accurate catheter reconstruction. This process begins with the placement of the catheters and their position relative to axial planes of the CT scan. Catheters that run perpendicular to the axial slices are easiest to reconstruct, but this is usually not the most efficient way to cover the target volume. Catheter placement exactly parallel to the axial slices of the CT scan is often easiest for the physician but provides the most difficult reconstruction since lengths of the catheters may be only partially imaged if they are placed between axial slices as imaged. Compromising by placing catheters at an oblique angle gives enough information to reconstruct the implant accurately while allowing the physician to use the patient’s anatomy, i.e., chest wall, to guide the catheter placement. The plastic catheters create minimal artifacts with the only problem being the identification of possible source locations. Observing how the x-ray marker dummies rest within the plastic catheters, one can determine where the effective endpoint of each catheter is for source loading and plan accordingly (Figure 16). In gynecological implants, e.g., tandem and ovoids applicators and vaginal cylinders, it is common to prescribe a given dose to a particular point relative to the applicator, such as point A for cervix cases, or to the surface of the applicator, such as the surface of a vaginal cylinder applicator. In these cases, it is less common to define specific target volumes and desired treatment volumes. Various weights or prescription percentages can be prescribed to different sections of the applicator, but this is usually achieved by individually modifying the prescription points on a case-by-case basis. As more of these cases are imaged and planned with CT and MR scans, more prescription options can be developed.
200
Jason Rownd
Figure 16. En face radiograph of a multiple plastic catheter breast brachytherapy simulation showing the location of the most distal source position near the end of each catheter within the support ball on the skin.
Quality Assurance Given that the scanners and their output are accurate with respect to the anatomy and applicator imaging, the final dosimetry should still be evaluated for accuracy and compliance with the prescription goals. The prescription compliance can be addressed by detailed inspection of the isodose distributions and DVHs. Additional inconsistencies can also be caught by comparing current plan results with previously treated cases or by explicit point-dose calculations using another treatment-planning system. In all treatment plans, the initial limitations on the imaging accuracy combined with the uncertainties of the applicator reconstructions and target volume delineations result in final uncertainties in the dosimetry which should be evaluated case by case.
References Judy, P.F., S. Balter, D. Bassano, E. C. Mccullough, J. T. Payne, and L. Rothenberg. (1977). “Phantoms for Performance Evaluation and Quality Assurance of CT Scanners.” AAPM Report No. 1. Chicago: AAPM, 1977. Lin, P.-J. P., T. J. Beck, C. Borras. G. Cohen, R. A. Jucius, R. J. Kriz, E. Nickoloff, L. Rothenberg, K. J. Strauss, and T. Villafana. “Specification and Acceptance Testing of Computed Tomography Scanners.” AAPM Report No. 39. New York: American Institute of Physics, 1993. Mutic, S., J.R. Palta, E. K., Butker, I. J. Das, M. S. Huq, L. D. Loo, B. J. Salter, C. H. McCollough, and J. Van Dyk. (2003). “Quality assurance for computed-tomography simulators and the computed-tomography simulation process: Report of the AAPM Radiation Therapy Committee Task Group No. 66.” Med Phys 30:2762–2792. Also available as AAPM Report No. 83. Price, R. R., L. Axel, T. Morgan, R. Newman, W. Perman, N. Schneiders, M. Selikson, M. Wood, and S. R. Thomas. (1990). “Quality assurance methods and phantoms for magnetic resonance imaging.” Med Phys 17(2):287–295. Also available as AAPM Report No. 28.
Chapter 14
Semiempirical Dose-Calculation Models in Brachytherapy Jeffrey F. Williamson, Ph.D. Department of Radiation Oncology Virginia Commonwealth University Richmond, Virginia Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 Biological Dosimetry Era: From Skin Erythema Dose to Exposure Rate (1900–1940) . . . . . . . . . 202 Classical Dosimetry Era (1940–1990) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202 Quantitative Dosimetry Era (1990–Present) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Source Strength Specification Quantities and Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Air-Kerma Strength and Absorbed Dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 Relationship Between Activity and Exposure Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 Obsolete Quantities for Specifying Source Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 Semiempirical Dose-Calculation Formalisms: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 Isotropic Point Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 Commonly Used Notation for Semi-empirical Dose Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . 209 Alternative Classical Model Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209 Relationship of Semi-empirical Dose-Calculation Model to TG-43 notation . . . . . . . . . . . . . . . 210 Application of Semiempirical Model to Practical Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211 Modeling of Source Anisotropy: The anisotropy factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211 Dose Calculation for Extended Sources: The 1-D Pathlength (Sievert Integral) Model . . . . . . . 212 Clinical Application of Semiempirical Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214 Implementation of Point-Source Models for 192Ir and 137Cs Seed Sources . . . . . . . . . . . . . . . . . . . . 214 Implementation of 1-D Pathlength Models for 137Cs Intracavitary Tubes . . . . . . . . . . . . . . . . . . . . 215 Application of 1-D Pathlength Models to Other Radionuclides . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218 Application of 1-D Pathlength Models to Internally Shielded Applicators . . . . . . . . . . . . . . . . . . . 221 Advanced Analytic Dose-Calculation Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222 Semiempirical Models in Brachytherapy Quality Assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 Example: Manual Calculation of Dwell Time for an HDR Vaginal Cuff Insertion . . . . . . . . . . . . . 224 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 Appendix: Derivation of Exposure Rate and Air-Kerma Rate Constants . . . . . . . . . . . . . . . . . 227 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Introduction Analytical and semiempirical dose-calculation models have a long and distinguished history in brachytherapy. As used in this review, a semiempirical model is a mathematical representation of the doserate distribution about a brachytherapy source characterized by (a) relatively few unknown parameters; (b) dependence of nongeometric model input parameters only on the radionuclide spectrum; and (c) methods of evaluating input parameters that do not require measured dose distributions for sources of the same type and construction as those being modeled. The most common semiempirical dose-computation models assume an idealized point source dose distribution or isotropic dose kernel, and derive dose rates around needle and tube sources by integrating the dose kernel over the extended radioactivity distributions. Such models were used almost exclusively in clinical brachytherapy through the 1980s because of the technical difficulties in performing high-resolution absorbed dose measurements around sources because of steep dose gradients, relatively low dose rates, and low-energy photons emitted by some brachytherapy sources. A form of the point-kernel model was introduced by Rolf Sievert in 1921 (Sievert 1921) which used an
201
202
Jeffrey F. Williamson
analytic formula, the Sievert integral, to perform one-dimensional (1-D) integration of the point-source dose kernel over the active length of a radium needle. Such models were quickly adopted by Quimby (1932) and Parker (Meredith 1967; Parker 1938) who incorporated this methodology into the QuimbyMemorial and Manchester implant systems, respectively. Upon the advent of computer-assisted treatment planning, Sievert integral-type models were widely used for 137Cs tubes (Krishnaswamy 1972), radium needles and tubes (Stovall and Shalek 1972), and even 125I seeds (Krishnaswamy 1979). While semiempirical models have largely been displaced in brachytherapy planning by more rigorous Monte Carlo and measurement-based methodologies, they are still useful for calculating dose-rate for 137Cs and 192Ir implants, and for manually verifying computerized dose calculations. Computation of absorbed dose and specification of source strength are tightly coupled together. To appreciate this relationship and to better understand the role of computational models in brachytherapy dose specification, it is useful to describe the history of brachytherapy in terms of three dosimetric eras: biological dosimetry (1900–1940), classical dosimetry (1940–1990), and quantitative dosimetry (1990–present).
Biological Dosimetry Era: from Skin Erythema Dose to Exposure Rate (1900–1940) During this time period, interstitial and intracavitary brachytherapy using radium needle and tubes became established as the first radiotherapeutic technique to successfully treat deep-seated tumors. By 1908, Marie Curie introduced the quantity activity and developed an approach for standardizing 226Ra source strength in terms of the mass of 226Ra contained in a source in 1913. By 1921, 226Ra standards, consisting of known masses of radium salt in glass vials, were available in Paris and Vienna for intercomparison with clinical sources. This allowed implant therapy to be specified in terms of “mg-h” (the product of mass of 226Ra in the implant and treatment duration). In the meantime, development of the free-air chamber [see Wyckoff (1980)] for an interesting historical review) allowed external beam-therapy to be specified in terms of the directly measurable quantity, exposure, by the late 1920s. A major problem was how to specify brachytherapy in terms of a comparable quantity. Some implant systems, e.g., the Manchester system, used Sievert-like dose-calculation algorithms in conjunctions with an estimated 226Ra exposure-rate constant, quantifying treatment delivery in terms of roentgens. However, rigorous physical measurements to confirm the exposure rate constant or otherwise establish the equivalence of 226Ra Roentgens (a megavoltage treatment modality) with the more rigorous external-beam exposure standard (limited to kV x-ray beams) was not possible in this era. The literature of this period documents numerous and often successful efforts to use biological dosimeters to intercompare different kilovoltage therapy beams and transfer instruments through the late 1920s and to intercompare radium treatment with orthovoltage therapy until about 1940. The most common endpoint was threshold erythema dose (TED) developed and promoted by Memorial Hospital (Quimby 1941). One unit of TED was that quantity of radiation needed to produce a detectable erythema in 80% of the subjects so irradiated. An exposure 1 TED of low dose-rate irradiation was biologically equivalent to about 11 Gy.
Classical Dosimetry Era (1940–1990) This era was accompanied by the development of mature classical systems for brachytherapy source implantation and the gradual replacement of 226Ra sources and artificially produced radionuclides such as 192Ir and 137Cs. It began with the successful application of Bragg-Gray cavity theory 22 to the calibration of 226Ra and other high-energy sources in terms of exposure (Attix and Ritz 1957), which allowed brachytherapy quantified using the same system of units and quantities as the external orthovoltage-beam therapy of the day. This development permitted brachytherapy dose delivery to be specified in terms of a rigorously defined quantity, exposure, which could, in principle, be measured by an instrument calibrated
14–Semiempirical Dose-Calculation Models in Brachytherapy
203
against a primary standard. In this chapter, we will discuss dose-calculation algorithms that seek to estimate absorbed dose in medium from sources that have been calibrated in terms of an exposure-rate or air-kerma rate standard.
Quantitative Dosimetry Era (1990–Present) The modern or quantitative era of brachytherapy dosimetry began to influence clinical practice in the 1990s and continues to the present. Quantitative dosimetry relies upon either dose-rate measurements around each specific source model or three-dimensional (3-D) radiation transport calculations based upon a validated geometric model of the source. In contrast to the semiempirical dose modeling philosophy, radionuclide-specific model parameters are broadened to include source-specific absorbed estimates derived from either measurements or from Monte Carlo simulations accounting fully for the geometry of the specific source. Quantitative dosimetry methodologies place less emphasis on uniformity of clinical dose-calculation practice in favor of physically accurate dose-rate estimates with well-defined uncertainties. The technical developments leading to clinical acceptance of quantitative dosimetry methods were motivated by concerns that semiempirical dose-calculation algorithms were not valid in the low-energy regime of 125I and 103Pd sources. To clinically utilize dose measurements and Monte Carlo calculations, a fully empirical dose-calculation formalism, the AAPM TG-43 protocol was developed and published in 1995 (Nath et al., 1995) and recently revised in 2004 (Rivard et al., 2004). Both the classical and quantitative dosimetry methods are based on the principle that brachytherapy source strength should be specified in terms of radiation output in free air. The goals of this chapter are to (1) review widely used and historical dose-calculation formalisms, (2) identify their dependence on the source-strength specification quantity used, (3) review the literature documenting the accuracy of semiempirical models, and (4) discuss their application in current clinical practice.
Source Strength Specification Quantities and Units Air-Kerma Strength and Absorbed Dose As discussed in more detail in chapter 11 of this book, photon-emitting brachytherapy sources are specified in terms of air-kerma strength, denoted by SK, in North America. The AAPM (Rivard et al., 2004) currently defines SK as the product of air-kerma rate, K" δ (d ) , in vacuo and due to photons of energy greater than ∂, at distance d, and the square of this distance, d2. SK = K" δ , air (d ) ⋅ d 2 .
(1)
The point of K" δ , air specification is located on the transverse-plane of the source (the plane normal to the longitudinal-axis of the source which bisects the radioactivity distribution). The unit of air kerma strength is 1 µGy·m2·h–1, and is often denoted in the literature by the symbol “U” where 1 U = 1 cGy·cm2·h–1 = 1 µGy·m2·h–1. Air-kerma strength is numerically (but not dimensionally) equal to the quantity reference air-kerma rate, K" ref , a very similar quantity defined by the International Commission on Radiation Units and Measurements (ICRU) (1998) and generally used outside North America. The quantity SK describes source strength in terms of kerma, Kx, which is the total kinetic energy transferred to charged particles by photon interactions with atoms per unit mass of material x. Although kerma can be specified in any medium x, usually air with (x = air), is assumed for radiation metrology and replaces the now obsolete quantity exposure, X. Absorbed dose to air, Dair, and SK are closely related:
204
Jeffrey F. Williamson
Dair = X ⋅
() W e
= SK ⋅ (1 − g ) ≈ SK ,
(2)
where (W/e) is the average energy imparted to air per ion pair created and is a constant, independent of photon energy: (W/e) = 33.97 eV/ion pair = 33.97 J/C = 0.876 cGy/R (Boutilon and Perroche-Rous 1987). The factor g is the fraction of kinetic energy transferred to the medium that is converted back to radiant energy (photons) by the Bremsstrahlung process. The quantity Kx · (1 – g) is often called “collisional kerma,” and refers to that component of the transferred energy which is ultimately absorbed by the medium via inelastic secondary charged particle collisions. Because g ≤ 0.001 at brachytherapy energies and tissuelike media, this radiative correction is usually ignored, further simplifying equation (2). Equation (2) assumes secondary charged particle equilibrium (CPE) (Attix 1986). CPE is realized when kerma remains relatively constant over the secondary electron range, assuring that the rates of energy absorption and energy transfer are approximately equal, and that kerma closely approximates absorbed dose. Virtually all brachytherapy dose-calculation algorithms and dosimetric analyses assume that CPE exists, and that D ≅ K everywhere. Although generally valid, CPE can be expected to break down in the presence of steep dose gradients very near sources (Roesch 1958), near metal-tissue interfaces (Nath, Yue, and Liu 1999), and within the active elements of thin, bounded detectors (Burlin 1966).
Activity To define the obsolete quantities for describing source output and to understand the classical formulations of semiempirical dose-calculation models, the quantity activity, A, must be introduced. It is defined as the rate of nuclear disintegration or transformation within a radioactive source. Its contemporary unit is the becquerel (1 Bq = 1 disintegration/s). We will freely use the more traditional but obsolete unit, the curie (1 Ci = 3.7 × 1010 disintegrations/s = 3.7 × 1010 Bq). Each disintegration represents the spontaneous transformation of an atom from one nuclear state to another. For most brachytherapy radionuclides, such nuclear state transformations are accompanied by emitted photons in the form of unconverted γ-rays, annihilation photons, characteristic x-rays, and Bremsstrahlung photons. Activity is measured by counting the number of photons, β particles, etc., emitted by an unencapsulated point source of the radionuclide by means of scintillation or proportional counters, from which its activity is inferred (NCRP 1985). For sealed brachytherapy sources, activity refers to that contained inside the encapsulation. As defined in this strict sense, activity is no longer used in brachytherapy dosimetry. However, activity continues to serve as the basis for treatment specification and dosimetry of unsealed radiopharmaceuticals used for diagnosis and therapy, and may play a future role in dosimetry of sealed beta-emitting sources for intravascular brachytherapy. NIST maintains activity standards for a wide variety of radionuclides in aqueous solution (Cavallo et al., 1973; Coursey et al., 1992).
Relationship Between Activity and Exposure Rate The activity of a photon-emitting radionuclide and the exposure rate in free space, X" δ (r ) , at distance r due to photons of energy greater than δ, are related by a fundamental quantity, the exposure rate constant, ( Γδ )x , defined as follows (ICRU 1998):
( Γ δ )x =
X" δ (r ) ⋅ r 2 A
.
(3)
14–Semiempirical Dose-Calculation Models in Brachytherapy
(Γ )
205
has units of R·cm2·mCi–1·h–1 and is equal to the exposure rate in R/h at 1 cm from a 1 mCi point source. It describes the rate at which air is ionized due to the emission of photons resulting from radioactive decay. The purpose of the energy cutoff δ in equations (3) and (1) is to eliminate low-energy Bremsstrahlung and characteristic x-rays that are always absorbed within any practical source. Because ( Γδ )x is defined in terms of an isotropic point source and exposure rates corrected for air attenuation and scattering, inverse-square law applies exactly. Thus, ( Γ δ )xis independent of the distance r used in equation (3). ( Γδ )x depends only on the number and energy of the photons emitted per disintegration. Suppose there are N different photons emitted per disintegration with energies E1, E2, . . ., EN in units of MeV. Each time an atom decays, suppose Pi photons of energy Ei are emitted where i = 1, . . ., N. The list {Ei , Pi }N constii =1 tutes the photon spectrum of the radionuclide. If the spectrum is known, then ( Γ δ ) can be calculated by: δ
x
x
(Γ ) δ
N
X
= 193.8 ⋅ ∑ Pi ⋅ Ei ⋅ ( µ en / ρ )iair ,
(4)
i =1
where ( µ en / ρ )iair is the mass energy absorption coefficient (in units of cm2/g) for air at energy Ei. A detailed derivation of this fundamental relationship is given in the appendix of this chapter. As explained in the appendix, the exposure-rate constant has been replaced by the air-kerma rate constant, ( Γ δ )K . As originally defined by the ICRU (1998), is a fundamental property of the unencapsulated radionuclide photon spectrum, applies only to idealized point sources, and neglects many significant properties of real sources such as self-absorption, filtration, and physical distribution.
Obsolete Quantities for Specifying Source Output Numerous quantities and units have been introduced over the decades for specifying strength of brachytherapy sources, many of which make misleading references to physical quantities having no direct connection to the primary standard from which the given calibration was derived. Confusion can be avoided by recognizing that all photon-emitting sealed brachytherapy sources are specified in terms of air-kerma strength, and that obsolete quantities are merely historically convenient means of restating SK. The oldest source-strength quantity, MRa, the mass of 226Ra contained inside a source excluding the nonradioactive core components and radioactive decay products, was introduced by Marie Curie in 1913 (Chavaudra 1995), well before the quantity activity was defined. Indeed, the unit curie was originally defined as the number of disintegrations produced by 1 g of 226Ra. MRa standards were prepared by carefully weighing pure 226Ra samples in an analytical balance. The National Institute of Science and Technology (NIST) MRa standard was introduced by Hönigschmidt in 1934 (Cavallo et al., 1973). To calibrate a user’s source in MRa, its radiation output is compared with that of the NIST radium standard by means of an ion chamber. NIST no longer offers an MRa calibration service. In contrast to the other radionuclides, exposure-rate constant of 226Ra (denoted by the special symbol (Γ δ)Ra,t in (Γ δ)Ra,t this chapter), is tabulated as a function of its effective capsule thickness, t, in mm of platinum. (Γ δ)Ra,t is normalized to the mass of radium contained in the source and has units of R·cm2·mg–1·h–1. Since MRa is limited to 226Ra sources, it is no longer used. The quantity is equivalent mass of radium (Meq), was introduced in the 1950s as a generalization of MRa, applicable to artificial radium substitutes as well as 226Ra. Meq is defined as that mass of 226Ra filtered by 0.5 mm Pt that has the same SK as that of the given source. Describing the strength of 60Co and 137Cs sources in terms of Meq allowed implant and radium needle dosimetry tables, which gave dose per mg-h of 226Ra, to be used without modification for these new sources. Because Meq is simply a statement of SK relative to that of a hypothetical radium needle, the given source being quantified need not contain 226Ra, be encapsulated in Pt, nor have a wall thickness of 0.5 mm. Since Kair = X · (W/e) and (Γ δ)Ra,0.5 = 8.25 R·cm2·mg–1·h–1 for 226Ra filtered by 0.5 mm Pt3, SK and Meq are related by:
206
Jeffrey F. Williamson
M eq =
(Γ ) δ
SK Ra , 0.5
⋅ (W / e )
=
SK
.
(5)
7.223
Meq continues to be widely used to specify strength of intracavitary and interstitial brachytherapy radium-substitute sources such as 137Cs and 192Ir. Similar to the approach of using Meq, apparent activity, Aapp, is a statement of source output relative to that of a hypothetical unfiltered point source. Aapp is the activity of a hypothetical unfiltered point source that has the same SK as that of the given source. Aapp =
SK
( Γ ) ⋅ (W / e ) δ
(6)
X
Apparent activity in units of mCi continues to be widely used for specifying strength for permanent interstitial implants (e.g., 125I and 103Pd sources). In contrast to Meq, which is based on the universally accepted (Γ δ)Ra,0.5 value of 8.25 R·cm2·mg–1·h–1, no consensus as to (Γ δ)X values for the other radionuclides exists. Often different vendors will assume different values for the same radionuclide. Thus, Aapp is an inherently ambiguous means of describing source strength. For 103Pd and 125I sources, the AAPM recommended (Williamson et al., 1999) that the (Γ δ)X values of 1.476 and 1.45 R·cm2·mg–1·h–1, respectively, be used for specification of Aapp.
Semiempirical Dose-Calculation Formalisms Isotropic Point Source Consider a hypothetical unencapsulated point source embedded in condensed medium, x, with an air· kerma strength of SK, illustrated in Figure 1. Our aim is to evaluate the dose rate, D (r), as a function of · radial distance, r. Because of spherical symmetry, D (r) has no angular dependence. Assuming CPE, equa· tion (2) implies that the dose rate to air in free space is equal to air-kerma strength: Dair (1 cm) = SK. Because
Figure 1. Unencapsulated point source of strength SK immersed in an unbounded water-equivalent medium.
14–Semiempirical Dose-Calculation Models in Brachytherapy
207
of the straight-line emission of photons in all directions (see Figure 2), photon intensity or fluence, Φ(r), at any point is proportional to the inverse square of its distance, r: Φ(r ) =
no. incident photons unit area irradiated
=
noo. photons emitted 4π r 2
∝ Dose(r ) ∝ Exposure(r ) ,
(7)
assuming that attenuation and scattering can be neglected. As a result of this purely geometric effect, the absorbed doses in free space, D(r1) and D(r2) at the two distances r1 and r2 are related by:
r = = 2 . D (r2 ) Φ(r2 ) r1
(8)
D" air (r ) = SK r 2 .
(9)
D (r1 )
2
Φ(r1 )
Letting r1 = 1 cm and r2 = r, we obtain
We now consider the effect of the tissue medium, x, on the dose distribution. To correct for difference in energy transfer efficiency between medium and air, we exploit the fundamental relationship between particle fluence and dose derived in the appendix, Dx(r) = Φ(ρ) · E · (µen,x/ρ). This relationship implies that that the dose to arbitrary x, and dose to air are related by: ,
D" x = µ en ρ D" air free space
(
) (µ ρ) ≡ (µ x
en
air
en
/ρ
)
(10)
x air
Figure 2. An isotropic point source of activity, A. To illustrate the derivation of inverse square law, the source is surrounded by vacuum and placed at the center of two concentric spherical surfaces of radii r1 and r2. By definition, an isotropic point source -Has no extension -Radiates photons with equal likelihood in all directions in straight-line paths.
208
Jeffrey F. Williamson
where the bar indicates that the corresponding quantity is the energy-fluence weighted average over of the photon spectrum. We now consider the effect of the condensed medium surrounding the source. The competition between photon attenuation and buildup of scattered photons can be described by the tissue attenuation and buildup fact, T(r), defined as
T (r ) =
Dx in medium Dx in free space
=
Exposure in mediuum at distance r Exposure in air
. r ce from a point sou
(11)
Sometimes this factor is called the “kerma-to-dose conversion factor,” depending on at which point in the derivation the CPE approximation is invoked. Substituting equations (10) and (11) into (9), and assuming that x = wat (water), we obtain the final result: ( µ / ρ )wat D" wat (r ) = SK ⋅ en 2 air ⋅ T (r ) . r
(12)
For all photon-emitting radionuclides with energies greater than 200 keV, including all radium substiwat tutes, ( µ en / ρ )air has the value 1.11 in water. Figure 3 shows T(r) for several radium-substitute radionuclides as well as for a few low-energy radionuclides. For 226Ra-equivalent radionuclides, T(r) deviates less than 5% from unity for r ≤ 5 cm. Numerous tabulations of T(r) are available in the literature; those of Meisberger and colleagues (Meisberger, Keller, and Shalek 1968), Berger (1968), and Van Kleffens and Star (1979) are among the best known. Most of these data are derived from theoretical photon
Figure 3. Photon attenuation and scatter factors, T(r), for a number of brachytherapy radionuclides. The left panel shows T(r) for a number of radium-equivalent radionuclides, as presented in the classic paper of Meisberger and Keller (1968). The right side shows T(r) for several low-energy radionuclides. Meisberger et al., fit their data to a third-degree polynomial T(r) = A + B · r + C · r2 + D · r3 that is widely used to represent T(r) in treatment planning systems.
14–Semiempirical Dose-Calculation Models in Brachytherapy
209
transport calculations. The classical semiempirical model assumes that T(r) is a function only of the radionuclide photon spectrum and that a single data set (e.g., for 192Ir) can be used for all 192Ir sources regardless of their construction. By solving equations (5) and (6) for SK and substituting the results in equation (12), one can derive equations relating the dose rate at distance r to equivalent mass of radium and apparent activity for the same unfiltered point source:
D" wat ( r ) = M eq ⋅ D" wat ( r ) = Aapp ⋅
(Γ ) δ
(Γ ) δ
Ra , 0.5
r X
r
2
⋅ fwat
2
⋅ T (r )
Equivalennt Mass of
226
Ra
(a )
(13)
⋅ fwat
⋅ T (r )
Apparent Activity
(b )
where fwat is the dose-to-exposure conversion factor given by: fwat =
Dwat X
= (W / e) ⋅ ( µ en / ρ )air = 0.876 wat
cGy R
⋅ ( µ en / ρ )air
wat
}
in free space .
(14)
For all 226Ra substitutes, fmed has the value 0.974 cGy·R–1 for water and 0.966 cGy·R–1 for muscle (Johns and Cunningham 1983). · Equations (12) and (13) give Dx (r), for a point source surrounded by water medium (easily generalized to an arbitrary medium) that has been specified in terms of Meq, Aapp, and SK. Assuming that the same exposure-rate constants, (Γ δ)Ra, 0.5 and (Γ δ)X were used to evaluated absorbed dose as were used to convert the measured air-kerma strength to Meq and Aapp via equations (5) and (6), all three equations should give numerically identical dose rates. This demonstrates that (Γδ) is, in fact, a “dummy” constant that plays no physical role in the dosimetry of output-calibrated sealed sources since any arbitrary, but consistently used, value will yield identical dose-rate distributions. These unit conversions may be performed by different individuals. For example, the vendor calibrates 125I sources by intercomparing them with the NIST SK standards. The vendor calculates Aapp from the measured SK by equation (6) using an assumed (Γ δ)X value and records the result on the source’s calibration certificate. The hospital physicist, in calculating dose rates by equation(13), must also use an assumed (Γ δ)X value. If the physicist fails to use the same value as the vendor, significant dose-calculation errors may result. Use of SK for clinical source-strength specification eliminates these dummy constants, thereby eliminating errors due to inconsistent conventional choices. For more details on the relationships between the various source-strength quantities and units and their impact on dose calculation, the reader is referred to the review by Williamson and Nath (1991).
Commonly Used Notation for Semiempirical Dose Calculation Numerous notational systems have been used in the literature to formulate equation(12). Only a few of the more common approaches will be outlined here. Alternative Classical Model Nomenclature One commonly used quantity in classical dose calculation is the energy-absorption build-up factor, B(µr). The build-up factor is defined as B( µ r ) =
total dose in medium at distance r
primary dose in medium m from a point source
,
(15)
210
Jeffrey F. Williamson
where µ [cm–1] is the total linear attenuation coefficient of the primary spectrum. The most commonly cited source of buildup factors is the classic article by Berger (1968) which is based upon a semianalytic solution of the Boltzmann transport equation. Meisberger et al. (Meisberger, Keller, and Shalek 1968) were the first to suggest that these theoretical data could be used to approximate the transverse-axis dose distribution around actual interstitial sources. For an ideal point source in water, equation (12) becomes ( µ / ρ )wat D" wat (r ) = SK ⋅ en 2 air ⋅ e− µ r ⋅ B( µ r ) , r
(16)
− µr which implies T (r ) = e ⋅ B( µ r ) . A related quantity, used in many modern analyses (Williamson 1996) is the scatter-to-primary ratio, SPR(r):
SPR(r ) ≡
scatter dose in medium at distance r
primary dose in meddium from a point source
(17)
T (r ) = e− µ r ⋅ (1 + SPR(r ))
A precursor of the radial dose function (RDF), as defined by the AAPM TG-43 report (Nath et al, 1995; Rivard et al., 2004), was introduced by Dale (1983). This function, also called RDF, is denoted by the symbol g– (r) to –distinguish it from the TG-43 quantity, gX(r). Dale defined it as g (r ) =
D (r, π / 2 ) ⋅ r 2
D (1 cm, π / 2 ) ⋅ (1 cm )
2
(18)
and allowed the possibility of using dose measurements or realistic Monte Carlo calculations to evaluate g– (r). As defined, g– (r) = gp(r). Late in the classical era, measured data were often used to evaluate g– (r), especially in the case of 125I seeds (Ling et al., 1983). More often, g– (r) was used to tabulate theoretical point-source data. Using this quantity, equation (12) becomes: wat ( µ / ρ )air ⋅ T (1 cm ) D" wat (r ) = SK ⋅ en ⋅ g (r ) 2 r
(19)
leading to the equivalence g (r ) = T (r ) T (1 cm ) . In the early 125I literature (Krishnaswamy 1978; Ling et al., 1983), the correction T(1 cm) was ignored and dose was calculated by D" wat (r ) = Aapp ⋅
(Γ ) δ
X 2
r
⋅ fwat
⋅ g (r ) .
(20)
Relationship of Semiempirical Dose-Calculation Model to AAPM TG-43 Notation Prior to the first AAPM TG-43 report published in 1995 (Nath et al., 1995), a variety of semiempirical and empirical schemes were in use, some of which are described above. To promote uniformity in dose-calculation practice, and to promote the use of measured or Monte Carlo dose-rate estimations, TG-43 endorsed a common dose-calculation formalism derived from the Interstitial Collaborative Working Group recommendations (discussed in chapter 11) (Anderson et al., 1990). The TG-43 dose-calculation protocol is a
14–Semiempirical Dose-Calculation Models in Brachytherapy
211
formalism for representing a single-source 2-D dose-rate distribution about an actual source, in a standardized tabular format. It assumes that a sparse matrix of dose-rate values has been previously measured or calculated by some other methodology (measurement or Monte Carlo). In contrast, the semiempirical dose-calculation formalism is theoretical method (admittedly a somewhat simplistic one) for calculating absorbed dose distributions. Thus in principle, classically calculated dose-rate values can be used to specify the TG-43 dosimetric ratios. For an idealized unfiltered point source, the following equivalences are easy to derive: wat Λ = ( µ en / ρ )air ⋅ T (1 cm )
(21)
gP (r ) = gL (r ) = T (r ) T (1 cm) . Let us now consider the more complex problem of relating semiempirical dose calculations to RDF · and Λ data derived from measurements or Monte Carlo estimates of dose rate, D (r,π /2), about actual sources having an effective active length L>0. If we assume that (i) the line-source geometry function accurately approximates the falloff of dose rate in free space outside the source and that (ii) the actual source spectrum in free space closely matches that assumed by equation (21), one would expect that the following approximations to apply with a high degree of accuracy. wat ( µ en / ρ )air ⋅ T (1 cm ) ⋅ GL (1 cm, π / 2 ) ≈ Λ ≡ D" (1 cm, π / 2 ) SK D" (r, π / 2 ) ⋅ GL (1 cm, π / 2 ) T (r ) T (1 cm ) ≈ gL (r ) ≡ D" (1 cm, π / 2 ) ⋅ GL (r, π / 2 )
(22)
The geometry correction factor, GL(1 cm,π/2), to the semiempirical approximation to the measured dose-rate constant can be quite significant for intracavitary tubes and other linear sources, deviating from unity by as much as 20%.
Application of Semiempirical Model to Practical Sources Modeling of Source Anisotropy: The Anisotropy Factor Despite its simplicity, the semiempirical isotropic point source model, equation (12), accurately predicts the transverse-axis dose-rate distributions of most actual radium-substitute sources. Simply by using an output quantity to calibrate the source, rather than A, the influence of its internal structure (filtration and self-absorption) has been implicitly accounted for. Had A instead of Aapp been used in equation (13)(b), then the expression for (Γ δ)X [equation (4)], would require correction for attenuation and scattering in the N radioactive core and surrounding encapsulation. Any uncertainties in {Ei , Pi }i =1 (which are large for many radionuclides) and filtration corrections would be directly degrade dose calculation accuracy. In addition, fundamental activity measurements are technically difficult for the high-intensity sources used in brachytherapy. For this reason, contained activity does not play a role in photon brachytherapy dosimetry. In contrast, equation (12) infers dose rate from a quantity measured outside the source, which is not influenced significantly influence by knowledge of the unfiltered photon spectrum. The required quantities, ( µ en / ρ )air and T(r), are ratios and are therefore largely insensitive to uncertainties in the assumed med
spectrum.
212
Jeffrey F. Williamson
Practically all brachytherapy sources are cylindrically symmetric, giving rise to anisotropic dose distributions. Additionally, some sources used in intracavitary brachytherapy have active lengths that are · comparable to typical calculation distances. Thus the dose rate, D (r, θ ), around a brachytherapy source · depends both on distance r, and polar angle, θ D (r, θ ) may deviate significantly from the transverse-axis · dose rate, D (r,π /2), predicted by equation (12) especially near the long axis of the source. In the case of implants consisting of many randomly oriented seeds with active lengths less than the minimum distance of interest, equation (12) will accurately represent the multiple seed dose distribution if an average correction for single-seed dose anisotropy is applied (Williamson 1986). This correction factor, called the anisotropy factor (1-D anisotropy function by the 2004 TG-43 update (Rivard et al., 2004), φan(r), is defined by averaging the dose at each fixed distance r with respect to solid angle, Ω:
φ an (r ) =
Average dose at r Transverse-axis dose at r
∫ =
4π
D" (r, θ ) ⋅ dΩ
4π D" (r, π / 2 )
∫ ⋅=
π
0
D" (r, θ ) ⋅ sinθ ⋅ dθ 2 ⋅ D" (r, π / 2 )
).
(23)
– Often a distance-independent average value of ƒan(r), called the “anisotropy constant,”φ an , is used. Incorporating this average correction into equation (12), leads to: wat S ⋅ ( µ en. / ρ )air D" wat (r ) = K ⋅ T (r ) ⋅ φ an r2
(24)
– Classically, φ an was evaluated by measuring relative photon fluence in air at relatively large distances (30 to 100 cm) using a NaI or GeLi scintillation detector (Ling, Anderson, and Shipley 1979; Ling et al., 1983). According to equation (24), source strength should be increased by a constant fraction ranging from 2% (192Ir seeds) (Nath et al., 1995) to 14% (103Pd seeds) to correct for polar anisotropy effects (Rivard et al., 2004). Lindsay et al. (Lindsay, Battista, and Van Dyk 2001) compared prostate implant 3-D dose distri· butions derived from the isotropic point-source model, D (r), to those derived from the full 2-D · single-source dose-calculation model, D (r, θ ). Based on voxel-by-voxel comparisons, they found that the isotropic point-source model introduced errors exceeding 10% of the D90 (see the chapter on interstitial implant dose specification) in 8% and 33% of the target volume for the Model 6711 125I and Model 200 103 Pd sources. Corbett et al. (2001) found that despite local changes to the dose distribution, inclusion of 2-D anisotropy effects did not alter the dose-volume histogram (DVH): neither the V100 nor the margin between D100 and prostate boundary were significantly altered. For volume implants consisting of parallel arrays of 192Ir seeds, Williamson (1986) reported a similar finding. Dose Calculation for Extended Sources: The 1-D Pathlength (Sievert Integral) Model Dose distributions around larger sources, such as intracavitary tubes and interstitial needles, are calculated by partitioning the extended source into a set of point sources to which corrections for distance, oblique filtration, attenuation, and scattering are separately applied. By summing these point source contributions, the dose at point P can be estimated. This class of algorithms, first described by Rolf Sievert in 1921 (Sievert 1921), is known as the 1-D pathlength model, because filtration and scatter corrections depend only on the distance between each source segment and the point of calculation. The term “Sievert integral,” which refers to a specific elliptic integral to which a 1-D integration of the point-dose kernel can be reduced, is often used to describe this class of algorithms.
14–Semiempirical Dose-Calculation Models in Brachytherapy
213
Consider the source of strength SK and contained activity A illustrated in Figure 4. The classical Sievert model approximates the cylindrical active core by a line of radioactivity positioned along its axis. The axial length of the core is called the active length, L. Oblique filtration is modeled by assuming that the capsule reduces dose by exponential attenuation using an effective filtration coefficient, µ′. The dose rate ∆D" ( x, y ) at point (x,y) from the incremental source ∆L located at angle θ is ∆D" ( x, y ) = A ⋅
∆L (Γ δ ) X ⋅ fmed ⋅ ⋅ T ( x / cos θ ) ⋅ e− µ '⋅t / cos θ , 2 L ( x / cos θ )
(25)
where is the exposure rate constant of the unfiltered source material. Since SK = A · (W/e) · (GÎ)X · e–µ¢t, equation (25) becomes: ∆D" ( x, y ) = SK ⋅
∆L L
⋅ e µ '⋅t ⋅
med ( µ en / ρ )air
( x / cos θ )2
⋅ T ( x / cos θ ) ⋅ e− µ '⋅t / cos θ .
(26)
By integrating with respect to Œ, and transforming to polar coordinates, we obtain the Sievert integral: SK ⋅ ( µ en / ρ )air ⋅ e µ ' t med
D" (r, θ ) =
L ⋅ r ⋅ cos θ
∫
θ2
θ1
e− µ ' t ⋅ssec θ ⋅T ( x ⋅ sec θ ) ⋅ dθ
(27)
The extra eº′·t term outside the integral is needed to avoid global “double correction” for filtration. Corrections for attenuation and scatter were first included in the model by Laurence (1935). The Sievert integral analytic solution was generalized to include point-kernel build-up factors by Batho and Young
Figure 4. A typical encapsulated line source, illustrating calculation of dose rate at point P at (x,y) relative to the source center by the Sievert integral method. L and t denote the active length and radial encapsulation thickness, respectively. The distances x and y are referred to as “distance away” and “distance along,” respectively in the literature.
214
Jeffrey F. Williamson
(1964), who also devised the numerical solution of (27) used on most early treatment planning computers (Young and Batho 1964).
Clinical Application of Semiempirical Models Implementation of Point-Source Models for 192Ir and 137Cs Seed Sources Since the introduction of computer-based treatment planning, the semiempirical point-source model with average correction for anisotropy, equation (24), has been used for evaluation of interstitial brachytherapy dose distributions for all types of high-energy interstitial seed sources, including 226Rn, 198Au, 60Co, 137 Cs, and 192Ir. Of these, only 192Ir seeds, available as interstitial seeds encapsulated in nylon ribbon for low dose-rate (LDR) brachytherapy and in the form of a cable-driven single-stepping source for high dose-rate (HDR) brachytherapy remain widely used in North America. A few centers continue to use 137 Cs spherical pellets in the Nucletron Selectron remote afterloading system (Grigsby, Williamson, and Perez 1992). The published literature suggests that the semiempirical point-source model is has an overall accuracy similar to that of Monte Carlo simulation for 192Ir and 137Cs seed-like (those with an active length of 1 cm or less) sources for calculation distances up to 5 cm. Most commonly, Monte Carlo simulation has been used to benchmark the semiempirical models in terms of accuracy, often using equations (22). For 137 Cs sources, Williamson (1988, 1998; Williamson and Seminoff 1987) noted that widely T(r) tabulations (Meisberger, Keller, and Shalek 1968; Berger 1968) agreed within 2% of Monte Carlo simulations, both for idealized point sources and transverse-axis dose-rate distributions for a wide range of intracavitary tube sources based on actual source geometry. One exception was a heavily-filtered 137Cs tube with a gold-wire active core (Williamson 1988), for which the large magnitude of oblique filtration resulted in breakdown of the line-source geometry function as a predictor of dose rate in free space. A more recent analysis (Perez-Calatayud 2004) demonstrates for a wide range of intracavitary brachytherapy sources and published Monte Carlo dosimetry studies that (i) geometry-function compensated dose-rate constants and RDFs are essentially independent of source design and (ii) semi-empirical approximations derived from Meisberger’s T(r) data closely approximate Monte Carlo-based dose transverse-axis dose distributions. Similar findings support use of classical semi-empirical models for 192Ir sources. Williamson (1991, 1996) noted that Monte Carlo simulation and semiempirical approximations yielded nearly identical doserate constants and RDFs for LDR 192Ir seeds. Wang and Sloboda (1998a) systematically investigated the dependence of the transverse-axis dose distribution on source geometry and filtration. They concluded that assumptions (i) and (ii) given in the paragraph above are valid within 2% to 3% for a wide range of HDR and LDR sources. However, especially for 192Ir, it is well known that RDFs are sensitive to phantom size (and hence minimum distance from tissue-air interfaces in a clinical implant) especially for distances over 5 cm (Meisberger, Keller, and Shalek 1968; Perez-Calatayud, Granero, and Ballester 2004; Williamson 1967). In summary, for 192Ir and 137Cs sources, the Meisberger T(r) data continue to provide an accurate approximation to transverse-axis dose-rate distribution provided tissue-air boundaries are 7 to 10 cm away from the source. 1-D anisotropy functions and anisotropy constants should be derived from angular doserate distributions at radial distances near 1 cm in condensed medium. Either measurements about the actual source or Monte Carlo simulations based upon an accurate 3-D geometric model of the source are acceptable. Monte Carlo-derived anisotropy functions, and other TG-43 parameters, are available for a variety of HDR (Daskalov, Loffler, and Williamson 1998; Perez-Calatayud et al. 2001; Wang and Sloboda 1998b; Williamson and Li 1995) and LDR (Ballester et al., 2004) sources.
14–Semiempirical Dose-Calculation Models in Brachytherapy
215
Implementation of 1-D Pathlength Models for 137Cs Intracavitary Tubes Most recently analyses of the 1-D pathlength model generalize equation (27) to a 3-D spatial integration over spatial volume. This formulation, first proposed by Williamson (19988), will be described here. Consider a source (Figure 5) in which the radionuclide is uniformly distributed #over the inner cylinder (active source), which has an active length, L, and a radius, S. The dose rate, D" (r ) , in cGy/h at location # r is given by:
(
SK ⋅ µ en / ρ # D" (r ) = # F (rc )
)
wat air
3 $ $ $ −2 ⋅ ∫ ρ (r ')( r − r ' ) ⋅ exp − ∑ µ j ⋅ λ ' j ⋅ [1 + SPR(λ ' 3 )] ⋅ dV V' V
j =1
(28)
$ where V (cm3) denotes space enclosed by the active source core and ρ (r ') is the relative radioactivity density (cm–3). The indices j = 1, 3 denote the media composing the active source core, the source encapsulation, and surrounding medium while λ′1, λ′2, and λ′3 denote the corresponding distances traversed by # primary photons passing from dV′ to r. The other symbols are defined as follows: SK = is the air-kerma strength of the source (in units of cGy·cm2·h-1), specifically the product of the air# # kerma rate in free space, K" air (rc ) , at the reference point rc and the square of the distance: # # 2 SK = K" air (rc )⋅ | rc | . # 2 # $ $ $ −2 F (rc ) = rc ⋅ ∫ ρ (r ') ( rc − r ' ) ⋅ exp ( − µ1 ⋅ S − µ 2 ⋅ t ) ⋅ dV ' is the self-absorption/filtration correction V for the reference air-kerma rate specification geometry (source immersed in vacuum). µj = filtration coefficients of the active source and filter media for j = 1 and 2. In all of the calculations described in this report, µj is treated as a thickness-independent parameter. For the case of the surrounding medium (assumed to be liquid water), j =3, µ3 is the average linear attenuation coefficient of water. Published tabulations often specify µj in units of mm–1. $ Usually, ρ (r ') is assumed to be constant, in which case its occurrences in the numerator and denominator of (28) cancel. A major ambiguity in implementing the 1-D pathlength model is specification of required filtration coefficients, µj. The literature describes at least four physical interpretations of the filtration coefficient: (1) The µj describe transmission of γ rays through cylindrical filters of various thicknesses, t. This is the classical experimental interpretation (Shalek and Stovall 1969) proposed by Keyser (1951) and Whyte (1955), who derived µj for 226Ra from γ-ray transmission measurements through cylindrical filters of various thicknesses in a free-air geometry (see Figure 6). (2) The µj are well approximated by the average air-kerma-weighted linear energy absorption coefficient, µen,j of the corresponding material (BIR/IPSM 1993; Krishnaswamy 1972,1978). (3) µ1 and µ2, are simply parameters of best fit, adjusted to optimize the agreement between the 1-D pathlength model predictions and an independent estimate of the dose-rate distribution. This interpretation was first introduced by Diffey (Diffey and Klevenhagen 1975) for optimizing agreement with measured dose-distribution data and by Williamson (1988) for matching Monte Carlo data. (4) The µj describe transmission in a near narrow-beam absorber geometry (Figure 6) and are given approximately by the mean linear attenuation coefficient of the material. This interpretation was introduced by Williamson (1996) to implement a model that combined the 1-D pathlength and scat-
216
Jeffrey F. Williamson
Figure 5. Simplified geometry of a cylindrically symmetric encapsulated source, illustrating the 3-D integration elements and ray tracing used to implement the 1-D pathlength algorithm. (Reproduced from Int J Radiat Oncol Biol Phys, vol. 36, “The Sievert integral revisited: Evaluation and extension to low energy brachytherapy sources,” J. F. Williamson, pp. 1239–1250. © 1996, with permission from Elsevier.)
Figure 6. Illustration of measurement geometry needed to realize the cylindrical transmission (Left: Interpretation (1) in text) and the narrow-beam contact absorber geometry (Right, Interpretation (4) in text). (Reproduced from Int J Radiat Oncol Biol Phys, vol. 36, “The Sievert integral revisited: Evaluation and extension to low energy brachytherapy sources,” J. F. Williamson, pp. 1239–1250. © 1996, with permission from Elsevier.)
14–Semiempirical Dose-Calculation Models in Brachytherapy
217
ter-separation (Williamson, Li, and Wong 1993) methods. Experimentally, µj describes transmission through a small block of filter material with a length equal to the active length and width equal to the physical diameter of the source. Figure 6 illustrates interpretations (1) and (4), while Figure 7 shows the substantial variation in filtration correction vs. thickness for these different interpretations. In the case of 137Cs γ rays filtered by stainless steel, interpretations (4), (1), and (2), yield µ values of 0.057 mm–1, 0.021 mm–1, and 0.0226 mm–1, respectively (Williamson 1988, 1998). Many commercial implementations of the 1-D pathlength model for interstitial needle or intracavitary tube sources impose various geometric approximation on the user, as illustrated by Figure 8 (right panel). The classical Sievert model was a 1-D line integral that used analytical approximations rather than exact ray tracing through an accurate 3-D geometric model of the source. A common approximation is a ceramic or steel line source embedded in a steel cylinder of diameter 2t. The dosimetric accuracy of the 1-D pathlength model and the optimal choice of filtration coefficients has been documented in a series of published papers by the senior author (Williamson 1988, 1998; Williamson and Seminoff 1987). For the two most commonly used intracavitary tubes (see Figure 8, left panel), 1-D pathlength model dose predictions were tested against more rigorous Monte Carlo-based dose calculations over the range of 0.25 cm to 7 cm away-and-along distances (Williamson 1998). Variants of the 1-D pathlength model tested included the µj ª µen,j approximation (interpretation (2)); treating µj as a fitting parameter (interpretation (3)); 3-D numerical integration over an accurate 3-D model of the source; and 1-D numerical integration over a simplified filtered line-source approximation to the source geometry. The results are shown in Table 1 and graphically in Figure 9 (Amersham, CDCS.J source) and Figure
Figure 7. Transmission, defined as the ratio of air-kerma from 192Ir g-rays in free space with a steel absorber in place to that emitted by an HDR brachytherapy source alone as a function of absorber thickness. Air-kerma is specified 25 cm from the source center on its transverse axis. Cylindrical and contact absorber transmission were calculated by Monte Carlo simulation, along with thickness-dependent air-kerma weighted mean linear attenuation and energy-absorption coefficients. (Reproduced from Int J Radiat Oncol Biol Phys, vol. 36, “The Sievert integral revisited: Evaluation and extension to low energy brachytherapy sources,” J. F. Williamson, pp. 1239–1250. © 1996, with permission from Elsevier.)
218
Jeffrey F. Williamson
Figure 8. Left: Drawings of the Amersham Model CSCS.J source (1992 version) and the 3M 6500/6D6C design. The active core of the former consists of 9 1.45 mm diameter borosilicate glass beads while the latter consists of zirconium phosphate glass microspheres. The packed array of microspheres was simulated as a solid cylindrical core of reduced density. Both sources are encapsulated in stainless steel. More details are given in Williamson (1998), consisting of a uniform glass cylinder placed symmetrically between source tips, as described in the vendor’s literature. Right: Illustration of three different geometric approximations in modeling the CDCS.J source dose distribution via the 1-D pathlength model.
10 (3M Model 6500 source). Generally, when the full 3-D geometry of the source (or at least a double concentric cylinder approximation) are used and best fit values of µj for steel and ceramic employed, the root mean-square (RMS) error is less than 1% and the maximum error no more than 4%. When the µj ≈ µen,j approximation along with a 3-D geometric model, the RMS error increases to about 3% with maximum errors of about 8% localized near the longitudinal axis of the source (see Figures 9 and 10). When the source geometry is approximated by a line-source on the axis of a cylindrical filter, this approximation dominates dose prediction error, so that regardless of the choice of filtration coefficients, RMS and maximum errors of 3% and 10%, respectively, are observed. In summary, the 3-D generalization of the classical Sievert-like 1-D pathlength dose-calculation algorithm estimates the 2-D dose-rate distribution about 137Cs steel-clad sources with low density cores, provided that the best filtration coefficients reproduced in Table 2 are used. This methodology can be used with confidence in place of Monte Carlo simulation or measurement with this class of sources. The author recommends that new sources be examined by means of both autoradiography and contact transmission radiography to verify the source geometry. Before relying upon a Sievert or 1-D pathlength algorithm clinically, it is recommended that the physicist first reproduce one of the published tables (Williamson 1998) to assess the adequacy of the algorithm implementation and to ensure that the input parameters are properly interpreted.
Application of 1-D Pathlength Models to Other Radionuclides Application of analytic models to predict dose distributions about sources that emit lower-energy photon spectra than 137Cs, is a topic of significant interest. Recent papers have focused on HDR 192Ir sources, of which several models are commercially available. Williamson (1996) evaluated the accuracy of the 1-D pathlength model, equation(28), against Monte Carlo simulation benchmarks for two 192Ir source models, a 169Yb seed, and the Model 6702 125I source for each of the four filtration coefficient evaluation methods
14–Semiempirical Dose-Calculation Models in Brachytherapy
219
Figure 9. Spatial distribution of fractional deviations of the 1-D pathlength model calculations from Monte Carlo dose-rate calculations for the Amersham CDCS.J source (1992 version, measured geometry). (a) 3-D geometry, approximating filtration coefficients by men. (b) 3-D geometry, approximating filtration coefficients by best fit values. (c) Ceramic line source model approximating filtration coefficients by men. (d) Ceramic line source model approximating filtration coefficients by best-fit values. (Reprinted from Int J Radiat Oncol Biol Phys, vol. 41. “Monte Carlo-based dose-rate tables for the Amersham CDCS.J and 3M model 6500 137Cs tubes,” J. F. Williamson, pp. 959–70. © 1998, with permission from Elsevier.)
described above. For the Nucletron “Classic” HDR 192Ir source (see Figure 11), the “best fit” and µj ª µen,j methods yielded RMS errors of 3.2% and 6.9%, respectively, and maximum errors as large as 20%. The best-fit filtration coefficients were found to vary significantly with source construction. Similar results for 192 Ir sources were found by other investigators (Baltas 1998; Cho and Muller-Runkel 1997). For 169Yb and 125 I sources, Williamson (1996) found RMS dose-prediction errors of 8% to 17%, and maximum errors as large as 46%, for conventional 1-D pathlength algorithm implementations.
220
Jeffrey F. Williamson
Figure 10. Dose-rate as a function of polar angle at a fixed 2 cm distance from the active core center of a 3M Model 6500 source. Data derived from MCPT simulation as well as several implementations of the 1-D pathlength model are shown. (Reprinted from Int J Radiat Oncol Biol Phys, vol. 41. “Monte Carlo-based dose-rate tables for the Amersham CDCS.J and 3M model 6500 137Cs tubes,” J. F. Williamson, pp. 959–70. © 1998, with permission from Elsevier.)
Table 1: Accuracy of various 1D pathlength dose-calculation algorithms, relative to Monte Carlo simulation, for the 3M Model 6500 and Amersham Model CDCS.J intracavitary 137Cs tubes Source
Model
%RMS Error
%Error Range
CDCS.J (1992 Model)
3D (best fit) 3D (µen) Cylindrical core (best fit) Ceramic line (best fit) Ceramic line (µen)
0.7% 2.8% 0.9% 2.7% 3.0%
–1.7% to 4.2% 0.3% to 8.1% –1.4% to 4.0% –16.0% to 5.2% –5.4% to 10.1%
3D (best fit) 3D (µen) Ceramic line (best fit) Ceramic line (µen)
0.7% 2.8% 3.2% 2.7%
–1.8% to 3.6% –1.2% to 7.6% –14.4% to 3.7% –5.2% to 9.2%
3M 6500
Clearly, the classical Sievert/1-D pathlength algorithms fail to deliver dose-prediction accuracy sufficient for clinical needs for radionuclides with mean photon energies below that of 137Cs. The mechanism of these errors is well known (Williamson 1988): the 1-D pathlength algorithm assumes that both primary and scattered-photon dose are perturbed by a factor that depends only on the thickness of source and filtration material traversed by primary photons. This approximation can be valid only if the scatter-toprimary ratio, SPR(r), is independent of polar angle. Even for 137Cs this assumption is false—because
14–Semiempirical Dose-Calculation Models in Brachytherapy
221
Table 2: Optimal filtration coefficients for steel-clad ceramic-core 137Cs Tubes Source Material
Linear energy absorption coefficient
Best fit filtration coefficient
Stainless steel - 3M6500 - CDCS-J
0.0226 mm–1 0.0226 mm–1
0.036 mm–1 0.037 mm–1
Source Core - 3M6500 (cylinder: 2.22 g/cc) - CDCS-J (spherical: 2.60 g/cc) - CDCS-J (Cylindrical: 1.73 g/cc)
0.0065 mm–1 0.0076 mm–1 0.0051 mm–1
0.0070 mm–1 (0.032 cm2 g–1) ————— 0.0046 mm–1 (0.027 cm2·g–1)
scattered-photon dose is approximately isotropically distributed about cylindrical sources (Williamson 1988, 1990), the SPR at fixed distance rapidly increases as the polar angle approaches zero. Based on the insight, that scatter dose is approximately isotropically distributed, Williamson (1996) introduced a revised 1-D pathlength model, called the “isotropic scattering model,” which is based upon separating the scattered and primary dose components. D" ( r, θ ) = D" pri (r, θ ) + D" sca (r, π / 2 ) ⋅ C ( r, θ )
(29)
= D" pri (r, θ ) + D" pri (r, π / 2 ) ⋅ C ( r, θ ) ⋅ SPR(r ) · The primary dose rate, D pri (r, θ ), is estimated by the conventional 1-D pathlength model, equation (28), using attenuation coefficients that approximate narrow beam attenuation, interpretation (4) from above. C(r,θ) is an empirical correction that accounts for deviations of the scatter-dose distribution from isotropy, introduced by Karaiskos et al. (2000) to improve the predictive accuracy of the model: Williamson assumed C(r,θ) = 1. As illustrated by Figure 11, the isotropic scattering model improved the predictive accuracy slightly for the HDR 192Ir source, yielding RMS and maximum errors of 2.9% and 8.9%, respectively. An advantage of equation (29) is that its input parameters are independent of the source design and in principle, can be specified without knowledge of the source dose distribution. The isotropic scattering model significantly improved predictive accuracy for lower energy sources as well, producing an impressive 6.6% RMS error (14% maximum error) for the Model 6702 seed. As noted, Karaiskos et al. (2000) improved the isotropic scattering model, for 192Ir sources only, by simplifying evaluation of the narrow-beam filtration coefficients and introducing a source-geometry independent scatter anisotropy correction, C(r,θ) = C(θ) = 1–(3×109) · (θ–90°). With these improvements, these authors obtained an RMS and maximum errors of 1.5% and 6%, respectively, for the Nucletron and VariSource HDR sources as well as LDR seeds and wires (Karaiskos et al., 2000; Pantelis et al., 2002). In summary, classical Sievert-like models should not be applied clinically to 192Ir or lower energy sources. The isotropic scattering model, as formulated by Karaiskos et al. (2000), is an alternative at least for 192Ir sources. However, this model is not available on commercial planning systems and should be used cautiously outside the domain in which it has been validated.
Application of 1-D Pathlength Models to Internally Shielded Applicators Internally shielded colpostats are widely used in intracavitary brachytherapy to spare bladder and rectal tissue anterior and posterior to the vaginal vault. Typically, such applicators contain 2 to 5 mm of lead or
222
Jeffrey F. Williamson
tungsten-alloy shielding in the medial aspects of the anterior and posterior applicator surfaces and can be produce dose reductions as large as 50% relative to the unshielded source dose distribution (Williamson 1990). Most commercially available planning systems neglect internal applicator shielding, resulting in systematic overestimates of bladder and rectal doses. The 1-D pathlength algorithm was first generalized to accommodate applicator shielding by Meertens and Van der Laarse (1985). The algorithm ray-tracing capability must be extended to the applicator materials. In clinical applications, this means the treatmentplanning system must have the capability of representing the applicator geometry and reconstructing its 3-D dose distributions in patients from planar radiographs or CT images (Lerma and Williamson 2002). Effective attenuation coefficients are then evaluated by optimizing the agreement between model predictions and measurements (or Monte Carlo calculations) about the shielded applicator. Other investigators who have contributed to these developments include Williamson (1990), who developed a scatter separation model for estimating dose around shielded applicators and Weeks (weeks 1998; Weeks and Dennett 1990), who developed a method of extracting optimal filtration coefficient values from Monte Carlo simulations of the applicator geometry. An example of the excellent agreement that can be achieved for 137 Cs-bearing shielded colpostats is shown in Figure 12. Few publications have assessed the accuracy of 1-D pathlength algorithms for predicting dose distributions about shielded applicators for lower energy sources. Lymperopoulou et al. (2004) used Monte Carlo simulation to evaluate the accuracy of the Nucletron/Plato implementation of the 1-D pathlength algorithm for a segmentally shielded HDR 192Ir vaginal/rectal applicator. The 1-D pathlength algorithm yielded acceptable accuracy at the prescription distance, but had dose-prediction errors as large as 15% at larger distances. Based on the high scatter-toprimary ratio of lower energy sources, and the substantial variation of attenuation corrections on shield cross-sectional area and location (Williamson et al., 1993), one would expect accuracy and generality of 1-D pathlength heterogeneity corrections to be quite limited.
Advanced Analytic Dose Calculation Algorithms A number of more sophisticated and general dose-calculation algorithms have been described in the literature. One promising approach is the 2-D scatter-integration algorithm developed by Kirov and Williamson
Figure 12. Dose rate (cGy/h) predicted by Weeks’ implementation (Weeks 1998) of the 1-D pathlength model (broken line) and as estimated by MCNP Monte Carlo calculations (solid line) for a CT-compatible Fletcher-Suit colpostats. The left and right panels illustrate good agreement between the two methods in planes normal to the source axis 2 cm anterior and posterior, respectively, of the applicator center. (Reprinted with permission from Med Phys, vol. 25, “Monte Carlo dose calculations for a new ovoid shield system for carcinoma of the uterine cervix,” K. J. Weeks, pp. 2288–2292. © 1998, with permission of AAPM.)
14–Semiempirical Dose-Calculation Models in Brachytherapy
223
(1997). Like the scatter subtraction method (Williamson, Li, and Wong 1993) of which it is a generalization, it is based upon a 2-D SPR table, SPR(r,θ), giving the scatter-to-primary ratio along the axis of an isotropic point-source collimated to a cone of half angle θ. By performing a Clarkson-like 2-D scatter integration, these authors were able to predict the 2-D dose distributions about HDR 192Ir and Model 6702 125I sources with maximum errors of 5%. This is equivalent to Karaiskos’ results (Karaislos et al., 2000) but without having to use an empirical scatter anisotropy correction. In principle, this method could be extended to more complex applicator and tissue heterogeneity problems. A more general approach is the superposition/convolution algorithm (Carlson and Ahnesjo 2000; Tedgren and ahnesjo 2003; Williamson, Baker, and Li 1991). In analogy to its external-beam counterpart, the superposition algorithm defines a 2-D scatter dose array that gives absorbed dose at any point from a colliding primary photon, in a coordinate system centered on the primary collision site and oriented along the primary photon trajectory. This dose-spread array is convolved with the primary TERMA distribution to give an estimate of the total scatter dose. Various methods have been proposed for scaling the scatterdose kernal to account for heterogeneities, to correct for the severe kernel tilt characteristic of brachytherapy, and to manage singularities in the convolution integration domain. Preliminary results suggest superposition is quite accurate but computationally complex. Another and more general approach to brachytherapy dose computation is use of deterministic solutions of the Boltzmann transport equation. Daskalov et al. (2000, 2002) have proposed using discrete ordinates solutions while Zhou et al. have suggested integral solutions of the transport equation (Zhou and Inanc 2003). These solutions approach the complexity, rigor, and computational efficiency of Monte Carlo simulation. Daskalov et al. (2002) have shown that single-processor 2-D discrete ordinates calculations are 1 to 2 orders of magnitude faster than EGS4 calculations of equivalent dimensionality and that single-source 3-D calculations using the PARTISN discrete ordinates code maybe as much as an order of magnitude faster than PTRAN Monte Carlo calculations (Williamson et al., 2001). One advantage of Monte Carlo calculations is that calculations can be cost-effectively parallelized using PC clusters with distributed memory while parallelizing discrete ordinates calculations requires much more expensive shared-memory multiple processor hardware.
Semiempirical Models in Brachytherapy Quality Assurance In addition to supporting computer-assisted treatment planning, semiempirical dose-calculation tools have significant value as a manual calculation aid, either for directly estimating treatment times for certain types of clinical implants or as tools for verifying the accuracy of computerized dose calculations. A particularly useful form of the classical Sievert equation (27) can be obtained by setting encapsulation thickness to zero (t = 0) and tissue attenuation-scatter to unity. Then, equation (27) reduces to a simple closed-form analytic expression: med ∆θ D" ( x, y ) = SK ⋅ ( µ en / ρ )air ⋅ L⋅x
(30)
where ∆θ is the angle, in radians, subtended by the active length, L, with respect to the point of interest (see Figure 13). When the interest point lies on the transverse axis (y = 0), then ∆θ = 2·tan–1 (L/2x), where tan–1 means inverse tangent or arctangent. Angles must be specified in units of radians rather than degrees. This approximation is extremely useful as a manual calculation aid and is highly accurate near the transverse axis of lightly encapsulated 137Cs sources or linear arrays of 192Ir seeds.
224
Jeffrey F. Williamson
Figure 13. Error in the isotropic point source-model relative to the line source model equation (30) as a function of transverse-axis distance expressed in multiples of active length.
Figure 13 shows that as the distance r = ( x 2 + y 2 ) becomes large in relation to active length, L, equation (30) reduces to the point-source equation (12), with T(r) = 1.0. For distances less than L (~1.5 cm for intracavitary tubes), use of equation (12) will yield errors of at least 10%. For distances greater than 1.5 L (2.0 to 2.5 cm for gynecologic tubes), the point-source approximation is accurate within 5%.
Example: Manual Calculation of Dwell Time for an HDR Vaginal Cuff Insertion Problem: Small (2 cm diameter) Fletcher colpostats are used to deliver 600 cGy to the vaginal apex at depth of 5 mm via a HDR 192Ir insertion as illustrated by Figure 14. Assume that the colpostats centers are separated by 2.5 cm and that the source SK = 2.50 cGy ∑m2 ∑h–1 at the time of treatment. Further assume that three dwell positions are activated uniformly and that the spacing is 5 mm. Calculate the dwell time/position needed to deliver the prescribed dose. We approach this problem in a stepwise fashion: · 1. Estimate the dose-rate, D (1 cm,π/2), at 1 cm distance. Using equation (22) and noting T(1 cm)·GL(1 cm,p/2)ª1.0, we obtain
(31) wat D" (1 cm,π /2) = SK Λ = SK ⋅ ( µ en / ρ )air 2 2. Adapt problem= to (30). ⋅ h -1 ⋅ 2.5equation cGy ⋅ cm
10 4 cm 2 /m 2 2 -1 3600 s/h ⋅ 1.11 = 7.71 cGy ⋅ cm ⋅ s
14–Semiempirical Dose-Calculation Models in Brachytherapy
225
Figure 14. Left: Geometry for estimating dwell time needed to deliver specified dose to the vaginal apex using HDR intracavitary brachytherapy Fletcher Colpostats. Right: Method for evaluating the active length, L, the gives a dose distribution equivalent to that of three equally spaced dwell positions.
As illustrated by Figure 14 (right panel) and discussed in more detail elsewhere (Williamson and Brenner 2003), we see that each colpostat loading, consisting of three dwell positions, can be approximated by a line source with active length, L = 1.5 cm. The perpendicular distance, r, between each linear source center and the apex prescription point can be calculated by r = 1.5 2 + (2.5 / 2 )2 = 1.95 cm . 3. Set up equation (30).
D (r, π / 2 ) = T ⋅ SK ⋅ ( µ en / ρ )air ⋅ med
= 6 ⋅ t ⋅ D" (r, π / 2 ) ⋅
∆θ L⋅r
(32)
−1
2 ⋅ tan ( L / 2 r ) L⋅r
where total dwell time T = 6·t. 4. Substitute numerical values into equation (32) and solve for t.
−1
600 cGy = 6 ⋅ t ⋅ 7.71 ⋅
Conclusion
2 ⋅ tan ( 0.75 / 1.95 )
(33)
1.5 ⋅ 1.95 2 ⋅ 0.3672 = 11.62 ⋅ t = t ⋅ 46.3 ⋅ 2.925 ⇒ t = 51.6 s/positiion
As the reader will note, semiempirical dose-calculation methodologies have had a long and venerable history in clinical brachytherapy. For high-energy photon-emitting seeds and linear 137Cs sources, properly implemented semiempirical dose-calculation algorithms have a predictive accuracy comparable to Monte Carlo simulation and produce far more consistent and precise results than direct measurement. Straightforward extensions of these models enable 2-D dose distributions to be accurately evaluated about
226 192
Jeffrey F. Williamson
Ir sources and shielded 137Cs-bearing applicators. Given the growing availability of Monte Carlo simulation, currently as a source of single-source TG-43 dose arrays and in the future as a patient-specific planning tool, what is the future of such algorithms? The author believes that semiempirical algorithms will continue to play an important role in LDR 137Cs-based brachytherapy and a role that deserves more consideration in HDR/LDR 192Ir-based brachytherapy for generating reference-quality single source dose arrays, whether they be packaged in the TG-43 formalism or in terms of classical dosimetry quantities. While Monte Carlo simulation can and should be used for this purpose, developing and publishing such tables is somewhat costly. Neither physicists nor journals have much motivation for pursuing such scientifically unrewarding work. Thus semiempirical dose calculations are cost-effective alternatives for locally generating dose tables using widely available software. Will Monte Carlo-based treatment planning make precalculated single-source arrays obsolete regardless of their origin? Several groups are developing accelerated Monte Carlo codes able to compute 3-D dose distributions in 23 mm3 voxels with computing times of 2 to 20 minutes on single PC or workstation processors. Thus, in principle, the answer is “yes”? However, with rare exceptions, planning system vendors have rarely made state-of-the-art brachytherapy software a priority. With the declining incidence of invasive cervix cancer in the developed world, it is by no means assured that commercially available Monte Carlo-based brachytherapy planning tools will be available for intracavitary brachytherapy in the foreseeable future. One important exception is low-energy seed implants for low-risk prostate cancer. As the number of permanent seed implants is continuing to rise worldwide, development of sophisticated but specialized image-guided intraoperative planning software packages continues. Incorporating Monte Carlo dose-calculation engines into such software packages is a logical step that could dramatically improve the accuracy of clinical dose estimation by rigorously accounting for tissue heterogeneities and interseed attenuation. However, before such dose-calculation tools can be effectively used, a number of problems must be solved, including extraction of seed orientation as well as location from CT images, suppression of streaking and other CT imaging artifacts, development of methods for assessing composition and density of soft-tissue heterogeneities. Thus, the author believes that for some time, classical dose-calculation methodologies will continue to play important roles in brachytherapy, both as planning and QA tools and as sources of reference-quality planning data in specific settings.
14–Semiempirical Dose-Calculation Models in Brachytherapy
227
Appendix : Derivation of Exposure Rate and Air-Kerma Rate Constants [Note: This appendix is adapted from Williamson and Brenner (2003) by kind permission of the publishers.]
The derivation of the exposure-rate constant, which relates contained activity of a point source to exposure rate given the photon spectrum, illustrates several fundamental dosimetry concepts. Consider an isotropic point source (Figure A1) of activity A (in mCi) in a vacuum surrounded by a thin spherical shell of radius r and thickness ∆r. The shell is assumed to contain air of density, ρ[g/cm2] where, by convention, the brackets specify the unit of preceding quantity. Further assume that the source emits P photons of energy E (in MeV) per disintegration. We first derive the fundamental relationship between photon fluence [defined by equation (7) in the text] at distance r, Φ(r) and absorbed dose, D(r). We assume that Φ (r) is known. By definition, the linear energy absorption coefficient, µen is given by:
µ en [cm -1 ] ≡
∆Eab Einc
⋅
1 ∆r
(A1)
where Einc is the photon energy incident upon the spherical surface of radius r and ∆Eab is the energy absorbed by the air in the spherical shell of thickness, ∆r. Since the source is isotropic, Einc(r) = 4πr2· Φ (r)·E. Substituting equation (2) and the expression for Einc(r) into the definition of absorbed dose, we obtain D (r ) ≡
∆Eab ∆m
=
Einc (r ) ⋅ µ en ⋅ ∆r 4π ⋅ r 2 ⋅ ∆r ⋅ ρ
= Φ(r ) ⋅ E ⋅ ( µ en ρ )
(A2)
where ∆m is the mass of the shell and (µen/ρ) is the mass energy absorption coefficient [cm2/g] for air medium evaluated at energy E. Note that all information regarding the density, thickness or geometry of the shell cancels out. As an aside, equation (3) describes the fundamental link between dosimetric quantities and more general radiation field descriptions. For radiation fields in which the photon trajectories # # have arbitrary angular and energy distributions as described by the angular fluence, Φ(r , Ω, E ) , equation (3) becomes # D (r ) =
∞
∫
# # # # Φ r , Ω , E ⋅ E ⋅ µ ρ r , E ⋅ d Ω ⋅ dE . ( ) ( ) ( ) en ∫
(A3)
0 4π
Returning to the problem at hand, the steps outlined in Figure A1 are followed to estimate exposure rate at distance r: " r ): Step 1: Calculate the particle flux (fluence rate), Φ( " r ) [ cm -2 ⋅ h -1 ] = A [ mCi] ⋅ P [ photons/dis] ⋅ 3.7 × 10 7 dis/s ⋅ 3600 s/h Φ( mCi 4π r 2
(A4)
" (r ) = 1.05997 × 1010 ⋅ A ⋅ P Φ r2 where r has units of cm. Step 2: Calculate the dose rate, D" air (r ) from the fundamental dose-fluence relationship, equation (A2).
228
Jeffrey F. Williamson -1 -1 " (r ) ⋅ E ⋅ ( µ ρ ) D" air ( r ) [ J ⋅ kg ⋅ h ] = Φ en
= 1.05997 × 10 ⋅ 10
A⋅P r
2
[ cm ⋅ h ] ⋅ E [ Mev] ⋅ ( µ en ρ ) [ cm /g ] 2
-1
× 1.602 × 10 =1.698 ⋅
A⋅P⋅E r
2
-13
2
[ J/MeV] ⋅ 1000 [ g/kg ]
(A5)
⋅ ( µ en ρ )
where “J” denotes the joule, the unit of energy. Step 3: Convert energy absorbed in air to ionization created and apply the definition of exposure. The energy imparted to air medium per unit ionization of like sign created is related by a constant, (W/e), which has a value 33.97 J/C. By definition, one unit of exposure = 1 R = 2.58×10–5 C/kg. Thus the exposure rate at distance r, Xδ (r ) , is: X" δ ( r ) [ R/h ] = 1.698 ⋅
A ⋅ P ⋅ E ⋅ ( µ en / ρ ) air
= 193.8 ⋅
r
2
1R J ⋅ 1C ⋅ kg ⋅ h 33.97 J 2.58 × 10 −4 C/kg
A ⋅ P ⋅ E ⋅ ( µ en / ρ ) air r
(A6)
2
2 Applying the definition of exposure rate constant (Γ δ ) X = X" δ (r ) ⋅ r / A , we finally obtain:
( Γ δ ) X = 193.8 ⋅ P ⋅ E ⋅ ( µ en / ρ )air
(A7)
Figure A1: Calculation of absorbed dose rate to air at distance r about an unencapsulated isotropic point source and air air-kerma rate constant,
14–Semiempirical Dose-Calculation Models in Brachytherapy
229
Generalizing this expression to a photon-emitting radionuclide of N different energies E1, E2,..., EN with corresponding emission rates (photons emitted per disintegration), P1, P2,..., PN, we obtain the final result: N
( Γ δ ) X = 193.8 ⋅
∑ P ⋅ E ⋅ (µ i
i
en
/ ρ )air , i
(A8)
i =1
where E has units of MeV and (Γ δ)X has units of R·cm2·mCi-1·h-1. Because the International System (SI) of units has replaced the quantity exposure with the quantity kerma, K, the ICRU has introduced the air-kerma rate constant, (Γ δ)K, to replace the exposure rate constant. Since K" δ (r ) = D" δ (r ) , we obtain from equation (A5)
(Γ δ )K =
K" δ (r ) ⋅ r 2 A
= 1.274 ⋅ 10
−15
N
⋅
∑ P ⋅ E [MeV] ⋅ (µ i
i
en
/ ρ )air , i [ cm 2 /g ]
(A9)
i =1
where (Γδ)K has units of Gy·m2·Bq-1·s-1.
References Anderson, L. L., R. Nath, K. A. Weaver et al. (Interstitial Collaborative Working Group). Interstitial Brachytherapy: Physical, Biological and Clinical Considerations. New York: Raven, 1990. Attix, F. H. Introduction to Radiological Physics and Radiation Dosimetry. New York: John Wiley and Sons, 1986. Attix, F. H., and V. H. Ritz. (1957). “A determination of the gamma-ray emission of radium.” J Res Natl Bureau Standards 59:293–305. Ballester, F., D. Granero, J. Perez-Calatayud, E. Casal, and V. Puchades. (2004). “Monte carlo dosimetric study of best industries and Alpha Omega Ir-192 brachytherapy seeds.” Med Phys 31(12):3298–3305. Baltas, D. S. Giannouli, A. Garbi, F. Diakonos, K. Gramani, G. T. Ioannidis, A. Tsalpatouros, N. Uzunoglu, C. Kolotas, and N. Zmboglou. (1998). “Application of the Monte Carlo integration (MCI) method for calculation of the anisotropy of 192Ir brachytherapy sources.” Phys Med Biol 43(6):1783–1801. Batho, H. R., and N. E. T. Young. (1964). “Tissue absorption corrections for linear radium source.” Br J Radiol 37:689. Berger, M. J. (1968). “Energy deposition in water by photons from point isotropic sources.” J Nucl Med Suppl 1:17–25. BIR/IPSM. “Recommendations for Brachytherapy Dosimetry: Report of a Joint Working Party. London: British Institute of Radiology and Institute of Physical Sciences in Medicine, 1993. Boutilon, M., and A. M. Perroche-Rous. (1987). “Re-evaluation of the W value for electrons in dry air.” Phys Med Biol 32:213–291. Burlin, T. E. (1966). “A general theory of cavity ionization.” Br J Radiol 39:361. Carlsson, A. K., and A. Ahnesjo. (2000). “The collapsed cone superposition algorithm applied to scatter dose calculations in brachytherapy.” Med Phys 27(10):2320–2332. Cavallo, L. M., B. M. Coursey, S. B. Garfinkel et al., (1973). “Needs for radioactivity standards and measurements in different fields.” Nucl Instr Methods 112:5–18. Chavaudra, J. (1995). “Pierre and Marie Curie-Sklodowska.” Med Phys 22:1877–1887. Cho, S. H., and R. Muller-Runkel. (1997). “Validity of the interval method for the determination of the anisotropy factor of high dose rate 192Ir sources.” Int J Radiat Oncol Biol Phys 37(2):483–487. Corbett, J. F., J. J. Jezioranski, J. Crook, T. Tran, and I. W. Yeung. (2001). “The effect of seed orientation deviations on the quality of 125I prostate implants.” Phys Med Biol 46(11):2785–2800.
230
Jeffrey F. Williamson
Coursey, B. M., L. J. Goodman, D. D. Hoppes et al. (1992). “The needs for brachytherapy source calibration in the united States.” Nucl Instr Methods Phys Res 312:246-250. Dale, R. G. (1983). “Some theoretical deviations relating to the tissue dosimetry of brachytherapy nuclides, with particular reference to iodine-125.” Med Phys 10:176. Daskalov, G. M., E. Loffler, and J. F. Williamson. (1998). “Monte Carlo-aided dosimetry of a new high dose-rate brachytherapy source.” Med Phys 25(11):2200–2208. Daskalov, G. M., R. S. Baker, D. W. Rogers, and J. F. Williamson. (2000). “Dosimetric modeling of the microselectron high-dose rate 192Ir source by the multigroup discrete ordinates method.” Med Phys 27(10):2307–2319. Daskalov, G. M., R. S. Baker, D. W. Rogers, and J. F. Williamson. (2002). “Multigroup discrete ordinates modeling of 125I 6702 seed dose distributions using a broad energy-group cross section representation.” Med Phys 29:113–124. Diffey, B. L., and S. C. Klevenhagen. (1975). “An experimental and calculated dose distribution in water around CDC-K type cesium-137 sources.” Med Phys 20:446. Gray, L. H. (1936). “An ionization method for the absolute measurement of gamma-ray energy.” Proc Roy Soc A156:578–596. Grigsby, P. W., J. F. Williamson, and C. A. Perez. (1992). “Source configuration and dose rates for the Selectron afterloading equipment for gynecologic applicators.” Int J Radiat Oncol Biol Phys 24(2):321–327. International Commission on Radiation Units and Measurements (ICRU). Report No. 58. “Dose and Volume Specification for Reporting Interstitial Therapy.” Bethesda, MD: ICRU, 1997. Johns, H. E., and J. R. Cunningham. The Physics of Radiology, 4th Edition. Springfield, IL: Charles C Thomas Publishers, 1983. Karaiskos, P., A. Angelopoulos, P. Baras, H. Rozaki-Mavrouli, P. Sandilos, L. Vlachos, and L. Sakelliou. (2000). “Dose rate calculations around 192Ir brachytherapy sources using a Sievert integration model.” Phys Med Biol 45(2):383-98. Keyser, G. M. (1951). “Absorption corrections for radium standardization.” Can J Phys 29:301. Kirov, A. S., and J. F. Williamson. (1997). “Two-dimensional scatter integration method for brachytherapy dose calculations in 3D geometry.” Phys Med Biol 42(11):2119–2135. Krishnaswamy, V. (1972). “Dose distribution about 137Cs sources in tissue.” Radiology 105:181–184. Krishnaswamy, V. (1978). “Dose distribution around an I-125 seed source in tissue.” Radiology 126:489–491. Krishnaswamy, V. (1979). “Dose tables for 125I seed implants.” Radiology 132:727–730. Laurence, G. C. (1935). “Intensity and dosage near radium needles.” Radiology 25:166–181. Lerma, F. A., and J. F. Williamson. (2002). “Accurate localization of intracavitary brachytherapy applicators from 3D CT imaging studies.” Med Phys 28:325–333. Lindsay, P., J. Battista, and J. Van Dyk (2001). “The effect of seed anisotrophy on brachytherapy dose distributions using 125I and 103Pd,” Med Phys 28(3):336–345. Ling, C. C., L. L. Anderson, and W. U. Shipley. (1979). “Dose inhomogeneity in interstitial implants using 125I seeds.” Int J Radiat Oncol Biol Phys 5:419–425. Ling, C. C., E. D. Yorke, I. J. Spiro, D. Kubiatowicz, and D. Bennet. (1983). “Physical dosimetry of I-125 seeds of a new design for interstitial implant.” Int J Radiat Oncol Biol Phys 9:1747–1752. Lymperopoulou, G., E. Pantelis, P. Papagiannis, P. Rozaki-Mavrouli, L. Sakelliou, D. Baltas, and P. Karaiskos. (2004). “A monte carlo dosimetry study of vaginal 192Ir brachytherapy applications with a shielded cylindrical applicator set.” Med Phys 31(11):3080–3086. Meertens, H., and R. van der Laarse. (1985). “Screens in ovoids of a Selectron cervix applicator.” Radiother Oncol 3(1):69–80. Meisberger, L. L., R. J. Keller, and R. J. Shalek. (1968). “The effective attenuation in water of the g-rays of gold198, iridium-192, cesium-137, radium-226, and cobalt-60.” Radiology 90:953. Meredith, W. J. (ed.). Radium Dosage: The Manchester System. Edinburgh: Livingston, pp. 31–136, 1967. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee, Task Group No. 43.” Med Phys 22:209–234. Also available as AAPM Report No. 51. Nath, R., N. Yue, and L. Liu. (1999). “On the depth of penetration of photons and electrons for intravascular brachytherapy.” Cardiovasc Radiat Med 1:72–79.
14–Semiempirical Dose-Calculation Models in Brachytherapy
231
National Council on Radiation Protection and Measurements (NCRP). NCRP Report No. 5. “A Handbook of Radioactivity Measurements Procedures.” Bethesda, MD: NCRP, 1985. Pantelis, E., D. Baltas, K. Dardoufas, P. Karaiskos, P. Papagiannis, H. Rozaki-Mavrouli, and L. Sakelliou. (2002). “On the dosimetric accuracy of a Sievert integration model in the proximity of 192Ir HDR sources.” Int J Radiat Oncol Biol Phys 53(4):1071–1084. Parker, H. M., “A dosage system for interstitial radium therapy. II. Physical aspects.” Br J Radiol 11:252–266. Perez-Calatayud, J., F. Ballester, M. A. Serrano-Andres, V. Puchades, J.L. Lluch, Y. Limami, and E. Casal. (2001c). “Dosimetry characteristics of the Plus and 12i GammaMed PDR 192Ir sources.” Med Phys 28(12):2576–2585. Perez-Calatayud, J., D. Granero, and F. Ballester (2004). “Phantom size in brachytherapy source dosimetric studies.” Med Phys 31(7):2075–2081. Perez-Calatayud, J., D. Granero, F. Ballester, V. Puchades, and E. Casal. (2004). “Monte Carlo dosimetric characterization of the Cs-137 selectron/LDR source: Evaluation of applicator attenuation and superposition approximation effects.” Med Phys 31(3):493–499. Quimby, E. H. (1932). “The grouping of radium tubes in packs and plaques to produce the desired distribution of radiation.” Am J Roentgenol Radium Ther 27:18–39. Quimby, E. H. (1941). “The specification of dosage in radium therapy.” Am J Roentgenol Radiat Ther 45:1–18. Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Also available as AAPM Report No. 84. Roesch, W. C. (1958). “Dose for nonelectronic equilibrium conditions.” Radiat Res 9:399–410. Shalek, R. J., and M. Stovall. “Dosimetry in Implant Therapy” in Radiation Dosimetry, vol III. F. H. Attix and E. Tochlin (eds.). New York: Academic Press, pp. 776–798, 1969. Sievert, R. M. (1921). “Die Intensitatsverteilung der primaren: Strahlung in der Nahe medizinischer Radiumpraparate.” Acta Radiol 1:89–128. Stovall, M., and R. J. Shalek. (1972). “A review of computer techniques for dosimetry of interstitial and intracavitary radiotherapy.” Comput Programs Biomed 1:125. Tedgren, A. K., and A. Ahnesjo. (2003). “Accounting for high Z shields in brachytherapy using collapsed cone superposition for scatter dose calculation.” Med Phys 30(8):2206–2217. Van Kleffens, H. J., and W. M. Sta. (1979. “Application of stereo x-ray photogrammetry (SRM) in the determination of absorbed dose values during intracavitary radiation therapy.” Int J Radiat Oncol Biol Phys 5:557. Wang, R., and R. S. Sloboda. (1998a). “Influence of source geometry and materials on the transverse axis dosimetry of 192Ir brachytherapy sources.” Phys Med Biol 43(1):37–48. Wang, R., and R. S. Sloboda. (1998b). “Monte Carlo dosimetry of the VariSource high dose rate 192Ir source.” Med Phys 25(4):415–423. Weeks, K. J. (1998). “Monte Carlo dose calculations for a new ovoid shield system for carcinoma of the uterine cervix.” Med Phys 25(12):2288–2292. Weeks, K. J., and J. C. Dennett. (1990). “Dose calculation and measurements for a CT-compatible version of the Fletcher applicator.” Int J Radiat Oncol Biol Phys 18:1191–1198. Whyte, G. H. (1955). “Attenuation of g radiation in cylindrical geometry.” Br J Radiol 28:635. Williamson, J. F. (1986). “The accuracy of the line and point dose approximation in Ir-192 dosimetry.” Int J Radiat Oncol Biol Phys 12:409. Williamson, J. F. (1988). “Monte Carlo and analytic calculation of absorbed dose near 137Cs intracavitary sources.” Int J Radiat Oncol Biol Phys 15:227–237. Williamson, J. F. (1990). “Dose calculations about shielded gynecological colpostats.” Int J Radiat Oncol Biol Phys 19:167–178. Williamson, J. F. (1991). “Comparison of measured and calculated dose rates in water near I-125 and Ir-192 seeds.” Med Phys 28:776–786. Williamson, J. F. (1996). “The Sievert integral revisited: Evaluation and extension to low energy brachytherapy sources.” Int J Radiat Oncol Biol Phys 36:1239–1250. Williamson, J. F. (1998). “Monte Carlo-based dose-rate tables for the Amersham CDCS.J and 3M model 6500 137Cs tubes.” Int J Radiat Oncol Biol Phys 41(4):959–70.
232
Jeffrey F. Williamson
Williamson, J. F., and D. A. Brenner. “Physics and Radiobiology of Brachytherapy” in Principles and Practice of Radiation Oncology, 4rd Edition. C. A. Perez, L. W. Brady, E. Halperin et al. (eds.). Philadelphia: J.B. Lippincott Company, pp. 472–537, 2003. Williamson, J. F., and Z. Li. (1995). “Monte Carlo aided dosimetry of the microselectron pulsed and high dose-rate 192 Ir sources.” Med Phys 22(6):809–820. Williamson, J. F., R. Baker, and Z. Li. (1991). “A convolution algorithm for brachytherapy dose computations in heterogeneous geometries.” Med Phys 18:1256–1265. Williamson, J. F., Z. Li, and J. W. Wong. (1993). “One-dimensional scatter-subtraction method for brachytherapy dose calculation near bounded heterogeneities.” Med Phys 20:233–244. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, R. Nath, M. J. Rivard, and G. Ibbott. (1999). “On the use of apparent activity (Aapp) for treatment planning of 125I and 103Pd interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Subcommittee on low-energy brachytherapy source dosimetry.” Med Phys 26:2529–2530. Williamson, J., G. Daskalov, R. Baker, D. Rogers, and I. Kawrakow. (2001). “Accuracy and efficiency comparisons between three-dimensional multigroup discrete ordinates and voxel based Monte Carlo methods for dosimetric modeling of the Model 6702 125I seed.” (Abstract). Med Phys 28(6):1229. Williamson, J. F., and R. Nath. (1991). “Clinical implementation of AAPM Task Group 32 recommendations on brachytherapy source strength specification.” Med Phys 18:439–448. Williamson, J. F., H. Perera, Z. Li, and W. R. Lutz. (1993). “Comparison of calculated and measured heterogeneity correction factors for 125I, 137Cs and 192Ir brachytherapy sources near localized heterogeneities.” Med Phys 20:209–222. Williamson, J. F., and T. Seminoff. (1987). “Template-guided interstitial implants: 137Cs-reusable sources as a substitute for 192Ir.” Radiology 165:265–269. Wyckoff, H. O. “From ‘Quantify of Radiation’ and ‘Dose’ to ‘Exposure’ and ‘Absorbed Dose:’An Historical Review.” Report No. Lauriston Taylor Lecture 4, National Council on Radiation Protection and Measurements, Washington, DC, 1980. Young, M. E. J., and H. F. Batho. (1964). “Dose tables for linear radium sources calculated by an electronic computer.” Br J Radiol 37:38–44. Zhou, C., and F. Inanc. (2003). “Integral-transport-based deterministic brachytherapy dose calculations.” Phys Med Biol 48(1):73–93.
Chapter 15
Quantitative Dosimetry Methods for Brachytherapy Jeffrey F. Williamson, Ph.D.1 and Mark J. Rivard, Ph.D.2 1 Department of Radiation Oncology Virginia Commonwealth University School of Medicine, Richmond, Virginia 2 Department of Radiation Oncology, Tufts University School of Medicine Boston, Massachusetts Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233 Brief History of Brachytherapy Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235 226 Ra Dosimetry: From Threshold Erythema Dose to Semiempirical Models . . . . . . . . . . . . . . . . 235 Rise of Modern Experimental Dosimetry Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236 Rise of Modern Computational Dosimetry Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236 Role of the AAPM in Standardizing Low-Energy Brachytherapy Dosimetry . . . . . . . . . . . . . . . . 237 Experimental Brachytherapy Dosimetry Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238 Reference-Quality Dose Measurement System Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . 239 Positional Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 240 Signal-to-Noise Ratio, Dosimeter Readout Precision, and Signal Stability . . . . . . . . . . . . . . . . 240 Energy-Response and Other Artifacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243 General Formalism for Absorbed Dose Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244 Brachytherapy Dose-Measurement Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248 Thermoluminescent Dosimetry (TLD) Techniques in Brachytherapy . . . . . . . . . . . . . . . . . . . . 248 Diode Dosimetry Techniques in Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254 Plastic Scintillator Dosimetry for Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 256 Radiochromic Film Dosimetry for Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258 Polymer Gel Dosimetry for Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261 Computational Dosimetry Methods in Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264 Monte Carlo-based Brachytherapy Dosimetry: General Considerations . . . . . . . . . . . . . . . . . . . . 265 Technical Aspects of Monte-Carlo Photon-Transport Simulation in Brachytherapy . . . . . . . . . . . 267 Physics of Photon Scattering and Choice of Cross-Section Library . . . . . . . . . . . . . . . . . . . . . 267 Geometric Modeling and Validation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271 Choice of Estimator and Other Variance Reduction Techniques . . . . . . . . . . . . . . . . . . . . . . . . 273 Application of the Monte Carlo Method to Calculation of Reference-Quality Dose-Rate Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275 Uncertainty of Monte Carlo Reference Dose-Rate Distributions . . . . . . . . . . . . . . . . . . . . . . . . 280 Additional Recommendations for Use of Monte Carlo Simulation in Preparing Reference-Quality Dose-Rate Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 Other Applications of Transport Codes in Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 Comparison of Theoretical and Experimental Dose-Rate Distributions for 125I and 103Pd Brachytherapy Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282 Scope of Seeds Included in Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282 125 I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282 103 Pd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283 Ratios of Monte Carlo and Measured Dose-Rate Distributions for 125I . . . . . . . . . . . . . . . . . . . . . 283 Ratios of Monte Carlo and Measured Dose-Rate Distributions for 103Pd . . . . . . . . . . . . . . . . . . . . 284 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
Introduction Computation of brachytherapy dose distributions for clinical treatment planning for a specified arrangement of sources is almost universally based upon the principle of source superposition. The dose rate,
233
234
Jeffrey F. Williamson and Mark J. Rivard
· D(r) , at each point r in the dose-calculation grid is the sum of individual source dose contributions, where each source is defined by its location ri and air-kerma strength SK,i N
D" (r ) = ∑ SK , i ⋅ D0 (r − ri )
(1)
i =1
The dose contribution from each source is a function only of its position ri relative to r and is given · by the single-source dose-rate distribution, D 0(r) which is defined as dose rate per unit SK from a single source with its center positioned at the origin. Prior to the publication of the original TG-43 report (Nath et al., 1995) in 1994, semiempirical or analytic models such as the Sievert integral were used to obtain · D 0(r). Current practice standards dictate a table-based approach. The TG-43 protocol (Nath et al., 1995; Rivrd et al., 2004) is essentially a standardized nomenclature for presenting sparse single-source dose arrays with an associated nonlinear interpolation scheme for estimating normalized dose rate between dose table entries. Clearly, any protocol based upon table lookup requires as input a table of measured or computed dose rates for each type of source. Dosimetry methods useful for estimating reference-quality dose-rate distributions needed to define · D 0(r) and other dosimetric input needed for clinical treatment constitute the main topic of this chapter. Reference-quality dosimetric data is intended for use in clinical treatment planning for sources of the type or model upon which the dose estimates were based. The AAPM recommends (Rivard et al., 2004; Williamson et al., 1998) that reference-quality data be derived from redundant independent dose estimates, be traceable to applicable NIST standards, and undergo a process of peer review before clinical acceptance by the community. This chapter will focus almost exclusively on photon-emitting sources. Although the main emphasis will be on estimation of single-source reference quality dose distributions, estimation of other types of dosimetric data, including direct use of quantitative dosimetry techniques in patient-specific treatment planning, will be briefly considered. Dosimetry is that branch of medical physics that concerns itself with the estimation of absorbed dose by means of experimental or fundamental theoretical techniques. An experimental dosimetry technique consists of a detector, known as a dosimeter, which produces a measurable signal that has a known relationship to absorbed dose at a specified point in the medium in the absence of the dosimeter. Well-defined procedures for dosimeter calibration, irradiation, readout and artifact correction are essential components of any experimental dosimetry method. Theoretical or computational dosimetry techniques must accurately model all physical phenomena that can potentially affect the brachytherapy dose distribution, including emission of radiation via radioactive decay and transport, scattering, and absorption of the emitted radiation by the source itself or surrounding media. For brachytherapy, the goal of most dosimetric techniques is estimation of absorbed dose at geometric points in condensed medium within which the radioactive sealed source or sources are immersed. Useful dosimetry methods must be quantitative: the estimated dose distribution must more or less accurately approximate the actual values of well-defined physical quantities and must be accompanied by a rigorous estimate of uncertainty. The estimated uncertainty specifies an interval about the measured or computed value within which the true absorbed dose may confidently be expected to lie (Taylor and Kuyatt 1994). Widely accepted brachytherapy dosimetry techniques support uncertainties ranging from 3% to 10% over the range of clinically relevant source-to-detector distances (0.5 mm to 10 cm). Dosimetry techniques with indeterminate uncertainties or uncertainties in excess of 10% are generally not acceptable for determination of reference quality dosimetry data (Rivard et al., 2004). Another feature of most quantitative dosimetric methods is that their results are traceable to the applicable NIST standards. For all techniques, this means that absolute dose-rate estimates must be normalized to the appropriate NIST SK standard for the source model under consideration. For experimental techniques using secondary dosimeters, their dose calibrations should be secondarily traceable to the appropriate absorbed dose or kerma
15–Quantitative Dosimetry Methods for Brachytherapy
235
primary standards. Other measured quantities, e.g., pressure, temperature, charge, mass, time, etc., should also have a clearly defined chain of traceability to the corresponding NIST standards.
Brief History of Brachytherapy Dosimetry 226
Ra Dosimetry: From Threshold Erythema Dose to Semiempirical Models
Experimental techniques for quantifying radiation fields arising from sealed radioactive sources have been used since the discovery of 226Ra in 1898 by Marie Curie. Roentgen is credited with introduction of the gold-leaf electroscope in 1900, which is essentially a primitive ion chamber that functions both as a detector and readout device. Villiard is believed to have introduced the first ionization chamber with electrodes held at constant potential soon thereafter (Villiard 1908). A wide variety of biological and chemical dosimeters were evaluated for use in mapping “dose” distributions about early brachytherapy sources, including bleaching butter and tissue necrosis in rabbits (Quimby 1928) (see also Quimby’s fascinating 1940 review (Quimby 1941)). By plotting the distance to effect (e.g., butter bleaching) as a function of exposure time and source strength, “dose” fall-off as a function of distance could be inferred. Perhaps the most widely used biological endpoint was threshold erythema dose (TED), which is defined as the dose needed to produce a barely detectable skin erythema in 80% of the subjects so irradiated. This technique, developed by Quimby and colleagues at Memorial Hospital, apparently had a reproducibility of about 10% (Dean 1923; Quimby 1928, 1941). The concept of absorbed dose, although first defined by Christen in 1914 (Christen 1914), was itself controversial during the first half of the 20th century. One problem was that no theory existed from which the absorbed dose could be inferred from response of a detector. The development of the free-air chamber (FAC) in the 1920s, from which the quantity exposure could be rigorously inferred, led to the widespread acceptance of exposure in free-space as the primary quantity for quantitatively describing all radiotherapy treatments [see Wyckoff’s historical review (Wyckoff 1980)]. The FAC and the quantity exposure provided a sound foundation for quantifying radiotherapy treatments of the 1930s, which were limited to x-ray spectra of 300 kVp. However, for the much higher energy (about 1.2 MeV on average) gamma rays emitted by 226Ra, exposure could not be measured accurately because of the difficulty in establishing secondary charged particle equilibrium in free-air FACs of manageable size. In her 1941 review of radium dosimetry, Quimby (1941) notes that experimental estimates of the 226Ra exposure-rate constant in the 1920–1940 period varied by a factor of two. For nearly 40 years, the radiological physics community struggled to solve the problem of how to specify brachytherapy treatments in terms of rigorously defined quantities, the measurements of which are traceable to well-defined primary standards. The problem was solved in the late 1930s following the introduction of quantum mechanical treatments of photon scattering (Gray 1929) and electron energy loss (Bethe 1930; Bloch 1933). This made possible the development of a practical cavity theory (Gray 1936; Laurence 1938), which allowed exposure to be rigorously inferred from small ion chambers with condensed matter walls thick enough to establish transient charged-particle equilibrium. By 1938, experimental estimates of (Γ δ)Ra,0.5 had converged to within a few percent of Attix’s definitive 1957 measurement (Attix and Ritz 1957), which yielded 8.25 R⋅cm2⋅mg−1⋅h−1 for radium needles filtered by 0.5 mm of Pt. NIST introduced exposure-based primary standards for 137Cs and 192Ir brachytherapy sources in 1974 and 1980 (Loftus 1970, 1980). To compute dose distributions about implanted 226Ra needles and tubes, physicists employed semiempirical computational dosimetry models, reviewed in more detail elsewhere in this monograph. Quimby (1922) developed a numerical technique for partitioning needles into a linear array of point sources, from which away-and-along exposure tables could be inferred. This was followed by Sievert’s more elegant analytical solution, the Sievert integral method, in 1923. By the late 1930s, the exposure/mg⋅h tables used
236
Jeffrey F. Williamson and Mark J. Rivard
by the Quimby (1952) and Manchester (Parker 1938) interstitial implant systems were based upon computed single-source dose distributions. This approach continued without major modification upon the introduction of artificial radium-substitute radionuclides. Direct measurement of absorbed dose near brachytherapy sources, using thermoluminescent dosimeters (TLDs), radiographic film, solid-state detectors, or small ion chambers was relatively rare even in the research laboratory and had little impact on clinical practice. Since there was no basis for assigning an uncertainty to these calculated dose-rate estimates, this approach does not satisfy our criteria, described above, for quantitative dosimetry. One may speculate that the reluctance to accept measured doses as the basis of dose prescription is due both to the technical difficulties that attend such measurements as well as the empirical character of clinical brachytherapy itself.
Rise of Modern Experimental Dosimetry Methods The classical, semi-empirical approach to brachytherapy dosimetry began to break down with the introduction of 125I interstitial seeds in the early 1970s, as this 30 keV x-ray emitter clearly fell outside the scope of validated analytic models. While 125I dose distributions derived from semiempirical models were published (Krishnaswamy 1978) and widely used, it was recognized (Ling et al., 1983) that internal seed structure could modulate the emitted photon spectrum and have significantly alter the absorbed dose distribution. The introduction of an exposure-based free air chamber (FAC) primary standard in 1985 (Loftus 1984), along with its increasing clinical use, led to a growing realization that more quantitative dosimetry methods were needed. For a more detailed discussion of early of 125I dosimetry, the reader is referred to appendix C of the first TG-43 report (Nath et al., 1995). In 1986, the National Cancer Institute funded a three-year, multi-institutional contract to perform a definitive review of low-energy seed dosimetry. The three institutions, collectively called the Interstitial Collaborative Working Group (ICWG) consisted of Memorial Sloan-Kettering, Yale University, and University of California at San Francisco led by principal investigators Lowell Anderson, Ravinder Nath, and Keith Weaver, respectively (Anderson et al., 1990). Using TLD-100 thermoluminescent chips and powder capsules, embedded in machined solid-water phantoms, the ICWG developed procedures, including TLD dose calibration and energy-response correction, for making quantitative estimates of absolute dose rates in water. Each of the three ICWG investigator groups independently measured (Chiu-Tsao et al., 190; Nath, Meigooni, and Meli 1990; Weaver et al., 1989) transverse-axis dose distributions for the 125 I and 192Ir than available to validate their TLD measurement methodology. This was followed by more complete 2-D dose distributions about 125I, 192Ir, and 103Pd brachytherapy sources then available ChiuTsao et al., 1991; Meigooni, Sabnis, and Nath 1990; Nath et al. 1993). The results showed good agreement among the different measurements and overall, substantial differences between measured and classically computed dose rates for 125I seeds, but good agreement between the classical and experimental approaches for 192Ir. As a result of the ICWG efforts along with many later investigators, TLD dosimetry came to be accepted as the most reliable and best validated experimental approach in brachytherapy and its results are widely used as the basis for clinical dose computation.
Rise of Modern Computational Dosimetry Approaches Independently of the ICWG initiative, other investigators were studying the use of Monte Carlo photontransport techniques as tools for quantitative evaluation of single-source dose distributions. Based on an accurate and detailed mathematical model of the internal structure of the source, photon histories can be generated and then evaluated to assess absorbed dose. 1-D Monte Carlo simulations and other solutions of the transport equation have been used since the 1960s to calculate radial dose distributions arising from isotropic point sources in medium, e.g., the widely
15–Quantitative Dosimetry Methods for Brachytherapy
237
used point-source buildup factors of Berger (1968) and the tissue-attenuation and scatter-buildup factors of Meisberger (Meisberger, Keller, and Shalek 1968). 1-D Monte Carlo techniques were first applied to 125 I by Dale (1983). However, Monte Carlo techniques were applied to more geometrically complex problems in brachytherapy only relatively recently. One of the earliest 3-D studies (Williamson, Morin, and Khan 1983) used Monte Carlo simulation to assess the accuracy of the Sievert model for platinum-encapsulated 226Ra and 192Ir sources suspended in free space. Burns and Raeside (1988) published one of the first Monte Carlo dosimetry studies based upon a geometrically realistic 125I seed model. Williamson’s 1988 Monte Carlo study (Williamson 1988) was the first to note that semiempirical dose calculation models overestimated absolute dose rates about 125I by 10% to 14% and demonstrated that low-energy titanium characteristic x-rays detected by the NIST primary standard accounted for most of this discrepancy. Comparisons of ICWG TLD measurements and Monte Carlo-based dose distributions (Anderson et al., 1990; Williamson 1991) showed excellent agreement. Williamson and his colleagues published a series of studies comparing Monte Carlo dose-rate predictions to silicon diode (Perera et al., 1994; Williamson, Perera, and Li 1993) and TLD (Das et al., 1996, 1997; Kirov et al., 1995; Valicenti et al., 1995) experimental benchmarks which established the reliability and accuracy of Monte Carlo-based dose predictions across the energy spectrum of brachytherapy in both heterogeneous and homogeneous geometries. Currently, Monte Carlo simulation is a widely used and accepted dosimetry tool.
Role of the AAPM in Standardizing Low-Energy Brachytherapy Dosimetry The AAPM has had a substantial impact on the clinical practice of brachytherapy dose calculation, promoting uniformity of dose-calculation practice that has benefited literally hundreds of thousands of patients. The AAPM’s role in brachytherapy dose computation began with the publication of the TG-43 report (Nath et al., 1995). Its proposed a dose-calculation formalism, derived from the ICWG recommendations (Anderson 1990) was based upon the concept of using measured or Monte Carlo-derived dose distributions, rather than de novo dose calculations using a semiempirical model. In addition, TG43 reviewed the published TLD and Monte Carlo data available for the models 6711 and 6702 125I sources, the model 200 103Pd source, and the stainless steel-clad 192Ir seed. For each of these sources, a consensus dataset was recommended, including dose-rate constants, anisotropy constants and factors, anisotropy functions, and radial dose functions. The 1995 TG-43 recommendations implied that 125I dose rates were 10% to 15% lower than those predicted by then current dose-calculation practices. In a related development, NIST had developed a new primary standard for low-energy seed air-kerma strength based upon the wide-angle free-air chamber (WAFAC) (Seltzer et al., 2003) which suppressed measurement of 4.5 keV Ti characteristic x-rays, which was responsible for much of the discrepancy between TG-43 and semiempirical dose calculations. Because of the potential for error at the user level was high, as both source strength and absorbed dose scales were changing by more than 10%, the AAPM created an Ad Hoc Working Group under the leadership of Dale Kubo to assist the community in adapting to these changes. The Working Group report (Kubo et al., 1998) developed a step-by-step procedure for implementing the new WAFAC standard and the TG-43 formalism and recommended that the 125I monotherapy prescribed dose be revised from 160 Gy to 145 Gy. In the authors’ opinion, the complexity of these changes motivated the community and planning software vendors to rapidly adopt the TG-43 dose-calculation formalism. Clinical brachytherapy practice patterns continued to rapidly evolve, motivating the AAPM to create a permanent Radiation Therapy Committee working group (later subcommittee) on Low-Energy Interstitial Brachytherapy Dosimetry (renamed “Photon-Emitting Brachytherapy Dosimetry Subcommittee,” or PEBD, in 2003), under the leadership of Jeffrey Williamson. The practice-pattern changes included a rapid shift from radical prostatectomy to permanent seed implantation as the dominant modality for treatment of low risk prostate cancer, a proliferation of new low-energy interstitial source products (from
238
Jeffrey F. Williamson and Mark J. Rivard
2 in 1996 to over 20 in 2000), extension of the WAFAC standard to 103Pd sources in 1999, and concerns over the stability of 103Pd dose specification over time. This subcommittee took an active role the introduction of new source strength standards and revised dosimetry practices, coordinating efforts by NIST, source vendors, the clinical community, and Accredited Dosimetry Calibration Laboratory (ADCL) secondary dosimetry laboratories often on a source model-by-model basis. PEBD devised a set of dosimetry guidelines (Williamson et al., 1998), recommending that every routinely used low-energy interstitial source product have a NIST-traceable SK calibration and be accompanied by peer-reviewed published and independent experimental and Monte Carlo dose-rate distributions. These guidelines have become de facto industry standards, accepted by nearly all vendors involved in the market. PEBD also took leadership in implementing the NIST WAFAC standard for the model 200 103Pd source in a coordinated fashion with the vendor including a historical assessment of “administered-to-prescribed” dose ratios that enabled users to duplicate dose delivery practices of the past in the face of multiple revisions to the 103Pd dosimetry parameters and calibration changes (Williamson et al., 2000, 2005). Another PEBD contribution was a major revision of the TG-43 dose-calculation formalism, the preparation of consensus dosimetry parameters for eight source models, and guidelines for the practice of TLD and Monte Carlo dosimetry (Rivard et al., 2004). A supplement to this report, including consensus dosimetry parameters for an additional ~10 source models, is expected to be submitted for AAPM approval in late 2005, with subsequent publication in Medical Physics in 2006. In 2005, the AAPM reorganized the Science Council committee structure, replacing PEBD with the Brachytherapy Subcommittee (BTSC), under the leadership of Zuofeng Li, and with responsibility for guidance in all areas of brachytherapy. Working groups under BTSC supervision extend the system of dosimetry recommendations to all classes of brachytherapy sources including intravascular brachytherapy sources.
Experimental Brachytherapy Dosimetry Methods Experimental measurement is the epistemic foundation of brachytherapy dosimetry as all other quantitative approaches must be directly or indirectly validated against measurement. Dose measurement requires accurately positioned detector that yields a measurable signal with a known relationship to relative or absolute absorbed dose in the medium in the absence of the detector. Detectors commonly used in this application include small TLDs or silicon diode. Emerging detector technologies include radiochromic film, plastic scintillator, and polymer gels. Until about 1980, direct measurement of dose around brachytherapy sources and applicators in support of clinical treatment planning and quality assurance was relatively uncommon even within the research setting, let alone the clinical environment. Even today, clinical quality assurance practice is limited to experimental confirmation of the strength and geometry of purchased sources and applicators (Nath et al., 1997). Historically, this is due not only to the difficulties and labor-intensity of such measurements, but to a consensus view that dose measurement was so difficult and intrinsically inaccurate at small distances from sealed sources that even simplistic theoretical models were more reliable. Indeed, brachytherapy dose measurement does place severe demands on detectors since the dose distributions are characterized by large dose gradients, a large range of dose rates, and relatively low photon energies. A suitable detector must have a wide dynamic range, flat energy response, small size, and high sensitivity. Nearly all detector systems explored to date in brachytherapy are secondary dosimeters, i.e., detectors that require calibration against radiation fields, usually low megavoltage photon beams, in which the dose rate is known at the detector location. The one exception to this generalization is the parallel plate extrapolation chamber, which has been developed into a primary standard of absorbed dose for sealed beta-emitting sources (Soares, Halpern, and Wang 1998).
15–Quantitative Dosimetry Methods for Brachytherapy
239
Reference-Quality Dose Measurement System Characteristics While there are no hard and fast criteria for what constitutes a suitable dose-measurement system for brachytherapy, systems that provide credible reference-quality dose-rate measurements generally share the following characteristics: (1) Support absolute as well as relative measurement of dose at a geometric point in the 0.5-cm to 10-cm distance range in liquid water or Solid Water-substitute media (2) Support spatial density of measurements of at least 5 mm or 5 degrees in polar angle (3) Support dose measurement uncertainties of no greater than 10%, preferably no more than 5%, in the 0.5-cm to 10-cm distance range. Total uncertainty (Taylor 1994) is assumed to include all known systematic error sources (type B) as well as statistical uncertainties (type A). (4) Support independent dose estimation, i.e., should require minimal knowledge of the dose distribution to correct for detector measurement artifacts. No detector system evaluated to date satisfies all these requirements. All currently used detectors, including organic detectors such as radiochromic film, plastic scintillator, polymer gels, and solid state detectors, such as diodes and TLDs, are subject to artifacts such as volume-averaging, self-attenuation, anisotropy, and variable energy-response. For secondary systems, criterion (1) implies that a dosimeter must have sufficient temporal reproducibility and stability of response that detector responses measured in a brachytherapy irradiation geometry can be compared to those irradiated by a calibrated megavoltage beam. For single element detectors such as ion chamber or diode detectors, intercomparison requires temporal stability of response. Multiple-element detectors such as TLD or film must have sufficiently uniform responses from detector-to-detector that the responses of detectors irradiated by different fields or detectors irradiated in to different doses in the same field can be compared. Criterion (2) addresses the issue of efficiency and labor intensiveness of dose-measurement technologies. The dose-measurement process must not be so labor intensive as to preclude repeated measurement needed to establish measurement precision and dose mapping over a sufficiently dense grid of measurement points, allowing high gradient dose distributions to be meaningfully characterized. For single-source reference-quality dose measurement, the 2004 TG-43 report sets forth general recommendations (Rivard et al., 2004). Approximately 50 to 100 measurement points are needed to characterize the 2-D dose distribution of a typical low energy interstitial seed. Criteria (3) and (4) are the most difficult to satisfy. The minimum uncertainty achievable by a # D" (r)/SK measurement is about 3% (Dempsey et al., 2000), the combined uncertainties of the air-kerma strength calibration (and transfer to the measured sources) and transfer of a megavoltage beam calibration to the detector system. Assuming our target uncertainty is 8%, this leaves a residual uncertainty budget of 7.4% for all other uncertainties associated with the measurement. Criterion (4) implies that detector artifact corrections should not be so large as that their estimations requires detailed knowledge of the dose distribution. For example, volume-averaging corrections over a 1 mm3 TLD detector can be estimated from simple inverse-square law calculations, while using a 0.03 cm3 air-filled volume may require extensive Monte Carlo calculations to derive displacement and gradient corrections. To achieve the target uncertainty, the uncertainty of these corrections, or error added by uncorrected or partially corrected artifacts, must be substantially less than 5%. The following paragraphs review the major types of measurement artifacts and uncertainty sources.
240
Jeffrey F. Williamson and Mark J. Rivard
Positional Accuracy The combination of short source-detector distances and inverse square-law gradients means that brachytherapy dose measurement is exquisitely sensitive to accuracy with which the detector can be positioned at known locations relative to the source. For example, to limit positioning uncertainties to 2% at distances of 2 mm, 5 mm, and 10 mm from a point source, requires positioning accuracies of 20, 50, and 100 µm, respectively. For typical single-element detector mechanical scanning systems, symmetry of orthogonal detector-reading profiles can be used to estimate the center of radioactivity in the detector coordinate system (Ling et al., 1985) as illustrated in Figure 1. This method results in a measurement coordinate system with the origin placed at the center of radioactivity (not physical source center) but does not account for location of the center of the active source volume in the detector housing. The accuracy achievable depends on the extent of hysteresis, deviation from orthogonality, and positioning precision of the detector transport mechanism. With micrometer-quality drives, 50 to 100 µm accuracies can be achieved. Alternatively, for a micrometer-positioned diode detector in a water phantom, Williamson has described a mechanical and optical alignment procedure which achieved an accuracy of 0.03 mm (Williamson, Perera, and Li 1993). For single-element integrating detectors such as TLD chips, solid water-substitute phantoms equipped with machined cavities and recesses for the source and array of TLD detectors is widely used (Anderson et al., 1990; Nath, Meigooni, and Meli 1990). The geometric accuracy achievable is determined by the machining accuracy and cavity size enlargement required for ready removal and insertion of source and detectors (Figure 2). It is estimated that geometric accuracies of 0.1 to 0.2 mm can be achieved by the solid phantom approach (Meigooni et al., 1995; Rivard et al., 2004). 2-D detectors [radiochromic film (Dempsey et al., 2000)] and 3-D detectors [polymer gels (Maryanski et al., 1996)] provide increased opportunity for improved geometric accuracy. Such detectors consist of a continuous sheet or volume of sensitive material. The detector coordinate system is imposed by the digital readout system [optical CT (Gore et al., 1996) or MRI for polymer gels and scanning densitometers (Dempsey et al., 1999) for radiochromic film], which assigns a signal intensity to each voxel or pixel of the detector medium. Digital readout systems often support very high spatial resolutions and geometric precisions (about one-third of the pixel dimension) assuming they are free of geometric distortion. Usually a combination of the orthogonal profile method and mechanical alignment of the detector with respect to the source is used to obtain localization accuracy of the order of 25 to 100 µm. Signal-to-Noise Ratio, Dosimeter Readout Precision, and Signal Stability Achieving a good balance between readout precision and spatial resolution is a challenge faced by all dosimeter systems in brachytherapy. Brachytherapy dose rates range from 50 cGy/s [5 mm from a high dose-rate (HDR) source] to much less than 1 cGy/h (5 cm from a clinical-strength 125I seed). Both small size, to better resolve spatially varying dose distributions and minimize volume-related detector artifacts, and high sensitivity are desired. The inherent sensitivity of a detector system is governed by its quantum efficiency (energy transferred from the radiation field per observed signal quantum) and the collection efficiency. Table 1 summarizes data for several commonly used single-element detectors. Because TLD fluorescent response is a second-order effect, its inherent response is much inferior to other detectors while silicon diode is the most efficient. Despite its poor intrinsic quantum yield, the practical response of TLD per unit volume approaches that of ion chamber, because TLD is solid. By irradiating TLDs long enough, integrated signals with good signal-to-noise ratio characteristics have been obtained over the distance range of 1 to 7.5 cm near low dose-rate (LDR) sources. Another consideration is electronic noise and background, which can dominate readout statistics at low signal intensities. For example, the signal from the ion chamber in Table 1 at a dose rate 0.1 cGy/h is only 1×10−17 amperes, which is significantly less than the leakage current from the best commercially available electrometer. In contrast
15–Quantitative Dosimetry Methods for Brachytherapy
(a)
(b)
(c)
(d)
241
Figure 1. Illustration of source symmetry method for identifying the coordinates, (X0,Y0) denoting the projection of the brachytherapy active source center in the plane (a) of a 2D detector. Panels (a) and (b) illustrate radiochromic film detector readings in an arbitrary detector frame of reference fixed to the upper left hand corner of the film. (X0,Y0) are located by finding the detector coordinates that bisect the distance intervals corresponding to the full-width half maximum of two orthogonal profiles, A and B, near the source center.
242
Jeffrey F. Williamson and Mark J. Rivard
Figure 2. Solid-water phantoms designed by A. S. Meigooni at the Mallinckrodt Institute of Radiology for measurement of transverse-axis (left panel) and polar (right panel) dose-rate distributions. At distances less than 2 cm, small (1 × 1 × 1 mm3) TLD-chips are used. Unused detector slots are filled with solid-water plugs.
Table 1. Absolute and Relative Sensitivity of Dosimeters Commonly Used in Brachytherapy. The relative signal and photon yield estimates assume 1 mm3 detector volume and absorbed dose of 0.1 cGy, respectively. Data from Perera et al. (1992). Detector
TLD-100 Silicon Diode Ion Chamber (air) Plastic Scintillator fiber
Energy (eV) dissipated/ observed quantum
No. quanta emitted
Typical quantum efficiency x geometric collection efficiency
Practical signal relative to ion chamber
8400
1.8 × 106
0.20 × 0.08
0.120
3.6
4.0 × 10
9
1.00 × 1.00
17,000
33.8
2.4 × 10
5
1.00 × 1.00
1.00
100
6.2 × 10
7
0.20 × 0.20
10.3
to the ion chamber, TLD is a practical dosimeter since (a) the TLD detector itself integrates the signal rather than an external readout system, (b) TLD background signal does not increase appreciably with time, and (c) long 6- to 48-hour irradiations are practical as many TLD detectors can be simultaneously irradiated in a single experiment. Similarly, model MD-55-2 radiochromic film, despite requiring nearly 80 Gy to achieve an optical density of 1.0, is practical for LDR brachytherapy dosimetry, because the signal maybe integrated for days (Monroe et al., 2001) or even weeks (Chiu-Tsao et al., 1994) to achieve good signal statistics. Small nonintegrating small diode and plastic scintillation detectors (Fluhs 1996) are useful because they have much higher effective sensitivities than ion chamber. For the purposes of reference-quality dose-rate measurements, the AAPM recommends using detectors that have 1s statistical uncertainties (type A) ≤ 5%. Stability of detector response through time is another problem that must be considered. For example, TLD detectors are subject to fading artifacts if not properly annealed (see below). Radiochromic film response continues to increase after exposure the rapidity and extent of which depends on a complex
15–Quantitative Dosimetry Methods for Brachytherapy
243
way on temperature (Reinstein and Gluckman 1999a) and dose (Ali et al., 2003, 2005) by careful temporal synchronization of the calibration and brachytherapy exposures (Dempsey et al., 2000). Energy-Response and Other Artifacts Another desirable feature of a brachytherapy dosimeter is a constant response as a function of energy. When this condition is present, sensitivity (net detector reading/dose in medium) is constant at all measurement points in the brachytherapy geometry. Because most dosimeters used in brachytherapy are composed of crystalline or organic media, their effective atomic numbers usually do not exactly match that of water or tissue. Point-to-point variations in detector sensitivity due to corresponding changes in the photon spectrum are known as energy-response artifacts. Energy-response artifacts are illustrated by Figure 3, which shows detector reading, relative to dose in water, as a function of energy for source-detector distances of 1 and 10 cm for several common detectors. Even for organic detectors (Zeff = 5.9) the measured reading/unit dose depends significantly on photon energy. In the low-energy photon range (20 to 30 keV), detector sensitivities tend to be independent of measurement distance, even though detector-response corrections relative to the megavoltage photon calibration maybe as large as 7 for silicon diodes. Thus, any of these detectors maybe used to map 103Pd and 125I dose distributions without detailed knowledge of the photon spectrum at each measurement point. However, for higher-energy sources, e.g., 192Ir at 400 keV, Figure 3 shows that diode response varies by 60%, a finding that has been confirmed experimentally (Williamson et al., 1993). Even when using TLD for 192Ir measurements (Meli, Megooni, and Nath 1988) or diode for 137Cs measurements (Williamson, Perera, and Li 1993), artifacts of approximately 10% to 15% have been observed. Hence for dosimetry of intermediate and high-energy brachytherapy sources, detectors should be matched to the composition of the surrounding medium as closely as possible.
Figure 3. (Left) Theoretical detector response (dose to detector/dose to water) of pure plastic scintillator (PVT), arsenic-doped plastic scintillator and TLD-100 (LiF) detectors at 1 cm and 10 cm distances in water from point sources of various energies. (Right) Same except TLD-100 detector is replaced by a silicon diode detector. The results of each detector type are normalized to unity at 1 cm from the 137Cs source. (Reprinted from Int J Radiat Oncol Biol Phys, vol 23, “Rapid two-dimensional dose measurement in brachytherapy using plastic scintillator sheet: linearity, signal-to-noise ratio, and energy response characteristics.” H. Perera, J. F. Williamson, S. P. Monthofer, W. R. Binns, J. Klarmann, G.L. Fuller, and J. W. Wong, pp. 1059–1069. © 1992, with permission from Elsevier.)
244
Jeffrey F. Williamson and Mark J. Rivard
General Formalism for Absorbed Dose Measurement Consider the general problem of inferring the absorbed dose rate per unit air-kerma strength (SK), # # BTx in the absence of the detector at location r from the detector reading, R(r ) the active [ Dwat (r# ) / SK ]meas center of which is located at the same point in space. As illustrated in Figure 4, the goal is to estimate dose rate to water at a geometric point in the reference liquid-water phantom (30-cm liquid water sphere) recommended by AAPM for specification of single-source dose distributions (Rivard et al., 2004). Clearly, many corrections are needed, including those for averaging of dose over the detector volume, displacement of the medium by the detector, and differences in the composition, size, and shape of the # BTx measurement phantom compared to the reference water sphere. [ D" wat (r ) / SK ]meas is given (Williamson and Meigooni 1995) by: #
#
R(r ) ⋅ g(T ) D" wat (r ) S = S ⋅ F ( R(r# )) ⋅ ε ⋅ E (r# ) K meas K lin λ where # R( r ) = g(T) =
ελ =
# E (r ) =
BTx
(2)
is the net detector reading integrated over the measurement time, T, in hours.
λ 1 − e− λT is the effective irradiation time for an #exposure time T near a source with a decay constant λ (h−1). Applying this correction to R(r ) converts the integrated reading into an hourly rate decayed to the beginning of the exposure.
the detector sensitivity, ε λ = R0 ( D0 )med meas where R0 is net the response of the detector to an arbitrary reference dose (D 0)med from the reference megavoltage photon beam λ used to calibrate the detectors. For detectors that have a linear response to absorbed dose, ελ is independent of the reference dose level. λ
[ R (r# ) / D (r# )] for Brachytherapy source = ε ( R , r# ) ε (R ) [ R / D ] for Calibration Beam λ 0
wat
0
med
BTx
0
λ
0
(3)
This quantity is the energy-response correction, defined as the linearized response of the detector per unit dose in water medium of the brachytherapy source relative to that of the reference beam λ. # # λ BTx # Dwat (r ) and Dmed are the doses necessary to achieve the reference reading R0 and E (r ) = Dmed Dwat (r ) at reading level R0.
Flin ( R ) =
[ R / D ] at a dose level correspoonding to reading R ε λ ( R) = [ R0 / D0 ] at a reference reading R0 λ ε λ ( R0 )
(4)
is the detector linearity correction, which is usually measured using the calibration radiation source. SK =
is the air-kerma strength of the source in units of U where 1 U = 1 mGy·m2·h–1 = 1 cGy·cm2·h–1 decayed to the beginning of the experiment.
15–Quantitative Dosimetry Methods for Brachytherapy
245
Figure 4. Illustration of the dose-measurement problem in brachytherapy. Given the reading of a volume detector located at (x,y,z) in a potentially non-water, nonspherical phantom, the goal is to estimate the dose rate at per unit SK at a geometric point when the source is placed at the center of liquid-water phantom 30 cm in diameter.
(
)
# # Flin R(r) ⋅ ε λ ⋅ E(r) in the denominator of equation (2) is equal to The product of the three terms # # its sensitivity ε B Tx (r) in the brachytherapy irradiation, i.e., the detector reading R, at position r in the brachytherapy measurement phantom per unit dose to water in the reference liquid water phantom. Of the above corrections, the relative energy response correction is the most difficult to specify accurately. Indeed, for low-energy source dosimetry, its uncertainty dominates the total dose-measurement # uncertainty (Rivard et al., 2004). Corrections for many phenomena are bundled into E(r), including detector artifacts such volume averaging, self attenuation, displacement of surrounding medium as well # as intrinsic energy response. In addition, E(r) accounts for the nonliquid water equivalence of the measurement phantom and while its denominator includes corrections to the megavoltage beam calibration, e.g., use of a solid-water–substitute phantom and any other deviations from the TG-51 protocol (Almond et al., 2005). To avoid systematic errors, the calibration conditions assumed by the denominator of equation (3) must match those of the detector calibration protocol used to measure # ελ. Both direct experimental and theoretical methods have been used to evaluate E(r) for low-energy source TLD dosimetry. If one assumes that the detector response is proportional to dose absorbed by the detector active volume, i.e., Rdet = α ⋅ D det λ
λ
(5)
where α is a constant independent of the photon energy, λ, of the source, then one can show that (Das et al., 1996) # # ∆D det(r)/∆D w at(r) # E(r)= ∆D det /∆D m ed MC
λ
MC
BRx
(6)
246
Jeffrey F. Williamson and Mark J. Rivard
where the delta, i.e., ∆D, refers to dose relative to the inherent normalization of the Monte Carlo code (usually some quantity proportional to Gy/simulated photon history). The terms in the numerator are derived from two parallel Monte Carlo simulations, center of a 30 cm liquid # one with the source at the # ∆D ( r ) water sphere yielding an estimate of dose, at the geometric point r , and a simulation includwat # ing the source and detector geometry in the measurement phantom, yielding an estimate of ∆Dwat (r ), # energy imparted to the detector active volume when centered at r , per unit mass of active material. The denominator requires calculation of the same quantities for the megavoltage photon beam calibration. This theoretical approach rigorously models such effects as volume averaging (displacement correction), displacement of the medium by the detector (replacement correction), and that component of energy response variations due to composition and density mismatches between the detector and the target water medium. Using photon Monte Carlo simulation, Williamson has evaluated relative energy response for TLD and other types of detectors using a modified form of equation (6): # # β ⋅ ∆K det (r ) / ∆K wat (r )]BTx [ # [ E (r )]MCPT = [ β ⋅ ∆K det / ∆K med ]λ
(7)
where absorbed dose has been replaced by collision kerma. The constants β are small corrections of the order of 1% that account for contribution to detector response by secondary electrons arising in the surrounding medium, which can be significant when the minimum linear dimension of the detector is comparable to the range of the secondary electrons. Usually, β can be evaluated with sufficient accuracy using Burlin cavity theory (Attix 1986; Valicenti et al., 1995). # The theoretical method of evaluating E (r ) is valid only if the detector response is inherently linear with respect to energy deposition, a characteristic often referred to as “LET linearity.” As discussed below, such LET linearity cannot be assumed for brachytherapy dosimeters, especially for low photon energies. The direct empirical approach consists of irradiating the detector to a known dose in free space in an x-ray beam (see Figure 5) with a spectrum that matches the brachytherapy source of interest. The criterion of equivalence is usually agreement between the effective energy derived from the beam half-value layer and the kerma-weighted mean energy of the brachytherapy source spectrum. Reference beams must be calibrated with an ion chamber having a NIST-traceable air-kerma calibration factor. During the development of quantitative TLD dosimetry, a number of investigators (Luxton et al., 1990; Meigooni, Meli, and Nath 1988a; Muench et al., 1991; Weaver 1984) used variations of this approach obtaining distance# independent E (r ) values ranging from 1.39 to 1.44 for TLD-100 detectors at 125I energies.
(
FS Measurements illustrated in Figure 5 provide an estimate of the free-air response, R K air
K
FS air
(R
)
hν
, where # is the air-kerma in free space at the detector center and hv is the beam energy. To estimate E (r ) from FS K air
)
hν Meas
Meas
for brachytherapy in-phantom measurements, a number of additional corrections are
required:
(
)
hν
(
)
hν
FS FS K FS D wa R K air air t thy # Meas ⋅ E (r) = λ # BTx ( R Dmed )Meas crepl ⋅ cdisp ( r ) thy
(8)
The quantities bearing the subscript “thy” must be evaluated by Monte Carlo or other theoretical method. The ratio
(
K FS D FS air wat
)
hν
thy
≈ ( µ en (hν ) / ρ )wat air
converts the air-kerma calibration to dose in water.
15–Quantitative Dosimetry Methods for Brachytherapy
247
Figure 5. Direct experimental measurement of energy-response correction by intercomparing the detector to a calibrated ion chamber in a beam the spectrum of which matches that of the brachytherapy source of interest.
# The replacement and displacement corrections, crepl and cdisp r, account for replacement of absorbing medium by the detector and averaging of dose over the detector volume. They depend on the brachytherapy dose-measurement geometry and are defined as: crepl =
D w at in m edium FS
K w at in em pty cavity occupied by detector
# # 1 # cdisp (r)= D w at (r) atpointr # ∫ D wat (r' )dV ' V (r)V (#r)
(9)
For 30 mg throw-away TLD powder-filled capsules, Weaver (Weaver et al., 1989) estimated crepl values of 0.965 to 0.970 for 125I dosimetry. By averaging 1/r2 over the TLD volume (Meigooni, Meli, and Nath 1988a; Weaver et al., 1989), yield cdisp values of 1.03 and 1.01 for 0.5 cm and 1 cm distances, respectively, for 3×3×0.9 mm3 TLDs and cdisp corrections of 1.06 and 1.02 for the larger TLD capsules at these distances. Using Monte Carlo simulation, Patel and Williamson (Patel et al., 2001) found that # the ratio of the product crepl ⋅ cdisp (r) to that of a perfect point detector to be about 1.05 for 1×1×1 mm3 TLD cubes at distances of 1 cm or greater for 125I seeds in Solid Water. Hence, for low-energy seed dosimetry, these corrections can be significant. The common practice of using published E values that may or may not match the experiment in question should be avoided and is inconsistent with the current TG-43 dosimetry recommendations (Rivard et al., 2004). The 2004 AAPM TG-43 protocol (Rivard et al., 2004) makes many detailed recommendations regarding the experimental determination of TG-43 dose-calculation parameters. While the recommendations are based largely on experience with TLD dosimetry of low-energy interstitial seeds, most are applicable to other detector systems. For measurement of dose-rate constant, AAPM recommends that dose
248
Jeffrey F. Williamson and Mark J. Rivard
measurements be obtained for 8 to 10 different sources, for a total of 15 individual readings at 1 cm on the transverse axis. At least one of the sources should have a directly traceable SK calibration (i.e., have been measured directly by NIST via the WAFAC) and the other experimental sources calibrated against this source by the investigator. For measurement of relative quantities (radial dose and 2-D anisotropy functions), at least three separate experimental runs are recommended using different sources with a total of about 12 readings, including readings in different quadrants assumed to be equivalent by cylindrical or planar symmetry. For most investigational sources, e.g., 241Am and 169Yb, as well the widely used highintensity 192Ir sources, no primary air-kerma strength standards are available. In these cases, it is the investigator’s responsibility to carefully measure SK for these sources using a large-volume ionization chamber that has direct traceability to NIST’s primary air-kerma standards for external beam therapy that bracket the spectrum of the experimental source. A number of investigators (Das et al., 1995; Goetsch et al., 1992) have described procedures for carrying out these measurements. With care, measurements with 2% to 3% uncertainty even for LDR sources are possible.
Brachytherapy Dose-Measurement Systems In the following sections, a number of dose-measurement techniques are briefly reviewed, with an emphasis on their potential for supporting measurement of reference-quality single-source dose-rate measurements. Thermoluminescent Dosimetry (TLD) Techniques in Brachytherapy Currently, TLD dosimetry is the most widely accepted experimental “gold standard” for reference-quality dose measurement in brachytherapy. Dr. John Cameron and his group (Cameron, Suntharalingam, and Kenney 1968) introduced many of the instruments, annealing and handling techniques, and radiation oncology applications that have made TLD a clinically practical dosimetry tool. In 1966, this group published the first TLD dosimetry study in brachytherapy (Ponnunni Kartha, Kenney, and Cameron 1966). As discussed earlier, acceptance of TLD-based brachytherapy dosimetry in the 1990s was due to the efforts of the NIH-supported ICWG (Anderson et al., 1990), the excellent agreement between TLD and Monte Carlo dose-estimation techniques (Williamson 1991), and the emphasis placed by AAPM on quantitative dosimetry in clinical practice standards (Nath et al., 1995; Rivard et al., 2004; Williamson et al., 1998). TLD Handling and Readout Procedures. Although instrumentation and TLD annealing practices vary from investigator to investigator, TLD-100 LiF extruded ribbons subjected to the classical pre-irradiation annealing protocol (annealing TLD detectors at 400 ° C for 1 hour followed by rapid cooling and 24 hours of 80 ° C annealing prior to irradiation) (Zimmerman, Rhyner, and Cameron 1966) is the most common approach. The low-temperature annealing eliminates the short-lived, low-temperature glow-curve peaks so that TLD readings do not depend on the time interval between exposure and reading. Weaver (1984) validated pre-annealed TLD-100 powder. Meigooni has published the most comprehensive study of the influence of annealing and readout procedures on the precision and linearity of TLD-100 detectors (Meigooni et al., 1995). Using a Harshaw 2000 reader, he demonstrated that nitrogen gas purging of the TLD readout chamber dramatically improved the reproducibility of repeated readings in the low dose range and minimized the differences in precision and linearity of TLD response between acute and protracted LDR exposures. In addition, Meigooni showed that the classical pre-irradiation annealing protocol was equivalent to a more time-efficient protocol consisting of 1 hour of 400° annealing before irradiation followed by 10 minutes of 100° annealing just prior to reading (Meigooni et al., 1995). With nitrogen flow and a conventional TLD reader, minimum exposures of 0.5 and 1 cGy are required to maintain precisions of ±5% and ±2%, respectively, for 3 × 3 × 0.9 mm3 detectors. For small (1 × 1 × 1 mm3) TLD chips, the corresponding low-dose limits at are 1 cGy and 15 cGy. Finally,
15–Quantitative Dosimetry Methods for Brachytherapy
249
Meigooni’s work shows that significant deviation from linearity can occur (Figure 6) for low dose exposures, as well as at high doses, especially for small 1 mm TLD cubes. Thus, linearity of TLD response should always be verified across the entire dose-measurement range and any significant corrections included in dose-rate estimates. Most investigators find it necessary to correct for variation of individual detector sensitivities (Cameron, Suntharalingam, and Kenney 1968; Weller et al., 1983) in order to achieve 2% to 4% standard deviations with 4 to 6 readings. Use of detector-specific sensitivity corrections, Si, requires that individual TLD chips be stored in an aluminum or glass-annealing planchette in individually numbered receptacles. In moving TLDs to and from the irradiation phantom to the planchette, care must be taken not to scramble the detectors with respect to their identifying numbers. The Si factors are measured by exposing the entire batch to an identical dose of megavoltage x-rays. The relative detector sensitivity, Si, is defined as Si = TLi
1 N
N
⋅ ∑ TLi
(10)
i= 1
where TLi denotes the reading of the i-th chip. Several sequential uniform irradiations and readings are required to accurately define a table of Si factors. Background readings should be subtracted from each TLi if significant. Some individual chips may never stabilize and should be discarded. With reference to # equation (2), the net mean reading at position r becomes n # # R(r)= 1/n∑ i= 1 (TLi(r)− TLbkgd )/Si
(11)
Linearity of the TLD detectors should never be assumed (Meigooni et al., 1995), especially when measuring doses on the order of a few cGy. Near the low-dose limit, the measured TL may not vary linearly with dose because of surface effects arising from nonradiation induced thermoluminescence or uncertainty in dark-current (zero-dose) TL measurements. Linearity corrections should be carefully
Figure 6. TLD-100 responses (left axis) and linearity corrections (right axis) for graded 4 MV x-ray doses for large (Left panel) and small (right panel) detectors. Nitrogen gas purging was used along with the standard 24 hour 80∞ pre-irradiation annealing protocol. (Reprinted from Med Phys, vol 22, “Instrumentation and dosimeter-size artifacts in quantitative thermoluminescence dosimetry of low-dose fields.” A. S. Meigooni, A. S., V. Mishra, H. Panth, and J. F. Williamson, pp. 555–561. © 1995, with permission from AAPM.)
250
Jeffrey F. Williamson and Mark J. Rivard
measured by irradiating small groups (n = 4–6 chips) of detectors to graded doses of megavoltage xrays over the brachytherapy dose range (generally 1 to 1000 cGy). A plastic phantom (polystyrene or solid-water), which lends itself to accurate determination of absorbed in medium, should be used. During the reading session, the background or “dark current” reading, TLbkgd, should be carefully measured and subtracted from every reading. The first author’s approach is to take TLbkgd as the reading with no detector positioned in the reader. At each dose level k, an average reading
TL
k
= 1/ n
∑ (TL
i
− TLbkgd ) / Si
can be calculated where i is the i-th detector in k-th dose group which is exposed to a dose Dk to the phantom material. From this data, the response ε k = TL k / Dk can be calculated and plotted as a function of TL. The linearity correction, Flin(TL), can be defined as Flin (TLnet ) =
[TL
net
/ D ] at a dose D corresponding too TLnet
[TL
0
/ D0 ] at a reference reading TL0
=
ε (TLnet ) ε (TL0 )
(12)
where TL0 is the net reading that results from an exposure in the linear region of detector response (D0 = 40–100 cGy) in which the correction is normalized to unity. The example in Figure 6 shows that this correction can approach 10% in the brachytherapy dose range. Relative Energy Response of TLD. According to the 2004 TG-43 protocol (Rivard et al., 2004), the relative energy response correction, E(r), is the largest single source of type B (systematic) uncertainty for TLD dosimetry in the low-energy range. For 125I and 103Pd dosimetry, most investigators treat E(r) as a distance-independent constant with a value near 1.4, although when volume-averaging and solidto-liquid water corrections are included, E(r) varies significantly with distance (Patel et al., 2001) (Figure 7, lower left). However, for higher-energy sources, in which the scatter-to-primary ratio rapidly increases with distance, variations in E(r) with distance as large as 10% to 15% have been reported (Meigooni, Meli, and Nath 1988b) (Figure 7, upper right). The standard value of 1.4 represents an average of reported TLD-100 responses to low-energy x-ray beams measured in air, illustrated in Figure 7 (upper left). Both Monte Carlo (Patel et al., 2001) and experimental (Meigooni, Meli, and Nath 1988a) data demonstrate that E(r) values inferred from in-air measurements depend on TLD size. Since freeair measurements relate TLD reading to dose in a void left by removing the chip, a replacement correction (2%-5%), is needed to correct for the phantom material displaced by the detector. Accurate measurement of E(r) is difficult because: (i) photons from the low-energy tail of the Bremsstrahlung spectrum bias the measurements to an unknown extent, (ii) the limited precision of TLD readout, and (iii) the relatively large uncertainty of ion-chamber dosimetry in this energy range. Because of the uncertainties associated with E(r) measurements, some authors have proposed calculating E(r) directly by Monte Carlo simulation (Patel et al., 2001; Vlicenti et al., 1995; Williamson and Meigooni 1995). Although volume-averaging, displacement, and detector self-attenuation corrections can be easily included, the method assumes that TLD response is proportional to energy imparted to the detector (intrinsic linearity), an assumption which has been questioned for some TLD phosphors and annealing and glow-curve analysis techniques (da Rosa and Natte 1988; Tochilin, Goldstein, and Lyman 1968). For the widely used TLD-100 chips, used TLD-100 chips, and Cameron annealing and readout techniques, the evidence for intrinsic linearity is controversial. Das et al. (1986) compared the Monte Carlo and the experimental free-air x-ray beam approaches for four beam qualities ranging down to 19 KeV effective energy. Their measured relative responses (1.42 to 1.50) were in good agreement (relative to stated 4% experimental precision) with measurements reported by other investigators and their own Monte Carlo calculations (Figure 7, lower left). However, a more recent and comprehensive paper by Davis et al. (2003) concludes the opposite, that the measured TLD-100 energy response correction is underestimated by Monte Carlo calculations by 10% to 5% in the 24 to 47 keV energy range. Their
15–Quantitative Dosimetry Methods for Brachytherapy
251
Figure 7. Several plots illustrating properties of the relative energy-response correction. Upper Left: Plot of early (1980-1992) energy-response data (Hartman, Schlegel, and Scharfenberg 1983; Meigooni, Meli, and Nath 1988a; Muench et al., 1991; Weaver 1984) for TLD-100 detectors using calibrated x-ray beams as function of effective energy. Upper Right: Relative energy-response as a function of transverse-axis distance for 192Ir and 169 Yb sources calculated by the first author’s Monte Carlo code. Lower Left: TLD-100 detector relative energy response corrections calculated by the code PTRAN_CCG as a function of transverse-axis distance from a Symmetra 125I source for two compositions of Solid Water™ and two detector sizes (Patel et al., 2001). Lower Right: Comparison of measured to calculated energy-response corrections as a function of effective x-ray beam energy for two relatively recent studies (Das et al., 1996; Davis et al., 2003).
E(r) values, measured for nine beam qualities, have a stated uncertainty of 0.6%. Their data, presented as normalized TL/Kair ratios, suggest that relative energy response corrections for 103Pd–125I spectra are in the range of 1.58 to 1.61, about 10% higher than previously reported values. Clearly, further research is needed to resolve the discrepancy between published E(r) values, to identify the appropriate role for transport calculations in TLD dosimetry, and to reduce the large uncertainty associated with relative energy-response corrections (Rivard et al., 2004). Using a 10% larger energy response correction would
252
Jeffrey F. Williamson and Mark J. Rivard
cause measure dose-rate constants to fall by 10%. This would improve the agreement between Monte Carlo and TLD for some sources (Hedtjarn, Carlsson, and Williamson 2000) but would clearly worsen agreement for others, e.g., 103Pd (Monroe and Williamson 2002). In utilizing measured or Monte Carlo E(r) estimates published by others, the recommendations of the 2004 TG-43 report should be heeded. TLD experimentalists should confirm that the measurement methodology assumed by a published E(r) ratio matches their experimental technique with regard to TLD detector type and size, annealing and readout technique, and megavoltage beam calibration technique. The latter requires accounting for differences in calibration phantom material and dose-specification media used by the experimentalist and assumed by the selected E(r) estimate. The experimentalist should confirm that appropriate volume averaging, displacement, and self-absorption corrections regardless of whether they are included in E(r) or applied separately. Choice of Phantom Material for Brachytherapy Dose Measurements. As noted earlier, one of the most significant sources of measurement error is uncertainty in the relative position of source and detector. Since TLD measurement commonly is based upon sparse sampling of the dose distribution using a relatively small number of independently positioned detectors, the symmetry method (Figure 1) cannot be used to identify the source center. The ICWG investigators controlled positioning artifacts by using plastic phantoms with precisely-machined detector cavities whose location relative to the source can be accurately specified by using digitally-controlled milling machines (see Figure 2). The phantom material selected by the ICWG, which remains widely used to this day, was a commercial water substitute, Solid Water™, manufactured by Gammex-RMI (Middleton, WI). It is a mixture of epoxy resin and calcium carbonate, the proportions of which are designed to reproduce megavoltage beam depth doses in water and the appearance of liquid water on CT imaging. Because early studies (Meigooni, Meli, and Nath 1988a) were interpreted as demonstrating that Solid Water and liquid water were equivalent for low-energy 125 I dose distributions, it was not immediately appreciated that this level of calcium additive overcompensated in the 30 keV range. Using Monte Carlo simulation, Williamson (1991) first showed that use of Solid Water measurement medium underestimated the dose rates in liquid water by factors ranging from 4% at 1 cm to 25% at 10 cm. The AAPM now recommends that all low-energy brachytherapy measurements be corrected for nonwater equivalence of the measurement medium (Rivard et al., 2004). Recently, the measured calcium concentration of Solid Water was found to deviate from the vendor’s specification by as much as 30% (1.6% Ca by weight compared to the vendor’s specification of 2.3%) (Patel et al., 2001), the consequences of which are illustrated by Figure 7 (lower left). Another measurement of Solid Water composition by the same author found that the calcium concentration was lower than the specification by 10% (Chiu-Tsao et al., 2003) which raises concerns over the constancy of its composition. Therefore, when Solid Water or other proprietary water substitutes are used in experimental dosimetry, the AAPM recommends (Rivard et al., 2004) that their atomic composition be measured and these results used via Monte Carlo simulation to derive distance-dependent phantom-to-liquid water corrections. The present authors recommend using high-purity commercial plastics such as polystyrene or polymethylmethacrylate, which have more uniform, reproducible, and better-characterized compositions, for future low-energy, photon-emitting brachytherapy dosimetry studies. However, the plastic-to-water conversion coefficients, which are much larger (Meigooni, Meli, and Nath 1988a) than corresponding Solid Water corrections, need to be carefully evaluated. For 192Ir and higher-energy source dosimetry, absorbed dose is independent of phantom composition (Meli, Meigooni, nd Nath 1988), allowing use of commercial plastics such as PMMA without significant corrections. Summary: Uncertainty of TLD and Role in Clinical Dosimetry. Relatively few of the brachytherapy experimental and computational investigations published prior to 1999 included a rigorous uncertainty analysis. Based upon the approach developed in several more recent studies (Gearheart et al., 2000; Monroe et al., 2001; Monroe and Williamson 2002; Nath and Yue 2000), the AAPM (Rivard et al., 2001) now recommends that all publications claiming to provide reference-quality brachytherapy
15–Quantitative Dosimetry Methods for Brachytherapy
253
dosimetry data include an uncertainty analysis adhering to the approach outlined in NIST Technical Note 1297 (Taylor and Kuyatt 1994). This report recommends using the Law of Propagation of Uncertainty (LPU) to estimate the uncertainty of a quantity y, that has a functional dependence on measured or estimated quantities x1,…xN, as follows: y = f ( x1 ,..., x N ) N −1 N ∂f 2 ∂f ∂f σ = ∑ σ 2 σ x ,x + ∑ ∑ x i = 1 j = i + 1 ∂xi ∂x j i = 1 ∂xi 2
N
2 y
i
i
(13) j
where σxi,xj (assumed zero here) represents the covariance of the two variables. By defining the relative uncertainty propagation factor as %
∂Y
∂x
≡
x ∂Y
Y ∂x
, equation (13) becomes
∂f %σ y = ∑ % %σ x ∂xi i =1 N
2
(
i
)
2
(14)
For the detector-to-dose conversion equation (2), in which the quantities xi are multiplied or divided #
only, equation (14) can be further simplified by noting that % ∂ [ D" wat (r ) / SK ]meas ∂xi = 1 BTx
. The cont-
ributing uncertainties, %σxi are classified by the NIST report as either type A or type B. The former are statistical uncertainties describing the reproducibility of repeated measurements. NIST recommends taking %σxi to be the %standard error about the mean (%standard deviation divided by square root of number of independent observations) for type A uncertainties. Type B uncertainties are defined as “other than type A,” including estimated systematic uncertainties. If a quantity xi is believed with confidence to lie in the interval [a1,a2], a common approach is to assume that xi is uniformly distributed over this
(
interval yielding σ x = ( a2 − a1 ) / 2 3 i
)
(Taylor and Kuyatt 1994). By following these recom-
mendations, the final %σy will approximate the half-width of the 68% confidence interval expressed as a percentage of the mean. Table 2 represents the authors’ estimate of TLD and Monte Carlo (discussed elsewhere in this chapter). The 4% estimate of TLD readout precision is that proposed by Gearheart (Gearheart et al., 2000) and Nath (Nath and Yue 2000), while the 5% uncertainty in E(d) is recommended by AAPM (Rivard et al., 2004). The medium (specific to RMI Solid Water) conversion uncertainty estimate is based upon Figure 7 (lower left) and the assumption that calcium content by weight is a uniform distribution140 bounded by [1.6%, 2.3%]. The final result, that TLD measurement uncertainty varies from 8% to 10%, depending on distance, agrees well with the uncertainty analyses of Gearheart, Nath, and the AAPM ((Gearheart et al., 2000; Nath and Yue 2000; Rivard et al., 2004). In summary, TLD dosimetry, using TLD-100 LiF chips or extruded ribbons, has emerged as the detector offering the best compromise between small size, sensitivity, flat energy response, and ease of accurate positioning: it is currently accepted as the experimental “gold standard” for measurement of absolute dose rates in brachytherapy. Its disadvantages are: labor intensiveness, relatively poor spatial resolution because of practical limits on the number of measurement points, and relatively large total uncertainty dominated by limited readout precision, uncertainties in available energy-response estimates, and what has turned out to be an unfortunate choice of measurement medium. While Monte Carlo simulation has,
254
Jeffrey F. Williamson and Mark J. Rivard Table 2: Uncertainties for 125I Transverse-Axis TLD and Monte Carlo Dose Estimation · → TLD uncertainties in measurement of Dwat (r )/SK for 125I in Solid Water
Component
1 cm distance
5 cm distance
%σxi
Type
%σxi
Type
Repetitive TLD measurements
4%
A
4%
A
TLD calibration (including Linac calibration)
3%
A+B
3%
A+B
Solid-to-liquid water conversion
2%
B
6%
B
Seed and TLD positioning errors (∆d = 100 µm)
2%
B
1%
B
Energy-response correction
5%
B
5%
B
ADCL SK measurement + transfer
2%
B
2%
B
Total combined uncertainty
7.9% 9.5% · → 125 Uncertainties for Monte Carlo estimates Dwat (r )/SK for I in liquid water
Statistics
0.3%
1.0%
Photo ionization cross-sections (∆σPE = 2.3%)
1.5%
4.5%
Seed geometry
2.0%
2.0%
Source energy spectrum
0.1%
0.3%
Total combined uncertainty
2.5%
5.0%
in principle, a much smaller uncertainty, its results depend on the accuracy of the input data. Monte Carlo can anticipate such phenomena as fail of the vendor or NIST calibration procedures to adhere to the definition of air-kerma strength, the presence of contaminant radionuclides on the source, or limited capacity to reconstruct the detailed internal structure of radioactive seeds. TLD measurement, on the other hand, is sensitive to all these effects, provided their dosimetric consequences are large enough to be detected as significant differences. For the reasons discussed above, the AAPM continues to recommend that all marketed low-energy brachytherapy products be subjected to both experimental (TLD or in some cases diode) characterization and Monte Carlo dosimetric characterization (Rivard et al. 2004;Williamson et al., 1998). In practice, Monte Carlo estimates of relative dosimetry parameters (radial dose function and anisotropy functions), verified by measurements within experimental uncertainty, are recommended for clinical use (Rivard et al. 2004) while an average of experimental and Monte Carlo dose-rate constants is recommended. The AAPM report (Rivard et al. 2004) assigns an average uncertainty of 4.8% to the averaged, or “consensus” dose-rate constant. Using its recommended consensus datasets, the AAPM report estimates that the total uncertainties of TG-43 dose calculations at 0.1 cm, 1 cm, and 5 cm are 6.7%, 5.7%, and 7.3%, respectively. This includes estimated uncertainties of the dose-rate constant and radial dose functions and an additional 3% uncertainty associated with transfer of the NIST-traceable SK values by the vendor’s calibration procedure. Diode Dosimetry Techniques in Brachytherapy Diode dosimetry was one of the first quantitative measurement techniques applied to low-energy seed dosimetry. One of the earliest applications by Ling et al. (1983) was measurement of the radial dose function for the newly introduced model 6711 125I seed in 1983. Anisotropy functions derived from diode
15–Quantitative Dosimetry Methods for Brachytherapy
255
measurements followed soon thereafter (Ling et al. 1985; Schell et al., 1987). These data were widely used in clinical practice until the late 1990s, when the community accepted the TG-43 approach. Comparisons of diode measurements to TLD measurements and Monte Carlo calculations (Ling et al., 1985; Williamson 1988, 1991; Williamson and Quintero 1988) revealed good agreement. The 2004 TG-43 report (Rivard et al., 2004) endorses use of diode measurements for the experimental determination of relative (radial dose and anisotropy functions) but not for absolute dose measurements (dose-rate constant). Diode dosimetry has also been applied to higher energy brachytherapy, most notably in the mapping of dose distributions about 137Cs-bearing shielded gynecological applicators (Mohan et al., 1985; Yorke et al., 1987). The most comprehensive set of diode measurements is that of Williamson et al. (Perera et al., 1994; Williamson, Perera, and Li 1993) who used diode measurements to benchmark Monte Carlo calculations for a wide range of brachytherapy photon spectra and geometries. As noted earlier, silicon diodes are among the most sensitive dosimeters used in brachytherapy, providing good signal-to-noise ratios at long distances from low strength brachytherapy sources. Active detector volumes are often very small, providing excellent spatial resolution. Finally, diode dosimeters are designed for use in liquid water medium, avoiding large corrections for non-water measurement media. Diode detectors are most often used with scanning water phantoms designed for external beam dosimetry. The P-type silicon diode, designed by Rikner (Rikner 1985; Rikner and Grusell 1987) for electron-beam measurements and marketed by Scanditronix as the “electron field diode,” or EFD, is the most frequently used detector. P-type diodes such as the EFD must be used cautiously. Because of their low impedance, most integrating electrometers used for ion chamber dosimetry are not appropriate. Most investigators have used the dedicated Scanditronix diode electrometer and readout system. Williamson (Williamson, Perera, and Li 1993) used a Keithley 602 electrometer in the fast current mode to readout EFD signals, obtaining a final reading by timer-driven integration of the strip chart output. Diode readings are very sensitive to small potential differences across the electrometer input. Williamson found that potential differences greater than 0.05 µV were associated with large sensitivity to temperature changes, detector nonlinearity, and unpredictable background currents. By using a constancy-check source, e.g., 90Sr ion chamber check source, a session-to-session reproducibility of 1% can be achieved. The mechanical scanning system must be very carefully calibrated so that the diode position relative to the source center can be estimated accurately. The source symmetry method (Figure 1) is widely used for this purpose. Williamson and colleagues (Williamson, Perera, and Li 1993) have described a mechanical and optical alignment technique that assures a positioning accuracy of 0.02 to 0.03 mm. Diode detectors do respond linearly the instantaneous doses characteristic of linac x-ray beams. Hence a 60Co teletherapy unit is preferable as a calibration source. While most investigators account for the distance-dependence of diode response due to spectral changes for high-energy sources, such as 137Cs, the effect is usually assumed to be negligible for 125I. Li (Li, Williamson, and Perera 1993) evaluated this hypothesis by means of Monte Carlo simulation, which included detailed geometric modeling of EFD internal structure and estimating energy deposition in its 60 µm thick active volume. The ratio of Ddet to silicon kerma for a point detector was found to be independent of distance from a model 6702 source within 1.5%, indicating that variation in the energy or anguMC
lar distribution of photons in water did not significantly affect results. In addition, the Rm eas D det
ratio was found to be constant within 3% over the 1 to 10 cm distance range, although the results show a definite trend towards lower responses at larger distances. In the same study described earlier in our discussion of TLD energy response, Das investigated the dependence of EFD diode response on photon energy (Das et al., 1996). For x-ray and γ-ray sources ranging from 19 keV to 60Co, these investigators found that the fundamental energy response, α = R/Ddet [see equation (5)], was constant within a 3% range.
256
Jeffrey F. Williamson and Mark J. Rivard
The most comprehensive study of diodes in brachytherapy is a series of papers published by the first author’s group (Das et al., 1996, 1997a; Kirov et al., 1995; Li, Williamson, and Perera 1993; Perera, Williamson, and Li 1994; Williamson, Li, and Perera 1993). The major goal of these works was not to characterize brachytherapy dose distributions, but to systematically assess the accuracy of Monte Carlo photon-transport simulation as a clinical dosimetry tool through systematic experimental benchmarking. As illustrated by Figure 8 (lower left and right panels), for relative measurements (Perera, Williamson, and Li 1994; Williamson, Perera, and Li 1993), excellent agreement was found. The average discrepancy between Monte Carlo simulation and EFD measurements ranged from 0.8% to 3% for 125 169 I, Yb, 192Ir, and 137Cs sources both in homogeneous water and in the presence of cylindrical air, titanium, aluminum, steel, and lead heterogeneities of various diameters and thicknesses. Figure 8 (lower left panel), addresses the accuracy of absolute EFD measurements in brachytherapy. For four different radionuclides, the inherent energy response, αBTx(d) is plotted as a function of transverse-axis distance, d, in water:
α BTx (d ) =
( R( d )
( ∆D
det
SK )meas BTx
(d ) ∆SK )MC
BTx
(15)
For different sources and distances, αBTx(d) is an energy-independent constant within ±2.5%. The average of the αBTx(d) agrees with Das’ measurements (Das et al., 1996) of intrinsic energy response within 3%. Thus, available data suggests that absolute dose-rate diode measurements with uncertainties of about 5% are possible, using Monte Carlo estimates of the relative energy response correction. In summary, diode dosimetry has a well-validated role in relative dose measurements about lowenergy brachytherapy sources. Although not widely used for absolute dose-rate measurements in this energy range, further investigation is warranted, as diode dosimetry may have significantly lower measurement uncertainties than those reported for TLD dosimetry. Because of the relatively large variation in detector sensitivity with measurement position, silicon diode is probably not a good choice for higher energy sources. Plastic Scintillator Dosimetry for Brachytherapy Plastic scintillator (PS) consists of a solid polystyrene or polyvinyl toluene base, along with one or more wave-shifting dyes. It converts about 3% of the energy absorbed from ionizing radiation to visible light through a complex multi-step process. PS detectors are approximately tissue equivalent and are much sensitive per unit detector volume than either TLD or ion chamber (see Table 1). In brachytherapy, singleelement PS probes has been intensively investigated by the Essen University group (Bambynek et al. 2000; Fluhs et al., 1996). These probes consist of a small PS detector mounted on the tips of a nonscintillating PMMA-clad light-conducting fiber which transports scintillation photons to a photomultiplier tube. Their system for measuring β-emitting eye plaque or intravascular source dose distributions consists of a 0.4 mm thick by 1 mm diameter BC-400 PS detector (Bambynek et al., 2000). To subtract optical background, which is dominated by Cerenkov radiation above the electron energy threshold of 170 keV, a second coaxial optical fiber that is optically isolated from the detector is readout by a second photomultiplier tube. Excellent linearity and reproducibility of 1.5% to 2% has been reported by several authors (Mourtada et al., 2003; Soares et al., 2001) using this system. Soares et al. (2001) and Mourtada et al, (2003) have calibrated single-element detectors against NIST-traceable sealed β-source standards and have compared absolute dose rates measured by PS to Monte Carlo calculations and other dose-rate measurements about a variety of therapeutic sealed β sources. Their intercomparison assigned an uncertainty of 10% to the PS system, and found that PS results agreed reasonably well with the other
257
Transverse axis dose distance2 (cGy·cm2·h–1)
15–Quantitative Dosimetry Methods for Brachytherapy
Figure 8. Comparison of brachytherapy diode measurements in water to Monte Carlo simulation for radionuclides ranging from 125I to 137Cs. Top: Relative transverse-axis dose rates as measured (data points) and as calculated by Monte Carlo simulation (lines) (Perera et al., 1994; Williamson, Perera, and Li 1993). Note that doses averaged over the active detector volume are plotted. Lower Left: Ratios of measured diode readings in water to dose in the active detector volume calculated by Monte Carlo simulation as a function of transverse axis distance for four different radionuclides (Das et al. 1996). The broken line indicating constant sensitivity is the average absolute detector response measured by Das using calibrated x-ray beams. (Reprinted from Phys Med Biol, vol 41, “Accuracy of Monte Carlo photon transport simulation in characterizing brachytherapy dosimeter energy response artifacts.” R. K. Das, Z. Li, H. Perera, and J. F. Williamson, pp. 995–1006. © 1996, with permission from IOP Publishing, Bristol.) Lower Right: Measured and calculated heterogeneity correction factors (HCF = dose in presence of heterogeneity/dose at same point in homogeneous geometry) for a 169Yb source perturbed by lead disks of various shapes and thicknesses. Replotted from data in Perera et al. (1994).
258
Jeffrey F. Williamson and Mark J. Rivard
measurements. In interpreting such intercomparisons, one must bear in mind that primary dose-rate standards for sealed β-sources utilizing extrapolation chambers themselves have rather large uncertainties of 6% to 10% (Soares, Halpern, and Wang 1998). The first author’s group has proposed using thin PS sheets as 2-D planar dosimeters (Perera et al., 1992) and organic scintillation cocktails to support 3-D dosimetry (Kirov et al., 2000). While the brachytherapy source irradiates the organic scintillator, low-noise CCD optical cameras are used to integrate the 2-D optical yield distribution (repeated at multiple viewing angles for 3-D dosimetry), from which the corresponding dose distribution is inferred. While this work is highly investigational and many problems remain to be solved, 2-D and 3-D organic scintillation offer the potential of on-line, efficient, high-resolution measurement of brachytherapy dose distributions. An important issue for photon-emitting brachytherapy PS dosimetry, is the linear energy transfer (LET) linearity, i.e., constancy of α = R/Ddet over the range of secondary electron energies encountered in brachytherapy. Reduction of PS sensitivity to high-LET radiations has been widely observed and is thought to be caused by ionization quenching (Birks 1970), in which alternative pathways compete with radiative de-excitation (fluorescence) are created by densely ionizing charged particle tracks. Whether ionization quenching or other forms of LET nonlinearity affect PS response to electrons under 100 keV is controversial [see Williamson et al. (1999) for a recent review]. Using methods similar to the previously discussed TLD and diode energy-response investigation (Das, Perera, and Williamson 1996), Williamson et al. (1999), measured the intrinsic response, α = R/Ddet of BC-400 and two experimental PS mixtures to photon sources ranging from 19 keV to 1.25 keV. After correcting for the radiological properties of PS, they found that PS response to 20 and 100 keV photons was 30% and 10% smaller, respectively, than the response to than the response to 192Ir photons. Thus, PS detectors should be used cautiously for measurement of absolute dose-rates about low-energy sources. In summary, plastic scintillator holds promise for brachytherapy dosimetry, both in its technologically well-developed single-element detector format and in the more investigational multidimensional format. However, more quantitative investigation of the accuracy and uncertainty of single-detector systems for photon-source dosimetry is needed before it can be accepted for reference-quality dose-rate or relative dose measurement. Radiochromic Film Dosimetry for Brachytherapy Radiochromic film (RCF) is emerging as a promising secondary dosimeter for 2-D brachytherapy dosimetry because of its high precision, excellent spatial resolution and approximate tissue equivalence. The RCF active sensor consists of diacetylene monomer crystals suspended in a water based gelatin, which in turn is sandwiched between protective Mylar™ sheets (Niroomand-Rad et al. 1998). Ionizing radiation initiates polymerization of the monomer, resulting in complex changes in the optical absorption spectrum, which are approximately proportional to absorbed dose. The signal is readout by means of a scanning densitometer, most commonly using a He-Ne monochromatic (633 nm) light source or redfiltered broadband light source with a mean wavelength of 680 nm. More recently, document scanners (Alva et al., 2002; Mack et al., 2003) has been explored for quantifying film response. The most widely investigated commercial film type, MD-55-2, has a thickness of 240 µm and two 12-µm thick active layers (Klassen, van der Zwan, and Cygler 1997). It is a rather insensitive detector, requiring a dose of nearly 80 Gy to produce a net optical density (OD) change of 1.0 at 633 nm. Recently, a nearly ten-fold more sensitive RCF version has been developed (Lynch et al., 2004), which reduces the useful dose range to under 4 Gy. Its application to quantitative brachytherapy dosimetry remains to be investigated. Most investigations of RCF in photon-emitting brachytherapy were based upon the double active layer RCF model MD-55-2. One of the earliest investigations (Chiu-Tsao et al., 2004) used a predecessor of MD-55-2 to map dose distributions from a 57 µ Gy⋅m2⋅h−1 model 6702 source. An exposure time of two weeks was used. Good qualitative agreement between relative dose rates measured by RCF and TLD
15–Quantitative Dosimetry Methods for Brachytherapy
259
and calculated by Monte Carlo simulation was observed. NIST (Soares 1991) has developed a procedure for using radiochromic film for mapping the relative dose distributions near sealed beta-emitting therapeutic sources. Piessens et al. (2002) found that RCF absolute dose-rate measurements agreed with plastic scintillator probe and TLD measurements within 3.5% at distances of 1 to 3 mm. This study accounted for RCF nonuniformity using the double exposure technique. In a comparison of measured absolute dose rates about an 90Y intravascular. RCF dosimetry exhibits a number of artifacts that must be meticulously controlled if accurate results are to be obtained. Zhu and colleagues at Washington University (Zhu et al., 1997) found that RCF strips exhibited 8% to 15% systematic variations in response for two different scanning wavelengths apparently due to non-uniform application of the sensitive emulsion on the film. These authors devised the so-called “double-exposure” non-uniformity correction technique. Prior to exposing a film to the unknown radiation source, the film is first exposed to a uniform dose of 20 Gy and scanned by a laser densitometer to obtain a pixel-by-pixel relative sensitivity map. The 2-D OD image obtained via scanning following the second exposure is then registered to the flat-field image. The flat-field image is subtracted from the second exposure image and the difference image corrected pixel-by-pixel by means of the relative sensitivity map. Zhu demonstrated that double-exposure densitometry eliminates structural non-uniformity, producing a constant response characterized by 2% to 5% random fluctuations. Despite the insensitivity of MD-55-2 film, the response above 3 Gy was relatively precise, with repeated readings exhibiting a standard deviation of 2% at a spatial resolution of 0.25 mm. Dempsey et al. (1999) characterized another source of artifact arising from the interaction of the film with the widely used 633 nm scanning laser densitometry systems, specifically the Molecular Dynamics Personal Densitometer. Nonreproducible interference fringes, due interference of light reflected from layer boundaries within the film and the glass scanning bed, as large as 7% were observed. These were managed by replacing the scanning bed by a defusing glass plate. Even more disturbing were severe light scattering/temporal system response artifacts. As illustrated by Figure 9, large OD gradients and discontinuities cause OD in dark regions to be underestimated by as much 40% and OD in high-transmission regions adjacent to darker regions to be substantially overestimated. For inverse square-law OD variations characteristic of brachytherapy, Dempsey showed that the system response artifact introduces dose
Figure 9. (a) Unprocessed Molecular Dynamics scans through uniform OD rectangular steps of varying widths, demonstrating OD underestimates as large as 30%. (b) The same scans after filtering and deconvolution (Dempsey et al., 1999).
260
Jeffrey F. Williamson and Mark J. Rivard
underestimates of about 10%. These artifacts were successfully eliminated by a combination of Wiener filtering and deconvolution in light-transmission space, based upon the measured point-spread function of the system. Since potentially any high-speed densitometry system and film (including silver halide) type combination could give rise to such artifacts, bar pattern scanning as recommended by Dempsey should be a routine quality assurance test for any quantitative application of film dosimetry. Once ionizing radiation exposure initiates polymerization, this chemical transformation persists upon removal of the radiation source, resulting in progressive increase of the OD as a function of post-exposure time and temperature (McLaughlin et al., 1991). The so-called “post-exposure OD growth effect” can persist for many months. The OD growth rate is initially quite fast (as large as 20% in 12 hours for 30 Gy exposures) varies systematically with post-exposure temperature history (Reinstein and Gluckman 1999b). In addition, both the growth rate and magnitude of growth vary systematically with dose (from 100% for small doses to 20% for large doses) (Ali et al., 2003; Dempsey et al., 2000). Published (Ali et al., 2005) and unpublished (Le and Williamson) data demonstrates that fractionated and protracted low dose-rate exposures, especially at low doses, result in moderately elevated (2% to 5%) OD responses, relative to single-fraction exposures, that take several hundred hours to dissipate. Thus the potential for significant dose-measurement errors is large if the acute calibration exposures, needed to measure the dose-response curve, and the brachytherapy exposures are not temporally synchronized. Despite these technical problems, the potential of radiochromic film to significantly reduce the uncertainty of brachytherapy dose measurements is high. After correcting for RCF nonuniformity and densitometer artifacts, as described above, the reproducibility (95% confidence interval) of uniformly irradiated MD-55-2 doubly-exposed film was 2% for a 0.4 × 0.4 mm2 pixel at 2 Gy and 1.5% % for a 0.8 × 0.8 mm2 pixel at 1 Gy (Dempsey et al., 2000). At doses greater than 10 Gy, even single 100 µm pixel readings have a precision better than 1%. Dempsey and colleagues (Dempsey et al., 2000) at Washington University benchmarked RCF absolute dose measurements against Monte Carlo calculations for an HDR 192Ir source. The HDR and 6 MV calibration exposures were temporally synchronized, minimizing post-exposure OD growth artifacts. Double-exposure densitometry (Zhu et al., 1997) and densitometer system response corrections (Dempsey et al., 1999) were employed. Between doses of 10 and 150 Gy, 95% of the measured doses (100 µm resolution) agreed with the Monte Carlo benchmark within 4%. Even at low doses of 1 to 5 Gy away from the source cable, the Monte Carlo-RCF was 1% to 2% (2σ). The authors estimated that the total uncertainty of the measured absolute dose rates was 3.6% to 4.2% (coverage factor of two) for single-pixel readings at doses above 10 Gy. This is the lowest uncertainty reported for a brachytherapy dose measurement using a secondary dosimeter. Two major problems must be solved before RCF dosimetry can be applied to determination of reference-quality dose-rate distributions for low-energy brachytherapy. Accurate relative energy-response measurements for low energy x-rays are needed to confirm their independence from LET established. In an investigation patterned after the Das et al. TLD-diode study (Das, Perera, and Williamson 1996), Bohm et al. (Bohm, Pearson, and Das 2001) found that [OD/Drcf] was a constant, within ±5%, independent of photon energy in the 30 keV to 1.25 MeV energy range. Although MCNP photon-electron simulations were used to calculate dose to the active film layer, the precision of the experiment was estimated to be about 7%. Monroe et al. (2001) compared doses measured by RCF dosimetry to Monte Carlo doses for a balloon applicator filled with organically bound 125I solution. They found excellent agreement between the two techniques for high doses, indirectly confirming the energy-response correction derived from Monte Carlo calculations. While there is no theoretical basis for questioning the linearity of RCF response with respect to energy deposition, more accurate measurements and more detailed Monte Carlo calculations are needed. Finally, as described above, the problem of post-exposure growth, which gives rise to discrepancies between acutely exposed calibration films and films subjected to protracted LDR exposures, must be solved. The solutions advanced to date include (a) modeling of postexposure growth artifacts (Ali et al., 2005), (b) annealing of RCF to accelerate convergence of
15–Quantitative Dosimetry Methods for Brachytherapy
261
post-exposure OD growth (Reinstein, Gluckman, and Meek 1998), and (c) selection of densitometry wavelengths that minimize growth artifacts (Meigooni and Williamson, unpublished data). As noted above, the new high-sensitivity film may address some of these issues. While some problems remain to be solved, RCF dosimetry is the best developed of the emerging dosimetry systems and has well-documented performance advantages compared to TLD. Polymer Gel Dosimetry for Brachytherapy Since the 1950s, liquid gelatin materials have been used as tissue equivalent media for experimental dosimetry (Frigerio 1962; White 1978). However, consideration of gels as 3-D radiation detectors did not occur until the 1980s (Gore, Kang, and Schulz 1984). In this first decade of use, a variety of radiotherapy applications were considered (Schulz et al., 1990; Maryanski et al., 1993). Readout techniques also developed with some groups using nuclear magnetic resonance (NMR) imaging while others pursuing optical imaging techniques (Gore et al., 1996; Oldham et al., 2003; Oldham and Kim 2004) due to concerns of widespread MRI scanner unavailability (Oldham et al., 2001). Furthermore, gel dosimeters may be divided into three classes of detectors: Fricke gels (MacDougall et al., 2002; De Deene, De Wagter, and Baldock 2003); polymer gels (Maryanski et al., 1994); and superheated drop detectors (Lamba et al., 1998; d’Errico et al., 2002). This section will review application of polymer gels to HDR and LDR brachytherapy sources using both readout techniques. Free radicals are produced during irradiation of polymer gel dosimeters, with average distances of approximately 0.2 µm and 0.02 µm at doses of 0.1 and 100 Gy, respectively. In less than 1 ns, these radicals then form cross-linked micropolymers which both increase the gel opacity and cause local changes in the polymer structure which subsequently disturb the NMR relaxation rates. Thus, optical and NMR imaging techniques are suitable to perform detector readout to glean the distribution of radiation damage. While there are a variety of polymer gel dosimeter mixtures available such as the various recipes of BANG (Maryanski et al., 1996; Maryanski 2005), VIPAR (Baras et al., 2002; Kipouros et al., 2003), PABIG (Fong et al., 2001), and MAGIC (Pantelis et al., 2004), the mechanism of damage and micropolymerization are all similar since they all strive for tissue-equivalence over the photon energy range (10 keV to 10 MeV) used in radiation therapy. Among conventional radiation detectors used in the field of radiation therapy, polymer gels have the unique capability to acquire and render 3-D radiation dose distributions. With modern polymer gels, spatial resolution is generally limited to 2 mm in all directions by the readout technique and the need to balance signal intensity with diminishing detector sensitivity. Optical readout may be performed using a variety of light sources such as He-Ne laser at 633 nm or a laser diode at 670 nm. However, maximum opacity peaks at approximately 450 nm. Optical computed tomography (optical CT) reconstruction techniques may be implemented as is commonly performed for diagnostic radiology CT (Xu, Wuu, and Marayanski 2004). A photo of an optical CT scanner is presented in Figure 10. Conventional MRI units (~1.5 T) can be used to readout the nuclear magnetic transverse relaxation rate, T2 signal, with 1 < TR < 8 s and 20 < TE < 1,600 ms. Depending on the desired spatial resolution and signal-to-noise ratio, high-quality results using MRI readout can be typically obtained in less than one hour. As most applied research in this field occurs in an academic setting, availability of MRI for detectors was not a limitation and the majority of studies performed to date used MRI instead of optical means for detector readout. The suitability of polymer gels as radiation dosimeters for brachytherapy is evidenced by considering detector response as a function of various parameters. As long as the irradiation is kept to less than one hour, polymer gel detectors typically exhibit negligible dose rate dependence over the range used in radiation therapy—including the high dose rates used in HDR 192Ir brachytherapy. Furthermore, detector response as a function of post-irradiation time to readout does not significantly as this time is kept less than 30 days (Farajollah et al., 1999; Xu, Wuu, and Marayanski 2004). With an offset to account
262
Jeffrey F. Williamson and Mark J. Rivard
Figure 10. The OCTOPUS™ optical CT scanner for polymer gel readout. (Photograph courtesy of M. Maryanski.)
for intrinsic opacity, polymer gel detector response as a function of delivered dose is quite linear. The explicit range depends on detector size and readout technique, but linearity ±3% over a range of 1 to 20 Gy may be expected. Slightly larger variations in flatness may be expected for photons with energy ranging from 10 keV to 10 MeV (Pantelis et al., 2004). Because of polymer thermodynamics, detector response is somewhat dependent on irradiation temperature. For a dose of 7 Gy and temperature change from 23 °C to 25 °C, polymer gel detector efficiency increases by approximately 7% when readout using MRI (Spevacek et al., 2001). The majority of studies performed using polymer gels as dosimeters for brachytherapy have been to compare dose distributions of HDR 192Ir sources using other experimental) or Monte Carlo techniques to benchmark the gel detector. McJury and colleagues (McJury et al., 1999) compared measured dose distributions using the BANG gel with calculated dose distributions using an HDR 192Ir brachytherapy treatment planning system (PLATO v.11.3 from Nucletron BV, Veenendaal, The Netherlands) based on Monte Carlo measurements by Williamson and Li (Williamson and Li 1995) and measurements by Mishra and colleagues (Mishra, Waterman, and Suntharalingam 1998). The HDR 192Ir source was centrally positioned in a 500 cm3 Pyrex flask (see Figure 11) to deliver 6 Gy at 1.0 cm over 111 seconds as determined using the treatment-planning system (TPS). Separate vials containing polymer gel were prepared for absolute dose calibration using a linac 6 MV photon beam. MRI readout occurred 24 hours after irradiation using TR = 2 s, TE = 0.05 s, 2 mm slice thickness, 1 × 1 mm2 pixel size, and lasted 25 minutes. Over a radial range of to 50 mm, the average difference between the measured and TPS results was 0.17 Gy with a standard deviation of 0.13 Gy. At a distance of 1.0 cm, a dose of 6.5±0.2 Gy was measured using the polymer gel detector. This was considered good agreement since a positional uncertainty of 0.5 pixel corresponded to an absorbed dose error of ±0.8 Gy. Papagiannis and colleagues (Papagiannis et al., 2001) performed a similar study using the same type of HDR 192Ir brachytherapy source, planning system, and calibration beam. Even with subtle differences (TR = 4 s, TE = 0.04 s, readout 48 hours post-irradiation, 0.43 × 0.43 mm2 pixel size, 3-mm slice thickness, etc.), similar results were obtained. At distances less than 0.5 cm, differences in Monte Carlo and poly gel doses ranged from 10% to 40%. At larger distances, substantially better agreement was observed.
15–Quantitative Dosimetry Methods for Brachytherapy
263
Figure 11. Polymer gel dosimeter flask for measurement of HDR 192Ir source dose distribution. (Reprinted from McJury et al., “Experimental 3D dosimetry around a high-dose-rate clinical 192Ir source using a polyamide gel (PAG) dosimeter,” Phys Med Biol, vol 44. Copyright 1999, with permission from IOP Publishing)
However, there was concern that a practical and lucid technique for performing an uncertainty analysis was not available for polymer gel dosimeters (Oldham et al., 1998; Baldock, Murray, and Kron 1999). This concern was later addressed (MacDougall et al., 2002; De Deene, De Wagter, and Baldock 2003). De Deene and colleagues (De Deene, Reynaert, and De Wagter 2001) also assessed the accuracy of using polymer gel dosimeters in close proximity to an HDR 192Ir source. The methodological details were similar to the previous two studies, but emphasis was placed on the impact of oxygen and its role on inhibiting radiation damage. They observed 2% lower dose at 1.0 cm that estimated using Monte Carlo methods (Williamson and Li 1995) with a root mean square (RMS) uncertainty of 5.6%. In retrospect, positioning of the source within 30 mm of the flask inner wall might have significantly minimized the scatter and resultant absorbed dose at distances greater than 10 mm (Perez-Calatayud, Granero, and Ballester 2004). For gels containing oxygen concentrations ranging from 0 to 0.8 mg O2 per liter of gel, a linear increase in the threshold dose was observed with a slope of 28.2 Gy⋅L⋅mg–1. Thus, the threshold dose may be kept under 0.2 Gy as long as the oxygen concentration remains less than 0.007 mg/L. Additional studies have been performed to obtain polymer gel detector response for other brachytherapy sources. Farajollah and colleagues (Farajollah et al., 1999) measured the dose rate from 36 Nucletron Selectron LDR 137Cs brachytherapy sources using TLDs and BANG gel with MRI readout. At 2 cm, the average gel dose rate was 10% larger than calculated. Dose rates measured using TLDs from two different institutions were 3% low and 2% high. Considering the source calibration uncertainty of ±5% and other uncertainties such as applicator tube attenuation, both the gel and TLD results were in good agreement with the calculated dose rate. Fragoso and colleagues (Fragoso et al., 2004) also measured dose rate in the vicinity of Selectron LDR 137Cs sources, but this time EGSnrc was used to complement gel measurements with Monte Carlo calculations to account for applicator tube attenuation. Monte Carlo voxels were 1 × 1 × 1 mm3, and 54 M photon histories were calculated. A 1.5 T MRI with TR = 2 s and 50 < TE < 800 ms was used for gel readout. At distances beyond 1 cm and doses less than
264
Jeffrey F. Williamson and Mark J. Rivard
10 Gy, agreement between Monte Carlo calculations and polymer gel measured doses were typically within 5%, which was well within the experimental uncertainties. Hasson (1998) also measured dose distributions near 192Ir and 137Cs sources, and obtained agreement within 3% for dose distributions when compared to radiochromic film and Monte Carlo methods. Farajollah performed a similar study (Farajollah 1999) and obtained similar results. Ibbott and colleagues (Ibbott et al., 1999) measured the dose rate of an LDR 125I source towards determining brachytherapy formalism parameters, but did not give quantitative information as to the agreement of results using other techniques. Hafeli and colleagues (Hafeli et al., 2000) compared polymer gel dose measurements of a 188W/188Re intravascular brachytherapy source (IVBT, Nath et al., 1999) to an EGS4-based Monte Carlo calculation and measurements using TLD rods and radiochromic film from 0.5 to 5.0 mm. Normalized to the calculated dose rate at 2.0 mm, the, gel, TLD, and film dose rates were 1.01, 0.89, and 1.20, respectively. Volume averaging and positioning uncertainties subtended the largest components of the overall experimental uncertainties. Chan and colleagues (Chan et al., 2001) measured dose distributions in the vicinity of a 15.5 mm diameter LDR 106Ru eye plaque. Calibrations used four 22 mm diameter vials and irradiation with 6 MV to doses of 0.00, 1.00, 2.00, and 3.00 Gy. MRI readout was performed using TR = 2 s, TE = 20ms or 100 ms, 3 mm slice thickness, and 0.31 mm in-plane pixel size. Though 3-D dose distribution data were obtained, only central axis dose data were presented, and results were not compared to established techniques. In summary, use of polymer gel dosimeters as relative instruments having impressive 3-D capabilities is well established for a variety of radiotherapy applications, including the diverse scope of brachytherapy. However, additional research is needed to validate polymer gel dosimeters as an absolute detector for the demanding application of brachytherapy dosimetry.
Computational Dosimetry Methods in Brachytherapy As noted in the introduction, computational dosimetry methods have come to play an important role in the estimation of reference-quality dose-rate distributions used to implement TG-43 dose calculations in clinical practice. Indeed, AAPM guidelines (Williamson et al., 1998) specify that every low-energy interstitial seed product used in routine clinical practice should be supported by both measurements and dose-rate distributions computed by means of Monte Carlo simulation. Eligible measured or computed reference-quality dose distributions must meet the standards of the peer-reviewed publication process. Most of the AAPM-endorsed consensus TG-43 datasets (Rivard et al., 2004) were based upon purely Monte Carlo estimates of radial dose and anisotropy functions. Computational dosimetry methods fall into one of two broad categories: semiempirical dose-calculation algorithms and numerical solutions of the Boltzmann transport equation (BTE). Semiempirical algorithms are reviewed in chapter 14 of this volume and are not discussed further in this chapter. These algorithms are based upon heuristic approximations that support very efficient dose calculations for treatment planning. While very useful within limited domains, the accuracy of such algorithms must be established by benchmarking them against more rigorous computational or empirical dosimetry methodologies. Rigorous numerical BTE solutions collectively form one such class of methodologies and is the topic of this section. In principle, the BTE gives a complete mathematical description of the motion of particles through a host medium when the particle density is small relative to that of the host medium. When applied to the interaction of ionizing radiation with matter, the stationary (time independent) BTE takes the following form (Duderstadt and Martin 1979; Williamson 1988): $ # # ˆ ⋅ ∇Φ(r$, Ω, E ) + µ ( r$, E ) Φ(r$, Ω, E ) Ω # # # $ # $ # = ∫∫ Φ(r , Ω ', E ') ⋅ µ ( E ', Ω ' → E , Ω | r ) ⋅ dΩ ' dE ' + S (r , Ω, E )
(16)
15–Quantitative Dosimetry Methods for Brachytherapy
265
$ # # Φ(r,Ω,E ) is the angular flux (particles per unit where angle and energy at location r, # area,#solid $ # µ r,E is the linear attenuation coefficient, µ(E ',Ω '→ E ,Ω |r) is the differential cross # # section E ' , Ω' t o E, Ω giving the number of particles scattering from energy and angular trajectory per unit $ # # fluence, and S(r,Ω,E ) is the source # term, giving the number of particles arising at r per unit volume with direction and energy E and Ω . Equation (16) is essentially a balance equation. Its four terms, proceeding from left to right describe passage of particles along straight-line trajectories without collision, losses of particles due to collisions, gains of particles due to inscattering, and gains due to primary particle# production. of the problem geometry (i.e., specifying cross sections # # Given a description $ µ(E ',Ω '→ E ,Ω |r) and µ r,E as a function of spatial location), the primary source term and other boundary conditions, equation (16) can be solved for angular fluence, giving a complete characterization of the radiation distribution that accounts for particle transport, scattering, and absorption. Unfortunately, equation (16) is too complex to be solved accurately by analytic methods in any but the simplest of 1- and 2-D geometries. General deterministic numerical solutions, e.g., discrete ordinates, have been developed (Lewis and Miller 1993) mainly for neutron transport applications. As deterministic BTE solutions are increasingly used for solving photon transport problems in brachytherapy and other radiotherapy applications, they will be briefly discussed in this chapter. However, the most widely used numerical methodology used in brachytherapy dosimetry is stochastic or Monte Carlo solution of the BTE. In this approach, (16) is converted to an integral equation, which is solved by summing the quantity of interest over a randomly generated set of integrand values, called particle histories. Even in the presence of a complex 3D geometry, Monte Carlo will give statistical uncertain but unbiased estimates of that converge to the “exact” BTE solution in the limit of large number of histories.
(
)
(
)
Monte Carlo-based Brachytherapy Dosimetry: General Considerations The basic principles of Monte Carlo simulation have been extensively described elsewhere (Jenkins, Nelson, and Rindi 1988; Williamson 1988) and will be only briefly reviewed here. A 3D geometric description of all sources and attenuating media in system is required. A discrete-event Monte Carlo simulation involves randomly constructing a set of photon tracks or “histories” (see Figure 12(b)), each consisting of a series of successive collisions or interactions of the photon with matter connected by straight-line flight paths. The point of origin of each history is randomly selected from the radioactive distribution (usually assumed to be uniformly distributed in some volume or surface). Its trajectory in space is randomly selected from an isotropic distribution and its energy from the photon emission spectrum of the radionuclide. By ray tracing along the trajectory to find mean free path as a function of distance, the site of the first collision can be selected from an exponential distribution. The type of collision is randomly selected from the relative frequencies of photoelectric, coherent scattering, incoherent scattering, and other competing interaction types (see Figure 12(a)). The photon direction and trajectory of the scattered photon emerging from the collision are sampled from the normalized differential cross section. This process of randomly simulating serial collisions continues until the photon history is terminated by absorption, escape from the system, or Russian roulette. This process is repeated for 100,000 to many millions of histories depending on the variance reduction techniques employed. During random construction of each collision, the history is scored, i.e., the actual or estimated energy deposition to each simulated “detector” is computed and tallied, a process called “estimation.” The simplest estimation process is the so-called analog estimator, illustrated by Figure 12(b), in which energy is deposited in a detector volume if and only if a collision occurs in that spatial volume. Both the mean dose and variance about the dose are tallied. By central limit theorem, the confidence interval, describing the statistical uncertainty of the sample mean dose, is inversely proportional to the square root of the number of histories. In principle, it is straightforward to include transport of secondary electrons in the simulation, which is clearly necessary to solve problems in the megavoltage photon domain. However, including electron
266
Jeffrey F. Williamson and Mark J. Rivard
Figure 12. (a) Principle mechanisms of photon scattering and energy deposition in the 10 keV to 10 meV energy range, illustrating the various photon scattering processes and secondary photon production processes that a Monte Carlo photon transport code must simulate. From Williamson (1988d) with permission. (b) 2-D illustration of a photon history in a plane bisecting a low-energy seed.
transport into the simulation reduces the efficiency of the calculation by two orders of magnitude, limiting the simulation to relatively simple 1-D and 2-D geometries. In contrast, Monte-Carlo photon-transport (MCPT) problems of great complexity can be readily solved on PCs or workstations within a few hours. MCPT approximates absorbed dose in medium by collision kerma, which is a valid only when secondary charged-particle equilibrium (CPE) obtains. Based on theoretical considerations, Roesch first hypothesized that inverse-square law gradients could induce a state of electron disequilibrium giving rise to a buildup effect at short distances from brachytherapy sources wherein collision kerma could significantly overestimate absorbed dose (Roesch 1958). Several recent Monte Carlo studies (Luxton and Jozsef 1999; Nath, Yue, and Liu 1999; Wang and Li 2002) have confirmed this hypothesis. For example, Luxton (Luxton and Jozsef 1999) compared buildup factors calculated by EGS4 with those calculated by Berger (1968). For 600 and 800 keV sources, the dose to kerma (D/K) ratio is about 0.5 at 0.5 mm distance and effects of 2% persist out to 1.5 and 2 mm, respectively. Such effects occur at lower energies (Ling and Yorke 1989). Because D/K deviations scale with secondary electron range, extremely thin scoring shells at very small distances are needed to resolve them. Wang (Wang and Li 2002) recently used EGSnrc to investigate secondary particle disequilibrium effects for two commercial HDR 192Ir sources, including the effects of encapsulation and β-ray transport as well as secondary electron transport. In the 0.5 to 1 mm distance range, the combined effects of two electron transport components was as large as 13% and 29% for the current Nucletron and old VariSource source models, respectively. The D/K ratio was 1.07 at 1 mm and essentially unity at distances ≥ 2 mm. When secondary electron transport alone was simulated D/K ranged from 1.03 to 1.08 in the 0.5 to 1.0 mm range. Thus simulation of both photons and secondary charged particles is important for clinical applications, e.g., intravascular brachytherapy, in which near-zone doses are important. In addition, the older orthovoltage and 60Co dosimetry literature (Dutreix and Bernard 1966; Gray 1940; Paterson 1942) documents that significant deviations from D/K = 1 occur at media interfaces, e.g., metal applicator-tissue interfaces. In addition to calculating clinically meaningful doses to soft-tissue cavities in bone, or mucosa in contact with metal applicators, secondaryelectron disequilibrium effects need to be considered when simulating the response of thin detectors. Aside from these important exceptions, we can conclude that secondary-electron transport and CPE-failure effects are relatively unimportant for interstitial and intracavitary brachytherapy, and that neglecting these effects, as does MCPT simulation, leads to less than 1% errors in most clinical settings.
15–Quantitative Dosimetry Methods for Brachytherapy
267
Technical Aspects of Monte-Carlo Photon-Transport Simulation in Brachytherapy Accurate and reliable MCPT-based dosimetry requires attention to many technical details such as use of appropriate physical models for photon interactions and cross-section library; accuracy of the underlying geometric model; and use of appropriate estimators and other variance-reduction techniques. Each of these issues will be reviewed in the following sections. Physics of Photon Scattering and Choice of Cross-Section Library As noted above, a MCPT code requires total cross sections for each photon collision process, tabulated as a function of incident photon energy and medium. In addition, mass-energy absorption coefficients maybe needed to convert energy fluence into estimates of absorbed dose in the charged-particle equilibrium approximation. Many public-domain codes (e.g., ITS and MORSE) neglect the influence of electron binding on photon scattering, approximating coherent scattering by uncollided primary photons (by subtracting µcoh from µtot) and incoherent scattering by the Klein-Nishina differential cross section, which assumes that the target electrons are free and at rest. Electron binding corrections are usually made by multiplying the Thompson and Klein-Nishina differential cross sections by atomic form factors (AFFs) and incoherent scattering factors (ISFs), respectively, giving rise to the incoherent and coherent scattering cross sections. The tabulations of AFFs and ISFs used by up-to-date cross section libraries are those of Hubbell, et al. (Hubbell and Øverbø 1979; Hubbell et al., 1975). Williamson (Williamson, Deibel, and Morin 1984) suggests that the free-electron approximation Klein-Nishina (KN) is of questionable accuracy below energies of 100 keV. The binding of target atomic electrons to nuclei significantly modifies the angular distribution of scattered photons, particularly in high-atomic number media. Although binding effects have less influence on energy deposition, the free-electron approximation was found to over estimate energy fluence transmitted through thick barriers by 2% to 15% in the low-energy range. A more clinically relevant comparison (Figure 13), shows that the choice of collisional physics
Figure 13: Ratio of transverse-axis dose for the model 6711 125I seed computed by various photon interaction models, to dose computed based upon incoherent scattering using the Hubbell’s incoherent scattering functions and coherent scattering in the atomic form-factor approximation. ISF only = incoherent scattering with coherent scattering omitted; KN only = free-electron Klein-Nishina scattering; KN + CS = electron Klein-Nishina scattering with coherent scattering; IA + CS and BIA + CS = coherent scattering with two approximations to Compton Doppler broadening. (unpublished data: C. Costescu and J. Williamson).
268
Jeffrey F. Williamson and Mark J. Rivard
model modifies model 6711 transverse-axis dose distribution by less than 2% at distances less than 4 cm. An interesting conclusion is that KN alone is much more accurate than KN with coherent scattering, the model used with EGS4 with KEK extensions. For 20 keV point sources and for doses near the longitudinal axis of 125I seeds, Costescu’s and Williamson’s unpublished data shows much larger 5% to 8% variations with model choice. Incorporating the more complex Doppler broadening model of incoherent scattering (broadening of the discrete Compton scattering energy at each angle into a continuous spectrum due to the orbital electron momentum distribution) influences brachytherapy dose distributions by less than 2%. However, mass-energy absorption coefficients used to convert energy fluence into dose must be consistent with the physical model used to transport photons to avoid systematic errors. For example, if KN only is used to transport photons, both the µ and µen/ρ tables must be recalculated for the KN model. Below 100 keV, both the authors and the AAPM recommend that coherent scattering be explicitly simulated and that the Compton free-electron distribution be corrected for orbital-electron binding effects via ISF factors. If for reasons of efficiency, a less accurate model, e.g. KN-only, must be used, this approach should be carefully benchmarked against the standard photon-collision model for the domain of interest. For liquid-water medium, biological tissues, and certain plastics, coherent scattering, in contrast to other interaction mechanisms, is strongly influenced by molecular bonds and inter-molecular forces. For water, these extra-atomic forces increase the value of the coherent-scattering cross section by 20% to 40% in the 10 keV to 40 keV energy region compared to the predictions of the mixture-rule model, which accounts only for atomic binding forces. For liquid-water, Morin’s (1982) AFF tabulation is still the most comprehensive tabulation. An important secondary scattering mechanism in brachytherapy is characteristic x-ray production following ejection of an orbital electron. While any collision process that ejects an orbital electron can initiate the atomic relaxation process and the resultant cascade of characteristic x-rays and Coster-Konig or Auger electrons, photoelectric collisions in medium- and high-atomic number materials dominates. For example, approximately 25% of the photon fluence emitted by the model 6711 125I seed is due to emission of characteristic x rays within the silver wire upon which the radioactive material is adsorbed (Ling et al., 1983; Williamson 2002). The significant contribution of low-energy (4.5 keV) characteristic x-rays emitted by the titanium encapsulation of 125I seeds to the free-air exposure rate has been experimentally observed by Kubo (1985) and theoretically demonstrated by Williamson (1988b) to significantly influence the 1985 NIST iodine seed calibration standard (Loftus 1984). Using a simplified atomic relaxation process and tabulated vacancy probabilities, transition probabilities, and fluorescent yields of Plechaty (Plechaty, Cullen, and Howerton 1978), Williamson has shown good agreement between his Monte Carlo calculations and spectroscopic measurements by NIST for Ag characteristic x-rays (Williamson 2002) and by Kubo (1985) for Ti x-rays. Recently, the more up-to-date and complete Lawrence Livermore Laboratory EADL data base (Perkins and Cullen 1991) has become available. EADL tabulates all radiative and nonradiative transitions for each subshell of each element from Z=1 to Z=100. Future low-energy brachytherapy simulations should utilize EADL-compatible data. Simulation of atomic relaxation in the EGSnrc code is now based upon EADL data (Kawrakow and Rogers 2003). Accurate MCPT-aided brachytherapy dosimetry would not be possible without the evolution of accurate photon cross-section libraries made over the last 30 years (Hubbell 1999). The success of these efforts owes much to John Hubbell of NIST who has systematically compiled and critically evaluated both measured and theoretically calculated cross-sections to provide evermore accurate and comprehensive cross-section libraries. During this evolution, photo ionization cross sections for light elements in the 20 to 40 keV energy range have changed significantly. Several investigators (Bohm, DeLuca, and DeWerd 2003; Demarco, Wallace, and Boedeker 2002; Williamson 1991) have demonstrated that 103Pd and 125I brachytherapy dose distributions are quite sensitive to choice of cross-section library and that
15–Quantitative Dosimetry Methods for Brachytherapy
269
use of obsolete cross sections can introduce errors of 5% to 10% or more. Hence we will briefly review the evolution of cross-section libraries and their use in popular public-domain codes. Virtually all modern libraries use very similar scattering cross sections, based upon the Hubbell’s nonrelativistic incoherent scattering and relativistic atomic form factors (Hubbell and Øverbø 1979; Hubbell et al., 1975) as described above. The major differences from one library to another reside in the origin of photo ionization cross sections. In describing the different libraries, we will use the Radiation Safety Information Computational Center (RSICC) Data Library Code (DLC) nomenclature for distinguishing the different libraries (see Table 3). RSICC, located at Oak Ridge National Laboratory, packages and distributes transport codes and data libraries for nominal fees. With regard to treatment of photo ionization, three broad groups of libraries are currently used in public-domain codes. The Storm and Israel (1970) photoelectric cross sections, DLC-15, are based upon the theoretical calculations of Rakavy and Ron (1967) using screened atom models of Schmickley and Pratt (1967). These cross sections continue to be used by EGS4 and EGSnrc, despite up to 4% deviations in the low-energy brachytherapy range relative to currently-recommended libraries (Figure 14). The second family of libraries, represented by DLC-7F, is based upon the semi-empirical photo-ionization cross sections of McMasters et.al. (1969), which were derived from least-squares curve fitting to an extensive collection of theoretical and measured data (Ca 1968) weighted according to their assessed accuracy. The experimental cross-section data had generally been obtained by subtracting the theoretically calculated scattering cross sections from the measured total cross sections. These photoeffect cross sections underestimate absorption by nearly 10% in the 20 to 30 keV energy range in water (Bohm, DeLuca, and DeWerd 2003; Demarco, Wallace, and Boedeker 2002; Williamson 1991) relative to more current libraries. Despite these large errors, these cross sections have been used in the MCNP cross-section libraries DLC-200 and DLC-189 (MCNP libraries MCPLIB01, MCPLIB02, and MCPLIB03) since 1982. The third, and currently recommended library family introduced in 1983, uses the theoretical photo-effect cross sections of Scofield (1973) based upon a relativistic Hartree-Slater model renormalized to the Hartree-Fock model for low atomic-number
Table 3. Major Characteristics of Photon Cross-Section Libraries Distributed by RSICC DLC designation
Year of origin
Data sources
Client Codes
DLC-15
1971
Storm and Israel. Older theoretical treatments of photo-ionization and scattering
EGS4, EGSnrc
DLC-7F
1975
Current scattering and semi-empirical MCNP 4b and 4c McMasters photo ionization cross-sections
DLC-189 DLC-200
1997 2000
RSICC MNCP libraries. DLC-7F cross sections. Includes MCPLIB 01 and 02
DLC-99
1983
HUGO library. Hubbell form factors, normalized Scofield photo-ionization cross sections
DLC-146
1989
Same as DLC-99 except photo-ionization normalization dropped
DLC-136 DLC-174
1989 1997
NIST PHOTX and XCOM databases. Identical to DLC-146
DLC-179
1997
EPDL97 library including EADL atomic relaxation data. Identical to DLC-146, except subshell cross sections included
MCNP 4b and 4c PTRAN prior to 2000
PTRAN after 2000
MCNP library MCPLIB04 (LANL website) since 2002
270
Jeffrey F. Williamson and Mark J. Rivard
Figure 14. (Left) Ratio of photo-ionization cross sections from various libraries to current NIST XCOM database for water as a function of photon energy. (Reprinted from Phys Med Biol, vol 47, “An analysis of MCNP cross-sections and tally methods for low-energy photon emitters,” J. J. Demarco, R. E. Wallace, and K. Boedeker, pp.1321–1332. © 2002, with permission from IOP Publishing, Bristol.) (Right) Ratio dose of dose rate on transverse-axis of a Model 2000 103Pd seed as calculated by MCNP/DLC-179 to MCNP/DLC-200. (Reprinted from Med Phys, vol 30, “Brachytherapy dosimetry of 125I and 103Pd sources using an updated cross section library for the MCNP Monte Carlo transport code,” T. D. Bohm, P. M. DeLuca Jr., and L. A. DeWerd, pp. 701–711. © 2003, with permission from AAPM.)
elements (Das et al., 1997b). Extensive comparisons (Saloman and Hubbell 1986, 1987) of Scofield’s calculations with the NIST experimental cross-section data base, including many recent “direct” measurements of photo-ionization, demonstrated that this theoretical approach was more accurate than competing approaches. The initial library, DLC-99 was used in the first author’s code, PTRAN_CCG through 2000. Further intercomparisons revealed that unnormalized Scofield calculations further improved the match between tabulated and measured cross sections Hubbell (1999). Hence, the current generation of crosslibraries (DLC-146, 136, and 189), introduced in 1989, incorporate this improvement. Below 1 MeV in water, dropping the normalization increases the photoelectric cross section by 2.7%. Essentially, all current libraries, including the NIST PHOTX/XCOM and EPDL97 tabulations all use the same theoretical approaches to calculating scattering and photo-ionization cross sections. The LANL website indicates that an MCNP EPDL97-equivalent library, MCPLIB04 has been available since 2002. Figure 14 (left) compares the three families of photoelectric cross sections to the current NIST XCOM cross sections. Mass-energy absorption coefficients, compatible with post-DLC-146 libraries, have been published by Seltzer (1993) and can be downloaded from the NIST website1. Several papers demonstrate that use of obsolete cross section libraries significantly compromises the accuracy of 125I and 103Pd TG-43 dosimetric parameters. Bohm et al. (Bohm, DeLuca, and DeWerd 2003) demonstrated that use of default MCNP cross sections (DLC-200), overestimates dose by 4% at 1 cm to 20% at 5 cm. For analog dose estimation, DeMarco et al. (Demarco, Wallace, and Boedeker 2002) show similar results (4% to 8% errors in dose-rate constant estimation in 20 to 40 keV energy range) when default MNCP cross sections are used and somewhat smaller (2.5%) errors when default EGS (DLC-15) cross sections are used. DeMarco also demonstrates that large and unpredictable dose-estimation errors arise when mass-energy absorption coefficients based upon DLC-146 are used to convert particle fluence scores into dose for MCNP simulations using DLC 15 or 200 for constructing histories. Hedtjarn et al. (Hedtjarn, Carlsson, and Williamson 2000) shows that the smaller differences between 1
http://physics.nist.gov/PhysRefData/XrayMassCoef/cover.html.
15–Quantitative Dosimetry Methods for Brachytherapy
271
the DLC 99 and 146 libraries introduce dose-calculation errors ranging from 1% at 1 cm to 3.5% at 5 cm for an 125I seed. The updated TG-43 report recommends (Rivard et al., 2004) using DLC146/XCOM/EPDL97-equivalent cross sections for reference-quality dosimetry and will not accept as candidate data sets Monte Carlo simulations based upon default MCNP cross sections. Geometric Modeling and Validation Accurate MCPT dose calculation requires a flexible and general system of geometric modeling. Brachytherapy sources and applicators have complex internal designs often leading to highly anisotropic dose distributions. In addition, dose distributions are sensitive to the shape and size of the surrounding scattering medium. Calculation of absolute dose rates may require simulation of the experimental geometry used to standardize air-kerma strength for the source. Finally, comparison of MCPT predictions with measured results may require simulation of detector response. Simulation of detectors such as commercially available silicon diodes poses a difficult modeling challenge, since these detectors are small and may have intricate internal structures. The approach of choice is surface- (Briesmeister 1997) or volume (Li and Williamson 1992)-based combinatorial modeling. This approach, sometimes called complex combinatorial geometry or CCG, defines complex spatial regions as set-theoretic intersections, unions and differences of elementary volumes such as cuboids, ellipsoids, elliptic cylinders, cones and half-planes. The modeling code must allow the composition, location, and dimensions of each geometric region to be independently specified, must allow complex structures to be nested inside of one another, and must support point classification, line-segment classification, and ray tracing. Most of the public-domain codes support CCG modeling packages. Exceptions are EGS4 and EGSnrc (Kawrakow and Rogers 2003) which effectively limit users to modeling relatively simple cylindrically symmetric and or 3-D Cartesian slab geometries supported by the user packages DOSrz and DOSxyz. Nevertheless, a number of high- and lowenergy sources can be adequately modeled by the DOSrz package. Of the codes widely used in brachytherapy, MCNP has the most general and sophisticated geometric modeling capability. Because low-energy seed dose distributions are so sensitive to the assumed internal structure of seed, it is essential to validate through measurements seed and, by extension, applicator and detector geometries. Both authors have found vendor-supplied mechanical drawings of sources to be misleading and sometimes inaccurate. Tools for verifying seed geometry include contact transmission radiographs, micrometer measurements, microscopic examination of unassembled seed components, and pin-hole autoradiography (Kirov et al. 1995b). Some investigators (Hedtjarn, Carlsson, and Williamson 2002; Rivard 2001) have used milling machines to longitudinally “slice” open seeds along their symmetry axis. Figure 15 illustrates three investigations in which such measurements significantly influenced the geometric model adopted. In Figure 15(a), transmission contact radiographs indicated that the inverted cup ends were spherical in shape rather than cylindrical as represented in the vendor’s literature. Examination of unassembled end-weld components in a dissection microscope enabled the authors to devise a detailed model (Monroe and Williamson 2002). In the case of the DRAXIMAGE LS-1 seed, Figure 15(b), such radiographs documented that the gold localization rod and the two radioactive spheres were mobile, assuming different configurations depending on the seed orientation. Because the thickness of hemispherical end caps (see Figure 15(c)) is thinner (0.05 mm) than the cylindrical sections of the capsule (0.10 mm), the motion of the radioactive beads significantly modulated the seed output. For clinical use, the first author (Williamson 2002) recommended a weighted average of dose distributions corresponding to different geometric configurations. This strategy was first proposed by Rivard (2001), who noted (Figure 15(d)) that the radioactive and nonradioactive spherical beads in the North American Scientific 125 I seed (Model MED3631-A/M) assumed a staggered array configuration that deviated from cylindrical symmetry. Impact of this effect using this principle on non-uniformly loaded sources was examined further (Rivard, Kirk, and Leal 2005).
272
Jeffrey F. Williamson and Mark J. Rivard
(a)
(b)
(c)
(d)
Figure 15. (a) Contact radiograph and final geometric model of the Theragenics Model 200 103Pd seed (Monroe and Williamson 2002); (b) contact transmission radiograph and final geometric model of the DRAXIMAGE LS-1 125I seed with its axis vertical illustrating mechanical shift of the internal structures (Williamson 2000); and (c) 200X magnified electron microscopy image of the end of a Ag rod from a Model 6711 125I seed illustrating that it illustrating previously unappreciated beveled rather than right circular cylinder shape. (d) Geometry of the model 3631-A/M seed when aligned vertically, illustrating the staggered configuration of the internal spherical elements and a cut-away photograph of the seed (in horizontal orientation) that has been mechanically sliced open (Rivard 2001).
15–Quantitative Dosimetry Methods for Brachytherapy
273
Choice of Estimator and Other Variance Reduction Techniques An estimator is a computational device for extracting a statistical estimate of the quantity of interest from a simulated photon trajectory. Since MCPT simulations assume that energy transferred from photon fields to matter is absorbed locally, the quantity of interest in brachytherapy is collision kerma. Kerma rates calculated by MCPT codes are generally expressed in units of cGy/simulated primary photon or history. Using the photon spectrum assumed by the simulation, it is straightforward to convert kerma/emitted primary photon into kerma rate per unit contained activity, represented by ∆K = K" /A with
cont
M C PT
units of cGy⋅mCi ⋅h . Thus the source strength specification quantity intrinsic to the MCPT method is contained activity. In the sections below, the process of normalizing MCPT ∆K estimates to clinically relevant source-strength specification quantities will be described. The choice of estimator and method of implementation depends on whether dose-at-a-geometric point or dose averaged over an active detector volume is desired. Simulation efficiency (statistical precision achieved for a given CPU time), the proximity of media boundaries to the point or volume of interest, and the spatial resolution desired are also important considerations. Each of the sections below will review the estimators of importance to brachytherapy and their principal applications. The simplest device for improving spatial resolution and computational efficiency, especially at short distances, is to deterministically calculate the primary-photon collision kerma component rather than to stochastically score randomly constructed primary-photon trajectories. This reduces the variance of the primary dose component to zero. The calculation consists of straightforward 3-D numerical integration # # − ∑ µ ⋅% # −2 ⋅ (µen /ρ )⋅ r , over the distribution of radioactivity, ρ(r), in terms of of the operator, ρ(r)⋅ e –1
−1
mCi/cm3. By limiting Monte-Carlo simulation to scoring of scattered-photon trajectories, the variance of the total-dose estimates near sources can be reduced by a factor of two. Numerous other variance-reduction techniques have been developed and implemented which can greatly reduce the CPU time required to achieve the desired statistical precision for complex problems. Common strategies (Jenkins, elson, and Rindi 1988; Lewis and Miller 1993; Williamson 1988c) include Russian Roulette and splitting, forced scattering, and biased angle sampling. Each of these techniques uses biased sampling distributions in conjunction with correction factors designed to eliminate any bias in the sample mean of the quantity of interest. These sampling distributions and photon-weight correction factors are cleverly designed so as to force the computer to preferentially focus CPU effort on simulated random events that contribute significantly to the quantity of interest rather than those that do not. Spherically Symmetric Tracklength Estimator. This estimator is applicable only to spherically symmetric 1-D problems in which the collision kerma rate depends only on the radial distance, r, from the origin. The most commonly studied problem is that of an isotropic point source positioned at the origin in unbounded homogeneous medium. ∆K(r) is estimated by tallying those photon trajectories that cross an imaginary spherical detector surface of radius r centered on the source. A complete technical description is given by Williamson (1987). Major advantages of this approach are simplicity, computational efficiency, and that it gives an unbiased estimate of the dose rate at a geometric point. Freedom from volume-averaging artifacts follows from the fact that an infinitely thin scoring surface rather than a finite-thickness scoring volume is used to simulate a point detector. Many of the early MCPT studies in brachytherapy, e.g., Dale (1983), are based upon the spherically-symmetric tracklength estimator. Analogue and Tracklength Estimators. Most MCPT applications utilize so-called analog estimators to estimate absorbed dose or collision kerma. This approach consists of tallying energy transfers from those photon collisions that occur in an detector volume, V, centered about the point of interest. The resultant estimated dose is simply the quotient of energy transferred to the detector volume by its # mass. Thus for collisions j such rj ∈V, the score is (Wj−1Ej−1 – WjEj)/ρV where Wj is the weight of the
274
Jeffrey F. Williamson and Mark J. Rivard
photon. In MCNP, analogue scoring is called the F4 tally. In heterogeneous geometries, care must be taken to ensure that the detector contains homogeneous medium, i.e., that its boundaries do not intersect media interfaces. To use analog estimation to evaluate dose at a point, the detector volume is centered on the point of interest and dose at the detector center approximated by dose averaged over its volume. This approach must be used cautiously in brachytherapy. If dose gradients are large relative to the detector dimensions, spatial resolution will be lost due to volume averaging. If the detector volume is constrained to be small by nearby media interfaces or the need for high spatial resolution, efficiency may be unacceptably low since only a small fraction of the simulated collisions will occur within the detector. The tracklength estimator, which, like the analogue estimator, also gives an unbiased estimate to the energy absorbed by the detector volume, V, greatly increases simulation efficiency by analytically estimating the energy transferred to the detector volume by all photon flight paths that cross the detector (Williamson 1987). Where ∆% is the distance traveled in V by a photon of energy Ej and weight Wj, the score is then W j E j ⋅ ∆% ⋅ ( µ en / ρ ) V . MCNP calls this estimator the *F8 tally. An even more efficient variant of tracklength scoring, expected-value tracklength estimation (Williamson 1987), scores all photons whose projected trajectories intersect the detector. In simple single spherical detector geometries, use of tracklength scoring can improve efficiency by factors of 7–10. In CT voxel detector grids, containing many thousands of detectors, expected-value tracklength scoring increases efficiency by factors of 20–50 relative to analogue scoring if an efficient voxel tray tracing algorithm (Siddon 1985) is available. If only scattered-photon trajectories are stochastically scored, volume averaging bias (average dose in V/dose at geometric center of V) for spherical detectors will be limited to 1% if the detector radius is limited to 60% of the source-detector distance or 15% of the primary photon mean-free path, whichever is smaller (Williamson 1987). Because of its greater efficiency, tracklength scoring should always be used in preference to analogue scoring in photon-transport simulations. The most important application of the volume detector estimators is simulation of detector response in brachytherapy dosimetry. In this situation, one wants to model detector artifacts such volume averaging and medium displacement. Thus an analogue or tracklength estimator is required. To simulate the response of a TLD dosimeter placed at a small distance d from a brachytherapy source, a LiF-filled rectangular prism detector having the same # dimensions as the actual TLD should be used. This MCPT simulation will yield the quantity ∆K det (r ) ; the energy transferred to a TLD-size cube#of LiF per unit # mass of LiF centered at r per contained mCi·h. By using analog estimation, the ∆K det (r ) estimate will include the effects of volume averaging, directional anisotropy, and self-attenuation. To correlate this theoretical endpoint with the measurable response of the TLD, the fundamental assumption, # # TLlin (r ) ∝ ∆Ddet (r ) , (see equation (5) must be made. Using Burlin cavity theory, as described above, to # # convert ∆K det (r ) into ∆Ddet (r ) , the relative energy response [equation (7)] and other quantities needed to convert detector readings into dose rates can be evaluated by Monte Carlo simulation. Analytic Point-Dose Estimators. In evaluating reference dose-rate distributions for brachytherapy sources, in contrast to simulating detector response, the quantity of interest is dose at a geometric point. In this application, the volume averaging characteristic of tracklength or analogue estimation presents a source of artifact no different than that faced by the experimentalist. Analytic estimators seek to overcome this conflict between spatial resolution and computational efficiency by forcing every simulated collision to contribute to the quantity of interest at the desired geometric point. This is done by analytically estimating the probable energy transfer contribution made by each simulated photon collision to each of the specified target points. Although analytic point-dose estimators are very complex to implement and greatly increase the computing time required/simulated collision, overall efficiency is often dramatically improved by increasing the number of simulated events that contribute to the sample mean. The simplest analytic point-dose scoring technique, next flight estimation (Kalos 1963), has been used successfully by the author and his colleagues for many years in computational brachytherapy dosimetry
15–Quantitative Dosimetry Methods for Brachytherapy
275
(Williamson 1987). For each simulated collision, next flight calculates the probability that the scattered photon emerging from the collision intersects the point detector and transfers energy to the medium at its location. This requires correction for loss of scattered photon flux emerging from the collision site by inverse square-law and attenuation through the intervening media separating the collision and the detector. Unfortunately the (collision-to-detector distance)−2 singularity in the estimator destroys the statistical stability of the dose estimate, since a single collision occurring close to the detector may make a huge but random dose contribution. Although the history-averaged estimate is bounded and unbiased, the second moment of the distribution does not exist, the variance of the estimate cannot be computed, and the simulation converges slower than central limit theorem (Kalos 1963). This problem can be solved by replacing the e−µr r−2 term in the estimator by its average value within a small critical sphere surrounding the point detector (Williamson 1987). This modified “bounded” next-flight estimator solves the problem of accurately calculating kerma at a point in a dilute medium such as air found in calibration geometries. In addition, bounded next flight works well in condensed media so long as the point of interest remains 1 to 2 mm from any media interface and the critical averaging sphere encloses a single medium. It significantly increases efficiency for low-energy sources and is the estimator of choice for 125I dosimetry. Williamson (1987) has given a rules of thumb for the critical radius, R: R should be set at 30% of the source-to-detector distance or 1 mean-free path (whichever is smallest) to ensure bias of less than 1% when primary dose is computed analytically. Although bounded next flight scoring is very powerful, it has several limitations. To eliminate the 1/r2 singularity, the detector point must be enclosed by a sphere of homogeneous medium. If this sphere is made too small, the variance will not converge (will not decrease with in proportion to number of simulated histories). Thus the standard deviation calculated over the number of histories simulated cannot be used to predict statistical precision of the mean dose rate, and the dose rate estimate will not be reliable. Problems in which media interfaces severely constrain the averaging radius, e.g., dose near an applicator-tissue interface and dose to a small detector, cannot be solved with bounded next flight estimation. A more powerful generalization of next flight estimation, once-more collided-flux estimation (OMCFE) (Kalli and Cashwell 1977), was modified and validated for use in brachytherapy MCPT simulations by Li et al. (Li, Williamson, and Perera 1993). This method avoids the need for averaging over a homogeneous region, allowing the point detector to be placed arbitrarily close to a media boundary. Whenever a photon collision occurs near the point of interest and results in an excessively high next-flight score, the position and trajectory of this collision are randomly resampled from biased probability distributions and next-flight estimation applied to the resampled collision site. The biased probability distributions are cleverly chosen so that the resultant weight correction factor cancels out the 1/r2 singularity. The result is an unbiased estimate of kerma at a point and a 30-fold improvement in efficiency compared to analog estimation applied to typical detector sizes. An important application of OMCFE is calculation of ∆Kdet for small TLD or diode detectors at large (>20 mm) source-detector distances where the efficiencies of analog and tracklength estimation are poor, volume averaging artifacts are small, and dose to the detector volume is well approximated by dose at its geometric center.
Application of the Monte Carlo Method to Calculation of Reference-Quality Dose-Rate Distributions # To calculate clinically relevant absolute dose rates, D" wat (r ), in water or other condensed media, the native # MCPT normalization of ∆K wat (r ) , dose per simulated collision (or equivalently dose per contained mCi-h), must be changed to air-kerma strength. This is achieved by running two simulations: one simulation with the seed model located at the center of a 30 cm liquid-water phantom for the purpose of calculating the kerma-rate distribution, ∆K wat (r, θ ), and a second simulation in which the seed model is placed in a vacuum or large air sphere for the purpose of calculating the air kerma-strength per mCi-h
276
Jeffrey F. Williamson and Mark J. Rivard
(or history), ∆SK. Given these two quantities, appropriately normalized absolute dose rates can be estimated:
[ D"
wat
(r, θ ) SK ]MCPT = BRx
∆K wat (r, θ ) ∆SK
(17)
assuming that ∆Dwat(r,θ) ≈ ∆Kwat(r,θ) is a sufficiently accurate approximation. Calculation of the various TG-43 dose ratios is straightforward:
Λ=
∆K wat (r = 1 cm, θ = π / 2 )
gL (r ) =
∆SK
(18)
∆K wat (r, π / 2 ) ⋅ GL (1 cm, π / 2 ) ∆K wat (1 cm, π / 2 ) ⋅ GL (r, π / 2 )
To calculate the distribution, ∆Kwat(r,θ), the senior author (JFW) typically calculates polar angle dose distributions at distances r = 0.25, 0.5, 0.75, 1.0, 1.5, 2.0, 3.0, 4.0, 5.0, and 7.0 cm using an angular mesh with 1° increments near the longitudinal and transverse axes to a maximum of 5° increments at intermediate angles. For transverse-axis dose distributions, a fine grid (0.1 to 0.2 cm increments) is used in the 0.1 to 1 cm distance range, and 0.5 to 2.5 cm steps used for larger distances. The recent TG-43 report specifies minimum spatial resolutions and ranges for Monte Carlo calculations (Rivard et al., 2004). Typically, the bounded next flight estimator is used at distances greater than 0.5 cm and the OMCFE at shorter distances to avoid intersection of the scoring spheres with the seed structure. Simulation of 500,000 to 20,000,000 histories, depending on distance and radionuclide, is sufficient to guarantee statistical precisions better than 0.5% near the source and 1.5%-2% far from the source. Generally, the author always calculates primary kerma rate deterministically. There are two approaches to estimating the effective air-kerma strength constant, ∆SK, for a source: the point-extrapolation method and the WAFAC method. In the more widely used “point-extrapolation” approach (Hedtjarn, Carlsson, and Williamson 2000; Williamson 2000), the seed model is placed at the center of 2-to 5-m diameter sphere of dry air. Then, using the bounded next flight estimator, the transverse-axis dose-rate distribution, ∆Kair(d), is calculated over the 1.0 to 150 cm at distances range. To ensure that these results are not contaminated by low energy characteristic x-rays from Ti, Ti characteristic x-ray production should be suppressed or a 5 keV energy cutoff employed by the estimator. Then, the quantity (∆SK)extr is evaluated by fitting the ∆Kair(d) data to the following equation:
∆ K air (d ) ⋅ d 2 = ( ∆SK )extr ⋅ (1 + α d + β d 2 ) ⋅ e− µ d
(19)
where the fitting parameter, µ, describes primary photon attenuation in air, while the best-fit parameters, α and β describe the buildup of scattered photons. This idealized geometry excludes all scattering surfaces and approximates the air-kerma detector by a geometric point, as specified by the definition of air-kerma strength. Figure 16(b) illustrates the fit of equation (19) to Monte Carlo data for the three seed geometries described by Figure 16(a). In the case of the Model 200 seed, there is a dramatic breakdown of inverse-square law that persists out to 20 cm distance. At long distances, oblique filtration through the thin palladium metal layer reduces the contribution of radioactivity distributed in the four circular end surfaces of the two carbon pellets. At shorter distances, primary photon trajectories are less oblique
15–Quantitative Dosimetry Methods for Brachytherapy
(a)
(b)
(c)
(d)
277
Figure 16. (a) Assumed geometry of the model 200 103Pd seed, the model 6702 125I seed, and the model 6711 125 I seed. The two graphite pellets inside the model 200 source are thought to be coated with palladium metal 2 µm in thickness for current accelerator-produced product. The model 6711 seed contains a solid Ag right circular cylinder, which is coated with radioiodine; (b) Fit of equation (19) to the transverse-axis air-kerma distribution; (c) the polar air-kerma profile in air at 30 cm distance; (d) the 2-D anisotropy function at 1 cm distance in liquid water.
278
Jeffrey F. Williamson and Mark J. Rivard
and traverse thinner layers of Pd, causing air-kerma to increase faster than inverse square law (Monroe and Williamson 2002; Williamson 2000). Screening of radioactivity as illustrated by Figure 16, can occur whenever radioactivity is dispersed within or on the surface of a high-density core with sharp corners or edges. When this phenomenon occurs, the polar dose profile may contain discontinuities or become extremely sensitive to alignment of the interior seed components. As illustrated by Figure 16(c), radioactivity screening introduces significant polar anisotropy (“overshoot”) near the transverse axes of the models 6711 and 200 seeds. This invalidates use of the point-extrapolation method since the WAFAC chamber integrates over a 16° cone centered on the transverse plane. As shown by Monroe and Williamson (Monroe and Williamson 2002; Williamson 2000), use of (∆SK)extr introduces a 15% variation into Λ with respect to Pd metal layer thickness and worsens the agreement with Λ measurements. Theoretically, the WAFAC appears not accurately realize air-kerma strength for sources with transverse-plane anisotropy, since SK is defined in terms of kerma to a geometric point. On the other hand, had source strength been specified in terms of point-kerma measurements, the calibration uncertainty would have been much larger. In cases where radioactivity screening introduces real or apparent anisotropy near the transverse plane, the AAPM (Rivard et al., 2004) recommends using the WAFAC method of ∆SK estimation, in which more of the geometric detail of the NIST primary standard is explicitly modeled in the simulation. As described in more detail in papers from Williamson’s group (Monroe and Williamson 2002; Williamson 2000, 2002), the WAFAC collimator, shielding barriers, and surrounding walls, floor and ceiling are added to source-air sphere geometry. As shown in Figure 17, the WAFAC chamber is operated in two active
Figure 17. Side view if the Wide Angle Free-Air Chamber (WAFAC) illustrating use of the short 11 mm and long 153 mm cylindrical electrodes. The chamber is cylindrically symmetric about the axis bisecting the seed center, which is placed on a rotating holder during measurements to average over any equatorial anisotropy. The primary collimator located just upstream of the Al filter is not illustrated.
15–Quantitative Dosimetry Methods for Brachytherapy
279
volume configurations, denoted by the subscripts “153” and “11,” which correspond to the length of the active ionization volume in mm as defined by the separation of its two circular electrodes. The subtraction of the charge measured by the 11 mm long chamber, from that measured by the full-length chamber geometry is intended to reduce the effects of wall scattering and electric field non-uniformity (Seltzer et al., 2003). Using an expected value track-length estimator, simulation of 25 million histories typically yields ab estimates of ∆ Et , the energy absorption rate within the WAFAC collection volume of length t, with statistical precisions of 0.3%. In this case, both primary and scattered photons must be scored stochastically, in order to include volume averaging and other affects. Then (∆SK)WFC can be evaluated by
( ∆SK )WFC =
( ∆ E ab − ∆ E ab )d 2 153
11
ρair (V153 − V11 )
kinv ⋅ katt = ( ∆ K air ⋅ d 2 )WFC ⋅ kinv ⋅ katt
(20)
where ρair is the density of air and kinv and katt are inverse square-law and attenuation corrections defined by Seltzer (Seltzer et al., 2003). Because Ti x-ray production is suppressed, no correction of the Al filter used by NIST is needed. Similarly, all of the other correction factors required by NIST’s WAFAC measurement protocol are unity by virtue of the Monte Carlo simulation design. The factor kinv is needed because the aperture is a flat plane rather than a spherical element. Thus, the transverse-axis fluence at distance d is underestimated by the fluence averaged over the WAFAC aperture area. By assuming an isotropic point source, the photon fluence can be analytically averaged over the aperture area yielding (Daskalov and Williamson (2001) −1
kinv
R 2 π R 2 2π %d % R2 . 2 = 2 ⋅ ∫ 2 = R d ⋅ ln 1 + 2 2 d d 0 d + %
(21)
The factor katt in (20) corrects for attenuation between the source and the aperture and within the collecting volume. It is evaluated by exploiting the fact that point-extrapolation and WAFAC ∆SK simulations yield identical results for any source with an isotropic fluence distribution over the WAFAC collection volume. Both WAFAC and point-extrapolation simulations are performed for a point source of 103Pd or 125I, yielding values of (∆SK)extr and (∆Kair⋅d2)WFC. Since, (∆SK)extr = (∆SK)WFC, katt is given by
katt =
( ∆SK )extr
for a poinnt source
(22)
kinv ⋅ ( ∆ K ⋅ d 2 )WFC Equation (22) yields katt values of 1.023 (Monroe and Williamson 2002) and 1.013 (Williamson 2002) for 103Pd and 125I, respectively, which are very close to the values of 1.028 and 1.013 published by NIST (Seltzer et al., 2003). Using the WAFAC method of SK normalization, excellent results have been achieved by the Williamson and his group. The Model 200 103Pd seed dose-rate constant is nearly independent of palladium metal layer thickness and is close to the theoretical unfiltered point-source Λ (Monroe and Williamson 2002; Williamson 2000). The final calculated Λ for the current light-seed geometry, (0.691) is very close to measured value (0.68) measured by Nath et al. (2000). For the model 6711 seed, the
280
Jeffrey F. Williamson and Mark J. Rivard
WAFAC method lowers the computed Λ by 3.7% to 0.925 compared to 0.959 for the point-extrapolation method (Williamson 2002). Uncertainty of Monte Carlo Reference Dose-Rate Distributions Relatively little work on estimating the total uncertainty of Monte Carlo results has been published. Based upon partial uncertainty analyses in Williamson’s papers, the updated TG-43 report (Rivard et al., 2004) endeavored to make a more comprehensive uncertainty analysis for 125I seed transverse-plane dose calculations using NIST’s uncertainty estimation protocol (Taylor and Kuyatt 1994). The reader is referred to section Summary: Uncertainty of TLD and Role in Clinical Dosimetry of this chapter for an introduction to the NIST Report 1297 (Taylor and Kuyatt 1994) methodology. For Monte Carlo estimates of quantity, Y (= Λ or g(r), etc.), TG-43 considered the total per cent uncertainty, %σY, to consist of three sources: type B uncertainty due to uncertainty of the underlying cross sections, %σYµ; type B uncertainties arising from uncertainty of the seed geometric model, %σYgeo; and the type A statistical uncertainty, %σYs inherent to the Monte Carlo technique. Applying equation (13), one obtains:
∂Y ∂Y %σ Y2 | geo + %σ Y2 | s = % %σ µ2 + % ∂µ ∂geo 2
%σ Y = %σ
2 Y |µ
+ %σ
2 Y | geo
+ %σ
2 Y |s
where the relative uncertainty propagation factor is defined as %∂Y /∂x ≡
2
(23)
x ∂Y
. The variable x denotes Y ∂x either the cross-section value, µ, or geometric dimension, geo, of interest. The uncertainties estimated here are standard uncertainties, having a coverage factor of unity, approximating a 68% level of confidence. The influence of cross-section uncertainty was derived from Hedtjarn’s Monte Carlo data (Hedtjarn, Carlsson, and Williamson 2000) which gives Monte Carlo estimates of Λ and g(r) calculated for two different cross-section libraries, DLC-99 (circa 1983) and DLC-146 (1995). The photoelectric cross sections of the two libraries differ by about 2% between 1 to 40 keV, corresponding to a 1.1% change in µ for the mean photon energy emitted by 125I. Using these data to numerically estimate the derivative in equation (17), one obtains %∂Λ/∂µ = 0.68. Assuming that %σΛ = 2% (Hubbell 1999), then uncertainty in Λ due only to cross-section uncertainty, %σΛµ, is 1.4%. Estimation of geometric uncertainty, %σΛG, is complex and poorly understood. Each source design is characterized by numerous and unique geometric parameters, most of which have unknown and potentially correlated probability distributions. However, a few papers in the literature report parametric studies, in which the sensitivity of dosimetric parameters to specified sources of geometric variability is documented. For example, Williamson (2002) has shown that range of distances between the two radioactive spherical pellets of the DraxImage 125I source (3.50 to 3.77 mm) corresponds to a 5% variation in calculated dose-rate constant. Rivard (2001) and Williamson (2000) published similar finding for the NASI model MED3631-A/M 125I and the model 200 103Pd sources, respectively. If this phenomenon is modeled by a Type B rectangular distribution bounded by the minimum and maximum Λ values, the standard uncertainty is given by:
%σ Λ | geo = 100
Λ max − Λ min
(24)
2Λ 3
For the DraxImage source, (equation (24) yields a %σΛ|geo = 1.4%. The TG-43 report took 2% to be a reasonable and conservative estimate of %σΛ|geo. The reported statistical precision of Monte Carlo Λ estimates ranges from 0.5% for Williamson’s recent studies to 3% for Rivard’s MED3631 A/M study.
15–Quantitative Dosimetry Methods for Brachytherapy
281
Thus for a typical Williamson study, one obtains a %σΛ of 2.5%. Using the %σΛ|s reported by each investigator for the sources included in TG-43, %σΛ varied from 2.5% to 3.7% for the eight seeds considered in that report. The methods above can be extended to other calculation distances. As distance increases from 2 to 5 cm, progressively increases from 0.2% to 4.6%, respectively. As summarized in Table 2, the contributions from geometric uncertainty are reduced but those due to statistical and spectral uncertainties increase. The final conclusion, summarized in Table 2, is that with Williamson-level statistical uncertainties, 1-σ uncertainties of MCPT simulation range from 2.5% at 1 cm to about 5% at 5 cm. % σthe D" (r )|µ However, much more systematic study of Monte Carlo uncertainties is indicated. Additional Recommendations for Use of Monte Carlo Simulation in Preparing Reference-Quality Dose-Rate Distributions The updated AAPM TG-43 protocol (Rivard et al., 2004) makes numerous recommendations regarding use of photon-transport simulations for low-energy brachytherapy dosimetry. Recommendations discussed above in detail include use of XCOM-compatible cross-section libraries, characteristics of an adequate collisional physics model, and the need to explicitly model seed internal geometry. The report also makes recommendations on maximum statistical uncertainties and volume-averaging artifacts, the need for a formal uncertainty analysis, and documentation to be included in peer-reviewed publications. Regarding choice of codes, AAPM recommends three particular codes (EGS, MCNP, and Williamson’s PTRAN code) that have been widely used in brachytherapy and benchmarked against experimental measurements or each other. Other Monte Carlo codes and other types of transport equation solutions (multigroup codes, discrete ordinates codes, and integral transport methods) not previously used in brachytherapy dosimetry, “should be more rigorously tested and documented in the peer-reviewed literature before proposing to use their results clinically.” Perhaps the single most important recommendation relevant to Monte Carlo is best quoted directly “However, regardless of the transport code chosen and its pedigree, all investigators should assure themselves that they are able to reproduce previously published dose distributions for at least one widely used brachytherapy source model. This exercise should be repeated whenever new features of the code are explored, upon installing a new code version, or as part of orienting a new user.”
Other Applications of Transport Codes in Brachytherapy This chapter has emphasized application of Monte Carlo simulation to estimation of single- low-energy source dose-rate distributions for use in treatment planning. However, radiation transport codes have been used much more broadly in brachytherapy. We will only briefly touch upon some of these uses. • Application of Monte Carlo simulation to estimation of single-source dose distributions for 192Ir (Williamson 1995) and 137Cs (Fragoso et al., 2004; Williamson 1998)sources • Validation and parameter estimation for semi-empirical dose-calculation models (Tedgren and Ahnesjo 2003; Williamson 1996) • Characterization of shielded applicator corrections for treatment planning (Lymperopoulou et al., 2004; Weeks 1998; Williamson 1988a) • Use condensed history Monte Carlo codes in intravascular brachytherapy dosimetry • Monte Carlo dose-calculation engines for patient-specific treatment planning A number of investigators have applied Monte Carlo simulation to clinically realistic multiple-source implants with the goals of exploring the clinical impact of interseed attenuation and tissue heterogeneities
282
Jeffrey F. Williamson and Mark J. Rivard
on interstitial implants (Christensen et al., 2001; DeMarco et al., 1999; Lerma and Williamson 2002; Markman et al., 2001). To date, such codes generally require at least several days of computing time and are not easy to interface to planning images and dose-display and -analysis software found in commercial planning systems. However, several groups are developing accelerated Monte Carlo codes especially designed to facilitate efficient and more accurate planning in brachytherapy. Thus, integration of transport codes into clinical planning and delivery of brachytherapy is an important research area. Monte Carlo simulation is not the only kind of Boltzmann transport equation solver that has been investigated for use in brachytherapy dosimetry and treatment planning. Deterministic transport codes that have been successfully used in brachytherapy include discrete ordinates codes (Daskalov et al., 2000, 2002) and integral transport methods (Zhou and Inanc 2003).
Comparison of Theoretical and Experimental Dose-Rate Distributions for 125I and 103Pd Brachytherapy Sources While it is commonplace to compare LDR low-energy photon-emitting brachytherapy dosimetry parameters, such as g(r) and F(r,θ) as used in treatment-planning systems, comparisons of the dose-rate distributions of brachytherapy sources in not readily available. In practice, the dose-falloff is dominated by the solid-angle (c.f., the geometry function), which masks the subtle differences between sources. Additionally, there are a multitude of brachytherapy source models available having marked different construction. As a consequence, variations in brachytherapy dosimetry parameters between source models are expected to be larger than parameters obtained by different investigators for the same source. Furthermore, not all investigators used the same active length towards determining the geometry function or other dependent parameters. Therefore, we analyzed the ratios of dose-rate distributions obtained using theoretical and experimental techniques for a variety of brachytherapy sources.
Scope of Seeds Included in Analysis Specifically, the ratios of the theoretically calculated dose-rate distribution were divided by the experimentally determined dose-rate distribution on the transverse plane. A total of 19 seeds were included in this analysis, divided into 14 125I seeds and 5 103Pd seeds. The scope of seeds included in this comparison were as follows: 125
I Amersham model 6702 Amersham model 6711 Amersham model 6733 Best model 2301 Draximage model LS-1 IBt model 1251L Imagyn model IS-12501 IsoAid model IAI-125A Mentor model SL-125/SH-125 NASI model MED3631-A/M Nucletron model 130.002 SourceTech Medical model STM1251 Theragenics model 125.SO6
15–Quantitative Dosimetry Methods for Brachytherapy
283
103
Pd Best model 2335 Draximage model Pd-1 IBt model 1031L NASI model MED3633 Theragenics model 200
While some of these seeds are no longer available for clinical use, their data were still included in this analysis toward increasing the scope of this comparison to glean subtle differences between theoretical and experimental dosimetry techniques. Candidate datasets encompassing Monte Carlo and measured datasets were considered and compared in a manner similar to that described in the 2004 update to the AAPM TG-43 Report (Rivard et al., 2004). There were 38 candidate datasets for the 125I seeds with 25 ratios, and 14 candidate datasets for the 103Pd seeds with 10 ratios. Many of the references for these 52 datasets can be found in either appendix A or the references section in Rivard et al., (2004). In addition to sorting by radionuclide type (e.g., 125I or 103Pd), agreement between calculated and measured dose-rate distributions was assessed as a function of distance, r, and based on whether or not the radionuclide-backing material was either highor low-Z.
Ratios of Monte Carlo and Measured Dose-Rate Distributions for 125I The 25 ratios for 125I are shown in Figure 18a. Due to the related uncertainties of the data and the tight clustering on this linear-linear graph, no attempt as been made to discern particular ratios since the project focus is a general assessment of dose-rate ratios for all 125I seeds. The ratios typically fall within ±20% for all distances. From the observed trends, the following hypotheses may be made: (i) at close distances such as r = 0.5 cm, detector positioning uncertainties become pronounced for some datasets; (ii) differences between calculated and measured dose rates are minimized 1 ≤ r ≤ 2 cm, possibly due to maximally increasing the ratio of detector signal strength and positioning uncertainties; and (iii) differences generally increase in a monotonic fashion, possibly due to uncertainties in the phantom material correction factors for measurements or possibly due to uncertainties in Monte Carlo cross-section libraries which become more pronounced at larger distances. Of these 25 ratios, 13 ratios were for 8 125I sources having low-Z radionuclide-backing materials such as graphite or resin, and 12 ratios were for the remaining 6 125I sources having high-Z backing materials such as silver or tungsten. These dose-rate ratios are shown in Figures 18b and Figures 18c, respectively. While the calculated-to-measured dose-rate ratios for seeds having high-Z backing materials seeds are somewhat more clustered for 2 ≤ r ≤ 5 cm, the differences are not definitive and it is evident that correlation of the ratios is not significantly dependent on the radionuclide-backing material. As the lines are all stacked, the data in Figure 18a may be plotted in a different manner as in Figure 18d. Here, the frequency distributions for the 25 125I ratios are presented at r = 1 cm and r = 5 cm with 4% intervals (e.g., 100% ± 2%). As expected from Figure 18a, it is now clearly apparent that the r = 1 cm data has a tighter distribution than the r = 5 cm data.
284
Jeffrey F. Williamson and Mark J. Rivard
Figure 18a. Twenty-five ratios of calculated-tomeasured dose-rate distributions on the transverse plane for 14 125I seeds and 38 candidate datasets.
Figure 18b. Ratios of calculated-to-measured dose-rate distributions on the transverse plane for 125 I seeds having low-Z backing materials.
Figure 18c. Ratios of calculated-to-measured dose-rate distributions on the transverse plane for 125 I seeds having high-Z backing materials.
Figure 18d. Frequency distribution of the number of calculated-to-measured 125I dose-rate distribution ratios at r = 1 cm and r = 5 cm.
Ratios of Monte Carlo and Measured Dose-Rate Distributions for 103Pd Figure 19a depicts the 10 ratios for 103Pd. Just as for 125I, the best agreement between calculated and measured results was at r = 1 cm. However for 103Pd, the disparity as a function of distance was even more pronounced than for 125I. None of the 103Pd sources had high-Z backing materials, therefore, a comparative analysis as performed in Figures 18b and 18c was not possible. However, the frequency distribution presented in Figure 18d may be compared to that of Figure 19b. Again, the same trend of improved correlation at smaller distances was observed.
Conclusions In the past decade, quantitative dosimetry methods have continued to improve and become more widely used as the basis for clinical treatment planning. Low-energy photon-emitting brachytherapy planning now depends almost exclusively on Monte Carlo-based or measured single-source dose distributions. It appears that quantitative dosimetry techniques are rapidly overtaking more traditional model-based dose
15–Quantitative Dosimetry Methods for Brachytherapy
Figure 19a. Ten ratios of calculated-to-measured dose-rate distributions on the transverse plane for five 103Pd seeds and 14 candidate datasets.
285
Figure 19b. Frequency distribution of the number of calculated-to-measured 103Pd dose-rate distribution ratios at r = 1 cm and r = 5 cm.
calculation in the higher energy domain as well. While no major breakthroughs have occurred in the last decade, the uncertainty of Monte Carlo and TLD dose-estimation has become better established: 3% to 5% for Monte Carlo and 7% to 10% for TLD. As quantitative dosimetry techniques have become more widely disseminated and applied to more source models, broader intercomparisons of theoretical and measured dose distributions are possible, further increasing confidence in the basic dosimetric infrastructure of brachytherapy. In the next decade, we can anticipate the following improvement: (a) Improved dose measurement techniques with uncertainties near the theoretical limit of 3% to 4%; and (b) development of more accurate dose-calculation engines based upon Monte Carlo simulation and other transport equation solvers for patient-specific treatment planning.
References 1st Line Plain Ali, I., C. Costescu, M. Vicic, J. F. Dempsey, and J. F. Williamson. (2003). “Dependence of radiochromic film optical density post-exposure kinetics on dose and dose fractionation.” Med Phys 30(8):1958–1967. Ali, I., J. F. Williamson, C. Costescu, and J. F. Dempsey. (2005). “Dependence of radiochromic film response kinetics on fractionated doses.” Appl Radiat Isot 62(4):609–617. Almond, P. R., P. J. Biggs, B. M. Coursey, W. F. Hanson, M. S. Huq, R. Nath, and D. W. O. Rogers. (1999). “AAPM’s TG-51 protocol for clinical reference dosimetry of high-energy photon and electron beams.” Med Phys 26(9):1847–1870. Also available as AAPM Report No. 67. Alva, H., H. Mercado-Uribe, M. Rodriguez-Villafuerte, and M. E. Brandon. (2002). “The use of a reflective scanner to study radiochromic film response.” Phys Med Biol 47(16):2925–2933. Anderson, L. L., R. Nath, K. A. Weaver, D. Nori, T. L. Philips, Y. H. Son, S.-T. Chiu-Tsao, A. S. Meigooni, J. A. Meli, and V. Smith. Interstitial Brachytherapy, Physical, Biological, and Clinical Considerations. New York: Raven Press, 1990. Attix, F. H. Introduction to Radiological Physics and Radiation Dosimetry. New York: John Wiley and Sons, 1986. Attix, F. H., and V. H. Ritz. (1957). “A determination of the gamma-ray emission of radium.” J Res Natl Bureau Standards 59:293–305. Baldock, C., P. Murray, and T. Kron. (1999). “Uncertainty analysis in polymer gel dosimetry.” Phys Med Biol 44:N243-N246. Baras, P., I. Seimenis, P. Kipouros, P. Papagiannis, A. Angelopoulos, L. Sakelliou, E. Pappas, D. Baltas, P. Karaiskos, P. Sandilos, and L. Vlachos. (2002). “Polymer gel dosimetry using a three-dimensional MRI acquisition technique.” Med Phys 29:2506-2516.
286
Jeffrey F. Williamson and Mark J. Rivard
Bambynek, M., D. Fluhs, U. Quast, D. Wegener, and C. G. Soares. (2000). “A high-precision, high-resolution and fast dosimetry system for beta sources applied in cardiovascular brachytherapy.” Med Phys 27(4):662–667. Berger, M. J. (1968). “Energy deposition in water by photons from point isotropic sources.” J Nucl Med Suppl 1:17–25. Bethe, H. A. (1930). “Zur Theorie des Durchgangs schneller Korpuskularstrahlen durch Materie.” Ann Physik 5:325–400. Birks, J. B. “Physics of the Liquid Scintillation Process” in The Current Status of Liquid Scintillation Counting. E. D. Bransome (ed.). New York: Grune and Stratton, pp. 3–12, 1970. Bloch, F. (1933). Z. f. Physik 81:363. Bohm, T. D., P. M. DeLuca Jr., and L. A. DeWerd. (2003). “Brachytherapy dosimetry of 125I and 103Pd sources using an updated cross section library for the MCNP Monte Carlo transport code.” Med Phys 30(4):701–711. Bohm, T. D., D. W. Pearson, and R. K. Das. (2001). “Measurements and Monte Carlo calculations to determine the absolute detector response of radiochromic film for brachytherapy dosimetry.” Med Phys 28(2):142–146. Briesmeister, J. F. “MCNP—a General Monte Carlo N-Particle Transport Code, Version 4B.” Report No. LA-12625, Los Alamos National Laboratory, Los Alamos, NM, 1997. Burns, G. S., and D. E. Raeside (1988). “Two-dimensional dose distribution around a commercial 125I seed.” Med Phys 15(1):56–60. Cameron, J. R., N. Suntharalingam, and G. N. Kenney. Thermoluminescent Dosimetry. Madison, WI: University of Wisconsin Press, 1968. Chan, M. F., A. Y. Fung, Y-C. Hu, C-S. Chui, H. Amols, M. Zaider, and D. Abramson. (2001). “The measurement of three dimensional dose distribution of a ruthenium-106 ophthalmological applicator using magnetic resonance imaging of BANG polymer gels.” J Appl Clin Med Phys 2:85–89. Chiu-Tsao, S.-T., and L. L. Anderson. (1991). “Thermoluminescent dosimetry for 103Pd seeds (model 200) in solid water phantom.” Med Phys 18:449–452. Chiu-Tsao, S.-T., L. L. Anderson, K. O’Brien, and R. Sanna. (1990). “Dose rate determination for 125I seeds.” Med Phys 17(5):815–825. Chiu-Tsao, S.-T., A. de la Zerda, J. Lin, and J. H. Kim. (1994). “High-sensitivity GafChromic film dosimetry for 125 I seed.” Med Phys 21(5):651–657. Chiu-Tsao, S. T., T. L. Duckworth, C. Y. Hsiung, Z. Li, J. Williamson, N. S. Patel, and L. B. Harrison. (2003). “Thermoluminescent dosimetry of the SourceTech Medical model STM1251 125I seed.” Med Phys 30(7):1732–1735. Christen, T. (1914). “Radiometry” Arch Roentgen Ray 29:210.. Christensen, G. E., B. Carlson, K. S. Chao, P. Yin, P. W. Grigsby, K. Nguyen, J. F. Dempsey, F. A. Lerma, K. T. Bae, M. W. Vannier, and J. F. Williamson. (2001). “Image-based dose planning of intracavitary brachytherapy: Registration of serial-imaging studies using deformable anatomic templates.” Int J Radiat Oncol Biol Phys 51(1):227–243. Dale, R. G. (1983). “Some theoretical deviations relating to the tissue dosimetry of brachytherapy nuclides, with particular reference to iodine-125.” Med Phys 10:176-183. daRosa, L., and H. Nette. (1988). “Thermoluminescent dosimeters for exposure assessment in gamma or x radiation fields with unknown spectral distribution.” Appl Radiat Isot 39:191–197. Das, R. K., V. Mishra, H. Perera, A. S. Meigooni, and J. F. Williamson. (1995). “A secondary air kerma strength standard for Yb-169 interstitial brachytherapy sources.” Phys Med Biol 40(5):741–756. Das, R. K., Z. Li, H. Perera, and J. F. Williamson. (1996). “Accuracy of Monte Carlo photon transport simulation in characterizing brachytherapy dosimeter energy response artifacts.” Phys Med Biol 41:995–1006. Das, R., A. S. Meigooni, V. Mishra, M. A. Langton, and J. F. Williamson. (1997a). “Dosimetric characteristics of the type 8 ytterbium-169 interstitial brachytherapy source.” J Brachytherapy Int 13:219–234. Das, R. K., D. Keleti,Y. Zhu, A. S. Kirov, A. S. Meigooni, and J. F. Williamson. (1997b). “Validation of Monte Carlo dose calculations near 125I sources in the presence of bounded heterogeneities.” Int J Radiat Oncol Biol Phys 38(4):843–853. Daskalov, G. M. and J. F. Williamson. (2001). “Monte Carlo-aided dosimetry of the new Bebig IsoSeed 103Pd interstitial brachytherapy seed.” Med Phys 28(10):2154–2161. Daskalov, G. M., R. S. Baker, D. W. Rogers, and J. F. Williamson. (2000). “Dosimetric modeling of the microselectron high-dose rate 192Ir source by the multigroup discrete ordinates method.” Med Phys 27(10):2307–2319.
15–Quantitative Dosimetry Methods for Brachytherapy
287
Daskalov, G. M., R. S. Baker, D. W. Rogers, and J. F. Williamson. (2002). “Multigroup discrete ordinates modeling of 125I 6702 seed dose distributions using a broad energy-group cross section representation.” Med Phys 29:113–124. Davis, S. D., C. K. Ross, P. N. Mobit, L. Van der Laarse, W. J. Chase, and K. R. Shortt. (2003). “The response of lif thermoluminescence dosemeters to photon beams in the energy range from 30 kV x rays to 60Co gamma rays.” Radiat Prot Dosim 106(1):33–43. Dean, A. L. (1923). “Results of skin tests made to determine an objective dose for radium radiations.” Am J Roentgenol 10:654–661. De Deene, Y., C. De Wagter, and C. Baldock. (2003). “Comment on ‘A systematic review of the precision and accuracy of dose measurements in photon radiotherapy using polymer and Fricke MRI gel dosimetry.” Phys Med Biol 48:L15–L18. De Deene, Y., N. Reynaert, and C. De Wagter. (2001). “On the accuracy of monomer/polymer gel dosimetry in the proximity of a high-dose-rate 192Ir source.” Phys Med Biol 46:2801–2825. d’Errico F., R. Nath, S. K. Holland, M. Lamba, S. Patz, and M. J. Rivard. (2002). “A position-sensitive neutron spectrometer/dosimeter based on pressurized superheated drop (bubble) detectors.” Nuc Instr Meth Phys Res A 476:113–118. DeMarco, J. J., J. B. Smathers, C. M. Burnison, Q. K. Ncube, and T. D. Solberg. (1999). “CT-based dosimetry calculations for 125I prostate implants.” Int J Radiat Oncol Biol Phys 45(5):1347–1353. Demarco, J. J., R. E. Wallace, and K. Boedeker. (2002). “An analysis of MCNP cross-sections and tally methods for low-energy photon emitters.” Phys Med Biol 47(8):1321–1332. Dempsey, J. F., D. A. Low, A. S. Kirov, and J. F. Williamson. (1999). “Quantitative optical densitometry with scanning-laser film digitizers.” Med Phys 26(8):1721–1731. Dempsey, J. F., D. A. Low, S. Mutic, J. Markman, A. S. Kirov, G. H. Nussbaum, an J. F. Williamson. (2000). “Validation of a precision radiochromic film dosimetry system for quantitative two-dimensional imaging of acute exposure dose distributions.” Med Phys 27(10):2462–2475. Duderstadt, J. J., and W. R. Martin. Transport Theory. New York: Wiley, 1979. Dutreix, J. and M. Bernard. (1966). “Dosimetry at interfaces for high energy x and gamma rays.” Br J Radiol 39:205–210. Farajollahi, A. R. (1999). “An investigation into applications of polymer gel dosimetry in radiotherapy.” Med Phys 26:493 [abstract]. Farajollahi, A. R., D. E. Bonnett, A. J. Ratcliffe, R. J. Aukett, and J. A. Mills. (1999). “An investigation into the use of polymer gel dosimetry in low dose rate brachytherapy.” Br J Radiol 72:1085-1092. Fluhs, D., M. Heintz, F. Indenkampen, and C. Wieczorek. (1996). “Direct reading measurement of absorbed dose with plastic scintillators—the general concept and applications to ophthalmic plaque dosimetry.” Med Phys 23(3):427–304. Fong, P. M., D. C. Keil, M. D. Does, and J. C. Gore. (2001). “Polymer gels for magnetic resonance imaging of radiation dose distributions at normal room temperature.” Phys Med Biol 46:3105-3113. Fragoso, M., P. A. Love, F. Verhaegen, C. Nalder, A.M. Bidmead, M. Leach, and S. Webb. (2004). “The dose distribution of low dose rate Cs-137 in intracavitary brachytherapy: Comparison of Monte Carlo simulation, treatment planning calculation, and polymer gel measurement.” Phys Med Biol 49:5459–5974. Frigerio, N.A. (1962). “Neutron penetration during neutron capture therapy.” Phys Med Biol 6:541–549. Gearheart, D. M., A. Drogin, K. Sowards, A. S. Meigooni, and G. S. Ibbott. (2000). “Dosimetric characteristics of a new 125I brachytherapy source.” Med Phys 27(10):2278–2285. Goetsch, S. J., F. H. Attix, L. A. DeWerd, and B. R. Thomadsen. (1992). “A new well ionization chamber for the calibration of iridium-192 high dose rate sources.” Int J Radiat Oncol Biol Phys 24:167–170. Gore, J. C.,Y. S. Kang, and R. J. Schulz. (1984). “Measurement of radiation-dose distributions by nuclear magneticresonance (NMR) imaging.” Phys Med Biol 29:1189–1197. Gore, J. C., M. Ranade, M. J. Maryanski, and R. J. Schulz. (1996). “Radiation dose distributions in three dimensions from tomographic optical density scanning of polymer gels: I. Development of an optical scanner.” Phys Med Biol 41(12):2695–2704. Gray, L. H. (1929). “The absorption of penetrating radiation.” Proc Royal Society A122:647–668.
288
Jeffrey F. Williamson and Mark J. Rivard
Gray, L. H. (1936). “An ionization method for the absolute measurement of gamma-ray energy.” Proc Roy Soc A156:578-596. Gray, L. H. (1940). “Physical investigation of the contribution of the photo-electrons from sulphur to x-ray ionization.” Br J Radiol 13:25–30. Hafeli, U. O., W. K. Roberts, D. S. Meier, J. P. Ciezki, G. J. Pauer, E. J. Lee, and M. S. Weinhous. (2000). “Dosimetry of a W-188/Re-188 beta line source for endovascular brachytherapy.” Med Phys 27:668–675. Hasson, B. F. (1998). “Chemical dosimetry in the near-zone of brachytherapy sources.” (Abstract). Med Phys 25:2076. Hartmann, G. H., W. Schlegel, and H. Scharfenberg. (1983). “The three-dimensional dose distribution of 125I seeds in tissue.” Phys Med Biol 28(6):693–699. Hedtjarn, H., G. A. Carlsson, and J. F. Williamson. (2000). “Monte Carlo-aided dosimetry of the Symmetra model I25.S06 125I, interstitial brachytherapy seed.” Med Phys 27(5):1076–1085. Hedtjarn, H., G. A. Carlsson, and J. F. Williamson. (2002). “Accelerated Monte Carlo based dose calculations for brachytherapy planning using correlated sampling.” Phys Med Biol 47(3):351–376. Hubbell, J. H. (1999). “Review of photon interaction cross section data in the medical and biological context.” Phys Med Biol 44:R1–R22. Hubbell, J. H., and I. Øverbø. (1979). “Relativistic atomic form factors and photon coherent scattering cross sections.” J Phys Chem Ref Data 8:69–105. Hubbell, J. H., W. J. Veigele, E. A. Briggs et al., “Atomic form factors, incoherent scattering functions, and proton scattering cross sections,” Journal Journal of Physics and Chemistry Reference Data 4, 471-538 (1975). Ibbott G. S., M. Maryanski, A. Drogin, D. Gearheart, T. Painter, and A. Meigooni. (1999). “Characterization of a new brachytherapy source by BANG gel dosimetry.” (Abstract). Int J Radiat Oncol Biol Phys 45:417–418. Jenkins, T. M., W. R. Nelson, and A. Rindi. Monte Carlo Transport of Electrons and Photons. New York: Plenum Press, 1988. Kalli, H. J., and E. D. Cashwell. “Evaluation of Three Monte Carlo Estimation Schemes for Flux at a Point.” Informal Report No. LA-6865-MS, Los Alamos Scientific Laboratory, Los Alamos, New Mexico, 1977. Kalos, M. H., (1963). “On the estimation of flux at a point by Monte Carlo.” Nucl Sci Engr 16: 111–117. Kawrakow, I. and D. W. O. Rogers. “The EGSnrc Code System: Monte Carlo Simulation of Electron and Photon Transport.” Report No. Technical Report PIRS-701 (4th printing), National Research Council of Canada, Ottawa, Canada, 2003. Kipouros, P., P. Papagiannis, L. Sakelliou, P. Karaiskos, P. Sandilos, P. Baras, I. Seimenis, M. Kozicki, G. Anagnostopoulos, and D. Baltas. (2003). “3D dose verification in 192Ir HDR prostate monotherapy using polymer gels and MRI.” Med Phys 30:2031–2039. Kirov, A. S., A. S. Meigooni, Y. Zhu, R. K. Valicenti, and J.F. Williamson. (1995a). “Quantitative verification of 192Ir PDR and HDR source structure by pin-hole autoradiography.” Med Phys 22(11 Pt 1):1753–1757. Kirov, A., J. F. Williamson, A. S. Meigooni, and Y. Zhu. (1995b). “TLD, diode and Monte Carlo dosimetry of an 192 Ir source for high dose-rate brachytherapy.” Phys Med Biol 40(12):2015–2036. Kirov, A. S., S. Shrinivas, C. Hurlbut, J. F. Dempsey, W. R. Binns, and J. L. Poblete. (2000). “New water equivalent liquid scintillation solutions for 3D dosimetry.” Med Phys 27(5):1156–1164. Klassen, N. V., L. van der Zwan, and J. Cygler. (1997). “GafChromic MD-55: Investigated as a precision dosimeter.” Med Phys 24(12):1924–1934. Krishnaswamy, V. (1978). “Dose distribution around an I-125 seed source in tissue.” Radiology 126:489-491. Kubo, H. (1985). “Exposure contribution from Ti K x rays produces in the titanium capsule of the clinical I-125 seed.” Med Phys 12:215–220. H. D. Kubo, B. M. Coursey, W. F. Hanson et al., “Report of the ad hoc committee of the AAPM radiation therapy committee on 125I sealed source dosimetry,” Int J Radiat Oncol Biol Phys 40 (3), 697-702 (1998). Kubo, H. D., B. M. Coursey, W. F. Hanson, R. W. Kline, S. M. Seltzer, R. E. Shuping, and J. F. Williamson. (1998). “Report of the ad hoc committee of the AAPM Radiation Therapy Committee on 125I sealed source dosimetry.” Int J Radiat Oncol Biol Phys 40:697–702. Lamba, M., S. K. Holland, H. Elson, F. d’Errico, and, R. Nath. (1998). “Magnetic resonance imaging of microbubbles in a superheated emulsion chamber for brachytherapy dosimetry.” Med Phys 25:2316–2325. Laurence, G. C. (1938). “Some problems of gamma-ray measurement in radiology.” Am J Roentgenol 40:92–103.
15–Quantitative Dosimetry Methods for Brachytherapy
289
Lerma, F. A., and J. F. Williamson. (2002). “Accurate localization of intracavitary brachytherapy applicators from 3D CT imaging studies.” Med Phys 28:325–333. Lewis, E. E., and W. F. Miller. “Computational Methods of Neutron Transport.” La Grange Park, IL: American Nuclear Society, 1993. Li, Z., and J. F. Williamson. (1992). “Volume-based geometric modeling for radiation transport calculations.” Med Phys 19:667–677. Li, Z., J. F. Williamson, and H. Perera. (1993). “Monte Carlo calculation of kerma to a point in the vicinity of media interfaces.” Phys Med Biol 38:1825–1840. Ling, C. C., and E. D. Yorke. (1989). “Interface dosimetry for I-125 sources.” Med Phys 16(3):376–381. Ling, C. C., M. C. Schell, E. D. Yorke, B. B. Palos, and D. O. Kubiatowicz. “Two-dimensional dose distribution of 125 I seeds.” Med Phys 12(5):652–655. Ling, C. C., E. D. Yorke, I. J. Spiro, D. Kubiatowicz, and D. Bennett. (1983). “Physical dosimetry of I-125 seeds of a new design for interstitial implant.” Int J Radiat Oncol Biol Phys 9:1747–1752. Loftus, T. P. (1970). “Standardization of 137Cs gamma-ray sources in terms of exposure units (roentgens).” J Res Nat Bur Stand 74A:1–6. Loftus, T. P. (1980). “Standardization of 192Ir gamma-ray sources in terms of exposure.” J Res Nat Bur Stand 85:19–25. Loftus, T. P. (1984). “Exposure standardization of 125I seeds used for brachytherapy.” J Res Nat Bur Stand 89:295–303. Loftus, T. P. (1988). “Calibration of gamma-ray emitting brachytherapy sources.” Nat Bur Stand Spec Pub 89:2250. Luxton, G., M. A. Astrahan, D. O. Findley, and Z. Petrovich. (1990). “Measurement of dose rate from exposurecalibrated 125I seeds.” Int J Radiat Oncol Biol Phys 18(5):1199–1207. Luxton, G., and G. Jozsef. (1999). “Radial dose distribution, dose to water and dose rate constant for monoenergetic photon point sources from 10 keV to 2 MeV:EGS4 Monte Carlo model calculation.” Med Phys 26(12):2531–2538. Lymperopoulou, G., E. Pantelis, P. Papagiannis, P. Rozaki-Mavrouli, L. Sakelliou, D. Baltas, and P. Karaiskos. (2004). “A monte carlo dosimetry study of vaginal 192Ir brachytherapy applications with a shielded cylindrical applicator set.” Med Phys 31(11):3080–3086. Lynch, B., M. Ranade, J. Li, and J. Dempsey, (2004). “Characteristics of a New Very High Sensitivity Radiochromic Film.” Abstract TH-C-303-3. Presented at the American Association of Physicists in Medicine Annual Meeting, Pittsburgh, PA, 2004 (unpublished). (Abstract). Med Phys 31(6):1837. MacDougall, N. D., W. G. Pitchford, and M. A. Smith. (2002). “A systematic review of the precision and accuracy of dose measurements in photon radiotherapy using polymer and Fricke MRI gel dosimetry.” Phys Med Biol 47:R107–R121. Mack, A. G. Mack, D. Weltz, S. G. Scheib, H. D. Bottcher, and V. Seifert. (2003). “High precision film dosimetry with GAFCHROMIC films for quality assurance especially when using small fields.” Med Phys 30(9):2399–2409. Markman, J., J. F. Williamson, J. F. Dempsey, and D. Low. (2001). “On the validity of the superposition principle in dose calculations for intracavitary implants with shielded vaginal colpostats.” Med Phys 28(2):147–155. Maryanski, M. (2005). “New polymer gel dosimeter with high refractive index for artifact-free optical scanning.” Med Phys 32:In press [abstract SU-FF-T-209]. Maryanski, M. J., J. C. Gore, R. P. Kennan, and R. J. Schulz. (1993). “NMR relaxation enhancement in gels polymerized and cross-linked by ionizing radiations; A new approach to 3-D dosimetry by MRI.” Magn Res Imag 11:253–258. Maryanski, M. J., R. J. Schulz, G. S. Ibbott, J. C. Gatenby, J. Xie, D. Horton, and J. C. Gore. (1994). “Magnetic resonance imaging of radiation dose distributions using a polymer-gel dosimeter.” Phys Med Biol 39:1437–1455. Maryanski, M. J., G. S. Ibbott, P. Eastman, R. J. Schulz, and J. C. Gore. (1996). “Radiation therapy dosimetry using magnetic resonance imaging of polymer gels.” Med Phys 23(5):699–705. McJury, M. P. D. Tapper, V. P. Cosgrove, P. S. Murphy, S. Griffin, M. O. Leach, S. Webb, and M. Oldham. (1999). “Experimental 3D dosimetry around a high-dose-rate clinical 192Ir source using a polyacrylamide gel (PAG) dosimeter.” Phys Med Biol 44:2431–2444.
290
Jeffrey F. Williamson and Mark J. Rivard
McLaughlin, W. L. C. Yun-Dong, C. G. Soares et al. (1991). “Sensitometry of the response of a new radiochromic film dosimeter to gamma radiation and electron beams.” Nucl Instr Meth Phys Res A302:165–176. McMasters, W. H., N. Kerr, J. H. Mallett et al. “Compilation of X-ray Cross Sections,” Report No. UCRL-50174, Sec. II, Rev. 1, Lawrence Livermore Laboratory, Livermore, CA, 1969. Meigooni, A. S., J. A. Meli, and R. Nath. (1988a). “A comparison of solid phantoms with water for dosimetry of 125 I brachytherapy sources.” Med Phys 15(5):695–701. Meigooni, A. S., J. A. Meli, and R. Nath. (1988b). “Influence of the variation of energy spectra with depth in the dosimetry of 192Ir using LiF TLD.” Phys Med Biol 33(10):1159–1170. Meigooni, A. S., V. Mishra, H. Panth, and J. F. Williamson. (1995). “Instrumentation and dosimeter-size artifacts in quantitative thermoluminescence dosimetry of low-dose fields.” Med Phys 22(5):555–561. Meigooni, A. S., S. Sabnis, and R. Nath. (1990). “Dosimetry of palladium-103 brachytherapy sources for permanent implants.” Endocuriether/Hypertherm Oncol 6:107–117. Meisberger, L. L., R. J. Keller, and R. J. Shalek. (1968) “The effective attenuation in water of the gamma rays of gold 198, iridium 192, cesium 137, radium 226, and cobalt 60.” Radiology 90:953–957. Meli, J. A., A. S. Meigooni, and R. Nath. (1988). “On the choice of phantom material for the dosimetry of 192Ir sources.” Int J Radiat Oncol Biol Phys 14(3):587–594. Mishra, V., F. M. Waterman, and N. Suntharalingam. (1997). “Anisotropy of an 192Ir high dose rate source measured with a miniature ionization chamber.” Med Phys 24:751–755. Mohan, R., I. Y. Ding, M. K. Martel, L. L. Anderson, and D. Nori. (1985). “Measurements of radiation dose distributions for shielded cervical applicators.” Int J Radiat Oncol Biol Phys 11(4):861–868. Monroe, J. I., J. F. Dempsey, J. A. Dorton, S. Mutic, J. B. Stubbs, J. Markman, and J. F. Williamson. (2001). “Experimental validation of dose calculation algorithms for the GliaSite RTS, a novel 125I liquid-filled balloon brachytherapy applicator.” Med Phys 28(1):73–85. Monroe, J. I., and J. F. Williamson. (2002). “Monte Carlo-aided dosimetry of the theragenics TheraSeed model 200 103 Pd interstitial brachytherapy seed.” Med Phys 29(4):609–621. Morin, L. R. M. (1982). “Molecular form factors and photon coherent scattering cross sections of water.” J Phys Chem Ref Data 11:1091–1098. Mourtada, F., C. G. Soares, S. M. Seltzer, P. M. Bergstrom Jr., J. M. Fernandez-Verea, J. Asenjo, and S H. Lott. (2003). “Dosimetry characterization of a 32P source wire used for intravascular brachytherapy with automated stepping.” Med Phys 30(5):959–971. Muench, P. J., A. S. Meigooni, R. Nath, and W. L. McLaughlin. (1991). “Photon energy dependence of the sensitivity of radiochromic film and comparison with silver halide film and LiF TLDs used for brachytherapy dosimetry.” Med Phys 18(4):769–775. Nath, R., and N. Yue. (2000). “Dose distribution along the transverse axis of a new 125I source for interstitial brachytherapy.” Med Phys 27(11):2536–2540. Nath, R., A. S. Meigooni, and J. A. Meli. (1990). “Dosimetry on transverse axes of 125I and 192Ir interstitial brachytherapy sources.” Med Phys 17(6):1032–1040. Nath, R., N. Yue, and L. Liu. (1999). “On the depth of penetration of photons and electrons for intravascular brachytherapy.” Cardiovasc Radiat Med 1:72–79. Nath, R., A. S. Meigooni, P. Muench, and A. Melillo. (1993). “Anisotropy functions for 103Pd, 125I, and 192Ir interstitial brachytherapy sources.” Med Phys 20(5):1465–1473. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee, Task Group No. 43.” Med Phys 22:209–234. Also available as AAPM Report No. 51. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24(10):1557–1598. Also available as AAPM Report No. 59. Nath, R., H. Amols, C. Coffey, D. Duggan, S. Jani, Z. Li, M. Schell, C. Soares, J. Whiting, P. E. Cole, I Crocker, and R. Schwartz. (1999). “Intravascular brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 60.” Med Phys 26:119–152. Also available as AAPM Report No. 66. Nath, R., N. Yue, K. Shahnazi, and P. J. Bongiorni. (2000). “Measurement of dose-rate constant for 103Pd seeds with air-kerma strength calibration based upon a primary national standard.” Med Phys 27:655–658.
15–Quantitative Dosimetry Methods for Brachytherapy
291
Niroomand-Rad, A., C. R. Blackwell, B. M. Coursey, K. P. Gall, J. M. Galvin, W. L. McLaughlin, A. S. Meigooni, R. Nath, and J. E. Rodgers. (1998). “Radiochromic film dosimetry: Recommendations of AAPM Radiation Therapy Committee Task Group 55. American Association of Physicists in Medicine.” Med Phys 25(11):2093–2115. Also available as AAPM Report No. 63. Oldham, M., and L, Kim, (2004). “Optical-CT gel dosimetry II: Optical artifacts and geometric distortion.” Med Phys 31:1093–1104. Oldham, M., M. McJury, I. B. Baustert, S. Webb, and M. O. Leach. (1998). “Improving calibration accuracy in gel dosimetry.” Phys Med Biol 43:2709–2720. Oldham, M., J. H. Siewerdsen, A. Shetty, and D. A. Jaffray. (2001). “High resolution gel-dosimetry by optical-CT and MR scanning.” Med Phys 28:1436–1445. Oldham, M., J. H. Siewerdsen, S. Kumar, J. Wong, and D. A. Jaffray. (2003). “Optical-CT gel dosimetry I: Basic investigations.” Med Phys 30:623–634. Pantelis, E., A. K. Karlis, M. Kozicki, P. Papagiannis, L. Sakelliou, and J. M. Rosiak. (2004). “Polymer gel water equivalence an relative energy response with emphasis on low photon energy dosimetry in brachytherapy.” Phys Med Biol 49:3495–3514. Papagiannis, P., E. Pappas, P. Kipouros, A. Angelopoulos, L. Sakelliou, P. Baras, P. Karaiskos, I. Seimenis, P. Sandilos, and D. Baltas. (2001). “Dosimetry close to an 192Ir HDR source using N-vinylpyrrolidone based polymer gels and magnetic resonance imaging.” Med Phys 28:1416–1426. Parker, H. M., “A dosage system for interstitial radium therapy. II. Physical aspects.” Br J Radiol 11:252–266. Patel, N. S., S.-T. Chiu-Tsao, J. F. Williamson, P. Fan, T. Duckworth, D. Shasha, and L. B. Harrison. (2001). “Thermoluminescent dosimetry of the Symmetra 125I model I25.S06 interstitial brachytherapy seed.” Med Phys 28(8):1761–1769. Paterson, E. (1942). “A comparison of the action of x and gamma radiation on fibroblasts.” Br J Radiol 14:302–306. Perera, H., J. F. Williamson, Z. Li, V. Mishra, and A. S. Meigooni. (1994). “Dosimetric characteristics, air-kerma strength calibration and verification of Monte Carlo simulation for a new ytterbium-169 brachytherapy source.” Int J Radiat Oncol Biol Phys 28:953–971. Perera, H., J. F. Williamson, S. P. Monthofer, W. R. Binns, J. Klarmann, G.L. Fuller, and J. W. Wong. (1992). “Rapid two-dimensional dose measurement in brachytherapy using plastic scintillator sheet: linearity, signal-to-noise ratio, and energy response characteristics.” Int J Radiat Oncol Biol Phys 23(5):1059–1069. Perez-Calatayud, J., D. Granero, and F. Ballester (2004). “Phantom size in brachytherapy source dosimetric studies.” Med Phys 31(7):2075–2081. Perkins, S. T., and D. E. Cullen. “Tables and Graphs of Atomic Subshell and RelaxationData Derived from the LLNL Evaluated Atomic Data Library (EADL), Z = 1 – 100.” Report No. UCRL-50400, Vol. 30, Lawrence Livermore National Laboratory, Livermore, CA, 1991. Piessens, M., N. Reynaert, J. Potempa, H. Thierens, W. Wijns, and L. Verbeke. (2002). “Dose distributions for 90Y intravascularbrachytherapy sources used with balloon catheters.” Med Phys 29(7):1562–1571. Plechaty, E. F., D. E. Cullen, and R. J. Howerton, “Tables and Graphs of Photo-Interaction Cross Sections from 0.1 keV to 100 MeV Derived from the LLL Evaluated-Nuclear-Data Library,” Report No. UCRL-50400, Vol. 6, Rev. 2, Lawrence Livermore Laboratory, Livermore, CA, 1978. Ponnunni Kartha, K. I., G. N. Kenney, and J. R. Cameron. (1966). “An experimental determination of the absorption and buildup factor in water for radium, cobalt 60, and cesium 137 gamma rays.” Am J Roentgenol Radium Ther Nucl Med 96(1):66–69. Quimby, E. H. (1922). “The effect of the size of radium applicators on skin doses.” Am J Roentgenol 9:671–683. Quimby, E. H. (1928). “The intensity of radiation in the vicinity of filtered radon implants.” Radiology 10:365–376. Quimby, E. H. (1941). “The specification of dosage in radium therapy: 1940 Janeway Lecture.” Am J Roentgenol 45:1–17. Quimby, E. H. “Dosage Calculations in Radium Therapy” in Physical Foundations of Radiology, O. Glasser, E. H. Quimby, L. S. Taylor et al. (eds.). New York: Paul B. Hoeker, pp. 339–372, 1952. Rakavy, G., and A. Ron. (1967). “Atomic photoeffect in the range Eg=1-2000 keV.” Phys Rev 159:50–56. Reinstein, L. E., and G. R. Gluckman. (1999a). “Optical density dependence on postirradiation temperature and time for MD-55-2 type radiochromic film.” Med Phys 26:474–478.
292
Jeffrey F. Williamson and Mark J. Rivard
Reinstein, L. E., and G. R. Gluckman. (1999b). “Optical density dependence on postirradiation temperature and time for MD-55-2 type radiochromic film.” Med Phys 26(3):478–484. Reinstein, L. E., G. R. Gluckman, and A. G. Meek. (1998). “A rapid colour stabilization technique for radiochromic film dosimetry.” Phys Med Biol 43(10):2703–2708. Rikner, G. (1985). “Characteristics of a p-Si detector in high energy electron fields.” Acta Radiol Oncol 24(1):71–74. Rikner, G., and E. Grusell, (1987). “General specifications for silicon semiconductors for use in radiation dosimetry.” Phys Med Biol 32(9):1109–1117. Rivard, M. J. (2001). “Monte Carlo calculations of AAPM Task Group Report No. 43 dosimetry parameters for the MED3631-A/M 125I source.” Med Phys 28(4):629–637. Rivard, M. J., B. L. Kirk, and L. C. Leal. (2005). “Impact of radionuclide physical distribution on brachytherapy dosimetry parameters.” Nucl Sci Engr 149:101–106. Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Also available as AAPM Report No. 84. Roesch, W. C. (1958). “Dose for nonelectronic equilibrium conditions.” Radiat Res 9:399–410. Saloman, E. B., and J. H. Hubbell. “X-Ray Attenuation Coefficients (Total Cross Sections): Comparison of the Experimental Data Base with the Recommended Values of Henke and Theoretical Values of Scofield for Energies between 0.1 keV-100 keV,” Report No. NBSIR 86-3431, National Bureau of Standards, Washington, DC, 1986. Saloman, E. B., and J. H. Hubbell. (1987). “Critical analysis of soft x-ray cross section data.” Nucl Instr Meth Phys Res A 255:38–42. Schell, M. C., C. C. Ling, Z. C. Gromadzki, and K. R. Working. (1987). “Dose distributions of model 6702 I-125 seeds in water.” Int J Radiat Oncol Biol Phys 13(5):795–799. Schmickley, R. D., and R. H. Pratt. (1967). “K-, L-, and M-shell atomic photoeffect for screened-potential models.” Phys Rev 164:1228–1236. Schulz, R. J., A. F. de Guzman, D. B. Nguyen, and J. C. Gore. (1990). “Dose-response curves for Fricke-infused agarose gels as obtained by nuclear magnetic resonance.” Phys Med Biol 35:1611–1622. Scofield, J. H. “Theoretical Photoionization Cross Sections from 1 to 1500 keV.” Report No. UCRL-51326, Lawrence Livermore Laboratory, Livermore, CA, 1973. Seltzer, S. M. (1993). “Calculation of photon mass energy-transfer and mass energy-absorption coefficients.” Radiat Res 136(2):147–170. Seltzer, S. M., P. J. Lamperti, R. Loevinger, M. G. Mitch, J. T. Weaver, and B. M. Coursey. (2003). “New national air-kerma-strength standards for 125I and 103Pd brachytherapy seeds.” J Res Nat Bur Stand 108:337–358. Siddon, R. L. (1985). “Fast calculation of the exact radiological path for a three-dimensional CT array.” Med Phys 12(2):252–255. Soares, C. G. (1991). “Calibration of ophthalmic applicators at NIST: A revised approach.” Med Phys 18:787–793. Soares, C. G., D. G. Halpern, and C.-K. Wang. (1998). “Calibration and characterization of beta-particle sources for intravascular brachytherapy.” Med Phys 25:339–346. Soares, C. G., S. Vynckier, H. Järvinen, W. G. Cross, P. Sipilä, D. Flühs, B. Schaeken, F. A. Mourtada, G. A. Bass, and T. T. Williams. (2001). “Dosimetry of beta-ray ophthalmic applicators: Comparison of different measurement methods.” Med Phys 28:1373-1384. Spevacek, V., J. Novotny Jr., P. Dvorak, J. Novotny, J. Vymazal, and T. Cechak. (2001). “Temperature dependence of polymer-gel dosimeter nuclear magnetic resonance response.” Med Phys 28:2370–2378. Storm, E., and H. Israel. (1970. “Photon cross sections from 1 keV to 100 MeV for elements Z=1 to Z=100.” Nucl Data Table A7:565–681. Taylor, B. N., and C. E. Kuyatt. “Guidelines for evaluating and expressing the uncertainty of NIST measurement results.” NIST Technical Note 1297, U.S. Government Printing Office, Washington, DC, 1994. Tedgren, A. K., and A. Ahnesjo. (2003). “Accounting for high Z shields in brachytherapy using collapsed cone superposition for scatter dose calculation.” Med Phys 30(8):2206–2217. Tochilin, E., N. Goldstein, and J. T. Lyman. “The Quality and LET Dependence of Three Thermoluminescent Dosimeters and their potential use as secondary standards.” Presented at the Proceedings of the Second International Conference on Luminescence Dosimetry. Report CONF-68 09 20. United States Atomic Energy Commission, Springfield, 1968 (unpublished).
15–Quantitative Dosimetry Methods for Brachytherapy
293
Valicenti, R. K., A. S. Kirov, A. S. Meigooni, V. Mishra, R. K. Das, and J. F. Williamson. (1995). “Experimental validation of Monte Carlo dose calculations about a high-intensity Ir-192 source for pulsed dose-rate brachytherapy.” Med Phys 22(6):821–829. Villiard, M. (1908). “Instruments de mesure a lecture directe pour les rayons X.” Arch d’Elec Med 16:325. Wang, R., and X. A. Li. (2002). “Dose characterization in the near-source region for two high dose rate brachytherapy sources.” Med Phys 29(8):1678–1686. Weaver, K. A. (1984). “Response of LiF powder to 125I photons.” Med Phys 11(6):850–854. Weaver, K. A., V. Smith, D. Huang, C. Barnett, M. C. Schell, and C. Ling. (1989). “Dose parameters of 125I and 192Ir seed sources.” Med Phys 16(4):636–643. Weeks, K. J. (1998). “Monte Carlo dose calculations for a new ovoid shield system for carcinoma of the uterine cervix.” Med Phys 25(12):2288–2292. Weller, M. K., E. D. Slessinger, J. W. Wong et al. (1983). “A practical method for precise thermoluminescent dosimetry.” Treat Plan 8:22–26. White D. R. (1978). “Tissue substitutes in experimental radiation physics.” Med Phys 5:467–479. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, and R. Nath. (1998). “Dosimetric prerequisites for clinical use of new low energy photon interstitial brachytherapy sources.” Med Phys 25:2269–2270. Williamson, J. F. (1987). “Monte Carlo evaluation of kerma at a point for photon transport problems.” Med Phys 14(4):567–576. Williamson, J. F. (1988a). “Monte Carlo and analytic calculation of absorbed dose near 137Cs intracavitary sources.” Int J Radiat Oncol Biol Phys 15:227–237. Williamson, J. F. (1988b). “Monte Carlo evaluation of specific dose constants in water for 125I seeds.” Med Phys 15:686–694. Williamson, J. F. “Monte Carlo Simulation of Photon Transport Phenomena” in Monte Carlo Simulation in the Radiological Sciences. R. L. Morin (ed.). Boca Raton, FL: CRC Press Inc., pp. 53-102, 1988c. Williamson, J. F. (1991). “Comparison of measured and calculated dose rates in water near I-125 and Ir-192 seeds.” Med Phys 28:776–786. Williamson, J. F. (1996). “The Sievert integral revisited: Evaluation and extension to low energy brachytherapy sources.” Int J Radiat Oncol Biol Phys 36:1239–1250. Williamson, J. F. (1998b). “Monte Carlo-based dose-rate tables for the Amersham CDCS.J and 3M model 3600 137Cs tubes.” Int J Radiat Oncol Biol Phys 41(4):959–970. Williamson, J. F. (2000). “Monte Carlo modeling of the transverse-axis dose distribution of the model 200 103Pd interstitial brachytherapy source.” Med Phys 27:643–654. Williamson, J. F. (2002). “Dosimetric characteristics of the DRAXIMAGE model LS-1 1-125 interstitial brachytherapy source design: a Monte Carlo investigation.” Med Phys 29(4):509–521. Williamson, J. F., and A. S. Meigooni. “Quantitative Dosimetry Methods for Brachytherapy” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.). Madison, WI: Medical Physics Publishing, pp. 87–134, 1995. Williamson, J. F., and F. J. Quintero. (1988). ”Theoretical evaluation of dose distributions in water about models 6711 and 6702 125I seeds.” Med Phys 15(6):891–897. Williamson, J. F., and Z. Li. (1995). “Monte Carlo aided dosimetry of the microselectron pulsed and high dose-rate 192 Ir sources.” Med Phys 22(6):809–819. Williamson, J. F., H. Perera, and Z. Li. (1993). “Comparison of calculated and measured heterogeneity correction factors for 125I, 137Cs, and 192Ir brachytherapy sources near localized heterogeneities.” Med Phys 20(1):209–222. Williamson, J. F., W. Butler, L. A. DeWerd et al. (2005). “Recommendations of the American Association of Physicists in Medicine regarding the impact of implementing the 2004 Task Group 43 report on dose specification for 103Pd and 125I interstitial brachytherapy.” Med Phys 32:In press. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, and R. Nath. (1998). “Dosimetric prerequisites for clinical use of new low energy photon interstitial brachytherapy sources.” Med Phys 25:2269–2270. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, R. Nath, M. J. Rivard, and G. Ibbott. (2000). “Recommendations of the American Association of Physicists in Medicine on 103Pd interstitial source calibration and dosimetry: Implications for dose specification and prescription.” Med Phys 27:634–642. Also available as AAPM Report No. 69.
294
Jeffrey F. Williamson and Mark J. Rivard
Williamson, J. F., F. C. Deibel, and R. L. Morin. (1984). “The significance of electron binding corrections in Monte Carlo photon transport calculations.” Phys Med Biol 29(9):1063–1073. Williamson, J. F., J. F. Dempsey, A. S. Kirov, J. I. Monroe, W. R. Binns, and H. Hedtjarn. (1999). “Plastic scintillator response to low-energy photons.” Phys Med Biol 44(4):857–871. Williamson, J. F., R. L. Morin, and F. M. Khan. (1983). “Monte Carlo evaluation of the Sievert integral for brachytherapy dosimetry.” Phys Med Biol 28(9):1021–1032. Wyckoff, H. O. “From ‘Quantify of Radiation’ and ‘Dose’ to ‘Exposure’ and ‘Absorbed Dose:’An Historical Review.” Report No. Lauriston Taylor Lecture 4, National Council on Radiation Protection and Measurements, Washington, DC, 1980. Xu, Y., C.-S. Wuu, and M. J. Maryanski. (2004). “Performance of a commercial optical CT scanner and polymer gel dosimeters for 3-D dose verification.” Med Phys 31:3024-3033. Yorke, E. D., M. C. Schell, J. W. Gaskill, and C. C. Ling. (1987). “Using measured dose distribution data of the Fletcher-Suit-Delclos colpostat in brachytherapy treatment planning.” Int J Radiat Oncol Biol Phys 13(9):1413–1419. Zhou, C., and F. Inanc. (2003). “Integral-transport-based deterministic brachytherapy dose calculations.” Phys Med Biol 48(1):73–93. Zhu, Y., A. S. Kirov, V. Mishra, A. S. Meigooni, and J. F. Williamson. (1997). “Quantitative evaluation of radiochromic film response for two-dimensional dosimetry.” Med Phys 24(2):223–231. Zimmerman, D. W., C. R. Rhyner, and J. R. Cameron. (1966). “Thermal annealing effects on the thermoluminescence of LiF.” Health Phys 12(4):525–531.
Chapter 16
The TG-43 Brachytherapy Dose Calculation Formalism Mark J. Rivard Department of Radiation Oncology, Tufts University School of Medicine Boston, Massachusetts Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296 Clinical Rationale for Accurate Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 Task Group # 43 Dosimetry Formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298 General 2-D Formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 Air-Kerma Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300 Dose-Rate Constant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301 Geometry Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301 Radial Dose Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302 2-D Anisotropy Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303 General 1-D Formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303 Consensus Datasets for Clinical Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304 Source Geometry Variations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306 General Discussion of TG-43 Dosimetry Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306 Air-Kerma Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306 Dose-Rate Constant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 Radial Dose Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 2-D Anisotrophy Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307 Uncertainty Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308 Recommended Dosimetry Parameter Acquisition Methodology . . . . . . . . . . . . . . . . . . . . . . . . 308 General Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 Preparation of Dosimetry Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 Air-Kerma Strength Standard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 Dose-Rate Constant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309 Radial Dose Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310 2-D Anisotrophy Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 1-D Anisotrophy Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 Reference Data and Conditions for Brachytherapy Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 Radionuclide Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 Reference Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 Methodological Recommendations for Experimental Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . 312 Detector Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312 Medium and Energy Response Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313 Specification of Measurement Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314 Methodological Recommendations for Monte Carlo-based Dosimetry . . . . . . . . . . . . . . . . . . . . . . 314 Specification of the Monte Carlo Calculation Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315 Good Practice for Monte Carlo Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316 Publication of Dosimetry Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316 Clinical Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317 Dose-Calculation Formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318 Acceptance Testing and Commissioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 Source Calibrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
295
296
Mark J. Rivard
Introduction In 2004, the American Association of Physicists in Medicine (AAPM) published an update (TG-43U1) by Rivard et al., (Rivard et al., 2004a,b) to the Task Group No. 43 brachytherapy dosimetry protocol (TG43) by Nath et al., (1995). In this joint AAPM/American Brachytherapy Society (ABS) Summer School text on brachytherapy physics, a synopsis of this updated protocol (AAPM TG-43U1) is included for completeness and internal reference. While the reader is referred to the original report, this chapter should not be construed as AAPM guidance or recommendations since some of the descriptions and sections are somewhat modified from the published AAPM report/protocol. Specifically, this chapter focuses on the formalism and background of the protocol, and leaves the reader to reference the 2004 update (TG-43U1) for seed-specific data. The original TG-43 report introduced a new brachytherapy dose calculation formalism in 1995 that was largely based on findings of the Interstitial Collaborative Working Group (ICWG, 1990). Previous calculation formalisms were based upon apparent activity, equivalent mass of radium, exposure-rate constants, and tissue-attenuation coefficients. These older formalisms did not account for source-to-source differences in encapsulation or internal construction. With the exception of radium, the exposure-rate constants and other input parameters to these algorithms depended only on the radionuclide (Williamson 1997). In contrast, TG-43 employed dose-rate constants and other dosimetric parameters that depended on the specific source design, and are directly measured or calculated for each source design. Additionally, TG-43 presented consensus dosimetry data, in terms of the recommended formalism, for the three low-energy photon emitting source models then available (Theragenics Corporation model 200 103Pd source and Amersham Health models 6702 and 6711 125I sources). These data were based upon a critical review of ICWG measured dose-rate distributions using LiF TLD as well as other measurements and Monte Carlo calculations available in the literature. Overall, the TG-43 protocol has resulted in significant improvements in the standardization of both dose-calculation methodologies as well as dose-rate distributions used for clinical implementation of brachytherapy. For example, the differences between the previously used dose-rate distributions and those recommended by TG-43 were as large as 17% for some sources. These changes have been exhaustively reviewed by the physics community and are generally accepted. Many treatment planning software vendors have implemented the TG-43 formalism and the recommended dosimetry parameters in their systems. LiF TLD dose measurements and Monte Carlo dose calculations have largely replaced the semi-empirical dose-calculation models of the past. Since the publication of the TG-43 protocol a decade ago, significant advances have taken place in the field of permanent source implantation and brachytherapy dosimetry, necessitating an update of this protocol for the following reasons: (a) To eliminate minor inconsistencies and omissions in the original TG-43 formalism and its implementation (Meigooni and Nath 2000; Kline 1996; Fung 1998). (b) To incorporate subsequent AAPM recommendations, addressing requirements for acquisition of dosimetry data as well as clinical implementation (Williamson et al., 2000). These recommendations, e.g., elimination of Aapp and description of minimum standards for dosimetric characterization of low-energy brachytherapy sources (Williamson et al., 1998, 1999b) needed to be consolidated in one convenient document to complement the joint AAPM/RPC Brachytherapy Source Registry (RPC 2005), (c) To critically reassess published brachytherapy dosimetry data for the 125I and 103Pd source models introduced both prior and subsequent to publication of the TG-43 protocol in 1995, and to recommend consensus datasets where appropriate.
16–The TG-43 Brachytherapy Dose Calculation Formalism
297
(d) To develop guidelines for the determination of reference-quality dose distributions by both experimental and Monte Carlo methods and to promote consistency in derivation of parameters used in TG-43 formalism. Approximately 20 new low-energy interstitial brachytherapy source models have been introduced to the market since publication of TG-43 in 1995. These commercial developments can be attributed to the rising popularity of permanent prostate brachytherapy. Some of these new brachytherapy sources were introduced into clinical practice without thorough scientific evaluation of the necessary dosimetric parameters. The AAPM addressed this issue in 1998, recommending that at least one experimental and one Monte Carlo determination of the TG-43 dosimetry parameters be published in the peer-reviewed literature before using new low-energy photon-emitting sources (those with average photon energies less than 50 keV) in routine clinical practice (Williamson et al., 1998). Thus, many source models are supported by multiple dosimetry datasets based upon a variety of basic dosimetry techniques. This confronts the clinical physicist with the problem of critically evaluating and selecting an appropriate dataset for clinical use. To address this problem, this AAPM protocol presented a critical review of dosimetry data for eight 125I and 103Pd source models which satisfied the aforementioned criteria as of July 15, 2001, including the three low-energy source models included in the original TG-43 protocol. A further update of this protocol is anticipated for 2006 in Medical Physics to provide consensus, single source dose distributions and dosimetry parameters for low-energy photon sources. The present protocol (TG-43U1) recommends a single, consensus dataset for each source model from which the full 2-D dose-rate distribution can be reconstructed. Beta- or neutron emitting sources such as 90Sr, 32P, or 252Cf are not considered in this protocol. Finally, methodological guidelines are presented for physicist-investigators aiming to obtain dosimetry parameters for brachytherapy sources using calculative methods or experimental techniques.
Clinical Rationale for Accurate Dosimetry While low-energy, photon-emitting brachytherapy sources have been used to treat cancers involving a variety of anatomical sites, including eye plaque therapy for choroidal melanoma and permanent lung implants (Ling et al., 1983; Lee et al., 2003), their most frequent indication today is for the treatment of organconfined prostate cancer (Yu et al., 1999). Prostate cancer is the most prevalent cancer in men in the United States with approximately 220,000 new cases incident per year and an annual death rate of about 29,000 (Mettlin et al., 1998). While approximately 60% of new cases are confined to the organ at time of diagnosis, only about 2.2% of these new cases were treated with brachytherapy in 1995. Since that time, the percentage has increased to about 30% of all eligible patients receiving implants in current practice. This increase was largely due to improvements in diagnosis and case selection facilitated by introduction of the prostate specific antigen (PSA) screening test, and to improved ultrasound-guided delivery techniques. In the United States, the pioneering work was performed by a group of investigators located in New York (Zelefsky and Whitmore 1997) and Seattle (Ragde et al., 1998) based on work conducted in Denmark (Holm et al., 1983). The ABS critically reviewed the current published data and published consensus guidelines in 2001 to direct appropriate selection of patients. Prostate brachytherapy is appropriate as monotherapy for men with low risk favorable disease (T1-2a, Gleason score ≤6 and PSA < 10). While a prospective, randomized trial will likely never be performed due to strong patient preferences, long term survival rates published thus far appear comparable between prostate brachytherapy and radical prostatectomy for men with organ-confined disease. The most widely used technique utilizes transrectal ultrasound (TRUS) guided implantation of either 125I or 103Pd brachytherapy sources using a templateguided needle delivery system to avoid open surgery required by the retropubic approach (Whitmore, Hilaris, and Grabstald 1972; Peschel, King, and Roberts 1998). Due to its higher dose rate, 103Pd is often prescribed for higher grade tumors with 125I reserved for their low grade counterparts, despite a lack of clinical evidence supporting this practice.
298
Mark J. Rivard
The ABS recommends that CT-based post-implant dosimetry be performed on every patient undergoing a permanent prostate seed implant. Multiple studies have shown that clinical outcomes in prostate brachytherapy, both for the retropubic approach and the TRUS-guided technique, correlate with dose coverage parameters. Similarly, there is evidence that toxicity correlates with urethral and rectal doses. The extensive clinical experience of Memorial Sloan-Kettering Institute (1078 patients with retropubic approach surgery) from 1970–1987 was reviewed by Zelefsky and Whitmore (Zelefsky and Whitmore 1997). Multivariate-analysis revealed a D90 implant dose of 140 Gy to be an independent predictor of recurrence-free local control at 5, 10, and 15 years (p=0.001). D90 is defined as the dose delivered to 90% of the prostate volume as outlined using post-implant CT images. Similarly, a review of 110 implants at Yale University by Nath et al. (1998) using the retropubic implant approach from 1976 to 1986 reported a correlation (p=0.02) of recurrence-free local control after 10 years with V100; V100 is defined as the percentage of target volume receiving 100% of the prescribed dose. Two recent retrospective studies of the TRUS technique demonstrate that the clinical outcome depends on dose delivered and prostate volume coverage. Stock et al. (1998) reported on an experience of 134 prostate cancer patients implanted with 125I and not treated with teletherapy or hormonal therapy. They assessed rates of freedom from biochemical failure as a function of the D90 dose. A significant increase in freedom from biochemical failure (92% vs. 68% after 4 years) was observed (p=0.02) for patients (n=69) where D90 > 140 Gy. Potters et al. (2001) recently reviewed the impact of various dosimetry parameters on biochemical control for their experience of 719 patients treated with permanent prostate brachytherapy. Many of these patients also received teletherapy (28%) or hormone therapy (35%). Furthermore, 84% of the implants used 103Pd with the remainder using 125I. Their results indicated that patient age, radionuclide selection, and use of teletherapy did not significantly affect biochemical relapse-free survival (PSA-RFS). The only dose-specification index that was predictive of PSA-RFS was D90. Like the other two studies, studies by Stock et al., and Potters et al., were based on pre-TG-43 prescription doses of 160 Gy, and both indicated a steep dependence of clinical outcome with dose in the range of 100 to 160 Gy. For example, Stock et al., reported freedom from biochemical failure rates of 53%, 82%, 80%, 95%, and 89% for patients receiving D90<100 Gy, 100≤D90<120 Gy, 120≤D90<140 Gy, 140≤D90<160 Gy and D90≥160 Gy, respectively. The close correlation between D90 and PSA-RFS, and a dose response in the clinical dose range of 100 to 160 Gy are strong justifications for improved accuracy in the dosimetry for interstitial brachytherapy. Based on the updated dosimetry formalism and changes in calibration standards, the clinical medical physicist is advised that guidance on prescribed-to-administered dose ratios for 125I and 103Pd will be forthcoming in a subsequent AAPM report (Williamson et al., 2005).
Task Group No. 43 Dosimetry Formalism In the TG-43U1 protocol, both 2-D (cylindrically symmetric line source) and 1-D (point source) dosecalculation formalisms are given. To correct small errors and better address implementation details neglected in the original protocol, all quantities are defined. Throughout this protocol, the following definitions are used: • A source is defined as any encapsulated radioactive material that may be used for brachytherapy. There are no restrictions on the size or on its symmetry. • A point source is a dosimetric approximation whereby radioactivity is assumed to subtend a dimensionless point with a dose distribution assumed to be spherically symmetric at a given radial distance r. The influence of inverse square law, for the purpose of interpolating between tabulated transverseplane dose-rate values, can be calculated using 1/r2. • The transverse-plane of a cylindrically symmetric source is that plane which is perpendicular to the longitudinal-axis of the source and bisects the radioactivity distribution.
16–The TG-43 Brachytherapy Dose Calculation Formalism
299
• A line source is a dosimetric approximation whereby radioactivity is assumed to be uniformly distributed along a 1-D line-segment with active length L. While not accurately characterizing the radioactivity distribution within an actual source, this approximation is useful in characterizing the influence of inverse square law on a source’s dose distribution for the purposes of interpolating between or extrapolating beyond tabulated TG-43 parameter values within clinical brachytherapy treatment planning systems. • A seed is defined as a cylindrical brachytherapy source with active length, L, or effective length, Leff (described later in greater detail) less than or equal to 0.5 cm. These parameters are utilized by the TG-43U1 formalism in the following sections.
General 2-D Formalism The general, two-dimensional (2-D) dose-rate equation from the 1995 TG-43 protocol is retained,
" ( r ,θ ) = SK ⋅ Λ ⋅ G L ( r ,θ ) ⋅ g L ( r ) ⋅ F ( r ,θ ) D G L ( r0 ,θ 0 )
(1)
where r denotes the distance (in centimeters) from the center of the active source to the point of interest, r 0 denotes the reference distance which is specified to be 1 cm in this protocol, and θ denotes the polar angle specifying the point-of-interest, P(r,θ), relative to the source longitudinal-axis. The reference angle, θ0, defines the source transverse-plane, and is specified to be 90° or π/2 radians (Figure 1). Each of the factors used in equation (1) will be discussed in turn in the sections below. Equation (1) includes additional notation compared with the corresponding equation in original TG-43 formalism, namely the subscript “L” has been added to denote the line source approximation used for the geometry function. For evaluation of dose rates at small and large distances, the reader is referred to the published TG-43U1 report. This formalism applies to sources with cylindrically symmetric dose distributions with respect to the source longitudinal-axis. In addition, the consensus datasets assume that dose distributions are symmetric with respect to the transverse-plane, i.e., that radioactivity distributions to either side of the
Figure 1. Coordinate system used for brachytherapy dosimetry calculations.
300
Mark J. Rivard
transverse-plane are mirror images of one another. However, this formalism is readily generalized to accommodate sources that are not symmetric with respect to the transverse-plane. In clinical practice, radiopaque markers are used to identify source position and orientation. Generally, these markers are positioned symmetrically within the source capsule such that the marker, the radioactivity distribution, and the capsule have the same geometric center on the symmetry-axis of the source. Thus, determination of the location of the radioisotope distribution is based upon identification of the radio-opaque markers. All sources discussed herein can be accurately represented by a capsule and radio-opaque markers that are symmetric with respect to the transverse plane, which by definition bisects the active source and specifies the origin of the dose-calculation formalism. However, equation (1) can accommodate sources that are asymmetric with respect to the transverse plane. For sources such that: (i) the radioactivity distribution is clearly asymmetric with respect to the planes bisecting the capsule or marker; and (ii) and the extent of asymmetry is known a priori or can be measured via imaging; and (iii) the source orientation under clinical implant circumstances (e.g., via CT or radiography), then the source coordinate system origin should be positioned at the geometric center of the radionuclide distribution (as determined using positioning information obtained from the markers), not the geometric center of the exterior surface of the capsule or marker. If radiopaque markers do not facilitate identification of source orientation and the asymmetrical distribution under clinical circumstances, then the geometric center of the source must be presumed to reside at the radio-opaque marker centroid as is conventionally performed. Air-Kerma Strength The AAPM has consistently taken the position that air-kerma strength (SK), first introduced in AAPM TG32 report (Nath et al., 1987), should be the quantity used for specifying brachytherapy source strength for the purpose of defining calibration standards, documenting source strength on calibration reports and for all aspects of dose calculation and treatment prescription. However, the quantity apparent activity (Aapp) has been widely used by vendors and customers alike to specify the strength of sealed radioactive sources. Using Aapp rather than SK suffers from a number of problems that are outlined in TG-43U1. Consequently, the AAPM continues to recommend that Aapp not be used for brachytherapy source strength specification. Air-kerma strength has units of µGy·m2·h–1 and is numerically identical to the quantity Reference Air Kerma Rate recommended by ICRU 38 and ICRU 60 (ICRU 38, 1985; ICRU 60, 1998). For convenience, these unit combinations are denoted by the symbol U where 1 U = 1 µGy·m2·h–1 = 1 cGy·cm2·h–1. The National Institute of Science and Technology (NIST) maintains the U.S. primary air-kerma standards for x-rays in the energy range of 10 to 300 keV and for photon-emitting radionuclides such as 137Cs, 192Ir, 103Pd · and 125I. Air-kerma strength, SK, is the air-kerma rate, Kδ (d), in vacuo and due to photons of energy greater than d, at distance d, multiplied by the square of this distance, d2. · SK = Kδ (d) · d2
(2)
· The quantity d is the distance from the source center to the point of Kδ (d) specification (usually but not necessarily associated with the point of measurement) which should be located on the transverse-plane of the source (the plane normal to the longitudinal-axis of the source which bisects the radioactivity distribution). The distance d can be any distance that is large relative to the maximum linear dimension of the · radioactivity distribution so that SK is independent of d. Kδ (d) is usually inferred from transverse-plane air-kerma rate measurements performed in a free-air geometry at distances large in relation to the maximum linear dimensions of the detector and source, typically of the order of 1 meter. The qualification “in vacuo” means that the measurements should be corrected for photon attenuation and scattering in air and any other medium interposed between the source and detector, as well as photon scattering from any nearby objects including walls, floors and ceilings. Of course, air-kerma rate may also be calculated to subvert
16–The TG-43 Brachytherapy Dose Calculation Formalism
301
some of the limitations imposed on practical measurements. The energy cutoff, d, is intended to exclude low-energy or contaminant photons (e.g., characteristic x-rays originating in the outer layers of steel or titanium source cladding) that increase without contributing significantly to dose at distances greater than 0.1 cm in tissue. The value of d is typically 5 keV for low-energy photon-emitting brachytherapy sources, and is dependent on the application (Williamson 2000). Dose-Rate Constant The definition of the dose-rate constant in water, Λ, is the ratio of dose rate at the reference position, P(r0,θ0), and SK. L has units of cGy·h-1·U-1 which reduces to cm–2. Λ=
D" ( r0 , θ 0 )
(3)
SK
The dose-rate constant depends on both the radionuclide and source model, and is influenced by both the source internal design and the experimental methodology used by the primary standard to realize SK. In 1999, a notation was introduced, ΛnnD,PqqS, to identify both the dose-rate measurements or calcula· tions used to determine D (r0 ,θ 0) and the calibration standard to which this dose rate was normalized (Williamson et al., 1999b). The subscript “D” denotes reference dose rate, “nn” denotes the year in which this reference dose rate was published (either measurement or calculation), “P” denotes the provider or origin of the source strength standard (e.g., P = “N” for NIST, or P = “T” for the in-house calibrationstandard of ‘Theragenics Corporation’), “qq” denotes the year in which this source strength standard was implemented, and the “S” subscript denotes the word standard (Williamson et al., 2000). For example, Λ97D,N99S indicates a dose-rate constant determined from dosimetry measurements published in 1997 and normalized to an SK traceable to the 1999 NIST standard. Additional notation may also be utilized such as TLD Λ 97D,N85S for the dose-rate constant for the model 6702 source published in 1997 using TLDs 6702
and the 1985 NIST standard. These notations are useful for comparing results from multiple investigators, and readily highlight features such as utilization of the calibration procedure and whether or not influence of titanium K-shell x-rays is included. Geometry Function The purpose of the geometry function is to improve the accuracy with which dose rates can be estimated by interpolation from data tabulated at discrete points. Physically, the geometry function neglects scattering and attenuation, and provides an effective inverse square-law correction based upon an approximate model of the spatial distribution of radioactivity within the source. Like all the other functions used in this formalism, the geometry function is used to provide a means of interpolation between points of calculated or measured dose-rate for approximations with sufficient accuracy as used in treatment planning applications. However, the geometry function is somewhat based on the physical effect of dose-falloff, and the nature of the function may be explicitly given. Therefore, the AAPM TG-43U1 report recommends use of point- and line-source models giving rise to the following geometry functions: GP (r ,θ )
=
r −2
β G L ( r ,θ ) = Lr sin θ r 2 − L2 / 4
(
point-source approximation
(4)
if θ ≠ 0°
)
−1
linne-source approximation if θ = 0°
302
Mark J. Rivard
where β is the angle, in radians, subtended by the tips of the hypothetical line source with respect to the calculation point, P(r,θ). The geometry function used for derivation of dose rates from TG-43 parameters should be identical to that used to prepare the radial dose function and 2-D anisotropy function data, including use of the same active length, L, used in G(r,θ). Under these conditions, TG-43 dose calculations will reproduce exactly the measured or Monte Carlo-derived dose rates from which g(r) and F(r,θ) tables were derived. This protocol recommends consistent use of the line-source geometry function for evaluation of 2-D dose distributions, and use of either point- or line-source geometry functions for evaluations of 1-D dose distributions. Use of such simple functions is warranted since their purpose is to facilitate interpolation between tabulated data entries for duplication of the original dosimetry results. In the case where the radioactivity is distributed over a right-cylindrical volume or annulus, this protocol recommends taking active length to be the length of this cylinder. For brachytherapy sources containing uniformly spaced multiple radioactive components, L should be taken as the effective length, Leff, given by, Leff = ∆S×(N),
(5)
where N represents the number of discrete, radioactive pellets contained in the source with a nominal radioactive pellet center-to-center spacing ∆S (Meigooni et al., 2005; Rivard et al., 2005). If Leff is greater than the physical length of the source capsule (usually ~4.5 mm), the maximum separation (distance between proximal and distal aspects of the radioactivity distribution) should be used as the active length, L. This technique avoids singularities in evaluating G(r,θ) for points of interest in tissue which are located on the hypothetical line source just beyond the tip and end of the physical source. More complex forms of the geometry function have a role in accurately estimating dose at small distances outside the tabulated data range, i.e., extrapolating g(r) and F(r,θ) to small distances (Rivard 1999a, 2000). Use of such expressions is permitted. However, most commercial brachytherapy treatment planning systems support only point- or line-source geometry functions (Kouwenhoven et al., 2001; Meli 2002; Rivard et al., 2002). Therefore, it is the responsibility of the physicist to transform the tabulated TG-43 parameters given in this protocol, which are based upon point- and line-source approximations, to a format consistent with more complex geometry functions that may be available on their treatment planning systems. Radial Dose Function The radial dose function, gX(r), accounts for dose fall-off on the transverse-plane due to photon scattering and attenuation, i.e., excluding fall-off included by the geometry function. gX(r) is defined by equation (6), and is equal to unity at r0 = 1 cm. g X (r) =
" θ ) G (r θ ) D(r, X " D( r ,θ ) G X (r,θ ) 0
0
0
0
0
(6)
0
The revised dose-calculation formalism has added the subscript “X” to the radial dose function and geometry function to indicate whether a point-source, “P”, or line-source, “L”, geometry function was used in formatting the data. Consequently, this protocol presents tables of both gP(r) and gL(r) values. It is sometimes convenient to approximate the radial dose function as a polynomial, where fitting parameters a0 through a5 should be determined so that they fit the data within ±2%. Also, the radial range over which the fit meets this specification should be clearly specified.
16–The TG-43 Brachytherapy Dose Calculation Formalism
303
gX(r) = a0 + a1r + a2r2 + a3r3 + a4r4+ a5r5.
(7)
Equation (7) corrects a typographical error (Rivard 1999b) in the original TG-43 protocol. While table lookup via linear interpolation or any appropriate mathematical model fit to the data may be used to evaluate gX(r), some commercial treatment planning systems currently accommodate a fifth-order polynomial fit to the tabulated g(r) data. Since this type of polynomial fit may produce erroneous results with large errors outside the radial range used to determine the fit, alternate fitting-equations (Moss 2000) have been proposed which are less susceptible to this effect. 2-D Anisotropy Function The 2-D anisotropy function, F(r,θ), is defined as: F(r,θ )=
D" (r,θ ) G L (r,θ ) . D" (r,θ ) G L (r,θ ) 0
(8)
0
The 2-D anisotropy function describes the variation in dose as a function of polar angle relative to the transverse-plane. While F(r,θ) on the transverse-plane is defined as unity, the value of F(r,θ) off the transverse-plane typically decreases as: (i) r decreases, (ii) as θ approaches 0° or 180°, (iii) as encapsulation thickness increases, (iv) and as photon energy decreases. However, F(r,θ) may exceed unity at |θ-90°| > ± arcsin(L/2r) for right-cylinder sources coated with low-energy photon emitters due to screening of photons by the active element at angles towards the transverse-plane. Active length, L, used to evaluate GL(r,θ) in equation (4) shall be the same L used to extract gL(r) and F(r,θ) from dose distributions via equations (6) and (8), respectively. Otherwise, significant errors in dosimetry results at small distances may arise. For example, at r = 0.5 cm, a change in L from 3 to 5 mm results in a 5% change in GL(r,θ0).
General 1-D Formalism While a 1-D isotropic point-source calculation [equation (9)] only provides an approximation the true 2D dose distribution, it simplifies source localization procedures by eliminating the need to determine the orientation of the source longitudinal-axis from imaging studies. D" (r)= SK ⋅ Λ ⋅
G X (r,θ0 )
⋅ g (r)⋅ φan (r) G X (r0 ,θ0 ) X
(9)
Users should adopt one of the following implementations of equation (9): 2
r " r = S ⋅ Λ ⋅ 0 ⋅ g r ⋅ φ (r) , D K P an r
()
()
(10)
or 2
r " ( r) = S ⋅ Λ ⋅ 0 ⋅ g ( r) ⋅ φ (r) . D K P an r
(11)
304
Mark J. Rivard
While commercial treatment planning systems generally utilize the formalism in equation (10), that presented in equation (11) is the recommended formalism due to its improved accuracy at small distances, e.g., less than r = 1 cm. These formulations require consistency between the geometry function used for dose calculation and the geometry function used for extracting gX(r) from the transverse-plane dose distribution. Furthermore, these revised formulations correct an inconsistency in equation (11) of the original TG-43 protocol that indirectly recommended the following incorrect equation: " = S Λ ⋅ GP (r, θ 0 ) ⋅ g (r) ⋅ φ (r ) . NOT RECOMMENDED D(r) K L an GP (r0 , θ 0 )
(12)
While use of the wrong gX(r) datasets will typically give errors in the calculated dose rate of less than 2% at distances beyond 1 cm, average errors of 3%, 15%, and 74% arise at distances of 0.5, 0.25, and 0.1 cm, respectively. Clinical utilization of the 1-D dosimetry formalism presented in equation (12), or other formalisms that inconsistently apply the geometry function, are not recommended. The 1-D anisotropy function, φan(r), is identical to the anisotropy factor defined by the original TG43 protocol. At a given radial distance, φan(r) is the ratio of the solid angle-weighted dose rate, averaged over the entire 4 π steradian space, to the dose rate at the same distance r on the transverse-plane, see equation (13). π
φan(r)=
∫ D" ( r,θ ) sin (θ ) dθ 0
(13)
2 D" ( r, θ 0 )
Note that one should integrate dose rate, not the values of the 2-D anisotropy function to arrive at φan(r). Equation (11) will exactly reproduce the solid-angle weighted dose rate at a given r and is recommended because the line-source geometry function will provide more accurate interpolation and extrapolation at small distances. The accuracy achievable using the 1-D formalism for prostate implants has been reported (Lindsay, Battista, and Van Dyk 2001; Corbett et al., 2001). For brachytherapy treatment planning systems that do not permit entry of φan(r), Equations (10) or (11) can still be implemented by carefully modifying gX(r) to include φan(r) as shown in equation (14). Linear interpolation may be used to match the gX(r) and φan(r) grid spacing. These modified dosimetry parame– ters, g´(r) and φ 'an, are defined as: g ' ( r ) = g X ( r ) ⋅ φan ( r ) φ 'an = 1
(14)
Consensus Datasets for Clinical Implementation In the TG-43U1 protocol and subsequent updates, a deadline was established for inclusion of sources which had satisfied the AAPM recommendations that comprehensive (2-D) reference-quality dose-rate distribution data be accepted for publication by a peer-reviewed scientific journal. Appropriate publications can report either Monte Carlo, or experimentally derived TG-43 dosimetry parameters. For each source model, a single consensus dataset was derived from multiple published datasets according to the
16–The TG-43 Brachytherapy Dose Calculation Formalism
305
following methodology (Nath et al., 2002). If items essential to critical evaluation were omitted, the authors were contacted for information or clarification. (a) Experimentally determined values for the dose-rate constant (EXPΛ) were averaged. Separately, Λ values obtained using Monte Carlo techniques (MCΛ) were averaged. The CONΛ value recommended in this protocol (CONΛ) is the equally weighted average of the separately averaged experimental and Monte Carlo L values. In cases where there is only 1 experimental result and 1 Monte Carlo result: CONΛ = [EXPΛ + MCΛ]/2. (b) Each candidate dataset was examined separately and eliminated from consideration if it was determined to have a problem, e.g., data inconsistency. Corrections for use of a non-liquid water measurement phantom were applied if not included in the original investigators’ analysis. (c) For the 2-D anisotropy function, F(r,θ), and the radial dose function, g(r), all candidate datasets for a given source model were transformed using identical line-source geometry functions to permit fair comparison. The radial dose function was corrected for non-liquid water measurement medium if necessary. Assuming that the different datasets agreed within experimental uncertainties, the consensus data were defined as the ideal candidate dataset having the highest resolution, covering the largest distance range, and having the highest degree of smoothness. For most source models examined in this protocol, the consensus F(r,θ) and g(r) data, CONF(r,θ) and CONg(r), were taken from the transformed Monte Carlo dataset. (d) A few entries in the tabulated consensus datasets were taken from the non-ideal candidate dataset(s) to cover a larger range of distances and angles. These data were italicized to indicate that they were not directly confirmed by other measurements or calculations. (e) The 1-D anisotropy function, φan(r), was derived using numerical integration of the dose rate, as – calculated from CONF(r,θ) dataset, with respect to solid angle. Use of the anisotropy constant, φ 'an, is discouraged. (f) When scientifically justified for a given source model, exceptions or modifications to these rules were made, and are described later. For example, if the datasets were too noisy, they were rejected. (g) Following tabulation of g(r) and F(r,θ) for all eight source models, overly dense datasets were down-sampled to permit reasonable comparisons. Removal of a dataset point was deemed reasonable if linear interpolation using adjacent points resulted in a difference no larger than ±2% of the dataset point in question. Similarly, because the various authors used different table grids, it was necessary to interpolate some of the data into the common mesh selected for presenting all eight datasets. Linear-linear interpolation was used for F(r,θ) datasets, and log-linear interpolation was used for g(r) datasets. In this case, interpolated data are indicated by boldface. The details used to evaluate dosimetry parameters for each source were: 1. Internal source geometry and a description of the source, 2. Review of the pertinent literature for the source, 3. Correction coefficients for 1999 anomaly in NIST air-kerma strength measurements (if applicable), 4. Solid water-to-liquid water corrections, 5. Experimental method used: TLD or diode, 6. Active length assumed for the geometry function line-source approximation,
306
Mark J. Rivard 7. Name and version the Monte Carlo transport code, 8. Cross-section library used by Monte Carlo simulation, 9. Monte Carlo estimator used to score kerma or dose, and
10. Agreement between Monte Carlo calculations and experimental measurement.
Source Geometry Variations Source geometry and internal construction are highly manufacturer-specific and can affect the dosimetric characteristics of the source. Source models vary with regard to weld thickness and type, radioactivity carrier construction, presence of radio-opaque material with sharp or rounded edges, the presence of silver (which produces characteristic x-rays that modify the photon spectrum), and capsule wall thickness. Radioactive carriers may consist of a radio-transparent matrix, a radio-opaque object coated with radioactivity, or a radiotransparent matrix with highly attenuating radioactive coating. For example, the Amersham model 6702 and NASI model 3631-A/M sources utilize spherical resin carriers coated or impregnated with radioactivity. The number of spheres varies from 3 or more per source. Other sources, such as the Amersham model 6711, utilize a silver rod carrier. The amount of silver, or the length of silver rod, varies by the source model. Graphite pellets are also used. For example, in the Theragenics Corporation model 200 103 Pd source, the pellets are coated with a mixture of radioactive and nonradioactive palladium. All 125I and 103Pd source models, except for the now-obsolete model 6702 source, contain some type of radio-opaque marker to facilitate radiographic localization. For example, the graphite pellets of the Theragenics Corporation source are separated by a cylindrical lead marker. Beside the obvious dependence of photon spectrum on the radioisotope used, the backing material (e.g., the radiopaque marker) may further perturb the spectrum. For the sources containing 125I deposited on silver, the resultant silver x-rays significantly modify the effective photon spectrum. These source construction features influence the resultant dose rate distribution and the TG-43 dosimetry parameters to varying degrees. Accurate knowledge of internal source geometry and construction details is especially important for Monte Carlo modeling. While the dosimetry formalism was presented, its applicability to the derivation of consensus datasets is given later.
General Discussion of TG-43 Dosimetry Parameters Air-Kerma Strength Standards The NIST Wide-Angle Free-Air Chamber or WAFAC-based primary standard became available in 1998, and was used to standardize the 125I sources then available. The WAFAC standard shifted for unknown reasons in 1999, and was corrected in the first half of 2000. In early 2000, NIST noticed a shift in wellchamber calibration coefficients, with an apparent downward shift in the WAFAC air-kerma strength of approximately 3% for 125I sources, 5% for 125I-on-Ag sources, and 5% for 103Pd sources. NIST conducted a thorough investigation of this anomaly, completely checking all systems, mechanical, electrical, environmental, software, etc., particularly any factor that could equally affect both WAFAC instruments. However, a convincing explanation for this temporary shift was not determined. Thus, seed calibrations performed at NIST after mid 1998 and before January 1, 2000, need correction due to this measurement anomaly present in 1999. These dates are subject to revision should changes in manufacturing procedures, source geometry, or the WAFAC standard itself occur that affect the accuracy of vendor or ADCL secondary standards. Future source model-specific revisions to the calibration standard could require corresponding corrections to the recommended dose-rate constant. For this reason, regular calibration comparisons among NIST, ADCL, and vendors are required.
16–The TG-43 Brachytherapy Dose Calculation Formalism
307
In summary, there were two possible situations regarding the calibration of seeds at NIST using the WAFAC-based air-kerma strength standard. First, seed calibrations performed at NIST during the 1999 calendar year need correction due to a measurement anomaly present in 1999 only. This correction was determined by another WAFAC measurement for each seed model made at a designated date in 2000 or later. Secondly, WAFAC measurements made prior to 1999 and after January 1, 2000, needed no correction (Williamson et al., 1999a). Since the notation SK,N99 represents the NIST WAFAC-based air-kerma strength standard as officially introduced on January 1, 1999, this notation is used for all WAFAC measurements, regardless of the date of calibration. Thus, all measured dose-rate constant data given in this protocol have been normalized to the SK,N99 standard. Any measured dose-rate constants normalized to NIST calibrations performed in 1999 have been appropriately corrected for WAFAC measurement anomalies. Dose-Rate Constant Specifying the dose-rate constant as accurately as possible is essential, as it is used to transform the relative TG-43 dose distribution into absolute dose rates given the air-kerma strength of the sources deployed in the patient. As discussed in more detail later, Monte Carlo simulations have a freedom from detector positioning and response artifacts, smaller estimated uncertainty, and can yield artifact-free dose-rate estimates at distances shorter or longer than those accessible by TLD measurement techniques. However, the accuracy of Monte Carlo is inherently limited by the investigators’ ability to accurately delineate the source internal geometry. Few Monte Carlo studies have systematically evaluated the effects of geometric uncertainty, internal component mobility, tolerances in the fabrication of sources, and small manufacturing changes on the uncertainty of calculated dose-rate distributions. Therefore, the use of Monte Carlo values without confirmation by experimental studies is highly undesirable. Drawbacks of TLD dosimetry include (a) limited precision of repeated readings and spatial resolution; (b) a large and somewhat uncertain relative energy response correction; (c) failure of most investigators to monitor or control the composition of the measurement medium (Williamson 1991). For these reasons, the low-energy interstitial brachytherapy dosimetry subcommittee (LIBD) recommends using an equally-weighted average of the average measured (e.g., using TLDs) and average calculated (e.g., Monte Carlo-derived) values since the two recommended dosimetry characterization techniques have complementary strengths and limitations. Radial Dose Function For each source, Monte Carlo values of g(r) were graphically compared with experimental values. A comparison of the Monte Carlo and experimental g(r) results were expected to show an average agreement of ±10%. While the observed differences were typically < 5% for r ≤ 5 cm, systematic differences as large as 10% were observed due to use of outdated Monte Carlo cross-section libraries. Experimental values are difficult to measure at r < 1 cm, but Monte Carlo calculation of dose rate values are often available at smaller distances. In each case, the most complete dataset (typically Monte Carlo values) was used since values were readily available over a larger range of distances (especially at clinically significant distances closer than 1 cm) than provided by experimental measurements. 2-D Anisotropy Function Because Monte Carlo based datasets generally have superior smoothness, spatial and angular resolution, and distance range, all anisotropy functions recommended in this protocol are derived from Monte Carlo results which have been validated by comparison to less complete experimental datasets. A graphical comparison of datasets was performed, and the agreement between the Monte Carlo datasets and the experimental datasets was again expected to be ±10%. For θ > 30°, observed differences between the datasets were typically < 5% with a maximum of about 9%. For θ ≤30°, differences were larger (typi-
308
Mark J. Rivard
cally ~10% with maximum ~17%), and are attributed to volume averaging and the high-dose-rate gradient near the source longitudinal-axis as well as uncertainties in the source geometry assumed by Monte Carlo simulations.
Uncertainty Analysis Most of the experimental and computational investigations, especially those published prior to 1999, failed to include a rigorous uncertainty analysis, and the AAPM recommends that henceforth dosimetry investigators include rigorous uncertainty analyses, specific to their methodology employed, in their published articles (i.e., Table 1). Based on results of Monroe and Williamson (Monroe and Williamson 2002), purely Monte Carlo estimates of the transverse-axis dose-rate per unit air-kerma strength typically have uncertainties of 2% to 3% at 1 cm and 3% to 5% at 5 cm, depending on the type and magnitude of internal seed geometric uncertainties. Since relatively little has been published on estimation of systematic (type B) uncertainties of Monte Carlo-based dose estimation, the following sections apply the principles of uncertainty analysis, as outlined in NIST Technical Note 1297 (Taylor and Kuyatt 1994), to estimation of total uncertainty of Monte Carlo dose-rate constants, MCΛ, Monte Carlo radial dose functions MCg(r), consensus dose-rate constants, CONΛ, and absolute transverse-axis dose as evaluated by the dosimetric parameters recommended in the report. Implementation of a practical uncertainty analysis (Rivard, Melhus, and Kirk 2004) may be considered.
Recommended Dosimetry Parameter Acquisition Methodology The AAPM recommends a list of methodological details that should be described in brachytherapy dosimetry publications based upon either experimental or theoretical methods, along with more prescriptive
Table 1. Generic Uncertainty Assessment for Experimental Measurements Using TLDs, and Monte Carlo Methods for Radiation Transport Calculations. Type A and B uncertainties correspond to statistical and systematic uncertainties, respectively. All values provided are for 1 σ. TLD uncertainties Component
Type
Repetitive measurements
4.5%
TLD dose calibration
A Type B
Monte Carlo uncertainties Component Statistics
2.0%
(including linac calibration)
r = 1 cm
r = 5 cm
Photo-ionization
0.3% · d (1,θ0) =
1.0% · d (5,θ0) =
cross-sections
1.5%
4.5% 2% · d (5,θ0) =
(2.3%) LiF energy correction
5.0%
Seed geometry
Measurement medium
3.0%
Source energy
2% · d (1,θ0) =
spectrum
0.1%
0.3%
Quadrature sum
2.5%
5.0%
Correction factor Seed/TLD positioning
4.0%
Quadrature sum
4.5%
Total uncertainty
8.6%
ADCL SK uncertainty
1.5%
Total combined uncertainty in Λ
8.7%
7.3%
16–The TG-43 Brachytherapy Dose Calculation Formalism
309
guidelines on performing such studies. To better appreciate results from a particular dosimetric measurement and its uncertainties, the reader is referred to a listing of parameters needed to assess data for TLD measurements (Kron et al., 1999). Unfortunately, this level of description was not realized in many of the papers cited. When key data or methodological details were missing from a published paper, the author was contacted and asked to provide the missing information.
General Recommendations Since publication of TG-43 (Nath et al., 1995), the AAPM has published guidelines on dosimetric prerequisites (Williamson et al., 1998) for low-energy photon-emitting interstitial brachytherapy sources. However, they provided few technical recommendations to investigators publishing reference-quality doserate distributions derived from measurements or radiation transport calculations. Based on the AAPM experience (Nath et al., 2002) of analyzing publications and preparing consensus dosimetry datasets, more detailed recommendations on dosimetry methodology and data analysis were presented to define minimum requirements for future source dosimetry studies.
Preparation of Dosimetry Parameters Dosimetric parameters should be tabulated for both 1-D and 2-D dose-calculation models. This will require the investigator to calculate the geometry function and the radial dose function using both point-source (1-D) and line-source (2-D) geometry functions. Consequently, the investigator should always specify the active length used for the 2-D line-source geometry function. It is necessary to recommend minimum spatial resolutions and ranges for which these parameters should be defined since specification at a few distances or angles will not allow a sufficiently complete or accurate dose reconstruction of the 2-D dose distribution. The underlying dose distribution may have high gradients and result in inaccurate interpolation by brachytherapy treatment-planning systems, unnecessarily giving rise to dose-delivery errors. Air-Kerma Strength For experimental measurement of absolute dose rates to water, at least one source should have direct traceability of SK to the 1999 NIST WAFAC calibration standard. Other sources should have a precisely transferred air-kerma strength using high-precision transfer devices such as well-characterized well-ionization chambers and secondary standards maintained by the investigator as well as the manufacturer’s laboratories. The investigator using experimental techniques should state the NIST SK calibration uncertainty in the evaluation of Λ. Dose-Rate Constant The experimental investigator should rigorously control and try to minimize all detector response artifacts such as: dose-rate dependence, dose response nonlinearity, energy dependence, volumetric averaging, temporal stability of readings and calibration coefficients, and accuracy of detector positioning both in the source measurement setup and the detector calibration setup. These issues should be discussed in the measurement methodology section of the published paper, and a rigorous uncertainty analysis should also be provided. · Experimentally, Λ is evaluated by taking the ratio of the absolute dose rate, D (r0 ,θ 0) (the only absolute dose rate required to define TG-43 dosimetry parameters) and the measured air-kerma strength of the source, decayed to the time of dose-rate measurement. Typically 8 to 10 sources are used, with at least · one source having direct traceability to a NIST calibration. At least 15 measurements of D (r0 ,θ 0) are · generally performed. For example, multiple measurements of D (r0 ,θ 0) around a single NIST WAFACcalibrated source could be made by placing multiple TLDs in different quadrants of the transverse plane.
310
Mark J. Rivard
Monte Carlo radiation transport codes commonly provide direct estimates of absorbed dose or collision kerma per number of histories simulated (or some other internal normalization quantity, e.g., number of disintegrations, proportional to the number of primary photons simulated). Two simulations are necessary: one with the source model embedded in a phantom, yielding estimates of dose at specified points, and a second simulation with the source model positioned within a vacuum or a large air sphere. The doserate constant can then be estimated using the following ratio [equation (15)] since the numerator and denominator are similarly normalized, and the normalization constant used by the Monte Carlo code is irrelevant (Williamson 1988). Λ=
d" (r0 , θ 0 )
(15)
sK
· The parameter d (r0 ,θ 0) is the dose rate per history estimated using Monte Carlo methods at the reference position, and sK is the air-kerma strength per history estimated using Monte Carlo methods. Note the lower-case notation used to differentiate the normalized parameter, e.g., dose rate per history [cGy·h–1·history–1] as compared to absolute dose rate [cGy·h–1]. Although Monte Carlo studies are potentially free from experimental artifacts such as positioning uncertainties, energy response corrections, and signal-to-noise ratio limitations, such simulations require an accurate and complete geometric model of the source, selection of an appropriate cross-section library, and careful selection of dose tallying (estimation) and variance-reduction strategies. As with experimental studies, Monte Carlo-based dosimetry studies should include a complete uncertainty analysis. Care must be taken in evaluating dose rates at distances falling outside the range of tabulated data, rmin and rmax, especially at r < 0.5 cm. While difficult to measure, modern Monte Carlo techniques can easily calculate dose rates at distances as small as 0.1 cm from the center of the source (Williamson and Li 1995; Rivard 2002). This protocol recommends that gX(r) data be extracted and tabulated from Monte Carloderived dose rates for r ≥ 0.1 cm if possible. Users are warned that when working at small distances it is essential to use the same geometry function and L for evaluating absorbed dose as when used in the conversion of measured or Monte Carlo data in the TG-43 parameter ratios. Often, FX(r,θ) and gX(r) will exhibit non-physical properties due to breakdown of the line-source GL(r,θ) very near the source. Radial Dose Function The minimum distance range over which transverse-plane dose-rate data shall be measured or calculated should be from 0.5 to 7 cm for 125I and from 0.5 to 5 cm for 103Pd. Special attention to accuracy and completeness should be given to the 0.5 cm to 1.5 cm distance range, which is the most critical for accurately calculating dose for typical prostate and other volume implants and for accurately relating absolute dose rates (via the dose-rate constant) to the relative dose distribution. However, accurate dose characterization at shorter distances is critical for some procedures (e.g., episcleral eye plaque therapy) and for estimating hotspots in all implants. Accurate dose distributions at larger distances also contribute to overall dose-calculation accuracy and are important for assessing dose to organs at risk. Thus dose rates should be estimated down to the smallest distance supported by the methodology used. Some investigators have reported g(r) data down to 1 mm or less. Monte Carlo simulation can easily estimate dose at sub-millimeter distances as well as distances of 10 to 15 cm. In addition, some investigators have reported TLD or diode measurements at distances less than 1 mm. Since a 5th order polynomial is frequently used for smoothing transverse-plane data, one should present a minimum of 6 data points to fully-constrain the fit. Because these fits tend to perform poorly at small distances where the dose rate is highest, care should be taken to assure the fit is in good agreement at these positions. Acceptable levels of agreement are outlined in the following section.
16–The TG-43 Brachytherapy Dose Calculation Formalism
311
2-D Anisotropy Function When reporting 2-D anisotropy function data, at a minimum, F(r,θ) should be tabulated at radial distances, r = {0.5, 1, 2, 3, 5, and 7 cm} for 125I and {0.5, 1, 2, 3, and 5 cm} for 103Pd, and from θ = {0° to 90° with 10° increments} for all sources that are symmetric about the transverse-plane. To minimize extrapolation errors, data should be determined over as wide a radial distance range as reasonably possible given the experimental method or calculation technique. To minimize interpolation errors, data should also be obtained such that bilinear interpolation between various F(r,θ) data points produce errors less than 2%. When measuring 2-D anisotropy function data with sources that are cylindrically symmetric and symmetric about the transverse-plane (four identical quadrants), it is recommended that investigators determine relative dose rates a minimum of three times at each position analyzed. For source designs that are asymmetric about the transverse-plane or exhibit internal component shifts that may result in asymmetric dose distributions, F(r,θ) should be similarly tabulated, except with θ = {0° to 180° with 10° increments}. Because of the increased sampling capabilities using Monte Carlo methods in comparison to experimental methods, investigators using Monte Carlo methods should consider calculating dose rates using much finer grids in high-gradient regions such as near the source ends (e.g., θ near 0° or 180°). Angular resolution of a few degrees near the ends may be needed, with 10° resolution elsewhere. For those source designs with anisotropic photon-fluence distributions near the transverse plane, measurements and calculations having higher angular resolution are required to ensure that experimentally determined anisotropy functions are accurately normalized and that air-kerma strength per simulated history for Monte Carlo simulations is accurately calculated. The anisotropy constant does not exactly reproduce either the measured or Monte Carlo dosimetry data for r < 1 cm. TG-43U1 provides a detailed explanation for implementation of dose-calculation errors at small distances, e.g., r ≤ 1 cm. 1-D Anisotropy Function To derive 1-D anisotropy function data, a solid-angle weighted-average of the relative dose rates, uncorrected by the geometry function, should be performed over all angles. When examining small radii where θ=0° or 180° would place the calculation point within the source, the weighting should exclude the capsule/source volume and include only the volume outside the encapsulation. This is easily calculated for radii, r, less than half the capsule length where r sin θ > rcap, where rcap is the radius of the capsule.
Reference Data and Conditions for Brachytherapy Dosimetry Radionuclide Data The currently recommended values for half-lives, abundances and energies of 125I and 103Pd are presented in Table 2 (NNDC 2005; Lund/LBNL 1999; Seltzer 1993, Seltzer and Hubbell 1995). These values should be used to interpret future experimental measurements and as source spectra in Monte Carlo calculations. The 125I spectrum should be described in terms of five different photon energies. The 103Pd emission spectrum should be described in terms of eight discrete photon emissions. Although the relative number of high-energy photons emitted by 103Pd is low, their contribution to dose at distances beyond 10 cm can be clinically relevant and should also be considered for shielding calculations and exposure-control procedures (Rivard, Waid, and Wierzbicki 1999). Reference Media Water continues to be the recommended medium for reference dosimetry of interstitial brachytherapy sources. For dosimetry calculations and measurements, it may be necessary to know the composition of various forms of water and air. Pure, degassed water is composed of two parts hydrogen atoms and one
312
Mark J. Rivard Table 2. Recommended Nuclear Data for 125I and 103Pd for Brachytherapy Dosimetry
125
I (half-life = 59.40 ± 0.01 days) photon energy (keV) photons per disintegration
103 Pd (half-life = 16.991 ± 0.019 days) photon energy (keV) Photons per disintegration
27.202
0.406
20.074
0.224
27.472
0.757
20.216
0.423
30.98
0.202
22.72
0.104
31.71
0.0439
23.18
0.0194
35.492
0.0668
39.75
0.00068
294.98
0.00003
357.5
0.00022
497.1
0.00004
Weighted mean energy = 28.37 keV
Total = 1.476
I Γ10 keV = 0.0355 µGy·m2·h-1·MBq-1
125
Weighted mean energy = 20.73 keV
Total = 0.7713
Pd Γ10 keV = 0.0361 µGy ·m2·h-1·MBq–1
103
part oxygen atoms, with a mass density of 0.998 g ·cm–3 at 22 °C. Reference conditions for dry air are taken as 22°C and 101.325 kPa (760 mm Hg) with a mass density of 0.001196 g · cm–3. The composition of air may change as a function of relative humidity, and one must account for this effect (ICRU 37, 1989; Seltzer 2001). The proportion by weight of water in air of 100% relative humidity varies only between 1% and 2%, for temperatures between 16 °C and 26 °C and pressures between 735 mm Hg and 780 mm Hg. The change in mass density of saturated air is no more than a 1% reduction with respect to that for dry air, over this range of temperatures and pressures. Thus, the mass density will be set at 0.00120 g·cm–3 for both dry and moist air. For Monte Carlo calculations, the recommended relative humidity is 40%, which corresponds to the relative humidity in an air-conditioned environment where measurements should be carried out (Table 3).
Methodological Recommendations for Experimental Dosimetry Compared to Monte Carlo theorists who may idealize reality by a theoretic construct, the experimental investigator should address the variability that represents the clinical environment. The experimental study should investigate a reasonably large sample of sources received from multiple shipments at different stages of the production stream from the manufacturer. Detector Choice LiF TLD remains the method of choice for the experimental determination of TG-43 dosimetry parameters for low-energy photon-emitting brachytherapy sources. While a variety of other experimental dosimeters such as diodes, diamond detectors, miniature ionization chambers, plastic scintillators, liquid ionization chambers, polymer gels, radiographic and radiochromic film, and chemical dosimeters have been used for brachytherapy dosimetry, their validity for obtaining brachytherapy dosimetry parameters has not yet been convincingly demonstrated. The reader is referred to the TG-43U1 report (Rivard et al., 2004a) for a more in-depth discussion with references justifying the current view.
16–The TG-43 Brachytherapy Dose Calculation Formalism
313
Table 3. Composition (percent mass) of air as a function of relative humidity at a pressure of 101.325 kPa. Relative humidity (%)
Hydrogen
Carbon
Nitrogen
Oxygen
Argon
0
0.0000
0.0124
75.5268
23.1781
1.2827
10
0.0181
0.0124
75.4048
23.2841
1.2806
40
0.0732
0.0123
75.0325
23.6077
1.2743
60
0.1101
0.0123
74.7837
23.8238
1.2701
100
0.1842
0.0122
74.2835
24.2585
1.2616
Use of the aforementioned experimental dosimeters is an area of future research of significant scientific value. For measuring brachytherapy dosimetry parameters, detectors should have the following properties: (a) Detectors should have a relatively small active volume such that averaging effects of high-gradient dose fields over this volume are negligible or are accurately accounted for by correction coefficients. (b) A well-characterized energy-response function such that differences between the calibration energy and experimentally measured energy are either negligible or may be quantitatively accounted for. (c) Sufficient precision and reproducibility to permit dose-rate estimation with 1s statistical (Type A) uncertainties ≤ 5%, and 1s systematic uncertainties ≤ 7%. For example, TLD statistical uncertainties may be improved through repeated measurement at a given location, and systematic uncertainties may be improved through measuring chip-specific calibration coefficients. Typical statistical and systematic uncertainties for 1 × 1 × 1 mm3 TLD-100 chips are 4% and 7%, respectively, with total combined uncertainties of 7% to 9% (Meigooni et al., 2001). Therefore, 1 × 1 × 1 mm3 TLD-100 chips are considered a valid detector to perform the aforementioned absolute and relative measurements. Medium and Energy Response Characterization The measurement medium should also be well characterized (Meigooni et al., 1994). While epoxy-based substitutes for water, such as Solid Water™ by Gammex-RMI or Virtual Water™ by MED-CAL Inc., have liquid-water conversion coefficients that are less than 5% for high-energy teletherapy beams, coefficients range from 5% to 15% for low-energy photon-emitting sources. Recently, the measured calcium concentration of Solid Water‘ was found to have deviated from the vendor’s specification by as much as 30% (Patel et al., 2001). Therefore, when Solid Water is used in experimental dosimetry, the atomic composition of the material used should be measured and correction coefficients based on the measured composition of Solid Water should be used. Although Solid Water is the most widely used material for TG-43 reference dosimetry, it has several shortcomings. In addition to concerns over the constancy of its composition, Solid Water and similar water substitutes require solid-to-liquid water conversion corrections ranging from 4% to 15% in the 1 to 5 cm range. Alternative materials including polystyrene, polymethylmethacrylate, or plastic water (model PW2030 by Computerized Imaging Reference Systems, Inc.) may be possible candidates for future low-energy photon-emitting brachytherapy dosimetry studies. The relative energy response correction, E(r), is the largest single source of type B (systematic) uncertainty for TLD and other secondary dosimeters used in brachytherapy dosimetry. It is defined as the ratio of TLD response per unit dose in water medium at position r in the brachytherapy source geometry, to its
314
Mark J. Rivard
response per unit dose in the calibration geometry, usually a calibrated 60Co or 6 MV x-ray beam. In general, E(r) depends on source-to-detector distance, r, and may include corrections for volume averaging (influence of dose gradients in the TLD volume), detector self-absorption, medium displacement, and conversion from the measurement medium to liquid water (Williamson and Meigooni 1995). Most investigators treat E(r) as a distance-independent constant, although when it includes volume-averaging and Solid-to-liquid water corrections, as is often the case for Monte Carlo estimates, E(r) varies significantly with distance (Patel et al., 2001). A more in-depth discussion of these issues is given in the TG-43U1 report (Rivard et al., 2004a). In utilizing measured or Monte Carlo E(r) estimates published by others, the AAPM recommends that TLD experimentalists confirm that the associated measurement methodology matches their dosimetry technique with regard to TLD detector type and size, annealing and readout technique, and megavoltage beam calibration technique. The latter requires accounting for differences in calibration phantom material and dose-specification media used by the experimentalist and assumed by the selected E(r) estimate. Finally, further research is needed to resolve the discrepancy between published E(r) values, to identify the appropriate role for transport calculations in TLD dosimetry, and to reduce the large uncertainty associated with relative energy-response corrections. Specification of Measurement Methodology The experimental investigator should describe the important features of the measurement materials and methods to permit assessment of the results: 1. Description of the external and internal source geometry, 2. Brachytherapy source irradiation geometry, orientation, and irradiation timeline, 3. Radiation detector calibration technique (including protocol from which the technique is derived) and energy response function, E(r), 4. Radiation detector (dimensions, model #, and vendor) and readout system (e.g., electrometer unit model # and settings, or TLD readout unit model #, vendor, time-temperature profiles, and annealing program), 5. Measurement phantom (composition, mass density, dimensions, model #, and vendor), 6. Phantom dimensions and use of backscatter (at least 5 cm backscatter is recommended for 125I and 103 Pd dosimetry measurements), 7. Estimation of the impact of volume averaging on the results at all detector positions, 8. Number of repeated readings at each position, the number of different sources used, and the standard deviation of the repeated readings, 9. NIST SK value and uncertainty used for the measured source(s), and 10. Uncertainty analysis section assessing statistical and systematic uncertainties and their cumulative impact.
Methodological Recommendations for Monte Carlo-based Dosimetry Monte Carlo codes used to model photon transport for brachytherapy dose calculation should be able to support detailed 3-D modeling of source geometry and appropriate dose-estimation techniques. In addition, they should be based upon modern cross-section libraries and a sufficiently complete model of photon scattering, absorption, and secondary photon creation. Codes that have been widely used for interstitial
16–The TG-43 Brachytherapy Dose Calculation Formalism
315
brachytherapy dosimetry include EGS, MCNP, and Williamson’s PTRAN code. These codes have been widely benchmarked against experimental measurements or each other, so that their appropriate operating parameters and limitations can be considered to be well understood (Williamson et al., 1993). In general, the AAPM recommends Monte Carlo investigators utilize such well-benchmarked codes for brachytherapy dosimetry studies intended to produce reference-quality dose-rate distributions for clinical use. However, regardless of the transport code chosen and its pedigree, all investigators should assure themselves that they are able to reproduce previously published dose distributions for at least one widely used brachytherapy source model. This exercise should be repeated whenever new features of the code are explored, upon installing a new code version, or as part of orienting a new user. Other radiation transport codes, including Monte Carlo codes not previously used in brachytherapy dosimetry, should be more rigorously tested and documented in the peer-reviewed literature before proposing to use their results clinically. This is especially true for other types of transport equation solutions, including multigroup Monte Carlo, discrete ordinates methods, and integral transport solutions that have been proposed for brachytherapy dosimetry. Due to the short range of the secondary electrons produced by interactions from photons emitted by the radionuclides covered in this protocol, electron transport is not required and collision kerma closely approximates absorbed dose. Since the investigator performing Monte Carlo analysis can control many features of the transport calculations, it is imperative that the salient details be described in publications presenting Monte Carlo-derived brachytherapy dosimetry data. For instance, the collisional physics model should be described. The standard model used by experienced Monte Carlo users includes incoherent scattering corrected for electron binding by means of the incoherent scattering factor, coherent scattering derived by applying the atomic form factor to the Thompson cross-section, and explicit simulation of characteristic x-ray emission following photoelectric absorption in medium- and high-atomic number media. For sources containing Ag or Pd, it is imperative that, if characteristic x-ray production is not explicitly simulated, the primary source spectrum be appropriately augmented to include their presence. Specification of Monte Carlo Calculation Methodology A list of key features that should be specified by the investigator in the publication follows: 1. Radiation transport code name and version number, and describe major options, if any, invoked, 2. Cross-section library name and version number, and describe customizations performed if any, 3. Radiation spectrum of the source, 4. Manner in which dose-to-water and air-kerma strength are calculated: specify name of estimator or tally, whether or not transport was performed in air and how attenuation correction coefficients · were applied, and how suppression of contaminant x-ray production for Kδ (d) calculations was performed to be compliant with the NIST SK,N99 standard, 5. Source geometry, phantom geometry, and sampling space within the phantom, 6. Composition and mass density of the materials used in the brachytherapy source, 7. Composition and mass density of the phantom media, 8. Physical distribution of the radioisotope within the source (point-source modeling is unacceptable), and 9. Uncertainty analysis section assessing statistical and systematic uncertainties and their cumulative impact.
316
Mark J. Rivard
Good Practice for Monte Carlo Calculations For calculating brachytherapy dosimetry parameters, the following requirements should be adhered to: 1. Primary dosimetry calculations should be performed in a 30-cm diameter liquid water phantom, but calculations in Solid Water may also be performed to supplement experimental results, e.g., calculation of E(r), performed in Solid Water or other solid water substitutes. Typical calculations will produce dosimetry results extending out to r ~ 10 cm, with at least 5 cm of backscatter material for 125I and 103Pd dosimetry calculations. 2. Enough histories should be calculated to ensure that dosimetry results have a 1 σ (k=1, 67% confidence index) ≤ 2% at r ≤ 5 cm, and that calculations for derivation of SK, have 1 σ ≤ 1 % at the point of interest. 3. Modern, post–1980 cross-section libraries should be used, preferably those equivalent to the current NIST XCOM database such as DLC-146 or EPDL97. Exclude or appropriately modify (DeMarco, Wallace, and Boedeker 2002) older cross-section libraries (Storm and Israel 1970). Note that EGS4, EGSnrc, and MCNP all currently require modification or replacement of their default photo-ionization cross-sections to meet this requirement. Furthermore, moist air bestdescribes experimental conditions in comparison to dry air, and mass-energy absorption coefficients for moist air are recommended to minimize systematic uncertainties. 4. Manufacturer-reported source dimensions and compositions of encapsulation and internal components should be verified through the use of physical measurements, transmission radiography, and autoradiography. Just as the TLD experimentalist should measure an appropriate sample of sources, the Monte Carlo investigator should quantify the geometric variations in a sample of similar size. 5. The impact of volume-averaging artifacts should be limited to ≤1% through the appropriate choice of estimators (tallies) and scoring voxels if used. · 6. Calculations of d (r0 ,θ 0) to derive F(r,θ) should include high-resolution sampling in high-gradient regions such as near the source ends or in regions where internal source shielding causes abrupt · changes in d (r0 ,θ 0) and subsequently F(r,θ). · 7. K(d) should be modeled as a function of polar angle for sK simulation and Λ derivation. Williamson has shown that for some sources, detectors with large angular sampling volumes (such as the NIST WAFAC) will have a significantly different response than point-kerma detectors positioned on the transverse-plane. When the radioactivity is dispersed within or on the surface of a high-density core with sharp corners and edges, it may be necessary to simulate, if only approximately, the WAFAC geometry (dimensions and composition) to permit investigators the opportunity to directly compare Monte Carlo calculations of Λ with NIST-based measurements of Λ. 8. Mechanical mobility of the internal source structures, which has the potential to significantly affect the dose distribution, should be considered by the Monte Carlo investigator in developing both the geometric model of the source and the uncertainty budget.
Publication of Dosimetry Results Previous AAPM recommendations (Williamson et al., 1998) stated that dosimetry results should be published preceding clinical implementation. However, the journal Medical Physics established a “seed policy” in 2001 that, in effect, limits printing of articles to Technical Notes unless they contain significant
16–The TG-43 Brachytherapy Dose Calculation Formalism
317
new science. In order to comply with this restriction imposed by the journal, the AAPM will accept technical notes with limited details as acceptable, provided the full details as listed above are available to the committee at the time of evaluation. This policy in no way prevents publication of the article in other journals, as other scientific journals of interest to medical physicists are appropriate venues for publication of these dosimetry parameters. In an AAPM report, Williamson et al. (1998) recommended that dosimetry results be published by independent investigators, but did not offer a strict definition of what this independence entails. The spirit of the initial recommendation was to prompt publication of multiple studies to assess all the TG-43 brachytherapy dosimetry parameters, e.g., Λ, g(r), F(r,θ), and φ(r). Through determining the consensus datasets for the brachytherapy sources evaluated in this protocol, a definition of “Independence Policy” was adopted. There are two aspects of this policy, and both shall be met for full compliance. The first meaning of “independent studies” is that they are performed, written, and published by investigators who are affiliated with institutions independent of the source vendor and who have no major conflicts-of-interest with that vendor. The second meaning of “independent studies” is that they are scientifically independent of one another. In the case of two measurement-based studies, this will usually mean that two different investigators have used their own methodologies for measuring Λ and sampling the relative dose distribution, as TLD dosimetry is highly technique and investigator dependent. In the case of an empirical study and a Monte Carlo study, if properly executed, they will yield scientifically independent estimates of the TG-43 parameters. Thus, so long as the two studies are successfully scrutinized by the peer-review process and satisfy the AAPM scientific requirements, the empirical and Monte Carlo investigator rosters can overlap or even be identical. It is permissible to publish the Monte Carlo and measured estimates in the same paper so long as the two datasets are independently tabulated. In this context, “Not independent” means that the one study is used to modify the outcomes and methods of the other to improve agreement between the two datasets in a manner that is not scientifically justified. When possible, the authors should cite previous publications where the measurement system or techniques were first described, and illustrate only the key features. It does not benefit either the reader or the journal in question to continually restate the definition of TG-43 parameters or their formalism. Simply citing this protocol or the original TG-43 publication will suffice.
Clinical Implementation Dose distributions in and around clinical interstitial implants are calculated using computerized radiotherapy treatment planning (RTP) systems. For sources with radio-opaque markers, the 3-D coordinates of the centers (or the two ends) of the markers in implanted sources are determined using multiple-view radiographs or CT scans. The dose-rate contributions from each source at the points of interest are calculated using a one-dimensional or two-dimensional dose-calculation algorithm. These contributions are then summed to determine the total dose rate. This procedure assumes that there are no source-to-source shielding effects, that all tissue is water equivalent, and that the scattering volume within the patient is equivalent to that used in the consensus datasets. The term equivalent in this context means at least 5 cm of waterequivalent material surrounds the point of calculation. Many RTP systems are available commercially and use a variety of methods to calculate clinical dose-rate distributions. Some of the RTP systems use the single-source dosimetry data in a tabular form as input, whereas others represent the data by means of a mathematical formula that requires input of certain coefficients. Some use the TG-43 dose-calculation formalism and others do not. Prior to adopting the recommendations presented in this report, the medical physicist should implement the dose-calculation data and technique recommended by this report on his/her treatment planning system and quantitatively assess the influence of this action on dose delivery. This is best performed by
318
Mark J. Rivard
comparing the dose distribution for typical implants based on the revised dose-calculation procedure with those based upon the currently implemented algorithm for the same seed locations, source strengths, and dose-calculation grid. The potential impact of these dose-calculation modifications on dose delivery relative to the current dose-calculation technique should be discussed with the appropriate radiation oncologist before clinically implementation. Finally, the comparison of old and new dose-calculation algorithms for the same seed input data, and the resultant decisions that may impact clinical dose delivery, should be documented for future reference and for regulatory purposes.
Dose-Calculation Formalism RTP systems are divided into those that comply fully with the TG-43 formalism, and those that do not. Full compliance is defined here as the use of equation (1) or one of its two approximations, given in equations (10) and (11). Full compliance also requires the use of the dose-rate constant, appropriate values of the radial dose function, and the 1-D or 2-D anisotropy functions that are provided in this protocol. For RTP systems that use the TG-43 dose-calculation formalism and permit customer input of dosimetry parameters, one should enter (or verify the correct entry of) the recommended parameters, and check the accuracy of the dose calculation algorithm. The tabulated data provided in this protocol should be used with such systems. In other cases, coefficients in an equation, e.g., 3rd to 5th order polynomial fits, describing the TG-43 parameters may be required. In these cases, the RTP-specific model or formula should be fit to the data provided by this protocol. For example, for systems that require a third-order (or higher) polynomial fit of the radial dose function, the clinical physicist is responsible for obtaining the best-fit coefficients by fitting the polynomial equation to appropriate gX(r) data from this protocol. Before implementing the dose-calculation model, it is necessary to evaluate the quality of the parametric fit. Deviation of the fitted data from those recommended in this protocol should be less than ±2% which will yield a dose-rate calculation accuracy of ±2% over the distance range of interest. The medical physicist shall take responsibility for verifying the accuracy of dosimetry data, whether the manufacturer or the user actually enters the data. The AAPM recommends using RTP systems that comply fully with the TG-43 formalism, whenever possible. However, some RTP systems do not use the TG-43 dose-calculation algorithm. In most cases, one can devise a method to force the algorithm to generate the single-source dose-rate distributions recommended here by using modified values for the dosimetry parameters required by the RTP system. This conversion should be performed with care. As with RTP systems based on the TG-43 dose calculation formalism, one should assure that the RTP system is generating correct single-source dose-rate data by creating a single-source treatment plan with the modified parameters before clinical use. Meigooni et al. (1995) have described an example of this approach. The methods used to arrive at modified data, as well as records of the evaluation of the RTP system, should be documented carefully and retained for use following installation of upgrades and for inspection by regulatory authorities. Extreme caution should be exercised whenever parameters should be entered or displayed that have units that do not match the units on documentation printed by the RTP system or displayed on its monitor. Procedures should be developed and documented to describe exactly how the modified data and parameters are related to the non-TG-43 parameters assumed by the RTP system. These procedures should address both clinical treatment planning practices and chart-checking procedures. Ratios of the unconventional units to the conventional units should be supplied, to facilitate review of the planning method. Because this approach is prone to errors in implementation or interpretation, this method should be used as the last resort.
16–The TG-43 Brachytherapy Dose Calculation Formalism
319
Acceptance Testing and Commissioning Thorough acceptance testing and commissioning shall be carried out prior to clinical implementation of a new RTP system or new source model. The user should document the results of these tests both for later reference, and for compliance with applicable regulations. As a minimum, calculations of the dose-rate distribution shall be performed for a single source of each type to be used clinically. The recommendations of the AAPM TG-40 (Kutcher et al., 1994), AAPM TG-53 (Fraass et al., 1998), and AAPM TG-56 (AAPM 1997) should be followed. The user should determine the range of distances from the source over which the RTP calculations meet the recommended level of accuracy (±2%). If deviations between calculated results and the published data exceed ±2%, the deficiencies of the RTP system should be documented and further investigated by the user. This is especially important for RTP systems that fit a model to published data, because such models are prone to large errors outside the range of the reference data. In the high-dose-gradient regions close to a source, particularly near the ends of a source, the acceptable error may need to be larger. These deficiencies should be considered when evaluating treatment plans, and when considering the purchase of an RTP system. To perform comparisons at the recommended level of precision, numerical point dose rates calculated by the RTP system, rather than measured diameters of plotted isodose contours, should be used. The user should evaluate any deviations from the recommended data, taking into consideration the uncertainty of calculations at very small distances. The calculation matrix should be adjusted to a value appropriate for the high gradients near a source. For low-energy photon-emitting sources, grid spacing should not exceed 1 × 1 × 1 mm3; this size generally represents a reasonable compromise between calculation speed and accuracy. Isodose curves should be generated as part of the commissioning and continuing quality assurance procedures, but should be viewed as a test of the spatial accuracy of the graphic dose display function of the RTP system, rather than a definitive test of the underlying dose-calculation algorithm. Because comparisons should include both point dose-rate calculations and the placement of isodose lines, the user should also ensure that the RTP system and its graphical output devices cause isodose curves to appear in the correct locations relative to corresponding point calculations.
Source Calibrations For calibrating radioactive sources, the AAPM has previously recommended that users not rely on the manufacturer’s calibrations, but instead confirm the accuracy of source-strength certificates themselves by making independent measurements of source-strength that are secondarily traceable to the primary standard maintained at NIST according to AAPM TG-40. For patient treatments, AAPM further recommended that all clinically used sources bear calibrations that are secondarily traceable to the primary standard. AAPM TG-56 (AAPM 1997) defines “direct traceability,” “secondary traceability,” and “secondary traceability with statistical inference” as follows: “Direct traceability is established when either a source or a transfer instrument (e.g., well chamber) is calibrated against a national standard at an ADCL or at NIST itself.” “Secondary traceability is established when the source is calibrated by comparison with the same radionuclide and design that has a directly traceable calibration or by a transfer instrument that bears a directly traceable calibration.” “Secondary traceability by statistical inference is established when a source is one of a group of sources of which a suitable random sample has direct or secondary traceability.”
320
Mark J. Rivard
This protocol, in accord with the previously published AAPM guidance (TG-56, TG-40), also recommends that all sources implanted into patients have measurements based upon secondary traceability. Normally, vendors should be expected to provide calibration certificates that document secondary traceability to NIST for their sources. Each institution should maintain a means for verifying vendor calibrations by air-kerma strength measurements with secondary traceability. Source sampling and instrument quality assurance guidelines are documented in TG-56. Source strength verification with secondary traceability can be achieved either by means of a chamber bearing a calibration that is directly traceable to the appropriate national standard, or by comparison to a source of the same model having a calibration that is directly traceable to the national standard. These methods are described. 1. Secondarily traceable calibrations at the institution using a transfer instrument. For brachytherapy sources, calibrating sources with secondary traceability is best done with a well-type ionization chamber having an ADCL-assigned, directly traceable calibration coefficient determined for the source model to be used. TG-56 recommends returning the chamber to the ADCL at intervals of two years for recalibration. 2. Secondarily traceable calibrations at the institution by source intercomparison. When NIST or an ADCL provides an air kerma strength value for a given source, that source is then said to have a directly traceable calibration. The user can then use this source to calibrate his/her transfer instrument, such as a well chamber. The well chamber in this situation is said to have a calibration coefficient bearing secondary traceability. To accomplish this, the user may obtain a source from a vendor, send it to NIST or an ADCL and obtain an air kerma strength value for that source. The user can then use this source to calibrate his or her well chamber. The well chamber can then be used to confirm the vendor-stated source strengths of other sources of the same model before they are used clinically. If the chamber is calibrated by the ADCL, the calibration coefficient is directly traceable. If the calibrated source method is used, the user is responsible for transferring the calibration to his or her instrument. Regardless of which method is used, the well chamber should be checked for constancy on a regular basis using a long-lived source such as 137Cs. The AAPM TG-40 has recommended that a constancy check be performed at each used that the well chamber exhibit constancy to within ±2%. The clinical user is cautioned that the secondary traceability calibration procedure should meet some minimum standard of quality, i.e., that for the intercomparisons method, the total uncertainty should be comparable to that achievable using an instrument with a directly traceable calibration. For example, for the calibrated chamber or the calibrated source approach, the total expanded uncertainty (2σ = 95% confidence level) is obtained by adding in quadrature the uncertainties of instrument or source calibration at the ADCL (typically 2.4% including the uncertainty at NIST). It is emphasized that the precision of measurement in the well chambers is better than this (generally within ±0.5%). The typical total, expanded uncertainty at the user facility using a source transfer to a chamber is typically 3.0%. This is the maximum uncertainty that is relevant for secondarily traceable calibration. Any additional steps in the intercomparison process will increase the total uncertainty and thus will not meet the minimum standard of quality recommended in this protocol. The user is further warned that under no circumstances should a vendor or user calibration be used as a basis of institutional verification calibrations. Finally, it is important not to confuse a source bearing a calibration with direct or secondary traceability with a “calibrated source” obtained from the source manufacturer. Use of a source calibrated by a manufacturer is not an acceptable alternative for providing in-house calibrations with traceability to NIST or an ADCL.
16–The TG-43 Brachytherapy Dose Calculation Formalism
321
References American Association of Physicists in Medicine (AAPM). (1997). “Code of practice for brachytherapy physics: AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24:1557–1598. Also available as AAPM Report No. 60. Corbett, J. F., J. J Jezioranski, J. Crook, T. Tran, and I. W. T. Yeung. (2001). “The effect of seed orientation deviations on the quality of 125I prostate implants.” Phys Med Biol 46: 2785–2800. DeMarco, J. J., R. E. Wallace, and K. Boedeker. (2002). “An analysis of MCNP cross-sections and tally methods for low-energy photon emitters.” Phys Med Biol 47:1321–1332. Fraass, B., K. Doppke, M. Hunt, G. Kutcher, G. Starkschall, R. Stern, and J. Van Dyk. (1998). “American Association of Physicists in Medicine Radiation Therapy Committee Task Group 53: Quality assurance for clinical radiotherapy treatment planning.” Med Phys 25:1773–1829. Fung, A. Y. C. (1998). “Comments on: ‘Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43.” Med Phys 25:2477. Holm, H. H., N. Juul, J. F. Pedersen, H. Hansen, and I. Strøyer. (1983). “Transperineal 125iodine seed implantation in prostatic cancer guided by transrectal ultrasonography.” J Urol 130:283–286. International Commission on Radiation Units and Measurements (ICRU). Report No. 37. “Stopping Powers for Electrons and Positrons.” Bethesda, MD: ICRU, 1989. International Commission on Radiation Units and Measurements (ICRU). Report No. 38. “Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology.” Bethesda, MD: ICRU, 1985. International Commission on Radiation Units and Measurements (ICRU). Report No. 60. “Fundamental Quantities and Units for Ionizing Radiation.” Bethesda, MD: ICRU, 1998. Interstitial Collaborative Working Group (ICWG), Interstitial Brachytherapy: Physical, Biological, and Clinical Considerations, edited by L. L. Anderson, R. Nath, K. A. Weaver, D. Nori, T. L. Phillips, Y. H. Son, S-T. ChiuTsao, A. S. Meigooni, J. A. Meli, and V. Smith (Raven Press, New York, 1990). Kline, R. W. (1996). “Comments on: Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43.” Med Phys 23:1579. Kouwenhoven, E., R. van der Laars, and D. R. Schaart. (2001). “Variation in interpretation of the AAPM TG-43 geometry factor leads to unclearness in brachytherapy dosimetry.” Med Phys 28:1965–1966. Kron, T., L. DeWerd, P. Mobit, J. Muniz, A. Pradhan, M. Toivonen, and M. Waligorski. (1999). “A checklist for reporting of thermoluminescence dosimetry (TLD) measurements.” Phys Med Biol 44:L15–L17. Kutcher, G. J. L. Coia, M. Gillin, W. F. Hanson, S. Liebel, R. J. Morton, J. R. Palta, J. A. Purdy, L. E. Reinstein, G. K. Svensson. M. Weller, and L. Wingfield. (1994). “Comprehensive QA for radiation oncology: Report of the AAPM Radiation Therapy Committee Task Group 40.” Med Phys 21:581–618. Also available as AAPM Report No. 46. Lee, W., B. D. T. Daly, T. A. DiPetrillo, D. M. Morelli, A. C. Neuschatz, J. Morr, and M. J. Rivard. (2003). “Limited resection for non-small cell lung cancer: Observed local control with implantation of I-125 brachytherapy seeds.” Ann Thorac Surg 75:237–243. Lindsay, P., J. Battista, and J. Van Dyk. (2001). “The effect of seed anisotropy on brachytherapy dose distributions using 125I and 103Pd.” Med Phys 28:336–345. Ling, C. C., E. D. Yorke, I. J. Spiro, D. Kubiatowicz, and D. Bennett. (1983). “Physical dosimetry of 125I seeds of a new design for interstitial implant.” Int J Radiat Oncol Biol Phys 9:1747–1752. Lund/LBNL Nuclear Data Search. (1999). Version 2.0, February 1999, by S. Y. F. Chu. L. P. Ekström, and R. B. Firestone, providing the WWW Table of Radioactive Isotopes, maintained by the Lawrence Berkeley National Laboratory, Berkeley, CA, USA, and the Department of Physics, Lund University, Sweden; http://nuclear data.nuclear.lu.se/nucleardata/toi/ last accessed May 1, 2005. Meigooni, A. S., and R. Nath. (2000). “Response to: ‘Comment on: Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43.” Med Phys 27:265. Meigooni, A. S., Z. Li, Vi. Mishra, and J. F. Williamson. (1994). “A comparative study of dosimetric properties of Plastic Water and Solid Water in brachytherapy applications.” Med Phys 21:1983–1987.
322
Mark J. Rivard
Meigooni, A. S., J. F. Williamson, and R. Nath. “Single Source Dosimetry for Interstitial Brachytherapy” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.) Madison, WI: Medical Physics Publishing, pp. 209–233, 1995. Meigooni, A. S., Z. Bharucha, M. Yoe-Sein, and K. Sowards. (2001). “Dosimetric characteristics of the Best® doublewall 103Pd brachytherapy source.” Med Phys 28:2568–2575. Meigooni, A. S., V. Rachabatthula, S. B. Awan, and R. A. Koona. (2005). “Comment on ‘Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 32:In press (for June issue). Meli, J. A. (2002). “Let’s abandon geometry factors other than that of a point source in brachytherapy dosimetry.” Med Phys 29:1917–1918. Mettlin, C. J., G. P. Murphy, D. S. Rosenthal, and H. R. Menck. (1998). “The national cancer data base report on prostate carcinoma after the peak in incidence rates in the U.S.” Cancer 83:1679–1684. Monroe J. I., and J. F. Williamson. (2002). “Monte Carlo-aided dosimetry of the Theragenics TheraSeed® Model 200 103Pd interstitial brachytherapy seed.” Med Phys 29:609–621. Moss, D.C. (2000). “Improved analytical fit to the TG-43 radial dose function, g(r).” Med Phys 27:659–661. Nath, R., L. Anderson, D. Jones, C. Ling, R. Loevinger, J. Williamson, and W. Hanson. “Specification of brachytherapy source strength: A report by Task Group 32 of the American Association of Physicists in Medicine.” AAPM Report No. 21 (American Institute of Physics, New York, 1987). Nath R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43.” Med Phys 22:209–234. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24:1557–1598. Nath, R., K. Roberts, M. Ng, R. Peschel, and Z. Chen. (1998). “Correlation of medical dosimetry quality indicators to the local tumor control in patients with prostate cancer treated with iodine-125 interstitial implants.” Med Phys 25:2293–2307. Nath. R., M. J. Rivard, B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. Ibbott, and J. F. Williamson. (2002). “Status of the American Association of Physicist in Medicine Radiation Therapy Committee Subcommittee’s Low-Energy Interstitial Brachytherapy Source Dosimetry: Procedure for the development of consensus single source dose distributions.” Med Phys 29:1349. National Nuclear Data Center (NNDC). Nuclear data from NuDat, a web-based database maintained by the National Nuclear Data Center, Brookhaven National Laboratory, Upton, NY, USA; last database update reported as August 12, 2000; http://www.nndc.bnl.gov/nndc/nudat/ last accessed May 1, 2005. Patel, N. S., S-T. Chiu-Tsao, J. F. Williamson, P. Fan, T. Duckworth, D. Shasha, and L. B. Harrison. (2001). “Thermoluminescent dosimetry of the SymmetraTM 125I model I25.S06 interstitial brachytherapy seed.” Med Phys 28:1761–1769. Peschel, R. E., C. R. King, and K. Roberts. (1998) “Pubic arch interference in permanent prostate implant patients.” J Brachytherapy Int 14:241–248 . Potters L., Y. Cao, E. Calugrau, T. Torre, P. Fearn, and X-H. Wang. (2001). “A comprehensive review of CT-based dosimetry parameters and biochemical control in patients treated with permanent prostate brachytherapy.” Int J Radiat Oncol Biol Phys 50:605–614. Radiological Physics Center (RPC), The M.D. Anderson Cancer Center, Houston, TX http://rpc.mdanderson. org/rpc/htm/Home_htm/Low-energy.html last accessed May 1, 2005. Ragde, H., A.-A. A. Elgamal, P. B. Snow, J. Brandt, A. A. Bartolucci, B. S. Nadir, and L. J. Korb. (1998). “Ten-year disease free survival after transperineal sonography-guided iodine-125 brachytherapy with or without 45-gray external beam irradiation in the treatment of patients with clinically localized, low to high Gleason grade prostate carcinoma.” Cancer 83:989–1001. Rivard, M. J. (1999a). “Refinements to the geometry factor used in the AAPM Task Group Report No. 43 necessary for brachytherapy dosimetry calculations.” Med Phys 26:2445–2450. Rivard, M. J. (1999b). Comments on: Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43.” Med Phys 26:2514.
16–The TG-43 Brachytherapy Dose Calculation Formalism
323
Rivard, M. J. (2000). “Neutron dosimetry for a general 252Cf brachytherapy source.” Med Phys 27:2803–2815. Rivard, M. J. (2002). “Comprehensive Monte Carlo calculations of AAPM Task Group Report No. 43 dosimetry parameters for the Model 3500 I-Plant 125I brachytherapy source.” Appl Radiat Isot 57:381–389. Rivard, M. J., C. S. Melhus, and B. L. Kirk. (2004). “Brachytherapy dosimetry parameters calculated for a new 103Pd source.” Med Phys 31:2466–2470. Rivard, M. J., D. S. Waid, and J. G. Wierzbicki. (1999). “Mass attenuation coefficients of Clear-Pb“ for photons from 125 103 I, Pd, 99mTc, 192Ir, 137Cs, and 60Co.” Health Phys 77:571–578. Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. Ibbott, R. Nath, and J. F. Williamson. (2002). “Comment on: Let’s abandon geometry factors other than that of a point source in brachytherapy dosimetry.” Med Phys 29:1919–1920. Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. (2004a). “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations (AAPM Report No. 84).” Med Phys 31:633–674. Rivard, M. J., W. M. Butler, L. A. DeWerd, M. S. Huq, G. S. Ibbott, Z. Li, M. G. Mitch, R. Nath, and J. F. Williamson. (2004b). “Erratum: ‘Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:3532–3533. Rivard, M. J., W. M. Butler, L. A. DeWerd, M. S. Huq, G. S. Ibbott, C. S. Melhus, M. G. Mitch, R. Nath, and J. F. Williamson, (2005) “Response to ‘Comment on “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.”’ Med Phys 2005,32:(in press for June issue). Seltzer, S. M. (1993) “Calculation of photon mass energy-transfer and mass energy-absorption coefficients.” Radiat Res 136:147–170. Seltzer, S. M., and J. H. Hubbell. (1995). “Tables and graphs of mass attenuation coefficients and mass energy-absorption coefficients for photon energies 1 keV to 20 MeV for elements Z = 1 to 92 and some dosimetric materials.” Japanese Society of Radiological Technology ISSN 1340-7716. Seltzer, S. M. of NIST, private communication with author, September 13 (2001). Stock, R. G., N. N. Stone, A. Tabert, C. Iammuzzi, and J. K. DeWyngaert. (1998). “A dose-response study for I-125 prostate implants.” Int J Radiat Oncol Biol Phys 41:101–108. Storm, E., and H. Israel. (1970). “Photon cross-sections from 1 keV to 100 MeV for elements Z=1 to Z=100.” Nucl Data Tables A 7:566–575. Taylor, B. N., and C. E. Kuyatt. “Guidelines for evaluating and expressing the uncertainty of NIST measurement results,” NIST Technical Note 1297, (U.S. Government Printing Office, Washington D.C., 1994). Whitmore, W. F. B. Hilaris, H. Grabstald. (1972). “Retropubic implantation of iodine-125 in the treatment of prostatic cancer.” J Urol 108:918–920. Williamson, J. F. (1988). “Monte Carlo evaluation of specific dose constants in water for 125I seeds.” Med Phys 15:686–694. Williamson, J. F. (1991). “Comparison of measured and calculated dose rates in water near I-125 and Ir-192 seeds.” Med Phys 18:776–786. Williamson, J. F. “Physics of Brachytherapy” in Principles and Practice of Radiation Oncology. 3rd edition. C. A. Perez and L. W. Brady (eds.). Philadelphia: Lippincott-Raven Press, pp. 405–467, 1997. Williamson, J. F., “Low Energy Photon Source Dosimetry” Proceedings of 2000 World Congress on Medical Physics and Biomedical Engineering, edited by G. D. Fullerton, IEEE Engineering Medicine and Biology Society, 2328 July 2000, Chicago, IL, MO A313-01. Williamson, J. F., and Z. Li. (1995). “Monte Carlo aided dosimetry of the microSelectron pulsed and high dose rate 192 Ir sources.” Med Phys 22:809–819. Williamson, J. F., and A. S. Meigooni. “Quantitative Dosimetry Methods in Brachytherapy” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.). Madison, WI: Medical Physics Publishing, pp. 87–133, 1995. Williamson, J. F., H. Perera, Z. Li, and W. R. Lutz. (1993). “Comparison of calculated and measured heterogeneity correction factors for 125I, 137Cs, and 192Ir brachytherapy sources near localized heterogeneities.” Med Phys 20:209–222. Williamson, J. F. B. M. Coursey, L. A. DeWerd, W. F. Hanson, and R. Nath. (1998). “Dosimetric prerequisites for routine clinical use of new low energy photon interstitial brachytherapy sources.” Med Phys 25:2269–2270.
324
Mark J. Rivard
Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. H. Hanson, R. Nath, and G. Ibbott. (1999a). “Guidance to users of Nycomed Amersham and North American Scientific, Inc., I-125 interstitial sources: Dosimetry and calibration changes: Recommendations of the American Association of Physicists in Medicine Radiation Therapy Committee Ad Hoc Subcommittee on Low-Energy Seed Dosimetry.” Med Phys 26:570–573. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, R. Nath, M. J. Rivard, and G. Ibbott. (1999b). “On the use of apparent activity (Aapp) for treatment planning of 125I and 103Pd interstitial brachytherapy sources: Recommendations of the American Association of Physicists in Medicine Radiation Therapy Subcommittee on Low-Energy Brachytherapy Source Dosimetry.” Med Phys 26:2529–2530. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, R. Nath, M. J. Rivard, and G. Ibbott. (2000). “Recommendations of the American Association of Physicists in Medicine on 103Pd interstitial source calibration and dosimetry: Implications for dose specification and prescription (AAPM Report No. 69).” Med Phys 27:634–642. Williamson, J. F., W. Butler, L. A. DeWerd, M. S. Huq, G. S. Ibbott, Z. Li, M. G. Mitch, R. Nath, M. J. Rivard, and D. Todor. (2005). “Recommendations of the American Association of Physicists in Medicine regarding the impact of implementing the 2004 Task Group 43 report on dose specification for 103Pd and 125I interstitial brachytherapy.” Med Phys 32:1424–1439. Yu,Y., L. L. Anderson, Z. Li, D. E. Mellenberg, R. Nath, M. Schell, F. M. Waterman, A. Wu, and J. C. Blasko. (1999). “Permanent prostate seed implant brachytherapy: Report of the American Association of Physicists in Medicine Task Group Report No. 64.” Med Phys 26:2054–2076. Zelefsky, M. J., and W. F. Whitmore, Jr. (1997). “Long-term results of retropubic permanent 125iodine implantation of the prostate for clinically localized prostatic cancer.” J Urol 158:23–30.
Chapter 17
Treatment Planning Considerations of Brachytherapy Procedures Ali S. Meigooni, Ph.D.,1 and Robert E. Wallace, Ph.D.2 1 University of Kentucky, Lexington, KY 2 Cedar-Sinai Medical Center, Los Angeles, CA Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326 Calculation Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326 Linear Source Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326 TG-43 Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 326 Sievert Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327 Interpolation Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 Point Source Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 TG-43 Formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329 Traditional Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330 Curvilinear Line Source Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 330 Line Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Stranded Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Source Data Entry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Linear Source Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 2-D TG43U1 Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Tissue Attenuation Coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333 Look-up Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333 Point Source Approximation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333 1-D TG-43U1 Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333 Gamma Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334 Look-up Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334 Specific Features in Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334 Calculation Geometries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334 Imaging Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335 Radiographic Reconstruction: Orthogonal Films . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335 Radiographic Reconstruction: Linear Stereo-Shift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335 Radiographic Reconstruction: Rotation Stereo-Shift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336 Radiographic Reconstruction: Three or More Film Noncoplanar Film and Fiducial Jigs . . . . . . 336 Volumetric Reconstruction: DICOM Image Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337 Volumetric Reconstruction from CT Image Series . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337 Real-Time Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338 Quality Control of Treatment Planning Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338 AAPM Task Group 43 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339 AAPM Task Group 53 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340 AAPM Task Group 56 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342 AAPM Task Group 64 Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342 Implementations and Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344 Recommendations for Data Entry in Varian Planning Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344 Recommendations for Data Entry in Prowess Planning System . . . . . . . . . . . . . . . . . . . . . . . . . . . 344 Recommendations for Data Entry in the ADAC Pinnacle Planning System . . . . . . . . . . . . . . . . . . 345 Recommendations for Data Entry in Nucletron TheraplanTM and SPOTTM Brachytherapy Planning Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345 Shortcomings and Recommendations in the Present Planning Systems . . . . . . . . . . . . . . . . . . 345 Linear Source Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
325
326
Ali S. Meigooni and Robert E. Wallace Point Source Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346 Interpolation and Extrapolations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346 Strategies To Implement TG43U1 Parameters in Systems that Do Not Support the TG-43 . . . . . . 346 Nomogram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
Introduction The continuous improvements of the brachytherapy treatment modality is partly due to the advancement of the technical aspects of source design and treatment delivery, and partly due to the development in the treatment planning systems and quality controls in the program (Russel and Blasko 1993; Nori 1992; Porrazzo et al., 1992; Blasko et al.,1995; Sommerkamp, Ruppercht, and Wannenmacher 1988; Willins and Wallner 1998; Merrick et al., 1998, 2000; Fuks et al., 1991; Kumar and Good 1986; Tapen et al., 1998). Although there are still several shortcomings in these aspects of brachytherapy, a review of the existing programs and protocols will give us guidelines for future activities. In this chapter, the existing treatment planning methods will be reviewed and the shortcomings of each technique will be presented.
Calculation Algorithm Dose calculations in brachytherapy are divided into two general categories of linear and point source approximations. The main distinctions between these two categories can be related to the fact that if the distribution of the activities within the source and the its self absorption can be neglected or it has to be considered in the dose calculations. Generally, when the calculation points was located at a distance greater than twice of the active length of the source, or the energy of the photon emission was greater than 100 keV, the source can be assumed as an isotropically emitting radiation source, and otherwise it will be assumed as a linear source. In the following sections, the dose calculation methods for each of these two categories are presented.
Linear Source Approximation Dose distributions around a linear source are generally calculated using one of the following three methods. TG-43 Algorithm In 1988, Task Group 43 (TG-43) of the American Association of Physicists in Medicine (AAPM) was started, and introduced a new protocol for dosimetric characterization and dose calculation formalism for brachytherapy sources (Nath et al., 1995). Since then, many investigators have determined the dosimetric characteristics of various source models and hence, some shortcomings of the original TG-43 protocol have been discovered and hence an updated AAPM TG-43 protocol (also known as AAPM TG-43U1) has been introduced (Rivard et al., 2004). The formalism introduced by this protocol is as follows:
D" (r, θ ) = SK ⋅ Λ ⋅
GL (r , θ ) GL (r0 , π / 2 )
⋅ gL (r ) ⋅ F (r, θ ),
(1)
17–Treatment Planning Considerations of Brachytherapy Procedures
327
where · D (r,θ ) is the dose rate at point P(r,θ) as shown in Figure 1; SK is the air kerma strength; Λ is the dose rate constant; GL(r,θ) is the geometric function; gL(r) is the radial dose function; and F(r,θ) is the 2-D anisotropy function. The readers are referred to the AAPM TG-43U1 protocol for a detailed description of these parameters (Rivard et al., 2004). Moreover, the members of this task group are reviewing the dosimetric characteristics of the different brachytherapy source models, and derive the consensus of the multiple measured and Monte Carlo simulated data for each source type. These consensus datasets for source parameters will be available in their future publications. The TG-43 formalism can easily be adapted in any computerized treatment planning systems. Sievert Integral For many years dose distributions around a linear brachytherapy source were calculated using the Sievert Integral (Sievert 1921). In this method, a linear source subdivided into several equal segments (Figure 2) where each segment was small enough that could be assumed as a point source. The integration of the exposure rate contributions from each segment to a given point was then calculated using point source approximation. The final integrated exposure rate was presented as: I ( x, y ) =
M eq .Γ Ra Ly
eµ t
'
∫
θ2
θ1
e− µ t sec θ dθ '
'
(2)
where Meq = Total source strength in term of mgRaEq. ΓRa = Exposure rate constant of 226Ra source, with 0.5 mm of platinum filter. L = Active length of the source. µ′ = Effective attenuation coefficient of the filter (source capsule). t = Thickness of the source capsule. If the source strength is specified in term of air kerma strength (SK), this equation will changed to I ( x, y ) =
SK eµ t W Ly( ) e
'
∫
θ2
θ1
e− µ t sec θ dθ ' , '
(3)
where W/e is defined as work function that is the average energy absorbed per unit charge of ionization in air. The original Sievert integral has corrected the exposure for the attenuation of the radiation by the source capsule and eliminated the self-absorption by the core of the source. However, Shalek and Stovall (1969) had modified this integral to consider the self-absorption of the source, as well. Depending on the type
328
Ali S. Meigooni and Robert E. Wallace
Figure 1. Schematic diagram of a brachytherapy source, used in calculation of dose rate at point P, from a linear source, using the AAPM TG-43 formalism.
Figure 2. Schematic diagram of a brachytherapy source, illustrating geometric relationships used in calculation of exposure at point P, from a linear source, using the Sievert Integral method.
of source and filtration, the energy spectrum may be significantly altered by the filter. This Sievert Integral can be evaluated by numerical analysis. Because this integral uses the energy absorption coefficient, the underlining assumption in this method is that the emitted energy fluence is approximately attenuated by filter thickness traversed by the photon. This is an approximation that has been shown to work well for 226Ra and 192Ir sources in region bounded by the active source ends (Shalek and Stovall 1969; Williamson, Morin, and Khan 1983). However, Monte Carlo simulations (Williamson, Morin, and Khan 1983) have shown that beyond the end of the active source region, the Sievert approach introduces significant error and practically breaks down in the extreme oblique directions. The treatment-planning systems using the Sievert integrals, acquire the exposure rate constant of the source, its geometric information such as the active length, physical length, capsule thickness, and the attenuation coefficients of the capsule and the core of the source. In addition, the values of exposure to dose conversion factors (f-factor), half-life, and the radial dose function or tissue air ratio of the source are needed to generate a dose distribution matrix in an along-away format. The tissue air ratio for some of the initial sources introduced for brachytherapy treatments were determined by Meisberger et al. (Meisberger, Keller, and Shalek 1968). These values were given in the forms of a third order polynomial coefficient, more commonly known as Meisberger coefficients. If the source strength is given in term of the air-kerma strength, the exposure rate constant value, Γ, should be replaced (W/e)–1= 1.142.
17–Treatment Planning Considerations of Brachytherapy Procedures
329
Interpolation Method A third method of dose calculations around a linear source is interpolation of a set of data points that has been determined using either experimental method or theoretical calculations. The along-and-away table of Krishnaswamy (1972) for the 137Cs source or tabulated data in Paterson-Parker’s system (Meredith 1967) and Quimby system (Glasser et al., 1961) for radium equivalent sources are among the most commonly used data in calculations of the dose rate from a linear source, using the interpolation methods. These tabulated data are in a matrix form with a given format that can easily be applied for clinical applications, particularly for a quick double check dose calculation with a simple calculator. It should be noted that some of these tables are only applicable for specific source geometry and radionuclides. For example, the along-and-away table of Krishnaswamy (1972) is only valid for the 137Cs radionuclide with the source geometry that has been specified in that publication. Although this method of calculation is simple, its accuracy is not warranted. Moreover, there are different tabulated data from various source models. Thus, it is not easy to prepare consensus datasets.
Point Source Approximation Although most of the brachytherapy sources are linear in shape, however, under the following conditions the source geometry can be ignored and an isotropic point source can be utilized during the treatment planning of an implant patient. These conditions are: (a) The distance from the source is far enough (r > 2L) that the distribution of the activity within the source will not have any impact to dose calculation at a given point. (b) The random orientation of the brachytherapy sources within the implant volume will minimize the effects of dose anisotropy within the implant volume. The following methods are describing the dose calculations using the point source approximation technique. TG-43 Formalism The 2-D formalism introduced in the AAPM TG-43U1 protocol can be simplified as D" (r ) = SK ⋅ Λ ⋅
GL (r, θ ) GL (r0 , π / 2 )
⋅ gL (r ) ⋅ ϕ ave (r ),
(4a)
Or
r D" (r ) = SK ⋅ Λ ⋅ o g p (r ) ⋅ ϕ ave (r ), r 2
(4b)
Where the ϕan(r) is called 1-D anisotropy function at the radial distance of r which is calculated using the following equation as: π •
ϕ an (r ) =
∫ D(r,θ ) Sin(θ ) dθ 0
•
2 D (r, θ o )
(5)
330
Ali S. Meigooni and Robert E. Wallace
where gp and gL are the radial dose functions of the source determined using point or line source approximations, respectively. Although the updated TG-43 formalism recommends use of equation (4a) in order to achieve better accuracy at short distances (r < 1 cm), commercially available treatment planning systems generally utilize equation (4b). Therefore, dose calculations with the point source approximation require the dose rate constant of the source, 1-D anisotropy function, and radial dose function (i.e., gp) of the source. In case of the permanent implant, the half-life of the source also would be required. Traditional Algorithm The traditional algorithm for dose calculations around brachytherapy sources were based on Γ of the source corrected for the source anisotropy constant measured in air, apparent source activity (App), and tissue air ratio (B(r)) as shown below: 1 D" (r ) = Gamma ⋅ App ⋅ 2 ⋅ f ⋅ T ⋅ DC ⋅ ⋅B(r ) ⋅ 100 , r
(6)
where, f is the exposure to dose conversion factor (R/cGy), T is the exposure time (hours), and DC is the source decay correction during the treatment defined as: DC = 1.443 ⋅ Half − Life ⋅ [1 − exp(0.693 ⋅ T / Half − Life)
B(r ) = [ exp(− µ ⋅ r )
]
]
(7)
(8)
For low-energy photon-emitting brachytherapy sources, this calculation method has been considered to be in error since the exposure rate measured in air and its conversion to dose to tissue has been inconsistent between the user and vendors. Moreover, the vendor determined Aapp by measuring the exposure rate of the source at a distance and dividing that by the gamma constant of the source, which was not necessarily the same as the Γ values used by the physicists.
Curvilinear Line Source Approximation Line Source Flexible sources such as 192Ir wires may be bent into arbitrary shape. There is no simple or general mathematical procedure to calculate the dose at a point. The most conventional method is to proceed by subdividing the curved source into a train of small source segments. The total dose to a given point is calculated by superposition of the dose contributions from each source segments (Rosenwald 1997; Marinello and Pierquin 1997; Meigooni et al., 2004). It is not necessary for all the segments to have the same length, rather the length of each segment should be selected in order to reproduce the curvature of the wire. Also, the tabulated dose distributions around a source wire, as a function of the length of the wire are available for some of the brachytherapy sources (Rosenwald 1997). One may also present such results graphically in an easy-to-use form. Figure 3 shows a sample of isodose curves (also known as isodose “snail”) for 192 Ir wires as a function of active lengths ranging from 1 to 7 cm (Pierquin and Marinello 1997; Schlienger et al., 1970). From these data, one can calculate the doses actually delivered by multiplying the tabulated data by the reference source strength and the treatment duration time. The isodose values obtained for the
17–Treatment Planning Considerations of Brachytherapy Procedures
331
Figure 3. The “Snail” isodose for 192Ir wires of length ranging from 1 to 7 cm. Each curve corresponds to a given dose rate (cGy/day) for sources of unit linear reference kerma rate in central plane of the wire. (Reproduced from Pierquin and Marinello (1997) with permission of Medical Physics Publishing.)
central plane are valid as long as the plane of calculations is at a distance less than one-quarter of the length of the wire. Stranded Source A stranded source with the center-to-center spacing between the adjacent sources of S, can be closely approximated by a continuous linear source of length L = N*S, where N = number of souces in the strand. It should be noted that the length L is larger that greater than the distance stranded sources. The excess distances on each separating the extreme point sources of the end of the equivalent continuous line source and stranded sources are S/2 (Figure 4).
Source Data Entry Linear Source Approximation 2-D TG43U1 Parameters The AAPM TG-43U1 2-D dosimetric characteristics (i.e., dose rate constant, radial dose function, and 2D anisotropy function) can be easily entered in the source files of any treatment planning systems. These parameters are given either as a tabulated data or in the form of polynomial coefficients. It should be noted that, the polynomial coefficients are only warranted within the range of the data that the coefficients have been fitted. Any extrapolation from these coefficients may lead to large and unexpected errors. The users are recommended to verify the algorithm and methodology of dose calculations outside of the given range of data. The 1995 TG-43 protocol recommends the 5th order polynomial fit [equation (9)] to the radial dose function of the sources as, g(r ) = a0 + a1r + a2 r 2 + a3r 3 + a4 r 4 + a5 r 5
(9)
332
Ali S. Meigooni and Robert E. Wallace
Figure 4. A tandem of N sources in a strand form compared with an 192Ir wire with continuous activity distribution. (Reproduced from Pierquin and Marinello (1997) with Permission of Medical Physics Publishing.)
However, several investigators have shown that the polynomial fit could not accurately represent the radial dose functions with large variations at short or extended distances (Furhang and Anderson 1999; Meigooni et al., 2003). Furhang and Anderson (1999) have introduced the double exponential fit [equation (10)] and Meigooni et al. (2003) have introduced the modified 5th order polynomial fit [equation (11)] for more accurate representation of these radial dose functions. g(r ) = C1e( − µ r ) + C2 e( − µ r ) 1
(
(10)
2
)
g(r ) = a0 + a1r + a2 r 2 + a3r 3 + a4 r 4 + a5 r 5 e− br
(11)
Figures 5a and 5b shows a sample of these types of curve fits. It should be noted that in equation (10), C2 was recommended to be a negative value. Therefore, introducing a positive number for the initial value of this parameter during the curve fit may lead to an erroneous result. These authors recommend using either equation (10) or (11) in the brachytherapy treatment planning systems. Meanwhile, the physicist could utilize these fitting routines in order to generate a larger number of data points from a coarse dataset. This dataset can be used in treatment planning systems, which incorporates interpolation methods in patient dosimetry. The 2-D anisotropy functions can be introduced in tabulated form. However, it should be noted that, some of the treatment planning systems require the 2-D anisotropy functions for the angles ranging from 0 to 90° degrees, and some require data from 0 to 180°. It is interesting to know that, although the 2-D dosimetric parameters of the source can be entered in the treatment planning systems, they may not be using that properly. For instance, some prostate implant treatment planning systems assume that the source directions are normal to the CT or ultrasound image plane, and do not use true source directions for the treatment planning. For true linear source calculations, the users should be able to introduce the location and direction of the seeds by digitizing either the two ends of each source, or introduce the center and orientation of the source (e.g., digitizing the catheter or applicators holding the source).
17–Treatment Planning Considerations of Brachytherapy Procedures
333
Figure 5. Comparison of the 5th order polynomial fit with (a) modified polynomial fit and (b) double exponential fit for radial dose function of BrachySeed Pd-103 model Pd-1 source.
Tissue Attenuation Coefficients The traditional treatment planning systems acquire the source geometric information such as the active length, physical length, capsule thickness, and the attenuation coefficients of the capsule and the core of the source. In addition to Γ, f-factors, and half-lives, Meisberger coefficients are needed to generate a matrix of dose distribution in an along-away. Look-up Table A look-up table configured in “along and away” Cartesian rectangular coordinates centered on a line source center can be used to represent the two-dimensional distribution of dose rate about a prototype source of a given design, having unit strength. Such tables are interderivable with the polar-coordinate based TG43 dose-rate tables for unit air-kerma strength. The advantage of such tables is that dose calculation points are typically distributed on rectangular grids in treatment planning systems. The process of mapping the dose-rate contributed by a given source is simplified by reduction to a process of interpolation, and the planning system speed improves. Printed look-up tables allow a clinical physicist to estimate the dose to selected points by a “hand calculation.” Look-up tables for other quantities than dose rate are possible. For example, a treatment planning system may not fully support TG-43 parameters for dose calculation. Suppose that such a system did not support the geometry factor, G(r,θ), and the anisotropy function, F(r,θ), as distinct entities but supported one normalized relative dose table in two dimensions (either “along and away” or the polar coordinate systems). Given the TG-43 parameters, one could compute the product, G(r,θ)*F(r,θ), and fill in the table. Thus, one may combine TG-43 parameters for a given source to fit the requirements of the specific treatment planning system.
Point Source Approximation 1-D TG-43U1 Parameters The 1-D TG43U1 dosimetric characteristics (i.e., dose-rate constant, radial dose function, and 1D anisotropy function) of the brachytherapy sources can be easily entered in the source files of any treatment planning systems. Similar to the linear source data entry, these parameters can be given either as a tabulated data or in the form of polynomial coefficients. Although the 2004 TG-43U1 report recommends use of equation (4a) in order to achieve a dose calculation accuracy at short distances (r < 1 cm), commercially available treatment planning systems generally utilize equation (4b). Therefore, dose calculations
334
Ali S. Meigooni and Robert E. Wallace
with this methodology require the point source approximation radial dose function (i.e., gp) of the source. Similar to the above description for data entry of the linear source models, care should be taken for introducing the radial dose function of the brachytherapy sources. Both, the original and updated TG-43 protocols, recommend the tabulated radial dose function using the 5th order polynomial fit. The 1-D anisotropy functions can be introduced in tabulated form. The updated TG-43U1 has given the 1-D anisotropy of brachytherapy seeds for radial distances ranging from 0.5 to 7 cm. This protocol – recommend using the modified dosimetry parameters, g′(r) and φ a′n defined as, g′(r ) = g (r ) ⋅ φ an
(12)
φ an′ = 1
(13)
Gamma Factors Γ values for a brachytherapy source are defined as the exposure rate at a reference distance along the transverse bisector of the source, per unit source strength. Γ has units of either R◊cm2/(h◊mCi), if the source strength is expressed in terms of Aapp, or R⋅cm2/(h⋅mg) if the source strength is expressed in terms milligram radium equivalent. Γ values for commonly available brachytherapy sources can be found in most of the radiation therapy physics textbooks (Meigooni, Saw, and Nath 1997; Khan 1992). Source strength units must be consistent with Γ units entered into a treatment-planning system. Look-up Table A look-up table configured in “along and away” Cartesian rectangular coordinates centered on a line source center can be used to represent the two-dimensional distribution of dose-rate about a prototype source of a given design, having unit strength. In principle, such a table could be derived for a point source. Such tables are interderivable with the polar-coordinate based TG-43 dose-rate tables for unit air-kerma strength. For the case of a line source, look-up tables for quantities other than dose-rate are possible. For example, a treatment planning system may not fully support TG-43 parameters for dose calculation. Suppose that such a system did not support the radial dose function g(r), and the anisotropy factor, φan(r), as distinct entities, but instead supported a relative dose-rate table in one dimension. Given the TG-43 data, one could compute the product, g(r)* φan(r), and fill in the table by calculating the distance r(x,y) for points on a rectangular grid. Thus, requirements of the treatment-planning system can be met with recommended TG43 consensus dataset values without requiring that the planning system be modified.
Specific Features in Planning Calculation Geometries The geometry of the brachytherapy implants is normally determined by the shape and location of the target volume. For instance, brachytherapy treatment of endobroncheal cancer is normally performed using a linear source, while an interstitial breast implant is most commonly either a single or double plane implant. Whether calculations are pursued at selected points, in a plane, in a set of parallel or orthogonal planes, or at a number of points regularly or randomly distributed in a volume, each approach will requires a coordinate system representing points that map from the virtual world of the plan to the real world of patient
17–Treatment Planning Considerations of Brachytherapy Procedures
335
anatomy. The dimensions, locations, and often the orientations of the brachytherapy sources relative to the treatment volume must therefore be clearly identified. The relation between the implant sources and target volume can be identified using one of the following imaging techniques. The relationship of the target volume to the plan coordinates can be similarly defined.
Imaging Support Radiographic Reconstruction: Orthogonal Films Orthogonal radiographs are taken at right angles, with the central axis of the x-ray beams meeting approximately in the middle of the implant. Usually, AP and lateral films, exposed isocentrically, provide such geometry. The coordinate system is conventionally established with the x-axis from the right to the left of the patient, the y-axis from inferior to superior, and z-axis from posterior to anterior. The AP film represents a magnified view of the implant image projected onto x-y plane, while the lateral film presents the image projected onto the z-y plane. The origin of the coordinate system is chosen to be a common point on both films such as one end of a source or applicator. For a typical treatment planning, the origin and the x-axis on the AP film and y-axis on the lateral film must be identified before brachytherapy source localization can be performed. The points of interest must be identified relative to the applicators (i.e., point A) or relative to the fiducial markers (i.e., bladder) placed in the patient. The orthogonal radiographic technique is not applicable for implants where the sources are masked by an object or skeletal bone. For example, the brachytherapy sources in the pelvic area for the GYN or prostate implant patients with femoral prostheses cannot be seen in the lateral film. Moreover, the brachytherapy seeds in the prostate implant patients are not clearly visible in the lateral film since the femoral bone easily blocks them. The alternative methods for these patients are linear or rotational stereoshift techniques in additional to CT, MRI, and ultrasound imaging techniques as described below. Radiographic Reconstruction: Linear Stereo-Shift The linear stereo-shift method of source localization consists of taking two radiographs of the patients from the same view, but shifting either the patient or x-ray tube by a certain distance (i.e., 20 cm) between the two exposures. Figure 6 shows the principle of the linear stereo-shift technique, for identification of two sources. Using the similar triangle rules for the two triangles of T1BT2 and B1BB2, the coordinates of point A can be related to those of point B as: d y1 + s − y2
=
F−Z Z
.
(14)
Also, from similar triangles of T1AT2 and A2AA1 it can be found that: d s
=
F− f f
.
(15)
From these two equations one can find: Z=F
( F − f )( y2 − y1 ) − d . f ( F − f )( y2 − y1 ) − d .F
.
336
Ali S. Meigooni and Robert E. Wallace
Figure 6. Schematic diagram of the source and target geometry in the linear stereo-shift radiographic technique.
Radiographic Reconstruction: Rotation Stereo-Shift The stereo-shift technique can be achieved by rotating the x-ray tube, instead of shifting that in a linear fashion. This procedure is performed with a x-ray machine with a rotational capability. Figure 7 demonstrates a typical rotational stereo-shift technique. The calculations of the coordinates is the same as the linear stereo-shift, except for the distances between the two target position, d, which is calculated as: d = 2 R .sin(θ / 2 )
(17)
Other rotational methods exist, using films that are orthogonal to the beam central axis (i.e., the dashed lines in Figure 7). An excellent resource for x-ray photogrammetry methods can be found in Hallert (1970). Radiographic Reconstruction: Three or More Noncoplanar Films Determination of dose distribution in complex implant geometry, such as prostate implant, may require more complex imaging technique than just orthogonal film imaging. In principle, a set of three (or more) non-coplanar images can be used with a fiducial jig having four or more radiopaque markers to reconstruct the locations of sources in three dimensions (Altschuler, Findla and Epperson 1983; Altschuler and Kassaee 1997; Amols and Rosen 1981; Biggs and Kelly 1983; Jackson 1983; Rosenthal and Nath 1983; Su et al., 2004; Todor et al., 2002). The non-coplanarity of the beams provides views that can be useful to sort seeds, which may be occluded in other images by anatomical structures or other sources. Use of a jig with embedded markers set out in a three-axis grid provides a coordinate system. The requirement of four or more markers improves the determination of the three-dimensional (3-D) coordinate system from the radiographic images of the fiducial markers. Since the geometry of the jig is known, the coordinate transformations can be determined via linear least squares or algebraic estimation. The greater the number of fiducial markers, the better is the estimation. The same is true of the number of films. No attempt
17–Treatment Planning Considerations of Brachytherapy Procedures
337
Figure 7. Schematic diagram of the source and target geometry in the rotational stereo-shift radiographic technique.
will be made here to provide details of such methods due to their complexity. However, there has been much work in the field on such methods and Su et al. (2004) is a clear point of departure. Volumetric Reconstruction: DICOM Image Sources The Digital Imaging and Communications in Medicine (DICOM) standard was created to aid the distribution and viewing of medical images, such as CT scans, MRIs, and ultrasound (NEMA 2004). The National Electrical Manufacturers Association (NEMA) introduced this format of distribution of images. A single DICOM file contains both a header (which stores information about the patient’s name, the type of scan, image dimensions, etc), as well as all of the image data (which can contain information in three dimensions). Most modern brachytherapy planning systems have facilities to import and use images (and image sets) that are presented via some DICOM mechanism. The DICOM standard specifies the format of files containing the medical (or other, dose and 3-D structures, for example) images and specifies the requirements for file transport and interchange using digital media. Exchange may be facilitated by copying to a portable, interchangeable Compact-Disk (CD) ROM, through the agency of Universal Serial Bus (USB) based “solid-state disks”, or by direct network connection, email, etc. While such systems may have complex design and implementation, their use is relatively simple by virtue of the DICOM standard. Volumetric Reconstruction from CT Image Series There has been much effort over many years in defining and automating seed detection, localization, and (unique) identification in CT axial series (Bice et al., 1999; Brinkmann and Kline 1998; Li et al., 2001; Liu et al., 2003; Roy et al., 1993; Tutar, Managuli, and Shamdasani 2003). Tubic et al. (2001) described simulated annealing methods to sort and cull possible seed locations as an optimization problem. Holupka et al. (2004) describe a machine vision application using the elliptical Hough transform to find and sort
338
Ali S. Meigooni and Robert E. Wallace
possible seed locations. 2004). The common denominator of all these methods is use of axial CT source images that are reconstructions using sampled raw CT detector data. The spatial resolution is therefore limited by CT voxel dimensions and the Nyquist rate for sampling in and among the axial images. In this setting, the achievable spatial resolution of detected seed positions is on the order of one to two millimeters with no information about seed orientation. This derives from the fact that the pixel sizes are approximately the same diameter of cylindrical brachytherapy sources (about 0.8 mm) and the interslice spacing is the order of the length of the sources (i.e., 5 mm). Recently, Tubic and Beaulieu have reported promising initial findings using raw CT detector sinogram data to provide unambiguous seed localization with positional accuracy of 0.2 mm and orientation determination to within 3° about the source principal axis (Tubic and Beaulieu 2005). This method has logistical issues in that CT sinogram data , while a likely DICOM object, is not typically provided for use beyond image reconstruction in CT scanners. Real-Time Planning Traditionally, the ultrasound guided prostate seed implants were performed in three separate steps: (1) Volume study, a series of ultrasound images were collected 1 or 2 weeks prior to the implant. These images were used to pre-plane the treatment, and determine the seed activity, number of seeds and needles, and locations of the seeds. (2) Brachytherapy implant, the patient was prepared in the same position as the volume study, using the patient landmarks such as urethra and bladder. Then the needles, which were loaded by the seeds following the pre-plane, were placed in the patient. (3) Post implants CT-based evaluation. The patient dosimetries were evaluated using the CT images taken a month after the implant procedure. However, with the advancement of the technological aspects of the brachytherapy procedures, such as the Mick Applicator (Mick Radio-Nuclear Instruments Inc., Mount Vernon, NY) also known as “Mick Gun,” the first two procedures are combined into one. In this case, seeds of fixed activities are purchased prior to the implant, in special magazine that can be placed in Mick applicator. In this implant, the patient dosimetry is been performed as each needle is inserted in the patients, which is called “read time.” With this technique, once the treatment is over, the patient dosimetry is also completed.
Quality Control of Treatment-Planning Systems Several reports of AAPM task groups provide guidance on the quality assurance (QA) of treatment-planning systems in general with some mention regarding such systems for brachytherapy. The general recommendation is that each component of the system be tested with appropriate independent and standard methods. In each clinical use of a brachytherapy planning system, the resulting plans should be verified using an independent, if idealized, system. Hand calculation of dose at selected points assuming exposure and/or dwell times, source strengths, realistic source positions should be performed to verify a specific plan. Alternatively, a second computer-based system that implements the hand calculation (i.e., in a general purpose spreadsheet program) could suffice for the check if it had itself been earlier verified. A patient treatment can be defined by specific data: the input data for source position (and orientation), shield position (and orientation), source strength, prescription dose to a point or to an isodose surface enclosing a volume, and the dose-rate data for single sources of the design used (i.e., TG-43 or similar data). Using appropriate patient and treatment specific information, the responsible medical physicist can perform an end-to-end secondary calculation that validates the more complex plans generated by a dedicated treatment-planning system. Thus, the general statement is one of prudence and caution. While the present discussion revolves around treatment-planning systems, the QA of a part of the brachytherapy system cannot realistically be performed in a vacuum except to prove consistency of that system. The goal is to ensure that all compo-
17–Treatment Planning Considerations of Brachytherapy Procedures
339
nents used in a brachytherapy treatment perform as intended and as expected. Each needs be tested, alone and as a part of the complex whole.
AAPM Task Group 43 Recommendations The updated TG-43 report (Rivard et al., 2004) suggests that, before clinical application of a new radiotherapy treatment planning (RTP) system or a new source model on an established RTP system, a thorough acceptance testing and commissioning must be carried out. This protocol recommends that the users document the results of their tests for later reference, as well as the compliance with applicable regulations. Moreover, it recommends that the dose rates calculated by the RTP system from a single source should be compared with the dose-rate distribution derived from the tabulated parameters and equations presented in the protocol. In addition, this recommendation suggests that the user should compare the RTP system calculations with the dose-rate distributions derived from the appropriate 1-D or 2-D data tables from the protocol. To facilitate this comparison for the brachytherapy sources included in this protocol, a tabulated transverse-plane dose rates as a function of distance was provided (Table 1). As has been recommended previously by the AAPM, this comparison should yield agreement within ±2% over all angles and over the range of radial distances commissioned (Kutcher et al. 1994; Fraass et al., 1998; Nath et al., 1997). If deviations between calculated results and the published data exceed ±2%, the deficiencies of the RTP system should be documented and further investigated by the user. This is especially important for RTP systems that fit a model to the published data, because such models are prone to large errors outside the range of the reference data. Moreover, this protocol suggests that, in the high dose-gradient regions close to a source, particularly near the ends of a source, the acceptable error may need to be larger. These deficiencies should be considered when evaluating treatment plans, and when considering the purchase of an RTP system.
Table 1. Dose Rates (cGy·h–1U–1) as a Function of Distance for Eight Brachytherapy Sources Using the 1-D Dosimetry Formalism of Equation (11) with Interpolation for gL(r) and φan(r) r (cm)
Amersham Amersham model model 6702 6711
Best model 2301
NASI model MED3631A/M
Bebig model I25.S06
Imagyn model IS-12501
Theragenics NASI model model 200 MED3633
0.5
4.119
3.937
3.978
4.112
3.922
3.426
3.014
3.184
1
0.995
0.911
1.004
0.986
0.95
0.815
0.587
0.626
1.5
0.413
0.368
0.419
0.42
0.398
0.334
0.199
0.215
2
0.213
0.186
0.217
0.207
0.205
0.169
0.0837
0.0914
3
0.0768
0.0643
0.0783
0.0746
0.0733
0.0582
0.0206
0.0227
4
0.0344
0.0284
0.0347
0.0325
0.0323
0.0246
0.00634
0.00697
5
0.0169
0.0134
0.0171
0.0157
0.0157
0.0118
0.00221
0.00247
6
0.0089
0.00688
0.00908
0.00811
0.0084
0.00592
0.000846
0.000933
7
0.0049
0.00373
0.00506
0.00429
0.00459
0.00328
0.000342
0.000364
(Reproduced from Table XV in the AAPM TG43U1 report: “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31: 633–674. M. J. Rivard, B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. Also available as AAPM Report No. 84. © 2004, with permission from AAPM.)
340
Ali S. Meigooni and Robert E. Wallace
To perform comparisons at the recommended level of precision, numerical point dose rates calculated by the RTP system, rather than measured diameters of plotted isodose contours, should be used. The user should evaluate any deviations from the recommended data, taking into consideration the uncertainty of calculations at very small distances. The calculation matrix should be adjusted to a value appropriate for the high gradients near a source. For low-energy photon-emitting sources, voxel size should not exceed 1×1×1 mm3; this size generally represents a reasonable compromise between calculation speed and accuracy. Isodose curves should be generated as part of the commissioning and continuing quality assurance procedures, but should be viewed as a test of the spatial accuracy of the graphic dose display function of the RTP system, rather than a definitive test of the underlying dose-calculation algorithm. Because comparisons should include both point dose-rate calculations and the placement of isodose lines, the user should also ensure that the RTP system and its graphical output devices cause isodose curves to appear in the correct locations relative to corresponding point calculations.
AAPM Task Group 53 Recommendations The AAPM Task Group 53 report assists the medical physicist in developing and implementing a comprehensive program of quality assurance for radiotherapy treatment planning (Fraass et al., 1998). This report was the first guidance on the topic of treatment planning QA from the AAPM, although there are several related reports, including the recent report from Task Group 40 on Comprehensive QA for Radiation Oncology (Kutcher et al., 1994). Further expansion of AAPM recommendations regarding treatment planning quality assurance is likely after the radiation oncology community accumulates some experience with the approach recommended in this report. In recent years, the increased complexity of the treatment-planning process required to support such procedures as conformal radiotherapy has led to the need for a comprehensive set of quality assurance guidelines that can be applied to treatment-planning systems that support this complex process. This task group has been charged by the AAPM to prepare this report recommending the scope and content of necessary quality assurance procedures and the frequency of tests, from acceptance testing, characterization and commissioning to routine QA of clinical system use. Radiotherapy treatment planning (RTP) has long been an important part of the radiotherapy treatment process, so assuring that the treatment-planning process is being performed correctly is thus an important responsibility of the radiation oncology physicist. In recent years, as 3-D and image-based treatment planning has begun to be practiced in numerous clinics, the need for a comprehensive program for treatment-planning QA has become even more apparent. The AAPM TG-40 report published an overall approach to QA for the therapy process (Kutcher et al., 1994), but this work includes only a very general discussion of brachytherapy treatment-planning QA issues. In this report, radiation oncology physicists were advised to create the appropriate QA program for the treatment-planning systems and processes used in their clinics. Although this QA program will vary widely between different clinics, use of this report should allow each clinic to concentrate its QA efforts on those areas of most importance. General Definitions and Aims The radiotherapy treatment-planning process is defined to be the process used to determine the number, orientation, type, and characteristics of the radiation beams or brachytherapy sources used to deliver a large dose of radiation to a patient in order to control or cure a cancerous tumor or other problem. Most often, treatment planning is performed with the assistance of a computerized treatment-planning system that helps the treatment planner and physician define the target volume, determine beam directions and shapes, calculate, and evaluate dose distributions. The AAPM TG-53 report codifies the planning process and identifies those that may require attention to maintain correct functionality. Their recommendations are found in Table 2.
17–Treatment Planning Considerations of Brachytherapy Procedures
341
Table 2. The Clinical Treatment Planning Process 1. Patient Positioning and Immobilization • Establish patient reference marks/patient coordinate system. 2. Image Acquisition and Input • Acquire and input CT, MR, and other imaging information into the planning system. 3. Anatomy Definition • Define and display contours and surfaces for normal and critical structures. • Geometrically register all input data (CT, MR), including registration with initial simulation contours, films, patient position, etc. • Define target contours, generate 3-D target surface using surface expansion, import target information from multiple imaging modalities. • Generate electron density representation from CT or from assigned bulk density information. 4. Beam/Source Technique • Determine beam or source arrangements. • Generate beam’s-eye-view displays. • Design field shape (blocks, MLC). • Determine beam modifiers (compensators, wedges). • Determine beam or source weighting. 5. Dose Calculations • Select dose calculation algorithm and methodology, calculation grid and window, etc. • Perform dose calculations. • Set relative and absolute dose normalizations. • Input the dose prescription. 6. Plan Evaluation • Generate 2-D and 3-D dose displays. • Perform visual comparisons. • Use DVH analysis. • Calculate NTCP/TCP values, and analyze. • Use automated optimization tools. 7. Plan Implementation • Align (register) the real patient with the plan (often performed at a plan verification simulation). • Calculate Monitor Units or implant duration. • Generate hardcopy output. • Transfer plan into record and verify system. • Transfer plan to treatment machine. 8. Plan Review • Perform overall review of all aspects of plan before implementation. (Reproduced from Table 1-2 in the Task Group 53 report: Med Phys 25:1773–1829, “American Association of Physicists in Medicine Radiation Therapy Committee Task Group 53: Quality assurance for clinical radiotherapy treatment planning,” B. Fraass, K. Doppke, M. Hunt, G. Kutcher, G. Starkschall, R. Stern, and J. Van Dyk. Also available as AAPM Report No. 62. © 1998, with permission from AAPM.)
342
Ali S. Meigooni and Robert E. Wallace
While general in definition, each step identified above has either identical or direct analogues in brachytherapy treatment planning. Other AAPM task groups provide more recommendations.
AAPM Task Group 56 Recommendations The recommendations of the AAPM Task Group 56 for quality assurance of treatment-planning systems are scant (Nath et al., 1997). In fact, the authors reported that they found that “Relatively little has been written on QA of clinical treatment planning systems in general and even less is available specifically for brachytherapy treatment planning systems.” This report identifies those areas specific to brachytherapy treatment planning that need to be addressed in a QA program: 1. Methods to reconstruct source positions and orientations in 3D from a number of imaging sources: radiograph sets, and 3D CT or ultrasound image sets. 2. Methods to predict or project catheter trajectory from the same image sets. 3. Graphics and image display system(s) linearity and correctness. 4. Mechanisms to assign source strengths and placement durations (HDR dwell through permanent implant) to each individual source. 5. Dose calculation algorithms. 6. Dose distribution optimization, evaluation, and representation algorithms. 7. Hard-copy documentation: accuracy of numeric and text information, geometric accuracy of printed representations of source locations and dose distributions. To these we add: 8. Methods to reconstruct shield or filter positions and orientations in 3-D from a number of imaging sources: radiographs, and 3-D CT, or ultrasound image sets. TG-56 report summarizes QA recommendations in its Table VIII, reproduced here at Table 3, and also recommends that treatment specifications, times, positions, and dose should be verified by a secondary calculation. This is similar to the practice of checking monitor unit calculations for each planned external beam therapy. An independent check of each treatment is often possible using a copy of the data that was entered into the treatment planning system (in the form of look-up tables of dose rate at an array of positions about a single source) and the data defining the specific treatment being checked.
AAPM Task Group 64 Recommendations The recommendations of the AAPM Task Group 64 regarding treatment-planning systems are that the medical physicist shall verify that the treatment planning system reproduces the values published in TG43 (et seq.) for single source dose at total decay for an 125I model 6711 or a 103Pd model 200 seed in the point source approximation (Yu et al., 1999). This serves to verify that the dose planning system complies with the dosimetric formalism of TG-43 (Nath et al., 1995). This corresponds to the recommendations of Task Group 56 for initial acceptance testing where the results of a planning system are compared to manual calculations. The Task Group 64 report further recommends that the medical physicist shall verify that the system similarly provides valid results for simple configurations of multiple seeds. This echoes the recommendations of the Task Group 56 report. Task Group 64 also recommends QA for imaging sources, equipment, implant templates, applicators, and accessories, and physical dosimeters. Well-chambers shall be ADCL calibrated on a schedule and
17–Treatment Planning Considerations of Brachytherapy Procedures
343
Table 3. Brachytherapy Computer Planning System Quality Assurance Function
Benchmark Data
Frequency
Verify geometric accuracy of I/O peripherals: digitizer, CT or ultrasound interface, and plotter.
Digitize/plot pattern of known geometry; for CT/US, image and reconstruct phantom implant
Monthly
Verify input parameters for all pre-calculated single source arrays
Published recommendations, source vendor’s mechanical drawings
Initially and annually
Verify dose, dwell time, and treatment time calculations at representative points for all source files
Published dose-rate tables; manual calculations
Initially, annually, new software version or source identity
Accuracy of single source isodoses
Point source output
Initially, new software version
Accuracy of multiple-source isodose contouring
Point source data for symmetric source arrays
Initially, new software version
Accuracy of plan rotation matrix
Constancy of point doses, source positions, and isodoses under repeated orthogonal rotations for symmetric source arrays
Initially, new software version
Consistency of printed plan documentation
Assumed input parameters
Every clinical use
Accuracy of coordinate reconstruction
Radiograph phantom with known catheter geometry (added this writing: phantom with known source position geometry)
Initially, new software version
Accuracy of electronic downloading of treatment parameters of afterloader
Comparison of treatment unit and planning system printed output
Initially, new software version, each treatment
Dose-volume histogram/implant figures of merit
* Use isotropic point source or segment of line source allowing analytic calculation of DVH
* Initially, new software version
* Constancy of test case DVH
* Annually
Optimization software
Run series of test cases based in idealized implant geometries of various sizes; Develop a sense of what optimization does tot an implant compared to uniform loading before trying it on patients
Initially, spot check when software changes by duplicating old cases
Overall system test
Run series of standardized plans to globally test all clinically used features
Initially, new software version, annually
(Reproduced from Table VIII in the Task Group 56 report: Med Phys 24:1557–1598, “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56,” Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997).” Also available as AAPM Report No. 59. © 1997, with permission from AAPM.)
344
Ali S. Meigooni and Robert E. Wallace
constancy checks using a long-lived radionuclide (preferably with energy comparable to the clinical source) should be performed prior to each use. Ion chambers, Geiger-Müller counters, and survey dosimeters should be calibrated on a schedule. Phantom-based verification of ultrasound imaging systems should be performed to ensure correct alignment of the physical template and the software generated template grid display. Fluoroscopy and radiographic units should be verified for linearity and minimized distortion on display. In particular, ultrasound systems QA should follow those laid out in the AAPM TG-1 report (Goodsitt et al., 1998).
Implementations and Factors Recommendations for Data Entry in Varian Planning Systems VariSeedTM 7.1, is the most recent version of this treatment planning system (Varian Medical Systems, Charlottesville, VA) available for clinical applications. This treatment-planning system utilizes 3-D dosimetry technique to generate dose distributions around a brachytherapy source using either the line or point source approximation. VariSeed accepts all the TG-43 dosimetric parameters (dose rate constant, radial dose function, 2-D anisotropy functions, 1-D anisotropy function, and anisotropy constant) for a particular source in a single library. This planning system only accepts one set of radial dose functions for point and linear source models, rather than two separate data, as recommended in TG-43U1. Therefore, two separate source files must be generated for each source, one for point source and one for linear source calculations. It should be noted that, as per TG-43U1 recommendations, one could use linear radial dose function for point source calculation, as long as the proper geometric function is used. At this stage, the VariSeed planning system doses not have this capability. EclipseTM v7.2.24 (Vision v6.5) is a general-purpose treatment planning system (Varian Medical Systems, Charlottesville, VA) that includes brachytherapy planning in its BrachyVisionTM module. Source data entry and considerations in Eclipse/BrachyVision mimic those for the VariSeed product except that the radial dose function is entered only by 5th order polynomial expansion.
Recommendations for Data Entry in Prowess Planning system There are three different versions of the Prowess planning systems (1370 Ridgewood Dr., Ste. 20 Chico, California) commonly in the United States. The Prowess 2D-planning system is a very user-friendly treatment-planning system for different types of implants using both points and linear sources. In this version of the treatment-planning systems, the source data can be entered for seeds or linear sources. The source parameters in this program can portion of this program required the geometric structures of the source, Γ values, exposure-to-dose conversion factors, and tissue-to-air ratio in the form of the 3rd order polynomial fit coefficients. For a linear source approximation, the program generates a 2-D matrix of dose distributions, using the Sievert Integral method. Although, the program has the option of the TG-43 formalism, this option is only used by replacing the gamma factors by the dose rate constant, placing f-factor=1, and including the anisotropy factors in the calculations. The attenuation coefficients for the source capsule and core must be introduced for linear source approximation. The tabulated data are not editable for the 2-D anisotropy of the radiation distribution around the source. Moreover, the tissue attenuation coefficients of the sources can be only introduced as 3rd order polynomial fits, rather than 5th order polynomial fit as suggested by the TG-43 recommendation. Recently the 3D-Version of the Prowess planning system has become available but provides only for point source approximation. The implementation for linear sources remains in development.
17–Treatment Planning Considerations of Brachytherapy Procedures
345
Recommendations for Data Entry in the ADAC Pinnacle Planning System The Pinnacle3 treatment planning system by Philips Medical Systems is a general-purpose radiation therapy-planning package that includes brachytherapy planning. Source data entry is by one of several methods that support legacy source data as well as TG-43U1 point and line source approximations for more modern sources. Once source data are entered, interconversion can be made between TG-43 parameters and other representations of dose rate about a generalized source. TG-43 radial dose functions may be entered in tabular form or via (up to) 5th order polynomial. The 1-D and 2-D anisotropy function data is entered in tabular TG43U1 form with a variable mesh size. Source strength in a wide variety of units and specific to the clinic inventory may also be entered. Once all data are entered and commissioned, an along-away table for each source type is generated.
Recommendations for Data Entry in Nucletron TheraplanTM and SPOTTM Brachytherapy Planning Systems The TheraplanTM software is a general-purpose RTP system that includes modules for brachytherapy planning. Sources are entered (“modeled” in Theraplan vernacular) by one of several methods providing back-compatibility with legacy source data and providing TG-43 point and line source(s). Once source data is entered, inter-conversion can be made between TG-43 style and other representations of dose-rate about a generalized source. Notably, Theraplan supports both the point and line source approximations of TG43U1 as separate sources. TG-43 radial dose functions may be entered in tabular form or via the Meisberger (or other) 5th order polynomial. Anisotropy data are entered in tabular TG43U1 form and the (outdated) anisotropy constant is supported for point source approximation modeled. The SPOT ProTM system is a special purpose intra-OP planning system for prostate implant planning available from Nucletron Corporation (Columbia, MD). Typical sources are 125I and 103Pd seeds and SPOT Pro takes only TG43U1 formatted data. The TG43U1 recommendation for having separate and different g(r) data for point and line source approximations is supported in SPOT Pro creating two versions of data for a given seed design, one using point source data, the other using line source data. All TG43U1 data are entered in tabular form.
Shortcomings and Recommendations in the Present Planning Systems Linear Source Calculations Despite the vast development of the TG-43 protocol in brachytherapy source dosimetry, the present algorithm is mostly applicable for short brachytherapy sources. Patel et al. (2001) have shown the use of cylindrical coordinates for treatment planning parameters of an elongated 192Ir source. Also they have introduced modified TG-43 algorithm for elongated sources, using the cylindrical coordinate system. Moreover, at the close vicinity of some brachytherapy sources, such as 192Ir, there may be significant contribution of dose from the transmitted electron from the core and the capsule of the source (Baltas et al., 2001; Patel et al., 2001). The present TG-43 algorithm does not cover the dosimetry of such mixed beams. In addition, the tissue heterogeneity corrections have not been adequately addressed in any brachytherapy dosimetry protocols or commercially available planning systems. The 5th order polynomial fit to the radial dose function fails in describing the data at short distances relative to the source axis, particularly for sources with activities distributed only at the two ends of the source. Double exponential method recommended by Furhang and Anderson (1999) and the modified polynomial fit recommended by Meigooni et al. (2003) are more accurate techniques for fitting the data.
346
Ali S. Meigooni and Robert E. Wallace
Point Source Calculations Application of point source approximation may be suitable for random orientation of the sources distributed within the implant volume. However, in implants with the fixed orientation of the sources using an applicator (i.e., eye-plaque) or ribbons (i.e., stranded seeds) this method is not suitable. Meanwhile, the limitation of the digitization of the source geometry in an implant volume, does not allow us to utilize the full source dosimetry, as it is needed.
Interpolation and Extrapolations There is no clear recommendation for the methods of extrapolation of the dose distribution, outside of the zone for which dosimetric data are available. There are only limited dosimetric evaluations of the brachytherapy sources, beyond the range that is commonly called “clinical range,” and more work is needed to clarify our techniques of dose extrapolation.
Strategies To Implement Dosimetry Parameters in Systems that Do Not Support TG-43 Format Treatment-planning systems do not always support the latest formulations dosimetry and data. This can nearly always be remedied by combining TG-43 parameters into composite factors and/or rendering into either “along and away” tables or TG-43–like tables. Such tables can be used in treatment-planning systems or for “hand calculation” of dose to selected points. In planning systems that provide point-source only models and that use the 1995 TG-43 anisotropy constant, there are two paths to use the TG43U1 (or equivalent) anisotropy factors, φan(r). The first is to combine this factor with the point source radial dose function, gP(r), to create an anisotropy weighted function, gP(r) φan(r), that one enters in the array provided for the radial dose function (Williamson 2002). When the planning system only supports polynomial expansions of the radial dose function (or gP(r) φan(r) ), then the g(r) function must be expanded into tabular form, the product (with φan(r)) calculated, and the resulting composite function refit to a polynomial of appropriate degree for use in the planning system. In this case, the anisotropy constant entered into the planning system would be unity. The second method is to calculate an anisotropy constant, φan, from the TG43U1 (or equivalent) anisotropy factor data as the inverse-square (of distance) weighted average of φan(r). Formulation and discussion of this option are discussed in greater detail in appendix D of TG-43U1.
Nomogram In general, a nomogram is a novel method of graphic correlation between the parameters in the field of radiation therapy for a quick determination of quantities such as treatment time. Early nomograms were developed to assist in planning 125I prostate brachytherapy (Anderson 1976). Anderson has further expanded the application of this methodology in brachytherapy for determination of the number of 192Ir ribbons and seeds, used in planar implant (Anderson 1984; Anderson, Hilaris, and Wagner 1985). Figure 8 shows a sample nomogram by Lowell Anderson for LDR 192Ir sources. Such nomograms exist for several different radionuclides, particularly for 125I and 103Pd brachytherapy sources (Anderson 1976). In commonly used clinical application, the nomogram, particularly in prostate implant with 125I and 103Pd sources, is one of the most common treatment aids. With an appropriate nomogram, one can readily find the number and strength of seeds to order for patient treatment within clinically acceptable accuracy. One also may find recommendations regarding source spacing (Anderson 1976).
17–Treatment Planning Considerations of Brachytherapy Procedures
347
Figure 8. A sample nomogram by Lowell Anderson for planar implant with 192Ir ribbons with peripheral dose rate of 10 Gy d–1. (Reproduced from Endocuriether/Hypertherm Oncol, vol 1, “A nomograph for planar implant planning,” L. L. Anderson, B. S. Hilaris, and L. K. Wagner, pp. 9–15. © 1985, with permission of the American Institute of Physics.)
These nomograms consist of multiple lines, each scaled for a particular set of implant parameters, such as volume, total and individual seed activities, number of seeds, and seed spacing. These nomograms are designed for a particular application with a particular prescribed dose. For instance, some of the earlier nomograms were designed for prostate implants with 160 Gy prescribed dose, using uniform distribution of the seeds. With some simple steps and calculations, one can determine the number of sources for the low-energy as well as the high-energy brachytherapy in day-to-day clinical applications. Therefore, this calculation method at a given point for a given brachytherapy source provides recommendations based on the accumulated experience of many cases and constitutes of a sample of this experience.
References Altschuler, M. D., P. A. Findlay, and R. D. Epperson. (1983). “Rapid, accurate, three-dimensional location of multiple seeds in implant radiotherapy treatment planning.” Phys Med Biol 28:1305–1318. Altschuler, M. D., and A. Kassaee. (1997). “Automated matching of corresponding seed images of three simulator radiographs to allow 3D triangulation of implanted seeds.” Phys Med Biol 42:293–302. Amols, H. I., and I. I. Rosen. (1981). “A three-film technique for reconstruction of radioactive seed implants.” Med Phys 8:210–214. Anderson, L. L. (1976). “Spacing nomogram for interstitial implants of 125I seeds.” Med Phys 3:48–51. Anderson, L. L. “Computerized Treatment Planning Systems.” F. Bagne (ed). Rockville, MD: US DHHS, 1984. Anderson, L. L., B. S. Hilaris, and L. K. Wagner. (1985). “A nomograph for planar implant planning.” Endocuriether/Hypertherm Oncol 1:9–15. Baltas, D., P. Karaiskos, P. Papagiannis. L. Sakelliou, E. Loeffler, and N. Zamboglou. (2001). “Beta versus gamma dosimetry close to Ir-192 brachytherapy sources.” Med Phys 28 (9): 1875–1882. Bice, W. S. J., D. F. Dubois, J. J. Prete, and B. R. Prestidge. (1999). “Source localization from axial image sets by iterative relaxation of the nearest neighbor criterion.” Med Phys 26:1919–1924. Biggs, P. J., and D. M. Kelly. (1983). “Geometric reconstruction of seed implants using a three-film technique.” Med Phys 10:701–704.
348
Ali S. Meigooni and Robert E. Wallace
Blasko, J. C., K. Wallner, P. D. Grimm, and R. Ragde. (1995). “Prostate specific antigen based control following ultrasound-guided 125I implantation for stage T1/T2 prostate carcinoma.” J Urol 154:1096. Brinkmann, D., and R. W. Kline. (1998). “Automated seed localization from CT datasets of the prostate.” Med Phys 25(9):1667–1672. Butler, W. M., and G. S. Merrick. (1996). “I-125 Rapid Strand™ loading technique.” Radiat Oncol Invest 4:48–49. Fraass, B., K. Doppke, M. Hunt, G. Kutcher, G. Starkschall, R. Stern, and J. Van Dyk. (1998). “American Association of Physicists in Medicine Radiation Therapy Committee Task Group 53: Quality assurance for clinical radiotherapy treatment planning.” Med Phys 25:1773–1829. Also available as AAPM Report No. 62. Fuks, Z., S. A. Leibel, K. E. Wallner, C. B. Begg, W. R. Fair, L. L. Anderson, B. S. Hilaris, and W. F. Whitmore. (1991). “The effect of local control on metastatic dissemination in carcinoma of the prostate: long-term results in patients treated with 125I implantation.” Int J Radiat Oncol Biol Phys 21:537. Furhang, E. E., and L. L. Anderson. (1999). “Functional fitting of interstitial brachytherapy dosimetry data recommended by the AAPM Radiation Therapy Committee Task Group 43.” Med Phys 26(2):153–160. Glasser, O., E. H. Quimby, L. S. Taylor, J. L. Weatherwax, and R. H. Morgan. Physical Foundation of Radiology, 3rd ed. New York: Harper & Row, 1961. Goodsitt, M. M., P. J. Carson, S. Witt, D. L. Hykes, and J. M. Kofler. (1998). “Real-time B-mode ultrasound quality control test procedures: Report of AAPM Ultrasound Task Group No. 1.” Med Phys 25:1385–1406. Hallert, B. X-Ray Photogrammetry: Basic Geometry and Quality. New York: Elsevier Publishing Company, 1970. Holupka, E. J., P. M. Meskell, E. C. Burdette, and I. D. Kaplan. (2004). “An automatic seed finder for brachytherapy CT postplans based on the Hough transform.” Med Phys 31(9):2672–2679. Jackson, D. D. (1983). “An automatic method for localizing radioactive seeds in implant dosimetry.” Med Phys 10:370–372. Khan, F. The Physics of Radiation Therapy. Baltimore, MD: Williams and Wilkins, 2003. Krishnaswamy, V. (1972). “Dose distribution about 137Cs sources in tissue.” Radiology 105:181–184. Kutcher, G. J., L. Coia, M. Gillin, W. F. Hanson, S. Leibel, R. J. Morton, J. R. Palta, J. A. Purdy, L. R. Reinstein, G. K. Svensson, M. Weller, and L. Wingfield. (1994). “Comprehensive QA for radiation oncology: Report of the AAPM Radiation Therapy Committee Task Group No. 40.” Med Phys 21:581–615. Also available as AAPM Report No. 46. Kumar, P. P., and R. R. Good. (1986). “Vicryl carrier for I-125 seeds: Percutanious transperineal insertion.” Radiology 159:276. Li, Z., I. A. Nalcacioglu, S. Ranka, S. K. Sahni, J. R. Palta, W. Tome, and S. Kim. (2001). “An algorithm for automatic, computed-tomography-based source localization after prostate implant.” Med Phys 28(7):1410–1415. Liu, H., G. Cheng, Y. Yu, R. Brasacchio, D. Rubens, J. Strang, L. Liao, and E. Messing. (2003). “Automatic localization of implanted seeds from postimplant CT images.” Phys Med Biol 48:1191–1203. Marinello, G., and B. Pierquin. “The Paris Systems, Optimization of Dose, and Calculation of Treatment Time.” Chapter 4 of A Practical Manual of Brachytherapy. B. Pierquin and G. Marinello. Translated by Frank Wilson, Beth Erickson, and Jack Cunningham. Madison, WI: Medical Physics Publishing, 1997. Meigooni, A. S., H. Zhang, C. Perry, S. A. Dini, and R. A. Koona. (2003). “Theoretical and experimental determination of dosimetric characteristics for BrachySeed Pd-103 model Pd-1 source.” Appl Radiat Isot 58:533–541. Meigooni, A., C. B. Saw, and R. Nath. “Basic Physics of Brachytherapy” in Principles and Practice of Brachytherapy. Subir Nag (ed.). Armonk, NY: Futura Publishing Company, Inc., 1997. Meigooni, A. S., V. Rachabatthula, S. B. Awan, and R. A. Koona. (2004). “Treatment planning considerations for prostate implants with the new linear RadioCoil™ 103Pd brachytherapy source.” To be submitted to Medical Physics. Meisberger, L. L., R. J. Keller, and R. J. Shalek. (1968) “The effective attenuation in water of the gamma rays of gold 198, iridium 192, cesium 137, radium 226, and cobalt 60.” Radiology 90:953–957. Meredith, W. J. (ed.). Radium Dosage: The Manchester System. Edinburgh: Livingston, 1967. Merrick, G. S., W. M. Butler, A. T. Dorsey, and H. L. Walbert. (1998). “The influence of timing on the dosimetric analysis of transperineal ultrasound guided prostatic conformal brachytherapy.” Radiat Oncol Invest 6:182. Merrick, G. S., W. M. Butler, A. T. Dorsey, J. H. Lief, and M. L. Benson. (2000). “Seed fixity in the prostate/periprostatic region following brachytherapy.” Int J Radiat Oncol Biol Phys 46(1):215–220.
17–Treatment Planning Considerations of Brachytherapy Procedures
349
Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee, Task Group No. 43.” Med Phys 22:209–234. Also available as AAPM Report No. 51. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24(10):1557–1598. Also available as AAPM Report No. 59. NEMA 2004. “The DICOM Standard” online at http://medical.nema.org/dicom/2004.html, National Electrical Manufacturers Association, Rosslyn, VA. Nori, D. (1992). “Conformal brachytherapy of prostate cancer: An effective outpatient treatment.” Cancer Watch 1:124). Patel, N. S., S.-T. Chiu-Tsao, P. Fan, H. S. Tsao, S. F. Liprie, and L. B. Harrison. (2001). “The use of cylindrical coordinates for treatment planning parameters of an elongated 192Ir source.” Int J Radiat Oncol Biol Phys 51(4):1093–1102. Pierquin, B., and G. Marinello. A Practical Manual of Brachytherapy. Translated by Frank Wilson, Beth Erickson, and Jack Cunningham. Madison, WI: Medical Physics Publishing, 1997. Porrazzo, M. S., B. S. Hilaris, C. R. Moorthy, A. E. Tchelebi, C. A. Mastoras, L. L. Shih, L. Stabile, and N. Salvaras. (1992). “Permanent interstitial implantation using palladium-103: The New York Medical College preliminary experience.” Int J Radiat Oncol Biol Phys 23:1033. Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. (2004). “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Also available as AAPM Report No. 84. Rosenthal, M. S., and R. Nath. (1983). “An automatic seed identification technique for interstitial implants using three isocentric radiographs.” Med Phys 10(4):475–479. Rosenwald, J. C. “Calculation of Dose Distribution,” Chapter 3 of A Practical Manual of Brachytherapy. B. Pierquin and G. Marinello. Translated by Frank Wilson, Beth Erickson, and Jack Cunningham. Madison, WI: Medical Physics Publishing, 1997. Roy, J. N., K. E. Wallner, P. J. Harrington, C. C. Ling, and L. L. Anderson. (1993). “A CT-based evaluation method for permanent implants: application to prostate.” Int J Radiat Oncol Biol Phys 26(1):163–169. Russel, K. J., and J. C. Blasko. (1993). “Recent advances in interstitial brachytherapy for localized prostate cancer.” Probl Urol 7:260. Schlienger, M., J. C. Rosenwald, M. Miclutia, R. Quint, and B. Pierquin. (1970). “Controle dosimetrique en brachycuritherapie par les isodoses,’escargot’.” Acta Radiol 9:282—288. Shalek, R. J., and M. Stovall. “Dosimetry in Implant Therapy” in Radiation Dosimetry, vol III. F. H. Attix and E. Tochlin (eds.). New York: Academic Press, pp. 776–798, 1969. Sievert, R. M. (1921). “Die Intensitätsverteilung der primären g-Strahlung in der Nähe medizinischer Radiumpräparate.” Acta Radiol 1:135. Sommerkamp, H., M. Ruppercht, and M. Wannenmacher. (1988). “Seed Loss in Interstitial radiotherapy of prostatic carcinoma with I-125.” Int J Radiat Oncol Biol Phys 14:389. Su, Y., B. J. Davis, M. G. Herman, and R. A. Robb. (2004). “Prostate brachytherapy seed localization by analysis of multiple projections: Identifying and addressing the seed overlap problem.” Med Phys 31(5):1277–1287. Tapen, E. M., J. C. Blasko, P. D. Grimm, H. Radge, L. Ray, S. Cliford, J. Sylvester, and T. Griffin. (1998). “Reduction of radioactive seed embolization to lung following prostate brachytherapy.” Int J Radiat Oncol Biol Phys 42:1063. Todor, D. A., G. N. Cohen, H. I. Amols, and M. Zaider. (2002). “Operator-free, film based 3D seed reconstruction in brachytherapy.” Phys Med Biol 47:2031–2048. Tubic, D., A. Zaccarin, J. Pouliot, and L. Beaulieu. (2001b). “Automated seed detection and three-dimensional reconstruction. II. Reconstruction of permanent prostate implants using simulated annealing.” Med Phys 28:2272–2279. Tubic, D., and L. Beaulieu. (2005). “Sliding slice: A novel approach for high accuracy and automatic 3D localization of seeds from CT scans.” Med Phys 32:163–174. Tutar, I. B., R. Managuli, and V. Shamdasani. (2003). “Tomothynthesis-based localization of radioactive seeds in prostate brachytherapy.” Med Phys 30:3135–3142.
350
Ali S. Meigooni and Robert E. Wallace
Williamson, J. F. “Dosimetric characteristics of the DRAXIMAGE model LS-1 I-125 interstitial brachytherapy source design: A Monte Carlo investigation.” Med Phys 29:509–521. Williamson, J. F., R. I. Morin, and F. M. Khan. (1983). “Monte Carlo evaluation of the Sievert integral for brachytherapy dosimetry.” Phys Med Biol 28:1021. Willins, J., and K. Wallner. (1998). “Time-dependent changes in CT-based dosimetry for I-125 prostate brachytherapy.” Radiat Oncol Invest 6:157–160. Yu, Y., L. L. Anderson, Z. Li, D. E. Mellenberg, R. Nath, M. C. Schell, F. M. Waterman, and A. Wu. (1999). “Permanent prostate seed implant brachytherapy: Report of the American Association of Physicists in Medicine Task Group No. 64.” Med Phys 26:2054–2076. Also available as AAPm Report No. 68.
Chapter 18
Systems 1B Manchester Planar and Volume Implants and the Paris System Michael T. Gillin, Ph.D., and Firas Mourtada, Ph.D. The University of Texas M.D. Anderson Cancer Center Department of Radiation Physics Houston, Texas Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 351 The Manchester System for Surface Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352 The Manchester Interstitial System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354 The Manchester Planar System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354 The Manchester Volume System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360 Paris System of Single and Double Plane Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363 Optimized Stepping Source Dosimetry System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369 Actual Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371
Introduction The development of systems for interstitial brachytherapy began early in the practice of radiation oncology. It is now a challenge for the contemporary radiation oncology physicist to comprehend the practice of radiation oncology before the current digital universe. Different priorities existed in the pre-digital era. Some of these priorities addressed the need to consistently produce clinically acceptable dose patterns in the absence of both visual dose distributions and significant medical physics support. Madame Curie prepared the first brachytherapy calibration source of radium salt. The unit of milligrams of 226Ra became a measure of source strength for intercomparison of brachytherapy sources. Sievert developed a calculation approach in 1921 for filtered sources. In 1932 Quimby published the first article on the system, which now bears her name. The Quimby system is based upon a uniform distribution of sources to produce a non-uniform radiation distribution. In 1934 the first paper, which described the Manchester System, also commonly known as the Paterson and Parker system, was published. In the Manchester System, the sources are arranged in a non-uniform pattern in order to produce a “uniform” dose distribution. In 1960 Pierquin published the first paper on the Paris System, which was based on the use of 192Ir. In some ways the Paris System, with its variable source spacing, can be viewed as an evolution from the Manchester System with its fixed source spacing. These systems were developed to serve as a guide to the oncologist. If the sources were arranged according to the rules of the system, then the oncologist had a sense of the resultant dose distribution and could provide consistency in patient care. Image-based computerized dose distributions developed later. These systems still have a role in the thought process, which should precede every implant. Given a specific volume to be treated, the systems suggest the pattern of the source placement, including the separation between the sources and the source length. The initial sources used in the Manchester System were radium (226Ra) sources. These sources had both a physical length and a shorter active length, which is addressed by the Manchester System. The classic Manchester dose prescription, translated to the modern concept of dose, was 65 Gy in 6 to 8 days. This lead to full strength 226Ra needles of 0.66 mg cm–1 and half strength 226Ra needles of 0.33 mg cm–1. Eventually 137Cs needles were available as a replacement for 226Ra needles. These needles were also characterized by the physical and active length. Many institutions possessed inventories of various needles that had different lengths and source strengths. There were needles with a linear activity distribution,
351
352
Michael T. Gillin and Firas Mourtada
needles with a high activity at one end in the event that one end could not be crossed (Indian club needles), and needles with high activity on both ends (dumbbell needles). Today, less descriptive but more politically correct names would most likely be used. Modern basic physics references for interstitial brachytherapy include the chapter by Shalek and Stovall, “Brachytherapy Dosimetry,” in volume III of The Dosimetry of Ionizing Radiation (Shalek and Stovall 1990) and Physical Aspects of Brachytherapy: Medical Physics Handbook 19 by Godden (1988). In addition, there is chapter 13 in The Physics of Radiology, 4th Edition by Johns and Cunningham (1983). There are multiple clinical/physics texts, which are devoted in whole or in part to brachytherapy. Chapter 12, “Physics of Brachytherapy” by Glasgow and Perez, in Principles and Practice of Radiation Oncology, Second Edition, by Perez and Brady is an outstanding summary of brachytherapy in a clinical textbook (Perez and Brady 1992). A Practical Manual of Brachytherapy by Pierquin and Marinello has been translated into English by Wilson, Erickson, and Cunningham (Pierquin and Marinello 1997) and is the most complete English language description of the Paris System. There are some classic books from the 1930s and 1940s, which describe the Manchester System, e.g., Radium Dosage, The Manchester System, edited by W. J. Meredith (Paterson et al. 1947). Readers of the older works should remember that some numeric values found in these texts have been updated, as various parameters have been refined. The importance of units does not have to be stressed to physicists. Units are especially important in brachytherapy. It may be necessary to be able to convert from one system of units to another, especially as they pertain to the statement of source activity. The definition of terms is also important. The various systems have specific terms, which are explicitly defined by that system.
The Manchester System for Surface Applicators The Manchester System for surface applicators is based upon the assumption that the source distribution rules are followed. The result is a ±10% dose uniformity over the treatment surface which is located at a distance, h, from the plane of the implant. The original dose tables from the 1930s gave the amount of radium in milligram-hours (mg-h) to give a “dose” of 1000 R to the treatment area, which is at a distance h from the surface of the applicator. Over the years, a number of authors have recommended corrections to convert the original tables to “modern” units. The resultant correction factor, which ranges between 1.064 (Johns and Cunningham) to 1.08 (Godden) and 1.10 (Shalek), was used to multiply the existing tables to create new tables, which provide the amount of milligram-hours to deliver 10 Gy at various treatment distances as a function of the area defined in square centimeters. For example, consider a 16-cm2 area and a treatment distance of 10 mm. The table as modified by Gibb and Massey (1980) shows that 611 mg-h are required to deliver 10 Gy. These tables can be used for other isotopes by substituting mCi-h or MBq-h for mg-h as determined by the ratio of the exposure rate constants. The exposure rate constant, Γ, is defined as the exposure rate in R/hr at a point 1 meter from a 1 Ci source. The exposure rate constant for 226Ra, filtered by 0.5 mm Pt, a classic number in brachytherapy, is 0.825 R m2 hr–1 Ci–1, or 8.25 mR cm2 hr–1 mg–1. The exposure rate constants for different sources can be found in the texts, which have been previously described. Thus, if the implant is to be performed using 192Ir, the 611 mg-h of radium would be converted to: 611 mg-h (radium) × 8.25 (226Ra)/4.69 (192Ir) = 1075 mCi-h (192Ir) , where Ra–226Γδ (8.25) and Ir–192Γδ (4.69) are the exposure rate constants in units of R cm2/mCi-h.
(1)
Systems 1B Manchester Planar and Volume Implants and the Paris System
353
If the prescribed dose of 10 Gy is to be delivered over 24 hours, then the required activity is: 611 mg-h/24 h
=
25.5 mg 226Ra
or
(2) 1075 mCi-h/24 h
=
44.8 mCi 192Ir .
It is also possible to correct the tables for rapidly decaying isotopes, such as 222Ra or 198Au. Paterson and Parker first described a dosimetry system for surface applicators. They then extended their work to propose a system for interstitial applications. There are some differences between these two systems. However, the basic approach is the same in that source distribution rules are defined and tables are presented that provide the product of the activity and treatment time as a function of area and distance to deliver a specific amount of dose. The Manchester System permits the sources to be arranged in a crossing pattern, i.e., the sources can be perpendicular to each other as they define the plane of interest. The source distribution rules for surface applicators, as presented by Godden (1988), which are based upon chapter II, “Mould Treatments” in the Paterson et al. (1947) book for rectangular applicators (a × b) and a treatment distance h, are as follows: 1. The distance between the active source ends should not be greater than 2h. 2. For rectangles and/or squares of area a × b, where b > a, sources should be placed around the periphery and possibly some in lines parallel to b, the number of lines being such that the area is divided into strips of width 2h. 3. If the activity per unit length on the periphery is ρ, then for a <2h all of the activity should be placed at the periphery, for 2h < a < 4h, one added line is required whose linear density is 1/2 ρ, for a >4h two or more lines are added whose linear density is 2/3 ρ. 4. The number of mg-h for 10 Gy must be increased by a correction factor to account for any elongation factor, i.e., the ratio of b to a. This elongation correction factor ranges from 1.025 for a b/a ratio of 1.5 to 1.12 for a ratio of 4 to 1. Consider a simple example for a rectangular applicator of a 2 cm × 3 cm applicator that is designed to treat to a depth of 5 mm. Thus twice the depth, 2h, is 10 mm, so that the sources would be placed 10 mm apart. The applicator is designed such that there are needles around the entire periphery (no uncrossed ends). For 6 cm2 at a treating distance of 5 mm, Table 5.1 in Godden (1988) gives 191 mg-h per 10 Gy, while Table 13-4 of Johns, which is Table 2 in this chapter, gives 188 mg-h per 10 Gy. If the prescribed dose is 60 Gy, then the number of milligram-hours required is: 191 mg-h/10 Gy × 60 Gy × 1.025 = 1175 mg-h .
(3)
If the patient will be asked to wear this applicator for 7 hours per day for a total of 5 days, then the number of milligrams is: 1175 mg-h/(7 hours/day × 5 days) = 33.6 mg .
(4)
Assume that the applicator is constructed with four needles (2 cm active length each), which are separated by 10 mm, and two needles (3 cm active length each), which are crossing (see Figure 1). The total
354
Michael T. Gillin and Firas Mourtada
Figure 1. Representation of an applicator constructed with four needles (separated by 10 mm) and two needles crossing.
active length is 14 cm. The total length of needles on the periphery is 10 cm and 4 cm of total needle length in the center. In this example, the shorter distance is exactly equal to 4h and the applicator is defined with two central sources. Following the rules for applicators, the linear intensity can be calculated as follows: (3 + 3 + 2 + 2)ρ + (2/3)ρ(2 + 2) = 12.67 ρ ,
(5)
where ρ = 2.65 mg/cm for 226Ra or 2.65 (8.25/4.69) 4.66 mCi (192Ir)/cm. Thus assuming the use of 192Ir sources, the four peripheral sources would have a linear activity of 4.67 mCi/cm and the two central sources would have a linear activity of 3.13 mCi/cm.
The Manchester Interstitial System In the interstitial system, the “Stated Dose” is defined as the dose, which is 10% greater than the minimum dose in the plane of the calculation. The maximum dose in the volume being treated will be substantially higher than the minimum dose as the point of calculation approaches a source. The Manchester System provides guidance for both planar and volume implants.
The Manchester Planar System For single plane implants, Manchester/Paterson and Parker (1947) considered the uniformity of the dose in a plane, which is located 5 mm from the source plane and is parallel to the source plane. Thus a 10mm thick volume of tissue with the sources located in the center of this volume is treated with a single plane implant. The area of the implant is defined as the width of the implant times the distance between the crossing needles, assuming that there are crossing needles on both ends of the implant. The area to be used in the Manchester System tables should be reduced by 10% for each uncrossed end. Figure 2 presents a drawing of the areas involved for implants with crossing needles, implants with one end crossed, and implants with no crossing sources. The Manchester System permits double plane implants of up to 2.5 cm thickness. Both Godden (1988, Table 6.1) and Johns and Cunningham (1983, p. 474) present separation factors for separation distances of 1.5 cm, 2.0 cm, and 2.5 cm. Johns suggests that these factors are only approximate and provides an alternative approach by calculating the doses in planes which are 5, 10, and 15 mm from the source planes. Chapter V, “Interstitial Treatments” of Paterson et al. (1947) presents the Manchester rules for planar and volume implants. Johns and Cunningham (1983) Table 13-5 is surprisingly weak on the Manchester rules, while Shalek presents a concise statement of the rules (Shalek and Stovall 1990). The following is a modern version of these rules as presented by Godden (1988):
Systems 1B Manchester Planar and Volume Implants and the Paris System
355
1. Sources should be placed in a single plane with a certain amount of the activity distributed on the periphery and the remaining amount distributed evenly in the center. The relative distribution depends upon the area, as shown below. 2. The needles should be in parallel rows at a spacing not greater than 1.0 cm. 3. Crossing needles are permitted and they should be placed across the active ends if possible, but not more than 1 cm from the active ends. 4. If the ends of the implant cannot be across, 10% of the area is deducted from each uncrossed end. 5. For implants, which use seeds or short sources, the distance between the active ends should not exceed 1 cm. 6. If two planes are used, the separate planes should be arranged as for a single implant, parallel to each other. If they differ in area, then the area used to determine the amount of activity is the average area and the activity is divided between the planes in a ratio of the areas of the planes. In addition, there are rules for the distribution of activity, which have been memorized by oncologists and physicists for decades (see Table 1). While the exact ratios may not be obvious, the fact that more activity is placed in the central part of the implant as the area being implanted increases seems intuitive. Consider a 2 cm × 3 cm interstitial implant (as opposed to the applicator, which was described above; see Figure 3). Assuming two crossed ends, one would use the same number of mg-h per 10 Gy, i.e., 191. The reason for this is that for a single plane implant, the prescribed dose is in a plane 5 mm from the plane of the implant. The implant is scheduled to be in place 7 days or 168 hours to deliver a total dose of 60 Gy. The number of milligrams required is:
Figure 2. Areas involved for implants with crossing needles, implants with one end crossed, and implants with no crossing sources. (Reprinted from Biomedical Uses of Radiation, W. Hendee (ed), Fig. 5.39. © 1999, with permission of Wiley-VCH, Weinheim.)
356
Michael T. Gillin and Firas Mourtada Table 1. Manchester Distribution of Activity for Planar Implants Area
Relative Proportion of Activity
Periphery
Center
< 25 cm2
2/3
1/3
> 25 cm2 and < 100 cm2
1/2
1/2
> 100 cm
1/3
2/3
2
161 mg-h/10 Gy × 60 Gy/168 h = 5.75 mg 226Ra or 10.1 mCi 192Ir .
(6)
The distribution rules call for two-thirds of the activity to be placed on the periphery or 6.8 mCi (192Ir) on the periphery and one-third of the activity in the center or 3.3 mCi (192Ir). The linear activity for the peripheral sources is 0.68 mCi/cm and for the central sources is 0.82 mCi/cm. Consider a classic single plane implant using 137Cs needles (physical length 42 mm and active length 30 mm.) As shown in Figure 4, this implant has one uncrossed plane. The inventory of the 137Cs needles contains three 2.00 mg Ra equivalent needles and two 1.00 mg Ra equivalent needles, which are available for use with this implant. The area of the implant is: 3 cm × 3 cm × 0.9 = 8.1 cm2.
(7)
For this area, approximately 221 mg-h/10 Gy is required (see Table 2). Two-thirds of the activity should be placed on the periphery and thus the three 2.00 mg Ra eq needles are used on the periphery. One third of the activity should be centrally located and thus the two 1.00 mg Ra equivalent needles are used in the center of this implant. The dose rate is: (8) 8 mg ⋅ 1000 cGy 221 mg-h
= 36.2 cGy/h.
Figure 3. 2 cm × 3 cm interstitial implant representation.
Systems 1B Manchester Planar and Volume Implants and the Paris System
357
Table 2. Surface Applicators and Planar Implants The table gives RA, the number of milligram hours required to deliver 10 Gy to muscle tissue for different areas and treatment distances. Filtration 0.5 mm Pt. The table may be used for planar implants by using a treatment distance of 0.5 cm. Treatment Distance (cm) 2
Area cm 0 1 2 3 4 5 6 7 8 9 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 60 70 80
.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
32 72 103 128 150 170 188 204 219 235 250 278 306 335 364 392 418 444 470 496 521 546 571 594 618 642 664 685 708 729 750 851 947 1044
127 182 227 263 296 326 354 382 409 434 461 511 557 602 644 682 717 752 784 816 846 876 909 935 967 994 1024 1053 1080 1110 1141 1283 1426 1567
285 343 399 448 492 531 570 603 637 667 697 755 813 866 918 968 1021 1072 1122 1170 1215 1261 1305 1349 1392 1432 1472 1511 1550 1585 1619 1790 1944 2092
506 571 632 689 743 787 832 870 910 946 982 1053 1120 1184 1245 1303 1362 1420 1477 1530 1582 1635 1688 1743 1793 1843 1894 1942 1990 2037 2083 2319 2532 2726
792 856 920 978 1032 1083 1134 1182 1229 1272 1314 1396 1475 1553 1622 1690 1755 1821 1881 1943 2000 2060 2119 2179 2234 2290 2344 2399 2452 2504 2556 2815 3059 3301
1139 1204 1274 1331 1388 1436 1495 1547 1596 1645 1692 1780 1865 1947 2027 2106 2180 2252 2328 2398 2468 2532 2598 2662 2726 2787 2848 2908 2966 3025 3082 3362 3628 3891
1551 1625 1697 1760 1823 1881 1938 1993 2047 2099 2149 2247 2341 2429 2514 2601 2683 2764 2841 2917 2997 3073 3145 3215 3285 3351 3421 3484 3548 3612 3676 3974 4257 4532
2026 2100 2172 2241 2307 2369 2432 2490 2548 2605 2660 2769 2870 2968 3063 3155 3242 3326 3405 3484 3562 3639 3713 3787 3859 3931 4003 4071 4139 4207 4275 4605 4913 5213
2566 2636 2708 2772 2835 2896 2956 3011 3067 3123 3178 3284 3389 3490 3585 3682 3777 3872 3962 4047 4131 4220 4306 4389 4466 4546 4626 4706 4781 4857 4929 5288 5632 5958
3166 3295 3349 3383 3450 3513 3575 3634 3694 3752 3809 3917 4027 4131 4240 4341 4441 4540 4634 4730 4824 4915 5000 5089 5174 5258 5341 5422 5505 5586 5668 6054 6419 6756
Filtration (mm Pt) Correction to mg hrs
0.3 –4%
0.5 0
0.6 +2%
0.8 +6%
1.0 +10%
This table was prepared from the original by Meredith (M12) by multiplying his values by C=1.064. (Reprinted from Johns, H.E., and J. R. Cunningham, The Physics of Radiology, 4th Edition, ©1983. Courtesy of Charles C Thomas Publisher, Ltd., Springfield, IL.)
1.5 +20%
358
Michael T. Gillin and Firas Mourtada
If the total dose to be delivered is 7000 cGy, then the total time of the implant is: 7000 cGy/(36.2 cGy/h) = 193.4 hours.
(9)
Plan a simple two-plane implant with a separation of 2.0 cm between the planes (see Figure 5). The separation factor is 1.4. The planned dose is 70 Gy to be delivered in 168 hours. Both planes have the same area, i.e., 4 cm × 3 cm. There are no crossing sources. The area to be used in the Manchester tables is: 4 cm × 3 cm × 0.8 = 9.6 cm2.
(10)
Interpolating the Manchester tables for this area yields 244 mg-h/10 Gy. This factor now needs to be adjusted by the separation factor. 244 mg-h ⋅ 1.4 10 Gy
=
342 mg-h
.
10 Gy
(11)
Note that the number of milligram hours has been increased as a result of the separation factor. 244 mg-h ⋅ 1.4 10 Gy
=
342 mg-h
.
10 Gy
Figure 4. Single plane implant with one crossing source.
(12)
Systems 1B Manchester Planar and Volume Implants and the Paris System
359
Each plane, which contains five sources, should contain a total 7.15 mg. The two peripheral sources contain two-thirds of this activity, i.e., 2.8 mg per source. The three central sources should contain one third of this activity, i.e., 0.79 mg per source. The next step in this planning exercise is to consult the inventory of 137Cs needles that are available to be used. Assume that the inventory contains 3-cm active length needles with activities of 1.0 mg Ra eq and 3.0 mg Ra eq. Thus, due to limitations in the inventory of sources, the total activity per plane is 9 mg. This still results in a source pattern, which obeys the Manchester rules, in that two-thirds of the activity is still on the periphery. If 192Ir was used for this implant, as opposed to 137Cs or 226Ra, approximately 25 mCi would be required using a conversion factor based on the ratio of the gamma rate constants (or the exposure rate constants, as has been previously demonstrated.) It is now possible to calculate the dose rate in various planes, using the Manchester surface applicators and planar implant table. 5 mm from a source plane:
10 mm from a source plane:
15 mm from a source plane:
10 Gy ⋅ 9 mg 244 mg h 10 Gy ⋅ 9 mg 450 mg-h 10 Gy ⋅ 9 mg 685 mg-h
=36.9 cGy/h
=20.0 cGy/h
(13)
=13.1cGy/h
The total dose rate in a plane, which is 10 mm from each source plane, is 40 cGy/h. The total dose rate in a plane, which is 5 mm from one source plane and 15 mm from the other source plane, is 50 cGy/h. The time required to deliver 70 Gy is 175 hours (7000 cGy/40 cGy/h).
Figure 5. Double plane implant.
360
Michael T. Gillin and Firas Mourtada
The Manchester Volume System In the Manchester Volume System, the “Stated Dose” is taken to be the dose, which is 10% higher than the absolute minimum dose in the effective volume. Consider a line radiating from a spherical or cylindrical volume, which passes between the sources. If the Manchester source distribution rules are followed, then the minimum dose at the surface of the volume is 10% below the stated dose. The maximum dose on the line increases as the point of interest moves toward the center of the volume and may be as much as 20% above the stated dose. Volumes are considered to be composed of a rind, which covers the entire surface, like a skin, and a central core. For a sphere, the rind is the surface of the sphere, while for a cylinder, the rind is a belt, the curved surface of the cylinder, plus the two flat ends. The total activity is divided into a number of equal parts and distributed according to the rules, which are presented below. This is summarized in Johns and Cunningham (1983, Table 13-8) and in Godden (1988, Table 6.5). See Table 3. Chapter V of Paterson et al. (1947) also contains the rules for volume implants. The following rules for volume implants are taken from Godden (1988): 1. The sources on each face must be spaced as evenly as possible. 2. The sources must be distributed throughout the core. 3. The sources on each surface must be spaced evenly and not more than 1.0 to 1.5 cm apart. 4. For a cylindrical implant, there must not be less than eight sources in the belt and four in the core. 5. For cuboid implants arranged as a multiplane implant, the distribution of sources is as for planar implants. 6. The geometric volume of the implant must be reduced by 7.5% for each uncrossed end. The geometric volume is defined as the cross-sectional area multiplied by the active length, which is the active length of the belt. There are separate tables, which provide the number of mg-h for 10 Gy for volume implants. Modern versions of these tables can be found in Johns and Cunningham (1983, Table 13-8) or in Godden (1988, Table 6.3). See Table 3. Consider planning a volume implant, using the Manchester approach, for a floor of the mouth lesion. For planning purposes, this lesion can be viewed as a square of 2.7 cm × 2.7 cm with a 3.0 cm height (see Figure 6). One approach would be to place needles in a 1 cm matrix with an active length of 3.0 cm. This would result in 12 needles in the belt and 4 needles in the core. The volume of the implant is 21.9 cm3. Based upon an interpolation of Table 13-8 (see Table 3), 284 mg-h/10 Gy would be required. In a 1967 monograph entitled Implant Dosimetry by Shalek and Stovall (1967) describes an implant resembling the above (see Figure 7). In this actual clinical example, 15 needles were used in total. Twelve of the needles had an active length of 3.5 cm and were of Indian club design with a total activity of 1.5 mg each. Two needles had an active length of 3.0 cm with a total activity of 1.0 mg each. One needle had an active length of 3.5 cm with a total activity of 1.16 mg. There were crossing needles on one end and the Indian club needles were implanted such that the higher activity portion of the needle was placed deeper into the patient, which essentially represented another crossed end. There were eight needles in the belt and four needles in the core. The total activity implanted was 20.74 mg. The dose rate for this implant was: 20.74 mg/(284 mg-h/10 Gy) = 73 cGy/h .
(14)
Systems 1B Manchester Planar and Volume Implants and the Paris System Table 3. Volume Implants Volume cm3 cm3
Rv mg hr
5 10 15 20 30 40 50 60 80 100 140 180 220 260 300 340 380
106 168 220 267 350 425 493 556 673 782 979 1156 1322 1479 1627 1768 1902
Distribution Rules Volume should be considered as a surface with 75% activity and core with 125% Rules for cylinders Belt—50% activity with minimum 8 needles Ends—12.5% of activity on each end Core—25% with minimum of 4 needles For each uncrossed end, reduce volume by 7.5% Length Diameter
1.5% 2.0
Increase mg hr 3%
6%
2.5
3.0
10%
15%
This table was prepared from the original by Meredith (M12) by multiplying his values by C=1.064. (Reprinted from Johns, H.E., and J. R. Cunningham, The Physics of Radiology, 4th Edition, ©1983. Courtesy of Charles C Thomas Publisher, Ltd., Springfield, IL.)
Figure 6. Theoretical volume implant.
361
362
Michael T. Gillin and Firas Mourtada
An example of a permanent implant is presented in chapter V, “Interstitial Treatments” in Paterson et al. (1947). This old Manchester example used radon (222Ra) seeds that were encapsulated in gold. The following example uses 198Au seeds, half-life of 2.698 days, for the same implant, which was a 4.0 cm diameter sphere, volume 33.5 cm3, designed to deliver 65 Gy. Interpolating the volume Manchester table, Johns (Johns and Cunningham 1983, Table 13-8) yields 376 mg-h/10 Gy. Thus the total number of mg-h needed to deliver 65 Gy is 2444 mg-h. It is now necessary to convert this to the appropriate 198Au activity. Using the classic exposure rate constants (226Ra 0.825 R m2 h–1 Ci–1 and 198Au 0.238 R m2 h–1 Ci–1), it can be determined that: 1 mCi (198Au) = 0.288 mg (226Ra) 1 mCi decayed (198Au) = 1.44 × 2.698 days × 24 hours/day × 0.288 mg
(15)
= 26.9 mg-h (226Ra) . To determine the required activity of 198Au: 2444 mg-h . = 90.9 mCi . mg-h 26.9 mCi
Figure 7. Implant using 15 needles (cf., Shalek and Stovall 1967).
(16)
Systems 1B Manchester Planar and Volume Implants and the Paris System
363
Following the classic Manchester rules, the rind would contain six parts of the activity (68 mCi) and the core 2 parts (23 mCi) (see Table 4). For planning purposes, the implant would consist of 12 seeds, 5.7 mCi per seed, on the surface of the 4 cm diameter sphere and 6 seeds, 3.8 mCi each, evenly distributed throughout the volume of the sphere.
Paris System of Single and Double Plane Implants The French, led by Pierquin and Chassagne, began proposing a different system of interstitial implants in the early 1960s. This system used 192Ir in the form of wire, straight pins, or double pins. In the 1970s in the United States, it was possible to purchase either wire or pins from Europe as well as 192Ir seeds from various American companies. Today, U.T. M.D. Anderson (MDACC) continues to use 192Ir wire by purchasing the wire from the United Kingdom and activating the wire in a reactor, which is located in Texas. Americans who made major early contributions to the use of 192Ir included both Henschke and Delclos. Customized 192Ir sources, which could be obtained in one or two days, meant that institutions were no longer dependent upon maintaining a large inventory of mock 226Ra sources, such as 137Cs. (In the 1970s, the federal government had a program that encouraged hospitals to dispose of their 226Ra sources.) The Paris System is defined for both single- and double plane implants. The maximum thickness allowed for a single plane implant is 12 mm and the maximum thickness for a double plane implant is 25 mm. The Paris System of dosimetry is based upon determining the minimum dose in the region between the sources as arranged in allowed geometries and then defining the reference dose as a percentage of this minimum dose. The net result is to define a prescription dose that encompasses the entire treatment volume, assuming that the rules of the Paris System have been followed. The allowed geometries are a single plane with equally spaced sources and a double plane with sources either in a triangular or a square pattern. The active length of the sources is larger than the treated length by a factor of 1.4. The spacing between the sources is varied, depending upon the thickness of the target to be treated. Basic concepts in the Paris System include: 1. The basal dose or basal dose rate, which is the minimum dose rate in the central plane between two or more sources. (The central plane is generally the plane, which bisects the sources.) 2. The reference dose or reference dose rate, which is taken as 85% of the average of the basal doses. (The 85% criterion is based upon the clinical experience of the team of oncologists and physicists who developed this system.) 3. The treatment volume, which is the volume enclosed by the reference dose.
Table 4. Distribution of Activity for Volume Implants Volume Shape
Distribution of Activity
Sphere Cylinder Crossed at active ends 1 end uncrossed 2 ends uncrossed Crossed at needle tips 1 end uncrossed
Rind 6 Parts Belt 4 4 4 4 4
Core 2 Parts Core 2 2 2 2 2
End A 1 1 0 2 2
End B 1 0 0 2 0
364
Michael T. Gillin and Firas Mourtada 4. A central plane, which is a plane perpendicular to the long axis of the sources and which bisects the sources. 5. A hyperdose volume, which is the volume surrounded by the 170% isodose surface. Note that it is twice the reference dose value. 6. The use of the same linear activity sources for the implant that is the same linear reference kerma rate or the same mCi/cm for the entire implant. Marinello (Marinello et al. 1985) has defined the Paris System criteria for the use of seeds, namely that the spacing between the seeds should be less than or equal to 1.5 times the active length of the seed. Thus for seeds which are 3 mm in length, the spacing between seeds, if the Paris System is being used, should be less than or equal to 4.5 mm.
Chapter 4 in Pierquin and Marinello (1997) contains a comprehensive description of the Paris System. There are multiple figures in this chapter that present the basic geometric concepts of the Paris System. Table 5 of this chapter presents the relationships between source length and source spacing for single and double plane implants. There is also a summary of these rules in an article by Gillin et al. (1984). An initial example using the Paris System is a simple single plane implant. Consider a target volume of the following dimensions, length 2.0 cm, width 1.2 cm, and thickness 0.5 cm (see Figure 8). The ratio of the treated thickness to the spacing for two sources is 0.5. Thus the separation between the sources is 1.0 cm. The ratio of the treated length to the active length is 0.7. Thus the source length is approximately 2.8 cm. The basal dose is calculated at one point, midway between the sources in the central plane. If the
Figure 8. Example of Paris system single plane implant.
Systems 1B Manchester Planar and Volume Implants and the Paris System
365
Table 5. Predictive Relationships in the Paris System for Rectilinear Sources Equal in Length Patterns
2 Lines N lines in 1 plane N lines in squares N lines in triangles
Ratio of Treated Length to Active Length
Ratio of Treated Thickness to Spacing
Ratio of Lateral Margin to Spacing
Ratio of Safety Margin to Spacing
0.7 0.7 0.7 0.7
0.5 0.6 1.55–1.60 1.3
0.37 0.33 — —
— — 0.27 0.20
(Reprinted from Marinello, G. and B. Pierquin, A Practical Manual of Brachytherapy, Table 4.1. © (1997) with permission of Medical Physics Publishing, Madison, WI.)
prescription dose is 60 Gy, delivered at 0.6 Gy/h, the required linear activity can be determined. Appendix 2 in Pierquin and Marinello (1997) provides absorbed dose rates in water for 192Ir wire of two different diameters, 0.3 mm and 0.5 mm for a 1 U/cm (1 µGy h–1 m2 cm–1) source as a function of wire length and distance. The calculated dose rate for a 1 U/cm source for this two-source implant is 11.4 cGy/h. To obtain the 60 cGy/h dose rate, a source of 5.24 U/cm is required. AAPM Task Group 43 (TG 43) report (Nath et al. 1995), Table IV provides a conversion factor of 4.2 U/mCi (apparent activity), so that in older units the linear apparent activity is 1.25 mCi/cm. However, in the United States 192Ir seeds would most likely be used as opposed to wire. Following the Paris rules, these seeds would be spaced at 5 mm center-tocenter spacing and each seed would contain 0.62 mCi. Historically, American vendors wanted the 192Ir activity expressed in terms of milligram radium equivalents (mg RA eq), so that each seed would contain approximately 0.35 mg Ra eq. Consider the Paris System approach to the 2.5 cm × 3.5 cm × 2.5 cm thick target volume, described in the Manchester System double plane implant by Godden. The solution offered by Godden involves the use of five needles per plane, spaced 1.0 cm apart. The separation between the source planes is 2.0 cm. There are no crossing needles (see Figure 9). The peripheral needles contain twice the activity of the central needles. A total of 18.5 mg Ra eq is used, which converts to 32.5 mCi of 192Ir. The Paris System would propose a double plane implant with a square pattern (see Figure 10). Using the relationships presented in Table 5, the Paris System would do the following: Active length = Treated Length/0.7 = 3.5/0.7 = 5.0 cm. Source Spacing = Treated Thickness/1.6 = 2.5/1.6 = 1.6 cm.
(17)
Safety margin = 0.27 to source spacing = 0.27 × 1.6 = 0.4 cm. Thus the Paris System requires eight sources, four in each plane, with a separation of 1.6 cm between the sources and 1.6 cm between the planes. The basal dose would be calculated in the center of the three squares that are formed by this implant, and averaged. The reference dose would be 0.85 times the basal dose. The 5.0 cm long sources will extend in each direction approximately 7.5 mm beyond the 3.5 cm long target. The outer two pairs of sources would be outside the target volume, while the inner two pairs would be centered in the target volume. What is the linear activity needed if the Paris System is to deliver the same dose, 60 Gy, over the same time period, 168 hours, using 192Ir wires? One could perform an unfiltered line source calculation, assuming a dose rate of 1 mCi/cm and then scale the result. Given the fixed geometries and the calculation in the central plane, the manual calculations for the three basal dose rates are quite easy.
366
Michael T. Gillin and Firas Mourtada
Figure 9. Example of Manchester System double plane implant.
Figure 10. Example of Paris system double plane implant.
Systems 1B Manchester Planar and Volume Implants and the Paris System
DR =
Γ A −1 l ⋅ ⋅ 2 ⋅ tan , 2h h l
367
(18)
where Γ is the dose rate constant, h is the perpendicular distance to the center of the square, A is the activity, and l is the active length, 5 cm. In this example, the perpendicular distances, h, are approximately 1.13 cm, 2.6 cm, and 4.2 cm. A linear activity of approximately 0.67 mCi/cm would be required. Thus, the total activity would be approximately 26.7 mCi, which is close to the 32.5 mCi if the Manchester approach were used. It is also possible to perform manual calculations using the TG-43 approach. " θ ) = D" (r ,θ ) [G(r, θ) / G(r , θ ) ] g(r)F(r,θ ) . D(r, 0 0 0 0 0
(19)
TG 43 (Nath et al. 1995) defines the geometry factor, G(r,Θ), which accounts for the spatial distribution of activity within the source, ignoring photon absorption and scattering in the source structure. For a line source, TG 43 presents the geometry factor as: G(r,θ ) =
β Lr sin(θ )
.
(20)
where L is the active length of the source, β is the angle subtended by the active source with respect to the point (r, Θ), and r is the distance from the point of interest to the center of the source. In chapter 9 of Pierquin and Marinello (1997), Calitchi and Marinello describe a single plane implant that was used to treat a recurrence in a chest wall following mastectomy. The thickness of this implant was 1.0 cm, while the length was 3.0 cm and the width 6.0 cm (see Figure 11). The spacing between the sources, as calculated by the Paris System rules for this 1.0 cm thick implant, is 1.0/0.6 or 1.7 cm. Four sources with an active length of 4.3 cm (3/0.7) would generously cover this implant. This example displays the flexibility of the Paris System with its variable source spacing. The following calculation uses a TG43 approach to calculate the dose rate in the central plane between sources 2 and 3. The determination of the dose rate constant in water, Λ, for a 4.3-cm long 192Ir source is a challenge. The dose rate constant presented in Nath et al. (1997, Table VI) for 192Ir, 1.12 cGy h–1 U–1 , is for a seed. Dose rate constants are a function of source composition and source length. Karaiskos et al. (2001) provide Monte Carlo and Sievert calculated dose rate constants for three different diameter 192Ir sources as a function of source length. Figure 3 of Karaiskos et al. (2001) can be interpolated to yield a dose rate constant of 0.6 cGy h–1 U–1 for a 4.3 cm long 192Ir source with an outer diameter of 0.3 mm. Ballester et al. (1997) earlier calculated Monte Carlo dose rate distributions around 192Ir wires. Table IV of that work can be interpolated to yield a dose rate constant for this 4.3 cm source of 0.61 cGy h–1 U–1.
368
Michael T. Gillin and Firas Mourtada
Figure 11. Paris System single plane implant with variable source spacing.
D(r,Θ) = SkΛ[G(r,Θ)/G(r0,Θ0]g( r)F(r,Θ) ,
(21)
where Sk equals 1 U, Λ equals approximately 0.60 cGy h–1 U–1, g(r ) is essentially 1.0 over the distances being considered, F(r,Θ) is 1.0 for calculations in the central plane, G(r,Θ) is equal to β/Lr sin θ, r0 equals 1.0 cm, Θ0 equals Π/2, r is the distance from the sources to the point of calculation, which is this case is 0.85 cm and 2.55 cm, G(0.85, Π/2) = 0.651 cm–2, G(1.00, Π/2) = 0.528 cm–2, G(2.55, Π/2) = 0.128 cm–2, D = 1 U × 0.6 cGy h–1 U–1 {2 × 0.651/0.528 + 2 × 0.128/0.528}, and D = 1.77 cGy/h for 1 U. Using a conversion factor of 4.205 mCi–1, the total activity of each source is 0.238 mCi. Assume that the desired activity is 1.0 mCi/cm. Then the TG-43 calculated dose rate is D = 1.77 × 4.3/.238 = 32 cGy/h . This calculated value is reassuringly close to the unfiltered line source result of 30.8 cGy/hr.
(22)
Systems 1B Manchester Planar and Volume Implants and the Paris System
369
What limits the separation between the sources? Why not use three sources, which are placed 3.0 cm apart, to cover the 6.0 cm width? The answer is that such a distribution would result in large hyperdose sleeves, which most likely would lead to necrosis. The Paris System recommendation, based upon clinical experience, is to limit the diameter of the hyperdose sleeve to 8 to 10 mm or less. In chapter 8 of Pierquin and Marinello (1997), Pernot and Marinello describe two mobile tongue implants. The larger implant has target volume dimensions of 4 cm in width by 3.5 cm in length by 2.5 cm in thickness. Their solution is to use three 192Ir loops in a pattern of squares. The suggested spacing is 1.6 cm. Each wire is approximately 9 cm in length. This would result in an active length of each leg being approximately 3.7 cm in length and a crossing piece of 1.6 cm. If the three loops are centered in the target volume, then there is a safety margin of 0.43 cm (0.27 × 1.6 cm) on each end of the implant. The total width of the implant, according to the resultant Paris System distribution, is 3.2 cm + 2 × 0.43 cm, or 4.06 cm. The target dimensions of the smaller implant are 2.5 cm width, 3.0 cm in length, and 1.5 cm in thickness. Such an implant is ideal for the use of 192Ir pins, which have a 1.2 cm separation between their legs. A pattern in 12-mm squares produces an appropriate solution. The only minor challenge is determining the length of the legs of the pin. In the Paris System, the sources should be 10+% longer than the target volume on each side. Thus, in order to cover the 3.0 cm length, the sources should extend approximately 3 mm longer on the side that does not have the crossing piece of the pin.
Optimized Stepping Source Dosimetry System An optimized stepping source dosimetry system has been described by van der Laarse and Prins from Nucletron Research (van der Laarse 1994). This stepping source system has evolved from the Paris System and includes computer-based planning. The significant features of this system include: 1. Equidistant dwell positions have optimized dwell times. 2. Dose specification points for the purposes of optimization are defined midway between the catheters along the active length of the catheter. 3. Active dwell positions remain inside the target volume. Major et al. (2002) compared the Paris System and the dose point system in terms of conformality and homogeneity of dose distributions. This paper addressed the active source length question, i.e., should active dwell positions remain within the target volume, as suggested in the optimized stepping source system, or should they extend past the target volume by some amount. They conclude that the optimal active lengths depend upon the catheter separation, the method of optimization, and the geometrical shape of the target volume. The distance between the outer source position in the catheter and the surface of the Planning Target Volume (PTV) is between 2.5 and 5 mm, with the source positions being in the PTV. They also conclude that in optimized systems an isodose value, which is higher than the 85% of the mean central dose, should be used to define the reference dose. As shown in Figure 12, a simple comparison, which involves two catheters separated by 1.0 cm, between the Paris System and the optimized stepping source dosimetry system has been performed. In the Paris System, the equal linear activity sources extended past the target volume and in the optimized stepping source system, the sources (dwell positions) were confined to the target volume. Dose points were defined halfway between the catheters and opposite each dwell position. The optimization was performed on all dose points on volume, which is equivalent to optimizing on dose points at a distance. The dwell times in the optimized system varied by approximately a factor of three, with the shortest times in the center and the longest times being at the first and last dwell positions.
370
Michael T. Gillin and Firas Mourtada
Figure 12. Comparison of optimized stepping source system (above) with the Paris System (below).
The introduction of remote afterloading devices, both high dose rate and pulsed dose rate, raises many clinical questions, especially with respect to time-dose relationships. A comprehensive article on head and neck interstitial implants by Mazeron et al. (2002) discusses some of these questions as part of an overall review.
Actual Practice Several recent interstitial implants from U.T. MDACC have been reviewed. These implants were performed by different oncologists. The first implant was a single plane implant for recurrent malignant fibrohisteocytoma of the right upper extremity. The prescription was 45 Gy at 60 cGy/h. Fifteen catheters were implanted with spacing that ranged between 5 mm and 15 mm. (The last 2 catheters on one side had the 15 mm separation, while the separation between the remaining 13 was 5 to 10 mm.) The active length varied between 4.5 cm to 5.5 cm. 192Ir-wire was used with a linear activity of approximately 1 mCi/cm. The total activity was 68.8 mCi. No particular system was used. However, the implant closely resembles a Paris System implant. A manual, unfiltered line source calculation produces a basal dose of approximately 77 cGy/h, while the computer calculation yields 82 cGy/h at the same location. The difference between the computer calculation and the manual calculation reflects distance approximations made in the manual calculation. The Paris System suggests a reference dose rate of 65 cGy/h, while in fact 60 cGy/h was used. Given the larger separation between the last two catheters, the reference dose surface is not contiguous. The reference dose volume breaks up around catheter 14 and 15. The hyperdose regions do not exceed the Paris guidelines. In the opinion of the oncologist, this is an acceptable implant. The second implant was essentially a two-plane implant plus an additional source for the nasal septum. There were two sources in one plane, three sources in the second plane, which was separated by 2 cm, and one source in the middle essentially at 1 cm from each plane. The source were 2.2 cm long Ir-wire with a linear activity of 1.01 mCi/cm. A prescription dose rate of 50 cGy/h was used to a total dose of 67 Gy. This implant did not conform to any system, but in the judgment of the oncologist covered the target
Systems 1B Manchester Planar and Volume Implants and the Paris System
371
volume. In the central plane, the prescription dose rate covered the target volume. This is also true at a distance of 5 mm in either direction from the central plane. However, at 10 mm in either direction, the prescription isodose no longer defined a contiguous volume. The hyperdose sleeves were within the Paris System criteria. The patient tolerated well and displayed minimum tissue changes one year later. The final implant was a permanent gold seed implant for unresectable rectal carcinoma. One hundred fifty-three seeds, at 1.81 mCi/seed, were implanted. CT-based dose distributions were calculated. The prescription isodose encompassed an approximately 6-cm-long volume, changing shape on each axial slice reflecting changes in the seed distribution. This implant did not follow a specific system. However, the oncologist was satisfied with this palliative procedure.
Summary The classical systems may not be widely used in current clinical practice. They do provide a reference for comparison of contemporary work. They also reflect significant clinical experience and suggest limits to fundamental parameters, such as separation between the sources. Interstitial brachytherapy still offers a unique therapeutic approach in selected circumstances. Knowledge of the classical systems is useful both in the planning of contemporary implants and in the evaluation of such implants. Interstitial brachytherapy has been an important element in the history of radiation oncology. It is important for contemporary medical physicists to know this history. As Robert Heinlein noted in The Notebooks of Lazarus Long, “A generation which ignores history has no past and no future.”
References Ballester, F., C. Hernandez, J. Perez-Calatayud, and F. Lliso. (1997). “Monte Carlo calculation of dose rate distributions around 192Ir wires.” Med Phys 24:1221–1228. Gibb, R., and J. B. Massey. (1980). “Radium dosage: SI units and the Manchester system.” Br J Radiol 53:1100–1101. Gillin, M. T., R. W. Kline, J. F. Wilson, and J. D. Cox. (1984). “Single and double plane implants: A comparison of the Manchester System with the Paris System.” Int J Radiat Oncol Biol Phys 10:921–925. Godden, T. J. The Physical Aspects of Brachytherapy. Medical Physics Handbook 19. Bristol: Adam Hilger, 1988. Johns, H. E., and J. R. Cunningham. The Physics of Radiology, 4th Edition. Springfield, IL: Charles C Thomas Publishers, 1983. Karaiskos, P., P. Papagiannis, A. Angelopoulos, L. Sakelliou, D. Baltas, P. Sandilos, and L. Vlachos. (2001). “Dosimetry of 192Ir wires for LDR interstitial brachytherapy following the AAPM TG-43 dosimetric formalism.” Med Phys 28:156–166. Major, T., C. Polgár, J. Fodor, A. Somogyi, and G. Németh. (2002). “Conformality and homogeneity of dose distributions in interstitial implants at idealized target volumes: A comparison between the Paris and dose-point optimized systems.” Radiother Oncol 62(1):103–111. Marinello, G., M. Valéro, S. Leung, and B. Pierquin. (1985). “Comparative dosimetry between iridium wires and seed ribbons.” Int J Radiat Oncol Biol Phys 11:1733–1739. Mazeron, J. J., G. Noël, and J. M. Simon. (2002). “Head and Neck Brachytherapy.” Semin Radiat Oncol 12(1):95–108. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43.” Med Phys 22(2):209–234. Also available as AAPM Report No. 51. Paterson, R., H. M. Parker, F. W. Spiers, M. C. Tod, S. K. Stephenson, and W. J. Meredith. Radium Dosage, The Manchester System. W. J. Meredith (ed). Baltimore, MD: The Williams and Wilkins Company, 1947. Perez, C. A., and L.W. Brady. (eds.). Principles and Practices of Radiation Oncology, Second Edition. Philadelphia: J. B. Lippincott Company, 1992.
372
Michael T. Gillin and Firas Mourtada
Pierquin, B., and G. Marinello. A Practical Manual of Brachytherapy. Madison, WI: Medical Physics Publishing, 1997. Shalek, R. J., and M. Stovall. Brachytherapy Dosimetry, The Dosimetry of Ionizing Radiation. Vol. III. K. R. Kase, B. E. Bjarngard, and F. H. Attix (eds). San Diego, CA: Academic Press, Inc., 1990. Shalek, R. J., and M. Stovall. Implant Dosimetry. The University of Texas M.D. Anderson Hospital and Tumor Institute at Houston, 1967. van der Laarse, R. (1994). “The stepping source dosimetry system as an extension of the Paris System.” Brachytherapy from Radium to Optimization 34:319–330.
Chapter 19
Quimby-Based Brachytherapy Systems Robert D. Zwicker, Ph.D. Department of Radiation Medicine University of Kentucky, Lexington, Kentucky Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373 The Quimby System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374 Planar Source Distribution (226Ra) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374 Dose Homogeneity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374 Planar Exposure Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374 Volume Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 The Kwan System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 Source Distribution (192Ir) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 Prescription Point Locations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376 Dose Tables/Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377 The Memorial System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379 Physical Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 379 Prescription Point Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380 192 Ir Nomograph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380 The Tufts/MCV System (Zwicker System) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381 Source Distribution/Dose Uniformity in Seed Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381 Prescription Surface Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381 Interplanar Spacing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382 Dose Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382 The Saw Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383 Optimization Indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383 Optimal Reference Dose Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383 Generalized Optimization Indices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
Introduction Systems of source distribution and dose determination for surface and interstitial brachytherapy were introduced decades ago by Quimby (1932) and by Paterson and Parker (1949), and these systems still influence the planning of brachytherapy treatments today. Low dose rate (LDR) interstitial implants most often have been carried out using uniform activity sources, yielding a Quimby-like dose distribution. With remote afterloading and variable dwell times, planar implants seeking dose uniformity in planes parallel to the implant planes will yield optimized activity distributions similar to those of the Manchester system. Even with more complex treatment target and source geometries, an understanding of the basic principles of source loading and consequent dose distribution, as illustrated by the early systems, is essential to effective interstitial brachytherapy treatment planning. The early implant systems, including those guiding intracavitary implants, were developed for use with 226 Ra, which was the primary isotope for use in brachytherapy from the earliest days of radiation therapy until its gradual replacement by 137Cs and 192Ir, completed in the United States about two decades ago. The use of active needles for interstitial brachytherapy was curtailed and generally replaced by the use of inserted catheters that could be afterloaded with 192Ir seeds or wires in the convenience of the patient’s room. The transition from radium needles to iridium seeds or wire did not, however, change the basic considerations of activity distribution and consequent dose distribution, and the early systems remained in use with minor modifications. In the following sections we will examine specifically the results of
373
374
Robert D. Zwicker
uniform planar source activity distribution, as examined first by Quimby (1932), and pursued in more detail in the ensuing years by a number of others (Anderson, Hilaris, and Wagner 1985; Anderson and Osian 1986; Kwan et al. 1983; Pierquin et al. 1978; Saw and Suntharalingam 1988; Zwicker, Schmidt-Ullrich, and Schiller 1985).
The Quimby System Planar Source Distribution (226Ra) The earliest planar brachytherapy system was described by E. H. Quimby in 1932. This in fact was not a true system in the sense of examining possible source distributions and providing consequent guidance to clinicians in the selection and placement of sources. It simply assumed an equally spaced array of uniform 226Ra source activities and presented the resulting exposure rates to be achieved at various distances from the source plane along a line perpendicular to and through the center of the implanted plane.
Dose Homogeneity It is of interest to examine the dose distributions resulting from a Quimby-like (uniform activity) planar implant, as shown for a single plane 192Ir implant in Figure 1. It is apparent and not surprising that in a parallel plane at a fixed distance from the implant plane, the dose is higher near the center of the implant than at the edges. This is expected because points near the center of the implant are in close proximity to more sources than are points near the edges. The isodose surfaces surrounding the implant are therefore thicker near the center than near the edges, as indicated in Figure 1, and this fact must be taken into consideration in planning a Quimby-type implant. Especially for large single-plane implants, it is important that appropriate planning should be carried out to ensure that doses delivered near the implant boundaries are sufficiently high to achieve the clinical intent of the procedure. The original Quimby system was applied only to the dosimetry of surface applicators, and did not address the problems encountered with implants of two or more planes. The characteristics of biplanar implants, however, are similar to those of single planes in that higher doses are typically delivered in the central portions of the target volume rather than near the edges, and the isodose surfaces of interest are thicker near the center of the implant. An example is given in Figure 2, which shows a 6×6 cm biplanar implant designed to treat a target volume of dimensions 6×6×2.5 cm along the symmetry axes. The reference isodose curve in this case is shown extending to the target boundaries along these axes, leaving the corners of the rectangular target volume at a lower dose. Points near the center of the implant, midway between the two implant planes, receive doses around 25% to 30% higher than the reference dose. If the chosen reference dose for this implant is extended outward to guarantee complete coverage of the target volume, the central dose will exceed the reference dose by an even greater amount. Elucidation and optimal exploitation of these characteristics of Quimby-like implants was a major impetus for the development of later implant systems specific to the use of 192Ir seeds or wires.
Planar Exposure Tables In the original Quimby work, tables were published which presented the exposure per milligram-hour to the center of the field at various distances from the 226Ra implant plane, for various implant areas. With the change to 192Ir and more recent changes in the basic calibration of these sources, the original tables of Quimby can no longer be regarded as directly useful for dose determination.
19–Quimby-Based Brachytherapy Systems
375
Figure 1. Isodose curves for a single plane 192Ir seed implant with seed and ribbon spacings of 1.0 cm, and uniform source activities. (Reprinted from Saw, C. B., and N. Suntharalingam, “Reference dose rates for single- and double-plane Ir-192 implants,” Med Phys 15:391–396, © 1988 with permission of AAPM.)
Volume Implants For Quimby volume implants sources were expected to be uniformly distributed and spaced 1 to 2 cm apart. A table was provided for varying implant volumes, giving the source strengths required to deliver 1000 cGy minimum dose inside the implanted volume (Anderson and Presser 1995). Corrections for elongated volumes were the same as for the Manchester system. As with Quimby planar implants, tissues near the center of the implanted volume are expected to receive doses significantly higher than the minimum dose delivered at the volume boundaries.
The Kwan System Source Distribution (192Ir) Kwan and colleagues (1983) took note of the considerable dose inhomogeneity that is unavoidable when uniform activity seeds in a planar distribution are used to treat rectangular target volumes. These researchers carried out detailed studies of the doses delivered at two points of particular interest in Quimby-
376
Robert D. Zwicker
Figure 2. Isodose curves for a biplanar 192Ir implant with seed and ribbon spacings of 1.0 cm, and uniform source activities. (Reprinted from Zwicker, R. D., and R. Schmidt-Ullrich, “Dose uniformity in a planar interstitial implant system,” Int J Radiat Oncol Biol Phys 31:149–155, © 1995 with permission from Elsevier.)
like monoplane and biplanar implants. The studies incorporated square source planes, with sources spaced at 1.0 cm intervals along the source ribbons. The intraplanar ribbon spacing and interplanar separation were both set at either 1.0 cm or 1.5 cm, and the resulting dose homogeneity was reported and evaluated.
Prescription Point Locations The first point of interest in this study is located on the boundary of the rectangular target volume, near a corner of the volume, and is associated with the minimum dose to be allowed to the target tissues (Figure 3). The second point is near the geometric center of the target volume, positioned midway between seeds, and is identified as a target dose maximum. With the reference dose defined in this manner, the ratio of the minimum to maximum target doses can be quite low, depending on implant construction (sources inside, outside, or just at the target boundaries) and the lateral dimensions of the implant plane or planes. Based on a combination of dose uniformity and normal tissue sparing, Kwan et al. (1983) concluded that the best source configuration is that illustrated in Figure 3b, with the sources implanted exactly to the edge
19–Quimby-Based Brachytherapy Systems
377
Figure 3. End-on view of biplanar implant showing interest points in Kwan system. (Reprinted from Kwan, D. K., A. R. Kagan, A. J. Olch, P. Y. M. Chan, B. Hintz, and M. Wollin,“Single and double-plane iridium-192 interstitial implants: Implantation guidelines and dosimetry,” Med Phys 10:456–461, © 1988 with permission of AAPM.)
of the target volume. With this configuration in place, the characteristics of implants of dimensions 2×2 cm to 10×10 cm were examined in this study.
Dose Tables/Graphs Figure 4 shows Kwan’s dose rate curves for the maximum and minimum target dose points, plotted as a function of implant area, for a 2.0 cm target thickness, with individual seed activities of 0.5 mCi. Results for the two possible ribbon and interplanar separations of 1.0 and 1.5 cm are shown. These curves can be used for preplanning of biplanar implants by scaling the seed activity to achieve a desired minimum dose
378
Robert D. Zwicker
rate for implants of thickness 2.0 cm constructed in this manner. For a 2.5 cm target thickness, the dose rates given in Figure 4 must be scaled downward by a factor of 0.87. Figure 5 shows the ratio of the minimum-to-maximum target dose for biplanar implants, again plotted as a function of implant area. Kwan et al. (1983) concluded from these curves that for biplanar implants treating target thicknesses of 2.0 or 2.5 cm, a ribbon and interplanar separation of 1.5 cm is preferred. Even then, however, the minimum-to-maximum dose ratio is seen to vary from 0.5 down to about 0.38, which suggests that if the prescription dose is identified as the minimum target dose for such an implant, a large fraction of the target volume near the center of the implant may receive doses at levels two or more times the prescription dose.
Figure 4. Dose rate curves from Kwan system showing maximum and minimum target dose points as a function of implant area. Target thickness is 2.0 cm and seed activities 0.5 mCi. (Reprinted from Kwan, D. K., A. R. Kagan, A. J. Olch, P. Y. M. Chan, B. Hintz, and M. Wollin,“Single and double-plane iridium-192 interstitial implants: Implantation guidelines and dosimetry,” Med Phys 10:456–461, © 1988 with permission of AAPM.)
19–Quimby-Based Brachytherapy Systems
379
Figure 5. Ratio of minimum to maximum target dose for biplanar implants, according to Kwan system. (Reprinted from Kwan, D. K., A. R. Kagan, A. J. Olch, P. Y. M. Chan, B. Hintz, and M. Wollin, “Single and double-plane iridium-192 interstitial implants: Implantation guidelines and dosimetry,” Med Phys 10:456–461, © 1988 with permission of AAPM.)
For rectangular implant areas, Kwan et al. (1983) proposed the use of an equivalent square, determined as for external beam dosimetry, to extract data from the square-area curves for planning purposes. Their studies showed this technique to be accurate to within about 3% for single-plane and 5% for two-plane implants. Studies of actual implants in which the source positions sometimes deviated by ±2 mm from their ideal locations yielded agreement in the minimum target doses within 5% between the computer treatment planning calculations and the published dose rate curves.
The Memorial System Physical Optimization Workers at Memorial Sloan-Kettering Cancer Center (MSKCC) first published guidelines for an 192Ir implant design over four decades ago (Laughlin et al. 1963). The system later introduced at MSKCC by Anderson, Hilaris, and Wagner (1985) and by Anderson and Osian (1986) was intended to make use of existing in-house stocks of 192Ir seeds in ribbons, and a nomograph was developed for staggered biplanar implants to determine the number of ribbons needed in the base plane for a fixed-seed activity to achieve a dose rate of about 10 Gy/day at a reference point near the periphery of the implant volume. It was expected that this choice of activity distribution and reference dose point selection would represent
380
Robert D. Zwicker
a “physical optimization” in some sense, achieving the best possible approximation to the desired dose rate distribution with the sources available. No assumptions are made concerning the intraplanar seed spacings. For these implants the superficial plane has one fewer ribbon than the base plane, and the interplanar separation is 1.5 cm.
Prescription Point Location The reference point for this system is described as being located 1.5 spacing units (usually 1.5 cm) inside end sources in the ribbon direction, midway between edge ribbons, and midway between source planes. As the reference point is near the implant corner, it is expected that the reference isodose surface will cover most of the target tissues if the target thickness is 1.5 cm. 192
Ir Nomograph
A nomograph was developed at MSKCC to make use of in-house stocks of 192Ir of known seed strength (Figure 6). The purpose of the nomograph was to determine, for a given seed activity, the number of 192Ir ribbons required to deliver a specified dose rate to the reference isodose surface. Input data needed to use the nomograph include the active ribbon length, base plane width, and seed strengths of available sources. A similar approach was later developed for volume implants using 125I, and still later for 103Pd implants.
Figure 6. Nomograph used to determine source data from the Memorial system. (Reprinted from Anderson, L. L., and A. D. Osian, “Brachytherapy optimization and evaluation,” Endocuriether Hypertherm Oncol (now called Brachytherapy) 2:S25–S32. © 1986 with permission from American Brachytherapy Society)
19–Quimby-Based Brachytherapy Systems
381
The Tufts/MCV System (Zwicker System) Source Distribution/Dose Uniformity in Seed Implants In the work of Zwicker, Schmidt-Ullrich, and Schiller (1985) it was recognized that the conditions of the Manchester system for higher activity at the implant periphery or extension of active sources outside the target volume could often not be realized. This was especially true of breast implants, in which anatomical constraints generally precluded the use of crossing needles or catheters, and extension of the implanted area beyond the target volume could leave active sources at the level of the skin, likely resulting in added discomfort for the patient and a poor cosmetic result. It was further recognized that the need for enclosing a rectangular volume in the prescription isodose surface was not clear, as tumors and surgical beds are not expected to assume a rectangular shape. In view of this, a system was developed for biplanar implants of target thickness 2.0 to 3.5 cm which would guarantee the extension of the reference isodose surface to the exact target boundaries along the three symmetry axes, but would allow the treated volume to be thinner than the stated target thickness (but no thinner than the interplanar source separation) near the implant corners. It was expected that this would lead to a more homogeneous dose distribution in terms of the central planar doses. This system also assumed intraplanar seed and ribbon spacings of 1.0 cm, but allowed the interplanar separation to vary in order to effect exact coverage of the target along the symmetry axes.
Prescription Surface Definition Figure 7 illustrates the concepts of this system for a 6×6×2.5 cm biplanar implant with an optimized interplanar separation of 1.3 cm. It is seen that the reference isodose provides exact coverage of the target volume along the implant symmetry axes. A greater interplanar separation for this implant would force the reference isodose curve out beyond the target boundary in the direction of the target thickness, while a smaller separation would increase unnecessarily the dose in the central plane relative to the reference dose.
Figure 7. Isodose curves for a biplanar 6×6×2.5 cm 192Ir implant planned by the Zwicker system, with seed and ribbon spacings of 1.0 cm, and optimized interplanar separation of 1.3 cm. (Reprinted from Zwicker, R. D., R. Schmidt-Ullrich, and B. Schiller, “Planning of Ir-192 seed implants for boost irradiation to the breast,” Int J Radiat Oncol Biol Phys 11:2163–2170, © 1985 with permission from Elsevier.)
382
Robert D. Zwicker
Interplanar Spacing Exact coverage of the target along all three symmetry axes is accomplished by finding the optimal interplanar separation that accomplishes this. Table 1 gives dose rates and optimized interplanar separations for biplanar implants with implant areas of 3×3 to 10×10 cm, target thicknesses of 2.0 to 3.5 cm, implanted with 1.0 mCi seeds. For implant planning purposes, the target dimensions are determined by any combination of direct examination, surgical report, films, and/or three-dimensional (3-D) imaging deemed appropriate. Once the target has been defined, the areas of the implant planes are known, and the interplanar separation needed for an optimal implant is taken from the table. As in the Kwan system, for rectangular implants an equivalent square is used to enter the data tables, but in this case the interplanar separation is reduced by 1 mm.
Dose Tables Table 1 shows the expected dose rates at the reference dose position for the range of biplanar implants described above, assuming seed activities of 1.0 mCi. For seed ordering purposes, these values can be scaled to find the required activity to achieve any desired dose rate. It is of interest that with the reference (prescription) dose defined in the manner described here and the interplanar separation varied for optimal target coverage, the dose near the geometric center of the implant, midway between planes, is only about 25% to 30% higher than the reference dose. A later work in this series (Zwicker and Schmidt-Ullrich 1995) investigated the effects of increasing the intraplanar ribbon spacing to 1.5 cm, leaving the seed spacing at 1.0 cm. It was shown that in an idealized point-source case, the small volume treated to very high dose in the vicinity of the seeds is increased by roughly the square root of the ratio of the number of seeds in the implant at 1.0 spacing to the number of seeds in the implant at 1.5 cm spacing. For typical cases, this translates to an increase of about 30% in the high dose volume if the intraplanar ribbon spacing is increased to 1.5 cm. In spite of this, circumstances sometimes dictate that the number of catheter insertions be minimized, so tables of dose rates and interplanar separations for 1.5 cm ribbon spacings were presented in this paper. This is reproduced in Table 2. Table 1. 192Ir Double-plane Implant Table. Ribbon Spacing = 1.0 cm. t(cm)
2.0
2.5
3.0
3.5
A(cm) 3×3
s 1.2
DR 61
s 1.5
DR 53
s 1.8
DR 46
s 2.1
DR 40
4×4
1.2
69
1.5
60
1.7
53
2.0
47
5×5
1.1
79
1.4
69
1.7
61
1.9
55
6×6
1.1
84
1.3
75
1.6
67
1.9
60
7×7
1.0
92
1.3
81
1.5
73
1.8
66
8×8
1.0
96
1.2
87
1.5
78
1,7
71
9×9
0.9
103
1.2
91
1.4
83
1.7
75
10×10
0.9
107
1.1
97
1.4
87
1.6
80
Double-plane implant, seed activity = 1.0 mCi; dose rate DR in cGy/h. Note: For rectangular implants, use equivalent square field size formula, then reduce s by 1.0 mm. (Reprinted from Zwicker, R. D., R. Schmidt-Ullrich, and B. Schiller, “Planning of Ir-192 seed implants for boost irradiation to the breast,” Int J Radiat Oncol Biol Phys 11:2163–2170, © 1985 with permission from Elsevier.)
19–Quimby-Based Brachytherapy Systems
383
Table 2. 192Ir Double-plane Implant Table. Ribbon Spacing = 1.5 cm. t(cm)
2.0
2.5
3.0
3.5
A(cm) 3×3
s 1.2
DR 38.5
s 1.5
DR 34.0
s 1.8
DR 30.1
s 2.1
DR 26.8
4.5×4.5
1.1
46.7
1.4
41.7
1.7
37.3
2.0
33.5
6×6
0.9
55.2
1.3
49.0
1.6
44.3
1.9
40.2
7.5×7.5
0.8
60.8
1.2
54.5
1.5
49.5
1.8
45.1
9×9
0.7
66.3
1.0
60.4
1.3
55.2
1.6
50.5
10.5×10.0
0.6
70.3
1.0
63.8
1.3
58.4
1.6
53.7
Double-plane implant, seed activity = 1.0 mCi; dose rate DR in cGy/h. (Reprinted from Zwicker, R. D., and R. Schmidt-Ullrich, “Dose uniformity in a planar interstitial implant system,” Int J Radiat Oncol Biol Phys 31:149–155, © 1995 with permission from Elsevier.)
The Saw Method Optimization Indices Saw and Suntharalingam (1988) recognized the influence of the choice of reference dose rate on the quality of an implant and carried out a detailed study of this effect using three volumetric indices to characterize implant quality. The coverage index (CI) was defined in terms of the percent of the target volume receiving dose rates equal to or greater than the reference dose rate. The relative dose homogeneity index (HI) gave the percent of the target volume receiving dose rates between 100% and 150% of the reference dose rate. The external volume index (EI) represented the percent (normalized to target volume) of volume outside the target volume receiving dose rates equal to or greater than the reference dose rate.
Optimal Reference Dose Rates Calculations were carried out for two implant geometries: a single plane treating a 1.0-cm-thick volume, and a biplanar implant treating a volume of thickness 2.5 cm. Implant areas varied from 3×3 cm to 12×12 cm. Intraplanar seed and ribbon spacings were 1.0 for the single-plane implants. For the two-plane implants, seed spacings were 1.0 cm within the ribbons, but both intraplanar ribbon spacing and interplanar separation were 1.5 cm. Seed activities were 1.0 mCi in both cases. The study examined the choice of reference dose rate for each implant geometry, and plotted the three indices as functions of reference dose. It was found in the cases presented that the HI reached a maximum, typically in the range where the external coverage index was at a low value. The plots for biplanar implants are shown in Figure 8. It was proposed, therefore, that the optimal dose rate for each geometry was that at which HI was maximum. Values of EI, CI, and HI were then plotted for optimized implants (Figure 9). The coverage index over the range of implant areas studied varied from 85% to 95%.
Generalized Optimization Indices Further studies of the effect of reference dose choice on dose uniformity in biplanar implants were carried out with variable interplanar separations and with the upper limit of the HI also allowed to vary (Zwicker
384
Robert D. Zwicker
Figure 8. External volume index (EI), coverage index (CI), and dose homogeneity index (HI) plotted as a function of reference dose, for a 6×6 cm biplanar implant, as described in the Saw system. Curves are for implants with seed spacing 1.0 cm, and intraplanar and interplanar ribbon spacings of 1.5 cm. Implant areas: (a) 3×3 cm; (b) 6×6 cm; (c) 9×9 cm; (d) 12×12 cm. (Reprinted from Saw, C. B., and N. Suntharalingam, “Reference dose rates for single- and double-plane Ir-192 implants,” Med Phys 15:391–396, © 1988 with permission of AAPM.)
and Schmidt-Ullrich 1995). It was shown here that for variable target thicknesses, optimization of the interplanar separation was an important element in achieving acceptable dose uniformity.
References Anderson, L. L., and A. D. Osian. (1986). “Brachytherapy optimization and evaluation.” Endocuriether Hypertherm Oncol 2:S25–S32. Anderson, L. L., and J. L. Presser. “Classical Systems I for Temporary Interstitial Implants: Manchester and Quimby Systems” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds). Madison, WI: Medical Physics Publishing, pp. 301–322, 1995. Anderson, L. L., B. S. Hilaris, and L. K. Wagner. (1985). “A nomograph for planar implant planning.” Endocuriether Hypertherm Oncol 1:9–15.
19–Quimby-Based Brachytherapy Systems
385
Figure 9. Optimized values of EI, CI, and HI for biplanar implants of implant area 3×3 to 12×12 cm. (Reprinted from Saw, C. B., and N. Suntharalingam, “Reference dose rates for single- and double-plane Ir-192 implants,” Med Phys 15:391–396, © 1988 with permission of AAPM.)
Kwan, D. K., A. R. Kagan, A. J. Olch, P. Y. M. Chan, B. Hintz, and M. Wollin. (1983). “Single and double-plane iridium-192 interstitial implants: Implantation guidelines and dosimetry.” Med Phys 10:456–461. Laughlin, J. S., W. M. Silver, E. I. Holodny, and F. W. Ritter. (1963). “A dose description system for interstitial radiation therapy.” Am J Roentgenol 89:470–490. Paterson, R., and H. M. Parker. “Interstitial Treatments” in Radium Dosage: The Manchester System. W. J. Meredith (ed.). Baltimore, MD: Williams and Wilkins Company, pp. 28–38, 1949. Pierquin, B., A. Dutreix, C. H. Paine, D. Chassagne, G. Marinello, and D. Ash. (1978). “The Paris system in interstitial radiation therapy.” Acta Radiol Oncol 17:33–48. Quimby, E. H. (1932). “The grouping of radium tubes in packs and plaques to produce the desired distribution of radiation.” Am J Roentgenol 27:18. Saw, C. B., and N. Suntharalingam. (1988). “Reference dose rates for single- and double-plane Ir-192 implants.” Med Phys 15:391–396. Zwicker, R. D., and R. Schmidt-Ullrich. (1995). “Dose uniformity in a planar interstitial implant system.” Int J Radiat Oncol Biol Phys 31:149–155. Zwicker, R. D., R. Schmidt-Ullrich, and B. Schiller. (1985). “Planning of Ir-192 seed implants for boost irradiation to the breast.” Int J Radiat Oncol Biol Phys 11:2163–2170.
Chapter 20
Implant Design and Execution Robert D. Zwicker, Ph.D. Department of Radiation Medicine University of Kentucky, Lexington, Kentucky Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387 Implant Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387 Steel Needle/Catheter Lead Insertion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387 Catheter Seating: Buttons and Clamps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 Blind-End Catheters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 388 Freehand Versus Template Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 Dose Homogeneity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 Dose Homogeneity Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389 Source Placement and Dose Uniformity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 390 Planar Implant Preplanning: Source Positioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391 Quimby-like Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391 Manchester Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 391 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 392
Introduction Early interstitial implants were carried out using radium needles, generally inserted into the patient using special clamps, or, frequently, using the fingers to hold the sources. The difficulty of handling the sources and the speed with which the implantation had to be carried out were factors in reducing the quality of these implants. The introduction of afterloading techniques allowed a more relaxed atmosphere for positioning of source guides in the target tissues. For low dose rate (LDR) implants, the radioactive material, usually 192Ir, is then loaded into the carefully pre-positioned source guides, usually in the safety and convenience of the patient’s hospital room. For high dose rate (HDR) treatment, the source guides are connected to the treatment machine in the shielded HDR treatment room, and the remote-controlled source is stepped through the guides while the operators are in a shielded location outside the room. Interstitial implants as used in breast brachytherapy may involve the use of any number of source guides, typically 10 to 25 or more. Techniques for placement of source guides include the use of plastic catheters and/or steel needles, placed either freehand or with the help of a template. These techniques are described below
Implant Techniques Steel Needle/Catheter Lead Insertion The most common method of source guide placement (Perez, Zwicker, and Williamson 2003) involves the use of a hollow steel needle (a stylette, typically 17 gauge), inserted into and through the target tissues in the desired location and threaded with the solid-lead end of a catheter of the same gauge as the needle, as shown in Figure 1. Often the inserted needles are left in place while the next and remaining needles are inserted, as this greatly facilitates the relative positioning of the source guides (catheters). Once the needles are positioned and the catheter leads threaded through them, the needles, with the leads, are then pulled through the tissue and out the other side. The catheter follows into the needle’s previous position, and is pulled until the motion is stopped by a button on the end of the catheter (see Figure 1). The open end of the catheter is then usually secured by means of a friction washer or clamp, and the catheter may
387
388
Robert D. Zwicker
Figure 1. Insertion of guide needles and implant catheters. (Reprinted from Neblett (1995), Figure 1, p. 283, with permission from Medical Physics Publishing, ©1995 American Association of Physicists in Medicine.)
then be cut to any desired length. The friction washer or clamp may be sutured to the skin. If it is desired to treat tissues up to the skin, spacers may be positioned between the button and the skin to allow the source to protrude to the level of the skin. This can be especially helpful in the treatment of base of tongue disease (Schmidt-Ullrich et al. 1991). For implants involving a large number of catheters, later identification of the catheters may be facilitated by the use of catheters of different colors. In conventional brachytherapy planning, the catheters are filled with dummy ribbons containing radioopaque marker seeds at 1.0 cm intervals, along with some other coding seed markers for ease of identification. With the markers in place, orthogonal films are usually taken and these are used to digitize the seed locations into the treatment-planning computer. The catheter buttons and washers may also be radio-opaque for easy visualization on radiographs. More details are included in another chapter in this monograph. For LDR implants, the 192Ir seed ribbons are cut to the required length (i.e., the planned number of seeds) and are inserted and secured in the catheters. For HDR implants, the open catheter ends are covered with protective caps that can be easily removed at treatment time.
Catheter Seating: Buttons and Clamps A variety of button, washer, and clamp types have been introduced over the years, and a number are still available. These include various sizes, shapes, and materials, often designed for patient comfort and to minimize the intrusion of the button or clamp into the patient’s skin. Some examples are shown in Figure 2. For conventional radiographic localization it is often convenient that the skin position be made visible in radiographs by the use radio-opaque buttons, washers, or clamps. For CT-based treatment planning, however, such radiographic contrast may cause harmful artifacts.
Blind-End Catheters Situations sometimes arise in which the interstitial source guides cannot be passed completely through the tissue, so that one end of the catheter or needle must remain inside the patient. Examples of such cases include brain implants, interstitial gynecological implants, and some deep organ brachytherapy treatments.
20–Implant Design and Execution
389
Figure 2. Button end (top) and blind end (bottom) catheters, washers, and buttons for securing catheters.
Special source guides have been developed to accommodate these circumstances. An example of these is shown in Figure 3.
Freehand Versus Template Implants The use of templates to guide the needles along the desired paths is expected to yield an implant in which the source locations are closely representative of the preplanned locations (Figure 3). This technique is common in gynecological and prostate implants, in which the path of the needles is easily predicted and fairly reliably followed. In some other treatment sites such as head and neck and breast, anatomical constraints may present problems that are more easily addressed with the flexibility of a freehand implant. Such constraints may be addressed to some extent with the use of a custom template, although this will lead to a higher cost and longer planning time prior to the implant. In some techniques the template is left in place for the duration of the dose delivery, while in others the template may be removed after the source guides are positioned. Some critics of template implants claim that they are associated with a greater degree of discomfort for the patient, as tissues placed under pressure by the needles will not be relaxed when the needles are released. The question of freehand versus template implants is still under debate at many centers.
Dose Homogeneity Dose Homogeneity Indicators Source guide placement is a critical step in the process of interstitial brachytherapy treatment. Even with the high degree of flexibility in dwell times available with HDR, a poor source guide placement can obviate the possibility of acceptable target coverage with good dose uniformity. In general, placement of source guides tightly inside the target volume will lead to unacceptably high doses in the central interior of the target, while placement well outside the target may lead to unnecessary treatment of surrounding healthy tissues to high doses. For characterization of dose uniformity, several dose homogeneity indicators have
390
Robert D. Zwicker
Figure 3. Template with blind-end catheters (Syed-Neblett design).
been introduced. The method of Paul (Paul et al. 1986) and others (Saw and Suntharalingam 1988; Wu, Ulin, and Sternick 1988) is to rate dose uniformity in terms of the relative volume of tissue treated to doses between the reference dose and some acceptable upper limit, usually 150% of the reference dose. The Radiation Therapy Oncology Group (RTOG) 95-17 breast brachytherapy protocol (Kuske, Bolton, and Hanson 1998) made use of a homogeneity indicator more closely related to the concepts of the Paris system (Pierquin et al. 1978), defining a mean central dose as an average minimum dose between the implant planes, in the central plane normal to and bisecting the source ribbons (Figure 4). The ratio of the prescription dose to the mean central dose was taken as the homogeneity index, and for an acceptable implant, this was expected to be greater than or equal to 75%. Some measure of dose uniformity should be adopted at all active clinics, and each implant evaluated against the in-house dose homogeneity standard. For further discussion on the topic, see chapter 25 in this monograph on post-procedure evaluation.
Source Placement and Dose Uniformity Some simple guidelines can help in effective source guide positioning. For example, the distance between seeds or dwell positions should be kept at a practical minimum. For simple point-source Quimby-like (uniform activity distribution) implants it has been shown that as the seed number in a planned implant decreases (and seed activity correspondingly increases by a factor f), the total volume enclosed in the highdose bubbles surrounding each source will increase by approximately the square root of the factor f (Zwicker and Schmidt-Ullrich 1995). Direct calculations have shown that for biplanar 192Ir implants, increasing the intraplanar ribbon spacing from 1.0 to 1.5 cm can increase the high dose volume by roughly 30%. Hence any increase in the intraplanar seed or ribbon spacing, as is often desired in order to reduce the number of needle insertions, must be considered in the light of the higher degree of dose inhomogeneity which follows. For biplanar implants the interplanar separation is also critical to the achievement of good target coverage and dose uniformity (Zwicker, Schmidt-Ullrich, and Schiller 1985). Guidelines for interplanar
20–Implant Design and Execution
391
Figure 4. Illustration of Mean Central Dose as average dose at the center of triangles connecting sources, in a plane normal to and bisecting the sources. (Adapted and reprinted from Hanson and Graves (1995), Figure 6, p. 371, with permission from Medical Physics Publishing, ©1995 American Association of Physicists in Medicine.)
separation have been given in the literature for both LDR and HDR implants, as summarized in the following section.
Planar Implant Preplanning: Source Positioning Quimby-like Implants For two-plane Quimby-like (Quimby 1932) implants, interplanar separations have been given for a range of target thicknesses, with the aim of guaranteeing coverage of the target along the symmetry axes. It has been shown that these guidelines lead to implants with dose homogeneity that is optimal in some sense. The details of these guidelines are contained in the publications of Zwicker, SchmidtUllrich, and Schiller (1985) and Zwicker and Schmidt-Ullrich (1995) and are described in the previous chapter on Quimby-based brachytherapy systems.
Manchester Implants For simple biplanar HDR implants planned to give a Manchester-type (Paterson and Parker 1949) dose distribution (uniform dose in a plane at a fixed distance from an implant plane), the optimal interplanar separations has been shown (Zwicker and Schmidt-Ullrich 1995) to stand in a simple relationship to the target thickness T: S = T/1.414.
392
Robert D. Zwicker
An idealized implant constructed according to this guideline will show a highly uniform dose in the central plane parallel to the implant planes, and that dose will be the same as the peripheral prescription dose. One problem with the Manchester approach is that it leads to a prescription isodose surface that is highly rectangular in shape, which seems a somewhat unnatural fit to the basically amorphous volume of disease typically encountered. In view of this, a careful preplanning of each implant should be carried out, making use of all the information available to characterize the target size and shape. In such cases, implant guidelines such as that given above can be expected to aid in predicting the result of any proposed source guide positioning scheme, but would not in general be rigidly applied to every implant.
References Hanson, W. F., and M. Graves. “ICRU Recommendations on Dose Specification in Brachytherapy” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds). Madison, WI: Medical Physics Publishing, pp. 361–378, 1995. Kuske, R. R., J. S. Bolton, and W. Hanson. (1998). RTOG 95-17: A Phase I/II Trial To Evaluate Brachytherapy as the Sole Method of Radiation Therapy for Stage I and II Breast Carcinoma. Philadelphia: Radiation Therapy Oncology Group 1998:1–34. Neblett, D. L. “Clinical Techniques and Applications Available for Interstitial Implantation” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds). Madison, WI: Medical Physics Publishing, pp. 281–300, 1995. Paterson, R., and H. M. Parker in Radium Dosage: The Manchester System. W. J. Meredith (ed.). Baltimore, MD: Williams and Wilkins Company, pp. 6–38, 1949. Paul, J. M., R. F. Koch, P. C. Philip, and F. R. Khan. (1986). “Uniformity of dose distribution in interstitial implants.” Endocuriether Hypertherm Oncol 2:107–118. Perez, C. A., R. Zwicker, and J. Williamson. “Clinical Applications of Brachytherapy. I. LDR and PDR” in Principles and Practice of Radiation Oncology. C. A. Perez, L. W. Brady, E. C. Halperin, and R. K. Schmidt-Ullrich (eds.). Philadelphia: Lippincott, Williams and Wilkins, pp. 538–603, 2003. Pierquin, B., A. Dutreix, C. H. Paine, D. Chassagne, G. Marinello, and D. Ash. (1978). “The Paris system in interstitial radiation therapy.” Acta Radiol Oncol 17:33–48. Quimby, E. H. (1932). “The grouping of radium tubes in packs and plaques to produce the desired distribution of radiation.” Am J Roentgenol 27:18. Saw, C. B., and N. Suntharalingam. (1988). “Reference dose rates for single- and double-plane Ir-192 implants.” Med Phys 15:391–396. Schmidt-Ullrich, R., R. D. Zwicker, A. Wu, and K. Kelly. (1991). “Interstitial Ir-192 implants of the oral cavity: The planning and construction of volume implants.” Int J Radiat Oncol Biol Phys 20:1079–1085. Wu, A., K. Ulin, and E. S. Sternick. (1988). “A dose uniformity index for evaluating Ir-192 interstitial breast implants.” Med Phys 15:104–107. Zwicker, R. D., and R. Schmidt-Ullrich. (1995). “Dose uniformity in a planar interstitial implant system.” Int J Radiat Oncol Biol Phys 31:149–155. Zwicker, R. D., R. Schmidt-Ullrich, and B. Schiller. (1985). “Planning of Ir-192 seed implants for boost irradiation to the breast.” Int J Radiat Oncol Biol Phys 11:2163–2170.
Chapter 21
Advanced 3-D Planning Jean Pouliot, Ph.D., Étienne Lessard, Ph.D., and I-Chow Hsu, M.D. Department of Radiation Oncology, UCSF Comprehensive Cancer Center San Francisco, California Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393 Conventional Brachytherapy Planning (Catheter-based Planning) . . . . . . . . . . . . . . . . . . . . . . . . . . 393 3-D Imaging in Brachytherapy (Anatomy-based Planning) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395 Advanced 3-D Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395 A Change of Paradigm (Inverse Planning) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395 Intensity-Modulated Brachytherapy (IMBT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396 Treatment Inverse Planning Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396 Dose Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397 Dose Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397 Surface and Volume Dose Constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398 Flexible Definition of the Clinical Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401 Applicable to All Anatomical Sites Accessible to Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . 401 Multiple Targets, Each with its Own Dose Prescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401 Multiple Organs at Risk, Each with its Own Dose Prescription . . . . . . . . . . . . . . . . . . . . . . . . . . 402 Facilitate Adjustment Based in Specific Clinical Situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402 Class Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403 Clinical Implementation and Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 403 Prostate HDR Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404 Comparison Between GEO and IPSA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404 Protection of the Penile Bulb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405 MRS-defined Boost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405 Number of Catheters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 406 Gynecological Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408 Interstitial Gynecological Implant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408 Vaginal Boost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 408 Different Choice of Weighting Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409 Permanent Prostate Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409 Forward vs. Inverse Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409 Variation of Seed Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 409 Boost of Positive Biopsy Areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 410 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 412
Introduction Conventional Brachytherapy Planning (Catheter-based planning) The last decade has seen major changes in the way radiation treatments are delivered. The century-old objective of radiation therapy, to deliver a curative dose to the target while preserving normal tissues, can now be aimed at with a high degree of sophistication. However, in spite of major improvements achieved with three-dimensional (3-D) imaging modalities that allow the anatomy to be properly defined, brachytherapy is slowly taking advantage of these important new pieces of information. The insertion of the seeds for permanent implants or of catheters for afterloading treatment is now image guided, improving greatly the localization of the applicators. At the dose planning stage however, not so long ago the
393
394
Jean Pouliot et al.
(a)
(b) Figure 1. Evolution of the type of information presented to the planning team over the years. (a) Traditional orthogonal pair of x-ray films showing the dummy catheters. (b) CT or MRI based planning where the anatomy and the catheters are defined in 3-D.
anatomy was only indirectly taken into account. For high dose rate (HDR) brachytherapy, catheters were placed in the target volume and it was assumed that if the dose distribution covers the catheters, it should also cover the anatomy. Imaging was commonly used to set the treatment margins, but optimized dose distributions were simply based on considerations such as catheter positions and desired dose to a few defined points. This necessarily resulted in an approximation of the shape of the anatomy. For the case of treatment of the prostate, volume optimization resulted in a dose distribution that was essentially cylindrical in shape. This may represent an “average” prostate, but each prostate has a very specific shape. Any approximation of its shape to a geometrical representation results in an overdosage of normal tissues surrounding the prostate. The organs at risk in the vicinity of the prostate may be included in this unnecessarily overdosed region. What do you see? The information provided to the team performing the dose planning has dramatically changed over time, both in nature and in quality. This evolution is illustrated in Figure 1. In Figure 1a, a conventional pair of orthogonal x-ray films is used for the planning of an interstitial HDR prostate implant. Dummy ribbons with radio-opaque markers are inserted in each catheter. The dosimetrist must digitize the dwell positions (after identifying which dummy sources belong to which catheter), and determine which of these dwell positions will be activated to cover the treatment length specified by the physician. With the exception of few patient dose points, a Foley catheter for the bladder, a rectum radioopaque marker, etc., the emphasis is placed on the catheters. Where are they, how many, how long should they be activated? This is why most, if not all, of the optimization tools developed were catheter based; i.e., they would produce a dose distribution closely related to the distribution of catheters. Because most of the anatomical information is unknown to the optimization tool, manual adjustments always follow to customize the plan.
Advanced 3-D Planning
395
3-D Imaging in Brachytherapy (Anatomy-based Planning) 3-D imaging modalities, mainly ultrasound and computed tomography (CT), but also magnetic resonance imaging (MRI) (Ménard et al. 2004; Pouliot, Lessard, and Hsu 2004) and functional imaging such as magnetic resonance spectroscopy (MRS) (Pouliot, Lessard, and Hsu 2004) or positron emission tomography (PET), are rapidly being adopted as a standard for dose planning in brachytherapy. Similarly to external beam radiation therapy, the anatomical structures of interest are contoured and used to define planning target volumes (PTVs) and organs at risk (OARs). Here, PTVs and CTVs (clinical target volumes) are considered the same in brachytherapy. A discussion to differentiate them is beyond the scope of this chapter. Equipped with these new imaging modalities, the dosimetrist is now provided with both anatomical and catheter position information. This is illustrated in Figure 1b. The anatomical structures (prostate, rectum, urethra, bladder, etc.) are visible, in addition to the catheters. Therefore, when sources are placed, or dwell times adjusted, the dose distribution can be calculated and the user can assess the amount of dose received by the organs, also opening the possibilities to compute organ-specific dose volume histograms (DVHs). Most of the optimization tools are catheter based and therefore do not take into account the relation of the dose coverage and the protection of OARs during the optimization process. Yet, a number of modern tools have made the final dose distribution adjustment a lot easier than before. Examples of such tools are the ability to drag an isodose with the mouse while dwell times are adjusted or source positions are modified on the fly. A variety of dose-point optimization tools are also available, where dose can be optimized at certain points. All those tools however can only aim at one objective at a time, therefore relying on the user to judge what is an appropriate compromise simply based on visual assessment of the dose distribution and DVH. This approach is time consuming and the final treatment planning quality is often limited by the tight schedule of the clinical staff.
Forward or Conventional Planning A method of radiation treatment planning where the treatment parameters are first chosen and then the resulting dose distribution is calculated and evaluated.
Advanced 3-D Planning A Change of Paradigm (Inverse Planning) The inverse planning approach can be defined as a method of radiation treatment planning where one starts with the desired dose distribution, or clinical objectives, and then determines the treatment parameters that will achieve it. This is opposed to the conventional forward planning approach where the treatment parameters are first chosen and then the resulting dose distribution is calculated and evaluated. Since inverse treatment planning begins with the description of the desired dose distribution, it represents a change of paradigm in the planning process. CT or MRI contours of the CTV and OAR are used not only to define the anatomy for visual assessment and DVH calculation, but also to optimize the dose distribution. Therefore, they provide the physician with added flexibility and control to shape the dose distribution. In inverse planning, the anatomical features together with the dose constraints constitute the starting point of the dose optimization process. This requires that the multiple targets and OARs are known, and that the admissible and required dose coverage is specified. The computer knows the possible source positions because they need to be previously digitized, but from the user’s perspective, the emphasis is on the anatomy and the dose constraints. At the planning stage, the dwell times (for afterloader) or source
396
Jean Pouliot et al.
positions (for permanent implants) become irrelevant to the user. This is different from forward planning where the dose distribution is iteratively adjusted by modifying the dwell times or the source positions until an acceptable dose distribution is produced. With the inverse planning approach, the dosimetrist works directly on the dose distribution and the compromise between target coverage, dose homogeneity, and OAR protection. This change of perspective brings the planning process closer to the real concern of the clinician, the dose delivered to the anatomy. The main benefit of the inverse planning approach is that all clinical requirements (dose coverage, dose homogeneity, OAR protection, etc.) are simultaneously and automatically taken into account in the planning process. As we move from the old two-dimensional (2-D) catheter-based planning to the new 3-D anatomy-based planning, the necessity of an inverse-planning approach becomes clear. With hundreds of source positions, irregularly shaped volumes, multiple target and organ volumes, multiple organ sensitivities, the chance of finding a treatment plan that would optimally satisfy all the requirements with a trial and error strategy becomes more and more remote. While the conventional forward planning approach may be adequate for uncomplicated cases, the adjustment of a dose distribution to respect different dose
Inverse Planning A method of radiation treatment planning where one starts with the desired dose distribution, or objectives, and then determines the treatment parameters that will fulfill it. Opposed to forward planning.
constraints on various targets and organs at risk in a reasonable time for clinical application is often beyond human capabilities. As the clinical expectation from brachytherapy escalates, treatment-planning requirements can only be met with an efficient approach such as inverse planning.
Intensity-Modulated Brachytherapy (IMBT) The concept of inverse planning for brachytherapy is the same as in external beam radiation therapy. The same three steps are required. First, imaging modalities must be used to provide 3-D anatomical structures. Then an inverse planning optimization tool, along with defined dose constraints, is used to determine the optimal dose distribution. Finally, a computerized delivery unit is used to deliver the complex intensity patterns, by controlling the leaf positions in external beam radiation therapy (EBRT) or the source positions and the dwell times in brachytherapy. Because of the close analogy between intensity-modulated radiation therapy (IMRT) and inverse-planned brachytherapy, the latter can be referred to as intensity-modulated brachytherapy (IMBT).
Treatment Inverse Planning Optimization The current treatment-planning optimization problem can be formalized as a combinatorial optimization problem. A combinatorial optimization problem is either a minimization or a maximization problem and can be described as a pair: a set of solutions and an objective function that assigns a quantitative value to each solution based on the optimization objectives. The solution space and objective function carry together a concise description of the problem. The goal is to find the optimal solution included in the solution space that possesses the optimal quantitative value from the objective function. The problem considered in the present case is the selection of the optimal treatment plan. The group of all possible treatment plans forms the solution space and the objective functions mathematically describe the ideal treatment
Advanced 3-D Planning
397
plan based on clinical objectives. Following the inverse planning approach, the clinical objectives are described by means of dose constraints specified to each anatomical structure. The anatomical structures are represented by planar contours digitized by the physician on parallel cross-sectional images (CT or MRI) using a treatment-planning system. The algorithm uses these digitized anatomical structures to define in the 3-D space a set of dose points from which the dose distribution is evaluated based on the clinical objectives. The multiple objectives sought are to deliver the curative dose to all the targets, reduce the high dose regions inside the targets, and reduce the dose to the OARs and the normal tissues. The algorithm also required a set of 3-D positions from which the radioactive source may deliver the dose. The source dwell positions for the afterloader and the source template positions for permanent implant are also defined on parallel cross-sectional images using the treatment-planning system. Dose Distribution Based on the information provided by the treatment-planning system, the dose rate contribution from a given source position to a given dose point is calculated according to the AAPM task group 43 (TG 43) formalism (Rivard et al. 2004). The dose rate contribution is saturated for the source positions closer than 2 mm from the dose points to eliminate excessive contributions. The choice of this saturation distance defines the size of the high-dose bubbles within the target that are considered clinically relevant. This distance also takes into account the outer diameter of a typical catheter. The group of all dose rates from every source position to every dose point forms a dose rate matrix. In order to reduce the calculation time within the optimization iteration loop, this dose rate matrix is stored in a look-up table. The total dose Di delivered to a dose point i can be calculated from this look-up table and the set of dwell time values for the afterloader or the set of source positions for permanent implant. Dose Constraints The calculated dose distribution Di has to be translated into a quantitative value in order to measure its degree of fulfillment with the clinical objectives. The first step toward the evaluation of the dose distribution is to mathematically describe the physician’s requests by means of dose constraints. A dose constraint is a set of numerical values that defines the acceptable dose limits for a specific organ and their relative importance over the other clinical criteria. They convert the dose delivered to a dose point i into a penalty value Wi. This conversion is defined by the following penalty relation and illustrated in Figure 2.
m min ( Di − D min ) if Di < D min Wi = m max ( Di − D max ) if Di > D max if D min ≤ Di ≤ D min 0
(1)
The coefficients Dmin and Dmax represent the lower and the upper range of acceptable doses. If the dose is within the permissible dose range, the penalty is null. If the dose goes below or above the range, the penalty increases at rates mmin and mmax, respectively. Adjustment of the weights mmin and mmax sets the relative importance between the clinical criteria. The bigger the weight, the stronger is the penalty. The standard notation used to abridge the penalty relation is given below. mmin [Dmin Dmax] mmax
standard dose constraint notation
The advantage of using this type of penalty relation is that the translation from the physician’s expectations of the ideal dose distribution to a mathematical form is straightforward. The relation defines a clear border between acceptable and unacceptable doses and the physical meaning of dose constraints is clear.
398
Jean Pouliot et al.
Figure 2. General form of a dose constraint.
The physicians have much clinical experience that can be translated into such dose constraints. In addition, this approach of defining the objectives induces an intuitive understanding of the optimization results. Moreover, this method is flexible and can describe different clinical objectives such as target minimal dose coverage and organ at risk protection. A set of dose constraints for HDR prostate brachytherapy is illustrated at Figure 3. In that case, dose constraints of two targets are defined, the prostate and a dominant intraprostatic lesion (DIL), and three OARs: the urethra, the rectum, and the bladder.
Dose Constraint A set of numerical values that defines the acceptable dose limits for a specific organ and their relative importance over the other clinical criteria.
Surface and Volume Dose Constraints In order to represent a wide variety of clinical objectives, the dose constraints applied to each digitized volume are divided into two types: one to evaluate the dose at the surface of the volume, the other to evaluate the dose inside the volume. In the case of a target, the first acts on dose points generated on the surface of the volume, thus forcing the dose distribution to be conformal to the volume. The second acts on dose points generated inside the volume to control the dose homogeneity. The following notation is used to describe the dose constraints of the each volume: on in
mmin [Dmin Dmax] mmax mmin [Dmin Dmax] mmax
surface dose constraints volume dose constraints
This set of dose constraints provides the ability to constrain the dose delivered to each volume independently and also permits defining different Dmin and Dmax limits on the surface and inside each volume.
Advanced 3-D Planning
399
Figure 3. Set of dose constraints for HDR prostate brachytherapy including two targets, the prostate and a dominant intraprostatic lesion (DIL), and three organs at risk: the urethra, the rectum, and the bladder
Moreover, the different weights open the possibility to balance the clinical importance between low and high dose on the surface and inside each volume. All the weights defined over all volumes taken into account are relative to each other and correspond to relative clinical priorities. The convention is to define the mmin of the principal target surface as the reference weight and to set it at the arbitrary value of 100. From Local to Global Dose Prescription The concept of prescribing at a given point no longer applies for inverse planning; the prescription is global and based on the anatomy. Hence, the set of all dose constraints over all the digitized volumes taken into account is considered as the definition of this global prescription. Objective Function Once the dose constraints are set, the penalty values Wi are evaluated for each dose points i generated by the algorithm. Each dose point is associated with a specific anatomical structure (prostate, urethra, bladder, rectum, etc.) and a specific region of this anatomical structure (surface, volume) and so to a specific dose constraint. Finally, the sum of the penalty values Wi over all dose points i is performed to obtain the global penalty also known as the “cost function” or the “objective function.” This objective function mathematically describes the clinical objectives and is used to evaluate the quality of a given dose distribution. The closer the dose distribution is from the ideal dose distribution, the smaller the objective function value.
400
Jean Pouliot et al.
The minimum value of this objective function corresponds to the optimal treatment plan as defined by the set of dose constraints. This objective function that simply considers the dose distribution is adequate for the afterloader for which the needles are implanted prior to the treatment planning being performed. At the planning stage, only the dwell times can be adjusted since the needles are already in place. However, for permanent implant the objective function should also include additional characteristics such as the number of needles that need to be implanted. Effectively, many plans may be clinically equivalent for a given dose distribution, but the one with the least number of needles should be favored. A previous study reported that minimizing the number of needles in the implant minimizes the edema associated with post-implant trauma (Speight et al. 2000). Hence, in order to complete the objective function, an additional term is included that constraints the number of needles. Optimization Engine Once the set of dose constraints is defined, an optimization engine is needed to minimize the objective function and converge toward the ideal treatment plan. The optimization engine is conceived to find the optimal dose distribution related to the clinical objectives among the extremely large number of possible solutions included in the solution space. Because the objective function is nonlinear and therefore may present multiple minima, an optimization technique able to avoid these local minima is required (Deasy 1997). The choice of this nonlinear objective function instead of a more simple linear objective function offers more flexibility in facilitating the description of sophisticated dose distribution. Different mathematical methods can be applied to minimize the objective function. They can be divided in two groups: deterministic methods and stochastic methods. Deterministic optimization algorithms travel downhill continuously on the objective function surface until a minimum is reached. Deterministic optimization algorithms have been applied for HDR brachytherapy planning optimization (Milickovic et al. 2002; Gabor, Streeter, and Astrahan 2003) and permanent implant planning optimization (Roy et al. 1991; Chen, Boyer, and Xing 2000). Although deterministic optimization algorithms are faster, they yield a solution at the nearest local minimum depending on the starting conditions and the best starting condition of the nonlinear objective function is unknown. Consequently, the solution may get trapped in a local minimum that is not the global or near the global minimum except by pure chance. An option to overcome this generic problem is to repeatedly execute the optimization several times using randomly selected starting conditions. This process generates a pool of solutions from which the final solution can be selected. However, this method does not guarantee that the pool of solutions contains the global or a near the global minimum. Moreover, the physician still has to select one treatment plan from a pool that may contain hundreds of treatment plans. Alternatively, stochastic optimization algorithms such as genetic algorithms (GA) or simulated annealing algorithms (SA) can process any form of objective functions, including nonlinear objective functions. Stochastic optimization algorithms apply a random search and therefore have the ability to overpass the local minima. A GA class algorithm applies the principles of natural selection for the computation of complex problems. A GA class algorithm is interesting in its own right and has been previously applied for HDR brachytherapy planning optimization (Lahanas, Baltas, and Zamboglou 1999, 2003; Yu et al. 2000) and permanent implant planning optimization (Yu and Schell 1996; Yang et al. 1998; Yamada et al. 2003). However, GA was not originally developed as an optimization algorithm (DeJong 1993), and GA does not offer any statistical guarantee of global convergence to an optimal solution (Forrest 1993). Nevertheless, it should be expected that GA might be better suited for some problems than SA (Ingber 1996). On the other hand, SA class algorithm does offer a statistical guarantee of global convergence to an optimal solution (Geman and Geman 1984; van Laarhoven and Aarts 1987; Aarts and Korst 1989). For that reason, an SA approach seems also to be a reasonable choice for the present combinational optimization problem. Previous investigations have shown that SA can be used to govern an optimization
Advanced 3-D Planning
401
process to automatically and rapidly produce plans for prostate permanent implant treatments (Sloboda 1992; Pouliot et al. 1996, 1999; Taschereau, Roy and Pouliot 1999; Redpath 2002). The same concept has also been applied for prostate HDR brachytherapy (Lessard and Pouliot 2001). Because it takes some time and effort to become familiar with a given code, the ability to tune a given SA algorithm for use in more than one problem should be considered an important feature of an SA algorithm (Ingber 1996). Therefore, the SA algorithm previously developed for prostate HDR brachytherapy was extended to optimize any type of permanent and temporary brachytherapy (Lessard 2004). This thesis concluded that SA is suitable for treatment-planning optimization and produced reliable high-quality treatment planning. Treatment-planning optimization does not aim to find the mathematically optimal treatment plan, but to reach a clinically acceptable treatment plan in a clinically acceptable computation time. This work also demonstrates that an SA optimization engine carefully designed for a specific task can find an adequate solution in a very short time, opening the door to intraoperative treatment-planning optimization. The algorithm produced by this research is a comprehensive inverse planning algorithm for brachytherapy planning based on simulated annealing (IPSA). The next sections will outline the clinical and technical advantages brought along with the introduction of this new technology in the clinic.
IPSA An automated treatment-planning tool based on a constraint satisfaction approach (inverse planning) and solved by a stochastic optimization engine (simulated annealing).
Flexible Definition of the Clinical Objectives Applicable to All Anatomical Sites Accessible to Brachytherapy The sophisticated nonlinear objective function described in the previous section offers sufficient flexibility to describe multiple types of clinical objectives for multiple types of anatomical structures. This approach has provided an efficient way of defining a wide variety of dose distributions from a standard implant procedure to new sophisticated implant procedure ever considered before. Therefore, a supple objective function simplified the clinical introduction of the inverse planning in the clinic starting with standard treatment procedure such as interstitial prostate temporary implant and facilitated the creation of a more sophisticated treatment procedure such as interstitial gynecological temporary implant. Hence, a wide variety of anatomical sites, some illustrated at Figure 4, have been treated since the clinical introduction of IPSA: prostate (Lachance et al. 2002; Hsu et al. 2004); penis, gynecological (uterus, cervix, vaginal wall, and vulva) (Lessard, Hsu, and Pouliot 2002; DeWitt et al. 2004); rectum, soft-tissues (sarcomas), breast, neck, nasopharynx, base of tongue and for surface applicators (Taschereau et al. 2004). Independently of the clinical site and the complexity of the case, the algorithm produces a plan in a short time for clinical application, generally in less than 30 seconds. Moreover, these plans are equal (for simple cases), even better (for complex cases), than any plans produced manually by an experienced medical physicist. Henceforth, this inverse planning tool is used for every single HDR brachytherapy case at University of California San Francisco (UCSF), from very simple vaginal cylinders to complex interstitial implants. Multiple Targets, Each with its Own Dose Prescription The ability to define multiple targets, each with its own dose prescription, permits the definition of complex dose gradient within one tumor and/or between different tumors. Examples of targets within target would
402
Jean Pouliot et al.
Figure 4. Examples of clinical sites optimized with IPSA.
be the boost (delivery of higher dose than the prescription dose) of cancer-validated areas in prostate defined by MRS (Pouliot et al. 2004) or of the positive biopsy areas. Multiple Organs at Risk, Each with its Own Dose Prescription The inverse planning approach always benefits from additional anatomical information without increasing the complexity of the planning process. One more advantage of inverse planning is that the burden on the physicist does not increase with the number of targets and OARs. This facilitates and therefore encourages the inclusion in the planning process of OARs that were neglected before. Examples of additional OAR are the penile bulb and the neurovascular bundles in the treatment of prostate cancer. Facilitate Adjustment Based in Specific Clinical Situation Another feature of this objective function is its ability to prioritize its objectives. For example, by changing the penalty weight, the inverse planning can produce a plan with a slightly more dose heterogeneity but improved target coverage or produce a plan with improved OAR protection but reduced target coverage. Because the optimization process is efficient and fast, the physician can review multiple plans before delivering the treatment. This facilitates the adjustment of the dose distribution based on specific clinical
Advanced 3-D Planning
403
situations. Note that no manual adjustment of the source positions and dwell times is needed; a simple change in the relative priorities of the clinical objective will produce a new treatment plan that respects the new requirements. This approach is straightforward and much more intuitive and efficient than the conventional forward planning approach. Class Solution Standard treatment planning procedures for standard treatment cases make the planning process more efficient and promote consistency between plans produced for individual patients and by different planning staff. In the context of inverse treatment planning, such treatment-planning standards are known as “class solutions.” A class solution is defined as a set of dose constraints, or clinical objectives, relevant to one anatomical site that has been tuned to cover variations over a wide range of patients. The aim of a class solution is not to produce the optimal treatment plan per patient but to offer a set of dose constraints that would lead to the best outcome for a cohort of patients. The class solution can be used as a starting point for every patient, reducing dramatically the time needed to plan individual patients. The acceptance of the final treatment plan must be based on clinical information and the physician’s judgment. A class solution is not required for treatment inverse planning but recommended. The advantage of such an approach is that an acceptable plan can be produced in a reasonable time frame and ensures that all patients are treated in a very similar way, making their comparison more straightforward. The concept of class solution is commonly used in external beam inverse planned IMRT. The establishment of an IMRT class solution has been demonstrated to significantly improve planning efficiency while keeping high-quality treatment plans (Khoo et al. 2003; Xia et al. 2004). In addition, carefully developed IMRT class solutions have been found to be robust, requiring minimal fine-tuning on a patient-to-patient basis (Mott, Livsey, and Logue 2004). Class solutions need to be robust enough to produce dose distributions within suitable clinical tolerances regardless of individual patient anatomy. In inversely planned brachytherapy, such a set of dose constraints for a particular type of implant that will produce a clinically acceptable plan can also be identified. The same set of dose constraints has been used over the last 3 years at UCSF for inversely planned prostate HDR brachytherapy, and it has consistently produced very good plans. Other institutions have also experienced similar conclusions with HDR prostate brachytherapy (Lachance et al. 2002; Ménard et al. 2004). Despite the wide variety in organ volumes and shapes between patients, this class solution produced acceptable plans in all patients with excellent consistency in the dose distribution among patients. Rarely do the dose constraints need to be adjusted. This is very encouraging for the use of a class solution approach in brachytherapy.
Class Solution A set of dose constraints, or clinical objectives, relevant to one particular anatomical site but tuned to cover variations over a wide range of patients.
Clinical Implementation and Implications IPSA was included as a beta version in PLATO-BPS (Nucletron, Veenendaal, The Netherlands) in 2000 and has been clinically evaluated at UCSF1 (Hsu et al. 2004) and CHUQ2 (Lachance et al. 2002) for CT1 2
University of California, San Francisco, UCSF Comprehensive Cancer Center. Centre hospitalier universitaire de Québec, Hôtel-Dieu de Québec.
404
Jean Pouliot et al.
guided HDR brachytherapy planning. This beta version was also introduced at NIH3 in 2002 for their MRIguided prostate HDR brachytherapy program (Ménard et al. 2004). This new technology has been successfully introduced in the clinic at these three institutions where several hundred patients have been treated with inversely planned HDR brachytherapy in the past 4 years. These independent clinical trials indicated that dosimetric indices and overall procedure time were improved with the introduction of IPSA for treatment planning. In addition to these three first North American sites, a first European institution (CAV4) specializing in pulse dose rate (PDR) brachytherapy is evaluating IPSA for this special purpose (Poupon, Castelain, and Lartigau 2003). More recently IPSA was also included in SPOT-PRO (Nucletron, Veenendaal, The Netherlands) and clinically evaluated at CHUQ in 2004 for intraoperative permanent prostate implant planning. In both cases (PLATO-BPS and SPOT-PRO), once the anatomical structures and the needle positions are digitized, the physician selects the class solution reflecting the clinical objectives and calls IPSA from a special button implemented in the treatment-planning system for the purpose of the clinical evaluation. The treatment-planning system gives the digitized anatomical structure and needle positions to IPSA, which performed the optimization and gives back the optimal treatment plan to the treatment-planning system. The physician finally evaluates the quality of the plan through visual assessment of the dose distribution and the DVH generated by the treatment-planning system. It is hoped that IPSA will be officially released in these treatment-planning systems and be accessible to the community in the near future. The following sections present several clinical applications and investigations of IPSA capabilities performed during the past 4 years. These clinical examples should clarify the new avenues available with this innovative treatment-planning tool and their clinical implications.
Prostate HDR Brachytherapy Comparison Between GEO and IPSA Geometric optimization (GEO) is a popular optimization tool available in commercial planning systems for HDR brachytherapy (Edmundson 1990). GEO determines the dwell time of a particular dwell position proportional to the sum of the inverse squared distances to other dwell positions. In a geometrically optimized implant, the dwell positions near the peripheral of the implant tend to get larger dwell times. The opposite is true for dwell positions located in the center of the implant. With this approach the dwell time values (or dwell weights in this case) are relative to each other. Afterward, the dwell weight values are normalized so that the prescribed dose covers the whole tumor volume. In general, GEO improves the coverage of the implanted volume and decreases hot spots within the implant. However, this optimization tool fails to use the anatomical information because it is only based on catheter positions. Therefore, time-consuming manual adjustments are required prior and after GEO to moderate the dose delivered to the OARs while keeping an acceptable coverage of the prostate. First, the final dose distribution obtained with GEO strongly depends on the selection of the dwell positions that contribute to deliver the dose. These dwell positions, known as the “active” dwell positions, are chosen from the available dwell positions within the implant prior to the GEO computation being performed. This is traditionally performed manually, based on visual assessment, keeping the dwell positions located inside or close to the tumor volume while excluding the dwell positions in the surrounding area of the OARs. The effectiveness of this manual method strongly depends on the physicist’s experience (Giannouli et al. 2000). In addition, even a meticulous selection of the active dwell positions may not be sufficient to produce the dose distribution wanted. Because GEO does not take into account the
3 4
National Institute of Health hospital, Bethesda, MD. Centre Alexis Vautrin, Nancy, France.
Advanced 3-D Planning
405
anatomy during the computation, it is frequent that manual dwell times are required after the GEO computation to reduce the dose delivered to the OARs and ensure an acceptable coverage of the prostate. Two independent clinical studies have concluded that IPSA produces superior treatment plans to GEO by improving coverage of the target while minimizing dose to normal structures (Lachance et al. 2002; Hsu et al. 2004). These studies also demonstrated that IPSA provides consistent results from patient to patient. Protection of the Penile Bulb It has been demonstrated that the dose of radiation received by the bulb of the penis correlates with the risk of impotence after radiotherapy for prostate cancer (Fisch et al. 2001; Merrick et al. 2001). In order to investigate if the dose received by the penile bulb could be significantly reduced, this anatomical structure has been contoured for several patients and included in the inverse planning process. For each case, IPSA was called twice; with and without taking into account this new volume. On average, the V50 of the penile bulb is reduced by 14% for a small penalty of 1% on the prostate V100. Moreover, the inclusion of this OAR eliminates all cases where the prescription isodose reached the penile bulb. This shows how the explicit inclusion of an additional OAR in the dose constraints can result in a substantial reduction of dose delivered to this organ at risk without a significant change of the target dose coverage. With MRI-based planning, the neurovascular bundles could be identified and contoured for inclusion as an organ to be protected. The irradiation of the penile bulb and the neurovascular bundles play a role in the loss of potency. Because of the proximity of the neurovascular bundles to the prostate, the level of protection might not be as significant as for the bulb. Constraints may nevertheless be defined to ensure prostatic coverage while avoiding undesirable hotspots to the neurovascular bundles. MRS-defined Boost The advent of functional imaging modalities such as PET or combined MRI/MRS imaging is an important step toward better cancer-validated tumor targeting in radiation therapy. MRS provides a non-invasive method of detecting small molecular markers (historically the metabolite choline, citrate, and polyamines) within the cytosol and extracellular spaces of the prostate. It has been demonstrated that the high specificity of MRS to metabolically identify cancer can also be used to improve the ability of MRI to identify the location and extent of cancer within the prostate (Kurhanewicz, Vigneron, and Nelson 2000). A study of 53 biopsy-proven prostate cancer patients prior to radical prostatectomy and step-section pathologic examination demonstrated a significant improvement in cancer localization to a prostatic sextant (left and right—base, mid-gland, and apex) using combined MRI/MRS versus MRI alone (Scheidler et al. 1999; Wefer et al. 2000). This information can be used to dose escalate only the cancer-validated area within the prostate also known as dominant intraprostatic lesions (DIL). Pouliot conducted a feasibility study using MRS guidance on 10 patients treated with HDR brachytherapy (Pouliot et al. 2004). MRI and MRS scans were obtained on a 1.5 Tesla system to determine the location of the DIL. Eight of the ten patients showed two distinct DIL. In all cases, the DIL were located in the peripheral zone of the gland. The population of patients included prostate volumes ranging from 30 to 45 cc. The DIL volumes ranged from 0.4 cc to 6.0 cc, representing 5% to 18 % of the prostate volumes. CT or MRI scans (five patients each) with the catheters in place were performed and transferred to the treatment planning system. The transversal CT/MRI slices were aligned manually to match the corresponding anatomical structures. In addition to the prostate, urethra, bladder, and rectum, the DIL were also contoured based on MRI and MRS. No restrictions on the relative distance between the DIL and the rectum or urethra were imposed in delineating the DIL contours. Therefore, on several occasions, the DIL was very close to the rectum. For each case, IPSA was used to optimize the dose distribution and to dose escalate different levels of DIL boosts. DVHs of the target and each OAR were computed and the results
406
Jean Pouliot et al.
compared with optimized plans without DIL boost. The D1cc of both the bladder and the rectum show a small increase of 5% from no boost to the 150% boosts. The V120 of the urethra shows no increase for the 120% boost compared to no boost. However, increases of up to 50% of the urethra volume are observed for the 150% DIL boost. This indicates that DIL boosts of 120% can be safely delivered. Larger boost values up to 150% need more investigation. Besides, it could be argued that only the DIL plus a margin should receive the full-prescribed dose, allowing one to reduce the dose delivered to the rest of the prostate. This would result in a strong reduction of dose delivered to the surrounding tissues and OARs. However, better radiobiological data, perhaps gathered from functional imaging, will need to be obtained before such dose reduction could be considered. Number of Catheters A study focused on the number of catheters needed for a proper implantation of prostate using remote afterloader and GEO was recently performed (Charra-Brunaud et al. 2003). One could infer that it would be of value to minimize the number of implant catheters, both to minimize trauma to the patient and to simplify the implant procedure, while still obtaining the desired dose distribution. The extension of this study to compare the impact of the number of catheters on the dose distribution in prostate HDR brachytherapy for GEO and IPSA optimizations was recently presented (Pouliot, Lessard, and Hsu 2004). Transrectal ultrasound images of the prostate from 24 patients were transferred into the planning system. Urethra and prostate contours were digitized on each axial slice, as well as hypothetical locations of the catheters (2/3 of the catheters along the prostate contour, 1/3 around the urethra). Each patient data set was implanted with 9, 12, 15, 18, and 21 catheters (examples for 9, 15, and 21 are shown in Figure 5). For each case, two dose plans were generated: A GEO with local prescription to ensure a PTV V100 = 95%, and an IPSA with global prescription defined by a class solution. The dosimetrical indices were analyzed as a function of the number of catheters and categorized by prostate volumes. The dosimetrical indices variations relative to the number of catheters are presented in Figure 6. IPSA V100 are superior to GEO V100 for all number of catheters except for nine catheters (V100, IPSA = 92%, GEO = 95%). With GEO, an increase of the urethra V150 is observed (from 4 to 40%) when the number of catheters decreases from 21 to 9. With IPSA, the increase is only from 0 to 2%. For plans optimized with GEO, V150 gradually increases when the number of catheters decreases from 15 to 9 (p < 0.0001). The same trend is observed for small and large prostates. For IPSA plans, V150 is independent of the number of catheters. As one might expect from a dose distribution optimized on dose constraints, the constant target coverage and high dose level observed with IPSA also correspond to constant dose homogeneity. For GEO, the homogeneity index (HI) decreases from 0.69 to 0.49 when the number of catheters decreases from 21 to 9. Only a small HI decrease, from 0.69 to 0.65, is observed with IPSA. The average IPSA HI for 9 catheters is larger (better) than with GEO for 15 catheters. Therefore, IPSA can achieve dosimetrically equivalent treatment plans using a smaller number of catheters than when using GEO. The HI value is relatively constant with IPSA because the dose distribution homogeneity is controlled by the same set of dose constraints independently of catheter locations and their quantity. On the other hand, GEO is closely dependent on the catheter locations and therefore on their quantity. IPSA can also achieve a better dose distribution homogeneity with 21 catheters by changing the dose constraints. However, this is at the cost of the target coverage and OAR protection. One must clearly understand that the compromise between dose homogeneity and the other dosimetric indices is now adjustable with IPSA and therefore needs to be defined by the physician through the dose constraints. This new degree of freedom brings added flexibility for shaping the dose distribution but also poses questions about what should be the ideal shape of the dose distribution.
Advanced 3-D Planning
407
Figure 5. Simulation of a prostate implanted with various numbers of catheters, from left to right: 9, 15, and 21 catheters.
Figure 6. Variation of the dosimetric indices of the prostate (V150, HI, COIN) and the urethra (V150) with the reduction of the number of catheters.
408
Jean Pouliot et al.
For plans optimized with GEO, V150 increased when fewer catheters were used. Fifteen to twentyone catheters cover the prostate adequately without creating excess hot spots. For plans optimized with IPSA, for the same prostate coverage as GEO plans, smaller volumes of hot spots were observed, independently of the number of catheters between 9 and 21, and better urethra protection was achieved. Plans obtained with IPSA show reduced dosimetrical sensitivity variability due to changes in the number of catheters, providing additional flexibility to the physician for the implantation technique. IPSA appears to compensate for variation in catheter position tested in the study and makes brachytherapy less operator dependent.
Gynecological Cancer Interstitial Gynecological Implant The tandem-ovoid intracavitary system is an integral part of radiation therapy for brachytherapy of gynecological malignancies. The classic Fletcher-Suit intracavitary applicator system includes an intrauterine tandem and two intravaginal ovoids. Using the standard loading, this system produces a pear-shaped highdose region centered on the cervix. This brachytherapy system allows a very high dose to be delivered to the cervix while sparing adjacent bladder and bowel. The flexibility of this applicator allows it to be tailored to a variety of different patient anatomy types. However, for cases where the tumor extends beyond the range of the tandem-ovoid applicator, the insertion of additional catheters has the potential to deliver a curative dose to tumors located away from an accessible anatomical cavity (Hsu et al. 2002). The catheters are inserted around the vaginal wall, into the parametrial, paravaginal, or paraurethral regions, in addition to the vaginal tandem in an attempt to increase the dose to the region that is outside the standard pear-shaped dose distribution. This interstitial gynecological implant technique is promising but implies complex treatment planning. The number of catheters implanted and the implant geometry varies widely with the different tumor volumes and shapes. Conventional forward planning starting with geometrical optimization requires intensive adjustments to ensure tumor coverage and OAR protection. Moreover, because the implant is very different for each patient, it is impossible to develop a forward treatment-planning methodology applicable to each patient that will accelerate the planning process. Alternatively, the inverse planning approach applies the same dose constraints to each patient and produces consistent treatment plans independently of the number of catheter implanted and the implant geometry. The introduction of the inverse planning in our clinic facilitated the development of this new treatment procedure because it gives enough flexibility to the physician during the implantation and enough flexibility to the physicist to manipulate the dose distribution as wanted. Complex dose distribution based on complex implant geometry is achievable in a clinically acceptable time only with an inverse planning approach (Lessard, Hsu, and Pouliot 2002). Vaginal Boost In addition to the tumor volume and the OARs, a region around the vaginal tandem is contoured to specify to the algorithm to ignore the dose within the vaginal cylinder, as doses delivered to the nylon composing the cylinder are clinically irrelevant. The dose constraint defining the dose delivered within that volume is simply neglected using null weight penalties. In addition, it is the author’s belief that it is advantageous to maintain a high-dose region around the tandem in a fashion similar to standard tandemovoid applications. To ensure the high dose wanted at the surface of the cylinder, the minimal dose constraint at the surface is usually boosted to 120% of the prescription dose.
Advanced 3-D Planning
409
Different Choice of Weighting Factors For the first gynecological implant performed at UCSF, the dose distribution was obtained with conventional geometric optimization followed by manual adjustment of the dwell times. In parallel, an inverse planning was performed with the same anatomical information. The set of dose constraints was determined such that the inverse planning would generate a dose distribution similar to the one produced with forward planning method. Afterward, several inverse plans were obtained with different emphasis given to the PTV coverage by varying the surface and volume minimal dose weight penalties from 70 to 500 (70 ≤ mmin ≤ 500) while the other dose constraint values were kept constant. DVHs were generated for the PTV and the OARs for all plans (Lessard, Hsu, and Pouliot 2002). The improvement in PTV coverage comes at the expense of an increased dose to the OARs. Also, while the coverage approaches 100% of the PTV, each OAR receives more doses. This is the old compromise of radiotherapy, to deliver a curative dose to the tumor while preserving surrounding normal tissues. Some inverse plans show improved dose coverage for an equivalent protection to OARs compared to the forward plan. The forward plan was performed by an experienced physician and took over an hour, while a graduate student produced all the inverse plans together in about 15 minutes. This illustrates the capacity of the inverse planning method to produce better treatment plans in a short time.
Permanent Prostate Implants Forward vs. Inverse Planning Radioactive seed implantation has become a popular alternative treatment option for men with localized prostate cancer. The success of this procedure depends on the proper selection of the seed distribution followed by the accurate placement of radioactive isotopes within a known volume of prostate cancer. Planning a 3-D seed array before implantation involves choosing a radioactive seed type and an appropriate activity that will adequately encompass the prostate gland while minimizing the dose to the surrounding adjacent structures. At UCSF, more than 750 prostate patients have been implanted with radioactive seeds since 1996. Over time, the clinical experience and post-implant analysis led to the establishment of a number of dosimetrical and technical rules that are followed to determine, using forward planning, the number of seeds and their localization within the prostate to rapidly produce a good dose distribution. Rules such as minimum target coverage with the prescribed dose, maximum dose to urethra, proximity of the seeds from the rectum wall, number of needles, etc., ensure that the prostate will be well treated and that the plan will remain effective even if seed misplacement or edema occur. With the advent of inverse planning, it was possible to generate a set of dose constraints and use IPSA to automatically generate plans that fulfill the rules established before. On Figure 7a, the experienced dosimetrist forward-planned a dose distribution following the rules. The dose distribution obtained with the inverse planning for the same prostate is shown on Figure 7b. The anatomical structures visible on this plane are the prostate and the urethra (in the center). The three isodoses displayed are the 100%, 150%, and 200%. IPSA consistently reproduces equivalent implants in terms of dosimetric indices and seed placement rules as the forwardplanned implants produced by an experienced physicist. In addition, the optimization is performed in a very short time, about 15 seconds for 400,000 iterations (2.5 GHz PC), opening the door to intraoperative treatment-planning optimization. Variation of Seed Activity In the case of a forward planning approach, changing the seed activity would force the dosimetrist to start over, just as if doing the plan for the first time. However, with inverse planning, once the dose constraints are specified, one can rapidly generate a new plan using a different seed activity. Figure 8 shows isodose
410
Jean Pouliot et al.
(a)
(b)
Figure 7. Implant performed by an experienced physicist (a) and obtained with the inverse planning for the same prostate (b). The anatomical structures visible on this plane are the prostate and the urethra (in the center). The three isodoses displayed are the 100%, 150%, and 200%.
distributions for seed activities ranging from 0.37 to 0.9 mCi. The dosimetric indices are equivalent independently of the seed activities. The seed activity is directly taken into account in the objective function through the dose distribution calculation. As a result, the inverse planning distributes the seeds and determines automatically the number of needles and seeds needed with regard to the seed activity. Other studies indicate that good coverage can be obtained with seed activities ranging from 0.3 mCi to 0.7 mCi (Wuu et al. 2000; D’Souza and Meyer 2001; Sloboda, Pederson, and Halperin 2003; Sloboda et al. 2003). Furthermore, Taschereau showed that forward planning of the seed positions, even with low activity seeds, tends to be less robust to misplacement than plans generated by an inverse planning method with higher activity seeds (Taschereau, Roy, and Pouliot 2000). However, those authors also raised the question of dose homogeneity. The interplay between dose coverage, OAR protection, and dose homogeneity in the target volume has been an important question in external beam IMRT. In a number of clinical situations, the requirement of dose homogeneity can be relaxed to obtain better dose coverage and OAR protection. The same issue is now playing with the use of inverse planning in permanent implants. In a recent study (Beaulieu et al. 2004), plans were shown to be robust to misplacement and migration of seeds over a wide range of seed activity and for various seed models. With a properly tuned inverse planning algorithm able to ensure the dose coverage and OAR protection in the presence of placement errors, the choice of a preferred seed activity is open. The upper part of this range offers the opportunity to reduce significantly the number of seeds and needles, thus reducing the trauma to the patient, saving time in an operating room planning setting, and reducing the cost of the procedure. Boost of Positive Biopsy Areas The combination of an optimization algorithm with MRS information has proven to be a safe approach to boost positive biopsy areas with permanent implant (Zaider et al. 2000). We have demonstrated the feasibility of using IPSA to boost DIL with HDR prostate brachytherapy. To evaluate its ability to boost a region within the target for permanent prostate implants, small volumes were defined in addition to the prostate
Advanced 3-D Planning
411
(a)
(b)
(c)
(d)
(e)
(f)
Figure 8. Implant obtained with the inverse planning for seed activities ranging from 0.37 to 0.9 mCi (a to f). The anatomical structures visible on this plane are the prostate and the urethra (in the center). The three isodoses displayed are the 100%, 150%, and 200%. The number of seeds counted in each implant is also displayed.
(a)
(b)
(c)
Figure 9. Implant obtained with the inverse planning without dose constraints applied on these new volumes (a), with a boost of 150% with (b) and without (c) maximal dose limit. The anatomical structures visible on this plane are the prostate, the urethra (in the center), and three small volumes that simulate positive biopsy areas. The three isodose levels displayed are the 100%, 150%, and 200%.
412
Jean Pouliot et al.
and the urethra already digitized by the physician over the transversal ultrasound images. Figure 9a presents the dose distribution for the optimization without dose constraints applied on these new volumes. Figures 9b and 9c correspond to two boosts of 150% with and without maximal dose limit. The anatomical structures visible on this plane are the prostate, the urethra (in the center), and three small volumes that simulate positive biopsy areas. The three isodose levels displayed are the 100%, 150%, and 200%. One can observe that the dose distributions are adequate for all optimizations. The inclusion of the boost to the positive biopsy areas clearly forces the 150% isodose to cover the regions. Further investigations are needed to establish a safe dose level for boost of positive biopsy areas in clinical applications.
Conclusion Modern planning tools, such as the anatomy-based IPSA presented in this chapter, efficiently use the information from the 3-D imaging modalities at all stages of the treatment procedure. IPSA automatically and rapidly produces conformal dose coverage of the target volume while limiting the dose to OARs in the delivery of HDR brachytherapy or permanent implants. With this inverse planning approach, the focus is on the physician’s prescription and dose constraints instead of on the technical limitations. Consequently, the physician’s control of the treatment is improved. With clear anatomical images, functional images to locate the cancerous tissues and rapid optimization to take all this information into account within the clinical time frame, plans can be generated with dose distribution that fulfills multi-objective dose constraints, sparing organs often not considered before, and personalized for each patient, a task inherently difficult to achieve with forward planning.
References Aarts, E. H. L., and J. H. M. Korst. Simulated Annealing and Boltzmann Machine. New York: John Wiley & Sons, 1989. Beaulieu, L., L. Archambault, S. Aubin, E. Oral, R. Taschereau, and J. Pouliot. (2004). “The robustness of dose distributions to displacement and migration of 125I permanent seed implants over a wide range of seed number, activity, and designs.” Int J Radiat Oncol Biol Phys 58:1298–1308. Charra-Brunaud, C., I.-C. Hsu, V. Weinberg, and J. Pouliot. (2003). “Analysis of the interaction between number of implant catheters and dose-volume histograms in prostate high-dose-rate brachytherapy using a computer model.” Int J Radiat Oncol Biol Phys 56(2):586–591. Chen, Y., A. L. Boyer, and L. Xing. (2000). “A dose-volume histogram based optimization algorithm for ultrasound guided prostate implants.” Med Phys 27:2286–2292. Deasy, J. O. (1997). “Multiple local minima in radiotherapy optimization problems with dose-volume constraints.” Med Phys 24:1157–1161. DeJong, K. A. “Genetic Algorithms Are NOT Function Optimizers” in Foundations of Genetic Algorithms: 2. E. D. Whitley (ed.). FOGA Proceedings July 24–29, 1992, Vail, Colorado. San Mateo, CA: Morgan Kaufman, pp. 5–18, 1993. DeWitt, K. D., I.-C. Hsu, V. K. Weinberg, É. Lessard, and J. Pouliot. (2004). “3-D inverse treatment planning for the tandem and ovoid applicator in cervical cancer.” Int J Radiat Oncol Biol Phys, In press. D’Souza, W. D., and R. R. Meyer. (2001). “Dose homogeneity as a function of source activity in optimized I-125 prostate implant treatment plans.” Int J Radiat Oncol Biol Phys 51:1120–1130. Edmundson, G. K. “Geometry-based Optimization for Stepping Source Implants” in Brachytherapy HDR and LDR. A. A. Martinez, C. G. Orton, and R. F. Mould, (eds.). Columbia, MD: Nucletron, pp. 184–192, 1990. Fisch, B. M., B. Pickett, V. Weinberg, and M. Roach. (2001). “Dose of radiation received by the bulb of the penis correlates with risk of impotence after three-dimensional conformal radiotherapy for prostate cancer.” Urology 57(5):955–959. Forrest, S. (1993). “Genetic algorithms: Principles of natural selection applied to computation.” Science 261:872–878.
Advanced 3-D Planning
413
Gabor, J., O. E. Streeter, and M. A. Astrahan. (2003). “The use of linear programming in optimization of HDR implant dose distributions.” Med Phys 30(5):751–760. Geman, D., and S. Geman. (1984). “Stochastic relaxation, Gibbs distribution and Bayesian restoration images.” IEEE Trans Pattern Anal. Mac Intell (PAMI) 6:721–741. Giannouli, S., N. Milickovic, D. Baltas, M. Lahanas, N. Uzunoglu, C. Kolotas, and N. Zamboglou. (2000). “Autoactivation of source dwell positions for HDR brachytherapy treatment planning.” Med Phys 27(11):2517–2520. Hsu, I.-C., É. Lessard, V. Weinberg, and J. Pouliot. (2004). ”Comparison of inverse planning simulated annealing and geometrical optimization for prostate high-dose-rate brachytherapy.” Brachytherapy 3:147–152. Hsu, I.-C., J. Speight, J. Hai, É. Vigneault, T. Phillips, and J. Pouliot. (2002). “A comparison between tandem and ovoids and interstitial gynecologic template brachytherapy dosimetry using a hypothetical computer model.” Int J Radiat Oncol Biol Phys 52(2):538–543. Ingber, L. (1996). “Adaptive simulated annealing (ASA): Lessons learned.” Control and Cybernetics 25(1):33–54. Khoo, V. S., J. L. Bedford, S. Webb, and D. P. Dearnaley. (2003). “Class solutions for conformal external beam prostate radiotherapy.” Int J Radiat Oncol Biol Phys 55(4):1109–1120. Kurhanewicz, J., D. Vigneron, and S. J. Nelson. (2000). “Three-dimensional magnetic resonance spectroscopic imaging of brain and prostate cancer.” Neoplasia 2(1-2):166–189. Lachance, B., D. Beliveau-Nadeau, É. Lessard, M. Chretien, I.-C. Hsu, J. Pouliot, L. Beaulieu, and É. Vigneault. (2002). “Early clinical experience with anatomy-based inverse planning dose optimization for high-dose-rate boost of the prostate.” Int J Radiat Oncol Biol Phys 54:86–100. Lahanas, M., D. Baltas, and N. Zamboglou. (1999). “Anatomy-based three-dimensional dose optimization in brachytherapy using multiobjective genetic algorithms.” Med Phys 26(9):1904–1918. Lahanas, M., D. Baltas, and N. Zamboglou. (2003). “A hybrid evolutionary algorithm for multi-objective anatomybased dose optimization in high-dose-rate brachytherapy.” Phys Med Biol 48:399–415. Lessard, É. (2004). “Development and clinical introduction of an inverse planning dose optimization by simulated annealing (IPSA) for high dose rate brachytherapy.” Thesis abstract. Med Phys 31(10):2935. Lessard, É., and J. Pouliot. (2001). “Inverse planning anatomy-based dose optimization for HDR-brachytherapy of the prostate using fast simulated annealing algorithm and dedicated objective function.” Med Phys 28:773–779. Lessard, É., I.-C. Hsu, and J. Pouliot. (2002). “Inverse planning for interstitial gynecological template brachytherapy: Truly anatomy-based planning.” Int J Radiat Oncol Biol Phys 54(5): 1243–1250. Ménard, C., R. C. Susil, P. Choyke, G. S. Gustafson, W. Kammerer, H. Ning, R. W. Miller, K. L. Ullman, N. S. Crouse, S. Smith, É. Lessard, J. Pouliot, V. Wright, E. McVeigh, C. N. Coleman, and K. Camphausen. (2004). ”MRI-guided HDR prostate brachytherapy in a standard 1.5T scanner.” Int J Radiat Oncol Biol Phys 59(5):1414–1423. Merrick, G. S., K. Wallner, W. M. Butler, R. W. Galbreath, J. H. Lief, and M. L. Benson (2001). “A comparison of radiation dose to the bulb of the penis in men with and without prostate brachytherapy-induced erectile dysfunction.” Int J Radiat Oncol Biol Phys 50(3):597–604. Milickovic, N., M. Lahanas, M. Papagiannopoulou, N. Zamboglou, and D. Baltas. (2002). “Multiobjective anatomybased dose optimization for HDR-brachytherapy with constraint free deterministic algorithms.” Phys Med Biol 47:2263–2280. Mott, J. H., J. E. Livsey, and J. P. Logue. (2004). “Development of a simultaneous boost IMRT class solution for a hypofractionated prostate cancer protocol.” Br J Radiol 77:377–386. Pouliot, J., É. Lessard, and I.-C. Hsu. “Number of catheters in prostate high dose rate brachytherapy: The role of inverse planning.” ESTRO Joint Brachytherapy Meeting GEC/ESTRO-ABS-GLAC, Barcelona, Spain, May 13–15, 2004. Pouliot, J., D. Tremblay, J. Roy, and S. Filice. (1996). “Optimization of permanent 125I prostate implants using fast simulated annealing.” Int J Radiat Oncol Biol Phys 36(3):711–720. Pouliot, J., R. Taschereau, C. Coté, J. Roy, and D. Tremblay. (1999). “Dosimetric aspects of permanents radioactive implants for the treatment of prostate cancer.” Physics in Canada 55(2):61–68. Pouliot, J., Y. Kim, É. Lessard, I.-C. Hsu, D. B. Vigneron, and J. Kurhanewicz. (2004). “Inverse planning for HDR prostate brachytherapy used to boost dominant intraprostatic lesions defined by magnetic resonance spectroscopy imaging.” Int J Radiat Oncol Biol Phys 59(4):1196–1207. Poupon, L., B. Castelain, E. Lartigau. (2003). “Pulse dose rate brachytherapy: Optimization and place of imaging.” Cancer Radiather 7:136–146 (French).
414
Jean Pouliot et al.
Redpath, A. T. (2002). “Automatic determination of needle and source positions for brachytherapy of the prostate using 125Iodine rapid strand.” Radiother Oncol 64:215–227. Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson (2004). “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Also available as AAPM Report No. 84. Roy, J. N., K. E. Wallner, S. T. Chiu-Tsao, L. L. Anderson, and C. C. Ling. (1991). “CT-based optimized planning for transperineal prostate implant with customized template.” Int J Radiat Oncol Biol Phys 21:483–489. Scheidler, J., H. Hricak, D. B. Vigneron, K. K. Yu, D. L. Sokolov, L. R. Huang, C. J. Zaloudek, S. J. Nelson, P. R. Carroll, and J. Kurhanewicz. “Prostate cancer: Localization with three-dimensional proton MR spectroscopic imaging—clinicopathologic study.” Radiology 213(2):473–480. Sloboda, R. S. (1992). “Optimization of brachytherapy dose distribution by simulated annealing.” Med Phys 19:955–964. Sloboda, R., J. E. Pedersen, and R. M. Halperin. (2003). “Is there a preferred strength for regularly spaced 125I seeds in inverse-planned prostate implants?” Int J Radiat Oncol Biol Phys 55:234–244. Sloboda, R., J. E. Pedersen, J. Hanson, and R. M. Halperin. (2003). “Dosimetric consequences of increased seed strength for I-125 prostate implants.” Radiother Oncol 68:295–297. Speight, J., K. Shinohara, B. Pickett, V. Weinberg, I.-C. Hsu, and M. Roach III. (2000). “Prostate volume change after radioactive seed implantation: Possible benefit of improved dose volume histogram with perioperative steroid.” Int J Radiat Oncol Biol Phys 48(5):1461–1467. Taschereau, R., J. Roy, and J. Pouliot. (1999). “Monte Carlo simulation of prostate implants to improve dosimetry and compare planning methods.” Med Phys 26(9):1952–1959. Taschereau, R., J. Roy, J. Pouliot, and D. Tremblay. (2000). “Seed misplacement and stabilizing needles in transperineal permanent prostate implants.” Radiother Oncol 55:59–63. Taschereau, R., P. Stauffer, I.-C. Hsu, J. L. Schlorff, A. J. Milligan, and J. Pouliot. (2004). “Radiation dosimetry of a conformal heat-brachytherapy applicator.” Technol Cancer Res Treat 3(4):347–358. van Laarhoven, P. J. M., and E. H. L. Aarts. Simulated Annealing: Theory and Applications. Dordrecht, The Netherlands: D. Reidel Publishing, 1987. Wefer, A. E., H. Hricak, D. B. Vigneron, F. V. Coakley, Y. Lu, J. Wefer, U. Mueller-Lisse, P. R. Carroll, and J. Kurhanewicz. (2000). “Sextant localization of prostate cancer: Comparison of sextant biopsy, magnetic resonance imaging and magnetic resonance spectroscopic imaging with step section histology.” J Urol 164:400–404. Wuu, C. S., R. D. Ennis, P. B. Schiff, E. K. Lee, and M. Zaider. (2000). “Dosimetric and volumetric criteria for selecting a source activity and a source type (125I or 103Pd) in the presence of irregular seed placement in permanent prostate implants.” Int J Radiat Oncol Biol Phys 47:815–820. Xia, P., N. Lee, Y.-M. Liu, I. Poon, V. Weinberg, E. Shin, J. M. Quivey, and L. J. Verhey. (2004). “A study of planning dose constraints for treatment of nasopharyngeal carcinoma using a commercial inverse treatment planning system.” Int J Radiat Oncol Biol Phys 59(3):886–896. Yamada, Y., L. Potters, M. Zaider, G. Cohen, E. Venkatraman, and M. Zelefsky. (2003). “Impact of intraoperative edema during transperineal permanent prostate brachytherapy on computer-optimized and preimplant planning techniques.” Am J Clin Oncol 26(5):130–135. Yang, G., L. E. Reinstein, S. Pai, Z. Xu, and D. L. Carroll. (1998). “A new genetic algorithm technique in optimization of permanent 125I prostate implants.” Med Phys 25(12):2308–2315. Yu, Y., and M. C. Schell. (1996). “A genetic algorithm for the optimization of prostate implants.” Med Phys 23:2085–2091. Yu, Y., J. B. Zhang, G. Cheng, M. C. Schell, and P. Okunieff. (2000). “Multi-objective optimization in radiotherapy: Applications to stereotactic radiosurgery and prostate brachytherapy” Artif Intell Med 19:39–51. Zaider, M., M. J. Zelefsky, E. K. Lee, K. L. Zakian, H. I. Amols, J. Dyke, G. Cohen, Y. C. Hu, A. K. Endi, C.-S. Chui, and J. A. Koutcher. (2000). “Treatment planning for prostate implants using magnetic-resonance spectroscopy imaging.” Int J Radiat Oncol Biol Phys 47(4):1085–1096.
Chapter 22
Optimization in Brachytherapy Gary A. Ezzell, Ph.D. Mayo Clinic Scottsdale Scottsdale, Arizona Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415 General Concepts of Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415 Classes of Optimization Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 425 Concepts Underlying Two Stochastic Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426 Simulated Annealing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 426 Genetic Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427 Specific Brachytherapy Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428 Permanent Prostate Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428 High Dose Rate Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 431 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
Introduction Mathematical optimization is a very wide field with a huge literature. A number of different techniques have been applied to brachytherapy, and to describe them all is beyond both my competence and time allotment. So, my plan here is to first discuss general concepts and principles of optimization, and then show examples of how some have been applied to brachytherapy. The discussion will not be particularly mathematical, but it will introduce some of the terminology used in papers related to optimizing radiation treatments. My goal is primarily to explain the most important concepts so that users of brachytherapy optimization software will find it less mysterious. Secondarily, readers with a deeper interest can use these concepts and terms to make the literature more accessible.
General Concepts of Optimization First, let us start with some specific applications to set the context. Consider prostate implants with permanent seeds. A common optimization problem is to decide which seed locations are to be loaded in order to meet some objectives related to target coverage, dose uniformity, and rectal/urethral sparing. Or, consider high dose rate (HDR) brachytherapy with variable dwell times. One might want to decide which dwell positions and times to use for a two-catheter endobrochial implant in order to deliver a sufficiently uniform dose at 1 cm from each of the catheters. Generalizing from these examples, the problem is always to design a distribution of source terms such that the resultant dose distribution satisfies certain constraints and meets certain objectives as well as possible. Let us discuss this in more detail and define some terms that are commonly used. In brachytherapy, the free variables are those elements of the problem that the planner can control. These may be source locations, source strengths, and/or dwell times. For low dose rate (LDR) prostate implants, the seed locations, as dictated by the template, can be selected, while the source strength may be fixed, as is the time, since the implant is permanent. For HDR treatment with implanted applicators, as for bronchus, breast, or cervix, the dwell locations within the applicator may be chosen as well as the dwell times, but the source strength is fixed. For some stereotactic brain implants, the location of each catheter may be optimized before the procedure. In this case the catheter locations are continuously variable, while the seed strengths may be chosen from an inventory. Each particular application has its own range of variables.
415
416
Gary A. Ezzell
Every problem also has its constraints. These may be hard constraints, meaning that they cannot be violated. Or they may be soft constraints, meaning that they may be violated but at some loss of plan quality. Constraints may be physical (e.g., dwell times cannot be negative) or clinical (e.g., “at least 98% of the target must be covered by the prescription dose”). Any solution that satisfies all of the constraints is feasible. Sometimes, finding a feasible solution is sufficient to meet the clinical need. More commonly, the goal is not just to satisfy the constraints, but also to find a solution that optimizes some objective. For example, “the dose to the surface of the prostate planning target volume (PTV) is to match the prescription dose as closely as possible.” This is generally stated as a minimization problem: minimize the variance of the doses Di at points i on the PTV surface from the prescription dose Dp. Minimize f = ∑( Di – Dp)2
(1)
The optimization problem is to find the feasible solution that is optimal, that produces the lowest value of the objective function, f. Frequently, there is more than one clinical objective, and typically they compete with each other: “minimize the dose variation on the surface of the PTV” and “minimize the dose to the adjacent rectum.” Multiple objectives cannot, in general, be simultaneously optimized, but must in some fashion be balanced against each other. Although there are techniques for multi-objective optimization, most commonly the different objectives are combined into a single mathematical function, or objective function, that is to be minimized (or maximized). The simplest way to combine them is to sum the individual terms, each multiplied by a weighting factor, or importance factor. For example, the previous two objectives could be combined into a single objective function: Minimize f = w1∑(Di – Dp)2 + w2∑(Dj – Lr , if >0, 0 otherwise) .
(2)
Here, the first sum is over the PTV surface points i and the second is over the rectal points j. (Note that the first term is squared, to penalize over- and underdosage of the target, while the second penalizes doses only if they exceed the rectal limit Lr .) Soft constraints and objectives are conceptually very similar and can be handled identically by incorporating appropriate terms in the objective function. A penalty based on the number of needles, N, might be added: Minimize f = w1∑(Di – Dp)2 + w2∑(Dj – Lr , if >0, 0 otherwise) + w3N .
(3)
As is perhaps becoming clear, the objective function is a mathematical model of the clinical goals. Whether or not the optimization process produces useful results depends, to a large extent, on how well that model captures the essence of the clinical thinking. Since no model is perfect, and no two clinicians think alike, the optimization system needs to allow the planner to influence the outcome of the optimization. In equation (3), the parameters that the planner could vary include the relative values of the weight factors, wi , and the dose objectives, Dp and Lr . The reader might object that the dose objectives are set by the physician, and so are not actually available as planning parameters. Nevertheless, it is useful to recognize that the dose objectives that the physician really wants may not be the best values to use as input to the optimization software. There are:
22–Optimization in Brachytherapy
417
(a) the dose objectives you actually want to achieve, (b) the dose parameters you give to the optimizer, and (c) the dose results you eventually accept. Ideally, these would all be the same. In some implementations, however, the user may not have tools like relative weights to adjust. In that case, the dose parameters become the tools used to steer the results in the desired direction. In order to illustrate these and some other concepts, let us look at an artificial problem, as shown in Figure 1. There are points 0–6 on the inside and outside of a target region, two sources, A and B, within the target, and a sensitive region above the target with a dose point R. To simplify the analysis, the “dose” is assumed to depend only on the source strength and the inverse square of the distance, so that a source strength of one unit produces one unit of dose at a unit distance. We have these simple relationships for the dose to the various points: D0 = A + B D1 = D5 = A + B/5 D2 = D4 = A/5 + B
(4)
D3 = A/9 + B D6 = A + B/9 DR = A/2 + B/2 . Let us take as our first dose objective that the target is to receive at least 5 units of dose. Our goal is to choose values of A and B that will satisfy that objective. It is clear from the geometry of the problem that point 0 will have the highest dose of the seven dose points specified and points 3 and 6 will have the least. If we plot on the [A,B] plane the equations defining these limiting doses, we can identify the region of solution space that meets this simple objective (Figure 2). Any combination of A and B in the shaded region will produce at least 5 units of dose in the target points 0–6.
Figure 1. Simple brachytherapy problem.
418
Gary A. Ezzell
Let us also assume that the dose to point R is not to exceed 5 units. Figure 3 shows that constraint plot. Any combination of A and B in the shaded region will produce no more than 5 units of dose to point R. Combining these target and sensitive region constraints identifies that region of solution space that would satisfy both types of objectives (Figure 4). The cross-hatched region shows the solutions that are feasible, giving at least 5 units of dose to all the target points and no more than 5 units to the sensitive structure point. (The scale of the axes has been changed to enlarge the feasible region on the plot.)
Figure 2. Solution space plot for the target dose constraints. The shaded region satisfies the constraints requiring all the target doses to be 5 units or more.
Figure 3. Solution space plot for sensitive region constraint. The shaded region satisfies the constraint that the dose to R be 5 units or less.
22–Optimization in Brachytherapy
419
Let us now make this an optimization problem and decide to minimize the dose to point R while still satisfying the target dose constraints. Graphically, reducing the acceptable value of DR shifts that line nearer the origin, leading to an optimal solution of A=B=4.5 units (Figure 5). It is possible, of course, to set up a problem with no feasible solutions. We cannot achieve a dose to point R less than 4.5 if we insist on the target points being at least 5. This simple example illustrates some other generalities. 1. Feasible solutions to sets of linear constraints are bounded by a convex multidimensional polyhedron. (This simple example is two-dimensional.)
Figure 4. Solution space plot for all the constraints showing the region of feasible solutions, those that give at least 5 units of dose to all the target points and no more than 5 units to the sensitive structure point.
Figure 5. Reducing the acceptable level for DR leads to an optimal solution at a vertex.
420
Gary A. Ezzell 2. Inequalities define regions of the space. Equalities define surfaces of lower dimensionality. (For example, in Figure 3, DR<5 defines a section of a plane, while DR=5 defines a line.) 3. When an optimum solution is sought by maximizing or minimizing a linear function of the variables, the optimal solution will lie on a vertex.
Knowing that the solution space is convex is important. Convexity here means that you can move from one feasible solution to another along a straight vector in the space without ever leaving the feasible region. This property is used by many efficient search algorithms. Let us now modify the problem to make it both more difficult and more interesting. Point R has been moved inward to 0.8 units from the origin (Figure 6). Now, let us set up two competing dose objectives, similar to that in equation (2). The first objective, T, relates to the average of the dose deviations on the target surface points from the desired value of 5.
T =
6
∑( D
i
i =1
− 5) / 6 . 2
(5)
The doses Di are as given in equations (4) above. The second, S, relates to the dose to the sensitive region, DR. S = DR = A/1.64 + B/1.64 .
(6)
This is similar to that from equation set (4) but altered because of the change in the location of point R. Now, let us combine these two functions into one objective function: F = (w1 T + w2 S)/(w1 + w2) ,
(7)
where w1 and w2 are two weighting factors. Note that dividing by the sum of the weights is a technique that ensures that the function depends on the relative weights, not their absolute values. Figure 7 shows how function T varies with source strengths A and B. There is a minimum at A = B ≈ 4.25. Figure 8 shows how function S varies with source strengths A and B: a simple monotonic decrease with decreasing values of both variables. Figures 9 through 11 show how the combined objective function F varies with A and B for three different combinations of the weighting factors. In Figure 9, w1 = 100, w2 = 1. In Figure 10, w1 = 5, w2 = 1. In Figure 11, w1 = 3, w2 = 1. The shape of function F changes as the relative weights change, and the location of the minimum also shifts. Since the symmetry of the problem ensures that the dose variation is minimized with A=B, we can easily plot the shift in optimal source strength as a function of relative weight, as seen in Figure 12. These shifts are not large for this artificial problem, but they serve to illustrate the point that changing the objective function by varying weighting factors alters the shape of the function and so changes the optimization result. A more important point can be illustrated by plotting the two competing objectives against each other for each of the optimal solutions and for some non-optimal solutions, as shown in Figure 13. Here, each point represents a solution obtained with a different combination of the weighting factors. The two compet-
22–Optimization in Brachytherapy
Figure 6. Revised brachytherapy problem with point R moved to 0.8 cm from the origin.
Figure 7. Target deviation function T.
Figure 8. Sensitive region function S.
421
422
Gary A. Ezzell
Figure 9. Optimization function F with w1=100, w2=1.
Figure 10. Optimization function F with w1=5, w2=1.
Figure 11. Optimization function F with w1=3, w2=1.
22–Optimization in Brachytherapy
423
Figure 12. Optimal source strength as a function of relative weight factors.
Figure 13. Illustration of the Pareto front for the sample brachytherapy problem. For a given dose to the organ, R, there is a minimum variation of dose to the target that cannot be improved.
ing objectives are the structure dose, function S, and the target uniformity, function T. Note that the nonoptimal solutions are all to the right of the optimal ones. Physically, this indicates that it is possible to have the same structure dose at different target uniformity values (by having unequal source strengths.) But, for a given structure dose, there is a target uniformity value that cannot be improved upon (which occurs when the sources have equal strength.) The line connecting the optimal solutions, called the Pareto front or Pareto limit, shows where the best possible trade-offs occur as the relative importance of the two objectives is varied. In the terminology of optimization theory, the non-optimal solutions are dominated by the optimal, or nondominated ones. To summarize: when there are competing objectives, an optimal solution occurs when no objective can be improved without worsening another one. This concept of a Pareto front has real practical consequences for planners of radiation treatments. If a physician wants to cover a prostate while sparing the rectum, the planner can provide a series of plans, with, for example, 98%, 95%, 93%, and 90% target coverage, each with the best obtainable rectal dose. Such a series of results represents the Pareto front for the problem at hand and shows the best available choices. A planner faced with a clinical optimization problem needs to:
424
Gary A. Ezzell (a) decide which objective is most important, (b) decide the limit, or range of limits, to which that objective may be pushed, and then (c) optimize the other objective by pushing the primary objective to this limit.
When there are more than two objectives, the Pareto front becomes multidimensional, which complicates both the search and the presentation of the results. This simple example has, for any combination of weights, a single minimum that is optimal solution. In general, such is the case for any objective function that is a linear combination of linear or quadratic terms. It is possible to create functions that have multiple minima; combinations of dose/volume constraints can have that characteristic (Deasy 1997). Figure 14 demonstrates the idea with an arbitrary objective function with two minima. The solution represented by the diamond sits at a local minimum and the circle at the global minimum. Some optimization techniques will quickly and efficiently find the local nearest local minimum and not move beyond that to a global minimum. Others have the ability to escape local traps. Therefore, knowing whether or not the problem is such that local minima may exist is useful for choosing which optimization technique to bring to bear. This is one of the points that will be discussed in the sections that follow regarding individual optimization techniques. Before moving on to specific optimization techniques, there are two more general issues to discuss. One might be called the “evaluation problem.” In practice, users of brachytherapy optimization software have to work with the tools provided in the software. The user cannot, in general, design the optimization function directly; he can only steer it by adjusting parameters such as relative weights. Once the plan is “optimized,” meaning that the software has provided a solution to the problem as stated, the planner then needs to evaluate the quality of the plan. This might be done by looking at isodose curves and dose volume histograms (DVHs). Alternatively, one might use quality indices such as “conformity index” and “homogeneity index” or biological functions such as equivalent uniform dose or tumor control probability. Ideally, the objective function would encode the evaluation tool preferred by the user. If you are going to judge a plan using DVHs, then why not optimize the DVHs directly? Unfortunately, the optimization problem is not usually coded in a way that mirrors the final evaluation process. Papers about optimization algorithms usually describe the method and its implementation and then discuss its merits in terms of evaluators such as these. The practical upshot is that the planner has two conceptual tasks. He should (1) carefully think through how to quantitatively evaluate the results before (2) using the mechanics of the optimizer software to control those results. He must not assume that what the computer declares is mathematically optimal according to its algorithm is actually clinically optimal. The second general issue relates to the robustness of the optimized solution in the face of uncertainty, either in the input data or in the execution of the plan. A good example is permanent prostate seed implants.
Figure 14. Arbitrary objective function illustrating a local (diamond) and global (circle) minimum.
22–Optimization in Brachytherapy
425
There is always uncertainty in where the seeds will actually be implanted, and so the final plan quality should not be overly dependent on following the preplan exactly. A plan with more seeds of somewhat lower activity may be more stable in the face of uncertainty than one with fewer seeds of higher activity. Analyses of optimization algorithms should include consideration of the effects of perturbations on the results.
Classes of Optimization Techniques Many approaches to optimization use iterative numerical techniques and share a common overall structure (Figure 15). Most of the important differences are in how the techniques create new solutions to evaluate. One way of classifying optimization methods is by how they choose new solutions and so move through solution space toward an answer. “Deterministic” methods (e.g., steepest descent, conjugate gradient, Nelder-Mead simplex) find their way from the starting point to a minimum by moving “downhill”; that is, each iteration searches for a solution that reduces the value of the objective function. Frequently, these methods use the mathematical characteristics of the problem to guide the search very efficiently. The gradient of solution space at the location of the current solution is determined, either numerically or through differentiation of the objective function, and the next solution is obtained by moving “downhill.” Such techniques need to solve two problems at each iteration: which direction to move and how far to move. For objective functions that are strictly linear or quadratic, it is possible to calculate both the direction and the step size so as to arrive at the nearest minimum very efficiently. [For a good description of the steepest descent and conjugate gradient methods, see Shewchuk (1994).] Note that these methods move rapidly downhill but cannot escape from a local minimum (Figure 16). They are “deterministic” in the sense that they always arrive at the same solution if restarted from the same beginning point. If it is known a priori that a single minimum exists, then there is no concern. In the absence of such knowledge, then restarting the optimization multiple times from a variety of initial conditions may be fruitful. “Stochastic” methods introduce an element of randomness into the search process. Each new prospective solution is created by altering the present solution according to an algorithm that includes some random choices. For example, in a prostate permanent implant problem, potential seed locations may be turned on or off randomly. Simulated annealing and genetic algorithms (discussed in more detail later) are examples that have been used in brachytherapy. Such methods can be applied to objective functions of any mathematical form. Since they need not move strictly downhill, they have the potential of escaping local minima. Conversely, because of their random nature, they typically cannot guarantee that the best solution found is at a true minimum, either local or global. However, in radiation therapy problems, it is seldom necessary to know that the proposed solution is formally optimal. They are less efficient than deterministic methods since they spend a good fraction of time creating and evaluating disadvantageous solutions. Note that stochastic methods can also move strictly downhill if solutions are rejected unless they improve the objective function.
Figure 15. Flow chart for a generalized, iterative optimization algorithm.
426
Gary A. Ezzell
Figure 16. Downhill search methods move from a starting point (S1 or S2) to the nearest minimum.
“Deductive” methods (e.g., branch and bound) test solutions in a systematic, logical sequence that is designed to eliminate subsets of potential solutions from consideration, and are thus faster than exhaustive search. Like stochastic methods, they can be applied to problems with a complicated structure. Depending on the implementation, they can either always find the global optimum or a solution that is adequate. “Heuristic” or “phenomenological” methods use related relationships to arrive at an adequate solution to the original problem. Edmundson’s “geometric optimization” is an example (Edmundson et al. 1993); it bases HDR dwell times on the relative locations of the dwell positions, not on dose calculations to points of interest. These related relationships are sometimes called “adjoint functions.” Such methods can be very fast, and they produce results that may be useful without being formally optimal. It is possible to combine these methods, of course. There are implementations in radiation therapy planning that use simulated annealing to initially move through solution space to find a profitable region, and then switch to downhill search to quickly find the nearby minimum. Implementations that use adjoint functions as surrogates for dose objectives may operate on those functions with any of these other optimization methods.
Concepts Underlying Two Stochastic Methods Two of the stochastic methods used for optimization mimic processes that occur in nature. As such, their underlying concepts are interesting and deserve additional explanation.
Simulated Annealing Simulated annealing is a stochastic optimization technique that searches for an optimal solution for an objective function in a fashion analogous to a crystal seeking its lowest energy state as it slowly cools from an initially high temperature (Metropolis et al. 1953). The following description is adapted from Ezzell and Luthmann (1995). Simulated annealing proceeds iteratively in the following sequence: 1. An initial solution (set of dwell times and/or source positions) is chosen and evaluated using the objective function. 2. Another solution is constructed from the current one by varying the free variables in a random direction and by an amount that is dependent on a parameter analogous to the temperature, initially large. If the new solution is better than the current one, it replaces the current. If it is worse, then
22–Optimization in Brachytherapy
427
it might still be accepted. The probability of acceptance depends on the temperature parameter and the magnitude of the deterioration. In the original papers, the Boltzmann distribution was used to determine the probability, just as in nature when a system may move to an energetically unfavorable state as a result of random thermal motions. If accepting the new solution would change the objective function F by ∆F, then the probability of accepting the solution is:
p = e-∆F/kT , where T is the temperature parameter and k is a scaling factor. Note that if the new solution is better, then ∆F is negative, –∆F is positive, and the probability is greater than unity (i.e., acceptance is assured.) If the new solution is worse, then the acceptance probability is less than unity. Large temperatures make accepting disadvantageous solutions more likely. 3. The process iterates with the temperature parameter being reduced, thus shortening the steps and reducing the likelihood of uphill moves. By reducing the temperature sufficiently slowly, the system is allowed find the region of solution space that contains the global optimum and converge to it (Figure 17). Geman and Geman (1984) proved that with an appropriate (but logarithmic and slow) cooling scheme, the system will always find the global optimum. There are fast simulated annealing algorithms that use quicker cooling rates and different probability distributions and converge more rapidly. They cannot guarantee finding the global optimum, however. One of the earliest applications of simulated annealing to brachytherapy planning was by Sloboda (1992). The VariSeed™ planning system (Varian Medical Systems, Palo Alto, CA) uses an implementation of simulated annealing for permanent seed implants.
Genetic Algorithms Genetic algorithms were originally developed as models of machine learning (Holland 1975) and were soon applied to optimization problems. They mimic the processes of natural evolution by operating on populations of potential solutions. Genetic algorithms proceed by:
Figure 17. Simulated annealing allows solutions to sometimes move “uphill.” In this illustration, the starting point is S. The third move, while the temperature is high, is allowed to be large and uphill, thus the solution escapes the local minimum.
428
Gary A. Ezzell 1. Constructing a population of individual solutions, each being a string of numbers encoding the values of the free variables. Typically, each solution is a bit string of concatenated binary numbers. 2. Evaluating each individual using the objective function, and then assigning a fitness value to each. 3. Producing a new population, or generation, from the old using a combination of operators, such as: (a) proportional replication, which selects a string for representation in the new generation with a probability dependent on its fitness, (b) crossover, in which pairs of selected strings “mate” and produce offspring that contain parts of each parent, and (c) mutation, that introduces variation by altering randomly selected bits in some of the strings.
With appropriate choices for the probabilities for selection, crossover, and mutation, the population can explore solution space and identify individual solutions that have excellent fitness, that is, that are near an optimal solution (Figure 18). One of the earliest applications of genetic algorithms to brachytherapy planning was by Yu and Schell (1996).
Specific Brachytherapy Applications Let us now turn to some of the applications of these techniques to specific brachytherapy problems. This is by no means an exhaustive review, but should serve to illustrate some interesting points.
Permanent Prostate Implants Prostate brachytherapy with permanently implanted seeds has been a fertile field for optimization researchers. The problem is to decide which seed locations to load in order to satisfy dose criteria for the prostate, urethra, and rectum. Obtaining solutions quickly enough to permit intraoperative planning has been another goal of some recent studies. Yu and Schell (1996) described a genetic algorithm that was later developed into a clinical tool, Prostate Implant Planning Engine for Radiotherapy (PIPER) (Yu et al. 1999), and used for intraoperative planning (Messing et al. 1999). One of the interesting aspects of this work is that it used truly multiobjective ranking of plans instead of combining the various objectives into a single objective function using relative weights. The objectives related to: (a) total source strength; (b) conformality of dose to the target periphery; (c) maximum critical structure dose to urethra, rectum, and bladder; (d) sensitivity of these dose objectives to random variation in seed placement; and (e) number of needles used. In order to rank the individual plans in the population, they used a measure of the “distance” of any plan from a hypothetical ideal plan and a hypothetical anti-ideal plan. If there are M plans to be ranked and N objectives, then each plan is characterized by a vector with N elements (O1, O2, O3, …, ON), where Oj is the value of objective j for that plan. The hypothetical ideal plan would have the vector composed of the best Oi to be found in all M plans (even though they could not all be simultaneously feasible.) Conversely, the hypothetical non-ideal plan would have the vector composed of the worst Oi to be found
22–Optimization in Brachytherapy
429
(a)
(b) Figure 18. Genetic algorithms begin (panel A) with a population of solutions. The better solutions are allowed to reproduce more frequently, exchanging their characteristics with others and also being subject to some random changes. Subsequent populations begin to cluster around minima (panel B).
in all M plans. Using the notation for the distance measure called the Lp metric, the distance of plan i from the ideal plan (designated with +) and non-ideal plan (designated with –) is:
N p L ( i ) = ∑ g j × x j ( i ) − x +j j =1 + p
1/ p
1/ p
N + p and L ( i ) = ∑ g j × x j ( i ) − x j , j =1 + p
(8)
where gj is the metric weight for objective j and p is the metric exponent that can run from 1, 2, …, ∞. The relative importance of the different objectives can be influenced by choosing the gj. The ranking for plan i is then Rp ( i ) = L−p ( i ) / , L−p ( i ) + L+p ( i ) ,
(9)
430
Gary A. Ezzell
and represents the relative distance of this plan from the anti-ideal. The ranking is numerically larger for preferred treatment plans. The basic scheme here was to let the gj vary and, for each combination of weights, find a solution that resides on or near the Pareto frontier for the problem, meaning that no objective could be improved without sacrificing at least one other. From this set of nondominated solutions, the ones having the highest ranking were presented for clinical review. Another group, at Memorial Sloan-Kettering Cancer Center, reported on a genetic algorithm for intraoperative prostate implant planning (Yang et al. 1998), and a subsequent report (Zelefsky et al. 2003) detailed the clinical applications of its use in 248 patients. Compared to the results for patients planned earlier with manual techniques, “…multivariate analysis demonstrated that the optimization technique was an independent predictor of improved target coverage, reduced urethral dose, and more rapid resolution of urinary-related symptoms.” Further, “…the improved dosimetric conformity with the optimization technique did not compromise the biochemical outcome, as the 4-year actuarial prostate-specific antigen relapse-free survival rate for this group was 97%.” Lee et al. (1999) and D’Souza et al. (2001) described approaches to planning prostate seed implants based on mixed-integer programming. In this general technique, some or all of the variables are constrained to be integers. In this specific case, the presence or absence of a seed at a particular location is a binary 0/1. Doses to points of interest are, of course, continuous functions. In mixed-integer programming, the objective function and constraints must be linear combinations of the integer and continuous variables. One method used to solve mixed-integer problems is the branch and bound algorithm, and both authors investigated implementations of that. Branch and bound algorithms iteratively partition the set of feasible solutions to form subproblems (i.e., branches are created). Each subproblem is solved by allowing all the variables to be continuous. If the result for that branch is inferior to that from another subproblem having feasible integer variables, then the first branch can be eliminated from consideration (i.e., bounds are set). D’Souza et al. (2001) based their objective function and linear constraints on factors associated with: (a) volume of target receiving the prescribed dose, (b) volume of critical organs exceeding the critical dose, (c) peak dose to critical organs, and (d) numbers of seeds and needles. They found that their “optimized approaches do not favour the peripheral loading pattern that is usually advocated in the manual approach. A sensitivity analysis that was performed indicates that, relative to manual plans, the quality of the plans obtained from the optimized approaches are [sic] less susceptible to random uncertainties in seed placement during the treatment delivery process.” The branch and bound technique is relatively slow and took approximately 50 minutes to find a solution (although it was still faster than some stochastic methods using the same computer speeds). A subsequent paper from that group (Yoo et al. 2003) compared it to an algorithm that used related adjoint functions and a “greedy heuristic” that produced clinically acceptable plans in less than 2 seconds. They defined a “region of interest (ROI) adjoint function” as the average dose in the ROI for a unit-strength source at a given location. The greedy heuristic used the ratio of target and ROI adjoint functions to rank seed positions according to their ability to irradiate the target ROI while sparing critical structure ROI. The heuristic: (a) selected the highest ranking seed, (b) created an exclusion zone around it, (c) selected as the next seed the highest-ranking seed outside the exclusion zone, and
22–Optimization in Brachytherapy
431
(d) iterated with a method of dynamically changing the size of the exclusion zone according to the number of seeds. Thus, this is an approximate method that spaces the seeds throughout the volume according to those that are most effective at irradiating target while sparing structures. They evaluated the results of the two algorithms using DVHs, dose non-uniformity ratios (Saw et al. 1993), conformation number (van’t Riet et al. 1997), and urethra and rectal doses above certain thresholds. The two methods produced plans that were very similar, but the heuristic method was 1500 times faster and so more applicable to intraoperative planning.
HDR Brachytherapy Optimization in HDR brachytherapy was described at the last AAPM summer school on brachytherapy (Ezzell and Luthmann 1995). That discussion highlighted two basic methods, one that applied when the desired doses to particular points are specified and another that used a heuristic method that does not require dose specification. The first method minimized the variance of the difference between desired and calculated doses using a numerical technique called singular value decomposition (SVD). Since SVD does not force solutions to be non-negative, the implementation by van der Laarse and De Boer (1990) used an objective function that combined a term for the dose variance with another based on the difference in dwell times between successive dwell positions. For an implant with N dose points and M dwell positions, the objective function, F, is: N
(
F = ∑ D p , i − Dc , i i =1
)
2
M
(
+ w∑ t j − t j +1 j =1
)
2
(10)
where Dp,i and Dc,i are the prescribed and calculated doses to point i, and t is the dwell time at dwell position j. The relative importance dwell time variation term is controlled by the weight parameter w. By increasing the weight of the penalty for large changes in neighboring times, the number and magnitude of negative dwell times could be reduced or eliminated. The second method, called “geometric optimization,” was a heuristic developed by Edmundson et al. (1993) that made the dwell time at each position inversely proportional to the sum of the inverse square of the distances from that point to all the others. The idea is to achieve dose uniformity in an implanted volume, as much as possible, by requiring that the dose contribution to every dwell position from the other sources be the same for all sources. Since, to first order, dose follows the inverse square law, the calculation depends only on the geometry of the implant. This method was applied to an intraoperative HDR prostate program (Edmundson et al. 1995; Martinez et al. 1995, 2001). The Plato treatment planning system (Nucletron, Columbia, MD) for HDR includes implementations of these methods described by van der Laarse and Edmundson. There are a number of optimization alternatives in that system, and they were discussed in some detail in Ezzell and Luthmann (1995). Some work best for implants designed to treat a volume (e.g., prostate or breast implants) and others for implants designed to treat tissue at a particular distance from each catheter (e.g., endobronchial implants). The method called “polynomial optimization on volume” is an example of the former. It uses the geometric algorithm to apportion total dwell times for each catheter or needle and then uses a parameterized form of the SVD algorithm to distribute the dwell times among the individual dwell positions in order to best satisfy the dose point constraints. The BrachyVision™ treatment planning system (Varian Medical Systems, Palo Alto, CA) includes an implementation of the Nelder-Mead simplex algorithm (Press et al. 1988). This is a relatively simple down-
432
Gary A. Ezzell
hill search algorithm that does not require any gradient computations. It works well for situations with a limited number of variables, and clinical HDR implants fall within that regime. More recently, Lessard and Pouliot (2001) reported an interesting application of fast simulated annealing to prostate implants with HDR brachytherapy. Here, the free variables were the dwell times for the HDR source as it moved through catheters that had been implanted. They did not optimize the locations of the catheters. Their objective function summed penalty factors for minimum target dose, dose uniformity within the target, urethra dose, and dose to tissues outside the target. The penalty factors increased linearly for doses outside the prescribed limits. They did not use relative weights to control the outcome; instead, they modified the limits and slope of the penalty terms and demonstrated the ability thereby to trade off dose uniformity and urethral dose. They also checked that it was necessary to use simulated annealing to escape from local minima by running the algorithm with the temperature parameter set to 0. This prevented the algorithm from ever accepting changes that increased the objective function value and so converted it to a strict downhill search. Those results were inferior to the simulated annealing results, thus demonstrating the existence of local minima. A group of investigators from Germany and Greece have published a range of papers in the past few years looking at how to accomplish multi-objective optimization of HDR implants with gradient-based optimizers, fast simulated annealing, and genetic algorithms (Milickovic et al. 2002; Lahanas et al. 2003a,b). A focus of this work has been to present efficiently to the planner a range of solutions that lie on the Pareto front for the implant being considered. Remember that whenever there are multiple objectives, there are multiple “optimal” solutions. Changing the relative importance of the objectives shifts the “best” solution. Using both DVH and dose variance objectives, this group developed a list of candidate solutions, each “best” according to a particular combination of importance factors. The DVHs and other evaluation parameters, such as the conformal index (COIN) (Baltas et al. 1998), for this set of plans are presented to the planner for review and ultimate selection.
Summary Optimization in brachytherapy always begins with a numerical model of the clinical problem. There is the model of the implant itself and the biological structures it resides in and near. Then there is the model of what is to be accomplished; this is encoded in the objective function(s). The optimizing algorithm searches through potential solutions to find one that best minimizes (or maximizes) the objective function. If the modeling has been done well, then that solution will be a good clinical solution, with luck the best available. However, since the models are always imperfect and approximate, the results need to be evaluated according to criteria both objective and subjective. There is seldom a single, best answer. Since there usually are multiple objectives that compete with each other, the planner uses whatever tools there are to steer the optimizer along the Pareto front, developing a set of possible solutions that cannot be improved upon in any one area without losing quality in another. Our job is to define those choices, and the clinician’s is to choose between them.
References Baltas, D., C. Kolotas, K. Geramani, R. F. Mould, G. Ioannidis, M. Kekchidi, and N. Zamboglou. (1998). “A conformal index (COIN) to evaluate implant quality and dose specification in brachytherapy.” Int J Radiat Oncol Biol Phys 40(2):515–524. Deasy, J. O. (1997). “Multiple local minima in radiotherapy optimization problems with dose-volume constraints.” Med Phys 24(7):1157–1161. D’Souza, W. D., R. R. Meyer, B. R. Thomadsen, and M. C. Ferris. (2001). “An iterative sequential mixed-integer approach to automated prostate brachytherapy treatment plan optimization.” Phys Med Biol 46(2):297–322.
22–Optimization in Brachytherapy
433
Edmundson, G. K., N. R. Rizzo, M. Teahan, D. Brabbins, F. A. Vicini, and A. Martinez. (1993). “Concurrent treatment planning for outpatient high dose rate prostate template implants.” Int J Radiat Oncol Biol Phys 27(5):1215–1223. Edmundson, G. K., D. Yan, and A. Martinez. (1995). “Intraoperative optimization of needle placement and dwell times for conformal prostate brachytherapy. Int J Radiat Oncol Biol Phys 33:1257–1263. Ezzell, G., and R. W. Luthmann. “Clinical Implementation of Dwell Time Optimization Techniques for Single Stepping-Source Remote Applicators” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.). Madison, WI: Medical Physics Publishing, pp. 617–639, 1995. Geman, S., and D. Geman. (1984). “Stochastic relaxation, Gibbs distributions, and Bayesian restoration of images.” IEEE Trans Pattern Anal Mach Intell (PAMI) 6:721–741. Holland, J. H. Adaptation in Natural and Artificial Systems. University of Michigan, Ann Arbor. 1975. Lahanas, M., D. Baltas, and S. Giannouli. (2003a). “Global convergence analysis of fast multiobjective gradientbased dose optimization algorithms for high-dose-rate brachytherapy.” Phys Med Biol 48(5):599–617. Lahanas, M., D. Baltas, and N. Zamboglou. (2003b). “A hybrid evolutionary algorithm for multi-objective anatomybased dose optimization in high-dose-rate brachytherapy.” Phys Med Biol 48(3):399–415. Lee, E. K., R. J. Gallagher, D. Silvern, C. S. Wuu, and M. Zaider. (1999). “Treatment planning for brachytherapy: An integer programming model, two computational approaches and experiments with permanent prostate implant planning.” Phys Med Biol 44(1):145–165. Lessard, É., and J. Pouliot. (2001). “Inverse planning anatomy-based dose optimization for HDR-brachytherapy of the prostate using fast simulated annealing algorithm and dedicated objective function.” Med Phys 28(5):773–779. Martinez, A., J. Gonzalez, J. Stromberg, G. Edmundson, M. Plunkett, G. Gustafson, D. Brown, D. Yan, F. Vicini, and D. Brabbins. (1995). “Conformal prostate brachytherapy: Initial experience of a phase I/II dose-escalating trial.” Int J Radiat Oncol Biol Phys 33(5):1019–1027. Martinez, A. A., I. Pataki, D. Edmundson, E. Sebastian, D. Brabbins, and G. Gustafson. (2001). “Phase II prospective study of the use of conformal high-dose-rate brachytherapy as monotherapy for the treatment of favorable stage prostate cancer: A feasibility report.” Int J Radiat Oncol Biol Phys 49(1):61–69. Messing, E. M., J. B. Zhang, D. J. Rubens, R. A. Brasacchio, J. G. Strang, A. Soni, M. C. Schell, P. G. Okunieff, and Y. Yu. (1999). “Intraoperative optimized inverse planning for prostate brachytherapy: Early experience.” Int J Radiat Oncol Biol Phys 44(4):801–808. Metropolis, N., A. Rosenbluth, N. Rosenbluth, A. Teller, and E. Teller. (1953). “Equation of state calculations by fast computing machines.” J Chem Phys 21:1087–1092. Milickovic, N. M. Lahanas, M. Papagiannopoulo, N. Zamboglou, and D. Baltas. (2002). “Multiobjective anatomybased dose optimization for HDR-brachytherapy with constraint free deterministic algorithms.” Phys Med Biol 47(13):2263–2280. Press, W. H., S. A. Teukolsky, W. T. Vettering, and B. P. Flannery. Numerical Recipes in C, Second Edition. New York: Cambridge University Press, 1988. Saw, C. B., N. Suntharalingam, and A. Wu. (1993). “Concept of dose nonuniformity in interstitial brachytherapy.” Int J Radiat Oncol Biol Phys 26(3):519–527. Shewchuk, J. R. (1994). “An introduction to the conjugate gradient method without the agonizing pain.” http://www2.cs.cmu.edu/~jrs/jrspapers.html. Sloboda, R. S. (1992). “Optimization of brachytherapy dose distributions by simulated annealing.” Med Phys 19(4):955–964. van der Laarse, R., and R. W. De Boer. “Computerized High Dose Rate Brachytherapy Treatment Planning” in Brachytherapy HDR and LDR. A. A. Martinez, C. G. Orton, and R. F. Mould (eds.). Columbia, MD: Nucletron Corporation, 1990. van’t Riet, A., A. C. Mak, M. A. Moerland. L. H. Elders, and W. van der Zee. (1997). “A conformation number to quantify the degree of conformality in brachytherapy and external beam irradiation: Application to the prostate.” Int J Radiat Oncol Biol Phys 37(3):731–736. Yang, G., L. E. Reinstein, S. Pai, Z. Xu, and D. L. Carroll. (1998). “A new genetic algorithm technique in optimization of permanent 125I prostate implants.” Med Phys 25(12):2308–2315.
434
Gary A. Ezzell
Yoo, S., M. E. Kowalok, B. R. Thomadsen, and D. L. Henderson. (2003). “Treatment planning for prostate brachytherapy using region of interest adjoint functions and a greedy heuristic.” Phys Med Biol 48(24):4077–4090. Yu, Y., and M. C. Schell. (1996). “A genetic algorithm for the optimization of prostate implants.” Med Phys 23(12):2085–2091. Yu,Y., J. B. Zhang, R. A. Brasacchio, P. G. Okunieff, D. J. Rubens, J. G. Strang, A. Soni, and E. M. Messing. (1999). “Automated treatment planning engine for prostate seed implant brachytherapy.” Int J Radiat Oncol Biol Phys 43(3):647–652. Zelefsky, M, J.,Y. Yamada, C. Marion, S. Sim, G. Cohen, L. Ben-Porat, D. Silvern, and M. Zaider. (2003). “Improved conformality and decreased toxicity with intraoperative computer-optimized transperineal ultrasound-guided prostate brachytherapy.” Int J Radiat Oncol Biol Phys 55(4):956–963.
Chapter 23
QA Review of Brachytherapy Treatment Plans Zuofeng Li, D.Sc. Department of Radiation Oncology Mallinckrodt Institute of Radiology Washington University School of Medicine St. Louis, Missouri Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 435 Review of Treatment Prescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 436 Review of Technical Aspects of Treatment Plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438 Patient Image Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438 Applicator Insertion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 438 Quality of Images and Accuracy of Imaging Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439 Target and Critical Organ Segmentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440 Review of Input Parameters to Treatment Plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441 Applicator/Catheter Geometry Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 441 Source/Applicator/Catheter Localization Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442 Point of Interest/Critical Organ Localization Accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442 Source Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 442 Plan Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443 Plan Quality Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443 Point-based Plan Quality Evaluation Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 443 Volume-based Plan Quality Evaluation Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444 Independent Plan Calculation Check . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448 Tools for Independent Calculation Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 449 Independent Calculation Check of Gynecological Intracavitary Implants . . . . . . . . . . . . . . . . . . . . 450 Independent Calculation Check of Intraluminal Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450 Independent Calculation Check of Planar Interstitial Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453 Independent Calculation Check of Volume Interstitial Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . 454 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
Introduction Brachytherapy treatment isodose plans are commonly used to determine the optimal source strength and distribution for the treatment to be delivered, to document the isodose distribution resulting from a delivered treatment, and to serve as a guide for subsequent management of a patient’s disease. While most brachytherapy treatment plans that a clinical medical physicist encounters on a routine basis serve the former two purposes, the use of brachytherapy treatment plans as a guidance for future patient treatment is equally important, especially in the cases of multifractionated brachytherapy treatments such as high dose-rate (HDR) intracavitary gynecological procedures; or in the case of post-permanent prostate implant dosimetry plans, where the quality of the implant is evaluated and used as guidance to determine the subsequent management of the patient’s disease, including potential re-implantation of the prostate to increase doses to the “cold-spots” of the initial implant. Accuracy of treatment plans therefore not only has direct impact on the delivery of an individual brachytherapy treatment, but also impacts on the entire process of the management of a patient’s disease. Institutional policies should be implemented that require physics review of all brachytherapy treatment plans, as well as the timing of the review relative to the start and/or completion of the treatment.
435
436
Zuofeng Li
A physics treatment plan quality assurance (QA) review scrutinizes a brachytherapy treatment plan for the following end points: 1. Treatment plan’s adherence to institutional treatment policies and/or national treatment guidelines. 2. Accuracy of technical parameters. 3. Suitability of the plan for the patient’s treatment in terms of dose/source distribution optimization, dose homogeneity, target coverage, and critical organ doses. In addition, an independent calculation is often performed to validate the correctness of the treatment. While this calculation may not be able to detect errors on the order of a few percent in the treatment plan, it serves to alert the physicist in the event that a significant error has occurred. Such errors can occur due to the entry of incorrect technical parameters, a misinterpretation of the treatment prescription, or a software bug that was not detected in the treatment planning system (TPS) acceptance testing and commissioning process. Treatment plan QA review, by default, assumes that the TPS has been properly acceptancetested/commissioned, that the source data are up-to-date, and that periodic TPS QA tests have been performed. However, TPS QA does not replace or in any manner diminish the importance of QA review of treatment plans, even for basic algorithmic accuracy or correctness. A state-of-the-art brachytherapy TPS is a complex software and hardware system, capable of electronic input and output data transfer; manipulation of volumetric patient data including image fusion, coordinate transform and rotation, and segmentation; automatic or semi-manual optimization of HDR dwell positions and times; and dose calculation using various formalisms. A typical TPS acceptance testing and commissioning process will not allow testing of all available functions in the TPS nor the different manners that these functions are combined to obtain a patient treatment plan. Vigilance and efforts in detecting any unexpected or abnormal TPS behavior are therefore a critical aspect of treatment plan QA review.
Review of Treatment Prescription While the preparation of a treatment prescription is the sole responsibility of the treating radiation oncologist, a careful review of the prescription is an integral part of any treatment plan QA review. Review of the prescription should aim at the following aspects: 1. Prescription is appropriate for the goals of the brachytherapy treatment. While a physicist cannot question every aspect of a treatment prescription, it is important that the physicist understand the goals of the treatment and how the prescription is designed to achieve these goals. Errors in prescription do occur, for example, due to a miscommunication between the departments of surgery, pathology, and radiation oncology, or simply in the process of transcribing information on patient disease or the prescription itself. A prescribed minimal peripheral dose of 145 Gy for the treatment of permanent prostate seed implant, for example, using 125I seeds, is only appropriate if the patient is to receive brachytherapy alone for his early stage prostate cancer. Variations in prescribed dose or isotope should be immediately brought to the attention of the treating physician. Review of the treatment prescription therefore necessitates a careful reading of the patient’s history and physical exam report, including the stage of the disease, as well as any previous treatment that the patient may have received. 2. Prescription adheres to institutional treatment policies and/or national treatment guidelines for the disease. Many institutions have treatment policies or protocols that aim to maintain consistent treatment prescriptions and techniques for patients with similar diseases. In addition, national organizations, such as the American Brachytherapy Society, have published guidelines and pattern
23–QA Review of Brachytherapy Treatment Plans
437
of care studies on the brachytherapy treatment of various cancers (Arthur et al. 2003a; Beyer et al. 2000; Gaspar et al. 1997; Nag 2000; Nag et al. 1993; 1999a,b; 2000a,b,c; 2001a,b,c; 2002; 2003a,b; Potter et al. 2001). These guidelines, in the event that no institutional treatment policies exist, should, in general, be followed. Random deviations from an existing protocol not only endanger the success of a patient’s treatment, but also carry medical-legal consequences. While individual treating physicians may choose to deviate from these protocols based on the particular patient’s disease and physical conditions, the physicist should communicate with the treating physician to assure such deviations are reasonable to the extent that that patient’s disease warrants. Frequent deviations in treatment prescription from institutional and/or national treatment guidelines should proceed either in a clinical trial setting, or should lead to a discussion on the need to revise treatment protocols. At Mallinckrodt Institute of Radiology, the treatment of cervical cancer follows a treatment protocol (see Table 1), using a combination of whole pelvis and split-field pelvis external beam fields, interlaced with brachytherapy tandem and ovoids intracavitary implants. The source-loading pattern for a given tandem and ovoid insertion also follows a strict protocol (Grigsby, Williamson, and Perez 1992). Physicists are expected to be familiar with these protocols and to review the prescriptions of a tandem and ovoids treatment for deviations. When starting a new treatment site or technique for which no institutional protocols exist, the physicist should work with the treating physician to develop such a protocol, based on literature reviews and attendance at workshops and training courses. Examples of such treatments include accelerated partial breast brachytherapy treatments using low dose rate (LDR), interstitial HDR, or MammoSite® HDR techniques. 3. The prescription should be complete and accurate for fulfillment of regulatory requirements. The Nuclear Regulatory Commission (NRC), in Title 10, Part 35 of the Code of Federal Regulations, specifies the required information to be completed in the “written directive” before and/or after Table 1. Mallinckrodt Institute of Radiology Treatment Policy for Cervical Cancer External beam treatment (Gy) Treatment scheme
Indication
Whole pelvis
Split field
Intracavitary treatment
Total: smallest to largest insertion
Maximum Target vaginal dose Point A mgRaeq·h (Gy) dose (Gy)
Point P dose (Gy)
mgRaeq·h
56–60
5580–7980
A
IB < 2 cm
0 Gy
45 Gy
7000
150
70–80
B
IB 2–4 cm
10
40
7500
150
80–85
61–66
5580–8550
C
IB/IIA/IIB/IIIA bulky (>4 cm), limited parametrial extension
20
30
8000
150
84–90
61–67
5600–9100
D
IIB/IIB bulky, extensive parametrial extension
20
40
8000
150
84–90
71–77
5600–9100
E
IIB, IIIB, IV, poor anatomy, poor regression
40
20
6500
150
92–94
69–74
4610–7410
(Reprinted from Williamson, J. F. “Clinical Brachytherapy Physics” in Principles and Practice of Radiation Oncology, 3rd ed. C. A. Perez, L. W. Brady, E. C. Halperin, and R. K. Schmidt-Ullrich (eds.). © 1998 with permission from Lippincott Williams & Wilkins.)
438
Zuofeng Li insertion of sources into the patient. While not inherently a part of a treatment plan review, it is often convenient for the physicist to review the prescription, if used as the written directive for the treatment, to ensure that all required information is entered correctly.
Review of Technical Aspects of Treatment Plans A careful review of technical aspects of brachytherapy treatment plans typically constitutes the most timeconsuming part of a comprehensive treatment plan QA review. It is also the part of a treatment plan QA review that requires the expertise of a qualified physicist. Once satisfied that the treatment prescription is appropriate for the treatment under review, the physicist can now concentrate on the treatment plan itself. This review should address all milestones in the process of treatment plan creation, starting from patient data acquisition, be they volumetric or planar images, to the adequacy of final isodose plots. A checklist is often helpful for review of complex treatment plans, especially for HDR treatments, where the plan review must be completed in a compressed timeframe.
Patient Image Review A review of patient imaging serves multiple purposes in the process of treatment plan QA review. It assures that the applicator insertion is adequate for the treatment, that the images were acquired with minimal artifacts, and that the images are optimized for accurate source/applicator localization and dose calculation. The very first step in patient image review needs to confirm that the images belong to the correct patient. While this may seem a trivial check, the consequences of using the wrong patient image for treatment planning can be significant. The patient name, ID, and date of acquisition on the images must be verified to ensure that the correct image data set is used for treatment planning. Once that is done, the following aspects of image review can proceed. Applicator Insertion The physicist should review the placement of applicators and catheters relative to the patient’s anatomy to ensure that their locations are adequate for the desired treatment. The applicator insertion should allow adequate dose coverage of the treatment target, and minimize critical organ doses. For cervical cancer treatments, cervical markers are inserted into the cervix for identification on planar x-ray films. The optimal insertion of tandem and ovoid applicators requires that the ovoid surfaces to be at close proximity to these markers. As Katz and Eifel (2000) reported, a typical distance between the surfaces of the ovoids to the markers at M. D. Anderson Cancer Center is approximately 7 mm. Figure 1 shows a diagram of applicator geometry. Additional elements of a “good” tandem and ovoids implant include the symmetric placement of ovoids relative to the tandem both on the antero-posterior (AP) and on the lateral films, adequate distance of tandem to the sacrum and the pubis on the lateral film, and adequate packing to push the bladder and rectum away from the applicators. An inverted tandem insertion, indicated by the tandem curving toward the sacrum, may indicate perforation of the uterus (Jhingran and Eifel 2000). While patient-to-patient variations will occur, significant deviations from such typical “good geometry” should be brought to the attention of the treating physician. Such deviations may be due to the inadequate insertion of the applicators, or more likely due to the “slipping” of the applicator system caudally following insertion, and may indicate a need to re-insert and repack the applicators before patient films are acquired again for proper treatment planning. The dosimetry of interstitial implants similarly is dependent on the adequate insertion of treatment catheters, for both coverage of target volume and minimization of critical organ doses. Clips and markers are often placed in the tumor resection margin during the surgical removal of the tumor. They should
23–QA Review of Brachytherapy Treatment Plans
439
Figure 1. Tandem and ovoids applicator system geometry. (Reprinted from Katz, A., and P. J. Eifel. (2000). “Quantification of intracavitary brachytherapy parameters and correlation with outcome in patients with carcinoma of the cervix.” Int J Radiat Oncol Biol Phys 48(5):1417–1425. © 2000 with permission from Elsevier.)
be used to help determine the adequacy of catheter insertion. Catheter separations or the distance between peripheral catheters and the target periphery should be evaluated, as they directly impact the quality of the implant in terms of both target coverage and dose inhomogeneity. Inadequate catheter insertion should be brought to the attention of the treating physician at the earliest possible time, so that remedial actions, such as insertion of additional catheters, may be taken. Catheters near critical organs, such as the rectum in gynecological interstitial implants, may need to be left unused to avoid excessive dose to this critical organ. Alternatively, “spacers” may need to be used in these catheters to allow adequate distances between the sources and the critical organs. The physicists performing treatment plan review should be familiar with all the applicator systems used in their institutions, especially in their physical dimensions and limitations. This is of particular relevance in HDR intraluminal treatments, such as endobronchial, esophageal, and bile duct treatments. As the HDR units typically have a limited range of distal-most and proximal-most treatable distances, the physicist should review the treatment planning films to ensure that the treatment target, as identified by the surgical markers, falls within this range, as can be identified by the x-ray markers inserted into the treatment catheters. Table 2 shows the maximum and minimum indexer lengths available in the Nucletron V2 HDR unit, together with the treatable target lengths and step sizes. Such mechanical limits are dependent on the HDR unit’s design and should be reviewed for each implant during patient imaging review. Quality of Images and Accuracy of Imaging Parameters Patient images, either volumetric or planar, need to be reviewed for their adequacy in allowing accurate applicator and catheter localization and dose calculations. For each type of brachytherapy implant, an imaging protocol should be developed and adhered to, with the parameters of these protocols chosen to minimize imaging artifacts, and to allow accurate applicator, target, and critical organ localization. Breathing artifacts in computed tomography (CT) scans or orthogonal radiographs will significantly
440
Zuofeng Li Table 2. Nucletron V2 HDR Unit Characteristics Parameter
Maximum Treatable Distance (from indexer faceplate) Minimum Treatable Distance (from indexer faceplate) Maximum number of dwell positions
Step sizes Gynecological transfer tube Flexible catheter transfer tube Stainless steel needle transfer tube
Nominal Values 1500 mm
Comment 4 mm additional catheter length for check cable test
725 mm 48
# of dwell positions multiplied by step size must be between min. and max. treatable distance
2.5 mm, 5 mm, 10 mm 1200 mm 1000 mm 1200 mm
increase the uncertainties in applicator and catheter localization accuracy, as well as in the delineation of treatment target and critical organs. Figure 2 shows an orthogonal pair of radiographs for the HDR treatment of bile duct, where the patient breathing motion artifact caused up to 1 cm differences in the y-coordinates of the x-ray markers. The physicist needs to evaluate the dosimetric consequences of such imaging artifacts, and communicate with the treating physician to arrive at the optimal actions to be taken in the patient’s treatment. When volumetric imaging is used for brachytherapy treatment planning, the field of view (FOV), slice thickness, table pitch, and gantry angle of the scans need to be reviewed to ensure that they are per the scanning protocol established for the particular brachytherapy treatment. Deviations from the established scanning parameters may reduce the accuracy of applicator, target, and critical organ localization, and could result in significantly increased dose calculation uncertainties. When orthogonal or other types of planar radiographs are used, the isocenter of the radiographs should be placed near the center of the target volume. The magnification factors of the radiographs and gantry angles need to be confirmed. The coordinate system used on the simulator, such as a Varian Ximatron unit, may be different from the International Electrotechnical Commission (IEC) coordinate system used in some treatment planning systems such as the Nucletron Plato TPS. The conversion of gantry angles between the two coordinate systems must be done correctly to avoid significant source localization errors. The filming technique used should allow clear and unambiguous recognition of applicators, surgical clips, and radiographic markers on the film. Target and Critical Organ Segmentation Image-based brachytherapy treatment planning has become common for some treatment sites, such as LDR and HDR transrectal ultrasound (TRUS) guided prostate brachytherapy, accelerated partial breast brachytherapy using either interstitial or MammoSite technique and gynecological intracavitary brachytherapy using CT/magnetic resonance (MR)-compatible applicators. The segmentation of treatment target volumes and critical organs for each patient may be performed by a radiation oncologist or a dosimetrist. The physicist reviewing the treatment plan should be familiar with the anatomy of the treated site, as well as the institutional and/or national treatment guidelines related to this part of treatment planning. The contours of segmented target and critical organs should be reviewed for both anatomical accuracy when applicable, such as the prostate, seminal vesicles, bladder, rectum for prostate implant, or the seroma and MammoSite balloon for breast implant. The expanded planning target volume (PTV) should be
23–QA Review of Brachytherapy Treatment Plans
(a)
441
(b)
Figure 2. AP and lateral films of a bile duct HDR treatment. (a) Patient instructed to breathe in and hold breath; (b) No patient instruction.
reviewed for appropriate application of expansion margins and for overlaps with critical organs such as skin of the breast.
Review of Input Parameters to Treatment Plans Once satisfied that the images used to construct treatment plans are of adequate quality for the purpose of treatment planning, and that the applicators and catheters are inserted to allow the desired dose distribution, the physicist can concentrate on the correctness and accuracy of applicator and catheter localization and input into the TPS. Applicator/Catheter Geometry Parameters The dimensions of brachytherapy applicators, catheter indexer lengths, and their numbering directly have an impact on the correctness and the accuracy of the treatment plan. Ovoid and vaginal cylinder diameters need to be verified by comparison with documented values or direct measurement from the patient images. Ovoids larger than 2 cm in diameters are often constructed by sliding a plastic cap on 2-cm diameter ovoids, and are therefore often indistinguishable from the 2 cm ovoids. The diameters of such ovoids, as well as CT/MR-compatible vaginal cylinders, therefore must be documented in the operating room immediately following their insertion. The physicist reviewing the treatment plan should review such documentation and compare the values used in the treatment plan to ensure that they are correctly recorded and used in the treatment plan. Diameters of vaginal cylinders with metal identification rings can be measured off the radiographs, and are thus readily verifiable. Indexer lengths of HDR catheters must be reviewed carefully, with full understanding of the type of catheters used, for example, stainless steel needles or plastic flexible catheters. Depending on the transfer tubes used, treatment distance errors of up to 50 cm have occurred. Every attempt is warranted in avoiding such systematic errors that resulting in significant dose delivery errors and medical events.
442
Zuofeng Li
Source/Applicator/Catheter Localization Accuracy The input of sources and applicators/catheters are typically performed in treatment planning using coded x-ray markers or CT-compatible markers. The physicist reviewing a treatment plan needs to be familiar with the coding scheme of the x-ray markers, or the physical characteristics of CT-compatible markers, so that the distance information encoded in these markers can be correctly entered into the TPS. When interstitial or intraluminal brachytherapy treatments using multiple catheters are planned, the individual catheters should be numbered, such that the source loading and/or dwell position and time calculations individualized for each catheter can be correctly reproduced in every treatment fraction. The physical numbering of the catheters needs to be compared with the data entered into the treatment plan. If coded x-ray markers are used instead of numbering of treatment catheters, the x-ray markers must remain in the treatment catheters until source loading time. The physicist should be familiar with their appearances on the radiographs or CT images. Treatment planning systems often report source localization errors by comparing the reconstructed source length values with their expected values. These reports should be carefully reviewed to ensure that no significant errors have occurred during source localization. The use of spacer and the value of step sizes in the treatment plan require special attention. Their locations and values need to be compared with those indicated on the patient images, and double-checked prior to treatment delivery. Errors in their use in treatment planning can result in geographically displaced dose distributions, or potentially treatment of wrong site. Point of Interest/Critical Organ Localization Accuracy Points of interest are often used in brachytherapy treatment plans to represent prescription point or critical organ doses. They are commonly measured on orthogonal films or from volumetric patient images, and entered manually into the TPS for calculation of doses to these points. The localization of points of interest is of particular importance in gynecological tandem and ovoids treatment planning, as they may directly affect the overall delivery of the treatment. The classic Manchester system definition of point A measures level of point A superiorly from vaginal fornices (typically assumed to be the superior ovoid surfaces). Errors in localization of point A often result in greater than 10% error in delivered dose. While point B, rectum and bladder points are typically not used for brachytherapy treatment prescription, their dose values are sometimes used to optimize treatment plans such as changing the source loading pattern and dwell times to reduce rectum and bladder doses, or to assist in subsequent patient management, such as the need to include parametrium boost external beam radiation treatment. In the treatment planning for MammoSite breast treatment, the balloon diameter and its center may be measured off radiograph films. These measurements need to be reviewed for accuracy, as they are directly used for dose prescription and delivery. For intraluminal implants or planar or volumetric interstitial implants, the doses are often prescribed to points at a given distance to the catheter(s). The placement of these points again directly has an impact on the total dose delivered to the patient. Such points may be automatically generated, such as in the case of the Nucletron Plato TPS, or manually entered for LDR treatments. The physicist reviewing a treatment plan should fully understand the mechanism through which these points are determined and localized relative to the treatment catheters and patient anatomy, and determine its appropriateness as applied to the plan under review as well as its adherence to institutional protocols. Source Characteristics TPS acceptance testing and commissioning usually includes a complete testing of the source characteristics, including benchmark data, physical and active source dimensions, as well as single and multiple
23–QA Review of Brachytherapy Treatment Plans
443
source dose distribution in a predetermined configuration. Treatment plan QA review therefore can concentrate on the particular sources used for a given patient treatment. Aspects of source characteristics review include verifying that the correct source model is used for the treatment plan, that the source decay was calculated accurately, and that the number of sources used in the treatment plan agrees with the prescription. Plan Optimization On a modern brachytherapy TPS, a number of automatic or semi-automatic plan optimization algorithms are typically available. While it is not the intention of this chapter to discuss these algorithms, a physicist reviewing a treatment plan, in which one of these algorithms is used, needs to have a detailed understanding of these algorithms. The selection of a particular algorithm should be specific to the goals of the treatment, in addition to keeping with institutional consistency. Automatic optimization algorithms most often are used for the preplanning of permanent prostate implants and various HDR treatment implants. For permanent prostate implants, the optimization algorithm may be a simple geometry model that places seeds at locations in or near the prostate, based on a set of geometry rules, such as distance between needles and seeds, columns to avoid, etc. Alternatively, the optimization algorithm may attempt to distribute the seeds based on dosimetric specifications in terms of target coverage and critical organ doses. The physicist should work with the treating physician in selecting an optimization algorithm for his or her institution’s treatment planning for prostate implants. Attempts should be made to use the same optimization algorithm and optimization parameters for all prostate implants. This effort in maintaining treatment planning consistency will help achieving more predictable treatment plans in the form of a relationship between prostate volume and the number and total activity of sources. Such information is critical in determining whether a plan “makes sense,” and should help maintain consistency of patient treatment. For HDR treatments, the selection of optimization algorithms and optimization parameters can directly impact the quality of the treatment plan as well. The Geometry Optimization algorithm (Edmundson 1990) and its variants are often used for interstitial planar and volume implants, while optimization based on dose points has been successfully used for gynecological implants (Stitt et al. 1992; Thomadsen et al. 1992). The placement of dose points relative to the applicators directly affects the resulting dose distribution. Physicists reviewing such treatment plans should pay special attention to how the dose points were placed. When dose volume histogram (DVH)-based optimization algorithms are used, the physicist should attempt to establish DVH optimization parameters for each treatment site, and subsequently review each treatment plan for its adherence to such protocols.
Plan Quality Evaluation The evaluation of a brachytherapy treatment plan quality includes the aspects of target coverage and normal organ doses, as well as dose homogeneity of the plan. Similar to external beam radiation therapy, the evaluation of the quality of a treatment plan is specific to a treatment site, as well as the delivery technique used in the treatment.
Point-based Plan Quality Evaluation Parameters Traditionally, brachytherapy plans based on two-dimensional radiographs do not have explicit quantitative representation of treatment target or critical organs beyond what can be estimated based on patient bony anatomy, surgical clips, or markers inserted into the patient, or reconstructed from Foley balloons and vaginal packing for gynecological treatments. When necessary, doses to these points can be calculated
444
Zuofeng Li
as surrogates to target coverage and critical organ doses. Significant over- or underdosage to those landmarks needs to be discussed with the treating physician. For interstitial implants, the International Commission on Radiation Units and Measurement (ICRU) (ICRU 1997) defines several parameters that are useful in the evaluation of the quality of a treatment plan. The report defines the central plane of an interstitial implant to be a plane through the major (long) direction of the implant. In practice, this often is taken to be the direction of the implant needles or catheters. For a complex implant, in which a single central plane cannot be defined with respect to all implanted catheters, subvolumes may be defined, such that a central plane may be defined for each subvolume. The mean central dose (MCD) is an extension of the basal dose of the Paris implant system, defined as the average of local minimum doses between the sources in the central plane(s) of an implant. For implants with parallel needles, the MCD can be determined geometrically as the midpoint between a group of neighboring needles. Figure 3 shows the definition of central planes and the corresponding MCD calculation for a complex interstitial implant. The Dose Homogeneity Index (DHI), defined as the ratio of the prescription dose to the MCD, has found frequent use in determining the whether a treatment plan is acceptable in terms of dose homogeneity.
Volume-based Plan Quality Evaluation Parameters When the three-dimensional volumetric dose distribution is calculated for a brachytherapy treatment plan, additional parameters may be defined for the plan quality evaluation. Dose distribution in the treated volume, which may be calculated with only the knowledge of source locations, has proven to be useful for dose homogeneity evaluations. When a set of volumetric images is available for a given patient, dose distribution to anatomically defined treatment target, as well as organs at risk (OARs), may be critically evaluated in the quality evaluation of a brachytherapy treatment plan. Given a dose distribution from implanted sources, the ICRU (1997) defines the high dose (HD) and low dose (LD) volumes, and the prescribed dose (PD) that help to evaluate the dose homogeneity and
Figure 3. A complex implant with two central planes defined. The Mean Central Dose (MCD) is the average of the doses at points A–G. (Reprinted from ICRU Report 58: Dose and Volume Specifications for Reporting Interstitial Therapy. Bethesda, MD: ICRU, © 1997 with permission from ICRU, Bethesda, MD.)
23–QA Review of Brachytherapy Treatment Plans
445
coverage of an implant. Details on the definitions and applications of those quantities can be found in the chapter on ICRU interstitial reporting recommendations, and are not repeated here. A useful parameter in brachytherapy plan evaluation is the maximum contiguous dose, defined to be the minimum dose that encloses all sources in an implant (Neblett et al. 1985). This can be estimated on a TPS by plotting consecutively decreasing isodose clouds until an isodose cloud is seen to enclose all sources. This all-source-enclosing enclosing isodose cloud should cover the intended treatment target. Many other tools exist that help to describe quantitatively the quality of a brachytherapy plan, especially in regard to dose homogeneity. Anderson (1995) and Thomadsen (1999) described them in detail. As many of the brachytherapy treatment planning systems today are able to calculate DVHs of various types (cumulative, differential, natural), DVH values of the treated volume, i.e., the volume of tissue receiving a given dose level, are often used for evaluation of dose homogeneity of a treatment plan, while DVH values for a given segmented organ, either target or OARs, are useful for evaluation of the target coverage of the treatment as well as risk of treatment complications. A simple plotting of the volume of an organ (the sum of voxels) receiving doses in a given range produces a differential DVH is shown in Figure 4. The cumulative DVH is calculated from the differential DVH by summing all voxels receiving doses up to a given dose level, then plotting these “cumulative” volumes against the dose levels, shown in Figure 5. A disadvantage of differential or cumulative DVH plots for brachytherapy is that, due to the high dependence of brachytherapy dose distribution on the distance between the sources and the points of interest, all DVH plots will appear similar, as demonstrated in Figure 6, which shows the cumulative DVH plot for a single HDR dwell position treatment (MammoSite), compared to the cumulative DVH of an optimized HDR interstitial breast brachytherapy treatment of Figure 5. Anderson (1986) developed a form of DVH, the natural DVH, which mathematically removes the inverse-square-dependence of dose
Figure 4. Differential DVH plot of the treated volume of a breast brachytherapy treatment.
446
Zuofeng Li
Figure 5. Cumulative DVH plot of the treated volume of an optimized breast brachytherapy treatment.
Figure 6. Cumulative DVH of the treated volume of a MammoSite® treatment.
23–QA Review of Brachytherapy Treatment Plans
447
distribution in a brachytherapy implant, resulting in a DVH plot as shown in Figure 7. Note that a peak appears near the prescription dose of 340 cGy. In comparison, for a MammoSite breast brachytherapy treatment using a single dwell position (effectively a point source), the natural DVH appears like a flat line, as seen in Figure 8. Various parameters can be calculated from a natural DVH plot, as seen in Figure 7, that may help comparative evaluation of a treatment plan. While it is difficult to generalize the acceptable values of these parameters, in general, sharper peaks and rapid fall-off in the high-dose side of a natural DVH indicate more homogeneous dose distribution. Saw and Suntharalingam (1991) described the dose non-uniformity ratio (DNR), defined as the ratio of a high-dose volume and the total volume of tissue receiving the prescription dose. As the magnitude of high-dose volume may be related to complications such as soft tissue necrosis, the value of DNR needs to be minimized. When volumetric imaging is used, such as in prostate or breast brachytherapy, the target coverage and critical organ doses of the treatment plan should be evaluated based on existing institutional protocols or national guidelines. The Coverage Index (CI), defined as the ratio of PTV/CTV (clinical target volume) receiving the prescribed dose to the volume of the PTV/CTV, is typically used to describe target coverage. Significant over- or underdosage to the target or the critical organs may require modification to the plan. The physicist performing plan review should have intimate knowledge of those protocols and guidelines, and be ready to recommend revision of a treatment plan or suggest remedial action. The use of DVH plots for brachytherapy plan evaluation can be largely empirical. Similar to external beam DVH plots, acceptable DVH values depend on the specific disease being treated, as well as the treatment technique adopted. For permanent prostate implants, the AAPM Task Group 64 (TG 64) report (Yu et al. 1999) recommended a set of DVH values. Stock et al. (2002) reported that keeping the dose that covers 90% of the PTV, D90, between 140 Gy to 180 Gy appears to be optimal, with D90 < 140 Gy
Figure 7. The Natural DVH plot of the optimized interstitial breast brachytherapy implant shown in Figure 6.
448
Zuofeng Li
Figure 8. Natural DVH of a MammoSite® breast brachytherapy treatment.
associated with increased biochemical failure, and D90 > 180 Gy with a small increase in long-term urinary symptoms. For interstitial breast HDR treatments, Arthur et al. (2003b) recommended a set of parameters extracted from the DVH of the treated volume for evaluation of HDR accelerated breast brachytherapy treatments, as shown in Table 3.
Independent Plan Calculation Check An immediate question that arises after all the detailed and tedious review of the plan for a given brachytherapy treatment is why perform an independent plan check calculation? The answer to this question determines the complexity of the check calculation process and procedure. The author submits the following goals for an independent brachytherapy plan calculation check: 1. Prevention of catastrophic data input errors. Such errors may be due to misidentification of dose prescription points, incorrect data transfer, or misinterpretation of treatment prescription.
Table 3. Accelerated Partial Breast Brachytherapy Plan Evaluation Criteria Volume of Treated Tissue Receiving 150% of Prescription Dose (V150)
Volume of Treated Tissue Receiving 200% of Prescription Dose (V200)
Dose Volume Ratio
50 cc
20 cc
0.75
V150 1 − V 100
23–QA Review of Brachytherapy Treatment Plans
449
2. Prevention of unexpected software errors. Even though a TPS has been subject to rigorous testing during its commissioning and acceptance testing process, it is conceivable that certain parts of the software were never tested, or the use of a specific function was never anticipated. The complexity of a modern brachytherapy TPS is such that it is impossible to test all possible functions and the sequences in which these functions are utilized to arrive at a treatment plan. As treatment planning systems are often designed with a certain operation sequence in mind, any deviation from the intended sequence may result in unexpected results. 3. Regulatory compliance. The NRC requires that all brachytherapy calculations, either manually performed or computer-generated, must be checked. While the NRC does not specify how such checks should be performed, a quantitative calculation check would appear to satisfy this requirement. An independent calculation check, therefore, should be designed to quickly detect significant/catastrophic errors, while not significantly slowing down the treatment planning and delivery process. This is especially important in HDR brachytherapy treatments, where everything must happen in a compressed time frame.
Tools for Independent Calculation Checks Many tools have been available, proposed, or can be modified for use for independent brachytherapy plan checks. Given the strengths or dwell times or the sources and their spatial distributions, an independent plan calculation check scheme calculates dose delivered to a given point. Considering that such algorithms have been available before the arrival of computerized treatment planning, it is no surprise that many traditional dose calculation formalisms or implant systems have found youth in the independent calculation check of brachytherapy plans. In addition to performing a gross dosimetry accuracy check, they may help provide an insight into the relationship of source strengths and resulting doses and their points of specification, compared to computerized number manipulations. An independent plan calculation check can be done, in the more complex form, using a second brachytherapy TPS or software systems specifically developed for this purpose. Saw et al. (1998) described the use of an LDR brachytherapy TPS for the independent calculation check of HDR brachytherapy treatment plans. The source positions are transferred into the LDR TPS via digitization, with appropriate scaling of LDR source strengths to reflect the varying dwell times of HDR source at the active dwell positions. The authors stated that all calculations between the HDR system and the LDR system agreed to within 10%, with 80% of them agreeing to within 5%. Such a scheme, by its completeness, helps to identify human errors in data input as well as potential algorithmic errors in the HDR system in the aspect of dose prescription point placement, coordinate system translation and rotation, and dwell time weight optimization. The authors stated that each independent calculation check took less than 20 minutes. Cohen et al. (Cohen, Amols, and Zaider 2000) and Lachaine et al. (Lachaine, Gorman, and Palisca 2003) developed their own in-house programs to perform independent dose calculations for HDR treatments, either based on the source position input into the HDR TPS (Cohen et al.) or on the HDR treatment unit control file exported from the HDR TPS (Lachaine et al.). Cohen et al. observed a typical discrepancy of up to 3%, while Lachaine et al. saw maximum differences of 2%, compared with the HDR TPS. It should however be noted that simply transferring the already-localized source and point of interest coordinates into a second TPS and performing a second dose calculation will only verify the dose calculation algorithm accuracy, but may not detect the arguably more serious and frequent errors in source localization. At the other end of the spectrum, classical dose calculation algorithms, such as away-and-along tables for cesium tube sources, Sievert integral (Williamson 2003) for unfiltered line sources, Paterson-Parker system (Williamson 2003) for planar and volume implants (with appropriate accounting for modern source
450
Zuofeng Li
strength units and correction factors, such as shown in Table 4), and prostate implant nomograms (Anderson 1976; Anderson, Moni, and Harrison 1993; Wang and Potters 2001), serve to provide quick and often as accurate estimates of the relation between total source strength and prescription dose rates. Other methods (Mayo and Ulin 2001; Rogus, Smith, and Kubo 1998; Ezzell 1994, 2000a,b; Miller, Davis, and Horton 1996; Das et al. 2004; Cohen et al. 2002) have been developed that facilitate the quick manual calculation for specific brachytherapy treatment sites. They are typically easy to use and achieve adequate accuracy that, depending on the application, may substitute for computerized treatment planning.
Independent Calculation Check of Gynecological Intracavitary Implants Gynecological intracavitary implants are often performed using tandem and ovoid applicators or cylinder applicators, either using cesium tubes for LDR applicators or HDR remote afterloaders. At Mallinckrodt Institute of Radiology, the loading patterns of tandem ovoids follow strict rules depending on the length of the tandem and diameter of the ovoids. Computer-generated dose values for points A, B, and P (at 6 cm away from patient midline and level of point A) can be compared to manual calculations based on the calculated, standard values in Table 5. The average of the dose values at left and right points A, B, and P, generated by the TPS, typically are within 10% of the predicted values. While the table was calculated for LDR implants, HDR treatments at Mallinckrodt Institute of Radiology have maintained the same loading pattern by weighting the dwell times accordingly. The table therefore is useful for both LDR and HDR tandem and ovoid treatments. Mayo and Ulin (2001) described a method for checking the treatment time calculation for HDR vaginal cylinder treatments. For dose prescription points located at 5 mm away from cylinder surface in the vaginal apex region, tapering off to cylinder surface at points along the tranverse vaginal apex, the authors proposed the determination of a scaling factor K, that relates the prescribed dose D, the source strength Sk , and the total treatment time TT, in the form of TT = K × D/Sk .
(1)
Given that the dose prescription points are at varying distances from the vaginal cylinder surface, the authors arrived at an equation that described K as a function of vaginal cylinder diameter, the length of the cylinder with prescription point at 5 mm away from cylinder surface, as well as the length of the cylinder with prescription points on cylinder surface. The authors observed a maximum discrepancy of 5% of their independent calculation from the HDR TPS calculations.
Independent Calculation Check of Intraluminal Treatments Intraluminal brachytherapy treatments, such as endobronchial, esophageal, and bile duct treatments, typically use no more than two treatment catheters with long active lengths. When a single catheter of relative little curvature is used for the treatment, such as is often the case for esophageal treatments, the Sievert unfiltered line source integral may be used. Let L be the active length, Sk the total source strength, Λ the single-source dose rate constant, x the away distance of the dose calculation point away from the bisector of the line source, and T the total treatment time, then the dose D at this point can be approximated by D = 2×T ×
Sk × Λ L×x
L . 2 × x
tan −1
(2)
23–QA Review of Brachytherapy Treatment Plans
451
Table 4. Paterson-Parker System Table for Plane and Volume Implants Volume implantsPlanar implants Volume (cm3) (cm2)
mgRaeq·h* 1000 ‘P-P’R
Minimum dose/ IRAK† cGy/(µGy·m2)
Area (cm2)
mgRaeq·h* 1000 ‘P-P’R
Minimum dose/ IRAK† cGy/(µGy·m2)
1 2 3 4 5 10 15 20 25 30 40 50 60 70 80 90 100 110 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400
34 54 70 85 99 158 207 251 291 329 398 462 522 579 633 684 734 782 829 919 1005 1087 1166 1242 1316 1389 1459 1528 1595 1661 1725 1788 1851
3.49 2.20 1.68 1.38 1.194 0.752 0.574 0.474 0.408 0.361 0.298 0.257 0.228 0.206 0.188 0.174 0.162 0.152 0.143 0.129 0.118 0.110 0.102 0.0958 0.0904 0.0857 0.0815 0.0779 0.0746 0.0716 0.0690 0.0665 0.0643
0 2 4 6 8 10 12 14 16 18 20 24 28 32 36 40 44 48 52 56 60 64 68 72 76 80 84 88 92 96 100 120 140 160 180 200 220 240 260 280 300
30 97 141 177 206 235 261 288 315 342 368 417 466 513 558 603 644 685 725 762 800 837 873 908 945 981 1016 1052 1087 1122 1155 1307 1463 1608 1746 1880 2008 2132 2256 2372 2495
3.97 1.23 0.844 0.672 0.578 0.506 0.456 0.413 0.378 0.348 0.323 0.285 0.255 0.232 0.213 0.197 0.185 0.174 0.164 0.156 0.149 0.142 0.136 0.131 0.126 0.121 0.117 0.113 0.109 0.106 0.103 0.0910 0.0813 0.0740 0.0682 0.0633 0.0593 0.0558 0.0527 0.0502 0.0477
1000 ‘P-P’R, 1000 Manchester system roentgens; IRAK, integrated reference air-kerma. * Original Manchester values from Paterson, R, Parker HM: A dosage system for interstitial radium therapy. Br J Radiol 11:313–339, 1938. † Modified from original values for 192Ir assuming 8.6 Gy minimum peripheral dose per 1000 ‘P-P’R and 7.227 µGy·m2/mgRaeq·h. (Reprinted from Williamson, J. F. “Clinical Brachytherapy Physics” in Principles and Practice of Radiation Oncology, 3rd ed. C. A. Perez, L. W. Brady, E. C. Halperin, and R. K. Schmidt-Ullrich (eds.). © 1998 with permission from Lippincott Williams & Wilkins.)
452
Zuofeng Li Table 5. Table for Tandem and Ovoid Plan Manual Calculation Check Used at Mallinckrodt Institute of Radiology. HDR Treatment Dwell Times Are Scaled to Achieve Identical Loading Patterns. Applicator Component
Small tandem Medium tandem Standard tandem Endometrial tandem 2.0 cm colpostats* 2.5 cm colpostats* 3.0 cm colpostats* Mini-ovoids (1.6 cm colpostats)
Dose Rate (cGy·h-1) per mgRaeq+
Loading (mgRaeq)
Point A
Point B
Point P
20 10-20 10-10-20 10-20-10 20-20 25-25 30-30 10-10
1.545 1.543 1.070 1.308 0.553 0.474 0.418 0.660
0.295 0.297 0.260 0.278 0.250 0.244 0.228 0.255
0.205 0.207 0.185 0.195 0.183 0.182 0.173 0.190
Multiply above numbers by total mgRaEq in each component. + 3M 137Cs tubes, 1.4 cm active length. No correction for decay. *Includes 6% applicator attenuation correction.
Modification of the Sievert integral for HDR treatments is necessary to account for significant variations of source dwell times. In a typical HDR optimized single catheter treatment plan, the dwell times at the ends of the catheters are often significantly higher than those in the middle section of the catheter, while those in the middle section remain mostly constant. The Sievert integral, in the form above, can then be combined with point source dose calculations for the dwell positions at the ends of the catheters, to arrive at a dose estimate often within 5% of the TPS calculations. Rogus et al. (Rogus, Smith, and Kubo 1998) investigated the relationship of catheter length and treatment distance for HDR single catheter treatments. Assuming straight-line or moderately curved catheters, the authors proposed a fitting equation for the total treatment time t as shown below:
(
)
t ( d , L )ref = −1.35 + 7.74 d + 0.322 d 2 +
L − 50 50
( −0.591 + 6.92 d + 0.0230 d ) , 2
(3)
for the reference condition of a prescription of 500 cGy and a source strength of 10 Ci, where d is the distance of prescription point away from the catheter, and L is the active length of the catheter. This equation is simply scaled to apply to other prescription doses and source strengths. Ezzell (2000a) studied the influence of catheter curvature on dose calculation accuracy based on straight-line assumption such as outlined above, and concluded that the ratio of the distance between end dwell positions (the cord) and the active length of a single, curved HDR treatment catheter can be used as a measure of the catheter curvature. Figure 9 shows the cord-to-active-length ratio required to maintain dose homogeneity within 10% (Ezzell 2000a). For treatments using two catheters, Miller et al. (Miller, Davis, and Horton 1996) and Ezzell (2000b) described methods to verify their treatment times. Miller et al. eliminated 50% of the active dwell positions located in areas where the two catheters overlap, such that the two catheters can be considered independently as if no overlap occurs. Ezzell’s formalism actually calculates the expected dwell time for each dwell positions, based on the distance of a dwell position on a given catheter to the second catheter.
23–QA Review of Brachytherapy Treatment Plans
453
Figure 9. Cord length relative to active length required to maintain 10% homogeneity for a single, curved catheter. (Reprinted from Ezzell, G. A. (2000). “Limitations of the straight-line assumption for endobronchial HDR brachytherapy treatments.” Med Phys 27(1):151–153. © 2000 with permission from AAPM.)
A sample Microsoft Excel spreadsheet is available that can be used for performing these calculations, available at ftp://ftp.aip.org/epaps/medical_phys/E-MPHYA6-27-008005/example_worksheets.pdf.
Independent Calculation Check of Planar Interstitial Implants In general, the Paterson-Parker system applies well to the independent calculation check of planar implants. The width and the length of the implant can be estimated from orthogonal radiographs of the implant, or measured from a TPS display, with the orientation of the implant rotated to facilitate the measurement. The area of the implant thus measured can then be used for table lookup on a standard Paterson-Parker planar implant table, such as presented by Williamson (2003). The resulting mgRaeq·hrs for the implant should then be corrected for the elongation factor of the implant, and compared to the mgRaeq·hrs calculated by the brachytherapy TPS for the implant, following appropriate unit conversion. It is important to keep in mind that the original Paterson-Parker system’s prescription dose for planar and volume implants is the so-called “modal-dose,” or the dose at 10% higher the minimum dose in the surface that contains the dose prescription point. If the intention is to cover the implant volume by at least the prescription dose, the mgRaeq·hrs looked up from such a table would need to be increased by 10% to be comparable with TPS calculations. Ezzell (1994) reviewed 66 rectangular HDR planar implants, and described a method to estimate the total treatment time required to deliver a prescription dose from the planar implant. Let I be the Dose area index, defined as I=
Dose × area Source strength × Total time
,
(4)
454
Zuofeng Li
then I = A(T) + B(T) × E + C(T) × E2 ,
(5)
where T is the thickness of treatment, E is the length of the equivalent square of implant. For source strength specified in cGy cm2 s-1, Ezzell gave the calculations of the fitting coefficients A, B, and C as follows: A = 3.245 – 1.269 × T + 0.1014 × T2 B = 1.030 – 0.0728 × T
(6)
C =-0.02083 + 0.001925 × T . Testing of this method on real patients showed that it mostly predicts the total treatment time to within 10%.
Independent Calculation Check of Volume Interstitial Implants Volume interstitial implants can typically be easily checked by use of the Paterson-Parker tables (Williamson 2003). Das et al. (2004) reported this use for accelerated partial breast brachytherapy treatments, where better than 7% agreement was observed when corrections were made for elongation factor, and the original Paterson-Parker table was corrected for modern units. The volume of tissue treated to the prescription dose is looked up on the DVH plot calculated by the HDR TPS. When a DVH is not available, the implanted volume can be estimated by measuring the width, length, and height of the volume bounded by the implanted catheters. The Paterson-Parker system was designed for radium sources, which are minimally attenuated in tissue. For low-energy sources such as 125I and 103Pd, used for permanent prostate seed implant, Paterson-Parker system is no longer applicable. Cohen et al. (2002) reported the use of nomograms for the independent calculation check of prostate seed implants. Let davg be the average distances between pre-planned needles/seeds in the lateral, anterior-posterior and superior-inferior directions; the authors reported the following nomograms: davg + 0.8 = 1.524 1.09 U cm Sk
2.2
(7)
for 125I seed implants with a prescription dose of 144 Gy, treated to the prostate volume with a 5 mm margin in all directions except the posterior, and davg + 0.8 = 5.395 1.09 U cm Sk
2.56
(8)
for 103Pd seed implants with a prescription of 140 Gy. Compared with computerized treatment plans generated using a genetic algorithm, the authors reported agreement of better than 10% in the total activity required.
23–QA Review of Brachytherapy Treatment Plans
455
Conclusions This chapter describes briefly the goals and methods in the QA review of brachytherapy treatment plans. As modern brachytherapy treatment plans become more complex, and as the use of brachytherapy becomes more widespread, physicists are challenged constantly to come up with efficient and accurate methods to quickly evaluate the quality of a brachytherapy treatment plan. The physicist reviewing a treatment plan needs to remain vigilant in performing this important task.
References Anderson, L. L. (1976). “Spacing nomograph for interstitial implant of 125I seeds.” Med Phys 3(1):48–51. Anderson, L. L. (1986). “A ‘natural’ volume-dose histogram for brachytherapy.” Med Phys 13:898–903. Anderson, L. L. (1995). “Dose Specification and Quantification of Implant Quality” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (ed.). Madison, WI: Medical Physics Publishing, 1995. Anderson, L. L., J. V. Moni, and L. B. Harrison. (1993). “A nomograph for permanent implants of palladium-103 seeds.” Int J Radiat Oncol Biol Phys 27(1):129–135. Arthur, D. W., F. A. Vicini, R. R. Kuske, D. E. Wazer, and S. Nag. (2003a). “Accelerated partial breast irradiation: An updated report from the American Brachytherapy Society.” Brachytherapy 2(2):124–130. Arthur, D. W., D. E. Wazer, D. Koo, L. Berle, L. Cuttino, M. Yunes, M. Rivard, D. Todor, S. Tong, T. Tenenholz, and T. DiPetrillo. (2003b). “The importance of dose volume histogram evaluation in partial breast brachytherapy: A study of dosimetric parameters.” Int J Radiat Oncol Biol Phys 57(2):S361–S362. Beyer, D., R. Nath, W. Butler, G. Merrick, J. Blasko, S. Nag, and C. Orton. (2000). “American brachytherapy society recommendations for clinical implementation of NIST-1999 standards for (103)palladium brachytherapy.” Int J Radiat Oncol Biol Phys 47(2):273–275. Cohen, G. N., H. I. Amols, and M. Zaider. (2000). “An independent dose-to-point calculation program for the verification of high-dose-rate brachytherapy treatment planning.” Int J Radiat Oncol Biol Phys 48(4):1251–1258. Cohen, G. N., H. I. Amols, M. J. Zelefsky, and M. Zaider. (2002). “The Anderson nomograms for permanent interstitial prostate implants: A briefing for practitioners.” Int J Radiat Oncol Biol Phys 53(2):504-511. Das, R. K., R. Patel, H. Shah, H. Odau, and R. R. Kuske. (2004). “3D CT-based high-dose-rate breast brachytherapy implants: treatment planning and quality assurance.” Int J Radiat Oncol Biol Phys 59(4):1224–1228. Edmundson, G. “Geometry Based Optimization for Stepping Source Implants” in Brachytherapy HDR and LDR. A. A. Martinez, C. G. Orton, and R. F. Mould (eds.). Columbia, MD: Nucletron Corporation, 1990. Ezzell, G. A. (1994). “Quality assurance of treatment plans for optimized high dose rate brachytherapy—planar implants.” Med Phys 21(5):659–661. Ezzell, G. A. (2000a). “Limitations of the straight-line assumption for endobronchial HDR brachytherapy treatments.” Med Phys 27(1):151–153. Ezzell, G. A. (2000b). “A manual algorithm for computing dwell times for two-catheter endobronchial treatments using HDR brachytherapy.” Med Phys 27(5):1030–1033. Gaspar, L. E., S. Nag, A. Herskovic, R. Mantravadi, and B. Speiser. (1997). “American Brachytherapy Society (ABS) consensus guidelines for brachytherapy of esophageal cancer.” Int J Radiat Oncol Biol Phys 38(1):127–132. Grigsby, P. W., J. F. Williamson, and C. A. Perez. (1992). “Source configuration and dose rates for the Selectron afterloading equipment for gynecologic applicators.” Int J Radiat Oncol Biol Phys 24(2):321–327. International Commission on Radiation Units and Measurements (ICRU). Report 58: Dose and Volume Specifications for Reporting Interstitial Therapy. Bethesda, MD: ICRU, 1997. Jhingran, A., and P. J. Eifel. (2000). “Perioperative and postoperative complications of intracavitary radiation for FIGO stage I-III carcinoma of the cervix.” Int J Radiat Oncol Biol Phys. 46(5):1177–1183. Katz, A., and P. J. Eifel. (2000). “Quantification of intracavitary brachytherapy parameters and correlation with outcome in patients with carcinoma of the cervix.” Int J Radiat Oncol Biol Phys 48(5):1417–1425. Lachaine, M. E., J. C. Gorman, and M. G. Palisca. (2003). “A fast, independent dose check of HDR plans.” J Appl Clin Med Phys 4(2):149–155.
456
Zuofeng Li
Mayo, C. S., and K. Ulin. (2001). “A method for checking high dose rate treatment times for vaginal applicators.” J Appl Clin Med Phys 2(4):184–190. Miller, A. V., M. G. Davis, and J. L. Horton. (1996). “A method for verifying treatment times for simple high-doserate endobronchial brachytherapy procedures.” Med Phys 23(11):1903–1908. Neblett, D. L., A. M. N. Syed, A. A. Puthawala, R. Harrop, H. S. Frey, and S. E. Hogan. (1985). “An interstitial implant technique evaluated by contiguous volume reference.” Endocuriether/Hypertherm 1:213–221. Nag, S. (2000). “Brachytherapy for prostate cancer: summary of American Brachytherapy Society recommendations.” Semin Urol Oncol 18(2):133–136. Nag, S., A. A. Abitbol, L. L. Anderson, J. C. Blasko, A. Flores, L. B. Harrison, B. S. Hilaris, A. A. Martinez, M. P. Mehta, and D. Nori. (1993). “Consensus guidelines for high dose rate remote brachytherapy in cervical endometrial, and endobronchial tumors.” Int J Radiat Oncol Biol Phys. 27(5):1241–1244. Nag, S., P. E. Cole, I. Crocker, S. K. Jani, K. V. Kishnasastry, V. Massullo, R. Nath, D. Nori, S. Parikh, P. Rubin, B. Speiser, P. S. Teirstein, P. Tripuraneni, R. Waksman, and J. F. Williamson. (1999a). “The American Brachytherapy Society perspective on intravascular brachytherapy.” Cardiovasc Radiat Med 1(1):8–19. Nag, S., D. Beyer, J. Friedland, P. Grimm, and R. Nath. (1999b). “American Brachytherapy Society (ABS) recommendations for transperineal permanent brachytherapy of prostate cancer.” Int J Radiat Oncol Biol Phys 44(4):789–799. Nag, S., W. Bice, K. DeWyngaert, B. Prestidge, R. Stock, and Y. Yu. (2000a). “The American Brachytherapy Society recommendations for permanent prostate brachytherapy postimplant dosimetric analysis.” Int J Radiat Oncol Biol Phys 46(1):221–230. Nag, S., B. Erickson, B. Thomadsen, C. Orton, J. D. Demanes, and D. Petereit. (2000b). “The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for carcinoma of the cervix.” Int J Radiat Oncol Biol Phys 48(1):201–211. Nag, S., B. Erickson, S. Parikh, N. Gupta, M. Varia, and G. Glasgow. (2000c). “The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for carcinoma of the endometrium.” Int J Radiat Oncol Biol Phys 48(3):779–790. Nag, S., R. R. Kuske, F. A. Vicini, D. W. Arthur, and R. D. Zwicker. (2001a). “Brachytherapy in the treatment of breast cancer.” Oncology (Huntingt) 15(2):195–205. Nag, S., D. Shasha, N. Janjan, I. Petersen, and M. Zaider, (2001b). “The American Brachytherapy Society recommendations for brachytherapy of soft tissue sarcomas.” Int J Radiat Oncol Biol Phys 49(4):1033–1043. Nag, S., E. R. Cano, D. J. Demanes, A. A. Puthawala, and B. Vikram. (2001c). “The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for head-and-neck carcinoma.” Int J Radiat Oncol Biol Phys 50(5):1190–1198. Nag, S., C. Chao, B. Erickson, J. Fowler, N. Gupta, A. Martinez, and B. Thomadsen. (2002). “The American Brachytherapy Society recommendations for low-dose-rate brachytherapy for carcinoma of the cervix.” Int J Radiat Oncol Biol Phys 52(1):33–48. Nag, S., J. M. Quivey, J. D. Earle, D. Followill, J. Fontanesi, and P. T. Finger. (2003a). “The American Brachytherapy Society recommendations for brachytherapy of uveal melanomas.” Int J Radiat Oncol Biol Phys 56(2):544–555. Nag, S., R. Dobelbower, G. Glasgow, G. Gustafson, N. Syed, B. Thomadsen, and J. F. Williamson. (2003b). “Intersociety standards for the performance of brachytherapy: A joint report from ABS, ACMP and ACRO.” Crit Rev Oncol Hematol 48(1):1–17. Paterson, R., and H. M. Parker. (1938). “A dosage system for interstitial radium therapy.” Br J Radiol 11:313–339. Potter, R., E. Van Limbergen, W. Dries, Y. Popowski, V. Coen, C. Fellner, D. Georg, C. Kirisits, P. Levendag, H. Marijnissen, H. Marsiglia, J. J. Mazeron, B. Pokrajac, P. Scalliet, and V. Tamburini. (2001). “Prescribing, recording, and reporting in endovascular brachytherapy. Quality assurance, equipment, personnel and education.” Radiother Oncol 59(3):339–360. Rogus, R. D., M. J. Smith, and H. D. Kubo. (1998). “An equation to QA check the total treatment time for singlecatheter HDR brachytherapy.” Int J Radiat Oncol Biol Phys 40(1):245–248. Saw, C. B., and N. Suntharalingam. (1991). “Quantitative assessment of interstitial implants.” Int J Radiat Oncol Biol Phys 20(1):135–139.
23–QA Review of Brachytherapy Treatment Plans
457
Saw, C .B., L. J. Korb, B. Darnell, K. V. Krishna, and D. Ulewicz. (1998). “Independent technique of verifying highdose rate (HDR) brachytherapy treatment plans.” Int J Radiat Oncol Biol Phys 40(3):747–750. Stitt, J. A., J. F. Fowler, B. R. Thomadsen, D. A. Buchler, B. P. Paliwal, and T. J. Kinsella. (1992). “High dose rate intracavitary brachytherapy for carcinoma of the cervix: The Madison system: I. Clinical and radiobiological considerations.” Int J Radiat Oncol Biol Phys. 24(2):335–348. Stock, R. G., N. N. Stone, M. Dahlal, and Y. C. Lo. (2002). “What is the optimal dose for 125I prostate implants? A dose-response analysis of biochemical control, posttreatment prostate biopsies, and long-term urinary symptoms.” Brachytherapy 1(2):83–89. Thomadsen, B. R., S. Shahabi, J. A. Stitt, D. A. Buchler, J. F. Fowler, B. R. Paliwal, and T. J. Kinsella. (1992). “High dose rate intracavitary brachytherapy for carcinoma of the cervix: The Madison system: II. Procedural and physical considerations.” Int J Radiat Oncol Biol Phys 24(2):349–357. Thomadsen, B. R. (1999). Achieving Quality in Brachytherapy. Bristol: Institute of Physics, 1999. Wang, X. H., and L. Potters. (2001). “A theoretical derivation of the nomograms for permanent prostate brachytherapy.” Med Phys 28(4):683–687. Williamson, J. F. “Clinical Brachytherapy Physics” in Principles and Practice of Radiation Oncology, 4th ed. C. A. Perez, L. W. Brady, E. C. Halperin, and R. K. Schmidt-Ullrich (eds.). Philadelphia: Lippincott Williams & Wilkins, 2003. Yu, Y., L. L. Anderson, Z. Li, D. E. Mellenberg, R. Nath, M. C. Schell, F. M. Waterman, A. Wu, and J. C. Blasko. (1999). “Permanent prostate seed implant brachytherapy: Report of the American Association of Physicists in Medicine Task Group No. 64.” Med Phys 26(10):2054–2076. Also available as AAPM Report No. 68.
Chapter 24
Quality Management for Interstitial Implants Anil Kumar Sharma, Ph.D. Radiation Oncology Department Long Beach Memorial Medical Center Long Beach, California Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 459 Permanent Interstitial Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461 Temporary Interstitial Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461 Quality Management for the Implantation Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461 Template-based Implants Using Rigid Guides or Needles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 461 Prostate Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 462 Gynecological Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463 Template-based Breast Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 463 Multiple Site Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464 Flexible Catheter Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464 Breast/Chest Wall Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464 Head and Neck Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 465 Other Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466 Practical Considerations of Implant Reconstruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466 Templates and Rigid Needles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466 HDR Prostate Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467 Gynecological Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 467 Flexible Catheters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468 Breast Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468 Head and Neck Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 468 Quality Management of Treatment Planning and Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . 468 Prostate Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469 Gynecological Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 470 Head and Neck Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471 Quality Assurance of Source Application and Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471 Interstitial Brachytherapy Treatments with LDR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 471 Interstitial Brachytherapy Treatments with HDR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 472 Needle or Catheter Movements and Possible Corrections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473
Introduction The aim of quality management in brachytherapy is to ensure that each individual radioactive source application to the area of interest is administered accurately according to the radiation oncologist’s intent, consistent with the written directives and within the scope of the institution’s radioactive material license. Also, the procedure followed for the treatment should be safe for the patient, staff, and others who may be exposed to radiation during the course of treatment. In order to provide satisfactory brachytherapy services to the patient with full compliance of the federal and state regulations, an institution is required to develop a quality management program (QMP), which becomes an essential part of their brachytherapy program. The federal and state regulations are meant for the safety of the patient, personnel, and general public; and without making any clinical judgments relative to the patient’s treatment, they help define the standard of practice for all those involved in delivering brachytherapy services. They form the basis of a necessary, but not sufficient, quality assurance (QA)
459
460
Anil Kumar Sharma
program in brachytherapy. The regulatory agencies, like Nuclear Regulatory Commission (NRC) or the agreement states’ Radiological Health Branch, require that the institution’s practices assure that the radiation from byproduct material will be administered as directed by the authorized user. So, like all other medical treatments, the prescription for interstitial brachytherapy (written directive) is an order in writing for a specific patient, dated and signed by the authorized user prior to the administration of radiation by brachytherapy. For low dose rate (LDR) applications, this prescription must contain the treatment site, name of the radioisotope, number of sources and source strength, and prior to completion of treatment, time of implant or total dose. For high dose rate (HDR) brachytherapy, prior to the treatment, treatment site, isotope, and total dose must be mentioned in the written directive (USNRC 1991a). The objectives of federally enforced QMP (the term “QMP” is dropped in the current part 35, but the requirements are retained) are all related to the written directive as follows: • Written directive before administration of radiation • Identification of the patient before administration of radiation (by more than one way) as the individual named in the written directive • Final plan and related calculations in accordance with the written directive • Administration of radiation in accordance with the written directive • Identification and evaluation of any unintended deviation from the written directive and implementation of appropriate action. NRC requires annual review of the actual treatment with respect to the written directive for a number of patients based upon statistical sampling principles. Also, NRC regulatory guide 8.33, Task DG-8001, contains requirements and recommendations of a QA nature to implement the QMP (USNRC 1991b). The recommendations in this guide can be adapted by individual licensees to arrive at a QMP suitable for their practice and needs, the net effect of which is to emphasize QA in brachytherapy. Besides NRC’s and the agreement states’ limited requirements and recommendations, detailed QA recommendations have come from several organizations. The brachytherapy section of the AAPM’s Task Group 40 (TG 40) (Kutcher et al. 1994) report, discusses both LDR and HDR brachytherapy QA in detail. For HDR brachytherapy practice, AAPM’s TG 59 (Kubo et al. 1998) is a comprehensive guide with several recommendations. Interstitial Collaborative Working Group (ICWG) of the American Brachytherapy Society (ABS) in their report (ICWG 1990) recommended source calibration procedures and other QA requirements in brachytherapy. From time to time ABS has come up with recommendations for several brachytherapy procedures (Nag et al. 2000, 2001, 2002). A major focus of the QA program is to assure accurate operation of all software and mechanical and radioactive devices used for planning, delivery, or QA of brachytherapy treatments. Each institution practicing interstitial brachytherapy should develop a written, periodic QA protocol defining the tests to be performed and their frequency (Nath et al. 1997). Most references cited in the previous paragraph deal with the QA program related to radiation safety in brachytherapy, source calibration, safe handling and application of radioisotopes either manually or through remote-controlled devices, QA of these units and source calibration devices, and QA of treatment planning systems. The goal of periodic device QA is to ensure that the operating characteristics of the device remain unchanged with time. The other equally important part of a brachytherapy QMP is the procedure-specific quality assurance, which is defined as a set of actions selected to ensure that each important step leading to delivery of a brachytherapy procedure is correctly carried out. In this chapter, quality management of interstitial brachytherapy will be discussed with special emphasis on this procedure-specific QA for manual and remote afterloaded interstitial implants. Interstitial brachytherapy, which is the direct application of radioactive isotopes into the tumor, can broadly be divided into two categories: permanent and interstitial brachytherapy.
24–Quality Management for Interstitial Implants
461
Permanent Interstitial Brachytherapy In permanent interstitial brachytherapy the radioactive sources are directly implanted into the tumor and left there permanently to decay. Normally, such applications are carried out in the operating room after careful preplanning. Permanent seed implants of the prostate, pancreas, nasopharynx, unresectable and unsuitable tumors of the head and neck, and stereotactic brain implants using 125I or 103Pd fall into this category. The radioactive sources are transported to the operating room (OR) and flashed and implanted by the radiation oncologist according to the extent of the tumor and safety of the surrounding structures, following the preplan carried out prior to the actual procedure. The number of sources used for such implants depends on the type of the radioactive isotope, size of the tumor, total dose to be delivered, and the initial activity of the source. Almost invariably all sources in permanent implants have the same air kerma strength. Quality management of such implants is dealt with elsewhere and will not be discussed in this chapter.
Temporary Interstitial Brachytherapy In temporary interstitial brachytherapy the radioactive sources are either directly implanted into the tumor or are afterloaded into a set of catheters, which may be rigid needles (metallic, ceramic or plastic) or flexible tubes (nylon or plastic) arranged in an appropriate geometrical pattern with an intention to cover the whole tumor while sparing surrounding normal tissue. In temporary implants, after an adequate amount of dose is delivered to the tumor, the sources are removed from the tissue or from the catheters into which they were sent. The total time of the implantation depends on the type of source, number and strength of each source, and the geometric pattern in which they are arranged. Most interstitial implants in the early days of brachytherapy were direct implantation of radium needles, which subsequently got replaced by 137Cs needles in late 1970s. Today, most interstitial implants are catheter based (rigid needles or flexible tubes) into which small radioactive sources (mostly 192Ir) are afterloaded either manually or through a remote afterloading machine. Quality management of such afterloaded implants is the main topic of this chapter.
Quality Management for the Implantation Procedure Interstitial brachytherapy techniques vary according to the site to be implanted, tumor volume, and whether the target is approachable from multiple sides or from just one side. Usually, the sites that can be approached from only one side are implanted with rigid guides using templates, whereas others can be implanted with flexible catheters with or without the aid of a template. Templates not only help in guiding the catheters into the target at the time of implantation, but also help in keeping the implant in position during the course of treatment. Given the wide variety of interstitial procedures performed in any institution, it is understandable that the philosophy of maintaining quality of the implantation is different for different types of applications. Selection of the type of applicators, catheters, or needles for placement in the patient’s tumor is under the control of the radiation oncologist in consultation with the medical physicist. Physics QA duties include documentation of the catheters or needles inserted and their correct correlation with target volume.
Template-based Implants Using Rigid Guides or Needles Template-based implants with rigid guides are very common for the treatment of prostate, rectum, breast, and gynecological tumors. Templates for such applications come in various shapes and sizes and can be re-useable or disposable after one application. It is very important to decide, before the team heads for the OR, which template is most suitable for the patient and what kind of guides are required for the proce-
462
Anil Kumar Sharma
dure. For example, if the treatment is planned to be given by LDR 192Ir ribbons, the set of needles required is different from those used for HDR brachytherapy. Also, if interstitial radiofrequency hyperthermia is part of the treatment strategy, steel needles need to be implanted and used as antennae for hyperthermia. In such situations, the number of implanted needles will be more than required just for brachytherapy alone because for better uniformity of hyperthermia, the desired pattern of needles may be different. Therefore, a checklist of items required for the implantation procedure is the first step of a QMP. This list may vary with the site to be implanted. Some examples follow. Prostate Implants Syed-Neblett prostate template and the Martinez Universal Perineal Interstitial Template (MUPIT) are most commonly used for the treatment of carcinoma of the prostate. Whereas the MUPIT is essentially a grid of allowable guides in a square pattern with holes at every 0.5 cm interval, the Syed-Neblett prostate template has a provision for introducing a maximum of 18 needles arranged in two concentric circles. Occasionally, in order to encompass a very large prostate, templates designed for rectal applications can be used. The rectal template (also called the large prostate template) has an additional outer concentric circle of 3-cm radius around the inner two circles (of radii 1 and 2 cm) of a regular prostate template. These templates are designed to hold 17-gauge needles perpendicular to the template, but in some cases the needles can either be flared out or coned in to fit onto the target (Syed et al. 1992, 1997). The number of needles used is usually fewer than 18 in most prostate cases, most often from 14 to 18. Pre-procedure volume studies are carried out to determine the gland size, any extracapsular extension, calcification, median lobe involvement, seminal vesicle involvement, transurethral resection of the prostate (TURP) etc., so that the appropriate size template and strategy for treatment can be chosen. A three-way Foley balloon, with 50% Hypaque™ (contrast) solution in the balloon is required for contrast injection directly into the bladder before planning computed tomography (CT) is done. Before implantation of needles under transrectal ultrasound (TRUS) guidance, two gold markers are inserted into the base and two into the apex of the prostate. These markers can be used in applicator-based planning. Quality management for prostate implantation would include the following practical considerations: • Determination of the type of implant (LDR or HDR), so that appropriate needles can be chosen for implantation. LDR needles cannot be connected to the afterloader because only HDR needles have special features for connection to the afterloader. • Verification of the template orientation at the time of implantation. If the template is unintentionally rotated by 90°, the configuration of inner circle (urethral circle) becomes different from the usual convention, resulting in confusion at several stages, from numbering the channels to loading with a set of dummy trains to treatment planning and finally to the actual treatment, because the written procedure for each of these stages is meant for conventional position of the template. The template is in its correct orientation when two holes in the first circle coincide with the 12 o’clock and 6 o’clock positions. • Verification of the proper functioning of the ultrasound unit, and its calibration with respect to the template (if it is a transducer fixed system). • Post procedure notes in the patient’s chart by the physician, the diagram of the implant with written directives clearly spelling out the total dose, dose rate for LDR, or dose per fraction for HDR. • Sending the completed implant loading form to the brachytherapy dosimetry team in radiation oncology. Recently, two commercial systems, SWIFT™ from Nucletron Corporation (Columbia, MD) and Vitesse™ from Varian Medical Systems (Palo Alto, CA) have become available for ultrasound image-
24–Quality Management for Interstitial Implants
463
based implantation and dosimetry. For such systems extensive QA of the ultrasound imaging system and calibration of the template for one-to-one correspondence with the template in the ultrasound image is required on the same lines as for the permanent prostate seed implants. For details, QA of permanent seed implant devices may be referred to elsewhere in this monograph. Gynecological Implants The most used template systems for interstitial implantation, Syed-Neblett GYN and MUPIT, were introduced in 1974. Both systems use 17-gauge needles, but differ in their pattern of allowable guides. MUPIT allows needles in its type I holes at 1-cm intervals and perpendicular to the template, and in type II holes at 1.25-cm intervals at a 13° laterally outward direction (Martinez, Cox, and Edmundson 1984). The vaginal and rectal cylinders, available in two sizes, compose the intracavitary component of the system, and can accommodate an intrauterine tandem or drainage tube. Each of the cylinders can also accommodate up to eight needles. MUPIT can be used for both LDR and HDR applications. Commercialized templates are available through Nucletron Corporation for HDR applications. The Syed-Neblett template (Syed et al. 1986; Neblett 1995) for gynecological malignancies has a pattern of holes (1 cm apart) in five concentric circles or arcs at 1-cm radial increments and a central grooved vaginal obturator of 2 cm diameter and 15 cm length, which can accommodate six needles and a centrally located interauterine tandem of medium curvature. These templates are available as reuseable or disposable from Alpha Omega Services, Bellflower, CA, and Best Industries, Springfield, VA. All LDR needles used with these templates except the first needle (also called “guide”) have small, raised metal rings near the proximal end to prevent them from sliding through the template. HDR needles also are shaped at their proximal end for connection to an afterloading system. The implantation procedure is discussed in detail elsewhere in this monograph. The following practical considerations form part of the quality management program: • Choice of template (MUPIT or Syed-Neblett GYN 1, 2 or 3) needs to be made after careful evaluation of the pre-procedure imaging study. CT and particularly magnetic resonance (MR) images are extremely helpful in defining the volume to be implanted, and providing knowledge about the central and peripheral disease. If the disease is extensive in the anterior-posterior (AP) direction, choice of GYN 1 or 2 templates will lead to undercoverage of the target. • Type of implant (LDR or HDR) must be determined, so that appropriate needles can be chosen. Also, from the imaging studies, the depth of needles can be determined and whether the vaginal length requires the long obturator and longer needles (25 cm). • Implantation of gold markers into the cervical os requires an accuracy of better than 2 mm, since these markers are used for applicator-based planning. • Even though insertion of all the allowable needles is not necessary and would depend on the extension of disease, enough needles (usually more than 10 in either side of the obturator) need to be implanted to provide stability and rigidity to the template. • In LDR implants, obturator needle insertion is discouraged in the presence of an intrauterine tandem; but for HDR implants where dwell-time optimization is possible, implantation of four lateral obturator needles is found to be helpful in adequate coverage of centrally bulky disease. These needles are now regularly implanted in HDR applications, but their loading is decided at the time of planning. Template-based Breast Implants Conventionally, freehand breast interstitial implants have been performed as boost using either 192Ir ribbons or 125I seeds for LDR implants or for afterloading HDR fractionated treatments using 192Ir. Recently, partial
464
Anil Kumar Sharma
breast irradiation by interstitial implants has shown its potential for selected patients with early stage disease (Vicini et al. 1999; Baglan et al. 2001). Implants are carried out using a standardized template, with interplane and intercatheter separation of about 1.5 cm. Intraoperative ultrasound-guided placement of afterloading catheters can also be used to ensure that the catheters encompass the biopsy cavity. Detailed description of breast brachytherapy is covered elsewhere in this monograph. Quality management for the template-based breast implants will include proper choice of the template and ensuring that it is secure at the time of implantation so that the needles are arranged with proper separation as per the pre-plan. If ultrasound guidance is used, proper QA of the ultrasound unit is required. Multiple Site Implants Multiple site implants in close proximity are usually mixed implants, i.e., a template-based implant and a freehand flexible catheter implant or an intracavitary application. Examples of the former are vulvar implants and breast and axillary lymph node implants, whereas interstitial-intracavitary cervical and some extensive maxillary implants fall in the latter category. Additional considerations about the possible overlap of two implants need to be taken into account. Quality management of flexible implants is discussed in the next section and its extent in the presence of a rigid implant requires clear demarcation of the boundaries of each implant, at least externally, before the start of the implantation process. If there is any intracavitary application within or close to the rigid needle implant, securing the intracavitary applicator in place within 2 mm is required.
Flexible Catheter Implants Several sites are easily accessible from at least two sides for freehand flexible catheter implants, which are normally performed in the OR under general anesthesia, except for breast implants, where local anesthesia may be used instead. General considerations for such implants are to ensure the patency of the implanted catheters, availability of sterile (angled) sharp needles (trocars), buttons (plastic buttons of different colors, metallic buttons to be used as stoppers, and gold buttons for shielding), appropriate catheters, sterile rulers, and marking pens. Breast/Chest Wall Implants Either a single-channel device, called MammoSite®, or multiple catheters implanted interstitially are used for breast implants. Quality assurance features of these implants are very different from each other. Single-Channel Breast Implants (MammoSite®) MammoSite applicator (Proxima Therapeutics Inc., Alpharetta, GA) is a single-channel applicator with a silicon inflatable balloon at its distal end and two ports; one for HDR source application and the other for injecting fluid into the balloon, on its proximal end. This device can either be implanted at the time of lumpectomy or later within a few weeks of lumpectomy. This subject is dealt with in depth elsewhere in this monograph. Before implantation, ensuring that the there is no leakage in the balloon is an important QA step. If the cavity size is not known, at least two sizes of the spherical/elliptical applicators should be taken to the OR. A small diameter catheter, such as a baby NG tube, may be implanted side by side with the MammoSite applicator, which may be used later to get rid of the air pockets. If implanted a few weeks after lumpectomy, ultrasound guidance is required in identifying the cavity. The ultrasound unit should be calibrated before its use. Ensuring that the applicator is properly secured in its place and the HDR port has the flexible stylet in it are part of the QA for this procedure.
24–Quality Management for Interstitial Implants
465
Freehand Planar Breast Implants After marking the intended entry and exit points of these needles on the breast surface, 17-gauge stainless steel trocars are inserted. Most often two or three planar implants are carried out, interdigitating needles in the neighboring planes. • Interplanar distances and intercatheter distances should follow Paris or Manchester system rules for LDR implants for 192Ir wires and ribbons, respectively. For HDR implants, this spacing, though desirable, is not very crucial. • Catheters should be secured by the buttons at both ends, so that the unintended movement of catheters is no more than 2 mm. • Catheters of each plane should be of different colors or at least a different color button should be used for securing them. • Flexible metallic or plastic stylets should be inserted after the implant is done to ensure patency of the catheters. • Single-plane chest wall implants require careful spacing to avoid hot and cold spots in the implant. Head and Neck Implants Using flexible afterloading plastic catheters, previously unimplantable lesions involving retromolar trigone, soft palate, tonsellar pillars, phrangeal walls, and base of tongue can be effectively implanted (Puthawala et al. 1986; Syed and Puthawala 1996). Planar Implants Mostly used for neck nodes, these planar implants have the same kind of requirements as mentioned above for the freehand breast implants. Gold Button/Button Technique This technique is used in most of the oral cavity and oropharyngeal cancers. In this technique, the plastic tubes are held between a gold and plastic button at its sealed end and a plastic button on the skin. The gold button protects the overlying tissues against irradiation and trauma. Loop Technique/Arch Technique This technique is used for cancers encroaching on the mandible and occasionally for base of tongue lesions. The technique involves insertion of one plastic tube through the intraoral ends of a pair of hollow needles implanted on either side of the mandible. When the needles are removed, the plastic tube is wrapped over the alveolar ridge. This technique is also used for lesions involving the soft and hard palate and tonsillar pillars. In this technique, one plastic tube is inserted through the intraoral ends of a pair of hollow needles implanted from either side of the neck or cheek through the palate. In most oral cavity and oropharyngeal cancers, a combination of various techniques is used to encompass the tumor adequately. Multiple Site Implants It is a very common practice in head and neck brachytherapy that multiple sites are implanted at the same time. Various combinations of techniques and sites could be very confusing for the implanter as well as for the planner. Documentation of each site with the number of catheters and the technique used for
466
Anil Kumar Sharma
implantation should be made. Whereever the catheters from one site overlap into the other, marking on the skin should be made as a cautionary measure. QA considerations for head and neck implants are as follows: • Before implantation, marks on skin for proper catheter separation • Different color buttons for different planes of an implant. Different color catheters for different implants, in the case of multiple-site implants • Each catheter secured in its place with a metallic button so that the catheter movement is secured within 2 mm. Distal end blocked to ensure source ribbon does not go through • The intraoral plastic button(s) which serve as a spacer to allow the 192Ir sources to project beyond the mucosa • Not using the loop applications as for LDR implants for HDR applications since the HDR source may not be able to take sharp turns in the loop • For LDR implants, measurement of the catheter length and determination of number of seeds for each catheter • Clear diagram of the color-coded implant for brachytherapy dosimetry team. Other Sites Many other sites can be implanted with flexible catheters, abdomino-pelvic implants using a HAM applicator, or freehand implantation of tumor bed with multiple catheters in a single plane for recurrent tumors of rectum, bladder, sarcomas, etc.
Practical Considerations for Implant Reconstruction Rigid needles or flexible catheters implanted in the tumor and the critical structures located in the vicinity of the implant need to be reconstructed for the purpose of treatment planning in both LDR and HDR applications. AAPM TG 40 recommends that the position of all applicators or catheters be verified either radiographically or by CT. The ABS recommends that target volume delineations include (in the order of preference) determination on CT or MR images, through surgical clips or by projection drawn on orthogonal radiographs. Template-based implants and freehand implanted catheters are considered separately.
Templates and Rigid Needles To calculate a dose distribution for an interstitial implant, coordinates of all the radioactive sources must be determined relative to an arbitrary reference point that is used as the origin. This can be accomplished by orthogonal radiographs, variable angle radiographs, stereo-shift films, 3-film technique, or more elegantly by CT or MR scanning. For identification of individual catheters, coded dummy trains having a unique pattern are loaded into the catheters. The following practical considerations can be used as guide for the quality management of implant reconstruction of template-based interstitial implants: • Each template should be named and its allowable set of holes numbered in a unique fashion. For example, numbering of Syed-Neblett templates, at our institution, starts from the top (1 o’clock position) needle of the innermost circle and increases clockwise, from innermost to the outer concentric circles or arcs.
24–Quality Management for Interstitial Implants
467
• After the patient is implanted, all the implanted channels are numbered on a printed diagram of the template following the above convention. • For film-based reconstruction, each channel is loaded with a uniquely defined dummy train and the loaded dummy’s identification is marked against the channel in which it is loaded. • For CT-based reconstruction, there is no need to load dummy trains in steel needles, but plastic catheters may be loaded with dummy markers to be able to see them in scout views. • All dummy markers should truly represent the source wire, with the cold tip or leader of the same size as that of the source train. • The patient should be positioned in a reproducible position (we use three-point set up) so that when radiographs or scout films are taken before each fraction, the patient is in the same position as for the planning radiographs or CT. It is extremely important to ensure patient immobilization during the entire procedure of localization films, because identification of various channels and calculation points may require data entry from different sets of films. HDR Prostate Implants • A set of radiographs may be taken in the OR to ensure that the implant is in the correct place and the needles have not moved after the patient is brought to supine position from the earlier modified lithotomy position. This is important for implants with the needle tips exactly below the bladder base. • For film-based planning, a set of orthogonal or stereo-shift films with dummy trains is required. It has been reported (Sharma, Williamson, and Cytacki 1982) that accuracy of localization is better for orthogonal views; a set of AP/Lat films is desirable. • Three main critical structures in the prostate cases are bladder, urethra, and rectum. The Foley balloon, introduced previously into the bladder, should be filled with diluted (50%) contrast solution, otherwise the balloon may obscure dummy markers in the needles. To view the implant position with respect to a full bladder position, diluted (20%) Hypaque solution is injected directly into the bladder (that’s why three-way Foley) allowing easier urethral marking. • For CT-based planning, 2-mm transverse slices are desirable to identify the distal ends of the needle tips accurately. Gynecological Implants • Gynecological implants usually are characterized with numerous channels that can extend laterally about 7 cm on either side of the central obturator. Lateral radiographs result in severe overlapping of the channels. Several sets of orthogonal radiographs with selective loading of channels with dummy markers will be required. However, if two oblique (45° from the anterior) are taken with all the channels loaded with dummy ribbons, it is sometimes possible to work with just one set of radiographs. • Identification of the bladder points can be done according to ICRU 38 (ICRU 1985), but the rectal points cannot be defined in the same way as there is no vaginal packing. However, using a rectal marker or contrast, rectal points can be assigned. CT-based planning is preferable, since it allows delineation of structures.
468
Anil Kumar Sharma
Flexible Catheters Even though most freehand catheters are implanted without the aid of any well-defined geometrical pattern, their numbering should follow a very consistent order to avoid any kind of confusion for treatment planning and source loading. It is imperative that a preplan is carried out in such cases and guidelines from classical systems like Manchester, Paris, or Quimby be observed. Once the implant is complete, a diagram closely resembling the implant may be drawn showing different catheters in at least two planes, e.g., transverse and coronal or sagittal planes. Breast Implants Localization of MammoSite is carried out by inserting a dummy wire through the HDR port right up to the tip of the applicator. Either orthogonal films can be taken or CT scan performed after ensuring that the balloon is inflated with the planned amount of fluid (20% contrast). Without CT, however, the dose to the skin cannot be assessed accurately, nor can the trapped air be visualized, two problems that would lead to aborting the procedure. On the other hand, radiographic verification of the size of the balloon before treatment (to assure that it has not leaked) should be a routine QA procedure. Numbering of catheters is crucial in the case of multichannel implants. It is important to follow a convention consistently to avoid loading the catheters incorrectly. We follow the convention of shallow to deep planes with numbering starting at the most superior catheter of the shallow plane. All catheters are marked and the diagram of these catheters is made for planning and dose delivery. Head and Neck Implants The following points may be useful in proper reconstruction of complex head and neck implants: • Mark the catheters from superior to inferior starting at the shallow plane. • If the implantation is carried out in sagittal planes, start from the left side plane, numbering the catheters from anterior to posterior. • If there are multiple implants in close proximity, catheters should be marked continuously without separating them. • Numbering should be done for the corresponding flexible stylets as well, because catheters and their corresponding stylets will generally be of different sizes. • Patients usually are in pain (or sometimes even disoriented); carefully positioning them on the table with a headrest and possibly a masking tape to stabilize the head may be useful. Immobilization is essential because several films may be required and the patient’s position reproducibility is crucial. • Dummy ribbons should be loaded and their position in the catheters noted on the diagram. • Orthogonal films or variable angle oblique films are preferable. • For CT-based planning, the dummy ribbons are generally not required, but in some cases artifacts due to dental fillings or the gold buttons (used for providing shielding) may obscure the catheter’s distal end, necessitating the need to use dummy ribbons.
Quality Management of Treatment Planning and Evaluation Treatment planning for interstitial implants should be carried out on planning systems, which use ICWG formulism adopted in AAPM’s TG 43 report (Nath et al. 1995). Source strength should be specified as air kerma strength of the source to be used, with proper corrections for the source anisotropy along its
24–Quality Management for Interstitial Implants
469
axis. The planning system should be thoroughly tested before its clinical implementation. Also, periodic performance re-evaluation should be carried out by using reproducible catheter placement in a simple geometrical pattern. Post implant dose calculations should be performed after localization of the catheters or needles with respect to the target volume and the organs at risk. Dose distribution should be obtained in multiple planes after volume dose calculations. All patient dose calculations should be verified by the brachytherapy physicist by reviewing that no gross errors have occurred and by performing an independent dose calculation to one critical point. Dosimetry should be performed by a dosimetrist or a medical physicist and should be reviewed with the attending radiation oncologist before the start of an HDR fraction or during the first half of an LDR treatment.
Prostate Implants • After proper reconstruction of the implant, loading of channels is carried out as per the delineated target volume or according to the gold seeds previously implanted into the prostate. • Activity per seed in the central six needles is half of that used for the outer needles for LDR implants, dose rate about 65 cGy per hour. • For HDR implants, catheter length parameter, i.e., the distance from the indexer to the tip of the needles should be measured for all channels and entered into the planning system. This should be with 1mm for all channels. • For HDR implants, geometrical optimization works very well, but for further urethral dose reduction, the central six needles’ dwell time should be decreased to about one-third of the original dwell time. • Dose to the bladder points and rectal points should be calculated and be less than the prescribed dose; urethral dose also should be controlled. At our institution, no point in the urethra gets more than 110% of prescribed dose. • For TURP and re-irradiation cases, the dose to urethra should further be reduced by decreasing the dwell time in the central six needles (producing a hollow cylinder or doughnut shaped distribution). • Dwell positions for HDR implants should remain within the target volume. • The dose prescription and dose limitation to organs at risk will however take precedence over the above suggestions. • Dose evaluation for LDR or HDR prostate implants should be carried out and reported as per the ABS and ICRU 58 (ICRU 1997) guidelines. • Geometrical optimization generally results in doses in the center of the implant that are higher than the peripheral (or reference) dose. However, homogenous dose distributions can be achieved by manually reducing the dwell time for the inner circle catheters or differentially loading the catheters for LDR implants (Thomadsen et al. 1990). • Reading point dose values, particularly in the center of the implant, and drawing dose profiles at various levels help in assessing the dose homogeneity in the implanted volume. • Dose distributions are visually evaluated by inspecting the isodose lines in multiple planes throughout the planned target volume. The foremost optimization goal for prostate implants is to have the target volume covered as closely as possible with the reference isodose surface. The smallest-volume isodose surface covering the entire target volume is taken to be the planned target dose (PTD).
470
Anil Kumar Sharma • Though high-dose gradients exist around the source dwell positions, the total volume of the hyper dose sleeve (200% PTD) should be restricted to less than 10% of the treatment volume • Further analysis of the dose distribution should be carried out using differential dose volume histograms (DVHs) of the optimized plans. If the prostate volume implant is optimized to the same dose midway between the catheters, then the differential DVH will show a peak for that dose. • This peak is somewhat distorted by an inverse square law effect. These obscurative effects are eliminated in the “natural” DVH (Anderson 1986), a differential histogram for which the derivative of the baseline variable with respect to dose is proportional to the –5/2 power of the dose. Using the natural DVH, evaluation of the dose distribution between the catheters and in the treated volume is evaluated and the effect of dose optimization is assessed. • From the point dose values and the shape of dose profiles and DVHs, the extent to which the dwell time values for the central catheters are to be lowered is inferred. Changes can be made in a graphical optimization mode by simply dragging the graphical bar view of dwell times with the mouse or by typing in the required values. • Since attempts are made to make the dose distribution in the target volume homogeneous, it should approach Manchester system type loading. The mg-hrs required from Paterson-Parker tables (Paterson and Parker 1938) for the reference isodose surface volume can be compared with the actual mg-hrs obtained after appropriate conversions from the planned total air kerma strength for second calculation. In our center for prostate implants, these values matched within 5% for more than 97% of the cases, in fact for 68% of the treatment plans the mg-hrs were within 2% (Sharma et al. 1998).
Gynecological Implants Planning of gynecological tumors also follows most of the considerations mentioned above for the prostate implants, because the channels tend to be straight and implanted almost equidistantly and, loaded with almost the same spacing, form a geometrical matrix which is most suitable for geometrical optimization. However, several features are different from the prostate implants: • Central dose should be higher, because most tumors are at or around the center. For LDR implants 125% higher dose in the center is acceptable for non-tandem cases. With tandem a in place, the central dose is much higher. • For LDR planning, the central obturator needles are not supposed to be loaded if tandem is present, which alone is loaded with 10-5-5 mgRaeq type of loading. However, for HDR planning four lateral obturator needles can also be lightly loaded for better uniformity of the dose. • Obturator needles should not be loaded inferior to the external os, whether or not a tandem is in place, unless clinically indicated for extended vaginal disease. • Usually, several catheters are implanted (up to 44, and almost all are loaded for LDR implants with 192 Ir seeds—inner channels either loaded with lesser activity per seed or differentially unloaded in time) and tandem (if present) with 137Cs tubes. • For HDR planning, most planning systems allow more channels for planning than the maximum number of channels treatable on the afterloader. If the treatment planning or treatment unit only allows a maximum number of catheters equal to the number of channels, a composite plan should be carried out and later divided for the sake of treatment on the HDR machine. Extreme vigilance is required in such cases, where two subplans are used for treatment as far as the channel numbers on the patient and their corresponding planning catheter numbers are concerned.
24–Quality Management for Interstitial Implants
471
• Treatment planning and reporting for gynecological tumors should be volume based, as per ICRU 38 (ICRU 1985) and ICRU 58 reports (ICRU 1997). • Modified Paterson-Parker tables can again be used for a second check of dose calculations. For plans without tandem, the results are usually within 7%, and for plans with tandem within 15%.
Head and Neck Implants • Treatment planning of flexible catheters is in general more difficult as there can be more than one simultaneous implant and each one requires a different type of planning. For example, for tongue implants, the most distal source needs to be loaded above the mucosa for LDR implants, and with much higher dwell time for HDR implants. Whereas for neck implants dwell times near the skin should be reduced and dwell positions kept at least 5 mm from the skin. Tonsillar fossa catheters are loaded only to or slightly beyond the mid line, and the dwell times in the distal positions will depend on the position of other catheters that are implanted in planes orthogonal to the tonsillar fossa catheters. • A composite plan for multiple implants always should be made for dosimetry and evaluation. For HDR treatments, if required, different plans can be exported separately to the treatment machine. • For HDR treatments, all catheter length parameters should be measured individually, at both the entrance to the skin and the exit (or most distal) ends. The length of the catheter within the tissue as obtained by radiographic or CT means should be compared with these measurements. Both should agree within 2 mm. • Dose prescription in these implants is volume based; the target volume is either delineated on the CT slices or taken from the surgical markers. Effective interaction between physicist and radiation oncologist forms the basis of an effective treatment plan. • Plan evaluation should be carried out in all the transverse and sagittal planes to check for any hot or cold spots and any overlapping of loaded catheters. Natural DVHs are of great help in picking up the right isodose surface and for evaluating the plan using its dose uniformity and quality indices.
Quality Assurance of Source Application and Removal The most common radionuclide for LDR temporary interstitial implants is 192Ir, in the form of seeds arranged at specific intervals (usually 1 cm) in ribbons or wires of desired lengths. Source calibration, leak testing, receiving, logging, safe handling, and returning to the suppliers, etc., are regulated by NRC and agreement states. The institutions are required to manage their radiation sources in storage and in the patients as per their Radioactive Material Licenses (RAMLs). An excellent review on this topic is can be found in Williamson, Thomadsen, and Nath (1995), Thomadsen (1995), Thomadsen et al. (1994), and in chapter 9 on regulations in this monograph. In this section, only the practical considerations pertaining to the LDR and HDR interstitial brachytherapy treatment are presented.
Interstitial Brachytherapy Treatments with LDR • Based on the treatment plan, 192Ir source ribbons of proper strength and size are ordered and received in the department. • The number of sources and their strengths are compared with the user’s order and logged.
472
Anil Kumar Sharma • All sources should be stored in a secure room, with a radioactive material sign posted on its door. When opening the package, it should be determined that there is no contamination due to damage during shipping. Exposure at 1 m and at surface, along with a wipe test of the outer container, should be noted and logged. • Ribbons should be transported to the patient’s room in a cart with proper labeling of radioactive material after calibration as per AAPM TG 56 report (Nath et al. 1997). • Loading should be carried out by the authorized user and physics staff using long-handled forceps. A radiation survey with a calibrated meter should be done after the sources are loaded and secured in the respective channels as per the plan. The time of source loading and survey results are noted down in the patient’s chart. • A “Caution: Radiation Area” sign is posted on the patient’s door with the type of sources and radiation dose at 1 m, at bed side, and at the door noted on the radiation label. • Emergency phone numbers of the radiation safety officer (RSO) and the physician should also be on the label. • After the completion of treatment, all sources should be removed using long forceps and placed in a shielded container. • A radiation survey should be carried out and its results noted down. When satisfied that all sources have been removed and secured in the container, all radiation labels are removed from the door and room. • Sources are removed from the container and stored in the hot lab for later shipment to the supplier.
Interstitial Brachytherapy Treatments with HDR The AAPM Report No. 41 (TG 41) (Glasgow et al. 1993), Report No. 61 (TG 59) (Kubo et al. 1998). and Report No. 46 (TG 40) (Kutcher et al. 1994) guidelines on remote afterloading technology should be followed for daily, monthly, and other periodic guidelines. • All HDR applications should be done after a proper treatment day QA, which includes position verification of proper operation of the dummy cable and positioning of the source wire to within 1 mm, and functioning of all interlocks including the door, indicator lights, radiation detectors, catheter integrity, timer, and source decay correction. • New source calibration and periodic source calibration as per the RAML must be carried out. Differences up to 5% from the manufacturer’s calibration certificate are generally acceptable, but it is good practice to check with the manufacturer if the difference is more 3%. • Authorized user and medical physicist or RSO must be present for all HDR treatments. • The treatment plan should be compared with that from the treatment planning system by checking the dwell times in each channel and total treatment time. For large implants with numerous dwell positions, random dwell position and dwell times may be sufficient. • The implanted needles or catheters should be connected by one person to their respective numbers on the afterloader turret and verified by a second. • There should be a proper audiovisual contact with the patient during the treatment.
24–Quality Management for Interstitial Implants
473
• Emergency kit should be available in the treatment room, and two radiation survey meters (at least one of them being ionization chamber type with high range (500 R) should be available at the console. • Once the treatment is finished, the physicist surveys the patient, again taking readings at the afterloader surface and at a distance (some states require at 20 cm, NRC requirement is at 1 m) from the afterloader surface. • Before any HDR fraction, the patient with template and needles is either scanned or radiographed to ensure that the applicator is in the right position. MammoSite patients should be radiographed (or scouted) to ensure that the balloon is intact. • For the MammoSite applicator, it is mandatory to ensure that the luer-lock is firmly connected. Also, for all MammoSite patients, endobronchials and endoesophageals should be treated in the scanning position, i.e., supine on a treatment table or a gurney. • All source transfer tubes should be as far off the patient as possible to reduce any contact dose to other areas of the patient. • HDR fractionated treatments should be at least 6 hours apart; any shorter gap between the fractions should be justified and accounted for through recalculation in the dose prescription. For head and neck tumors, the ABS recommends at least a 6-hour gap (Nag et al. 2001). • Any treatment plans with more than the number of available channels on an older treatment planning system that requires the division into two separate plans should be checked very carefully by the authorized user, physicist, and therapist to ensure the planned channel is connected to the treatment channel.
Needle or Catheter Movements and Possible Corrections HDR brachytherapy is delivered over a number of fractions, allowing for potential movement of afterloading needles between fractions. Our own experience has shown that about 25% of the patients require adjustment of the catheter length parameter (Damore et al. 2000). Most movements take place for the second fraction after an overnight stay in the hospital. By keeping adequate margins for treatment, i.e., implanting needles deeper than required, but treating the target by leaving required space at the distal end, one can have enough room for applying corrections in the cleft lip and palate. Inferior displacement of interstitial needles between fractions is a potential source of error in the delivery of HDR brachytherapy for prostate, gynecological, and rectal cancers. Localization films or CT scouts are necessary before any treatment and corrections based on these films can be made so that the target volume is given the same dose pattern approved after planning every time.
References Anderson, L. L. (1986). “A ‘natural’ volume-dose histogram for brachytherapy.” Med Phys 13:898–903. Baglan, K. L., A. A. Martinez, R. C. Frazier, V. R. Kini, L. L. Kestin, P. Y. Chen, G. Edmundson, E. Mele, D. Jaffray, and F. A. Vicini. (2001). “The use of high-dose-rate brachytherapy alone after lumpectomy in patients with earlystage breast cancer treated with breast-conserving therapy.” Int J Radiat Oncol Biol Phys 50:1003–1011. Damore, S. J., A. M. N. Syed, A. Puthawala, and A. Sharma. (2000). “Needle displacement during HDR brachytherapy in the treatment of prostate cancer.” Int J Radiat Oncol Biol Phys 46(5):1205–1211. Glasgow, G. P., J. D. Bourland, P. W. Grisby, J. A. Meli, and K. A. Weaver. “Remote Afterloading Technology.” AAPM Report No. 41. New York: American Institute of Physics, 1993.
474
Anil Kumar Sharma
International Commission on Radiation Units and Measurement (ICRU) Report No. 38. “Dose Volume Specification for Reporting Intracavitary Therapy in Gynecology. Bethesda, MD: ICRU, 1985. International Commission on Radiation Units and Measurement (ICRU) Report No. 58. “Dose and Volume Specifications for Reporting Interstitial Therapy.” Bethesda, MD: ICRU, 1997. Interstitial Collaborative Working Group (ICWG). Interstitial Brachytherapy, Physical, Biological and Clinical Considerations. New York: Raven Press, 1990. Kubo, H. D., G. P. Glasgow, T. D. Pethel, B. R. Thomadsen, and J. F. Williamson. (1998). “High dose-rate brachytherapy treatment delivery: Report of the AAPM Radiation Therapy Committee Task Group 59.” Med Phys 25(4):375–403. Also available as AAPM Report No. 61. Kutcher, G. J., L. Coia, M. Gillin, W. F. Hanson, S. Leibel, R. J. Morton, J. R. Palta, J. A. Purdy, L. E. Reinstein, G. K. Svensson, M. Weller, and L. Wingfield. (1994). “Comprehensive QA for radiation oncology: Report of AAPM Radiation Therapy Committee Task Group 40.” Med Phys 21:581–618. Also available as AAPM Report No. 46. Martinez, A, R. Cox, and G. Edmundson. (1984). “A multiple-site perineal applicator (MUPIT) for treatment of prostate, anorectal, and gynecological malignancies.” Int J Radiat Oncol Biol Phys 10:297–305. Nag, S., B. Erickson, S. Parikh, N. Gupta, M. Varia, and G. Glasgow. (2000). “The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for carcinoma of the endometrium.” Int J Radiat Oncol Biol Phys 48(3):779–790. Nag, S., E. Cano, D. J. Demanes, A. A. Puthawala, B. Vikram; American Brachytherapy Society. (2001). “The American Brachytherapy Society recommendations high-dose-rate brachytherapy for head and neck carcinoma.” Int J Radiat Oncol Biol Phys 50(5):1190–1198. Nag, S., C. Chao, B. Erickson, J. Fowler, N. Gupta, A. Martinez, B. Thomadsen; American Brachytherapy Society. (2002). “The American Brachytherapy Society recommendations for low-dose-rate brachytherapy for carcinoma of the cervix, Int J Radiat Oncol Biol Phys 52(1):33–48. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. ( 1997). “Code of practice for brachytherapy physics: Report of AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24(10):1557–1598. Also available as AAPM Report No. 59. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43.”Med Phys 22:209–234. Also available as AAPM Report No. 51. Neblett, D. C. (1995). “Clinical Techniques on Applicators Available for interstitial Implantation” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.) Madison, WI: Medical Physics Publishing, pp. 282–300, 1995. Paterson, R., and H. M. Parker. (1938). “A dosage system for interstitial radium therapy.” Br J Radiol 11:252–266. Puthawala, A. A., A. M. N. Syed, D. L. Eads, D. Neblett, L. Gillin, and T. C. Gates. (1985). “Limited external irradiation and interstitial 192-iridium implant in the treatment of squamous cell carcinoma of the tonsillar region.” Int J Radiat Oncol Biol Phys 11(9):1595–1602. Sharma, S. H., J. F. Williamson, and E. Cytacki. (1982). “Dosimetric analysis of stereo and orthogonal reconstruction of interstitial implants.” Int J Radiat Oncol Biol Phys 8:1803. Sharma, A., A.M.N. Syed, A. Puthawala, and A. A. Farid. (1998). “HDR planning for prostate cancer implants using Varisource remote afterloader.” J Brachytherapy International 14:1–14. Syed, A. M. N., and A. A. Puthawala. Proceedings of the Workshop on HDR and LDR Brachytherapy Techniques for Prostate, Head and Neck and GYN Malignancies. Long Beach, CA: Long Beach Endocuritherapy Foundation, 1996. Syed, A. M. N., A. Puthawala, P. Austin, J. Cherlow, J. Perley, L. Tansey, A. Shanberg, D. Sawyer, R. Baghdassarian, B. Wachs, et al. (1992). “Temporary iridium-192 implant in the management of carcinoma of the prostate.” Cancer 69:2515–2524. Syed, A. M. N., A. Puthawala, N. Barth, A. Sharma, and A. Londrc. (1997). “High dose rate brachytherapy in the treatment of carcinoma of prostate: Preliminary results.“ J Brachytherapy No vol given:1–14. Syed, A. M. N., A. Puthawala, D. Neblett et al. (1986). “Transperineal interstitial/intracavitary ‘Syed-Neblett’ applicator in the treatment of carcinoma of the uterine cervix.” Endocurither/Hypertherm 2:1–13.
24–Quality Management for Interstitial Implants
475
Thomadsen, B. R., P. V. Houdek, R. van der Laarse, G. Edmundson, I. K. Kolkman-Deurloo, and A. G. Visser. “Treatment Planning and Optimization” in High Dose Rate Brachytherapy: A Textbook. S. Nag (ed.). Armonk, NY: Futura Publishing, pp. 79–145, 1994. Thomadsen, B. R., S. Shahabi, D. Buchler, W. Giese, and M. Mehta. (1990). “Differential loadings of brachytherapy templates.” Endocuriether/Hypertherm Oncol 6:197–202. Thomadsen, B. R. “Clinical Implementation of Remote Afterloading Interstitial Brachytherapy” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.). Madison, WI: Medical Physics Publishing, pp. 679–698, 1995. U.S. Nuclear Regulatory Commission. (1991a). Code of Federal Regulations, title 10, Rules and Regulations. Washington, DC: Nuclear Regulatory Commission. U.S. Nuclear Regulatory Commission. (1991b) Code of Federal Regulations, title 10, part 35, para 35.32, Energy, Rules and Regulations. Washington, DC: Nuclear Regulatory Commission. U.S. Nuclear Regulatory Commission. (1991c). Regulatory Guide 8.33 (Task DG-8001), Quality Management Program, para 3.1.6 and 3.2.10. Washington, DC: Nuclear Regulatory Commission. Vicini, F. A., V. R. Kini, P. Y. Chen, E. Horwitz, G. Gustafson, P. Benitez, G. Edmundson, N. Goldstein, K. McCarthy, and A. Martinez. (1999). “Irradiation of tumor bed alone after lumpectomy in selected patients with early-stage breast cancer treated with breast-conserving therapy.” J Surg Oncol 70:33–40. Williamson, J. F., G. Ezzell, A. Olch, and B. R. Thomadsen. “Quality Assurance of High Dose Rate Brachytherapy” in High Dose Rate Brachytherapy: A Textbook. S. Nag (ed.), Armonk, NY: Futura Publishing, pp. 147–212, 1994.
Chapter 25
Post-Implant Evaluation William S. Bice, Jr., Ph.D. University of Texas Health Science Center at San Antonio International Medical Physics Services San Antonio, Texas Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477 The Post Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477 Essential Characteristics of the Treatment-Planning System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477 Desirable Characteristics of the Treatment-Planning System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478 The Dose Evaluation Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 478 Individual Implant Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 479 Procedural Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480 Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482 Author’s Note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 483
Background Post-implant evaluation is performed after the implant procedure. While this may sound inane, or at least obvious, restricting this evaluation to completion of the implant has a purpose. Completion of the implant means after the removal of sources in temporary implant or after the implantation of sources for permanent brachytherapy. The intent is to distinguish post-implant evaluation from implant design or adjustment, performed during the implant procedure. Post-implant evaluation then is performed for two purposes: (1) determining whether the individual implant meets the desired goals and (2) evaluating the performance of the implant team.
The Post Plan The heart of the post-implant evaluation is the post plan. Even though the evaluation includes much more than the radiation dose distribution—examples might be an assessment of methods and techniques or a review of appropriate patient selection, the focus remains on achieving optimal dosimetry. The post plan provides immediate quantitative feedback about the implant, indicative of clinical success and a measure of procedural success.
Essential Characteristics of the Treatment-Planning System Modern treatment planning is based upon three-dimensional (3-D) imagery, typically a set of contiguous axial images. An image set or sets that show the patient’s anatomy as well as the source locations are essential to realistic depiction of implant quality. Knowing where the dose was delivered is just as important as knowing how much dose was delivered. Dose calculations for brachytherapy, because the dose changes so rapidly over short distances, should be performed on a grid of voxels no larger than 2 mm on a side (Yu et al., 1999; Nag et al., 2000). Calculations must be performed in three dimensions using the formalism and values described by AAPM Task Group 43 (TG-43) (Nath et al., 1995) or the more recent update (TG-43U1) (Rivard et al., 2004). Linear and point source calculations should be available where appropriate. For permanent seed implants, where the number of sources precludes individual source
477
478
William S. Bice
identification, an automated seed sorting algorithm is essential (Brinkman and Kline 1998; Bice et al., 1999; Li et al., 2001; Liu et al., 2003; Davis et al., 2004; Lam et al., 2004; Holupka et al., 2004). Automated localization and sorting algorithms have improved dramatically since their introduction in the mid-1990s. As a time-saving device their utility is obvious. More importantly, from the perspective of providing high-quality information, the results of these routines are consistent and repeatable. A planning system that can produce a digitally reconstructed radiograph (DRR), or even just an antero-posterior (AP) localization plot of the source locations, provides an extremely useful tool in evaluation of the automated procedures; properly scaled, the sorted seed locations should match an AP radiograph of the patient postimplant. Dose and structures must be available in the same display, preferably with both 2-D and 3-D available. The planning system must be capable of calculating and displaying structure-based dose-volume histograms (DVHs). Individual point placement, calculation, and display are essential.
Desirable Characteristics of the Treatment-Planning System Other tools that may be considered useful for post planning include dose-surface histograms (DSHs), line plots, or dose traces (Anagnostopoulos et al., 2003). Automated generation of quantifiers from the DVH/DSH is a time saver as is automated location of possible source positions. In contrast to seed sorting, where the number of possible source locations is reduced to the number of sources in the implant, automated location of possible source positions is dependent upon contrast differences in the image between the sources and the surrounding tissues. The computer, able to distinguish sources in the image, picks out these possible locations automatically. There has been some work on techniques that allow retention of some of the spatial information that has been lost in the DVH by dividing the target into sectors and generating a DVH for each sector (Bice, Prestidge, and Sarosdy 2001). Projection imagery, digitally reconstructed radiography, is useful for comparing to projection radiographs. Localization from multiple projection radiography can be useful, particularly in situations where this localization is used in conjunction with—fusion with—an image set where anatomy is displayed (Gong et al., 2002).
The Dose Evaluation Hierarchy Figure 1 shows the dose evaluation displays available from the treatment-planning systems described above. Isodose displays show the location of the dose distribution with the size and location of regions of higher and lower dose shown graphically. DVHs are derived from the isodose distributions, removing the spatial information pertaining to the location of the dose within a structure. In return, the DVH reduces the complex information available in the isodose display to a single graph. Further simplification of the information contained in the DVH leads to the generation of quantifiers, single numbers that describe a corresponding characteristic of the dose distribution. Thus there are three tiers to the dose evaluation hierarchy; in order of decreasing information but increasing simplicity, isodose curves, DVHs, and quantifiers. There is a fourth, limited tier that consists of linear dose displays or dose traces. Linear displays, from which quantifiers may also be drawn, sacrifice none of the spatial information. Such displays are appropriate only for linear structures. Examples include the urethra and the neurovascular bundles (DiBiase et al., 2000). Post-implant evaluation at the top of the hierarchy (isodose curves) is usually restricted to analysis of individual implants. For the individual implant it is important to know as much dosimetric information as possible. While it may be desirable to compare quantifiers for single implants, assessing the location of dose is vastly more important. An example of this is the implant that adequately covers the site of positive biopsy results, but does not quite cover the entire planning target volume (PTV). The coverage quantifiers may not be quite up to par, but the brachytherapist may conclude that the coverage is adequate to preclude further therapeutic intervention.
25–Post-Implant Evaluation
Single implants
Information
Analytical tool Isodose displays Dose-Volume Histograms Dose Trace Displays Quantifiers
Spatial information X
Dose information X X
X
X
479
Dose analysis X X X
Simplicity Group analysis
X
Figure 1. The dose-evaluation hierarchy. Dose information from an implant can be presented in several ways. Dose calculations are usually performed on a grid of voxels whose resolution depends on the desired accuracy of the calculation. Isodose curves and surfaces can be derived from this volumetric calculation, detailing the location of dose—if anatomical information is available from an image set, the relationship of the dose distribution to the patient’s anatomy can be determined. DVHs contain no spatial information over and above the fact that the stated dose-volume relationship pertains to the structure for which the DVH was generated, i.e., all the calculated dose points are located within the structure. A quantifier is a single number that describes some aspect of the implant, coverage, high dose regions, etc. Because of the large amount of information available from an isodose display, detailed analysis using isodose displays is usually restricted to single implants. Group analysis is better suited to the use of quantifiers due to their simplicity.
Group analysis, that used to evaluate the procedural methods and the skill of the implant team, is usually restricted to the use of quantifiers. A set of well-chosen quantifiers can be very revealing about the institution’s implant procedure, even with just a few implants in the database.
Individual Implant Assessment While the radiation following an implant procedure cannot be taken back, it is still important to assess the dose delivered to the target volume and the dose delivered to any critical structures. Figure 2 shows a paradigm that can be used in evaluating post-implant dosimetry for an individual implant. The decision scheme shown can be used after the post-implant dosimetry has been created. If the target volume received an inadequate dose, or inadequate coverage, then adjuvant therapies should be considered. An additional implant procedure may be considered, an additional application in the case of temporary implants or additional sources implanted in the case of permanent implants. External beam radiotherapy may be used, as well as other adjuvant therapies, chemotherapy or surgery, for instance. This is in contrast to salvage procedures—those performed following a procedure that has failed clinically rather than just dosimetrically. In evaluating the dose to critical structures the task of post-implant analysis becomes, arguably, more difficult. Ideally, no dose would be delivered to normal tissues. In addition to determining the dose to these structures then, the evaluator must determine how much dose is too much. The fuzzy world of risk versus benefits must be inhabited, where bowing to reality is a necessity. Worse, the radiation has been delivered, usually with attendant physical and biological uncertainties. Too some extent the brachytherapist can only watch while the results of his handiwork play out before him. Increased watchfulness can take the form of more frequent follow-up visits. In some cases, the use of therapeutic modalities—even as prophylactic—can be prescribed. An example of this is hyperbaric oxygen used to avoid radiation-induced rectal injury.
480
William S. Bice
Figure 2. A paradigm that may be used to evaluate an individual implant. This decision tree outlines a technique for conducting an evaluation on a single implant chart. Two evaluations are performed, one to determine the adequacy of coverage of the target volume and the other to adjudge the dose and the effect of dose on the critical structures. A “no” with regard to the first results in consideration of adjuvant therapies or at least a level of increased watchfulness on the part of the caregiver. A “no” concerning acceptable critical structure doses can result in the same increased watchfulness if complications are possible from the doses to these structures. All decisions eventually lead to filing the chart for eventual review as part of a group review (labeled programmatic view in the figure).
Procedural Assessment Analyzing a series of implants to study the quality of the implant program can be of enormous value. Any effort in this regard is worthwhile if the evaluation is performed honestly; publishable results are not the goal. A large database is not required. Series with as few as 10 brachytherapy procedures can still be useful (Bice et al., 1998). A suggested paradigm for evaluating the post-implant dosimetry on a series of implants is shown in Figure 3. Note the dissimilarity between this figure and the technique presented in Figure 2 showing individual implant assessment. In the group assessment shown in Figure 3 implant characteristics and dosimetric results—as described by quantifiers—are categorized for each implant. The choices of the quantifiers and the implant characteristics to evaluate are made by the implant team. The quantifiers are used to discover group traits, the characteristics correlated with the quantifier values in order to determine what kinds of choices can be made to improve the dosimetric results. Two points about this evaluation should be mentioned. First, the list of implant characteristics and implant quantifiers needs to be short. A long, comprehensive list will prove burdensome and the process will not be used, particularly in busy clinic. Additionally. a full set of characteristics and quantifiers has the tendency to be confusing particularly in terms of answering a specific question. Planning and a little thought can promote an intelligent choice of characteristics and quantifiers. Consider the following scenario. Suppose that the brachytherapist notices, anecdotally, that there appears to be an unusual amount of rectal bleeding in his permanent prostate brachytherapy patients
25–Post-Implant Evaluation
481
Figure 3. A paradigm that may be used to evaluate a series of implants. This is a possible technique for conducting a group or programmatic review. Checking for regulatory compliance is separated from the quantifier analysis. This was done to highlight the difference between the two and the intended effect of the quantifier analysis on improving implant quality. While the number of patients has been widened from one to many in comparing the review to the previous one, the scope of the analysis has been severely restricted. The implant team should determine what characteristics and quantifiers to extract. The text describes an example illustrating the utility of a limited scope in performing the post-implant evaluation. The waiting periods shown in the figure can be set as time periods or periods determined by the number of patients treated before another evaluation needs to be performed.
implanted with iodine. A review of the literature shows several papers that correlate dose and rectal symptoms. Choosing one that appears reputable, the brachytherapy team decides to keep track of the volume of rectum that receives more than 160 Gy (R160 Gy). The team can then compare their dosimetric results to those seen at the publishing institution, which recommended keeping R160 Gy to less than 1.3 cm3 to achieve less than a 5% chance of inducing Grade 2 proctitis (Snyder et al., 2001). Now a reasonable choice for implant characteristics to correlate might be needles placed in row one of the middle three columns of the implant template. Note that this kind of analysis can be limited to a dosimetric comparison. It can be done quickly on a relatively small number of patients, looking only at post-implant dosimetry. An analysis of actual bleeding rates would take longer and can be done using this technique, but remember the idea is not to be able to publish the results, but to perform better implants. If a large number of the institution’s patients had R160 Gy greater than 1.3 cm3, the team might choose to modify their implant technique. If the number of needles placed in the offending rows correlated to increased rectal doses, they might want to modify their preplans to move sources further away from the rectum. In this case, it would be appropriate to follow the path shown at the bottom of Figure 3, wait until another series of implants with this new design technique has been collected, and analyze the results on this new series. This finally brings us to the second point. Correlation of quantifier results to implant characteristics need not introduce sophisticated statistical analysis. A Fisher test can be used to perform this correlation, or even a table showing the data can highlight correlations. Just taking averages are often quite telling. A Student’s t-test is simple and can be used. In the example above, the correlation could probably have even been omitted: it is pretty clear that moving needles and sources away from the rectum would lower the dose. Remember, the idea is to improve implant quality, not to display mathematical prowess.
482
William S. Bice
Reporting From the physicist’s point of view, the implant record must at least contain a written directive; order, shipment, and receipt records; assay and other quality control results; survey results; nursing and patient release instructions and records; pre- and post-planning records and calculation checks; and an implant report (Anderson et al., 1991, Nath et al., 1997). The role of reporting is central to post-implant evaluation efforts. A complete implant report includes the results of post-implant dosimetry. At our institution, as at most, the implant report is part of the special physics consult. From the perspective of the post-implant evaluators, this report is an organizational tool, ensuring completeness of the information conveyed. If properly developed, this report can serve as the sole, or at least the primary, source of data for the post-implant evaluation used in group analysis. Suggested formats for implant reporting change with the type of implant. Not all dose specification methods are applicable to all implant types. Regardless of the form of the dose specification, volumetric, surface or points, the criteria need to be clearly stated, as well as the localization method used. Doses used in the written directive need to state the relationship between the directive and the specified dose. The specification needs to be consistent and reproducible. The format for reporting should conform to that recommended by consensus groups (Nag et al., 2003). Table 1 lists references for reporting formats for different kinds of implants. While every institution should feel free to revise these formats to fit their post-implant-evaluation needs, using the referenced formats will help convey information among institutions in a clear and consistent manner.
Conclusions Post-implant evaluation is essential to determine the quality of the implant, first to assess the outcome of the implant for the patient, second to appraise the implant technique employed by the brachytherapy team. The post plan is key in this evaluation. Proper post planning requires the proper tools, specifically a modern, image-based treatment-planning system. Proper, consistent reporting is essential to post implant evaluation, completely conveying the intent and result of the implant. Group, or programmatic, analysis is greatly facilitated by well-chosen report formats.
Author’s Note This chapter on post-implant evaluation is intentionally general in nature, in an attempt to address all types of implants. A subsequent chapter in this book deals with a specific type of implant, low dose rate (LDR) permanent prostate brachytherapy as an example of post-implant evaluation methods. LDR prostate implants were chosen presumably because of the impact that post-implant evaluation has had
Table 1. Suggested reporting formats. Brachytherapy Procedure
Reporting Format
Gynecological brachytherapy
ICRU 38 (1985)
Gynecological brachytherapy (HDR)
ICRU 38 (1985), Nag et al. (2000)
Interstitial implants
ICRU 58 (1997)
Prostate brachytherapy
Gillin et al. (1990), Yu et al. (1999), Nag et al., (2002)
General Brachytherapy
Anderson et al. (1991), Nath et al. (1997)
25–Post-Implant Evaluation
483
on improving the modality and the available tools. The detail presented in the subsequent chapter can be applied to other types of implants as well.
References Anagnostopoulos, G., D. Baltas, A. Geretschlaeger, T. Martin, P. Papagiannis, N. Tselis, and N. Zamboglou. (2003). “In vivo thermoluminescence dosimetry dose verification of transperineal 192Ir high-dose-rate brachytherapy using CT-based planning for the treatment of prostate cancer.” Int J Radiat Oncol Biol Phys 57(4):1183–1191. Anderson, L., R. Nath, A. J. Olch, and J. Roy. (1991). “American Endocurietherapy Society recommendations for dose specification in brachytherapy.” Endocuriether/Hypertherm Oncol 7:1–12. Bice, W. S., Jr., B. R. Prestidge, and M. F. Sarosdy. (2001). “Sector analysis of prostate implants.” Med Phys 28(12):2561–2567. Bice, W. S., Jr., B. R. Prestidge, P. D. Grimm, J. L. Friedland, V. Feygelman, M. Roach 3rd, J. J. Prete, D. F. Dubois, and J. C. Blasko. (1998). “Centralized multiinstitutional postimplant analysis for interstitial prostate brachytherapy.” Int J Radiat Oncol Biol Phys 41(4): 921–927. Bice, W. S., Jr., D. F. Dubois, J. J. Prete, and B. R. Prestidge. (1999). “Source localization from axial image sets by iterative relaxation of the nearest neighbor criterion.” Med Phys 26(9):1919–1924. Brinkmann, D. H., and R. W. Kline. (1998). “Automated seed localization from CT datasets of the prostate.” Med Phys 25(9):1667–1672. Davis, B. J., D. H. Brinkmann, J. J. Kruse, M. G. Herman, W. N. LaJoie, D. J. Schwartz, T. M. Pisansky, and R. W. Kline. (2004). “Selective identification of different brachytherapy sources, ferromagnetic seeds, and fiducials in the prostate using an automated seed sorting algorithm.” Brachytherapy 3(2):106–112. DiBiase, S. J., K. Wallner, K. Tralins, and S. Sutlief. (2000). “Brachytherapy radiation doses to the neurovascular bundles.” Int J Radiat Oncol Biol Phys 46(5):1301–1307. Gillin, M.T., D.L. Zellmer, D.F. Grimm and K. Sherwood. (1992). “Practical Considerations for Interstitial Brachytherapy” in Advances in Radiation Oncology Physics, Dosimetry, Treatment Planning and Brachytherapy. J. A. Purdy (ed.). AAPM Medical Physics Monograph Number 19. New York: American Institute of Physics, pp. 703-727, 1992. Gong, L., P. S. Cho, B. H. Han, K. E. Wallner, S. G. Sutlief, S. D. Pathak, D. R. Haynor, and Y. Kim. (2002). “Ultrasonography and fluoroscopic fusion for prostate brachytherapy dosimetry.” Int J Radiat Oncol Biol Phys 54(5):1322–1330. Holupka, E. J., P. M. Meskell, E. C. Burdette, and I. D. Kaplan. (2004). “An automatic seed finder for brachytherapy CT postplans based on the Hough transform.” Med Phys 31(9):2672–2679. International Commission on Radiation Units and Measurements (ICRU). Report No. 38. “Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology.” Bethesda, MD: ICRU, 1985. International Commission on Radiation Units and Measurements (ICRU). Report No. 50. “Prescribing, Recording and Reporting Photon Beam Therapy.” Bethesda, MD: ICRU, 1993. International Commission on Radiation Units and Measurements (ICRU). Report No. 58. “Dose and Volume Specification for Reporting Interstitial Therapy.” Bethesda, MD: ICRU, 1997. International Commission on Radiation Units and Measurements (ICRU). Report No. 62. “Prescribing, Recording and Reporting Photon Beam Therapy (Supplement to ICRU Report 50).” Bethesda, MD: ICRU, 1999. Lam, S. T., P. S. Cho, R. J. Marks II, and S. Narayanan. (2004). “Three-dimensional seed reconstruction for prostate brachytherapy using Hough trajectories.” Phys Med Biol 49(4):557–569. Li, Z., I. A. Nalcacioglu, S. Ranka, S. K. Sahni, J. R. Palta, W. Tome, and S. Kim. (2001). “An algorithm for automatic, computed-tomography-based source localization after prostate implant.” Med Phys 28(7):1410–1415. Liu, H., G. Cheng, Y. Yu, R. Brasacchio, D. Rubens, J. Strang, L. Liao, and E. Messing. (2003). “Automatic localization of implanted seeds from post-implant CT images.” Phys Med Biol 48(9):1191–1203. Nag, S., W. Bice, K. DeWyngaert, B. Prestidge, R. Stock, R., and Y. Yu. (2000). “The American brachytherapy society recommendations for permanent prostate brachytherapy postimplant dosimetric analysis.” Int J Radiat Oncol Biol Phys 46:221–230.
484
William S. Bice
Nag, S., R. J. Ellis, S. Merrick, R. Bahnson, K. Wallner, and R. Stock. (2002). “American Brachytherapy Society recommendations for reporting morbidity after prostate brachytherapy.” Int J Radiat Oncol Biol Phys 54(2):462–470. Nag, S., R. Dobelbower, G. Glasgow, G. Gustafson, N. Syed, B. Thomadsen, and J. F. Williamson. (2003b). “Inter-society standards for the performance of brachytherapy: A joint report from ABS, ACMP and ACRO.” Crit Rev Oncol Hematol 48(1):1–17. Narayanan, S., P. S. Cho, and R. J. Marks II. (2004). “Three-dimensional seed reconstruction from an incomplete data set for prostate brachytherapy.” Phys Med Biol 49(15):3483–3494. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee, Task Group No. 43.” Med Phys 22:209–234. Also available as AAPM Report No. 51. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24(10):1557–1598. Also available as AAPM Report No. 59. Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. (2004). “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Also available as AAPM Report No. 84. Snyder, K. M., R. G. Stock, S. M. Hong, Y. C. Lo, and N. N. Stone. (2001). “Defining the risk of developing grade 2 proctitis following 125I prostate brachytherapy using a rectal dose-volume histogram analysis.” Int J Radiat Oncol Biol Phys 50(2):335–341. Yu, Y., L. L. Anderson, Z. Li, D. E. Mellenberg, R. Nath, M. C. Schell, F. M. Waterman, A. Wu, and J. C. Blasko. (1999). “Permanent prostate seed implant brachytherapy: Report of the American Association of Physicists in Medicine Task Group No. 64.” Med Phys 26(10):2054–2076. Also available as AAPM Report No. 68.
Chapter 26
ICRU Recommended Dose and Volume Specifications for Reporting Interstitial Brachytherapy Ali S. Meigooni, Ph.D. Department of Radiation Medicine University of Kentucky, Lexington, Kentucky Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486 Definitions of Terms and Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486 Temporary and Permanent Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486 Source Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487 Description of Source Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487 Total Reference Air Kerma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487 Volume and Planes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488 Gross Tumor Volume (GTV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488 Clinical Target Volume (CTV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488 Planning Target Volume (PTV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488 Treatment Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489 Central Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489 Description of Dose Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489 General Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489 Dose Distribution in One or More Planes Through the Implant . . . . . . . . . . . . . . . . . . . . . . . . . . 489 Prescription Dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490 Minimum Target Dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490 Mean Central Dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 490 High-Dose Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493 Low-Dose Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494 Dose Uniformity Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494 Additional Representations of the Dose Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494 Time-Dose Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495 General Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 495 Time and Dose Rates for Temporary Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496 Time-Dose Pattern for Temporary Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496 Recommendations for Recording and Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498 Parameters Required for Recording and Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498 Description of Volumes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498 Description of Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498 Description of Technique and Source Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498 Description of Time Pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 498 Total Reference Air Kerma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499 Description of Dose Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499 Priority . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500 Practical Applications of the Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500 Temporary Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500 Permanent Implant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500 Single Stationary Source Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500 Moving Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500 Surface Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500 Appendix: Quantities and Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506 Basic Quantities and Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506 Definition of Quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506
485
486
Ali S. Meigooni Relationship Between the Quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507 Reference Air Kerma Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 507
Introduction One of the important technological developments in brachytherapy over the past two decades was the introduction of miniaturized and highly flexible sources. These sources can be used in afterloading devices with radionuclides of different activities that can produce a wide range of dose distribution within the treatment volume. At the same time, the sophisticated three-dimensional (3-D) source localization methods have been developed in this field and can be linked to computerized methods for dose calculation and representation of dose distribution. These developments have led many clinicians to depart from the longestablished implant systems and to utilize new methodologies for their patient treatments. Therefore, a common language is valuable to provide a method of dose specification and reporting which can be used for implants of all types and can be common to all those involved in interstitial brachytherapy. The International Commission on Radiation Units and Measurements (ICRU) has previously published reports dealing with dose and volume specification for reporting intracavitary therapy in gynecology (ICRU Report 38) (ICRU 1985). However, in 1997 a new report was published to address the problem of absorbed dose specification for reporting interstitial therapy (ICRU Report 58) (ICRU 1997). The aim of report 58 was to develop a common language based on the presently existing concepts. This report introduced concepts to describe what has been done in a way that can be more closely related to the outcome of treatment and one that was generally understood. This chapter reviews the ICRU 58 recommended concepts and procedures for dose specifications and reporting of the interstitial brachytherapy implants.
Definitions of Terms and Concepts Temporary and Permanent Implants Interstitial implants generally fall into two categories, temporary or permanent. In permanent implants, the brachytherapy sources remain in the patient, and they will not be removed. In contrast, in temporary implants, the radioactive sources are removed from the tissue after the treatment is completed. Traditionally, the temporary implants were performed with the linear (wires) radioactive sources or seeds arranged in a linear fashion (i.e., ribbons, etc.), whereas in permanent implants, multiple loose sources were distributed in random orientations. With the design of the low-energy radioactive sources in the form of stranded seeds or radioactive wires, now the permanent implant can also be performed with linear or pre-arranged ribbons. The loose seed implants are considered separately in this report. In planning temporary implants, the total time of implantation depends on the number of sources, their strengths, and the pattern of source distribution within the implant volume. In contrast, in permanent implants, the number of sources depends on their initial strength and type of the radioisotope. Both temporary and permanent implants are occasionally used as boost radiation therapy for external beam irradiation. In that case, the number of seeds and the initial activity of the seeds are adjusted upon the external beam radiation dose. In temporary implants, in the event of a non-ideal source pattern, there may be the possibility of improving the associated dose distribution through manipulation of the dwell time of some sources in the implant. However, at this time, such adjustment is not possible for permanent implants.
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
487
Source Specification The apparent source activity (mCi) and/or equivalent milligram of radium were two commonly used units in specifying the source strength. However, the recent American Association of Physicists in Medicine (AAPM) Task Group 43 (TG 43) (Nath et al. 1995) has recommended using air kerma (kinetic energy released in matter) strength for specification of the brachytherapy source strengths. The ICRU recommendation is also consistent with the TG 43 recommendation for specification of source strength. The reference air kerma rate of a brachytherapy source is the kerma rate to air, in air, at a reference distance of 1 m, corrected for air attenuation and scattering. For this purpose, this quantity is expressed in µGy·h–1 at 1 m, or mGy·h–1 at 1 m. For a linear brachytherapy source, the source strength is defined as the reference air kerma rate of the source divided by the equivalent active length of the source. Therefore, the quantity is expressed as µGy·h–1·cm–1. Please note that 1 µGy·h–1 at 1 m is equivalent to 1 cGy·h–1 at 1 cm, which is more commonly used in clinical applications.
Description of Source Patterns Since all the implants are essentially irradiating a volume of tissues, the term “volume implant” should not be used to describe a specific implant. Therefore, a more accurate description of the source pattern is briefly described as follows: 1. A single-plane implant is defined as an implant containing two or more sources that lie in the same plane. In some instances, the sources lie in a single curved surface. 2. A two-plane implant contains two planes, which are generally parallel to each other. 3. A multiplane implant contains three or more planes, and can often be described according to the number of planes of sources used. 4. A non-planar implant, which is not formed in recognizable planes, may be described by the location of the sources relative to a plane passing through the center of the implant or by a specific geometrical configuration (e.g., sphere or cylinder).
Total Reference Air Kerma The total reference air kerma (TRAK) is the sum of the products of the reference air kerma rate (Si) and the irradiation time (ti) for each source, as: TRAK = ∑ ti ⋅ Si .
(1)
This quantity is analogous to mg·h, which is proportional to the integral dose to the patient. Also, TRAK can serve as a useful index for radiation protection of personnel. The simple determination of the total reference air kerma does not, however, allow one to derive, even approximately, the absorbed dose in the immediate vicinity of the sources (i.e., in the tumor or target volume).
488
Ali S. Meigooni
Volume and Planes The gross tumor volume (GTV) and clinical target volume (CTV) in brachytherapy are defined entirely based on general oncological principles. Therefore, these definitions are identical to those given for external beam radiotherapy (see ICRU Report 50) (ICRU 1993). Gross Tumor Volume (GTV) The GTV is the gross palpable or visible/demonstrable extent and location of the malignant growth. According to this definition, there is no GTV after complete “gross” surgical resection. There is no GTV when there are only a few individual cells or “subclinical” involvement. Clinical Target Volume (CTV) The CTV is defined as the volume of the tissue that contains a GTV and/or subclinical microscopic malignant disease which has to be eliminated. Therefore, this volume of tissue has to be treated adequately in order to achieve the aim of therapy: cure or palliation. The CTV must always be described, independent of the dose distribution, in terms of the patient’s anatomy and the tumor volume. The CTV is a volume of tissue that needs to be irradiated, according to a specified dose-time pattern. As a minimum, the physical dimensions of the CTV are described in terms of its maximum dimension (cm) in three orthogonal directions (see Figure 1). Unlike the external therapy, the CTV in interstitial brachytherapy is sometimes selected at the time of implantation, on the assumption that it is contained within the minimum target isodose (see the section herein, Description of Dose Distribution, Minimum Target Dose). This procedure is not recommended by ICRU. The CTV should be clearly described in the patient chart before the implant is planned. Planning Target Volume (PTV) In external therapy, in order to ensure that all the tissues included in the CTV receive the appropriate dose, it is necessary to plan for a larger volume to be irradiated. This volume of tissue receiving the irradiation is defined as the planning target volume (PTV). In interstitial brachytherapy, the PTV is, in general, identical to the CTV with very few exceptions. For example, with some techniques in which there are uncertainties of consistency of source positions [high dose rate (HDR), moving sources, fractionated tech-
Figure 1. The physical dimensions of the CTV to be reported are the three maximum diameters measured in orthogonal directions. They are, in general, noted as width W, length L, and height H. (Reproduced from Figure 2-1, ICRU Report No. 58. © 1997, with permission from ICRU.)
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
489
niques] or alteration of source position (permanent implants) during the application, the PTV may be larger than the CTV to take these factors into account. However, in ICRU report 58, the term “CTV” is used rather than “PTV.” Treatment Volume The treatment volume is defined as the volume of tissue that is encompassed by an isodose surface that has been selected or specified by the radiation oncologist for a particular treatment. This isodose surface should, ideally, encompass the CTV. The dose value at this isodose surface is the minimum target dose (see the section herein, Description of Dose Distribution, Minimum Target Dose). Central Plane For source patterns in which the source lines are straight, parallel, of equal length, and with centers, which lie in a plane perpendicular to the direction of the source lines, this plane is the central plane (see the section herein, Description of Dose Distribution, Mean Central Dose). In an actual implant, all source lines may not necessarily be straight, parallel, and of equal length. In such cases, the central plane should be chosen perpendicular to the main direction of the source lines and passing through the estimated center of the implant. For more complex implants, it may be necessary to subdivide the target volume into two or more subvolumes for dose evaluation. In this event, a central plane may be defined for each of these subvolumes (see the section herein, Description of Dose Distribution, Mean Central Dose). The calculation of dose distributions in multiple planes throughout the target volume shows that a variation of a few millimeters, in the position of the central plane, is not critical.
Description of Dose Distribution General Concepts In brachytherapy, the dose distribution is nonhomogenous and includes steep dose gradients and regions of high dose surrounding each source. However, within the volume of the implant, there are regions where the dose gradient approximates a plateau (see Figure 2). In an interstitial implant, the regions of plateau dose are equidistant between adjacent neighboring sources, for sources of identical linear activity. They are regions of local minimum doses. Variations in the dose between the different plateau doses can be used to describe the dose uniformity of an implant. A region of plateau dose is the place where the dose can be calculated most reproducibly and compared easily by different departments. Dose Distribution in One or More Planes Through the Implant Although, in modern computer systems, the 3-D dose distribution can be computed and presented as isodose surfaces, these facilities are not yet available in many departments. In order to provide the minimum of information needed about the dose or dose-rate distribution, the calculation of isodose curves in at least one chosen plane is necessary. Methods to present dose information, either in tabular form or by graphical presentation have been discussed in ICRU report 42 (ICRU 1987). If only one plane is chosen for isodose calculation, the central plane of the implant (as defined in the previous section, Volume and Planes) should be chosen for this purpose. To assess the dose distribution in other areas of the implant, multiple planes for isodose calculation can be chosen, either parallel or perpendicular to the central plane.
490
Ali S. Meigooni
Figure 2. Plateau dose region between radioactive sources. The dose distribution in a plane perpendicular to linear and parallel sources shows a plateau dose region of low-dose gradient. (Reproduced from Figure 2-2, ICRU Report No. 58. © 1997, with permission from ICRU.)
Prescription Dose For purposes of this ICRU report, the prescribed dose is defined as the dose, which the physician intends to give, and enters in the patient’s treatment chart. Depending on the system used, the approach for dose prescription may be different. It is not the intention of this report to encourage users to depart from their normal practice of dose prescription. Minimum Target Dose The minimum target dose (MTD) is the minimum dose at the periphery of the CTV. It should be equal to the minimum dose decided upon by the clinician as adequate to treat the CTV. The minimum target isodose is the isodose surface corresponding to the MTD. It defines the treatment volume and should entirely encompass the clinical target volume (see the previous section, Volume and Planes). The MTD is known in some American centers as the “minimum peripheral dose.” The word “peripheral” is not recommended as being too vague and leading to confusion with the concept of peripheral dose in external therapy referring to the dose to healthy structures outside of the target volume. The MTD is known as the “reference dose” in the Paris system. The MTD is equal to about 90% of the prescribed dose in the Manchester system for interstitial therapy. Mean Central Dose In the field of brachytherapy, the mean central dose (MCD) is taken to be the arithmetic mean of the local minimum doses between sources, in the central plane, or in the central planes if there is more than one. In the case of a single-plane implant, the MCD is, in the central plane, the arithmetic mean of the doses at mid distance between each pair of adjacent source lines, taking into account the dose contribution at that point from all sources in the pattern (see Figures 3a and 3c). In the case of implants with line sources in more than one plane, the MCD is the arithmetic mean of the local minimum doses between each set of three adjacent source lines within the source pattern (Figures 3.b and 4). The minimum dose lies at the intersection of perpendicular bisectors of the sides of the triangles (geometric center) formed by these source lines. This point is equidistant from all three-source lines (see Figure 2).
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
(a)
491
(b)
(c) Figure. 3. In an implant where the source lines are rectilinear, parallel, and of equal length, the central plane is perpendicular to the direction of the source lines and passes through their centers. The mean central dose Dm is the arithmetic mean of the local minimum doses D in the plateau dose regions; (a) a planar implant; (b) a two-plane implant; (c) an actual single-plane implant where sources are not rectilinear. (Reproduced from Figure 2-3, ICRU Report No. 58. © 1997, with permission from ICRU.)
In some complex implants, a single central plane may not bisect or even include all the sources. In these cases, a MCD based on one plane can be misleading and it is advisable to subdivide the volume and to choose a separate central plane for each subvolume (see Figure 4). Three practical methods are acceptable for determining MCD. They include the following: 1. In the case of implants with parallel lines, identify triangles consisting of three adjacent source lines for all the sources, so that the triangles formed constitute as many acute triangles as possible. Determine the intersection points of the perpendicular bisectors of each triangle and calculate the local minimum dose at each of these points. The mean of these local minimum doses is the MCD (see Figures 3 and 4). This method is the most precise one for parallel lines. 2. Evaluation of dose profiles: Calculate dose profiles for one or more axes through the center of the implant expected to pass through as many local minima as possible. Determine, by inspection, the local minimum doses. The mean of these local minimum values is the MCD (see Figure 5). In a single surface implant performed following a curved surface, a profile may lead to an underestimation of the MCD. In a complex implant, it may be difficult to find axes passing through the minima and profiles may lead to an overestimation of the MCD. However, examples show that
492
Ali S. Meigooni
Figure 4. Two central planes (a) for the longest source lines and (b) for the shortest ones, in an implant. Two mean central doses are determined in the two subvolumes (a) Dma and (b) Dmb. Open circles are the intersections of sources with central plane, and closed circles are the points where the local minimum doses are calculated. (Reproduced from Figure 2-4, ICRU Report No. 58. © 1997, with permission from ICRU.)
(a)
(b)
Figure 5. Three profiles (b) are drawn along two orthogonal directions through a two-plane implant (a) with 8 parallel line sources, 10 cm long, 1.8 cm spacing. The profiles are calculated in percentage of the minimum target dose (thick line) along axes XX, YY, and Y’Y’ in the central plane. The profile along the axis YY is the most representative to estimate the MCD. The mean of the local minimum doses is the MCD. (Reproduced from Figure 2-5, ICRU Report No. 58. © 1997, with permission from ICRU.)
the error lies within acceptable limits. This method is some times preferred for seed implants. In a seed implant, such as the one presented in Figure 6, the dose should be calculated along several random profiles passing through the implant. 3. Inspection of dose distribution: Plot the dose distribution in the central plane. With isodose lines varying by 5% (at most 10%) of the local dose in the central region, the local minima can be deter-
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
493
mined by inspection. The mean of these local minima is the MCD (see Figure 7). This method is often preferred for complex implants with line sources. High-Dose Volume In order to correlate radiation dose with late damage, the high-dose volumes around sources should be assessed. There will inevitably be a high-dose zone around each source. Although it is often small and well tolerated, the exact tolerance dose and volume for interstitial therapy are not known. However, it is necessary, for intercomparison purposes, to agree on a way to describe the high dose volumes. It is suggested that a dose of approximately 100 Gy is likely to be the amount used in determining late effects. In those patients who receive 50 to 60 Gy as minimum target dose or 60 to 70 Gy as MCD, 100 Gy corresponds approximately to 150% of the MCD. It is therefore recommended that the size of the region receiving more than 150% of the MCD be reported. This will indicate that the high-dose volumes should be defined as the volumes encompassed by the isodose corresponding to 150% of the mean central dose around the sources in any plane parallel to the central plane where a high-dose region is suspected. The maximum dimension of the largest region in all planes calculated should be reported (see Figure 8).
Figure 6. The central plane is perpendicular to the main direction of the lines of implantation and passes through the center of the implant. (Reproduced from Figure 2-6, ICRU Report No. 58. © 1997, with permission from ICRU.)
Figure 7. Dose distribution in the central plane of an implant with six parallel iridium line sources, 6 cm long, 1.5 cm spacing, reference air kerma-rate 14.5 µGy·h–1 at 1 m. The dose varies by 5% between plotted isodose lines in the region of interest (A, B, C, D). The local minima, A, B, C, and D, can be easily estimated by inspection. DA and DD approximate 31 cGy·h–1 and DB and DC approximate 34 cGy·h–1. The estimated mean central dose is Dm = 32.5 cGy·h–1. (Reproduced from Figure 2-7, ICRU Report No. 58. © 1997, with permission from ICRU.)
494
Ali S. Meigooni
(a)
(b)
Figure 8. Dose distributions in two planes with three nonparallel loops loaded with seed ribbons. In (a), in a plane YY’ parallel to the main direction of the implant, the dose distribution does not display any unexpected high-dose region. In (b), in the central plane XX’ of the implant, the maximum dimension of the 150% isodose line (dotted line), varies from 6 mm to 27 mm. In lower perpendicular planes, the dimension of the high-dose region increases slowly to 33 mm. (Reproduced from Figure 2-8, ICRU Report No. 58. © 1997, with permission from ICRU.)
Low-Dose Volume A low-dose volume should be defined as a volume within the CTV, encompassed by an isodose corresponding to 90% of the prescribed dose. The maximum dimension of the low-dose volume in any plane calculated should be reported. In implants where the CTV is included within the minimum target isodose, the occurrence of a lowdose region is exceptional. If the CTV is not covered by the minimum target isodose, there will be low-dose regions due to geographical miss. To correlate the local recurrence rate with the dose distribution, it is recommended that low-dose volumes be reported. Dose Uniformity Parameters Several indices quantifying the homogeneity of the dose distribution have been proposed (see, for example, Paul, Koch, and Philip 1988; Wu, Ulin, and Sternick 1988; Saw and Suntharalingam, 1991). In this ICRU report, two parameters describing dose uniformity for interstitial implants are recommended. They can be derived directly from the concepts of MTD and MCD (see Figure 9): 1. The spread in the individual minimum doses used to calculate the MCD in the central plane expressed as a percentage of the MCD. 2. The dose homogeneity index (DHI), defined as the ratio of MTD to the MCD. Additional Representations of the Dose Distribution To obtain a full perception of the dose distribution of an implant, the use of volume-dose calculations has been advocated (see, for example, Neblett et al., 1985; Anderson 1986; Bridier et al., 1988; McCrae,
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
495
Figure. 9. Dose distribution in the central plane of a two-plane breast implant, with seven line sources, 10 cm long, 2 cm spacing, 90 cGy·h–1 at 1 m. The MCD is 70.9 cGy·h (the local minima in cGy·h are DA = 65.4, D = DD = 74.4, and D = 75.3). The MTD (100%) is 58.1 cGy·h–1. The spread in the individual minimum doses is from –8% to +6%. The dose homogeneity index, expressed as the ratio of the minimum target dose and the mean central dose, is 0.82 = 58.1/70.9. (Reproduced from Figure 2-9, ICRU Report No. 58. © 1997, with permission from ICRU.)
Rogers, and Dritschilo 1987). For this purpose, the CTV (or a larger volume including an additional margin) is subdivided in subvolumes (e.g., voxels) and the dose rate is calculated at the center of each subvolume. The volume receiving at least a specified dose is then defined as the sum of all subvolumes where, at the center at least, that dose is received. Examples of results are shown in Figure 10. Because of high-dose gradients, significant differences in calculated volumes can be observed, depending upon the size of the elementary subvolumes. The size of the grid and of the elementary subvolumes (voxels) used in dose and volume calculations should be clearly stated. Volume-dose data can also be represented by means of histograms, showing the distribution of fractions of the CTV receiving doses within chosen intervals. The value of these alternative representations of the dose distribution as possible prognostic factors for treatment outcome has still to be established in clinical research.
Time-Dose Factors General Considerations Considerable experience has been gained over many years with conventional dose rates. For removable implants, 60 to 70 Gy has usually been delivered in 4 to 8 days at a dose rate of 30 to 90 cGy·h. For permanent implants with 125I, doses of 120 to 160 Gy have been delivered with 50% of the dose received in the first 2 months and the majority of the remainder over the succeeding 6 months. Even with these conventional treatments, it has been recognized that the dose rates within and adjacent to the target volume vary considerably as a function of the distance from the sources and that these variations may be significant in determining effects on both tumor and normal tissues. The development of new afterloading techniques and, in particular, the use of HDR introduces new dose-time patterns that require evaluation. These include: • Continuous low dose rate (LDR); with short scheduled or unscheduled interruptions • Single moving source HDR used to treat several channels of an implant over several days to simulate continuous LDR (See the next section under Time-Dose Pattern for Temporary Implants, Pulsed irradiation.) • HDR; in a single fraction • Fractionated HDR
496
Ali S. Meigooni
Figure 10. Volume (sum of subvolumes receiving at least a certain dose) vs. dose, for two different patterns of parallel source lines: a two-plane implant with six sources 5 cm long (upper curve), and a cylindrical implant with seven sources 4 cm long (lower curve). The dose is expressed in percentage of the minimum target dose. The size of the voxel used for calculations is 1 mm (Bridier et al., 1988). (Reproduced from Figure 2-10, ICRU Report No. 58. © 1997, with permission from ICRU.)
• Combinations of external beam radiation with any type of brachytherapy with varying intervals between the two The overall treatment time for brachytherapy, and the duration of the interval(s) between treatments, can have an important effect on outcome. Therefore, the dose-time pattern should be recorded, and it is sometimes necessary to group different patterns for the purpose of analysis of outcome and intercomparisons. Time and Dose Rates for Temporary Implants Irradiation time is the time during which a radioactive source is present in the patient. The overall treatment time is the total time elapsed from the beginning of the first irradiation to the end of the last one. Instantaneous dose rate is the quotient of the dose and the irradiation time, for a given fraction or pulse. Average overall treatment dose rate is the quotient of the total dose and the overall treatment time. Average overall treatment dose rate is a concept useful for continuous LDR irradiations with or without short interruptions and for some pulsed irradiations (see the next section). Time-Dose Pattern for Temporary Implants Continuous irradiation. The overall treatment time does not differ from the irradiation time: only the instantaneous dose rate is considered. When the irradiation times of individual sources are different, the instantaneous dose rate varies with time and the average overall treatment dose rate is, in general, meaningless (Figure 11a). Noncontinuous irradiation. With the advent of remote afterloading, in most instances the overall treatment time is greater than the total irradiation time (which is the sum of the partial irradiation times) due to incidental or planned short interruptions during the treatment. The instantaneous dose rate is greater than the average overall treatment dose rate. In LDR irradiations, when the duration of one given
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
497
Figure 11. Overall treatment time and irradiation time for different types of treatment: A, continuous irradiation; B, noncontinuous irradiation; C, fractionated irradiation; D, hyperfractionated irradiation; E, pulsed irradiation in two fractions. (Reproduced from Figure 2-11, ICRU Report No. 58. © 1997, with permission from ICRU.)
interruption is longer than 10% of the total irradiation time, the irradiation should be considered “fractionated” irradiation. In this type of treatment, irradiation time is subdivided into multiple fractions. In fractionated irradiation, the overall treatment time is much greater than the total irradiation time. For fractionated irradiation, the instantaneous dose rate is the ratio of the dose per fraction and the irradiation time per fraction, and the aver age overall treatment dose rate is, in general, meaningless. Although the time interval between fractions is usually of the order of magnitude of a day or days, an LDR irradiation is considered fractionated when one given interruption is longer than 10% of the total irradiation time. The special case of multiple short irradiations with HDR source(s) is considered in the following paragraph. Hyperfractionated irradiation. When two or more fractions are given per day, the irradiation is considered a “hyperfractionated” irradiation. When the time interval between short HDR irradiations reaches or exceeds 4 hours, the irradiation should be considered a hyperfractionated irradiation. It should be considered as fractionated when the time interval is equal to one or several days. Pulsed irradiation. When a single HDR source is used to give a sequence of short irradiations (pulses) to simulate continuous LDR irradiation, the irradiation should be considered a “pulsed” irradiation as long as the time interval is shorter than 4 hours.
498
Ali S. Meigooni
Recommendations for Recording and Reporting It is recommended that adequate information be recorded to give a consistent description of any implant. The guidelines for reporting dose will make it possible to compare results of future brachytherapy practice and to better relate outcome to treatment. To report an implant the following should be recorded.
Parameters Required for Recording and Reporting Description of Volumes The description of the volumes should include, as a minimum, the GTV, the CTV, and the treatment volume. Description of Sources The description of the sources employed should include details of: 1. Radionuclide used, including filtration, if relevant. 2. Type of source used, i.e., wire, seeds, seed ribbon, hairpin, needle, etc. 3. Length of each source line used. 4. Reference air kerma rate of each source or source line. 5. The distribution of the strength within the source should be described (uniform or differential loading, etc.). Description of Technique and Source Pattern If the source distribution rules of a standard system have been followed, this shall be specified; if not, the source pattern should be described as explained previously in the section, Description of Source Patterns In addition, the following data should also be recorded: 1. Number of sources or source lines. 2. Separation between source lines and between planes. 3. Geometrical pattern formed by the sources with the central plane of the implant (e.g., triangles, squares), where relevant. 4. The surfaces in which the implant lies, i.e., planes or curved surfaces. 5. Whether crossing sources are placed at one or more ends of a group of linear sources. 6. The material of the inactive vector used to carry the radioactive sources, if any (e.g., flexible or rigid), whether rigid templates are used at one or both ends. 7. Type of remote afterloading, if used. Description of Time Pattern The description of the time pattern should include the type of irradiation with the necessary data on treatment and irradiation times as described below. The information on dose and time should provide the necessary data to calculate instantaneous and average dose rates. 1. Continuous irradiation: the overall treatment time should be recorded.
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
499
2. Noncontinuous irradiation: both the overall treatment time and the total irradiation time should be recorded, together with information about lengths of gaps. 3. Fractionated and hyperfractionated irradiation: the irradiation time of each fraction or pulse, the interval between fractions or pulses, and the overall treatment time should be recorded. 4. When the irradiation times of the different sources are not identical, they should be recorded. 5. Moving sources: a. Stepping sources: Step size and dwell time should be recorded if constant. Variation of the dwell times of a stepping source can be used for manipulating the dose distribution. If such a dose optimization is applied, this should be specified (e.g., optimization on dose points defined in the implant or geometrical optimization (Kolkman-Deurloo et al., 1994). For pulsed irradiation, at least two statements of dose rate may be necessary. One is the “pulse-average dose-rate,” which is the quotient of the pulse dose by the time from beginning to end of the pulse. The other is the maximum local dose rate at 1 cm from the stepping source. b. Oscillating sources: Speed in different sections of the vectors should be recorded. Total Reference Air Kerma The Total Reference Air Kerma (TRAK) for the total irradiation time should be recorded (see the previous discussion in the Definitions of Terms and Concepts section herein). Description of Dose Distribution The following doses should be recorded: 1. Prescribed Dose. If the dose is not prescribed at the level of either the MTD or the MCD, the method of dose prescription should be recorded. If, for clinical or technical reasons, the dose received differs from the prescribed dose, it should be noted (see the section Description of Dose Distribution, Prescription Dose). 2. The MTD should be recorded (see the section Description of Dose Distribution, Minimum Target Dose). 3. The MCD should be recorded (see the section Description of Dose Distribution, Mean Central Dose). 4. The following additional information, when available, should be recorded: a. Dimension of high-dose volume(s) (see the section Description of Dose Distribution, High-Dose Volume). b. Dimension of any low-dose volume (see the section Description of Dose Distribution, LowDose Volume). c. Any dose uniformity data (see the section Description of Dose Distribution, Dose Uniformity Parameters). d. Additional representation of dose distribution, if any (see the section Description of Dose Distribution, Additional Representations of the Dose Distribution).
500
Ali S. Meigooni
Priority Three levels of priority are recognized for reporting an interstitial therapy application. They are linked to the different levels of dose computation sophistication needed to fulfill the reporting requirements (Visser 1989). The wide variation in the availability of treatment-planning systems is recognized and taken into account in Tables 1 to 5.
Practical Applications of the Recommendations It is not the intention of this report and is not the task of the ICRU to encourage radiation oncologists to depart from their current practice of dose prescription or technique of application. Application of the reporting recommendations to existing systems and techniques is developed below.
Temporary Implants The recommended hierarchy of dose reporting, for a temporary implant is presented in Table 1.
Permanent Implant The majority of permanent implants are not done with recognizable source lines and it is therefore difficult to identify a central plane or to calculate a mean central dose. The recommended hierarchy of dose reporting for a permanent implant is presented in Table 2.
Single Stationary Source Line The source can be intraluminal or sometimes interstitial: LDR or HDR techniques can be applied. The recommended hierarchy of dose reporting, for a single source line, is presented in Table 3.
Moving Sources In addition to simulating a uniform line source, a moving source can be used to modify the dose distribution by changing the dwell time between moves or the speed of movement. The hierarchy of dose reporting of a moving source takes the form presented in Table 4.
Surface Applicators Although surface applicators are not interstitial implants, the physical factors which govern dose distribution from surface applicators are similar. At present, surface applicators are most commonly used for treating lesions involving skin or mucosal surfaces and the choroidal layer of the eye. When describing treatment by a surface applicator, the recommended hierarchy of reporting for surface applicators is presented in Table 5.
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
501
Table 1. Levels of Priority for Reporting Temporary Interstitial Implants Parameters for reporting temporary interstitial implants Description of Volume Gross Tumor Volume Clinical Target Volume Treated Volume Description of Source and Technique Radionuclide, type of source Source size and shape, source pattern Reference air kerma rate Inactivity vector (applicator), if any Description of Time Pattern Total Reference Air Kerma Description of Dose Prescribed dose including point or surface of prescriptionc. Reference Dose in Central Plane a. Mean Central Dose b. Minimum Target Dose Description of High and Low Dose Volume Uniformity Parameters Alternative Representation of Dose Distribution Dose Rates at Point or Surface of Prescription a
Prioritya
Levelb of computation
1 1 1 1
1 1 3 1
1 1
1 1
1
1
1 2 3 3 3
2 2 3 3 3 3
Priority 1. Concerned with doses in the central plan. 2. Required calculations outside the central plan. If this is not available, then a more detailed description of source pattern under priority 1 is required. 3. Additional information mostly of clinical research interest. b Level of computation 1. No computer needed. 2. Hand calculation and/or computer calculation in central plane. 3. 3-D computation needed. c Essential to establish consistent reporting and related to past experience, necessary for comparison of brachytherapy data and for relating outcome to treatment. If a classical system is used, the system should be identified. (Reproduced from Table 4.1 of ICRU Report 58, © 1997, with permission of ICRU.)
502
Ali S. Meigooni Table 2. Levels of Priority for Reporting Permanent Interstitial Implants Parameters for reporting permanent interstitial implants
Description of Volume Gross Tumor Volume Clinical Target Volume Treated Volume Description of Source and Technique Radionuclide, type of source Source size and shape, source pattern Reference air kerma rate Inactivity vector (applicator), if any Total Reference Air Kermac Description of Dose Prescribed dose including method of prescription Reference Dose in Central Plane c. Mean Central Dose d. Minimum Target Dose Description of High and Low Dose Volume Uniformity Parameters Alternative Representation of Dose Distribution a
Prioritya
Levelb of computation
1 1 1 1
1 1 3 1
1
1
1
1
1 1 2 3 3
2 2 3 3 3
Priority 1. Concerned with doses in the central plan. 2. Required calculations outside the central plan. If this is not available, then a more detailed description of source pattern under priority 1 is required. 3. Additional information mostly of clinical research interest. b Level of computation 1. No computer needed. 2. Hand calculation and/or computer calculation in central plane. 3. 3-D computation needed. c For permanent implants, the TRAK is calculated as the product of the total air kerma rate at the time of the implantation and the decay time constant λ, also known as the mean life time (λ = T/ln 2, where T is the half life). (Reproduced from Table 4.2 of ICRU Report 58, © 1997, with permission of ICRU.)
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
503
Table 3. Levels of Priority for Reporting Implants with a Single Stationary Source Line Parameters for reporting implants with a single stationary source line Description of Volume Gross Tumor Volume Clinical Target Volume Treated Volume Description of Source and Technique Radionuclide Length Shape (Straight/Curved) Reference air kerma rate Strength distribution (uniform linear strength is assumed, if not, distribution must be specified.) Diameter of inactive vector (applicator) Description of Time Pattern Total Reference Air Kerma Description of Dose and Prescription Point (Distance from the source line position along the source line) Minimum target dose if different from prescribed dose Dose at 1 cm from axis of the source line at its center Dose at the surface of applicator in contact with tissue Additional representation of dose distribution Dose Rate Average overall treated dose rate at the point or surface of dose prescription a
Prioritya
Levelb of computation
1 1 1 1
1 1 3 1
1 1 1
1 1 1
1 1 3 3
1 1 3 3
3
1
Priority 1. Concerned with doses in the central plan. 2. Required calculations outside the central plan. If this is not available, then a more detailed description of source pattern under priority 1 is required. 3. Additional information mostly of clinical research interest. b Level of computation 1. No computer needed. 2. Hand calculation and/or computer calculation in central plane. 3. 3-D computation needed. (Reproduced from Table 4.3 of ICRU Report 58, © 1997, with permission of ICRU.)
504
Ali S. Meigooni Table 4. Levels of Priority for Reporting Implants with Moving Source Parameters for reporting implants with moving source
Description of Volume Gross Tumor Volume Clinical Target Volume Treated Volume Description of Source and Technique Radionuclide, source type and source Type of movement Range of motion (effective length of source) Applicator-including diameter of inactive vector. Number of inactive vectors Reference air kerma rate Description of Time Pattern Intervals between fractions Irradiation time per fraction Total Reference Air Kerma Description of Dose Prescribed dose Minimum target dose For single source line or bifurcationc, dose at 1 cm For complex implant, mean central dose Method of dose optimization, if applicable Description of high and low dose volume Uniformity Parameters Additional representation of dose distribution Alternative Representation of Dose Distribution Dose Rates at point or surface of prescription a
Prioritya
Levelb of computation
1 1 1 1
1 1 3 1
1
1
1
1
1 1 1 1 2 2
1 2 2 2 3 2
3 3 3
3 3 3
Priority 1. Concerned with doses in the central plan. 2. Required calculations outside the central plan. If this is not available, then a more detailed description of source pattern under priority 1 is required. 3. Additional information mostly of clinical research interest. b Level of computation 1. No computer needed. 2. Hand calculation and/or computer calculation in central plane. 3. 3-D computation needed. c If there is a bifurcation, more than one target volume should be considered. d Continuous/step wise, size of step, unidirectional/oscillating. Uniform motion is assumed; if not, motion must be specified. (Reproduced from Table 4.4 of ICRU Report 58, © 1997, with permission of ICRU.)
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
505
Table 5. Levels of Priority for Reporting Use of Surface Applicators Parameters for reporting use of surface applicators Description of Clinical Target Volume Treated Volume Description of Applicator Shape (flat/curved, round/square, etc.) Size Description of Source Radionuclide and chemical form Concentration (seeds/tubes/plated)c Description of Time Pattern Intervals between fractions Irradiation time per fraction Total Reference Air Kerma Description of Dose Prescription Dose and point of dose prescription Dose at 5 mm in tissue at the center of the applicatord Minimum target dosee Description of high dose at tissue surface in contact with applicator, usually near the center of the applicator. Uniformity Parameters Additional representation of dose distribution Dose Rates Average dose rate at the point of dose prescription a
Prioritya computation
Levelb of
1 1 1
1 3 1
1
1
1
1
1
1
1
1
1 1 1
1 2 2
2
2
3
3
3
3
Priority 1. Concerned with doses in the central plan. 2. Required calculations outside the central plan. If this is not available, then a more detailed description of source pattern under priority 1 is required. 3. Additional information mostly of clinical research interest. b Level of computation 1. No computer needed. 2. Hand calculation and/or computer calculation in central plane. 3. 3-D computation needed. c Including distance from source(s) to the surface of the applicator. d For eye plaque, 5 mm from the internal sclera. e Dose at the distal extent of the clinical target volume. (Reproduced from Table 4.5 of ICRU Report 58, © 1997, with permission of ICRU.)
506
Ali S. Meigooni
Appendix: Quantities and Units Basic Quantities and Units The quantities and units used in brachytherapy are generally related to the dosimetric aspects of this treatment modality. ICRU report 33 (ICRU 1980) has a detail description of these parameters. The following information provides most relevant information for this chapter.
Definition of Quantities 1. Absorbed dose: The mean energy imparted by the radiation to a unit mass of medium. D=
dε dm
The SI unit of absorbed dose is J.kg–1. The special unit of the absorbed dose is gray (Gy) where 1 Gy = 1 J·kg–1. The subunit of absorbed dose is cGy, where 1 cGy = 1/100 Gy. 2. Kerma (kinetic energy released in matter): The sum of the initial kinetic energies of all the ionizing charged particles liberated by uncharged ionizing particles. K=
dEtr dm
The SI unit of kerma is J·kg–1. The special unit of the kerma is the gray (Gy) where 1 Gy = 1 J·kg–1. 3. The activity, A, is the Number of disintegration of a radioactive nucleus per second (dps). A=
dN dt
The SI unit of kerma is s–1. The special unit of the activity is the becquerel (Bq) where 1 Bq = 1 s–1. The more commonly used unit of activity is the curie (Ci) where 1 Ci = 2.7 × 1010 dps. 4. The air kerma-rate constant, Γδ, of a radioactive nuclide emitting photon is defined as: Γδ =
" 2 K# δ
,
A
where K# δ is the air kerma rate due to photons of energy greater than δ, at a distance, ", from a point source, with activity of A. The SI unit of air kerma constant is J.kg–1m2. The special unit of the air kerma constant is U, where 1 U = 1 µGy.m2.h–1 = 1 cGy.cm2.h–1.
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
507
Relationship Between the Quantities Absorbed dose and kerma in a reference material m
µ Dm = K air tr (1 − g ) ρ air where g is the small part of the kerma which is lost to bremsstrahlung in medium, and (µtr/r ) is the energy transfer coefficient of the radiation for a specific medium.
Reference Air Kerma Rate At the time when radium was used as brachytherapy sources, their strength was determined by measuring the mass of the radium. Subsequently, 1 gram of radium was specified as 1 Ci. As other radioactive materials become available, the concept of apparent activity (the activity of an unencapsulated point source, which could produce the same exposure rate as the capsulated source, at a given point) has been introduced. However, for low-energy photon emitters, the accuracy of the geometric structure of the sources was affecting this parameter and introducing some discrepancies among various dosimetric information. To eliminate these discrepancies, air-kerma rate was introduced by the AAPM (AAPM, 1987; Nath et al. 1995) for brachytherapy source determinations. The unit of air kerma strength is U, where where 1U = 1µGy·m2·h–1 = 1cGy·cm2·h–1.
References American Association of Physicists in Medicine (AAPM). Report No. 21. “Specification of Brachytherapy Source Strength.” New York: American Institute of Physics, New York, 1987. Anderson, L. L. (1986). “A ‘natural’ volume-dose histogram for brachytherapy.” Med Phys 13:898–903. Bridier, A., H. Kafrouni, J. P. Houlard, and A. Dutreix. “Comparison des distributions de dose en curietherapie interstitielle autour des sources continues et discontinues.” Tnt. Symp. on Dosimetry in Radiotherapie. IAEA SM-298/23. Vienna: IAEA, 1988. International Commission on Radiation Units and Measurements (ICRU). Report No. 33. “Radiation Quantities and Units.” Bethesda, MD: ICRU, 1980. International Commission on Radiation Units and Measurements (ICRU). Report No. 38. “Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology.” Bethesda, MD: ICRU, 1985. International Commission on Radiation Units and Measurements (ICRU). Report No. 42. “Use of Computers in External Beam Radiotherapy Procedures with High-Energy Photons and Electrons.” Bethesda, MD: ICRU, 1987. International Commission on Radiation Units and Measurements (ICRU). Report No. 50. “Prescribing, Recording and Reporting Photon Beam Therapy.” Bethesda, MD: ICRU, 1993. International Commission on Radiation Units and Measurements (ICRU). Report No. 58. “Dose and Volume Specification for Reporting Interstitial Therapy.” Bethesda, MD: ICRU, 1997. Kolkman-Deurloo, I. K., A. G. Visser, C. G. Niel, N. Driver, and P. C. Levendag. (1994). “Optimization of interstitial volume implants.” Radiother Oncol 31:229–239. McCrae, D., J. Rogers, and A. Dritschilo. (1987). “Dose-volume and complication in interstitial implants for breast carcinoma.” Int J Radiat Oncol Biol Phys 13:525–529. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee, Task Group No. 43.” Med Phys 22:209–234. Also available as AAPM Report No. 51. Neblett, D., S. A. M.Nisar, A. A. Puthawala, R. Habrop, H. S. Frey, and S. E. Hogan. (1985). “An interstitial implant technique evaluated by contiguous volume analysis.” Hypertherm Oncol 1:213–221.
508
Ali S. Meigooni
Paul, J. M., R. F. Koch, and P. C. Philip. (1988). “Uniform analysis of dose distribution in intersti tial brachytherapy dosimetry systems.” Radiother Oncol 13:105–125. Saw, C. B., and N. Suntharalingam. (1991). “Quantitative assessment of interstitial implants.” Int J Radiat Oncol Biol Phys 20:135–139. Visser, A. G. (1989). “An intercomparison of the accuracy of computer planning systems for brachytherapy.” Radiother Oncol 15:245–258.
Bibliography American Association of Physicists in Medicine (AAPM). Report No. 41. “Remote Afterloading Technology.” New York: American Institute of Physics, New York, 1993. Bernard, M., B. Guille, and G. Duvalet. (1975). “Mesure du debit d’exposition linéique nominal des sources a une dimension, utilisées en curiethérapie.” J Radiol Electrol 56:785–790. British Institute of Radiology (BIR). “Recommendations for Brachytherapy Dosimetry.” Report of a joint working party of the BIR and the Institute of Physical Sciences in Medicine (IPSM). London: British Institute of Radiology, 1993. British Committee on Radiation Units and Measurements (BCRU). (1984). “Specification of brachytherapy sources, memorandum from the British Committee on Radiation Units and Measurements.” Br J Radiol 57:941–942. Boutillon, M. “Values of g for Photon Energies” in CCEMRI Report (1)185-18. Paris: Consultative Committee for Standards of Measuring Ionizing Radiations, 1985. Cance, M., and J. P. Simoen. (1983). “Etalonnage des sources de curiethérapie” Compte-Rendus du Congrès de la Société Française des Physiciens d’Hépitaux. Paris: French Society of Hospital Physicists (SFPH), pp. 87–104, 1983. Comité Francais de Mesure des Rayonnements Ionisants (CMFRI). “Recommendations pour la determination des doses absorbees en curiethérapie. Rapport du Comite Francais ‘Mesure des Rayonnements Ionisants’ No. 1.” Paris: Bureau National de Metrologie, 1983. Dutreix, A., and A. Wambersie. (1968). “Étude de la repartition des doses autour de sources ponctu elles alignées.”Acta Radiol 7:389–400. Dutreix, A., and A. Wambersie. (1975). “Specification of gamma-ray brachytherapy sources.” Br J Radiol 48:1034. Dutreix, A., G. Marinello, and A. Wambersie. Dosimetrie en Curiethérapie. Paris: Masson, 1982. Goetsch, S. J., F. H. Attix, D. W. Pearson, and B. R. Thomadsen. (1991). “Calibration of 192Ir high dose-rate afterloading systems.” Med Phys 18:462–467. International Commission on Radiation Units and Measurements (ICRU). Report No. 24. “Determination of Absorbed Dose in a Patient Irradiated by Beams of X or Gamma Rays in Radiotherapy Procedures.” Bethesda, MD: ICRU, 1976. International Commission on Radiation Units and Measurements (ICRU). Report No. 29. “Dose Specification for Reporting External Beam Therapy with Photons and Electrons.” Bethesda, MD: ICRU, 1978. International Commission on Radiation Units and Measurements (ICRU). Report No. 35. “Radiation Dosimetry: Electron Beams with Energies Between 1 and 50 MeV.” Bethesda, MD: ICRU, 1984. Kutcher, G. J., L. Coia, M. Gillin, W. F. Hanson, S. Leibel, R. J. Morton, J. R. Palta, J. A. Purdy, L. E. Reinstein, G. K. Svensson, M. Weller, and L. Wingfield. (1994). “Comprehensive QA for radiation oncology: Report of AAPM radiation therapy committee task group 40.” Med Phys 21:581–618. Also available as AAPM Report No. 46. Ling, C. C., and Z. C. Gromadzki. (1981). “Activity uniformity of 192Ir seeds.” Int J Radiat Oncol Biol Phys 7:665–669. Ling, C. C., E. D. Yorke, I. J. Spiro, D. Kubiatowicz, and D. Bennett. (1983). “Physical dosimetry of I-125 seeds of a new design for interstitial implant.” Int J Radiat Oncol Biol Phys 9:1747–1752. Marinello, G., M. Valéro, S. Leung, and B. Pierquin. (1985). “Comparative dosimetry between iridium wires and seed ribbons.” Int J Radiat Oncol Biol Phys 11:1733–1739. Meredith, W. J. (ed.). Radium Dosage: The Manchester System. Edinburgh: Livingston, 1967. National Council on Radiation Protection and Measurements (NCRP). Report No. 41. “Specification of GammaRay Brachytherapy Sources.” Bethesda, MD: NCRP, 1974.
26–ICRU Recommended Dose and Volume Specifications for Brachytherapy
509
Netherlands Commission on Radiation Dosimetry (NCS). Report No. 4. “Recommendations for Dosimetry and Quality Control of Radioactive Sources Used in Brachytherapy.” Bilthoven: (Netherlands Commission on Radiation Dosimetry, 1991. Pierquin, B., A. Dutreix, C. H. Paine, D. Chassagne, G. Marinello, and D. Ash. (1978). “The Paris system in interstitial radiation therapy.” Acta Radiol Oncol 17:33–48. Pierquin, B., J. F. Wilson, and D. Chassagne. Modern Brachytherapy. New York: Masson, 1987. Quimby, E. H., and V. Castro. (1953). “The calculation of dosage in interstitial radium therapy.” Am J Roentgenol 70:739–749. Sralek, R. J., and M. Stovall. “Brachytherapy Dosimetry” in The Dosimetry of Ionizing Radiation. K. R. Kase, B. E. Bjarngard and F. H. Attix (eds.). New York: Academic Press, pp. 259–321, 1990. Steggerda, M. J., and B. J. Mijnheer. (1994). “Replacement corrections of a Farmer-type ionization chamber for the calibration of Cs-137 and Ir-192 sources in a solid phantom.” Radiother Oncol 31:76–84. Thomason, C., T. R. Mackie, and M. J. Lindstrom. (1991). “Effect of source encapsulation on the energy spectra of 192 Ir and 137Cs seed sources.” Phys Med Biol 36:496–506. Venselaar, J. L., W. F. Brouwer, B. H. van Straaten, and A. H. Aalbers. (1994). “Intercomparison of calibration procedures for Ir-192 HDR sources in The Netherlands and Belgium.” Radiother Oncol 30:155–161. Williamson, J. F. (1988). “Monte Carlo evaluation of specific dose constants in water for 125Iseeds.” Med Phys 15:686–694. Williamson, J. F., F. M. Khan, S. C. Sharma, and G. D. Fullerton. (1982). “Methods for the routine calibration of brachytherapy sources,” Radiology 142:511–516. Wu, A., K. Ulin, and E. S. Sternick. (1988). “A dose homogeneity index for evaluating Ir-192 interstitial implants.” Med Phys 15:104–107. Young, M. E. J., and H. F. Batho. (1964). “Dose tables for linear radium sources calculated by an electronic computer.” Br J Radiol 37:38–44.
Chapter 27
Intraoperative Radiation Therapy (IORT) Gil’ad N. Cohen, M.S., and Marco Zaider, Ph.D. Department of Medical Physics Memorial Sloan-Kettering Cancer Center New York, New York Clinical Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 511 Hardware. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512 Afterloading Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512 Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 512 Cost of IORT Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513 Treatment Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513 Dose Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513 Treatment-Planning Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514 Computer-Assisted Plan Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 514 Determination of Implant Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515 Dose Calculation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516 Classical Dosimetric Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 516 QA Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517 Facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517 Design and Shielding Considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 517 Personnel Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519 Emergency Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519
Clinical Overview Ideally, radiation treatments would deliver a curative dose to the tumor and completely spare surrounding healthy tissue. In the clinic, however, delivering the dose required for local tumor control may result in unacceptable normal tissue complications, in effect placing an upper limit on the dose deliverable using external beam radiation therapy (EBRT). This is of particular relevance for locally recurrent tumors, some of which may have been treated previously, and pediatric tumors, where normal tissue toxicity requires special consideration.Various treatment techniques have been proposed to overcome this limitation, including external beam (EB) intensity-modulated radiation therapy (IMRT) and brachytherapy. Briefly, IORT refers to a single fraction treatment delivered to a surgically exposed target area. Two competing approaches to IORT are currently in practice: the first is linac based using electron beams; the second employs a high dose rate (HDR) 192Ir afterloader. The brachytherapy-based intraoperative approach is of particular interest because it is best suited to maximize the therapeutic ratio and deliver doses higher than any other treatment modality. For instance: • The treatment is performed at the time of surgery, when the target area (the tumor bed) is exposed and the applicator can be placed directly over the target. • Organs at risk may be retracted and shielded as necessary. • This technique is free of any accessibility limitations: the applicator can be used in virtually any anatomic location. This is most important in the treatment of colorectal malignancies where the tumor bed is often inaccessible to the cones of a linac-based system (Harrison et al. 1998).
511
512
Gil’ad N. Cohen and Marco Zaider
• The flexible applicator used for HDR-IORT easily conforms to the target area, ensuring uniform dose delivery throughout the target (avoiding cold spots and hot spots often encountered when using the electron beam approach as a result of angle of beam incidence and field matching). Finally, beyond the clinical and technical considerations, IORT offers the convenience and cost effectiveness of accelerated-course radiotherapy used more recently for the treatment of breast cancer. In this chapter it is assumed that the reader is familiar with the general requirements for HDR brachytherapy. What follows is a summary of the special needs and peculiarities of brachytherapy based IORT.
Hardware Afterloading Devices The three HDR afterloading machines most commonly used are manufactured by Nucletron, BV (The Netherlands) and by Varian Medical Systems, Inc. (GammaMed® and VariSource™), and are all suitable for use in IORT.
Applicators Several applicators are mentioned in the literature (Nag et al. 1994), however, some may not have Food and Drug Administration (FDA) 510K approval. The user is urged to contact the vendor of the afterloader for further information. Described here is the Memorial Sloan-Kettering Cancer Center (MSKCC) experience with the Harrison-Anderson-Mick (HAM) applicator (Mick Radio-Nuclear Instruments, Mount Vernon, NY), as used with a GammaMed 12i afterloader. Versions of this applicator are also available for the VariSource and Nucletron afterloaders. The standard HAM applicator (Figure 1) consists of standard-length (130 cm) catheters embedded at 1 cm spacing in 8 mm thick silastic rubber: with 5 mm from the center of catheters to the front of the appli-
Figure 1. Picture of a standard HAM applicator (courtesy of Mick Radio Nuclear Instruments, Inc., Mount Vernon, NY).
27–Intraoperative Radiation Therapy (IORT)
513
cator and 3 mm to the back. The asymmetry, reducing the overall thickness of the applicator, was chosen for added flexibility. The number of catheters varies from 2 to 24, and the length of the silastic is 22 cm, resulting in a treatment area up to 23 cm in width and 20 cm in length. The prescription point is usually 0.5 cm from the surface of the applicator (or 1 cm from the plane of the catheters). A modified HAM applicator for use in breast-IORT is shown in Figure 2. Here, the silastic rubber is molded symmetrically with at least 1 cm from the surface to any of the catheters. This applicator is fitted with a tungsten shield, to protect the skin at the incision. Prescription using this applicator is usually 1 cm from the surface of the applicator (2 cm from the source plane).
Cost of IORT Brachytherapy On the average, the applicator cost is about $100.00/channel. Source exchange and periodic maintenance (assuming quarterly source exchanges) are approximately $50,000.00/year. With an average treatment width of five channels, and one treatment per week, the cost per treatment, excluding initial investments, is approximately $1,500.00.
Treatment Planning Dose Specification Being in an operating room (OR), this is arguably one of a few instances in which a medical physicist will be working upon oral instructions rather than on written directive (however, standard OR procedures should be followed: details of the prescription should be clearly repeated by the physicist and confirmed by the physician). An IORT prescription typically specifies the width (or number of channels), the length (or number of positions if the source stepping distance is known), the dose required, and the prescription point. Typically, a homogenous dose distribution is desired throughout the treatment area, with doses in the range of 15 to 20 Gy. Lower doses (10 to 15 Gy) are usually used for pediatric cases, and when IORT is used as a boost in conjunction with EBRT treatment. Occasionally, nonrectangular target areas and critical organs may be specified. The orientation of the applicator must be established to implement such dose prescriptions correctly.
Figure 2. The breast applicator modeled after the HAM applicator (Figure 1) is 2 cm thick and is used in tandem with a matching tungsten shield to protect the skin at the incision area.
514
Gil’ad N. Cohen and Marco Zaider
Treatment-Planning Approaches Most IORT treatment plans may assume fixed applicator geometry and be generated from pre-calculated plans or template-based plans. These are essentially fixed treatment geometries that are stored in, or automatically generated by, the treatment-planning system. These treatment-planning schemes are attractive mainly for their simplicity, reliability, and fast output (Anderson et al. 1997). Occasionally, custom plans are needed to accommodate regions of dose escalation or dose sparing. Figure 3 is an illustration of a plan with a dose escalation region in the middle of the treated area. Standard IORT plans are symmetric by nature. However, this often is not the case for custom treatments. In those cases, one should take care to note the orientation of the applicator with respect to the patient (e.g., channel 1-Med; applicator tip-Post); channel numbering tags are mounted on the catheters for this purpose (see Figure 1). While the most efficient of dose sparing methods employ a combination of retraction and shielding of the organs at risk, in some cases anatomic/geometric constraints require that the dose be tailored by means of treatment plan optimization. Breast IORT, for example, aims at delivering 20 Gy to a surface 1 cm away from the applicator while sparing the skin (ideally the skin dose should not exceed 10 Gy). In this type of geometry the skin is at the applicator surface, and very close to the treatment volume. It is usually possible to achieve a skin dose of 15 Gy by means of dose optimization, and further reduce the skin dose to 10 Gy using shielding (Figure 2). It is assumed that most facilities will be using the treatment planning system supplied with the afterloader, and that the physicist involved is familiar with that planning system. The details of the planning are left to be worked out by the physicist for the planning system available to him/her in a manner compatible with the clinical needs of the respective hospital.
Computer-Assisted Plan Verification Plan verification, unlike treatment planning, is not implemented in commercial software packages. Several schemes have been suggested (Ezzell 1994; Venselaer, Bierhuizen, and Klop 1995; Rogus, Smith, and Kubo 1998; Miller, Davis, and Horton 1996; Saw et al. 1998), most of which address only singlecatheter treatments. HDR brachytherapy plans are complex in nature; they involve many source-stopping positions with widely varying dwell times, and the time allotted for planning and verification is limited.
Figure 3. An example of a custom plan generated using Abacus (Varian Medical Systems, Charlottesville, VA) treating the center of the target to 17.5 Gy and the periphery of the target to 12.5 Gy.
27–Intraoperative Radiation Therapy (IORT)
515
Consequently, it is impractical to manually review the plan for errors. To address this problem, various implementations of computerized quality assurance (QA) using spreadsheet calculations are also used. Typically, the physicist will type dwell times into a spreadsheet form and calculate the dose at one or more points of interest. However, manual data entry slows the QA process, and makes this method susceptible to errors, especially if used for large implants. As a means for independent plan verification, it is suggested that a computer program be used to assist in this task (Cohen, Amols, and Zaider 2000). Here, we describe in detail the program used for HDR plan verification at MSKCC. For completeness, we shall broaden the scope of this discussion to any procedure (whether or not invasive, or under anesthesia) for which a patient is on the treatment table, waiting for the plan to be completed and the treatment to commence. This definition includes (at our hospital) cylinders for vaginal treatment, ring and tandem for cervical cancer, and endobronchial treatments of the lung. The QA of HDR treatment planning in general has been discussed extensively in the report of AAPM TG-59 (Kubo et al. 1998). This program is an efficient tool assisting the physicist in implementing AAPM recommendations. Once a treatment plan is completed, the treatment-planning system generates a data file, which contains all the parameters necessary for treatment. This information, which is transferred to the afterloader console for treatment delivery, serves as the main data source for the independent verification program. Use of this file for secondary-dose calculations offers several obvious advantages. It provides an orderly way to: (a) read treatment parameters into the dose verification program, (b) confirm that the treatment data file corresponds to the actual treatment plan, (c) ensure that the data file is intact, and (d) use actual treatment parameters to both verify treatment geometry and perform a secondary dose calculation. The secondary verification consists of two steps: (a) determination/validation of treatment geometry, and (b) dose calculation based on that geometry. Once these two steps are completed the independent reviewer can verify that the treatment file conforms to the desired treatment parameters. User input is minimal. Determination of Implant Geometry The main task in the determination of implant geometry is the correct reconstruction of dwell positions (DPs). Each DP is defined by a nominal dwell time and a set of coordinates in the patient’s coordinate space (X, Y, Z). For fixed geometry implants, DP coordinates are implicitly defined by the configuration of active channels. In general, for other implant geometries such as template-based implants, DP coordinates may be approximated using explicit user input. However, this is beyond the scope of this discussion. Applicator geometry for all procedures mentioned above is assumed fixed and can be encoded in the software. The program detects automatically the type of applicator being used, virtually eliminating user input. Specifically, the number of active channels and channel index indicate which applicator is used. Because for most applicators the location of the prescription point is also fixed, the complete implant geometry can be automatically reconstructed. In the case of IORT, the active treatment area is defined by the length L (Y-axis) and width W (X-axis), with the origin at the center of the implant (Figure 1). Assuming a flat applicator (Z = 0 for all DPs) with N channels and M positions per channel, the coordinates of DP(n,m) are given by:
N −1
(X ,Y ) n. m =
2
− (n − 1) × ChSp,
M −1 − (m −1) × Stepn − Shiftn 2
(1.1)
where “Shift” is the distance the first dwell position is shifted from the tip of the applicator, “Step” is the distance between dwell point positions, and “ChSp” is the spacing between channels (which in the case of the HAM applicator is fixed and equals 1 cm; see Figure 1). In the case of a single channel application, where N=1, Xn,m reduces to X1,m = 0. It should be noted that in the specific case of the HAM applicator, each “atlas” treatment plan uses a fixed length (i.e., M is constant) for all channels. This, however, need
516
Gil’ad N. Cohen and Marco Zaider
not be the case (see description of custom plans above), because the program reads all 40 dwell times for each of the 24 channels, whether or not they are used. The detection of the applicator type and the calculation of the dose at the reference point(s) are performed as follows: 1.Single channel applications: (channel no. < = 19) indicates a vaginal applicator; (channel no. > = 20) indicates a bronchial/esophageal applicator. [The GammaMed afterloader automatically tests the length of channels 1 to 19 (of 24 channels) to ensure that the first dwell position corresponds with the end of the catheter/tube. Unlike vaginal applicators, endoesophageal and endobronchial applicators are subject to distortion and variations in length, and are typically attached to channel no. 24 of the afterloader.] At our institution, the dose reference point for a vaginal procedure is set at Pref = (0.0, 0.0, d/2+0.5) cm, where d is the cylinder diameter. Because d is variable, the distance of Pref from the source channel is entered manually. For bronchial or esophageal procedures, Pref = (0.0, 0.0, 1.0) cm. 2.Two active channels indicate GYN ring and tandem applicator. Here, the user is required to enter the length of the tandem. With the tandem orthogonal to the plane of the ring, and the channel in the ring section of the applicator also fixed, the applicator geometry is defined. In some cases (applicable only when the tandem is made of two sections), the angle of the sleeve may not be perpendicular to the plane of the ring; then user input is required. The reference points defined as points Aright and Aleft are fixed relative to the applicator hardware. The outer diameter of the ring (i.e., of the caps) is also entered, to calculate the dose to the vaginal mucosa. Uterine and cervix points1 are also fixed with respect to the applicator and are automatically calculated. 3.Three or more active channels indicate a HAM applicator. The assumption made here is that the applicator is flat. We found this to be a good approximation for most procedures. For HAM-IORT, we set the dose reference point at Pref = (0.0, 0.0, 1.0) cm (1.0 cm from the source plane, or 0.5 cm from the surface of the applicator) and for breast IORT at Pref = (0.0, 0.0, 2.0) cm. In the case of a convex curved application (e.g., rectal IORT) the dose calculation will predict an overdose of 5% to 10%. While this is still within regulatory requirements, our treatment planning “Atlas” and dose verification program can both accommodate the curved geometry. Dose Calculation Dose-to-point calculations consist of summation of dose contributions from all dwell positions, using a two-dimensional dose table, F(r,σ), for the 192Ir source, and the fifth-order Meisberger polynomial, M(r) (Meisberger, Keller, and Shalek 1986). Both the dose table and polynomial are hard coded in the program, making it self contained. Dose calculations are typically within 1% to 2% of the planned dose. By the end of plan verification, the total treatment time is also known. The total treatment time (including dummy tests) should be communicated to the anesthesiologist, to ensure proper medication of the patient. Also at this time, surgeons, the radiation oncologist, and the OR staff are ready to start treatment. The last step in the plan verification process is (oral) approval of the plan by the physician. It is important to “pause” and review the plan with the physician: this is the last chance to catch any miscommunications or other errors in an otherwise busy and time-pressured environment.
Classical Dosimetric Systems To complement the plan verification discussed above, the classical Manchester system may be used. The original Manchester tables were designed to give the total amount of radium (mg) times application time (h) required to deliver 1000 R for various implants, given the implant’s volume or area, and elongation (Merredith 1967). This quantity, M, differs only by a constant from total reference air kerma (TRAK) 1
The cervix and uterine points are defined in a similar manner as point A: the cervix is 1 cm superior to the flange and 1 cm lateral to the tandem, and the uterine points are 1 cm inferior to the tip of the tandem and 2 cm lateral to the tandem.
27–Intraoperative Radiation Therapy (IORT)
517
expressed in gray (ICRU 1997). Because mg·h is still most commonly used, M is expressed as the total mg·h that produce 10 Gy in water. Using the ratio of exposure rate constants for radium and iridium and adjusting for units, the total dwell time for a nominal source activity of 370 GBq, expressed in s/Gy, is interpolated from the Manchester table, M (expressed in mg·h), as follows:
0.5 2 h 2 3/ 4 T = 0.0659 M A , 0.5 exp(0.05[E − 1] ) , h 0.5
(1.2)
where h is the distance (cm) from the plane of source to the treatment plane, A is the treatment area (cm2), and E is the elongation of A. The original Manchester tables were written for a treatment distance of 0.5 cm from the plane of the source. The equation above reflects the use of the table at other treatment distances as well. Predictions of the total nominal dwell time using this method are usually within 5% of the planned time.
QA Considerations In general, QA procedures should follow standard recommendations (Nath et al. 1997). However, it is imperative to remember that IORT is a single-fraction treatment. Proper QA of the hardware, software, and each treatment plan is essential for a successful IORT program. Special attention should be given to applicator QA: • The applicator is a single-use device that requires in-house sterilization. Each applicator should undergo QA prior to sterilization. Specifically, the QA procedure should include checks of catheter integrity and lengths, catheter labeling, and overall applicator integrity. If source transfer tubes are used, the QA should include them and they should be sterilized as well. • It is equally important to examine the applicator after the procedure is complete and to establish its integrity; if lead shields or source transfer guides were used, ensure that all are intact and accounted for. • Initial QA at the time of the IORT program is being set up should be performed to ensure the applicator and source transfer tubes are not damaged during the sterilization process. The operation room is a high-pressure environment. It is recommended that standardized forms and treatment protocols be used whenever possible. An example of a treatment setup form is provided in Figure 4.
Facility Design and Shielding Considerations The requirement of a shielded operating room is for the most part identical to those of a standard HDR treatment suite. A few design considerations are noteworthy: • Remote monitoring of the patient, usually placed at the treatment console, should include a monitoring station for anesthesia staff. • Essential OR staff and the radiation oncologist must remain “scrubbed” (i.e., they may not break sterility) during the procedure in case an emergency intervention is required. Room should be provided for them to wait throughout the procedure, preferably near the entrance to the OR.
518
Gil’ad N. Cohen and Marco Zaider
Figure 4. A template for an IORT worksheet. The redundancy of numerical entry (in the header of the form) and graphical representation helps ensure the oral directives are followed correctly.
27–Intraoperative Radiation Therapy (IORT)
519
• In the event of a radiation emergency, the patient may not leave the OR. Instead, the facility should include a shielded enclosure into which the afterloader and source may be retracted and isolated, to enable completion of surgery. • In general, operating rooms tend to become crowded very easily, occasionally hampering access to the patient. To ensure easy setup of the remote afterloader, ensure all cables are long enough to reach all sides of the patient/room.
Personnel Requirements Radiation oncology staff required to be present for each procedure include: a physician, a therapist, and two physicists (one to plan and deliver the treatment, and one for independent verification). Surgery often starts early morning and in some cases may continue until late at night. When the patient will be ready for IORT is not always predictable, which may place yet an additional strain on busy physics staff. Good planning may often help, but do realize that long, waiting hours will sometimes be unavoidable.
Emergency Procedures Emergency procedures during IORT follow the same guidelines as other HDR emergency procedures but for one crucial exception: the patient may not leave the operating room. As discussed above, the source and afterloader should be isolated from the patient, preferably into a shielded housing in the corner of the operating room.
References Anderson, L. L., M. R. Hoffman, P. J. Harrington, and G. Starkschall. (1997). “Atlas generation for intraoperative high dose rate brachytherapy.” J Brachyther Int 13:333–340. Cohen, G. N., H. I. Amols, and M. Zaider. (2000). “An independednt dose-to-point calculation program for the verification of high-dose-rate brachytherapy treatment planning.” Int J Radiat Oncol Biol Phys 48(4):1251–1258. Ezzell, G. A. (1994). “Quality assurance of treatment plans for optimized high dose rate brachytherapy—planar implants.” Med Phys 21:659–661. Harrison, L.B., B. D. Minsky, W. E. Enker, B. Mychalczak, J. Guillem, P. B. Paty, L. L. Anderson, C. White, and A. M. Cohen. (1998). “High dose rate intraoperative radiation therapy (HDR-IORT) as part of the management strategy for locally advanced primary and recurrent rectal cancer.” Int J Radiat Oncol Biol Phys 42(2):325–330. International Commission on Radiation Units and Measurements (ICRU). Report No. 58. Dose and Volume Specification for Reporting Interstitial Therapy. Bethesda, MD: ICRU, 1997. Kubo, H. D., G. P. Glasgow, T. D. Pethel, B. R. Thomadsen, and J. F. Williamson. (1998). “High dose-rate brachytherapy treatment delivery: Report of the AAPM Radiation Therapy Committee Task Group No. 59.” Med Phys 25(4):375–403. Also available as AAPM Report No. 61. Meisberger, L., R. Keller, and R. Shalek. (1986). “The effective attenuation in water of the gamma rays of gold 198, iridium 192, cesium 137, radium 226 and cobalt 60.” Radiology 90:953–957. Meredith, W. J., (ed). Radium Dosage. The Manchester System. Edinburgh: Livingston, 1967. Miller, A. V., M. G. Davis, and J. L. Horton. (1996). “A method for verifying treatment times for simple high-doserate endobronchial brachytherapy procedures.” Med Phys 23:1903–1908. Nag, S., P. Lukas, D. S. Thomas, and L. B. Harrison. “Intraoperative High Dose Rate Remote Brachytherapy” in High Dose Rate Brachytherapy: A Textbook. S. Nag. Armonk NY: Futura Publishing Company, 1994. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24(10):1557–1598. Also available as AAPM Report No. 59.
520
Gil’ad N. Cohen and Marco Zaider
Rogus, R. D., M. J. Smith, and H. D. Kubo. (1998). “An equation to QA the total treatment time for single-catheter HDR brachytherapy.” Int J Radiat Oncol Biol Phys 40:245–248. Saw, C. B., L. J. Korb, B. Darnell, K. V. Krishna, and D. Ulewicz. (1998). “Independent technique of verifying highdose rate (HDR) brachytherapy treatment plans.” Int J Radiat Oncol Biol Phys 40:747–750. Venselaar, J. L., H. W. Bierhuizen, and R. Klop. (1995). “A method to check treatment time calculations in Ir-192 high-dose-rate volume implants.” Med Phys 22:1499–1500.
Chapter 28
Introduction to Prostate Brachytherapy Wayne M. Butler, Ph.D. and Gregory S. Merrick, M.D. Schiffler Cancer Center Wheeling Hospital Wheeling, West Virginia Prevalence and Mortality of Prostate Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522 Frequency of Occurrence and Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522 Trends Over Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522 Effect of PSA Screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 523 Combined PSA and Digital Rectal Exam (DRE) Screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524 Clinical Characteristics of Prostate Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524 PSA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 524 Gleason Score . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525 Clinical Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525 Imaging Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 525 Other Prognostic Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527 Patient Selection and Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527 Partin Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527 Risk Group Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 527 Patient Quality of Life Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528 Preimplant Urinary Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528 Erectile Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528 Rectal Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528 Other Criteria Sometimes Used to Select or Reject Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . 528 Defining Clinical Success . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532 Randomized Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532 Survival . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 532 Biochemical Failure as a Survival Marker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533 When To Declare a PSA Failure? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 533 Therapy Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535 Radical Prostatectomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535 3-D Conformal External-Beam Radiation Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 535 Interstitial Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536 Combined Modality Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536 Hormonal Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 536 Biochemical, Recurrence-Free Survival . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537 Kaplan-Meier Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537 Comparative Single Institution Survival Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537 Inter-institution and Inter-modality Comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 Actuarial Survival for Low-Risk Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538 Actuarial Survival for Intermediate-Risk Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540 Actuarial Survival for High-Risk Patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540 Treatment-related Morbidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543 Urinary Morbidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544 The Role of Alpha-blockers on IPSS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544 Urethral Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546 Rectal Dosimetry and Morbidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546 Erectile Dysfunction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548 Mechanism of Brachytherapy-Induced Erectile Dysfunction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548 Radiation Dose to the Prostate Gland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 548
521
522
Wayne M. Butler and Gregory S. Merrick Dose to the Neurovascular Bundles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549 Dose to Penile Erectile Bodies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 549 Potency Preservation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551
Prevalence and Mortality of Prostate Cancer Frequency of Occurrence and Mortality Prostate cancer is one of very few cancers for which there is still a debate not just on how to treat but whether there should be any treatment at all. Today, in a well-screened population, few men diagnosed with prostate cancer should die from that disease. Current screening practice has increased the fraction of prostate cancer which is clinically localized at diagnosis to 90%, which is much higher than rates of locally confined disease found for other cancers with effective population screens—breast, cervix, and colon/rectum at 63%, 55%, and 39%, respectively (Carroll 2005). Nevertheless, the American Cancer Society, using models from the National Cancer Institute, estimates that in 2005 over 30,000 men in the United States will die from prostate cancer and over 230,000 new cases will be diagnosed (Jemal et al., 2005). Among men, prostate cancer accounts for one-third of all new cancers diagnosed—far exceeding lung and bronchus at 13%—but the death rate from prostate cancer, at 10% of all male cancer deaths, is second to lung and bronchus at 31% of all male cancer deaths. In Sweden, active cancer screening is not advocated, but citizens have essentially free access to necessary health care resources if they have symptoms or if they request a test such as prostate-specific antigen (PSA) or a digital rectal examination (DRE). In a prospective analysis of 8887 Swedish men diagnosed with prostate cancer, the disease specific survival rate at 15 years was 44% (Aus et al., 2005). In this largely unscreened population, the median age at diagnosis was 75 years and the median age at death was 80 years. These data represent a worst-case scenario for well-developed countries with high life expectancy. In the Swedish study, treatment with curative intent was given to only 11% of the population, and 37% did not even receive palliative treatment. In that portion of the population with localized cancer for which four fifths were treated curatively, the 10-year and 15-year disease specific survival rates were 93% and 80%, respectively.
Trends Over Time The median age at diagnosis in the United States is about 66 years, but has been decreasing as PSA screening is applied to ever-younger age cohorts. When PSA screening was introduced in the late 1980s, incidence rates increased dramatically, but the rates have subsequently decreased or leveled off for age groups over 60 years (American Cancer Society 2005) and overall the rate is only slightly greater than before the PSA era. Prostate cancer remains extremely rare in men under the age of 40 years. Because of the relatively advanced age at diagnosis and the typically slow progression of disease, prostate cancer patients have an average years of life lost of 6.1 years—the lowest years of life lost of 17 common cancers (Burnet et al., 2005). Brain cancer patients suffer the greatest loss at over 20 years. Cancer statistics show a significant trend for improved 5-year relative prostate cancer survival rates, which have increased from 67% in 1974–1976, to 75% in 1983–1985, to 99% in 1995–2000. Figure 1 plots prostate cancer death rates per 100,000 men. Cancer death rates rose in the late 1980s and peaked in 1992 when the number of deaths exceeded 40,000, and the rate has been declining ever since. Using analytical methods responsive to rapidly changing patterns of care, a continual improvement in survival
28–Introduction to Prostate Brachytherapy
523
Figure 1. Prostate cancer mortality rate per 100,000 male population. The results are age-adjusted to the distribution of ages in the 2000 census. (Adapted from American Cancer Society, 2005.)
has been documented (Brenner and Arndt 2005). For men with well- or moderately differentiated tumors and a localized or regional stage, their calculated 5-year and 10-year relative survival rates exceeded 100% indicating no excess mortality in that cohort of men, which compromised two thirds of the Surveillance, Epidemiology, and End Results (SEER) study population.
Effect of PSA Screening Although most of us who treat prostate cancer would be happy to accept credit for such increasingly favorable survival results, we should be aware of competing explanations. Screening contributes to diagnosis of disease at earlier stages (Harris and Lohr 2002). Certainly, some men are diagnosed with disease that may never have become clinically significant, but this effect is likely to be small because current incidence rates mimic that of the prescreening era. Earlier diagnosis also means that patients will inevitably live with their disease longer. PSA screening may also selectively enhance the detection of prognostically favorable tumors. Indeed, there has been a reduction in the incidence at diagnosis of metastatic disease, positive lymph nodes, and seminal vesicle involvement, but there has been no significant change in microscopic measures of aggressiveness. The value of PSA screening has been called into question by one of the founders of the test (Stamey et al., 2004) because of a purported loss over time of its positive predictive value. Prostate cancer readily meets two of the criteria for a disease amenable to mass population screening, namely, a high prevalence within the population and a high mortality rate if left untreated (Wilson and Crawford 2004). Therefore, the utility of a prostate cancer-screening test depends on its cost effectiveness. Cost effectiveness requires that the test have reasonably sensitivity (high true positive detection rate) and reasonably specificity (high true negative detection rate). The cost of the test itself should be relatively low, and the societal costs associated with interventions implemented after a positive test (such as morbidity due to biopsy) should be proportionate to years of life saved. For a PSA threshold of 4.0 ng/mL, the sensitivity of PSA is 75% and the specificity is 60% (Wilson and Crawford 2004). The sensitivity of any quantitative diagnostic test may be improved by lowering the threshold, but the specificity will always diminish, meaning that more men would be subjected to unnec-
524
Wayne M. Butler and Gregory S. Merrick
essary biopsies. As part of a cancer prevention trial, biopsies were performed on 2950 men who had completed 7 years in the placebo arm and whose PSA had never exceeded 4.0 ng/mL (Thompson et al., 2003). The biopsies were pathologically positive for prostate cancer in 15.2% of the men, and 15% of the cancers detected had a Gleason score of 7 or greater, which is almost certainly an indicator of clinically significant disease. Lowering the PSA threshold that prompts a prostate biopsy to 2.5 ng/mL will detect cancers with more favorable characteristics. In an analysis of PSA and cancer detection rate in over 36,316 biopsies, the detection rate in men with PSAs in the range 2.5 to 4.0 ng/mL was 27.5% while the rate for men in the 4.0 to 10.0 ng/mL range was similar at 30.1% (Gilbert et al., 2005). The effect of changing the threshold from 4.0 to 2.5 ng/mL improved the sensitivity from 87% to 94% while the specificity declined from 20% to 12%. One way to improve the sensitivity without greatly affecting the specificity of the PSA test has been to create age-specific thresholds. Other effective enhancements include determining the PSA velocity in ng/mL/yr, measuring PSA density in ng/mL/cm3 of prostate, measuring free PSA, complex PSA, and their ratios to total PSA, and measuring precursor forms of PSA (Han, Gann, and Catalona 2004). The frequency of screening may also be adjusted for patient risk factors, most notably, the PSA velocity or PSA doubling time. In a large screening population study, men who remained free of cancer over a 10-year period had an average PSA velocity of 0.03 ng/mL/yr (Berger et al., 2005). Also of significance was each total PSA in the previous annual exam. Men with an antecedent PSA of <1 ng/mL had only a 0.6% chance of being diagnosed with prostate cancer 2 years later. Extending the screening interval beyond annually in men with initial serum PSA less than 2 ng/mL would result in delays in prostate cancer detection and possibly affect treatment outcomes. Carter and Pearson (1999) found that there was a negligible risk of a PSA level of 2.0 ng/mL converting to a value greater than 4.0 ng/mL after 2 years, and the risk remained less than 1% for conversion to a value greater than 5.0 ng/mL after 4 years. A hypothetical 2year screening interval was found to delay detection by 4 to 20 months while a 4-year screening interval would have delayed detection by 4 to 44 months (Kundu et al., 2005).
Combined PSA and Digital Rectal Exam (DRE) Screening In a large cancer screening trial in which 76,705 men between 55 to 74 years of age were randomized to either the screening or control arm, screening was performed by PSA and DRE (Andriole et al., 2005). Patients in the screening arm were informed of a positive DRE or a PSA >4 ng/mL and referred to their primary care physician for discretionary follow-up. Men testing positive for either DRE or PSA were 14% of the total, but there was little overlap between the tests with 7.5% positive for DRE and 7.9% above the PSA threshold. Ultimately, 32% of the men with suspicious screening results underwent prostatic biopsy within 1 year of the screen, and 37% of those had biopsy-proven cancer. An abnormal PSA was nearly twice as likely as a positive DRE to result in biopsy, and the yield of positive biopsies was less in DRE positive men (34%) than in PSA suspicious men (44%). Nevertheless, a DRE is considered necessary to detect prostate cancers, which produce little or no PSA.
Clinical Characteristics of Prostate Cancer PSA PSA testing unequivocally revolutionized the diagnosis and treatment of prostate cancer. Prior to 1988 when that test became widely available, diagnosis was usually based on an abnormal digital rectal exam or an incidental pathological finding after some other procedure. Urologists and radiation oncologists had been smugly confident in the efficacy of either surgery or radiation therapy because most prostate cancers
28–Introduction to Prostate Brachytherapy
525
are inherently indolent and any primary therapy of localized disease will delay the time to progression to frank symptoms such as painful bone metastases. Among the 18% of men in one large series who developed bone metastases, the median time from surgery to that end point was 9 years (Pound et al., 1999). PSA demolished false perceptions of treatment efficacy because the test may also be used as a marker of success or a harbinger of symptomatic failure. A rising or non-nadiring PSA is now referred to as a biochemical failure, and such failures were found to be rife in older surgical and radiation therapy series. The dismal situation prompted the famous conundrum posed by the urologist Willet Whitmore (1990), “Is cure possible in those for whom it is necessary, and is cure necessary in those for whom it is possible?” Because disease detected by PSA screening and pathologically confirmed is very rarely classified as insignificant, the answer to the latter half of the question is yes. The answer to the former part is also yes, but the disease must be loco/regionally confined and the choice and quality of the treatment must be appropriate for success to occur.
Gleason Score Gleason score is one of the most important predictors of biologic behavior, extracapsular extension, and outcome after treatment, and often plays an important role in determining patient management. From microscopic examination of stained prostate tissue, a pathologist assigns a Gleason grade to a specimen based on the glandular architecture from well differentiated (grade 1) to poorly differentiated (grade 5). The grades of the two most prevalent architectural patterns (primary and secondary) are added together to produce the composite Gleason score, e.g., grade 3 + grade 4 = GS 7. Although Gleason scores on prostate needle biopsies often differ by ±1 from those assigned after radical prostatectomy, the differences are most common in lower grades (Bostwick 1994) for which there is an emerging consensus for treatment approach. Of greater concern is the spotty concordance between Gleason scores assigned by community based pathologists and those from prostate cancer specialists (Merrick et al., 1998). Community-based pathologists tend to undergrade Gleason scores relative to their more experienced counterparts. The distribution of Gleason scores within an institution should be similar to those reported in the literature: <1%, GS 2–4; 15%, GS 5; 65%, GS 6–7; and 19%, GS 8–10. An institution which has a large fraction (>10%) of patients diagnosed with Gleason score <5 cancers should consider sending slides from some of these patients for an outside review of pathology. In a recent survey of Gleason scoring practices, only 13% of pathologists ever reported a Gleason score ≤4, and of those assigning such low scores, 88% reported that they comprised <1% of cancers (Egevad, Allsbrook, and Epstein 2005). Diagnoses of prostatic intraepithelial neoplasia (PIN) or atypical small acinar proliferation (ASAP), which are considered precancerous, should precipitate further biopsies.
Clinical Stage The AJCC staging system for clinically detected prostate cancer is detailed in Table 1 (American Joint Committee on Cancer 2002). The correlation between clinical (digital rectal exam based) stage and pathology stage after radical prostatectomy is not strong (Grossfeld et al., 2001). As with undergrading of Gleason scores, there is also a tendency to understage prostate cancer due to reluctance by urologists to classify nonpalpable lesions that are visible on transrectal ultrasound (TRUS) as T2 lesions (Augustin et al., 2003). Imaging Modalities Other imaging modalities are sometimes invoked to supplement the digital rectal exam and TRUS in clinical staging of asymptomatic patients. A positive nuclear medicine bone scan, for example, indicates metastatic spread for which there is no current curative therapy. However, positive bone scans are
526
Wayne M. Butler and Gregory S. Merrick Table 1. Prostate Cancer Clinical Staging by the 2002 American Joint Committee on Cancer
Stage
Description
cT1
Microscopic disease neither palpable nor visible on TRUS
cT2
Palpable tumor apparently confined within the prostate
cT3
Tumor protrudes through the prostate capsule Tumor is fixed or invades beyond SV
cT4
Substage
Description
cT1a cT1b cT1c cT2a cT2b cT2c cT3a cT3b cT4
Incidental finding in ≤ 5% of tissue sample Incidental finding in > 5% of tissue sample Found on needle biopsy due to ↑ PSA Involves ≤ half of one lobe of the prostate Involves > half of one lobe of the prostate Involves both lobes of the prostate Extracapsular extension of one or both lobes Seminal vesicle invasion Invades bladder neck, muscle, pelvic wall or other
becoming increasingly rare in modern screened populations. In one study by Lee et al. (2000), the rate of positive bone scans was only 1% in men with a PSA <50 ng/mL, GS <8, and stage
28–Introduction to Prostate Brachytherapy
527
Other Prognostic Factors The tumor volume burden within the prostate measured as either the number of involved cores— such as 6 of 12—or the percentage of positive biopsies—taken as the length of involved cores relative to the total length of cores—has been shown to significantly affect the outcome of radical surgery and external beam therapy (Freedland et al., 2003; D’Amico et al., 2004). Perineural invasion found in the biopsy specimens has also been related to prognosis in surgery and external beam therapy (Beard et al., 2004; D’Amico et al., 2001). Findings of extensive tumor volume and/or perineural invasion appear to be indicators of extracapsular extension, and these are of little consequence in determining the outcome of permanent seed brachytherapy (Merrick et al., 2001, 2004).
Patient Selection and Classification At the time a patient must make an informed decision regarding his course of treatment, the only prognostic information normally available is the clinical stage, laboratory data regarding PSA and its variants, and pathology statements regarding the biopsy specimens. Any additional testing would be undertaken in response to a notably adverse prognosticator, and the result used to rule in or rule out any role for curative therapy. Almost all men who are regularly screened present with potentially curable disease. These men need information regarding their likelihood of cure and complications under a multitude of treatment scenarios.
Partin Tables In terms of the likelihood of cure, one of the most important predictors is whether or not the disease is organ confined, and if not, what is the likelihood of extraprostatic extension, seminal vesicle involvement, and lymph node involvement. Although there are competing tables, nomograms (Kattan 2003), and regression equations (Roach et al., 2000) generated by other groups, the tables published by Partin and colleagues at Johns Hopkins are the most commonly used to identify specific patient risk (Partin et al., 2001). For example, it takes but a moment to determine that a patient who presents with clinical stage T1c, a PSA of 7.3, and a Gleason score of 6 has a median likelihood of organ confined disease of 75%, extraprostatic extension of 23%, seminal vesicle involvement of 2%, and lymph node involvement of 0% (within a 95% confidence interval).
Risk Group Classification Personalized risk factors are ideal for individual guidance and decision making, but the 100 subgroups of the Partin tables subdivides populations treated at most institutions too finely for meaningful statistical analysis or inter-institutional comparisons. Some institutions report their results in terms of several PSA ranges or a few Gleason score groupings, but the logical approach to creating a few robust predictive groups is to use all three components of the Partin tables. Unfortunately, there is no consensus on how to best construct multifactorial risk groups. One commonly used three-risk group classification scheme is described in Table 2. The last column of the table lists the range of organ-confined disease found from the Partin tables that is allowed by the risk group definition. The advantage of this scheme is its simplicity and statistically robust stratification of biochemical survival curves. Local modifications sometimes redefine intermediate risk patients to have either stage T2b, or PSA between 10 and 20 ng/mL, or Gleason score 7. High risk patients then are defined as stage T2c, or PSA > 20 ng/mL, or Gleason score 8.
528
Wayne M. Butler and Gregory S. Merrick Table 2. A Commonly Used Risk Group Stratification Scheme Risk Group
Clinical Stage
PSA
Gleason Score
% Organ Confined †
% Lymph Node Positive †
Low
≤ T2a
and
≤ 10
and
≤6
95% – 58%
0% – 1%
Intermediate §
≥ T2b
or
> 10
or
≥7
88% – 21%
0% – 5%
High *
≥ T2b
and/or
> 10
and/or
≥7
57% – 6%
4% – 38%
†
Ranges of organ confined disease and lymph node positivity within risk group from Partin et al. (2001). Intermediate risk allows only one unfavorable factor. * High risk requires 2 or 3 unfavorable factors. §
Patient Quality of Life Factors Preimplant Urinary Function A baseline, pretreatment International Prostate Symptom Score (IPSS) should be obtained for all patients for use as either a selection or advisory tool and to track changes post-implantation. Questions on the IPSS are summarized in Table 3. Although patients with elevated IPS scores are considered to be at higher risk of post-implantation urinary retention, the relationship between IPSS and urinary obstruction is not clear. Terk, Stock, and Stone (1998) reported that pre-implantation IPSS predicted post-implantation urinary retention, while Merrick et al. (2002) found that with the use of prophylactic and prolonged alpha-blockers, there was no correlation of urinary retention with pre-implantation IPSS. Merrick and colleagues have also found that post-void residual urine volume—another marker for potential obstructive symptomatology—does not correlate with any post-implant indicators of urinary morbidity. Erectile Function About 50% to 70% of prostate cancer patients are potent at the time of diagnosis (Laumann, Paik, and Rosen 1999). Because maintenance of sexual function is an overriding concern of many patients, a baseline International Index of Erectile Function (IIEF) score should be obtained for all patients. We use a subset of the full questionnaire, the IIEF-6, supplemented with three of our own questions specific for brachytherapy induced erectile dysfunction as shown in Table 4. Rectal Function Although significant rectal morbidity is quite rare in prostate brachytherapy, quality of life for affected patients is greatly diminished. One indication of the severity of distress is that the overwhelming majority of medical malpractice cases brought against brachytherapists pertain to rectal injury (Elliott et al., 2004). We developed the Rectal Function Assessment Score (RFAS) shown in Table 5 to establish a baseline for bowel function and to track changes in function that may be attributable to the brachytherapy implant. Other Criteria Sometimes Used to Select or Reject Patients There are a number of additional presenting patient features that have been used at various times by various institutions as selection criteria. It is wasteful to expend healthcare resources on individuals with prostate cancer if their life expectancy from concomitant morbidities is much less than that from their untreated prostate cancer. Patient age, therefore, is a consideration, albeit a highly subjective one. And who would refuse treatment to a patient who desires treatment—so long as his risk of morbidity from the treatment was not great? At the other end of the spectrum, younger men have been dissuaded from
28–Introduction to Prostate Brachytherapy
529
Table 3. International Prostate Symptom Score (IPSS) Summary *
Question: Over the past month, how often have you had …
Incomplete emptying
… a sensation of not emptying your bladder completely after you finished urinating?
Frequency
… to urinate again less than 2 hours after you finished urinating?
Intermittency
… found you stopped and started again several times when you urinated?
Urgency
… found it difficult to postpone urination?
Weak stream
… a weak urinary stream?
Straining
… to push or strain to begin urination?
Nocturia
… to typically get up to urinate from the time you went to bed at night until the time you got up in the morning?
* Each item is scored from 0 (never) – 5 (almost always or 5 times/night nocturia) for a total IPSS range of 0 – 35.
brachytherapy due to mistaken notions of efficacy and a misguided concern for late radiation-induced malignancies. Younger men have a lower risk of biochemical failure than older men because they tend to have lower Gleason scores and a lower likelihood of extracapsular extension and distant metastases (Carter, Epstein, and Partin 1999). These authors found age to be an independent risk factor and at least as good a predictor of outcome as pretreatment PSA. In men less than 63 years old, the biochemical freedom from failure 5 years after brachytherapy exceeded 97% for all risk groups (Merrick et al., 2001a). Anatomic considerations such as prostate size, prostate volume, pubic arch interference, and obesity have been putative contraindications to prostate brachytherapy. Prostate dimensions and volume are justifiably of concern because of the technical obstacles presented by a gland that is much wider than the standard needle template grid (6 cm) or a gland of such great volume that the cost of the seeds needed to cover the volume is prohibitive. Many such patients may become treatable by conventional seed brachytherapy after 3 to 4 months of neoadjuvant hormonal therapy, which typically reduces the prostate volume by 35% to 40%. Preimplant hormonal cytoreduction is also frequently implemented in patients with elevated IPSS. Large prostate size was found to correlate with improved prostate cancer survival (Lehrer, Stone, and Stock 2005), and large prostate size was found to have no correlation with morbidity following brachytherapy (Merrick et al., 2000). Although pubic arch interference often presents a technical problem in implanting prostate glands that are greater than 4 cm tall in the anterior-posterior direction and more than a couple of centimeters wide at the anterior aspect, small glands may also be difficult to approach if the patient’s pubic arch is narrow. Maneuvering the patient into an extended lithotomy position makes the prostate accessible in almost all cases of interference, and in the remaining cases, either needle steering or an increased angle of needle approach provides adequate seed coverage of the anterior gland. Obese patients present technical difficulties for treatment by radical prostatectomy and external-beam radiation therapy, but they present only a modest challenge to brachytherapy. There was no significant difference in dosimetric quality, biochemical disease free survival, and quality of life measures between men with grades II and III obesity and non-obese patients (Merrick et al., 2002a, 2005). Reactive clinical diagnoses such as prostatitis and inflammatory bowel disease have also been alleged as contraindications to prostate brachytherapy despite a complete dearth of published support. Hughes et al. (2001) discerned no relation between prostatitis and postimplant urinary morbidity. In a similar study, Grann and Wallner (1998) found that inflammatory bowel disease such as ulcerative colitis and Crohn’s disease led to no unusual gastrointestinal or rectal side effects following brachytherapy implants that were not explicitly rectal sparing.
530
Wayne M. Butler and Gregory S. Merrick Table 4. Schiffler Cancer Center Modified International Index of Erectile Function (IIEF 6) + 3 Answer these questions for your situation over the last month and without the assistance of Viagra or mechanical aids.
1. How often are you able to get an erection during sexual activity? ❑ ❑ ❑ ❑ ❑ ❑
0 0 0 0 0 0
= = = = = =
No sexual activity Almost never/never A few times (much less than half the time) Sometimes (about half the time) Most times (much more than half the time) Almost always/always
2. When you had erections with sexual stimulation, how often were your erections hard enough for penetration? ❑ ❑ ❑ ❑ ❑ ❑
0 0 0 0 0 0
= = = = = =
No sexual activity Almost never/never A few times (much less than half the time) Sometimes (about half the time) Most times (much more than half the time) Almost always/always
3. When you attempted sexual intercourse, how often were you able to penetrate (enter) your partner? ❑ ❑ ❑ ❑ ❑ ❑
0 0 0 0 0 0
= = = = = =
Did not attempt intercourse Almost never/never A few times (much less than half the time) Sometimes (about half the time) Most times (much more than half the time) Almost always/always
4. During sexual intercourse, how often were you able to maintain your erection after you had penetrated (entered) your partner? ❑ ❑ ❑ ❑ ❑ ❑
0 0 0 0 0 0
= = = = = =
Did not attempt intercourse Almost never/never A few times (much less than half the time) Sometimes (about half the time) Most times (much more than half the time) Almost always/always q
5. During sexual intercourse, how difficult was it to maintain your erection to completion of intercourse? ❑ ❑ ❑ ❑ ❑ ❑
0 0 0 0 0 0
= = = = = =
6. How do you rate your confidence that you could get and keep an erection? ❑ ❑ ❑ ❑ ❑
0 0 0 0 0
= = = = =
Date
Very low Low Moderate High Very high
7. How often have you experienced pain at the time of orgasm? ❑ ❑ ❑ ❑ ❑ ❑
0 0 0 0 0 0
= = = = = =
No sexual activity Almost never/never A few times (much less than half the time) Sometimes (about half the time) Most times (much more than half the time) Almost always/always
8. How often have you noted blood in the ejaculate? ❑ ❑ ❑ ❑ ❑ ❑
0 0 0 0 0 0
= = = = = =
No sexual activity Almost never/never A few times (much less than half the time) Sometimes (about half the time) Most times (much more than half the time) Almost always/always
9. Have you noted a change in the intensity of your orgasm? ❑ Yes ❑ No
Patient Name
Did not attempt intercourse Extremely difficult Very difficult Difficult Slightly difficult Not difficult
28–Introduction to Prostate Brachytherapy
531
Table 5. Schiffler Cancer Center Rectal Function Assessment Score (RFAS) Respond regarding your experience during the past week 1. Frequency of stools per day:
7. Continence:
❑ 0 – 1 stool per day
❑ Normal continence; able to control stool movements at all times ❑ Gas incontinence only; able to control stool movements but not gas ❑ Minor spotting or leakage of stool (up to coin size) about once per week ❑ Minor spotting or leakage of stool (up to coin size) more than once per week ❑ Significant leakage of stool (larger than coin size) about once per week ❑ Significant leakage of stool (larger than coin size) more than once per week
❑ 2 stools per day ❑ 3 stools per day ❑ 4 or more stools per day 2. Consistency of stools: ❑ All stools formed ❑ Stools formed and loose ❑ Stools loose ❑ Watery stools 3. Urgency of stools: ❑ Somewhat urgent
8. Nighttime bowel movements (total number of nights in last week that you had to get up from bed to have a bowel movement):
❑ Urgent
❑ 0
❑ Very urgent
❑ 1
❑ No urgency
❑ 2 4. Abdominal discomfort:
❑ 3
❑ No discomfort
❑ 4
❑ Mild to moderate discomfort
❑ More than 4
❑ Somewhat severe discomfort ❑ Very severe discomfort 5. Hemorrhoidal discomfort: ❑ No discomfort ❑ Requires mild treatment (i.e. tucks, sitz baths) ❑ Requires topical medication (i.e. Prep H, etc.) ❑ Requires oral analgesics or narcotics for relief 6. Rectal bleeding:
9. Completeness of evacuation: ❑ Complete evacuation (requires one movement to completely empty bowel or feel you’re “all done”) ❑ Occasional multiple evacuations (about once a week feel like you’re not “all done” or it takes more than one movement to finish) ❑ Frequent multiple evacuations (more than once a week feel like you’re not “all done” or it takes more than one movement to finish) ❑ Requires enema to obtain complete emptying
❑ No rectal bleeding ❑ Blood on toilet paper: 1 time per week ❑ 2-3 times per week ❑ ≥ 4 times per week
10. My bowel movements after the implant are: ❑ Better ❑ Worse ❑ Same
Responses to questions 7 and 8 are scaled to 0 – 4 points; question 10 is not counted in the total score.
532
Wayne M. Butler and Gregory S. Merrick
There are substantive reports in the early prostate brachytherapy literature of greatly increased incontinence in patients who had undergone a transurethral resection of the prostate (TURP) prior to implantation. However, those patients were not treated with urethral sparing techniques. By limiting the dose to the epithelium of the TURP defect to 100% to 115% of the prescribed dose, the incidence of incontinence is not significantly different from non-TURP patients. In 100 TURP patients administered the UCLA Prostate Cancer Index, the mean urinary function and bother scores both exceeded 82 (quite favorable), which were comparable to untreated controls (Moran, Stutz, and Gurel 2004). Today, a preimplant TURP is readily accommodated in treatment planning and is of much less consequence than a post implant TURP, which will be discussed later.
Defining Clinical Success Randomized Trials The most convincing evidence to support the efficacy of a treatment is a randomized trial. There has been only one such contemporary trial in prostate cancer, and it compared the outcomes of 695 Scandinavian men randomized to either radical prostatectomy or watchful waiting (Holmberg et al., 2002). With a median follow-up of 6 years, there was a significant difference between the two cohorts in favor of intervention both in terms of prostate cancer specific mortality and the occurrence of distant metastases. Our rules of scientific evidence require that any conclusion be collaborated or replicated before it is included in the canon of medicine. However, it is unlikely that another trial comparing treatment with non-treatment will ever be completed. The trend in European medical literature is to justify doing less, while the trend in American medical literature is to justify doing more. Even though many American institutions would like to verify the Scandinavian study, the impetus among physicians and patients to take positive action upon the diagnosis of prostate cancer is overwhelming. Without the gold standard of randomized clinical trials, we are left with less convincing retrospective studies. In light of those results, it would be unethical to conduct a trial of any therapy against non-treatment or a placebo. There have been sporadic attempts to organize clinical trials comparing therapy modalities, but these languish from poor enrollment—patients develop a strong preference for one arm and refuse randomization—and the perception by physicians that the quality of practice in each arm may be unequal. The large, well-funded Surgical Prostatectomy vs. Interstitial Radiation Intervention Trial (SPIRIT, American College of Surgeons Oncology Group trial Z0070) offered hospitals $5,000 reimbursement per patient recruited, yet closed for lack of accrual. Even the most successful center found that only 1 in 30 patients would agree to randomization. Without the gold standard of randomized clinical trials, we are left with less convincing inter-institutional inter-modality comparisons.
Survival Death by prostate cancer is the ultimate failure marker. However, because of the indolent nature of most prostate cancer, large numbers of patients must be followed for a long period of time to make any statistically valid conclusions. When such results are published, they rarely are relevant to current practice because of the pace of change in medicine over the ensuing years. At best, they bolster trends that are evident in the current paradigm but give little guidance as to what the next step should be. Prior to the PSA era, Radiation Therapy Oncology Group (RTOG) trials did not demonstrate a prostate dose response above 70 Gy. We now know that treating prostate cancer is better than not treating it, and that patients treated to higher radiation doses by the standards of 20 years ago have fared better in prostate cancerspecific survival. In retrospect those 20-year survival rates are depressing. Should we have known that
28–Introduction to Prostate Brachytherapy
533
dose escalation was important and possible? Do those results tell us how to improve today’s practice or bring us any closer to knowing when enough is enough in dosimetry? Biochemical Failure as a Survival Marker The development of distant metastases is an unequivocal marker, but metastases develop too slowly in all but high-risk patients to inform changes in practice. Local progression as measured by physical examination or biopsies would be a dandy marker, but the tests are simply not sensitive enough. By default, therefore, PSA has become the harbinger and marker of failure. Although PSA is accepted as a legitimate marker of success or failure, its relationship to disease-specific survival has not yet been definitively answered. For a radical prostatectomy series of 1132 patients at the Cleveland Clinic, the overall 10-year survival for patients with a PSA recurrence was equivalent to those without a detectable PSA (Jhaveri et al., 1999). A subsequent report from the same institution on 936 external beam radiotherapy patients came to the same conclusion that biochemical failure does not lead to increased mortality (Kupelian et al., 2002). The mean follow-up was 56 months in the former paper and the median follow-up was 58 months in the latter. Among the external-beam patients, there was a trend for worse outcome in the biochemical failure cohort, and the overall survival curves were increasingly divergent after 8 years. The patients in the external-beam study were considerably older and had worse clinical stage, PSA, and biopsy Gleason score than those in the radical prostatectomy study. The authors suggest that longer follow-up is likely to show biochemical failure becoming significant in predicting overall survival. Conclusive evidence linking biochemical survival and overall survival will most likely occur in high-risk populations. In a study of 942 patients treated at the Fox Chase Cancer Center, biochemical failure was the strongest determinate of distant metastasis and also strongly correlated with cause-specific death (Pollack et al., 2003). Although biochemical failure may be of little real consequence to some patients, to those who progress to metastases the consequences are grave indeed. Pretreatment predictors for failure and death from prostate cancer have been exhaustively reported over the past decade, and of course they are essentially our classification of patients into risk groups. Favorable statistical odds are of cold comfort to the low- or intermediate-risk patient who progresses to metastases. However, the pattern of PSA failure does have considerable meaning for the individual, and these patterns have been well studied. In a multi-institutional retrospective analysis of 8669 patients treated with either surgery or external beam radiation, a PSA doubling time of less than 3 months was strongly predictive of prostate cancer specific mortality, and this cause specific mortality was virtually the same as all cause mortality (D’Amico et al., 2004). Even in these patients with rapid PSA doubling times, the median time to death from the date of biochemical failure was greater than 6 years. Among patients with PSA doubling times greater than 3 months, the actual doubling time was a significant predictor of cause specific death, but no median could be determined because the cause specific survival in this group exceeded 90% at 10 years post failure. In addition to providing a surrogate end point for mortality, rapid PSA doubling time indicates those patients most likely to benefit from early, aggressive therapy—although we do not appear to have as yet any effective salvage therapy. When To Declare a PSA Failure? Because radical prostatectomy should remove all prostate tissue, any measurable PSA after surgery indicates at best an incomplete resection and at worst residual disease. What constitutes a measurable PSA has evolved over the years as the sensitivity of the immuno-histochemical analysis has improved to the current decision threshold of 0.01 ng/mL. Even though the detection limit of sensitive PSA assays is 0.002 ng/mL, surgeons contend that the use of various failure declaration thresholds between 0.2 ng/mL and 0.5 ng/mL is justified by the residual presence of microscopic viable prostatic tissue. Radiation therapy, which was not considered ablative of the prostate gland, has used a different definition of biochemical failure. The American Society for Therapeutic Radiology and Oncology Consensus
534
Wayne M. Butler and Gregory S. Merrick
Panel (ASTRO 1997) recommends that biochemical failure be defined as three consecutive increases in PSA with readings taken at 3 to 4 month intervals during the first 2 years and every 6 months thereafter. The date of failure is to be taken as the midpoint between the post-irradiation nadir PSA and the first of the three consecutive rises. Because of the lead-time and backdating, patients must be followed with sequential PSA readings a minimum of 2 years for inclusion in survival analysis. As simple and straightforward as the ASTRO consensus definition of biochemical failure may appear, the recommended interval between PSA tests is often interpreted as an ideal minimum, which is subject to the patient’s availability and the institution’s preference. Because the backdating convention does not kick in until the third rise, very long intervals leave patients censored (not counted as failure at their most recent PSA) at one institution who would have been backdated as failures at another institution rigorously following the timing recommendations. Identification of the post-irradiation PSA nadir creates additional ambiguities that may significantly affect biochemical survival rates. This may be construed as the date of the lowest PSA measurement or the date of the last non-rising PSA. The consensus panel noted the phenomenon of rising then falling PSAs, “bouncing PSAs” in their terminology, but did not address the issue of multiple plateaus on an individual’s PSA versus time graph and whether several stable values should reset the rise count. Analyzing 1050 external beam patients, Williams (2004) found a relative 12% increase in 5-year survival when nadir was defined as the last non-rising PSA compared to defining nadir as the lowest PSA, but the 10-year progression-free survival was 13% less when using the last non-rising PSA definition. When PSA tests collected within ≤3 months of the previous test were excluded, thereby forcing a longer test interval, the 10-year survival changed from 32.1% to 47.7%. Similarly large changes in calculated survival rates were induced by detection bias when the PSA lower detection limit was increased above 0.1 ng/mL or the resolution of the test was decreased. Many alternate PSA testing regimens that would improve the ASTRO definition’s sensitivity and specificity for identifying clinical failures have been suggested, such as those by Horwitz et al., (2005). Few of the suggested fixes or replacements to the ASTRO definition of biochemical failure fully address unequivocally how to handle variable testing intervals, failure dating ambiguities, and detection bias. Permanent seed brachytherapy appears to be ablative of the prostate and capable of achieving PSA nadirs over time comparable to those produced by surgical resection. Using magnetic resonance spectroscopic imaging (MRSI), Pickett et al. (2004) found complete prostate metabolic atrophy within 48 months of 125I implantation in a population selected for low post-implant isodose coverage. The mean time to metabolic atrophy in these relatively lower quality implants was 29 months, and the MRSI-detected atrophy preceded the time of PSA nadir by 14 months. In higher quality seed implants, durable PSA nadirs are achieved within the first few tests after treatment, and those nadirs are usually near the detection limit. Measurements near the detection limit are inherently “noisy,” so there are frequent opportunities for rises and potential failure dates. The etiology of such PSA spikes or bounces remains unknown, but they occur in one-fourth to one-third of the implanted population (depending on the definition of bounce), and they have no apparent effect on biochemical survival (Merrick et al., 2003; Stock, Stone, and Cesaretti 2003). We might expect the dose escalation entailed by intensity-modulated radiation therapy (IMRT) to be ablative of the prostate like brachytherapy, but there are no reports of median post treatment PSA in IMRT series <0.5 ng/mL. Support for the suppression of prostate function from even conventional doses of radiation comes from studies of PSA changes in healthy prostates included in the treatment fields for rectal or bladder cancer (Zietman, Zehr, and Shipley 1999). With a median follow-up of 6 years, the median PSA in these patients with incidentally irradiated prostates was ≤0.5 ng/mL. If all or nearly all prostate tissue is made nonviable by radiation therapy, shouldn’t the definition of biochemical failure be made congruent with that used in radical prostatectomy series? Critz and Levinson (2004) applied a 0.2 ng/mL cut point to a 1469 patient population treated with combined brachytherapy and external beam and found the 10-year outcomes for low-, intermediate-, and high-risk groups were
28–Introduction to Prostate Brachytherapy
535
93%, 80%, and 61%, respectively. Using a cutpoint of 0.4 ng/mL, Merrick and colleagues (2004a) found a durable plateau on the survival curve of high-risk patients treated with combined modality therapy and whose Gleason scores were from 8–10 and PSA <20 ng/mL. From 24 to 93 months post implant, both the ASTRO definition and the PSA cutpoint survival curves were identical at 85%.
Therapy Modalities Locally confined prostate cancer may be treated with great success by a number of modalities with the primary difference being the type and severity of side effects and complications. In men with non-organ confined non-metastatic disease, treatments vary in aggressiveness, and there is considerable divergence in survival and morbidity between and within therapy modalities. Confronted with the lack of consensus on treatment approaches, patients must make their decision in the face of subtle or overt physician bias in favor of their specialty and the patient’s fear of cancer and inchoate concerns about treatment related side effects. There are four standard therapies that have patient follow-up extending back to the beginning of the PSA era: radical prostatectomy, 3-D conformal external-beam therapy, permanent seed interstitial brachytherapy, and watchful waiting. In addition, hormonal therapy is often used as an adjunct with the four standard therapies. External beam therapy may be adjunctive to radical prostatectomy—particularly in the case of positive surgical margins or other indicators of locally advanced disease—and to brachytherapy, both permanent seed and high dose-rate (HDR) brachytherapy.
Radical Prostatectomy Surgery excises the entire prostate with a minimal margin of extraprostatic tissue, typically <2 mm. The prostatic urethra is removed with the prostate, and the membranous urethra inferior to the prostate is reconnected to the bladder. A transperineal approach, which may result in a shorter hospital stay, is sometimes used, but the retropubic approach is the most common open procedure. In the latter approach, the seminal vesicles are removed and lymph nodes may be sampled in higher risk patients. The prostate resection may be aborted if the lymph nodes are grossly enlarged or if pathologic frozen section shows metastatic disease. Laparoscopic radical prostatectomies, using either direct manual or remote robotic control, are becoming increasingly common. The slight advantage in reduced operative morbidity and shorter hospital stays is offset by a considerably greater incidence of positive margins (Guillonneau et al., 2003) compared to retropubic results (Hull et al., 2002). The frequency of positive margins decreases with surgeon laparoscopic experience, but there is no difference in patient quality of life at 3, 6, or 12 months later compared to radical retropubic prostatectomy.
3-D Conformal External-Beam Radiation Therapy 3-D conformal external beam radiation therapy (3DCRT) delivers megavoltage photons from a linear accelerator to the target volume from a number of angles. The seminal innovation in 3DCRT was CT-based treatment planning, but successful treatment requires using patient immobilization devices and detailed field shaping with blocks or multileaf collimators. In this way, radiation dose to the target may be safely escalated and dose to normal tissue and sensitive nearby organs minimized. With dose escalation, the typical treatment course lasts 6 to 8 weeks depending on the total prescribed dose and daily fraction size. Because higher doses yield a significant survival benefit (Valicenti et al., 2000), further means of dose escalation have been introduced. IMRT has the potential to reduce morbidity and improve cure rates, but there is not yet sufficient data to make meaningful comparisons (Zelefsky et al., 2002).
536
Wayne M. Butler and Gregory S. Merrick
Interstitial Brachytherapy Beginning in 1970, the first permanent seed prostate implants were performed at Memorial Sloan-Kettering Cancer Center using an open retropubic approach and nomogram-based needle and seed placement. Long-term follow-up of over 1000 patients implanted at Memorial revealed disappointingly few men free of local progression (Zelefsky and Whitmore 1997). The only bright spot in the dismal results was that greater doses and better coverage were associated with a better chance of success. As the open retropubic era of prostate implants fizzled to a halt in the mid-1980s, an improved approach was developed in Scandinavia. Holm et al. (1983) used transrectal ultrasound visualization and a needle-guiding template to implant 125I seeds into the prostate through the perineum. Holm’s techniques were augmented in Seattle by Blasko, Ragde, and Grimm (1991) who used detailed preimplant dosimetric planning, improved stabilization, and preloaded needles. Permanent seed implants are now performed as outpatient surgery and they rarely require a hospital stay. HDR temporary implants are being tested at a number of sites as monotherapy for prostate, but currently the follow-up is too short and the data too sparse for further discussion in this chapter.
Combined Modality Treatments No monotherapy has been shown to be as effective in curing high-risk patients as combining different therapies. A strong case can be made for monotherapy in low-risk patients, while the choice of mono- or combined modality therapy for intermediate-risk patients is not yet settled. 3DCRT is the supplement of choice, usually preceding permanent seed implants to sterilize any microscopic disease in the lymph nodes and seminal vesicles and usually following radical prostatectomy to kill extraprostatic cells not removed by the surgery. HDR brachytherapy and external beam are similarly combined. IMRT is often added to elevate the prostate dose above 80 Gy after 3DCRT has delivered about 45 Gy to the whole pelvis. Some centers use pelvic IMRT for the lymph nodes, but treating the entire course as IMRT monotherapy is a higher cost option that is awaiting supporting evidence.
Hormonal Therapy Although rarely used as a curative regimen, hormonal therapy is initially effective at suppressing prostate cancer growth in most cases. Male hormones (androgens), most notably testosterone, stimulate growth of prostate cancer. Androgen deprivation therapy, ADT, interrupts biochemical pathways for either producing or utilizing testosterone. Most testosterone is produced in the testicles, so removal of the testicles (orchiectomy) was the first effective hormonal therapy. Drug therapy has now largely replaced orchiectomy for ADT. Table 6 lists some of the more common ADT drugs and their mechanism of action. Because no single agent is totally effective, combinations are sometimes used. The combination of an antiandrogen and a luteinizing hormone releasing hormone (LHRH) agonist is called “total androgen blockade.” Female hormones such as estrogen are also effective at suppressing testosterone production but are infrequently used because of feminizing side effects and cardiovascular mortality. Within a few years after the start of hormonal therapy, prostate cancer cells typically become resistant to the therapy, i.e., the tumor becomes hormone refractory, and the disease progresses. Because of the high probability of developing drug resistance and the pronounced side effects of hormones, ADT is rarely used for longer than 3 years in men with high risk prostate cancer. In external beam therapy and brachytherapy, ADT is used to shrink the prostate and appears to act as a radiation sensitizer in addition to its direct tumoricidal effect.
28–Introduction to Prostate Brachytherapy
537
Table 6. Selected Drugs Used In Androgen Deprivation Therapy Drug type
Drug target
Brand name
Generic name
5-alpha reductase
Block the enzyme that converts testosterone to the active form dihydroxytestosterone (DHT)
Proscar Avodart
finasteride dutasteride
Antiandrogen
Block the binding of DHT to testicular cellular receptors
Casodex Eulexin
bicalutamide flutamide
LHRH or GnRH agonist
Block pituitary production of luteinizing hormone needed to produce testosterone in the testicles
Lupron Zoladex
leuprolide goserelin
GnRH antagonist
Bind to luteinizing hormone receptor in testicular cells to block gonadotropin releasing hormone
Plenaxis
abarelix
Biochemical, Recurrence-Free Survival Kaplan-Meier Curves Most survival estimates follow the Kaplan-Meier statistics calculation. Each patient is a point along a time line, and that point is either the elapsed time from treatment to a defined event or from treatment to the last follow-up check. Since we are concerned with biochemical progression-free survival, the signal event will be the date of PSA failure under one of the various formal definitions. The statistical procedure refers to patients who have not failed as “censored,” and they are responsible for holding up the curve. For example, suppose you treated 1 patient each month, and at the end of 4 years you analyze the outcome of your 48 patients. Ideally, there would be one patient point each month out to 48 months, but several events distort the spacing. If a patient is lost to follow-up or dies without evidence of prostate cancer, that patient will be censored at the time from treatment to last follow-up or death. So if a patient you treated 3 years ago died in an accident 6 months after treatment, his data point will be stuck forever at 6 months. If your earliest PSA failure was in a man treated 3 1⁄2 years ago who failed 12 months after treatment, his data point will be stuck at 12 months even though he may still be alive. That first failure will drop the survival curve from 100% to the level represented by the ratio of survivors to the total population at risk at that point in time, in this case, 34/35 or 97.1%. If the second biochemical failure occurs at 36 months where there happens to be 10 patients at risk, the survival curve drops to 97.1% × 9/10 = 87.4%. One of the criticisms of the ASTRO definition is its backdating provision. The patient may have been declared at failure at 36 months, but the midpoint between nadir and first rise may have been at 18 months where there were 28 patients alive and at risk of failure. By moving the effective failure date back to where there are more patients at risk, the survival curve falls to 97.1% × 27/28 = 93.7% rather than 87.4%.
Comparative Single Institution Survival Results In the absence of randomized clinical trials, how can the outcomes of different modalities be compared? One way to minimize the effect of disparate definitions is to look at single institution reports comparing modalities. The Cleveland Clinic found no significant difference between brachytherapy, external-beam radiation therapy, or radical prostatectomy in low- and intermediate-risk patients followed for a median of 48 months and a minimum of 24 months (Ciezki et al., 2004). Survival curves for intermediate-risk patients stratified by the three modalities are shown in Figure 2. The authors used the ASTRO definition of failure for the radiation arms and 0.5 ng/mL for the surgical arm. They did not indicate separate mean patient age for each modality, but did report that short duration androgen deprivation therapy was employed
538
Wayne M. Butler and Gregory S. Merrick
Figure 2. Single institution biochemical progression free survival in intermediate-risk patients stratified by treatment modality. (Adapted from Ciezki et al., 2004.)
in 64%, 59%, and 17% of patients in the brachytherapy, 3DCRT, and surgery cohorts, respectively. ADT was not found to affect biochemical outcomes, and the known short-term suppression of PSA by ADT was not a factor because of the minimum 2-year follow-up. Another single practice employing either brachytherapy or surgery also found no significant difference between modalities for low-risk patients (Sharkey et al., 2005). The authors used the ASTRO definition of failure for brachytherapy and 0.4 ng/mL for the surgery. However, as shown in Figure 3, the differences were pronounced for intermediate- and high-risk patients. The surgery patients were considerably younger than the brachytherapy patients, 63 versus 72 years. ADT and external-beam therapy were also not shown to affect biochemical survival in any risk group.
Inter-institution and Inter-modality Comparisons How do we know whether published results are representative of nationwide clinical practice? We don’t know. However, by searching for the best reported outcomes that have been replicated by several institutions for each modality we can explores the realm of what is achievable within a given technique. In the low-, intermediate-, and high-risk survival sections below, only studies that classified patients into risk groups and which had follow-up exceeding 8 years were considered for inclusion. Long follow-up minimizes the confounding effect of early ADT and reduces the boost or penalty afforded by various definitions of failure. The studies and institutions referenced are not comprehensive and are biased toward larger populations over studies of similar quality but concerning smaller cohorts. Searching for the worst reported outcomes was not deemed worthwhile because there is little incentive for investigators to demonstrate that their results are as bad or worse than someone else’s.
Actuarial Survival for Low-Risk Patients Figure 4 illustrates comparative biochemical progression free survival results for patients at low risk as defined in Table 2. Data curves were selected as representative of the best-published results within a given modality for hormone naïve patients treated with a monotherapy approach. Radical prostatectomy results are from Baylor (Hull et al., 2002); 3-D conformal radiation therapy results are from Fox Chase (Feigenberg et al., 2005); permanent seed brachytherapy results are from Seattle (Blasko et al., 2000). The median
28–Introduction to Prostate Brachytherapy
539
Figure 3. Single institution biochemical progression free survival in intermediate and high risk patients treated with either brachytherapy or surgery. The difference between brachytherapy and radical prostatectomy for each risk group was statistically significant (p < 0.05). (Adapted from Sharkey et al., 2005.)
Figure 4. Comparative biochemical progression free survival results for patients at low risk, defined as initial PSA ≤ 10 ng/mL, Gleason score ≤ 6, and clinical stage ≤ T2a. Data curves were selected as representative of the best-published results for hormone naïve patients treated with a monotherapy approach. Radical prostatectomy results are from Hull et al. (2002); 3-D conformal radiation therapy results are from Feigenberg et al. (2005); permanent seed brachytherapy results are from Blasko et al. (2000).
540
Wayne M. Butler and Gregory S. Merrick
age in the 3DCRT and brachytherapy series was 69 years and 68 years, respectively, but the median age in the surgical series was only 62 years. Surgeons usually contend that there is no substantive difference between radical prostatectomy and brachytherapy in low-risk patients. Radiation oncologists may concede that the differences between 3DCRT and brachytherapy are substantial, but they also could contend that the treatment dose in the Fox Chase series was not high enough—the median dose was 76 Gy. However, the Fox Chase study also found that increasing the 3DCRT dose above 76 Gy had no significant effect on the biochemical survival of low risk men. On the other hand, the 4-year biochemical survival rates using intensity modulated radiation therapy with a minimum dose of 81 Gy at Memorial Sloan-Kettering (Zelefsky et al., 2002) are quite similar to reported results in the brachytherapy literature at similar maturity. Enthusiasm that any type of external-beam therapy is finally demonstrably competitive with other monotherapies must be tempered by the observation that the median follow-up in the MSKCC series was only 2 years and the curves were heavily front-loaded. Figure 5 illustrates the results in low dose-rate (LDR) and HDR brachytherapy when combined with external-beam therapy. The permanent seed brachytherapy data are from Critz and Levinson (2004), and the HDR brachytherapy results are from William Beaumont (Galalae et al., 2004). The permanent seed implant results are certainly no better with the addition of external beam therapy than the survival results of Figure 4 following monotherapy implants.
Actuarial Survival for Intermediate-Risk Patients Biochemical progression-free survival for hormone naïve intermediate risk patients stratified by monotherapy treatment modality is shown in Figure 6. Radical prostatectomy results are from Baylor (Hull et al., 2002), 3DCRT results are from M.D. Anderson Cancer Center (Cheung et al., 2005), and permanent seed brachytherapy results are from Merrick et al. (2005a). There is no consensus on the role of supplemental external beam radiation therapy and androgen deprivation therapy in intermediate risk patients. Figure 7 illustrates the results for LDR and HDR brachytherapy when combined with external-beam therapy. The permanent seed brachytherapy data are from Critz and Levinson (2004) whose patients were hormone naïve and treated with ≥45 Gy of external-beam radiation. The HDR brachytherapy results are from a multi-institution retrospective analysis that included some patients receiving ADT and various HDR fractionations but with external beam doses between 45 and 50 Gy in all cases (Galalae et al., 2004). Once again, note that the permanent seed implant results are no better after adding external beam than the biochemical survival following monotherapy implants as shown in Figures 3 and 6. It is somewhat surprising that the biochemical survival rate of intermediate-risk brachytherapy patients is virtually the same as for low-risk patients. This phenomenon is common in the permanent seed brachytherapy literature and has also begun to appear in HDR brachytherapy studies (Aström et al., 2005; Demanes et al., 2005). One possible explanation is based on data from the Partin tables (see Table 2). The primary difference between low- and intermediate-risk patients is the percentage of disease organ confined, which ranges from as high as 95% in very low-risk patients to as low as 21% in intermediate-risk patients with several borderline unfavorable features. All the while, the likelihood of finding disease in the lymph nodes, which are not treated in brachytherapy, remains below 5%. Brachytherapy implants that consistently treat a wide margin sufficient to sterilize extraprostatic disease extension should result in disease-free survival rates that are nearly indistinguishable between the first two risk groups.
Actuarial Survival for High-Risk Patients There are relatively few outcome studies of monotherapy in high-risk patients, and the numbers of patients in those studies tend to be small. The radical prostatectomy and 3DCRT patients in Figure 8 were hormone
28–Introduction to Prostate Brachytherapy
541
Figure 5. Comparative biochemical progression free survival results for patients at low risk, defined as initial PSA ≤ 10 ng/mL, Gleason score ≤ 6, and clinical stage ≤ T2a. Data curves were selected as representative of the best-published results for hormone naïve patients treated with a combined modality approach. HDR brachytherapy results are from Galalae et al. (2004); permanent seed brachytherapy results are from Critz and Levinson (2004).
Figure 6. Comparative biochemical progression free survival results for patients at intermediate risk. Data curves were selected as representative of the best-published results for hormone naïve patients treated with a monotherapy approach. Radical prostatectomy results are from Hull et al. (2002); 3-D conformal radiation therapy results are from Cheung et al. (2005); permanent seed brachytherapy results are from Merrick et al. (2005a).
542
Wayne M. Butler and Gregory S. Merrick
Figure 7. Comparative biochemical progression-free survival results for patients at intermediate risk. Data curves were selected as representative of the best-published results for hormone naïve patients treated with a combined modality approach. High dose rate brachytherapy results are from Galalae et al. (2004); permanent seed brachytherapy results are from Critz and Levinson (2004).
Figure 8. Comparative biochemical progression free survival results for patients at high risk. Data curves were selected as representative of the best-published results for patients treated with a monotherapy approach. Radical prostatectomy results are from Hull et al. (2002); 3-D conformal radiation therapy results with a median dose of 74 Gy and excluding patients with PSA > 20 ng/mL are from Hanks et al. (2002); the brachytherapy survival results exclude patients whose day 30 D90 dose was < 88% of the prescribed dose and are not purely monotherapy because some patients were treated with short term hormones (Stone, Stock, and Unger 2005).
naïve and truly monotherapy, while the brachytherapy series is not purely monotherapy because some of the patients received short-term ADT. The median dose in the external-beam study was 74 Gy, and the survival curve does not include patients with PSA > 20 ng/mL (Hanks et al., 2002). The brachytherapy survival curve from Mt. Sinai excludes patients whose minimal dose covering 90% of the prostate (D90) on day 30 dosimetry was <88% of the prescribed 125I dose of 160 Gy (Stone, Stock, and Unger 2005).
28–Introduction to Prostate Brachytherapy
543
Most of the published brachytherapy studies of high-risk patients have used an external-beam boost as well as ADT. Of the patients in the permanent seed brachytherapy curve of Figure 9, 40% received ADT (Merrick et al., 2005a). The HDR high-risk patients were hormone naïve, but the authors uniquely defined high-risk patients as having one or more adverse prognosticators: clinical stage T3, PSA > 20 ng/mL, or Gleason score 8–10 (Demanes et al., 2005). The outstanding progression free survival results from the Mt. Sinai group represent a selected cohort of high-risk patients, and those results have not yet been duplicated in any unstratified cohort of truly monotherapy high-risk patients. There is a broad consensus that high-risk patients treated with a permanent seed implant should also receive external-beam therapy. Androgen suppression may also be appropriate in patients at high risk of metastatic disease.
Treatment-related Morbidity Every effective cancer treatment has side effects and complications. Any therapy that claims to cure cancer without any morbidity either has not been sufficiently studied or the proponents are misinformed or selling worthless quackery. Yes, prostate brachytherapy does create patient morbidity, as do all other conventional treatments. The morbidity profiles of brachytherapy, radical surgery, and 3DCRT all differ in frequency and intensity of each effect, so much so that a comparison between modalities is beyond the scope of this survey. Prostate brachytherapy has often been oversold in terms of its low morbidity, and patients frequently come to the clinic with unrealistic expectations. A perfunctory reiteration of items listed on the informed consent form may be all that is legally required, but the patient is ill served unless the physician places each item in a context meaningful to the patient. In prostate brachytherapy, detailed dosimetric analyses have demonstrated that side effects, complications, and quality of life outcomes (QOL) are dependent on implant quality and are related to specific source placement patterns and the subsequent dose gradients produced (Butler and Merrick 2003).
Figure 9. Comparative biochemical progression-free survival results for patients at high risk. Data curves were selected as representative of the best-published results for hormone naïve patients treated with a combined modality approach. The permanent seed brachytherapy results included some patients who received short term ADT (Merrick et al. 2005a). The HDR patients were hormone naïve, but high risk was defined as one or more adverse prognosticators: clinical stage T3, PSA > 20 ng/mL, or Gleason score 8–10 (Demanes et al. 2005).
544
Wayne M. Butler and Gregory S. Merrick
Brachytherapy-related morbidity can be positively affected by refinements in patient selection, planning and intraoperative technique, and medical intervention (Merrick, Wallner, and Butler 2003).
Urinary Morbidity Urinary morbidity in the immediate post-implant setting has been well documented, with preimplant IPSS correlating with the duration of post implant obstructive symptoms (Desai et al., 1998). Approximately 12 months was required for the mean IPSS to return to the preimplant baseline. Certain interventions may substantially accelerate the recovery of urinary function. The Role of Alpha-blockers on IPSS Alpha-blockers such as tamsulosin (Flomax™) are widely used to ameliorate brachytherapy-related urinary morbidity. The initiation of alpha-blockers 2 to 3 weeks prior to implantation and continuation until IPSS normalization appears to maximize their beneficial effect. With prophylactic and prolonged alpha-blocker usage, Niehaus et al. (2005) found that the mean population IPSS returned to baseline values 4 months following brachytherapy (Figure 10), and half the patients returned to their baseline score at two weeks (Figure 11). However, alpha-blocker usage did not correlate with patients’ catheter dependency to relieve acute obstructive symptoms nor the need for post-implant surgical intervention. Fortunately, catheter dependence and the need for a post-implant TURP/TUIP manifest in less than 2% of patients (Figure 12). Niehaus and colleagues also reported that regardless of prostate size, radionuclide choice did not impact IPSS resolution, catheter dependency or the need for post brachytherapy surgical intervention. In terms of radionuclide choice, a prospective randomized trial comparing 103Pd with 125I reported a statistically faster rate of IPSS resolution in the 103Pd arm (Wallner et al., 2002).
Figure 10. Mean International Prostate Symptom Score (IPSS) difference from pre-implantation baseline (defined as 0) for 976 patients. All patients were treated prophylactically with alpha-blockers, and there were an average of 21 IPSS questionnaires returned per patient. The maximum increase over baseline occurred at 2 weeks post implant, and after 4 months the mean score difference dropped below the preimplant baseline. Data from Niehaus et al. (2005).
28–Introduction to Prostate Brachytherapy
545
Figure 11. Kaplan-Meier curve of the rate of return of 976 patients to their preimplant International Prostate Symptom Score. All patients were treated prophylactically with alpha-blockers. At 2 weeks post implant, half of the patients had returned to their preimplant baseline score. Data from Niehaus et al. (2005).
Figure 12. Distribution of postimplant day of urinary catheter removal for 516 consecutive patients treated prophylactically with alpha-blockers. Less than 2% of patients had any catheter dependence beyond 4 days post implant. Data from Niehaus et al. (2005).
546
Wayne M. Butler and Gregory S. Merrick
Dysuria is a relatively common event during the first few years following brachytherapy, but there are no items on the IPSS questionnaire (Table 3) pertaining to pain or burning on urination. Only rarely is dysuria severe in frequency or intensity (Merrick et al., 2003a). No clinically significant factors for the prediction of dysuria have been identified nor has the use of alpha-blockers significantly diminished the duration of dysuria (Merrick et al., 2005c). However, anti-inflammatory agents such as low-dose prednisone (5 to 10 mg daily) and the avoidance of bladder irritants such as caffeine or alcohol may alleviate the intensity and frequency of treatment-related dysuria. Urethral Dosimetry Wallner, Roy, and Harrison (1995) reported an association between urinary morbidity and urethral doses greater than 250% of the prescribed minimum peripheral dose (mPD). In contemporary series, urethral doses have not correlated with urinary morbidity because sophisticated treatment planning has maintained urethral sparing doses at 100% to 140% mPD. While urethral doses greater than 150% mPD should be minimized, under dosage of the urethra (>100% mPD) should be avoided. Leibovich and colleagues (2000) have measured the mean distance from the urethra to the nearest foci of cancer to be 3 mm and that 17% of all prostate cancer abuts the urethra. Although it is conceivable that certain segments of the urethra may be more sensitive to radiationinduced morbidity, detailed urethral dosimetry did not substantially improve the ability to predict urinary morbidity (Allen et al., 2005). Radiation doses of 100% to 140% mPD were well tolerated by all segments of the prostatic urethra. It is also possible that the relative insensitivity of segmental urethral doses may have been a result of prophylactic and prolonged alpha-blocker usage. Urethral strictures occur in 4% to 12% of patients and are directly related to over implantation of the periapical region and the use of supplemental external beam radiation therapy (Merrick et al., 2002b, 2005d). Typically, strictures involve the bulbomembranous urethra and are easily managed with dilatation or optical internal urethrotomy. Day 0 CT-based dosimetry has demonstrated that radiation doses to the bulbomebranous urethra are significantly greater in patients with strictures than in those without. With careful attention to implant technique including extensive use of sagittal ultrasonography, it is possible to implant the apex with a 5-mm margin without “dragging” seeds into the bulbomembranous urethral region.
Rectal Dosimetry and Morbidity Rectal complications primarily consist of mild self-limited proctitis (incidence 4% to 12%), and they usually resolve spontaneously (Gelblum and Potters 2000). Bowel function assessment by patient-administered questionnaires has documented that long-term bowel function is minimally affected by brachytherapy (Talcott et al., 2001). Using the RFAS questionnaire (Table 5), Merrick et al. (2003c) found that bowel function worsened slightly and remained elevated. The mean preimplant RFAS was 2.5 and the post implant RFAS was 4 out of the maximum score of 27 (Figure 13). On a retrospective summary question that asked patients to compare their current bowel function to what they recall from before the implant, a large majority perceived no change and those who felt their function had improved or worsened were about equal in number (Figure 14). Rectal dose is important in more severe complications, but it is difficult to compare rectal dosimetry between institutions because the timing of postimplant CT and how the rectum or rectal wall was evaluated varies widely. Waterman and Dicker (1999) reported that the minimum dose that encompassed 10% of the surface area of the rectum (rectal area D10) increased by 68% from day 0 through day 30. Using day 30 CT-based dosimetry, Snyder and colleagues (2001) reported grade II proctitis to be rectal volume dependent for any given dose and that no cases developed more than 36 months following implantation. They documented an 8% rate of grade II proctitis when less than 1.8 cm3 of rectum was exposed to 160 Gy following 125I monotherapy, while the risk increased to 25% when more than 1.8% of the rectum was
28–Introduction to Prostate Brachytherapy
547
Figure 13. Rectal function assessment score (RFAS) changes over time for 189 patients surveyed in 1999 (open circles) and again in 2002 (filled circles). The dark solid line indicates an insignificant improvement during the 3-year interval (Merrick et al. 2003c).
Figure 14. Patient perception of changes in bowel function relative to preimplant status. Questionnaires were administered to 189 patients in 1999 and again in 2002 (Merrick et al. 2003c).
so exposed. In a prospective randomized trial, the minimum dose received by 5% of the rectum best correlated with brachytherapy-related rectal morbidity (Merrick et al., 2003b). Although it has been assumed that perirectal sources increase rectal morbidity, a limited number of errant perirectal sources did not increase the risk of rectal bleeding provided the overall rectal wall doses were within acceptable values (Mueller et al., 2004). Following brachytherapy, rectal ulceration and fistula formation have occasionally been reported. Although dose is associated with rectal bleeding, Howard et al. (2001) found no correlation between rectal dose and the development of a fistula. They concluded that severe complications might occur in
548
Wayne M. Butler and Gregory S. Merrick
an unpredictable manner unrelated to known clinical, treatment, or dosimetric parameters. A reasonable conjecture is that a patient experiencing rectal bleeding sees his family physician, and that visit precipitates invasive rectal procedures such as biopsy, cauterization, laser coagulation, etc. These procedures exacerbate the problem because of the limited healing capacity of highly irradiated tissue. Although no studies have correlated constipation with rectal toxicity, constipation significantly increases the radiation dose to the rectum (Merrick et al., 2000b). For patients prone to constipation, postimplant attention to bowel habits and the use of stool softeners and laxatives for two half-lives of the implanted source will minimize rectal distention and decrease the dose to the anterior rectal wall. Intraoperatively, careful attention to implant technique and ultrasound anatomy can reduce the dose to the anterior rectal wall and minimize bowel dysfunction. Extensive use of both transverse and sagittal ultrasonography to confirm needle placement and the use of multiple ultrasound frequencies helps ensure proper seed placement. Higher transducer frequencies result in clearer definition of anatomy closer to the probe. Posterior row needle placements executed with the 7.5 MHz frequency help ensure that needles are placed within the posterior prostate capsule and not in the rectal wall.
Erectile Dysfunction Erectile dysfunction (ED) is a common sequela of all potentially curative local treatments. The reported rates of post-implant ED vary widely reflecting substantial differences in follow-up, patient selection, implant technique, and the mode of data collection. Early brachytherapy studies reported exceptionally high rates of potency preservation (Wallner, Roy, and Harrison 1996). However, patient-administered quality of life questionnaires differ markedly from data collected by physician interview, especially on sensitive sexual issues (Litwin et al., 1998). Following all treatment approaches, potency preservation rates are significantly lower when patients fill out a form rather than answering oral questions from their treating physician. Although the perception exists that preservation of erectile function is more likely following brachytherapy than other local treatments, the incidence of brachytherapy-induced ED is substantially greater than initially reported (Merrick et al., 2002c, 2005e). Subgroup analyses documented ED in 6% to 90% of patients undergoing brachytherapy with or without supplemental external-beam radiation therapy and/or androgen deprivation therapy. Mechanism of Brachytherapy-Induced Erectile Dysfunction Radiation-induced ED likely represents a multifactorial process including neurogenic compromise, vascular insufficiency, local trauma, and psychogenic causes with microvascular damage representing the most dominant factor (Zelefsky et al., 1999). There are several candidates for structures at risk. Emerging evidence strongly suggests that brachytherapy-related ED is technique-related and may be minimized with careful attention to radioactive source placement. Radiation Dose to the Prostate Gland Conflicting results have been reported regarding the relationship between radiation dose to the prostate gland and the development of brachytherapy-related ED. In multivariate analysis, Stock, Kao, and Stone (2001) reported that a D90 > 160 Gy for 125I implants and >100 Gy for 103Pd implants predicted for brachytherapy-induced ED. However, the absolute differences in potency preservation were minimal— 58% versus 64%, p = 0.02. In contrast, in both retrospective and prospective evaluations, Merrick and colleagues were unable to discern a relationship between radiation dose to the prostate gland and brachytherapy-related ED (Merrick et al., 2002c, 2005e).
28–Introduction to Prostate Brachytherapy
549
Dose to the Neurovascular Bundles Following radical prostatectomy, ED has been correlated with surgical trauma to the neurovascular bundles. It is conceivable that excessive radiation doses to the neurovascular bundles (NVB) could also represent a potential mechanism of brachytherapy-induced ED (Zelefsky and Eid 1998). Radiation doses to the NVB were calculated using an idealized 2-dimensional geometric model derived from day 0 CTbased images (DiBiase et al, 2000). Using the same methodology in both prospective and retrospective analyses, Merrick et al., (2000c) found no relationship between radiation dose to the NVB and the development of brachytherapy-induced ED. A recent study from Wake Forest University confirmed these findings (Kiteley et al., 2002). Although initial reports have shown no correlation between ED and NVB dose, it is possible that with additional follow-up those doses will be found to contribute to brachytherapy-related ED. Dose to Penile Erectile Bodies The penile erectile bodies—the paired corpora cavernosa and the midline corpus spongiosum—are potential targets for structure-specific radiation-associated ED. Figure 15 is a cartoon of the anatomy. Although the site-specific structure in the proximal penis remains unclear, radiation doses to the corpora cavernosa may be more important than those to the penile bulb because the corpora cavernosa represents the true erectile tissue whereas the corpus spongiosum plays little role in the development or maintenance of erectile rigidity (Mulhall et al., 2002). Detailed reports illustrating the image-based anatomy of the proximal penis have been published by Wallner et al., 2002a). The penile bulb is best visualized on T2-weighted magnetic resonance images and appears as an oval-shaped, hyper-dense midline structure located approximately 10 to 15 mm inferior to the apex of the prostate gland. In both a laboratory rat model and in clinical studies, investigators at the University of California at San Francisco reported that the external-beam radiation dose to the bulb of the penis correlated directly with post-treatment ED (Fisch et al., 2001). A higher incidence of ED was reported with increasing radiation doses to the proximal penis. The authors hypothesized that external-beam radiation directly or indirectly damaged the vascular supply of the erectile tissue as well as the nerves that supply the cavernosal smooth muscles. Subsequent histologic examination of the proximal penile shaft specimens demonstrated that the numbers of nitric oxide synthase-containing nerve fibers per corpus cavernosum were significantly decreased in the irradiated groups.
Figure 15. Sagittal schematic of the prostate and nearby structures.
550
Wayne M. Butler and Gregory S. Merrick
Prostate brachytherapy-induced ED was found to be highly dependent on the radiation dose to the proximal penis (Merrick et al., 2001b, 2002d). Using day 0 CT-based dosimetry, maximum potency preservation was dependent on limiting the penile bulb D50 to ≤40% of prescription dose and the proximal crura D50 to ≤50% of the prescription dose (Merrick et al., 2005e). When stratified by these cutpoints, greater than 70% of patients with dose below the cutpoint maintained potency versus only 30% in patients who received higher doses (Figure 16). In contrast, a patient series from Kiteley et al., (2002) failed to establish a relationship between radiation dose to the proximal penis and the development of brachytherapy-related ED. Ultimately, the proximal penis may prove to be the most important site-specific structure for the development of radiation therapy-induced ED. In the case of brachytherapy, suboptimal placement (either due to poor planning or poor implantation) of periapical radiation sources results in excessive radiation doses to the proximal penis. Refinements in implant technique to avoid over-aggressive periapical implantation along with extensive use of sagittal ultrasonography during the implant procedure will decrease the radiation dose to the proximal penis and may improve potency preservation. Potency Preservation ED following definitive therapy increases with time after treatment, and series with longer follow-up have uniformly reported lower rates of potency preservation. Erectile function gradually declines with age, and older patients are less likely to maintain erectile function following any definitive procedure, including brachytherapy. In a recent evaluation of low-risk patients on a prospective randomized trial, Merrick and colleagues (2005e) found that 50% of patients maintained potency 4 years following prostate brachytherapy. Preimplant erectile function—measured by the IIEF (Table 4)—represents the strongest predictor of post-treatment erectile function (Merrick et al., 2005e). That study also discovered that men who experienced nocturnal erections prior to treatment were more likely to maintain potency. Supplemental external-beam radiation therapy and androgen deprivation therapy may be detrimental when combined with brachytherapy in their impact on potency preservation, but there have been conflicting results regarding their effect. Merrick et al. (2005e) found no significant difference between the potency preservation rates between men treated with or without 3DCRT, or with or without ADT, or with 125I or 103Pd. Their
Figure 16. Potency preservation stratified by a penile bulb D50 dose of 30% of the prescribed minimal peripheral dose (Merrick et al. 2005e).
28–Introduction to Prostate Brachytherapy
551
3DCRT and brachytherapy treatment planning were designed to limit the dose to the proximal penis. On the other hand, Potters and colleagues (2001) found that use of ADT was a strong predictor for ED and that men treated with 3DCRT and permanent implant were much less likely to retain potency. Fortunately, the majority of patients with brachytherapy-induced ED respond favorably to sildenafil citrate (Viagra) (Merrick et al., 1999) or the other PDE-5 inhibitors (type 5 phosphodiesterase) vardenafil (Levitra) and tadalafil (Cialis). A 6-year actuarial rate of potency preservation of 92% was reported when potent patients were analyzed together with erectile dysfunction patients who used sildenafil (Merrick et al., 2002c).
Conclusions There is a growing volume of evidence that permanent prostate brachytherapy produces very favorable and durable biochemical survival for patients with early-stage carcinoma and who have either low-, intermediate-, or high-risk features. There is also increasing evidence in support of the trend to rely less on other modalities—3DCRT and ADT—in combination with brachytherapy in favor of a monotherapy approach. Biochemical progression-free survival is now so high in low- and intermediate-risk patients that only a very large, highly powered randomized trial could prove any innovation or competing modality to be superior to current permanent seed brachytherapy. With feedback from post-implant evaluation, brachytherapy planning philosophies and operative techniques have been refined not only to further enhance disease-free survival but also to further reduce ED and urinary and rectal morbidity. Significant urinary and rectal morbidity is becoming increasingly rare, and most patients with brachytherapy-induced ED respond positively to sildenafil.
References Abuzallouf, S., I. Dayes, and H. Lukka. (2004). “Baseline staging of newly diagnosed prostate cancer: a summary of the literature.” J Urol 171:2122–2127. Allen, Z. A., G. S. Merrick, W. M. Butler, K. E. Wallner, B. Kurko, R. L. Anderson, B. C. Murray, R. W. Galbreath. (2005). “Detailed urethral dosimetry in the evaluation of prostate brachytherapy-related urinary morbidity.” Int J Radiat Oncol Biol Phys. In press. American Cancer Society. Cancer Facts and Figures 2005. Atlanta, GA: American Cancer Society, 2005. American Joint Committee on Cancer. AJCC Cancer Staging Manual, 6th Ed. New York: Springer-Verlag, 2002. American Society for Therapeutic Radiology and Oncology Consensus Panel. (1997). “Consensus statement: guidelines for PSA following radiation therapy.” Int J Radiat Oncol Biol Phys 37:1035–1041. Andriole, G. L., D. L. Levin, E. D. Crawford, E. P. Gelmann, P. F. Pinsky, D. Chia, B. S. Kramer, D. Reding, T. R. Church, R. L. Grubb, G. Izmirlian, L. R. Ragard, J. D. Clapp, P. C. Prorok, and J. K. Gohagan. (2005). “Prostate Cancer Screening in the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial: Findings from the initial screening round of a randomized trial.” J Natl Cancer Inst 97:433–438. Aström, L., D. Pedersen, C. Mercke, S. Holmäng, and K. A. Johansson. (2005). “Long-term outcome of high dose rate brachytherapy in radiotherapy of localised prostate cancer.” Radiother Oncol 74:157–161. Augustin, H., M. Graefen, J. Palisaar, J. Blonski, A. Erbersdobler, F. Daghofer, H. Huland, and P. G. Hammerer. (2003). “Prognostic significance of visible lesions on transrectal ultrasound in impalpable prostate cancers: implications for staging.” J Clin Oncol 21:2860–2868. Aus, G., D. Robinson, J. Rosell, G. Sandblom, and E. Varenhorst. (2005). “Survival in prostate carcinoma—outcomes from a prospective, population-based cohort of 8887 men with up to 15 years of follow-up: Results from three counties in the population-based National Prostate Cancer Registry of Sweden. Cancer 103:943–951. Beard, C. J., M. H. Chen, K. Cote, M. Loffredo, A. A. Renshaw, M. Hurwitz, and A.V. D’Amico. (2004). “Perineural invasion is associated with increased relapse after external beam radiotherapy for men with low-risk prostate cancer and may be a marker for occult, high-grade cancer.” Int J Radiat Oncol Biol Phys 58:19-24.
552
Wayne M. Butler and Gregory S. Merrick
Berger, A. P., M. Deibl, H. Steiner, J. Bektic, A. Pelzer, R. Spranger, H. Klocker, G. Bartsch, and W. Horninger. (2005). “Longitudinal PSA changes in men with and without prostate cancer: Assessment of prostate cancer risk.” Prostate Feb 14, epub ahead of print. Blasko, J. C., H. Ragde, and P. D. Grimm. (1991). “Transperineal ultrasound-guided implantation of the prostate: morbidity and complications.” Scand J Urol Nephrol Suppl 137:113–118. Blasko, J. C., P. D. Grimm, J. E. Sylvester, and W. Cavanagh. (2000). “The role of external beam radiotherapy with I-125/Pd-103 brachytherapy for prostate carcinoma.” Radiother Oncol 57:273–278. Bostwick, D. G. (1994). “Gleason grading of prostatic needle biopsies. Correlation with grade in 316 matched prostatectomies.” Am J Surg Pathol 18:796–803. Brenner, H., and V. Arndt. (2005). “Long term survival rates of patients with prostate cancer in the prostate-specific antigen screening era: Population-based estimates for the year 2000 by period analysis.” J Clin Oncol 23:441–447. Burnet, N. G., S. J. Jefferies, R.J. Benson, D.P. Hunt, and F.P. Treasure. (2005). “Years of life lost (YLL) from cancer is an important measure of population burden—and should be considered when allocating research funds.” Br J Cancer 92:241–245. Butler, W. M., and G. S. Merrick. (2003). “Myths and fallacies in permanent prostate brachytherapy.” Med Dosim 28:155–165. Carroll, P. R. (2005). “Early stage prostate cancer-do we have a problem with over-detection, overtreatment or both?” J Urol 173:1061–1062. Carter, H. B., J. I. Epstein, and A. W. Partin. (1999). “Influence of age and prostate-specific antigen on the chance of curable prostate cancer among men with nonpalpable disease.” Urol 53:126–130. Carter, H. B., and J. D. Pearson. (1999). “Prostate-specific antigen testing for early diagnosis of prostate cancer: formulation of guidelines.” Urol 54:780–786. Cheung, R., S. L. Tucker, A. K. Lee, R. de Crevoisier, L. Dong, A. Kamat, L. Pisters, and D. Kuban. (2005). “Doseresponse characteristics of low- and intermediate-risk prostate cancer treated with external beam radiotherapy.” Int J Radiat Oncol Biol Phys 61:993–1002. Ciezki, J. P., E. A. Klein, K. Angermeier, J. Ulchaker, N. Chehade, A. Altman, Mahadevan, and C. A. Reddy. (2004). “A retrospective comparison of androgen deprivation (AD) vs. no AD among low-risk and intermediate-risk prostate cancer patients treated with brachytherapy, external beam radiotherapy, or radical prostatectomy.” Int J Radiat Oncol Biol Phys 60:1347–1350. Clarke, D. H., S. J. Banks, A. R. Wiederhorn, J. W. Klousia, J. M. Lissy, M. Miller, A. M. Able, C. Artiles, W. V. Hindle, D. N. Blair, R. R. Houk, and M. J. Sheridan. (2002). “The role of endorectal coil MRI in patient selection and treatment planning for prostate seed implants.” Int J Radiat Oncol Biol Phys 52:903–910. Critz, F. A., and K. Levinson. (2004). “10-year disease-free survival rates after simultaneous irradiation for prostate cancer with a focus on calculation methodology.” J Urol 172:2232–2238. D’Amico, A, A. V., R. Whittington, B. Malkowicz, M. Schnall, D. Schultz, K. Cote, J. E. Tomaszewski, and A. Wein. (2000). “Endorectal magnetic resonance imaging as a predictor of biochemical outcome after radical prostatectomy in men with clinically localized prostate cancer.” J Urol 164:759–763. D’Amico, A. V., J. Moul, P. R. Carroll, L. Sun, D. Lubeck, and M. H. Chen. (2004). “Prostate specific antigen doubling time as a surrogate end point for prostate cancer specific mortality following radical prostatectomy or radiation therapy.” J Urol 172:S42–S46; D’Amico, A. V., Y. Wu, M. H. Chen, M. Nash, A. A. Renshaw, and J. P. Richie. (2001). “Perineural invasion as a predictor of biochemical outcome following radical prostatectomy for select men with clinically localized prostate cancer.” J Urol 165:126–129. D’Amico, A. V., A. A. Renshaw, K. Cote, M. Hurwitz, C. Beard, M. Loffredo, and M. H. Chen. (2004). “Impact of the percentage of positive prostate cores on prostate cancer-specific mortality for patients with low or favorable intermediate-risk disease.” J Clin Oncol 22:3726–3732. Demanes, D. J., R. R. Rodriguez, L. Schour, D. Brandt, and G. Altieri. (2005). “High-dose-rate intensity-modulated brachytherapy with external beam radiotherapy for prostate cancer: California endocurietherapy’s 10-year results.” Int J Radiat Oncol Biol Phys 61:1306–1316. Desai, J., R. G. Stock, N. N. Stone, C. Iannuzzi, J. K. DeWyngaert. (1998). “Acute urinary morbidity following I-125 interstitial implantation of the prostate gland.” Radiat Oncol Investig 6:135–141.
28–Introduction to Prostate Brachytherapy
553
DiBiase, S. J., K. Wallner, K. Tralins, and S. Sutlief. (2000). “Brachytherapy radiation doses to the neurovascular bundles.” Int J Radiat Oncol Biol Phys 46:1301–1307. Egevad, L., W. C. Allsbrook, Jr., and J. I. Epstein. (2005). “Current practice of Gleason grading among genitourinary pathologists.” Hum Pathol 36:5–9. Elliott, K., K. Wallner, G. Merrick, and P. Herstein. (2004). “Medical malpractice of prostate brachytherapy.” Brachytherapy 3:231–236. Feigenberg, S. J., A. L. Hanlon, E. M. Horwitz, R. G. Uzzo, D. F. Eisenberg, and A. Pollack. (2005). “What pretreatment prostate-specific antigen level warrants long-term androgen deprivation?” Int J Radiat Oncol Biol Phys 61:1003–1010. Fisch, B. M., B. Pickett, V. Weinberg, M. Roach. (2001). “Dose of radiation received by the bulb of the penis correlates with risk of impotence after three-dimensional conformal radiotherapy for prostate cancer. Urol 57:955–959. Freedland, S. J., W. J. Aronson, G. S. Csathy, C. J. Kane, C. L. Amling, J. C. Presti Jr., F. Dorey, and M. K. Terris. (2003). “Comparison of percentage of total prostate needle biopsy tissue with cancer to percentage of cores with cancer for predicting PSA recurrence after radical prostatectomy: results from the SEARCH database.” Urol 61:742–747. Galalae. R. M., A. Martinez, T. Mate, C. Mitchell, G. Edmundson, N. Nuernberg, S. Eulau, G. Gustafson, M. Gribble, and G. Kovács. (2004). “Long-term outcome by risk factors using conformal highdose-rate brachytherapy (HDR-BT) boost with or without neoadjuvant androgen suppression for localized prostate cancer.” Int J Radiat Oncol Biol Phys 58:1048–1055. Gelblum, D. Y., and L. Potters. (2000). “Rectal complications associated with transperineal interstitial brachytherapy for prostate cancer.” Int J Radiat Oncol Biol Phys 48:119–124. Gilbert, S. M., C. B. Cavallo, H. Kahane, and F. C. Lowe. (2005). “Evidence suggesting PSA cutpoint of 2.5 ng/mL for prompting prostate biopsy: Review of 36,316 biopsies.” Urol 65:549–553. Grann, A., and K. Wallner. (1998). “Prostate brachytherapy in patients with inflammatory bowel disease.” Int J Radiat Oncol Biol Phys 40:135–138. Grossfeld, G. D., J. J. Chang, J. M. Broering, Y. P. Li, D. P. Lubeck, S. C. Flanders, and P. R. Carroll. (2001). “Under staging and under grading in a contemporary series of patients undergoing radical prostatectomy: Results from the Cancer of the Prostate Strategic Urologic Research Endeavor database.” J Urol 165:851–856. Guillonneau, B., H. el-Fettouh, H. Baumert, X. Cathelineau, J. D. Doublet, G. Fromont, and G. Vallancien. (2003). “Laparoscopic radical prostatectomy: oncological evaluation after 1,000 cases at Montsouris Institute.” J Urol 169:1261–1266. Han, M., P. H. Gann, and W. J. Catalona. (2004). “Prostate-specific antigen and screening for prostate cancer.” Med Clin North Am 88:245–265. Hanks, G. E., A. L. Hanlon, B. Epstein, and E. M. Horwitz. (2002). “Dose response in prostate cancer with 8-12 years’ follow-up.” Int J Radiat Oncol Biol Phys 54:427–435. Harris, R., and K. N. Lohr. (2002). “Screening for prostate cancer: an update of the evidence for the U.S. Preventive Services Task Force.” Ann Intern Med 137:917–929. Holm, H. H., N. Juul, J. F. Pedersen, H. Hansen, and I. Strøyer. (1983). “Transperineal 125iodine seed implantation in prostatic cancer guided by transrectal ultrasonography.” J Urol 130:283–286. Holmberg, L., A. Bill-Axelson, F. Helgesen, J. O. Salo, P. Folmerz, M. Haggman, S. O. Andersson, A. Spangberg, C. Busch, S. Nordling, J. Palmgren, H. O. Adami, J. E. Johansson, and B. J. Norlen. (2002). “A randomized trial comparing radical prostatectomy with watchful waiting in early prostate cancer.” N Engl J Med 347:781–789. Horwitz, E. M., H. D. Thames, D. A. Kuban, L. B. Levy, P. A. Kupelian, A. A. Martinez, J. M. Michalski, T. M. Pisansky, H. M. Sandler, W. U. Shipley, M. J. Zelefsky, G. E. Hanks, and A. L. Zietman. (2005). “Definitions of biochemical failure that best predict clinical failure in patients with prostate cancer treated with external beam radiation alone: a multi-institutional pooled analysis.” J Urol 173:797–802. Howard, A., K. Wallner, and B. Han. (2001). “Clinical course and dosimetry of rectal fistulas after prostate brachytherapy.” J Brachytherapy Int 17:37–42. Hughes, S., K. Wallner, G. Merrick, S. Miller, and L. D. True. (2001). “Preexisting histologic evidence of prostatitis is unrelated to postimplant urinary morbidity.” Int J Cancer 96 Suppl:79–82.
554
Wayne M. Butler and Gregory S. Merrick
Hull, G. W., F. Rabbani, F. Abbas, T. M. Wheeler, M. W. Kattan, and P. T. Scardino. (2002). “Cancer control with radical prostatectomy alone in 1,000 consecutive patients.” J Urol 167:528–534. Jemal, A., T. Murray, E. Ward, A. Samuels, R. C. Tiwari, A. Ghafoor, E. J. Feuer, and M. J. Thun. (2005). “Cancer statistics, 2005.” CA Cancer J Clin 55:10–30. Jhaveri, F. M., C. D. Zippe, E. A. Klein, and P. A. Kupelian. (1999). “Biochemical failure does not predict overall survival after radical prostatectomy for localized prostate cancer: 10-year results.” Urol 54:884–890. Kattan, M. W. (2003). “Nomograms are superior to staging and risk grouping systems for identifying high-risk patients: preoperative application in prostate cancer.” Curr Opin Urol 13:111–116. Kiteley, R. A., W. R. Lee, A. F. deGuzman, M. Mirzaei, and D. L. McCullough. (2002). “Radiation dose to the neurovascular bundles or penile bulb does not predict erectile dysfunction after prostate brachytherapy.” Brachytherapy 1:90–94. Kundu, S. D., R. L. Grubb, K. A. Roehl, J. A. Antenor, M. Han, and W. J. Catalona. (2005). “Delays in cancer detection using 2 and 4-year screening intervals for prostate cancer screening with initial prostate specific antigen less than 2 ng/ml.” J Urol 173:1116–1120. Kupelian, P. A., J. C. Buchsbaum, C. Patel, M. Elshaikh, C. A. Reddy, C. Zippe, and E. A. Klein. (2002). “Impact of biochemical failure on overall survival after radiation therapy for localized prostate cancer in the PSA era.” Int J Radiat Oncol Biol Phys 52:704–711. Laumann, E. O., A. Paik, and R. C. Rosen. (1999). “Sexual dysfunction in the United States: prevalence and predictions.” JAMA 281:537–544. Lee, N., R. Fawaaz, C. A. Olsson, M. C. Benson, D. P. Petrylak, P. B. Schiff, E. Bagiella, A. Singh, R. D. Ennis. (2000). “Which patients with newly diagnosed prostate cancer need a radionuclide bone scan? An analysis based on 631 patients.” Int J Radiat Oncol Biol Phys 48:1443–1446. Lehrer, S., N. N. Stone, and R. G. Stock. (2005). “Prostate cancer in a large prostate is associated with a decreased prostate specific antigen failure rate after brachytherapy.” J Urol 173:79–81. Leibovich, B. C., M. L. Blute, D. G. Bostwick, T. M. Wilson, T. M. Pisansky, B. J. Davis, D. M. Ramnani, L. Cheng, T. J. Sebo, and H. Zincke. (2000). “Proximity of prostate cancer to the urethra: Implications for minimally invasive ablative therapies.” Urol 56:726–729. Litwin, M. S., D. P. Lubeck, J. M. Henning, and P. R. Carroll. (1998). “Differences in urologist and patient assessments of health related quality of life in men with prostate cancer: Results of the CaPSURE database.” J Urol 159:1988–1992. Merrick, G. S., W. M. Butler, W. H. Farthing, A. T. Dorsey, and E. Adamovich. (1998) “The impact of Gleason score accuracy as a criterion for patient selection in prostate brachytherapy.” J Brachytherapy Int 14:113–121. Merrick, G. S., W. M. Butler, J. H. Lief, R. L. Stipetich, L. J. Abel, and A. T. Dorsey. (1999). “Efficacy of sildenafil citrate in prostate brachytherapy patients with erectile dysfunction.” Urol 53:1112–1116. Merrick, G. S., W. M. Butler, J. H. Lief, and A. T. Dorsey. (2000). “Temporal resolution of urinary morbidity following prostate brachytherapy.” Int J Radiat Oncol Biol Phys 47:121–128. Merrick, G. S., W. M. Butler, A. T. Dorsey, and J. T. Dorsey III. (2000b). “The effect of constipation on rectal dosimetry following prostate brachytherapy.” Med Dosim 25:237–241. Merrick, G. S., W. M. Butler, R. W. Galbreath, J. H. Lief, and J. G. Donzella. (2000c). “A comparison of radiation dose to the neurovascular bundles in men with and without prostate brachytherapy induced erectile dysfunction.” Int J Radiat Oncol Biol Phys 46:1069–1074. Merrick, G. S., W. M. Butler, R. W. Galbreath, J. H. Lief, and E. Adamovich. (2001). “Perineural invasion is not predictive of biochemical outcome following prostate brachytherapy.” Cancer J 7:404–412. Merrick, G. S., W. M. Butler, J. H. Lief, and R. W. Galbreath. (2001a). “Five-year biochemical outcome after prostate brachytherapy for hormone-naive men < or = 62 years of age.” Int J Radiat Oncol Biol Phys 50:1253–1257. Merrick, G. S., K. E. Wallner, W. M. Butler, R. W. Galbreath, J. H. Lief, M. J. Benson. (2001b). “A comparison of radiation dose to the bulb of the penis in men with and without prostate brachytherapy induced erectile dysfunction.” Int J Radiat Oncol Biol Phys 50:597–604. Merrick, G. S., K. E. Wallner, W. M. Butler, J. H. Lief, and R. W. Galbreath. (2002). “Prophylactic versus therapeutic alpha-blockers in the management of urinary morbidity following permanent prostate brachytherapy.” Urol 60:650–655.
28–Introduction to Prostate Brachytherapy
555
Merrick, G. S., W. M. Butler, K. Wallner, R. W. Galbreath, R. L. Anderson, B. S. Kurko, and J. H. Lief. (2002a). “Permanent prostate brachytherapy-induced morbidity in patients with grade II and III obesity.” Urol60:104–108. Merrick, G. S., W. M. Butler, B. G. Tollenaar, J. H. Lief, R. L. Anderson, B. J. Smeiles, R. W. Galbreath, M. J. Benson. (2002b). “The dosimetry of prostate brachytherapy-induced urethral strictures.” Int J Radiat Oncol Biol Phys 52:461–468. Merrick, G. S., W. M. Butler, R. W. Galbreath, R. L. Stipetich, L. J. Abel, and J. H. Lief. (2002c). “Erectile function after permanent prostate brachytherapy. Int J Radiat Oncol Biol Phys 52:893–902. Merrick, G. S., W. M. Butler, K. E. Wallner, J. H. Lief, R. L. Anderson, B. J. Smeiles, R. W. Galbreath, and M. L. Benson. (2002d). “The importance of radiation doses to the penile bulb vs. crura in the development of post brachytherapy erectile dysfunction.” Int J Radiat Oncol Biol Phys 54:1055–1062. Merrick, G. S., W. M. Butler, K. E. Wallner, J. H. Lief, A. Mulroy, and R. W. Galbreath. (2003). “Prostate specific antigen (PSA) velocity and benign prostate hypertrophy predict for PSA spikes following prostate brachytherapy.” Brachytherapy 2:181–188. Merrick, G. S., W. M. Butler, K. E. Wallner, R. W. Galbreath, B. C. Murray, D. J. Zeroski, and J. H. Lief. (2003a). “Dysuria following permanent prostate brachytherapy.” Int J Radiat Oncol Biol Phys 55:979–985. Merrick, G. S., W. M. Butler, K. E. Wallner, R. W. Galbreath, B. Kurko, and S. Cleavinger. (2003b). “Rectal function following brachytherapy with or without supplemental external beam radiation: Results of two prospective randomized trials.” Brachytherapy 2:147–157. Merrick, G. S., W. M. Butler, K. E. Wallner, A. L. Hines, and Z. Allen. (2003c). “Late rectal function after prostate brachytherapy.” Int J Radiat Oncol Biol Phys 57:42–48. Merrick, G. S., W. M. Butler, K. E. Wallner, R. W. Galbreath, J. H. Lief, and E. Adamovich. (2004). “Prognostic significance of percent positive biopsies in clinically organ-confined prostate cancer treated with permanent prostate brachytherapy with or without supplemental external-beam radiation.” Cancer J 10:54–60. Merrick, G. S., W. M. Butler, K. E. Wallner, R. W. Galbreath, and E. Adamovich. (2004a). “Permanent interstitial brachytherapy for clinically organ-confined high-grade prostate cancer with a pretreatment PSA < 20 ng/mL.” Am J Clin Oncol 27:611–615. Merrick, G. S., W. M. Butler, K. E. Wallner, R. W. Galbreath, Z. Allen, J. H. Lief, and E. Adamovich. (2005). “Influence of body mass index on biochemical outcome after permanent prostate brachytherapy.” Urol 65:95–100. Merrick, G. S., W. M. Butler, K. E. Wallner, R. W. Galbreath, J. H. Lief, Z. Allen, and E. Adamovich. (2005a). “Impact of supplemental external beam radiotherapy and/or androgen deprivation therapy on biochemical outcome after permanent prostate brachytherapy.” Int J Radiat Oncol Biol Phys 61:32–43. Merrick, G. S., W. M. Butler, K. E. Wallner, Z. Allen, J. L. DeFilippo, and E. Adamovich. (2005b). “The role of enzymatic prostatic acid phosphatase in the clinical staging evaluation of patients with newly diagnosed, untreated prostate cancer.” W V Med J. In press. Merrick, G. S., W. M. Butler, K. E. Wallner, Z. A. Allen, R. W. Galbreath, and J. H. Lief. (2005c). “Brachytherapyrelated dysuria.” BJU Int 95:597–602. Merrick, G. S., W. M. Butler, K. E. Wallner, R. W. Galbreath, R. L. Anderson, and Z. A. Allen. (2005d). “Prostate brachytherapy-induced urethral strictures.” Submitted Merrick, G. S., W. M. Butler, K. E. Wallner, R. W. Galbreath, R. L. Anderson, B. S. Kurko, J. H. Lief, Z. A. Allen. (2005e). “Erectile function after prostate brachytherapy.” Int J Radiat Oncol Biol Phys. In press. Merrick, G. S., K. E. Wallner, and W. M. Butler. (2003). “Minimizing prostate brachytherapy-related morbidity.” Urol 62:786–792. Moran, B. J., M. A. Stutz, and M. H. Gurel. (2004). “Prostate brachytherapy can be performed in selected patients after transurethral resection of the prostate.” Int J Radiat Oncol Biol Phys 59:392–396. Mueller, A., K. Wallner, G. Merrick, E. Ford, S. Sutlief, W. Cavanagh, and W. Butler. (2004). Perirectal seeds as a risk factor for prostate brachytherapy-related rectal bleeding.” Int J Radiat Oncol Biol Phys 59:1047–1052. Mulhall, J. P., P. Yonover, A. Sethi, G. Yasuda, and N. Mohideen. (2002). “Radiation exposure to the corporeal bodies during 3-dimensional conformal radiation therapy for prostate cancer.” J Urol 167:539–542. Niehaus, A., G. S. Merrick, W. M. Butler, K. E. Wallner, Z. A. Allen, R. W. Galbreath, and E. Adamovich. (2005). “Relationship between isotope, prostate volume and urinary morbidity following prostate brachytherapy.” Int J Radiat Oncol Biol Phys. Submitted.
556
Wayne M. Butler and Gregory S. Merrick
Partin, A. W., L. A. Mangold, D. M. Lamm, P. C. Walsh, J. I. Epstein, and J. D. Pearson. (2001). “Contemporary update of prostate cancer staging nomograms (Partin Tables) for the new millennium.” Urol 58:843–848. Pickett, B., R. K. Ten Haken, J. Kurhanewicz, A. Qayyum, K. Shinohara, B. Fein, and M. Roach. (2004). “Time to metabolic atrophy after permanent prostate seed implantation based on magnetic resonance spectroscopic imaging.” Int J Radiat Oncol Biol Phys 59:665–673. Pollack, A., A. L. Hanlon, B. Movsas, G. E. Hanks, R. Uzzo, and E. M. Horwitz. (2003). “Biochemical failure as a determinant of distant metastasis and death in prostate cancer treated with radiotherapy.” Int J Radiat Oncol Biol Phys 57:19–23. Ponsky, L. E., E. E. Cherullo, R. Starkey, D. Nelson, D. Neumann, and C. D. Zippe. (2002). “Evaluation of preoperative ProstaScint scans in the prediction of nodal disease.” Prostate Cancer Prostatic Dis 5:132–135. Potters, L., T. Torre, P. A. Fearn, S. A. Leibel, and M. W. Kattan. (2001). “Potency after permanent prostate brachytherapy for localized prostate cancer.” Int J Radiat Oncol Biol Phys 50:1235–1242. Pound, C. R., A. W. Partin, M. A. Eisenberger, D. W. Chan, J. D. Pearson, and P. C. Walsh. (1999). “Natural history of progression after PSA elevation following radical prostatectomy.” JAMA 281:1591–1597. Roach, M., 3rd, A. Chen, J. Song, A. Diaz, J. Presti, Jr., and P. Carroll. (2000). “Pretreatment prostate-specific antigen and Gleason score predict the risk of extracapsular extension and the risk of failure following radiotherapy in patients with clinically localized prostate cancer.” Semin Urol Oncol 18:108–114. Sharkey, J., A. Cantor, Z. Solc, W. Huff, S. D. Chovnick, R. J. Behar, R. Perez, J. Otheguy, and R. Rabinowitz. (2005). “(103)Pd brachytherapy versus radical prostatectomy in patients with clinically localized prostate cancer: A 12-year experience from a single group practice.” Brachytherapy 4:34–44. Snyder, K. M., R. G. Stock, S. M. Hong, Y. C. Lo, and N. N. Stone. (2001). “Defining the risk of developing grade 2 proctitis following 125I prostate brachytherapy using a rectal dose-volume histogram analysis.” Int J Radiat Oncol Biol Phys 50:335–341. Stamey, T. A., M. Caldwell, J. E. McNeal, R. Nolley, M. Hemenez, and J. Downs. (2004). “The prostate specific antigen era in the United States is over for prostate cancer: what happened in the last 20 years?” J Urol 172:1297–1301. Stock, R. G., J. Kao, and N. N. Stone. (2001). “Penile erectile function after permanent radioactive seed implantation for treatment of prostate cancer.” J Urol 165:436–439. Stock, R. G., N. N. Stone, and J. A. Cesaretti. (2003). “Prostate-specific antigen bounce after prostate seed implantation for localized prostate cancer: descriptions and implications.” Int J Radiat Oncol Biol Phys 56:448–453. Stone, N. N., R. G. Stock, and P. Unger. (2005). “Intermediate term biochemical-free progression and local control following 125iodine brachytherapy for prostate cancer.” J Urol 173 803–807. Talcott, J. A., J. A. Clark, P. C. Stark, and S. P. Mitchell. (2001). “Long-term treatment related complications of brachytherapy for early prostate cancer: A survey of patients previously treated.” J Urol 166:494–499. Terk, M. D., R. G. Stock, and N. N. Stone. (1998). “Identification of patients at increased risk for prolonged urinary retention following radioactive seed implantation of the prostate.” J Urol 160:1379–1382. Thompson, I. M., P. J. Goodman, C. M. Tangen, M. S. Lucia, G. J. Miller, L. G. Ford, M. M. Lieber, R. D. Cespedes, J. N. Atkins, S. M. Lippman, S. M. Carlin, A. Ryan, C. M. Szczepanek, J. J. Crowley, and C. A. Coltman Jr. (2003). “The influence of finasteride on the development of prostate cancer.” N Engl J Med 349:215–224. Valicenti, R., J. Lu, M. Pilepich, S. Asbell, and D. Grignon. (2000). “Survival advantage from higher-dose radiation therapy for clinically localized prostate cancer treated on the Radiation Therapy Oncology Group trials.” J Clin Oncol 18:2740–2746. Wallner, K. (2002). “MR imaging for prostate cancer staging: Beauty or beast?” Int J Radiat Oncol Biol Phys 52:886–886. Wallner, K., J. Roy, and L. Harrison. (1995). “Dosimetry guidelines to minimize urethral and rectal morbidity following transperineal I-125 prostate brachytherapy.” Int J Radiat Oncol Biol Phys 32:465–471. Wallner, K. E., J. Roy, and L. Harrison. (1996). “Tumor control and morbidity following transperineal iodine 125 implantation for stage T1/T2 prostatic carcinoma.” J Clin Oncol 14:449–453. Wallner, K., G. Merrick, L. True, W. Cavanagh, C. Simpson, and W. Butler. (2002). “I-125 versus Pd-103 for low risk prostate cancer: Morbidity outcomes from a prospective randomized multicenter trial.” Cancer J 8:67–73. Wallner, K. E., G. S. Merrick, M. L. Benson, W. M. Butler, J. Maki, B. G. Tollenaar. (2002a). “Penile bulb imaging.” Int J Radiat Oncol Biol Phys 53:928–933.
28–Introduction to Prostate Brachytherapy
557
Waterman, F. M., and A. P. Dicker. (1999). “Effect of post-implant edema on rectal dose in prostate brachytherapy.” Int J Radiat Oncol Biol Phys 45:571–576. Whitmore, W. F., Jr. (1990). “Natural history of low-stage prostatic cancer and the impact of early detection.” Urol Clin North Am 17:689–697. Williams, S. G. (2004). “Ambiguities within the ASTRO consensus definition of biochemical failure: never assume all is equal.” Int J Radiat Oncol Biol Phys 58:1083–1092. Wilson, S. S., and E. D. Crawford. (2004). “Screening for prostate cancer.” Clin Prostate Cancer 3:21–25. Zietman, A. L., E. M. Zehr, and W. U. Shipley. (1999). “The long-term effect on PSA values of incidental prostatic irradiation in patients with pelvic malignancies other than prostate cancer.” nt J Radiat Oncol Biol Phys 43:715–718. Zelefsky, M. J., and J. F. Eid. (1998). “Elucidating the etiology of erectile dysfunction after definitive therapy for prostate cancer.” Int J Radiat Oncol Biol Phys 40:129–133. Zelefsky, M. J., and W. F. Whitmore Jr. (1997). “Long-term results of retropubic permanent 125iodine implantation of the prostate for clinically localized prostatic cancer.” J Urol 158:23–29. Zelefsky, M. J., K. E. Wallner, C. C. Ling, A. Raben, T. Hollister, T. Wolfe, A. Grann, P. Gaudin, Z. Fuks, and S. A. Leibel. (1999). “Comparison of the 5-year outcome and morbidity of three-dimensional conformal radiotherapy versus transperineal permanent iodine-125 implantation for early stage prostate cancer.” J Clin Oncol 17:517–522. Zelefsky, M. J., Z. Fuks, M. Hunt,Y. Yamada, C. Marion, C. C. Ling, H. Amols, E. S. Venkatraman, and S. A. Leibel. (2002). “High-dose intensity modulated radiation therapy for prostate cancer: early toxicity and biochemical outcome in 772 patients.” Int J Radiat Oncol Biol Phys 53:1111–1116.
Chapter 29
Permanent Prostate Brachytherapy Treatment Planning Wayne M. Butler, Ph.D. and Gregory S. Merrick, M.D. Schiffler Cancer Center Wheeling Hospital Wheeling, West Virginia Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559 Target Volume Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560 Imaging Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560 What Is the Target Volume? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560 Prostate Only—or Less . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 562 Planning with a Dosimetric Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 562 Explicit Planning Target Volume (PTV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563 Should the Seminal Vesicles Be Included? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563 Clinical Necessity for Dosimetric Margins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564 How Much Margin Is Necessary? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564 Creating the PTV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 564 Symmetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565 Planning Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565 Pre-OR Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567 Intraoperative and Interactive Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567 Nomogram Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567 Computer Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 568 Dynamic Dose Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 568 Needle and Seed Placement Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569 General Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569 Uniform Loading: Idealism vs. Reality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570 Manual Loading Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570 Modified Uniform Loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572 How to Cover the Apex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572 Addition of Peripheral Needles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572 Removal of Selected Central Seeds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 572 A Sample Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573 Peripheral Loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576 Computer-Assisted and Optimized Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 577 Convergence of Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578 Choice of Radionuclide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578 Appropriate Seed Strengths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 578 Preplan Evaluation Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579 Review of Contemporary Planning Variations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 581 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 585
Introduction The ability to learn and plan is part of human nature, which is a good thing because we certainly have no hard-wired instinct for constructing a brachytherapy implant. What does come in handy in designing permanent prostate brachytherapy implants is our built-in aesthetic sensibility of symmetry, balance, and proportion. Effective treatment planning requires an understanding not only of the characteristics of various brachytherapy sources, but also an understanding of the disease being treated.
559
560
Wayne M. Butler and Gregory S. Merrick
Target Volume Definition Identification of the target volume in prostate brachytherapy initiates the treatment-planning process. Because physicists and dosimetrists rarely have expertise in urogenital anatomy, the radiation oncologist is usually tasked with target delineation; but the task is sometimes delegated to the physics staff. The physics staff may not be charged with outlining the prostate, but they are usually expected to define the planning target volume (PTV). The expansion of the clinical target volume (CTV) to the PTV, whether performed explicitly or implicitly, should be performed initially with guidance from the radiation oncologist and always with insight.
Imaging Modalities Prostate brachytherapy would probably not have become ascendant were it not for the relatively low cost, simplicity, and utility of transrectal ultrasound. Ultrasound systems usually have several ways to export images for treatment planning: hard copy prints which may be digitized or electronic transfer via video tape, stored sets of bitmap or jpeg images, or network DICOM transfer. Quality control checks and measurements should be periodically performed on the ultrasound system to identify image quality degradation before it affects patient scans (Goodsitt et al., 1998). Because the needle template grid is projected onto transverse ultrasound images, proper scaling and geometric accuracy of the projected grid should also be checked by phantom measurements. Magnetic resonance imaging (MRI) provides excellent visualization and is used instead of ultrasound at a few centers (Cormack et al., 2000), but the additional implant time under anesthesia and expense has not been justified by improved patient dosimetry or biochemical control and quality of life outcomes compared with centers performing conventional ultrasound based implants. Computerized tomographic (CT) imaging has also been used for planning and design of customized templates (Roy et al., 1991). Prostate visualization with CT is inferior to that achieved by ultrasound, and the approach has fallen into disuse.
What Is the Target Volume? Most commonly, a transrectal ultrasound volumetric study of the prostate gland is obtained in the ultrasound suite, preferably in the radiation oncology department, prior to the operative procedure. The preimplant volume study should not be performed on patients undergoing hormonal therapy unless the initiation of androgen deprivation therapy was at least 2 months prior. Such agents are cytoreductive, and the initial prostate volume, V0, will be diminished by the number of weeks of androgen deprivation therapy. For seed ordering purposes, we use an empirically derived formula to predict the volume change from combined use of a luteinizing hormone-releasing hormone (LH-RH) agonist and an antiandrogen approximately as: Vfinal =
weeks 7 1 + 2 2
V0
(1)
Equation (1) is plotted in Figure 1. There will be a 30% volume reduction of the prostate after 2 months and about a 40% reduction after 3 months. Implanting a patient during the early course of hormonal therapy will result in a considerable dose escalation as the seeds draw closer together. It is imperative that each patient and his referring physicians be queried regarding the use of hormones. The volume study is acquired with the patient in the dorsal lithotomy position using a transrectal ultrasound probe mounted securely to a stepper and stabilizer unit (Figure 2). The patient is scanned at 3 to 5
29–Permanent Prostate Brachytherapy Treatment Planning
561
Figure 1. Plot of relative prostate volume as a function of the duration of hormonal therapy. Avodart® is an example of a class of drugs called 5-alpha-reductases which block the conversion of testosterone to its active form, dihydroxytestosterone. Lupron® is a luteinizing hormone-releasing hormone agonist that denies the testicles the hormone necessary to produce testosterone. For total androgen blockade, these drugs are usually combined with an antiandrogen such as Casodex®. Androgen deprivation therapy, if used at all, should be initiated at least 2 months before a permanent prostate implant, and if not used prior to implant, should be delayed after the implant for at least 2 half-lives of the radionuclide because of the inevitable dose escalation in a shrinking prostate.
Figure 2. A floor-stand stabilizer with an adjustable stepper unit attached. The ultrasound probe and needle template are attached to the stepper. Virtually all stepper/stabilizer units allow independent 6-axis probe movement. Table-mounted stabilizers are also in common use, but they are not as rigid as a floor-mounted unit. The floor stand allows the apparatus to be easily wheeled out of the way for post implant cystoscopy and reproducible reinsertion of the probe for any last minute changes. (Illustration courtesy of CMS, Inc.)
562
Wayne M. Butler and Gregory S. Merrick
mm intervals starting from the proximal seminal vesicles just superior to the bladder neck/base of the prostate gland and extending to the apex. The vertical midline of the prostate should be in the center of the transverse field from base to apex. The posterior border of the prostate should be under sufficient pressure from the ultrasound probe so that it is flat at midgland, neither concave up nor down, and the border should be positioned at a reference row of template markers or no more than 2 mm below that row. The importance of consistency in volume studies cannot be overstressed. To ensure that no slices were omitted or duplicated, a sagittal image of the prostate should be recorded, and the measuring tools available in the ultrasound software used to determine the distance from the base to the apex. That distance should correspond to the separation between the transverse base and apex images. The path of the urethra should be visualized on each ultrasound slice by either infusing an aerated gel via the penile urethra or by placing a urinary catheter prior to the procedure. It is very important to ensure that the catheter bulb is well within the bladder and that there is no tension applied to the catheter in order to minimize catheter distortion of the prostate. Other structures frequently contoured are the rectum and bladder. Long-term serious brachytherapy-induced urinary and rectal morbidity occur rarely (~1%) in good programs, and that portion of morbidity attributable to high radiation dose can be reduced to very rare status by careful planning and execution. Prostate Only—or Less In institutions where radical prostatectomy is considered the “gold standard” or the prevailing treatment paradigm, the target volume is frequently defined as the prostate only with minimal margins in analogy with surgery. If the goal is to duplicate the surgical experience, then the brachytherapy results will be no better than those reported for surgical series. As illustrated in the previous chapter, the best-reported biochemical survival results following surgery are good, but not as good as the best brachytherapy results. The centers reporting the best brachytherapy results almost all treat with dosimetric margins, and one study reported that a dosimetric margin >3 mm was almost as good a predictor of freedom from biochemical failure as the widely used minimum dose covering 90% of the prostate, D90 (Choi et al., 2004). With currently available technology, an even riskier approach is to attempt the brachytherapy equivalent of a lumpectomy. Using the location of positive biopsies supplemented by tumor sensitive imaging such as MRI spectroscopy or immunoscintigraphy and single-photon emission computed tomography, only cancer foci or prostate zonal anatomy, typically the peripheral zone, are targeted. A sextant needle biopsy sampling, with each core 1 mm in diameter and 2 cm long, samples <0.05 cm? of the prostate. That can hardly be considered representative of the average 35 cm? prostate. No prostate tumor sensitive imaging modality can detect less than a few mm? of cancer, yet most prostate cancers are microscopically multifocal. Prostate lumpectomy via surgery or radiation may be the next phase of therapy, but that advance will require better technology. Planning with a Dosimetric Margin One way to add a treatment margin to the prostate is to plan the implant so that the source positions and source strength are selected to maintain the 100% isodose line at the desired distance from the prostate. The shortcoming of this approach is that contour of the prescribed dose is lumpy in three-dimensions and scalloped in two-dimensions. Some treatment-planning systems are capable of determining the margin on each transverse slice at any angle around the center of gravity. By documenting these values, the planner may then modify the implant so that the margin to the prescribed isodose meets some specified minimum or average value. Margin calculation technology is not yet widely available, so practitioners of dosimetric margining are forced to rely on crude manual measurements or subjective estimates. This inevitably leads to inconsistencies in the interpretation and application of dosimetric margins.
29–Permanent Prostate Brachytherapy Treatment Planning
563
Explicit Planning Target Volume (PTV) Creating an explicitly drawn PTV clearly documents the volumetric planning goals and allows detailed dose calculations for that structure. By requiring the prescription isodose to completely cover the PTV, the dosimetric margin everywhere will be at least as great as the chosen physical margin. In a recent survey of preimplant practices at 8 institutions that regularly perform monotherapy and combined modality brachytherapy with both 125I and 103Pd, Merrick et al. (2005a) found that 6 of 8 planned with an explicit PTV. Should the Seminal Vesicles be Included? Classifying patients into risk groups essentially stratifies them according to the likelihood of extracapsular extension, seminal vesicle involvement, and lymph node involvement. Low risk men have a negligible risk of seminal vesicle involvement and they are those whose primary clinical parameters are prostate specific antigen (PSA) ≤10 ng/mL, Gleason score <7, and stage ≤T2b. Intermediate risk men are categorized by a single unfavorable parameter: PSA > 10 ng/mL, Gleason score ≥7, or clinical stage ≥T2b. Their risk of seminal vesicle involvement is >2% (Partin et al., 2001). For high-risk patients with two or more unfavorable parameters, the risk of seminal vesicle involvement is >6%. In radical prostatectomy series where the seminal vesicles are taken, seminal vesicle disease extension is often confined to the first centimeter proximal to the prostate (Kestin et al., 2002). For this reason, patients treated with externalbeam radiation therapy (XRT) adjunctive to brachytherapy have their seminal vesicles included in the treatment portals. Some brachytherapists go further and explicitly place seeds in the vesicles as illustrated in Figure 3. However, it is less common that the seminal vesicles are included in either the planning or post-implant PTV.
Figure 3. Transverse ultrasound image of the seminal vesicles with four implanted seeds visible (arrow). Typically, four needles containing two seeds each, 1 cm apart, are used to implant the proximal seminal vesicles of patients treated at the Schiffler Cancer Center.
564
Wayne M. Butler and Gregory S. Merrick
Clinical Necessity for Dosimetric Margins In patients with a pretreatment PSA < 10 ng/mL, approximately 50% manifest extracapsular disease at the time of radical prostatectomy (Partin et al., 2001). Increasing PSA, Gleason score, and clinical stage are markers for an increased likelihood that the disease is not organ-confined. Many surgical series have noted an adverse effect on biochemical progression free survival due to additional factors that are also surrogates for increased likelihood and extent of extracapsular disease: increasing percent positive biopsies, perineural invasion, and a 4+3 pattern in Gleason score 7 rather than a 3+4 pattern. Unlike urologic surgeons, brachytherapists are not constrained in their efforts to treat the periprostatic region. In brachytherapy implants designed to eradicate extraprostatic disease, adverse factors are of much less consequence (Merrick et al., 2001, 2002, 2004). It is our contention that treating with adequate dosimetric margins results in the superior survival results seen in the best brachytherapy series compared with the best radical prostatectomy series. How Much Margin Is Necessary? The mean extent of extraprostatic extension in radical prostatectomy specimens has been reported by the Mayo Clinic (Davis et al., 1999) and the Cleveland Clinic (Sohayda et al., 2000) to be 0.5 mm and 1.1 mm, respectively. The maximum extracapsular extension at the Mayo and Cleveland Clinics was 4.4 mm and 10.0 mm, respectively. Both clinics concluded that brachytherapy margins of 5.0 mm would encompass about 99% of cases amenable to local control. One caveat to this conclusion is that the conventional H&E (haematoxylin-eosin) staining used in these studies may not completely define the radial extraprostatic extension. The determination of molecular margins after radical prostatectomy by reverse transcriptase-polymerase chain reaction for prostate-specific antigen has questioned the validity of extracapsular extension and surgical margin assessment by standard staining (Theodorescu, Frierson, and Sikes 1999; Straub et al., 2001). However, even if H&E staining does underestimate the extent of extracapsular disease, it is likely that in contemporary patient populations disease extending beyond 5 mm still occurs rarely. Capsular penetration beyond 5 mm was an extremely significant predictor of failure in univariate analysis of a prostatectomy series and was so strongly correlated with lymph node involvement that it was not an independent predictor in multivariate failure analysis (Stamey et al., 2000). Figure 4 displays biochemical PSA progression-free survival curves for implants with and without margin and with and without XRT. Treating high-risk prostate cancer with brachytherapy combined with XRT is considered standard of care, because omitting XRT will reduce the 8-year biochemical survival from about 85% down to 65%. Omitting XRT and treating less than the entire prostate and the likely extracapsular disease extension is profoundly detrimental—all the patients (n = 19) in the D’Amico study (D’Amico et al., 1998) failed within 3 years. The other data in the survival figure are from Dattoli (2005), Merrick et al., (2005), and Sylvester et al., (2003).
Creating the PTV The auto-margin tools available in modern treatment planning software are sufficient to create the PTV from the defined prostate. Such tools allow specification of separate margins for at least the 4 Cardinal directions in the transverse plane as well as superior and inferior margins. If the superior and inferior margins are set equal to the slice separation, one can duplicate the shape of the prostate base on a superior slice and duplicate the prostate apex on an inferior slice. All other intervening slices will have a PTV at least as large as the prostate on any adjacent slice and will be further expanded by the chosen transverse margins in areas where the prostate on the current slice is larger than the prostate on adjacent slices. By choosing the superior and inferior margins to be slightly greater than the slice separation, the PTV on each slice will be at least equal to the size of the prostate on any adjacent slice with a slight additional
29–Permanent Prostate Brachytherapy Treatment Planning
565
Figure 4. Biochemical PSA progression-free survival in high-risk patients stratified by the use of planning margins and external beam radiation therapy. Data are from Merrick et al. (2005), Dattoli (2005), Blasko (Sylvester et al., 2003), and D’Amico et al. (1998).
margin. Margin settings of 4 to 6 mm laterally and anteriorly, 0 mm posteriorly, and 6 to 7 mm superiorly and inferiorly, work well with our patients. Planning with margin has been shown to result in similar margins on post implant analysis, as shown in Figure 5 (Merrick et al., 2003). At the base, and to a much lesser degree at the apex, the posterior border of the prostate is separated from the rectum. Posterior margin may be added to those slices, but the margin should not extend below the lowest implant row. The closest approach of the rectum to the prostate is near the apex, and almost all brachytherapy-induced as well as radical prostatectomy induced rectal injuries occur at the apex. Symmetry One deficiency of auto-margin brachytherapy software is the inability to create a more generous margin at either the base or the apical regions as per the brachytherapist’s request. For this, manual intervention is needed. Another problem is how to cope with asymmetry in the prostate volume study. The prostate of the anesthetized patient in the operating room can usually be aligned so that it is bilaterally symmetric. Patient movement or spasms often cause the asymmetry observed in the volume study obtained in the clinic. In anticipation of the likely situation in the operating room and to make treatment planning, needle loading, and plan execution simpler and more robust, the PTV should be manually adjusted to make it bilaterally symmetric.
Planning Types Table 1 lists the type of planning applicable to prostate brachytherapy as defined by the American Brachytherapy Society (ABS) (Nag et al., 2001). It is likely that most centers will continue to plan permanent seed implants before entering the operating room (OR) even if technology advances to the ultimate level of dynamic dose calculation and most centers adopt it. The intervening types are currently available technologies for intraoperative planning.
566
Wayne M. Butler and Gregory S. Merrick
Figure 5. Sagittal views of the dosimetric margin from the prostate (shaded interior) to the 100% isodose line (solid curve), the 90% isodose line (dashed curve), and the 75% isodose line (dotted curve). The plots are a composite of 13 implants each of 125I (a) and 103Pd (b). (Reproduced from Am J Clin Oncol, vol 26, “Extracapsular radiation dose distribution after permanent prostate brachytherapy,” G. S. Merrick, W. M. Butler, K. E. Wallner, L. R. Burden, and J. E. Dougherty, pp. e179–e189. © 2003, with permission from Lippincott Williams and Wilkins.)
29–Permanent Prostate Brachytherapy Treatment Planning
567
Table 1. American Brachytherapy Society (ABS) Nomenclature for Different Types of Planning Used in Permanent Prostate Brachytherapy (Nag et al., 2001). Planning approach
Definition
Preplanning
Creation of a plan outside the operating room (OR) hours, days or weeks before the implant procedure.
Intraoperative
Plan created in the OR. The patient remains stationary between the time of the volume study and the implant procedure.
Interactive
The treatment plan is revised periodically during the implant procedure using image-based feedback of needle position to recalculate dose.
Dynamic dose calculation
Dose distribution continuously updated using deposited seed position feedback.
(Reproduced from Int J Radiat Oncol Biol Phys, vol 51, “Intraoperative planning and evaluation of permanent prostate brachytherapy: Report of the American Brachytherapy Society.” S. Nag, J. P. Ciezki, R. Cormack, S. Doggett, K. DeWyngaert, G. K. Edmundson, R. G. Stock, N. N. Stone, Y. Yu, and M. J. Zelefsky, pp. 1422–1430. © 2001, with permission from Elsevier.)
Pre-OR Planning Among the virtues of planning prior to encountering the patient in the OR is cost effectiveness. With the volume study performed by the brachytherapist no more than a few weeks prior to the implant, there will be negligible change in the prostate, and fewer than 1% of patients exhibit changes from the time of the volume study to the operating room volume capture that necessitate modification of the plan. The plan, therefore, will entail little or no seed waste. Having a plan also saves valuable OR time and allows the use of pre-loaded needles. This planning approach should result in an implant of better dosimetric quality and one that is more resistant to seed loss or misplacement. Taking the time for a thoughtful, insightful consideration of alternatives, optimizing the fit between seed strength and prescribed dose, manually verifying calculations, and second checking assumptions are all worthwhile activities that minimize errors and lead to good outcomes. Mistakes are made most often when established procedures are not followed, and shortcuts are most likely taken when staff is under deadline pressure. Prostate cancer does not require emergency, heroic measures. Foregoing a pre-OR plan treats prostate brachytherapy as emergency medicine, and the attendant time stress on the planning and implant team appears to invite errors. Frequent mismatches between the prostate volume, shape, or position in the pre-OR study and that performed in the OR led some institutions to forego the former for the latter. The anesthetized patient has much less movement artifact, but the root causes of much of the difference seen in studies performed at the two times are differences in technique, apparatus, and patient positioning. It is important that the same individual, whether an ultrasonographer or brachytherapist, set up the patient in both environments and use similar tables and patient leg support in both.
Intraoperative and Interactive Planning Nomogram Approaches Intraoperative prostate brachytherapy planning has a long history, and most early practitioners employed Lowell Anderson’s nomogram (Anderson 1976) or modified versions of it (Stock et al., 1995). In the Stock implementation, the prostate volume was estimated using a urologist’s measurement of length, width, and height (l, w, h, respectively) and determining the ellipsoidal volume from
568
Wayne M. Butler and Gregory S. Merrick
Ve =
π 6
⋅l ⋅w⋅h
(2)
An activity nomogram is then used to determine the total seed strength to bring to the OR. In the OR, an ultrasound planimetry determined volume is used to determine the total activity to implant. Dividing by the available individual seed strength gives the total number of seeds to be implanted. Their nomogram calls for 75% of the total strength to be placed on the periphery of the gland. Stock and colleagues accordingly place needles spaced about 1 cm apart around the periphery of the prostate so that each needle lies about 0.5 cm inside the border of the prostate on the largest mid-gland transverse image. Dividing the number of seeds designated for the periphery by the number of needles placed, the number of seeds per peripheral needle is calculated. The spacing between seeds along each needle track is determined by the base to apex length of the gland with the proviso that the center-to-center spacing be ≤1 cm. Interior needles are placed in a similar fashion. These needles are placed about 1 cm from the periphery and 1 cm from each other. Urethral sparing is achieved by requiring that no needle be closer than 0.5 cm to the urethra. Although users of the nomogram approach to intraoperative planning treat only the prostate rather than a PTV that includes margins, there is no reason why the approach could not be modified to treat the PTV. Auto-margining and calculation of the new planning volume would only take a minute. However, current nomograms make no allowance for the effect of prostate shape on total seed strength needed. Treatment planning dose optimization produces significantly different total strengths when the same volume is shaped as a cylindrical, triangular, or rectangular column. Because nonstandard seed spacing is difficult to implement in the OR, nomogram-based planning almost mandates proficiency with the Mick seed applicator. A recent analysis of the post-operative dosimetry results from practitioners at 35 centers trained by Stock and Stone in their “ProSeed” nomogram approach indicates that 90% of patients implanted with a prescribed 160 Gy monotherapy 125I achieved their day 30 dosimetry goal of D90 ≥ 140 Gy (Stone et al., 2005). Computer Planning All the elements of treatment planning can be moved from the clinic to the OR, but that consumes too much valuable OR time for most, so various hybrid approaches have been developed. Beyer, Shapiro, and Puente (2000) save time by using standard needle positions for all patients. Treatment planning is performed based on those default needle positions while the needles are being placed. The next step up the hierarchy is to evaluate the plan based on the actual needle location. This was first reported by Zelefsky and colleagues at Memorial Sloan-Kettering Cancer Center (Zelefsky et al., 2000; Todor et al., 2003). The positions of the needles on ultrasound were captured by the planning system, and seed placement in the needles was determined by in-house optimization software. Users of commercial software can accomplish this by identifying the needle tips on the transverse ultrasound image or the entire needle on sagittal projections. The planning software then either modifies an existing plan to account for a changed needle position or augments a plan being developed sequentially with each additional needle.
Dynamic Dose Calculation Although an approach similar to needle projection discussed above can be applied to identify and localize seeds as they are extruded from each needle, the process is tedious and time-consuming and does not account for perturbations caused by later needle insertions or seed migration or reorientation (Van Gellekom et al., 2004; ;Rivard, Evans, and Kay 2005). Real-time dynamic dose calculation requires automatic seed identification and registration with prostate imaging. Most of the effort on the seed identification
29–Permanent Prostate Brachytherapy Treatment Planning
569
task has focused on the use of a limited number of x-ray images (Su et al., 2004; Narayanan, Cho, and Marks 2004). The seed coordinates would then be fused to the ultrasound image set. The latter step is proving to be as challenging as the first. Perhaps the breakthrough will come in ultrasound imaging as the sole modality for anatomy and seeds. Trans-urethral ultrasound has been shown to more clearly identify seeds with better spatial resolution and better soft tissue differentiation than trans rectal ultrasound (Holmes et al., 2003).
Needle and Seed Placement Approaches General Guidelines In all treatment planning philosophies, it is considered poor technique to use single seed needles as in Figure 6. If that single seed is intended to address a local dosimetric defect, consider the imaging uncertainties and the seed placement uncertainties. Roberson and colleagues (1997) found the placement uncertainty to be about 3 mm transversely and 5 mm longitudinally. In an implant with 24 needles and 100 seeds, is that single seed worth the trauma of an additional needle? Most immediate post-operative morbidity from prostate brachytherapy is due to needle trauma. As a corollary to our single seed needle advice, the planner should minimize the number of specially loaded needles, i.e., those with other than a standard seed to seed spacing. Increasing the complexity of the plan increases the likelihood of errors not only in loading the needles (or placing seeds via the Mick applicator), but also because the altered loading is designed to modulate the dose in a specific pattern. Consider placement uncertainty above and whether that modulation would be tolerable if shifted cranially or caudally by 5 mm.
Figure 6. Targeting crucial areas with single seeds as in (a) is risky because of seed placement and visualization uncertainties. A more robust approach is shown in (b) where two or three seeds are deployed to cover the areas of interest.
570
Wayne M. Butler and Gregory S. Merrick
As a final general rule, be sparing in your use of back-to-back seeds. These are typically used to build up dose at the prostate base and apex. Double sources create contiguous high dose regions that may place the bladder neck and the bulbomembranous urethra at risk. Incontinence may result if the bladder sphincter is damaged, and urethral strictures may occur if the bulbomembranous urethra is overdosed (Merrick et al., 2002a).
Uniform Loading: Idealism vs. Reality Haakon Ragde, John Blasko, and Peter Grimm began transperineal, ultrasound guided prostate brachytherapy in Seattle, Washington in 1985. They used a uniform seed loading approach in which a relatively large number of low strength seeds were evenly distributed throughout the prostate (Sylvester 2005). An ideal prostate brachytherapy implant would take the form of microscopically dispersed particles of a beta emitting or low-energy photon-emitting radionuclide. Of course this is the same ideal as the radiopharmaceutical industry to find a molecular carrier with exquisite binding specificity to cell membrane molecules that are unique to the target tissue. Alas, the radionuclides used in prostate brachytherapy are encased in macroscopic seeds, and even if they were microscopically dispersible, the emitted photons are not of sufficiently low energy for cumulative, long-range effects to be negligible. In any prostate volume in which seeds are spaced at the corners of a 1 cm cubic grid, the central dose will be much higher than the peripheral dose because of such cumulative effects. Central prostate and urethral doses frequently exceeded 300% of the prescribed dose while the coarse 1 cm cubic grid produced a peripheral scalloping of the 100% isodose that would fail to cover the anterior-lateral curvature of the prostate completely. Unacceptably high urinary morbidity led the Seattle group to abandon uniform loading in favor of a modified version within 2 years after the start of their program. The uniform loading approach is not just an historical curiosity, but forms the basis for most manually planned implants today. Figure 7 illustrates a schematic prostate and typical template markings. The template grid rows and columns are spaced 0.5 cm apart, but the markings are 1 cm apart. The rows are labeled with numbers and the columns with letters. The dotted cross hairs on each image are formed by the D-column and the 3-row. The template columns adjacent to the uppercase letter columns are labeled with lower case letters a…f. The unmarked rows are half-integer numbers, 1.5, 2.5, etc. Manual Loading Algorithm The following paragraphs describe a manual needle and seed placement algorithm that starts with uniform loading, then applies modifications. The modifications create a seed distribution that is urethra sparing and relatively homogeneous—for brachytherapy. Such a uniform dose distribution is very tolerant of systematic seed placement errors (Butler et al., 2000). Because the urethra usually aligns along the D-column, most plans avoid that location. To cover the prostate at mid-gland with a uniform 1 cm square grid, one is obligated to use lowercase columns. The most posterior row implanted in Figure 7 is the 1.5 row, and the most anterior is the 5.5 row. The needles in the figure are placed on the intersection of every lowercase column with every half-integer row that intersects the PTV or is no more than 1 or 2 mm outside the PTV on the largest mid-gland image. Seeds are placed along these valid needle tracks at 1 cm increments from the base (0.0 cm plane) to the apex. If the needle is not within the PTV on the base plane but is within the PTV on the 0.5 cm plane, start the seed train on the base plane. If the needle does not intersect the PTV until the 1 cm plane, start the seed train on either the 1 cm plane or the 0.0 cm plane, and continue placing the seeds at 1 cm increments until they either fall on the last plane of the PTV intersected by the needle. Seeds that are more than 5 mm outside the PTV are omitted.
29–Permanent Prostate Brachytherapy Treatment Planning
571
Figure 7. Schematic of the prostate (thin solid line) and the Wheeling approach to the definition of the planning target volume (PTV, heavy solid line) at the base, midgland and apex of the prostate. The central open circle marks the urethra and the small open or filled circles indicate needle paths and seed positions in a uniform loading approach. The prostate on each slice is expanded 4 to 5 mm to create the PTV. Additional expansion accounts for the size of the prostate on the next larger adjacent slice along with uncertainties in the position of the needle tip and ultrasound transducer plane. Note the bilateral symmetry imposed on the PTV in the midgland slices rather than maintain the skew or asymmetry of the prostate. The PTV is also extended inferior to the apex.
572
Wayne M. Butler and Gregory S. Merrick
Modified Uniform Loading How to Cover the Apex The first modification to the uniform seed loading considers variations in how to treat the apex. If the apical PTV ends at a whole cm increment such as 4.0 cm, then there will be no problem with coverage. Note that the prostate apex of the model in Figure 7 occurs at 40 mm but that the apical portion of the PTV extends to 45 mm from the base and contains no seeds, making it dosimetrically cold. How you correct that deficiency depends on the measured sagittal length of the gland and the concurrence of the radiation oncologist on how to proceed. One solution is to extend the uniform loaded seed trains to the 5.0 cm plane. This is the favored option if the sagittal length of the prostate is nearly 4.5 cm long, implying that the 4.5 cm plane just barely missed seeing prostate. Placing seeds on the 5.0 cm plane, therefore, does not create an excessive inferior margin, and the 5.0 cm seed plane will sandwich the apical PTV between two heavily loaded seed planes to provide coverage. On the model prostate, repeating the 4 ¥ 4 seed pattern of the 4.0 plane on the 5.0 plane would be a good start. The 4 corner seeds of that 16-seed pattern would be unnecessary because the area of the PTV on the 4.5 cm plane is small. The other solution for apical PTV coverage is to select a few of the uniform loading pattern needles for back-to-back seed spacing so that some seeds will lie on the 4.5 cm plane. A pattern using this approach is discussed in detail below. Truncating plans at half integer transverse planes was once rare in our practice but has become very common since our discovery of a strong correlation between radiation dose to the bulbomembranous urethra and urinary strictures (Merrick et al., 2002a) and dose to the proximal penis and patient preservation of erectile function (Merrick et al., 2001a, 2002b). Addition of Peripheral Needles The second modification is to place additional needles along the lateral and anterior periphery on template grid positions that are offset 5 mm in the y direction and 0 or 5 mm horizontally from the main uniformly loaded grid. Usually, six or fewer peripheral needles are sufficient to deliver seeds to cover the anterior and lateral curvature of the prostate completely. As with the basic uniform needle pattern, no peripheral needle should be placed in template grid holes that are more than a couple of millimeters outside the PTV on the largest transverse slice. Unlike the basic uniform needle pattern, no peripheral needle should be placed more than 6 mm inside the PTV. For the schematic prostate of Figure 7, uppercase column and integer row coordinates A3, G3, B5, and F5 are used. Along these needle paths, seeds are placed at 1 cm increments on transverse planes that begin with a 5 mm offset from the base plane. Valid seed positions are no more than 0.6 cm inside or 0.5 cm outside the PTV. Although in Figure 7 only one seed is shown within the A3 and G3 needles (at the 2.5 cm plane), the needle would not be used unless there were valid seed positions on either or both the 1.5 cm and 3.5 cm planes. If the 2.5 cm plane is the only valid seed position, then the needle should be moved inward 0.5 cm to pick up more prostate tissue. Removal of Selected Central Seeds The final modification to uniform seed loading is to reduce the linear seed density in the central needles. Urethral sparing is implicit in this action, but explicit urethral sparing is not necessary if the brachytherapist tracks the path of the urethra in the OR. The number of needles to consider for seed removal and the removal pattern is highly dependent on the seed strength and the radionuclide. Low-strength seeds may require removal of only 1 or 2 seeds from the four needles closest to the urethra. High-strength seeds usually require removal of two or more seeds from about a dozen needles. Due to differences in their radial dose functions, 125I implants are more tolerant of large interior holes while 103Pd implants are less tolerant and need more careful attention to prevent a central low-dose region.
29–Permanent Prostate Brachytherapy Treatment Planning
573
A Sample Plan To illustrate these principles, a plan for a typical prostate, 31 cm3 in this case, is developed and analyzed in Figures 8 through 11. The prostate was first enlarged with 4 to 5 mm margins laterally and anteriorly and 7 mm superiorly and inferiorly to a bilaterally symmetric PTV. Margin was added posteriorly near the base and apex, and the radiation oncologist requested more generous margins at the apex so that the resulting PTV was 61 cm3, an enlargement factor of 1.97, which exceeds our standard enlargement factor of 1.8 for this size prostate. The apical PTV drawn for this patient is shown in Figure 9a based on the measured sagittal length of the gland. Had the prostate been a few mm longer, the alternate ending of Figure 9b would have been employed. The plan was created with uniformly loaded needles on lowercase columns and half-integer rows. On the base (0.0 cm) plane of Figure 8, there is only the PTV, no prostate. Because of the tilt of the prostate/bladder interface, seeds in the two most anterior rows (3.5 and 4.5) would likely end up in the bladder or bladder wall if executed exactly as planned. In the reality of the OR, those needle tips are not
Figure 8. Seed distribution and isodose plan for a 31 cm3 prostate (heavy grey line) after enlargement to a 61 cm3 PTV (heavy dark line). The base is at the upper left (0.0 cm offset) and the slices proceed at 0.5 cm intervals left to right culminating with the apex at lower right (4.0 cm offset). Two alternate ways of treating the apex are illustrated in Figure 9. The 100% and 150% isodose curves are shown, with the 100% as the outermost. Seeds present on a slice are small filled circles, and needle tracks empty on a slice are small open circles. The urethra appears at D3.5 at the base and curves downward to D3.8 at the base. The seed removal pattern from four central needles each on the 2.5 and 3.5 rows was designed to lower V150 to the desired range with urethral sparing as a secondary effect. The 150% isodose still touches the urethra in places, but additional sparing of the urethra was left for the brachytherapist in the OR as shown in Figure 10.
574
Wayne M. Butler and Gregory S. Merrick
Figure 9. Alternate treatments of the apex. In both rows (a) and (b), the first image is of the prostate apex, which is the last image in Figure 8. Row (a) was applied to this patient because the measured sagittal length of the prostate was <4.0 cm. Eight seed trains were extended to place seeds on the 4.5 cm plane—6 of those seeds are back-to-back. Row (b) would apply if the sagittal length of the prostate measured between 4.0 and 4.5 cm. Seed trains are extended to place seeds on the 5.0 cm plane, and the 4.5 cm plane is covered by the prescribed dose even though no seeds are on that plane.
pushed into the bladder but are brought into contact with the bladder wall via sagittal ultrasound and fluoroscopic guidance and the seed trains appropriately shortened depending on the tilt of the bladder/prostate interface. Seeds more than 5 mm outside the PTV were omitted from the uniformly loaded seed trains with one exception. The six seeds in the 1.5 row at the base were left in this plan—and in all our plans, despite being more than 5 mm outside the PTV—in order to augment the dose to the proximal seminal vesicles and ensure thorough coverage of the base. Only three additional peripheral needles were deemed necessary. One of these was placed at D5.0. The midline column should be avoided, but at a high anterior position and offset from the base, the seeds in this needle will not harm the bladder neck or urethra. The other two peripheral needles were placed at B4.0 and F4.0. These peripheral needles and seed trains do not extend to the base but are offset 5 mm from the base plane. As with the uniformly loaded needles, seeds more than 5 mm outside the PTV in the peripheral needles were omitted. Because the strength of the seeds was high relative to the prescribed dose—3.0 U 103Pd, 125 Gy— the seed removal pattern extended well beyond those closest to the urethra. Uniformly loaded seeds were removed, starting with the 1.0 cm plane, in an alternating zigzag pattern spread over eight central needles. The goal in all plans should be not only to spare the urethra, but also to lower V150 and V200 to the desired range. The actual apical plan for this patient is shown in Figure 9a along with an alternate plan that would have been employed had his prostate been a few mm longer. The zigzag vacancy pattern of the 4.0 cm plane was aliased on the final 4.5 cm plane with seeds. Four other uniform load seed trains were extended to place seeds on the 4.5 cm plane leaving the last seeds in those needles back-to-back.
29–Permanent Prostate Brachytherapy Treatment Planning
575
Figure 10. The same implant of Figures 8 and 9 but with four needles near the urethra shifted 2 mm away from the plan template coordinates to create a generous urethral sparing region. Two needles on the posterior row were moved upward 2 mm for rectal sparing. Upper figure: needle placement diagram with the six needle shifts. Lower image: the PTV, urethra, and seeds covered with the 150% isodose volume.
Because the 150% isodose line still touches the urethra, additional sparing of the urethra may be necessary if the brachytherapist executes the plan without adjustment or modification. In the OR, the brachytherapist can track the urethra more accurately and reproducibly than is possible from an in-office volume study. In our practice, additional urethra sparing would not be applied to the plan, but intentional deviation of periurethral needles from the plan positions is an expected result of OR tracking of the urethral path. Figure 10 shows how much change modest movement of a few needles can effect. Figure 11 displays the urethra, prostate, and PTV dose-volume histograms (DVHs) from the plan of Figures 8 and 9. DVHs are also shown for the urethra and PTV after those slight needle shifts. A great deal of power to control the implant lies in the radiation oncologist’s hands, and it is incumbent on the physicist to understand the preferences of the radiation oncologist before planning a complex loading scheme with superb urethral sparing if such a plan is unnecessary. Finally, one should check the plan for robustness. There are sophisticated software programs available that can apply to each seed in the implant a Gaussian spread of perturbations along each major
576
Wayne M. Butler and Gregory S. Merrick
Figure 11. DVHs of the plans in Figures 8, 9, and 10. The solid lines show the urethra, prostate, and PTV curves for the unmodified plan. Note that the PTV curve is shifted to higher doses and indicates less homogeneity than the prostate curve because of needle and seed additions around the periphery of the PTV. Needle shifts of 2 mm outward on the four needles around the urethra (Figure 8) result in a pronounced shift of the urethral DVH (left dotted line) to lower doses and slightly more homogeneity. Movement of those needles plus two needles near the rectum degrades the PTV curve (right dotted line) so that it is less homogeneous. The needle-shifted prostate DVH behaves similarly and is omitted to reduce clutter.
axis (Wuu et al., 2000; Beaulieu et al., 2004). Tallying the results of hundreds of runs allows calculation of the standard deviation of each dosimetric parameter of interest. If clinically relevant mean seed displacements are used as input to perturbations of the plan, the only results to be concerned about are those that indicate coverage failures and not high doses because high doses to the urethra or rectum are obviated by needle dithering in the OR. By planning with margin and an explicit PTV, perturbation trials of a well-designed plan may find instances where the PTV is not fully covered, but never the prostate. If the prostate itself is under dosed, the cool region will be centrally located and attributable to overly aggressive urethral sparing. A simpler test of robustness is to check for the effect of seed loss. About 1% to 2% of seeds are lost to dosimetry during the first half-life of the radionuclide (Merrick et al., 2000), and seed loss has a greater effect on dosimetric quality than random perturbations of all seeds by a couple of millimeters. Random deletion of a single seed on the periphery typically reduces the coverage of the PTV by the 100% prescription isodose, V100, by <2%. For the patient plan of Figures 8 and 9, no loss of a single internal seed creates a cool spot centrally, and the worse case scenario of a two-seed loss reduces V100 from 100% to 98.4%.
Peripheral Loading Just as uniform loading was predicated on the theoretical advantages of radionuclides with rapidly attenuated emissions, peripheral loading is predicated on an ideal source, albeit one with opposite properties. Radium and its analogs, 137Cs and 192Ir, exhibit little attenuation other than inverse square effects over the range of distances of interest to brachytherapists. For example, the radial dose function of 192Ir is 1.0±0.02 out to a distance of 5 cm. Cumulative effects from multiple sources are an important consideration. An ideal peripherally loaded prostate would have all the sources 5 to 10 mm inside the prostate capsule to spare the rectum and bladder while providing an adequate dose to the prostate plus margin. Judicious placement of the peripheral sources can fill the interior with a dose above the prescription threshold and below the level that will damage the urethra. Nevertheless, peripheral loading stands, like uniform loading, as an ideal construct to be modified rather than executed in its pure form. High dose-rate (HDR) prostate
29–Permanent Prostate Brachytherapy Treatment Planning
577
brachytherapy, which places a cylinder of catheters within the prostate, starts off ideally, but modifies the distribution in the interest of homogeneity by placing a few interior needles. See the chapter on HDR prostate brachytherapy for examples. Although no definitive recommendations have been proposed for preimplant prostate dosimetry, AAPM TG-56 suggested that treatment plans be “designed to place seeds peripherally to improve dose homogeneity and to avoid unnecessary radiation damage to the urethra” (Nath et al., 1997). Consistent with this statement, the ABS recommended that modified peripheral loading techniques be used to minimize the length of urethra receiving >200% of prescription dose (Nag et al., 1999). With low dose-rate (LDR) permanent seed radionuclides, substantial modifications to the peripheral ideal are necessary. Except for the smallest prostates, the approach begins with two or more concentric cylinders of seeds. Because the outermost cylinder should follow the shape of the patient’s PTV, this approach is not as amenable to algorithmic design. Most practitioners try to increase long-range cumulative effects by using higher strength seeds or frequent use of back-to-back seeds. Their approach is characterized by placing virtually no seeds outside the PTV because the higher seed strength will push the dose out to the margins. To spare the rectum, the posterior needle plane should be established 5 mm anterior to the posterior border of the prostate. With the posterior plane of needles established on a mid-gland slice, an arc of needles is placed 4 to 6 mm inside the PTV following the lateral and anterior periphery and with each needle at least 1 cm from its nearest neighbor. As with the modified uniform loading approach discussed above, these needles should have seeds extending symmetrically from the base to the apex (or 0.5 cm inferior to the apex when the apex falls on a half integer plane). For most prostate PTVs, a second cylinder along with several interior needles must be placed to build up the dose centrally. To create a more homogeneous dose, the tips of most of these needles should extend only to the transverse plane 0.5 cm inferior to the base plane. If these seed trains follow the standard 1-cm spacing, seeds dropped by these needles occupy the half integer planes. Needles and seeds should also be kept at least 0.5 cm from the urethra.
Computer-Assisted and Optimized Planning Almost from the beginning, dedicated prostate implant planning systems contained software to speed up and enhance the planning process. Among manual planners, the most widely used feature is the ability to drop whole needles full of seeds along any chosen path. Another feature allows users to specify a set of needle and seed placement rules such as the manual algorithm discussed above for uniform/modified uniform seed loading. Such geometry rules based software should allow the user to specify allowed coordinates—rows and columns—for at least two types of needles—uniform, interior and peripheral. Each type of needle should provide an editable list of target specification criteria such as maximum and minimum outer and inner margins for placement of needles relative to the PTV and the urethra as well as bilateral symmetry constraints and the avoidance of single seed needles. Geometry rules never produce an optimum plan. Users must be acutely aware of what the isodose plots and DVH curves are showing them. Inverse planning optimization of prostate seed implants has provided thesis material for a substantial fraction of a generation of graduate students. The computer software tries to satisfy multiple dosimetric (DVH based) and geometric objectives. Some of the algorithms used have been based on genetic optimization (Yu 2005), simulated annealing (Pouliot et al., 1996), and mixed integer iteration (D’Souza et al., 2001). All these and more are detailed in the chapter on brachytherapy optimization. Users quickly discover that the optimization engine cannot meet all dosimetric goals even when priorities are changed. Few of the resulting solutions cannot be further improved by manual intervention, usually by relaxing geometric boundaries on seed placement.
578
Wayne M. Butler and Gregory S. Merrick
Convergence of Approaches The manual planning approaches, modified uniform loading and modified peripheral loading, reach convergence when identical seed strengths are used. Even the inversely planned implants approach the same convergence when the geometric constraints are applied with an aesthetic sensibility and symmetry embodied in the manual approaches. All three approaches can accomplish the goals of good coverage of the treatment volume, acceptable homogeneity, plus urethral and rectal sparing. As seed strength increases, homogeneity decreases, but there is no a priori reason to favor homogeneous implants over inhomogeneous implants other than a conservative impulse to not waste dose. Because current brachytherapy doses appear sufficient to completely ablate all prostate tissue (Pickett et al., 2004), it is unlikely that any homogeneity indices will correlate with biochemical survival. Jones et al., (2002) found no correlation between dose inhomogeneity and urinary morbidity in a population where the mean V200 and V300 were 35% and 14%, respectively. However, a seed distribution that places high dose regions in close proximity to critical structures is unnecessarily risky. Maintaining the planned dosimetric demarcation between cancer and critical structure is difficult in practice.
Choice of Radionuclide There is scant published evidence to support the choice of one radionuclide over the other. The traditional radiobiological argument was that the shorter half-life of 103Pd makes it favored over 125I for faster growing, more aggressive tumors. This logic led many centers to adopt 125I for low-risk, monotherapy patients and 103Pd for intermediate- and high-risk patients. However, there are many centers that use either one isotope or the other for all patients and who report good results. In very preliminary results from a randomized trial comparing 125I with 103Pd in low-risk patients, there was a slight trend toward enhanced biochemical survival in the 103Pd arm (Wallner et al., 2003). PSA levels of patients implanted with 103Pd also fall to undetectable levels faster than those implanted with 125I (Merrick et al., 2004a). In terms of morbidity, there was a significantly faster resolution of urinary symptoms among patients in the 103Pd arm of a randomized trial (Wallner et al., 2002). Operative edema has a more deleterious effect on 103Pd implants than on 125I implants because of the shorter half-life and faster drop off of the radial dose function in the former (Butler et al., 2000a). The radial dose function of 103Pd falls off more rapidly with distance because 103Pd has a lower average photon energy than 125I. Edema increases the average distance between seeds in an implant, and this will have a greater effect on the coverage parameters D90 and V100 of 103Pd implants. Further exacerbating the loss of coverage is the 17-day half-life of 103Pd compared to 60 days for 125I. An implanted prostate will be edematous during a greater portion of the effective life of 103Pd than for 125I.
Appropriate Seed Strengths Increasing the seed strength will reduce the cost of an implant as long as manufacturers and hospitals bill by the number of seeds ordered and not by the total strength. However, the number of seeds required is not a linear function of seed strength. The number of 103Pd seeds used at the Schiffler Cancer Center fits an empirical equation based on the TRUS volume (V) in cm?, the prescribed dose (mPD) in Gy, and the seed strength (U): Seeds = 4.9
V ⋅ mPD U
0.44
(3)
29–Permanent Prostate Brachytherapy Treatment Planning
579
Doubling the seed strength from 2 U to 4 U to deliver a monotherapy dose to an average size prostate only reduces the seed count by about 25%. Table 2 lists the nondosimetric treatment parameters at our center averaged over the past 3 years. By maintaining a constant ratio between prescribed dose and individual seed strength, the planner can change the prescribed dose for an already-planned patient if the radiation oncologist desires. By adjusting the seed strength, the planner can use exactly the same seed distribution as in the original plan and duplicate the relative dosimetric quality parameters of the original plan. Reducing the number of seeds by increasing individual seed strength to deliver the same prescribed dose as with lower strength seeds does not necessarily mean that seed placement errors or even seed loss will have greater adverse prostate dosimetric consequences. Narayana and colleagues (2005) performed a randomized trial comparing high- versus low-strength 125I seeds and found that the high-strength implants were more robust and produced higher values of postoperative dosimetric parameters than the low strength implants when comparing prostate D80 and V100,150,200. In the high-strength arm, rectal and urethral doses were not significantly higher, but the total volume covered by the prescription isodose was about 18% greater. For both arms, the 2-week MRI-based mean V200 was about 1.4 times the plan V200, but the mean D90, D95, and D99 values were all considerably lower than the plan values for both arms. Part of the loss of Dxx coverage may be due to residual edema at 2 weeks post implant, but part of the loss in Dxx and the increase in V200 may have been due to their use of stranded seeds for all the implants. Figure 12 illustrates schematically the cause of central dose escalation and the loss of peripheral coverage as edema waxes and wanes in a stranded seed implant compared to a loose seed implant. Using strands that are longer than the expected extent of linear edema can prevent loss of coverage in a stranded implant, but the dose escalation is unavoidable.
Preplan Evaluation Criteria Before using a planning system for the first time, a physicist should acceptance test the system with partic· ular attention to geometric fidelity and dosimetric accuracy. The initial dose rate, D (0), at selected points around a seed may be calculated from the seed’s tabulated basic dosimetry parameters and the equations of AAPM Task Group 43 and its successors (Nath et al., 1995; Rivard et al., 2004). Most dedicated brachytherapy planning systems report either dose rate or dose to total decay for permanent prostate implants, but some report only total dose. The total dose delivered, D(t), is given by: T D(t) = D" (0 ) ⋅ 1/ 2 ln (2 )
(4)
where T1/2 is the half-life of the radionuclide and dividing by ln(2) gives the average life of the radionuclide. Every plan should undergo some independent check. Rather than verify that the computer is still calculating correctly each time, it is probably more useful to verify one’s planning consistency and to track trends. A simple check that is totally independent of the computer is to calculate the total strength needed for an implant from the average dimension, davg, of the PTV from the maximum length, width, and height, davg = 1 ⁄3 (l·w·h). By gathering dimensions for a few dozen implants and scaling the total strength used in boost implant to monotherapy implants, the total strength can be fit to an empirical power equation such as: Sk = α ⋅ β
davg
For the 103Pd implants in our program, a = 61 U and b = 1.46.
(5)
580
Wayne M. Butler and Gregory S. Merrick Table 2. Mean treatment-Planning Parameters Used at the Schiffler Cancer Center. Patients who receive a 90 Gy or a 115 Gy 103Pd implant receive 45 Gy or 20 Gy, respectively, via 3-D conformal external beam therapy. 125
Parameter
Overall
103
I
Pd
145 Gy
90 Gy
115 Gy
125 Gy
Ultrasound volume (cm3)
33.6 ±8.8
37.1 ±8.9
29.0 ±9.3
33.1 ±8.1
35.2 ±7.1
Planning volume (cm3)
65.3 ±12.7
70.3 ±11.8
58.3 ±14.9
63.8 ±11.9
67.6 ±10.5
Planning margin (mm)
5.02 ±0.54
4.97 ±0.53
5.03 ±0.55
4.92 ±0.41
4.96 ±0.49
Enlargement factor
1.97 ±0.21
1.93 ±0.20
2.06 ±0.21
1.96 ±0.18
1.94 ±0.17
Number of seeds *
123 ±16
126 ±13
110 ±17
118 ±15
125 ±13
Number of needles
26 ±3
26 ±3
25 ±3
26 ±3
27 ±3
0.54 ±0.3
2.28 ±0.09
2.85 ±0.13
3.01 ±0.12
Seed strength (U)
* This is the number of seeds actually implanted, which exceeds the number of seeds in the preplan by 6 to 10 extra seeds. These extras are primarily used to implant the proximal seminal vesicles.
Figure 12. Behavior of loose seeds versus stranded seeds (dotted outline). (a) At the time of implant, the loose seeds (first row) and stranded seeds (second row) extend from the base to the apex (heavy arcs). (b) As the prostate undergoes edema, loose seeds move with the swelling tissue, but strands do not stretch and prostate tissue moves away from the seeds at either or both the base and apex. (c) As edema resolves, loose seeds will follow and maintain full-length coverage. Because strands are not rigid, the seeds will move inward with the shrinking tissue resulting in a closer seed separation than was planned. The stranded seeds will overdose the center of the prostate while under dosing the ends.
Table 3 lists quality factors used as guidelines to judge all implants, whether monotherapy or boost therapy, 125I or 103Pd. One factor conspicuous for its absence is rectal dose. Every prostate has its posterior border placed in the same position and implanted in exactly the same way—uniformly loaded needles occupying all valid positions on the template 1.5 row of Figures 8 and 9. Contouring the rectum to obtain a DVH would not produce any actionable information.
29–Permanent Prostate Brachytherapy Treatment Planning
581
Table 3. Prostate implant plan evaluation criteria used at the Schiffler Cancer Center Evaluated quantity
Parameter
Value
Patient specific needs
PTV, TURP dose, etc.
Primary importance
Planning volume coverage
V100
>99.8% volume
Urethral volume coverage
Urethral V125 Urethral V150
10% – 50% volume <15% volume
Urethra dose
Mean
110% – 140% mPD
Homogeneity
V150
35% – 45% plan vol, 125I 45% – 55% plan vol, 103Pd
High dose volume
V200
<15% plan volume, 125I <25% plan volume, 125I
Number of needles
Minimize
30±4 *, †
Number of seeds
Minimize
130±18 *, †
No. of especially loaded needles
Minimize
5±2 *
Target volume / US volume
Ratio
1.75 ± 0.22 *
V100, V150 and V200 are the percentage of the planning volume covered by 100, 150, and 200% of the prescribed dose (mPD), respectively. * There is no statistically significant difference in these parameters between radionuclides, monotherapy, or boost therapy. † Typically, 3 or 4 extra needles and 7% extra seeds are used at the time of implant beyond those called for in the plan.
Review of Contemporary Planning Variations An ABS survey in 1998 noted that 61% of brachytherapists added periprostatic planning margins and 63% of brachytherapists implanted seeds in periprostatic locations (Prete et al., 1998). To investigate the state of prostate planning today in detail, we invited eight experienced brachytherapists who implant with both 125 I and 103Pd to submit monotherapeutic and boost preimplant dosimetry plans for central review (Merrick et al., 2005a). All 32 plans utilized the same transrectal ultrasound volumetric study with the prostate outlined. Slices superior to the base and inferior to the apex were included as was a sagittal image that had the prostate length marked. Each brachytherapist was asked to plan in his usual fashion by moving the prostate, enlarging it to create a PTV if desired, and using seeds of any strength of both radionuclides. Central dosimetric analysis included evaluation of treatment and dosimetric parameters. Table 4 lists some of the variable brachytherapy planning parameters as applied to the prostate for 103Pd monotherapy, and Table 5 lists dosimetric quality and margin parameters. The mean PTV to prostate volume ratio varied dramatically (mean 1.29, range 0.99 to 1.76), and the target length ranged from 3.5 to 4.5 cm. Two brachytherapists did not draw an explicit PTV and one drew a PTV that did not cover the entire prostate but instead included the seminal vesicles. Across the four treatment modalities involving both radionuclides, the prostate V100 was >95% in all cases, but the V150 ranged from 30% to 92% and the V200 from 7% to 52%. Prostate DVHs are shown in Figure 13 for monotherapy 125I and 103Pd plans. The urethra V100 was 100% in all cases, with six of the eight brachytherapists limiting the UV150 to <3%. Mean dosimetric treatment margins from the prostate to the 100% isodose surface also varied significantly (average 4 mm, range 0.3 to 7.7 mm). All brachytherapists used extracapsular seeds, and five implanted >25% of the seeds in extracapsular locations (range 6% to 58%). In addition, significant vari-
1.60
0.99
1.17
1.00
1.76
1.57
1.21
1.00
1.29
A (JM)
B (BM)
C (PG)
D (KW)
E (GM)
F (JB)
G (PM)
H (JA)
Mean
3.94
3.5
3.5
4.0
4.5
3.5
4.0
4.0
4.5
PTV Length (cm)
21.8
24
29
24
22
18
23
14
20
Number Needles
7.2
15
3
4
9
16
2
6
3
Specially Loaded Needles
1.0
0
6
0
0
2
0
0
0
Single Seed Needles
8.8
4*
4
0
4
52
0
6
0
Back to Back Seed Pairs
2.43
2.20
2.59
1.81
3.05
2.95
2.00
2.47
2.33
Seed Strength (U)
82.5
78
74
100
91
78
91
62
86
Number of Seeds
27.5
5
8
41
50
24
34
17
41
(32.1)
(6.4)
(10.8)
(41.0)
(54.9)
(30.8)
(37.4)
(27.4)
(47.7)
Extracapsular Seeds (Total) (%)n
4.25
0
0
8
10
0
0
10
6
Sem. Ves. Seeds
(Reproduced from Brachytherapy, In press. “Prostate brachytherapy preimplant dosimetry: a multi-institutional analysis,” G. S. Merrick, W. M. Butler, K. E. Wallner, J. C. Blasko, J. Michalski, J. Aronowicz, P. Grimm, B. Moran, P. McLaughlin, J. Usher, J. Lief, and Z. Allen. © 2005, with permission from Elsevier.)
* The variability for each parameter was similar for 103Pd boost therapy and for 125I monotherapy and boost therapy.
PTV / Prostate ratio
Investigator
Table 4. Variable Brachytherapy Planning Parameters Stratified by Investigator as Applied to the Prostate for 103Pd Monotherapy *
582 Wayne M. Butler and Gregory S. Merrick
29–Permanent Prostate Brachytherapy Treatment Planning
583
Table 5. Dosimetric Parameters Stratified by Investigator. Vxx and UVxx Are % Prostate and Urethra Volume Covered by xx Percent of the mPD, Respectively. Dyy and UDyy are the Minimum Doses, as Percentage of the Minimal Peripheral Dose (mPD), Covering yy Percent of the Volume Investigator
V100
V150
V200
D100
D90
UV150
UD50
UD10
Mean margin *
± SD
A B C D E F G H Mean
98.8 97.6 99.9 100.0 100.0 99.8 99.9 100.0 99.5
43.1 48.1 50.4 86.0 58.1 53.0 67.4 68.5 59.3
15.1 20.2 14.5 52.5 16.2 13.3 37.9 35.0 25.6
81.5 68.3 98.8 99.2 111.7 96.5 94.5 95.1 93.2
124.5 111.5 117.3 144.5 130.7 120.5 128.2 123.7 125.1
23.4 1.3 1.4 33.2 2.9 1.3 0.0 0.3 8.0
134.8 108.7 127.9 141.1 129.9 123.3 124.9 111.8 125.3
159.0 133.6 140.1 161.9 140.4 137.1 134.7 137.3 143.0
3.39 1.13 3.29 5.50 7.68 3.56 3.28 2.65 3.81
3.90 2.99 2.73 3.03 2.44 3.09 2.05 1.87 2.76
* Dosimetric margins were measured in mm from the prostate to the 100% isodose line at 8 angles (45° increments) on each of 8 slices.
(a)
(b) Figure 13. Prostate DVHs stratified by brachytherapist from highest to lowest D60 dose (the dose at 60% of the prostate volume). (a) monotherapy, 125 Gy 103Pd; (b) monotherapy, 145 Gy 125I.
584
Wayne M. Butler and Gregory S. Merrick
(a)
(b)
(c)
(d)
Figure 14. Composite monotherapy 103Pd isodose plan incorporating all seeds from each brachytherapist. The seed strengths in the eight plans weighted by 1/8, and the 660 seeds used resulted in 375 unique seed positions and a maximum redundancy of 5 seeds on one coordinate. The isodose lines from outermost to innermost are 100%, 150%, and 200% of mPD. (a) 5 mm superior to the base; (b) prostate base; (c) mid-gland; (d) prostate apex.
ability existed in the number of needles, the number of seeds, and seed strength. One investigator used the same number of needles and seeds for all four plans by maintaining a constant seed strength to mPD ratio. The 660 seeds placed by individual investigators for 103Pd monotherapy resulted in 376 unique seed positions, and the maximum redundancy at any position was five seeds in four instances. Figure 14 shows a composite isodose plan of all eight 103Pd monotherapy submissions. Except for a few cases, individual brachytherapists were more consistent across modalities than was the consistency across investigators for each modality.
29–Permanent Prostate Brachytherapy Treatment Planning
585
Conclusions Although prostate brachytherapy prescription doses are fairly uniform, there are multiple seed placement philosophies currently in use. Substantial variability exists regarding target volume, seed strength, dose homogeneity, treatment margins, and extracapsular seed placement. Without standardization of planning and post implant dosimetry techniques, there can be no meaningful multi-institutional comparisons of biochemical outcomes and morbidity.
References Anderson, L. L. “Spacing nomogram for interstitial implants of 125I seeds.” (1976). Med Phys 3:48–51. Beaulieu, L., L. Archambault, S. Aubin, E. Oral, R. Taschereau, and J. Pouliot. (2004). “The robustness of dose distributions to displacement and migration of 125I permanent seed implants over a wide range of seed number, activity, and designs.” Int J Radiat Oncol Biol Phys 58:1298–1308. Beyer, D. C., R. H. Shapiro, and F. Puente. (2000). “Real-time optimized intraoperative dosimetry for prostate brachytherapy: a pilot study.” Int J Radiat Oncol Biol Phys 48:1583–1589. Butler, W. M., G. S. Merrick, A. T. Dorsey, and J. H. Lief. (2000). “Comparison of seed loading approaches in prostate brachytherapy.” Med Phys 27:381–392. Butler, W. M., G. S. Merrick, J. H. Lief, and A. T. Dorsey. (2000a). “Isotope choice and the effect of edema on prostate brachytherapy.” Med Phys 27:1067–1075. Choi, S., K. Wallner, G. Merrick, and W. Butler. (2004). “Treatment margins predict biochemical outcomes after prostate brachytherapy.” Cancer J 10:175–180. Cormack, R.A., H. Kooy, C. M. Tempany, A. V. D’Amico. (2000). “A clinical method for real-time dosimetric guidance of transperineal 125I prostate implants using interventional magnetic resonance imaging.” Int J Radiat Oncol Biol Phys 46:207–214. D’Amico, A. V., R. Whittington, S. B. Malkowicz, D. Schultz, K. Blank, G. A. Broderick, J. E. Tomaszewski, A. A. Renshaw, I. Kaplan, C. J. Beard, and A. Wein. (1998). “Biochemical outcome after radical prostatectomy, external beam radiation therapy, or interstitial radiation therapy for clinically localized prostate cancer.” JAMA 280:969–974. Dattoli, M. “103Pd Brachytherapy: Rationale, Design, and Evaluation” in Basic and Advanced Techniques in Prostate Brachytherapy. A. P. Dicker, G. S. Merrick, F. M. Waterman, R. K. Valicenti, and L. G. Gomella (eds.). London, UK: Taylor and Francis, 2005. Davis, B.J., T. M. Pisansky, T. M. Wilson, H. J. Rothenberg, A. Pacelli, D. W. Hillman, D. J. Sargent, and D. G. Bostwick. (1999). “The radial distance of extraprostatic extension of prostate carcinoma: implications for prostate brachytherapy.” Cancer 85:2630–2637. D’Souza, W. D., R. R. Meyer, B. R. Thomadsen, and M. C. Ferris. (2001). “An iterative sequential mixed-integer approach to automated prostate brachytherapy treatment plan optimization.” Phys Med Biol 46:297–322. Goodsitt, M.M., P. L. Carson, S. Witt, D. L. Hykes, and J. M. Kofler, Jr. (1998) “Real-time B-mode ultrasound quality control test procedures. Report of AAPM Ultrasound Task Group No. 1.” Med Phys 25:1385–1406. Holmes, D. R. 3rd, B. J. Davis, C. J. Bruce, and R. A. Robb. (2003). “3D visualization, analysis, and treatment of the prostate using trans-urethral ultrasound.” Comput Med Imaging Graph 27:339–349. Jones, S., K. Wallner, G. Merrick, J. Corriveau, S. Sutlief, L. True, and W. Butler. (2002) “Clinical correlates of high intraprostatic brachytherapy dose regions.” Int J Radiat Oncol Biol Phys 53:328–333. Kestin, L., N. Goldstein, F. Vicini, D. Yan, H. Korman, and A. Martinez. (2002). “Treatment of prostate cancer with radiotherapy: should the entire seminal vesicles be included in the clinical target volume?” Int J Radiat Oncol Biol Phys 54:686–697. Merrick, G. S., W. M. Butler, A. T. Dorsey, M. L. Benson, and J. H. Lief. (2000). “Seed fixity in the prostate/periprostatic region following brachytherapy.” Int J Radiat Oncol Biol Phys 46:215–220. Merrick, G. S., W. M. Butler, R. W. Galbreath, J. H. Lief, and E. Adamovich. (2001). “Perineural invasion is not predictive of biochemical outcome following prostate brachytherapy.” Cancer J Sci Am 7:404–412.
586
Wayne M. Butler and Gregory S. Merrick
Merrick, G. S., K. E. Wallner, W. M. Butler, R. W. Galbreath, J. H. Lief, M. J. Benson. (2001a). “A comparison of radiation dose to the bulb of the penis in men with and without prostate brachytherapy induced erectile dysfunction.” Int J Radiat Oncol Biol Phys 50:597–604. Merrick, G. S., W. M. Butler, R. W. Galbreath, J. H. Lief, and E. Adamovich. (2002). “Biochemical outcome for hormone naïve patients with Gleason score 3+4 versus 4+3 prostate cancer undergoing permanent prostate brachytherapy.” Urol 60:98–103. Merrick, G. S., W. M. Butler, B. G. Tollenaar, R. W. Galbreath, and J. H. Lief. (2002a). “The dosimetry of prostate brachytherapy-induced urethral strictures.” Int J Radiat Oncol Biol Phys 52:461–468. Merrick, G. S., W. M. Butler, K. E. Wallner, J. H. Lief, R. L. Anderson, B. J. Smeiles, R. W. Galbreath, and M. L. Benson. (2002b). “The importance of radiation doses to the penile bulb vs. crura in the development of post brachytherapy erectile dysfunction.” Int J Radiat Oncol Biol Phys 54:1055–1062. Merrick, G. S., W. M. Butler, K. E. Wallner, L. R. Burden, and J. E. Dougherty. (2003). “Extracapsular radiation dose distribution after permanent prostate brachytherapy.” Am J Clin Oncol 26:e179–e189. Merrick, G. S., W. M. Butler, K. E. Wallner, R. W. Galbreath, J. Lief, and E. Adamovich. (2004). “Prognostic significance of percent positive biopsies in clinically organ-confined prostate cancer treated with permanent prostate brachytherapy with or without supplemental external-beam radiation.” Cancer J Sci Am 10:54–60. Merrick, G. S., W. M. Butler, K. E. Wallner, R. W. Galbreath, Z. Allen, and E. Adamovich. (2004a). “Temporal effect of neoadjuvant androgen deprivation therapy on PSA kinetics following permanent prostate brachytherapy with or without supplemental beam radiation.” Brachytherapy 3:141–146. Merrick, G. S., W. M. Butler, K. E. Wallner, R. W. Galbreath, J. H. Lief, Z. Allen, and E. Adamovich. (2005). “Impact of supplemental external beam radiotherapy and/or androgen deprivation therapy on biochemical outcome after permanent prostate brachytherapy.” Int J Radiat Oncol Biol Phys 61:32–43. Merrick, G. S., W. M. Butler, K. E. Wallner, J. C. Blasko, J. Michalski, J. Aronowicz, P. Grimm, B. Moran, P. McLaughlin, J. Usher, J. Lief, and Z. Allen. (2005a). “Prostate brachytherapy preimplant dosimetry: a multiinstitutional analysis.” Brachytherapy. In press. Nag, S., D. Beyer, J. Friedland, P. Grimm, and R. Nath. (1999). “American Brachytherapy Society (ABS) recommendations for transperineal permanent brachytherapy of prostate cancer.” Int J Radiat Oncol Biol Phys 44:789–799. Nag, S., J. P. Ciezki, R. Cormack, S. Doggett, K. DeWyngaert, G. K. Edmundson, R. G. Stock, N. N. Stone, Y. Yu, and M. J. Zelefsky. (2001). “Intraoperative planning and evaluation of permanent prostate brachytherapy: Report of the American Brachytherapy Society.” Int J Radiat Oncol Biol Phys 51:1422–1430. Narayana, V., S. Troyer, V. Evans, R. J. Winfield, P. L. Roberson, and P. W. McLaughlin. (2005). “Randomized trial of high- and low-source strength 125I prostate seed implants.” Int J Radiat Oncol Biol Phys 61:44–51. Narayanan, S., P. S. Cho, and R. J. Marks 2nd. (2004). “Three-dimensional seed reconstruction from an incomplete data set for prostate brachytherapy.” Phys Med Biol 49:3483–3494. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43.” Med Phys 22:209–234. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24:1557-1598. Also available as AAPM Report No. 51. Partin, A. W., L. A. Mangold, D. M. Lamm, P. C. Walsh, J. I. Epstein, and J. D. Pearson. (2001). “Contemporary update of prostate cancer staging nomograms (Partin Tables) for the new millennium.” Urol 58:843–848. Pickett, B., R. K. Ten Haken, J. Kurhanewicz, A. Qayyum, K. Shinohara, B. Fein, and M. Roach. (2004). “Time to metabolic atrophy after permanent prostate seed implantation based on magnetic resonance spectroscopic imaging.” Int J Radiat Oncol Biol Phys 59:665–673. Prete, J. J., B. R. Prestidge, W. S. Bice, J. L. Friedland, R. G. Stock, and P. D. Grimm. (1998). “A survey of physics and dosimetry practice of permanent prostate brachytherapy in the United States.” Int J Radiat Oncol Biol Phys 40:1001–1005. Pouliot, J., D. Tremblay, J. Roy, and S. Filice. (1996). “Optimization of permanent 125I prostate implants using fast simulated annealing.” Int J Radiat Oncol Biol Phys 36:711–720.
29–Permanent Prostate Brachytherapy Treatment Planning
587
Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. (2004). “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Also available as AAPM Report No. 84. Rivard, M. J., D.-A. Evans, and I. Kay. (2005). “A technical evaluation of the Nucletron FIRST® system: Conformance of a remote afterloading brachytherapy seed implantation system to manufacturer specifications and AAPM Task Group report recommendations.” J Appl Clin Med Phys 6:22–50. Roberson, P. L., V. Narayana, D. L. McShan, R. J. Winfield, and P. W. McLaughlin. (1997). “Source placement error for permanent implant of the prostate.” Med Phys 24:251–257. Roy J. N., K. E. Wallner, S. T. Chiu-Tsao, L. L. Anderson, and C. C. Ling. (1991). “CT-based optimized planning for transperineal prostate implant with customized template.” Int J Radiat Oncol Biol Phys 21:483–489. Sohayda, C., P. A. Kupelian, H. S. Levin, and E. A. Klein. (2000). “Extent of extraprostatic extension in localized prostate cancer.” Urol 55:382–386. Stamey, T. A., C. M. Yemoto, J. E. McNeal, B. M. Sigal, and I. M. Johnstone. (2000). “Prostate cancer is highly predictable: a prognostic equation based on all morphological variables in radical prostatectomy specimens.” J Urol 163:1155–1160. Stock, R. G., N. N. Stone, M. F. Wesson, and J. K. DeWyngaert. (1995). “A modified technique allowing interactive ultrasound-guided three-dimensional transperineal prostate implantation.” Int J Radiat Oncol Biol Phys 32:219–225. Stone, N. N., J. H. Chircus, R. G. Stock, and J. Presser. “The ProSeed Approach: A Multicenter Study of the Results of Brachytherapy Training” in Basic and Advanced Techniques in Prostate Brachytherapy. A. P. Dicker, G. S. Merrick, F. M. Waterman, R. K. Valicenti, and L. G. Gomella (eds.). London, UK: Taylor and Francis, 2005. Straub, B., M. Muller, H. Krause, C. Goessl, M. Schrader, R. Heicappell, and K. Miller. (2001). “Reverse transcriptase-polymerase chain reaction for prostate-specific antigen in the molecular staging of pelvic surgical margins after radical prostatectomy.” Urol 57:1006–1011. Su, Y., B. J. Davis, M. G. Herman, and R. A. Robb. (2004). “Prostate brachytherapy seed localization by analysis of multiple projections: Identifying and addressing the seed overlap problem.” Med Phys 31:1277–1287. Sylvester, J. E. “The Seattle Prostate Institute Approach to Treatment Planning for Permanent Implants” in Basic and Advanced Techniques in Prostate Brachytherapy. A. P. Dicker, G. S. Merrick, F. M. Waterman, R. K. Valicenti, and L. G. Gomella (eds.). London, UK: Taylor and Francis, 2005. Sylvester, J. E., J. C. Blasko, P. D. Grimm, R. Meier, and J. A. Malmgren. (2003). “Ten-year biochemical relapsefree survival after external beam radiation and brachytherapy for localized prostate cancer: the Seattle experience.” Int J Radiat Oncol Biol Phys 57:944–952. Theodorescu, D., H. F. Frierson, and R. A. Sikes. (1999). “Molecular determination of surgical margins using fossa biopsies at radical prostatectomy.” J Urol 161:1442–1448. Todor, D. A., M. Zaider, G. N. Cohen, M. F. Worman, and M. J. Zelefsky. (2003). “Intraoperative dynamic dosimetry for prostate implants.” Phys Med Biol 48:1153–1171. Van Gellekom, M. P. R., M. A. Moerland, H. K. Wijrdeman, and J. J. Batterman. (2004) “Quality of permanent prostate implants using automated delivery with seedSelectron‰ versus manual insertion of RAPID Strands‰.” Radiother Oncol 73:49–56. Wallner, K., G. Merrick, L. True, M. W. Kattan, W. Cavanagh, C. Simpson, and W. Butler. (2002). “125I versus 103Pd for low risk prostate cancer: morbidity outcomes from a prospective randomized multicenter trial.” Cancer J 8:67–73. Wallner, K., G. Merrick, L. True, S. Sutlief, W. Cavanagh, and W. Butler. (2003). “125I versus 103Pd for low risk prostate cancer: preliminary PSA outcomes from a prospective randomized multicenter trial.” Int J Radiat Oncol Biol Phys 57:1297–1303. Wuu, C. S., R. D. Ennis, P. B. Schiff, E. K. Lee, and M. Zaider. (2000). “Dosimetric and volumetric criteria for selecting a source activity and a source type (125I or 103Pd) in the presence of irregular seed placement in permanent prostate implants.” Int J Radiat Oncol Biol Phys 47:815–820. Yu, Y. “The PIPER Prostate Brachytherapy Planning System” in Basic and Advanced Techniques in Prostate Brachytherapy. A. P. Dicker, G. S. Merrick, F. M. Waterman, R. K. Valicenti, and L. G. Gomella (eds.). London, UK: Taylor and Francis, 2005.
588
Wayne M. Butler and Gregory S. Merrick
Zelefsky, M. J.,Y. Yamada, G. Cohen, E. S. Venkatraman, A. Y. Fung, E. Furhang, D. Silvern, and M. Zaider. (2000). “Postimplantation dosimetric analysis of permanent transperineal prostate implantation: improved dose distributions with an intraoperative computer-optimized conformal planning technique.” Int J Radiat Oncol Biol Phys 48:601–608.
Chapter 30
Treatment Delivery in Prostate LDR Brachytherapy Eugene P. Lief, Ph.D. Maimonides Comprehensive Cancer Center Brooklyn, New York Types of Surgical Procedures Used for LDR Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589 Earlier Approaches to the Brachytherapy of the Prostate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 589 Transperineal Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590 Patient Positioning, Anesthesia, and Implant Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 590 Imaging Techniques and Prostate Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592 Fluoroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592 Ultrasound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593 Magnetic Resonance Imaging and Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593 Cone-Beam CT-Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 593 Equipment and Accessories for Permanent Prostate Implants . . . . . . . . . . . . . . . . . . . . . . . . . 594 Ultrasound and Accessories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594 Mick Applicator™ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594 Pre-loaded needles and Rapid Strand™ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594 Radioactive Isotopes and Seeds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596 Iodine: 125I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596 Palladium: 103Pd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596 Radioactive Seed Assay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597 Radiation Safety in Permanent Prostate Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597 Receiving, Handling, and Loading Radioactive Seeds Before the Procedure . . . . . . . . . . . . . . . . . 597 Radiation Survey in the Radioactive Source Laboratory, Operating and Recovery Rooms . . . . . . . 598 Discharge Instructions to the Patient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 598 Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 599
Types of Surgical Procedures Used for LDR Delivery Earlier Approaches to the Brachytherapy of the Prostate Prostate brachytherapy was attempted for the first time at the beginning of the 20th century (Pasteau and Degrais 1914), when 226Ra sources were inserted initially via urethra using a silver tube and subsequently with the aid of the cystoscope. Due to the high urethral dose, there was severe morbidity after the procedure. Later on, radioactive gold (198Au) was used in the form of a colloid suspension and eventually as seeds (Flocks et al. 1952). The isotope 198Au has a short half-life (less than 3 days) and less penetrating radiation than 226Ra. The results of the treatment with radioactive gold were no better than with the external beam from high-energy x-ray machines; however, the radiation safety hazard was higher. As a result, the popularity of the prostate brachytherapy declined in the 1960s. A retropubic prostate brachytherapy technique using 125I seeds was introduced at Memorial SloanKettering Cancer Center (Whitmore, Hillaris, and Grabstald 1972, 2002). The isotope 125I offered radiobiological and radioprotective advantages due to the relatively short half-life (59.4 days) and low effective energy of its radiation (27.4 keV). The number of needles and seeds required for the treatment were determined intraoperatively after laparotomy, staging, and measurement of the prostate. Over 1000 patients were treated using this technique. Due to the deep location of the prostate gland within the pelvis and obstruction by the pubic symphysis, it was difficult to implant the seeds precisely.
589
590
Eugene P. Lief
Operative complications, such as prostatic or paraprostatic hemorrhage, obturators or femoral nerve injury, and retained foreign body, were reported for 6% of 300 patients (Fowler et al. 1979). Twenty-three percent in the same group of patients developed pelvic (lymphocele or hematoma, abscess, or cellulitis), cardiovascular (pulmonary embolus, superficial or deep vein thrombophelebitis, and one cerebrovascular accident), wound infection or hematoma, urinary (retention, fistula, acute epididymitis, acute bacterial prostatitis), or other post-operative complications. Of 177 patients who were followed for 6 months or more and who did not receive supplemental external beam radiation therapy, 28% experienced protracted or delayed morbidity owing to the operative procedure or radiation (Fowler et al. 1979). The symptoms of such late complications included urinary (voiding symptoms, urgency incontinence, stress incontinence), edema, rectal discomfort and bleeding, and wound sinus or hernia.
Transperineal Method Because of the problems with other earlier types of prostate implants, the transperineal interstitial implant was developed in combination with the external beam radiation therapy (Charyulu 1980). The technique was first improved by introducing C-arm fluoroscopy for seed and needle visualization and guidance (Kumar and Bartone 1981). Use of the transrectal ultrasonic probe, attached to a rigid needle template, enabled real-time needle visualization with respect to the prostate (Holm et al. 1983). Addition of the fluoroscopy to transrectal ultrasound (TRUS) allowed three-dimensional (3-D) on-line visualization and intraoperative interactive treatment planning (Nag 1985). In the 1980s to 1990s, there were numerous innovations of the transperineal technique including pretreatment planning and popularizing the procedure (Blasko et al. 1987; Blasko, Ragde, and Grimm 1991; Roy et al. 1993; Merrick et al. 1997; Nag et al. 1999, 2000). The latest innovations in transperineal method include intraoperative dosimetry (Zelefsky et al. 2000), intraoperative dynamic dose optimization (Lee and Zaider 2003; Todor et al. 2003), and innovative use of magnetic resonance imaging suggested for the image-guided prostate brachytherapy (Kooy et al. 2000). Intraoperative dosimetry allows treatment planning in the operating room based on the images acquired at the beginning of the procedure. The dynamic dose optimization goes one step further and performs the implant evaluation in the operating room before the end of the procedure. The dose distribution is based on the actual rather than on the planned seed positions. This opportunity allows correcting dose distribution by re-planning the case and implanting the remaining seeds as suggested by the planning system. We will return to these innovative techniques in chapter 33, “Modern Advances in Brachytherapy of the Prostate.”
Patient Positioning, Anesthesia, and Implant Delivery Typically, patients are told to drink only clear fluids in the 48 hours and to fast for at least 8 hours before the procedure. The patient either undergoes bowel preparation the night before or an enema on the morning of the procedure to obtain good TRUS images. After general or spinal anesthesia is administered, the patient is placed in the dorsal lithotomy position. A stabilizer, either floor-mounted or couch-mounted, rigidly holds a stepper to which the template and ultrasound transducer is attached. An example of the couch-mounted unit is the Radiation Therapy Products (Seattle, WA) precision stabilizer. The stabilizer consists of several connected bars made of aluminum and steel. The position of the bars may be adjusted to accommodate the patient and may be locked by a single trigger that tightens special nuts. Some groups additionally use a wedge-shaped cushion beneath the gluteal region (Stock et al. 1995) and gas or hydraulic stirrups to keep the patient’s legs in the proper position. The ultrasound probe holder with a reusable or disposable template (Figure 1) is securely attached to the stepper or stabilizer. It is important to distinguish between systems in which the ultrasound transducer moves independently from the template and
Treatment Delivery in Prostate LDR Brachytherapy
591
Figure 1. Transperineal disposable template for inserting needles in prostate LDR brachytherapy.
those where they move in synchrony. The physical distance between the template and the needle tip when visible on the transverse ultrasound monitor screen is constant in the latter situation but not the former. A Foley catheter is inserted, and the scrotum is retracted anteriorly and sutured or taped to the anterior abdominal wall. If the planning ultrasound volume was not obtained and a plan not designed previously but is done intraoperatively, then the probe is positioned in the rectum and the normal anatomy is identified. After that, the prostate is scanned in 0.5-cm probe position intervals, and the images are sent to the treatment planning system. The prostate contours and urethra should be visible in every slice. Using the images obtained, a plan may be designed manually or with computer assistance. The optimal seed location may be computed using a genetic algorithm (Zaider et al. 2000; Zelefsky et al. 2000) or other approach. After the plan is done, the implant can proceed as described below. Implants may be performed using a nomogram determination of the required activity instead of detailed and individualized planning. In this instance, a urologist determines the dimensions of the prostate based on the preoperative ultrasound scan. Using nomograms (Anderson 1976; Anderson, Moni, and Harrison 1993), one can find the total seed activity required for the target coverage. Based on the total activity and the activities of the individual seeds available, the radiation oncologist orders the required number of seeds. Sometimes, the number of seeds ordered exceeds the number of seeds required by 10% to 15%, to compensate for possible inaccuracies in the seed placement. During the procedure, the radiation oncologist acquires prostate images and inserts needles in possible locations in the periphery of the prostate, avoiding the urethra, rectum, and pubic arch (Nag, Fernandes, and Bahnson 1998; Vicini et al. 1999). Using the total number of seeds and the total number of needles implanted, the radiation oncologist decides how many seeds should be administered through each needle. Some authors suggest implanting more seeds through the central than through peripheral needles because of increased length of intersection with the prostate tissue (Stock et al. 1995). Additional information about the tumor location in the prostate and proximity of the rectum, urethra, or bladder can affect the loading of some needles. If following the coordinates of a plan, the urologist or radiation oncologist inserts either preloaded needles or hollow needles with stylets inside under ultrasound guidance or fluoroscopy. Due to high mobility of the prostate, some authors recommend stabilizing the prostate position by several “anchor” needles before source implantation (Stone et al. 1995; Feygelman et al. 1996a; Dattoli and Walker 1997; Pedley 2002). One method is to insert a pair of the stabilization needles simultaneously into the central holes in the left and right sides of the template, directed medially into the prostate gland. After that, other needles
592
Eugene P. Lief
are inserted and their position noted on the TRUS image. Needle placement begins from the most anterior, which minimizes ultrasound interference of previously implanted seeds and needles. Needles are inserted on alternate side of the template to reduce possible asymmetry due to hemorrhage and edema. Some needles, especially on the anterolateral edges, may require special care to avoid interference with the pubic arch. Moving the patient’s legs to a more extended dorsal lithotomy position almost always eliminates pubic arch interference. The posterior row of needles should not approach the anterior rectal mucosa closer than 3 to 5 mm. Special care should be taken for the needles to avoid the urethra. If preloaded needles are used, the needles may be inserted to their requisite position and depth and the seed trains deposited sequentially or one or more rows may be inserted prior to seed deposition. If an afterloading approach is used, once the “cold” needles are implanted, the needles’ obturators are removed to see if blood or urine drips out. The Mick Applicator (described in detail later in this chapter in the section “Equipment and Accessories for Permanent Prostate Implants”) is then attached to the needle hub and the template ring is advanced to the surface of the template. A magazine loaded with radioactive seeds is inserted into the applicator’s magazine receptor for sequential insertion of seeds. Seeds are implanted one after another; once the first seed is deposited, the Mick TP Needle is retracted, usually by l cm, and the next seed is transferred and deposited by means of the applicator’s stylet plunger. After the insertion of the required number of seeds, the applicator is detached and the procedure is repeated with the next needle. The number of seeds implanted through each needle depends on the plan chosen. If the plan was calculated prior to the procedure, the patient’s treatment position and the organ locations with respect to the ultrasound probe and the template should be as close to the planned positions as possible. This has to be checked before starting the seed placement. The implant coordinates are defined from the template and the depth of insertion. Due to possible deviations from the planned positions, the implant is visually evaluated before the end of the procedure. If some of the seeds end up at a location different than planned, sometimes it is possible to place a few extra seeds in the originally planned locations. After the procedure, cystoscopy is performed by the urologist, to find and remove seeds that were possible dislodged and ended up in the bladder or urethra. Blood clots could be also removed during the cystoscopy.
Imaging Techniques and Prostate Localization Fluoroscopy Fluoroscopy may be used for imaging of permanent prostate implantation. As fluoroscopic images do not show the prostate, CT– or other 3-D–based planning has to be performed preoperatively. Imaging in the operating room can be performed using a C-arm, which can be rotated for an antero-posterior (AP) or a lateral projection. Although this approach has been largely superseded by ultrasound imaging, we should briefly describe it. Prior to the procedure, a CT scan with a rectal marker and Foley catheter is acquired with 3 or 5 mm spacing of contiguous slices. The radiation oncologist identifies the target on each slice. After digitizing the contours, a physicist, with the help of a special computer program, finds possible needle positions and directions in three dimensions, taking into account the needles’ angles and the location of the urethra and the pubic symphysis on each CT-slice. Using possible needle directions, one can generate needle intersections with the contour of the target on each particular slice and perform planning with the objective of covering the target with the prescription dose while sparing the rectum and urethra. Based on the plan, a physicist can chart the needle directions with respect to the bony anatomy and urethra marker location. In the operating room, the needles have to be inserted according to the plan under fluoroscopy guidance. Before implanting seeds, the radiation oncologist should make sure that the AP and the lateral fluoroscopic images of the inserted needles match the planning needle insertion diagram. After that, the plan can proceed as usual, using the fluoroscopic guidance that can give a good quality image of the seeds and needles but not the prostate. Nowadays, fluoroscopy is used primarily as an adjunct to ultrasound guid-
Treatment Delivery in Prostate LDR Brachytherapy
593
ance. An AP fluoroscopic view activated momentarily during placement of each needle allows visualization of needles gently “tenting” the bladder wall, gives an overview of the developing implant, and helps identify misplaced seeds or those moving out of its planned position.
Ultrasound Ultrasound imaging is the most popular modality for permanent prostate implants (Stock et al. 1995; Nag, Fernandes, and Bahnson 1998). It provides good prostate visualization without any ionizing radiation dose. The ultrasound probe snugly fits into a holder that can move smoothly craniocaudally or in fixed increments of 0.5 cm to scan the prostate. The template rests on the top of the probe holder at a fixed position with respect to it. The metal frame with the probe holder is secured to the table or floor stabilizer to minimize its movement. The position of the probe is adjusted to obtain the prostate image within the grid screen on the monitor. Software within the ultrasound system projects a grid corresponding to template coordinates onto the transverse prostate image. Sometimes ultrasound imaging is used in conjunction with the fluoroscopy. Dilute contrast medium can be inserted into the bladder to assist visualization on fluoroscopy, and an air-filled gel placed in the catheter or by direct filling of the penile and prostatic urethra allows seeing the urethra on ultrasound (Ash et al. 2000). The position of the prostatic apex may be identified on fluoroscopy by means of a retrograde urethrogram.
Magnetic Resonance Imaging and Spectroscopy Magnetic resonance imaging (MRI) is considered to provide the best visualization of the prostate and seeds at the same time. Possible scale distortion does not increase the average error of the source localization [see more information and references in the localization chapter in this book (chapter 12)]. Besides postimplant evaluation (see chapter 12 herein), the MRI can be used intraoperatively for MRI-guided prostate brachytherapy (Kooy et al. 2000). Magnetic resonance spectroscopy (MRS) brings the use of magnetic resonance one step further. By distinguishing several typical lines in the magnetic resonance spectra, one can derive the spatial distribution of the malignant cells based on the magnetic resonance spectra in various portions of the prostate (Zaider et al. 2000). For the clinical use, this technology requires advanced magnetic resonance spectroscopy and detailed information about magnetic resonance spectra of various malignancies.
Cone-Beam CT-Imaging Cone-beam CT-imaging can be installed on linear accelerators (linacs) and simulators such as the Acuity simulator from Varian Medical Systems, Inc. (Palo Alto, CA). Some pioneering work has been already performed using cone-beam CT-simulator imaging for brachytherapy visualization (Prestidge 2004). There are several advantages of this approach. Cone-beam CT-imaging (especially using simulators which offer much greater separation between the imager and treatment head than linacs) provides greater clearance than a conventional CT, so the patient can be scanned in the treatment position. Scanning and reconstruction times are short enough, since the whole area of interest can be scanned in a single turn of the gantry in a few minutes. Usually, simulator rooms are more available for scheduling long brachytherapy procedures than linac rooms. All this makes the simulator rooms preferable locations for brachytherapy procedures because the use of a cone-beam CT will allow the evaluation of the quality of an implant while still being able to make corrections. Intraoperative use of more advanced imaging techniques, such as cone-beam CT-imaging and MRI, has certain drawbacks. The procedure time typically increases compared to the conventional ultrasoundbased technique. The most efficient teams can perform a thorough TRUS- and fluoro-based procedure
594
Eugene P. Lief
with 30 preloaded needles in 30 to 40 minutes, starting from the probe insertion and ending with postoperative cystoscopy (Butler 2005). This time can increase by 10 to 15 minutes, if the Mick Applicator is used instead of the preloaded needles. CT-based procedures typically last 2 hours and more, while MRIbased techniques can further increase the operation time to 3 hours or more. In addition to the increased procedure time, creating sterile environment of the operating room in the cone-beam CT or MRI areas as well as scheduling large blocks of time can be more difficult than using an existing operating room with dedicated C-arm and ultrasound units.
Equipment and Accessories for Permanent Prostate Implants Ultrasound and Accessories The main equipment for the prostate implants typically includes the ultrasound device with the rectal probe, the stepping device carrying the probe, the perineal template (Figure 1) and the stabilizing assembly, which is either floor or couch mounted. The ultrasound machine is typically a portable device with special software that displays the grid pattern on the screen. The stepping device attaches to both the stabilizing assembly and the rectal probe, restricting the motion of the latter to move only in and out of the patient with the precise steps. The needle template has holes for 17 or 18 gauge needles arranged typically in a 13x13 square matrix with 5-mm spacing between numbered rows and columns. Figure 1 shows a disposable template; however, some of the templates are reusable. The reusable templates have to be properly cleaned and sterilized after each procedure.
Mick Applicator™ A special applicator for loading 125I and 103Pd seeds of various models was developed by Felix W. Mick, Mick Radionuclear Instruments, Inc. (Bronx, NY) (Figure 2). The Mick Applicator comprises a body, to which a seed magazine is attached, stepping device, and a stylet plunger. The magazines can be of two kinds: a reusable one that can hold 10 seeds (Figure 3a) or a disposable shielded magazine holding up to 15 seeds (Figure 3b). The magazines can be taken apart for manual seed loading. A small, springloaded device holds the seeds in the magazine tightly together. The stepping mechanism in the Mick Applicator allows retracting the applicator by a fixed distance (usually 1 cm) for the next seed placement. The plunger in the Mick Applicator is used to deposit a seed, once the applicator with the needle arrives at the desired location. The magazines have to be loaded prior to sterilization before the procedure. Typically, this is done in a radioactive source laboratory (“hot lab”) on a day preceding the procedure. The loading pad (Figure 4) can make the manual magazine loading significantly simpler, especially if reverse action tweezers are also used. The work can be performed in a radioactive source laboratory behind a radioprotective transparent screen, such as leaded glass.
Pre-loaded needles and Rapid Strand™ An alternative to Mick Applicator is to use preloaded needles, i. e., the needles that are filled with radioactive seeds and spacers between them. The needle loading should be performed accordingly to the plan. Each needle would contain the required number of seeds, with each of the seeds occupying the specified position, based on the plan recommendations. Missing seeds in the needle can be replaced by the spacers. The sterilized needles can be preloaded before the procedure. The needle tip is plugged with surgical bone wax or rectal suppository (Yu et al. 1999). The loaded needles are kept in a sterilized needle box
Treatment Delivery in Prostate LDR Brachytherapy
595
Figure 2. Mick Applicators™ from Mick Radionuclear, Inc., with attached magazines and 10 cm to 25 cm needles.
Figure 3a. A reusable magazine for Mick Applicator™, for implanting 125I and 103Pd seeds. The magazine can be taken apart for loading, similar to Figure 3b.
Figure 3b. Disposable shielded magazines for Mick Applicator™. The magazine on the left was taken apart for loading, the one in the center is loaded, and the one on the right is empty.
until the operation. During the procedure, each needle with seeds is inserted into the corresponding position in the template, according to the preoperative plan. In 1995 Amersham Healthcare (Arlington Heights, IL) introduced Rapid Strand™, in which the 125I seeds, located 1 cm apart, are enclosed within a stiff Vicryl® suture material. The purpose of this arrangement is to improve the seed fixity after the procedure (Merrick et al. 2000) and to improve the dose distribution by avoiding “seed clusterization.” Vicryl used as a suture material is hygroscopic and softens and swells in the presence of moisture from body fluids. If the swelling occurs before the ejection of the strand, it can jam the implant needle and make expelling it impossible. Therefore, plugging the tip of the loaded needle is critically important. The bone wax, used for the needle loading with conventional seeds and spacers, was too hard for the original version of Rapid Strand, and needles were plugged by dipping into melted suppositories such as Anusol-HC™. This material partially melts at 37° C and allows expulsion in 2 to 3 minutes after the needle is implanted. Newer versions of Rapid Strand are mechanically
596
Eugene P. Lief
Figure 4. “V” block for loading magazines with radioactive seeds has five threaded holes for magazines, a loading slot, a seed well for loose seeds, and a seed gauge for checking the seed dimensions.
strong enough to push through bone wax. In recent years, several other manufacturers have developed products to help reduce seed migration where seeds are squeezed into double female sockets on site or purchased custom-stranded or encased in biocompatible material.
Radioactive Isotopes and Seeds Iodine: 125I 125
I and 103Pd are the isotopes most often used for permanent prostate implants. Seeds of both isotopes are usually encapsulated in 4.5-mm long titanium shells. The outer diameter of the shells is about 0.8 mm. Model 200 103Pd seeds from Theragenics (Buford, GA) have slightly larger diameters of 0.81 mm. Some models of seeds have a unique appearance on radiographs because of differences in the length, density, or number of radio-opaque markers used. The average energy of photons emitted by 125I source is 27.4 keV, and the half-life is 59.4 days. The half value layer (HVL) in lead is only 0.025 mm. Ninety percent of the total dose is delivered in 197 days. Typical air kerma strength of 125I sources used for prostate brachytherapy is 0.4 to 1.0 U, or 0.3 to 0.8 mCi.
Palladium: 103Pd Radioactive 103Pd emits characteristic x-rays of 20.1 keV and 23.0 keV, as well as Auger electrons. The HVL in lead for all types of radiation from the 103Pd isotope is only 0.008 mm. Typical air kerma strength for 103Pd seeds used for prostate brachytherapy is between 1.4 and 3.5 U (1.1 to 2.7 mCi per seed). The half-life of 103Pd is 16.97 days, and 90% of the total dose is delivered in 56 days.
Treatment Delivery in Prostate LDR Brachytherapy
597
Radioactive Seed Assay It is recommended by AAPM Task Group No. 56 (Nath et al. 1997) that the user should assay 10% of the shipped seeds. Discrepancies of 3% or more between the mean of the assay and the manufacturer’s calibration should be investigated. Unresolved discrepancies of 5% or more should be reported to the manufacturer. The most accurate way to perform the assay is to use a well chamber with a suitable seed holder. Some groups assay seeds in bulk, or in cartridges. There are special cartridge holders available for this purpose. In addition to other assay methods, an autoradiograph could be taken with the seeds lying on the ready-pack film at 1 cm or more distance from each other. The purpose of such a radiograph is to visually compare the film density to make sure that the seeds are roughly of the same strength. For assaying sterile seeds in sterile suture, one can order some more seeds from the same batch, just for the assay. Alternatively, a sterile insert to the standard dose calibrator can be used for the direct assay of sterile seeds in suture (Feygelman et al. 1996b; Butler et al. 1998).
Radiation Safety in Permanent Prostate Implants Receiving, Handling, and Loading Radioactive Seeds Before the Procedure Radioactive seeds are ordered in advance and are shipped as radioactive material via licensed carriers. Usually, the small glass vial with seeds is shipped in a small lead container, with walls few millimeters thick (compare with the HVLs listed in the previous section). Using packing material, the lead container is placed in the middle of a larger cardboard box that is used for shipping. After the lead container is received and unpacked, it should be stored in a radioactive source laboratory until the seed assay and loading begins. The lead container should be opened by the physicist only when he or she is behind a shielding screen with lead glass. Due to the relatively low energies of 125I and 103Pd radiation, a plain screen offers sufficient protection. A standard 0.5-mm lead-equivalent apron with thyroid shield also offers protection to most organs at risk. A ring badge has to be worn all the time while working with radioactive sources to accurately monitor the radiation dose to hands. After opening the lead container, it is convenient to tightly pack the container with a piece of foam that comes with it, snugly securing the glass vial inside the open lead container. After that, all the operations with the vial can be performed while it remains inside the container, to minimize the dose. The vial can be opened and slightly tilted over the well of the loading block, so about a dozen seeds can be transferred to the well of the loading pad for assaying and loading in magazines. Magazine housings are shielded, so as soon as a magazine is loaded, it should be put together using a special magazine holder and transferred to the sterilization block (Figure 5). The magazine housing and the block offer significant protection from radiation of the seeds inside the magazine. When all the seeds are loaded, it is useful to compare the positions of the plungers in all magazines, e.g., by tilting the block and visually aligning its handle with the tips of the plungers on one side. The tips should be aligned, if all the magazines are identical and contain the same number of seeds. A lower position of a plunger could mean that it contains fewer seeds than is intended. A higher position could mean loading too many seeds or misalignments of the loaded seeds. The sterilization block is loaded into the sterilization container (Figure 5), wrapped into a sterilization pad, labeled with the patient’s information and a special tag with the radioactivity sign, and delivered into the sterilization area at the operating room. After the sterilization, the wrapped container is brought to the operating room. The radioactivity tag stays attached to the container all the time that it contains seeds. After that, the sign should be either destroyed or returned back to the radioactive source laboratory for the future use.
598
Eugene P. Lief
Figure 5. Seed sterilizer consists of the outer stainless steel case and the inner solid shielding block which could hold up to 10 loaded magazines.
Radiation Survey in the Radioactive Source Laboratory, Operating and Recovery Rooms A radiation survey, using a sensitive survey meter, should follow seed receiving, loading, and implanting. The purpose of the survey is to make sure that there are no seeds lost, and that there is no leakage of radioactivity through a broken seed shell. Since the amount of radioactivity in one seed is very small and confined to a small area, the survey meter should be moved around very slowly, allowing sufficient time for displaying the higher reading in a small area around a loose seed. In the absence of loose seeds, the readings in noncontaminated areas, such as at the surface of the shipping box, should be at the background level (below 0.1 mR/hr). In the operating room, the initial survey should be performed before the seeds are brought in at the beginning of the procedure. After the procedure, the survey starts in the patient area and the stretcher that is about to leave the room with the patient. The presence of the patient with implanted seeds in close proximity will affect the survey meter readings. After the patient leaves the operating room, the whole operation room has to be surveyed, especially the trash bucket, space around the implantation area, and surgical instruments used for the implant. Before discharging the patient, the recovery room should be also surveyed, especially the patient’s bed, stretcher, Foley catheter, urine bag, and toilet. If a loose seed is found, it should be placed in the lead container and brought back to the radioactive source laboratory. The survey meter is also used to measure the radiation dose rate near the pelvis and at a 1-m distance at four points in front, back, and left and right sides.
Discharge Instructions to the Patient After the discharge, the institution is no longer accountable for the radioactive sources implanted. Therefore, before discharging, a medical or a health physicist should give instructions to the patient about the radiation safety aspects of his implant. Due to the low energy of radiation from 125I and 103Pd seeds, the dose rate at the skin is very small. Therefore, a patient may stay in the same bed with his partner and
Treatment Delivery in Prostate LDR Brachytherapy
599
be in the same room with children (Ash et al. 2000); however, it is advised to avoid prolonged close contacts (such as sitting on the lap) with children under 18 years and pregnant women for at least one half-life of the radionuclide (Nag et al. 1999). National Commission on Radiation Protection and Measurements (NCRP) Commentary No. 11, “Dose Limits for Individuals Who Receive Exposure from Radionuclide Therapy Patients” provides additional information in this context (NCRP 1995). Usually, the patient receives a letter (or a small card) from the hospital with information about the implant, medical doctor, and contact information. This document can be produced to explain the radioactivity, which can be detected by screening devices at the airports or nuclear power plants, or seed images that can be found on a subsequent radiological exam. The patient is advised that there is a small chance of a loose seed escaping during urination or ejaculation. Some hospitals provide patients with special kits containing a strainer for urination, tweezers, and a lead pouch. If a loose seed is found, it should be handled with tweezers, placed in a protective lead pouch, and brought back to the hospital during the next visit. According to the AAPM TG-64 recommendations, this requirement is not necessary (Yu et al. 1999). In extremely rare cases, a seed can be expelled in the semen during ejaculation. Although the risk of radiation injury from a single seed is minimal, some centers advise patients to abstain from intercourse for 1 to 2 months, or to wear condoms, or masturbate for the first few ejaculations to prevent the unlikely passage of a seed into the partner (Nag et al. 1999). Patients should be advised that prostate brachytherapy may affect potency and fertility, but that pregnancy remains possible. Patients should be advised that there is a low risk of a seed migration to the lung, but this has not been shown to have any clinically adverse effects (Gupta, Nag, and Gupta 1993; Steinfeld, Donahue, and Plaine 1991). The non-fixed seeds take approximately 1 day to migrate to the lungs (Nag, Vivekanandam, and Martinez-Monge 1997). Therefore, a chest radiograph is advised at the first follow-up visit (rather than on the day of implant) to scan the lungs for embolized seeds. The patient should be advised of their presence if detected (Nag et al. 1999).
Acknowledgment Figures 1–5 are used with permission of Felix W. Mick, Mick Radionuclear, Inc., Bronx, NY.
References Anderson, L. L. (1976). “Spacing nomograph for interstitial implants of 125I seeds.” Med Phys 3:48–51. Anderson, L. L., J. V. Moni, and L. B. Harrison. (1993). “A nomograph for permanent implants of palladium-103 seeds.” Int J Radiat Oncol Biol Phys 27:129–136. Ash, D., A. Flynn, J. Battermann, T. de Reijke, P. Lavagnini, and L. Blank. (2000). “ESTRO/EAU/EORTC recommendations on permanent seed implantation for localized prostate cancer.” Radiother Oncol 57:315–321. Blasko, J. C., H. Ragde, and P. D. Grimm. (1991). “Transperineal ultrasound-guided implantation of the prostate: morbidity and complications.” Scand J Urol Nephrol Suppl 137:113–118. Blasko, J. C., H. Ragde, and D. Schumacher. (1987). “Transperineal percutaneous iodine-125 implantation for prostatic carcinoma using transrectal ultrasound and template guidance.” Endocuriether/Hypertherm Oncol 3:131–139. Butler, W. M., A. T. Dorsey, K. R. Nelson, and G. M. Merrick. (1998). “Quality assurance calibration of I-125 Rapid Strand in a sterile environment.” Int J Radiat Oncol Biol Phys 41:217–222. Butler, W. M. (2005). Personal communication. Charyulu, K. K. N. (1980). “Transperineal interstitial implantation of prostate cancer: A new method.” Int J Radiat Oncol Biol Phys 6:1261–1266. Dattoli, M., and K. A. Walker. (1997). “Simple method to stabilize the prostate during transperineal prostate brachytherapy.” Int J Radiat Oncol Biol Phys 388:341–342.
600
Eugene P. Lief
Feygelman, V., J. L. Friedland, R. M. Sanders, B. K. Noriega, and J. M. Pow-Sang. (1996a). “Improvement in dosimetry of ultrasound-guided prostate implants with the use of multiple stabilization needles.” Med Dosim 21(2):109–112. Feygelman, V., B. K. Noriega, R. M. Sanders, and J. L. Friedland. (1996b). “A simple method for verifying activity of iodine-125 seeds in rigid absorbable suture.” Med Dosim 21:261–262. Flocks, R. H., H. D. Kerr, H. B. Elkins, and D. Culp. (1952). “Treatment of carcinoma of the prostate by interstitial radiation with radioactive gold (Au-198): A preliminary report.” J Urol 68:510–522. Fowler, J. E., W. Barzell, B. S. Hilaris, and W. F. Whitmore. (1979). “Complications of 125iodine implantation and pelvic lymphadenectomy in the treatment of prostatic cancer.” J Urol 121:447–451. Gupta, S., S. Nag, and J. Gupta. (1993). “Pulmonary embolism of permanently implanted radioactive iodine-125 seeds.” Endocuriether/Hypertherm Oncol 9:27–31. Holm, H. H., N. Juul, J. F. Pedersen, H. Hansen, and I. Str?yer. (1983). “Transperineal 125iodine seed implantation in prostatic cancer guided by transrectal ultrasonography.” J Urol 130:283–286. Kooy, H. M., R. A. Cormack, G. Mathiowitz, C. Tempany, and A. V. D’Amico. (2000). “A software system for interventional magnetic resonance image-guided prostate brachytherapy.” Comput Aided Surg 5:401–413. Kumar, P. P., and F. F. Bartone. (1981). “Transperineal percutaneous I-125 implant of prostate.” Urology 130:283–286. Lee, E. K., and M. Zaider. (2003). “Intraoperative dynamic dose optimization in permanent prostate implants.” Int J Radiat Oncol Biol Phys 56(3):854–861. Merrick, G. S., W. M. Butler, A. T. Dorsey, and H. L. Walbert. (1997). “Prostatic conformal brachytherapy: 125I/103Pd postoperative dosimetric analysis.” Radiat Oncol Investig 5(6):305–313. Merrick, G. S., W. M. Butler, A. T. Dorsey, J. H. Lief, and M. L. Benson. (2000). “Seed fixity in the prostate/periprostatic region following brachytherapy.” Int J Radiat Oncol Biol Phys 46(1):215–220. Nag, S. (1985). “Transperineal iodine-125 implantation of the prostate under transrectal ultrasound and fluoroscopic control.” Endocuriether/Hypertherm Oncol 1:207–211. Nag, S., P. S. Fernandes, and R. Bahnson. (1998). “Transperineal image-guided permanent brachytherapy for localized cancer of the prostate. Review article.” Urologic Oncol 4:191–202. Nag, S., S. Vivekanandam, and R. Martinez-Monge. (1997). “Pulmonary embolization of permanently implanted radioactive palladium-103 seeds for carcinoma of the prostate.” Int J Radiat Oncol Biol Phys 39:667–670. Nag, S., D. Beyer, J. Friedland, P. Grimm, and R. Nath. (1999). “The American brachytherapy society (ABS) recommendations for transperineal permanent brachytherapy of prostate cancer.” Int J Radiat Oncol Biol Phys 44:789–799. Nag, S., W. Bice, K. DeWyngaert, B. Prestidge, R. Stock, R., and Y. Yu. (2000). “The American brachytherapy society recommendations for permanent prostate brachytherapy postimplant dosimetric analysis.” Int J Radiat Oncol Biol Phys 46:221–230. Nath, R., L. L. Anderson, J. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24:1557–1598. Also available as AAPM Report No. 59. National Council on Radiation Protection and Measurements (NCRP) Commentary No. 11, Dose Limits for Individuals Who Receive Exposure from Radionuclide Therapy Patients. Bethesda, MD: NCRP, 1995. Pasteau, O., and P. Degrais. (1914). “The radium treatment of cancer of the prostate.” Arch Roentgenol Ray 28:396–410. Pedley, I. D. (2002). “Transperineal interstitial permanent prostate brachytherapy for carcinoma of the prostate.” Surg Oncol 11:25–34. Prestidge, B. P. (2004). “Early Experience with Cone-Beam CT for Brachytherapy.” Materials of 8th Annual International Conference and Workshop “New And Future Developments In Radiotherapy,” sponsored by Wayne State University, School of Medicine, Detroit, MI. November 12–14, 2004, San Diego, CA. Roy, J. N., K. E. Wallner, P. J. Harrington, C. C. Ling, and L. L. Anderson. (1993). “A CT-based evaluation method for permanent implants: application to prostate.” Int J Radiat Oncol Biol Phys 26(1):163–169. Steinfeld, A. D., B. R. Donahue, and L. Plaine. (1991). “Pulmonary embolization of iodine-125 seeds following prostate implantation.” Urology 37:149–150.
Treatment Delivery in Prostate LDR Brachytherapy
601
Stock, R. G., N. N. Stone, M. F. Wesson, and J. K. DeWyngaert. (1995). “A modified technique allowing interactive ultrasound-guided three-dimensional transperineal prostate implantation.” Int J Radiat Oncol Biol Phys 32(1):219–225. Stone, N. N., R. G. Stock, J. K. DeWyngaert, and A. Tabert. (1995). “Prostate brachytherapy: Improvements in prostate volume measurements and dose distribution using interactive ultrasound guided implantation and threedimensional dosimetry.” Radiat Oncol Invest 3:185–195. Todor, D. A., M. Zaider, G. N. Cohen, M. F. Worman, and M. J. Zelefsky. (2003). “Intraoperative dynamic dosimetry for prostate implant.” Phys Med Biol 48:1153–1171. Vicini, F. A., V. R. Kini, G. Edmundson, G. S. Gustafson, J. Stromberg, and A. Martinez. (1999). “A comprehensive review of prostate cancer brachytherapy: Defining an optimal technique.” Int J Radiat Oncol Biol Phys 44:483–491. Whitmore, W. F., Jr., B. Hilaris, and H. Grabstald, H. (1972). “Retropubic implantation of iodine-125 in the treatment of prostatic cancer.” J Urol 108:918–920. Whitmore, W. F., Jr., B. Hilaris, and H. Grabstald, H. (2002). “Retropubic implantation of iodine-125 in the treatment of prostatic cancer. 1972.” J Urol 167(2Pt2):981–983; discussion 984. Yu,Y., L. L. Anderson, Z. Li, D. Mellenberg, R. Nath, M. C. Schell, F. M. Waterman, A. Wu, and J. C. Blasko. (1999). “Permanent prostate seed implant brachytherapy: Report of the American Association of Physicists in Medicine Task Group No. 64.” Med Phys 26(10):2054–2075. Also available as AAPM Report No. 68. Zaider, M., M. J. Zelefsky, E. K. Lee, K. L. Zakian, H. I. Amols, J. Dyke, G. Cohen, Y. Hu, A. K. Erdi, C. Chui, and J. A. Koutcher. (2000). “Treatment planning for prostate implants using magnetic-resonance spectroscopy imaging.” Int J Radiat Oncol Biol Phys 47:1085–1096. Zelefsky, M. J., Y. Yamada, G. Cohen, E. S. Venkatraman, A. Y. C. Fung, E. Furhang, D. Silvern, and M. Zaider. (2000). “Postimplantation dosimetric analysis of permanent transperineal prostate implantation: Improved dose distributions with an intraoperative computer-optimized conformal planning technique.” Int J Radiat Oncol Biol Phys 48(2):601–608.
Chapter 31
Post Implant Evaluation William S. Bice, Jr., Ph.D. University of Texas Health Science Center at San Antonio International Medical Physics Services San Antonio, Texas Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604 Films for Record . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604 Paired Film Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605 Meaningful Post-Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605 Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605 Plane Radiography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605 Computed Tomography (CT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606 Magnetic Resonance Imaging (MRI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606 Other Imaging Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607 Image Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607 Image Timing: Edema . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608 Source Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608 Calculations and Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608 Calculation Formalism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 608 The Dose Evaluation Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 611 Isodoses and Gridded Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 612 Dose-Volume Histograms (DVHs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613 Structure- and Nonstructure-Based Histograms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613 Integral Dose-Volume Histograms (IDVHs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613 Differential Dose-Volume Histograms (DDVHs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615 Natural Dose-Volume Histograms (NDVHs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615 Dose-Surface Histograms (DSHs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 616 Dose Traces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 616 DVH Sampling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618 Quantifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618 Coverage Quantifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618 Uniformity Quantifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 621 Critical Structure Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622 Dosimetric Uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 622 Dynamic Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625 Advanced Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625 Subglandular Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 625 Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626 Correlation of Clinical Results to Post-Implant Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 626 Cures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 628 Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 628 Post-Implant Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 630 The Post-Implant Evaluation Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 630 Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 630 Program Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 631 Multicenter Trials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 632 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 632 Appendix. Sample Reporting Format for Permanent Prostate Brachytherapy . . . . . . . . . . . . 633 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635
603
604
William S. Bice, Jr.
Introduction Only the most irresponsible implant team would enter an operating room without a plan. This plan could be a formal one, calculated, printed, and approved by the physician, or could be as informal as an envisioned ideal procedure, based on the anatomy of the patient and the practiced skill of the brachytherapist. But, no matter how skilled the physician, how practiced the team, how ideal the patient, the implant never, ever matches the plan. Properly performed afterloading techniques, e.g., high dose-rate (HDR) procedures, are less affected by this discrepancy than implants performed entirely during the operative procedure, e.g., permanent prostate brachytherapy. In the case of seed implants this has been the impetus for in-OR (inoperating room) treatment planning and in-OR dosimetry. Post-operative evaluation of the implant can then be defined as narrowly as that performed after the entire procedure or as broadly as any analysis performed after placement of the delivery apparatus, for example, afterloading catheters or gynecological applicators. In this chapter the analysis is restricted to the more narrow definition: post-implant evaluation is analysis performed after the implant has been completed and the radiation dose delivered (or seeds have been implanted). Dosimetry and design performed after the applicators are in place is considered part of the implant procedure, and has been described in other chapters of the monograph. So if we restrict post-implant evaluation to the analysis performed after the procedure is finished, why do it? The radiation dose cannot be taken back; looking afterward only invites trouble! There are two compelling reasons for conducting post-implant evaluation. Clearly, the quality of the implant needs to be evaluated to determine if there is any need for additional treatment or if unwanted complications might result. Just as importantly however, the implant should be evaluated in terms of the quality of the procedure itself: a measure of the performance of the implant team. This is more than an evaluation of the skill of the physician. Post-implant evaluation should include the adequacy of the implant design, skill of the operating room staff, and appropriateness of the equipment used in the procedure. Nothing contributes to making the next implant better like reviewing the last twenty. Most post-implant evaluation is performed to assess permanent prostate brachytherapy (PPB) procedures. This will, of necessity then, be the slant of this chapter. It does not mean that the same methods cannot be used to evaluate other types of implants. What procedure would not be improved by a detailed review of dose delivered, the patient outcome, and the techniques employed? But it is in PPB that the need for formal post-implant planning and procedural evaluation first became apparent. The gains were immediate and dramatic. A series of brachytherapy “tools” were developed to aid in, or due to, post-implant analysis for PPB. Ultrasound and computed tomography (CT)-based three-dimensional (3-D) imaging, automated source identification and localization, refinement and development of implant quantifiers, electronic transfer formats, and standards for brachytherapy all came about because of the success and popularity of PPB. The street is traveled in two directions: no treatment modality has seen such rapid, widespread improvement due to the development of quality evaluation tools. PPB and post-implant evaluation were made for each other; each getting better, feeding off the successes of the other.
Background Films for Record Fifteen years ago meaningful post-implant dosimetry did not exist. The best that most clinical physicists could offer was something called “films for record”. This usually consisted of a single radiograph, or at best a pair of radiographs, taken on a simulator, which showed little more than source locations and bony anatomy. These we dutifully filed in the patient’s film jacket along with the physician’s note, confident that we had, indeed, performed the implant. How depressingly inadequate!
31–Post Implant Evaluation
605
Paired Film Localization Of course, one could perform source localization from a pair of films or, even better, a triplet. Most treatment-planning systems, even those primarily dedicated to external-beam planning (this was all of them during this time) had the capacity to localize points and sources from a paired film set. This was primarily to satisfy the demand for gynecological brachytherapy planning; five or six large sources in a more or less fixed geometry with a relatively fixed set of four or five calculation points. Using this system to identify and localize a prostate implant with 100 seeds was euphemistically described as burdensome. Nevertheless, it was possible, and one could generate isodose planes, or even clouds, to depict the distribution of radiation dose. Without anatomical reference, of course, the information from paired-film dosimetry is really of little more use than from films for record. More information to file in the patient’s record, this time convincing ourselves that, not only did we perform the implant, but there was also radiation involved. This is not to say that source localization from paired films is useless. Paired film localization is still in use today and, rightfully, a topic in this book. But the best use of this technique has become in conjunction with another imaging modality, usually tomographic—one that depicts soft tissues. Fusing the source information from the paired radiographs with the soft-tissue imaging modality marries the dose and structures, essential for determining the quality of the implant.
Meaningful Post-Planning Post-planning then is a function of the imaging modality and the treatment-planning system. Post-implant evaluation performed without meaningful post-planning is an empty process, drawing conclusions with no data. Characteristics of meaningful post-planning include: 1. the use of tomographic imaging modalities with automated source localization capabilities; 2. performance of 3-D calculation, analysis and, preferably, display; 3. calculation and analysis based upon localized structures—targets and critical structures; 4. a calculation grid size that is appropriately small enough to distinguish dosimetric changes over either small volumes of interest or in areas of rapidly changing dose; 5. structure-based dose-volume or dose-surface analysis with sufficient sampling of dose within the structure to meet statistical requirements; and 6. implant quantifier determination.
Imaging Plane Radiography Fluoroscopy and radiography have been used in the post-implant setting. The lack of anatomical information is clearly a significant drawback. As described above paired films or, alternatively, multiple-projection fluoroscopy can be used to determine source locations (Tubic et al., 2001; Tutar et al., 2003; Narayanan et al., 2004; Su et al., 2004). A plane radiograph can always be taken to verify the number of sources in an implant. This is especially important for prostate brachytherapy, where the sources may migrate within, sometimes eliminated from, the patient (Merrick et al. 2000a). Visibility of the sources depends on the source configuration and the marker type. Plane films may be used in conjunction with modalities where sources appear on multiple images and the source count may be required to reliably
606
William S. Bice, Jr.
reduce the available source positions to true source locations. Seed sorting will be discussed later in this chapter.
Computed Tomography (CT) The first published use of CT for post-implant dosimetry was in the early 1990s at Memorial Sloan-Kettering Cancer Center (Roy et al., 1993b). Structures and sources were localized on the axial image sets and this information fed into the in-house planning system. This was the first display of organ and dose location together for post implant dosimetry. It soon became apparent that the concept of designing implants based upon matched peripheral dose (MPD), where choosing a reference dose volume to match the volume of the gland, was inadequate. Post-implant evaluation quickly achieved its intended purpose: provide enough feedback to the implant team to allow them to change the design and delivery system to perform better implants. By far, the majority of the post-implant imaging performed today is CT based. CT units are ubiquitous in radiotherapy, the expense is relatively low, and the source locations are easily visualized. Almost the entire experience in the medical community with evaluating the quality of post-implant dosimetry is based upon CT. The lack of ability to visualize the glandular borders clearly, particularly in the edematous post-implant patient, is disconcerting. Large uncertainties exist in the results because of the uncertainties in glandular borders. The variance of the post-implant dosimetry due to the inability to clearly delineate the prostate gland appears to be related to implant quality; better implants have less dependence on visualization of the gland (Bice et al., 1998; Merrick et al., 1999b; Lee et al., 2002; Crook et al., 2002; Lindsay, Van Dyk, and Battista 2003; Han et al., 2003b). The imaging sequence used to generate the tomographic CT-image set varies widely among practitioners. Scan parameters are usually chosen based upon the ability to best determine source locations. This means a reduced field of view, usually about 15 cm, just large enough to encompass the meaningful dose regions and the critical structures. Slice spacing varies from 1 mm to 5 mm; slice width also varies from that which provides abutting slices to something smaller (1-mm slices at 5-mm spacing, for example). When detecting 4.5-mm sources with “random” orientations, 5-mm spacing between slices is almost certainly under sampling, resulting in unwanted uncertainties with regard to localization and subsequent dosimetry. The reasons for such large spacing, elimination of the source location redundancy problem (seed sorting), and general time savings with regard to contouring and location of the sources on each slice, have been adequately addressed with modern planning systems (Brinkman and Kline 1998; Bice et al., 1999; Li et al., 2001; Liu et al., 2003; Davis et al., 2004; Lam et al., 2004; Holupka et al., 2004). Automated detection and seed sorting, combined with the ability to interpolate contoured structures obviates the need for the introduction of such uncertainties. Although no rigorous study pertaining to the choice of CT imaging sequence has been published, source location uncertainties can dramatically affect the dosimetric analysis and possibly the biological outcome (Lindsay, Van Dyk, and Battista 2003). Slice spacing greater than 3 mm is not recommended (Nag et al., 2000). The effect on source location of acquiring the CT image set with helical scans as opposed to axial scans is similarly unstudied but seems to make little difference clinically.
Magnetic Resonance Imaging (MRI) CT provides excellent visualization of the sources, but very poor delineation of the soft tissue boundaries. Magnetic resonance imaging (MRI), with excellent soft tissue contrast, has also been used for post-implant imaging (Moerland et al., 1997; Dubois 1999; Dubois et al., 1998). The ability to visualize glandular borders, particularly at the base and apex of the gland, is of significant advantage. A pelvic coil is much preferred to an endorectal coil, as the latter tends to cause great discomfort in the post-operative patient.
31–Post Implant Evaluation
607
For MRI, several authors have published viable imaging sequences (Dubois 1999; Amdur et al., 1999; Kooy et al., 2000; McLaughlin et al., 2002; Polo et al., 2004). A reduced field of view, similar to CT, is always used. Scans may be acquired with sequences that provide T1- or T2-weighted images. Care must be taken to ensure image distortion in MR data acquisition is kept to a minimum. A comprehensive quality assurance (QA) program is a must. Dosimetry based upon MRI alone, however is probably inadequate due to the inability to distinguish source locations—which do not image—from other objects which show up as holes: blood vessels, calcifications, spacers. Source localization methods based on MRI alone tend to reduce the distance between implanted sources, drawing the isodose volumes in toward the center of the implant (Dubois et al., 1997; Prete et al., 1998). Dubois and Davis have done some work with regard to developing an MR-visible source by doping the source with paramagnetic elements (Dubois 1999, Davis et al., 2004).
Other Imaging Modalities Glandular borders are readily imaged with ultrasound. Tomographic image sets, collected in either axial or parasagittal directions, can be used to build up 3-D images. The use of a rectal ultrasound probe is uncomfortable, of course, and the use of ultrasound for post-implant imaging has been mostly limited to acquisition in the OR following the procedure while the patient is still anesthetized. The real issue with regard to using ultrasound for post-implant dosimetry is the inability to perceive source locations on the ultrasound image. One manufacturer sells an echogenic source, but reliable post-implant dosimetry from an ultrasound image set using this source is not currently viable (Han et al., 2003a). Transurethral ultrasound has shown promise for delineation of the prostatic borders and seed localization but is currently limited to phantom studies (Holmes, Davis, and Robb 2001). There has been some preliminary work done with positron emission tomography (PET) imaging of the sources (Sajo 2003). Although PET images metabolic activity, it has not been shown to be useful in prostate cancer. Single photon emission computed tomography (SPECT) using 111In has had some success in imaging the disease both locally and as a measure of nodal or metastatic involvement (Sodee et al., 1998; Texter and Neal 1998; Fang et al., 2000; Ellis et al., 2003; Ellis, Kim, and Foor 2004; DeWyngaert et al., 2004). Additionally, anatomic information is only marginally visible on PET or SPECT images; fusion with CT is regularly used to localize the metabolic information. MR spectroscopy (MRS) offers both anatomic information and localization of the disease (Zaider et al., 2000; Mizowaki et al., 2002; Claus, Hricak, and Hattery 2004). Again though, accurately visualizing the sources is difficult—requiring fusion of the MR and MRS image sets to CT.
Image Fusion Various image fusion methods have been used in post-implant imaging. The idea is to co-register the image data from a modality in which the sources are clearly visible, like CT or fluoroscopy, with another modality that images soft tissue well, such as MRI or ultrasound. The advantages are clear; the major drawbacks are cost and patient trauma. Of the available modalities, CT-MR fusion has been most extensively studied for post-implant dosimetry. See figure 3. Co-registration of the image sets has been performed by eye (Amdur et al., 1999) or by other more quantifiable and reproducible methods, such as available source locations (Dubois, Bice, and Prestidge 2001) or mutual information (McLaughlin et al., 2002; Tsai et al., 2004). Combining pre-implant volume study images with post-implant CT images has been employed by the Seattle group to help delineate the gland on CT (Badiozamani et al., 1999; Smith et al., 2003), but there is some debate as to the utility of this type of image fusion as edema resulting from the operative procedure clearly changes the structure of the prostate. Use of the pre-implant volume study to draw the gland on a post-implant CT
608
William S. Bice, Jr.
clearly underestimates the size of the gland, but the quantitative affect on coverage quantifiers appears to be small. This is shown in Figure 1. Ultrasound-CT fusion has been used successfully for post-implant evaluation and to perform planning for subsequent salvage therapy (Narayana et al., 1997; Bice et al., 2000). Ultrasound has also been fused to fluoroscopic projection imagery for prostate brachytherapy (Gong et al., 2002). The cost of and time associated with acquiring and fusing two different image sets currently prohibits the use of image fusion on a routine basis, although recent developments in CT-US fusion may change this thinking. Several programs have found the use of CT-MR fusion valuable, performed in a selected number of cases, as a learning tool. The brachytherapist may use the MR to determine what changes need to be made in his CT-image-based contouring techniques (McLaughlin et al., 2002).
Image Timing: Edema Various investigators have described the edematous change in the prostate gland during the operative procedure and post-implant resolution of the edema. Consequently, timing of post-implant image acquisition affects dosimetry. Imaging performed immediately post implant, also known as “day-0 imaging,” results in lower doses being depicted on the post-implant treatment plan than imaging performed at any time thereafter. The typical change is shown in Figure 2. Several authors have shown that the most representative time to image the patient is 2 to 3 weeks following the procedure. Nevertheless day-0 imaging is still performed at many institutions, the advantages being immediate feedback and convenience for the patient. At institutions where a catheter is placed during the imaging procedure in order to be able to better localize the urethra, day-0 imaging precludes an additional catheter placement. Whatever time is chosen to perform the imaging procedure, it should be consistent. Analyzing a series of implants where the postimplant imaging procedure was performed at different times adds an unnecessary level of complexity and uncertainty to the results (Prestidge et al., 1998; Merrick et al., 1998; Willins and Wallner 1998; Yue et al. 1999a,b,c; Butler et al., 2000b; Speight et al., 2000; Waterman and Dicker 1999, 2000; Dogan et al., 2002).
Source Localization Automated localization of sources in the tomographic image set, particularly for permanent prostate brachytherapy, has become essential for post-implant analysis. Covered elsewhere in this book, the source detection routine is divided into two parts: detection of source images on each slice and, since the sources may appear on more than one slice, reduction of the source image locations from all of the slices into possible source positions. The second task, also known as “seed sorting,” is by far the more onerous to perform by hand. Multiple authors have provided automated solutions to both the problem of source image detection and seed sorting. The advantages of automation are not just limited to time savings. Having a logical and reproducible method of interpreting the information from the image set is paramount (Brinkman and Kline 1998; Bice et al., 1999; Li et al., 2001; Liu et al., 2003; Davis et al., 2004; Lam et al., 2004; Holupka et al., 2004).
Calculations and Analysis Calculation Formalism In February 1995, the recommendations of the AAPM task group formed to address problems with interstitial dosimetry were published (Nath et al., 1995). Commonly known simply as TG43, these recommendations provided a straightforward formalism for performing interstitial source dosimetry calcu-
31–Post Implant Evaluation
609
Figure 1. Change in coverage quantifiers with change in glandular borders.
Figure 1.a. V100.
Figure 1.b. D90. Figure 1. The effect of misestimating gland size on post-implant imagery is shown in Figures 1a and 1b. In both figures the glandular borders for 50 implants were expanded and contracted by 2-mm increments, except in the posterior direction where the glandular border is most visible. In Figure 1a V100 is plotted, showing less change in V100 if the gland is drawn too small as compared to being drawn too large. A similar effect is shown on D90 in Figure 1b.
610
William S. Bice, Jr. Figure 2. Change in volume and implant coverage due to edema.
Figure 2.a. Change in volume with resolution of edema.
Figure 2.b. Change in coverage with resolution of edema. This is data from 20 iodine monotherapy patients. The volume of the gland is reduced as the edema from the implant procedure resolves. The coverage of the gland correspondingly improves. Integrating the dose over the life of the implant indicates that the most representative time to image the gland is about day 25. Also note that these implants were performed early in the transperineal implant era. The coverage values shown above would, at best, be considered only marginally acceptable today. [Reprinted from Int J Radiat Oncol Biol Phys, vol. 40, “Timing of computed tomography-based postimplant assessment following permanent transperineal prostate brachytherapy,” B. R. Prestidge, W. S. Bice, E. J. Kiefer, and J. J. Prete, pp. 1111–1115. © 1998, with permission from Elsevier.]
31–Post Implant Evaluation
611
lations, published an historical analysis and a set of consistent dose calculation parameters for each known interstitial source type, and provided a framework for future changes in calibration and calculation methods and results. Subsequently, an update, known as TG43U1, was published which precluded the mixing of line and point source measurements and calculations, established a mechanism for determining calculation parameters from different data sets, and extended the number of sources characterized according this revised formalism (Rivard et al., 2004). TG43 and TG43U1 are covered in detail elsewhere in this monograph. It suffices to state that the publication of these two sets of recommendations had a profound impact on brachytherapy in general and that post-implant analysis should only be performed using these calculation protocols.
The Dose Evaluation Hierarchy Table 1 shows a hierarchy of tools that are used in evaluation of post-implant dosimetry. Each will be discussed in some detail. As shown, moving down the table each tool sacrifices information for simplicity. At the top of the hierarchy are isodose curve displays. Isodose curves retain all the dose and spatial information in an implant. Regions of low and high dose are readily seen and localized. At the next level, dose-volume histogram (DVH) displays have reduced the information available by limiting the spatial information to simply being inside the structure, but have simplified the dosimetric information within the implant to a single two-dimensional (2-D) plot. Line plots or dose traces are a special case for linear type structures where spatial information can be retained in a 2-D plot. At the lowest level of the hierarchy are dosimetric quantifiers. Quantifiers are the result of distilling information from the DVH to represent a characteristic of the implant as a single value. Quantifiers are simple to use and compare, but the information from a single quantifier is really very sparse. Keeping in mind that the purpose of post-implant evaluation is really twofold, one can see how the dose evaluation hierarchy has grown to fit nicely into its purpose. For single implant analysis, most of the time in an evaluation should be spent at the top levels of the hierarchy—the isodose displays and the DVH. Adequate dose to specific locations within the gland, like the site of positive biopsies, can be determined.
Table 1. The Dose Evaluation Hierarchy Modern treatment planning systems offer a variety of tools with which to analyze the dose from an implant. Isodose curves and surfaces contain all the spatial and dosimetric information, dose-volume histograms sacrifice the spatial information, quantifiers are an attempt to describe a characteristic of the implant with a single value. Single implants are usually analyzed at the top of the hierarchy with display isodose structures, group analysis is usually performed at the bottom of the hierarchy with quantifier evaluation.
612
William S. Bice, Jr.
The locations of high-dose regions in critical structures that might cause complications become apparent. Judging the sufficiency of a specific implant is best performed at the top of the hierarchy. Evaluating the effectiveness of the implant team and the implant procedure is the subject of more than a single implant, often a great number of implants. This is too much information to glean from the tools at the top of the hierarchy; hence this sort of analysis is best performed with implant quantifiers, located at the bottom. Looking at a trend in coverage, for instance, can be falsely colored by a single bad implant. Compiling coverage quantifiers from the last 50 implants gives a pretty clear picture of how the team has progressed over time or the true effect of a procedural or equipment change.
Isodoses and Gridded Calculations Isodose values can be shown as curves on a single slice or set of slices as shown in Figure 3 or as 3-D surfaces as shown in Figure 4. The use of tomographic image sets allows the calculation and display of the dose distribution registered to the structures of interest: the target volume and any critical structures. Brachytherapy calculations must be performed in three dimensions. To display isodose lines or surfaces, computerized calculations are performed on a regular grid, with voxel size determined by the grid spacing. Several authors have recommended this grid spacing for brachytherapy be no larger than 2 mm (Yu et al., 1999; Nag et al., 2000). Figures 5a, 5b, and 5c show the effect of grid size selection on a simple dose distribution generated using sources and spacing commonly found in permanent brachytherapy.
Figure 3. Isodose curves on a MR-CT fused image set. Six consecutive, from base towards the apex, axial slices from a prostate post plan depicting delivered dose overlaid upon an MR image set in which the prostate, urethra, and rectum have been outlined. The source locations are from an axial CT data set. The image sets were fused using an iterative least squares fit to available source positions. The reference dose for this implant was 144 Gy. (Dubois, Bice, and Prestidge 2001)
31–Post Implant Evaluation
613
Figure 4. Three dimensional isodose surface. Shown is the reference isodose surface on an implant where coverage of the base and anterior gland appears to be inadequate.
Dose-Volume Histograms (DVHs) Structure- and Nonstructure-Based Histograms Dose-volume histograms (DVHs) are used in brachytherapy to display graphical representations of radiation dose to structures. With the exception that the dose values calculated pertain to volume elements that lie somewhere within the structure for which the calculation is performed, spatial information is lost in the DVH. In the DVH, volume is usually the dependent variable and is plotted against dose, usually the independent variable. When convenient, percent volume may be shown as opposed to volume. The DVH for the prostate gland for an implant is shown in Figure 6. The low-dose region at the anterior base can be seen as a drop off of the curve below 100% of the volume at a dose less than the prescription or reference dose (vertical line). The specific location of high-dose and low-dose regions within the gland is not discernible in the DVH. Less occasionally, the DVH may be presented as nonstructure based. This is a DVH of the entire calculation volume. Whereas this presentation was used extensively prior to being able to determine target locations in the image sets, it is rarely used now. Analysis is predominately restricted to structure-based DVH, even if there are gross uncertainties in the precise location of the structure. Integral Dose-Volume Histograms (IDVHs) Figure 6 shows an integral or cumulative DVH (IDVH). The ordinate for each point on the curve represents the volume or percent of volume of the structure that achieved greater than or equal to the dose on
614
William S. Bice, Jr. Figure 5. Effect of grid size on isodose calculation and display.
Figure 5a. Slice 0 of the RTOG prostate phantom calculated with a 4-mm grid size. This is a common default resolution for external beam radiotherapy planning systems.
Figure 5b. The same slice but with a 2-mm grid size.
Figure 5c. The same slice with a 1-mm grid size. Note the marginal improvement over the 2-mm calculation grid. This figure illustrates the effect that grid size has on isodose calculation and display. A 2-millimeter grid size has been considered an adequate compromise between speed and accuracy and is recommended as the largest grid appropriate for prostate brachytherapy. Most modern treatment planning systems easily achieve rapid calculation times even with grid sizes as small as 0.5 mm.
the abscissa. For example in the DVH shown, 50% of the prostate received at least 180 Gy. The IDVH is by far the most prevalent DVH display; practitioners commonly refer to the integral DVH as simply the DVH, reserving the terms integral or cumulative only when required to differentiate between other forms of dose-volume display.
31–Post Implant Evaluation
615
Figure 6. Dose-volume histogram. This is the dose-volume histogram, or integral dose-volume histogram, for the prostate gland following an iodine monotherapy implant. Percent volume of the gland that achieved that dose or greater is plotted against the dose in Gy. The reference dose of 144 Gy is shown as a vertical line. V100 for the implant was 97%. Also shown are the dose-surface histograms for the urethra and the rectum.
Differential Dose-Volume Histograms (DDVHs) One alternative form of the DVH is the differential DVH (DDVH). The DDVH for the prostate IDVH of Figure 6 is shown in Figure 7. In the DDVH, each point on the curve represents the volume or percent of volume of the structure that received the dose value encompassed by the bin values, neither more nor less. The DDVH in Figure 7, for example, shows that 0.32 cm3 of prostate volume received a dose between 179 and 180 Gy (the bin range for the 179.5 Gy bin). Because the plot was normalized, this information is obtained from the inset box. Also shown in this figure is the DDVH for the entire calculation volume. Without a structure in the image set, this would have been all that was available to the early practitioners of brachytherapy dosimetry. Note the large volumes in the calculation volume DDVH. The low doses overwhelm the high-dose regions, largely due to the inverse square law. Hence there exists the need for normalizing this plot to the largest value for each curve. How frustrating it must have been for these early practitioners to realize that the information of interest, the information about the target region, is in the calculation volume DDVH, but the structure of the curve is hidden by displaying the low-dose regions. Natural Dose-Volume Histograms (NDVHs) A third type of DVH is used in brachytherapy, particularly in situations where the target volume is unknown or not displayed in the image set. The natural DVH (NDVH) was developed by Anderson to remove the inverse square law effect from the DDVH (Anderson 1986). Instead of graphing simple volume in the DDVH, the NDVH plots volume divided by dose raised to the -3/2 power against dose. This has the very desirable effect of enhancing the higher-dose regions of the dose display while depressing the lower-dose regions. In the simple DDVH the graph is overwhelmed by the larger volume of the low-dose
616
William S. Bice, Jr.
Figure 7. Differential dose-volume histogram. These are the differential dose-volume histograms, the prostate and the entire calculation volume, for the same implant as shown in Figure 6. Each point on the curve represents the volume, or percent volume, that received the dose indicated by the 1 Gy bin. The displays have been normalized due to large differences in the volumes. The inset shows actual values. See the text.
regions, obscuring the information where the doses are the greatest, usually within the target structure. Figure 8 illustrates this effect and how the NDVH provides a remedy. Note the lack of visible information at doses greater than the reference dose (the vertical line) in Figure 7. Figure 8 shows the NDVH for both the calculation grid and the prostate gland. Note the similarity at doses greater than the reference dose, the major discrepancy being dose regions greater than the reference dose but outside the gland. Anderson and other investigators used the NDVH to see dose structure at high-dose regions in and around the implant, presumably where the target volume is located, to evaluate the quality of implants, even though the target volume is not seen in the image set. Some commercial planning systems have applied the NDVH approach to defined anatomic structures or target volumes (Moerland et al., 2000). Dose-Surface Histograms (DSHs) Dose-surface histograms (DSHs) can also be used to portray the dose on the surface of structure where volumetric dose may be less meaningful. An example of this is the rectum, where dose to the rectal wall is probably more important than dose to anything within the rectum. DSHs plot surface area or percent surface area on the ordinate. As with the DVH, a differential DSH can be developed as well as a natural DSH, although the utility of the latter is not entirely apparent. Dose Traces The dose trace or line plot is an appropriate graphical depiction of dose for small linear structures. For prostate brachytherapy, this display has been used as a tool for analyzing the dose to the urethra and the neurovascular bundles. In HDR brachytherapy, comparative displays of measured versus calculated dose at points within the urethra have been the subject of dose traces (Anagnostopoulos et al., 2003). A urethral dose trace for the implant shown in Figure 6 is shown in Figure 9.
31–Post Implant Evaluation
617
Figure 8. The natural dose-volume histogram. The natural dose-volume histogram removes the effect of lower doses having larger volumes because of the inverse square law. Shown are the NDVH for the entire calculation volume and the prostate volume as in Figure 7. The similarity of the curves in the regions above the reference dose indicate that the NDVH can be used to tease out information about the target doses when there is no target visible in the image set. This relies of course upon the assumption that the high dose regions were delivered to the target.
Figure 9. Urethral dose trace. This is a dose trace along the center of the urethra for the implant of Figure 6. The distance is measured along the urethra from base to apex slices. Only the first 2.7 cm of the length of the urethra is shown. Slice positions are indicated at the top of the vertical lines. The maximum dose of 189 Gy can be seen to occur at slice position 48 (1.8 cm from the base).
618
William S. Bice, Jr.
DVH Sampling DVHs are created by sampling the target volume, calculating the dose to sample point (representing a sample volume), and binning the result. The DVH created from a grid calculation is an example of sampling at regular intervals. The optimal sampling technique and the uncertainty due to sampling of the target volume have been the matter of some debate (Lu and Chin 1993; Niemierko and Goitein 1994; Karouzakis et al., 2002). This uncertainty can best be seen in the DDVH. Figures 10a through 10e are examples of DDVH created by regular grid sampling, random sampling with varying numbers of samples, and the sampling noise removed from regular grid sampling by convolving the result with a simple box filter (moving averages in the financial world). As can be seen, generating a less noisy DDVH is possible with any of these methods, but the uncertainty with regard to the result, particularly for brachytherapy DVH, has yet to be studied and quantified.
Quantifiers Dosimetric quantifiers are used to further distill the information from the DVH down to a single value that describes some aspect of the implant. Like the DVH from which they are derived, quantifiers may be structure based or nonstructure based. Several quantifiers have both forms: for instance, the nonstructure-based dose non-uniformity ratio (DNR) (Saw and Suntharalingam 1991) and V150 that is structure based. Both are calculated similarly, the volume encompassed by the surface described by 150% of the reference dose divided by the volume of the surface encompassed by the surface described by the reference dose (the DNR) or else divided by the volume of the target (V150). Quantifiers are used to describe certain attributes of the implant. These can be classified as describing either structure coverage or dose uniformity. Coverage Quantifiers Coverage quantifiers, for obvious reasons, are usually restricted to structure-based DVH analysis. A possible exception is the use of matched peripheral dose (MPD), which may be derived from a non-structure based DVH. The MPD, originally developed as a design tool, is the dose that encompasses the same volume as the volume of the target. Although MPD does not require locating the target in the image set, as a coverage quantifier this really is a fatal flaw, indicating nothing about the relative location between the dose distribution and the target. MPD is rarely used for post-implant evaluation. Coverage quantifiers can be grouped according to the Volume quantifiers (V) and the Dose quantifiers (D). V quantifiers describe the volume of the structure that receives at least a certain dose. Both the volume and the dose may be stated as a percentage, the first as a percentage of the target volume, the second as a percentage of the reference dose. The V quantifier is subscripted to indicate the dose, or percentage of dose, to which the quantifier refers. For instance, V150 refers to the volume, or percentage of the volume, that receives at least 150% of the reference dose. Note that V150 could also refer to the volume, or percentage of the volume, which receives 150 Gy, or 150 cGy, or some other expression of a dose of 150 of some unit. In this regard, the nomenclature is indistinct. The majority of the time, both the volume and the dose are expressed as percentages, allowing comparison between implants that were performed on different volume targets and which had different reference doses—even different isotopes. Common usage has modified, to a degree, the format of the V quantifiers. Used to describe coverage of the target, a V is almost always used, but in describing dose to critical structures, the V may be replaced with a letter pertaining to the critical structure. For instance R100 may refer to the volume, or percent of volume, of the rectum which receives 100% of the reference dose or greater; U200 the volume, or percent of volume, of the urethra which receives 200% of the reference dose or greater.
31–Post Implant Evaluation Figure 10. Sampling errors in the dose-volume histogram.
Figure 10a. The DDVH calculated directly from a 2-mm grid.
Figure 10b. The DDVH calculated directly from a 1-mm grid.
Figure 10c. The DDVH calculated from a 2-mm grid with a simple box filter smoothing routine.
619
620
William S. Bice, Jr.
Figure 10d. The DDVH calculated from 100,000 random samples.
Figure 10e. The DDVH calculated from 1,000,000 random samples. This series of differential dose-volume histograms shows the noise inherent in creating the DDVH. Most treatment-planning systems that show the DDVH use some sort of smoothing routine to generate a less noisy plot. There is some difference between Figures 10c (under sampling and smoothing) and 10e (heavily sampled), but the clinical significance is probably marginal.
It may also be misleading or inappropriate to normalize the value of the volume in presenting the quantifier. Stating a rectal dose quantifier like R100, in terms of a percentage of the rectal volume may portray no useful information. Rectal complications are most likely related to the volume of the rectum that receives a certain dose, not a percentage of the rectum that received that dose. The added complication of using an indeterminate length to describe the rectum is also thus avoided by quoting the value of the volumetric quantifier in volume rather than as a percentage, 3.8 cm3 versus 20%, for example. D quantifiers describe the dose at which a given volume of the target receives that dose or greater. Like V quantifiers, D quantifiers are subscripted, this time with the volume rather than the dose. Again the value of the quantifier and the subscript may be normalized or not and the distinction is not always clear. D90 for instance most commonly means the dose, or percentage of dose, that encompasses 90% of the target. It may also pertain to 90 cm3 or 90 mm3 or 90 of some other unit of volume. Unlike the V quantifiers, D quantifiers usually normalize the volume but rarely normalize the dose. This allows comparison between targets of different volume, but not between implants with different reference doses. Occasionally the need
31–Post Implant Evaluation
621
does exist. For example, when performing comparisons between implants performed with different isotopes or when including implants performed as monotherapy and implants performed as boost therapy in the same data set, different reference doses are used. In these cases the D quantifier must be normalized to the percentage of each implant reference dose. Because the V quantifiers and the D quantifiers both describe coverage of the target, it would be surprising they were independent of each other. Figure 11 is a graph of this relationship between implants performed at a single institution. For the implant design and delivery methods used at this institution, V100 = 87% was equivalent to D90 = 100% of the reference dose. This held true for both iodine implants with a reference dose of 145 Gy and palladium implants with a reference dose of 125 Gy (115 Gy, prior to 2000). This relationship is probably strongly related to the practices adhered to by a specific implant team, and is not therefore generally applicable. Uniformity Quantifiers The doses in brachytherapy are, by nature, inhomogeneous. Nevertheless ignoring the size of the highdose regions would be folly. Excess dose that does not contribute to curing the disease enhances the risk of complications. The question of wasted dose in prostate cancer has been studied by Ling resulting in the conclusion that doses in excess of 125% of the prescription dose probably do not contribute to curing the disease (Ling et al., 1994). Inhomogeneity quantifiers fall into two categories: those that quantify the inhomogeneous nature of the dose and those that simply quantify the size of the high-dose regions. The best example of a quantifier that describes the homogeneity of an implant is the full width at half maximum (FWHM) of the DDVH; the smaller this value, the more homogeneous the implant. V150 is the most common example of a quantifier that describes the size of the high-dose regions in the implant. As described previously, this is the percentage volume of the gland that receives greater than 150% of the reference dose.
Figure 11. The relationship between V100 and D90 at a single institution. V100 and D90 both describe coverage of the gland and it is not surprising that they correlate. For this institution, a V100 of 87% may be interpreted as about the same coverage as a D90 equal to the reference dose. This was true for both iodine and palladium implants.
622
William S. Bice, Jr.
Uniformity and the size of the high-dose region are not necessarily related. This is shown in Figure 12, where the DDVH for two implants with similar V150 have dramatically different uniformity as measured by the FWHM DDVH. It should be noted that the value of V150 is only meaningful as a quantifier of true dose uniformity if used when comparing implants with the same coverage by the reference dose, V100, or if V150 is normalized to V100. Despite this, V150 is more commonly used than the dose homogeneity index (DHI) because complication rates are thought to be much more likely due to excessive dose rather than non-uniformity of the dose distribution. The DHI is equal to the [V(TDR) – V(HDR)]/V(TDR), where V(TDR) is the volume covered by the target dose-rate and V(HDR) is the high dose-rate volume—usually the volume covered by 1.5 times the target dose-rate (Wu, Ulin, and Sternick 1988). Because V150 is the percentage of the target volume that has a dose greater than 150% of the reference dose, one might assume that V150 is equal to (1 – DHI). This is not, in general, true for two reasons. First, the denominator for V150 is the entire gland volume, not just the volume covered by the reference dose. Second, the DHI, originally developed for iridium implants, is still usually considered a nonstructure-based quantifier.
Critical Structure Dosimetry Dosimetry for critical structures varies according to the structure type. In an effort to calculate doses that will correlate to complications that arise from the dose delivered, most brachytherapists have strived to tailor the dose calculation, display, and evaluation to the form of the structure. Consider the possible critical structures in prostate brachytherapy. The bladder and rectum might be more amenable to DSH analysis; the urethra and neurovascular bundles, line dosimetry; the penile bulb, a DVH; the bladder neck; a point calculation. Efforts have been made to standardize critical structure dosimetry and reporting, but with much less success than in standardizing target dosimetry (Yu et al., 1999; Nag et al., 2000, 2002). For instance, the dose to the rectum has been characterized by different investigators as a DVH, an annular DVH, a DSH, a dose trace, and as a single point value. Urethra doses have been characterized in terms of length, area, and volume. The West Virginia group found that within some limitations all of these methods were viable and comparable as analysis tools for post implant evaluation (Merrick et al., 1999a; Butler et al., 2000a). An excellent investigation of the optimal replication of rectal surface dose measurements was performed by Hilts et al. (Hilts, Spadinger, and Keyes 2002). Looking at length of rectum, average anterior dose, and DVH of the rectum, they found that the DVH of the anterior half of the rectum was the optimal alternative to DSHs.
Dosimetric Uncertainty The vast majority of our credible knowledge pertaining to post-implant evaluation of prostate implants comes from CT imaging. As previously noted, it is very difficult to precisely delineate glandular borders on CT. The effect on post-implant dosimetric quantifiers has been reported by various authors, but without consensus. Contouring recommendations have been developed by various groups (McLaughlin et al., 2002; Crook et al., 2002). Figure 1 showed the results in the coverage quantifiers for expansion and contraction in 2 mm increments of the prostatic contours for a series of implants as shown in Figure 13. No expansion was performed in the posterior direction as the extent of the gland is more easily visualized due to the presence of the rectal wall. The effect of this expansion and contraction scheme on 50 125I implants is graphed for V100 in Figure 1a and D90 in Figure 1b. An analysis of 30 palladium implants gave very similar results, as do other expansion methods used to perform the same analysis. As can be seen, drawing the gland too large makes the coverage quantifiers smaller, drawing the gland too small makes the coverage quantifiers larger, but to a lesser extent. One would conclude that erring in the smaller direction would likely be more accurate. One practitioner uses the pre-implant ultrasound volume, without edema, to overlay and contour the post-
31–Post Implant Evaluation
623
Figure 12. The difference between V150 and uniformity.
Figure 12a. The DDVH for an implant with V150 = 55% and FWHM = 96 Gy.
Figure 12.b. The DDVH for an implant with V150 = 55% and FWHM = 77 Gy . Although commonly used as such, V150 is not really a measure of implant uniformity. The implant in Figure 12b. is more uniform than that in 12a. (FWHM of 77 Gy versus 96 Gy), but both have a V150 of 55%. The relationship between V150 and FWHM is dependent upon target volume, source strength, and implant design.
624
William S. Bice, Jr.
Figure 13. Expansion and contraction of the glandular volume. One possible method of expansion and contraction of prostatic contours from post-implant CT scans is shown here. The gland was grown or contracted in 2-mm increments about the center of mass, except in the posterior direction, where the glandular borders are more visible than in the other directions. The analysis was performed in order to estimate the effects of uncertainty in drawing the glandular borders. The results are shown in Figures 1a. and 1b.
operative, edematous gland on CT. While this almost certainly overestimates the prostatic coverage, it has the advantage of consistency, being based on an imaging modality in which the glandular borders are readily visible, and the advantage of erring in the too-small direction, the direction of smaller error (Badiozamani et al. 1999; Smith et al., 2003). Finding and contouring the location and extent of critical structures on a CT can also present difficulties. The urethra is best visualized following catheterization, not necessarily a problem if the post implant imaging is performed immediately following the procedure, but at significant discomfort to the patient if performed at a later time. A retrograde urethrogram, although less invasive, is inadequate due to the inability of the dye to cross the urogenital diaphragm. Neurovascular bundles (NVB) may or may not be critical structures for retention of potency; the uncertainty is probably due in part to the inability to visualize them on CT. Work has been performed recently using MRI to localize the NVB in prostate brachytherapy (McLaughlin et al., 2005). The rectal wall, penile bulb, and bladder are usually of sufficiently dissimilar contrast when compared to adjacent structures to be adequately visualized in CT image sets.
31–Post Implant Evaluation
625
Dynamic Dosimetry While prostate brachytherapy offers a unique opportunity to evaluate the dose being delivered from sources that are visible alongside the target and critical structures, it should be remembered that the image set still only provides a single snapshot of the dose delivered over the life of the implant. The resolution of postimplant edema, seed migration, and anatomical changes can all occur while the radiation dose is being delivered. Roy and Ling have performed extensive radiobiological analyses for iodine prostate brachytherapy, calculating the effective life of an iodine implant to be about 275 days (Roy et al., 1993a; Ling et al., 1994). During this time the relative change between source locations and the structures of interest has not been rigorously studied. Models which characterize edematous changes have been developed which portray the effect of these changes in dosimetric coverage of the gland, but very little has been done to quantify the resolution of edema on the dose to critical structures. Waterman and Dicker (1999, 2000) did show a continuous, exponential increase in rectal dose R10 and in urethral dose as edema resolved. Most seed loss probably occurs within the first week following the implant. The most common mechanism for this is likely migration through the bladder or urethra wall and subsequent elimination with the urine. Occasionally seeds work their way into the periprostatic arteriovenous complexes and can later be found in the lung or heart tissues. The migration of seeds is less for stranded seeds than for loose seeds. This sort of seed loss is thought to have little effect on the dose delivered in permanent brachytherapy, but temporal studies have yet to be performed which illustrate this (Merrick et al., 2000a; Fuller, Koziol, and Feng 2004).
Advanced Techniques Subglandular Analysis Several techniques have been used to perform subglandular targeting of dose. One institution implants only the peripheral zone of the prostate, based upon extensive biopsy data that shows the likelihood of cancer occurring elsewhere within the gland to be extremely small. These implants are performed under MRI-guidance so that the peripheral zone may be imaged (D’Amico et al., 2000; Kooy et al., 2000). Other institutions have used MR spectroscopy or 111In imaging (see Figure 14) to locate most likely foci of disease within the gland (Ellis, Kim, and Foor 2004). These brachytherapists still target the whole gland but increase the dose locally at the cancer sites. Whether the intent is to reduce complications or to boost the dose to sites of active disease, there is an increased challenge to post-implant evaluation. First, imaging of the subglandular target must be performed with a modality in which this target can be seen. If imaging using this modality is performed pre-implant only, registration of the pre-implant image set to the post-implant image set is plagued by changes not only in the shape and location of the edematous gland, but also possibly uncorrelated changes in the subglandular target volume. MRI and nuclear medicine imaging are inadequate for seed localization, which implies the need for a second imaging modality, along with the inherent problems of fusion methods and added cost. Creation of the DVH for a structure removes all the spatial information with regard to the distribution of dose within the target. Sector analysis was developed to retain some of this information while still maintaining the simplicity of DVH and quantifier comparison. In this technique the gland is divided into 12 sectors, 4 each at the base, midgland, and apex, as shown in Figure 15. A DVH is then created for each sector, each with a corresponding set of quantifiers. Sector analysis is particularly useful for analyzing a series of implants. Trends in implant technique can be teased from a very large data set (Bice, Prestidge, and Sarosdy 2001),
626
William S. Bice, Jr.
Figure 14. SPECT image using 111In. Partial volume brachytherapy boosts are possible using SPECT imaging or magnetic resonance spectroscopy. This image was taken 5 days following the injection of 111In, showing uptake on the left side of the gland in the peripheral zone. This presents a unique challenge with regard to post-implant dosimetry; target subvolumes have not been studied with regard to what doses are appropriate or achievable. Localization of the subvolume almost certainly requires the fusion of two or more image sets, some acquired before the implant, some after.
Sensitivity Analysis The inability of the brachytherapist to position the sources precisely as planned has been of concern since the beginning of post-implant planning and analysis (Roberson et al., 1997). Several investigators have used source displacement analysis to model the effect of the source misplacement and/or local migration on the quality of the implant (Roberson et al., 1997; Wuu et al., 2000; Beaulieu et al., 2004). These techniques rely on post-implant evaluation to develop a model for source displacement and then apply this model to either pre-implant design or to temporal analysis over the life of the implant. Table 2 shows the utility of displacement analysis for determining the sensitivity of an implant design to misplacement of the sources. In this particular table the sensitivity of glandular coverage to source displacement is quantified as the average and standard deviation of D90 based upon 50 iterations of source displacement using randomly distributed values of displacement. Practitioners often refer to this sensitivity as a measure of design robustness.
Correlation of Clinical Results to Post-Implant Dosimetry In citing clinical results for PPB, note the difficulties associated with conducting and publishing results for this disease. Despite attempts to automate and standardize the methods used to design and perform the implant, dosimetric results are still largely dependent upon the skill of the practitioner—a skill which changes with the acquisition of experience. Evaluation tools are not truly standardized, nor are analysis and reporting techniques. Prostate cancer is an extremely indolent disease. With regard to cure, databases are not really mature until the 10- or 15-year mark. Complications usually occur within the first 1 or 2
31–Post Implant Evaluation
627
Figure 15. Sector analysis. Sector analysis allows retention of some of the spatial information dose to certain regions of the gland while affording the advantages of dose-volume histogram and quantifier analysis. Shown above is the sector results for an implant that was adequately covered well at the base (sectors 1–4) and the midgland (sectors 5–8), but the coverage at the apex was questionable. While the practitioner would probably be able to glean this sort of information from a set of isodose curves, such an approach has the disadvantage of being difficult to quantify and would be inappropriate for analysis of a large database.
Table 2. Displacement Analysis Results Design
Number of Sources
Source Strength (U)
Total Strength (U)
V100 Design (Displaced) %
V150 Design (Displaced) %
D90 Design (Displaced) Gy
Design #1
66
0.432
28.5
99.5 (90.4 ± 3.5)
65.2 (37.6 ± 8.5)
186 (147 ± 6)
Design #2
45
0.635
28.6
99.4 (87.6 ± 4.1)
68.4 (44.6 ± 9.1)
183 (139 ± 9)
This is an example of source displacement analysis. It can be used for pre-planning, as was done here, or postplanning. This table compares two types of implant design for iodine monotherapy on a 22.2 cm3 gland, one high activity, one low activity, but both with the same total source strength. Each design was submitted to 20 trials of random displacement (Gaussian displacement, isotropic, mean of 0.0 mm, one standard deviation of 5.0 mm). The ability of Design #1 to better withstand the rigors of displacement would make us tend to label it as the more robust of the two.
628
William S. Bice, Jr.
years following the implant, but statistical significance is often difficult to achieve simply because serious complications from the procedure are infrequent. Studies characterizing complication rates often entail following large patient populations for a significant period of time.
Cures Early studies describing adequate prostate brachytherapy implants centered on coverage of the target volume, but lacked any correlation with clinical outcomes (Willins and Wallner 1997). The first published dose response study for permanent prostate brachytherapy was based on the post-implant evaluation of 134 early-stage prostate cancer patients at Mt. Sinai Hospital in New York (Stock et al., 1998) who had undergone iodine implants. The analysis was performed by correlating post-implant, CT-image generated D90 for these patients to biochemical control of the disease: the decrease of prostate specific antigen (PSA) levels following the implant. Stock found a statistically significant (p=0.02) difference in control rates between patients that had a post-implant D90 of greater that 140 Gy and those that had a D90 of less than 140 Gy. The first group had an actuarial control rate of 85% at 6 years while those patients in the second group had a control rate of 45%. These results were later confirmed by Potters et al. (2001), and extended to palladium monotherapy implants with a cutoff D90 of 125 Gy. As was shown in Figure 9, D90 of 145 Gy for iodine monotherapy and D90 of 125 Gy for palladium monotherapy both correspond, for the institution shown, a V100 of 87%.
Complications The most common rectal complication from prostate brachytherapy is proctitis, resulting in bleeding. Many authors have successfully correlated rectal complications with post-implant dose to the rectum (Han and Wallner 2001; Snyder et al., 2001; Merrick and Butler 2004). But as previously mentioned, the nonstandardization of determining the rectal target volume has made comparison of results difficult. The earliest rectal dose post-implant analysis was published by Wallner in an analysis of 45 iodine monotherapy patients (Wallner et al. 1995). The conclusion of this work was that doses to significant portions of the rectum, when converted to TG43 calculation formalism, be limited to 90 Gy. Merrick, calculating doses to the length and surface area of the anterior mucosal surface, determining mean, median, and maximum values showed that a more reasonable estimate of the allowable dose to 1 cm length of the rectum should be 100% of the reference dose, 145 Gy for iodine, 125 Gy for palladium. The maximum dose should be limited to 120% of the reference dose. This study included 45 patients (Merrick et al., 1999a). Snyder et al. (2001), in a 212-patient study, quantified results from an annular DVH described by the outer rectal wall and the inner, mucosal surface of the rectum and showed that keeping all but 1.3 cm3 of the resultant volume below the reference dose of 160 Gy resulted in less than 5% chance of Grade 2 proctitis. These results were tabulated and are graphed in Figure 16 as an example of a complication DVH; dose-volumes to the right of the line result in 5% chance of Grade 2 proctitis. Similar findings were published by Han and Wallner (2001) in studying results from 160 patients using both surface area and a DVH generated simply from the outer rectal wall. This study, using the reference dose as a cutoff found significant differences between bleeders and nonbleeders at 3 cm2 versus 7 cm2 and at 0.6 cm3 versus 2.5 cm3. Similar analyses of urethral complications have achieved less success. Urethral complications consist of both irritative and obstructive symptoms with the latter being more morbid. Wallner recommended keeping the dose to the urethra below 360 Gy (TG43) (Wallner et al., 1995). By 1998 Desai et al. (1998) had been able to correlate an increase in urinary symptoms with dose to gland and dose to the urethra, but made no specific recommendations. Stock and Stone (2002) found a correlation between excessive prostate D90 values and urethral complications and recommended keeping D90 below 180 Gy for iodine monotherapy. Excessive doses, greater than the reference dose, to the membranous urethra were found to result in
31–Post Implant Evaluation
629
Figure 16. Dose-volume analysis of rectal complications. Snyder et al. (2001) analyzed Grade II rectal proctitis rates from iodine monotherapy patients and tabulated the results. The analysis was unique in that, by holding the probability of occurrence constant, the authors were able to determine the probability of occurrence over a range of dose-volume values. The figure above is a portrayal of these data in the form of a complication DVH. If the rectal DVH were calculated in the same fashion as in the reference and plotted on this graph, a line that crossed over or laid to the right of the curve shown would presumably indicate a 5% chance of Grade 2 proctitis.
urethral obstruction in a retrospective, matched-patient study by Merrick et al. (2002). This same study also was able to correlate D90 and V150 to urethral obstruction. Multiple authors have attempted to correlate critical structure doses to erectile dysfunction following the implant. Early work by DiBiase showed a weak correlation between impotence and dose to the neurovascular bundles (NVB) (DiBiase et al., 2000). The NVB are not readily seen in CT images so a surrogate location for these structures was used. This location was based upon prostatic border delineation, adding another layer of uncertainty to the position of the NVB. Other, similar studies were unable to substantiate this correlation (Merrick et al., 2000b). Stock was able to correlate D90 values to erectile dysfunction, but it is doubtful that the gland itself is the critical structure in this regard. It is more likely that the high doses delivered to the prostate were correlates to high doses to some other critical structure. In any case, keeping D90 less than 160 Gy for iodine and 100 Gy for palladium monotherapy were recommended in their results (Stock, Kao, and Stone 2001). A strong correlation was noted between impotence and excessive dose to the penile bulb. The penile bulb lies 1 to 2 cm inferior to the apex of the prostate gland. It has been strongly recommended that the dose to 50% of this structure, D50, be kept below 50 Gy. Most implant design techniques employed today easily achieve this recommendation (Merrick et al., 2001, 2002).
630
William S. Bice, Jr.
Post-Implant Targets Table 3 lists suggested implant evaluation targets for PPB based upon the listed references. It should be noted that these targets are highly subjective and include the author’s bias. They should be tailored to the needs of the brachytherapist and the practice. Balancing the optimal chance for a cure against the risk of complications for the individual patient is the art of medicine, always within the realm of the patient’s physician.
The Post-Implant Evaluation Program A post-implant evaluation should be conducted for each patient. Target coverage and dose to critical structures should be calculated and evaluated in terms of implant targets, like those in Table 3, which have been set by the brachytherapist and modified according to the patient’s needs. A practitioner who has never had a suboptimal implant has never looked. Such implants should be evaluated for additional intervention in the case of inadequate coverage, or for added therapeutic intervention or increased follow up and monitoring in the case of excessive doses to critical structures (Yu et al., 1999; Nag et al. 2000, 2002).
Reporting Suggested post-implant reporting formats vary according to the type of implant. In addition to the demographic data, as a minimum each implant evaluation report should contain the type of implant, the target volume—as formulated in International Commission on Radiation Units and measurements (ICRU) reports 50 and 62, if possible, a description of the target and implanted volume; a description of the implant plan, to include the type and form of sources, the air kerma strength; and an evaluation of the implant results, a description of the dose to the target and to the critical structures (Anderson et al., 1991; Nath et al., 1997). Several dedicated prostate brachytherapy treatment-planning systems offer automated report generation; this can reduce dramatically the amount of time required. Table 4 lists the recommended reportTable 3. Suggested Post-Implant Dosimetry Targets Structure
Intent
Goal
References
Prostate Gland
Cure
D90 for iodine monotherapy > 140 Gy D90 for palladium monotherapy > 125 Gy D90 for boosts > reference dose
Stock et al. (1998) Potters et al. (2001)
Prostate Gland
Urethral complications
D90 < 180 Gy V150 < 60% reference dose
Stock and Stone (2002) Merrick et al. (2002)
Membranous Urethra
Urethral complications
Dose to the membranous urethra < reference dose
Merrick et al. (2002)
Rectum
Rectal complications
Dose to > 1 cm length of anterior mucosal wall < reference dose Max dose to anterior mucosal wall < 120% of reference dose
Merrick et al. (1999a)
Rectum
Rectal complications
Annular DVH of rectum < 1.3 cm3 to 160 Gy (iodine)
Snyder et al. (2001)
Rectum
Rectal complications
Surface area of outer rectal wall < 5 cm2 to reference dose
Han and Wallner (2001)
31–Post Implant Evaluation
631
Table 4. Suggested Reporting Formats Brachytherapy procedure
Reporting format
Gynecological brachytherapy
ICRU 38
Gynecological brachytherapy (HDR)
ICRU 38, Nag et al. (2000)
Interstitial implants
ICRU 58
Prostate brachytherapy
Gillin 1990, Yu et al. (1999), Nag et al. (2002)
General Bbrachytherapy
Anderson et al. (1991), Nath et al. (1997)
ing formats for several implant types. A suggested format for prostate brachytherapy implants is attached as an appendix. Compiling these results into a useful format for analyzing a series of implants can be quite cumbersome. Due to the popularity of the procedure, software that automatically downloads and stores the implant descriptors in a database is available for prostate brachytherapy; compiling the results for other types of implants has yet to reach this level of ease.
Program Review A review of the implant program should be performed at least annually (Nath et al., 1997; Nag et al. 2000). Additional reviews may be conducted following the introduction of a new technique or a change of equipment (Bice, Prestidge, and Sarosdy 2001). Program reviews can be very useful in guiding the implant team into taking the necessary steps to improve implant quality. While there is no specific format for this review, suggestions have been made in regulatory documents with regard to the adequacy of record keeping (U.S. NRC 1991). Such analysis is useful, of course, but does little to improve the performance of the implant team in achieving better implants. In this regard the review should include an analysis of implant dosimetry quantifiers and how well these quantifiers achieve the stated goals. These quantifiers should not be limited to target coverage but should also include dosimetry information pertaining to critical structures. Additionally, some sort of tracking with regard to clinical outcome is extremely useful. While it is impractical to expect most implant programs to maintain a full database with regard to patient follow-up, tracking a single type of complication and attempting to draw a correlation to the post-implant dosimetry is both possible and useful. For instance, knowing which patients had urinary complications following prostate brachytherapy allows the practitioner to focus on the dosimetry for these patients and to determine how the implant characteristics for these patients differ from the rest. Quantifiers to focus first might be those listed in Table 3, or other correlates for urinary complications—D90 values, for instance. It is also possible to choose other dosimetric characteristics to analyze, dose to the bladder neck or number of needle passes, for example. Of course, the evaluation is not the goal; we hope for better implants. Improving implant performance can be achieved only if the results are critically analyzed by the implant team members. Changing implant design, equipment, or delivery techniques should all be considered in order to improve. The implant team needs to digest and learn from this type of evaluation, possible only when egos are swallowed and frank discussion ensues. This sort of analysis can become difficult and entail hard work, but the results can be quite dramatic.
632
William S. Bice, Jr.
Multicenter Trials Multicenter trials are essential to the advancement of medical knowledge. These trials have been performed using post-implant evaluation results (Bice et al., 1998; Wallner et al., 2002; Al-Qaisieh 2003; Mangili et al., 2004). Such studies are especially difficult due to the uncertainties associated with target and critical structure delineation. The lack of a standard protocol for calculating and analyzing dosimetry results is also problematic. The quality control for research protocols that rely on post-implant evaluation to categorize and compare results often must be quite detailed, rely on extensive over reading of subjective interpretation, and require centralized data storage and analysis (RTOG 2002).
Summary In that the results can affect more than just a single implant, post-implant evaluation may be considered the most critical element of the brachytherapy procedure. Skipping this evaluation is dangerous for the patient and for the implant program (Stock and Stone 2002). It is truly impossible to deliver high-quality care consistently without an aggressive evaluation program. The basis for any brachytherapy evaluation program is the post plan. Post-implant evaluation provides the foundation upon which our current knowledge base is laid. Only by looking at the results from large series of implants have we been able to make sense of clinical outcomes. How the implant procedure is performed does make a difference. Problems do exist. There is no perfect imaging modality; what works well for delineating the soft tissues is notoriously poor at localizing the sources; and vice versa. Combining imaging modalities has the added uncertainty associated with image co-registration and is expensive. Our analytical tools are admittedly immature. Modern calculation methods and treatment planning systems have been available for less than 10 years. Which quantifiers are valuable and how they relate to clinical outcomes will continue to be the subject of study for many years. Analysis and display methods are nonstandard; which tier on the post-implant evaluation hierarchy is appropriate for evaluation of dosimetry will continue to be a matter of debate. Three or four years ago there was a noticeable shift of the focus of clinical studies from cures to complications. This is highly encouraging, indicative of a satisfaction with our ability to cure the disease, concentrating instead on how to make the procedure less morbid. We know we deliver better implants, but how much better we can get and whether this improvement in quality can be generally attained remain uncertain.
31–Post Implant Evaluation
633
Appendix. Sample Reporting Format for Permanent Prostate Brachytherapy Brachytherapy Report Date: Wednesday, January 1, 2000 Patient: Mr. Truly Anonymous, MR#00001 Physician: Frank Lee Neerdowell, MD
General On Tuesday, November 14, 1999 an ultrasound volume study was performed on Mr. Anonymous following a biopsy positive for prostate cancer. From this ultrasound study, the size and location of the prostate gland was determined and a radioactive implant (brachytherapy) using iodine-125 (GreedySeedVendorModel)(NIST99) radioactive sources was planned. Usually termed a preplan, this treatment plan was used to guide the placement of seeds in the operating room during the implant procedure performed on Thursday, December 1, 1999. During the implant procedure 106 sources of iodine-125 (GreedySeedVendorModel) (NIST99) were implanted in and around the prostate with the intent of delivering a curative dose of radiation to the gland. Subsequently, on Thursday, December 15, 1999, a CT scan was performed to evaluate the quality of the implant; by noting the position of the seeds a dosimetric evaluation of the implant was conducted. This treatment plan is termed a “post plan” in that it was performed after the procedure.
Description of Volumes Mr. Anonymous has clinical stage T1c adenocarcinoma of the prostate, a Gleason sum score of 6/10, and a pre-implant prostate specific antigen level of 4.6 ng/ml. The gross tumor volume includes the left apex and left base of the prostate gland, which were positive from the biopsy. The clinical target volume is the entire prostate gland, which had a volume of 43.1 cm3 based upon the ultrasound volume study. The planning volume extends several millimeters outside the gland lateral and anterior with a larger margin at the base and apex of the gland than in the central regions. The posterior margin of the PTV is smaller to minimize rectal dose.
Sources and Techniques GreedySeedVendor I125 sources, model GSVModel#, were used to perform the implant. The implemented preplan consisted of an array of 24 needles with a total of 104 sources arranged to deliver dose in as uniform a pattern as possible throughout the planning target volume. Most needles had sources placed at 10 mm spacing, the exceptions being at the periphery of the gland, particularly at the base and apex, to ensure adequate dosimetric coverage. The implant design technique is termed “modified-peripheral” with the exact locations detailed in the loading map attached to the preplan. The source positions used in this implant were uniquely designed for this patient in conjunction with the physician’s desire to achieve a specific dose distribution. The implant was carried out with 2 extra sources. The reason that extra sources were implanted was based upon the perceived source distribution at the end of the implant procedure as determined via fluoroscopy. The total number of sources implanted was 106. 105, I-125 (GreedySeedVendor) (NIST99) sources of 0.34 mCi (0.429 U) each were used in the post plan evaluation. This number was derived from an AP radiograph with exact locations determined from the CT scan.
634
William S. Bice, Jr.
Total Reference Air Kerma At the time of the implant the total reference air kerma rate for this implant was 45.50 microGy per hour at 1 meter. Source strength was verified by assay to within 5% of this value by using an independent measuring system that has a calibration traceable to the National Institute for Standards and Technology.
Description of the Dose Distribution The prescribed dose for this implant was 145 Gy. Sometimes termed the “reference dose,” this is the minimum dose to be achieved in the design of the implant (or preplan). Isodose curves and dose volume histograms (DVHs) attached to the preplan illustrate this. The achieved dose distribution is detailed in the post plan isodose display and DVHs. Because of post-implant edema and uncertainties in the delineation of the prostatic volume on CT, the resultant doses are best described in terms of several quantifiers generated form analysis of the DVH. The percentage of the gland that received 100% of the reference dose or greater, termed V100, for this implant was 96.7%. Similarly, that percentage of the gland that received 80% of the reference dose or greater, V80, was 99.8%. D90, or the lowest dose that covers 90% of the gland, is 167 Gy. With regard to uniformity, the dose non-uniformity ratio (DNR) for this implant was 0.471. This is the ratio of the tissue volume that received 150% or greater of the reference dose divided by the volume that received 100% of the reference dose or greater. The DNR is a measure of the uniformity of the implant with a range from 0 to 1, smaller numbers indicative of a more uniform dose distribution. The DNR is independent of prostate volume. It is a possible indicator of the chance of complications and is used in the same manner as V150, which was 55.2% for this implant. A urethra dose trace (dose along the center of the urethra) is provided in the post plan as is a dose surface histogram of the urethral wall. These indicate that none of the urethra received more than 200% of the reference dose and none received more than 250% of the reference dose. The dose to the bladder neck is estimated to be 113 Gy. A DSH of the rectum indicates that 2.78 square centimeters of the rectum wall received 100% of the reference dose or greater. Similarly, 0.05 square centimeters of the rectal wall received 150% of the reference dose or greater. Submitted:
Reviewed by:
______________________________
______________________________
William S. Bice, Jr., Ph.D., CSMP Medical Physicist
Frank Lee Neerdowell, MD Radiation Oncologist
31–Post Implant Evaluation
635
References Al-Qaisieh, B. (2003). “Pre- and post-implant dosimetry: An inter-comparison between UK prostate brachytherapy centres.” Radiother Oncol 66(2):181-183. Amdur, R. J., D. Gladstone, K. A. Leopold, and R. D. Harris. (1999). “Prostate seed implant quality assessment using MR and CT image fusion.” Int J Radiat Oncol Biol Phys 43(1):67–72. Anagnostopoulos, G., D. Baltas, A. Geretschlaeger, T. Martin, P. Papagiannis, N. Tselis, and N. Zamboglou. (2003). “In vivo thermoluminescence dosimetry dose verification of transperineal 192Ir high-dose-rate brachytherapy using CT-based planning for the treatment of prostate cancer.” Int J Radiat Oncol Biol Phys 57(4):1183–1191. Anderson, L. L. (1986). “A ‘natural’ volume-dose histogram for brachytherapy.” Med Phys 13:898–903. Anderson, L., R. Nath, A. J. Olch, and J. Roy. (1991). “American Endocurietherapy Society recommendations for dose specification in brachytherapy.” Endocuriether/Hypertherm Oncol 7:1–12. Badiozamani, K. R., K. Wallner, W. Cavanagh, and J. Blasko. (1999). “Comparability of CT-based and TRUS-based prostate volumes.” Int J Radiat Oncol Biol Phys 43(2):375–378. Beaulieu, L., L. Archambault, S. Aubin, E. Oral, R. Taschereau, and J. Pouliot. (2004a). “The robustness of dose distributions to displacement and migration of 125I permanent seed implants over a wide range of seed number, activity, and designs.” Int J Radiat Oncol Biol Phys 58:1298–1308. Bice, W. S., Jr., B. R. Prestidge, and M. F. Sarosdy. (2001). “Sector analysis of prostate implants.” Med Phys 28(12):2561–2567. Bice, W. S., Jr., B. R. Prestidge, P. D. Grimm, J. L. Friedland, V. Feygelman, M. Roach 3rd, J. J. Prete, D. F. Dubois, and J. C. Blasko. (1998). “Centralized multiinstitutional postimplant analysis for interstitial prostate brachytherapy.” Int J Radiat Oncol Biol Phys 41(4): 921–927. Bice, W. S., Jr., D. F. Dubois, J. J. Prete, and B. R. Prestidge. (1999). “Source localization from axial image sets by iterative relaxation of the nearest neighbor criterion.” Med Phys 26(9):1919–1924. Bice, W. S., Jr., J. E. Freeman, L. F. Russell, Jr., G. D. Case, D. F. Dubois, J. J. Prete, and B. R. Prestidge. (2000). “Use of image coregistration in salvage prostate brachytherapy.” Tech Urol 6(2):151–156. Brinkmann, D., and R. W. Kline. (1998). “Automated seed localization from CT datasets of the prostate.” Med Phys 25(9):1667–1672. Butler, W. M., G. S. Merrick, A. T. Dorsey, and B. M. Hagedorn. (2000a). “Comparison of dose length, area, and volume histograms as quantifiers of urethral dose in prostate brachytherapy.” Int J Radiat Oncol Biol Phys 48(5):1575–1582. Butler, W. M., G. S. Merrick, A. T. Dorsey, and J. H. Lief. (2000b). “Isotope choice and the effect of edema on prostate brachytherapy dosimetry.” Med Phys 27(5):1067–1075. Claus, F. G., H. Hricak, and R. R. Hattery. (2004). “Pretreatment evaluation of prostate cancer: Role of MR imaging and 1H MR spectroscopy.” Radiographics 24(Suppl 1):S167–S180. Crook, J., M. Milosevic, P. Catton, I. Yeung, T. Haycocks, T. Tran, C. Catton, M. McLean, T. Panzarella, and M. A. Haider. (2002). “Interobserver variation in postimplant computed tomography contouring affects quality assessment of prostate brachytherapy.” Brachytherapy 1(2):66–73. D’Amico, A., R. Cormack, S. Kumar, and C. M. Tempany. (2000). “Real-time magnetic resonance imaging-guided brachytherapy in the treatment of selected patients with clinically localized prostate cancer.” J Endourol 14(4):367–370. Davis, B. J., Brinkmann, D. H., Kruse, J. J., Herman, M. G., LaJoie, W. N., Schwart, D. J., Pisansky, T. M., and R. W. Kline. (2004). “Selective identification of different brachytherapy sources, ferromagnetic seeds, and fiducials in the prostate using an automated seed sorting algorithm.” Brachytherapy 3:106–112. DeWyngaert, J. K., M. E. Noz, B. Ellerin, E. L. Kramer, G. Q. Maguire Jr., and M. P. Zeleznik. (2004). “Procedure for unmasking localization information from ProstaScint scans for prostate radiation therapy treatment planning.” Int J Radiat Oncol Biol Phys 60(2):654–662. Desai, J., R. G. Stock, N. N. Stone, C. Iannuzzi, and J. K. DeWyngaert. (1998). “Acute urinary morbidity following I-125 interstitial implantation of the prostate gland.” Radiat Oncol Investig 6(3):135–141. DiBiase, S. J., K. Wallner, K. Tralins, and S. Sutlief. (2000). “Brachytherapy radiation doses to the neurovascular bundles.” Int J Radiat Oncol Biol Phys 46(5):1301–1307.
636
William S. Bice, Jr.
Dogan, N., N. Mohideen, G. P. Glasgow, K. Keys, and R. C. Flanigan. (2002). “Effect of prostatic edema on CT-based postimplant dosimetry.” Int J Radiat Oncol Biol Phys 53(2):483–489. Dubois, D. F. (1999). Magnetic Resonance Imaging in Permanent Prostate Brachytherapy. Department of Radiology. San Antonio, Texas, University of Texas Health Science Center at San Antonio. Dubois, D. F., W. S. Bice Jr., and B. R. Prestidge. (2001). “CT and MRI derived source localization error in a custom prostate phantom using automated image coregistration.” Med Phys 28(11):2280–2284. Dubois, D. F., B. R. Prestidge, L. A. Hotchkiss, W. S. Bice Jr., and J. J. Prete. (1997). “Source localization following permanent transperineal prostate interstitial brachytherapy using magnetic resonance imaging.” Int J Radiat Oncol Biol Phys 39(5):1037–1041. Dubois, D. F., B. R. Prestidge, L. A. Hotchkiss, J. J. Prete, and W. S. Bice Jr. (1998). “Intraobserver and interobserver variability of MR imaging- and CT-derived prostate volumes after transperineal interstitial permanent prostate brachytherapy.” Radiology 207(3):785–789. Ellis, R. J., E. Kim, and R. Foor. (2004). “Role of ProstaScint for brachytherapy in localized prostate adenocarcinoma.” Expert Rev Mol Diagn 4(4):435–441. Ellis, R. J., A. Vertocnik, E. Kim, H. Zhou, B. Young, B. Sodee, P. Fu, S. Beddar, V. Colussi, J. P. Spirnak, K. H. Dinchman, M. Resnick, and T. J. Kinsella. (2003). “Four-year biochemical outcome after radioimmunoguided transperineal brachytherapy for patients with prostate adenocarcinoma.” Int J Radiat Oncol Biol Phys 57(2):362–370. Fang, D. X., R. G. Stock, N. N. Stone, B. R. Krynyckyi, C. K. Kim, and J. Machac. (2000). “Use of radioimmunoscintigraphy with indium-111-labeled CYT-356 (ProstaScint) scan for evaluation of patients for salvage brachytherapy.” Tech Urol 6(2):146–150. Fuller, D. B., J. A. Koziol, and A. C. Feng. (2004). “Prostate brachytherapy seed migration and dosimetry: analysis of stranded sources and other potential predictive factors.” Brachytherapy 3(1):10–19. Gillin, M.T., D.L. Zellmer, D.F. Grimm and K. Sherwood. (1992). “Practical Considerations for Interstitial Brachytherapy” in Advances in Radiation Oncology Physics, Dosimetry, Treatment Planning and Brachytherapy. J. A. Purdy (ed.). AAPM Medical Physics Monograph Number 19. New York: American Institute of Physics, pp. 703-727, 1992. Gong, L., P. S. Cho, B. H. Han, K. E. Wallner, S. G. Sutlief, S. D. Pathak, D. R. Haynor, and Y. Kim. (2002). “Ultrasonography and fluoroscopic fusion for prostate brachytherapy dosimetry.” Int J Radiat Oncol Biol Phys 54(5):1322–1330. Han, B. H., and K. E. Wallner. (2001). “Dosimetric and radiographic correlates to prostate brachytherapy-related rectal complications.” Int J Cancer 96(6):372–378. Han, B. H., K. Wallner, G. Merrick, K. Badiozamani, and W. Butler. (2003a). “The effect of interobserver differences in post-implant prostate CT image interpretation on dosimetric parameters.” Med Phys 30(6):1096–1102. Han, B. H., K. Wallner, G. Merrick, W. Butler, S. Sutlief, and J. Sylvester. (2003b). “Prostate brachytherapy seed identification on post-implant TRUS images.” Med Phys 30(5):898–900. Hilts, M., I. Spadinger, and M. Keyes. (2002). “Comparison of methods for calculating rectal dose after 125I prostate brachytherapy implants.” Int J Radiat Oncol Biol Phys 53(3):775–785. Holmes, D. R., 3rd, B. J. Davis, and R. A. Robb. (2001). “3D localization of implanted radioactive sources in the prostate using trans-urethral ultrasound.” Stud Health Technol Inform 81:199–205. Holupka, E. J., P. M. Meskell, E. C. Burdette, and I. D. Kaplan. (2004). “An automatic seed finder for brachytherapy CT postplans based on the Hough transform.” Med Phys 31(9):2672–2679. International Commission on Radiation Units and Measurements (ICRU). Report No. 38. “Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology.” Bethesda, MD: ICRU, 1985. International Commission on Radiation Units and Measurements (ICRU). Report No. 50. “Prescribing, Recording and Reporting Photon Beam Therapy.” Bethesda, MD: ICRU, 1993. International Commission on Radiation Units and Measurements (ICRU). Report No. 58. “Dose and Volume Specification for Reporting Interstitial Therapy.” Bethesda, MD: ICRU, 1997. International Commission on Radiation Units and Measurements (ICRU). Report No. 62: “Prescribing, Recording and Reporting Photon Beam Therapy (Supplement to ICRU Report 50).” Bethesda, MD: ICRU, 1999. Karouzakis, K., M. Lahanas, N. Milickovic, S. Giannouli, D. Baltas. and N. Zamboglou. (2002). “Brachytherapy dose-volume histogram computations using optimized stratified sampling methods.” Med Phys 29(3):424–432.
31–Post Implant Evaluation
637
Kooy, H. M., R. A. Cormack, G. Mathiowitz, C. Tempany, and A. V. D’Amico. (2000). “A software system for interventional magnetic resonance image-guided prostate brachytherapy.” Comput Aided Surg 5:401–413. Lam, S. T., P. S. Cho, R. J. Marks II, and S. Narayanan. (2004). “Three-dimensional seed reconstruction for prostate brachytherapy using Hough trajectories.” Phys Med Biol 49(4):557–569. Lee, W. R., M. Roach 3rd, J. Michalski, B. Moran, and D. Beyer. (2002). “Interobserver variability leads to significant differences in quantifiers of prostate implant adequacy.” Int J Radiat Oncol Biol Phys 54(2):457–461. Li, Z., I. A. Nalcacioglu, S. Ranka, S. K. Sahni, J. R. Palta, W. Tome, and S. Kim. (2001). “An algorithm for automatic, computed-tomography-based source localization after prostate implant.” Med Phys 28(7):1410–1415. Lindsay, P. E., J. Van Dyk, and J. J. Battista. (2003). “A systematic study of imaging uncertainties and their impact on 125I prostate brachytherapy dose evaluation.” Med Phys 30(7):1897–1908. Ling, C. C., J. Roy, N. Sahoo, K. Wallner, and L. Anderson. (1994). “Quantifying the effect of dose inhomogeneity in brachytherapy: Application to permanent prostatic implant with 125I seeds.” Int J Radiat Oncol Biol Phys 28(4):971–978. Liu, H., G. Cheng, Y. Yu, R. Brasacchio, D. Rubens, J. Strang, L. Liao, and E. Messing. (2003). “Automatic localization of implanted seeds from postimplant CT images.” Phys Med Biol 48:1191–1203. Lu, X. Q., and L. M. Chin. (1993). “Sampling techniques for the evaluation of treatment plans.” Med Phys 20(1):151–161. Mangili, P., L. Stea, F. Cattani, S. Lappi, F. Giglioli, E. Calamia, F. Ziglio, R. Martinelli, and B. Longobardi. (2004). “Comparative study of permanent interstitial prostate brachytherapy post-implant evaluation among seven Italian institutes.” Radiother Oncol 71(1):13–21. McLaughlin, P. W., V. Narayana, D. G. Drake, B. M. Miller, L. Marsh, J. Chan, R. Gonda Jr., R. J. Winfield, and P. L. Roberson. (2002). “Comparison of MRI pulse sequences in defining prostate volume after permanent implantation.” Int J Radiat Oncol Biol Phys 54(3):703–711. McLaughlin, P. W., V. Narayana, A. Meriowitz, S. Troyer, P. L. Roberson, R. Gonda Jr., H. Sandler, L. Marsh, T. Lawrence, and M. Kessler. (2005). “Vessel-sparing prostate radiotherapy: dose limitation to critical erectile vascular structures (internal pudendal artery and corpus cavernosum) defined by MRI.” Int J Radiat Oncol Biol Phys 61(1):20–31. Merrick, G. S., and W. M. Butler, (2004). “Rectal function following permanent prostate brachytherapy.” W V Med J 100(1):18–20. Merrick, G. S., W. M. Butler, A. T. Dorsey, and H. L. Walbert. (1998). “Influence of timing on the dosimetric analysis of transperineal ultrasound-guided, prostatic conformal brachytherapy.” Radiat Oncol Investig 6(4):182–190. Merrick, G. S., W. M. Butler, A. T. Dorsey, J. H. Lief, H. L. Walbert, and H. J. Blatt. (1999a). “Rectal dosimetric analysis following prostate brachytherapy.” Int J Radiat Oncol Biol Phys 43(5):1021–1027. Merrick, G. S., W. M. Butler, A. T. Dorsey, and J. H. Lief. (1999b). “The dependence of prostate postimplant dosimetric quality on CT volume determination.” Int J Radiat Oncol Biol Phys 44(5):1111–1117. Merrick, G. S., W. M. Butler, A. T. Dorsey, J. H. Lief, and M. L. Benson. (2000a). “Seed fixity in the prostate/periprostatic region following brachytherapy.” Int J Radiat Oncol Biol Phys 46(1):215–220. Merrick, G. S., W. M. Butler, A. T. Dorsey, J. H. Lief, and J. G. Donzella. (2000b). “A comparison of radiation dose to the neurovascular bundles in men with and without prostate brachytherapy-induced erectile dysfunction.” Int J Radiat Oncol Biol Phys 48(4):1069–1074. Merrick, G. S., K. Wallner, R. W. Galbreath, J. H. Lief, and M. L. Benson. (2001). “A comparison of radiation dose to the bulb of the penis in men with and without prostate brachytherapy-induced erectile dysfunction.” Int J Radiat Oncol Biol Phys 50(3):597–604. Merrick, G. S., W. M. Butler, K. E. Wallner, J. H. Lief, R. L. Anderson, B. J. Smeiles, R. W. Galbreath, and M. L. Benson. (2002). “The importance of radiation doses to the penile bulb vs. crura in the development of postbrachytherapy erectile dysfunction.” Int J Radiat Oncol Biol Phys 54(4):1055–1062. Mizowaki, T., G. N. Cohen, A. Y. Fung, and M. Zaider. (2002). “Towards integrating functional imaging in the treatment of prostate cancer with radiation: the registration of the MR spectroscopy imaging to ultrasound/CT images and its implementation in treatment planning.” Int J Radiat Oncol Biol Phys 54(5):1558–1564. Moerland, M. A., H. K. Wijrdeman, R. Beersma, C. J. Bakker, and J. J. Battermann. (1997). “Evaluation of permanent I-125 prostate implants using radiography and magnetic resonance imaging.” Int J Radiat Oncol Biol Phys 37(4):927–933.
638
William S. Bice, Jr.
Moerland, M. A., R. van der Laarse, R. W. Luthmann, H. K. Wijderman, and J. J. Battermann. (2000). “The combined use of the natural and the cumulative dose-volume histograms in planning and evaluation of permanent prostatic seed implants.” Radiother Oncol 57:279–284. Nag, S., W. Bice, K. DeWyngaert, B. Prestidge, R. Stock, R., and Y. Yu. (2000). “The American brachytherapy society recommendations for permanent prostate brachytherapy postimplant dosimetric analysis.” Int J Radiat Oncol Biol Phys 46:221–230. Nag, S., R. J. Ellis, S. Merrick, R. Bahnson, K. Wallner, and R. Stock. (2002). “American Brachytherapy Society recommendations for reporting morbidity after prostate brachytherapy.” Int J Radiat Oncol Biol Phys 54(2):462–470. Narayana, V., P. L. Roberson, R. J. Winfield, and P. W. McLaughlin. (1997). “Impact of ultrasound and computed tomography prostate volume registration on evaluation of permanent prostate implants.” Int J Radiat Oncol Biol Phys 39(2):341–346. Narayanan, S., P. S. Cho, and R. J. Marks II. (2004). “Three-dimensional seed reconstruction from an incomplete data set for prostate brachytherapy.” Phys Med Biol 49(15):3483–3494. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee, Task Group No. 43.” Med Phys 22:209–234. Also available as AAPM Report No. 51. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24(10):1557–1598. Also available as AAPM Report No. 59. Niemierko, A., and M. Goitein. (1994). “Dose-volume distributions: A new approach to dose-volume histograms in three-dimensional treatment planning.” Med Phys 21(1):3–11. Polo, A., F. Cattani, A. Vavassori, D. Origgi, G. Villa, H. Marsiglia, M. Bellomi, G. Tosi, O. De Cobelli, and R. Orecchia. (2004). “MR and CT image fusion for postimplant analysis in permanent prostate seed implants.” Int J Radiat Oncol Biol Phys 60(5):1572–1579. Potters, L., Y. Cao, E. Calugaru, T. Torre, P. Fearn, and X. H. Wang. (2001). “A comprehensive review of CT-based dosimetry parameters and biochemical control in patients treated with permanent prostate brachytherapy.” Int J Radiat Oncol Biol Phys 50(3):605–614. Prestidge, B. R., W. S. Bice, E. J. Kiefer, and J. J. Prete. (1998). “Timing of computed tomography-based postimplant assessment following permanent transperineal prostate brachytherapy.” Int J Radiat Oncol Biol Phys 40(5):1111–1115. Prete, J. J., B. R. Prestidge, W. S. Bice, D. F. Dubois, and L. A. Hotchkiss. (1998). “Comparison of MRI- and CT-based post-implant dosimetric analysis of transperineal interstitial permanent prostate brachytherapy.” Radiat Oncol Investig 6(2):90–96. Radiation Therapy Oncology Group (RTOG) P0232 (2002). “A phase III study comparing combined external beam radiation and transperineal interstitial permanent brachytherapy with brachytherapy alone for selected patients with intermediate risk prostatic carcinoma.” B. Prestidge, (Principal Investigator). Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. (2004). “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Also available as AAPM Report No. 84. Roberson, P. L., V. Narayana, D. L. McShan, R. J. Winfield, and P. W. McLaughlin. (1997). “Source placement error for permanent implant of the prostate.” Med Phys 24(2):251–257. Roy, J. N., L. L. Anderson, K. E. Wallner, Z. Fuks, and C. C. Ling. (1993). “Tumor control probability for permanent implants in prostate.” Radiother Oncol 28(1):72–75. Roy, J. N., K. E. Wallner, P. J. Harrington, C. C. Ling, and L. L. Anderson. (1993). “A CT-based evaluation method for permanent implants: application to prostate.” Int J Radiat Oncol Biol Phys 26(1):163–169. Sajo, E. (2003). Private communication. The use of PET to localize sources in permanent prostate brachytherapy. W. Bice. Baton Rouge, LA. Saw, C. B., and N. Suntharalingam. (1991). “Quantitative assessment of interstitial implants.” Int J Radiat Oncol Biol Phys 20(1):135–139. Smith, S., K. Wallner, G. Merrick, W. Butler, S. Sutlief, and P. Grimm. (2003). “Interpretation of pre- versus postimplant TRUS images.” Med Phys 30(5):920–924.
31–Post Implant Evaluation
639
Snyder, K. M., R. G. Stock, S. M. Hong, Y. C. Lo, and N. N. Stone. (2001). “Defining the risk of developing grade 2 proctitis following 125I prostate brachytherapy using a rectal dose-volume histogram analysis.” Int J Radiat Oncol Biol Phys 50(2):335–341. Sodee, D. B., R. J. Ellis, M. A. Samuels, J. P. Spirnak, W. F. Poole, C. Riester, D. M. Martanovic, R. Stonecipher, and E. M. Bellon. (1998). “Prostate cancer and prostate bed SPECT imaging with ProstaScint: Semiquantitative correlation with prostatic biopsy results.” Prostate 37(3):140–148. Speight, J., K. Shinohara, B. Pickett, V. Weinberg, I.-C. Hsu, and M. Roach III. (2000). “Prostate volume change after radioactive seed implantation: Possible benefit of improved dose volume histogram with perioperative steroid.” Int J Radiat Oncol Biol Phys 48(5):1461–1467. Stock, R. G., and N. N. Stone. (2002). “Importance of post-implant dosimetry in permanent prostate brachytherapy.” Eur Urol 41(4):434–439. Stock, R. G., J. Kao, and N. N. Stone. (2001). “Penile erectile function after permanent radioactive seed implantation for treatment of prostate cancer.” J Urol 165(2):436–439. Stock, R. G., N. N. Stone, A. Tabert, C. Iannuzzi, and J. K. DeWyngaert. (1998). “A dose-response study for I-125 prostate implants.” Int J Radiat Oncol Biol Phys 41(1):101–108. Su, Y., B. J. Davis, M. G. Herman, and R. A. Robb. (2004). “Prostate brachytherapy seed localization by analysis of multiple projections: Identifying and addressing the seed overlap problem.” Med Phys 31(5):1277–1287. Texter, J. H., Jr., and C. E. Neal. (1998). “The role of monoclonal antibody in the management of prostate adenocarcinoma.” J Urol 160(6 Pt 2):2393–2395. Tsai, A., W. Wells, C. Tempany, E. Grimson, and A. Willsky. (2004). “Mutual information in coupled multi-shape model for medical image segmentation.” Med Image Anal 8(4):429–445. Tubic, D., A. Zaccarin, J. Pouliot, and L. Beaulieu. (2001a). “Automated seed detection and three-dimensional reconstruction. I. Seed localization from fluoroscopic images or radiographs.” Med Phys 28:2265–2271. Tutar, I. B., R. Managuli, V. Shamdasani, P. S. Cho, S. D. Pathak, and Y. Kim. (2003). “Tomosynthesis-based localization of radioactive seeds in prostate brachytherapy.” Med Phys 30(12):3135–3142. U.S. Nuclear Regulatory Commission. (1991). Regulatory Guide 8.33, Quality Management Program. Washington D.C.: Nuclear Regulatory Commission. Wallner, K., J. Roy, L. True, W. Cavanagh, C. Simpson, and W. Butler. (1995). “Dosimetry guidelines to minimize urethral and rectal morbidity following transperineal I-125 prostate brachytherapy.” Int J Radiat Oncol Biol Phys 32(2):465–471. Wallner, K., G. Merrick, L. True, W. Cavanaugh, C. Simpson, and W. Butler. (2002). “I-125 versus Pd-103 for low-risk prostate cancer: morbidity outcomes from a prospective randomized multicenter trial.” Cancer J 8(1):67–73. Waterman, F. M., and A. P. Dicker. (1999). “Effect of post-implant edema on the rectal dose in prostate brachytherapy.” Int J Radiat Oncol Biol Phys 45(3):571–576. Waterman, F. M., and A. P. Dicker. (2000). “The impact of postimplant edema on the urethral dose in prostate brachytherapy.” Int J Radiat Oncol Biol Phys 47(3):661–664. Willins, J., and K. Wallner. (1997). “CT-based dosimetry for transperineal I-125 prostate brachytherapy.” Int J Radiat Oncol Biol Phys 39(2):347–353. Willins, J., and K. Wallner. (1998). “Time-dependent changes in CT-based dosimetry of I-125 prostate brachytherapy.” Radiat Oncol Investig 6(4):157–160. Wu, A., K. Ulin, and E. S. Sternick. (1988). “A dose homogeneity index for evaluating 192Ir interstitial breast implants.” Med Phys 15:104–107. Wuu, C. S., R. D. Ennis, P. B. Schiff, E. K. Lee, and M. Zaider. (2000). “Dosimetric and volumetric criteria for selecting a source activity and a source type (125I or 103Pd) in the presence of irregular seed placement in permanent prostate implants.” Int J Radiat Oncol Biol Phys 47:815–820. Yu, Y., L. L. Anderson, Z. Li, D. E. Mellenberg, R. Nath, M. C. Schell, F. M. Waterman, A. Wu, and J. C. Blasko. (1999). “Permanent prostate seed implant brachytherapy: Report of the American Association of Physicists in Medicine Task Group No. 64.” Med Phys 26(10):2054–2076. Also available as AAPM Report No. 68. Yue, N., Z. Chen, R. Peschel, A. P. Dicker, F. M. Waterman, and R. Nath. (1999a). “Optimum timing for image-based dose evaluation of 125I and 103PD prostate seed implants.” Int J Radiat Oncol Biol Phys 45(4):1063–1072.
640
William S. Bice, Jr.
Yue, N., A. P. Dicker, B. W. Corn, R. Nath, and F. M. Waterman. (1999b). “A dynamic model for the estimation of optimum timing of computed tomography scan for dose evaluation of 125I or 103Pd seed implant of prostate.” Int J Radiat Oncol Biol Phys 43(2):447–454. Yue, N., A. P. Dicker, R. Nath, and F. M. Waterman. (1999c). “The impact of edema on planning 125I and 103Pd prostate implants.” Med Phys 26(5):763–767. Zaider, M., M. J. Zelefsky, E. K. Lee, K. L. Zakian, H. I. Amols, J. Dyke, G. Cohen, Y. C. Hu, A. K. Endi, C.-S. Chui, and J. A. Koutcher. (2000). “Treatment planning for prostate implants using magnetic-resonance spectroscopy imaging.” Int J Radiat Oncol Biol Phys 47(4):1085–1096.
Chapter 32
HDR Brachytherapy for Prostate Zoubir Ouhib, M.S. Lynn Regional Cancer Center of Boca Raton Community Hospital Boca Raton, Florida History of Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642 Permanent Seed Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642 Temporary LDR Approach with 192Ir . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642 Transrectal Ultrasonographic Permanent Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642 Ultrasound-Guided HDR Implant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643 Ultrasound-Guided SWIFT™ or Real-Time HDR Prostate Procedure . . . . . . . . . . . . . . . . . . . . . . 643 Factors Affecting the Choice of HDR Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644 Individual Decision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644 Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644 Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644 Other . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645 Patient Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645 HDR Brachytherapy as Monotherapy Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645 HDR Brachytherapy as a Boost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645 Equipment Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645 HDR Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645 Template and Needles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645 Imaging System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645 Items for Patient Comfort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646 Procedure in the OR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646 Items Needed for OR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646 Real Time vs. Traditional CT HDR Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 646 Patient Positioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647 Template Positioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 647 Ultrasound Volume Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649 Needle Placement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649 Post Implant Cystoscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649 Simulation Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 Traditional CT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 Final Needle Positioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 Reference Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 Image-Guided Brachytherapy for “Real Time” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 Treatment Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 Target Volume Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 Critical Organs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651 Dosimetry Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652 Dose Prescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652 HDR Boost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652 Monotherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652 Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652 Needles Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652 Treatment Plan Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652 Independent Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 Visual Inspection of the Implant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 Patient Positioning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 Quality Assurance (QA) Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654
641
642
Zoubir Ouhib Pretreatment Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 Items Related to the Patient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 Items Related to the HDR Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 Items Related to the Treatment Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 Items Between Treatment Plan and HDR Treatment Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654 Treatment Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656 Post Treatment Stage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
History of Brachytherapy Prostate brachytherapy, like any other medical technology, has seen its share of evolution over the past years. These changes are associated with the use of different isotopes, new techniques, innovative treatment planning software, and better imaging modalities.
Permanent Seed Implants In 1903 Alexander Graham Bell postulated that one way to effectively control tumors is by placing radioactive sources directly into the tumors. In the early 1970s, the open retropubic technique was introduced by Willet Whitmore and colleagues at Memorial Sloan-Kettering Cancer Center (MSKCC) (Whitmore et al., 1974). This approach did not allow the radiation oncologist and the surgeon to visualize the entire prostate clearly or to plan with accuracy the placement of the needles and seeds into the gland. The planning was performed with the use of a standard nomogram, which provided the user basic information such as the number of needles, the seeds, and their spacing. Seeds were placed using a Mick applicator (Mick RadioNuclear Instruments, Inc., Mt. Vernon, NY). This technique did not provide a distribution of seeds that was consistent in quality. Because the results of the open retropubic implant technique were inferior to most radical prostatectomy and external beam series, this technique was essentially abandoned in the early 1980s.
Temporary LDR Approach with 192Ir Low dose rate (LDR) temporary implants utilize 192Ir in the form of a wire or seeds embedded in nylon ribbon. This implant technique introduced the use of a prostate template (Syed et al., 1992). Plastic needles were implanted through the template. Nomograms were used to balance and optimize the activity per seed, number of seeds, and number of needles. Unfortunately, the results were more or less similar to the MSKCC method described above.
Transrectal Ultrasonographic Permanent Implants In the 1980s, Holm et al. (1983, 2003) showed that transrectal sonography could be used to more accurately implant the prostate. This technique was soon thereafter implemented in the United States. The two radionuclides of choice were 125I and 103Pd. This transrectal sonography enabled the brachytherapist to reconstruct the prostate gland three-dimensionally, creating an ideal model for treatment. With a template for needle guidance, it also allowed for precise, real-time visualization of needle placement during surgery. Treatment planning software designed specifically for prostate implants became available. Ultrasound images (volume studies) were acquired prior to the implant and a preplan was generated. The preplanning allowed the brachytherapy staff to evaluate the plan using a given seed strength prior to the procedure and to order the appropriate number of seeds. The use of needles preloaded with seeds and spacers became
32–HDR Brachytherapy for Prostate
643
an alternative to the Mick applicator. Other centers planned their implants in the operating room (OR) prior to the implant. This was initially done with the use of nomograms (total activity based on gland volume) and available seed activities. This became a real-time technique and has evolved to where computer optimization in the operating room is displacing nomogram use. Despite the good visualization from the ultrasound imaging system and improved techniques, there was an apparent difference in dosimetric quality when the pre-implant and post implant plans were compared. Among the factors that contributed to this difference were the needle positions (intended vs. actual location), seed migration, and prostate motion during needle insertion.
Ultrasound-Guided HDR Implant To improve the dosimetry for prostate implants, the use of remote afterloading with high dose rate (HDR) 192 Ir was introduced in the late 1980s. HDR brachytherapy has been used for more than 40 years to treat cancer of the cervix and endometrium. HDR remote afterloading replaced the manually loaded sources in which the radiation dose delivered to the prostate and surrounding normal tissues is very difficult to control. The needle insertion is performed in the OR with the use of the transrectal ultrasound and fluoroscopy. The placement of the needles is done using a fixed template (Figure 1) or the Syed free-style template technique (Syed et al., 1992). The afterloader is able to move the 192Ir source in specified steps called “dwell positions” (Figure 2) within each catheter. These dwell positions can be programmed to occur at different spacing, from 2.5 mm to 10.0 mm apart. This method can be used as a monotherapy treatment or as a boost before or after external beam radiotherapy treatment (EBRT).
Ultrasound-Guided SWIFT™ or Real-Time HDR Prostate Procedure Nucletron’s SWIFT™(Nucletron Corporation, Columbia, MD) treatment-planning approach is similar to the conventional ultrasound-guided method described above with some significant advantages. Some of these advantages are:
Figure 1. Template and needles.
644
Zoubir Ouhib
Figure 2. Dwell points.
- Reduce treatment-planning time-(compared to conventional HDR planning) - Planning is based on 3-D ultrasound results - Perform implantation simultaneously with treatment planning - Immediate feedback on implant quality - Generate optimal final implant geometry - Patient and staff benefit from faster, more efficient procedure - The SWIFT allows the user to access multiple 2-D ultrasound images and generates a 3-D volume to determine optimal implant geometry and to facilitate needle placement
Factors Affecting the Choice of HDR Treatment Individual Decision Patients are becoming more and more informed and involved with their treatment options and final choice. Some of their decisions are based on conversations with friends or relatives. Access to information via the Internet has become one of their tools to search and confirm their options. News headlines regarding prominent individuals making their final choice of treatment also play a role. Age There is no defined lower or upper limit, but typically the ages of prostate cancer patients treated with HDR range from 48 to 78 years old. These patients have to be physically able to go through a surgical procedure and be confined to a bed for 30 hours. Patients with heart disease should be cleared by a cardiologist, who may discourage them from going through this procedure due to their health condition. Stage The stage of the disease is an important factor in determining if the patient should have HDR brachytherapy as a boost, following a course of EBRT, or as a monotherapy treatment.
32–HDR Brachytherapy for Prostate
645
Other Other factors affecting the choice of therapy include the availability of HDR technology in the local area and the financial burden on some patients.
Patient Selection HDR Brachytherapy as Monotherapy Treatment - Clinical stage II, T1c or T2a - PSA (prostate-specific antigen) ≤ 10 ng/mL - Gleason score ≤ 6 or less
HDR Brachytherapy as a Boost - Clinical stage ≥ T2b - Gleason score ≥ 7 - Pretreatment PSA ≥ 10
Equipment Selection HDR Unit HDR units are available through several vendors such as Nucletron Corporation and Varian Medical Systems (Palo Alto, CA). The unit consists of the delivery unit, the control console that interfaces with the unit, and the treatment-planning system.
Template and Needles Some manufacturers of HDR systems provide templates and needles designed for their unit and their needles (Figure 1). This ensures compatibility and reliability. These items are also available from other third-party vendors. It is the responsibility of the user to confirm that they are safe and adaptable to their system. The needles consist of either 15 gauge (1.9 mm diameter) or 17 gauge stainless steel needles. The number of needles required is about 16 to 20. In addition, Flexiguide plastic needles (Flexiguide Ltd., Devon, UK) of the same gauge as the steel needles are invariably used when the treatment involves more than one fraction per implant. Flexineedles are more comfortable for the patient and have no computed tomography (CT) artifact.
Imaging System • Ultrasound unit and ultrasound probe: to view target and needle location in a transverse view • Probe stabilization device to hold the ultrasound transducer rigidly • Fluoroscopy unit such as a C-arm to ensure that needles are parallel once implanted
646
Zoubir Ouhib • CT unit to acquire images for treatment planning purpose. This is not necessary if real-time planning is performed.
Items for Patient Comfort One important item, sometimes neglected, is patient comfort once the needles are inserted. Discomfort can contribute to the needles and template shifting while the patient is searching for a comfortable position or a less painful one. Needle movement affects the dose distribution. Items for considerations are: • Leg support to elevate the patient’s knees to improve comfort (Figure 3) • “Prostate mattress,” which is a mattress with a square cutout just below the patient’s perineum to improve access to the template and catheters. This mattress adds comfort to the patient by eliminating friction of the needles with the mattress (Figure 4).
Procedure in the OR Items Needed for the Implant Table 1 contains a list of items needed in the OR or surgical suite. These are imaging devices and accessories. Figure 5 shows a template for HDR implant coordinates.
Real-Time vs. Traditional CT HDR Brachytherapy The real time ultrasound-based method eliminates the use of traditional CT following the implantation for verification of needle positions and acquisition of images for planning purposes. Ultrasound using the SWIFT system is used to verify the needle positions before each fraction.
Figure 3. Knee support for patient.
32–HDR Brachytherapy for Prostate
647
Figure 4. Prostate mattress.
Patient Positioning Following anesthesia, the patient is positioned in a dorsal lithotomy position (Figure 6). The perineum is usually shaved. This area plus the suprapubic and lower abdominal area are then scrubbed and prepped. A catheter is used to drain the bladder of urine and 150 cm3 of sterile saline are reinfused into the bladder to improve ultrasound imaging. A traumatic tower clip or suture is used to pull the scrotum and testicles over the base of the penis onto the anterior suprapubic area.
Template Positioning The ultrasound probe is inserted into the rectum and the prostate is viewed from base to apex (Figure 7). The probe stabilization is adjusted from left to right and anterior to posterior to center the target volume on the ultrasound unit.
Table 1. Items Needed in the OR for the HDR Prostate Procedure Imaging devices
Accessories
Ultrasound unit
Prostate needles
Ultrasound probe
Flexiguide plastic needles
Probe stabilization device
Prostate template and associated tools
Fluoroscopy unit or C-Arm
Template coordinate configuration for needle placement (Figure 5) Nonradioactive seeds to be used as markers Prostate mattress
648
Zoubir Ouhib
Figure 5. Template coordinate.
Figure 6. Patient in lithotomy position.
32–HDR Brachytherapy for Prostate
649
Figure 7. Probe insertion.
Ultrasound Volume Evaluation The prostate is scanned from base to apex to get the overall gland size and its geometry. The presence and position of any calcifications are noted. This information can help distinguish needles from calcifications on the ultrasound display because of the similar appearance of needles and calcifications on ultrasound images.
Needle Placement Closed-end stainless steel needles are used for the implant. The gland can be stabilized with the insertion of two to four needles. Once the prostate is stabilized, needles are inserted in a pattern that will maintain a good ultrasound image of the prostate. This is usually done from anterior to posterior and medial to lateral. All the needles are implanted within the gland relative to the largest transverse slice (Figure 8). The needles should not to be advanced too far superiorly until the cystoscope is in place. This will prevent the possibility of inadvertently penetrating the bladder mucosa with the anterior rows of needles. As discussed previously, plastic flexible needles are much more comfortable for the patients; and once the stylets are removed during CT, imaging these Flexiguides shows no artifact. The detachable portion of the template is removed and the remaining portion of the template is sutured to the perineum (Figure 9). A digital rectal examination should be performed to feel where the posterior needles are in relation to the rectal mucosa.
Post Implant Cystoscopy Prior to cystoscopy, the patient’s legs are taken out of the lithotomy position and positioned in a slight frogleg on the table. The flexible cystoscope guides and defines the final needle tip placement in relation to the
650
Zoubir Ouhib
Figure 8. Completion of needle insertion.
Figure 9. Template being sutured to the skin.
32–HDR Brachytherapy for Prostate
651
bladder. Needles are advanced superiorly to a submucosal position within the bladder. This will ensure adequate dose to the prostatic base during treatment.
Simulation Procedure Traditional CT Final Needle Positioning Since there is no preplanning done prior to the implant, patients must have a post implant CT. Contrast (Hypaque™, 50 to 100 cc) is infused into the bladder lumen via the catheter. Antero-posterior (AP) and lateral scout views of the pelvis are first obtained to verify needle location. The metal stylets are then removed from the Flexiguide catheters, and CT images are obtained from a level superior to the base to several centimeters inferior to the apex at 3-mm intervals. The implanted flexiguide needles will appear as air-filled holes within the gland. Reference Plane All catheters should be visible at the base in order to deliver an adequate dose to that location. Adjustments can be made by reinserting the stylet and advancing the needle appropriately to the desired location. A final scout with all the stylets in place is taken to be used as a quick reference for subsequent fractions. The needles and/or the templates can then be marked to identify any future shift. The transverse CT images at 3-mm intervals are then acquired and used for treatment planning. The most superior location of each needle tip in relation to the base of the prostate gland should be verified prior to each fraction for possible shift (they should be identical to the first planned fraction).
Image-Guided Brachytherapy for “Real Time” The real-time method does not require CT image acquisition since the plan is generated from the ultrasound volume study. The ultrasound-based verification using the SWIFT is done before each fraction. It is a good idea for those who are just starting (part of the learning curve process) or switching to this method to do a few dosimetric comparisons with the CT planning to confirm their results.
Treatment Planning Target Volume Definition The target volume is defined to be the prostate’s peripheral margin as drawn on the post needle placement CT or ultrasound images. Some users include a small portion of the seminal vesicles to ensure adequate dose to the base and also to treat one of the more likely pathways of extracapsular extension.
Critical Organs The urethra, rectum, and bladder are drawn on the images. Dwell times may need to be modified to adjust the dose to these organs. Needles too close to these organs at risk may be left unused. This is one of the reasons why additional needles inserted at the time of implantation (maximum of two) are useful.
652
Zoubir Ouhib
Dosimetry Goals The following guidelines should be considered while generating the plan: 1. The D90 (dose to 90% of the volume) should be between 105% and 115% of the prescribed dose. In addition, the V100 (volume receiving 100% of prescribed dose) should be between 98% and 100%. 2. The prescribed or minimum peripheral dose is delivered to the target volume. 3. The dose to any segment of the rectum wall should be limited to less than 75% of the prescription dose. 4. The bladder neck should be limited to 80% to 85% of the prescription dose. 5. The urethral dose in transurethral resection of the prostate (TURP) patients should be less than 103% for monotherapy or 103% to 110% for combined HDR and EBRT. For non-TURP cases, the urethral dose should be limited to 120%.
Dose Prescription The dose prescription, number of fractions, and interval between implants varies among institutions. HDR Boost A variety of HDR fractionation schemes exist. In addition, the number of implantations performed also varies. Table 2 shows the published combined EBRT and HDR brachytherapy dose prescription protocols for several groups. The minimum time between fractions is usually 6 hours. Monotherapy Very few institutions have adopted the concept of HDR brachytherapy as a monotherapy modality. Table 3 shows some published dose schemes.
Treatment Needles Identification Prior to the first treatment and any subsequent fraction, needles have to be identified, labeled, and confirmed by a second individual. This will serve two purposes: (1) to avoid a misadministration if transfer tubes are connected to the wrong Flexigude needles; and (2) to save time when treatment is delivered.
Treatment Plan Verification The treatment plan should be verified by a second member of the physics staff, and special attention should given to parameters such as prescription dose, source index, source activity, and source-stepping (2.5 mm, 5 mm, 10 mm).
32–HDR Brachytherapy for Prostate
653
Table 2. External Beam Radiotherapy (EBRT) and HDR Dose Prescriptions Group
EBRT (Gy)
Total dose (Gy)
Interval between implants
Seattle Prostate Institute Seattle, WA (1989–1995)
50
4.0 × 4
1
16
N/A
Seattle Prostate Institute Seattle, WA (1996–present)
45
5.5 × 3
1
16.5
N/A
William Beaumont Hospital Royal Oak, MI
46
5.5 × 1 6.0 × 1 6.5 × 1
3
16.5 18.0 19.5
1 week
California Endocurietherapy Cancer Center Oakland, CA
36
6.0 × 3
2
36.0
1 week
University Hospital Brachytherapy Center Kiel, Germany
50
15.0 × 1
2
30.0
1 week
Goteburg, Sweden
50
10.2 × 1
2
20.4
2 weeks
39.60
6.0 × 4
1
24.0
N/A
Long Beach Memorial Medical Center Long Beach, CA
HDR fractionation Number (Gy × fx) of implants
(Reprinted from Brachytherapy, vol. 13, issue 3, “High dose rate brachytherapy in the treatment of prostate cancer,” R. R. Rodriguez, J. D. Demanes, and G. A. Altieri, pp. 503–523. © 1999, with permission of Elsevier.)
Table 3. HDR Monotherapy Dose Scheme Institution William Beaumont Hospital Royal Oak, MI Long Beach Memorial Medical Center Long Beach, CA
HDR Fractionation (Gy × fx)
Number of implants
Total dose (Gy)
Interval between implants
7×3 9.5 × 2
2 2
42 38
2 weeks 2 weeks
6×2
2
24
2 to 3 weeks
654
Zoubir Ouhib
Independent Calculation An independent calculation should be performed to verify the accuracy of the plan. Individual needle treatment time or total time should be the focus.
Visual Inspection of the Implant Prior to each treatment following CT or ultrasound for real time, template and needles should be visually checked for any possible shift. Marks on the template in relation to the skin and others on the needles in relation to the template are useful as reference points. It has been reported that there is significant caudal displacement of the needles between HDR fractions (Damore et al., 2000). The needles need to be readjusted to their initial location (either advanced or retracted).
Patient Positioning The patient position should be kept comparable to the one during CT if real-time planning is not being used. Modification of the patient’s leg position between the time of image acquisition and treatment delivery could result into a modified dose to the target and organs at risk.
Quality Assurance (QA) Procedures Quality assurance (QA) procedures are used throughout the duration of the implant. They can be divided into three different stages:
Pretreatment Stage Items Related to the Patient • Patient identification: the patient must be identified by more than one method • Available complete prescription: the prescription should include the area of treatment, the dose/fraction, total dose, number of fractions, treatment area, elapsed time between fractions and implants, dose to critical organs, and any other important directives related to the plan of treatment • Prior radiation therapy (EBRT or brachytherapy) treatment should clearly be stated or communicated in writing to the planning physics staff Items Related to the HDR Unit Several items must be checked for malfunction and possible repair or replacement prior to treatment. Table 4 shows the list of those items. Items Related to the Treatment Plan The list of items related to the treatment-planning portion that needs to be evaluated is provided in Table 5. Items Between Treatment Plan and HDR Treatment Unit Several items that need to be checked for agreement between the plan and the HDR console treatment unit are listed in Table 6.
32–HDR Brachytherapy for Prostate Table 4. QA Checklist of Items Related to HDR Unit Communication equipment: camera and intercom Door interlock system Warning light outside the room Radiation monitor inside and outside the HDR room Hand-held radiation monitor Survey meter Treatment interrupt system Emergency stop Timer termination Console source position indicator Treatment complete indicator Transfer tubes condition Catheter attachment lock system Date and time on control unit Autoradiograph confirming the source position Visual inspection of the HDR unit
Table 5. QA Checklist for the Treatment Planning System Dose/Fraction/Total dose comparison between plan and prescription Source index between plan and measured value for the Flexiguide/needles Critical organ(s) dose is acceptable Step size used is the intended one Correct source decay Correct prescription line selected for dose assignment Independent check confirmation Reasonable dwell time for each needle and all needles combined based on past experience
Table 6. QA Items To Be Compared between the Treatment Plan and the HDR Treatment Console Patient name is identical Plan identification number and appropriate fraction Source strength Source index Step size Dwell positions and appropriate dwell time as well as total treatment time Channel number verification Appropriate transfer tube used Channels and needles number matching
655
656
Zoubir Ouhib
Treatment Stage • Monitor patient movement and comfort (view and listen) • Monitor treatment (channel number, dwell time, source position) • Avoid any distraction and be prepared to respond to a possible emergency • Compare backup timer to control unit timer
Post Treatment Stage • Verify that treatment was complete • Confirm with room monitors • Survey the patient • Perform visual inspection of the template and needles
Conclusion Because the HDR brachytherapy treatments are administered within a short period of time (minutes) and on an outpatient basis, the procedure is well tolerated by patients. This procedure has allowed the brachytherapist to deliver a tumoricidal dose of radiation conformally to the prostate while minimizing the dose to the bladder, urethra, and rectum. The radiation safety aspect of it for all hospital staff is optimal. Published data have shown high local clinical and biochemical control rates. While the preplan method is still widely used, there is a growing interest in the real-time method. This is driven mainly by the ability to complete the treatment within a reasonable time (first treatment can be given while patient is still in the OR) and also to create a conformal treatment plan during the implantation procedure. Interim results suggest biochemical control rates as good as or better than other forms of radiotherapy. Formal comparison or randomized, controlled trials to compare this approach with other therapies (single or combined) are needed in the near future.
References Damore, S. J., A. M. N. Syed, A. Puthawala, and A. Sharma. (2000). “Needle displacement during HDR brachytherapy in the treatment of prostate cancer.” Int J Radiat Oncol Biol Phys 46(5):1205–1211. Holm, H. H., N. Juul, J. F. Pedersen, H. Hansen, and I Stroyer. (1983). “Transperineal 125-iodine seed implantation in prostatic cancer guided by transrectal ultrasonography.” J Urol 130:283–286. (2000) J Urol 167:985–988. Discussion 988–989. Rodriguez, R. R., J. D. Demanes, and G. A. Altieri. (1999). “High dose rate brachytherapy in the treatment of prostate cancer.” Brachytherapy 13(3):503–523. Syed, A. M. N., A. Puthawala, P. Austin, J. Cherlow, J. Perley, L. Tansey, A. Shanberg, D. Sawyer, R. Baghdassarian, B. Wachs, et al. (1992). “Temporary iridium-192 implant in the management of carcinoma of the prostate.” Cancer 69:2515–2524. Whitmore, W. F., B. S. Hilaris, H. Grabstald, and M. Batata. (1974). “Implantation of 125I in prostatic cancer.” Surg Clin North Am 54:887–895.
Chapter 33
Modern Advances in Prostate Brachytherapy Eugene P. Lief, Ph. D. Maimonides Comprehensive Cancer Center Brooklyn, New York New Treatment Isotopes and Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657 Seeds with the 131Cs Radioactive Isotope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657 RadioCoilTM 103Pd Wire Line Source for Permanent Prostate Implants . . . . . . . . . . . . . . . . . . . . . . . 658 169 Ytterbium Low Energy Source for LDR and HDR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660 Innovative Planning of the Permanent Prostate Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664 Genetic Algorithm for Intraoperative Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664 Intraoperative Dynamic Dose Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665 Biologically Based Implant Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665 Recent Developments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666 New Type of LDR Sources for Better Ultrasound Localization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666 Robotically Assisted Permanent Prostate Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 666 Shielded Mick® 200-TPV Applicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669
Brachytherapy of the prostate is a rapidly developing field. Some of the new advances include use of the new isotopes and sources, innovative planning, and use of modern imaging tools. In addition, commercial treatment-planning systems are becoming much more versatile and user-friendly. Most of them have a significantly improved graphic user interface, and some systems are now capable of intraoperative planning. All major software packages can quickly perform calculations for various types of sources, both for planning and post-implant evaluation. We have already mentioned some of the modern advances in the chapter on the treatment delivery. In this chapter, we will discuss several most recent developments.
New Treatment Isotopes and Sources Seeds with the 131Cs Radioactive Isotope A new source with the 131Cs radioactive isotope has been developed by IsoRay, Inc. (Richland, WA), based on the invention made in 1967 by Donald C. Lawrence. The seed received U. S. Food and Drug Administration (FDA) 510(k) clearance in 2003. The 131Cs isotope was suggested for brachytherapy use back in 1960s (Henschke and Lawrence 1965). 131Cs is a γ-ray emitter with the most prominent peaks in the 29 to 34 keV regions. The main distinction of the new isotope from 125I and 103Pd conventionally used in permanent prostate implants is a significantly shorter half-life of 9.7 days versus 60 and 17 days for 125I and 103Pd, respectively. According to some recent radiobiological data (Brenner et al., 2002; Armpilia et al., 2003), use of an isotope with a shorter half-life is much more biologically effective in the prostate cancer treatment. Other possible applications include treatment of malignant tumors in the breast, head and neck, lung, and pancreas. Pacific Northwest National Laboratory’s (PNNL’s) Medical Seed Lab and the National Institute of Standards and Technology (NIST) evaluated dosimetric properties of the new seed model CS-1 with radioactive 131Cs isotope (Murphy et al., 2004) using AAPM Task Group 43 (TG-43) recommendations (Nath et al., 1995; Williamson et al., 1998). Model CS-1 of 131Cs brachytherapy seed (Figure 1) is assembled in a 4.3-mm long titanium tube with the outer diameter of 0.8 mm and a wall thickness of 0.05 mm (Murphy et al., 2004). Two laser-welded 0.1 mm caps on each end of the tube make the total length equal
657
658
Eugene P. Lief
Figure 1. Scheme of the new seed with 131Cs isotope developed at IsoRay, Inc. (Richland, WA). [Courtesy of John T. Hrobsky, IsoRay, Inc.]
to 4.5 mm. The x-ray marker is a centrally located gold wire, 4.1 mm long and 0.25 mm in diameter. The wire is placed in a glass and ceramic tube with 0.4 mm inside and 0.65 mm outside diameters. The length of the core is 4.2 mm to allow clearance inside the tube for fit-up and welding. Results of PNNL and NIST study of the CS-1 seed with radioactive 131Cs (Murphy et al., 2004) yielded a dose rate constant of (0.888 ±0.019) cGy/(h·U) in Virtual Water™ from Med-Cal, Inc. (Verona, WI), corresponding to (0.915±0.020) cGy/(h·U) in real water, when the conversion factor of 1.031 is used. The value 1.031 of the conversion factor was calculated for the distance of 1.0 cm from the source, using the Monte Carlo N-Particle code (MNCP), version 4C with photon cross-section library DLC-2000 (Briesmeister 2000; DeMarco, Hugo, and Solberg 2002; Rivard 2001). Over the range of distances from 0.5 to 7.0 cm, the value of the calculated conversion factor changes from 1.004 to 1.24, respectively. For seeds with a typical internal activity of 4.5 mCi (0.637 U/mCi), the dose rate constants listed above result in absorbed dose rates at 1.0 cm from the source equal to 1.72 cGy/h and 1.78 cGy/h for Virtual Water and liquid water, respectively. The radial dose function of the new seed was similar to that of a number of different seeds with the radioactive 125I. Measurements yielded an anisotropy function that maximizes at a value of 0.71±0.01 in the region that is defined by 0° to 10° and 0.5 to 1.0 cm (Murphy et al., 2004). This anisotropy compares favorably to the 0° anisotropy of 0.3 to 0.8 for most 125I and 103Pd seeds currently available (Weaver 1998; Luxton and Jozsef 1999; Kirov and Williamson 2001).
RadioCoil™ 103Pd Wire Line Source for Permanent Prostate Implants Prostate seed implantation has several difficulties. Seed migration can change the target dose distribution and also move the seed to undesirable areas, such as the lungs (Ankem et al., 2002). In addition, during the implantation, seeds may cluster together creating high- and low-dose regions. There were several attempts to resolve these problems. Amersham Healthcare, Inc. (Nycomed Amersham, United Kingdom) has developed RapidStrand™ 125I source composed of Model 6711 125I seeds enclosed within a stiff, braided, absorbable Vicryl™ suture material 1 cm from each other (Kumar and Good 1986; Butler and
33–Modern Advances in Prostate Brachytherapy
659
Merrick 1996). Although this invention was an improvement (Tapen et al., 1998; Lee et al., 2002), the Vicryl is hygroscopic. It can soften and swell when exposed to body fluids, which may cause jamming in the needle if not handled properly (see the chapter on prostate implant delivery). Another innovation, Vari-Strand™, a flexible monofilament strand, was introduced by Advanced Care Pharmacy (Oxford, CT). These strands are composed of selectable seeds embedded into a synthetic bio-absorbable material. Seeds, selected by the user, are embedded in the strands with variable spacing by pharmacy, with additional time and cost required. Recently, RadioMed Corporation (Tyngsboro, MA) introduced a new 103Pd linear source, RadioCoil™, in the form of a 0.2-mm wide and 0.05-mm thick ribbon coiled in a dense helix with the diameter of 0.35 mm (Meigooni et al., 2004). The active length of the new source is variable: it could be 0.5 cm (for research purposes) or any integer number from 1 cm to 5 cm ± 0.5 mm (for clinical use). Such a design of the source improves its fixity and possibly its ultrasonic visibility. The device received FDA 510(k) clearance in December 2001. Individual wires are delivered to users in stainless steel cartridges, which interface directly with a specially designed 19-gauge delivery needle with thicker walls to improve needle rigidity and to ease the insertion. The RadioCoil source is fabricated from a ribbon of high-purity rhodium activated in a cyclotron to produce 103Pd uniformly distributed in the material, coiled in the helix, and cut to length. The targeted activity is usually 0.1 to 2.5 mCi (0.13 to 3.2 U) per linear centimeter. Measurements in Solid Water™ yielded the dose rate constant Λ values of 0.641 and 0.600 cGy/(h·U) for 0.5-cm and 1-cm sources, respectively (Meigooni et al., 2004). Monte Carlo simulations for the same sources in liquid water suggested the values of 0.650 and 0.597 cGy/(h·U) for the same sources, respectively. For 2.0 to 5.0 cm sources in liquid water, the dose rate constants found by Monte Carlo calculations were 0.457, 0.350, 0.278, and 0.230 cGy/(h·U), respectively. Unfortunately, sources longer than 1 cm cannot be calibrated by NIST using the Wide Angle Free Air Chamber (WAFAC) instrument; therefore, a well chamber is the only way to verify the strength of the 2 to 5 cm wires. The Monte Carlo calculated radial dose function in liquid water was fitted to a modified fifth-order polynomial using the formula: g(r ) = (a0 + a1r + a2 r 2 + a3r 3 + a4 r 4 + a5 r 5 ) ⋅ e− br.
(1)
Parameters of the fit (1) for 0.5-cm and 1-cm sources (Meigooni et al., 2003) are listed in Table 1. Comparison of the dose distribution calculations performed by a commercial treatment-planning system Table 1. Numerical values of parameters of the polynomial fit (1) of the radial dose function g(r) for 0.5-cm and 1.0-cm long RadioCoil™ 103Pd linear sources, calculated by the Monte Carlo method (After Meigooni et al., 2003 and Meigooni et al., 2004). Parameters
Values for the source length 0.5 cm
1.0 cm
a0
1.1081
1.0698
a1
3.8193
2.8626
a2
–3.0605
–1.2438
a3
2.6018
1.0687
a4
–0.7296
–0.2644
a5
0.0853
0.0316
b
1.3521
1.2630
660
Eugene P. Lief
and a Monte Carlo code PTRAN, version 7.43 (Williamson 1988) showed a discrepancy of up to 16% (Meigooni et al., 2004). The difference can be attributed to the treatment planning neglect of the self-attenuation for the linear source model and a lack of two dimensional (2-D) anisotropy considerations for the point source model. Further investigation is necessary for more precise calculation of the dose distribution from RadioCoil 103Pd source by commercial treatment-planning systems. 169
Ytterbium Low-Energy Sources for LDR and HDR
169
Yb is a rare-earth element which decays to stable thulium, 169Tm, by electron capture with a half-life of 32.0147±0.0093 days (Unterweger, Hoppes, and Schima 1992). 169Yb emits a spectrum of photons with energies ranging from 49.8 to 307.7 keV (Figure 2). Calculation of transmission (Munro 2004) shows that gold foil with thickness of only 0.2 mm reduces the dose rate by more than 50%, and a 1-mm thickness reduces it by a factor of 10 (Figure 3). Comparison with 192Ir on the shielding thickness, required in tungsten, lead, and concrete (Figures 4 through 6), shows that the requirements for 169Yb are significantly less stringent (Munro 2004). Due to these properties of the new isotope, it was considered for brachytherapy. First, the innovative use of 169Yb for brachytherapy seeds was found advantageous over 125I due to better dose uniformity and higher initial dose rate (MacPherson and Battista 1995). Radiobiologically, 169Yb can be used for permanent implants for slowly proliferating tumors if the initial dose rate was sufficiently high (Orton 1974; Lazarescu and Battista 1997). Because 169Yb has an average energy of 93 kV, radiation safety would be problematic in a permanent seed implant compared to 125I, which has a lead half-value thickness of 0.02 mm and a tissue tenth-value thickness of only a few centimeters. Two types of low dose rate (LDR) sources with 169Yb were produced by Amersham International, Inc.: type 6 and type 8. Type 6 consists of four Yb2O3 spheres with density 8.5 g/cm3, 0.5 mm in diameter, sealed within a titanium shell 4.5 mm long and 0.838 mm in outer diameter, with a wall thickness of 0.076 mm (MacPherson and Battista 1995). Hemispherical titanium end caps are laser-welded to the ends of the shell. The ytterbium oxide spheres are immobilized within the seed by two titanium spacers (0.4 mm length × 0.6 mm diameter) placed near the ends of the seed. The newer type 8 seed uses a
Figure 2. Measured energy spectrum of 169Yb decay from the new low-energy source developed by Implant Sciences Corporation. [Courtesy of John J. Munro III, Ph.D., Implant Sciences Corporation, Wakefield, MA.]
33–Modern Advances in Prostate Brachytherapy
661
Figure 3. Broad-beam transmission of photons from 169Yb through platinum (ρ=21.45 g/cm3) and gold (ρ=19.32 g/cm3). The lines (solid: platinum and dashed: gold) represent the calculated transmission and the solid circles, the measured data for platinum. [Courtesy of John J. Munro III, Ph.D., Implant Sciences Corporation, Wakefield, MA.]
Figure 4. Calculated broad-beam transmission of photons from 169Yb (solid curve) and 192Ir (dashed curve) through tungsten (ρ=17.00 g/cm3). [Courtesy of John J. Munro III, Ph.D., Implant Sciences Corporation, Wakefield, MA.]
662
Eugene P. Lief
Figure 5. Broad-beam transmission of photons from 169Yb (solid curve and solid circles) and 192Ir (dashed curve) through lead (ρ=11.34 g/cm3). The lines represent calculated transmission and the circles, measurements performed for 169Yb. [Courtesy of John J. Munro III, Ph.D., Implant Sciences Corporation, Wakefield, MA.]
Figure 6. Calculated broad-beam transmission of photons from 169Yb (solid curve) and 192Ir (dashed curve) through concrete (ρ=2.30 g/cm3). [Courtesy of John J. Munro III, Ph.D., Implant Sciences Corporation, Wakefield, MA.]
33–Modern Advances in Prostate Brachytherapy
663
cylinder of pure ytterbium metal, density 6.5 g/cm3, 2.5 mm long and 0.3 mm in diameter, encapsulated in a titanium tube 4 mm long, 0.075 mm thick, and 0.51 mm in diameter. There are no end spacers for this type of seed. Monte Carlo calculations suggest the value of the dose rate constant Λ0 to be equal to 1.25 ±0.05 cGy/U (MacPherson and Battista 1995). In 2004, Implant Sciences Corporation (Wakefield, MA) developed a low-energy/high-activity source of 169Yb for use in high dose rate (HDR) brachytherapy (Munro 2004; Medich and Munro 2005). The new source has several advantages. First, due to a much lower γ-ray energy than for a conventional 192Ir HDR source, the new source will need significantly less shielding. This difference can drastically reduce the cost and difficulties of room preparation and make operating the HDR unit significantly easier. Thus, a 10 Ciapp 192Ir source located at a distance of two meters from an exterior wall would require 47 cm of concrete (Figure 6) or 36 mm of lead (Figure 5) to reduce the exposure rate to less than 2 mR/hr. A 20 Ciapp 169Yb source, which provides the same treatment time as a 10 Ciapp 192Ir source in the same location, would require only 31 cm of concrete or 11 mm of lead for the same protection. Second, the new source has the possibility of shielded applicators used for treatment that can produce a significant modulation of the delivered dose distribution. The increased shielding in the applicators and rapidly falling radial dose function (Figure 7) could better protect such critical organs as the urethra in the prostate treatment, skin in the treatment of the breast, as well as rectum and bladder in gynecological applications. Third, the use of local shielding around the treatment site will reduce the dose to the distant areas, such as contralateral breast and will possibly enable new types of treatments, e.g., pediatric tumors. The 169Yb HDR source is designed compatible with typical HDR afterloaders (Figure 8). The source is doubly encapsulated with 4.5 mg/mm3 titanium inner capsule and a stainless steel 7.8 mg/mm3 outer capsule and is attached to a 210-cm long stainless steel cable to permit manipulation by the remote afterloader. The active element, made of ytterbium oxide with the density of 6.90 mg/mm3, has a diameter of 0.65 mm and a length of 4 mm (Munro 2004; Medich and Munro 2005). The outer capsule is less than 1 mm in diameter and less than 6 mm long. The dose rate constant Λ was calculated to be 1.179 cGy/(U·h).
Figure 7. Comparison of the radial dose functions for photons from radioactive 169Yb (solid curve) and 192Ir (dashed curve). [Courtesy of John J. Munro III, Ph.D., Implant Sciences Corporation, Wakefield, MA.]
664
Eugene P. Lief
The value of the anisotropy constant φ an was obtained by using 1/r2-weighting of φ an (r ≥ 1 cm ). The result of this calculation was φ an = 0.93.
Innovative Planning of the Permanent Prostate Implants Genetic Algorithm for Intraoperative Planning Intraoperative planning of the permanent prostate implants is based on the ultrasound images acquired at the beginning of the procedure. The advantages of such planning include “reproducibility” of the planned position, significant simplification of the process, and time saving. The reproducibility is achieved by definition, since unlike the computed tomography (CT)-based planning, the images are acquired in the treatment position and the patient does not move between the image acquisition and the implant delivery. This eliminates the alignment of the patient in the treatment position with the planned position and simplifies the overall process. From the patient’s prospective, this approach makes the preoperative CT or ultrasound unnecessary and saves the patient an extra trip to the hospital. The preoperative planning is typically computerized and takes about 10 to 15 minutes. As an additional advantage, this type of planning leads to the next more advanced step—intraoperative dynamic dose optimization, which we will describe later. Earlier approaches to the intraoperative planning were based on a genetic algorithm used for computerized seed placement optimization (Yu and Schell 1996; Lee et al., 1999). The idea of the integer linear programming model is to consider a three-dimensional (3-D) grid defined by possible discrete needle
Figure 8. Scheme of the new 169Yb HDR source developed by Implant Sciences Corporation. The assembly consists of an inner source encapsulated in titanium and placed into the outer source capsule made of stainless steel. The capsule is securely attached to a stainless steel cable that is driven by an HDR remote afterloader. [Courtesy of John J. Munro III, Ph.D., Implant Sciences Corporation, Wakefield, MA.]
33–Modern Advances in Prostate Brachytherapy
665
positions defined by the template and possible discrete positions of the implanted seeds along these needles defined by the transverse image planes. Number 1 or 0 is assigned to each possible seed position, depending whether the seed was or was not implanted there. The radiation dose to the target and surrounding healthy tissue and critical organs can be modeled as a linear combination of the indicator variables. A system of linear constraints is employed to keep the dose at each point within the desired limits. Since it is generally impossible to exactly satisfy all the conditions simultaneously, the program calculates the objective function, which it tries to optimize. Two computational approaches: a branchand-bound algorithm and a genetic algorithm were considered for the optimization (Lee et al., 1999). Both approaches were found practically useable and capable of performing calculations in 5 to 15 minutes; however, the algorithms were quite sensitive to the model parameters, which therefore have to be carefully chosen. The computer-based optimization resulted in better plans and superior dose distributions found in post-implant analysis (Yu and Schell 1996; Kaplan et al., 2000; Zelefsky et al., 2000; Nag et al., 2001; Stone et al., 2003; Raben et al., 2004).
Intraoperative Dynamic Dose Optimization The next development in the operative planning was dynamic dose optimization, which is performed during the delivery of the implant (Zaider et al., 2000a; Lee and Zaider 2003; Todor et al. 2003). Generally, the advantage of this approach as compared with the previous is the dosimetric feedback during the procedure. The feedback shows the difference between the planned and the actual locations of the deposited seeds. This deviation can be expressed dosimetrically by showing the difference between the intended and the actual dose distributions. The program also suggests how to compensate for this difference by implanting additional seeds and/or changing planned positions of the seeds still remaining to be implanted. By following the suggested modified plan, the radiation oncologist can achieve the actual dose distribution as close as possible to the planned one. Specifically, the dynamic dose optimization implementation can achieve two main goals: (1) performing both planning and post-implant evaluation during the procedure and making adjustments as necessary; and (2) performing the plan and its evaluation using the same imaging technique—ultrasound with possible addition of fluoroscopy for seed and needle localization (Todor et al., 2003). Necessary elements of this approach include automatic seed detection and registration of fluoroscopic and ultrasound images. It is recommended to check the implant dosimetry and to reoptimize the plan 2 to 4 times during the procedure (Lee and Zaider 2003). When only two reoptimizations are performed, it is suggested to perform the first one after 50% of seeds are implanted, and the second one after finishing delivery of the reoptimized plan. At that point, the radiation oncologist checks possible “cold spots” in the target coverage and if necessary implants more seeds suggested by the second plan reoptimization. Similarly for the three-reoptimization technique when the dosimetric feedback is obtained after one-third of the initially planned seeds have been implanted, than after one-half of the reoptimized plan has been delivered, and the last one—after all the seeds from the twice reoptimized plan have been delivered. The four-reoptimization technique is performed similarly to the previous two. Use of the intraoperative plan reoptimizations significantly improved the dose distribution achieved (Lee and Zaider 2003; Potters et al., 2003).
Biologically Based Implant Planning The idea of biologically based planning for brachytherapy of the prostate cancer (Zaider et al., 2000b, 2005; Mizowaki et al., 2002; Susil et al. 2003, 2005a,b; Ménard et al., 2004) is to define the target volume based on the regions of cancer in the prostate while sparing regions with normal prostatic tissue. As it was demonstrated, the magnetic resonance spectroscopic imaging (MRSI) of the prostate can distinguish both
666
Eugene P. Lief
malignant and benign regions in parenchyma. The principle of this detection is the increase in the choline plus creatine versus citrate ratio found in malignant regions of the prostate. MRS is capable of measuring this ratio and identifying suspicious regions in the prostate. Knowing the probable location of the tumor within the prostate allows designing the plan better covering the suspicious regions while better sparing normal tissue of the prostate and the adjacent critical organs. After the plan is finished, it could be delivered using the dynamic dose optimization as described above.
Recent Developments New Type of LDR Sources for Better Ultrasound Localization Nowadays, ultrasound is the most popular intraoperative imaging modality for the permanent prostate implant (Merrick, Wallner, and Butler 2003). It allows prostate visualization and by minimizing the use of fluoroscopy, significantly reduces total radiation exposure to the operating room personnel (Schwartz et al., 2003). The problem with the ultrasound is the limited visibility of the implanted seeds. Part of the problem is the dependency of the magnitude of the backscattered signal from the seed orientation. Seeds that are oriented with their long axis perpendicular to the incident ultrasound beam are more apt to be detectable by transrectal ultrasound (Davis et al., 2003). In contrast, seeds oriented parallel to the incident ultrasound beam will likely be undetectable using conventional transrectal ultrasound due to signal amplitude reduction and the noise of backscatter from typical anatomy. In order to improve seed echogenicity, a new model, EchoSeed™ (Amersham Health, Inc., Arlington Heights, IL), was introduced. The EchoSeed has a corrugated appearance with multiple orientations of its surfaces (Tornes and Eriksen 2001). It incorporates six alternating ridges and grooves in a regular sinusoidal pattern along the length of the seed. This form produces ultrasound backscatter that is more omnidirectional and less sensitive to seed orientation than for conventional seeds. A recent study (Davis et al., 2003) evaluated ultrasound-scattering properties for three different types of sources from Amersham, Inc.: OncoSeed™ (model 6711), EchoSeed™, and RapidStrand. Ultrasound images for angles of incidence varying from 90° (perpendicular) to 20° at 5 MHz and 7.5 MHz were produced by raster scanning the seeds in a degassed water bath. It was found that the corrugated seeds appeared as contiguous objects over the range of the experimental conditions, whereas for the standard seeds that occurred only from 90° to 80°. The ranges of the backscattered integrated-optical-density ratio of the seeds from 85° to 20° for the frequency of 7.5 MHz were: 1.26 to 3.77 (corrugated vs. standard) and 1.008 to 10.86 (corrugated vs. RapidStrand). Similarly, for 5 MHz, the ranges were 1.48 to 3.72 (corrugated vs. standard) and 1.21 to 9.53 (corrugated vs. RapidStrand). Backscattered signal increase ranged from 1.66 dB to 20.7 dB for the corrugated seeds compared with other models. Corrugated seeds produced greater backscatter and were more readily identifiable by the ultrasound than seeds of conventional shapes. The advantages of EchoSeed are not so marked in clinical studies, and it remains difficult to distinguish seeds from calcifications.
Robotically Assisted Permanent Prostate Implants An idea of the prostate brachytherapy performed by a robot was explored by a group of computer scientists from Johns Hopkins University (Masamune et al., 2001; Fichtinger et al., 2002, 2004, 2005a,b; Shen et al., 2004; Cleary et al., 2005). The group also studied a possibility of automatic seed reconstruction in post-implant evaluation (Jain et al., 2005). An automatic system has been developed and evaluated for transrectal ultrasound-guided prostate brachytherapy. The system consists of a new robotic brachytherapy delivery system integrated with an FDA-approved brachytherapy planning and monitoring suit (Fichtinger et al., 2005b). The planning computer prepares the intraoperative plan based on the ultrasound images.
33–Modern Advances in Prostate Brachytherapy
667
The robot inserts preloaded needles into the prostate, the physician approves the positions, and the seeds are implanted. The sources are tracked with ultrasound in real time, giving the opportunity to reoptimize the plan before the procedure is finished. In training phantoms, all clinically relevant locations inside the prostate were reached. The average transverse placement error was about 2 mm, with 2.5 mm in the worst case. The group claims technical viability of the robotic brachytherapy, which has been demonstrated in mechanical and anthropomorphic phantoms. The first clinical implementation of robotic seed implantation was accomplished using equipment from Nucletron Corporation (Columbia, MD). The FIRST™ (Fully Integrated Real-time Seed Treatment) system includes the SeedSelectron® (Figure 9)—an automated unit for seed implantation, an ultrasound unit with the probe, and a laptop computer (Rivard, Evans, and Kay 2004). The process starts with ultrasound image acquisition in multiple planes and building a 3-D image. The next step is contouring, which the radiation oncologist performs manually. After that, inverse planning is performed by a computer program, SPOT PRO™, using a simulated annealing method for the given set of volumes and constraints. Based on the accepted plan, the SeedSelectron automatically builds trains with any combination of radioactive seeds and spacers and automatically implants them through the needle manually inserted in the prostate. As the needle is implanted, a real-time correction for the deviation between the planned and the actual needle positions can be entered in the system if necessary. After all seeds are implanted, SPOT PRO calculates the cumulative dose distribution based on the actual seed positions. If necessary, additional seeds can be implanted before the procedure is over. The final step is CT-based seed localization and dose calculation for the post-implant evaluation. This is done based on post-operative CT. The SeedSelectron uses proprietary cartridges with seeds and spacers manufactured by Nucletron Corporation (Figure 9). A special dosimetric diode built into the SeedSelectron checks the activity of each seed before the implantation. If the measured value is out of a predefined range, the seed can be discarded in a container instead of implanted. Another array of diodes verifies that the sequence of radioactive seeds
Figure 9. SeedSelectron® from Nucletron Corporation (Columbia, MD) performs automatic prostate seed implantation based on a 3-D ultrasound image and inverse planning results. In the picture, the operator inserts two special magazines (one for seeds and the other for spacers) in the device prior to use. [Courtesy of Raymond Horn, Nucletron Corporaton.]
668
Eugene P. Lief
and nonradioactive spacers in any train is correct. As an additional quality assurance procedure, some seeds can be extracted from the magazine and assayed using a well chamber. If necessary, these seeds can be quickly sterilized after the assay and implanted using a conventional needle (see chapter 30 on the delivery techniques). For seed accounting purposes, an x-ray of the loaded magazine can be taken before and after the procedure. Seeds distinctly appear on conventional film images of the magazines.
Shielded Mick® 200-TPV Applicator As a result of constant development of the company’s products, Mick Radio-Nuclear Instruments, Inc. (Mount Vernon, NY) came up with a new applicator 200-TPV (Figure 10). There are several new features in this design. First, the receptor area of the applicator is shielded by heavy walls to improve the radiation protection for the users—first of all, to reduce the exposure to the user’s hands. Dose rate measurements at the top of the loaded magazine inserted in the applicator, yielded only 0.02 mR/h. Second, the dial-in seed spacing selector can be rotated to five distinct locations indicating the chosen distance between the consecutively implanted seeds: 7.5, 8.0, 10.0, 11.0, and 12.0 mm. The distance can be changed any time during the implant by just rotating the selector to another fixed position. This provides the flexibility necessary for a good implant. The variable distance, for instance, could be helpful to implant seeds closer to each other at the periphery of the gland, while keeping the sources further apart in the vicinity of the rectum and the urethra. As it was mentioned at the beginning, the prostate brachytherapy is a rapidly developing field. We wrote about several innovative approaches developed in the recent years but there are many more being designed. The new emerging advances include the cone-beam CT for HDR and LDR planning (Prestidge 2004), automated seed sorting algorithm for selective identification of the radioactive sources, ferromagnetic seeds, and fiducials (Davis et al., 2004), magnetic resonance imaging (MRI)-compatible robotic manipulator for the image-guided prostate treatment (Krieger et al., 2005) and many other topics that would further develop the practice of the prostate brachytherapy.
Figure 10. A new shielded Mick® 200-TPV applicator from Mick Radionuclear Instruments, Inc. (Mount Vernon, NY). The receptor area is shielded by heavy walls to minimize the exposure to the personnel. The dial-in seed spacing selector can be rotated to five distinct locations to select the spacing of 7.5, 8.0, 10.0, 11.0, and 12.0 mm. [Courtesy of Felix W. Mick.]
33–Modern Advances in Prostate Brachytherapy
669
References Ankem, M. K., V. S. DeCarvalho, A. M. Harandgozo, V. H. Hartano, M. Perrotti, K. Han, W. J. Shih, E. Malka, E. C. White, R. Maggio, R. Ioffreda, J. W. Goldsmith, and R. E. Weiss. (2002). “Implications of radioactive seed migration to the lungs after prostate brachytherapy.” Urology 59:555–559. Armpilia, C. I., R. G. Dale, I. P. Coles, B. Jones, and V. Antipas. (2003). “The determination of radiobiologically optimized half-lives for radionuclides used in permanent brachytherapy implants.” Int J Radiat Oncol Biol Phys 55:378–385. Brenner, D. J., A. A. Martinez, G. K. Edmundson, C. Mitchell, H. D. Thames, and E. P. Armour. (2002). “Direct evidence that prostate tumors show high sensitivity to fractionation (low α/β ratio) similar to late responding normal tissue.” Int J Radiat Oncol Biol Phys 52:6–13. Briesmeister, J. F. (2000). “MCNP—A general Monte Carlo N-particle transport code—Version 4C.” Los Alamos National Laboratory Report LA-13709-M. March 2000. Butler, W. M., and G. S. Merrick. (1996). “I-125 Rapid Strand™ loading technique.” Radiat Oncol Invest 4:48–49. Cleary, K., V. Watson, D. Lindisch, D. Stoianovici, A. Patriciu, D. Mazilu, R. H. Taylor, G. Fichtinger, S. Xu, C. S. White, J. Donlon, and M. Taylor. (2005). “Precision placement of instruments for minimally invasive procedures using a needle driver robot.” Int J Med Robot Comput Assist Surg 1:40–47. Davis, B. J., R. R. Kinnick, M. Fatemi, E. P. Lief, R. A. Robb, and J. F. Greenleaf. (2003). “Measurement of the ultrasound backscatter signal from three seed types as a function of incidence angle: application to permanent prostate brachytherapy.” Int J Radiat Oncol Biol Phys 57:1174–1182. Davis, B. J., Brinkmann, D. H., Kruse, J. J., Herman, M. G., LaJoie, W. N., Schwart, D. J., Pisansky, T. M., and R. W. Kline. (2004). “Selective identification of different brachytherapy sources, ferromagnetic seeds, and fiducials in the prostate using an automated seed sorting algorithm.” Brachytherapy 3:106–112. DeMarco, J. J., G. Hugo, and T. D. Solberg. (2002). “Dosimetric parameters for three low-energy brachytherapy sources using the Monte Carlo N-particle code.” Med Phys 29:662–668. Fichtinger, G., A. Krieger, R. C. Susil, A. Tanacs, L. L. Whitcomb, and E. Atalar. “Transrectal Prostate Biopsy Inside Closed MRI Scanner with Remote Actuation, Under Real-Time Image Guidance.” Proceedings to Medical Image Computing and Computer-Assisted Interventions 2002. Lecture Notes in Computer Science, Vol. 2488, Part 1. New York: Springer-Verlag, pp. 91–98, 2002. Fichtinger, G., T. L. DeWeese, A. Patriciu, A. Tanacs, A. Mazilu, J. H. Anderson, K. Masamune, R. H. Taylor, and D. Stoianovici. (2004). “Robotically assisted prostate biopsy and therapy with intra-operative CT guidance.” Acad Radiol 9:60–74. Fichtinger, G., A. Deguet, G. Fischer, E. Balogh, K. Masamune, R. H. Taylor, L. M. Fayad, and S. J. Zinreich. (2005a). “CT image overlay for percutaneous needle insertions.” Comput Assist Surg (in review). Fichtinger, G., E. C. Burdette, A. Tanacs, A. Patriciu, D. Mazilu, L. L. Whitcomb, and D. Stoianovici. (2005b). “Robotically assisted prostate brachytherapy with transrectal ultrasound guidance – preliminary experiments.” Brachytherapy (in review). Henschke, U. K., and D. C. Lawrence. (1965). “Cesium-131 seeds for permanent implants.” Radiology 85:1117–1119. Jain, A., Y. Zhou, T. Mustufa, E. C. Burdette, G. S. Chirikjian, and G. Fichtinger. (2005). “Matching and Reconstruction of Brachytherapy Seeds using the Hungarian Algorithm (MARSHAL).” Med Phys (in review). Kaplan, I. D., E. J. Holupka, P. Meskell, S. J. Soon, B. Saltzman, P. Church, and G. P. Kearney. (2000). “Intraoperative treatment planning for radioactive seed implant therapy for prostate cancer.” Urology 56:492–495. Kirov, A. S., and J. F. Williamson. (2001). “Monte Carlo-aided dosimetry of the Source Tech Medical Model STM125I 125I interstitial brachytherapy source.” Med Phys 28:764–772. Krieger, A., R. C. Susil, C. Menard, J. A. Coleman, G. Fichtinger, E. Atalar, and L. L. Whitcomb. (2005). “Design of a novel MRI compatible manipulator for image guided prostate intervention.” IEEE Trans Biomed Engr 52:306–313. Kumar, P. P., and R. R. Good. (1986). “Vicryl carrier for I-125 seeds: Percutaneous transperineal insertion.” Radiology 159:276. Lazarescu, G., and J. J. Battista. (1997). “Analysis of the radiobiology of ytterbium-169 and iodine-125 permanent brachytherapy implants.” Phys Med Biol 42:1727–1736.
670
Eugene P. Lief
Lee, E. K., and M. Zaider. (2003). “Intraoperative dynamic dose optimization in permanent prostate implants.” Int J Radiat Oncol Biol Phys 56:854–861. Lee, E. K., R. J. Gallagher, D. Silvern, C. S. Wuu, and M. Zaider. (1999). “Treatment planning for brachytherapy: An integer programming model, two computational approaches and experiments with permanent prostate implant planning.” Phys Med Biol 44:145–165. Lee, W. R., A. F. deGuzman, S. K. Tomlinson, and D. L. McCullough. (2002). “Radioactive sources embedded in suture are associated with improved postimplant dosimetry in men treated with prostate brachytherapy.” Radiother Oncol 65:123–127. Luxton, G., and G. Jozsef. (1999). “Radial dose distribution, dose to water and dose rate constant for monoenergetic photon point sources from 10 keV to 2 MeV: EGS4 Monte Carlo model calculation.” Med Phys 26:2531–2538. MacPherson, M.S., and J. J. Battista. (1995). “Dose distributions and dose rate constants for new ytterbium-169 brachytherapy seeds.” Med Phys 22:89–96. Masamune, K., G. Fichtinger, A. Patriciu, R. C. Susil, R. H. Taylor, L. R. Kavoussi, J. H. Anderson, I. Sakuma, T. Dohi, and D. Stoianovici. (2001). “System for robotically assisted percutaneous procedures with computed tomography guidance.” Comput Assist Surg 6:370–383. Medich, D. C., and J. J. Munro III. (2005). “Monte Carlo dosimetric characterization of an ytterbium-169 high dose rate brachytherapy source.” Med Phys (in review). Meigooni, A. S., H. Zhang, C. Perry, S. A. Dini, and R. A. Koona. (2003). “Theoretical and experimental determination of dosimetric characteristics for BrachySeed Pd-103 model Pd-1 source.” Appl Radiat Isot 58:533–541. Meigooni, A. S., H. Zhang, J. R. Clark, V. Rachabatthula, and R. A. Koona (2004). “Dosimetric characteristics of the RadioCoil™ 103Pd wire line source for use in permanent brachytherapy implants.” Med Phys 31:3095–3105. Ménard, C., R. C. Susil, P. Choyke, J. Coleman, R. Grubb, A. Gharib, A. Krieger, P. Guion, D. Thomasson, K. Ullman, S. Gupta, V. Espina, L. Liotta, E. Petricoin, L. L. Whitcomb, G. Fichtinger, E. Atalar, N. C. Coleman, and K. Camphausen. (2004). “An interventional MRI technique for the molecular characterization of intra-prostatic dynamic contrast enhancement.” Molecular Imag (in press). Merrick, G. S., K. E. Wallner, and W. M. Butler. (2003). “Permanent interstitial brachytherapy for the management of carcinoma of the prostate gland.” J Urol 169:1643–1652. Mizowaki, T., G. N. Cohen, A. Y. Fung, and M. Zaider. (2002). “Towards integrating functional imaging in the treatment of prostate cancer with radiation: The registrations of the MR spectroscopy imaging to ultrasound/CT images and its implementation in treatment planning.” Int J Radiat Oncol Biol Phys 54:1558–1564. Munro III, J. J. (2004). “169Ytterbium low energy gamma source for high dose rate brachytherapy.” Implant Sciences Corporation Edition. Wakefield, MA. Murphy, M. K., K. Piper, L. R. Greenwood, M. G. Mitch, P. J. Lamperti, S. M. Seltzer, M. J., Bales, and M. H. Philips. (2004). “Evaluation of the new cesium-131 seed for use in low-energy x-ray brachytherapy.” Med Phys 31:1529–1538. Nag, S., J. P. Ciezki, R. Cormack, S. Doggett, J. K. DeWyngaert, G. K. Edmundson, R. G. Stock, N. N. Stone, Y. Yu, and M. J. Zelefsky. (2001). “Intraoperative planning and evaluation of permanent prostate brachytherapy: Report of the American Brachytherapy Society.” Int J Radiat Oncol Biol Phys 51:1422–1430. Nath R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43. American Association of Physicists in Medicine.” Med Phys 22:209–234. Also available as AAPM Report No. 51. Orton, C. (1974). “Time-dose factors in brachytherapy.” Br J Radiol 47:603–607. Potters, L., E. Calugaru, K. B. Thornton, T. Jackson, and D. Huang. (2003). “Toward a dynamic real-time intraoperative permanent prostate brachytherapy methodology.” Brachytherapy 2:172–180. Prestidge, B. P. “Early Experience with Cone-Beam CT for Brachytherapy.” Materials of 8th Annual International Conference and Workshop “New And Future Developments In Radiotherapy.” Sponsored by Wayne State University, School of Medicine, Detroit, MI. November 12–14, 2004, San Diego, CA, 2004. Raben, A., H. Chen, A. Grebler, J. Geltzeiler, M. Geltzeiler, I. Keselman, S. Litvin, S. Sim, A. Hanlon, and J. Yang. (2004). “Prostate seed implantation using 3D-computer assisted intraoperative planning vs. a standard look-up nomogram: Improved target conformality with reduction in urethral and rectal wall dose.” Int J Radiat Oncol Biol Phys 60:1631–1638.
33–Modern Advances in Prostate Brachytherapy
671
Rivard, M. J. (2001). “Monte Carlo calculations of AAPM Task Group Report No. 43 dosimetry parameters for the MED3631-A/M 125I source.” Med Phys 28:629–637. Rivard, M. J., D.-A. R. Evans, and I. Kay. (2005). “A technical evaluation of the Nucletron FIRST system: Conformance of a remote afterloading brachytherapy seed implantation system to manufacturer specifications and AAPM Task Group report recommendations.” J Appl Clin Med Phys 6:22–50. Schwartz, D. J., B. J. Davis, R. J. Vetter, T. M. Pizansky, M. G. Herman, T. M. Wilson, W. N. LaJoie, and A. L. Oberg. (2003). “Radiation exposure to operating room personnel during transperineal interstitial permanent prostate brachytherapy.” Brachytherapy 2:98–102. Shen, D., Z. Lao, J. Zeng, W. Zhang, I. Sesterhenn, L. Sun, J. W. Moul, E. H. Herskovits, G. Fichtinger, and C. Davatzikos. (2004). “Optimization of biopsy strategy by a statistical atlas of prostate cancer distribution.” Med Image Anal 8:139–150. Stone, N. N., S. Hong, Y.-C. Lo, V. Howard, and R. G. Stock. (2003). “Comparison of intraoperative dosimetric implant representation with postimplant dosimetry in patients receiving prostate brachytherapy.” Brachytherapy 2:17–25. Susil, R. C., A. Krieger, J. A. Derbyshire, A. Tanacs, L. L. Whitcomb, E. R. McVeigh, G. Fichtinger, and E. Atalar. (2003). “System for MR Image–guided prostate interventions: Canine study.” Radiology 228:886–894. Susil, R. C., C. Ménard, A. Krieger, J. A. Coleman, K. Camphausen, P. Choyke, K. Ullman, S. Smith, G. Fichtinger, L. L. Whitcomb, N. C. Coleman, and E. Atalar. (2005a). “Transrectal prostate biopsy and fiducial marker placement in a standard 1.5T MRI scanner.” Urology (in review). Susil, R. C., A. Krieger, J. A. Derbyshire, A. Tanacs, L. L. Whitcomb, E. R. McVeigh, G. Fichtinger, E. Atalar. (2005b). “A system for MRI-guided diagnostic and therapeutic prostate interventions.” Radiology (in press). Tapen, E. M., J. C. Blasko, P. D. Grimm, H. Ragde, R. Luse, S. Clifford, J. Sylvester, and T. W. Griffin. (1998). “Reduction of radioactive seed embolization to the lung following prostate brachytherapy.” Int J Radiat Oncol Biol Phys 42:1063–1067. Todor, D. A., M. Zaider, G. N. Cohen, M. F. Worman, and M. J. Zelefsky. (2003). “Intraoperative dynamic dosimetry for prostate implants.” Phys Med Biol 48:1153–1171. Tornes, A., and M. Eriksen. (2001). “Improved visualization of prostate brachytherapy sources.” J Brachytherapy Int 17:126. Unterweger, M. P., D. D. Hoppes, and F. J. Schima. (1992). “New and revised half-life measurements results.” Nucl Instrum Meth Phys Res A312:349–352. http://www.physics.nist.gov/PhysRefData/Halflife/halflife.html. Weaver, K. (1998). “Anisotropy functions for I-125 and Pd-103 sources.” Med Phys 25:2271–2278. Williamson, J. F. “Monte Carlo Simulation of Photon Transport Phenomena” in Monte Carlo Simulation in the Radiological Sciences. R. L. Morin (ed.). Boca Raton, FL: CRC, pp. 53–102, 1988. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, and R. Nath. (1998). “Dosimetric prerequisites for routine clinical use of new low energy photon interstitial brachytherapy sources.” Med Phys 25:2269–2270. Yu, Y., and M. C. Schell. (1996). “A genetic algorithm for optimization of prostate implants.” Med Phys 23:2085–2091. Zaider, M., C.-S. Chui, G. Cohen, D. A. Silvern, and P. Gajiwala. (2000a). “Real-time optimization for prostate LDR brachytherapy.” Int J Radiat Oncol Biol Phys 48:601–608. Zaider, M., M. J. Zelefsky, E. K. Lee, K. L. Zakian, H. I. Amols, J. Dyke, G. N. Cohen, Y. Hu, A. K. Erdi, C. Chui, and J. A. Koutcher. (2000b). “Treatment planning for prostate implants using magnetic-resonance spectroscopy imaging.” Int J Radiat Oncol Biol Phys 47:1085–1096. Zaider, M., M. J. Zelefsky, G. N. Cohen, C.-S. Chui, E. D. Yorke, L. Ben-Porat, and L. Happersett. (2005). “Methodology for biologically-based treatment planning for combined low-dose-rate (permanent implant) and high-dose-rate (fractionated) treatment of prostate cancer.” Int J Radiat Oncol Biol Phys 61:702–713. Zelefsky, M. J., Y. Yamada, G. Cohen, E. S. Venkatraman, A. Y. C. Fung, E. Furhang, D. Silvern, and M. Zaider. (2000). “Postimplantation dosimetric analysis of permanent transperineal prostate implantation: improved dose distributions with an intraoperative computer-optimized conformal planning technique.” Int J Radiat Oncol Biol Phys 48:601–608.
Chapter 34
Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases Sou-Tung Chiu-Tsao, Ph.D. Beth Israel Medical Center & St. Luke’s-Roosevelt Hospital Center New York, New York Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674 Eye anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674 Intraocular Malignancies and Benign Diseases in Eye . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674 Episcleral Eye Plaques for Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674 125 I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 674 103 Pd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677 106 Ru . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677 60 Co . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677 90 Sr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 677 Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 679 Involvement of Ophthalmologist, Radiation Oncologist, and Physicist . . . . . . . . . . . . . . . . . . . . . . 679 Guidelines of Collaborative Ocular Melanoma Study (COMS) Group. . . . . . . . . . . . . . . . . . . . . . . 679 Guidelines of Radiological Physics Center (RPC). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 680 Procedures Involving Physicists, Radiation Oncologists, and Ophthalmologists . . . . . . . . . . . 681 Tumor Localization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681 Fundoscopy and Fundus Photography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 681 Ultrasound Echography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 682 Computed Tomography (CT) Scan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683 Magnetic Resonance Imaging (MRI) Scan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684 Correlation of the Localization Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685 Tumor Drawing on Fundus Diagram Relative to Critical Normal Structures . . . . . . . . . . . . . . . . . . 685 Tumor Diagram in Eye . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 686 2-D Cross-Sectional Diagram of Tumor in Eye . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 688 3-D Perspective of Tumor in Eye . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689 Treatment Planning. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689 Transferring Tumor Information from the Fundus Diagram and Ultrasound A-Scan and B-Scans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689 Criteria on Dose and Dose Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689 Decision Making. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 690 Checking and Confirming Source Availability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 690 Dose Calculation Real vs. Apparent Dosimetry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691 Dose Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 691 Quality Assurance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 693 Initial Steps for All Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 693 Pre-Op Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694 Operating Room . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694 Post-Op Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 697 Plaque Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 700 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 700
673
674
Sou-Tung Chiu-Tsao, Ph.D.
Introduction Intraocular malignant tumors are the most common eye cancer. Radiation treatments have proven successful in controlling these tumors (Bosworth et al., 1988; COMS 2001a,b,c; Finger 2000; Fontanesi et al., 1993 Freire et al., 2002; Garretson, Robertson, and Earle 1987; Hungerford 2003; Jampol et al., 2002; Nag et al., 2003a; Packer 1987; Packer, Rotman, and Salanitro 1984; Packer et al.,1980, 1992; Rotman et al., 1983). The National Eye Institute has launched the Collaborative Ocular Melanoma Study (COMS) to evaluate the efficacy of using 125I eye plaques compared with enucleation (eye removal) for medium size tumors (Jampol et al., 2002; COMS 2001a,b,c; Earle, Kline, and Robertson 1987; Fine, Straatsma, and Hawkins 1987). There are also benign eye diseases that may respond to radiation treatment (Aizman et al., 2004; Finger et al., 2003). This chapter addresses the issues involving the eye plaque treatments.
Eye Anatomy The normal eye anatomy can be depicted by the diagram in Figure 1a. There are three layers enclosing the eye: the retina (inner), the choroid (middle), and the sclera (outer). The choroid layer is rich in vasculature. The ciliary body is the anterior extension of choroid. The uveal tract consists of the iris, ciliary body, and choroid. Both sclera and choroid are 1 mm thick on average. The retina, macula, lens, optic nerve, optic disc are all critical for vision. The macula is the posterior portion of the retina crucial for visual acuity. The foveola is the floor of the central pit (fovea) of the macula. The other landmarks in the eye are the equator, the ora serrata, and the limbus.
Intraocular Malignancies and Benign Diseases in Eye Malignant melanomas that grow in the uveal layer of choroid and ciliary body push the overlaying retina toward the central portion of the eyeball, thereby distorting the retinal layer. Retinoblastoma is another type of intraocular malignant tumor that grows in the retinal layer. Figure 1b is a schematic diagram showing a choroidal melanoma in an eye. The majority of choroidal melanomas are dome shaped and located in the posterior portion of the eye. The tumor base is the region adjacent to the underlying interior sclera. The basal diameter is the chord length of tumor base. Tumor apex is the apex of the dome. The apical height is the distance between the apex and the interior scleral surface. The portion of the retina over the tumor is pushed inward toward the vitreous body. Retinal detachment may occur around the tumor periphery. Macula degeneration is a benign ophthalmic condition characterized by progressive destruction and dysfunction of the central retina (macula). Age-related macular degeneration is the leading cause of blindness in the United States (Finger et al., 2003).
Episcleral Eye Plaques for Treatment 125
I
An eye plaque loaded with 125I seeds (Anderson and Chiu-Tsao 1993) is applied externally to the scleral surface over the tumor base for treatment of the intraocular malignancies, as shown also in Figure 2. Gold alloy is generally used as a substrate for 125I seeds. It also serves as radiation shield for other parts of head (Alberti et al., 1991; Anderson and Chiu-Tsao 1993; Harnett and Thomson 1988; Hilaris et al., 1988; Karolis, Frost, and Billson 1990; Kline and Yeakel 1987; Luxton et al., 1988; Packer et al., 1980; Schell et al., 1989; Weaver, Peksens, and Barnett 1987). The COMS standard plaque design (Kline and Yeakel 1987) is the most popular type, which is currently available in seven standard sizes, with diameters ranging from 10 mm to 22 mm in 2 mm increments. (The plaques of 10 and 22 mm diameter have become available in recent years.) The elemental composition of gold alloy used in the COMS plaque is 77% gold, 14%
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 675
Figure 1(a). Diagram of a normal eye.
Figure 1(b). Diagram of a choroidal melanoma in an eye with an eye plaque applied on sclera over the tumor base. The other anatomical structures and landmarks are also indicated.
silver, 8% copper, and 1% palladium by weight. The seeds are loaded into molded troughs of a silastic seed carrier insert that fits snugly in the concave aspect of the gold backing. Figure 2a shows all five sizes of COMS plaque. The insert, gold backing, and dummy plaque for a 16-mm COMS plaque are shown in Figure 2b. Figure 3 shows the top and side view of a 14-mm COMS plaque. The seed troughs are arranged in concentric rings about the plaque’s central axis. There are 5, 8, 13, 13, 21, 24, 21 seed slots in the plaques with diameters 10, 12, 14, 16, 18, 20, and 22 mm, respectively. The seed arrangements in these plaques, except the 14-mm plaque, are shown in Figures 4a and 4b. The silastic layer between the bottom of the seed trough and the eye surface is 1 mm thick; serving as a spacer material to avoid extreme hot spots in scleral layer. The concave aspect of the silastic insert has radius of curvature of 12.3 mm designed to conform to the eye surface curvature. The gold backing is in the form of a segment of spherical shell (0.5 mm thick and 15.05 mm in radius of curvature) terminated
676
Sou-Tung Chiu-Tsao, Ph.D.
(a)
(b) Figure 2. (a) COMS standard eye plaques in five different sizes, 20, 18, 16, 14, and 12 mm in diameter, from left to right. Row 1: gold backings, row 2: silastic seed carrier inserts with seeds loaded, row 3: silastic inserts without seeds, row 4: acrylic pieces to fit in silver rims of dummy plaques. (b) 125I seeds, a silastic seed carrier insert, gold backing, and dummy plaque for a 16-mm COMS plaque, from left to right.
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 677
Figure 3. Seed arrangement in the silastic seed carrier insert and the gold plaque design. (a) Top view and (b) side view of a 14-mm COMS standard plaque.
by a cylindrical segment, which is also called the “lip.” The lip provides collimation of radiation to protect some normal structures in the eye (Kline and Yeakel 1987). There are also notched plaques available for fitting over tumors close to optical nerve (see Figure 4c.) 103
Pd
The COMS plaque design has also been used to accommodate 103Pd seeds with or without silastic carrier for ocular melanoma treatments. Thermoluminescent dosimetry (TLD) for 103Pd seed in COMS plaque with silastic carrier has been reported (Chiu-Tsao et al., 1995). The attenuation of radiation dose by the silastic carrier was found to be 15%. Dense packing of 103Pd seeds in the COMS design plaque without the silastic insert has also been used (Finger et al., 2002, 2003.) 106
Ru
106
Ru is a beta-emitter with maximum energy of 3.5 MeV and a half-life of 368 days. 106Ru eye plaques are available from Bebig (Berlin, Germany) in 16 different types and sizes with activities ranging from 10 to 50 MBq. Figures 5a and 5b display some 106Ru eye plaques and a schematic drawing. This modality of treatment has gained popularity in recent years, more in Europe than in the United States (Chan et al., 2001; Taccini et al., 2004; Astrahan 2003; Kaiserman et al., 2002; Lommatzsch, Werschnik, and Schuster 2000). 60
Co
Historically, 60Co eye plaques were available for treatment before 125I, 103Pd, and 106Ru plaques became widely used. They were also manufactured in different diameters (Augsburger et al., 1986’ Beddoe 1975; Beitler et al., 1990; Casebow 1971). Due to the higher-energy photon radiation emitted by 60Co (1.25 MeV), there is no radiation shielding of normal ocular structures provided by the plaque (Earle, Kline, and Robertson 1987). The use of 60Co eye plaques has been greatly reduced in recent years. 90
Sr
90
Sr eye plaques, usually used for pterygium treatment, have also been reported for treatment of choroidal melanoma by Missotten et al. (Missotten 1998)
678
Sou-Tung Chiu-Tsao, Ph.D.
Figure 4(a). Seed trough diagrams for four sizes of COMS standard plaque, 12, 16, 18, and 20 mm in diameter.
Figure 4(b). Seed trough diagrams for two sizes of COMS standard plaque, 10 mm (left) and 22 mm (right) in diameter. (Photo courtesy Dr. Mark Rivard.)
Figure 4(c). Notched COMS plaque (right), silastic insert (center) and dummy (left). (Photo courtesy Dr. Mark Rivard.)
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 679
Figure 5(a). 106Ru eye plaques from Bebig. (Courtesy Bebig GmBH, Berlin, Germany.)
Figure 5(b). Schematic drawing of a 106Ru eye plaque, showing the radioactive part sandwiched between the window and backing layers. (Courtesy Bebig GmBH, Berlin, Germany.)
Protocol Involvement of Ophthalmologist, Radiation Oncologist, and Physicist After an ophthalmologist makes the diagnosis of intraocular melanoma or benign disease, he/she refers the patient to the radiation oncologist for consultation. When eye plaque treatment is chosen as the treatment modality, the physicist participates in and organizes the treatment planning, quality assurance of treatment delivery, and radiation safety procedures.
Guidelines of Collaborative Ocular Melanoma Study (COMS) Group Patients with medium size tumors between 2.5 and 10 mm in apical height, and 16 mm or less in basal diameter are randomized between enucleation and 125I eye plaque treatment. Tumor base:
all pigmented areas of tumor adjacent to the underlying sclera
Tumor apex:
the highest elevation point of the tumor
Sclera thickness:
1 mm
Interior sclera:
center of tumor base at 1 mm from the exterior surface of sclera
Apical height:
distance between interior sclera and tumor apex
Basal diameter:
diameter of the tumor base
680
Sou-Tung Chiu-Tsao, Ph.D.
Guidelines of Radiological Physics Center (RPC) I-125 seeds
model 6711 or 6702 from MediPhysics (Amersham, Arlington, IL) (model 6702 has been discontinued, and new models of 125I seeds have become available from other vendors)
Seed activity
0.5 to 5 mCi
COMS standard plaques
10, 12, 14, 16, 18, 20 and 22 mm in diameter (Figures 2, 3, and 4) gold plaque with silastic seed carrier insert
Selection of plaque size
longest tumor base diameter in mm plus 4 to 6 mm
Dummy plaques
12, 14, 16, 18 and 20 mm in diameter silver rim, acrylic in center
For a medium size tumor receiving 125I eye plaque treatment: Prescribed dose:
100 Gy delivered in 5 to 12 consecutive days (later revised to 85 Gy based on the TG-43 recommendation) (Nath et al., 1995; Hanson 1996; Kline and Earle 1996)
Dose rate:
between 0.5 and 1.25 Gy/hr.
Prescription point:
(a) 5 mm from tumor base center (interior sclera) for a tumor with height 5 mm or less, (b) tumor apex point for height over 5 mm.
Clinical interest points for point dose calculation and reporting: • Tumor apex • 5 mm depth from interior sclera (if different from apex) • Tumor base center (interior sclera) • Center of optic disc • Center of lens • Opposite retina (22 mm from interior sclera) • Foveola (center of macula) Critical normal structures: • Sclera • Macula • Optic nerve • Optic disc • Retina • Lens The plaque size is chosen so that the tumor base, with a tumor-free margin of 2 to 3 mm, is covered entirely by the plaque. If the tumor lies within 2 mm of the optic nerve and does not subtend more than 90°-angle around the optic nerve, the posterior edge of the plaque should lie between the optic nerve and the posterior edge of the tumor. A notched plaque of the same design, including a lip around the notch, may be used. The COMS design plaques (circular and notched) are available from Trachsel Dental Studio, Inc., (507-288-2362, Rochester, MN).
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 681 Quality assurance review: Pre-treatment: • • • •
Ultrasound A-scan and B-scan polaroid photos Fundus photos Fundus diagram Tumor dimension and location
Post-treatment: • Plaque dosimetry data including dose estimates for critical interest points • Plaque diagram (or photograph) showing placemen of seeds and their activities and distribution • Fundus diagram showing tumor and critical structures • Total time of irradiation • Isodose curve plot in one central plane perpendicular to the plaque through the tumor apex Dose calculation: • Point source approximation • Specific gamma ray constant G = 1.45 R⋅cm2⋅h-1⋅mCi-1 • f factor = 0.91 cGy/R • Specific dose rate constant L = 1.45 × 0.91 = 1.32 cGy⋅h-1⋅mCi-1 • Anisotropy corrections: ignored • Influence of gold and silastic: neglected for the time being due to the lack of consensus among published data • Tissue attenuation and scatter: use radial dose function, g(r) based on the data of Ling et al (1983) for model 6711 seeds and Schell et al. (1987) for 6702 seeds. The values of g(r) at 0.5 cm are also used for r < 0.5 cm for each seed model. After the TG-43 report (Nath et al., 1995) was published, revised recommendation from the RPC was presented (Hanson 1996; Kline and Earle 1996). The further revised data recommended by TG43U1 (Rivard et al., 2004) should be evaluated for eye plaque dosimetry.
Procedures Involving Physicists, Radiation Oncologists, and Ophthalmologists Tumor Localization Fundoscopy and Fundus Photography Direct and indirect ophthalmoscopy are performed with patient’s pupil dilated. The inner part of the eye is visible with the aid of an ophthalmoscope. The normal eye appears orange in color. Photographs taken with an ophthalmoscope and fundus camera are called “fundus photographs.” Wide angle and narrow angle can be achieved to view a wide range and narrow detail of certain structures of interest, respectively (Yannuzzi et al., 2004). Figures 6a and 6b show fundus photos of a normal eye in narrow and wide angle views, respectively. When ocular melanoma is present, a greyish shadow appears in the fundus photo (see Figure 6c). One way to measure the tumor basal diameter is to compare it with the visible optic disc diameter. The optic disc and the foveola are used as the reference landmarks. The diameter of optic disc (referred to as the “disc diameter” (DD) is about 1.5 mm, even though patient to patient variation is possible. The
682
Sou-Tung Chiu-Tsao, Ph.D.
Figure 6. Fundus photo of an eye with ocular melanoma.
distance between any two interest points is estimated in multiples of DD. When the dimension is within 10 mm and the measured object is in the posterior portion of the eye globe, this way of estimation is usually reliable. When the tumor extends anterior to the equator, it is challenging to determine the dimensions based on ophthalmoscopy alone. Ora serrata can be used as a reference landmark in this case. The shape, location, and dimensions of the tumor base are then drawn in the fundus diagram which will be described in a later section (Tumor Drawing on Fundus Diagram Relative to Critical Normal Structures). Ultrasound Echography Ultrasound echography has proven to be a useful tool in determining some dimensions and the location of a tumor in eye relative to the critical normal structures (Char et al., 1980; Peyster et al., 1985; Pavlin et al., 1987, 1989). In order to achieve the high spatial resolution required for these determinations, specialized devices operating at up to 200 MHz have become available to achieve 10 micron resolution (Bridal et al., 2003; Daftari et al., 2001; Coleman et al., 2004a,b; Foster et al., 1998; Hewick et al., 2004; Pavlin and Foster 1995). One of the ultrasound scanners recommended by the COMS group is “Ophthascan S” by Biophysic Medical, Inc. (Pleasant Hill, CA). A-Scan. An A-scan is a one-dimensional (1-D) scan. Each spike of the curve in an A-scan represents an interface of two media with different sound speeds (see Figures 7a and 7b). The main purpose is to determine the distance between various interfaces, and hence the thickness of individual structures. For tumor localization in the eye, an A-scan is first performed to identify the structure boundaries and to localize the tumor apex. Once the apex is localized, exact perpendicularity with the tumor surface is achieved to display steeply rising spikes at the apex and at the interior scleral surface. In this way, the apical height (tumor thickness) can be determined accurately. Figure 8a shows an A-scan image for a melanoma. B-Scan. To obtain a complete topographic overview, a B-scan is also performed. It is achieved by moving the transducer in a linear fashion across the conjunctiva to form a two-dimensional image (2-D) of a cross section of the eye (see Figure 7c). Figure 8b shows a B-scan image of an eye with a tumor very close to the optic nerve. The B-scan image through the tumor is indicative of its geometry relative to other landmarks. However, it is limited in yielding accurate quantitative measurements of tumor dimensions,
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 683
(a)
(b) Figure 7. Schematic diagrams of an (a) ultrasound A-scan and (b) for ultrasound B-scan of an eye with ocular melanoma.
because it is not easy to assure the plane of B-scan and the retinal detachment may obscure the periphery of the tumor base. In addition, both A-scans and B-scans are difficult to perform for anterior tumors. COMS guidelines recommend B-scans in three different planes: transverse, longitudinal, and axial. The transverse scan is done across the meridians through the tumor center. The longitudinal scan is within the same meridian where the tumor is located. The axial scan is performed to include the optic nerve and the tumor in the same scam image (COMS 1999). Computed Tomography (CT) Scan CT scans with and without contrast have been shown to be successful in the diagnosis of ocular melanomas and the measurements of the tumor thicknesses and basal diameters. CT scans allow highly reproducible measurements of basal diameters by displaying the extremities of the tumors and using the measure
684
Sou-Tung Chiu-Tsao, Ph.D.
(a)
(b) Figure 8. (a) An ultrasound A-scan image for a choroidal melanoma. (b) An ultrasound B-scan image of the same melanoma. (Reprinted from “Brachytherapy of Ocular Melanoma” by B. S. Hilaris, D. Nori, L. Anderson, and S. Chiu-Tsao in An Atlas of Brachytherapy. B. S. Hilaris, D. Nori, and L. Anderson (eds.), pp. 304–310. © 1988, with permission from Macmillan Publishing Company).
distance function of the scan analysis algorithm (see Figure 9). The accuracy of the tumor thickness determination using CT is comparable to ultrasound A-scan (Augsburger et al., 1987; Char et al., 1980; Peyster et al., 1985, 1988). CT scans have also been used to localize melanoma and identify possible suspected extraocular extension that may be difficult to see in an ultrasound echogram (Hesselink 2005). The advantage of the multiple thin-slice CT scan images is in its capability of allowing reconstruction of a three-dimensional (3-D) object, which is very helpful in visualizing the geometry of the tumor relative to the critical normal structures (Astrahan et al., 1990b). Magnetic Resonance Imaging (MRI) Scan MRI scans of ocular melanomas in eye have been successfully performed with the aid of special surface coils (Bilaniuk et al., 1985; Gomori et al., 1986; Hesselink 2005; Hosten et al., 1997; Houdek et al., 1989; Mafee 1996; Stroszczynski et al., 1998; Peyman and Mafee 1987; Peyster et al., 1988; Sullivan and Harms
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 685
Figure 9. CT scan images of an eye with choroidal melanoma in transverse cross section. (Reprinted from Augsburger, J. J., R. G. Peyster, A. M. Markoe, E. G. Guillet, J. A. Shields, and M. E. Haskin. “Computed tomography of posterior uveal melanoma.” Arch Ophthalmol 105:1512–1516. © 1987, with permission from American Medical Association.)
1986; Zimmerman and Bilaniuk 1988). Both T1-weighted and T2-weighted images were obtained based on the fast relaxation of melanin present in the tumor. The melanoma appears hyperintense and hypointense in the T1- and T2-weighted images, respectively, relative to the vitreous body. The advantage of MRI over the CT scan is in its ability to provide precise spatial definition of the target and to offer better diagnostic specificity. In addition, the direct sagittal MR images give graphic information not available with CT (Houdek et al., 1989) reported MRI scans using a series of 3 mm slices in transverse, sagittal, and coronal planes to determine precise location and dimensions of ocular tumors (see Figure 10). Similar to CT scanning, reconstruction of 3-D structures is achievable with thin slices of MRI images. Due to the higher cost (compared with ultrasound), both CT and MRI scans have not been routinely done for all plaque patients. The COMS protocol does not require CT and/or MRI scans for plaque patients. Correlation of the Localization Modalities Studies have been performed to evaluate the correlations between fundoscopy, ultrasound, CT, MRI, and gross pathologic analysis, in the measurements of tumor basal diameters and thicknesses (Peyster et al., 1985, 1988). For the measurements of tumor thickness, there is good correlation between the ultrasound A-scan, CT, MRI, and pathology. For basal diameter measurements, the fundoscopy, CT, MRI, and pathology correlate well. Tumors can be distinguished from the retinal detachment on CT and MRI scans, but not easily on ultrasound B-scans.
Tumor Drawing on Fundus Diagram Relative to Critical Normal Structures The fundus diagram, also referred to as the “retinal diagram,” is a distorted way of representing the entire inner layer of an eye from anterior ciliary body to the posterior pole, which is located at the center of the diagram. As shown in Figure 1, the optic axis (the line connecting the centers of lens and eye globe) extends
686
Sou-Tung Chiu-Tsao, Ph.D.
(a)
(b)
(c)
(d)
Figure 10. MRI scan images of an eye with choroidal melanoma; (a) and (b) pre-operative sagittal and coronal images; (c) and (d) corresponding post-implant images showing the plaque over the tumor base. (Reprinted from Int J Radiat Oncol Biol Phys, vol 17, “MR technique for localization and verification procedures in episcleral brachytherapy,” P. V. Houdek, J. G. Schwade, A. J. Medina, C. A. Poole, K. R. Olsen, D. H. Nicholson, S. Byrne, R. Quencer, R. S. Hinks, and V. Pisciotta, pp. 1111–1114. © 1989, with permission from Elsevier.)
anteriorly and posteriorly. The intersection of the optic axis with the posterior retinal surface is the posterior pole. The center of the macula, foveola, is very close to the posterior pole. The optic disc, about 1.5 mm in diameter, is centered medially at about 4 mm from the posterior pole. A fundus diagram looks like an analog clock face, with 1 to 12 o’clock positions indicating the meridians of an eye globe. Figure 11a shows the fundus diagrams of both right and left eyes as viewed anteriorly by a health care worker. The 12 and 6 o’clock positions are superior and inferior to the eye center, respectively. Each radial line in the fundus diagram represents an arc on the great circle of an eye in a meridian extending from the posterior pole anteriorly. The length of each radial line segment corresponds to the arc length between two interest points on that clock hour. (Figure 11b shows a fundus diagram of a left eye, with chord and arc lengths indicated.) The equator, ora serrata, and limbus are represented by the concentric circles of increasing radii in the fundus diagram. Since the latitudinal circles decrease in diameter as one moves anteriorly from the
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 687
(a)
(b) Figure 11. (a) Fundus diagrams (retinal charts) of a right and a left eye. (b) Fundus diagram showing the chord and arc lengths (in parentheses) for a left standard eye with 22-mm inner diameter.
688
Sou-Tung Chiu-Tsao, Ph.D.
equator, the magnification factors for the true latitudinal distances between two points on different meridians increase rapidly with the increasing radial distance in the fundus diagram. When the tumor is relatively small (10 mm in diameter or less) and posterior to the equator, the basal diameter usually corresponds to the straight-line length on the fundus diagram. For large tumors, the distortion of this diagram makes it less straightforward to relate the tumor base dimension and the drawing of tumor base. There is a lot of distortion in the fundus photo of a large tumor that extends from posterior to anterior portion of the eye. In addition, it is also a challenge to draw an anterior tumor accurately on the fundus diagram due to the distortion. These are the cases for which we would greatly appreciate a 3dimensional representation of the tumor in eye (Astrahan et al., 1990b; Evans, Astrahan, and Bate 1993; Kepka, Johnson, and Kline 1988). Example fundus diagrams for a right and a left eye with tumor are shown in Figures 12a and 12b, respectively. The locations, dimensions, and shapes of the example tumor bases are depicted by the crosshatched regions. In Figure 12a, the tumor is on the temporal side of the right eye and in the 9 o’clock meridian. The tumor is posterior to the equator, 10 mm × 7 mm in basal dimensions. The posterior margin of the tumor is 5 mm from the foveola, the posterior pole. It is 9 mm from the optic disc edge. In Figure 12b, the tumor is on the superior side of the left eye, and centered in the 12 o’clock meridian. It is posterior to the limbus and mostly anterior to the equator. Its largest (longitudinal) basal diameter is 12 mm. The transverse basal diameter in the perpendicular direction to the largest diameter is 10 mm, which appears to be of longer length than the largest diameter on this fundus diagram, due to the builtin distortion. The posterior margin of the tumor is 13 mm from the foveola and 14 mm from the optic disc edge.
Tumor Diagram in Eye 2-D Cross-Sectional Diagram of Tumor in Eye The tumors drawn in the fundus diagrams, Figures 12a and 12b, are shown in Figures 13a and 13b as domes in the 2-dimensional cross-sectional diagrams through their respective meridians in which the tumor centers were located. The posterior pole and lens are always present in this type of diagrams. However, depending on the meridian in which the tumor is located, optic disc may or may not be on that same meridian. In order to display the distance between the tumor’s posterior margin and the optic disc edge, it has
Figure 12. Tumor drawing on a fundus diagram, showing the tumor base dimensions; (a) right eye, (b) left eye.
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 689
Figure 13. Two-dimensional cross sectional diagrams in the meridian that the tumor is located, (a) and (b) corresponding to Figures 12a and 12b.
been a practice to draw the optic disc and optic nerve in this same diagram, as if the optic disc were in that same meridian. The tumor protrusion into the vitreous body and the longitudinal basal diameter can be depicted in such 2-D diagrams. The tumor example shown in Figures 13a and 13b are 4.5 mm and 5.5 mm in apical heights, and 10 mm and 12 mm in longitudinal basal diameters, respectively. 3-D Perspective of Tumor in Eye A 3-D perspective of tumor in eye is very important in visualizing its geometrical relationship relative to the various normal structures. The tumors in Figures 13a and 13b are shown by 3-D cone-shaped domes in Figures 14a and 14b, using the program by Astrahan et al. (1990a,b). While the computer display of the 3-D perspective is very helpful, it is my experience that the plastic model of an eye (e.g., from Edmund Scientific) with some modeling clay (e.g., Silly Putty™ or equivalent) simulating a melanoma would be more enlightening. Since both fundus photography and ultrasound echography have some limitations in determining certain tumor dimensions accurately, CT and MRI scans serve as complementary tools and provide 3-D reconstruction and visualization. So far, both CT and MRI are still more costly imaging modalities, hence less utilized in routine practice.
Treatment Planning Transferring Tumor Information from the Fundus Diagram and Ultrasound A-Scan and B-Scans From the fundus diagram and the ultrasound A-scans and B-scans for a patient case, as provided by the ophthalmologist, the physicist transfers the information onto the 2-D and 3-D diagrams, manually or using the Astrahan program (Astrahan et al., 1990a). Criteria on Dose and Dose Rate Assure that the prescription (85 Gy) isodose surface passes through the prescription point, encompasses the tumor and extend to or beyond the edge of the gold plaque. The dose may be delivered over a duration
690
Sou-Tung Chiu-Tsao, Ph.D.
(a)
(b)
Figure 14. Three-dimensional perspectives of the exterior of the eye with the eye plaque applied over the tumor base, corresponding to the fundus diagrams above.
of 5 to 12 days. The dose rate should be between 0.5 and 1.25 cGy/h to the prescription point, which is defined as the tumor apex, if apical height is 5 mm or greater, or 5 mm depth from interior sclera, if apex is less than 5 mm high. With the seed distributions in the COMS standard design plaques, the dose to tumor base center is generally within 1000 Gy, i.e., 10 times of the prescription dose for medium size tumors. It would be helpful that each institution decide on its own protocol of initial dose rate at the prescription point based on a workable time schedule feasible for operating room, ophthalmologist, radiation oncologist, and physicist. For example, an initial dose rate of 62 cGy/h would be suitable for a 7-day implant if operating room is available the same day of each week. Decision Making Plaque Selection. Based on the tumor base dimensions determined, the eye plaque diameter is selected. Usually the tumor base is irregular in shape. The plaque diameter must be equal or slightly larger than the longest basal dimension plus 4 to 6 mm. For example, treatment of a tumor of 9 mm in its longest dimension would require a 14-mm plaque. If the posterior margin of the tumor is within 2 mm from the optic disc edge, a notched plaque may be indicated. The dimension of the notch is determined.(Kline, personal communication). A customized notched plaque may be ordered from Trachsel Dental Studio. Seed Model. Medium activity model 6711 125I seeds are generally available in the range between 1 and 5 mCi from MediPhysics (Amersham). These seeds can be ordered as needed and are used in most of eye plaque treatments. Some centers also used model 6702 seeds (previously available in 10 to 40 mCi from MediPhysics, but discontinued now), which are decayed after usage in brain implants. Since the introduction of many new models of 125I and 103Pd seeds in recent years, these new models are also feasible for eye plaque treatment, provided that the activities fit in the range between 1 and 5 mCi. A. Joint AAPM/RPC registry of low-energy brachytherapy seeds has been established. Interested users can check the RPC website for information. Activity Distribution (Uniform Loading; Differential Loading; Dense Packing). Uniform loading of seed activities in all slots in the silastic insert is the most straightforward way, provided that all the 125I
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 691 seeds are available in one activity group (within about 10% of one another). Some radiation oncologists prefer to achieve a smaller hot spot near tumor base center by leaving one or a few central slots empty. This is the simplest scheme of differential loading. When the seeds are only available in two or more activity groups (due to supply limitation), differential loading may be considered, distributing higher activity seeds in the peripheral ring/s and lower activity seeds in the inner ring/s. Dense packing of 103Pd seeds in the plaque has been used partially due to the lack of high activity 103Pd seeds. A treatment-planning calculation is performed to decide the seed activity group/s needed for planned initial dose rate at the prescription point on the day of implant. To comply with the COMS/RPC requirements on total dose delivered and to accommodate your operating room schedule, appropriate adjustment in source activity distribution can be made. Checking and Confirming Seed Availability The user calls the seed vendor to confirm that the desired activity group/s of the seeds are available to get ready for a certain operating room schedule. Dose Calculation, Real vs. Apparent Dosimetry The point source approximation and dose calculation parameters (Ling et al., 1983; Schell et al., 1987) previously recommended by the RPC are required in the reporting of doses to COMS database. This calculation algorithm is referred to as “apparent dosimetry,” which is recommended in order to maintain consistency among all COMS centers for future comparison purposes. The “real dosimetry” should incorporate the influence of source anisotropy, gold and silastic, gold plaque lip collimation, anterior air interface, interseed interference, and bony orbit, etc. There are numerous reports on these studies, published, presented, and ongoing (Anderson and Chiu-Tsao 1993; Chiu-Tsao et al., 1986, 1990, 1993: Cygler et al., 1990; de la Zerda et al., 1992, 1996; Harnett and Thomson 1988; Jin et al., 1993; Kanna et al., 1993; Ling et al., 1983, 1985, 1989; Luxton et al., 1988; Luxton, Astrahan, and Petrovich 1988; Meli and Motakabbir 1993; Alberti et al., 1991; Nath, Meigooni, and Meli 1990; Schell et al., 1987; Weaver 1986; Weaver et al., 1989; Williamson 1988a, 1988b, 1991;Wu, Sternick, and Muise 1988; Wu and Krasin 1990). These modifications are obviously a lot more complicated than the simplified algorithm currently recommended and required by the RPC. A comparison between the apparent dosimetry and the real dosimetry (de la Zerda et al., 1992, 1996) is shown in Figure 15a. The apparent dosimetry overestimates the dose around a COMS plaque by about 30%. Recent results by de la Zerda et al. (1992, 1996) using TLD dosimetry for a centrally loaded 125I 6711 seed in a 20 mm COMS plaque are shown in Figure 15b. For comparison, the isodose curves in the corresponding plane for an 125I seed in the homogeneous phantom (Chiu-Tsao et al., 1990) are also plotted. The dose reduction and the lip collimation effects due to the gold backing and silastic insert of the 20-mm COMS eye plaque are clearly demonstrated by the two sets of solid and broken curves in Figure 15b. Caution must also be taken by users that the real dosimetry in a homogeneous phantom for the two models (6711 and 6702) of 125I seeds are different by as much as 10% (Anderson et al., 1990), even though the COMS/RPC guidelines require that the same specific dose rate constant, 1.32 cGy⋅h-1⋅mCi-1 at 1 cm, be used in the dose calculations. After some consensus is reached in the near future on the real dosimetry algorithm that will be implemented on 3-D treatment planning computer calculations, more accurate dose estimates will be available, allowing more accurate correlation of tumor response and normal tissue complications with absorbed dose. The treatment planning computer system named “Plaque Simulator” from Bebig1 (Berlin, Germany) (Astrahan 2003, 2005; Astrahan et al., 1990a,b, 1997) has features that would allow the inclusion of anisotropy function, the shielding effect by the lip and/or gold backing, silastic filtration, etc., chosen by 1
Tel. (+49 30) 98 10 84-0, Fax (+49 30) 94 10 84-150, URL: www.bebig.de)
692
Sou-Tung Chiu-Tsao, Ph.D.
(a)
(b) Figure 15. (a) Graph showing Dr2/S vs r, comparing the “apparent” and “real” dosimetry for an 125I model 6711 seed in homogeneous medium; also the “real” dosimetry along the plaque’s central axis with the influence of gold and silastic taken into account. (b) Isodose curves in the transverse central plane of a 20 mm COMS eye plaque (solid curves) for a centrally loaded 125I 6711 seed, and in a homogeneous phantom (broken curves) based on TLD measurement results. (Reprinted from de la Zerda, A., S. Chiu-Tsao, H. S. Tsao, J. Lin, L. Boulay, I. Kanna, and J. H. Kim, “125I eye plaque dose distribution including penumbra characteristics.” Med Phys 23:407–418. © 1996, with permission from AAPM.)
the user. The accuracy of Plaque Simulator calculations for 125I seeds has been verified (Knutsen et al., 2001, Krintz et al. 2002). Figure 16a shows an isodose curve plot in a cross section for an eye plaque treatment of the tumor shown in Figure 12a, using the COMS-endorsed apparent dosimetry. Figure 16b shows the isodose curves for a 14-mm COMS plaque with the lip collimation effect calculated using “partial visibility algorithm” (Astrahan et al., 1990a). The heterogeneity effects observed for model 6711 seed may not be the same for other models of 125I seeds, which have different interior components. Such effect should be studied and included in the treatment-planning calculation for individual seed model.
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 693
(a)
(b)
Figure 16. (a) Isodose curves in a cross section for an eye plaque treatment of the tumor using the “apparent” dosimetry. (b) Isodose curves for an 20-mm COMS plaque with the lip collimation effect calculated using “partial visibility algorithm.” (Reprinted from Astrahan, M. A., G. Luxton, G. Jozsef, P. E. Liggett, and A. Petrovich, “Optimization of 125I ophthalmic plaque brachytherapy.” Med Phys 17:1053–1057. © 1990, with permission from AAPM.)
Quality Assurance Initial Steps for All Cases At the initial stage of setting up an eye plaque treatment protocol in your institution, there are many issues that should be addressed. Proper communication channels among physicists, radiation oncologists, ophthalmologists, and operating room personnel, floor nurses must be formed to assure smooth operations. The quality control steps recommended for the physicists to take are described below. Quality Control of Plaques and Silastic Inserts. Upon receiving the plaques and/or silastic inserts, you want to verify (1) the seed positions in each seed carrier insert and (2) the radii of curvature of the gold backings and silastic inserts. This is particularly important for custom plaques (nonstandard design). 1. Take AP radiographs of the seed carrier inserts loaded with the dummy seeds containing nonradioactive core. The patterns shown in Figures 3 and 4 are expected. Verify the seed coordinates and ring diameters. 2. Take lateral radiographs of the gold backings and silastic inserts to confirm their radii of curvature. Quality Control of Treatment-Planning Computer. Verify that the single-seed calculational results using your treatment-planning computer agree with those of “apparent dosimetry” recommended by the COMS/RPC guidelines. Until the consensus is reached in the future regarding the incorporation of updated “real” dose distribution data, the COMS-endorsed “apparent dosimetry” data for point-source approximation may be used. The TG43U1 recommended dosimetric data are available for some models of 125I and 103Pd seeds (Rivard et al., 2004). Quality Control of Well-type Ionization Chamber. (1) Verify that the well-type ionization chamber is calibrated with an 125I seed traceable to the National Institute of Standards and Technology (NIST) or an Accredited Dose Calibration Lab (ADCL).
694
Sou-Tung Chiu-Tsao, Ph.D.
(2) Check the ionization chamber factor constancy using a low-activity long-lived check source (e.g., Am or 137Cs) prior to each use.
241
Pre-Op Preparation Seed Ordering and Receiving. After the seed availability is confirmed, the seeds ordered can generally be delivered to your center on the next business day (in the continental United States). Upon receiving the seeds, the physicist in charge needs at least one, preferably two, business days to prepare the plaque and send it for overnight gas sterilization. Seed Activity Verification. It is important for the user to independently verify the individual seed activity and compare with those quoted by the vendor. A well-type ionization chamber is suitable for this purpose. Seed Loading and Plaque Preparation. The silastic insert, with its convex side up, is mounted on top of a jig with hemispherical protrusion (about 25 mm in diameter) (see Figure 17a). (A steel ball bearing 1 inch in diameter can also serve the purpose.) An 125I seed is then picked up by a tweezer or vacuum pickup device and inserted into a seed trough in the seed carrier, one at a time. After all seeds are loaded, cover the seed carrier insert by the gold plaque which has had its inner rim coated with a thin layer of bonding agent (Dow Corning SILASTIC Medical Adhesive Silicone Type A, cat. No. 891). Make sure that the concave inner surface of the gold plaque is free of glue. Let the plaque assembly sit for at least a few hours to allow the glue to cure (see Figure 17b). Gas Sterilization of Plaque and Container. The prepared plaque assembly is transferred to a small lead container (see Figure 17c). This container is punctured with a few holes to allow sterilizing gas to pass through. Each end of the container with punctured holes is lined inside with a thin layer of lead (not gas-tight) to provide radiation shielding. A dummy plaque of the same diameter is also put into the same container (see Figure 17d). The container is wrapped with a radiation caution label, then put into a gas sterilization bag, and sent for gas sterilization (see Figure 17e showing photo of container in bag). This procedure generally requires 24 hours turnaround time. Note that cidex sterilization is not recommended, because of the possibility of patient reaction to trapped cidex between gold and silastic. Transport of Plaque Container. After sterilization, the bag containing loaded plaque is transported by a physicist to the operating room shortly before the procedure. The operating room nurse takes charge of this bag from that point on. Operating Room Transillumination. This procedure is schematically shown in Figure 18a. In the clinical examination, a fiberoptic light source is placed under the tumor. The examiner views through a dilated pupil or through the opposite sclera. Malignant melanoma presents a greyish area that dose not transilluminate. In the operating room, the involved eye is first rotated such that the surface of the eye corresponding to the tumor is exposed with a retractor. A fiberoptic light source is then located as close as possible to 180° opposite the tumor. A greyish shadow corresponding to the tumor base is visible to the eye surgeon (see Figure 18b.) Marking of Tumor Base. The boundary of the tumor base is then marked on the sclera with a surgical marking pencil (see Figure 18b). Using a caliper, the dimensions of the tumor base are measured. It is also helpful, if possible, to measure the distance between the anterior tumor margin and limbus (or ora serrata) in that meridian with caliper (see Figure 13a). Dummy Plaque Suturing. The dummy plaque (same size as the therapeutic radioactive plaque) is placed, with suture threads in place, on the sclera covering the scleral marks that identify the tumor base. The dummy plaque must completely cover the tumor base as well as a tumor-free perimeter of 2 mm or more (see Figure 18c). [Exception: If the tumor is within one disc diameter (1.5 mm) from the optic nerve,
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 695
(a)
(b)
(c)
(d)
(e) Figure 17. (a) A jig designed for seed loading into a silastic insert. (b) Gold backing covering the silastic insert from above. (c) Lead container with punctured holes and interior lining in its ends. (d) Both the active and dummy plaques in the container. (e) Gas sterilization bag with lead container inside.
696
Sou-Tung Chiu-Tsao, Ph.D.
(a)
(b)
(c)
(d)
Figure 18. (a) Schematic diagram for transillumination. [(a) reprinted from Atlas of Clinical Ophthalmology, 1st Edition, D. J. Spalton, R. A. Hitchings, and P. A. Hunter (eds.). London: Gower Medical Publishing Ltd., ©1984, with permission from J. P. Lippincott.] (b) Photo of an eye with scleral marking of tumor base. (c) A dummy plaque sutured in place. (d) An active therapeutic plaque sutured in place.
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 697 the posterior edge of the plaque may be placed between the optic nerve and the posterior margin of the tumor. A tumor-free perimeter less than 1.5 mm in that meridian is covered by the plaque.] The boundary of the dummy plaque is then marked on the sclera with a surgical marking pencil. The suture lugs used are identified. Active Plaque Suturing. The dummy plaque is removed with the suture threads left in place. The opaque therapeutic plaque is placed within the outer ring of the scleral marks for the removed dummy plaque. The suture threads originally left in place are used to go through the same suture lugs as identified for the dummy plaque (see Figure 18d). Transillumination with Sutured Plaque. After the active plaque is secured in place, confirmation of its position is determined with the aid of focal illumination from a fiberoptic light source. If the active plaque placement is found to be unsatisfactory, the sutures can be adjusted to provide optimal positioning (Robertson, Fuller, and Anderson 1987). Intraoperative Ultrasound Scans. Intraoperative ultrasound B-scans have been used to examine the active plaque placement relative to the tumor (Pavlin et al., 1987, 1989). Since this requires additional time spent in the operating room, it is not commonly done in most centers. Caliper Measurements of Some More Dimensions. Caliper measurements of the following dimensions are recommended after the therapeutic plaque placement: 1. Distance between the limbus and the anterior boundary of the plaque in that meridian, 2. Cornea diameter (distance between opposite limbus) in that meridian. Refer to the cross-sectional diagram in Figure 13a with these dimensions indicated. Handling of Accidentally Contaminated Plaque. In the case of accidental contamination of the plaque in the operating room, autoclaving can be done. The silastic, gold backing, and seeds would survive occasional autoclaving. The silver dummy plaque with acrylic insert, however, would not survive; acrylic melts in the high temperature used for autoclaving. You can first push out the acrylic insert and then send the silver rim for autoclaving. Post-Op Procedures Radiation Safety Procedure and Labels. Radiation safety labels are attached to the patient’s chart. A radiation survey is performed in patient”s room and recorded. An eye patch with thin lead foil could be used, although is not essential. Since the radiation exposure distributions depend on the plaque location in the eye, the size and location of the eye patch should be adjusted accordingly to achieve the most effective protection. There are published reports of radiation survey results for some patient cases (Myers and Abramson 1988; Pavlin et al., 1989). The nursing aspect has been discussed by Myers (Myers and Abramson 1988). Some hospitals are allowed to send the patient home for a few days before plaque removal. This generally requires a license amendment from the Nuclear Regulatory Commission (NRC) or its agreement state counterpart approved for that hospital by the regulatory agency. Verification with Ultrasound and MRI. Both ultrasound echography and MRI scan have been used to verify the plaque placement. Pavlin et al. (1987, 1989) reported their determination of the relationship of eye plaque to the tumor base using ultrasound. They monitored the accuracy of plaque placement postoperatively. The relationship of plaque margin to the optic nerve can also be determined using this method. Figures 19a and 19c show, respectively, B-scans of a well-centered plaque and a plaque covering the posterior edge of a tumor very close to optic nerve. A malpositioned plaque is shown in Figure 19b. An MRI scan can image sutured eye plaque on the tumor after implantation. Figure 20 shows a series of sagittal slices, 3 mm in thickness, obtained by Houdek et al. (1989), for an eye plaque case. The artifact caused by gold in the CT scan is not present in the MRI scan. Post-implant verification of the plaque placement using MRI scan would allow us to obtain actual geometrical relationships between tumor,
698
Sou-Tung Chiu-Tsao, Ph.D.
(a)
(b)
(c) Figure 19. Ultrasound B-scan images showing tumors and overlying plaques. (a) A plaque well centered over the tumor. (b) A malpositioned plaque. (c) A posterior tumor and plaque edge very close to optic nerve. [(c) reprinted from Pavlin, C. J., B. Japp, E. R. Simpson, H. D. McGowan, and P. J. Fitzpatrick, “Ultrasound determination of the relationship of radioactive plaques to the base of choroidal melanomas,” Ophthalmol 96:538–542, © 1989, with permission from American Academy of Ophthalmology.)
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 699
Figure 20. A series of sagittal MRI images of the implanted plaque applied over the tumor base. (Reprinted from Int J Radiat Oncol Biol Phys, vol 17, “MR technique for localization and verification procedures in episcleral brachytherapy,” P. V. Houdek, J. G. Schwade, A. J. Medina, C. A. Poole, K. R. Olsen, D. H. Nicholson, S. Byrne, R. Quencer, R. S. Hinks, and V. Pisciotta, pp. 1111–1114. © 1989, with permission from Elsevier.)
plaque, and normal structures, and hence arrive at accurate dose estimates. Since the dimensions of all the structures of interest are of the order of millimeters, it would be advisable to use very thin slices of 3 mm or less. Final Dose Calculation. After all the appropriate parameters are identified and measured in the operating room and in the post-op procedures, the physicist will incorporate such information to calculate the dose distributions in and around the eye and prepare the dosimetry report. Plaque Removal Another operating room procedure is required for the plaque removal. This is much faster than the plaque placement. The removed plaque is placed into the original container together with the dummy plaque used for the same patient. Retrieval of Plaque from Operating Room. Physicist is notified of the removal and makes arrangement to retrieve the plaques. Seed Unloading and Identification. In the radioactive source handling area, the silastic insert and gold plaque is separated with a knife along the inner rim of the gold plaque (see Figure 21). Individual
700
Sou-Tung Chiu-Tsao, Ph.D.
Figure 21. Removal of silastic insert from the gold backing.
seeds are removed from the silastic seed carrier insert and identified. These seeds could be reused for another patient case, plaque or interstitial, if timing is right.
Conclusion The issues involved in the eye plaque treatments of intraocular tumors have been presented in this chapter. This manuscript is meant to be a starting point for any physicist who is planning an eye plaque treatment in his/her hospital. Hands-on training by a COMS certified physicist or by the COMS training center is essential for you to run an eye plaque treatment planning successfully.
References Aizman, A., P. T. Finger, U. Shabto, A. Szechter, and A. Berson. (2004). “Palladium 103 (103Pd) plaque radiation therapy for circumscribed choroidal hemangioma with retinal detachment.” Arch Ophthalmol 122:1652–1656. Alberti, W., B. Pothmann, P. Tabor, K. Muskalla, K. P. Hermann, and D. Harder. (1991). “Dosimetry and physical treatment planning for iodine eye plaque therapy.” Int J Radiat Oncol Biol Phys 20:1087–1092. Anderson, L., and S. Chiu-Tsao. “Physical Aspects of Eye Plaque Brachytherapy Using Photon Emitters” in Radiotherapy of Intraocular and Orbital Tumors. R. Sagerman and W. Alberti (eds.). New York: Springer-Verlag, pp. 347–361, 1993. Anderson, L. L., R. Nath, K. A. Weaver, D. Nori, T. L. Phillips, Y. H. Son, S. Chiu-Tsao, A. S. Meigooni, J. A. Meli, and V. Smith. Interstitial Brachytherapy, Physical, Biological and Clinical Considerations. New York: Raven Press, 1990. Astrahan, M. A. (2003). “A patch source model for treatment planning of ruthenium ophthalmic applicators.” Med Phys 30:1219-1228. Astrahan, M. A. (2005). “Improved treatment planning for COMS eye plaques.” Int J Radiat Oncol Biol Phys 61(4):1227–1242. Astrahan, M. A., G. Luxton, G. Jozsef, P. E. Liggett, and A. Petrovich. (1990a). “Optimization of 125I ophthalmic plaque brachytherapy.” Med Phys 17:1053–1057. Astrahan, M. A., G. Luxton, G. Jozsef, T. D. Kampp, P. E. Liggett, M. D. Sapozink, and Z. Petrovich. (1990b). “An interactive treatment planning system for ophthalmic plaque radiotherapy.” Int J Radiat Oncol Biol Phys 18:679–687. Astrahan, M. A., G. Luxton, Q. Pu, and Z. Petrovich. (1997). “Conformal episcleral plaque therapy.” Int J Radiat Oncol Biol Phys 39(2):505–519.
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 701 Augsburger, J. J., J. W.Gamel, V. F. Sardi et al. (1986). “Enucleation vs. cobalt plaque radiotherapy for malignant melanomas of the choroid and ciliary body.” Arch Ophthalmol 104:655–661. Augsburger, J. J., R. G. Peyster, A. M. Markoe, E. G. Guillet, J. A. Shields, and M. E. Haskin. (1987). “Computed tomography of posterior uveal melanoma.” Arch Ophthalmol 105:1512–1516. Beddoe, A. H. (1975). “Isoexposure curves for Co-60 ophthalmic applicators.” Aust Radiol 19:145-151 Beitler, J., B. McCormick, and R. Ellsworth. (1990). “Ocular melanoma: Total dose and dose rate effects with Co-60 plaque therapy.” Radiology 176:275–278. Bilaniuk, L. T., J. F. Schenck, R. A. Zimmerman, H. R. Hart, T. H. Foster, W. A. Edelstein, H. I. Goldberg, and R. I. Grossmam. (1985). “Ocular and orbital lesions: Surface coil MR imaging.” Radiology 156:669–674. Bosworth, J. L., S. Packer, M. Rotman, T. Ho, and P. T. Finger. (1988). “Choroidal melanoma: I-125 plaque therapy.” Radiology 169:249-251. Bridal, S. L., J. M. Correas, A. Saied, and P. Laugier. (2003) “Milestones on the road to hier resolution, quantitative, functional ultrasonic imaging.” IEEE Proceedings 91:1543–1561. Casebow, M. P. (1971). “The calculation and measurement of exposure distributions from Co-60 ophthalmic applicators.” Br J Radiol 44:618–624. Chan, M. F., A. Y. C. Fung, Y.-C. Hu, C.-S. Chui, H. Amols, M. Zaider, and D. Abramson. (2001). “The measurement of three dimensional dose distribution of a ruthenium-106 ophthalmologicalapplicatorusing magnetic resonance imaging of BANG polymer gels.” JACMP 2:85–89. Char, D. H., R. D. Stone, A. R. Irvine, J. B. Crawford, G. F. Hilton, L. I. Lonn, and A. Schwartz. (1980). “Diagnostic modalities in choroidal melanoma.” Am J Ophthalmol 89:223–230. Chiu-Tsao, S., K. O’Brien, R. Sanna, H. S. Tsao, C. Vialotti,Y. S. Chang, M. Rotman, and S. Packer. (1986). “Monte Carlo dosimetry for 125I and 60Co in eye plaque therapy.” Med Phys 13:678–682. Chiu-Tsao, S., L. Boulay, I. Kanna, A. de la Zerda, and J. H. Kim . (1995). “Dose distribution in the eye for 103Pd seed in COMS eye plaque.” Med Phys 22:923. Chiu-Tsao, S., L. Anderson, K. O’Brien, and R. Sanna. (1990). “Dose rate determination for 125I seed.” Med Phys 17:815–825. Chiu-Tsao, S., L. Anderson, K. O’Brien, L. Stabile, and J. Liu. (1993). “Dosimetry for 125I seed (model 6711) in eye plaques.” Med Phys 20:383–389. Coleman, D. J., R. H. Silverman, A. Chabi, M. J. Rondeau, K. K. Shung, J. Cannata, and H. Lincoff. (2004b). “High-resolution ultrasonic imaging of the posterior segment.” Ophthalmology 111(7):1344–1351. Coleman, D. J., R. H. Silverman, M. J. Rondeau, H. C. Boldt, H. O. Lloyd, F. L. Lizzi, T. A. Weingeist, X. Chen, S. Vangveeravong, and R. Folberg. (2004a). “Noninvasive in vivo detection of prognostic indicators for highrisk uveal melanoma: Ultrasound parameter imaging.” Ophthalmology 111(3):558–564. Collaborative Ocular Melanoma Study Group. (1999). “Echography (ultrasound) procedures for the collaborative ocular melanoma study. COMS Report No. 12.” J Ophth Nurs Technol Part I 18(4):143–149, Part II 18(5):219–232. Collaborative Ocular Melanoma Study Group. (2001a). “Collaborative Ocular Melanoma Study (COMS) randomized trial of I-125 Brachytherapy for medium choroidal melanoma: I. Visual acuity after 3 years. COMS Report No. 16.” Ophthalmology 108:348–366. Collaborative ocular melanoma study group. (2001b). “the coms Randomized Trial Of Iodine 125 brachytherapy for choroidal melanoma, II: Characteristics of patients enrolled and not enrolled: COMS Report No. 17.” Arch Ophthalmol 119:951–965. Collaborative Ocular Melanoma Study Group. (2001c). “The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma, III: Initial mortality findings; COMS Report No. 18.” Arch Ophthalmol 119:969–982. Cygler, J., J. Szanto, M. Soubra, and D. Rogers. (1990). “Effects of gold and silver backings on the dose rate around an 125I seed.” Med Phys 17:172–178. Daftari, I., D. Barash, S. Lin, and J. O’Brien. (2001). “Use of high-frequency ultrasound imaging to improve delineation of anterior uveal melanoma for proton irradiation.” Phys Med Biol 46(2):579–590. de la Zerda, A., S. Chiu-Tsao, H. S. Tsao, J. Lin, and J. H. Kim. (1992). “Effect of a COMS eye plaque on 125I dose distribution, in particular reference to the penumbra characteristics.” (Abstract). Med Phys 19:776. de la Zerda, A., S. Chiu-Tsao, H. S. Tsao, J. Lin, L. Boulay, I. Kanna, and J. H. Kim. (1996). “125I eye plaque dose distribution including penumbra characteristics.” Med Phys 23:407–418.
702
Sou-Tung Chiu-Tsao, Ph.D.
Earle, J., R. W. Kline, and D. M. Robertson. (1987). “Selection of iodine-125 for the Collaborative Ocular Melanoma Study.” Arch Ophthalmol 105:763–764. Evans, M. D. C., M. A. Astrahan, and R. Bate. (1993). “Tumor localization using fundus view photography for episcleral plaque therapy.” Med Phys 20:769–775. Fiedler, J. A., V. J. Pisciotta, X. Wu, A. M. Markoe, C. F. Serago, J. G. Schwade, and P. V. Houdek. (1993). “A magnetic resonance imaging-based treatment planning method for episcleral brachytherapy.” Endocuriether/Hypertherm Oncol 9:201–208. Fine, S. L., B. R. Straatsma, and B. S. Hawkins. (1986). “Collaborative Ocular Melanoma Study.” Invest Ophthalmol Visual Sci 139:1987. Finger, P. T. (1997). “Radiation therapy for choroidal melanoma (therapeutic review).” Surv Ophthalmol 42:215–232. Finger, P. T. (2000). “Tumour location affects the incidence of cataract and retinopathy after ophthalmic plaque radiation therapy.” Br J Ophthalmol 84:1068–1070. Finger, P. T., Y. P. Gelman, A. M. Berson, and A. Szechter. (2003). “Palladium-103 plaque radiation therapy for macular degeneration: Results of a 7 year study.” Br J Ophthalmol 87:1497–1503. Finger, P. T., D. F. Lu, A. Buffa et al. (1993). “Palladium-103 versus iodine-125 for ophthalmic plaque radiation therapy.” Int J Radiat Oncol Biol Phys 27:849–854. Finger, P.T., A. Berson, T. Ng, and A. Szechter. (2002). “Palladium-103 plaque radiotherapy for choroidal melanoma: An 11-year study.” Int J Radiat Oncol Biol Phys 54:1438–1445. Fontanesi, J., D. Meyer, S. Xu, and D. Tai. (1993). “Treatment of choroidal melanoma with I-125 plaque.” Int J Radiat Oncol Biol Phys 26:619–623. Foster, F., K. A. Harasiewicz, C. J. Pavlin, G. R. Lockwood, D. H. Turnbull, and D. A. Christopher. (1998). “High frequency ultrasound B-scan imaging.” JEMU (Journal d’Echographie Et De Medicine Par Ultrasons) 19:189–194. Freire, J. E., L. T. Komarnicky, J. A. Garcia-Young et al. (2002). “Custom designed plaque radiotherapy for nonresectable irismelanoma in 38 patients. Tumor control and ocular complications. Int J Radiat Oncol Biol Phys 54:343. Garretson, B. R., D. M. Robertson, and J. D. Earle.(1987). “Choroidal melanoma treatment with iodine-125 brachytherapy.” Arch Ophthalmol 105:1394–1397. Gomori, J. M., R. I. Grossman, J.A. Shields, J. J. Augsburger, P. M. Joseph, and D. Desimeone. (1986). “Choroidal melanomas: Correlation of NMR spectroscopy and MR imaging.” Radiology 158:443–445. Hanson, W. (1996). “Implementation of the new I-125 standard and TG-43 recommendations in a cooperative clinical trial.” Med Phys 23:1020. Harnett, A. N., and F. S. Thomson. (1988). “An iodine-125 plaque for radiotherapy of the eye: manufacture and dosimetric considerations.” [Published Erratum Br J Radiol 61(732):1191] Br J Radiol 61(729):835–838. Hewick, S. A., A. C. Fairhead, J. C. Culy, and H. R. Atta. (2004). “A comparison of 10 MHz and 20 MHz ultrasound probes in imaging the eye and orbit.” Br J Ophthalmol 88(4):551–555. Hesselink, J. R. (2005). “Imaging the eye and orbit. Http://Spinwarp.Ucsd.Edu/Neuroweb/ Text/Orb-220.Htm. Hilaris, B. S., D. Nori, L. Anderson, and S. Chiu-Tsao. “Brachytherapy of Ocular Melanoma” in An Atlas of Brachytherapy. B. S. Hilaris, D. Nori, and L. Anderson (eds.) New York: Macmillan Publishing Co, pp. 304–310, 1988. Hosten, N., A. J. Lemke, B. Sander, R. Wassmuth, K. Terstegge, N. Bornfeld, and R. Felix. (1997). “MR anatomy and small lesions of the eye: Improved Delineation with a special surface coil.” Eur Radiol 7(4):459–463. Houdek, P. V., J. G. Schwade, A. J. Medina, C. A. Poole, K. R. Olsen, D. H. Nicholson, S. Byrne, R. Quencer, R. S. Hinks, and V. Pisciotta. (1989). “MR technique for localization and verification procedures in episcleral brachytherapy.” Int J Radiat Oncol Biol Phys 17:1111–1114. Hungerford, J. L. (2003). “Current trends in the treatment of ocular melanoma by radiotherapy.” Clin Exper Ophthalmol 31:8–13. Jampol, L. M., C. S. Moy, T. G. Murray, S. M. Reynolds, D. M. Albert, A. P. Schachat, K. R. Diddie, R. E. Engstrom, P. T. Finger, K. R. Hovland, L. Joffe, K. R. Olsen, and C. G. Wells. (2002). “For the COMS Follow-up of plaqued eyes working group. The COMS randomized trial of iodine-125 brachytherapy for choroidal melanoma: IV, Local treatment failure and enucleation in the first 5 years after brachytherapy. COMS Report No. 19.” Ophthalmol 109(12):2197–2206.
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 703 Jin, Z., S. Chiu-Tsao, J. A. Shih, and J. H. Kim. (1993). “Bone heterogeneity effect on the 125I dose distribution.” (Abstract). Med Phys 20:896. Kaiserman, I., I. Anteby. I. Chowers, E. Z. Blumenthal, I. Kliers, and J. Pe’er. (2002). “Changes in ultrasound findings in posterior uveal melanoma after ruthenium 106 brachytherapy.” Ophthalmol 109(6):1137–1141. Kanna, I., S. Chiu-Tsao, A. de la Zerda, and J. H. Kim. (1993). “Interseed interference effect on the 125I eye plaque dose distribution.” (Abstract.) Med Phys 20:900. Karolis, C., R. B. Frost, and F. A. Billson. (1990). “A thin I-125 seed eye plaque to treat intraocular tumors using an acrylic insert to precisely position the sources.” Int J Radiat Oncol Biol Phys 18:1209–1213. Kepka, A. G., P. M. Johnson, and R. W. Kline. (1988). “The generalized geometry of eye plaque therapy.” Med Phys 15:375–379. Kline, R., and J. Earle. (1996). “Implications of TG-43 for dose prescription and calculation for I-125 eye plaques.” Med Phys 23:1020. Kline, R. W., and P. D. Yeakel. (1987). “Ocular melanoma, I-125 plaques.” Med Phys 14:475. Knutsen, S., R. Hafslund, O. R. Monge, H. Valen, L. P. Muren, B. L. Reestad, J. Krohn, and O. Dahl. (2001). “Dosimetric verification of a dedicated 3D treatment planning system for episcleral plaque therapy.” Int J Radiat Oncol Biol Phys 51:1159–1166. Krintz, A. L., W. F. Hanson, G. S. Ibbott, et al. (2002). “Verification of plaque simulator dose distribution using radiochromic film.” Med Phys 29:1220–1221. Krintz, A. L., W. F. Hanson, G. S. Ibbott, and D. S. Followill. (2003). “A reanalysis of the collaborative ocular melanoma study medium tumor trial eye plaque dosimetry.” Int J Radiat Oncol Biol Phys 56:889–898. Ling, C. C., E. D. Yorke, I. J. Spiro, D. Kubiatowicz, and D. Bennett. (1983). “Physical dosimetry of 125I seeds of a new design for interstitial implant.” Int J Radiat Oncol Biol Phys 9:1747–1752. Ling, C. C., M. C. Schell, E. D. Yorke, B. B. Palos, and D. Kubiatowicz. (1985). “Two-dimensional dose distribution of 125I seeds.” Med Phys 12:652–655. Ling, C. C., G. T. Chen, J. W. Boothby, K. Weaver, A. Stuart, C. Barnett, D. Char, and T. L. Phillips. (1989). “Computer assisted treatment planning for 125I ophthalmic plaque radiotherapy.” Int J Radiat Oncol Biol Phys 17:405–410. Lommatzsch, P. K., C. Werschnik, and E. Schuster. (2000). “Long-term follow-up of Ru-106/Rh-106 brachytherapy for posterior uveal melanoma.” Graefe’s Arch Clin Exp Ophthalmol 238:129–137. Luxton, G., M. A. Astrahan, P. E. Liggett, D. L. Neblett, D. M. Cohen, and Z. Petrovich. (1988). “Dosimetric calculations and measurements of gold plaque ophthalmic irradiators using iridium-192 and iodine-125 seeds.” Int J Radiat Oncol Biol Phys 15:167–176. Luxton, G., M. A. Astrahan, and Z. Petrovich. (1988). “Backscatter measurements from a single seed of 125I for ophthalmic plaque dosimetry.” Med Phys 15:397–400. Mafee, M. F. “Orbital and Intraocular Lesions” in Clinical Magnetic Resonance Imaging, 2nd Edition. R. R. Edelman, J. R. Hesselink, and M. B. Zlatkin. (eds.). Philadelphia: W.B. Saunders Company, pp. 985–1020, 1996. Meli, J.A., and K. A. Motakabbir. (1993). “The effect of lead, gold and silver backings on dose near 125I seeds.” Med Phys 20:1251–1256. Missotten, L., W. Dirven, A. Van Der Schueren, A. Leys, G. De Meester, and E. Van Limbergen. (1998) “Results Of treatment of choroidal malignant melanoma with high-dose-rate strontium-90 brachytherapy. A retrospective study of 46 patients treated between 1983 and 1995.” Graefe’s Arch Clin Exp Ophthalmol 236:164–173. Myers C.A., and D. H. Abramson. (1988). “Radiation protection, choroidal melanoma and iodine-125 plaques.” J Ophthalmic Nurs Technol 7:103–107. Nag, S., J. M. Quivey, J. D. Earle, D. Followill, J. Fontanesi, and P. T. Finger. (2003a). “The American Brachytherapy Society recommendations for brachytherapy of uveal melanomas.” Int J Radiat Oncol Biol Phys 56(2):544–555. Nag, S., D. Wang, H. Wu, C. J. Bauer, R. B. Chambers, and F. H. Davidorf. (2003b). “Custom-made “nag” eye plaques for 125I brachytherapy.” Int J Radiat Oncol Biol Phys 56(5):1373–1380. Nath, R., A. S. Meigooni, and J. A. Meli. (1990). “Dosimetry on transverse axes of 125I and 192Ir interstitial brachytherapy sources.” Med Phys 17:1032–1040.
704
Sou-Tung Chiu-Tsao, Ph.D.
Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee, Task Group No. 43.” Med Phys 22:209–234. Also available as AAPM Report No. 51. Packer, S. (1987). “Iodine-125 radiation of posterior uveal melanoma.” Ophthalmol 94:1621–1626. Packer, S., M. Rotman, and P. Salanitro. (1984). “Iodine-125 irradiation of choroidal melanoma, clinical experience.” Ophthalmol 91:1700–1708. Packer, S., M. Rotman, R. G. Fairchild, D. M. Albert, H. L. Atkins, and B. Chan. (1980). “Irradiation of choroidal melanoma with iodine 125 ophthalmic plaque.” Arch Ophthalmol 98:1453–1457. Packer, S., S. Stoller, M. L. Lessor, F. S. Mandel, and P. T. Finger. (1992). “Long-term results of iodine-125 irradiation of uveal melanoma.” Ophthalmol 99:767–774. Pavlin, C. J., and F. S. Foster. (1995). Ultrasound Biomicroscopy of the Eye. New York: Springer-Verlag, 1995. Pavlin, C. J., B. Japp, D. G. Payne et al. “Intraoperative Use of Ultrasound in the Management of Choroidal Melanomas” in Ophthalmic Echography. K. C. Ossoinig (ed.). Proceedings of the 10th SIDUO Congress. Dortrecht, The Netherlands: Martinus Nijhoff/ Dr. W. Junk; pp. 391–399, 1987. Pavlin, C. J., B. Japp, E. R. Simpson, H. D. McGowan, and P. J. Fitzpatrick. (1989). “Ultrasound determination of the relationship of radioactive plaques to the base of choroidal melanomas.” Ophthalmol 96:538–542. Peyman, G. A., and M. F. Mafee. (1987). “Uveal melanoma and similar lesions: The role of magnetic resonance imaging.” Radiol Clin No Am 25(3):471–486. Peyster, R. G., J. J. Augsburger, J. A. Shields, T. V. Satchell, A. M. Markoe K. Clarke, and M. E. Haskin. (1985). “Choroidal melanoma: Comparison of CT, fundoscopy, and US.” Radiology 156:675–680. Peyster, R. G., J. J. Augsburger, J. A. Shields, B. L. Hershey, R. Eagle, and M. E. Haskin. (1988). “Intraocular tumors: Evaluation with MR imaging.” Radiology 168:773–779. Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. (2004). “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Also available as AAPM Report No. 84. Rivard, M. J.. W. M. Butler, L. A. Dewerd, M. S. Huq, G. S. Ibbott, Z. Li, M. G. Mitch, R. Nath, and J. F. Williamson. (2004). “Erratum: ‘Update Of AAPM Task Group No. 43 Report: A Revised AAPM Protocol For Brachytherapy Dose Calculations.’” [Med Phys 31:633–674 (2004)]. Med Phys 31:3532–3533. Robertson, D. M., D. G. Fuller, and R. E. Anderson. (1987). “A technique for accurate placement of episcleral iodine125 plaques.” Am J Ophthalmol 103:63–65. Rotman, M., S. Packer, R. Long, S. Chiu-Tsao, and L. Z. Sedhom. “Ophthalmic Plaque Irradiation of Choroidal Melanoma” in Intraocular Tumors. P. K. Lommatzsch and F. C. Blodi (eds.). Berlin: Springer-Verlag, pp. 341–346, 1983. Schell, M. C., C. C. Ling, Z. C. Gramadzki, and K. R. Working. (1987). “Dose distributions of model 6702 I-125 seeds in water.” Int J Radiat Oncol Biol Phys 13:795–799. Schell, M. C., K. A. Weaver, T. L. Phillips, D. H. Char, J. M. Quivey, C. Barnett, and C. C. Ling. (1989). “Design of iodine-125 eye plaques for radiation therapy.” Endocuriether/ Hypertherm Oncol 5:83–90. Spalton, D. J., R. A. Hitchings, and P. A. Hunter. Atlas of Clinical Ophthalmology. 1st Edition. London: Gower Medical Publishing Ltd., 1984. Stroszczynski, C., N. Hosten, N. Bornfeld, T. Wiegel, A. Schueler, P. Foerster, A. J. Lemke, K. T. Hoffmann, and R. Felix. (1998). “Choroidal hemangioma: MR findings And differentiation from uveal melanoma.” Am J Neuroradiol 19(8):1441–1447. Sullivan, J. A., and S. E. Harms. (1986). “Surface-coil MR imaging of orbital neoplasms.” Am J Neuroradiol 7:29–34. Taccini, G., F. Cavagnetto, G. Coscia, S. Garelli, and A. Pilot. (2004). “The determination of dose characteristics of ruthenium ophthalmic applicators using radiochromic film.” Med Phys 24:2034–2037. Weaver. K. A. (1986). “The dosimetry of 125I seed eye plaques.” Med Phys 13:78–83. Weaver K. A., R. Peksens, and C. Barnett. “Special Brachytherapy Procedures and Dosimetry for Tumors of the Brain, Eye, Head and Neck, and Perineum” in Radiation Oncology Physics 1986. J. G. Kereiakes, H. R. Elson, and C. G. Born. (eds.). AAPM Medical Physics Monograph No. 15. New York: American Institute of Physics, pp. 677–699, 1987. Weaver, K. A., V. Smith, D. Huang, C. Barnett, M. Schell, and C. C. Ling. (1989). “Dose parameters of 125i and 192ir seed sources.” Med Phys 18:636–643.
34–Episcleral Eye Plaques for Treatment of Intraocular Malignancies and Benign Diseases 705 Williamson, J. F. (1988a). “Monte Carlo evaluation of specific dose constants in water for 125I seeds.” Med Phys 15:686–694. Williamson, J. F. (1988b). “Theoretical evaluation of dose distributions in water about models 6711 and 6702 125I seeds.” Med Phys 15:891–897. Williamson, J. F. (1991). “Comparison of measured and calculated dose rates in water near I-125 and Ir-192 seeds.” Med Phys 18:776–786. Wu, A., E. S. Sternick, and D. J. Muise. (1988). “Effect of gold shielding on the dosimetry of an I-125 seed at close range.” Med Phys 15: 627–628. Wu, A., and F. Krasin. (1990). “Film dosimetry analyses on the effect of gold shielding for iodine-125 eye plaque therapy for choroidal melanoma.” Med Phys 17:843–846. Yannuzzi, L. A., M. D. Ober, J. S. Slakter, R. F. Spaide, Y. L. Fisher, R. W. Flower , and R. Rosen. (2004). “Ophthalmic fundus imaging: Today and beyond.” Am J Ophthalmol 137(3):511–524. Zimmerman, R. A., and L. T. Bilaniuk. (1988). “Ocular MR imaging.” Radiology 168:875–876.
Chapter 35
Pterygium Brachytherapy Physics Sou-Tung Chiu-Tsao, Ph.D. Beth Israel Medical Center & St. Luke’s-Roosevelt Hospital Center New York, New York Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707 Eye Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707 Pterygium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707 Radiation Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707 90 Sr Eye Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708 Nuclear Data, Energy Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 708 Geometrical/Physical Configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709 Dose-Rate Calibration and Dose Distribution Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 709 National Institute of Standards and Technology (NIST) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713 Accredited Dosimetry Calibration Laboratory (ADCL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713 Regulatory Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 713 Treatment Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715 Prescription Dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715 Treatment Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715 Treatment Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 715 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716 Quality Assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 716
Introduction Eye Anatomy The normal eye anatomy can be depicted by the diagram in Figure 1, and is about 2.5 cm in diameter. The normal adult eye is about 2.5 cm in diameter. There are three layers enclosing the eye; namely, the retina (inner), the choroid (middle), and the sclera (outer). The choroid layer is rich in vasculature. The ciliary body is the anterior extension of the choroid. The uveal tract consists of the iris, the ciliary body, and the choroid. The cornea is the transparent window covering the anterior surface of the eye. The conjunctiva is the thin transparent layer covering the outer surface of the eye. It begins at the outer rim of the cornea, covering the visible portion of the sclera, and lining the inner surface of the eyelids (see Figure 2a.)
Pterygium Pterygium is a raised, wedge-shaped growth of the conjunctiva (see Figure 2b). When it grows over the central cornea, it can affect vision. It may also alter the shape of the cornea, causing astigmatism. It is generally removed surgically.
Radiation Treatment Post-operative radiation treatment is generally delivered using a 90Sr eye applicator, which is a hand-held device (see Figure 3) (IAEA 2002). The detail features and characteristics of the 90Sr eye applicator are described in the next section. The prescription dose is generally specified at the surface of the applicator, which is in contact with the pterygium surgical site. The dose rate at the applicator surface is about 100
707
708
Sou-Tung Chiu-Tsao
Figure 1. Normal eye anatomy diagram. [Reprinted from Brachytherapy Physics, J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.). Figure 1, page 454. © 1995, with permission from Medical Physics Publishing.]
(a)
(b)
Figure 2. (a) Normal conjunctiva. (b) Pterygium growth over cornea. [Illustration by Mark Erickson. Courtesy of St. Luke’s Cataract & Laser Institute, http://www.stlukeseye.com.]
300 cGy per second (cGy/s), depending on the individual applicator and the decay from the date of manufacture. There have been many studies of 90Sr eye applicator treatment using different dose fractionation schemes and various waiting time periods after surgery (Beyer 1991; Cooper 1978; Fukushima et al., 1999” Jürgenliemk-Schulz et al., 2004; Monteiro-Grillo et al., 2000; Nakamatsu et al., 2004; Nishimura et al., 2000; Pajic et al., 2004; Schulz et al., 2003; Wilder et al., 1992). Post-operative radiation treatment indeed provides better treatment outcomes, compared with surgery alone, as demonstrated in Figure 4 (Jürgenliemk-Schulz et al., 2004). However there is no conclusive statement about the best dose fractionation scheme. From some reported studies, it has been found that the waiting time post-op may affect the treatment outcome (Cooper and Lerch 1980). 90
Sr Eye Applicators
Nuclear Data, Energy Spectrum The beta radiation emitted by the radionuclide 90Sr/Y has a wide energy spectrum, with maximum energy 90 Sr is 28.8 years of 2.27 MeV. The half-life of
35–Pterygium Brachytherapy Physics
709
Figure 3. A hand-held 90Sr eye applicator is shown positioned on a model eye sphere. The plastic plate provides radiation shielding to protect the hand. [Reprinted from IAEA TECDOC-1274. “Dosimetry and Medical Radiation Physics Section, IAEA, Calibration of photon and beta ray sources used in brachytherapy, Guidelines on Standardized Procedures at Secondary Standards Dosimetry Laboratories (SSDLs) and Hospitals.” © 2002, with permission from the International Atomic Energy Agency, Vienna, Austria.]
[http://www.nndc.bn1.gov/nudat2/reCenter.jsp?z=38&n=52.]. The daughter radionuclide is 90Y with a 64hour half-life (Soares et al., 2001). Because the radionuclide 90Y remains in equilibrium with the radionuclide 90Sr, the whole unit decays with the half-life of 90Sr.
Geometrical/Physical Configurations The eye applicators come in two basic geometries, with curved or flat applicator surface, as described below. Diagrammed example for a curved applicator and a flat applicator are shown in Figures 5a and 5b, respectively. The detailed construction of a typical flat 90Sr eye applicator is shown in Figure 6. A ceramic disk containing radioactive material fits into a cup with a thin window, 0.1 mm thick. A steel plug fits above the ceramic disk in the cup.
Dose-Rate Calibration And Dose Distribution Measurement The eye applicator sources are calibrated in terms of surface absorbed-dose-rate to water (Cross et al., 2001; Deasy and Soares 1994; Goetsch and Sunderland 1991; Soares 1991, 1995; Soares et al. 2001). Surface dose rate calibration procedure is given in chapter 11 by Dr. Larry DeWerd. The surface dose rate in the central area of 4 mm diameter is reported in the calibration report, but this varies widely by model. The depth dose curve for a planar 90Sr eye applicator is shown in Figure 7.
710
Sou-Tung Chiu-Tsao
Figure 4. Comparison of treatment outcomes using postoperative radiation (left panels) with those with surgery alone (right panels). [Reproduced from Int J Radiat Oncol Biol Phys, “Prevention of pterygium recurrence by postoperative single-dose beta-irradiation: A prospective randomized clinical double-blind trial.” I. M. Jürgenliemk-Schulz, L. J. Hartman, J. M. Roesink, R. J. Tersteeg, I. van Der Tweel, H. B. Kal, M. P. Mourits, and H. K. Wyrdeman, vol 59, pp. 1138–1147. © 2004, permission from Elsevier.]
35–Pterygium Brachytherapy Physics
711
(a)
(b) Figure 5. (a) Configuration of a flat 90Sr eye applicator, model SIA20. With 12 mm total diameter, the active area is 9 mm in diameter. The applicator has a thick backing and a stainless steel filter of 0.05 mm. Left: front view; right, side view showing the handle. (b) Configuration of a concave 90Sr eye applicator, model SIA6. The average diameter is 12 mm, with 15 mm total diameter. The applicator is 1 mm thick and has a stainless steel filter of 0.1 mm. The radius of curvature is 10 mm. Left, front view; right, side view showing the handle. [Reprinted with permission from NCS (Netherlands Commision on Radiation Dosimetry). “Quality control of sealed beta sources in brachytherapy. Recommendations on detectors, measurement procedures and quality control of beta sources.” Report 14, 2004. This report is available through http://www.ncs-dos.org.]
Figure 6. Side view diagram of a typical 90Sr planar eye applicator construction. [Courtesy of Dr. Christopher Soares, NIST (Soares 2004).]
712
Sou-Tung Chiu-Tsao
Figure 7. Relative depth dose curve (normalized to 1 mm depth) for 90Sr planar eye applicator is shown by circle symbols. [Reprinted from IAEA TECDOC-1274. “Dosimetry and Medical Radiation Physics Section, IAEA, Calibration of photon and beta ray sources used in brachytherapy, Guidelines on Standardized Procedures at Secondary Standards Dosimetry Laboratories (SSDLs) and Hospitals.” © 2002, with permission from the International Atomic Energy Agency, Vienna, Austria.]
Figure 8. An example plot of isodose curves in mGy/s for an 90Sr eye applicator as measured by NIST using radiochromic film dosimetry. [Courtesy of Dr. Christopher Soares, NIST (Soares 2004).]
35–Pterygium Brachytherapy Physics
713
National Institute of Standards and Technology (NIST) NIST offers a service for the calibration1 of 90Sr eye applicators that provides: (1) the calculation of the emission rate, and (2) a mapping of the 90Sr distribution across the surface of the applicator in order to ascertain uniformity of dose, using an extrapolation chamber and radiochromic film dosimetry (Soares 2004). Accredited Dosimetry Calibration Laboratory (ADCL) The ADCL at the University of Wisconsin-Madison also offers a similar service.
Regulatory Guidelines The Nuclear Regulatory Commission (NRC) issued NRC Information Notice 94-17 on March 11, 1994 (NRC 1994) regarding the regulatory guidelines of establishing a Quality Management Program (QMP) on the use of 90Sr eye applicator in the treatment of superficial eye conditions. They informed licensees that: (1) Researchers at the National Institute of Standards and Technology (NIST) recognized large discrepancies among calibrated outputs assigned to 90Sr eye applicators. (2) Original manufacturer calibrations were expressed in older (traditional) units, which differed from the International System (SI) units. (3) Calibration values were not comparable for units from different manufacturers. (4) Discrepancies larger than 10% could exist when comparing output measurements between competent measurement laboratories using state-of-the-art techniques.
The QMP that is submitted to NRC should include written policies and procedures that meet the five objectives, as described in Title 10, Code of Federal Regulations,10 CFR 35.32(a): 1. That prior to administration, a written directive, signed and dated by an authorized user, is prepared for each applicable administration. (A written directive for 90Sr eye applicators means an order, in writing, for a specific patient, dated and signed by an authorized user prior to administration of radiation. It must include the radioisotope, the treatment site, source strength (corrected for decay), and exposure time (or equivalently, the total dose). 2. That prior to each administration, the patient is identified by more than one method as the individual named in the written directive. 3. That final plans of treatment and related calculations are in accordance with the respective written directive. 4. That each administration is in accordance with the written directive. 5. That any unintended deviation from the written directive is identified and evaluated, and appropriate action taken.
1
For further information on the NIST calibration service, contact Dr. Christopher Soares at (301) 975-5589.
714
Sou-Tung Chiu-Tsao
Then, the Advisory Committee on the Medical Uses of Isotopes (ACMUI) advised the NRC staff that calibration was not a critical factor in the use of 90Sr eye applicators for treating pterygium because licensees treated for response rather than to tolerance. Members of the ACMUI recommended that the 90Sr eye applicator licensees continue to treat patients as they had using the original manufacturer’s 90Sr activity values and treatment times derived from decay correction charts. The ACMUI further recommended cautioning licensees that if they were to use an applicator other than the one currently in their possession, or if they were to buy a new one, their current technique might not be applicable to another device because of variances in stated and actual exposure rates for the different applicators. Effective October 24, 2002, NRC amended its regulations at 10 CFR 35.432, “Calibration measurements of brachytherapy sources.” This section of the regulations requires, among other things, that brachytherapy sources be calibrated before first medical use on or after the effective date of the rule. The effect of this requirement on 90Sr eye applicator licensees and permittees was that they all had to have their applicators calibrated using the new techniques that more accurately measure the actual exposure rate for the eye applicator and provide the calibration results in SI units. Thus, the calibration values and units may be significantly different from those provided in the original manufacturer’s calibrations or more recent calibrations. NRC expects each 90Sr eye applicator licensee and permittee to carefully review the results of the new calibration with its authorized medical physicist to assure proper interpretation of the calibration results. This review should include discussion of appropriate changes to the written directives so that the patient treatments are based on the new calibration information. As described below, the authorized user has two options and should select the best option, based on his/her medical judgment, for his/her practice. It is important for the licensee or permittee to keep in mind that, although the units and calibration values may be different, the actual amount of radioactive material (corrected for decreases from radioactive decay) contained in each applicator and its distribution in the applicator remains the same. The written directive regulations for eye applicator procedures require that the licensee (or permittee) include the following information in the written directive: • Prior to implantation: treatment site, radionuclide, and dose • After implantation, but prior to completion of procedure: treatment site, radionuclide, number of sources, and total source strength, and exposure time (or total dose). The following two optional approaches have been recommended by the NRC. Option 1 Maintaining the same treatment regimen—revising total dose in the written directive. If the licensee’s or permittee’s medical experience with its 90Sr treatment regimen before the October 2002 required recalibration indicated that the treatments provided appropriate medical results, the authorized user may elect to administer the same amount of radiation to the treatment site and provide the same medical results after recalibration, as before. Even though the units and calibration values may be different from those of an earlier calibration, the actual exposure rate (corrected for decreases from radioactive decay) remains the same. Therefore, to administer the same amount of radiation from a specific eye applicator, the authorized user should keep the treatment time the same and adjust the total dose to a new value, based on the new calibration exposure rate. For example, based on the original manufacturer’s calibration data, the authorized user believes that the exposure rate is 0.42 Gray per second (Gy/s), but the exposure rate based on the new calibration certificate is really 0.55 Gy/s, a value 31% higher. The authorized user’s medical experience is that the treatment times used in the past provided good medical results. To achieve the same medical results, the authorized
35–Pterygium Brachytherapy Physics
715
user would keep the administration time the same and increase the value of the total dose documented in the written directive by 31%. The treatment times will change with time, because of the normal radioactive decay of the 90Sr. Option 2 Changing the treatment regimen—retaining the same written directive total dose value. Although the authorized user’s medical experience with his/her 90Sr treatment regimen before the October 2002 required calibration indicated that the treatments provided appropriate medical results, the authorized user decides he/she wants to keep the value of the total dose the same in future written directives. In this case, the authorized user will adjust the treatment time so that the total dose value recorded in the written directive does not change. Because the treatment regimen is changed, the authorized user should monitor his/her patients to see if the expected medical results stay the same or change. For example, based on the original manufacturer’s calibration data, the authorized user believes that the exposure rate is 0.42 Gy/s, but the exposure rate based on the new calibration certificate is really 0.55 Gy/s, a value 31% higher. In this example, the authorized user decides to keep the total dose value the same in the written directive. To achieve the same value for the total dose, the authorized user would have to reduce the administration time by 31%. NRC issued Information Notice 2004-02 on February 5, 2004, regarding the use of newly calibrated strontium-90 (90Sr) eye applicators (NRC 2004). This notice alerts medical-use licensees and permittees to discrepancies associated with new and older calibration values for 90Sr eye applicators and effects this may have on their use under the regulatory requirements in 10 CFR Part 35, “Medical Use of Byproduct Material.”
Treatment Planning Treatment planning calculation is generally done manually with the aid of a calculator to calculate the treatment time.
Prescription Dose Prescription dose per fraction and total dose at the applicator surface is decided, written, and dated on the patient chart by the radiation oncologist (authorized user) on the case.
Treatment Time Based on the prescription dose per fraction and the calibrated surface dose rate on the day of application, the treatment time is calculated by a physicist or dosimetrist, and checked by a second physics team member. The calculation and check are documented in the patient chart.
Treatment Delivery Treatment delivery is carried out by the team of radiation oncologist, physicist, and nurse. Before the application of the eye applicator, the nurse would prepare the patient and help the patient lying supine on a treatment table.
716
Sou-Tung Chiu-Tsao
Procedure The radiation oncologist on the case holds the eye applicator and positions the applicator surface on the surgical site of pterygium for the duration of the calculated treatment time to deliver the prescribed dose. The treatment delivery time can be based on a stopwatch held by a physicist or a nurse. This delivery time should be checked by using a digital timer.
Quality Assurance The user should establish a Quality Management Program (QMP) and follow the procedure. Assure that: • the dose rate calibration is performed and reported by NIST or ADCL • the prescription dose is properly written in the treatment chart and given to physicist for calculation of treatment time • the treatment time is calculated according to the prescription dose and the dose rate (incorporating decay) as calibrated by NIST or ADCL • the treatment time calculation is checked and properly documented • the patient’s identity is checked prior to treatment delivery using at least two different methods • the treatment site is accurate according to the surgeon’s description • the treatment delivery time per fraction is accurate, using the proper timer device • the treatment delivery record is properly documented.
References Beyer, D. C. (1991). “Pterygia: single-fraction postoperative beta irradiation.” Radiology 178:569–571. Cooper, J. S. (1978). “Postoperative irradiation of pterygia: Ten more years of experience.” Radiology 128:753–756. Cooper, J. S., and I. A. Lerch. “Postoperative irradiation of pterygia: An unexpected effect of the time/dose relationship.” Radiology 135(3):743–745. Cross, W. G., J. Hokkanen, H. Järvinen, F. Mourtada, P. Sipilä, C. G. Soares, and S. Vynckier. (2001). “Calculation of beta-ray dose distributions from ophthalmic applicators and comparison with measurements in a model eye.” Med Phys 28:1385. Deasy, J. O., and C. G. Soares. (1994). “Extrapolation chamber measurements of 90Sr+90Y beta-particle ophthalmic applicator dose rates.” Med Phys 21:91. Fukushima, S., T. Inoue, T. Inoue, and S. Ozeki. (1999). “Postoperative irradiation of pterygium with 90Sr eye applicator.” Int J Radiat Oncol Biol Phys 43:597–600. Goetsch, S. J., and K. A. Sunderland. (1991). “Surface dose rate calibration of Sr-90 plane ophthalmic applicators.” Med Phys 18:161. International Atomic Energy Agency (IAEA). TECDOC-1274. “Dosimetry and Medical Radiation Physics Section, IAEA, Calibration of photon and beta ray sources used in brachytherapy, Guidelines on Standardized Procedures at Secondary Standards Dosimetry Laboratories (SSDLs) and Hospitals.” Vienna: IAEA, 2002. Jürgenliemk-Schulz, I. M., L. J. Hartman, J. M. Roesink, R. J. Tersteeg, I. van Der Tweel, H. B. Kal, M. P. Mourits, and H. K. Wyrdeman. (2004). “Prevention of pterygium recurrence by postoperative single-dose beta-irradiation: A prospective randomized clinical double-blind trial.” Int J Radiat Oncol Biol Phys 59:1138–1147. Monteiro-Grillo, I., L. Gaspar, M. Monteiro-Grillo, F. Pires, and J. M. Ribeiro da Silva. (2000). “Postoperative irradiation of primary or recurrent pterygium: Results and sequelae.” Int J Radiat Oncol Biol Phys 48:865–869.
35–Pterygium Brachytherapy Physics
717
Nakamatsu, K., Y. Nishimura, Y. Yagyu, R. Koike, and S. Kanamori. (2004). “A randomized trial of 30Gy/3 fractions versus 40Gy/4 fractions in postoperative strontium-90 radiation therapy (RT) for pterygia.” Int J Radiat Oncol Biol Phys 60:S551. NCS (Netherlands Commision on Radiation Dosimetry). “Quality control of sealed beta sources in brachytherapy. Recommendations on detectors, measurement procedures and quality control of beta sources.” Report 14, 2004. This report is available through http://www.ncs-dos.org. Nishimura, Y., A. Nakai, T. Yoshimatsu, Y. Yagyu, K. Nakamatsu, H. Shindo, and O. Ishida. (2000). “Long-term results of fractionated strontium-90 radiation therapy for pterygia.” Int J Radiat Oncol Biol Phys 46:137–141. Nuclear Regulatory Commission (NRC) (1994). NRC Information Notice 94-17: Submission of Quality Management Plan (QMP), Calibration, and Use. http://www.nrc.gov/reading-rm/doc-collections/gen-comm/info-notices/ 1994/in94017.html Nuclear Regulatory Commission (NRC) (2004). NRC Information Notice 2004-02: Strontium-90 Eye Applicators: New Calibration Values and Use.http://www.nrc.gov/reading-rm/doc-collections/gen-comm/info-notices/ 2004/in200402.pdf Pajic, B., A. Pallas, D. Aebersold, G. Gruber, and R. H. Greiner. (2004). “Prospective study on exclusive, nonsurgical strontium-/yttrium-90 irradiation of pterygia.” Strahlenther Onkol 180:510–516. Schulz, I. M., L. J. Hartman, J. M. Roesink, R. J. Tersteeg, I. van Der Tweel, H. B. Kal, M. P. Mourits, and H. K. Wyrdeman. (2003). “Prevention of pterygium recurrence by postoperative single dose beta-irradiation: A prospective double-blind trial.” Int J Radiat Oncol Biol Phys 57:S252. Soares, C. G. (1991). “Calibration of ophthalmic applicators at NIST: A revised approach.” Med Phys 18:787. Soares, C. G. (1995). “Comparison of NIST and manufacturer calibrations of 90Sr+90Y ophthalmic applicators.” Med Phys 22:1487. Soares, C. G. (2004). “Calibration of Ophthalmic Applicators.” IRD-P-09 report. http://physics.nist.gov/Divisions/ Div846/QualMan/Procdures/WebProcedure09v200.pdf. Soares, C. G., S. Vynckier, H. Järvinen, W. G. Cross, P. Sipilä, D. Flühs, B. Schaeken, F. A. Mourtada, G. A. Bass, and T. T. Williams. (2001). “Dosimetry of beta-ray ophthalmic applicators: Comparison of different measurement methods.” Med Phys 28:1373. Wilder, R. B., J. M. Buatti, J. M. Kittelson, D. S. Shimm, P. M. Harari, E. E. Rogoff, and J. R. Cassady. (1992). “Pterygium treated with excision and postoperative beta irradiation.” Int J Radiat Oncol Biol Phys 23:533–537.
Chapter 36
Brachytherapy for Brain Tumors Shirish K. Jani, Ph.D.1 and Patrick Hitchon, M.D.2 1 Department of Radiation Oncology Sharp Memorial Medical Campus San Diego, California 2 Department of Neurosurgery The University of Iowa Hospitals and Clinics Iowa City, Iowa Rationale for Brain Tumor Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 720 Outcome of Brachytherapy Treatment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722 Future of Brain Brachytherapy in the Era of Radiosurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . 722 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 724
Rationale for Brain Tumor Brachytherapy Radiation therapy plays a central role in the treatment of brain tumors in adults. It is the most effective nonsurgical therapy for patients with malignant gliomas and also has an important role in the treatment of patients with low-grade gliomas. Historically, the whole-brain irradiation of up to a dose of 43 to 45 Gy was followed by a limited-field treatment. For low-grade astrocytoma, the role of postoperative radiotherapy in the management of incompletely resected low-grade astrocytomas has not been firmly established. However, when a complete surgical resection is not performed, postoperative radiation may be recommended, particularly when the astrocytoma has a high proliferative potential. Some patients with incompletely resected low-grade gliomas can be followed and radiation therapy deferred until clinical or radiographic progression occurs. Radiation therapy usually includes 1.7- to 2-Gy daily fractions to a total dose ≤50 Gy. The treatment fields include the primary tumor volume only, as defined by magnetic resonance imaging (MRI), and do not encompass the whole brain. In low-grade astrocytomas, radiation therapy can be expected to produce a 5-year survival rate of 50% and a 10-year survival rate of 20%. For high-grade glioma, an analysis of three studies done by the Brain Tumor Study Group (BTSG) showed that postoperative radiotherapy doses >50 Gy were significantly better in improving survival than no postoperative treatment, and that 60 Gy resulted in significantly prolonged survival compared with 50 Gy (Sheline et al. 1980). Based on these data, a standard dose for high-grade glioma is considered to be 60 Gy in 30 to 33 fractions, which corresponds to a dose just above the threshold for radionecrosis. Approximately half of patients with anaplastic astrocytomas exhibit radiographic evidence of response following 60 Gy of radiation, as compared with 25% of patients with glioblastoma multiforme (GBM). Complete radiographic response is rare in either case. The results of standard external beam radiotherapy in patients with malignant gliomas have been poor. Patients with glioblastoma multiforme have a median survival of 9 to 12 months, while patients with anaplastic astrocytomas survive a median of 3 years. In an attempt to improve these poor results, a number of new approaches have been tried, including the use of radiosensitizers, boron neutron capture therapy (BNCT), hyper-fractionated radiotherapy (HFRT), focal dose escalation with interstitial brachytherapy, and radiosurgery. Brachytherapy became an attractive choice based upon the patterns of failure in these patients. Almost 90% of patients die within 18 months after therapy, most commonly because of local persistence of the tumor, which may be controlled if a sufficient amount of irradiation can be delivered. Postoperative radiation therapy of 60 Gy offered the best median survival rate. However, the response to external-beam
719
720
Shirish K. Jani and Patrick Hitchon
radiation therapy had reached a plateau because of the intolerance of healthy brain tissue to excessive irradiation. To treat these tumors with high local doses, interstitial brachytherapy has been used with debulking surgery as far back as the early 1960s (Bond et al. 1965, Chase et al. 1961, Correll et al. 1961, Gutin and Dormandy 1982, Hosobuchi et al. 1980, Mundinger et al. 1978, Mundinger 1979). In later years, stereotactic techniques were employed to place the radioactive seeds precisely, whereby the tumor received the highest dose and surrounding tissues were spared, as a result of the rapid fall-off in dose with distance from the sources. In this chapter, we provide a brief review of the physics of brain tumor brachytherapy including a description of radioisotopes, dosimetry, and the current status of the role brachytherapy plays in the era of stereotactic radiosurgery.
Methods Early brachytherapy for malignant brain tumors was performed with high activity 125I seed sources in a temporary setting. The University of California San Francisco (UCSF) method utilized a single source placed with the help of stereotactic frame (Gutin et al. 1981, Prados et al. 1992). The Brain Tumor Cooperative Group (BTCG) randomized the patients between external radiotherapy (60 Gy) and brachytherapy (60 Gy) plus external radiotherapy (Green et al. 1994). The brachytherapy was performed with a multicatheter implant using high-activity 125I seed sources (Hitchon et al. 1987, 1992; Jani et al. 1987a,b,c). 125 I became the choice of radioisotope because of its low energy of emission (Jani 1993). The source handling and patient hospitalization became easy to deal with. The implant procedure called for debulking of the tumor and placement of the catheters using a stereotactic frame. Pre-implant planning and dose calculation were carried out using a treatment-planning computer that utilized the computed tomography (CT) scan. Dose was given at a rate of approximately 40 cGy/hour and the implant time was approximately 6 days. Figure 1(a) shows the CT scan of a patient undergoing volumetric 125I seed implant with the help of a plastic template for optimal placement of the catheters. A follow-up scan of this patient is shown in Figure 1(b). Permanent implants with 125I have also been tried in treating metastatic brain tumors (Huang et al. 2002). Multiple iodine seeds were implanted to deliver a dose of about 350 Gy over the lifetime of the radioisotope. 131 I tagged monoclonal antibody has been used in a phase-I study of patients with newly diagnosed malignant glioma (Akabani et al. 2000). Studies are underway to optimize the dose in mCi of 131I. Liquid 32 P has been used in treating cystic brain lesions (Taasan et al. 1985). The 32P chromic phosphate with an activity of 0.1 to 3 mCi has been injected into the cyst with barbotage to deliver a dosage of about 200 Gy to the cystic wall. Knowing the very poor outcome of treatment in malignant glioma, some researchers have utilized neutron brachytherapy using 252Cf sources (Maruyama et al. 1984). 192Ir seed sources have also been used in temporary implant of cystic brain tumor (Matsumoto et al. 1998). Recently, the liquid-based 125I has been utilized in a Food and Drug Administration (FDA)-approved device called the “GliaSite” (Dempsey et al. 1998, Proxima 2004). As shown in Figure 2, it is a radiopaque silicone tube with a low-profile infusion port at the proximal end and a silicon balloon at the distal end. This distal balloon is a dual-layer configuration. The inner balloon serves as a reservoir for the liquidbased 125I radionuclide called “Iotrex.” The outer balloon is a safety reservoir. The inflatable balloon catheter is placed in the resection cavity at the time of surgical resection of the tumor. The balloon is inflated to a diameter of 2 to 4 cm, depending upon its conformity to the resected cavity. The balloon is then filled with Iotrex, which is an aqueous solution of organically bound 125I. Prescription doses of 40 to 60 Gy at 0.5 cm to 1 cm from the balloon surface have been used so far in treating patients with previously radiated malignant glioma. Figures 3 and 4 illustrate the use of The Gliasite. [Note: The materials of Figures 2, 3, and 4 have been reproduced with permission from Proxima Therapeutics, the manufacturer of GliaSite.]
36–Brachytherapy for Brain Tumors
Figure 1(a). Afterloading catheters are implanted as part of the postoperative brachytherapy using I for primary treatment of malignant glioma under BTCG protocol (The University of Iowa Hospitals and Clinics). The catheters are glued to the buttons, which are sutured to the scalp.
125
Figure 1(b). Dose distribution in one of the BTCG patients who underwent 125I implant. The tumor is encompassed within 6000 cGy isodose volume with a total treatment time of 150 hours at a dose rate of 40 cGy per hour.
Figure 1(c). An 125I seed is seen within the enhancing tumor margin of a patient with glioblastoma multiforme (left). At 8 months’ follow-up, the original tumor bed is free of disease; however, there is evidence of enhancing tumor recurring in the contra-lateral midbrain.
721
722
Shirish K. Jani and Patrick Hitchon
Outcome of Brachytherapy Treatment Brachytherapy has yet to prove its role as a part of overall management in treating malignant brain tumor (Laperriere et al. 1998, Mayr et al. 2002). Studies have shown limited benefit of brachytherapy in these patients. We, as physicists, play an important role in minimizing complications of brachytherapy. It has been demonstrated that careful planning is required to avoid serious complications of brachytherapy (Sneed et al. 1996). Brachytherapy can induce focal radionecrosis. This complication produces symptoms of mass effect in about 50% of patients with malignant glioma, requiring resection to remove the necrotic debris. Occasionally, treatment with corticosteroids can control the edema around the radionecrotic area, but often the patient becomes steroid dependent, with all of the attendant complications of chronic steroid use. Radionecrosis can be a significant limitation of the focal radiotherapy techniques.
Future of Brain Brachytherapy in the Era of Radiosurgery Focal irradiation with brachytherapy has resulted in modestly improved local control and survival in selected patients with recurrent or primary glioblastoma multiforme. Brachytherapy may be appropriate for treating lesions that are ≤4 cm in patients with performance status ≥70% and no subependymal spread. However, it has largely been replaced by radiosurgery. Over the last several years, there has been growing interest in the use of radiosurgery in the treatment of primary and recurrent malignant brain tumors. Radiosurgery is currently performed with one of three technologies: high-energy photons produced by linear accelerators, the Gamma Knife (AB Elekta, Stockholm, Sweden) or, less frequently, charged particles, such as protons or other ions produced by cyclotrons or synchrotrons. Radiosurgery is a relatively safe, effective means for improving both local control and survival in patients with newly diagnosed glioblastoma multiforme. Patient age and tumor volume are highly predictive of outcome. However, randomized trials are needed to further evaluate both the efficacy of stereotactic radiosurgery as a primary adjuvant modality for glioblastoma multiforme and the relative prognostic significance of patient and tumor characteristics. There appears to be no significant advantage to radiosurgery as a primary therapy for anaplastic astrocytoma or low-grade gliomas.
Figure 2. The GliaSite balloon catheter with a bottle containing Iotrex liquid as a source of 125 I radioisotope (Reprinted with permission from Proxima Therapeutics, Alpharetta, GA).
36–Brachytherapy for Brain Tumors
(a)
(b)
723
(c)
Figure 3. Application of the GliaSite in treating a metastatic brain tumor. (a) Pre-implant scan of 2/14/02. (b) The GliaSite implant scan of 2/20/02. (c) Post brachytherapy and GliaSite removal: scan of 4/3/02. (Reprinted with permission from Proxima Therapeutics, Alpharetta, GA).
(a)
(b)
(c)
(d)
Figure 4. Application of the GliaSite in treating a cystic glioblastoma multiforme. (a) Post-operative scan revealed a recurrent GBM as well as significant cerebral edema. (b) Post-operative MRI showed proper size and placement of the GliaSite balloon. (c) Four-month follow-up scan showed no evidence of disease and edema associated with original tumor. (d) Six-month scan showed similar absence of disease and edema. (Reprinted with permission from Proxima Therapeutics, Alpharetta, GA).
724
Shirish K. Jani and Patrick Hitchon
The survival rates, patterns of recurrence, and rates of complications (including radionecrosis) of radiosurgery and brachytherapy are similar in treating malignant glioma. Radiosurgery may be a more appealing approach than brachytherapy for the management of highly focal malignant gliomas because it is a noninvasive, single-day procedure that can usually be performed in an outpatient setting.
References Akabani, G., I. Cokgor, R. E. Coleman, D. González Trotter, T. Wong, H. S. Friedman, A. Garcia-Turner, J. E. Herndon II, D. DeLong, R. E. McLendon, X.-G. Zhao, C.N. Pegram, D.D. Bigner, and M. R. Zalutsky. (2000). “Dosimetry and dose-response relationships in newly diagnosed patients treated with iodine-131-labeled antitenacin monoclonal antibody therapy.” Int J Radiat Oncol Biol Phys 46:947–958. Bond, W. H., D. Richards, and E. Turner. (1965). “Experience with radioactive gold in the treatment of craniopharyngioma.” J Neurol Neurosurg Psychiat 28:30–38. Chase, N.E., H. L. Atkins, and J.W. Correll. (1961). “Interstitial irradiation of brain tumors with iridium 192.” Radiology 77:842–843. Correll, J. W., N. E. Chase, and H. L. Atkins. (1961). “New technique for interstitial irradiation of brain tumors.” J Neurosurg 18:800–803. Dempsey, J. F., J. A. Williams, J. B. Stubbs, T. J. Patrick, and J. F. Williamson. (1998). “Dosimetric properties of a novel brachytherapy balloon applicator for the treatment of malignant brain-tumor resection-cavity margins.” Int J Radiat Oncol Biol Phys 42:421–429. Green, S. B., W. R. Shapiro, P. C. Burger, R. G. Selker, J. C. VanGilder, S. Saris, M. G. Malkin, J. Mealy, J. Neal, J. T. Robertson, and J. Olson. “A Randomized Trial of Interstitial Radiotherapy (RT) Boost for Newly Diagnosed Malignant Glioma: Brain Tumor Cooperative Group (BTCG) Trial 8701.” Proceedings of the American Society of Clinical Oncology, Dallas, TX, p. 174, 1994. Gutin, P. H., and R. H. Dormandy. (1982). “A coaxial catheter system for afterloading radioactive sources for interstitial irradiation of brain tumors.” J Neurosurg 56:734–735. Gutin, P. H., T. L. Phillips, and Y. Hosobuchi. (1981). “Permanent and removable implants for the brachytherapy of brain tumors.” Int J Radiat Oncol Biol Phys 7:1371–1381. Hitchon, P. W., S. K. Jani, J. C. Van Gilder, J. C. Godersky, and J. F. Doornbos. (1987). “Interstitial radiation in recurrent gliomas.” Appl Neurophysiol 50:292–294. Hitchon, P. W., J. C. VanGilder, B. C. Wen, and S. K. Jani. (1992). “Brachytherapy for malignant recurrent and untreated gliomas.” Stereotact Funct Neurosurg 59:174–178. Hosobuchi,Y., T. L. Philips, T. A. Stupar, and P. H. Gutin. (1980). “Interstitial brachytherapy of primary brain tumors. preliminary report.” J Neurosurg 53:613–617. Huang, K., P. K. Sneed, M. S. Berger, S. Kunwar, D. A. Larson, W. M. Wara, and M. W. McDermott. (2002). “Permanent iodine-125 brachytherapy for brain metastasis resection cavities.” Int J Radiat Oncol Biol Phys 50:80 (Poster presentation). Jani, S. K. Handbook of Dosimetry Data for Radiotherapy. Boca Raton, FL: CRC Press, 1993. Jani, S. K., P. W. Hitchon, J. C. Van Gilder, E. C. Pennington, and D. H. Hussey. (1987a). “Choice of radioisotope in stereotactic interstitial radiotherapy of small brain tumors.” Appl Neurophysiol 50:295–301. Jani, S. K., J. E. Wacha, P. W. Hitchon, and J. C. VanGilder. (1987b). “Dosimetric comparison between stereotaxic and volumetric brain implants.” Med Dosimetry 12:23–26. Jani, S. K., P. W. Hitchon, J. C. Van Gilder, and B. C. Wen. (1987). “Normal brain irradiation during stereotactic brain implants using radioactive iodine-125.” Appl Neurophysiol 50:295–301. Laperriere, N. J., P. M. Leung, S. McKenzie, M. Milosevic, S. Wong, J. Glen, M. Pintilie, and M. Bernstein. (2002). “Randomized study of brachytherapy in the initial management of patients with malignant astrocytoma.” Int J Oncol 21(4):817–23. Maruyama, Y., H. W. Chin, A. B. Young, J. Beach, P. Tibbs, S. Goldstein, and W. Markesbery. (1984). “CT and MR for brain tumor implant therapy using CF-252 neutrons.” J Neurooncol 2:349–360. Matsumoto, K., E. Tada, N. Tamesa, S. Tomita, and T. Ohmoto. (1998). “Stereotactic brachytherapy for a cystic metastatic brain tumor in the midbrain.” Case report. J Neurosurg 88:141–144.
36–Brachytherapy for Brain Tumors
725
Mayr, M. T., I. R. Crocker, E. K. Butker, H. Williams, G. A. Cotsonis, J. J. Olson. (2002). “Results of interstitial brachytherapy for malignant brain tumors.” Int J Oncol 21:817–823. Mundinger, F. “Rationale and Methods of Interstitial Iridium 192 or Iodine 125 Protracted Long-term Irradiation” in Stereotactic Cerebral Irradiation. Inserm Symposium No. 12. G. Szikla (ed.). Amsterdam: Elsevier North Holland, 1979. Mundinger, F., W. Birg, and C. B. Ostertag. (1978). “Treatment of small cerebral gliomas with CT-aided stereotaxic curietherapy.” Neuroradiology 16:564–567. Prados, M. D., P. H. Gutin, T. L. Phillips, W. M. Wara, P. K. Sneed, and D. A. Larson. (1992). “Interstitial brachytherapy for newly diagnosed patients with malignant gliomas: The UCSF experience.” Int J Radiat Oncol Biol Phys 24:593–597. Proxima Therapeutics, Alpharetta, GA, www: http://www.proximatherapeutics.com, 2004. Sheline, G. E., W. M. Wara, and V. Smith. (1980). “Therapeutic irradiation and brain injury.” Int J Radiat Oncol Biol Phys 6:1215–1228. Sneed, P. K., K. R. Lamborn, D. A. Larson, and M. D. Prados. (1996). “Demonstration of brachytherapy boost doseresponse relationships in glioblastoma multiforme.” Int J Radiat Oncol Biol Phys 35:37–44. Taasan, V., B. Shapiro, J. A. Taren, and W. H. Beierwaltes. (1985). “Phosphorous–32 therapy of cystic grade IV astrocytoma: technique and preliminary application.” J Nucl Med 26:1335.
Chapter 37
Review of Head and Neck Brachytherapy Zuofeng Li, D.Sc. Department of Radiation Oncology Washington University School of Medicine St. Louis, Missouri Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 727 Sources and Dosimetry Systems Used in Head and Neck Brachytherapy . . . . . . . . . . . . . . . . 727 Cesium Tubes and Needle Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 728 Low Dose-Rate (LDR) 192Ir Ribbons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 729 Remote Afterloading High Dose-Rate (HDR) and Pulsed Dose-Rate (PDR) Sources . . . . . . . . . . 730 Other Sources Used in Head and Neck Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730 Dosimetry Systems and Dose Prescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730 Target Delineation and Implant Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 731 Basic Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732 Hairpins and Looping Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 732 Non-Looping Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 735 Treatment Complications and Pretreatment Dental Evaluation . . . . . . . . . . . . . . . . . . . . . . . . 736 Treatment Complications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 736 Dental Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 736 Clinical Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737 Nasopharynx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737 Oral Cavity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 737 Lip . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 738 Nasal Vestibule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 741 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 742 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 744
Introduction The American Cancer Society (ACS) estimates that cancers of the head and neck strike approximately 30,000 Americans per year, comprising slightly greater than 2% of all cancer incidences in the United States (ACS 2005). Head and neck cancers are often treated using brachytherapy, either alone or in combination with external beam radiation therapy. They include cancers of the oral cavity (base of tongue, oral tongue, floor of mouth, lip, and buccal mucosa), oropharynx, nasopharynx, and nose (nasal vestibule), as shown in Figure 1. Compared to external beam radiotherapy, brachytherapy delivers high tumor doses while allowing easy minimization of radiation doses to normal organs. Brachytherapy is not associated with the tissue deficits from surgery, thereby allowing improved maintenance of cosmetic and functional treatment outcomes. Brachytherapy also allows use of other radiation delivery techniques in the treatment of tumor recurrence or secondary malignancies, because of the minimal radiation doses to critical organs from brachytherapy treatment of the primary cancers.
Sources and Dosimetry Systems Used for Head and Neck Brachytherapy Brachytherapy for head and neck cancers most commonly uses the interstitial technique, with the intracavitary technique available for treatment of nasopharynx cancer. 137Cs needles replaced 226Ra needles for temporary interstitial implants. 192Ir sources, in the form of low activity 192Ir seeds in ribbons or high dose rate (HDR) remote afterloader sources, can be used for both interstitial and intracavitary techniques, while
727
728
Zuofeng Li
Figure 1. Head and neck anatomy. (Reprinted from Cox, J.D. (ed.), Moss’ Radiation Oncology: Rationale, Technique, Results, 7th edition, Figs. 8-6 and 8-8. © 1994, with permission of Mosby-Year Book.) 137
Cs tubes may be used for the intracavitary technique. The dosimetric characteristics of these sources have been subjects of many publications (Meisberger, Keller, and Shalek 1968; Krishnaswamy 1972; Williamson 1998b; Ballester et al. 1997, 2000, 2001; Casal et al. 2000; Perez-Calatayud 1999, 2001a,b,c, 2002; Williamson and Li 1995; Daskalov, Loffler, and Williamson 1998; Karaiskos et al. 2001, 2003; Angelopoulos et al. 2000).
Cesium Tubes and Needle Sources While these sources, shown in Figure 2, are currently unavailable from commercial vendors, they are part of the source inventory at many clinics. The works of Meisberger, Keller, and Shalek (1968) and Krishnaswamy (1972) have been traditionally adopted as the sources of dosimetry data for these sources. The AAPM Task Group 56 (TG 56) report (Nath et al. 1997) recommended that Sievert integral formalism may be used for the dosimetric calculations of these sources. Williamson (1998a) gave a detailed description of the Sievert integral equation, and indicated that a source encapsulation filtration coefficient may be approximately by a linear absorption coefficient of 0.023/mm for stainless-steel encapsulated cesium sources, with resulting maximum errors of less than 8%, and less than 3% errors near source transverse axis. Williamson (1998b) has since published Monte Carlo dose rate tables for the Amersham CDCS J-series and the 3M model 6500 137Cs tube sources. Perez-Catalayud and colleagues (Ballester et al. 2000; Casal et al. 2000; Perez-Calatayud et al. 2001a,b, 2002) have performed Monte Carlo calculations for many similar sources. Most of the existing data for the 137Cs sources, however, have been in the form of away-along tables or Sievert integral parameters. Many modern brachytherapy treatment-planning systems have since adopted the AAPM TG 43 recommended dose calculation formalism (Nath et al. 1995). Zhang et al. (Zhang, Beddar, and Sibata 2004) and Liu et al. (Liu, Prasad, and Bassano 2004) have performed conversions for typical 137Cs sources to obtain their dosimetric parameters in the TG 43 recommended formalism.
37–Review of Head and Neck Brachytherapy
729
Figure 2. Nycomed/Amersham CDCS A-type cesium needle and CDCS J-type tube sources. (Reprinted from Williamson, J. F. “Physics of Brachytherapy” in Principles and Practice of Radiation Oncology, 3rd Edition, C. A. Perez and L. W. Brady (eds.). © 1998, with permission from Lippincott-Raven.)
Low Dose-Rate (LDR) 192Ir Ribbons The AAPM TG 43 report provided a complete set of dosimetry parameters for the Best Medical International, Inc. (Springfield, VA) 192Ir seed. While the TG 43 report does not include the source dosimetry data for the Alpha-Omega Services, Inc, (Bellflower, CA) 192Ir seed, Ballester et al. (2004) have reported Monte Carlo calculated dosimetry data for this particular seed source.
730
Zuofeng Li
Remote Afterloading High Dose-Rate (HDR) and Pulsed Dose-Rate (PDR) Sources In the United States, Nucletron (Columbia, MD) and Varian Medical Systems (Charlottesville, VA) provide the Nucletron Classic and V2 units, the VariSource™ and GammaMed units, in both HDR and PDR flavors. Williamson and Li (1995) and Daskalov, Loffler, and Williamson (1998) performed Monte Carlo calculations for the Nucletron Classic HDR and PDR sources and Nucletron V2 HDR source, respectively, while Karaiskos et al. (2003) provide similar data for a new Nucletron PDR source. Angelopoulos et al. (2000) calculated the dosimetry parameters for the new Varian VariSource source. Ballester et al. (2001) and Perez-Catalayud et al. (2001c) presented corresponding data sets for the GammaMed 12i and Plus HDR and PDR sources, respectively.
Other Sources Used in Head and Neck Brachytherapy 192
Ir wires are commonly used for interstitial brachytherapy in Europe, but have been largely replaced by ribbons containing 192Ir seeds in the United States. Karaiskos et al. (2001), Ballester et al. (1997), and Perez-Calatayud et al. (1999) provided dosimetry data sets for 192Ir wires ranging in length from 0.5 cm to 12 cm. 198Au seeds have been used for head and neck brachytherapy (Horiuchi et al. 1991; Shimizutani et al. 1995), though there have been not recent dosimetry studies on this source. The Meisberger et al. (1968) publication remains the pre-eminent source dosimetry data for this source. 125I seeds have also been used for permanent as well as temporary head and neck brachytherapy (Vikram and Mishra 1994; Horwitz et al. 1997). Model-specific dosimetric parameters for the 125I seeds should be used for their dose calculations (Nath et al. 1995; Rivard et al. 2004).
Dosimetry Systems and Dose Prescription For head and neck LDR brachytherapy treatments, preplanning is in general performed. In the case of LDR interstitial treatments, the Manchester system (Williamson 1998a), the Quimby system (Williamson 1998a), and the Paris System (Pierquin and Marinello 1997) are often used for such preplanning. Depending on the volume to be treated, a single plane, double plane, or volume implant may be planned. Using the Manchester system, the needles/catheters are inserted at 1 cm to 1.5 cm apart. The prescription isodose line is typically chosen at 5 mm away from the implanted needles for planar implants. For volume implant, the Manchester system specifies prescription dose within the volume bounded by the implanted needles. The treating physician, based on the patient’s disease extent, may chose an isodose line a few millimeters outside this surface. Computerized preplanning may be performed to determine the required source strengths for the implant. The Manchester system recommends a dose rate of 45 cGy/h, though a choice of prescription dose rate either higher or lower than this recommended dose rate may be necessary based on the quality of implant and the target extent. Paterson (1952) recommended adjusting the total prescription dose for a given brachytherapy treatment based on the prescription dose rate, in order to deliver a biologically equivalent dose to the target. To deliver a treatment equivalent to 60 Gy in 7 days, Paterson proposed a correction curve ranging from 46 Gy in 3 days (64 cGy/h) to 62 Gy in 9 days (29 cGy/h). Pierquin and colleagues (Pierquin and Marinello 1997; Pierquin et al. 1973) reported that dose rates ranging from 30 cGy/h to 90 cGy/h had no effect on the rates of necrosis or tumor control for cancer of floor of the month and oral tongue, while Pernot et al. (1997) reported dose rates higher than 70 cGy/h resulted in significantly increased rates of soft tissue and bone necrosis. Parsons (1994a) reported that, at the University of Florida, no adjustment in total prescribed dose is recommended for prescription dose rates between 30 cGy/h and 80 cGy/h. These dose rates are achieved by using the “full-strength” cesium needles or iridium ribbons (0.66 mgRaEq/cm) for single plane implants, and “half-strength” cesium needles or iridium ribbons (0.33 mgRaEq/cm) for double plane implants. Volume implants use sources of still lower
37–Review of Head and Neck Brachytherapy
731
strengths, typically in the 0.22 mgRaEq/cm to 0.25 mgRaEq/cm range. Crossing needles at the ends of the implant should be used, whenever possible, to improve target coverage. While the simpler form of the Paris system can be found in standard text books, for example in Perez and Brady (1998), of particular relevance to head and neck brachytherapy is its adaptation for use of 192Ir hairpins and 192Ir wires/ribbons in a looping technique, shown in Figure 3, as described by Pierquin and Marinello (1997). The treated length is 0.8 times half the total active length; the treated thickness is 1.55 times the spacing between the branches of the hairpin or the loop; and the treated width is equal to the distance between the most distant loops or hairpins plus 0.5 times the spacing between the branches. Following the general Paris system, the prescribed dose rate is 0.85 times the average basal dose rate. The “hyperdose” region of the implants, i.e., the volume receiving 200% of the prescription dose, should be minimized. The Paris System limits the hyperdose volume by restricting the maximum needle/catheter spacing to less than or equal to 2.2 cm, though for head and neck cancers needle/catheter spacing larger than 1.4 cm has been correlated with increased rate of complications (Simon et al. 1993; Mazeron et al. 1990). The use of HDR or PDR remote afterloading techniques for head and neck brachytherapy allows improved flexibility in selecting source dwell positions and dwell times, which potentially results in better target coverage and reduced critical organ doses. While preplanning is not absolutely necessary for such treatments, the insertion of needles and catheters should follow the classical dosimetry systems outlined above, in terms of choosing between a single-plane implant vs. multiple-plane implant, and in selecting the needle/catheter spacing.
Target Delineation and Implant Techniques The delineation of a clinical target volume (CTV) for head and neck brachytherapy may be based on the physician’s clinical examination, implanted radiographic markers, and computed tomography (CT), magnetic resonance imaging (MRI), and ultrasound imaging. The planning target volume (PTV) is assumed to be the same as the CTV, as it is assumed that the implanted needles and catheters are fixed relative to the target. The CTV should include all palpable and/or visible tumor volumes, as determined
Figure 3. Paris System adapted for hairpins and loops. (Reprinted from Pierquin, B., and G. Marinello, A Practical Manual of Brachytherapy. © 1997, with permission from Medical Physics Publishing.)
732
Zuofeng Li
from radiographic or ultrasound images, plus an adequate margin, typically ranging from a few millimeters to 1 cm. It is important that the tumor volumes are determined based on pre-operative clinical examinations or images, for post-operative brachytherapy treatments, or pre-radiation treatment exams or images, for patients who have received external radiation therapy. It is therefore often necessary that radiopaque markers be implanted pre-surgery or pre-irradiation to facilitate the delineation of the CTV. Based on the tumor location and ease of access, the technique for inserting the needles and catheters may vary. Hollow stainless needles of 1.9 mm outer diameter (15 gauge) are often used for introducing the catheters, which typically have a thin lead at one end, with the other end ending either in a button with suturing holes or another thin lead, depending on the applications. In all cases, the locations of needle entrance and exit points should be measured using a sterile ruler and marked with a surgical pen to maintain accurate needle spacing. Radiopaque or CT-compatible buttons as appropriate for the catheter localization imaging technique (radiographs or CT scans) should be used to fasten the catheters to patient skin. Common interstitial implant techniques include the following:
Basic Technique As shown in Figure 4, the basic technique requires access from both sides of the target, such as for neck, buccal mucosa, or lip implants. A hollow stainless steel needle is inserted through the target. The thin leader end of a nylon catheter is threaded through the catheter. The catheter is then pulled through the target, together with the steel needle.
Hairpins and Looping Techniques The hairpin technique, shown in Figure 5, is used to deliver a high surface dose, in the presence of tumor extension to the mucosa or exophytic tumors. Gutter guides are inserted into the target, followed by the 192 Ir hairpins. The hairpins are then held down using a pair of forceps while the gutter guides are removed.
Figure 4. Basic interstitial brachytherapy catheter implant technique. (Reprinted from Hilaris, B., D. Nori, and L. L. Anderson, Atlas of Brachytherapy. © 1988, with permission from B. Hilaris.)
37–Review of Head and Neck Brachytherapy
733
In the United States the looping technique, shown in Figure 6, is commonly used to achieve comparable dose distributions. Two needles are inserted into the target through the same direction, followed by threading the leader of a catheter through one of the needles. The needle is pulled, together with the catheter, out of the target, leaving the catheter inside the target. The leader of the catheter is then threaded through the second needle and pulled out of the target together with it. A finger is placed in the loop to prevent kinking the catheters until the loop is pulled tight against the mucosa. Figure 7 shows an alternative loop-
Figure 5. Gutter guide and 192Ir hairpin insertion technique. (Reprinted from Hilaris, B., D. Nori, and L. L. Anderson, Atlas of Brachytherapy. © 1988, with permission from B. Hilaris.)
Figure 6. Looping technique. (Reprinted from Hilaris, B., D. Nori, and L. L. Anderson, Atlas of Brachytherapy. © 1988, with permission from B. Hilaris.)
734
Zuofeng Li
Figure 7. Looping technique using a wire. (Reprinted from Hilaris, B., D. Nori, and L. L. Anderson, Atlas of Brachytherapy. © 1988, with permission from B. Hilaris.)
Figure 8. Non-looping technique using a crossing catheter. (Reprinted from Hilaris, B., D. Nori, and L. L. Anderson, Atlas of Brachytherapy. © 1988, with permission from B. Hilaris.)
37–Review of Head and Neck Brachytherapy
735
ing technique, where a pair of needles is inserted into the target. A wire is then inserted through the needles. The needles are pulled out of the patient, leaving the wire inside. The catheter is then tied to the wire and pulled through the target.
Non-Looping Techniques For tumors deep inside the oral cavity, such as base of the tongue, the looping technique is often technically difficult to perform. Furthermore, for treatment using HDR or PDR units, the source cable may not be able to negotiate the loops. Vikram and Hilaris (1981) introduced a non-looping technique that uses a crossing catheter tied to the ends of a pair of catheters, shown in Figure 8. Nag et al. (1998) presented a non-looping technique, shown in Figure 9, where rubber sleeves are used to provide a distance between the ends of the catheters, and the catheters are tied together to form a loop. When a remote afterloader unit with a stepping source, such as an HDR or a PDR unit, is used to deliver brachytherapy in the oral cavity which requires adequate dose coverage of the mucosa, the dwell positions near the mucosa in each treatment catheter may be weighted more heavily so as to “push” the isodose
Figure 9. Non-looping technique of Nag et al. (1998). (a) A silk suture is tied to a button-ended catheter and a piece of 0.5 cm long rubber spacer is threaded over the button-ended catheter. (b) The two catheters are tied together at the dorsum, thus making a functioning loop. (Reprinted from (1998). Radiother Oncol 48, “Simplified non-looping functional loop technique for HDR brachytherapy,” Nag, S., R. Martinez-Monge, H. Zhang, and N. Gupta, pp. 339–341. © 1998, with permission from Elsevier.)
736
Zuofeng Li
distribution outside the tissue, thus achieving a dose distribution similar to those obtained from the hairpin, looping, or other non-looping techniques. Sethi et al. (1996) studied the relation between the dwell time weight at the three distal-most dwell positions with 2.5 mm step size, assumed to start at 1 mm below the mucosa, and the distance of prescription isodose line rises above the mucosa. For catheter spacing of 1 cm to 2 cm, inserted according to the Paris system, it was observed that the prescription isodose line rises above the mucosa by 1 to 2 mm if these three dwell positions receive 200% of the weight of the rest of the equally weighted dwell positions. The calculated dose homogeneity index values, defined as the ratio of volume receiving 150% prescription dose to volume receiving 100% prescription dose, were no different from those of the hairpin technique. It should be noted that this technique requires the catheters to rise above the mucosa by approximately 4 mm to allow the correct functioning of check source test for the HDR/PDR units, as the check source needs to travel 4 mm further than the distal-most dwell position.
Treatment Complications and Pretreatment Dental Evaluation Treatment Complications Soft tissue and bone necrosis (osteoradionecrosis) are common treatment complications associated with head and neck brachytherapy. Soft tissue necrosis is the destruction of the soft tissue that received a high radiation dose, and often appears as an ulcer. If the soft tissue necrosis overlies bony structures, such as the maxilla and the mandible, without obvious underlying bone necrosis, it is called “bone exposure.” Bone necrosis is due to the radiation damage to the cellular and vascular elements of the bone. Most soft tissue necrosis develops within 5 to 8 months after interstitial brachytherapy. Parsons (1994b) reported about 20% incidence rate of mild soft tissue necrosis in head and neck patients receiving combined external beam radiation therapy and brachytherapy. Similar incidence rate was reported for mild or moderately severe bone necrosis, which heals spontaneously in less than 6 months or more than 6 months, respectively. Severe bone necrosis requiring surgical intervention has been reported to occur in 1% to 6% of the patients (Mendenhall 2004). The rate of incidence of bone necrosis depends on the total dose received as well as, in the case of brachytherapy, the dose rate. Jereczek-Fossa and Orecchia (2002) indicated that the incidence rate increased steeply for total dose above 66 Gy. Fujita and colleagues (Fujita et al. 1996) reported a significantly increased rate of bone necrosis when a brachytherapy dose of 60 Gy delivered at higher than 55 cGy/h is combined with an external beam radiation dose of 30 Gy. The American Brachytherapy Society (ABS) (Nag et al. 2001) suggested use of spacers between the implanted volume and the neighboring bony structures, such as the mandible, to increase the distance of brachytherapy sources to the bony structures, in order to reduce the incidence of bone necrosis. Pernot et al. (1997) observed that the use of lead shield, as well as a treated surface area smaller than 12 cm2, were significantly correlated with reduced rates of soft tissue as well as bone necrosis in the treatment of cancers of the oral cavity and oropharynx. To prevent skin fibrosis, the skin dose in a head and neck brachytherapy treatment should be limited to less than the prescribed dose. If planar radiographs are used for treatment planning, radiopaque marker wires should be placed at the entrance and exit points of the treatment catheters to identify the skin on the films. Alternatively, radiopaque buttons may be used for this purpose, after assuring that they are pushed all the way to skin surface.
Dental Evaluation As recommended by the ABS (Nag et al. 2001), head and neck brachytherapy patients should receive a complete dental evaluation prior to the treatment. The incidence of bone exposure and necrosis increases
37–Review of Head and Neck Brachytherapy
737
with the overall poorer health of a patient’s teeth. Teeth in poor or questionable condition should be extracted, with 14 to 21 days allowed for healing, before start of treatment.
Clinical Sites Nasopharynx Levendag et al. (1997) discussed the design of the Rotterdam Nasopharynx applicator (RNA), shown in Figure 10, for delivery of HDR nasopharyx treatments, together with its dosimetric applications. The treatment protocol calls for delivery of 3 Gy/fx twice a day for 6 fractions after 60 Gy external beam dose for T1–T3 tumors, and 3 Gy/fx twice a day for 4 fractions after 70 Gy external beam dose for T4 tumors. Treatment planning is based on orthogonal radiographs, as shown in Figure 11. Target points include the nasopharynx (Na), the node of Rouviere (R), and the base of skull (BOS), while normal organ points for the optic chiasm (OC), the pituitary gland (P), the retina (Re), and the cord (C) are defined as well. An additional normal tissue point Pa is defined to represent the junction of the hard and soft palate. The dose distribution is optimized such that the nasophynx point Na, and if possible R, the node of Rouviere, receives a dose of 3 Gy, while the normal tissue dose points receive doses as low as possible. Table 1 shows examples of non-optimized and optimized dose to these points using this technique.
Oral Cavity Nag et al. (2001) provided a summary of HDR oral cavity treatment protocols for brachytherapy used either as sole treatment or boost treatment, shown in Table 2. The implant technique, depending on the tumor location, may be selected from among those discussed previously. Templates may be used to facilitate the accurate insertion of treatment catheters, as shown in Figures 12a and 12b and 13a and 13b. Mazeron et al. (1998) recommended that brachytherapy alone be used for T1–T2 squamous cell carcinomas of the oral cavity, using 192Ir LDR implants, to deliver a total dose of 65 to 70 Gy. It was further
Figure 10. HDR nasopharynx applicator. (Reprinted from Radiother Oncol 45, “A new applicator design for endocavitary brachytherapy of cancer in the nasopharynx,” Levendag, P. C., R. Peters, C. A. Meeuwis, L. L. Visch, D. Sipkema, C. de Pan, and P. I. Schmitz, pp. 95–98. © 1997, with permission from Elsevier.)
738
Zuofeng Li
Figure 11. Lateral (Lat) and anterior-posterior (AP) x-ray simulation films of an NPC patient. After placing lead markers on the (contralateral) outer canthus (1) and tragus (2), a reference line is drawn from (1) to (2); 1 cm posterior from (l), a normal tissue point for the retina (Re) is depicted. A second reference line is drawn between the anterior clinoid process (3) and the position of the node of Rouviere (R) (ventral part corpus C-I). The pituitary gland (P) is positioned 0.5 cm from (3) (center of sella) on the second reference line; the intersection of line (3-R) with the first reference line (l-2) represents the base of the skull target point (BOS). The optic chiasm (OC) is located on the skull base, 1.5 cm ventrally from (3). At the junction of the hard and soft palate, a normal tissue point (Pa) is indicated. The intersection of the Pa-BOS line with the bony outline of the base of the skull is taken as the target point for prescribing the dose to the nasopharynx (Na). Directly posterior to R, at the posterior border of corpus C-I, a normal tissue point for the cord (C) is indicated. The OC, P, C and R points are single midline patient points. BOS, Re, Na and Pa are bilateral patient points at 1.5, 2.5, 1.5 and 1 cm from the midline, respectively. The stipulated line on the lateral and AP X-ray films represents the 3 Gy (reference) isodose line. (Reprinted from Radiother Oncol 45, “A new applicator design for endocavitary brachytherapy of cancer in the nasopharynx,” Levendag, P. C., R. Peters, C. A. Meeuwis, L. L. Visch, D. Sipkema, C. de Pan, and P. I. Schmitz, pp. 95–98. © 1997, with permission from Elsevier.)
suggested that the dose rate should be limited to 0.3 to 0.6 cGy/h, with interneedle/catheter spacing of 1 to 1.4 cm, and that lead shielding of the mandible be consistently used.
Lip Due to its easy access, lip cancer can be treated often by a single plane implant, as shown in Figure 14. Crossing needles are often necessary. Million, Cassici, and Mancuso (1994) reported a combination of 30 Gy external beam dose and 35 Gy LDR brachytherapy dose, prescribed to 0.5 cm away from the source needles, for lip cancer. Guinot et al. (2003) used twice a day (BID) fractions of 5 to 5.5 Gy per fraction for 8 to 10 fractions of HDR brachytherapy sole treatment for lip cancer, shown in Figure 15, totaling 40.5 to 45 Gy, with the dose prescribed at the 90% isodose line following the application of the geometry optimization algorithm (Edmundson 1994). Lead shields are often placed between the lip and the gingiva to reduce the dose to the gingival mucosa.
37–Review of Head and Neck Brachytherapy
739
Table 1. Example of brachytherapy dose distribution, non-optimized and optimized, in target and normal tissue patient points for a single patient treated by HDR nasopharynx brachytherapy Patient Points
Non-optimized Dose (cGy) (%)
Optimized Dose (cGy) (%)
NA (R)
268 (89)
292 (97)
NA (L)
332 (111)
308 (103)
BOS (R)
146 (49)
146 (49)
BOS (L)
128 (43)
128 (43)
R
366 (122)
303 (101)
OC
45 (15)
46 (15)
P
69 (23)
70 (23)
Re (R)
48 (16)
52 (17)
Re (L)
60 (20)
61 (20)
Pa (R)
213 (71)
238 (79)
Pa (L)
253 (84)
248 (83)
C
90 (30)
78 (26)
(Reprinted from Radiother Oncol 45, “A new applicator design for endocavitary brachytherapy of cancer in the nasopharynx,” Levendag, P. C., R. Peters, C. A. Meeuwis, L. L. Visch, D. Sipkema, C. de Pan, and P. I. Schmitz, pp. 95–98. © 1997, with permission from Elsevier.)
Figure 12a. A template for oral tongue cancer implant with cesium needles mounted on the template. (Reprinted from Million, R. R., N. J. Cassisi, and A. A. Mancuso, “Oral Cavity” in Management of Head and Neck Cancer: A Multidisciplinary Approach, R. R. Million and N. J. Cassisi (eds.). © 1994, with permission from Lippincott.)
Figure 12b. Treatment of oral tongue cancer with two rows of needles mounted on templates. (Reprinted from Million, R. R., N. J. Cassisi, and A. A. Mancuso, “Oral Cavity” in Management of Head and Neck Cancer: A Multidisciplinary Approach, R. R. Million and N. J. Cassisi (eds.). © 1994, with permission from Lippincott.)
740
Zuofeng Li Table 2a. HDR brachytherapy as sole treatment for oral cavity cancers EBRT
Fx Size (Gy)
# fx
Equiv. dose (Gy)
# pts.
L.C.
0
3
20
65
3
—
0
6.5
7
63
27
53%
0
6
10
80
14
100%
0
4.5–5
10
54–63
13
90%
0
5.5–6
10
71–80
13
100%
Abbreviations: Fx = fractions; equiv. = equialent; Pts. = patients; L. C. = local control; EBRT = external beam radiation therapy. * Equivalent dose for tumor effects as if given at 2 Gy/day using the linear quadratic model with an α/β ratio of 10 (25). See appendix. Table 2b. HDR brachytherapy as boost treatment for oral cavity cancers EBRT dose (Gy)
HDR dose/fx (Gy)
# fx
Equiv. dose* (Gy)
# Pts.
L. C.
Survival
50
2.7
6
67
12
79%
45%
40–48
3
7
63–71
18
80%
Abbreviations: Fx = fractions; equiv. = equialent; Pts. = patients; L. C. = local control; HDR = high dose rate; EBRT = external beam radiation therapy. * Equivalent dose for tumor effects as if given at 2 Gy/day using the linear quadratic model with an α/β ratio of 10 (25). See appendix. (Tables 2a, and 2b modified and reprinted from Int J Radiat Oncol Biol Phys 50, “The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for head-and-neck carcinoma,” Nag, S., E. R. Cano, D. J. Demanes, A. A. Puthawala, and B. Vikram, pp. 1190–1198. © 2001, with permission from Elsevier.)
Figure 13a. Custom template for floor of mouth cancer brachytherapy with cesium needles mounted on template. (Reprinted from Million, R. R., N. J. Cassisi, and A. A. Mancuso, “Oral Cavity” in Management of Head and Neck Cancer: A Multidisciplinary Approach, R. R. Million and N. J. Cassisi (eds.). © 1994, with permission from Lippincott.)
37–Review of Head and Neck Brachytherapy
741
Figure 13b. Inserted floor of mouth template. (Reprinted from Million, R. R., N. J. Cassisi, and A. A. Mancuso, “Oral Cavity” in Management of Head and Neck Cancer: A Multidisciplinary Approach, R. R. Million and N. J. Cassisi (eds.). © 1994, with permission from Lippincott.)
Nasal Vestibule Parsons et al. (1994) summarized the treatment of nasal vestibule cancer using interstitial brachytherapy at the University of Florida, as shown in Figure 16. The implant consists of 2 to 4 planes of implanted needles, and computerized treatment planning is used to deliver a total dose of 55 to 75 Gy, depending on the size of the lesion. When combined with an external beam dose of 50 Gy, the brachytherapy dose is reduced to 20 to 30 Gy. As the implant does not follow the Manchester system, the prescribed dose rate may be significantly higher than 45 cGy/h. If only a small lesion in one of the nasal cavities is to be treated, a single wax mold, with a cesium source embedded in it, may be used.
742
Zuofeng Li
(a)
(b)
(c) Figure 14. LDR lip single plane implant with crossing needles; (a) implanted needles; (b) and (c) AP and lateral radiographs. (Reprinted from Million, R. R., N. J. Cassisi, and A. A. Mancuso, “Oral Cavity” in Management of Head and Neck Cancer: A Multidisciplinary Approach, R. R. Million and N. J. Cassisi (eds.). © 1994, with permission from Lippincott.)
Summary Brachytherapy is often used for treatment of head and neck cancers, either as a sole treatment or in combination with external beam treatments. Due to the anatomical limitations, the simple “through-and-through” implant technique may not be appropriate. Various implant techniques, together with their associated applicators and templates, have been developed for delivery of head and neck brachytherapy. While the fundamental dosimetric principles for head and heck brachytherapy do not differ from the interstitial brachytherapy treatments of other sites, modified dosimetry systems have been developed to accommodate the use of, for example, gutter pins or looped catheters, or to allow the implant of tumors in extremely
37–Review of Head and Neck Brachytherapy
743
Figure 15. An HDR single plane lip implant. (Reprinted from Radiother Oncol 69, “Lip cancer treatment with high dose rate brachytherapy,” Guinot, J. L., L. Arribas, M. L. Chust, J. L. Mengual, E. Garcia Miragall, M. Carrascosa, P. P. Escolar, V. Crispin, and C. Guardino, pp. 113–115. © 2003, with permission from Elsevier.)
Figure 16. Basic configuration of nasal vestibule implant used at the University of Florida. The needles near the nasal tip usually have a 2-cm active length, with the more cephalad needles becoming longer, growing up to 4.5 cm active length. (Reprinted from Million, R. R., N. J. Cassisi, and A. A. Mancuso, “Oral Cavity” in Management of Head and Neck Cancer: A Multidisciplinary Approach, R. R. Million and N. J. Cassisi (eds.). © 1994, with permission from Lippincott.)
744
Zuofeng Li
confined spaces such as the nasal vestibule. The incidence rates of soft tissue and bone necrosis can be high with the use of brachytherapy for head and neck cancers. These can be reduced by improving the dose distribution homogeneity and by use of spacers to reduce doses to the maxilla and the mandible.
References American Cancer Society (ACS). Cancer Facts and Figures 2005. Atlanta, GA: ACS, 2005. Angelopoulos, A., P. Baras, L. Sakelliou, P. Karaiskos, and P. Sandilos. (2000). “Monte Carlo dosimetry of a new 192 Ir high dose rate brachytherapy source.” Med Phys 27(11):2521–2527. Ballester, F., C. Hernandez, J. Perez-Calatayud, and F. Lliso. (1997). “Monte Carlo calculation of dose rate distributions around 192Ir wires.” Med Phys 24(8):1221–1228. Ballester, F., J. L. Lluch, Y. Limami, M. A. Serrano, E. Casal, J. Perez-Calatayud, and F. Lliso. (2000). “A Monte Carlo investigation of the dosimetric characteristics of the CSM11 137Cs source from CIS.” Med Phys 27(9):2182–2189. Ballester, F., V. Puchades, I. L. Lluch, M. A. Serrano-Andres, Y. Limami, J. Perez-Calatayud, and E. Casal. (2001). “Monte-Carlo dosimetry of the HDR 12i and Plus 192Ir sources.” Med Phys 28(12):2586–2591. Ballester, F., D. Granero, J. Perez-Calatayud, E. Casal, and V. Puchades. (2004). “Monte carlo dosimetric study of best industries and Alpha Omega Ir-192 brachytherapy seeds.” Med Phys 31(12):3298–3305. Casal, E., F. Ballester, J. L. Lluch, J. Perez-Calatayud, and F. Lliso. (2000). “Monte Carlo calculations of dose rate distributions around the Amersham CDCS-M-type 137Cs source. Med Phys 27(1):132–140. Cox, J.D. (ed.). Moss’Radiation Oncology: Rationale, Technique, Results, 7th edition. St. Louis: Mosby-Year Book, 1994. Daskalov, G. M., E. Loffler, and J. F. Williamson. (1998). “Monte Carlo-aided dosimetry of a new high dose-rate brachytherapy source.” Med Phys 25(11):2200–2208. Edmundson, G. K. “Volume Optimization: An American Viewpoint” in Brachytherapy: From Radium to Optimization. R. J. Mould and J. Batterman (eds.). Columbia, MD: Nucletron, pp. 314-318, 1994. Fujita, M.,Y. Hirokawa, K. Kashiwado,Y. Akagi, K. Kashimoto, H. Kiriu, K. Ohtani, and T. Wada. (1996). “An analysis of mandibular bone complications in radiotherapy for T1 and T2 carcinoma of the oral tongue.” Int J Radiat Oncol Biol Phys 24:333–339. Guinot, J. L., L. Arribas, M. L. Chust, J. L. Mengual, E. Garcia Miragall, M. Carrascosa, P. P. Escolar, V. Crispin, and C. Guardino. (2003). “Lip cancer treatment with high dose rate brachytherapy.” Radiother Oncol 69(1):113–115. Hilaris, B., D. Nori, and L. L. Anderson. Atlas of Brachytherapy. New York: Macmillan Publishing Company, 1988. Horiuchi, J., M. Takeda, H. Shibuya, S. Matsumoto, M. Hoshina, and S. Suzuki. (1991). “Usefulness of 198Au grain implants in the treatment of oral and oropharyngeal cancer.” Radiother Oncol 21(1):29–38. Horwitz, E. M., A. J. Frazier, F. A. Vicini, D. H. Clarke, G. K. Edmundson, R. D. Keidan, G. S. Gustafson, C. F. Dmuchowski, and A. A. Martinez. (1997). “The impact of temporary iodine-125 interstitial implant boost in the primary management of squamous cell carcinoma of the oropharynx.” Head Neck 19(3):219–226. Jereczek-Fossa, B. A., and R. Orecchia. (2002). “Radiotherapy-induced mandibular bone complications.” Cancer Treat Rev 28(1):65–74. Karaiskos, P., P. Papagiannis, A. Angelopoulos, L. Sakelliou, D. Baltas, P. Sandilos, and L. Vlachos. (2001). “Dosimetry of 192Ir wires for LDR interstitial brachytherapy following the AAPM TG-43 dosimetric formalism.” Med Phys 28(2):156–166. Karaiskos, P., A. Angelopoulos, E. Pantelis, P. Papagiannis, L. Sakelliou, E. Kouwenhoven, and D. Baltas. (2003). “Monte Carlo dosimetry of a new 192Ir pulsed dose rate brachytherapy source.” Med Phys 30(1):9–16. Krishnaswamy, V. (1972). “Dose distribution about 137Cs sources in tissue.” Radiology 105:181–184. Levendag, P. C., R. Peters, C. A. Meeuwis, L. L. Visch, D. Sipkema, C. de Pan, and P. I. Schmitz. (1997). “A new applicator design for endocavitary brachytherapy of cancer in the nasopharynx.” Radiother Oncol 45(1):95–98. Liu, L., S. C. Prasad, and D. A. Bassano. (2004). “Determination of 137Cs dosimetry parameters according to the AAPM TG-43 formalism.” Med Phys 31(3):477–483.
37–Review of Head and Neck Brachytherapy
745
Mazeron, J. J., J. M. Crook, G. Marinello, W. Walop, and B. Pierquin. (1990). “Prognostic factors of local outcome for T1, T2 carcinomas of oral tongue treated by iridium-192 implantation.” Int J Radiat Oncol Biol Phys 19(2):281–285. Mazeron, J. J., A. Gerbaulet, J. M. Simon, and C. Hardiman. (1998). “How to optimize therapeutic ratio in brachytherapy of head and neck squamous cell carcinoma?” Acta Oncol 37:583–591. Meisberger, L .L., R. J. Keller, and R. J. Shalek. (1968) “The effective attenuation in water of the gamma rays of gold 198, iridium 192, cesium 137, radium 226, and cobalt 60.” Radiology 90:953–957. Mendenhall, W. M. (2004). “Mandibular osteoradionecrosis.” J Clin Oncol 22(24):4867–4868. Million, R. R., N. J. Cassisi, and A. A. Mancuso. “Oral Cavity” in Management of Head and Neck Cancer: A Multidisciplinary Approach. R. R. Million and N. J. Cassisi (eds.). Philadelphia: Lippincott, 1994. Nag, S., R. Martinez-Monge, H. Zhang, and N. Gupta. (1998). “Simplified non-looping functional loop technique for HDR brachytherapy.” Radiother Oncol 48:339–341. Nag, S., E. R. Cano, D. J. Demanes, A. A. Puthawala, and B. Vikram. (2001). “The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for head-and-neck carcinoma.” Int J Radiat Oncol Biol Phys 50(5):1190–1198. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Task Group No. 43.” Med Phys 22(2):209–234. Also available as AAPM Report No. 51. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24(10):1557–1598. Also available as AAPM Report No. 59. Parsons, J. T. “Time-Dose-Volume Relations in Radiation Therapy” in Management of Head and Neck Cancer: A Multidisciplinary Approach. R. R. Million and N. J. Cassisi (eds.). Philadelphia: Lippincott, 1994a. Parsons, J. T. “The Effect of Radiation on Normal Tissues of the Head and Neck” in Management of Head and Neck Cancer: A Multidisciplinary Approach. R. R. Million and N. J. Cassisi (eds.). Philadelphia: Lippincott, 1994b. Parsons, J. T., S. P. Stringer, A. A. Mancuso, and R. R. Million. “Nasal Vestibule, Nasal Cavity, and Paranasal Sinuses” in Management of Head and Neck Cancer: A Multidisciplinary Approach. R. R. Million and N. J. Cassisi (eds.). Philadelphia: Lippincott, 1994. Paterson, R. (1952) “Studies in optimum dosage: the Mackenzie Davidson Memorial Lecture.” Br J Radiol 25:505–516. Perez, C. A., and L. W. Brady (eds.). Principles and Practice of Radiation Oncology, 3rd Edition. Philadelphia: Lippincott-Raven, 1998. Perez-Calatayud, J., F. Lliso, V. Carmona, F. Ballester, and C. Hernandez. (1999). “Monte Carlo calculation of dose rate distributions around 0.5 and 0.6 mm in diameter 192Ir wires.” Med Phys 26(3):395–401. Perez-Calatayud, J., F. Ballester, J. L. Lluch, M. A. Serrano-Andres, E. Casal, V. Puchades, and Y. Limami. (2001a). “Monte Carlo calculation of dose rate distributions around the Walstam CDC.K-type 137Cs sources.” Phys Med Biol 46(7):2029–2040. Perez-Calatayud, J., F. Lliso, F. Ballester, M. A. Serrano, J. L. Lluch,Y. Limami, V. Puchades, and E. Casal. (2001b). “A Monte Carlo study of dose rate distribution around the specially asymmetric CSM3-a 137Cs source.” Phys Med Biol 46(7):N169–174. Perez-Calatayud, J., F. Ballester, M. A. Serrano-Andres, V. Puchades, J.L. Lluch, Y. Limami, and E. Casal. (2001c). “Dosimetry characteristics of the Plus and 12i GammaMed PDR 192Ir sources.” Med Phys 28(12):2576–2585. Perez-Calatayud, J., F. Ballester, M. A. Serrano-Andres, J. L. Lluch, V. Puchades, Y. Limami, and E. Casal. (2002). “Dosimetric characteristics of the CDC-type miniature cylindrical 137Cs brachytherapy sources.” Med Phys 29(4):538–543. Pernot, M., E. Luporsi, S. Hoffstetter, D. Peiffert, P. Aletti, C. Marchal, P. Kozminski, A. Noel, and P. Bey. (1997). “Complications following definitive irradiation for cancers of the oral cavity and the oropharynx (in a series of 1134 patients).” Int J Radiat Oncol Biol Phys 37(3):577–585. Pierquin, B., and G. Marinello. A Practical Manual of Brachytherapy. Madison, WI: Medical Physics Publishing, 1997. Pierquin, B., D. Chassagne, F. Baillet, and C. Paine. (1973). “Clinical observations on the time factor in interstitial radiotherapy using iridium-192.” Clin Radiol 24:506–509.
746
Zuofeng Li
Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. (2004). “Update of AAPM Task Group No. 43 report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31(3):633–674. Also available as AAPM Report No. 84. Sethi, T., D. Ash, A. Flynn, and G. Workman. (1996). “Replacement of hairpin and loop implants by optimized straight line sources.” Radiother Oncol 39:117–121. Shimizutani, K., Y. Koseki, T. Inoue, T. Teshima, S. Furukawa, K. Kubo, H. Fuchihata, N. Masaki, H. Ikeda, and Y. Tanaka. (1995). “Application of 198Au grains for carcinoma of oral cavity.” Strahlenther Onkol 171(1):29–34. Simon, J. M., J. J. Mazeron, S. Pohar, C. Le Pechoux, J. M. Crook, L. Grimard, P. Piedbois, J. P. Le Bourgeois, and B. Pierquin. (1993). “Effect of intersource spacing on local control and complications in brachytherapy of mobile tongue and floor of mouth.” Radiother Oncol 26(1):19–25. Vikram, B., and B. S. Hilaris. (1981). “A non-looping afterloading technique for interstitial implants at the base of the tongue.” Int J Radiat Oncol Biol Phys 7:419–422. Vikram, B., and S. Mishra. (1994). “Permanent iodine-125 implants in postoperative radiotherapy for head and neck cancer with positive surgical margins.” Head Neck 16(2):155–157. Williamson, J. F. “Physics of Brachytherapy” in Principles and Practice of Radiation Oncology, 3rd Edition. C. A. Perez and L. W. Brady (eds.). Philadelphia: Lippincott-Raven, 1998a. Williamson, J. F. (1998b). “Monte Carlo-based dose-rate tables for the Amersham CDCS.J and 3M model 3600 137Cs tubes.” Int J Radiat Oncol Biol Phys 41(4):959–970. Williamson, J. F., and Z. Li. (1995). “Monte Carlo aided dosimetry of the microselectron pulsed and high dose-rate 192 Ir sources.” Med Phys 22(6):809–819. Zhang, P., A. S. Beddar, and C. H. Sibata. (2004). “AAPM TG-43 formalism for brachytherapy dose calculation of a 137Cs tube source.” Med Phys 31(4):755–759.
Chapter 38
History and Rationale for Accelerated Partial Breast Irradiation (APBI) Zoubir Ouhib, M.S. Lynn Regional Cancer Center of Boca Raton Community Hospital Boca Raton, Florida History of APBI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747 Rationale for APBI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 747 Patient Selection Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748 MammoSite® Device Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748 Monotherapy Eligibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748 Boost Therapy Eligibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749 Interstitial Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749 Monotherapy Eligibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749 Boost Therapy Eligibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749 Clinical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 750 Interstitial Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 750 MammoSite Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 750 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 752
History of APBI Over the past two decades, the combination of breast-conserving therapy (BCT) and radiotherapy has become an attractive and important alternative to mastectomy for in situ (DCIS) stages I and II breast cancers. A standard course of whole-breast external beam radiotherapy (EBRT) followed by a boost to the tumor bed generally requires 5 to 7 weeks to complete. For some patients, this time frame could be a logistical problem, particularly for working women, elderly patients, and those who live at a significant distance from a radiation therapy center. Additionally, EBRT must be integrated with systemic adjuvant chemotherapy, potentially creating a delay in the initiation of systemic therapy. Patients and their physicians might identify this as another barrier to the widespread acceptance of breast conservation.
Rationale for APBI The importance of breast irradiation to achieve local tumor control has been demonstrated by the National Surgical Adjuvant Breast Project (NSABP) protocol B-06 trial. Among the results derived from this protocol were the risk of tumor recurrence in the ipsilateral breast was decreased by 20% to 30% when EBRT followed lumpectomy alone (Kuske 2003). Other studies (Holland et al., 1990) have also confirmed these findings. EBRT to whole breast after segmental mastectomy (SM) is postulated to reduce local breast recurrence by eliminating residual cancer at the surgical site as well as eliminating cancer in remote areas of the breast. Studies of the pathologic anatomy of breast cancer suggest that remote foci of cancer exist in 26% to 36% of patients (Rosen et al., 1975). However, the relevance of these foci to ipsilateral breast failure rates after BCT was debatable as 65% to 100% of recurrences develop in the same quadrant as the initial tumor (Wazer et al., 2002). The proportion of recurrences near the SM site varies with the extent of surgical resec-
747
748
Zoubir Ouhib
tion and the amount of radiation therapy given, but in all cases no more than 3.8% of patients develop a recurrence in a remote location in the breast (Fisher et al., 1989; Clarke et al., 1994). These findings suggested that the major benefit of EBRT is derived from the radiation directed at the breast tissue immediately surrounding the SM site. This has led several investigators to question whether radiation therapy must be administrated to the entire breast. Furthermore, in trials of BCT in which EBRT was omitted from the treatment plan, recurrence at or near the SM site has been much more common than remote in-breast recurrence suggesting that multicentricity may not be a major cause of ipsilateral breast recurrence. These findings have led to the idea of delivering a relatively high dose of radiation to the “high risk” tissue immediately surrounding the SM site while effort is being made to reduce the dose to the surrounding normal tissue. Thus, interstitial breast brachytherapy became an obvious choice. With the clinically proven local-control rates of this method, a new technique was then introduced in APBI with the MammoSite® device from Proxima Therapeutics (Alpharetta, GA). This is a simple method that can provide accelerated brachytherapy at institutions that do not perform interstitial brachytherapy.
Patient Selection Criteria MammoSite® Device Cases Monotherapy Eligibility Patient selection criteria must meet the American Brachytherapy Society (ABS) and the American Society of Breast Surgeons (ASBS) criteria (Table 1). They include the following parameters: • Age ≥ 45 years. • Tumor size < 2 cm. • Invasive ductal histology • Negative nodal status • Negative marginal status • Applicator placement within 10 weeks of final lumpectomy procedure • Cavity post-lumpectomy with one dimension of at least 3.0 cm Ineligibility criteria include the following: • Patient-related Factors • Extensive intraductal component Table 1. Patient Selection Criteria Parameters
ABS Recommendations
ASBS Recommendations
Age
≥45
≥50
Diagnosis
Unifocal, invasive ductal carcinoma
Invasive ductal carcinoma or DCIS
Tumor size
≤3 cm
≤2 cm
Surgical margins
Negative microscopic surgical margins
Negative microscopic surgical margins >2 mm in all directions
Nodal status
N0
N0
38–History and Rationale for Accelerated Partial Breast Irradiation (APBI)
749
• Pure intraductal cancer • Lobular histology • Collagen vascular disease • Technical Factors • Excessive cavity size • Inadequate balloon-skin distance • Poor balloon-cavity conformance • Poor symmetry Boost Eligibility Any patient deemed appropriate for conventional boost therapy would be an acceptable MammoSite boost candidate. They have to meet all technical requirements stated above.
Interstitial Cases Monotherapy Eligibility Eligibility criteria are as follows: • Age > 40 years • Maximum tumor dimension ≤ 4.0 cm • Infiltrating ductal carcinoma or histologic variant (tubular, medullary, mucinous/colloid) • Negative surgical margins by at least 2 mm • Axillary lymph nodes dissection of levels I and II with up to three involved nodes • Implant must be performed within 8 weeks of the breast surgery Ineligibility criteria include the following: • Tumor histology with invasive or in situ lobular carcinoma or pure ductal carcinoma in situ • Skin involvement • Breast unsatisfactory for brachytherapy • Last breast surgery more than 8 weeks prior to planned brachytherapy Boost Therapy Eligibility Like the MammoSite cases, any patient qualified for conventional boost therapy would be considered for interstitial brachytherapy. They must meet some of the criteria stated above such as size of tumor, skin involvement, and time of surgery.
750
Zoubir Ouhib
Clinical Results Interstitial Brachytherapy Several groups have explored restricting radiotherapy to the tumor bed alone after lumpectomy (Vicini et al., 1990). Table 2 lists the major trials and their results. The Oschner Clinic has treated 52 patients with either low dose rate (LDR) or high dose rate (HDR) brachytherapy. The LDR patients received 45 Gy delivered over 3.5 to 6 days, whereas HDR patients received 32 Gy in 8 fractions over 4 days. After a median follow-up of 20 months, no patient had demonstrated a local recurrence, and the cosmetic results were good to excellent in 67% to 78%. The Royal Devon and Exeter Hospital used three different fractionation schemes in 45 patients (Clark et al., 1996). Here, 22 patients received 20 Gy in 2 fractions, 17 patients received 28 Gy in 4 fractions, and 8 patients received 32 Gy in 6 fractions. All patients that were treated underwent HDR implant after lumpectomy with negative or close margins. With a median follow-up of 18 months, the local recurrence rate was 8.8%. The London Regional Cancer Center, London, Ontario, treated 39 patients with clinical stage T1-2 breast cancer with HDR brachytherapy. Patients received a total dose of 37.2 Gy in 10 fractions administered twice daily (Perera et al., 1995). A median follow-up of 20 months showed a local recurrence of 2.6%. William Beaumont Hospital published their experience (Vicini et al., 2001) with 192Ir seed LDR brachytherapy in 50 patients. A minimum dose of 49.92 Gy was delivered over 4.0 days to the tumor bed plus a 1 to 2 cm margin using 125I seeds. A follow-up of 47 months showed no local failures, and the cosmetic outcome was good to excellent in 98%. They were two published studies that reported poor local control rates following limited-field radiotherapy. The first one was at Guy’s Hospital (Fentiman et al., 1991) where 27 patients were treated with LDR using 192Ir wires. The dose delivered was 55 Gy in 5.5 days to the tumor bed plus a margin of 2 cm. After a median follow-up of 72 months, 37% of the patients had failed in the treated breast. The second study was at Christie Hospital (Ribeiro et al., 1990; Magee et al., 1996) where a prospective randomized trial was conducted between 1982 and 1987 comparing post lumpectomy radiotherapy (RT) directed only to the tumor bed using electron field with whole-breast irradiation. All patients less than 70 years old with a ≤4 cm breast cancer and clinically negative axillary nodes were eligible, including those with invasive lobular histology, extensive intraductal component (EIC), and lymphovascular invasion. Surgery included gross tumor removal without microscopic margin assessment and no axillary lymph node dissection. The whole breast RT group also received radiation to the axillary and supraclavicular regions. An excess number of ipsilateral breast and nodal recurrences occurred in the limited-field RT group (8 years actuarial rate of 25% to 13%). This difference was pronounced for those with invasive lobular histology and lymphovascular space invasion. These poor results at Christie Hospital were associated with a lack of careful patient selection and inadequate surgical technique. Only gross tumor removal was performed and no attempt was made to achieve a wide excision or to assess margin status microscopically.
MammoSite Results Long-term follow-up data on these patients is still being collected for analysis. The short-term results indicate good to excellent results. Some parameters associated with the inherent dosimetry of the system indicate that the results should be as good as or perhaps better than interstitial brachytherapy. Among those is the D90 coverage. The coverage of the margins of the resection cavity was dramatically improved when analyzed in terms of the D90 coverage. The mean values were 90% [standard deviation (SD) 0.5, range: 89.2 to 90.8] for MammoSite and 69.8% (SD 7.3, range: 61.1 to 83.5) for interstitial brachytherapy. These values seem to
38–History and Rationale for Accelerated Partial Breast Irradiation (APBI)
751
Table 2. Breast-Conserving Therapy with Lumpectomy Plus Brachytherapy Alone Institutions
No. of Cases
Follow-up (Months)
Schema Gy × Fx
Total Dose (Gy)
% Local % Recurrence Good/Excellent Cosmetic Result
26
75
4.0 × 8
32
2
75
45
18
10 × 2
200
8.8
95
39
20
3.72 × 10
37.2
2.6
Not stated
45
57
5.20 × 7 4.33 × 7
36.4 30.3
4.4
97
126
30
5.2 × 7 (HDR)
36.4
0
Not stated
41
17
5.2 × 7
36.4
2.4
Not stated
54
36
4.0 × 8
32
0
100
66
32
3.4 × 10
340
—
—
32
33
3.4 × 10
34
3
88
26
75
>0.40 Gy/h
45
2
75
27
72
0.40 Gy/h
55
37
83
90
27
Not stated
50–60
4.4
Not stated
25
47
—
20–25
0
100
HDR Series Oschner Clinic, New Orleans, LA (King et al. 2000) Royal Devon/ Exeter England (Perera et al. 1995) London Regional Cancer Center London, Ontario (Perera et al. 1995) National Institute of Oncology, Budapest, Hungary Phase I/II Trial (Polgar et al. 2002) National Institute of Oncology, Budapest, Hungary Phase I/II Trial (Polgar et al. 2002) Orszagos Onkologiai Intezet, Budapest Hungary (Polgar et al. 2002) William Beaumont Hospital (Vicini et al. 2001) Radiation Therapy Oncology Group (Kuske 2003) Tufts-New England Medical Cent. (Wazer et al. 2002) LDR Series Oschner Clinic (King et al. 2000) Guy’s Hospital (Fentiman et al. 1991) Florence, Italy (Cionini et al. 1993) University of Kansas (Krishnan et al. 1992)
continued
752
Zoubir Ouhib Table 2. Breast-Conserving Therapy with Lumpectomy Plus Brachytherapy Alone (continued)
William Beaumont Hospital (Vicini et al. 1990) Radiation Therapy Oncology Group (Kuske 2003)
120
36
0.52 Gy/h
49.92
0
98
33
36
—
45
—
—
353
96
5.0 × 8
40
2.5
—
126
30
2.0 × 25
50
0
Not stated
External Beam Series Christie Hospital (Ribeiro et al. 1990) National Institute of Oncology, Budapest, Hungary Phase III trial (Polgar et al. 2002)
indicate an improvement both in coverage and in reproducibility (Edmundson et al. 2002). However the MammoSite device resulted in a treatment dose that was less uniform, with a mean dose homogeneity index (DHI) of 0.77 vs. 0.93. One of the issues with the use of the MammoSite balloon catheter that continues to be unresolved is the timing of the device placement. To assess its appropriateness, a review of all patients operated upon by two users of this modality between July 2001 and November 2002 was performed (Zannis, Walker, and Barclay-White 2003). Of the 343 patients reviewed, 137 (40%) were found to be intraoperative candidates for the placement of the balloon catheter. Their eligibility requirements were met with the information known in the operating room. However, postoperative final histology disqualified 40 of these 137 (29%). These findings would have caused the premature removal of the device in these 40 patients. Thus, patients receiving postoperative MammoSite placement should be considered after final pathology results are available. This practice will minimize the disappointment to these patients, and diminish wasted catheters.
Conclusion APBI using interstitial implant or the MammoSite device to deliver radiation to the tumor bed alone for carefully selected breast cancer patients over 4 to 5 days offers an attractive and viable alternative to conventional whole-breast irradiation. Both modalities of brachytherapy seem to produce preliminary results equivalent to those achieved with conventional EBRT. Extended follow-up is needed to determine the longterm efficacy of these two treatment modalities and the potential advantages and disadvantages of APBI.
References Cionini, L., P. Pacini, S. Marzano, et al. (1993). “Exclusive brachytherapy after conservative surgery in cancer of the breast.” (Abstract). Lyon Chir 89:128. Clark, R. M., T. Whelan, M. Levine, R. Roberts, A. Willan, P. McCulloch, M. Lipa, R. H. Wilkinson, and L. J. Mahoney. (1996). “Randomized clinical trial of breast irradiation following lumpectomy and axillary dissection for node-negative breast cancer: An update. Ontario Clinical Oncology Group.” J Natl Cancer Inst 88:1659–1664.
38–History and Rationale for Accelerated Partial Breast Irradiation (APBI)
753
Clarke, D. H., F. A. Vicini, and H. Jacobs. “High Dose Rate Brachytherapy for Breast Cancer” in High Dose Rate Brachytherapy: A Textbook. S. Nag (ed.). Armonk, NY: Futura Publishing, pp. 321–329, 1994. Edmundson, G. K., F. A. Vicini, P. Y. Chen, C. Mitchell, and A. A. Martinez. (2002). “Dosimetric characteristics of the MammoSite RTS, a new breast brachytherapy applicator.” Int J Radiat Oncol Biol Phys 52(4):1132–1139. Fentiman, I. S., C. Poole, D. Tong, P. J. Wintr, H. M. Mayles, P. Turner, M. A. Chaudary, and R. D. Rubens. (1991). “Iridium implant treatment without external radiotherapy for operable breast cancer; a pilot study.” Eur J Cancer 27:447–450. Fisher, B., C. Redmond, R. Poisson, R. Margolese, N. Wolmark, L. Wickerham, E. Fisher, M. Deutsch, R. Caplan, Y. Pilch, et al. (1989). “Eight-year results of a randomized clinical trial comparing total mastectomy and lumpectomy with or without radiation in the treatment of breast cancer.” N Engl J Med 320:822–828. Holland, R., J. L. Connolly, R. Gelman, M. Mravunac, J. H. Hendriks, A. L., Verbeek, S. J. Schnitt, B. Silver, J. Boyages, and J. R. Harris. (1990). “The presence of an extensive intraductal component following a limited excision correlates with prominent residual disease in the remainder of the breast.” J Clin Oncol 8:113–118. King, T.A., J. S. Bolton, R. R. Kuske, G. M. Fuhrman, T. G. Scroggins, and X. Z. Jiang. (2000). “Long-term results of wide-field brachytherapy as the sole method of radiation therapy after segmental mastectomy for Tis,1,2 breast cancer.” Am J Surg 180:299–304. Kuske, R. R. American Brachytherapy Society. School of breast brachytherapy. January 17–19, 2003, Las Vegas, Nevada. Krishnan, L., W. R. Jewell, E. C. Krishnan, R. Cherian, and F. Lin. (1992). “Breast cancer with extensive intraductal component: Treatment with immediate interstitial boost irradiation.” Radiology 183:273–276. Magee, B., R. Swindell, M. Harris, and S. S. Banerjee. (1996). “Prognostic factors for breast recurrence after conservative breast surgery and radiotherapy: Results from a randomised trial.” Radiother Oncol 39:223–227. Perera, F., F. Chisela, J. Engel, and V. Ventkatesan. (1995). “Method of localization and implantation of the lumpectomy site for high dose rate brachytherapy after conservative surgery for T1 and T2 breast cancer.” Int J Radiat Oncol Biol Phys 31:959–965. Polgar, C., Z. Sulyok, J. Fodor, Z. Orosz, T. Major, Z. Takacsi-Nagy, L. C. Mangel, A. Somogyi, M. Kasler, and G. Nemeth. (2002). “Sole brachytherapy of the tumor bed after conservative surgery for T1 breast cancer: Fiveyear results of a phase I-II study and initial findings of a randomised phase III trial.” J Surg Oncol 80:121–128. Ribeiro, G. G., G. Dunn, R. Swindell, M. Harris, and S. S. Banerjee. (1990). “Conservation of the breast using two different radiotherapy techniques: Interim report of a clinical trial.” Clin Oncol 2:27–34. Rosen, P. P., A. A. Fracchia, J. A. Urban, D. Schottenfeld, and G. F. Robbins. (1975). “‘Residual’ mammory carcinoma following simulated partial mastectomy.” Cancer 35:739–747. Vicini, F. A., V. R. Kini, P. Y. Chen, E. Horwitz, G. Gustafson, P. Benitez, G. Edmundson, N. Goldstein, K. McCarthy, and A. Martinez. (1999). “Irradiation of tumor bed alone after lumpectomy in selected patients with early-stage breast cancer treated with breast-conserving therapy.” J Surg Oncol 70:33–40. Vicini, F. A., K. L. Baglan, L. L. Kestin, C. Mitchell, P. Y. Chen, R. C. Frazier, G. Edmundson, N. S. Goldstein, P. Benitez, and R. R. Huang. (2001). “Accelerated treatment of breast cancer.” J Clin Oncol 19(7):1993–2001. Wazer, D. E., L. Berle, R. Graham, M. Chung, J. Rothschold, T. Graves, B. Cady, K. Ulin, R. Ruthazer, and T. A. DiPetrillo. (2002). “Preliminary results of a phase I/II study of HDR brachytherapy alone for T1/T2 breast cancer.” Int J Radiat Oncol Phys 53(4):889–897. Zannis, V. J., L. C. Walker, B. Barclay-White, and C. A. Quiet. (2003). “Postoperative ultrasound-guided percutaneous placement of a new breast brachytherapy balloon catheter.” Am J Surg 186:383–385.
Chapter 39
Breast Interstitial Implant and Treatment Planning Rupak K. Das, Ph.D. and Rakesh Patel, M.D. Department of Human Oncology University of Wisconsin, Madison, Wisconsin Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 755 Rationale and Selection Criteria for Partial Breast Irradiation . . . . . . . . . . . . . . . . . . . . . . . . . 756 Treatment Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 Low Dose-Rate APBI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 High Dose-Rate APBI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 Interstitial Volume Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 Template-Guided Implant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 757 Freehand Implant: Ultrasound Guided Supine Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758 Comparison of Implant Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758 2-D Treatment-Planning Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758 Imaging and Catheter Reconstruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 758 Planning Target Volume (PTV) Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 759 Dose Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 759 3-D Treatment-Planning Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 759 Imaging and Catheter Reconstruction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 759 PTV Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760 Dose Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 760 Evaluation of Treatment Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 761 Quality Assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 763 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 765
Introduction The treatment of breast cancer using radium needles started only a few years after Madam Curie discovered radium in 1896. By 1915, Janeway was treating primary breast cancer using interstitial radium needles (Shank 1986). The first published series of breast conservation therapy using radium needles was reported by Geoffrey Keynes in 1929 (Keynes 1929). In this study Keynes placed radium needles within the entire breast, internal mammary, supraclavicular, infraclavicular, and axillary lymph nodes. In this study he reported high local and regional control rates. Comparison of survival rates was found to be as favorable as the Halstead radical mastectomy, which was the common practice during that time. Based on this experience, he concluded that the insertion of radium needles was the preferred treatment for breast cancer, since it produced the same survival rates as that of mastectomy, while preserving the breast. Even with Keynes’ success in interstitial breast brachytherapy, breast cancer was treated predominantly by mastectomy for the next 50 years. In the 1970s, a new concept of surgical removal of the gross tumor within the breast along with axillary lymph nodes, followed by whole breast radiation and supplemental boost dose by 192Ir became the standard practice. With the availability of electron beam from linear accelerators in the 1980s, the brachytherapy boost was replaced by electron boost. A wide range of electrons energies gave the radiation oncologists the choice of selecting a beam whose practical range is less than the depth to the lung or heart and since the treatment is non-invasive, cosmesis outcome was better than brachytherapy boost. Recently, accelerated partial breast irradiation (APBI) for breast cancer patients with brachytherapy as the sole radiation modality following lumpectomy has shown promising results for select early-stage breast cancer patients (Arthur et al. 2003a; Kuske et al. 1994; Patel et al. 2003). In this technique, the radiation
755
756
Rupak K. Das and Rakesh Patel
dose is delivered within and surrounding the original tumor site over 4 to 5 days, instead of the traditional 6 weeks of external beam to the entire breast. Both high dose rate (HDR) and low dose rate (LDR) brachytherapy have been used for APBI.
Rationale and Selection Criteria for Partial Breast Irradiation Breast conservation therapy (BCT) is now widely accepted as a treatment option for most women with stages I and II invasive breast cancer and most patients with ductal carcinoma in situ (DCIS) (Veronesi et al. 1986; Blichert-Toft et al. 1988; van Dongen et al. 1992; Sarrazin et al. 1989; Fisher et al. 1989). Despite superior cosmetic outcome, BCT is more complex and requires a protracted treatment regimen comprising 6 weeks of daily external beam radiation therapy to the whole breast. This often proves prohibitive for the working woman, elderly patients, and those who live a significant distance from a radiation treatment center. In addition, with the more frequent use of adjuvant chemotherapy in both node-negative and node-positive patients, substantial delays can be incurred prior to the initiation of systemic chemotherapy if a conventional fractionated course of irradiation (XRT) is given first or in delivery of loco-regional XRT if chemotherapy is delivered beforehand. Thus, despite the potential benefits, only 30% to 50% of patients who are candidates for breast conservation actually receive it (Farrow, Hunt, and Samet 1992). Most of the logistical problems associated with BCT relate to the protracted course of external beam XRT delivered to the whole breast. Standard therapy after tumor excision generally includes 5 weeks of external beam XRT to the whole breast (45 to 50 Gy) followed by an additional 10 to 15 Gy boost to the tumor bed. The rationale for this approach is based upon two principles. First, higher doses of XRT are given to the “tumor bed” in an attempt to control residual small foci of cancer that may be left behind after excision alone. Second, whole breast XRT is used to eliminate possible areas of occult multicentric in situ or infiltrating cancer in remote areas of the breast. That such remote, multicentric areas of cancer exist has long been established. However, the biological significance of these areas of occult cancer is unknown and the necessity to prophylactically treat the entire breast has recently been questioned. There are many prospective randomized trials that have been conducted comparing the outcome of patients treated with lumpectomy alone or followed by whole breast XRT (Liljegren et al. 1994; Clark et al. 1992; Veronesi et al. 1993; Forrest et al. 1996; Schnitt et al. 1996). Across these studies, there is a similar threefold reduction in the ipsilateral breast failure rate with the addition of radiotherapy. For instance, in the National Surgical Adjuvant Breast Project (NSABP) B-06 randomized prospective trial, which included patients with invasive cancers up to 4 cm in size with negative surgical margins, the breast recurrence rate of 43% with lumpectomy alone is considerably higher than the 10% breast failure rate of the lumpectomy and whole breast irradiation arm (Fisher et al. 1989). In these studies, the percentage of patients receiving whole breast irradiation who recur in a remote location of the breast potentially not covered by partial breast irradiation was low (<3.5%). Is this absence of the expected number of remote breast recurrences due to sterilization of other quadrants by the whole breast radiation therapy employed in these series or to the biological insignificance of occult cancer foci in remote quadrants of the breast? If the former were true, then one would expect increased remote relapses in the conservative surgery-alone patients. To address this question, one must examine patients treated with surgery only and evaluate the number of remote recurrences. In fact, the proportion of patients who recur in a portion of the breast that would not be covered by partial breast irradiation is the same as patients receiving whole breast irradiation. This observation is validated by studies such as those conducted by Liljegren (Liljegren et al. 1994) and by Crile and Esselstyn (1990), who also demonstrated that 85% and 84% of local recurrences were in the immediate vicinity of the lumpectomy site defined as the surgical scar and the skin directly over the surgical field translating to an absolute risk of remote recurrence of 3.0% and 1.7%, respectively. From these data, one can infer that radiation therapy following lumpectomy has as its maximal effect on the reduction of breast
39–Breast Interstitial Implant and Treatment Planning
757
cancer recurrence at or very near the lumpectomy site casting doubt on the belief that remote recurrences originate from multicentric disease at the time of the index lesion demonstrated on histopathology. If the above observations are valid, radiation therapy can be confidently directed to the tissue surrounding the excision cavity of the breast. There are currently several groups studying the efficacy of lumpectomy bed irradiation alone in the management of early-stage breast cancer patients treated with BCT. Both interstitial brachytherapy techniques as well as external beam irradiation protocols have been implemented. Preliminary results from these trials are very encouraging and the techniques have been shown to be safe, tolerable, and highly reproducible (Arthur et al. 2003b).
Treatment Modalities Low Dose-Rate APBI LDR breast brachytherapy for partial breast irradiation has been successfully done both by 125I and 192Ir seeds (Arthur et al. 2003b; Vicini et al. 1997). In the former study 192Ir ribbons were used to deliver a dose of 45 Gy at a rate of 0.5 Gy/h while in the latter study 125I ribbons were used to deliver a dose of 50 Gy at a rate of 0.52 Gy/h. Catheters were placed under image guidance (fluoroscope, ultrasound) with a template as guidance or freehand, details of which are described in the implant section. In both the studies the treatments were delivered as in-patient for 4 to 5 days. At the end of the treatment the radioactive seeds in ribbons were removed followed by the removal of the catheters, and the patient discharged with skin-care instructions.
High Dose-Rate APBI In the HDR remote afterloader, a computer-driven, single cable with a source (192Ir, with an activity of 10 Ci) at the tip, moves from each programmed treatment position (dwell position) in a catheter after the position-specific treatment time (dwell time) has elapsed. After treating each position in a given catheter, the source is retracted into the machine and transmitted into the next treatment catheter. Such a system, also known as “stepping-source remote afterloader,” enables the planner to maximize the dose uniformity by varying the dwell time in each dwell position, while minimizing the implant volume needed to adequately cover the target volume.
Interstitial Volume Implants Template-Guided Implant Within 8 weeks of lumpectomy and axillary nodal evaluation, patients undergo an interstitial implant with one of two methods: a prone, stereotactic, template method with digital mammographic guidance or a supine, ultrasound-guided technique, both under local anesthesia. In the prone method, an ultrasound is initially performed allowing visualization and aspiration of the seroma contents followed by injection of 4 to 5 cm3 of non-ionic contrast and 3 to 4 cm3 of air, resulting in an air-fluid cavity that is well-visualized on subsequent pre-implant mammography. Next, the patient is positioned prone on the stereotactic table and the template applied, such that the surgical scar is located between the two templates and the contrast-enhanced cavity is centered on the template. After a digital mammogram is obtained and the target volume is demarcated, the respective template holes are numbed with local anesthetic. The excision cavity is defined by surgical clips or radiopaque contrast filling the seroma as described above with the target volume being defined as the volume encompassed by a 2-cm margin outside the lumpectomy cavity in all dimensions. A thick guiding template is then attached to the system, ensuring parallel
758
Rupak K. Das and Rakesh Patel
needle placement without deviation. Freehand needles are then placed in regions not covered by the template for optimal target volume coverage.
Freehand Implant: Ultrasound-Guided Supine Technique An alternate approach, the supine ultrasound-guided method, is used for situations where the template is difficult to use (medial lesions and smaller breast size). In this method, the target volume is first mapped out on the skin surface with ultrasound visualization of the lumpectomy cavity. Needle placement is marked on the skin with a 1-cm interval. No template or contrast injection is utilized. Needles are placed in a similar fashion as above with generous use of a local anesthetic mixture. Initially, a deep plane of needles is placed on the pectoralis major fascia with real-time ultrasound guidance. This is followed by at least one superficial plane resulting in a multiplane implant for adequate geometric coverage of the target volume. In both procedures, the final step requires replacing the needles with polyethylene tubing with a hemispherical button at each end. Extra attention is given to make sure that the button on the connector side of the remote afterloader is tight and flush to the skin.
Comparison of Implant Techniques The table below compares the two types of implant techniques: Supine/Ultrasound Technique
Prone/Digital Mammogram Technique
Placement of catheter is radiation oncologist dependent Less number of catheters Irradiation time same since planning target volume (PTV) is the same Since less catheters, simulation/planning time is less Total treatment time is less since less catheters and hence less time in check cable runs
Template-guided makes it radiation oncologist independent More number of catheters Irradiation time is the same More catheters, hence more simulation/planning time More treatment time with more catheters and more check cable runs
2-D Treatment-Planning Dosimetry Imaging and Catheter Reconstruction Traditionally, two-dimensional (2-D) dosimetric planning has always been initiated by the acquisition of orthogonal film radiographs of the implant. In order to differentiate and construct the catheters, dummy radiopaque ribbons, mimicking the ribbons with radioactive seeds (1-cm spacing), with markers for ribbon identification, are inserted in each catheter. Orthogonal radiographs are then acquired for source localization. Individual seeds in each ribbon are identified in the films and the data are then transferred to the treatment-planning system through the digitizer for computerized reconstructions. Since wide volume breast implants can have a large number of catheters (N>20), differentiating all catheters in one set of orthogonal films can be very challenging and may lead to errors. Dividing the simulation of the breast implant into two sets requires the patient and the breast implant to be positioned and oriented in the same way for both sets of orthogonal films. Error in the position and orientation of the implant in the two sets of films gives rise to errors in the volume of the implant. For HDR implants each catheter is trimmed and
39–Breast Interstitial Implant and Treatment Planning
759
numbered during simulation and is then hooked up to the HDR source simulator with the transfer tube. The length of each catheter is recorded, which is later used for programming the length of each channel of the HDR remote afterloader.
Planning Target Volume (PTV) Definition Once the catheters have been reconstructed, a PTV needs to be defined. Surgical clips placed at the time of the breast conserving surgery help delineate the lumpectomy cavity in the films. Computed tomography (CT) images with contrast agents filled in the lumpectomy taken after the implant also help in the delineation of the lumpectomy cavity. With the help of bony landmarks and surgical clips, the user has to translate the surgical cavity from the CT-based three-dimensional (3-D) information to the 2-D radiographic films. Although feasible, the process can be time-intensive, challenging, and inaccurate for the physicist and physician in defining a 3-D target volume on 2-D films. Once the lumpectomy cavity is defined, a 1- to 2-cm margin is added all around the cavity to define the PTV on the films.
Dose Optimization Before the widespread use of computer-aided dosimetry in implant therapy, classical implants were the main tools that could predict the dose distribution in and around an implantation system (Paterson and Parker 1934, 1938; Quimby 1935, 1941; Dutreix, Marinello, and Wambersie 1982). With the current standard of fast computers with large random access memories (RAMs), the location and the strength of the sources can be easily manipulated to produce a dose distribution that best conforms to the physician’s clinical goals. In single stepping-source modality, the optimization technique varies the active dwell positions and the relative dwell times to achieve the goal. All brachytherapy implants can be classified into two methods: (1) distance implant, consisting of one or two catheters or a singleplane implant and (2) volume implant. A detailed description on the theory of optimization has been described by Ezzell and Luthmann (1995). For breast implants, volume optimization, to make the dose in between the catheters as homogeneous as possible by dose point optimization and geometric optimization to determine the total relative dwell time to be used for each catheter, is a very common practice adhered to at the University of Wisconsin for breast implant dosimetry. Using the PTV on the films as a guidance, the physician can accordingly choose a normalizing factor to increase or decrease the 100% isodose volume to cover the target volume.
3-D Treatment-Planning Dosimetry Imaging and Catheter Reconstruction CT-based simulation of the patient is performed with the patient in the supine position within 24 hours of the implant. An en face picture of the implant along with the breast is obtained. After each catheter is trimmed and numbered, it is hooked up to the HDR source simulator with the transfer tube. The length of each catheter is recorded, which is later used for programming the length of each channel of the HDR remote afterloader. Simultaneously, each numbered catheter/button is also identified in the en face picture, which is subsequently used to identify and reconstruct the catheters in the treatment-planning system. Scanning is performed after each catheter is loaded with CT-compatible wires and the images are then transferred to the treatment-planning system. In the treatment-planning computer, the CT images are loaded in the planning system and catheter reconstruction is done by scrolling the CT slices from cephalad to caudad or vice versa while identifying the buttons on the CT slices with the help of the en face picture. Each catheter is then reconstructed from
760
Rupak K. Das and Rakesh Patel
one button end to the other by following the CT-compatible wires that were inserted in each catheter prior to the scan. An en face view of the implant, similar to the en face picture taken at the simulator, is generated. Configuration of the buttons between the two images is compared and used as a quality assurance check for the reconstruction of the catheters.
PTV Definition In defining the target volume, first the lumpectomy cavity is delineated on each axial CT slice. Radiopaque material like Omnipaque™ or Hypaque™ of suitable dilution, if injected in the lumpectomy cavity before the CT scanning or during the catheter placement, enhances the visualization of the cavity, and helps in the delineation. Target volume is defined as the lumpectomy cavity with a 2-cm margin, modified to 5 mm deep to the skin surface and also along the pectoral muscle as shown in Figure 1. Dwell positions in each catheter are then autoactivated by the treatment plan with some additional margins of 5 to 8 mm on the target volume, thus giving more degrees of freedom to conform the dose to the PTV. It is more efficient to add extra dwell positions at this phase of the planning process since, at the end, many of these extra dwell positions can be manually forced to zero dwell time rather than going back and adding more dwell positions—most planning systems will automatically discard all the optimizations that has been done before. Dwell positions within 5 mm of the skin are never activated to protect the patient skin.
Dose Optimization Geometric optimization for volume implant is done, the theory of which has been detailed elsewhere (Ezzell and Luthman 1995), and basal dose points on the central plane are generated with the help of the
Figure 1. Axial image of an interstitial breast implant with the lumpectomy cavity. A 2-cm margin was added to the lumpectomy cavity to generate the target volume that was then modified to 5 mm deep to the skin surface and also along the pectoral muscle.
39–Breast Interstitial Implant and Treatment Planning
761
treatment-planning software. Prescription of 3.4 Gy for 10 fractions or 4 Gy for 8 fractions are prescribed to the basal dose points and an isodose line selected to cover the entire target as optimally as possible. Manual optimization on each CT slice is then done interactively by dragging the 100% isodose line to cover the target volume as conformally as possible while adjusting the 150% isodose line to minimize the hot spots. A typical plan with the isodose distribution of the central plane is shown in Figure 2. Each catheter length is then entered manually.
Evaluation of Treatment Plan Compared to 2-D dosimetry, the use of CT scans in a treatment-planning system allows 3-D visualization of the actual relationship of the implanted region to the target volume and other critical structures, as shown in Figure 3. This information becomes even more critical when defining the lumpectomy cavity or target volume as well as a nearby critical structure that affords dose-volume analysis of the skin, contralateral breast, lung, and heart. Additionally, dose optimization of implants by interactive graphics allows excellent target volume coverage and assessment of dosimetric quality concurrently, thereby instilling confidence that dose is delivered to the desired partial breast region. The primary objective of the dosimetry is as follows: 1. Ideally, 100% of the lumpectomy cavity and the target volume should be covered by the prescription isodose line. Figure 4 shows one such plan with 100% of the lumpectomy cavity with a volume of 19.9 cm3 and 95.4% of the target with a volume of 230.5 cm3, covered by the prescription dose of 3.4 Gy per fraction. Critical structures like heart, lung, skin, and contralateral breast can also be delineated and its dose volume histogram (DVH) can be generated to aid the physician in the decision of the treatment plan.
Figure 2. Isodose distribution (100%) for an implant showing the lumpectomy cavity along with the modified target volume.
762
Rupak K. Das and Rakesh Patel
Figure 3. 3-D view of dose distribution (100% dose cloud) along with catheter reconstruction and target volume.
Figure 4. Dosimetry plan with 100% of the lumpectomy cavity with a volume of 19.9 cm3 and 95.4% of the target with a volume of 230.5 cm3, covered by the prescription dose of 3.4 Gy per fraction.
39–Breast Interstitial Implant and Treatment Planning
763
2. Target volume and the volume covered by the 100% isodose line V100 should be as congruent as possible or mathematically Target Volume ∩ V100 Target Volume ∪ V100
~1
(1)
should be equal unity for an ideal implant, where the 100% isodose lines conform exactly to the target volume. 3. Dose Homogeneity Index (DHI) as defined by DHI =
V100 − V150 V100
(2)
where V100 and V150 is the volume covered by the 100% and 150% isodose lines, respectively, and Wu, Ulin, and Sternick (1988) has been used to determine that the level of dose homogeneity for the implant should be as high as possible. A DVH for the implant is generated to record V100 and V150 to calculate DHI. For an ideal implant, DHI should also have a value of unity. However, this is unrealistic and impossible since there will be some hot spots around the source due to dose falloff. At the University of Wisconsin, a thumb rule is not to let the 150% isodose line from each dwell positions touch each other. At our institution, 0.75 < DHI < 0.90 is considered acceptable. All the above three characteristics should be taken into account in order to achieve the best dosimetry achievable for a given implant. For example, while planning a given implant, the absence of adequate catheters in and around the target volume might prompt the planner to cover the target volume at the cost of dose homogeneity by increasing the hot spots around some dwell positions. Even though the anatomical location of interstitial breast implants requires less concern for dose-limiting structures like heart and lungs (Kestin et al. 2000), telangiectasia and fat necrosis have always been dose-limiting concerns for this kind of procedure (Roston and El-Sayed 1987; Clarke et al. 1983). Extra attention should be given for dosimetric quantity like DHI, which accounts for the local “hot spots” in the implants since it can reduce the incidence of fat necrosis and other complications and tissue toxicities. Under these circumstances and on most occasions, we at the University of Wisconsin have added more catheters rather than accept a plan with a higher DHI. Contrary to this scenario, there can be situations where more than adequate catheters have been implanted. In these situations, adequate catheters might improve the DHI at the cost of treating more breast tissue, which leads to a higher ratio of the target volume to V100 congruency ratio. It is imperative that the physician and the physicist work closely together, keeping in mind the clinical goals, in achieving a successful APBI brachytherapy program.
Quality Assurance With the introduction of computerized treatment-planning systems and with new optimization algorithms as explained above, it has become difficult for the physicist to perform a meaningful check of the treatment plan output to gain the needed confidence that the forthcoming treatment will be appropriate. A commonly used method is to incorporate a point far from the implant and to evaluate the dose in the treatment-planning system. Using point source approximation, the dose to the same point is calculated manually using the total time from the plan along with the source strength and then compared with the dose to the point from the plan. To make this verification work for large implants like breast implants,
764
Rupak K. Das and Rakesh Patel
where the average treatment volume is about 200 cm3 with dimensions on the order of 6 cm, a point farther than 12 cm is needed to approximate this volume (within 5%) as a point source. Published radial function data greater than 12 cm either are not available or have large uncertainties, and hence a precise point dose calculation could not be performed by the treatment plan and compared to a hand calculation. A couple of other standard techniques (calculate some characteristic parameter or indices of the plan and compare it to an expected value) has been described by different authors. Thomadsen et al. (1992) used indices for different applicators and published the expected value of the indices for those applicators. Kubo and Chin (1992) published a single formula for checking treatment time with some constraints on the distance between the prescribed treatment and the center of the source train. Thomadsen (1993) offered a method for a double check for volume implants. This study is built on Thomadsen’s work, and provides an easy and quick calculation check for multicatheter breast implants. As explained above, for each plan, a DVH was generated to find the volume of the implant covered by the prescription dose, V100. RV as defined by Equation (3) for this volume was then calculated from R V ( Ci ⋅ s/Gy ) =
Time(s) ⋅ Activity (Ci) Elongation Factor ⋅ Prescribed Dose(Gy)
(3)
where Time(s) is the irradiated time from the plan, Activity (Ci) is the activity of the HDR source, and the Elongation Factor has been derived from the original Manchester volume implant table (Paterson and Parker 1938). This derived RV can then be compared to the original Manchester RV, corrected for modern units and factors and conversion from mgRaEq-h to Ci-s. Conversely, using the original Manchester RV, one can calculate the predicted time and compare it to the time from the computer plan (Das et al. 2004). Figure 5 represents the correlation between the implant volume (V100) to RV. In Figure 5a, the + represents this study plotted as a function of the implant volume. It was then compared to the original Manchester volume implant table (line) corrected for modern units and factors. To magnify the correlation, the ratio of this study and the Manchester calculation wasplotted as a function of the implant volume.
Figure 5. (a) Comparison of RV (Irradiated Time ⋅ Activity/Elongation Factor ⋅ Prescribed Dose) for this study with that of Manchester system. (b) Ratio of this study to that of the Manchester system.
39–Breast Interstitial Implant and Treatment Planning
765
As evident from the ratio study, Figure 5b, the RV values from this study agree within ±7% of the Manchester volume implant table. In our institution the quality assurance program for manual verification of HDR computer irradiation time calculation is based on this principle. A double check of the length of each catheter in the plan is also compared to the length of each catheter measured in the simulator.
Summary Traditionally, brachytherapy dosimetry has been performed by simulating a patient by inserting dummy ribbons in the catheters of the interstitial implant and obtaining orthogonal films. Aided by pre-implant CT images, radiation oncologists have tried to define the target volume on these simulation films. Since it is very challenging to define a 3-D target volume on a 2-D film, definition of the target volume has always been highly physician dependent. For breast interstitial implants, the average number of catheters used is quite large (20 in our institution). For many patients, at least two sets of orthogonal films need to be obtained to reduce the complexities of identification of the catheters. Positioning the patient in the exact same location twice is challenging and can result in inaccuracies in the definition of the target volume. With 2-D dosimetry, standardization of breast interstitial implants then becomes a daunting task and hence an objective clinical study of lumpectomy cavity coverage along with a variable margin and correlation to clinical outcome among multi-institutions has been almost impossible. CT-based interstitial breast implants with stepping-source allows improved treatment volume conformality. Today, this planning approach is being implemented in many institutions since it allows improved target volume delineation and optimal coverage and is much easier and more accurate than the conventional orthogonal film dosimetry. Using dose volume analysis of target volume coverage surely will be a powerful tool for the radiation oncologist.
References Arthur, D.W., D. Koo, R. D. Zwicker, S. Tong, H. D. Bear, B. J. Kaplan, B. D. Kavanagh. L. A. Warwicke, D. Holdford, C. Amir, K. J. Archer, and R. K. Schmidt-Ullrich. (2003a). “Partial breast brachytherapy after lumpectomy: Low dose-rate and high dose-rate experience.” Int J Radiat Oncol Biol Phys 56:681–689. Arthur, D. W., F. A. Vicini, R. R. Kuske, D. E. Wazer, and S. Nag. (2003b). “Accelerated partial breast irradiation: An updated report from the American Brachytherapy Society.” Brachytherapy 2:124–130. Blichert-Toft, M., M. Brincker, J. A. Andersen, K. W. Andersen, C. K. Axelsson, H. T. Mouridsen, P. Dombernowsky, M. Overgaard, C. Gadeberg, G. Knudsen, et al. (1988). “A Danish randomized trial comparing breast-preserving therapy with mastectomy in mammary carcinoma.” Acta Oncol 27:671–677. Clark, R. M., P. B. McCulloch, M. N. Levine, M. Lipa, R. H. Wilkinson, L. J. Mahoney, V. R. Basrur, B. Nair, R. S. McDermot, and C. S. Wong. (1992). “Randomized clinical trial to assess the effectiveness of breast irradiation following lumpectomy and axillary dissection for node-negative breast cancer.” J Natl Cancer Inst 84:683–689. Clarke, D., J. L. Curtis, A. Martinez, L. Fajardo, and D. Goffinet. (1983). “Fat necrosis of the breast simulating recurrent carcinoma after primary radiotherapy in the management of early stage breast carcinoma.” Cancer 52:442–445. Crile Jr., G., and C. B. Esselstyn Jr. (1990). “Factors influencing local recurrence of cancer after partial mastectomy.” Cleveland Clin J Med 57:143–146. Das, R. K., R. Patel, H. Shah, H. Odau, and R. R. Kuske. (2004). “3D CT-based high-dose-rate breast brachytherapy implants: treatment planning and quality assurance.” Int J Radiat Oncol Biol Phys 59:1224–1228. Dutreix, A., G. Marinello, and A. Wambersie. Dosimetrie en Curiethérapie. Paris: Masson, 1982. Ezzell, G. A., and R. W. Luthmann. “Clinical Implementation of Dwell Time Optimization Techniques for Single Stepping-Source Remote Applicators” in Brachytherapy Physics, J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.). AAPM Summer School 1994. Madison, WI: Medical Physics Publishing, pp. 617–639, 1995.
766
Rupak K. Das and Rakesh Patel
Farrow, D. C., W. C. Hunt, and J. M. Samet. (1992). “Geographic variation in the treatment of localized breast cancer.” N Engl J Med 326:1097–1101. Fisher, B., C. Redmond, R. Poisson, R. Margolese, N. Wolmark, L. Wickerham, E. Fisher, M. Deutsch, R. Caplan, Y. Pilch, et al. (1989). “Eight-year results of a randomized clinical trial comparing total mastectomy and lumpectomy with or without radiation in the treatment of breast cancer.” N Engl J Med 320:822–828. Forrest, A. P., H. J. Stewart, D. Everington, R. J. Prescott, C. S. McArdle, A. N. Harnett, D. C. Smith, and W. D. George. (1996). “Randomised controlled trial of conservation therapy for breast cancer: 6-year analysis of the Scottish trial. Scottish Cancer Trials Breast Group.” Lancet 348:708–713. Kestin, L. L., D. A. Jaffray, G. K. Edmundson, A. A. Martinez, J. W. Wong, V. R. Kini, P. Y. Chen, and F. A. Vicini. (2000). “Improving the dosimetric coverage of interstitial high-dose rate breast implants.” Int J Radiat Oncol Biol Phys 46:35–43. Keynes, G. (1929). “The treatment of primary carcinoma of the breast with radium.” Acta Radiol 10:393–402. Kubo, H., and R. B. Chin. (1992). “Simple mathematical formulas for quick checking of single catheter high dose rate brachytherapy treatment plans.” Endocuriether/Hypertherm Oncol 8:165–169. Kuske, R. R., J. S. Bolton, R. M. Wilenzick, W. M. P. McKinnon, B. Pullen, T. G. Scroggins, E. L. Zakris, and B. B. Fineberg. (1994). “Brachytherapy as the sole method of breast irradiation in Tis, T1, T2, N0,1 breast cancer.” ASTRO, San Francisco, CA, Oct 2–7, 1994. (Abstr.) Int J Radiat Oncol Biol Phys 30(S1):245. Liljegren, G., L. Holmberg, H. O. Adami, G. Westman, S. Graffman, and J. Bergh. (1994). “Sector resection with or without postoperative radiotherapy for stage I breast cancer: Five-year results of a randomized trial. UppsalaOrebro Breast Cancer Study Group.” J Natl Cancer Inst 86:717–722. Patel, R. R., H. K. Shah, R. K. Das, A. Forouzannia, H. Odau, and R. R. Kuske. “Interstitial High Dose Rate Accelerated Partial Breast Irradiation Alone in Breast Conservation Therapy: The University of Wisconsin Experience.” Proceedings of the 89th Scientific Assembly and Annual Meeting of the Radiological Society of North America, #012-RO. p. 147, 2003. Paterson, R., and H. M. Parker. (1934). “A dosage system for gamma-ray therapy.” Br J Radiol 7:592–632. Paterson, R., and H. M. Parker. (1938). “A dosage system for interstitial radium therapy.” Br J Radiol 11:252–266, 313–340. Quimby, E. H. (1935). “Physical factors in interstitial radium therapy.” Am J Roentgenol Radiat Ther 33:306–316. Quimby, E. H. (1941). “The specification of dosage in radium therapy.” Am J Roentgenol Radiat Ther 45:1–18. Roston, A. Y., and M. E. El-Sayed. (1987). “Fat necrosis of the breast: An unusual complication of lumpectomy and radiotherapy in breast cancer.” Clin Radiol 38:31. Sarrazin, D., M. G. Le, R. Arriagada, G. Contessa, F. Fontaine, M. Spielmann, F. Rochard, T. Le Chevalier, and J. Lacour. (1989). “Ten-year results of a randomized trial comparing a conservative treatment to mastectomy in early breast cancer.” Radiother Oncol 14:177–184. Schnitt, S., J. Hayman, R. Gelman, T. J. Eberlein, S. M. Love, K. Mayzel, R. T. Osteen, A. J. Nixon, S. Pierce, J. L. Connolly, P. Cohen, L. Schneider, B. Silver, A. Recht, and J. R. Harris. (1996). “A prospective study of conservative surgery alone in the treatment of selected patients with stage I breast cancer.” Cancer 77:1094–1100. Shank, B. (1986). “Breast preservation and brachytherapy.” Endocuriether/Hypertherm Oncol 2:17–24. Thomadsen, B. R. Calculation checks for volume implants (personal communication). University of Wisconsin, Madison, WI, 1993. Thomadsen, B. R., S. Shahabi, J. A. Stitt, D. A. Buchler, J. F. Fowler, B. R. Paliwal, and T. J. Kinsella. (1992). “High dose rate intracavitary brachytherapy for carcinoma of the cervix: The Madison system: II. Procedural and physical considerations.” Int J Radiat Oncol Biol Phys 24:349–357. van Dongen, J. A., H. Bartelink, I. S. Fentiman, T. Lerut, F. Mignolet, G. Olthuis, E. van der Schueren, R. Sylvester, J. Winter, K. van Zijl, et al. (1992). “Randomized clinical trial to assess the value of breast-conserving therapy in stage I and II breast cancer, EORTC 10801 trial.” J Natl Cancer Inst Monogr 11:l5–18. Veronesi, U., A. Banfi, M. Del Vecchio, R. Saccozzi, C. Clemente, M. Greco, A. Luini, E. Marubini, G. Muscolino, F. Rilke, et al. (1986). “Comparison of Halsted mastectomy with quadrantectomy, axillary dissection, and radiotherapy in early breast cancer: Long-term results.” Eur J Cancer Clin Oncol 9:1085–1089. Veronesi, U., A. Luini, M. Del Vecchio, M. Greco, V. Galimberti, M. Merson, F. Rilke, V. Sacchini, R. Saccozzi, T. Savio, et al. (1993). “Radiotherapy after breast-preserving surgery in women with localized cancer of the breast.” N Engl J Med 328:1587–1591.
39–Breast Interstitial Implant and Treatment Planning
767
Vicini, F. A., P. Y. Chen, M. Fraile, G. S. Gustafson, G. K. Edmundson, D. A. Jaffray, P. Benitez, J. Pettinga, B. Madrazo, J. A. Ingold, N. S. Goldstein, R. C. Matter, and A. A. Martinez. (1997). “Low-dose-rate brachytherapy as the sole radiation modality in the management of patients with early-stage breast cancer treated with breas-conserving therapy: Preliminary results of a pilot trial.” Int J Radiat Oncol Biol Phys 38(2):301–310. Wu, A., K. Ulin, and E. S. Sternick. (1988). “A dose homogeneity index for evaluating 192Ir interstitial breast implants.” Med Phys 15:104–107.
Chapter 40
Partial Breast Irradiation Using the MammoSite® Radiation Therapy System Wayne M. Butler, Ph.D., and Ernest G. Butler, B.S. Schiffler Cancer Center Wheeling Hospital Wheeling, West Virginia Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769 How a Single Dwell Site May Be Equivalent to Many . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 770 MammoSite Implantation Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 770 When to Perform the Implant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 771 Implant Evaluation and Planning from X-Rays and CT Scans . . . . . . . . . . . . . . . . . . . . . . . . . 772 Finding the Balloon Center and the Volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772 2-D Plan from Orthogonal Films . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772 3-D CT Skin Spacing Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 772 Conformance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773 3-D CT Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 775 Treatment Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779 QA Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779 Fallback Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779 Follow-Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 779 Ongoing and Future Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 781 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 781
Introduction The MammoSite® device is a balloon catheter that is positioned into the lumpectomy cavity for breast cancer treatments. The device is made by Proxima Therapeutics, Inc. (Alpharetta, GA) and is available in three different models. The original, which received Food and Drug Administration (FDA) clearance in May 2002, was a 4 to 5 cm diameter spherical balloon (see Figure 1). Since then, a 5 to 6 cm diameter spherical balloon and a 4 by 6 cm ellipsoidal balloon have also become available. The high dose rate (HDR), 5-day course of treatment for early-stage breast cancer requires a remote afterloading system. An 192 Ir source, which is attached to a cable, is mechanically positioned to the center of the balloon where it delivers the prescribed dose to an equatorial point 1 cm outside the surface of the balloon. Accelerated partial breast irradiation (APBI) was originally developed using interstitial needles to meet certain patient needs by offering a reduction in time and inconvenience of standard external beam radiation therapy (XRT). The short APBI treatment course may also minimize chemotherapy scheduling conflicts and may even reduce cosmetic problems and chronic toxicity. However, the real justification for APBI comes from recent studies published by Imamura et al. (2000), Ohtake et al. (1995), and Faverly et al. (1994). These groups demonstrated that a 1-cm margin around the lumpectomy cavity should provide adequate coverage of invasive adenocarcinoma extension in about 90% of cases. In carefully selected patients with early-stage breast cancer, treatment of the lumpectomy cavity only with HDR brachytherapy was found to be technically feasible and well tolerated (Baglan et al. 2001; Arthur et al. 2002). The MammoSite balloon brachytherapy device was conceived to satisfy the same clinical and patient convenience criteria as APBI, while being more appealing to patients based on the assumption that most women would prefer a single, large puncture site to the multiple needle paths necessary for conventional, interstitial HDR breast brachytherapy. Proxima also hoped that patient enthusiasm would encourage radiation oncologists to evaluate the device, and that the apparent simplicity of the device and the short learning
769
770
Wayne M. Butler and Ernest G. Butler
Figure 1. The MammoSite® device with various components identified. (Image is printed with permission from Proxima Therapeutics, Inc.)
curve necessary to implement a single HDR dwell site in the center of a spherical cavity would lead many to use the device. The rapid and wide adoption of the MammoSite device has justified the company’s business plan.
How a Single Dwell Site May Be Equivalent to Many The obvious difference between interstitial implants and MammoSite is the use of multiple dwell positions in multiple catheters versus one dwell site in a single catheter. How could the dosimetry between the two techniques be comparable? By restricting the acceptable size of the lumpectomy cavity to a relatively small volume and using a pressurized balloon to approximately double the cavity volume, adjacent tissue has to stretch to accommodate the balloon. In this way, breast tissue that may contain microscopic disease as far away as 2 cm from the incision may be pulled to within 1 cm of the balloon surface. Note that the balloon pressure does not compress breast tissue, which is essentially incompressible, but forces tissue around the cavity to conform to the increased volume and surface area of the balloon.
MammoSite Implantation Considerations The MammoSite is currently used to treat infiltrating ductal carcinomas that are less than 3.0 cm in diameter. However, phase III trials are pending which will include treatment of larger tumors, ductal carcinoma in situ (DCIS), lobular carcinomas, intraductal carcinomas, and even women with positive nodes. To determine patient eligibility for MammoSite brachytherapy, there must be an evaluation of the minimum lumpectomy cavity distance from the patient’s skin surface and an evaluation of the breast size and cavity location. Every center should create a list of guidelines based on clinical judgment and results from other institutions such as those in Table 1. Before offering MammoSite to patients, it is necessary for the oncology team (typically a surgeon, radiation oncologist, and a physicist) to learn the techniques required. The requisite skills are taught in a mandatory training course offered by Proxima Therapeutics. This course includes topics such as when to
40–Partial Breast Irradiation Using the MammoSite® Radiation Therapy System
771
Table 1. An Example of Accelerated Partial Breast Irradiation Selection Criteria Used for Spherical MammoSite Treatments at the Schiffler Cancer Center Parameter Tumor size Nodal status
Criterion
Metastatic status Surgical margins Extensive intraductal component Lobular histology DCIS alone Age Tumor location Breast morphology Skin spacing Balloon symmetry Air pocket volume
< 3 cm N0, N1 hematoxylin & eosin stain negative (May be pan keratin positive) M0 >1 mm Negative No No Clinical judgment Clinical judgment Clinical judgment >7 mm L = W ± 3 mm < 7 % of Treatment Volume
Off axis
≤ 3 mm
perform the implant procedure (at the time of lumpectomy or later) and how to choose the best size and shape of the balloon for the cavity. The development of a cooperative team is helpful for seamless procedure scheduling, because coordination between the radiation oncology center and the operating room (OR) can avoid technical complications and difficulties. Communications between the radiation oncology center and the OR often pertain to the inventory of MammoSite devices and billing, surgical scheduling and future appointments with radiation oncology, preparation of the 5–10% non-ionic contrast solution, and recording of the volume used to fill the device.
When to Perform the Implant The patient should be informed about the possibility of infrequent and typically minor side effects, such as infection of the catheter wound site and various skin reactions associated with the MammoSite radiation procedure. Once these topics have been discussed and the patient has consented to the procedure, a decision must be made as to when to perform the implant. The MammoSite balloon catheter may be placed during the lumpectomy or nodal dissection, but the balloon may need to be removed prior to the start of treatment if pathology reports an unsatisfactory nodal status. Although in principle one may orient the catheter so that source anisotropy contributes to dose reduction on the chest wall and at the skin surface, surgeon preference, patient convenience, and cosmetic considerations carry more weight in clinical practice. Placement of the MammoSite in a separate procedure after the pathology is complete may seem more practical because it does not waste an expensive device and lessens patient disappointment if the procedure must be aborted. However, insertion delays can cause difficulty in finding the tumor bed if too much time has elapsed. With ultrasound localization, the lumpectomy site is normally still visible after a couple of weeks. Of course, calibration of the directional and spatial accuracy of the ultrasound system is essential.
772
Wayne M. Butler and Ernest G. Butler
Implant Evaluation and Planning from X-Rays and CT scans Finding the Balloon Center and the Volume Treatment planning using a set of orthogonal x-rays and a computed tomography (CT) scan enhances accuracy and maximizes the dosimetric detail possible. The planning can be performed on either a two-dimensional (2-D) platform using the plain films or on a three-dimensional (3-D) platform using the CT scan or both. X-ray fluoroscopy and films are used to find the center of the balloon. A radiopaque dummy source position indicator is initially positioned in the approximate balloon center position (or a position that seems reasonable under fluoroscopy). The balloon is drawn on the films. This often requires patience and/or a hot lamp. The center is determined by positioning a 60° angle bisector with the angle sides tangent to the balloon and then drawing a line segment with the bisector. The angle template is then repositioned and the process repeated to find the center. Measuring the difference between the physical source position indicator and the geometrically determined center (with a magnification ruler or otherwise accounting for the film magnification) will then determine the necessary shift of the source position indicator. To determine the approximate volume, the length of the MammoSite balloon should be measured through the center parallel to the lumen and the width measured through the center perpendicular to the lumen. Using these measurements and Table 2, the nominal volume of the MammoSite may be estimated and the value verified against the reported fill volume measured on the syringe. The x-ray filming and centering process should be repeated until the MammoSite is clearly visible with the source position indicator in the center. The films can also be used to measure any off-axis deviation of the source and any asymmetry of the balloon. The simulation work sheet provided in Figure 2 is a useful tool to collect patient data and is very helpful during the planning stage.
2-D Plan from Orthogonal Films The 2-D plan [we use Nucletron’s Plato brachytherapy planning system (Columbia, MD)] may be created by digitizing closely spaced points around the perimeter of the MammoSite on the first orthogonal film, and then digitizing the coinciding points along the axis of the MammoSite on the second orthogonal film. The lumen and the center source position are also digitized. With the active source placed in the center position, the treatment time can be optimized to deliver the prescription dose to at least one point that is 1 cm outside the MammoSite. One solution is to align the lumen with the z-axis and then prescribe and optimize the dose to four points, which correspond to balloon radii along the x- and y-axes and 1 cm away from the balloon surface. This four-point optimization will compensate for any off-axis deviation of the lumen and will determine an accurate treatment time.
3-D CT Skin Spacing Evaluation The CT scan is an essential tool that is necessary for evaluating the distance between the surface of the skin and the device (see Figure 3), the conformance of the balloon to the cavity (see Figure 4), and the distance between the source and nearby critical structures. Proxima now recommends that the skin spacing be at least 7 mm and this distance must be determined from a 3-D perspective. Sagittal or coronal views often present situations not appreciable from the transverse slices (Edmundson et al. 2002). Creation of a new, temporary structure with a 7-mm autoradius expansion of the MammoSite, which can then be examined with a multiplane view, is quite helpful in evaluating adequate skin margin. If the skin spacing is not sufficient, the spacing as well as doses to critical structures and other dosimetric values can often be improved by rolling the patient onto her side into a decubitus position as demonstrated in Table 3.
40–Partial Breast Irradiation Using the MammoSite® Radiation Therapy System
773
Table 2. Physical and Dosimetric Characteristics for the Variably Inflated MammoSite Balloons 4–5 cm MammoSite Nominal fill volume (cm3) 34 36 38 40 42 44 46 48 50 52 54 56 58 60 62 64 66 68 70
Width (cm)
Length (cm)
4.00 4.05 4.15 4.20 4.30 4.35 4.45 4.50 4.55 4.65 4.70 4.75 4.85 4.90 4.95 5.00 5.05 5.10 5.15
4.00 4.05 4.10 4.10 4.15 4.20 4.25 4.30 4.30 4.35 4.35 4.40 4.40 4.45 4.50 4.55 4.60 4.60 4.65
5–6 cm MammoSite Dose Rate * (cGy/min/Ci) @ 1 cm 8.43 8.20 7.98 7.79 7.58 7.44 7.27 7.10 6.97 6.83 6.70 6.58 6.44 6.35 6.26 6.15 6.05 5.97 5.89
Nominal fill volume (cm3) 70 75 80 85 90 95 100 105 110 115 120 125 — — — — — — —
Width (cm)
Length (cm)
4.87 4.96 5.06 5.15 5.24 5.34 5.43 5.52 5.62 5.71 5.80 5.90 — — — — — — —
5.11 5.17 5.22 5.28 5.33 5.39 5.45 5.50 5.56 5.61 5.67 5.72 — — — — — — —
Dose Rate * (cGy/min/Ci) @ 1 cm 6.37 6.21 6.03 5.88 5.74 5.58 5.45 5.32 5.18 5.06 4.94 4.82 — — — — — — —
*Dose Rate calculation is at 1 cm outside of the balloon surface.
Conformance Conformance is measured by contouring the volume of any air- or fluid-filled cavities abutting the balloon and comparing that volume with the total treatment volume. The total treatment volume is the annular volume from the balloon surface to the prescription depth 1 cm beyond the surface. Total treatment volume, TTV = 4/3 π(3r2 + 3r + 1) .
(1)
Conformance = 1 – (air or fluid volume)/TTV .
(2)
A typical conformance threshold requires that the volume of cavities not occupied by breast tissue or the balloon be less than 10% of the total treatment volume. Another way to determine conformance is to compare the surface area of the balloon in contact with tissue with the total balloon surface area (Rivard
774
Wayne M. Butler and Ernest G. Butler
MammoSite Placement Log Patient Name _______________________________ Patient ID___________________ Implant date _______________________
Simulation date ______________________
(FDA mandates a maximum of 28 days in the patient from implant to last treatment.)
Radiation Oncologist______________________ Surgeon________________________ Lot number ___________
Implant length ____________ Breast
Exact volume of fluid in MammoSite _____________
cm_3
(35
q Right
3 cm_
q
Left
3 ≤ fill volume ≤ 70 cm_)
(Fill with 5% non-ionic contrast solution.)
Patient setup_____________________________________________________________ Initial position _______________mm Length retracted/extended ______________mm Source position ________________mm (The value determined as the center) Balloon length _____________mm Balloon width ____________mm (L = W ± 3 mm) MEASUREMENTS FROM CT: Balloon size
______________mm (Measure diameter perpendicular to lumen.)
Balloon to skin
______________mm (Need at least 7 mm)
Balloon to rib
______________mm
Balloon to lung
______________mm
Balloon to heart
______________mm
Conformance
______________%
Off-axis distance
______________mm (The symmetry requirement is ± £ 3 mm.)
(Conformance must be at least 90%)
CALCULATED DISTANCES: Prescription 1
______________mm (Balloon radius + 10 mm)
Prescription 2
______________mm – (Balloon radius + 10 mm)
Source to skin
______________mm (Balloon radius + Balloon to skin)
Source to rib
______________mm – (Balloon radius + Balloon to rib)
Source to lung
______________mm – (Balloon radius + Balloon to lung)
Source to heart
______________mm – (Balloon radius + Balloon to heart)
Calculation point
______100_____mm
DOSIMETRY INFORMATION: D90 ________cGy D100 ________cGy V100 _______ V150 _______ V200 ________ Max skin dose ______________cGy
Dose Homogeneity Index _____________
Figure 2. A sample simulation work sheet.
et al. 2003). Measuring perimeters of the catheter in contact with air and converting to area is simpler than contouring air volumes, but the two approaches may yield different values of conformance. The conformance will normally improve over time as the air or fluid has a chance to resorb or otherwise dissipate. Another solution practiced at some institutions is to place a small diameter catheter alongside the balloon at the time of implant so that air or fluid pockets may be drawn down should they occur.
40–Partial Breast Irradiation Using the MammoSite® Radiation Therapy System
775
(a)
(b) Figure 3. An example of inadequate spacing between the surface of the balloon and the skin that was dramatically improved and made acceptable for treatment by rolling the patient into the decubitus position. (a) Unsatisfactory minimum skin spacing (5.1 mm) on the patient in the supine position. (b) Skin spacing increased to an acceptable 7.55 mm by moving the patient to a rolled position.
3-D CT Plan The 3-D plan [we use the Xio brachytherapy planning system from Computerized Medical Systems (St. Louis, MO)] is created by positioning the source on the dummy seed marked by the source position indicator or in the center of the balloon. The treatment time can be used from that calculated for the 2-D plan
776
Wayne M. Butler and Ernest G. Butler
(a)
(b) Figure 4. Conformance improves with the passage of time as air- and fluid-filled cavities dissipate. (a) The initial conformity is 90.8%, with an 8 cm3 air pocket. (b) One week later, the conformity is 96.4%, with a 3 cm3 air pocket.
or optimized from the 3-D planning program. The benefit of 3-D planning is the construction of dose volume histograms, which provide a more in-depth appreciation of the dose distribution. Furthermore, the effects of the balloon being asymmetrical, patient breathing artifacts, and the lumen being off-axis are much more dramatically represented when viewed in 3-D (see Figure 5). Dosimetric results are provided in Table 4 and demonstrate the differences that may occur between institutions based on balloon size preferences, planning software, etc.
2
1
5 7.5
41.7 9.4
Mean 2
42.7 0.3
47.4 3.7 88.6 9.8
— —
3269 3289
D90 (cGy) 1866 2270
D100 (cGy) 84 84
V100 (%)
Dosimetric Data
25 22
V150 (%) 2 1
V200 (%) 0.702 0.738
DHI
7105 5472
95.5 10.8
112.1 15.7
PTV from CT (cm3)
989.9 1.5
— — 13.0 7.0
8.5 3.0
Center Skin position distance (mm) (mm)
97.8 2.1
— —
Tissue conformance (%)
0.94 0.57
— —
Off-axis distance (mm)
3190.5 137.8
— —
D90 (cGy)
2137.6 143.7
— —
D100 (cGy)
81.1 5.1
84.9 12.7
V100 (%)
25.4 3.7
19.8 2.3
V150 (%)
2774 1475
4.5 2.0
0.5 0.5
V200 (%)
198 249
0.69 0.03
0.77 0.04
Dose Homogeneity Index
2086 1765
Max Skin Max Lt. Max Rt. Max Heart Dose Lung Dose Lung Dose Dose (cGy) (cGy) (cGy) (cGy)
Table 4. Dosimetric Data for the MammoSite Breast Brachytherapy Applicator PTV from width (cm3)
From Edmundson et al.(2002). From the Schiffler Cancer Center
(SD)
(SD)
58.3 12.7
Mean 1
107.8 114.0
PTV from CT (cm3)
Balloon width (mm)
Skin Gap from CT (mm)
Balloon Volume (cm3)
Supine Rolled Lt.
Position of Patient
Table 3. A Comparison of Dosimetric Data for the Same Patient in a Supine vs. a Rolled Position
40–Partial Breast Irradiation Using the MammoSite® Radiation Therapy System 777
778
Wayne M. Butler and Ernest G. Butler
(a)
(b) Figure 5. Some factors of concern are often more visible when viewed in various planar reconstructions from a 3-D dataset. There are substantial artifacts in the anatomy, and the radiopaque dummy source markers are due to the patient’s breathing period not being greater than the total scan time. (a) An asymmetrical balloon. (b) The catheter lumen displaced off axis.
40–Partial Breast Irradiation Using the MammoSite® Radiation Therapy System
779
Treatment Procedures The generally accepted prescription for a MammoSite treatment is 3.4 Gy per fraction performed twice daily for 10 fractions with a minimum 6-hour interval between treatments. Once the treatment is planned, the same plan can be used repeatedly because the geometry is highly consistent from day to day. However, treatment time adjustments are necessary due to radionuclide decay. The MammoSite device should be checked before every treatment to verify that the balloon has not leaked or changed shape and that the center is still in the same position. Comparing x-rays or even CT scout films can verify that the volume has not changed and that the source position indicator does not require any adjustments to be in the center. It is possible for the device to leak, or for the patient’s daily routine to affect the symmetry, or to cause the center to shift. If the length or width of the device decreases by more than 2 mm on both films, then the volume should be checked.
QA Procedures All HDR brachytherapy treatments require some form of daily quality assurance (QA) as detailed in previous chapters. Each physicist must follow institutional policy and procedures towards being in compliance with state and federal regulations. The treatment plan should always be verified by hand calculations. The dose rate from obtained from Table 2, the prescribed dose, the source strength, and the treatment time can be used as input in the manual check equations of the from illustrated in Figure 6. The variation determined by either of the two equations should be less than 5%. The newer ellipsoidal balloons require a more complex calculation because of the need to treat with multiple dwell positions. During treatment the patient should be observed with monitors [closed-circuit TV (CCTV)], and after treatment the patient must be checked with a survey instrument to be certain the source has been returned to the safe and that there is no patient contamination. It is essential to create and regularly test a plan in case of an emergency where the source needs to be manually removed. Removal of the MammoSite occurs after the 10th treatment fraction, typically in the HDR 192Ir treatment room. Patients may be administered an anti-anxiety reducing drug such as lorazepam (Ativan) 1 hour before this treatment, and an acute pain medication such as morphine sulfate (Roxanol) may be orally administered just prior to the last treatment. The balloon is deflated, and the fill volume is verified. The device is then pulled out with a swift motion. Most patients consider the removal only mildly unpleasant. Patients are given a follow-up schedule before they leave, and they should be accompanied by a friend or relative.
Fallback Procedure A procedure should be formulated in the event that a patient cannot undergo the full cycle of 10 treatments. If the balloon breaks or the patient’s situation or willingness to continue changes and removal is required, then an alternative therapy should be offered. Some treatment options to consider are the implantation of a new MammoSite device, conversion to interstitial breast brachytherapy, or whole or partial breast external beam therapy. Rare events do occur, and it is a good idea to brainstorm and prepare for the unexpected.
Follow-Up All patients should be photographed after the final treatment for future cosmetic evaluation. On a regular basis such as a 6-month follow-up schedule, patients should complete a quality-of-life survey and undergo a brief examination. Follow-up mammograms are necessary, and both breasts should be imaged on an annual basis one year after treatment completion.
780
Wayne M. Butler and Ernest G. Butler
Schiffler Cancer Center HDR Manual Calculation Check Breast Brachytherapy
Patient: ______________________________________
Date: _____________________ Treatment Number: _________
Using the Proxima tables:
Dose ( cGy )
Total Dwell Time (min) = Dose Rate
_____________(min) =
Variation = 1 −
cGy * Strength (Ci ) min⋅ Ci
_________(cGy) cGy ___________ * __________(Ci ) min⋅ Ci
Manual Calc Value ___________(min) =1− = ______________ % ___________(min) Computer Value
Using the TG-43 based distance equation:
Dose @10 cm ( cGy ) = 2.75 ⋅ 10
−–66 cGy * Strength (U ) * Time ( s ) U ⋅ s
cGy __________(cGy) = 2.75 ⋅ 10 −6 * _________( U ) * _________(s) U⋅s
Variation = 1 −
Manual CalcValue Computer Value
= 1 −
___________( cGy ) ___________( cGy )
= ______________ %
DO NOT PROCEED WITH TREATMENT IF THE VARIATION USING EITHER EQUATION IS > 5%.
Calculated by: _________________________
Checked by: _________________________
Figure 6. Checksheet for comparing manually calculated vs. computer-calculated parameters.
40–Partial Breast Irradiation Using the MammoSite® Radiation Therapy System
781
Ongoing and Future Development The results from the initial clinical experience with the MammoSite reported on 54 patients implanted, of which 43 were treated. These patients experienced mild to moderate side effects, such as skin erythema (57%), dry desquamation (13%), and moist desquamation (5%) (Keisch et al. 2003). These patients have had good to excellent short-term cosmesis. A presentation by Keisch et al., at the 2003 American Society for Therapeutic Radiology and Oncology (ASTRO) conference correlated the cosmetic outcome to the skin spacing from the balloon surface. The report also showed that the coverage and reproducibility was improved compared to interstitial breast brachytherapy in terms of D90; however, the dose homogeneity index (DHI) was lower with the MammoSite patients. Shah and colleagues (2004) compared toxicity and cosmesis in 75 women who received interstitial brachytherapy and a subsequent cohort of 28 women treated with MammoSite. The authors found mild skin erythema to be more common in MammoSite patients than those receiving interstitial treatments; however, subcutaneous toxicity was lower for the MammoSite patients. Although MammoSite is still a young, developing procedure, there have been some very interesting studies examining the device and optimization of its use. A couple of studies have examined the effects of contrast concentration on dose perturbation and demonstrated that for reasonable concentrations of contrast (around 10% or less) the attenuation was less than 3% regardless of the balloon’s treatment size (Kassa et al. 2004; Kirk et al. 2004). There have also been several reports and presentations regarding the use of multiple dwell positions with the MammoSite. The use of multiple dwell positions can improve dose coverage but often results in less uniform treatment—the D90 goes up but the DHI goes down (Dickler et al. 2004). The spherical balloon was originally intended to be a single dwell form of treatment; however, the new ellipsoidal balloons require multiple dwell positions. Additional areas of interest include other possibilities for the prescription or limiting doses for nearby critical structures. Finally, a randomized phase III trial has begun to compare survival, cosmetic, and quality of life between women undergoing conventional whole breast irradiation versus accelerated partial breast irradiation—Radiation Therapy Oncology Group (RTOG) Protocol 0413. There are three APBI arms that include intensity-modulated radiation therapy (IMRT), interstitial breast brachytherapy, and MammoSite brachytherapy. The pathologic tumor size must still be less than 30 cm3 for stage II tumors and the specimen must have clear margins, but the lumpectomy cavity may occupy up to 30% of the whole breast volume in lower stage disease and DCIS. Patients may be as young as 18 years, they may have up to three positive lymph nodes, and their disease may be multifocal.
References Arthur, D. W., F. A. Vicini, R. R. Kuske, D. E. Wazer, and S. Nag. (2002). “Accelerated partial breast irradiation: An updated report from the American Brachytherapy Society.” Brachytherapy 1:184–190. Baglan, K. A., A. A. Martinez, R. C. Frazier, V. R. Kini, L. L. Kestin, P. Y. Chen, G. Edmundson, E. Mele, D. Jaffray, and F. A. Vicini. (2001). “The use of high-dose rate brachytherapy alone after lumpectomy in patients with earlystage breast cancer treated with breast-conserving therapy.” Int J Radiat Oncol Biol Phys 50:1003–1011. Dickler, A., M. Kirk, J. Choo, W. C. Hsi, J. Chu, K. Dowlatshahi, D. Francescatti, and C. Nguyen. (2004). “Treatment volume and dose optimization of MammoSite breast brachytherapy applicator.” Int J Radiat Oncol Biol Phys 59:469–474. Edmundson, G. K., F. A. Vicini, P. Y. Chen, C. Mitchell, and A. A. Martinez. (2002). “Dosimetric characteristics of the MammoSite RTS, a new breast brachytherapy applicator.” Int J Radiat Oncol Biol Phys 52:1132–1139. Faverly, D. R., L. Burgers, P. Bult, and R. Holland. (1994). “Three dimensional imaging of mammary ductal carcinoma in situ: Clinical implications.” Semin Diagn Pathol 11:193–198. Imamura, H., S. Haga, T. Shimizu, O. Watanabe, J. Kinoshita, H. Nagumo, T. Kajiwara, and M. Aiba. (2000). “Relationship between the morphological and biological characteristics of intraductal components accompanying invasive ductal breast carcinoma and patient age.” Breast Cancer Res Treat 62:177–184.
782
Wayne M. Butler and Ernest G. Butler
Kassa, B., F. Mourtada, J. L. Horton, and R. G. Lane. (2004). “Contrast effects on dosimetry of a partial breast irradiation system.” Med Phys 31:1976–1979. Keisch, M., F. A. Vicini, R. R. Kuske, M. Herbert, J. White, C. Quiet, D. Arthur, T. Scroggins, and O. Streerer. (2003). “Initial clinical experience with the MammoSite breast brachytherapy applicator in women with early stage breast cancer treated with breast-conserving therapy.” Int J Radiat Oncol Biol Phys 55:289–293. Kirk, M. C., W. C. Hsi, J. C. H. Chu, H. Niu, Z. Hu, D. Bernard, A. Dickler, and C. Nguyen. (2004). “Dose perturbation induced by radiographic contrast inside brachytherapy balloon applicators.” Med Phys 31:1219–1224. Ohtake, T., R. Abe, I. Kimijima, T. Fukushima, A. Tsuchiya, K. Hoshi, and H. Wakasa. (1995). “Intraductal extension of primary invasive breast carcinoma treated by breast-conservative surgery. Computer graphic three-dimensional reconstruction of the mammary duct-lobular systems.” Cancer 76:32–45. Rivard, M. J., J. Sganga, G. A. Cardarelli, B. A. Bornstein, M. J. Engler, J. Tsai, S. A. Kaufman, T. A. DiPetrillo, R. Graham, and D. E. Wazer. (2003). “A quantitative methodology for measuring conformance index of MammoSite catheter treatments.” Brachytherapy 2:43. Shah, N. M., T. Tenenholz, D. Arthur, T. DiPetrillo, B. Bornstein, G. Cardarelli, M. J. Zheng, M. J. Rivard, S. Kaufman, and D. E. Wazer. (2004). “MammoSite and interstitial brachytherapy for accelerated partial breast irradiation: Factors that affect toxicity and cosmesis.” Cancer 101:727–734.
Chapter 41
Brachytherapy for Gynecological Malignancies John L. Horton, Ph.D. University of Texas M. D. Anderson Cancer Center Department of Radiation Physics Houston, Texas
Introduction Brachytherapy for gynecological cancer has a long and successful history. Wickham treated cervical cancer with radium as early as 1906 and reported results for 1000 patients by 1913 (Janeway 1919). Today, intracavitary brachytherapy combined with megavoltage external beam therapy is the standard treatment for cervical cancer. Five-year disease free survival for patients with cervical cancer treated with radiation therapy ranges from 70% to 90% for FIGO stages I & II, 25% to 48% for stage III, and 5% to 34% for stage IV (Perez, Brady, and Halperin 2004). Eifel and colleagues reported in a retrospective study of 3489 patients with cervical cancer treated with radiation therapy, a 3.3% incidence of major late bladder complications, 3.2%, for rectum and 4.2%, for small bowel (Eifel et al. 2002). In this section we will review the anatomy of the female pelvis, various gynecological cancerous diseases, and the staging of these diseases. We will reexamine the past, discuss the present, and provide our predictions of the future of brachytherapy for gynecological cancers. The early treatment for gynecological malignancies was with radium. In the 1970s 137Cs was widely adopted as a radium substitute. Applicators evolved and manual afterloading of implants became the standard replacing preloaded applicators. Applicators now include templates for interstitial implants for vaginal, vulvar, and more advanced cervical disease. The initial radium treatments were low dose rate (LDR) treatments with the patient confined to the hospital. Today, the majority of patients continue to receive LDR treatments with 137Cs; however, there is a continuing increase in the percentage of outpatients receiving high dose rate (HDR) treatments with remote afterloader units. With recent changes in the U.S. Nuclear Regulatory Commission requirements for pulsed dose rate (PDR) remote afterloading units, more attention and consideration is being directed at PDR as a replacement for LDR 137Cs. Early treatments were performed following semi-empirically derived guidelines and standardized implant rules based on dosimetry calculations for idealized implants with doses determined at geometrically identified points on plane radiographs serving as surrogates for tumor and critical structure doses. The International Commission of Radiation Units and Measurements (ICRU) developed recommendations in an attempt to standardize the dose reporting for brachytherapy implants. Today, a remote afterloading unit with a stepping source provides the ability to achieve a higher degree of dose conformation to the target volume with computerized optimization of the treatment plan. Imaging is an essential component for any meaningful computerized optimization. Three-dimensional imaging with CT and/or MR is becoming evermore important in the evaluation and planning of implants. These topics are discussed in the following chapters.
References Eifel, P., A. Jhingran, D. Bodurka, C. Levenback, and H. Thames. (2002). “Correlation of smoking history and other patient characteristics with major complications of pelvic radiation therapy for cervical cancer.” J Clin Oncol 20:3651–3657. Janeway, H. (1919). “The treatment of uterine cancer by radium.” Surg Gynecol Obstet 29:242–265. Perez, C., L. Brady, and E. Halperin (eds.). Principles and Practice of Radiation Oncology. Philadelphia: Lippincott Williams and Wilkins, 2004.
783
Chapter 42
Volume Imaging in Gynecological Brachytherapy Bruce R. Thomadsen, Ph.D. Department of Medical Physics University of Wisconsin, Madison, Wisconsin Reasons to Use Volume Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785 Delineation of Normal Tissue Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 785 Delineation of Target Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786 Dose Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786 Available Modalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786 Positron Emission Tomography (PET) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786 Proven Usefulness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 786 Treatment Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787 Ultrasound (US) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787 Usefulness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787 Treatment Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 787 Computed Tomography (CT) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788 Delineation of Dose-Limiting Normal Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 788 Delineation of Target Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789 Applicators for Use with CT Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 789 Treatment Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 790 Magnetic Resonance Imaging (MRI) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 790 Delineation of Normal Tissue Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 790 Delineation of Target Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 790 Spatial Distortion, Image Registration, and Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 791 MRI-Compatible Applicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793 GOG/GEC-ESTRO Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 793 Summary of Proposed GOG Guidelines (Nag et al., 2004) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 794 GEC-ESTRO Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 795
Reasons to Use Volume Imaging Conventional localization for gynecological intracavitary brachytherapy used plane radiography. While excellent for localizing and reconstructing source positions based on radiographic markers, radiographs cannot show soft tissue differences such as differentiation between target tissues and surrounding uterus, or even readily discern the neighboring structures such as the rectum or bladder. While contrast in the normal structures, such as barium in the rectum and iodine in the bladder, certainly helps to visualize these organs, reconstructing their contours in three dimensions still proved impossible. Only those imaging modalities that differentiate between various soft tissues allow true, three-dimensional characterization of the dose distribution with respect to targets and normal structures.
Delineation of Normal Tissue Structures Radiographic reconstruction required inference of normal tissue structures from skeletal anatomy or from markers or contrast studies. Many of the conventions used for determining the positions of normal tissue structures are summarized and illustrated in chapter 46, ICRU Reporting for Cervical Brachytherapy. As noted in that chapter, using a Foley bulb to define a point to represent the dose to the bladder, or a point based on the vaginal wall for the rectum, fails to give representations of the maximum dose to those organs.
785
786
Bruce R. Thomadsen
The bladder tends to drape over the uterus with some superior aspects falling closer to the sources than the trigone, where the Foley bulb sits. Filling the bladder with contrast does show the volume of the bladder as it projects in the image plane. However, none of the triangulation reconstruction methods (such as orthogonal or shift images) can deduce the irregular shape of the organ in three dimensions. The rectum fares a bit better because its shape tends to be a little more regular than the bladder. With experience (and it takes a considerable amount of experience), one can determine the path of the rectum on a pair of images, its anterior-most aspect along its path. As the rectum moves to one side of the uterus, though, finding an oblique point that falls closest to the applicator becomes similar to finding a point on the bladder surface: essentially impossible. One of the locations at greatest risk for complications is the superior bowel. Even with energetic instillation of contrast, the superior bowel cannot be opacified.
Delineation of Target Tissues As poorly as radiographic methods depict the shapes of normal structures, the delineation of the target tissues is even worse. The uterus, tumor tissue, and the periuterine pelvic tissues all appear the same on radiographic images, allowing no differentiation, let alone establishing any surface contours.
Dose Distributions Three-dimensional dose distributions have been common in brachytherapy for many decades—long before external-beam dosimetry incorporated the necessary tools. However, the three-dimensional dose distributions did not relate to any targets or normal structures other than individually identified points, such as marked by surgical clips. Thus, with conventional, radiographic reconstruction, the capabilities of dose calculation failed to fulfill its potential. Only with volume imaging modalities that differentiate types of soft tissues and can still also depict the positions of sources can three-dimensional dose calculations become three-dimensional dosimetry.
Available Modalities Several imaging modalities provide volume information on patient anatomy. No modality provides everything desired for three-dimensional dosimetry. This situation mirrors the same problem facing external-beam treatment planning, and will be dealt with after a discussion of the various imaging modalities relevant to intracavitary brachytherapy. The mechanisms by which each modality images the patient fall outside the scope of this chapter and can be found in modern imaging science textbooks. This chapter assumes a familiarity with the imaging modalities on the part of the reader, and will discuss their mechanisms only as necessary to understand the nature of the images.
Positron Emission Tomography (PET) Proven Usefulness Fluorodeoxyglucose (FDG) tagged with18F has proven useful in cervical cancer evaluations in assessing metastases to pelvic and para-aortic lymph nodes metastases from cervical carcinoma (Rose et al., 1999; Grigsby, Siegel, and Dehdashti 2001), and tumor volume assessment, particularly before and after radiation (Narayan et al, 2001; Miller and Grigsby 2002), although once radiation damage begins in the tissues, the utility of PET scans diminishes (Nakamoto et al. 2002).
42–Volume Imaging in Gynecological Brachytherapy
787
Treatment Planning Mutic et al. (2002) demonstrated that treatment planning and dose calculations could be performed using only PET images. Standard applicators can be used, a distinct advantage over other volume imaging modalities (see below). Such localization requires positron-emitting dummy sources for localization, which are not difficult to manufacture. The resolution limits and uncertainty of positioning of such dummy source indicators fall outside the minimum tolerances usually accepted as adequate for dose calculation. However, the geometry of the source indicators would be well known; their true positions could be projected with adequate precision. As the role of PET expands in cervical cancer treatment planning, such localization techniques may come into play.
Ultrasound (US) US plays a dominant role in prostate brachytherapy, in target delineation, applicator placement guidance, and dosimetry. Some of these functions transcend into gynecological intracavitary brachytherapy. Usefulness US provides invaluable assistance in guiding the insertion of tandems into the uterine canal, particularly when the cervical os is obscured by tumor or in cases of retroverted uteri. Figure 1 shows examples of such images. Treatment Planning Even in the imaging context discussed, US images cannot determine the position of sources inside intracavitary applicators. Any of the currently used applicators shield the contents of the applicator lumen from the US waves, preventing imaging of source position indicators. This problem could be solved simply by knowing well the relative geometries of the applicator and the source positions in the applicator. However, beyond the problem of localizing the source positions, current external US units mostly provide twodimensional information in free-hand oriented planes, with no way to produce stacked images from which volume images could be constructed. Interestingly, very old b-scan units that produced imaging using three articulating arms often did just that—produced images in planes perpendicular to a screw drive that moved
(a)
(b)
Figure 1. Ultrasound images used for guidance during insertion of a tandem. [Image (a) courtesy of Scott Tannehill, St. Mary’s Medical Center, Milwaukee, WI. Image (b) courtesy of Daniel Petereit, Rapid City Regional Cancer Center, Rapid City, SD.]
788
Bruce R. Thomadsen
the unit along the patient’s body axis. For prostate implants, the transrectal US probes move in one dimension, producing planar images perpendicular to that motion, allowing three-dimensional reconstruction. There are newer US units that produce three-dimensional images by sweeping the ultrasound beam, mechanically or electronically, though a block of tissue (Rivard, Evans, and Kay 2005). The volume imaged tends to be a little small for intracavitary dosimetry, but in time that limitation is likely to fall away. In the situation for prostate implants, the distance from the probe to the region of interest remains small, on the order of 5 cm. Small distances allow imaging with high frequencies that yield higher resolution images. Using external transducers for cervical imaging requires lower frequency probes and result in lower resolution imaging, compromising again the dosimetric quality of the images. Transrectal ultrasound would not solve the problem because the uterus falls much farther superior compared to the relative position of the prostate. At these superior positions, the rectum diverges from the uterus, again increasing the distance from the probe to the structures of interest. While US is very useful in treatment planning in the larger sense of the term, and can more readily determine target locations, it is not likely that US will serve as an imaging modality for dosimetry without major advancements in ultrasound technology that drive subsequent development of US-based reconstruction.
Computed Tomography (CT) Computed tomography (CT) has served as the workhorse imaging modality for modern external-beam treatment planning. Three reasons explain the widespread use of CT for this purpose: 1. CT provides relatively high-resolution images that generally show sufficient soft-tissue contrast to delineate targets and normal structures, presented with little distortion. 2. CT images yield a quantity that relates to radiographic density—a quantity necessary for the calculation of the dose distribution. 3. CT scanners have become common enough, and the scans inexpensive enough, for use on all patients. While magnetic resonance imaging (MRI) (discussed below) may improve the first item (resolution) albeit with questionable distortion, the images fail to provide radiographic parameters and the availability remains relatively scarce as of this writing. With all this in mind, the discussion below considers the application of CT for treatment planning of cervical intracavitary brachytherapy. Figure 2 presents a typical CT image of a female pelvis. Delineation of Dose-Limiting Normal Structures CT provides a very good delineation of the bladder and rectum, two of the main neighboring normal structures of the uterus (Schoeppel et al., 1994). While CT also depicts the superior bowel well, identifying particular parts of the bowel, both on a given treatment and between treatments proves difficult at best. Instilling dilute contrast into the bladder often assists in differentiating it from the surrounding tissues, although excessive concentrations can produce artifacts. Filling the bladder obviously changes the position of the bladder in the pelvis, and moves the other pelvic structures. While true, that the posterior wall of the bladder essentially lies in contact with the uterus either full or empty, the rest of the bladder, particularly the sides, can change position markedly with respect to the uterus. Of course, the anterior wall moves well away for the uterus when full. Thus, if dose distributions are calculated with the bladder full, the treatment should keep the bladder full. Doing so changes little the dose to the posterior wall, but reduces the dose to the anterior wall greatly, thus providing a benefit as well as facilitating visualization. This practice only applies to high dose rate (HDR) treatments, since the bladder could not be kept at a particular
42–Volume Imaging in Gynecological Brachytherapy
789
Figure 2. A CT image of a female pelvis.
filling volume over the 2 or 3 days of low dose rate (LDR) treatment. For LDR treatments, the indwelling bladder catheter keeps the bladder empty, and the dose-calculation imaging should also apply to that situation for consistency. Rectal imaging poses slightly different problems. The rectum, while running just under the vagina, does not stay in contact with the uterus, but diverges posteriorly at about the level of the cervix. Contents in the rectum often will move the anterior rectal wall nearer to the uterus and the posterior wall away. Contents in the rectum generally just degrade the certainty of the position of the rectum, particularly since the contents move with time. Standard practice is to cleanse the bowel before LDR treatments and to place the patient on a low residue diet during treatment, although that does not assure that the rectum will remain content free. Imaging the empty rectum becomes much more difficult than when it contains solid or gaseous contents. When the walls collapse, the radiographic density of the rectum becomes very similar to the surrounding pelvic tissues. A very light coating of contrast (for example 5% barium agent in water) instilled into the rectum, and then withdrawn, can provide sufficient tissue contrast to allow identification of the rectal lumen. Delineation of Target Tissues For the most part, CT does not differentiate between the uterus, uterine tumors, and the surrounding pelvic contents. This lack of differentiation makes CT particularly unsuitable for determining target volumes. CT does depict grossly involved lymph nodes, but distinguishing apparently normal lymph nodes from other structures often proves challenging. Applicators for Use with CT Imaging Because of the metal parts, conventional intracavitary appliances such as tandems and colpostats, either for LDR or HDR applications, produce considerable artifacts on CT images, to the extent of obscuring rectal and bladder contours and prohibiting determination of the source positions within the appliance.
790
Bruce R. Thomadsen
Special applicators have been developed compatible for CT imaging (Schoeppel et al., 1989). Applicators compatible with CT imaging use low atomic-number materials to minimize photoelectric interactions and the subsequent sudden attenuation discontinuity at the tissue/applicator interface. Making a uterine tandem of a low atomic-number material becomes a challenge because, in general, such materials are not as strong as metals, which tend toward the higher atomic numbers. Graphite is an exception, which can form long and relatively strong tubes. However, the tube walls must be thicker than for metals. For example, commercially available graphite tandems for HDR applications are 7 mm in diameter compared to the normal 3 mm metallic units. Plastics exhibit this size inflation to a greater extent than graphite. Some metals with a relatively low atomic number, potentially titanium, may be able to be made thin enough while strong enough not to produce serious artifacts in CT images, but there are none available at this writing. For LDR applications, developers may wish to consider using LDR iridium sources instead of cesium sources to reduce the required diameter of the inner lumen. The vaginal portion of the appliance poses little difficulty in construction for CT compatibility. Size is less of a limiting factor, and there are already ovoids with afterloading shielding (Weeks, Montana, and Bentel 1991). Treatment Planning Treatment planning and dosimetry have been carried out using CT reconstruction for cervical intracavitary brachytherapy. Schoeppel evaluated the dose distribution for 10 patients using CT imaging. Schoeppel et al. (1994) Found that the maximum bladder dose fell an average of 2.3 cm superior to the International commission on Radiation Units and Measurements (ICRU)-defined point and the maximum rectal dose 1.3 cm superior. They also identified the cervix (noting that identification is easier on sagittal reconstructions) and found that the minimum dose averaged 0.6 of the point A dose, and in their series, was always less.
Magnetic Resonance Imaging (MRI) Magnetic resonance (MR) images tissues based on nonradiological properties, unlike CT. Because the images are based on the aligned-spin relaxation times, very different images result from the variety of settings and protocols that can be used. Some of the imaging sequences differentiate various tissues of interest in the female pelvis. Figure 3 displays an MR image of the cervix. Delineation of Normal Tissue Structures As with CT, MR distinguishes well the bladder and rectum, and thus can serve for evaluating the dose distribution to these organs. However, the bony anatomy may not be as evident on MR as with CT. Delineation of Target Tissues The advantage of MRI for cervical treatment planning lies in its ability to image the uterus and to differentiate between the uterus and uterine tumors (Pötter et al. 1991; Schoeppel et al., 1992; Hricak and Yu 1996). It has been shown that MRI can well demonstrate the extent of cervical cancer (Burghardt et al., 1989; Lien et al., 1993; Yu et al., 1995; Subak et al., 1995). MRI for cervical cases provides the best images using T2-weighted sequences, where the mucosa of the endocervix produces a high signal intensity, the stroma a low signal intensity, and the peripheral cervical tissue an intermediate signal intensity; cervical cancer is seen as a very intense signal (Nag et al., 2004). Fast spin echo imaging reduces motion artifact. The following description of cervical applications using MRI is summarized from Hawighorst et al. (1998), Hricak et al. (1991), Scheidler et al. (1998), and Sironi et al. (1993).
42–Volume Imaging in Gynecological Brachytherapy
791
Figure 3. An MR image in the region of the uterine cervix. [Image courtesy of Jason Rownd, Medical College of Wisconsin, Milwaukee, WI.]
On gadolinium contrast-enhanced T1-weighted images, the inner mucosal epithelium and the pericervical tissue enhance more than the inner cervical stroma. The parametrium, connective soft tissue that is adjacent and lateral to the uterus and not covered by peritoneum, is vascular and contains many efferent lymphatics. The uterine arteries and the distal ureters pass through the lateral parametrial tissues. The parametrium demonstrates intermediate signal intensity on T1-weighted images and varying degrees of high signal intensity on T2-weighted images. With dynamic contrast imaging [gadolinium chelates 0.1 mmole/kg], cervical cancer shows early contrast enhancement. Although contrast enhancement is also helpful in differentiating viable tumor from debris and areas of necrosis, it has not been shown to improve overall staging accuracy. Contrast enhancement may also assist in assessment of bladder or rectal involvement. T1 images and T1-weighted-fat saturation sequence help differentiate soft tissue and fat, and assess for lymphadenopathy. The appendix to Nag et al. (2004) gives a protocol, essentially a step-by-step guide, for MRI for cervical cancer brachytherapy. Table 1, adapted from that article, summarizes the images required for treatment planning. Spatial Distortion, Image Registration, and Fusion Treatment planning with MRI requires caution because the images may exhibit distortion, particularly near the edges of the patient due to the rapid change in magnetic susceptibility between the air and the patient tissues. This poses less of a problem for brachytherapy than external-beam treatments because most applications fall far from the skin. However, prudence dictates assurance of the position of structures.
pelvic
pelvic
pelvic
pelvic
1. FSE T2
2. SE T1
3. FSE T2
4. FSE T2
pelvic
pelvic
pelvic
1.SSFSE T2
2. FSE T2
3. SE T1 post Gd
OPTIONS
COIL
SEQUENCE
Sag+ax
Obl Cor
Sag+/-Ax
Sagittal
Axial
Axial
Sagittal
PLANE
5-700
4000
∞
4000
4000
5-700
2000
TR (ms)
10-15
85
100
85
85
10-15
85
TE (ms)
8
8
8
FLIP ETL
16
16
62.5
16
16
16
16
BW (kHz)
24-28
24-28
24-32
24-28
24-28
24-28
36-48
FOV (cm)
5
5
5
5
5
5
10
SLICE (mm)
1
1
1
1
1
1
0
GAP
256/256
512/256
256/256
512/256
512/256
256/256
256/256
MATRIX Frq/phase
Table 1. Protocol for Cervical Carcinoma MRI for Staging and Brachytherapy
2
2
0.5
2
2
2
1
NEX
St:SIap RC Post Gd
St:Siap SPF NPW
St:SIap SPF NPW
StSIap RC
IMAGING OPTIONS
NPW for sag. Sequence, if suspect bladder & rectal involvement.
If cervical mass not well seen on ax or cor.
Breath hold salvage T2.
Frq AP to reduce and wall motion.
Frq AP to reduce and wall motion artifact.
Extend to renal hilum if pelvic nodes present.
Localizer.
COMMENTS
792 Bruce R. Thomadsen
42–Volume Imaging in Gynecological Brachytherapy
793
Ameliorating the potential distortion problem usually comes from comparing MR images with those from CT. Meaningful comparison of the two images sets requires juxtaposition, a process called “registration,” where both image sets display simultaneously on a computer display and are translated and rotated until corresponding structures in each image fall on top of each other. This congruence requires that the patient assume the same position during both imaging sessions. While this condition never truly holds, immobilization molds, such as commonly used in external-beam treatments, minimize differences in positioning. Most modern treatment-planning systems have utilities that facilitate image registration and the combining of information from each image set (fusion), although some do not enable these capabilities in brachytherapy calculation modules. Assumedly, this situation will soon improve with time. Following image fusion, MR image distortion can be identified and targets and structures of interest can be contoured using the most relevant image information. In the absence of CT images or of image fusion software, treatment planning for intracavitary brachytherapy can proceed using just the MR image. In the event that the treatment-planning software will not permit the use of MR images, see the next section, GOG/GECESTRO Recommendations1. MRI-Compatible Applicator Similar to CT, MRI requires specially designed applicators. The requirement for MRI compatibility is the absence of all ferromagnetic materials. Practically, this mostly becomes similar to the CT-compatible applicators, using graphite and plastics. This criterion allows some metals, for example, titanium and tungsten. As noted above, titanium may be compatible with CT imaging if thin enough. Given the slight difference in the requirements for the two modalities, emerging applicators likely will satisfy either. As of this writing, commercially available MRI- and CT-compatible appliances remain very expensive—a major deterrent to their wider use. Some of the applicators have major design problems. For example, one HDR model comes with the typical set of small, medium, and large ovoid diameters. However, as the diameter increases, the source track always falls 1 cm from the later surface of the ovoid. Such an arrangement not only fails to provide the increased penetration that normally accompanies larger ovoids, but can result in a markedly lower dose to the cervix than delivered using more standard designs.
GOG/GEC-ESTRO Recommendations Volume imaging adds important information to treatment planning for cervical cancer intracavitary brachytherapy in establishing the dose distributions to imaged tumor and to normal structures. However, it probably is premature to begin prescribing treatments based on tumors as visualized on MRI for two reasons: 1. The dose needed to control cervical tumors is not known. 2. The actual target is not known. Amazing as it may seem after a century of cervical intracavitary brachytherapy, these two fundamental aspects of the treatment remain unknown. The reason behind the ignorance is clear: because for the vast preponderance of the history of the treatment modality, the true patient anatomy remained unknown in each case. MR images now show tumors, and it is tempting to approach these treatments as treatment planning for external-beam or interstitial brachytherapy, by defining the target based on the tumor and prescribing the customary point A dose to the tumor. The first problem with this approach is that through the years while doses were prescribed to point A, the extent of the tumor, and patient anatomies, have all been different. In some patients, the tumor dose may have been higher than the point A dose (for example, if the cervix 1
Gynecological Oncology Group/Groupe Européen de Curiethérapie-European Society of Therapeutic Radiology and Oncology.
794
Bruce R. Thomadsen
were small and the tumor confined to the cervix), or lower (for larger cervixes or tumors). The control rates for prescribing the dose to point A are known, but some of the failures may have been for tumors that extended beyond point A. Notwithstanding, because very little data exist to establish the dose that controls cervical tumors, prescribing treatments as a dose to the imaged tumor is risky at the present. Complicating determining the dose to prescribe to the tumor as a target, the success of intracavitary brachytherapy in treating cervical cancer may also depend on the peculiar dose distribution characteristic of a line source. Unlike interstitial brachytherapy or external-beam radiotherapy when the dose distribution is characterized by a region of mostly uniform dose, surrounded by a very rapid decrease, intracavitary dose distributions deliver very high doses very close to the sources and carry significant doses well beyond any conceivable target tissue. Thus, it becomes unclear that the single target, such as the distal edge of the visible tumor, carries all the important prescription information. And then, maybe it does. At the present, we just do not know. To try to prevent this situation from perpetuating, and to forestall treatment failures due to premature target-based prescription, a multi-organizational committee, led by the Gynecological Oncology Group (GOG), drafted suggested guidelines for volume-imaging based treatment planning, particularly for researchers and cooperative groups, to facilitate the gathering and exchange of data that may lead to establishing eventual guidance for future, image-based prescription. The first tenet of the guideline is that treatment prescription should follow the current practice, and doses to target tissues and normal structures would be calculated based on the volume images. In this way, the dosimetry resulting from the current practices can be determined. Target delineation may only use MR images. Bladder and rectal delineation may be performed using either MR or fused MR:CT images. A summary of the guidelines follows.
Summary of Proposed GOG Guidelines (Nag et al., 2004) 1. Treatments should be prescribed as in the current practice, and the integrated reference air kerma and the dose to point A as defined by the American Brachytherapy Society (ABS) (Nag et al. 2000; Nag et al. 2002) should be reported. 2. Cooperative Groups collect the data listed below and correlate the information with patient outcomes. 3. MRI is to be used for image-based cervical brachytherapy. T2-weighted MRI using a pelvic surface coil should be performed with the patient in the treatment position with brachytherapy applicators in place for intracavitary implants. PET/CT fusion should be investigated for treatment planning. 4. The following information should be gathered on each patient: a. Target Dosimetry: DVH of GTVI andGTV, as well as the D100, D95, D90 (dose to 100%, 95%, or 90% for the various GTV and CTV) and V100 (% of GTV covered by point A dose) using the following definitions: i. GTVI: gross tumor volume as defined through imaging plus any palpable or visual tumor. ii. GTV: the GTVI plus the entire cervix. iii. pCTV: the primary tumor clinical target volume, which equals the GTV plus a 2-cm margin. iv. rCTV: the pCTV plus a 1.5-cm margin around regional lymph nodes. v. CTV: the pCTV and the rCTV, all of which are to be included in the external beam radiotherapy field.
42–Volume Imaging in Gynecological Brachytherapy
795
b. Normal Structure Dosimetry: i. absolute DVH of the organ wall alone, which requires contouring of the inner and outer wall of the bladder and rectum. ii. for bladder doses: the ICRU bladder point dose, the maximum bladder dose, and the maximum doses to contiguous 1 cc and 5 cc volumes of bladder (BladderV1cc, BladderV5cc). iii. for rectal doses: the ICRU rectal point dose, maximum rectal dose, and the maximum doses to contiguous 1 cm3 and 5 cm3 volumes of rectum (RectumV1cc, RectumV5cc) are to be reported. iv. for small bowel doses: the maximum doses to contiguous 1 cm3 and 5 cm3 volumes of small bowel (Small BowelV1cc, Small BowelV5cc) are to be reported. Several issues remain unclear in the GOG protocol. For example, with fractionated HDR brachytherapy, should the target volumes be defined for each fraction, at the beginning of treatment, or before any external beam? This group and others are still grappling with these questions.
GEC-ESTRO Protocol The European brachytherapy group, GEC-ESTRO, is developing a similar set of guidelines. While many of their recommendations are the same or very similar to those of the GOG, some important differences exist, particularly in the definition of target volumes, on some doses to report, and on volumes of normal structures considered to be significant when irradiated to very high doses. Both groups are looking to harmonize the recommendations toward promoting consistency and advancing our knowledge of the necessary prescription dose for volume-based treatment planning of cervical intracavitary brachytherapy.
References Burghardt, E., H. M. Hofmann, F. Ebner, J. Haas, K. Tamussino, and E. Justich. (1989). “Magnetic resonance imaging in cervical cancer: A basis for objective classification.” Gynecol Oncol 33:61–67. Grigsby, P, B. Siegel, and F. Dehdashti. (2001). “Lymph node staging by positron emission tomography in patients with carcinoma of the cervix.” J Clinical Oncol 19:3745–3749. Hawighorst, H., S. O. Schoenber, P. G. Knapstein, M. V. Knopp, U. Schaeffer, M. Essig, and G. van Kaick. (1998). “Staging of invasive cervical carcinoma and of pelvic lymph nodes by high resolution MRI with a phased-array coil in comparison with pathological findings.” J Comput Assist Tomogr 22:75–81. Hricak, H., B. Hamm, R. C. Semelka, C. E. Cann, T. Nauert, E. Secaf, J. L. Stern, and K. J. Wolf. (1991). “Carcinoma of the uterus: Use of gadopentetate dimeglumine in MR imaging.” Radiology 181:95–106. Syed, N, Puthawala, A, Interstitial implant techniques, in Brachytherapy into the 21st Century, Endocurietherapy Research Foundation, Long BeachHricak, H., and K. Yu. (1996). “Radiology in invasive cervical cancer.” Am J Roentgenol 167:1101–1108. Lien, H. H., V. Blomlie, T. Iversen, C. Trope, K. Sundfor, and V. M. Abeler. (1993). “Clinical stage I carcinoma of the cervix: value of MR imaging in determining invasion into the parametrium.” Acta Radiol 34:130–132. Miller, T., and P. Grigsby. (2002). “Measurement of tumor volume by PET to evaluate prognosis in patients with advanced cervical cancer treated by radiation therapy.” Int J Radiat Oncol Biol Phys 53:353–359. Mutic, S., P. Grigsby, D. Low, J. Dempsey, W. Harms, R. Laforest, W. Bosch, and T. Miller. (2002) “PET-guided three-dimensional treatment planning of intracavitary gynecologic implants.” Int J Radiat Oncol Biol Phys 52:1104–1110. Nag, S., H. Cardenes, S. Chang, I. Das, B. Erickson, G. Ibbott, J. Lowenstein, J. Roll, B. Thomadsen, and M. Varia. (2004). “Proposed guidelines for image-based intracavitary brachytherapy for cervical carcinoma: Report from Image-Guided Brachytherapy Working Group.” Int J Radiat Oncol Biol Phys 60:1160–1172.
796
Bruce R. Thomadsen
Nag, S., C. Chao, B. Erickson, J. Fowler, N. Gupta, A. Martinez, B. Thomadsen; American Brachytherapy Society. (2002). “The American Brachytherapy Society recommendations for low-dose-rate brachytherapy for carcinoma of the cervix, Int J Radiat Oncol Biol Phys 52(1):33–48. Nag, S., B. Erickson, B. Thomadsen, C. Orton, J. D. Demanes, and D. Petereit. (2000). “The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for carcinoma of the cervix.” Int J Radiat Oncol Biol Phys 48(1):201–211. Nakamoto, Y., A. Eisbruch E. D. Achtyes, Y. Sugawara, K. R. Reynolds, C. M. Johnston, and R. L. Wahl. (2002). “Prognostic value of positron emission tomography using F-18-fluordeoxyglucose in patients with cervical cancer undergoing radiotherapy.” Gynecol Oncol 84:289–295. Narayan, K., R. J. Hicks, T. Jobling, T. Bernshaw, and A. F. McKenzie. (2001). “A comparison of MRI and PET scanning in surgically staged loco-regionally advanced cervical cancer: Potential impact on treatment.” Int J Gynecol Cancer 11:263–271. Pötter R, Kovacs G, Lenzen B, et al. (1991). “Technique of MRI assisted brachytherapy treatment planning.” Brachytherapy Journal 5:145–148. Rivard, M. J., D-A. R. Evans, and I. A. Kay. (2005). “A technical evaluation of the Nucletron FIRST® system: Conformance of a remote afterloading brachytherapy seed implantation system to manufacturer specifications and AAPM Task Group report recommendations.” J Appl Clin Med Phys 6:22–50. Rose, P., L. Adler, M. Rodriquez, P. F. Faulhaber, F W. Abdul-Karim, and F. Miraldi. (1999). “Positron emission tomography for evaluating para-aortic nodal metastasis in locally advanced cervical cancer before surgical staging: A surgicopathologic study.” J Clin Oncol 17(1):41–45. Scheidler, J., A. F. Heuck, M. Steinborn, R. Kimmig, and M. F. Reiser. (1998). “Parametrial invasion in cervical carcinoma: Evaluation of detection at MR imaging with fat suppression.” Radiology 206:125–129. Schoeppel, S. L., J. H. Ellis, M. L. LaVigne, R. A. Schea, and J. A. Roberts. (1992). “Magnetic resonance imaging during intracavitary gynecologic brachytherapy.” Int J Radiat Oncol Biol Phys 23(1):169–174. Schoeppel, S. L., B. A. Fraass, M. P. Hopkins, M. L. La Vigne, A. S. Lichter, D. L. McShan, S. Noffsinger, C. PerezTamayo, and J. A. Roberts. (1989). “A CT-compatible version of the Fletcher system intracavitary applicator: clinical application and 3-dimensional treatment planning.” Int J Radiat Oncol Biol Phys 17:1103–1109. Schoeppel, S. L., M. L. LaVigne, M. K. Martel, D. L. McShan, B. A. Fraass, and J. A. Roberts. (1994). “Three-dimensional treatment planning of intracavitary gynecologic implants: analysis of ten cases and implications for dose specification.” Int J Radiat Oncol Biol Phys 28:277–283. Sironi, S., F. De Cobelli, G. Scarfone, E. Colombo, G. Bolis, A. Ferrari, and A. DelMaschio. (1993). “Carcinoma of the cervix: Value of plain and gadolinium-enhanced MR imaging in assessing degree of invasiveness.” Radiology 188:797–801. Subak, L., H. Hricak, B. Powell, L. Azizi, and J. L. Storm. (1995). “Cervical carcinoma: Computed tomography and magnetic resonance imaging for preoperative staging.” Obstet Gynecol 86(1):43–50. Weeks, K. J., G. S. Montana, and G. C. Bentel. (1991). “Design of a plastic minicolpostat applicator with shields.” Int J Radiat Oncol Biol Phys 21(4):1045-1052. Yu, K. K., H. Hricak, L. L. Subak, C. J. Zaloudek, and C. B. Powell. (1998). “Preoperative staging of cervical carcinoma: Phased array coil fast spin-echo versus body coil spin-echo T2-weighted MR imaging.” Am J Roentgenol 171(3):707–711.
Chapter 43
Applicator Design and Dose Distributions Jason Rownd, M.S. Medical College of Wisconsin Milwaukee, Wisconsin Brachytherapy Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797 Tandem and Ovoid Applicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 797 Tandem and Ring Applicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 798 Tandem and Cylinder Applicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 799 CT/MR-Compatible Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 800 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 804 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 804
Brachytherapy Applicators High dose rate (HDR) brachytherapy applicators, of which there are many, have several intrinsic properties that make them suitable for successful implants. The specifics of material and construction are beyond the scope of this chapter. Applicators should be sturdy and robust since they will be used, sterilized, and reused countless times over the course of years. Applicators should be easy to insert and assemble within the patient. The relative comfort of the insertions is patient dependent but often involves sedation, depending also on the specific procedure. Applicators should be easily visualized by radiographs, or at least have suitable markers available to aid in the computer reconstruction of the implant for dosimetry. The standard steel and plastic applicators fit these criteria and the addition of x-ray markers or dummies allows for accurate applicator reconstruction. The carbon fiber and plastic nature of the computed tomography/magnetic resonance (CT/MR)-compatible applicators improve on some of the imaging aspects at the expense of other issues. They are less robust, larger (and therefore more uncomfortable), and harder to image with radiographs (a side effect of their improved compatibility with other imaging modalities). This improved compatibility justifies the other restrictions since it will allow for better visualization through multiple modalities.
Tandem and Ovoid Applicator The tandem and ovoids applicator (Figure 1) is a typical applicator used for treatment of cervical cancer with HDR brachytherapy. The tandem passes through the cervical os and extends into the uterine canal nearly the full length of the uterus. The colpostats or ovoids are placed laterally on both sides of the tandem, resting within the vaginal fornices. The proper insertion of the tandem and ovoids applicator is as much an art as it is a science. The position of the tandem axis relative to the midline of the ovoids is dependent on the patient’s anatomy and the physician’s skill in placing the applicator. Additionally, rectal retractors, vaginal packing, and other aids to the implant procedure can dramatically affect the applicator placement and the resulting dose, both to the tumor and to normal tissues, such as the rectum and the bladder. The major dose to the uterus is delivered from the tandem with the vaginal surface dose coming from the ovoids. These sections of the applicator are not completely separated with respect to their dose contributions to the implant. The proximal dwell positions in tandem will contribute some dose to the vaginal mucosa and the ovoid dwell positions will contribute some dose to the uterus. Typical prescriptions specify an amount of dose to point A, with an additional prescription constraint to the vaginal surface dose. Implant-specific applications of this general prescription concept can lead to some unintentional differences in delivered dose. Depending on the position from where one measures the 2-cm superior and 2-cm
797
798
Jason Rownd
Figure 1. HDR tandem and ovoids applicator.
lateral shifts to describe point A, the actual location can fall within a high gradient dose region where small changes in distance can have significant changes in delivered dose. This unintended variation in delivered doses for similar implants has resulted in the American Brachytherapy Society standardizing the location of the prescription point A (Nag et al. 2000). Point A should be referenced from the center of the dwell positions of the ovoid, located superiorly 2 cm plus the radius of the ovoid cap and 2 cm laterally from the tandem axis. This localization method correctly assumes that the centers of the ovoids, as defined by radiographic dummies, are more easily visualized by radiographs than the surface of the ovoids next to the vaginal surface. Also, the largest ovoid that can be accommodated by patient anatomy should be used to get the optimal dose distribution.
Tandem and Ring Applicator The tandem and ring applicator is one of the more popular applicators for HDR brachytherapy treatments of cervical carcinomas (Figure 2). With proper consideration, it can be thought of as a suitable alternative to the more traditional tandem and ovoids applicator (Erickson et al. 2000). The usual prescription point is also point A, as mentioned above. Similarly, in the case of a tandem and ring insertion, the superior shift to locate point A is 2 cm plus the ring cap thickness, starting from the plane of the ring dwell positions. Patients with narrow proximal vagina, or partial or complete loss of the vaginal fornices would have suboptimal geometry for the tandem and ovoid applicator. If these restrictions are too severe, then a tandem and cylinder applicator could be justified. In more moderate conditions, the tandem and ring provides a pear-shaped dose distribution that is similar to a tandem and ovoid applicator (Figure 3). In addition, the tandem and ring applicator has strong reproducibility and ease of insertion due to the designed geometry fixing the tandem to the center of the ring section of the applicator, unlike the tandem and ovoid applicator. This fixed geometry yields a predictable isodose distribution relative to the physical applicator for each insertion. However, the dose distribution with respect to normal tissues is influenced by the location and geometry of these normal tissues relative to the applicator. Specifically, the rectum and sigmoid tissues can move significantly between fractions, altering the isodose distribution within those tissues. Another benefit of the tandem and ring applicator geometry is variable activation of ring dwell positions anteriorly and posteriorly if required by disease progression (Figure 4). One dosimetric difference between the tandem and ovoid and the tandem and ring applicators, independent of the fixed-geometry considerations, is the relative depth dose distribution in the vaginal mucosa. The ovoids have a standard cap thickness of 10 mm from the source dwell location to the cap surface, whereas the ring has a standard thickness of 6 mm from the source dwell location to the ring surface. The
48–Applicator Design and Dose Distributions
799
Figure 2. HDR tandem and ring applicator.
Figure 3. Pear-shaped isodose distribution of an HDR tandem and ring treatment.
increased distance of the ovoid creates an increased depth dose within the vaginal mucosa. For bulky vaginal disease, this may be desirable. In other situations, this increased depth dose can increase the irradiated volume of vaginal tissue. As in the case of the ovoids, the largest ring diameter that can be accommodated by the patient’s anatomy should be used to get the optimal dose distribution.
Tandem and Cylinder Applicator In cases where the tandem and ring cannot be inserted properly due to constrictive patient anatomy, a possible alternative is the tandem and cylinder applicator. In this applicator, the vaginal surface is treated by
800
Jason Rownd
Figure 4. Isodose distribution in the plane of the ring section of an HDR tandem and ring treatment.
activating dwells within the length of the vaginal cylinder, in-line with the tandem section of the applicator. Different vaginal cylinder diameters are available to conform to patient anatomy (Figure 5). There are some advantages provided by this applicator over both the tandem and ovoid and the tandem and ring applicators. The tandem and cylinder applicator can be used in the aforementioned restricted anatomy case. The tandem and cylinder is an even simpler geometry, with correspondingly easier and reproducible insertions. Additionally, the vaginal surface can be treated to both a varying length and a varying dose along the length with suitable optimization, depending on disease progression. However, proper caution must be taken with the increased risk of rectal complications (Crook et al. 1987) because of the possibility of treating more volume of the rectum with the increased vaginal treatment lengths. One can control the length of the rectum treated by reducing the length of the vaginal cylinder treated. However, the increased depth dose caused by the increased distance of the vaginal cylinder surface from the dwell positions, relative to either the ring or the ovoid surfaces, is unavoidable (Figure 6). In all three of these applicators, the dwell positions within the vagina can and will contribute to the dose distribution at the cervix. As a result, one should control and measure the complete dose distribution from the proximal aspects of the vagina to the distal aspects of the uterus, as well as the nearby normal tissues. Packing and physical displacement can reduce the dose to normal tissues, but the complex relationship among the various sections of the applicators, the dwell positions therein, and the specific prescription requirements necessitates good geometry of implant and accurate dosimetry for each insertion.
CT/MR-Compatible Applicators In our institution, we have been using Nucletron’s CT/MR-compatible tandem and ring applicator for all of tandem and ring insertions (Figure 7). We also have available the tandem and ovoids (Figure 8) but prefer to use the tandem and ring applicators. We have also begun obtaining CT scans for planning for all insertions. In addition, MR scans are taken for retrospective analysis with the first fraction.
48–Applicator Design and Dose Distributions
Figure 5. HDR tandem and cylinder applicator.
Figure 6. HDR tandem and cylinder dosimetry.
Figure 7. CT/MR-compatible tandem and ring applicator.
801
802
Jason Rownd
There are new considerations as well as old benefits when using these applicators. The same fixed geometry, ease of insertion, and variable ring dwell activation still exist. These applicators are composed of carbon fiber and plastic. The material in these applicators means that they produce fewer artifacts in both the CT and the MR scans. In the MR scans, the applicator is basically an empty space in the image (Figure 9). In the CT scans, the applicator material shows up slightly brighter than the surrounding tissue and the dummy markers help to enhance potential dwell locations (Figure 10). In neither case is it completely trivial to identify possible dwell positions for treatment planning because the applicators themselves do not appear clearly on radiographs or CT scout images. There is currently no commercially available MR imaging dummies for complete dwell position reconstruction. We have been able to use traditional x-ray dummies for dwell position identification within the CT scans by using various image reconstruction views within the planning software. By reconstructing views through the
Figure 8. CT/MR-compatible tandem and ovoids applicator
Figure 9. MR reconstructed image of CT/MR-compatible HDR tandem and ring applicator.
48–Applicator Design and Dose Distributions
803
Figure 10. CT reconstructed image of CT/MR-compatible HDR tandem and ring applicator.
tandem and separately through the ring section of the applicator, we can easily identify the 1-cm spaced dummies and thus identify potential dwell positions. The metal x-ray dummies are clearly not suitable for MR imaging; but in the case of the CT images, they do not provide significant artifacts within the axial views. As a result of the diminished artifacts in the CT and MR imaging studies, the CT/MR-compatible applicators have become our standard applicators. Each fraction of a tandem and ring insertion is planned based upon CT reconstructions with normal tissues clearly defined slice by slice, sometimes with the aid of injected contrast. Specifically, we have been using a mixture of 40 cc sterile water and 10 cc contrast injected directly into the bladder via a catheter, with the catheter clamped during the imaging process. The resulting contrast clearly identifies the location of the bladder and also the relative thickness of the bladder wall near the applicator (Figure 11). Ultimately, isodose planning with the MR image set is desirable because of the enhanced tumor visibility on these images. Lack of a reasonable and easily identifiable dummy marker system for identifying potential dwell positions is making this direct MR-based planning difficult if not impossible. One common solution is image fusion or co-registration of MR images with some other imaging modality. You plan on one modality and co-register the isodose distribution with the MR image set. You could also use image fusion tools to identify target contours on the MR image set and directly map those contours upon the fused CT image set, with the resulting plan performed using the CT image set. Brachytherapy treatment planning systems are only now catching up with the external beam treatment planning systems in this area. As mentioned previously, there are some problems with the current crop of CT/MR-compatible applicators. Carbon fiber and plastic are reasonably strong, but not as strong as steel. To compensate for the relative durability of the materials, the carbon fiber applicators are physically larger, resulting in more difficult insertions, both in terms of physician handling and patient comfort. The discomfort from dilation of the cervix for the duration of the implant as well as the overall larger size of these applicators suggests that increased sedation may be necessary for each treatment fraction. Additionally, while the designs are
804
Jason Rownd
Figure 11. Axial slice of CT image set showing the tandem and its proximity to the bladder wall and bladder mucosa.
similar to their steel counterparts, they are not identical. This requires relearning the process of assembling the applicator within the patient as well as careful observation of the resulting dose distributions. In particular, the ovoids, from Nucletron (Columbia, MD), have a fixed lateral distance of 1 cm to their surface and larger ovoid choices might allow for a decreased cervical dose as the physical dwell positions are pushed laterally away from the central canal within the vagina.
Summary Each of the applicators discussed in this chapter have their benefits as well as difficulties associated with the use in HDR brachytherapy. In some cases, the benefits are simply a matter of ease of use. In other situations, it is the improved visualization of the tumor and normal tissues. Finally, the patient anatomy may restrict your choices of applicators to treat the cancer. Proper handling, imaging, prescription, planning, and treatment are all linked to a successful implant.
References Crook, J. M., B. A. Esche, G. Chaplain, J. Isturiz, I. Sentenac, and J. C. Horiot. (1987). “Dose-volume analysis and the prevention of radiation sequelae in cervical cancer.” Radiother Oncol 8:321–332. Erickson, B., R. Jones, J. Rownd, K. Albano, and M. Gillan. (2000). “Is the tandem and ring applicator a suitable alternative to the high dose rate Selectron tandem and ovoid applicator?” J Brachytherapy Int 16:131–144. Nag, S., B. Erickson, B. Thomadsen, C. Orton, J. D. Demanes, and D. Petereit. (2000). “The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for carcinoma of the cervix.” Int J Radiat Oncol Biol Phys 48(1):201–211.
Chapter 44
Manchester System for Gynecological Applications Anil Kumar Sharma, Ph.D. Radiation Oncology Department Long Beach Memorial Medical Center Long Beach, California Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 805 Radium Milligram-Hour Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 805 Rationale of Manchester Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806 Applicators and Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806 Intrauterine Tubes and Ovoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 806 Radium Sources and Their Loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807 Radium Substitutes and mgRaeq Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807 Dose Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 807 Original Point A Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808 Modified Point A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 808 Point B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809 Other Dose Specification Points as Variation of Point A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 809 Other Dose Limiting Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810 Vaginal Mucosa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810 Rectovaginal Septum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810 Relevance of the Manchester System Today . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810 Afterloading Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 810 Computerized Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811 Dwell Time Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811 Interstitial-Intracavitary Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 811 ABS Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812 ICRU 38 Revision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 812
Introduction Out of all the intracavitary procedures, treatment of uterine cervix is most widely used. The use of radium sources for this form of treatment started as early as 1903 in the United States as well as in Europe (Vahrson and Glaser 1988). In those early days, there was hardly any knowledge about the biological effects of radiation on the normal tissues and the tumor. There was little understanding about the dose, dose distribution, and the duration of implant. The dose prescription was entirely empirical. Then came the dosimetric “systems,” which basically are a set of rules, specific to a radioisotope and its spatial distribution in the applicator, to deliver a defined dose to a designated region. Within any system, specification of treatment in terms of dose, timing, and administration is necessary so as to implement prescription in a reproducible manner.
Radium Milligram-Hour Approach Two “systems”—Stockholm (Heyman 1924) and Paris (Lenz 1927), which came into clinical practice in 1911 and 1919, respectively—used the product of the total mass (mg) of radium used in the application and duration (hours) of the application (denoted as mg•h) for reporting the treatment. Both systems used a fixed number of mg•h, with the premise that for any given geometric arrangement of specified sources, dose at any point is directly proportional to the amount of activity in the sources and the duration of the
805
806
Anil Kumar Sharma
implant. The Stockholm system was a fractionated delivery system of three applications delivered within one month. Each application lasted 20 to 30 hours and was separated by a week or two from the other fraction. The amount of radium was unequal in uterus (30 to 90 mg, in linear tube) and in vagina (60 to 80 mg, in shielded silver or lead boxes), with more radium in the vagina than that in the uterus. Vaginal and uterine applicators were not fixed together. Total mg•h were usually 6500 to 7100, of which 4500 mg•h were in vagina. In the Paris system, however, almost equal amounts of radium were used in uterus and in vagina, in smaller amounts were than those used in the Stockholm system. They believed that better results can be obtained with small amounts of radium acting over a longer period of time than with large amounts during a short period, because more cells are irradiated during mitosis if the time period is longer. The Paris system comprised just one single application that lasted 5 days to deliver 7200 to 8000 mg•h. The intrauterine tube contained three sources in the ratio of 1:1:0.5 and two cork intravaginal cylinders (colpostats) one source each of the same strength as the top intrauterine source. Uterine sources in both systems were arranged in a line extending from the external os to nearly the top of the uterine cavity. Both systems preferred the longest possible intrauterine tube to increase the dose to the paracervical region and the pelvic lymph nodes. There was a limited use of external beam therapy in the Stockholm system, whereas the Paris system used external beam therapy before the implant.
Rationale of Manchester Approach When intracavitary therapy, specified in mg•h, is used in conjunction with external beam therapy, specified in terms of absorbed dose, overall radiation treatment cannot be adequately defined. So, in order to define the actual dose delivered in “fixed mg•h systems” in a more meaningful way, Tod and Meredith (1938) began to calculate the dose in roentgens to various sites in the pelvis by defining a series of points anatomically comparable from patient to patient. Due to rapid dose variations, the anatomical points were not assigned to obviously important sites like external os or the vaginal vault, but rather where the dosage was not too sensitive to small variations in applicator position. Tod concluded that the limiting radiation dose was not the dose to the critical structures, such as the rectum or bladder, but to the area in the medial edge of the broad ligament where uterine vessels cross the ureter (Meredith 1967). To this pyramid-shaped area, the base of which rests on the lateral vaginal fornices and apex curves around the anteverted uterus, the name “Paracervical Triangle” was given. It is considered that the tolerance of this paracervical triangle, is the main limiting factor in the irradiation of the uterine cervix. This led to the designation of a point for specifying the dose and a system of sources to deliver a constant dose rate to this point irrespective of the variation in size and shape of the uterus and vagina. The rationale of this approach, which came to be known as the Manchester approach, was to define treatment in terms of dose to a point representative of the target, i.e., uterus, which was a dose limiting point, more or less reproducible from patient to patient. To make things simple, the applicators and their loading were designed to enable the same dose rate to this point regardless of which combination of applicators is being used. To achieve the consistent dose rates, a set of strict rules dictating the relationship, position, and activity of radium sources in the uterine and vaginal applicators was devised. The Manchester system is a modification of the Paris and Stockholm systems, adapting the source loading technique of the Paris system and the fractionated delivery of dose from the Stockholm system.
Applicators and Sources Intrauterine Tubes and Ovoids The intrauterine tubes in the Manchester system are made of thin molded rubber or plastic with one end closed and supporting a flange at the other end for aiding fixation. These tubes are available in three
44–Manchester System for Gynecological Applications
807
lengths, meant for one, two, or three 20-mm long radium tubes. The applicators used in the vagina are a modification of the cork colpostats used in the Paris technique. They are made of hard rubber or plastic and have diameters of 20, 25, or 30 mm. They are approximately ellipsoids of revolution, mimicking the shape of the isodose surface around a radium tube of 15 mm length, and are named “ovoids.” They are used in pairs, one ovoid in each lateral vaginal fornix at the level of the cervix, and are locked in place by using a spacer or a washer. While the spacer separates the ovoids by 10 mm, the washer allows them to be almost in contact with each other. The ovoids are designed not only to be adaptable to the different sizes of the vagina, but also to take advantage of vaginal capacity to carry the radium laterally. Both the dimensions and the shapes of the applicators proved to be convenient and adaptable in clinical practice.
Radium Sources and Their Loading The radium tubes originally used in the Manchester system were of 15 mm active length. Later, 13.5-mm active length tubes were put to use necessitating the change in ovoid design to conform to the isodose surface of this source tube. A “unit” of radium containing 2.5 mg of 1 mm Pt-filtered radium was defined, and all loadings in the intrauterine tube and vaginal ovoids were made integral multiples of this unit. A long intrauterine tube with three sources contained 4, 4, 6 units, and medium intrauterine tube with 4, 6, and short with 8 units. Large, medium, and small ovoids were assigned 9, 8, and 7 units in each ovoid.
Radium Substitutes and mgRaeq Concept Radium, because of its radiation protection concerns is virtually not used these days. Radioisotopes like 137 Cs and 192Ir require simplified protection both in terms of thickness of barrier required to provide adequate protection and as well as the absence of gaseous radioactive daughter product. Dose distributions in tissue from these isotopes are not much different from those produced by radium because they too have energy higher than 300 keV and their dose distribution is not greatly affected by their energy, essentially following the inverse square law. These sources, called “radium substitutes,” can be calibrated in terms of a quantity that allows the use of radium tables without modification. The source strength is specified in terms of the mass of radium that would produce an equivalent exposure rate at a distance 1 m on the transverse axis of the source. One effective equivalent mass of radium, mgRaeq, of a radium substitute yields an exposure rate, at 1 m, of 0.825 mR/h. So, although this unit appears to be a unit of mass, it actually is a specification of the exposure rate at a distance. However, the mgRaeq of a source is not necessarily the ratio of the exposure rate constant of the radium substitute to radium. Source geometry and filtration affect the equivalency.
Dose Specification It was recognized, before the Manchester system, that the dose prescription in terms of mg•h ignored anatomical targets and tolerance organs. Also, it was clear that uniform dose distribution in the whole target, comprising the cervix itself, uterus, vaginal vault, and the parametria, was not going to be achieved, but certain points in the tolerance area, the paracervical triangle, could be given the same dose using the ingenious applicator design and source loading following the loading rules. These points lying on either side of the uterus were called “point A.” In the Manchester system, optimum dose was taken to be 8000 R, delivered to this point A (defined in the following section) in two sessions of about 72 hours each with a 4- to 7-day interval between. This implied a dose rate of 55 R per hour, which was achieved by the strict loading rules of this system.
808
Anil Kumar Sharma
Original Point A Definition Initially, this point was defined as being 2 cm lateral to the uterine canal and 2 cm from the mucous membrane of the superior fornix of the vagina in the plane of the uterus (see Figure 1). The value of point A for this dosage system was confirmed by a study of over 500 cases that showed a clear relationship between the tolerance of normal tissues and the dose this area received. This concept of the statement of dosage to a single point made this system the most acceptable brachytherapy technique for the treatment of cervical cancer. The source-loading rules were defined in a way that point A received the same dose rate no matter which ovoid and intrauterine combination is used. Experience with the system also showed that not more than about one-third of the total dose at point A should be delivered from radium in the vaginal ovoids.
Modified Point A Although point A is assumed to be defined in relation to important anatomical structures, they cannot be revealed in a normal radiograph, necessitating a convention by which the position of point A can be determined. For practical purposes, it was established by experience that the lower end of the intrauterine tube could be taken to be level with the lateral fornix (Tod and Meredith 1953), thus point A was redefined to be 2 cm along the axis of the central tube from the lower end and 2 cm from it laterally. In some cases, the uterus may not lie in the midline of the body and symmetric to the ovoids due to anatomical distortions due to disease or otherwise. In such cases, it is assumed that point A is carried with the uterus and still lies 2 cm superiorly along the axis of intrauterine sources and 2 cm perpendicular to the radium line on either side. See Figure 2.
Figure 1. Original definition of point A; 2 cm superior to the vaginal fornix and 2 cm lateral to the uterine canal. Point B is situated 3 cm lateral to point A.
44–Manchester System for Gynecological Applications
809
Figure 2. Modified point A is carried with the uterus and it lies 2 cm superiorly from the lower end of the intrauterine source (along the axis of intrauterine sources) and 2 cm perpendicular to the radium line on either side. Point B, which does not directly depend on the uterus, remains as a fixed point, 5 cm laterally from a point 2 cm up the midline from the end of the radium tube.
Point B While the dose to point A is the most useful index of limiting dosage that can be given, the lateral falloff of the dose is also of importance. For this reason, a further reference point B, was also defined to be 5 cm from the midline and 2 cm up from the mucus membrane of the lateral fornix. This point was chosen since it gives not only the dose in the vicinity of the pelvic wall near the obturator nodes, but also a good measure of the lateral spread of the effective dose. The dose at point B depends very little on the actual geometrical distribution of radium, such as the size of the ovoids and intrauterine tubes, but entirely on the total amount of the radium used. In those cases where the uterus does not lie in the midline of the body, the tissues in which point A lies are considered to be carried with the uterus, but point B, which does not directly depend on the uterus, remains as a fixed point, 5 cm laterally from a point 2 cm up the midline from the end of the radium tube. In the loading rules of the Manchester system, it was recommended that, if possible, the largest ovoids be used to carry the radium close to point B and increase the depth dose. It was advised to place the ovoids as far laterally as possible in the fornices for the same reason.
Other Dose Specification Points as Variation of Point A Points A and B are widely used although their exact meaning and their definition have been interpreted in different ways and vary by the institution. Some relate point A to anatomical references in the patient and others to the geometry of the sources. Different methods of definition provide different values for the calculated dose rate to point A. In the Manchester system, definition of point A, original point A, and the modified point A coincided with each other and its location was such that there was little variation in dose rate from one source arrangement to the other. So, if the Manchester system rules were followed, the dose rate to this point was predictable and more or less constant. Over the years, point A has been defined in many ways. Point Av (“v” stands for vagina) was proposed as 2 cm lateral to the mid point of the cervical collar and 2 cm above the top of the colpostats, measured at their intersection with the tandem midpoint on the lateral radiograph (Potish and Gerbi 1987). Also, the modified point A (modified in 1953) is sometimes denoted as Ao (“o” stands for external os). The
810
Anil Kumar Sharma
Fletcher system, described elsewhere in this monograph, defined several other points to account for the dose to regional lymph nodes and pelvic points (Fletcher and Hamberger 1980). In 1993, for a specially designed system (Madison system for HDR brachytherapy of uterine cervix), point M was defined (Thomadsen et al. 1992). It lies 2 cm cephalad along the tandem from a line connecting the center points of the vaginal ovoids and 2 cm perpendicular to the tandem, when using 1-cm radius ovoid caps. In this system, the uterus is held lower in the pelvis (using tanaculum) to lower the small bowel dose superior to the uterus. In this situation, this point M approximately coincides with original point A of the Manchester system. The American Brachytherapy Society (ABS), in its recommendations for low dose rate (LDR) brachytherapy of cervical cancer denoted the original point A as Ao (Nag, 2002) and the modified point A as Af (“f” stands for flange). The effects of the definition of doses are discussed below.
Other Dose Limiting Structures Besides the all-important paracervical triangle as dose limiting structure in the treatment of uterine cervix two more areas were considered to be vulnerable: the vaginal mucosa and the rectovaginal septum.
Vaginal Mucosa The tolerance of vaginal mucosa is such that not more than about 40% of the total dose to point A can safely be delivered through the vaginal ovoids and this should be taken into account in planning the differential loadings. A similar ratio was chosen after studying the intrauterine and vaginal loadings in a series of cases treated by the Stockholm technique.
Rectovaginal Septum The probable dose on the rectovaginal septum for any technique should be less than that at point A. In their experience in Manchester, dose to this area could be reduced to less than 80% of the dose to point A by carefully packing gauze to a thickness of at least 1.5 cm to pack ovoids away from the rectum.
Relevance of the Manchester System Today The Manchester system was meant for radium as the radioisotope and applicators specially designed to accommodate those sources following a set of rules to deliver almost a constant dose rate to its dose specification point A. Any variations in the selection of source, applicator, or the set of rules will result in dose delivery which most likely be different from that dictated by the Manchester system. With radium being all but replaced by 137Cs [LDR and medium dose rate (MDR)] and 192Ir [high dose rate (HDR)] and to a lesser extent by 60Co (LDR and HDR), it is imperative to look into the relevance of the Manchester system in modern times.
Afterloading Technique In the beginning, manual afterloading was the direct extension of the conventional preloaded applicator system using radium sources. But today, manual afterloading applicators are modified to accommodate 137Cs sources, but still follow the dosimetry associated with their original pre-loaded form. Most systems such as Fletcher, Henschke, or their modified versions still use point A for dose prescription. Only one applicator system, from Amersham, follows closely the recommendations of the Manchester system but uses 137 Cs sources. For a standard insertion, tables were designed for this system for various combinations of vaginal and uterine applicators. Use of point A in this system is close to its modified definition in the
44–Manchester System for Gynecological Applications
811
Manchester system, but for all other systems, prescription to point A should be used with caution. For example, it was reported that the calculated dose contribution from the ovoid sources can be in error by as much as 25% unless correction is made for the different absorption of 137Cs gamma rays in the Fletcher/Suit applicators designed for radium (Godden 1988). Caution also needs to be used when tables outlining mg•h and application times used for the Fletcher system are used with other applicators, especially with those in which the source axis is along the vagina rather than at a right angle to the intrauterine tube.
Computerized Dosimetry Today, computers are routinely used for dosimetry and treatment time calculations. Applicators are reconstructed from radiographs or from a computed tomography (CT) data set. Applicator-based brachytherapy still requires prescription points based on the applicator position. Most institutions still use points A and B although their meaning and their definition may be interpreted in different ways. Applicator geometry affects the dose to the modified point A. An afterloading applicator’s vaginal ovoids may not sit in the natural position for the ovoids, and may get pushed high, leaving point A in high dose gradient. Even though computerized dosimetry helps in eliminating reliance on the mg•h dose tables, error in establishing the dose prescription point may lead to serious dose delivery problems. In image-based brachytherapy, where dose prescription is according to the target volume, dose prescription to a point may not be relevant, but for intercomparison purposes many institutions still use them.
Dwell Time Optimization HDR remote afterloading systems use dwell time optimization, which may be useful in loading intracavitary applicator and vaginal ovoids optimally to deliver dose to either prescription points or to a target volume. In the intracavitary applications, with just three catheters and a limited number of allowable dwell positions, dwell time optimization cannot be used to its full advantage. Nevertheless, dose to critical organs like bladder and rectum can be reduced in some situations. The Manchester system’s role in such situations is limited to second check and to ensure proper proportion of the source activities used in intrauterine catheter and vaginal ovoids.
Interstitial-Intracavitary Implants Intracavitary applications can deliver adequate dose to the target volume in patients with early-stage disease and with normal anatomy. However, in patients with advanced stage disease and or distorted anatomy, intracavitary application may not be able to deliver tumoricidal dose to the target volume, i.e., parametria and lower vagina, without sparing the organs at risk, i.e., rectum and bladder from higher doses of radiation. Interstitial transperineal implants with rigid needles (which serve as interstitial ovoids) and a vaginal obturator with or without an intrauterine tandem are used in such situations to deliver adequate dose to the target volume with relative sparing of the critical organs. Interstitial techniques are designed to deliver prescribed dose to the target volume, generally taken to be the volume enclosed by the implanted needles. In early days of the development of these techniques, the dose was, however, prescribed and reported for point A. For interstitial implants, dose prescription to point A cannot be justified, because it may lie right on the loaded position of the needle. Also, only in those cases, where a tandem has been used, may one think of assigning point A using the modified Manchester system definition. Point A in such situations may be thought of lying perpendicular to the tandem (at a point 2 cm cephalad to the flange) and 2 cm from its axis. Also, this point should be midway between the needles (which may not be possible sometimes) so as to avoid rapid dose gradients. In those situations where intrauterine tandem has not
812
Anil Kumar Sharma
been used, point A definition becomes all the more difficult to apply. So, point A seems to be of no relevance in the case of interstitial-intracavitary implants.
ABS Recommendations ABS recommendations (Nag et al. 2002) for LDR intracavitary applications for the treatment of cancer of cervix retained original Manchester system point A. Also, for modern applicators and applications with vaginal cylinder, ABS recommends an alternate procedure of locating point A. For tandem and ovoids, localization of point A can be carried out using radiographs as follows: Draw a line connecting the middle of the sources in the vaginal ovoids on the antero-posterior (AP) radiograph and move 2 cm (plus radius of the ovoid) superiorly along the tandem from the intersection of this line with the intrauterine source line and then 2 cm lateral on either side of the tandem. For tandem and vaginal cylinder, the localization of point A can be carried out as follows: From the flange of the tandem, move 2 cm superiorly along the tandem and then laterally 2 cm perpendicular to the tandem on both sides of the AP radiograph.
ICRU 38 Revision International Commission on Radiation Units and Measurement (ICRU) report 38 (ICRU 1985) discouraged the use of points A and B because the exact meaning and their definitions have not always been interpreted in the same way in different centers and even in the same center over a period of time. The different methods of definition provide different values for the calculated dose rate to point A. Therefore, if the prescribed dose to point A is used to calculate the total time of application, different values of time will be obtained for different methods used to assign the prescription point. This report encourages the use of target volume for dose prescription and reporting along with the reference volume for 60 Gy absorbed dose prescription. This report is being revised and may include some dose points similar to the classical systems. Details of the ICRU 38 revision are discussed elsewhere in this monograph.
References Fletcher, G. H., and A. D. Hamberger. “Squamous Cell Carcinoma of the Uterine Cervix” in Textbook of Radiotherapy, 3rd Edition. G. H. Fletcher (ed.). Philadelphia: Lea & Febiger, pp. 720–773, 1980. Godden, T. L. Physical Aspects of Brachytherapy. Philadelphia: Adam Hilger, 1988. Heyman, J. (1924). “Technique and results in the treatment of carcinoma of the uterine cervix at “Radiumhemmet” Stockholm.” J Obstet Gynecol Brit Emp 31:1–19. International Commission on Radiation Units and Measurements (ICRU) Report No. 38. “Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology.” Bethesda, MD: ICRU, 1985. Lenz, M. (1927). “Radiotherapy of cancer of cervix at the radium institute, Paris, France.” Am J Roentegenol Radiat Ther 17:335–342. Meredith, W. J. “Dosage for Cancer of Cervix Uteri. Manchester System” in Radium Dosage: The Manchester System. Edinburgh: Livingstone, 1967. Nag, S., C. Chao, B. Erickson, J. Fowler, N. Gupta, A. Martinez, B. Thomadsen; American Brachytherapy Society. (2002). “The American Brachytherapy Society recommendations for low-dose-rate brachytherapy for carcinoma of the cervix, Int J Radiat Oncol Biol Phys 52(1):33–48. Potish, R. A., and B. J. Gerbi. (1987). “Cervical cancer: Intracavitary dose specification and prescription.” Radiology 165:555–560. Thomadsen, B. R., S. Shahabi, J. A. Stitt, D. A. Buchler, J. F. Fowler, B. R. Paliwal, and T. J. Kinsella. (1992). “High dose rate intracavitary brachytherapy for the carcinoma of the cervix: The Madison System: II. Procedural and physical considerations.” Int J Radiat Oncol Biol Phys 24:349–357.
44–Manchester System for Gynecological Applications
813
Tod, M. C., and W. J. Meredith. (1938). “A dosage system for use in the treatment of cancer of the uterine cervix.” Br J Radiol 11:809–824. Tod, M. C., and W. J. Meredith. (1953). “Treatment of cancer of cervix uteri—a revised ‘Manchester method.’” Br J Radiol 26:252–257. Vahrson, H., and F. H. Glaser. (1988). “History of HDR afterloading in brachytherapy.” Strahlenther Onkol 82 (suppl):2–6.
Chapter 45
Intracavitary Brachytherapy Applicators and UT M. D. Anderson Cancer Center Intracavitary Brachytherapy Techniques John L. Horton, Ph.D., Ann Lawyer, M.S., Paula Berner, B.S., Mandy Cunningham, B.S., and Firas Mourtada, Ph.D. Radiation Physics Department University of Texas M. D. Anderson Cancer Center Houston, Texas Intracavitary Brachytherapy Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815 Fletcher Family of Intracavitary Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815 Fletcher Suit Delclos Tandem and Ovoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 815 Tandem and Cylinders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817 Dome and Cylinders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 817 Henschke Applicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 818 Mold Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 820 Créteil Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 820 Institut Gustave Roussy Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 821 MR/CT Compatible Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 822 UT M. D. Anderson Cancer Center Intracavitary Brachytherapy Technique . . . . . . . . . . . . . 823 Historical Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 823 Placement of Applicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 824 Loading of Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 825 Treatment Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 826 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 827
Intracavitary Brachytherapy Applicators Fletcher Family of Intracavitary Applicators Fletcher Suit Delclos Tandem and Ovoids Probably the most widely used low dose rate (LDR) applicators for intracavitary cervical treatments in the United States are the Fletcher Suit Delclos (FSD) tandem and ovoid system. These applicators evolved from the original Manchester system gynecological applicators. The Manchester design used rubber colpostats and a rubber tandem. The colpostats were oval (prolate spheroid) in shape to conform to the isodose surface of a 226Ra tube source. The ovoid nomenclature was derived from the shape of these applicators. A rubber spacer separated the two ovoids. Fletcher changed to metal applicators and designed the ovoids to be cylindrical in shape. Each ovoid was attached to an individual tube that served as a handle. Since the ovoids are not attached to the tandem, they may be displaced anteriorly or posteriorly as needed, based on the patient anatomy and location of the disease. A yoke is also available to attach the ovoids to the tandem, if the anatomy and disease require it. Alternately, the ovoids may be held together with a scissors-type joint that allows a separation of up to 4 cm at the joint. The anterior and posterior poles of ovoids are equipped with small tungsten shields to protect the bladder trigone and anterior rectal wall without decreasing the dose to the uterosacral and broad ligaments. Tandems are designed with different angulations. The straight tandem is designated as a #0 tandem with increasing angulations designated #1, #2, and #3.
815
816
John L. Horton
Similar to the Manchester system, the Fletcher system was preloaded. Suit introduced afterloading for the system. However, Suit encountered difficulty afterloading the ovoids because the longitudinal axis of the source is aligned with the longitudinal axis of the ovoid. The Fletcher Suit square handle ovoids used a double hinge system to get the sources around the bend at the point where the handle attached to the ovoid proper. It was also necessary to make the ovoids 1 mm longer and to rotate the shielding to afterload the sources. Delclos (Delclos et al. 1980) developed round handle ovoids and placed the sources on springs to maneuver them around the bend where the ovoid handle attaches to the ovoid body. With this design the shields could be returned to the positions in the original Fletcher design. The FSD ovoids are 2 cm in diameter with plastic caps that extend the diameter to 2.5 cm or 3 cm. The 2-cm ovoids are referred to as small ovoids, the 2.5-cm as medium, and the 3-cm as large ovoids. Delclos also contributed the miniovoids that are 8 mm radius on the lateral aspect. However, these ovoids are not cylindrical as the medial aspect has a flat surface forming a “D” shape. The miniovoids are not shielded. The #1, #2, and # 3 FSD tandems are seen in the center of Figure 1. On the left of the figure is a set of small FSD ovoids with the large and medium caps above them. To the right of the small ovoids is a set of miniovoids. Between the miniovoids and the tandems are two flanges. A flange is positioned on the tandem to place the flange against the external os when the tandem is in position. This flange has a keel on it. Packing around the keel minimizes rotation and displacement of the tandem during treatment. A set of vaginal cylinders discussed in the next section is to the right of the tandems. The handles used to afterload the ovoids are to the right of the cylinders, with the tube used to contain the tandem sources on the far right side of the picture. The straight device at the bottom of the figure is used to insert radio-opaque seeds 5 mm deep
Figure 1. A set of Fletcher Suit Delclos (FSD) intracavitary brachytherapy applicators. On the left is a set of small ovoids. A set of large caps is above the ovoids and a set of medium caps above the large caps. To the right of the small ovoids is a set of minimovoids. Three tandems are to the right of the miniovoids with two flanges between the miniovoids and the tandems. Next to the tandems is a set of two source carriers for the ovoids and a source carrier for the tandem sources. A seed inserter is in a horizontal orientation at the bottom of the figure.
Intracavitary Brachytherapy Applicators and Intracavitary Brachytherapy Techniques 817 into the anterior and posterior aspects of the cervix (12 o’clock and 6 o’clock positions). This seed is used to visualize the position of the cervix and verifies the flange is positioned against the os. The shielding in the standard ovoids may reduce the dose at 2 mm from the ovoid surface by up to 45% over a small volume. A transmission curve through the ovoid shield is seen in Figure 2. The contours represent the amount of dose transmitted through the shields at various positions in a plane passing through the poles of the ovoid and intersecting the shields. The position of the bladder shield is at the top of the ovoid surface in the figure and the rectal shield is at the bottom of the ovoid surface in the figure. The contours were generated with a Monte Carlo calculation performed with MCNPX (Gifford et al. 2005). Tandem and Cylinders The Fletcher Suit Delclos armamentarium of applicators also includes tandem and cylinders (Delclos et al. 1980). The tandem and cylinders are used when the patient’s anatomy is too small to allow the insertion of the miniovoids or it is desired to treat disease that has vaginal extension. The cylinders are 2.5 cm in height and range in diameter from 1 cm to 5 cm. Multiple cylinders are used depending on the vaginal length. Cylinders are also available that included shielding that subtends an angle of approximately 60° to protect the rectum. The amount of shielding that may be placed into a cylinder depends on the diameter of the cylinder. Several cylinders are seen to the right of center in Figure 1. Dome and Cylinders Delclos (Delclos et al. 1980) also contributed to the development of the afterloaded dome and cylinder applicators. These applicators are used for vaginal cuff irradiation alone or for irradiation of both the vaginal cuff disease plus disease extending into the vagina for post-hysterectomy patients. A Walstam capsule
Figure 2. A plot of the radiation transmitted through the shields of a small FSD ovoid. The bladder shield is at the top surface of the ovoid and the rectal shield is at the bottom surface.
818
John L. Horton
8 mm in physical length containing either a 2.1-mm or 4.2-mm active length 137Cs source is used in the dome for the cuff irradiation. A Walstam capsule is seen is Figure 3. Standard 137Cs tube sources are added in the cylinders if required to treat disease extension into the vagina. The Walstam capsule delivers a fairly uniform dose to the surface of the cylinder. Figure 4 shows the dose rates in cGy per hour per mgRaeq at the surface of a 3-cm dome. If it is desired to treat vaginal extension of the disease, the activity of the tube source and its spacing is manually optimized based on the prescription. Figure 5 is an example of a sheet used for planning a dome and cylinder treatment. The figure shows the dose rates in cGy per hour per mgRaeq for both the Walstam capsule in the dome and 137Cs tube source in the cylinder with a 15 mm spacer between the Walstam capsule and the tube source. The dose rates displayed in Figures 4 and 5 are only valid for the sources and applicators currently in use at University of Texas M.D. Anderson Cancer Center (UTMDACC) and must be verified before using for other applicators and sources.
Henschke Applicator Henschke (Henschke 1960) designed the first afterloaded intracavitary applicator in the late 1950s. A Henschke applicator is seen in Figure 6. It was based on the Manchester intracavitary system and consisted of a tandem and two ovoids attached to individual handles. The ovoids of the original design were 2-cmdiameter nylon spheres. Commercially available systems have hemispherical ovoids, 2 cm in diameter. The ovoid diameter may be increased to 2.5 cm or 3 cm with the addition of nylon caps. The ovoids are secured to the tandem with a flange that fixes the position of the ovoids relative to the tandem. The sources were afterloaded through the handles of the ovoids. The sources were placed in the ovoid approximately parallel to the handle and to the tandem. Recall in the FSD applicator the ovoid sources are at approximately 90° to the handle and to the tandem. Suit faced a design problem achieving this 90° rotation for afterloading the FSD applicator. Because of the different orientation of the ovoid sources the dose distribution is different between the Henschke and FSD applicators. In 1978 Delclos (Delclos et al. 1978) discussed these differences in the dose distributions between the FSD family of applicators and other applicators including the Henschke applicator. This publication by Delclos was prompted by radiation oncologists’ reports of unexpected complications when using Fletcher’s technique. One of the reasons given for these complications was the use of Fletcher’s technique with the Henschke applicator. Delclos pointed out that Fletcher’s rules could not be extrapolated to other applicator systems. Another problem pointed out by Delclos was the change from 226Ra to 137Cs without making appropriate corrections for the differences in the dose distributions from the sources. A
Figure 3. A Walstam capsule, 8 mm in physical length with a 2.1 mm active length of 137Cs.
Intracavitary Brachytherapy Applicators and Intracavitary Brachytherapy Techniques 819
Figure 4. The dose rates in cGy/mgRaeq-hr on the surface of a 3.0 cm vaginal dome containing a Walstam capsule. Note the relatively uniform dose.
Figure 5. The dose rates in cGy/hr-mgRaeq on the surfaces of a 3.0 cm vaginal dome and vaginal cylinder. The dose rates are for a vaginal dome containing a Walstam capsule and the cylinder contains a 137Cs tube source manufactured by 3M. A 15-mm spacer separates the Walstam capsule from the tube source.
820
John L. Horton
Figure 6. A Henschke HDR applicator from Nucletron.
final problem discussed resulted because Fletcher had not patented the applicator and had no control over its manufacture or sale by commercial vendors. Delclos found ovoids with non-uniform thickness walls. Ovoids that did not center the radiation source on the axis of the ovoid were also implicated in the reported complications.
Mold Applicators Vaginal acrylic resin mold (moulage) applicators are used extensively in France. The mold is produced from an impression of the vagina taken with a quick setting alginate compound. The mold conforms to the cervix. A central opening in the mold accommodates the uterine source. Molds may be either custom fabricated for each patient or purchased in standard sizes of 2.5, 3.5, or 4.5 cm in diameter. The two principal systems based on these applicators are the Créteil method and the Institut Gustave Roussy method. Créteil Method The Créteil method (Pierquin and Marinello 1997) is based on flexible, linear sources or an array of sources of uniform linear activity throughout the implant. 192Iridium is typically used, but 137Cs or 60Co may also be used if the sources are small in diameter and flexible. There are two vaginal sources. The length of each vaginal source is 0.8 times the diameter of the cervix. These sources are afterloaded though plastic tubes closed at the distal end and open at the proximal end. The vaginal sources are positioned circumferentially at 7 mm from the right and left external lateral walls of the mold. The distances between the ends of the sources are equal in both the anterior and posterior aspects, with these separations positioned in the region of the bladder and anterior rectal wall. An outline diagram illustrating the geometry of the vaginal sources is seen in Figure 7. The tube carrying the uterine source is 35 cm long, sealed at the distal end and open outside the vagina to allow afterloading of the uterine source. The source extends distally to 5 mm from the superior aspect of the uterus. This source extends beyond the plane of the vaginal sources by 5 mm if the diameter of the cervix is less than 4 cm and by 1.5 cm if the cervix is larger. Two lead pellets are placed in the mold. One pellet is placed anterior to the external os and the other posterior to the external os. These pellets identify the position to calculate the cervical dose.
Intracavitary Brachytherapy Applicators and Intracavitary Brachytherapy Techniques 821 When the rules of the system are followed, the dose is prescribed to the reference isodose surface. This surface is pear shaped and at the level of the vaginal sources extends 7 mm outside of the surface of the mold. This reference isodose surface also extends 7 mm superior to the end of the uterine source. The reference isodose surface for the Créteil method equals 0.46 Gy per day for a linear activity of 1.0 U per cm. A dose of 60 Gy in a single application of 4 to 8 days is delivered to this reference isodose line. To deliver 60 Gy to the reference isodose surface in 4 days requires a linear activity of 32.6 U per cm. This linear activity is comparable to a typical UTMDACC tandem loading of 28.9 U per cm to 43.4 U per cm. The Créteil method delivers a dose at the cervix that is twice the dose prescribed to this reference isodose surface with the rectal dose less than or equal to the prescribed dose. Institut Gustave Roussy Method The Institut Gustave Roussy (IGR) method (Gerbaulet and Bridier 1997) is based on the use of 137Cs in an acrylic resin vaginal mold. The vaginal sources are positioned with the long axis of each source oriented in the anterior-posterior (AP) direction and placed equidistance from the cervical os with the os centered on the active length. The distance between the vaginal sources is equal to the average value of the active length of the individual sources. Figure 8 is an outline drawing of the vaginal source positions for the IGR technique. The superior end of the uterine sources is located at the junction of the upper and middle third of the uterine cavity. The inferior end extends to the plane between the vaginal sources. Computerized treatment planning is used to determine the implant time based on the treatment volume and the dose to critical structures.
Figure 7. An outline diagram of the Créteil vaginal mold demonstrating the orientation of the vaginal sources. The sources are 7 mm from the lateral vaginal walls. The spaces between the ends of the sources at the top and bottom of the diagram are equal and are oriented opposite the bladder and anterior rectal wall.
Figure 8. An outline diagram of the Institut Gustave Roussy vaginal mold illustrating the orientation of the vaginal sources. The sources are separated by a distance equal to their average of their active lengths and equidistance from the os.
822
John L. Horton
MR/CT Compatible Applicators With the increasing use of computed tomography (CT) and magnetic resonance imaging (MRI) for evaluation of the quality of implants, it is very important that brachytherapy applicators are CT/MR compatible. The CT artifacts produced by the shields are an important issue that must be addressed for Fletcher type LDR applicators. Weeks (Weeks et al. 1989 and Weeks and Montana 1997) has designed two models of CT-compatible applicators. The first model was based on the FSD applicator, but was constructed of acrylic with afterloaded shields. This design allowed the patient to receive a CT scan following the placement of the applicator, but prior to afterloading the shields and active sources. The problem with this design was that the size of the ovoid handles increased from 8 mm to 16 mm to allow the shields to be afterloaded. This increase in size made it more difficult for the radiation oncologist to observe the placement of the ovoids. In his second design Weeks returned to a metal applicator. The tandem and ovoids are constructed of black anodized aluminum and the handles are stainless steel. The external dimensions of the colpostat are the same as the Delclos miniovoids. Small, medium, and large ovoid dimensions are achieved by adding caps to this design. Weeks avoided the artifacts generated by the shields by again afterloading the shields. However, Weeks cleverly designed the source carriers used for afterloading the sources to be his shields. He constructed these source carriers from high-density material and designed them to provide the same shielding by aligning the shielding material longitudinally along the source and placing the shield in closer proximity to the source. Ling (Ling et al. 1987) used standard shielded FSD applicators, but changed the CT window and level values to reduce the appearance of the artifact on the CT image. This strategy allowed clinicians to visualize the applicators, outline the critical structures, and produce a volumetric analysis of the implants. Dose distributions were calculated assuming the patient was homogeneous and water equivalent. Pelloski (Pelloski et al. 2005) pursued a similar strategy of changing the CT window and level values to reduce the artifacts in an analysis of 93 implants. An example of the effect of the artifacts may be seen in Figure 9. Doses were calculated with a conventional treatment planning system ignoring the effect of the shields. He found that the International Commission on Radiation Units and Measurements (ICRU) bladder point correlated with, but underestimated, the dose to the 2 cc of bladder receiving the maximum dose. Pelloski also found that the ICRU rectal point correlated with the 2 cc of rectum receiving the maximum dose and was a reasonable approximation. Gifford (Gifford 2004) performed a Monte Carlo analysis on a subset of 12 of these patients to account for the effects of the shields. He found that the effect of the bladder shields was small, but that the rectal shields provided an average reduction in dose of 10% to the 2 cc of rectum receiving the highest dose. In this small subset he also observed that the dose was reduced by 405 cGy to 5% of the rectal surface for one patient. We have recently filed a patent application on a new applicator. This applicator is similar in external dimensions to the FSD applicator, but the position of the shield may be changed dynamically during CT scanning. The CT scan acquisition begins with the shield moved to a position that causes minimal interference with the CT scan. The CT scan is then interrupted and the shield repositioned without moving the patient or the applicator. The CT scan is then continued with the shield in a new position that again causes minimal interference with the remainder of the CT scans. MR compatible applicators are also important. For MRI, our preliminary analysis indicates, as expected, stainless steel will be the major material in generating MR artifacts. Tungsten generates artifacts but it appears that by judicious selection of MR scan parameters, we may be able to image an applicator made of nonferrous materials while continuing to use tungsten shields. However, our work is very preliminary in this area.
Intracavitary Brachytherapy Applicators and Intracavitary Brachytherapy Techniques 823
Figure 9. A transverse CT view and two reconstructed orthogonal planes of a patient with an FSD tandem and ovoids application. The artifacts created by the ovoid shields are visible in these views. The appearance of these artifacts has been reduced by optimizing the window and level parameters of the CT imager.
UT M. D. Anderson Cancer Center Intracavitary Brachytherapy Technique Historical Perspective The UTMDACC gynecological cancer database dates to 1960 and contains approximately 7000 patients treated with radiation therapy. The majority of these patients received both 40 to 45 Gy external beam treatment and intracavitary radiation with continuous LDR brachytherapy. Fletcher first described the system for intracavitary treatment of gynecological cancer at UTMDACC in Textbook of Radiotherapy (Fletcher 1980). The system was designed for 226Ra tube sources, 22 mm in physical length and 15 mm in active length with 1 mm Pt filtration. Presently, treatments are administered with the Nucletron continuous LDR system (SelectronTM) containing 137Cs pellets 2.5 mm in diameter or manually afterloaded 137Cs tubes, with a physical length of 20 mm and an active length of 13.5 mm. Because the system evolved based
824
John L. Horton
on 226Ra, our physicians continue to prescribe in mgRaeq. The discussion in the remainder of this chapter is in terms of mgRaeq rather than U. To convert from mgRaeq to U, just remember that 1 mgRaeq equals 7.227 U.
Placement of Applicators The UTMDACC system is not based on prescription to point A or dose rates to critical structures. The system is based on the geometry of the placement of the applicators in the uterus and vagina and the placement of the sources within the applicators (Katz and Eifel 2000). As illustrated in the lateral film in Figure 10, the tandem should be one third of the way between the junction of S1-S2 and the symphysis pubis. For the 808 insertions evaluated by Katz and Eifel (2000) the median distance between the sacrum and the tandem was 4.0 cm with an interquartile (IQ) range of 3.3 to 4.9 cm. The distance between the tandem and pubis was 8.4 cm with an IQ range of 7.6 to 9.1 cm. The tandem should be halfway between the bladder and the junction of S1-S2. Marker seeds inserted with the seed inserter (discussed above and seen in Figure 1) at the 12 o’clock and 6 o’clock positions can be seen in Figure 10. These seeds are 5 mm beneath the surface of the cervix. The ovoids should be against the cervix and the tandem should bisect the ovoids. The bladder and rectum should be packed away from the implant. The AP film in Figure 11 demonstrates that the ovoids should fill the vaginal fornices and they should be separated by 5 to 10 mm with the flange between them. The largest ovoids that can be accommodated without caudal displacement are used to fill
Figure 10. Lateral radiograph of an FSD tandem and ovoids application. Note the tandem is one third of the distance between the sacrum and the pubis, being closer to the sacrum. Tandem is halfway between the bladder and the sacrum. Tandem bisects the ovoids. The ovoids are against the marker seed in the cervix. The anterior rectal wall and bladder are packed away from the ovoids. The applicators are for a Selectron™ LDR remote afterloading unit. The position of the active pellets are marked on the film.
Intracavitary Brachytherapy Applicators and Intracavitary Brachytherapy Techniques 825 the fornices. Typically, the ovoids are separated just enough to accommodate the flange on the tandem. The tandem should be halfway between the ovoids. We always film the patient in the operating room to verify that these conditions are fulfilled. If these conditions are not met, the physicians will spend considerable time repacking the implant to achieve the desired results. On the rare occasion when these conditions are not fulfilled, we may spend extensive time planning the patient before the sources are loaded into the applicators.
Loading of Activity The activity loaded into the tandem ranges from 4 mgRaeq per cm to 6 mgRaeq per cm. If needed, spacers are used to fill the length of the tandem from tip to flange. In most cases, the physical length of the most inferior tandem source should not extend beyond the flange. If the ovoids are displaced inferiorly from the cervix as determined by the marker seeds, it may be necessary to extend a tandem source beyond the flange to avoid a cold spot between the tandem and ovoids. Typically, we place the inferior end of the physical length of the most inferior tandem source 2 to 3 mm above the anterior aspect of the ovoids in this case. We usually plan these cases before loading the sources. The activity in the tip of the tandem is usually greater than in the rest of the tandem to increase lateral coverage. The most common loading is 15 mgRaeq at the tip, followed by two sources of 10 mgRaeq. If significant endocervical disease exists, the tandem may be loaded with the middle source being a 15 mgRaeq. Typical tandem loadings range from 30 to 45 mgRaeq over a length of 6 to 8 cm. The small ovoids are 2 cm in diameter and contain rectal and bladder shields. Small ovoids are loaded 10 to 15 mgRaeq. When medium ovoid caps are added, the ovoid diameter is 2.5 cm and the ovoids are typically loaded 15 to 20 mgRaeq. Large ovoid caps increase the ovoid diameter to 3.0 cm. In this instance the ovoids are loaded 20 mgRaeq. When the vagina is too small to accommodate small ovoids, miniovoids are used. The miniovoids contain no shielding and are typically loaded 5 mgRaeq; on rare occasion they
Figure 11. An AP radiograph of an FSD tandem and ovoids application. Note the axis of the tandem is centered between the ovoids and the ovoids are separated by approximately 0.5 cm to 1 cm with just enough space to admit the flange on the tandem.
826
John L. Horton
are loaded with 10 mgRaeq. The clinical decision on the specific loading for each case is based on the extent of the disease, the dose received from external beam treatment, patient-applicator geometry, and mgRaeq-hours following the rules first proposed by Fletcher (Fletcher 1980).
Treatment Planning The majority of the tandem and ovoid treatments do not have computer-planned isodose curves and surfaces generated until after the completion of the treatment. The plane of most interest to the physicians is the lateral throw off plane. This plane passes through the internal os and the “center of activity” in the ovoids. The center of activity of the ovoids is found by drawing two lines. Each of these lines connects the posterior end of one ovoid source to the anterior end of the opposite ovoid source. The intersection of these two lines is defined to be the “center of the ovoid activity.” The treatment is evaluated on the basis of the 3150 cGy line in the lateral throw off plane. The choice of the 3150 cGy line dates back to the Manchester system that frequently prescribed 7000 R in 7 days. Conversion from R to cGy gives a value of 6300 cGy. The typical UTMDACC intracavitary brachytherapy treatment is given as two 48-hour applications. The 6300 cGy line is divided by two to give 3150 cGy for each treatment. For the two intracavitary applications and the external beam treatment, the bladder dose is limited to 75 to 80 Gy, the rectal dose is limited to 70 to 75 Gy, and the vaginal surface dose is limited to 120 to 140 Gy. A limit of 6000 to 6500 mgRaeq-hours is also observed. By examining the AP and lateral films from the operating room, the experience of the physicians allows them to predict in most instances whether the bladder and rectal dose limits will be exceeded on the second application. If the physicians have any questions, the treatment is planned before the sources are loaded into the applicators. When the vagina is too narrow to accommodate miniovoids, tandem and cylinders are used. In this case, sources are loaded in the tandem, continuing into the section of the tandem passing through the cylinders. When disease extends along the vaginal wall, sources are loaded further into the tandem along the cylinders. For these cases the dose is prescribed to the surface of the cylinder and a plan is obtained before the treatment is administered. For very extensive vaginal wall disease the tandem and cylinder treatments are supplemented with an interstitial 192Ir wire implant during the tandem and cylinder treatment. A treatment plan is obtained to determine appropriate source loadings in this situation. The treatments of most of the patients in the last 50 years have been planned from orthogonal films taken in the operating room to evaluate the quality of the implant. Isodose curves have been generated for the pelvic brim plane, the Fletcher trapezoid or nodal plane, and the sagittal plane in addition to the lateral throw off plane. These planes are standard ICRU planes and are discussed in this monograph in the chapter (B. Thomadsen) on ICRU specifications (ICRU 1985). In the last few years we have begun a research project to evaluate delivered doses with CT planning. The greatest problem with CT planning is the CT artifacts caused by the shields in the ovoids. We have reduced these artifacts by appropriate adjustment of the window and level parameters of the CT image. This adjustment has allowed the physicians to outline the anatomy and we have been able to generate threedimensional plans with commercially available software for the intracavitary patients. The problem with the commercially available software is that it does not adequately account for the ovoid shields. To overcome this problem, we have also evaluated a limited number of patients by using Monte Carlo code MCNPX (Los Alamos National Laboratory, Los Alamos, NM) to calculate the dose distributions for individual patients based on the actual placement of the applicators. This Monte Carlo evaluation requires an extreme amount of computational overhead and personnel time that prevents us from doing this on a routine basis. We are currently investigating the possibility of evaluating these implants using deterministic techniques for radiation transport for this evaluation. Our early results indicate that we may be able to reduce the time to perform these calculations by at least an order of magnitude (Mourtada et al. 2004).
Intracavitary Brachytherapy Applicators and Intracavitary Brachytherapy Techniques 827
References Delclos, L., G. H. Fletcher, V. Sampiere, and W. H. Grant. (1978). “Can the Fletcher gamma ray colpostat system be extrapolated to other systems?” Cancer 41:970–979. Delclos, L., G. H. Fletcher, E. B. Moore, and V. A. Sampiere. (1980). “Minicolpostats, dome cylinders, other additions and improvements of the Fletcher-Suit afterloadable system: Indications and limitations of their use.” Int J Radiat Oncol Biol Phys 6:1195–1206. Fletcher, G. H. Textbook of Radiotherapy. Third Edition. Philadelphia, PA: Lea & Febiger, 1980. Gerbaulet, A. and A. Bridier. “Institut Gustave Roussy Method” in A Practical Manual of Brachytherapy by B. Pierquin and G. Marinello. F. Wilson, B. Erickson, and J. Cunningham (trans.). Madison, WI: Medical Physics Publishing, pp. 176–181, 1997. Gifford, K. A 3-D CT Assisted Monte Carlo Evaluation of Intracavitary Brachytherapy Implants. The University of Texas Graduate School of Biomedical Sciences. Ph.D. dissertation. Houston, TX, 2004. Gifford, K., J. Horton, E. Jackson, T. Steger, M. Heard, F. Mourtada, A. Lawyer, and G. Ibbott. (2005). “Verification of Monte Carlo calculations around a Fletcher Suit Delclos ovoid with radiochromic film and normoxic polymer gel dosimetry.” Med Phys (Accepted for publication). Henschke, U. K. (1960). “‘Afterloading’ applicator for radiation therapy of carcinoma of the uterus.” Radiology 74:834. International Commission of Radiation Units and Measurements (ICRU). Report No. 38. “Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology.” Bethesda, MD: ICRU, 1985. Katz, A., and P. J. Eifel. (2000). “Quantification of intracavitary brachytherapy parameters and correlation with outcome in patients with carcinoma of the cervix.” Int J Radiat Oncol Biol Phys 48:1417–1425. Ling, C., M. Schell, K. Working K. Jentzch, L. Harisiadis, S. Carabell, and C. Rodgers. (1987). “CT-assisted assessment of bladder and rectum dose in gynecological implants.” Int J Radiat Oncol Biol Phys 13:1577–1582. Mourtada, F., T. Wareing, J. Horton, J. McGhee, D. Barnett, K. Gifford, G. Failla, and R. A. Mohan. (2004). “A deterministic dose calculation method applied to the dosimetry of shielded intracavitary brachytherapy applicators.” Med Phys 31:1807. Pelloski, C. E., M. Palmer, G. M. Chronowski, A. Jhingran, J. Horton, and P. J. Eifel. (2005). “Comparison between computed tomography-based volumetric calculations and ICRU reference point estimates of the doses of radiation delivered to the bladder and rectum during intracavitary radiation therapy for cervical cancer.” Int J Radiat Oncol Biol Phys (Accepted for publication). Pierquin, B., and G. Marinello. “Créteil Method” Chapter 10 Part B, in A Practical Manual of Brachytherapy by B. Pierquin and G. Marinello. F. Wilson, B. Erickson, and J. Cunningham (trans.). Madison, WI: Medical Physics Publishing, pp. 167–175, 1997. Weeks, K. J., and G. S. Montana. (1997). “Three-dimensional applicator system for carcinoma of the uterine cervix.” Int J Radiat Oncol Biol Phys 37:455–463. Weeks, K. J., S. L. Schoeppel, K. Pruss, M. Hopkins, and C. Perez-Tamayo. (1989). “Computer tomography-compatible afterloading Fletcher-Suit-Delclos colpostat with adjustable shielding.” Endocurie Hypertherm Oncol 5:169–174.
Chapter 46
ICRU Reporting for Cervical Brachytherapy Bruce R. Thomadsen, Ph.D. Department of Medical Physics University of Wisconsin, Madison, Wisconsin Reporting and Prescribing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 829 Conventional Brachytherapy (ICRU Report 38) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 830 Treatment dose—Reference Isodose Dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 830 Doses to Regional Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832 Bladder—Foley Balloon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832 Rectum—0.5 cm Below Vaginal Wall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 832 Lymph Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 833 Physical Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 833 Three-Dimensional Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 834 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 836
Reporting and Prescribing “Prescription” describes a treatment in an unambiguous manner to the person who will deliver the treatment.1 Often, prescriptions provide the necessary information for treatment in terminology understood by the facility personnel, but would not convey the total thrust of the treatment to an outsider. An example from the chapter on Manchester and Paris interstitial implants would be a prescription to give a patient 20 Gy with the implant. Were the implant planned using one or the other system, the dose delivered would be considerably different. The treating facility would know the intention of the implant, but someone from another facility that may see the patient at a later date would find the prescription insufficient to know what had been done to the patient. Reporting protocols provide a structure to facilitate exchanging information on patients. Reporting protocols may specify how to describe an individual’s treatment or may address how to summarize a group of patients. Any reporting protocol needs to describe the dose distribution in terms of the target tissues and the neighboring regional structures at risk. Obviously, the isodose distribution and all information about the source strength and the time-course of the treatment could be passed along to provide the information about a patient’s treatment. Such an approach would not facilitate comparing that patient’s treatment with other patients, nor would it be feasible in the context of summarizing a group of patients for an analysis or study. Some sort of abstract of the data would serve much better—quantities that prove important in the outcome of treatment. Report 38 of the International Commission on Radiation Units and Measurements (ICRU 1985) lays a foundation for describing an individual patient’s treatment. It does not attempt to propose a formalism for prescribing treatments, only for communicating important aspects of the treatment. While addressing a single patient, some of the quantities specified could be used in aggregate during analyses of a series of patients. ICRU report 38 considers only treatments planned using conventional, radiographic localization. Thus, because there is no volume imaging information, assessment of the adequacy of target coverage remains impossible, as does any true evaluation of doses to regional structures. However, the real limitation of the report, and the reason that few adopted many of its reporting recommendations, was that it failed to use terminology that was already common in describing treatments and, instead, presented a protocol that many practitioners considered irrelevant. While not particularly wise in the selection of the reporting 1
It may seem strange at times in brachytherapy when the person delivering the treatment may be the same as the person prescribing it.
829
830
Bruce R. Thomadsen
approach, the rationale behind the commission’s recommendations does give a different insight into intracavitary dosimetry than normally encountered.
Conventional Brachytherapy (ICRU Report 38) Treatment Dose—Reference Isodose Dimensions Most intracavitary cervical brachytherapy practiced in the English-speaking world either follows the convention of the Manchester system, and specifies the amount of dose given to the patient as the dose to points A, or alternatively, the practitioner may follow the University of Texas M.D. Anderson system, and not report the dose in a target-oriented manner, but specify the treatment in terms of mg·h or the equivalent, and the loading of the applicator. The French custom was different. Consider Figure 1. Shown are two treatments, identical except that the one on the right delivers more radiation to the patient. In the treatment on the left, the light gray isodose line, which may correspond to 60 Gy, passes through point A, while on the right, the 85 Gy isodose line passes through point A. Thus, one method of comparing these two applications entails simply observing these two doses. Thus, the patient on the right receives (85 Gy/60 Gy) = 1.42 times the dose to the patient on the left. This holds true at any position. If one is used to thinking in terms of the dose to point A, knowing that one patient received 60 Gy and the other 85 Gy conveys an idea about the treatment each received. Similarly, the total reference air kerma (TRAK) for the treatment, that is, the total amount of radiation emitted from the sources in terms of the sum of each source’s integrated reference air kerma over the duration of the treatment, also would be in proportion of 1:1.42. An alternative method of understanding the increased dose received by the patient on the right comes from looking at the projection of the dose away from the applicator. The horizontal arrows in the figure show the distance from the center of the applicator, at the position of point A, that the 60-Gy line falls. In the patient on the right, this distance is 2.3 cm, while on the left it is 2.0 cm. Obviously, the more radiation the patient receives, the farther from the applicator the isodose line projects. The distance this line
Figure 1. Two applications, identical except for the duration of the treatment. The application on the left delivered 60 Gy to point A, while that on the right gave 85 Gy. The horizontal arrow shows a different way of expressing the relative exposures, with the 60 Gy isodose surface falling 2 cm from the applicator and for the one on the right, 2.3 cm.
46–ICRU Reporting for Cervical Brachytherapy
831
projects then also can serve as a measure of the amount of treatment, assuming that one is used to looking at the projection distances for this purpose. Unlike the dose-at-a-point method, the relative projections distances in different directions do not remain constant, and it is not always clear from where the measurement should be made. Instead of measuring the distances from the applicator, a less ambiguous approach measures the thicknesses of the isodose lines. One technique first looks at the length of the specified isodose surface in the plane containing the tandem (d l, see Figure 2). Next, the maximum width of the surface is taken perpendicular to the height, almost always through the ovoids (dw). Finally, the maximal height in the direction orthogonal to the length and width is measured (d h). These three dimensions give the idea of how much radiation the patient receives. It is important to note that the three dimensions are never to be multiplied, as if to form a volume, since the actual volume of the isodose surface is much smaller and more complex. In addition to the lack of consistency in the projection of the isodose surface with respect to direction for a given increase in dose, the maximal-dimension approach also complicates the interpretation of the dimensions because much, if not most, of the length and width of the isodose surface is due to the physical size of the applicator within the isodose surface. In part, advocates would say that that gives information about the application and the size of the dose distribution in the patient. However, a further problem is that as the distance to the isodose surface increases, the shape of the surface tends to become more round, losing the information about the shape of the applicator. While the particular isodose surface selected to describe treatments remains arbitrary, although it must be constant over the group of patients considered—and ideally universally agreed upon, values too high simply take on the shape of the appliance with a margin, and too low, as noted, lose shape information. In ICRU report 38, the commission selected to describe the amount of radiation given to the patient in terms of the three dimensions of the 60 Gy isodose surface, and forego any reference to doses at points. They also required reporting the TRAK. This reporting method followed closely the practice common in some centers in France. The selection of the 60 Gy isodose surface corresponded to the dose of most interest in these practices, which mostly treated patients preoperatively, and thus delivered relatively low doses. For most curative regimens, doses (to point A) more usually fell in the range of 85 Gy, and the 60 Gy line projected far from the applicator. The concept was further complicated because the actual dose of the isodose surface to be used to describe the brachytherapy was 60 Gy minus the dose delivered by any external beam treatment. If a whole
Figure 2. The three standard, orthogonal dimensions of an intracavitary cervical application: on the left, showing the reference width, dw; on the right the reference length, dl, and the reference height, dh.
832
Bruce R. Thomadsen
pelvis field of 45 Gy was treated before the brachytherapy, then the brachytherapy isodose surface of (60 Gy – 45 Gy) = 15 Gy would provide the basis for the describing dimensions. For those facilities using a dose of 85 Gy for their patients, following the external beam, the 15 Gy isodose surface would correspond to the (15 Gy/40 Gy) = 37.5% isodose curve, well outside of the normal range of doses considered. Because the specification methodology seemed strange to most practitioners, and the isodose surface used as the basis for the specification dimension fell outside that normally of interest, few practitioners adopted the ICRU report 38 recommendations for treatment specification.
Doses to Regional Structures While the treatment specification formalism failed to take hold in the gynecological radiotherapy community, the recommendations for specifying the dose to regional structures did. Probably the most important contribution of the report was standardization of the reporting of the dose to the bladder and the rectum. Bladder—Foley balloon Of course, the bladder proper does not show on radiographs. Filling the bladder with contrast will provide an image of its volume, but without means of correlating a point on one image to the same point on the requisite second image needed to calculate a dose. To provide a calculation point, the report suggested placement of a Foley balloon in the bladder filled with ~7 cm3 of contrast. The location of the bladder dose point falls on the posterior aspect of the balloon (Figure 3). Rectum—0.5 cm Below Vaginal Wall Specifying a point to represent the rectum posed a greater challenge. Placing any device in the rectum unlikely would show the anterior wall, the anatomy of interest, because gravity would pull it to the posterior side. Anything that inflates to push against the anterior wall would cause distortion. Coating the rectal wall with an appropriate concentration of contrast so as not to obscure the source information and yet visu-
Figure 3. Radiograph showing the ICRU report 38 bladder and rectal points.
46–ICRU Reporting for Cervical Brachytherapy
833
alize the rectum takes a great deal of skill. The final recommendation for localizing a rectal point began with opacifying the posterior surface of the vagina (for example, with radiopaque gauze). The next step entailed drawing a line from the center of the ovoid sources (or the inferior-most tandem source—either was allowed) in the posterior direction on the lateral radiograph. The rectal point fell on that line, 0.5 cm beyond the posterior vaginal wall. (See Figure 3.) Lymph Nodes The contribution of the intracavitary brachytherapy to the regional lymph nodes, while not curative, needs to be included in their tally. Two methods were described, both based on skeletal anatomy. Pelvic-Wall Reference Point. Following Chassagne and Horiot (1977), the pelvic-wall reference points, intending to locate the obturator nodes, were found on orthogonal radiographs by first drawing a line on the antero-posterior (AP) radiograph connecting the superior aspects of the acetabula. Lines were then drawn perpendicular to this initial line, tangential to the medial aspects of the acetabula. Their intersection gave the positions of the nodes on the AP radiograph. The location on the lateral radiograph was the midpoint of the line connecting the superior aspects of the acetabula. This approach as described is not quite consistent. Figure 4 shows that the correct projection on the lateral image would be from the respective acetabulum a distance (dp/sAP)•slat towards the opposite acetabulum. Where dp is the distance from the superior aspect of the acetabulum to the pelvic wall point on the AP projection; sAP is the separation of the superior aspects of the acetabula on the AP projection; and slat is their separation on the lateral projection. Lymphoidal Trapezoid. Fletcher proposed a method to locate several other lymph node positions using the lymphatic trapezoid (Durrance and Fletcher 1968). Figure 5 shows the process.
Physical Parameters The ICRU recommendations also require reporting the physical input aspects of a treatment. These would include: • Sources - Radionuclide - Reference air kerma for each source - Model of each source - If point-like sources or moving sources, the pattern of sources or movement. • Applicator - Model, if description has been published - Otherwise, ° If rigid as a whole ° Whether the tandem is rigid with a fixed curvature ° Connection between the tandem and vaginal appliance ° Number and orientation of vaginal sources ° Shielding in the vaginal appliance. • Total Reference Air Kerma
834
Bruce R. Thomadsen
Figure 4. Locating the pelvic-wall points: on the left on the AP radiograph, and on the right on the lateral.
Figure 5. The lymphatic trapezoid: on the left on the AP radiograph, and on the right on the lateral.
Three-Dimensional Brachytherapy Early in the use of volume imaging for cervical brachytherapy, it became evident that the location indicated by the ICRU bladder and rectum points failed to correspond with the maximum doses to these structures (c.f., Schoeppel et al, 1989, 1994). In fact, the maximum dose may be several times the dose to the nominal points, particularly for the bladder that tends to droop over and around the uterus superiorly. With the ability to perform image-guided intracavitary brachytherapy becoming more common, there has been a tendency to try to treat brachytherapy more like either external-beam or interstitial implants, by locating a target volume and specifying a dose that conforms to that volume. Two problems hinder that approach: (1) that the actual target for cervical cases is not well known; and (2) that the dose the target requires is also not known. The dose in intracavitary brachytherapy continuously decreases with distance from the applicator. Thus, unlike the sharply defined beam in external radiotherapy or even in interstitial brachytherapy, there is no clear location that defines where the treatment stops. It might well be that the dose that carries beyond the cervix affects importantly the outcome of the treatment. Since we have also not known in the past where the actual edge of the cervix or a tumor lies on a treatment plan (there is no
46–ICRU Reporting for Cervical Brachytherapy
835
reason to think that point A demarcated that location), we do not know what doses we have been giving to whatever we might designate as a target. Given the uncertainties, it seems premature to jump into targetdriven treatments. For that reason, an intersociety group, lead by the Gynecological Oncology Group, has suggested a process to prepare for the eventual true image-guided intracavitary brachytherapy. (Nag et al., 2004) The first step is to perform volume imaging that will allow dose distributions to be calculated for contoured targets and organs at risk, but to treat the patient in the conventional manner. The data from these treatments will be gathered in a national registry, and when sufficient experience has accrued, correlated with outcome, the target and the doses to prescribe to targets can be determined. Thus, the reporting guidelines from this group are aimed at researchers who will publish results of their studies. One important restriction in these guidelines is that they only allow magnetic resonance imaging (MRI) for delineation of the targets. The reasons for this are discussed in this monograph in chapter 42, Imaging in Gynecological Brachytherapy, but in summary, only MRI differentiates the uterus and uterine tumors from the surrounding pelvic material. A summary of the recommendations follows. 1. The dose prescription method presently used in a particular center should not be changed. 2. The integrated reference air kerma and the dose to point A should be reported. 3. For research purposes, individual centers and cooperative groups should collect the image-based dosimetry information proposed in this report. 4. T2-weighted MRI using a pelvic surface coil is to be used for image-based cervical brachytherapy. Positron emission tomography/computed tomography (PET/CT) fusion should be investigated for treatment planning. 5. MRI-compatible applicators and appropriate software will be used. 6. The following terminology and approach should be reported: a. GTV(I): Gross tumor volume as defined through imaging. b. GTV: Defined as GTV(I), plus any clinically visualized or palpable tumor extension. c. GTV(CX): Defined as the GTV plus the entire cervix. d. Dose volume histogram (DVH) of GTV, GTV(I), GTV(CX) should be calculated, and the D100, D95, D90 (dose to 100%, 95%, or 90% of GTV) reported. e. V100 (% of GTV covered by point A dose) is to be reported. f. pCTV: is defined as the primary tumor clinical target volume, which equals the GTV(CX) plus a 1-cm margin. g. rCTV: is defined as the pCTV plus regional lymph nodes. h. CTV: includes the pCTV and the rCTV, all of which are to be included in the external beam radiotherapy field. i. For normal tissue doses, the inner and outer wall of the bladder and rectum are to be contoured and absolute DVH of the organ wall is to be performed. j. For bladder doses: The ICRU report 38 bladder point dose, the maximum bladder dose, and the maximum doses to contiguous 1 cm3 and 5 cm3 volumes of bladder (BladderV1cc, BladderV5cc) are to be reported. k. For rectal doses: The ICRU report 38 rectal point dose, maximum rectal dose, and the maximum doses to contiguous 1 cm3 and 5 cm3 volumes of rectum (RectumV1cc, RectumV5cc) are to be reported. l. For small bowel doses: The maximum doses to contiguous 1 cc and 5 cc volumes of small bowel (Small BowelV1cc, Small BowelV5cc) are to be reported. 7.
MR images should be fused with CT images if required for dose calculation.
836
Bruce R. Thomadsen
The European brachytherapy group, GEC-ESTRO, also is currently working on a proposed protocol for imaged-based cervical treatment planning. The future prospects for a unified reporting approach appear promising.
References Chassagne, D., and J. C. Horiot. (1977). “Proposals for common definitions of reference points in gynecological brachytherapy.” J Radiol Electrol Med Nucl 58:371–373. (French.) Durrance, F. Y., and G. H. Fletcher. (1968). “Computer calculation of dose contribution to regional lymphatics from gynecological radium insertions.” Radiology 91:140–148. International Commission of Radiation Units and Measurements (ICRU). Report No. 38. Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology. Bethesda, MD: ICRU, 1985. Nag, S., H. Cardenes, S. Chang, I. Das, B. Erickson, G. Ibbott, J. Lowenstein, J. Roll, B. Thomadsen, and M. Varia. (2004). “Proposed guidelines for image-based intracavitary brachytherapy for cervical carcinoma: Report from Image-Guided Brachytherapy Working Group.” Int J Radiat Oncol Biol Phys 60:1160–1172. Schoeppel, S. L., B. A. Frass, M. P. Hopkins, M. L. La Vigne, A. S. Lichter, D. L. McShan, S. Noffsinger, C. PerezTamayo, and J. A. Roberts. (1989). “A CT-compatible version of the Fletcher system intracavitary applicator: clinical application and 3-dimensional treatment planning.” Int J Radiat Oncol Biol Phys 17:1103–1109. Schoeppel, S. L., M. L. LaVigne, M. K. Martel, D. L. McShan, B. A. Fraass, and J. A. Roberts. (1994). “Three-dimensional treatment planning of intracavitary gynecologic implants: analysis of ten cases and implications for dose specification.” Int J Radiat Oncol Biol Phys 28:277–283.
Chapter 47
Physics of Uterine Corpus Brachytherapy Bruce R. Thomadsen, Ph.D. Department of Medical Physics University of Wisconsin, Madison, Wisconsin Current Treatment Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 837 Postoperative Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 837 Rationale: Prophylactic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 837 Requirements for the Dose Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 838 Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 838 Vaginal Cylinder . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 838 Vaginal Ovoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 839 Comparisons Between Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 840 Brachytherapy for Inoperative Corpus Cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844 Patient Characteristics and Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 844 Target . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845 Techniques and Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845 Historical Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845 Heyman Capsules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 845 Optimization for HDR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 850 Preoperative Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851 Rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851 Target and Treatment Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 851
Current Treatment Approaches The uterus consists of two sets of two regions: the superior aspect of the uterus called the “corpus,” or body of the uterus, and the inferior part where the organ narrows called the “cervix,” or neck of the uterus. Looking at it differently, the inside, mucosal lining of the uterus, is the endometrium, while the outer, muscular coat is the myometrium (see Figure 1). In this chapter, we will be concerned with treatment of cancer of the endometrium in the corpus. Cancer of the myometrium is a very different disease. In general, cancer of the endometrium is a disease treated primarily with surgical removal of the uterus, a hysterectomy. The nature of radiotherapy in this disease is very strongly shaped by the goals of the treatment.
Postoperative Brachytherapy Rationale: Prophylactic As noted, the vast majority of patients receive curative hysterectomies. However, with no further therapy, approximately 12% of the patients are likely to suffer recurrences (Graham 1971). The anastomosis (where the vagina is closed after the excision of the uterus, call the “vaginal cuff”) forms the most likely site of recurrence. In Graham’s series, patients who received prophylactic radiation to the vaginal cuff had no recurrences.
837
838
Bruce R. Thomadsen
Figure 1. A schematic illustration of the uterus and vagina, on the left in a coronal section and on the right in a sagittal. The dotted line in the coronal indicates the rough division between the endometrium and the myometrium.
Requirements for the Dose Distribution The target need only encompass the dome of the vagina—a very small volume compared to most radiotherapy. In delivering the treatment, care must be taken to avoid delivering excessive dose to the bladder or rectum, particularly since the therapy is prophylactic.
Techniques Depending on the extent of myometrium invasion, radiotherapy may include external beam and brachytherapy (often considered for more than 50% invasion) or brachytherapy alone (less than 50%). In the later case, low dose rate (LDR) doses to the vaginal surface fall in the range between 50 Gy in two fractions to 70 Gy in a single fraction. For high dose rate (HDR) brachytherapy, a regimen of three fractions of 12.2 Gy has been used at the University of Wisconsin. When combined with external beam, the dose for the brachytherapy component varies with the external-beam dose.
Vaginal Cylinder Probably the most common approach to vaginal cuff irradiation uses sources in a plastic cylinder placed in the vagina. Figure 2 shows some typical examples. The cylinders come in a variety of diameters and lengths to better conform to the patient’s anatomy and to maximally extend the mucosal lining for decreasing dose gradients. One attraction of the cylinders is the simplicity of use—insertion requires little skill or time. The cylinder should closely fit to the superior aspect of the vaginal cuff and contact the lateral surfaces of the vagina.
47–Physics of Uterine Corpus Brachytherapy
839
Figure 2. A variety of LDR and HDR vaginal cylinders. The three on the left are for LDR, preloaded treatments, while the one on the right is used with HDR treatments. Because the insertion is so quick, and post-procedural imaging is often not necessary, using the loaded, LDR 137Cs applicator seldom results in significant exposures to personnel.
For LDR applications, the lips often are sewn closed to hold the cylinder in place. Other options for maintaining LDR applicators in place include a base-plate and support jig or specially designed pants. For HDR applications, the tube that guides the source through the tandem provides fixation by locking to a brace on the table or baseplate.
Vaginal Ovoids Treating the vaginal cuff with ovoids uses part of the same treatment appliance used for cervical cancer; a tandem and ovoid, simply without the tandem because surgery removed the uterus into which the tandem would have been inserted. The ovoids usually come in three or four sizes, as shown in Figure 3 and described in Table 1. The sizes tend to be the same for LDR and HDR applications, although some sets do differ. Just as with the vaginal cylinders, the largest size ovoids that fit in the vagina should be used. Mixing sizes of ovoids in a given treatment should be avoided to give the most uniform dose distribution. Because there is no tumor present, there seldom is need to mix ovoid size. During the insertion, the ovoids are placed in contact with the superior and lateral surfaces of the fornices, and touching each other medially. Separations between the ovoids must be avoided due to “cool spots” in the dose distribution. The decrease in the dose from the ovoid surface into the tissue mostly follows between an inverse square and an inverse linear relationship with the distance. Table 1 gives the relationship between the doses at the surface of the ovoids and those 0.5 cm deep to the surface. Because of the constant applicator geometry and the usually low doses to the bladder and rectum, ovoid treatment of vaginal cuff may proceed following localization using a standard set of dwell times (corrected for source decay) without dose distribution calculations and/or optimization. The treatment duration or dwell times come from performing the normal dosimetry procedure on an idealized case—possibly films
840
Bruce R. Thomadsen
Table 1. Ovoid Sizes and the Dose 0.5 cm Deep and Lateral to the Surface of Vaginal Ovoids Relative to the Dose at the Ovoid Surface for Post-hysterectomy Treatments (Including Contributions from Both Ovoids) Ovoid type
Ovoid radius* [cm]
Dose at 0.5 cm relative to the surface dose
Mini
0.8 or 1
0.52
Small
1
0.52
Medium
1.25
0.56
Large
1.5
0.62
*Radius is the distance from the source to the lateral surface. The distance to the medial surface is smaller for the “mini” ovoids to allow insertion into very small vaginas. The distance to the lateral surface for “mini” ovoids can vary by appliance design.
Figure 3. Vaginal ovoids: leftmost, for HDR applications, showing only the small size caps; the middle three, the Fletcher-Suit ovoids in the three sizes for LDR applications; right, the Fletcher-Suit-Delclos in the mini configuration with no caps, also for LDR cases.
taken of the ovoids without a patient. The actual localization films for a patient serve first to assure that the applicator appears properly positioned, for example, near a surgical clip placed at the superior aspect of the vaginal cuff, and then to check the displacement of the bladder and rectum. The same films later allow post-procedure evaluation of the doses to the cuff and other structures.
Comparisons Between Techniques The two approaches have different dosimetric characteristics due to their geometries.
47–Physics of Uterine Corpus Brachytherapy
841
Loading and Dosimetry. Typical LDR loadings for vaginal cylinders place two 137Cs sources coaxially in the cylinder as shown in Figure 4, with a 0.5 to 2 cm of plastic between the ends of the source and cylinder. The strengths of the sources depend on the diameter of the cylinder. Typically, because of the relatively small volume treated, dose rates run around 0.8 to 1 Gy/h. HDR applications use cylinders of the same general design, and dwell positions that tend to simulate the two LDR sources. The step size may be up to 0.5 cm, although finer step sizes can sometimes improve the optimization. Because of source anisotropy, the dose distribution pulls inward toward the superior, axial direction. The decrease is likely greater for LDR applications than that shown by the dose calculation program because few programs consider intersource or applicator attenuation. LDR ovoids each carry one source with the strength dependent on the ovoid diameter. Table 2 gives the common loadings and the resultant dose rates. The dose rates in the table are specified at the lateral surface. HDR applications follow loadings as in Table 3, which produce isodose distributions typical of those in Figure 5. HDR applications generally use dwell positions that fall approximately 1 cm of the ends of the ovoid (microSelectron dwell positions 2–8 for 2.5 mm step sizes; VariSource, dwell positions covering 1.5 cm
Figure 4. An example of LDR loading for a vaginal cylinder.
Table 2. LDR Loadings and Lateral Dose Rates for Vaginal Ovoids Ovoid Size
Radius [cm]
Typical Loading [mgRaeq]
Approximate lateral Dose rate [Gy/h]
Small
1
10
0.82
Medium
1.25
15
0.79
Large
1.5
20
0.73
842
Bruce R. Thomadsen Table 3. Typical Loadings for HDR Applications Values give the Ci·s/Gy for dwell positions 2, 4, 6, and 8 for a Nucletron microSelectron using a standard unshielded (Williamson) applicator. These positions fall ±2.5 mm and ±7.5 mm from the center of the ovoid, respectively. Ovoid Size
Radius to lateral surface [cm]
Typical Loading [Ci·s/Gy]
Mini
1
310.2, 306.9, 302.0, 299.2
Small
1
370.1, 362.8, 352.2, 346.4
Medium
1.25
532.7, 527.8, 520.6, 516.1
Large
1.5
737.0, 733.5, 727.5, 723.6
Figure 5. Isodose distributions for vaginal ovoids.
starting 1 cm from the end of the tube). Activating dwell positions closer to the ends serves no therapeutic function but increases the doses to the bladder or rectum. As with the cylinders, the step size can be up to 0.5 cm between dwells. For treatments using either applicator, the prescription may be specified at the surface of the applicator or at 0.5 cm deep to the surface. The usual prescription specifies the dose at the lateral surface of the applicator, even though the crucial target falls superior. With the cylinders, as noted above, the superior directions reflect the decreased dose due to source anisotropy, which should be accounted for if prescribing laterally. Ovoids tend to have a very slight decrease midway between the two sides in the superior direction, but less than the cylinder. If prescribing at 0.5 cm deep to the lateral surface, the dose in the superior direction is increased slightly from the lateral. To compensate for the undesired isodose distribution in the superior direction with cylinders, pre-insertion-loaded LDR applicators often are oriented the cephalad-most source perpendicular to the cylinder axis, as shown in Figure 6. While such applicators were common with preloaded appliances, the technology to accomplish the same geometry with an afterloading applicator was more difficult. Wang (1975) developed an afterloading applicator that accomplished this orientation, as shown in Figure 7.
47–Physics of Uterine Corpus Brachytherapy
843
Figure 6. A pre-insertion loaded cylinder with the superior source perpendicular to the cylinder axis.
Figure 7. The Wang applicator. (Figure courtesy of Mick Radio-Nuclear Instruments, Inc.)
With HDR applications as described above, the decrease in dose superiorly is less than with LDR applications, and in part, increasing the cephalad dwell time can compensate for the decrease, albeit at the expense of normal tissues. Dose to Non-Target Vagina, Rectum, and Bladder. The anisotropy of the sources and their orientation in the cylinder projects the most intense doses in the direction of the rectum and the bladder. In addition, the location of the cylinder in the center of the vagina minimizes the distance to these normal tissue structures. Contrarily, the ovoids place the sources with the maximum dose projected in the direction of the
844
Bruce R. Thomadsen
vaginal cuff and minimum dose towards the rectum and bladder. The location of the ovoids to the sides of the vagina adds distance to the rectum and bladder, and allows the insertion of a speculum posterior to the ovoids. The inserted speculum pushes the rectum away from the sources. The addition of packing anterior to the ovoids adds distance the bladder. Neither form of spacing is possible with a well-placed cylinder. Thus, for the same dose to the prescription point, the cylinder delivers increased doses to the rectum and bladder compared with ovoids. In addition to the natural difference in dose distributions between the two applicators, ovoids often incorporate partial shielding in the direction of the rectum and bladder (Figure 8). The actual shape of the shielding varies with the model of the applicator. Incorporating shielding into a cylinder becomes more challenging. LDR cylinders may rotate in the vagina, so the shielding may not shadow the intended structures, and target structures may be mistakenly protected. HDR cylinders can lock into position, keeping the shielding appropriately in place. No treatment-planning system available at the time of writing correctly incorporates the effect of shielding into the dose distributions, and isodose distributions that show shielding should not be believed. For the most part, the programs simply use a constant, in-line attenuation through the shielding material, greatly underestimating the true dose behind the shielding. For an analysis of the effect of the shielding in the Fletcher family of applicators, see Haas et al. (1985).
Brachytherapy for Inoperative Corpus Cancer Patient Characteristics and Rationale While surgery forms the treatment of choice for corpus cancer, sometimes patients present with other conditions that contraindicate surgery. Because cancer of the endometrium is phenotypically related to
Figure 8. Linacogram of Fletcher-Suit ovoids showing shielding.
47–Physics of Uterine Corpus Brachytherapy
845
obesity, often the causes for inoperability follow the obesity. The reasons seldom have to do directly with the cancer.
Target The target tissues include the endometrial lining of the corpus, the entire cervix, and the vagina near the cervix. The sensitive neighboring organs include the obvious rectum and bladder, but possibly the superior bowel is at greater risk for radiation injury.
Techniques and Dosimetry Historical Approaches Some of the earliest techniques for treating the intact uterus simply used LDR radium sources in a tube in the uterine canal. The challenge to this approach is to deliver sufficient dose to the lateral cornu before giving the canal mucosa an intolerable dose. In Manchester, they preferred using a single, 1-cm thick, rubber tandem in the uterus (Paterson and Cole 1963) to reduce the mucosal dose. While they specified the dose to point A (recognizing that the thick tandem moved point A 0.5 cm laterally, but not commenting on the 0.3 cm lateral shift due to the thin tandem in the cervical cases) as “8000 R” (~76.8 Gy), the dose to a point 3 cm lateral near the top source in the tandem received about 46 Gy. Many authors recognized that placing sources in a single tandem delivered an insufficient dose to much of the upper uterus [see Nolan and Natoli (1948) and Freed and Pendergrass (1954) for reviews]. Several approaches evolved to increase the amount of radium in the uterine cavity, as shown in Figure 9. Some of the techniques included placing radium sources in a sheath tied with loops, and placed into the cornu, with a tandem in the middle of the uterus (Crossen 1946); radium sources attached to long sticks by hinges to allow the sources to fall into the cornu and fill the uterine cavity (Martin 1940); the “hysterostat,” consisting of a radium source on a remotely controlled swivel that rested flat against the fundus and two to four variablelength tandems to fill the volume of the uterus (Friedman 1940); a “Y” shaped applicator where the angle of the arms of the Y could be adjusted using a handle that protruded through the vagina so that the arms fit into the cornu (Schmitz and Schmitz 1935); a flexible, spring ring that holds radium sources (more or less) in contact with the uterine wall (Kaplan 1942); and several others. Nori and the group from Memorial Sloan-Kettering (Nori et al., 1982) adapted the three-tandem approach, with one straight tandem in the middle of the organ, and two curved tandems with the tips pointed laterally to facilitate afterloading. Heyman Capsules Development and description. The main reason for developing the Heyman capsule-type approach to treat the endometrium lay in shifting the dose from the cervix, where the traditional tandem and ovoids concentrate the dose (the wide part of the “pear-shaped” distribution), toward the fundus (an upside-down pear). With the Heyman technique, short LDR sources were placed in metallic capsules and inserted into the uterus (Heyman, Reuterwall, and Benner 1941). As many capsules as would fit into the uterus were used, first to increase the activity in the uterine cavity, and second to stretch and thin the uterine wall. A tandem often was also used in the cervical canal with ovoids in the vaginal fornices. Each capsule had a numbered, attached steel wire to facilitate its removal. Removal would be in the reverse order from insertion to prevent tangling. Figure 10 shows a radiograph of a patient containing Heyman capsules. Over time, the design of the capsules available commercially has changed from the intentionally bulky style used by Heyman, through the sleeker Campbell type (Campbell 1946). Sharma and MacDonald (1980) demonstrated that the original tables of treatment times developed by Benner applied as well to the Campbell-type Heyman capsules (Heyman and Benner 1946). Afterloading became possible with the
846
Bruce R. Thomadsen
(a)
(b)
(c)
Figure 9. Schematics of various applicators used for treatment of intact uteri: (a) Crossen’s radium sausages; (b) (left) Martin’s hinged applicator, (right) Martin’s applicators in a uterus; (c) Friedman’s hysterostat with the swiveling central source and the two bent side sources; and (d) Schmitz’s “Y” applicator.
47–Physics of Uterine Corpus Brachytherapy
847
Figure 10. Radiograph of an application using Heyman capsule.
development of miniature 137Cs sources and Simon-type Heyman capsules (Simon, Silverstone, and Roach 1971). Figure 11 shows an application using such applicators. These nylon capsules provided tissue-equivalent attenuation, very unlike the steel and lead of the original capsules. These afterloading capsules changed the practice in several ways. First, the straws that led the sources to the capsule in the uterine cavity took more space in the uterine canal than the cables that connected the preloaded capsules to the outside world, and the uterine canal required more dilatation than before and accommodated fewer capsules. The new capsules on the straws were more rigid than the old capsules on the cable, so the capsules in the uterine cavity could not lie sideways to make room for more capsules as the packing proceeded. The old capsules could, and did, lie sideways, flush against the fundus, while the newer capsules could now do so. The fatter straws, however, almost eliminated the possibility of the dreaded tangling of cables inside the patient.
848
Bruce R. Thomadsen
Figure 11. Radiograph of an application using Simon-Silverstone afterloading Heyman capsules.
Dosimetry. Of all the types of applicators, Nolan and Natoli (1948) demonstrated that the Heyman packing technique gave the most uniform dose distribution to the target [following Heyman, Reutrman, and Benner (1941), Heyman and Benner (1946), and Heyman (1947)]. The original Heyman capsules contained a 2 cm long by 0.28 cm diameter, 8 mg radium source enclosed by lead and stainless steel for a total filtration of 3-mm lead equivalent. Several sizes of capsule diameter allowed for adapting the application to the size of the uterine cavity. The treatment consisted of inserting as many of the capsules as would fit into the cavity with the intention of stretching the wall of the uterus. Stretching the uterine wall served three functions: (1) to give a more uniform dose to the tumor and cells in the wall (presumably by thinning the target); (2) to give a higher dose to the target by decreasing the distance to the lateralmost extent of the uterus; and (3) to reduce the uterine folds and ridges (Fletcher 1966). All three functions actually appear to be simply that the target becomes thinner as the capsules stretch the uterus. Benner, the
47–Physics of Uterine Corpus Brachytherapy
849
physicist for Heyman’s group, calculated a table based on ionization chamber readings of treatment durations to deliver a specific dose to the myometrium 1.5 cm from the edge of the mass of capsules. The time in the table varied with the size of the capsule and the number of capsules used. Benner emphasized that the treatment duration could not be calculated by simply using a constant number of mg·h. In spite of this warning, at least one influential text says to use a constant 2500 mg·h twice, separated by 3 weeks preoperatively, or 3000 mg·h twice without a hysterectomy (or 2500 mg·h three times 2 weeks apart for a large uterus) (Fletcher 1966). These mg·h amount to approximately twice those suggested originally by Heyman, but over the years, the group at the Radiumhemmet in Stockholm (Heyman’s institution) increased their applications to about this same level. Nolan and Natoli calculated that about 6000 to 8000 mg·h were needed (depending on the number of capsules) to deliver the minimum curative “dose” of 7000 “r” (about 67 Gy) to all of the uterus. They, and other authors, also suggest that the maximum dose remain below 15,000 r to avoid serious complications. One of the biggest changes in Heyman-type applications accompanying the shift to the afterloading capsules came in the dosimetry. The heavy, steel capsules masked the locations of individual sources on the localization x-ray images. Thus, values of dose mainly derived from experiments in vitro. Because of the tissue-equivalent construction of the afterloading capsules, the localization images clearly show the dummies indicating the positions the sources occupy during treatment, allowing the calculation of dose distributions. However, the ability to calculate dose preceded the prescription of dose. For decades the prescription consisted of mg·h (due to the inability to calculate anything else). The experimental results of Benner only gave the number of mg·h for the same dose given different loadings, but left the absolute dose unspecified. Using the same mgRaeq·h as with the old capsules was not an attractive option because the dose would be different for several reasons: (a) the markedly decreased attenuation through the afterloading capsules and sources compared with the old capsules; (b) differences in the source filtration; (c) difference in the shape of the dose distribution around the new, microcesium sources compared to the old radium sources; (d) orientation of the new sources, that followed the body axis more than the old sources; and (e) more dilatation required of the cervical canal to accommodate the wider straws while passing the capsules. Notwithstanding all the differences, most practitioners seemed to have continued to use the same mg·h for the treatment with the afterloaded capsules as with the preloaded. Where to specify the dose also raised questions. Heyman’s original article stated that surgical specimens of the uterine wall measured 1.5 cm on the average. More recently, Bastin (Bastin et al., 1993), also measuring uterine wall thickness during surgery, found, on an average, that a point approximately two-thirds through the wall (i.e., in mid myometrium) fell approximately 3 cm from the uterine midline for normalsized uteri, and about 4 cm from mid uterus for large uteri. Ultrasound measurements in vivo tend to give smaller dimensions, with the mid-myometrium appearing to fall about 2 cm from mid uterus, more in line with the measurements presented by Heyman. Herbolsheimer and colleagues (1991) reported on transferring the Heyman capsule technique to HDR treatments. HDR brachytherapy delivered in fractions lasting a few hours as an outpatient, Herbolsheimer points out, resulted in fewer patients developing venous thrombosis, a common problem occurring in the obese patient population required to lie fairly still in bed for the duration of the LDR treatments. The group at Würzburg, following techniques developed by several authors (Rotte 1975; Herbolsheimer 1988, 1989; Rauthe, Vahrson, and Giers 1988;) delivers 50 Gy in five 10-Gy fractions to a point 2 cm inferior and 2 cm lateral to the bottom of the midpoint of the fundus. Bauer et al. (1981) redesigned the multiple-tandem
850
Bruce R. Thomadsen
applicator with soft tandems (3 or 4, depending on the anatomy of the uterus), pushed apart by a balloon once the applicator clears the cervical canal.
Optimization for HDR With the variable dwell time allowed with an HDR unit, shifting the dose distribution cephalad and outward only requires using relatively longer dwell times toward the tip of a tandem. Bastin et al. (1993) described the Madison system for HDR treatment of corpus cancer. The dose was specified for four regions, shown schematically in Figure 12. In the original publication, the distance to the points was based on measurements on surgical specimens. Since then, the current practice developed into using ultrasound at the time of the procedure to determine the anatomical dimensions. In almost all cases, the in vivo measurements yield smaller distances than those from the specimens. The system specifies the dose at four regions: Point M:
defined similarly to point A as used for cervical brachytherapy, except at the lateral extent of the cervical wall.
Point S:
(superior) defined as cephalad of the tip of the tandem, two-thirds of the thickness of the fundus.
Points W:
(wall) defined as 2 cm caudal from the tip of the tandem and lateral to the tandem, two-thirds of the thickness of the uterine wall.
Lateral vaginal surface:
defined as the cephalad 2.5 cm of the lateral surface of the vaginal cylinder.
Figure 12. Illustration of the dose prescription points for treatment of cancer of the corpus with the Madison system (after Bastin et al., 1993). The figure shows the points for a large uterus, defined as accommodating two tandems rotated as shown, and a small uterus, only allowing the placement of a single tandem.
47–Physics of Uterine Corpus Brachytherapy
851
According to Bastin et al., a small uterus requires a single tandem, as used for treatment of cervical cancer. A “large” uterus, by definition, requires two tandems, inserted with a 90° rotation from the normal position, with the tips pointed outwardly, and falling into the cornu. Either 15° or 30° tandems work well, but 45° tandems leave too much space between the tips to cover the fundus with the treatment dose. During optimization, a single optimization point defines the dose at point S. For the dose at points W, the cephalad-most optimization point falls at the lateral distance at the level of the first dwell position. Points W actually represent a region, so the optimization points continue at 1-cm increments toward the vagina for a total of 2 cm. A gap must separate the points specifying the doses to points W and those for points M due to the relatively large change in geometry between the two. Points M specification begins about 1 to 2 cm cephalad of points M proper, with optimization points placed 2 cm lateral to the uterine midline, and continues at 0.5 cm intervals through points M. The vaginal optimization follows the same procedure as for cervical carcinoma, except for the specified dose. With the PLATO treatment-planning system by Nucletron (Columbia, MD), the optimization program produces a tendency for the single tandem treatments to run point S too low for the given dose at points W, and the opposite for two tandem treatments. Obesity often complicates the dosimetry for these patients by degrading the localization images. Particularly, many x-ray units fail to penetrate adequately the patients laterally. In such cases, orthogonal films taken at angles such as 45˚ to the patient’s right and left may prove adequate, or a different method (such as stereo shift) may be required.
Preoperative Brachytherapy Rationale Preoperative brachytherapy for corpus cancer has fallen out of vogue, but is still performed in some institutions, and may return to the fore in the future. The rationale for preoperative radiotherapy is that the irradiation kills the more loosely bound peripheral (and oxygenated) cells in a tumor, so at the time of surgery, any cells left behind or dropped along an incision track will be less likely to grow (having been sterilized).
Target and Treatment Approach Since the surgery will remove the uterus, the goal in preoperative brachytherapy for corpus cancer is sterilization of tissues through which the surgeon will cut, and the tissues that will be left behind near the surgical site. The radiation target thus is the region around the cervix and the top of the vagina, and the brachytherapy treatments and planning become just like treatments for cervical cancer.
References Bastin, K. T., D. A. Buchler, B. R. Thomadsen, J. A. Stitt. (1993). “Heyman’s-based model for high dose rate multitandem intracavitary brachytherapy.” Endocuriether/Hypertherm Oncol 9:9–13. Bauer, M., D. von Fournier, F. Fehrentz, H. Kuttig, K. Winkel, and F. Neldner. (1981). “Afterloading-methode zur simulation der intrauternen packmethode beim korpuskarzinom.” Strahlenther 157:793–800. Campbell, L. A. (1946). “Capsules and inserter for the use of radium cells in the treatment of cancer of the uterine fundus.” Am J Roentgenol 56:208–210. Crossen, R. J. (1946). “Carcinoma of the fundus uteri.” South Med J 39:445–451. Fletcher, G. H. “Uterine cervix” in Textbook of Radiotherapy. G. H. Fletcher (ed.). Philadelphia: Lea and Febiger, pp. 434–474, 1966. Freed, J. H., and E. P. Pendergrass. (1954). “An evaluation of the efficiency of various intrauterine radium techniques in the treatment of cancer of the corpus uteri.” Am J Roentgenol 71:253–266.
852
Bruce R. Thomadsen
Friedman, M. (1940). “The treatment of carcinoma of the corpus uteri.” Radiology 35:28–35. Graham, J. (1971). “The value of preoperative or postoperative treatment by radium for carcinoma of the uterine body.” Surg Gynecol Obstet 132:855–860. Haas J. S., R. D. Dean, and C. M. Mansfield. (1985). “Dosimetric comparison of the Fletcher ffamily of gynecologic colpostats 1950–1980.” Int J Radiat Oncol Biol Phys 11:1317–1321. Herbolsheimer, M. “Intrauterine Packing by Remote HDR Afterloading in Endometrial Carcinoma” in Changes in Brachytherapy. K. Rotte, and J. Kiffer (eds.). Nurnburg: Wachholz-Verlag, 1989. Herbolsheimer, M., K. Baier, P. Gall, E. Löffler, K. Rotte (1988). “Ferngesteuerte intrauterine Nachlade-Packmethode beim Endometriumkarzinom.” Röntgenberichte 17:226, cited in Krieger, H., D. Baltas, P. Kneschaurek. (1999). “Dosisspezifikation in der HDR-Brachytherapie” in Deutsche Gesellschaft für Medizinische Physik-Bericht Nr. 14. (ISBN 3-925218-13-0). Herbolsheimer, M., K. Baier, U. Götz-Gersitz, M. L. Güllenstern, K. Rotte, and O. Sauer. (1991). “Endometrial carcinoma remote HDR afterloading using modified Heyman packing.” Activity: Selectron Brachytherapy Journal 5:17–20. Heyman, J. (1947). “The radiotherapeutic treatment of cancer corporis uteri.” Br J Radiol 20:85–91. Heyman, J., and S. Benner. (1946). “Further experience with radiotherapy in cancer of the corpus of the uterus.” Acta Radiol 27:328–333. Heyman, J., O. Reuterwall, and S. Benner. (1941). “The Radiumhemmet experience with radiotherapy in cancer of the corpus of the uterus.” Acta Radiol 22:11–94. Kaplan, I. I. (1942). “Radiation in cancer of the corpus uteri.” Radiology 39:135–142. Martin, C. L. (1940). “Radiation therapy in carcinoma of the fundus of the uterus.” South Med J 33:135–143. McGinn, C. J., J. A. Stitt, D. A. Buchler, and T. J. Kinsella. (1991). “Intraoperative ultrasound guidance during high dose-rate intracavitary brachytherapy of the uterine cervix and corpus.” (Abstract). Endocuriether/Hypertherm Oncol 7:226. Nolan, J. F., and W. Natoli. (1948). “Dosage measurements for various methods of intrauterine radium applications in cancer of the endometrium.” Am J Roentgenol 59:786–795. Nori, D. A., B. Hilaris, L. Anderson, and L. L. Lewis. (1982). “A new endometrial applicator.” Int J Radiat Oncol Biol Phys 8:941–945. Noyes, W. R., K. T. Bastin, S. A., Edwards, D. A. Buchler, J. A. Stitt, B. R. Thomadsen, J. F. Fowler, and T. J. Kinsella. (1995). “Postoperative vaginal cuff irradiation using high dose rate remote afterloading: A phase II clinical protocol.” Int J Radiat Oncol Biol Phys 35:1439–1443. Paterson, R., and M. Cole. “The Body of the Uterus” in The Treatment of Malignant Disease by Radiotherapy. R. Paterson. London: Edward Arnold (Publishers) Ltd., 1963. Petereit, D. G., J. N. Sarkaria, J. Schink, S. Springman, and T. J. Kinsella. (1995). “High dose rate brachytherapy for medically inoperable stage I endometrial cancer.” (Abstract). Int J Radiat Oncol Biol Phys 35:227. Rauthe, G., H. Vahrson, and G. Giers. “Five-year Results and Complications in Endometrial Cancer: HDR Afterloading Versus Conventional Radium Therapy” in High Dose Rate Afterloading in the Treatment of Cancer of the Uterus, Breast, and Rectum. H. Vahrson and G. Rauthe (eds.). Munich: Urban and Schwarzenberg, pp. 240–245, 1988. Rotte, K. (1975). “Technik, strahlenbiologie und ergenbnisse der afterloading-behandlung gynakologischer karzinome.” Rongtgen-Berichte 4:251–266. Sarkaria, J. N., D. G. Petereit, T. J. Kinsella, and D. A. Buchler. (1995). “An analysis of acute complications and perioperative morbidity from high dose rate brachytherapy in the treatment of gynecological malignancies.” (Abstract). Int J Radiat Oncol Biol Phys 35:224. Schmitz, H., and H. Schmitz. (1935). “An improved technique for radium treatment of carcinoma of the uterine body.” Am J Roentgenol 34:759–765. Sharma, P. D., and J. C. F. MacDonald. (1980). “Dose distribution around radium arrays used in the treatment of uterine carcinoma.” Acta Radiol Oncol 19:229–233. Simon, N., S. M. Silverstone, and L. C. Roach. (1971). “Afterloading Heyman applicators.” Acta Radiol 10:232–238. Wang, C. C. (1975). “An afterloading applicator for intracavitary vaginal irradiation.” Radiology 117:225.
Chapter 48
Changing from Low Dose Rate to High Dose Rate Intracavitary Gynecological Brachytherapy Jason Rownd, M.S. Medical College of Wisconsin Milwaukee, Wisconsin Applicator Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 853 Radiobiology Primer: Linear Quadratic Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 854 Radiobiological Effect and Prescriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 856 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 857 Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 857
Applicator Considerations The tradition “Fletcher-like” low dose rate (LDR) tandem and ovoid applicator utilized 226Ra or 137Cs as the radiation source. Because of the large physical source sizes, the applicators themselves were bulky and uncomfortable for the patient to endure, as shown in Figure 1. Active loading of the applicator and relative size of these applicators made the insertions more difficult for both the physician and the patient. Additionally, with typically only five source positions within these applicators, optimization options were limited to the number of possible source positions and the source on hand in the clinic. After insertion, patient and applicator immobilization become issues. The patient is on enforced bed rest for the duration of the implant. However, the applicator position is only constrained by packing. Over the course of the implant, this packing may be dislodged, allowing some motion of the applicator within the patient. Additionally, the normal motion of the internal anatomy means the dose distribution throughout these tissues can vary from the planned values. A study by Grigsby et al. (1993) showed average movement of 2 cm. The move to high dose rate (HDR) tandem and ovoid applicators allowed remote afterloading of the radioactive source as well as better patient immobilization and plan optimization. Other HDR-based applicators were also developed (tandem and ring, tandem and cylinder, mold, etc.). The relative size of the 192Ir HDR source allowed the physical size of the applicators to be reduced, specifically the tandem diameter, shown in Figure 1. The reduction in the tandem diameter, from approximately 6 mm in LDR applicators to 3 mm in HDR applicators, meant less discomfort for the patient as a result of dilation of the cervix during the implant. The overall reduction in physical size of the applicator also made the insertions easier for the physician to place the applicators accurately. Whichever applicator is used, each HDR brachytherapy insertion, now handled as a shorter outpatient procedure, allows for better patient immobilization during the treatment. Because of the short duration of the implant, not only is more consistent patient immobilization easier to achieve, but it is more necessary. The combination of higher dose rates with correspondingly shorter treatment times and the greater radiobiological effectiveness of HDR brachytherapy treatments means that any deviation in placement will have a strong effect on the dosimetry. In addition, patient and fraction-specific optimization can be done with computer planning on each insertion, so physical changes in applicator placement, if documented, can be accounted for with a new treatment plan for each fraction. Regardless, optimization will not turn a bad implant into a good one.
853
854
Jason Rownd
Figure 1. LDR and HDR tandem and ovoid applicators.
Radiobiology Primer: Linear Quadratic Model A more detailed discussion on radiobiology is presented elsewhere in these proceedings. Therefore, only review or primer will be discussed here. There are four basic components that are often referenced when speaking of the basics of radiobiology, the four R’s: repair, repopulation, reassortment, and reoxygenation. Repair is a measure of how easily the irradiated cell can repair the damage received. The basis of the linear quadratic model is that low doses and low dose rates favor late-responding normal tissues relative to acute-responding tissues and tumors. Repopulation is a measure of how fast the tissues can produce new cells. Rapidly proliferating tumors and tissues tend to fare better than slower growing cells, especially when long treatment times are involved. Reassortment has to do with cell cycle variations in the irradiated tissues. Not every cell is at the same point in its cell cycle and, as a result, may respond differently to irradiation. Finally, reoxygenation refers to the fact that individual oxygen starved cells are more radioresistant. A useful model for understanding and interpreting the radiobiological differences of LDR and HDR brachytherapy is the linear quadratic (LQ). There are two fundamental assumptions that go into the LQ model that have to do with chromosome (DNA) breaks as the result of ionizing radiation. Single-strand breaks are when a single DNA strand is damaged by a single ionizing event. These breaks are assumed to be more easily repairable. Double-strand breaks to DNA are usually from two different ionizing events. These events are less frequent, less easily repaired, and usually the result of high doses or dose rates. Single-strand breaks have a probability of interaction proportional to dose, while double-strand breaks have a probability of interaction proportional to the square of the dose. The hypothesis is that an unrepaired double-strand break, either by two single-strand breaks in a sufficiently short time frame, or a double-strand break, will lead to cell death.
Low Dose Rate to High Dose Rate Intracavitary Gynecological Brachytherapy
855
The basic LQ model is described by equation (1): 2
S = e − (α D + β D ) ,
(1)
where S is the surviving cell fraction or overall probability of survival, α is the single event coefficient resulting in a double strand break, β is the double event coefficient resulting in a double strand break, and D is the delivered dose. Per the LQ model, single-strand breaks tend to occur as a result of lower doses than do the doublestrand breaks. This is illustrated in Figure 2 where the low dose part of the survival is governed by the α coefficient and the higher dose part of the curve is controlled by the β coefficient. Tumor cells tend to have an α/β ratio of 3 Gy compared to 10 Gy for late-responding normal tissues. Using this relationship, it is possible to choose a dose or dose rate that is more effective at killing tumor cells relative to late responding normal tissues, as seen in Figure 2. Note that there is a crossover point for the two curves in Figure 2 where αD = βD2. Doses below this crossover point are better at killing tumor cells and at allowing a higher survival of late-responding normal tissues. This same type of curve can be shown for dose rate dependence with LDR irradiation providing the same late tissue sparing as low doses.
Figure 2. Linear quadratic model of cell survival as a function of dose.
856
Jason Rownd
Radiobiological Effect and Prescriptions To understand the different treatment doses for both LDR and HDR brachytherapy, we must find a way of relating the effectiveness of the two dose rates with respect to cell survival. The biological effect, E, for HDR treatments, equation (2), is a good starting point: E = (nD ) * (α + β D ) ,
(2)
where (α + βD) is single acute dose and n is the number of well separated fractions. The biological effective dose (BED) for HDR treatments, also called the “extrapolated response dose (ERD)”, is equal to the total dose times the relative effectiveness, as in equation (3). ERD = BED =
E
α
= nD 1 +
D
,
α / β
(3)
Next, we can write the same type of equation for the LDR version of BED, equation (4): ERD = BED = nRt [1 +
2R
µ (α / β )
(1 −
1 − e− µt
µt
)] ,
(4)
where R is the dose rate, t is the implant time for each LDR fraction, and µ, is a rate of repopulation. Note that Rt is the LDR total dose for the implant assuming long half-life isotopes. By equating these variations of the BED, you can find the relationship between LDR dose rate, total dose, and HDR fractions and dose per fraction. A simple spreadsheet illustrating this relationship is shown in Table 1. With this type of equivalence, you can adjust HDR fraction numbers and sizes to best fit your desired LDR implant schedule. A simple rule of thumb that is reasonably accurate for the prescription points is that the total HDR dose is approximately 60% of the total LDR dose. This type of information will allow you to relate LDR and HDR doses for a particular cell type and a particular dose rate. This means you can relate the radiobiological equivalence of different doses at a single distance in a single cell type, and introduces the concept of isoeffect lines instead of isodose lines for the implant. Because the relationship between dose and BED is not linear, the relationship between isodose and isoeffect lines is also not linear. Specifically, the isoeffect gradient is higher for higher doses and dose rates. The pear-shaped isodose distribution that the classical LDR tandem and ovoid applicator delivers to the patient will not necessarily map completely to the same pear shaped isoeffect distribution when the switch to HDR applicators is completed. At the prescription dose level, the equivalence between dose and biological effect for the various dose rates should be maintained, but in other areas of the implant, they may not be. For example, the vaginal surface dose in an LDR tandem and ovoid implant might be 145% of the prescription dose at point A. However, to maintain the same biological effect in the HDR implant, a relative dose of 133% would be necessary for the vaginal surface dose. So, while the general shape of the isodose distributions between LDR and HDR implants will be similar, to maintain the radiobiological effect accurately, the specific details of the isodose shape will have to be modified. Additionally, when referencing the dose limitations to normal tissues, the radiobiological effectiveness should be recomputed for proper comparison with the observed dose rates and doses for LDR and HDR treatments.
Low Dose Rate to High Dose Rate Intracavitary Gynecological Brachytherapy
857
Table 1. Calculation of HDR Fraction Size Based upon Total LDR Dose, Dose Rate, and Number of HDR Fractions “Variable” (V) n = # of LDR fxs R = LDR Rate (cGy/h) T = Time in hours (LDR Dose/Rate) µ = 1.4 h for tumors α/β = 10 for tumors RT = Dose given by LDR (cGy) N = # of HDR fractions
Value of V
ERD LDR
1 60 83.3
54.25
Total Dose = 5000 cGy
1.4 10 5000 5 ERD HDR D = Dose/fraction
6.55 Gy
Total Dose
3277 cGy
Summary In summary, switching between LDR and HDR applicators involves more than just a convenient change of applicators. You should be aware of the dose changes required to maintain radiobiological equivalence of tumor control as well as normal tissue sparing when treating with HDR brachytherapy. In most cases, you will achieve more normal tissue sparing through detailed optimization of plans and geometric sparing of the normal tissues by proper brachytherapy insertions than by straight modification of the prescription doses.
Reference Grigsby, P.W., A. Georgiou, J. F. Williamson, and C. A. Perez. (1993). “Anatomic variation of gynecological brachytherapy prescription points.” Int J Radiat Oncol Biol Phys 27:725–729.
Chapter 49
Gynecological Interstitial Implants Anil Kumar Sharma, Ph.D. Radiation Oncology Department Long Beach Memorial Medical Center Long Beach, California Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 859 Applicators, Sources, and Accessories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860 LDR Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860 LDR Templates and Needles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 860 LDR Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 861 HDR Implants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 861 HDR Templates and Needles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 861 Afterloader . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 861 Source Transfer Tubes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862 Types of Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862 Cervical Interstitial-Intracavitary Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 862 Vaginal Interstitial-Intracavitary Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 863 Interstitial Implants of the Vulva . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865 Recurrent Gynecological Implants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 865 Reconstruction of Implant and Organs at Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 866 Film Based . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 866 CT Data Based . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 866 Treatment Planning/Dosimetry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867 Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 867 LDR Implant Dosimetry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868 Dose Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868 Differential Unloading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868 HDR Implant Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 868 Dose Fractionation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 869 Dose Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 869 Plan Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 869 Dose Delivery. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 869 Applicator Movement During the Course of Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 870 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 870
Introduction Conventional intracavitary applications deliver adequate dose to the target volume in patients with earlystage disease and with normal anatomy. However, in patients with advanced stage disease and/or distorted anatomy, intracavitary application may not be able to deliver tumoricidal dose to the target volume, i.e., parametria and lower vagina, without sparing the organs at risk, i.e., rectum and bladder from higher doses of radiation. Interstitial transperineal implants with rigid needles (which serve as interstitial ovoids) and a vaginal obturator, with or without an intrauterine tandem, can deliver adequate dose to the target volume with relative sparing of the critical organs. Figure 1 shows adequate coverage of a small lesion by intracavitary technique and the potential of an interstitial implant to cover the lateral extent of the advanced disease.
859
860
Anil Kumar Sharma
Figure 1. Conventional intracavitary application (left) with an intrauterine tandem and vaginal ovoids, though adequate for early-stage disease, may not be able to deliver tumericidal dose in advanced cases. Interstitial implant (right) in which the conventional intravaginal ovoids are replaced with interstitial needles implanted through paravaginal and parametrial tissues can deliver adequate dose to the target volume with relative sparing of the critical organs.
Applicators, Sources, and Accessories The most commonly used template systems for gynecological interstitial implantation are the Syed-Neblett template and the Martinez Universal Perineal Interstitial Template (MUPIT), both introduced in 1974. Both systems use 17-gauge needles, but differ in their pattern of allowable needle position. MUPIT allows needles in its type I holes at 1-cm intervals and perpendicular to the template, and needles at a 1.25-cm interval at a 13° laterally outward direction in type II holes (Martinez, Cox, and Edmundson 1984). A volume of 4 cm on either side of the template can be covered using type I holes and 7 cm with type II holes. The vaginal and rectal cylinders, available in two sizes, compose the intracavitary component of the system, and can accommodate an intrauterine tandem or drainage tube. Each of the cylinders can also accommodate up to eight needles. MUPIT can be used for both low dose rate (LDR) and high dose rate (HDR) applications. Commercialized templates are available through Nucletron Corporation (Columbia, MD) for HDR applications. The Syed-Neblett gynecological template (Syed and Neblett 1979), shown in Figure 2, has a pattern of holes (1 cm apart) in five concentric circles or arcs in 1-cm radial increments and a central grooved vaginal obturator of 2 cm diameter and 15 cm length, which can accommodate six needles and a centrally located intrauterine tandem of medium curvature. These templates are available as reuseable or more commonly as disposable from Alpha Omega Services (Bellflower, CA) and Best Industries (Springfield, VA). HDR needles are shaped at the proximal end for connection to an afterloading system.
LDR Implants Two LDR implants are performed about 2 to 3 weeks apart, delivering 20 to 30 Gy per implant after a course of external beam irradiation of about 50 Gy in 5 to 6 weeks. LDR implants require templates, steel or plastic (Flexiglass) needles with rigid metallic stylets and radiation sources for the needles and tandem. The activities used for these implants warrant proper shielding between the patients and the caregivers. LDR Templates and Needles Both templates (Syed-Neblett and MUPIT) were originally developed for LDR applications. Both systems use 17-gauge needles for LDR applications. All but the first LDR needle used in the Syed-Neblett system
49–Gynecological Interstitial Implants
861
Figure 2. The Syed-Neblett gynecological template with interstitial needles and intrauterine tandem (left). Photograph on the right is without the tandem but all grooves (6) in the vaginal obturator are loaded with needles.
have small metal rings near the proximal end to prevent them from sliding through the template. The distal ends of these needles are closed to prevent accidental source slippage into the patient. LDR Sources In the United States, 192Ir seed ribbons are used for interstitial LDR applications. 192Ir seeds are 3 mm long, 0.5 mm in diameter, and are arranged in nylon ribbons of 0.8 mm outer diameter with center-to-center separation of 1 cm. In the early days of interstitial brachytherapy, most LDR treatments used uniform loading along the ribbon length, since differential loading was more expensive and limited by source strength availability. But now differential loading is common, and the state-of-the-practice. 192Ir wires, formerly popular in the Europe, were 0.3 mm in diameter and encased in a plastic sheath of about 0.6 mm outer diameter. The wires are no longer available. For the tandem loading, however, 137Cs sources are used.
HDR Implants Over the years, HDR remote-afterloading brachytherapy has been gaining popularity because it can provide better radiation protection and dose optimization in the planned target volume. The dwell-time programming feature of HDR brachytherapy offers a much easier method of achieving a source distribution that would result in a desired dose distribution. HDR Templates and Needles Templates and needles used for HDR application are afterloader dependent (more so, on the outer diameter of the HDR source cable). The VariSource™ has a thinner source cable (0.59 cm outer diameter) compared to the Nucletron and GammaMed systems. Templates allowing the use of 17-gauge needles (as shown in Figure 3) are used for VariSource afterloader, and 15- or 16-gauge templates and needles are often used with the Nucletron and GammaMed systems. Afterloader The three most commonly used afterloaders are: GammaMed and VariSource from Varian Medical Systems (Charlottesville, VA) and MicroSelectron from Nucletron Corporation. All HDR afterloaders use 192 Ir high activity (10 Ci max) source of about 0.5 cm length. The same source is used for both tandem and needle loading during fractionated HDR treatments.
862
Anil Kumar Sharma
Figure 3. The Syed-Neblett gynecological template can be used for both LDR and HDR applications. The figure shows the template, vaginal obturator, tandem, and HDR needles, which are slightly modified from the original LDR needles by providing a threaded funnel at the proximal ends.
Source Transfer Tubes Source transfer tubes for Nucletron and GammaMed are reuseable guide tubes of a fixed length to connect the implant needle to the treatment unit. VariSource uses either disposable or reuseable source transfer tubes that need to be first connected to an intermediate catheter and then to the needle (Figure 4). The “catheter length parameter” (CLP) of VariSource is variable with a maximum extension of 150 cm from indexer to tip of the needle. GammaMed and Nucletron afterloaders can be connected directly with the needles through the source transfer tubes without any intermediate connector.
Types of Implants Using interstitial techniques, several types of gynecological implants are performed (Syed and Puthawala 1996). The pre-implant preparation is about the same as for the conventional intracavitary applications. Even though an epidural or spinal block can be used for interstitial implantation, most patients undergo general anesthesia. For all types of interstitial implants, the patient is placed in the lithotomy position and a Foley catheter is inserted into the bladder; its balloon is filled with 7 cc of 50% Hypaque solution.
Cervical Interstitial-Intracavitary Implants The interstitial-intracavitary implant technique utilizes an intracavitary tandem (whenever possible), as in conventional intracavitary application, but the conventional intravaginal ovoids are replaced with interstitial needles implanted through paravaginal and parametrial tissues. If a tandem cannot be used, six obturator needles are implanted into the uterus. Gold seeds are implanted into the cervix at the 2 o’clock and 8 o’clock positions. The endometrial canal is sounded and its depth is noted. The endocervical canal is then dilated and a tandem with medium curvature and a metal flange fixed at the appropriate level to
49–Gynecological Interstitial Implants
863
Figure 4. HDR needles are connected to the intermediate connectors, which in turn are connected to the VariSource™ afterloader indexer using source transfer tubes for fractionated HDR treatments.
be against the cervix is inserted into the endometrial canal. The first guide needle is inserted into the posterior or the anterior lip of the cervix to a depth of 3 to 4 cm. The vaginal obturator is then threaded over the tandem to abut against the flange. The first guide needle is then inserted into the groove on the surface of the vaginal obturator. In the Syed-Neblett system (Syed et al., 1986), a rubber O-ring of 2 cm diameter is then threaded over the vaginal obturator and placed into the groove on the template to secure the obturator and the first needle into position. The template is then held against the perineum, and the remaining needles are implanted transperineally through the holes in the template into the parametria and paravaginal tissues to the same depth as the first guide needle. The needles are dipped in alcohol for easy insertion through the template, which acts as a lubricant, and later as the alcohol evaporates quickly, the needles are held in place due to friction. Once the desired number of needles is implanted and the tandem is secured in place by tightening the screw at the obturator end, the template is secured in position by 2.0 silk sutures through the perineal skin and anterior two corners of the template. Figure 5 shows a typical intracavitary-interstitial HDR implant with 20 needles and intrauterine tandem; no needles are implanted in the first (obturator) circle, and in the second circle only six needles are used, leaving three anterior and three posterior positions empty to reduce dose to bladder and rectum. The space between the perineum and the template is filled with vaginal gauze soaked in antibiotic cream.
Vaginal Interstitial-Intracavitary Implants Primary carcinoma of the vagina comprises less than 2% of the gynecological malignancies, but its local control in relatively advanced disease remains unsatisfactory with intracavitary irradiation. Fleming et al. (1980) reported that with interstitial-intracavitary application, the disease free survival can be significantly increased. This is achieved because of the relatively homogeneous radiation dose to the entire paracolpium and parametrium at risk, with relative sparing of bladder and rectum in spite of the anatomic distortion associated with the extensive disease. The extent of the disease is determined generally by computed tomography (CT) scanning the pelvis and abdomen of the patient (see Figure 6). In some cases, interstitial implant is carried out in conjunction with a laparotomy, which permits ascertaining precise size and extent of the disease. Multiple needles are inserted into the tumor-bearing regions of the vagina and lateral tissue. The implants are individualized with respect to the number of needles and the depth of their insertion. If possible, a medium curvature tandem is also used and its loading depends on the extent of the
864
Anil Kumar Sharma
Figure 5. Interstitial-intracavitary HDR implant with a tandem and 20 needles for a stage IIIB cervical cancer.
Figure 6. Vaginal implant with 16 needles; 6 obturator needles and 10 needles in the second circle are implanted.
49–Gynecological Interstitial Implants
865
disease. Even though needles in up to three circles or arcs are implanted using the Syed-Neblett applicator, only a limited number of them are actually loaded for treatment. Some of these needles are used for stabilization of the template. When a tandem is used, the obturator needles are loaded in only that sector of the template where gross disease is present.
Interstitial Implants of the Vulva Incidence of vulvar cancer accounts for about 8% of all gynecological malignancies, with up to 50% of these patients having advanced disease at presentation (Tewari et al., 1999), with involvement by direct extension to adjacent structures, including, the vagina, urethra, and anus, and by lymphatic embolization to regional groin lymph nodes. Advanced vulvar disease can be controlled by implanting the tumor with flexible afterloading catheters along with an interstitial implant using template and needles. Flexible catheter implant is usually a two-plane implant, whereas the size of the template-based interstitial implant depends on the extent of the disease. Satisfactory implant technique requires both tubes and buttons and template and needles to encompass the disease along with a course of external beam irradiation (see Figure 7).
Recurrent Gynecological Implants Salvage interstitial implantation of temporary 192Ir and permanent 125I can be performed for recurrent cervical, vaginal, endometrial, and ovarian malignancies (Syed et al., 1975). In most cases, laparotomy is performed which facilitates: (i) direct observation of tumor recurrence; (ii) separation of bowel adhesions;
Figure 7. Interstitial implant of the vulva is a combination of template-based needle implant and a flexible-catheters-and-buttons implant. Usually flexible-catheter implant is a two-plane implant (8 catheters in this example) and only the obturator and second circle needles are implanted (12 in this example) in the template-based needle implant.
866
Anil Kumar Sharma
(iii) tumor reduction surgery; (iv) placement of interstitial needles under direct observation; (v) ability to implant 125I seeds into the target not encompassed by needles; and (vi) ability to separate small bowel from the needles with an omental pedicle graft. Either two interstitial implants are performed with a 2-week interval or one interstitial implant followed by external beam radiation therapy is the course of treatment for these recurrent tumors. If a radiofrequency hyperthermia system is available, these patients with recurrent disease may benefit from the radiopotentiating effect of interstitial hyperthermia delivered for an hour just before or after the HDR fraction.
Reconstruction of Implant and Organs at Risk Gynecological implants are characterized with numerous channels that can extend laterally about 7 cm either side of the central obturator. Two main critical structures in these cases are the bladder and the rectum. Reconstruction of the implant and the organs at risk can either be carried out using radiographs or by CT.
Film Based Neblett recommended diagonal loading with coded dummy ribbons (Figure 8) and several sets of orthogonal radiographs for the reconstruction of large gynecological implants (Neblett 1994). With just one set of antero-posterior (AP) and lateral (LAT) radiographs, severe overlapping of the channels takes place in the lateral radiograph. In our practice, however, we have found that if two oblique (45° from the anterior) radiographs are taken with all the channels loaded with dummy ribbons, it possible to work with just one set of radiographs. Identification of the bladder points can be done according to International Commission on Radiation Units and Measurements (ICRU) report 38 (ICRU 1985), but the rectal points cannot be defined in the same way as there is no packing in the vagina. However, by using a rectal marker (bariumsoaked gauze, as shown in Figure 8), rectal points can be assigned on the anterior surface of the rectal marker to represent the anterior rectal wall for determining the maximum rectal dose.
CT Data Based Applicator-based or image-based brachytherapy can be carried out using a CT data set (Figure 9). For reconstruction of the steel needle implant, dummy markers are not required; however, if plastic needles are implanted, it is better to put dummy markers to visualize the catheters in the scout films. CT images
Figure 8. Dummy markers (left) are required to be inserted in each needle for reconstruction of the implant in film-based reconstruction. AP and lateral radiographs (right) with dummy markers in the needles, dye in the Foley balloon, and barium-soaked gauze in the rectum.
49–Gynecological Interstitial Implants
867
Figure 9. Reconstruction of the implant and organs at risk using CT data. The implant and target volume are reconstructed independent of each other in image-based brachytherapy.
have limitations for clear demarcation of the tumor from other nearby tissue. For better visualization of the tumor, and parametrial and vaginal infiltration of the disease, magnetic resonance imaging (MRI) offers significant improvements over CT (Hricak and Yu 1996). MRI-based planning requires compatible needles and tandem, which may produce distortion during the implant making it difficult to clearly demarcate the structures. Bladder and rectum can be delineated on transverse CT slices for their three-dimensional (3D) reconstruction.
Treatment Planning/Dosimetry Treatment planning for the interstitial template technique should not be considered just an extension of the intracavitary technique because there is no vaginal packing in interstitial implants and the “ovoids” are interstitial needles with extensive lateral spread. There may or may not be a central tandem and depending on the extent of the disease, needles may be quite close to the rectum and bladder. Point A may not play any significant role, and the dose to point B may not be very different from that of the point A for large implants.
Planning Planning should be carried out on the target volume ascertained from clinical examination, CT, and MR information. Cervical tumors display increased signal intensity on the T2-weighted images relative to normal cervical stroma, and the paracervical soft tissues demonstrate high intensity on the T1- and T2weighted images. MRI helps to define tumor size and spread of the disease to parametrial tissue. However, bony anatomy is not well differentiated on MRI; but by fusing it on CT, preplanning can be carried out using an idealized geometry of the allowable needles through the template. This helps in identifying the
868
Anil Kumar Sharma
number, location, and depth of the channels to be implanted. Based on this plan, seed activities, number of source ribbons, and their lengths can also be estimated for LDR implants.
LDR Implant Dosimetry Several guidelines are available for LDR dosimetry of the Syed-Neblett technique: • If the intrauterine tandem is used, none of the obturator needles should be loaded • If there is no central bulky disease, anterior three and posterior three needles in the second circle should be avoided to reduce dose to bladder and rectum • If intrauterine tandem cannot be used, then six obturator needles should be loaded with ribbons containing four or five seeds per ribbon • The activity per seeds should be in the range of 0.25 to 0.35 mgRaeq • Cervical tandem loading of 10, 5, 5 mgRaeq should be used and, depending on the length of the intrauterine tandem, the number of 137Cs tubes in the tandem should be decided • From the reconstructed geometry of the implant and the extent of the disease, position of the first source in the tandem and first seed in the guides should be decided. Dose Rate Average dose rate to point A should optimally be in the range of 50 to 65 cGy per hour. The seed activities should be chosen in such a way that the point A dose rate does not exceed 80 cGy per hour. Differential Unloading Differential unloading of the tandem and the central needles should be carried out whenever necessary to achieve higher doses to the point B and for sparing the rectum and bladder, and to keep the volume raised to 150% of the prescription dose from surrounding more than one needle.
HDR Implant Dosimetry The following guidelines help in planning HDR interstitial implants: • Tandem loading with the HDR source should be continuous and optimized along with other needles to cover the target volume adequately for image-based brachytherapy • For HDR planning four lateral obturator needles can also be lightly loaded (about 4 cm) for better uniformity of the dose • Central dose should be higher, because most tumors are at or around the center. For LDR implants, 125% of the prescribed dose is acceptable in the center for nontandem cases. The same should be acceptable for HDR planning. Again, biologically, with HDR, 125% is significantly higher than 125% at LDR. With the tandem in place, a much higher central dose will result. Though high-dose gradients exist around the source dwell positions, the total volume of the hyperdose sleeve (200% TD) should be restricted (Sharma et al., 1998). In most interstitial applications without the tandem, hyperdose sleeve should be restricted to less than 10% of the volume covered by the prescribed dose. But such restrictions may not apply in those cases where a central tandem is used because of the tandem loading, which is much more than the individual needles.
49–Gynecological Interstitial Implants
869
Dose Fractionation HDR brachytherapy is used as boost after a course of external beam therapy. At our institute, two HDR implants, separated by an interval of 2 to 3 weeks, are carried out after an external beam dose of 50.4 Gy in 28 fractions, with midline block placed at 39.6 Gy. For each implant 18 Gy in three equal fractions are delivered to the target volume. Dose Optimization The straight channels implanted almost equidistantly and loaded with almost the same spacing form a geometrical matrix which is most suitable for geometrical optimization (Edmundson 1994). Geometrical optimization generally results in doses in the center of the implant that are higher than the peripheral (or reference) dose. This is preferable for gynecological malignancies with bulky central disease. Dwell time optimization, which can be achieved either by geometrical optimization or optimization by constraints followed by manual adjustments, should not only provide adequate coverage to the target volume, but also take into consideration the dose delivered to the organs at risk, i.e., rectum and bladder.
Plan Evaluation Further analysis of the dose distribution can be carried out using contiguous (Neblett 1994) or differential dose volume histograms (DVHs) of the optimized plans. A DVH is an excellent representation in 2-D of the complex 3-D dose distribution of brachytherapy plans. If the implant is optimized to the same dose midway between the catheters, then the differential DVH will show a sharp peak for that dose. Based on this property, it is used for assessing the homogeneity of dose distribution. The height and narrowness of the peak are useful in assessing dose homogeneity. For this type of differential histogram, where differential volume is plotted against the dose, the peak is somewhat distorted by an inverse square law effect. At very large distances (low doses), the entire implant is seen as a single source and the volume per unit dose increases (as –5/2 power of dose) as the dose decreases, masking to some extent the result of relatively uniform dose regions between the catheters. A similar effect occurs, at very small distances from the sources, in the high-dose regions to the right of the peak. These obscurative effects are eliminated in the “natural” DVH (Anderson 1986), a differential histogram for which the derivative of the baseline variable with respect to dose is proportional to the –5/2 power of the dose (Figure 10). Using the natural DVH, the dose distribution between the catheters and in the treated volume is evaluated and the effect of dose optimization is assessed. Other uniformity and quality indices can be derived from this histogram for further analysis of the dose distribution. The mg-hrs required from Paterson-Parker tables (Paterson and Parker 1938) for the reference isodose surface volume can be compared with the actual mg-hrs obtained after appropriate conversions from the planned total air kerma strength for second calculation.
Dose Delivery Patients are admitted in individual rooms for LDR applications and proper shields are provided near the pelvic area for the caregivers and visitors. Source loading and unloading is carried out by the radiation oncologist in the presence of a physics staff member. A radiation survey is performed at the start of the treatment and at the end of the treatment. For HDR applications, a pretreatment check is made by comparing the dwell positions and dwell times of the planning computer printout with those of the afterloader computer. A quick check of radiation levels at the afterloader surface and at 1 m is made before the start of the treatment. Also, in the same manner, the patient is surveyed using a calibrated ionization survey meter before the first fraction. The patient is monitored during the treatment, which is delivered in the
870
Anil Kumar Sharma
Figure 10. “Natural” DVH of a template-based interstitial vaginal implant. The base of the peak is at 100%TD; the sharp peak indicates good dose uniformity and less spread of the dose in the target volume.
presence of the radiation oncologist and the medical physicist. At the end of the treatment, the medical physicist enters the room first with the radiation survey meter and again the patient is surveyed and radiation levels at the surface and at 1 m from the afterloader are noted down. The treatment fraction completion printout is compared with the planned treatment printout to ensure that the treatment delivered is actually the treatment intended.
Applicator Movement During the Course of Treatment Continuous LDR and fractionated HDR brachytherapy are delivered over a period of 25 to 40 hours, allowing for potential movement of afterloading needles between fractions as reported in the literature (Damore et al. 2000). Most movements take place during the overnight stay in the hospital, and are inherent during the LDR delivery or for the second fraction of the HDR brachytherapy. By keeping adequate margins for treatment, i.e., implanting needles deeper than required, but treating the target by leaving required space at the distal end, one can have enough room for applying corrections in the CLP. Inferior displacement of interstitial needles is a potential source of error in the delivery of LDR brachytherapy for gynecological cancers. Localization films or CT scouts are necessary before fractionated treatments and corrections based on these films can be made so that every time the target volume is given the same dose pattern as planned. Similarly for LDR treatments, mobile C-arms may be used for confirming the implant position.
References Anderson, L. L. (1986). “A ‘natural’ volume-dose histogram for brachytherapy.” Med Phys 13:898–903. Damore, S. J., A. M. N. Syed, A. Puthawala, and A. Sharma. (2000). “Needle displacement during HDR brachytherapy in the treatment of prostate cancer.” Int J Radiat Oncol Biol Phys 46(5):1205–1211.
49–Gynecological Interstitial Implants
871
Edmundson, G. K. “Volume Optimization: An American Viewpoint” in Brachytherapy: From Radium to Optimization. R. J. Mould and J. Batterman (eds.). Columbia, MD: Nucletron, pp. 314–318, 1994. Fleming, P., A. M. Nisar Syed, D. Neblett, A. Puthawala, F. W. George 3rd, and D. Townsend. (1980). “Description of an afterloading 192Ir interstitial-intracavitary technique in the treatment of carcinoma of vagina.” Obstet Gynecol 55:525–535. Hricak, H., and K. Yu. (1996). “Radiology in invasive cervical cancer.” Am J Roentgenol 167:1101–1108. International Commission on Radiation Units and Measurements (ICRU) Report No. 38. “Dose Volume Specification for Reporting Intracavitary Therapy in Gynecology.” Bethesda, MD: ICRU, 1985. Martinez, A, R. Cox, and G. Edmundson. (1984). “A multiple-site perineal applicator (MUPIT) for treatment of prostate, anorectal, and gynecological malignancies.” Int J Radiat Oncol Biol Phys 10:297–305. Neblett, D. L. “Clinical Techniques and Applications Available for Interstitial Implantation” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds). Madison, WI: Medical Physics Publishing, pp. 281–300, 1995. Neblett, D. L., A. M. N. Syed, A. A. Puthawala, R. Harrop, H. S. Frey, and S. E. Hogan. (1985). “An interstitial implant technique evaluated by contiguous volume reference.” Endocuriether/Hypertherm 1:213–221. Paterson, R., and H. M. Parker. (1938). “A dosage system for interstitial radium therapy.” Br J Radiol 11:252–266. Sharma, A., A. M. N. Syed, A. Puthawala, and A. A. Farid. (1998). “HDR planning for prostate cancer implants using Varisource remote afterloader.” J Brachytherapy International 14:1–14. Syed, A.M.N., and D. L. Neblett. “Interstitial-Intracavitary Applicator in the Management of Gynecological Malignancies” in proceedings of the Pacific Endocurietherapy Society winter meeting, Dec 5–7, 1979. Syed, A. M. N., and A. Puthawala. “Interstitial Implant Techniques” in Brachytherapy into the 21st Century. Long Beach, CA: Endocurietherapy Research Foundation, 1996. Syed, A. M. N., A. Puthawala, D. Neblett, (1986). “Transperineal interstitial/intracavitary ‘Syed-Neblett’ applicator in the treatment of carcinoma of the uterine cervix.” Endocurither/Hypertherm 2:1–13. Syed, A. M. N., N. Feder, F. George, et al. (1975). “Management of Extensive Residual Cancer with Interstitial Iridium Implant: A Preliminary Report” in Afterloading: 20 Years of Experience 1955-1975. New York, NY, pp. 119–124, 1975. Tewari, K., F. Cappuccini, A. M. N. Syed, A. Puthawala, P. J. DiSaia, M. L. Berman, A. Manetta, and B. J. Monk. (1999). “Interstitial brachytherapy in the treatment of advanced and recurrent vulvar cancer.” Am J Obstet Gynecol 181(1):91-98.
Chapter 50
Intravascular Brachytherapy Shirish K. Jani, Ph.D.1, Gerard B. Huppe, M.S.2, Vincent Massullo, M.D.3, and Paul Teirstein, M.D.4 1 Department of Radiation Oncology Sharp Memorial Medical Campus San Diego, California 2 Department of Radiation Oncology Grossmont Hospital San Diego, California 3 Department of Radiation Oncology John Muir-Mount Diablo Health System Concord, California 4 Department of Interventional Cardiology Scripps Clinic La Jolla, California Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 873 Gamma-Emitting Isotopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 874 Dose to Surrounding Normal Tissues from Gamma Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . 877 Dose to Other Body Organs from Gamma Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 877 Beta-Emitting Isotopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 878 Sources for Intravascular Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 879 Iridium-192 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 879 Strontium-90 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880 Phosphorus-32 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880 Yttrium-90 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 880 Phosphorus-32 Coated Stent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881 Vanadium-48 Stent . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881 Yttrium-90–Filled Balloon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 881 Delivery Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884 Current Status of Intravascular Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 886
Background Böttcher et al. (1994) reported one of the earliest uses of endovascular radiation therapy for prevention of restenosis from Frankfurt. This study, conducted in the early 1990s, is often referred to as the “Frankfurt Trial.” Following successful percutaneous transluminal angioplasty (PTA) of the superficial femoral artery, the patients were treated with a high dose rate (HDR) remote afterloader employing a 10 Ci 192Ir source; a dose of 12 Gy was delivered to tissue up to 3 mm from the source. A total of 30 patients with 31 restenotic lesions were treated using this method, with a median follow-up of 33 months. In the 28 patients available for follow-up, the 5-year patency rate (determined clinically and by ultrasound) was 82%. Follow-up CT/magnetic resonance (MR) imaging did not reveal any adverse reactions in the perivascular tissues. One of the earliest randomized trials in brachytherapy of human coronary artery was conducted at the Scripps Clinic in 1994 where 56 patients with in-stent restenosis were randomized between stenting and stenting plus 192Ir brachytherapy (Teirstein et al. 1997). A dose of about 14 Gy was delivered to the exter-
873
874
Shirish K. Jani et al.
nal elastic lamina. Although the trial had a small number of patients, it showed statistically significant reduction in the rate of restenosis. The Washington Radiation for In-Stent Restenosis Trial (WRIST) at Washington Hospital under the direction of Ron Waksman showed similar results of 192Ir brachytherapy among 130 patients (Waksman et al. 2000). Subsequently, several multi-institutional randomized trials in vascular brachytherapy showed remarkable reduction in restenosis among patients who underwent localized radiotherapy. The Gamma-1 trial sponsored by Cordis Corporation had randomized 251 patients for Ir-192 brachytherapy (Leon et al. 2001). The Proliferation Reduction with Vascular Energy Trial (PREVENT) and Intimal Hyperplasia Inhibition with Beta In-stent Trial (INHIBIT) sponsored by Guidant Corporation looked at the efficacy of 32P beta radiation for prevention of coronary restenosis (Raizner et al. 2000, Waksman et al. 2002). The Stents and Radiation Therapy Trial (START) sponsored by Novoste Corporation had randomized 400 patients and utilized an 90Sr source train to deliver beta radiation (Popma et al. 2002). The Vienna trials 1 through 5 were conducted to study the role of brachytherapy in preventing restenosis in femoral artery (Minar et al. 1998). In 1999, Waksman (1999) had reported an excellent overview of the status of clinical trials in vascular brachytherapy. Intravascular brachytherapy became one of the most successful interdisciplinary fields in radiation medicine. The physicists collaborated with radiation oncologists and interventional cardiologists/radiologists to explore the role of brachytherapy in preventing vascular restenosis. The physicists were challenged the most in terms of determining the dose at close distances; something our community has traditionally ignored. It was quite an exciting time in terms of exploring new territory not visited before. In this chapter, we have reviewed the devices that were employed in conducting clinical trials in intravascular brachytherapy (IVBT). We describe the physics and dosimetry of the radioisotopes. Finally, we discuss the future of IVBT in the era of drug-coated stents.
Gamma-Emitting Isotopes Iodine (atomic number Z=53; atomic mass A=125) nuclei decay via electron capture to the excited state of 125Te (tellurium, Z=52). The 125Te reaches its ground state by either internal conversion (93% of the time) or gamma emission. As shown in Figure 1, both of these processes lead to the production of characteristic x-rays, mainly in the range of 27.4 to 35 keV (Jani 1993).
Figure 1. Decay scheme for 125I radioisotope.
Intravascular Brachytherapy
875
Iridium (atomic number Z=77; atomic mass A=192) nuclei decay via negative beta emission (95.6% of the time) to 192Pt (platinum, Z=78) and via electron capture (4.4%) to 192Os (osmium, Z=76). The excited nuclei of 192Pt and 192Os emit a number of gamma rays in the energy range of 0.136 to 1.06 MeV; the primary emission being in the 0.3 to 0.6 MeV range (Figure 2). Palladium (atomic number Z=46, atomic mass A=103) nuclei decay via electron capture, largely to the first (90%) and second (10%) excited states of 103Rh (rhodium, Z=45). The characteristic x-rays from 103 Rh are mainly in the range of 20 to 23 keV (Figure 3).
Figure 2. Decay scheme for 192Ir radioisotope.
Figure 3. Decay scheme for 103Pd radioisotope.
876
Shirish K. Jani et al. Table 1. Physical Properties of Gamma Isotopes Isotope
Property
125
192
I
Average gamma energy (keV) Half-life (days) Half-value layer in mm of lead Commercially available source size (mm)
103
Ir
145
Pd
Sm
28 60 0.025
370 74.2 3
21 17 0.01
41 340 0.06
0.8 × 4.5
0.5 × 3
0.8 × 4.5
0.8 × 4.5*
mCi/hr Dose rate constant
1.45
4.69
1.48
0.88
(cGy/hr mCi)
1.16
4.55
0.95
1.1
Exposure rate constant R cm2
*Not available commercially. Table 2. Process of Energy Absorption for Gamma Isotopes 125
I
a
% energy transferred in water (Relative importance of interactions) f-factor in (cGy/R)b (See text for details)
a b
192
Ir
103
Pd
145
Sm
PE CS
93% 7%
0% 100%
99% 1%
81% 19%
Water Muscle Bone
0.88 0.92 4.44
0.97 0.96 0.93
0.89 0.92 4.3
0.89 0.93 4.2
Page 163, Johns and Cunningham (1983). Page 739, Johns and Cunningham (1983).
Samarium (atomic number Z=62, atomic mass A=145) nuclei decay via electron capture to 145Pm (promethium, Z=61), which, in turn, decays to stable 145Nd (neodymium, Z=60) by the same mechanism. As shown in Figure 4, the x-rays in the energy range of 38 to 61 keV are emitted from a sealed 145Sm source. Table 1 gives important physical properties of the gamma isotopes discussed in this chapter. Dose rate constant is defined as the dose rate in cGy/hr at 1 cm from a 1-mCi source in water. Table 2 describes the relative importance of photoelectric effect (PE) and compton scatter (CS) interactions for the gamma isotopes. The f-factor given here is defined as the amount of dose delivered (cGy) per unit exposure (roentgen, or simply R). As shown in the table, for example, the absorption of energy in bone for 125I gammas is 4.44/0.92 = 4.82 times more than in muscle. This is because PE interaction is very dominant here due to lower gamma energy and high Z difference between bone and muscle. In vascular brachytherapy, a calcified plaque (with calcium having Z=20 compared to water with Z~7.5) is highly likely to absorb low-energy gamma from iodine or palladium compared to surrounding muscle cell layers in arterial walls. As a result, a calcified plaque could cast a dose shadow behind a plaque with calcium. If the intent of vascular radiotherapy is to deliver a certain dose to treat media, then portions of media shadowed by plaque could receive decreased doses. This is an unwanted feature of low-energy gammas in vascular brachytherapy.
Intravascular Brachytherapy
877
Figure 4. Decay scheme for 145Sm radioisotope.
Dose to Surrounding Normal Tissues from Gamma Sources Dose distribution around sealed sources depends, in general, on four factors: distance, source size and make, absorption in tissues, and scattering by tissues (Nath 1995). Absorption decreases the dose at a point, whereas scatter increases it. At distances very close to a source, the absorption and scattering in tissues have opposite and equal effects and therefore nullify each other’s effects. Therefore, at distances less than about 10 mm, the dose depends primarily on distance and secondarily on source size and housing. Further away from the source, the dose begins to be influenced by absorption and scatter; these two being not equal in magnitude. Since these two depend on gamma energy, the dose at several centimeters is energy-dependent. For low-energy gammas, the absorption by tissues is greater than scatter; hence, dose drops off faster than inverse square effect. For high-energy gammas this effect is not as pronounced. The dosimetric parameter that accounts for absorption and scatter within the medium is called “radial dose function g(r).” It is defined as the ratio of dose at distance r to dose at 1 cm in tissues with the effect of the inverse square law removed; both points being on the transverse axis of the source. The g(r) has conventionally been obtained experimentally from the depth dose data of small sealed sources. It is unity at 1 cm by definition. Table 3 lists the radial dose function for four gamma isotopes of interest for vascular application. Variation of dose with distance beyond 1 cm for small (point-like) sealed sources is given in Table 4. These depth dose data show that dose beyond 1 cm is still greatly influenced by distance. Also, the dose at these distances is very little. It is important to note that dose values given in Table 4 are normalized to 1 cm distance. For vascular brachytherapy, the dose may be prescribed at, let us say, 3 mm distance. The values relative to this dose prescription would be much smaller than those of Table 4. Furthermore, the data given here are for small sources. In clinical applications where linear sources of several centimeters in length are utilized, it would be necessary to employ the measured depth dose data.
Dose to Other Body Organs from Gamma Sources Dose to tissues within 5 cm from irradiated artery was described in the previous section. The gamma source that is widely used in many brachytherapy procedures is 192Ir. It is used in manually loaded implants as well as remotely loaded high dose rate (HDR) units, such as microSelectron™ machine by Nucletron Corporation (Columbia, MD) in the use of a peripheral vascular irradiation study. At present, this isotope happens to be the most frequently used gamma source in coronary irradiation studies as well. Since it is
878
Shirish K. Jani et al. Table 3. The Radial Dose Function of Small Gamma Sources Radial Distance r (cm) 1 2 3 4 5 6 7 8 9 10
Radial Dose Function, g(r)* 125
I
1.00 0.86 0.68 0.53 0.38 0.29 0.23 0.20 — —
192
Ir
1.00 1.03 1.00 0.97 0.97 0.97 0.95 0.94 0.91 0.87
103
Pd
1.00 0.54 0.29 0.16 0.09 — — — — —
145
Sm
1.00 1.02 0.97 0.92 0.87 0.78 0.69 0.61 0.52 0.44
*Page 161, Jani (1993). Table 4. Dose to Surrounding Normal Tissues from Gamma Sources Distance from Source (cm) 1 2 3 4 5
Relative Dose (%) 125
I
100 21.5 7.6 3.3 1.5
192
Ir
100 25.7 11.1 6.1 3.9
103
Pd
100 13.5 3.2 6.1 3.9
145
Sm
100 25.5 10.8 5.8 3.5
a medium- to high-energy gamma emitter, it may be of interest to examine dose to peripheral tissues and organs from an 192Ir brachytherapy. Measurements made in humanoid phantom show that tissue attenuation begins to affect the dose as point of interest moves further away from a source (Venselaar et al. 1996). This is partly because the influence of inverse square law is not as dominant as it is at close distances. Using calculated dose gradient from an 192Ir linear source (Kline et al. 1985), for up to 1 cm distance and measured values beyond this distance (Venselaar et al. 1996), we have estimated the dose to tissues, which may be tens of centimeters away from a vascular brachytherapy site. Table 5 shows that dose to far away tissues within a patient is small. In fact, the dose to body tissues may be significantly higher from a standard interventional procedure the patient must undergo during balloon angioplasty and/or stenting than these 192Ir values. The issue of the biological effect of these small doses is beyond the scope of this chapter.
Beta-Emitting Isotopes Beta emitting isotopes used in recent clinical trials are listed in Table 6. Phosphorous (32P) nucleus emits a negative beta particle and permanently changes its physical form to become a sulfur nucleus (32S). The 32 P is called a parent element and the 32S is called a daughter element. The 32P beta may have any energy from almost zero to a maximum of 1.7 MeV.
Intravascular Brachytherapy
879
Table 5. Estimated Dose to Peripheral Organs from an Iridium-192 Vascular Application.
A Dose of 800 cGy Is Assumed To Be Delivered at 0.3 cm from a 5 cm Linear Source Distance from Source (cm)
Estimated Dose (cGy)*
0.3 1.0 — 10 20 30 40 50
800 179 — 2.5 0.44 0.12 0.04 0.01
*Values derived from data of Venselaar et al. (1996) and Kline et al. (1985). Table 6. Beta-Emitting Isotopes Employed in Clinical Trials Isotope (Symbol)
Maximum energy (MeV) Average energy (MeV)
Phosphorus (32P)
Rhenium (188Re)
Strontium (90Sr)
Yttrium (90Y)
1.7 0.60
2.2 0.78
0.5 0.02
2.3 0.90
Rhenium (188Re) decays to Osmium (188Os) by emitting betas of up to 2.2 MeV in energy. By itself, the 188Re has a very short life span (a half-life of 17 hours). It may be used in combination (equilibrium) with its parent isotope, tungsten (188W), which has a half-life of 69 days. Strontium (90Sr) decays to a daughter element called yttrium (90Y) by emitting a beta particle of 0.5 MeV in energy. A unique situation occurs in this decay process. The 90Y daughter also happens to be radioactive and decays to zirconium (90Zr) by emitting betas of up to 2.3 MeV in energy. The parent 90Sr and its daughter 90Y both are beta emitters but they have different energy characteristics. They also decay at vastly different rates, a characteristic called the “half-life” of an isotope. In a sealed radioactive 90Sr source, the 90Y daughter is also present. Therefore, a sealed 90Sr source not only emits its own betas, but also betas of its daughter, which happen to be more energetic than its own betas. Therefore, a seated 90Sr source is often referred to as a 90Sr/ 90Y source.
Sources for Intravascular Brachytherapy Iridium-192 Ir is a sealed gamma source in the form of a small seed, 0.3 cm × 0.05 cm in size (Best Medical International, Springfield, VA). A linear source of any length may be simulated by arranging these cylindrical seeds in an array embedded in nylon ribbon (Figure 5). The first randomized trials on coronary irradiation were carried out using 192Ir. Cordis Corporation (Miami, FL) has utilized this source in their delivery device called Checkmate™, as described later in the chapter. Its radioactive half-life is 74 days and hence 192
880
Shirish K. Jani et al.
Figure 5. Schematic drawing of 192Ir seed array (Best Medical International, Springfield, VA).
may be used for vascular brachytherapy for a few weeks after its receipt. Further details on this isotope are described elsewhere (Jani 1997, 1998).
Strontium-90 Sr is a sealed beta source in the form of small cylindrical seeds, 0.25 cm long and <0.2 cm in diameter (Novoste Corp., Norcross, GA). Currently, it is available only in the form of a “train,” i.e., an array of backto-back arrangement of 12, 16, or 24 seeds, resulting in a total active length of 3, 4, and 6 cm (Hillstead et al. 1998, Jani 1999). The 90Sr radioisotope emits beta particles with energy of up to 2.25 MeV. It has a half-life of 28 years and hence a source train developed for vascular irradiation may be used repeatedly for many years. The Beta Energy Restenosis Trial (BERT) and START trial have utilized this source in their device called Beta Cath™, which is described later in this chapter.
90
Phosphorus-32 P wire is a sealed beta source with 0.27 cm × 0.03 cm diameter embedded at the end of a Ni-Ti wire (Guidant Corp. Vascular Intervention, Houston, TX). It emits beta particles with energy of up to 1.71 MeV, and exhibits a half-life of 14 days (Raizner and Calfee 1998). Useful lifespan of 32P source for vascular radiotherapy would be short compared to 90Sr source train. The INHIBIT and PREVENT trials have utilized this source within a device called Galileo™, which is described later in the chapter. 32
Yttrium-90 90
Y wire is a sealed beta source with 0.29 cm length and 0.034 cm overall diameter (Schneider-AG, Geneva, Switzerland) (Verin et al. 1998). It emits beta particles of up to 2.28 MeV energy with a half-life of 64 hours. The useful lifespan of a 90Y wire source for vascular brachytherapy would be no more than a few days.
Intravascular Brachytherapy
881
Figure 6. Schematic drawing of 32P linear source (Guidant Corporation, Santa Clara, CA).
Phosphorus-32 Coated Stent The 32P coated stent is a sealed beta source in the form of an expandable stent (Isostent, Inc., Belmont, CA). The Strecker biliary stent coated with 32P is 0.20 cm long and about 0.22 cm in diameter. The PalmazSchatz coronary stent coated with 32P may be 0.65 cm long and expandable to 0.35 cm diameter (Fischell et al. 1998).
Vanadium-48 Stent The 48V stent is a sealed gamma/beta source in the form of a nickel-titanium (nitinol) stent with dimensions of 0.15 cm × 0.03 cm (Progressive Angioplasty Systems, Menlo Park, CA). The nitinol stent is activated by a high-energy proton beam from a cyclotron to produce 48V from 48Ti. The 48V radioisotope has a half-life of 16 days and decays by emitting up to 0.7 MeV positrons and numerous x-ray energy emissions (Li et al. 1998). The stent may be implanted at the disease site permanently or may be removed after a certain time when the desired dose of radiation is delivered to the arterial wall.
Yttrium-90–Filled Balloon The 90Y-filled balloon is a dilatation catheter balloon (Mansfield/Boston Scientific, Watertown, MA) about 0.3 cm in diameter and 2 cm in length, filled with yttrium-chloride solution (Nordion International, Ottawa, Table 7. Depth Dose Comparison of Sealed Linear Sources Distance from Source Center (mm)
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Relative Dose Along Transverse Axis 92
Ir Seed Array (27 mm length) (Kline et al. 1985) 162 100 71 55 49 37 31 27 24
90
32
92
Sr Seed Array (30 mm length) (Soares et al. 1998)
P Wire (27 mm length) (Mourtada et al. 2000)
Ir Wire (20 mm length) (Popowski et al. 1995)
173 100 63 42 29 19 13 9 5
183 100 57 30 18 10 5 3 1
— 100 56 31 20 13 8 5 —
882
Shirish K. Jani et al.
Canada) (Amols et al. 1996a). As discussed above, the 90Y radionuclide is a pure beta emitter with a halflife of 64 hours and transition energy of 2.28 MeV. A balloon filled with radioactive 90Y liquid would automatically be centered within the lumen and hence could possibly provide improved homogeneity in dose around the source within the arterial wall. Tables 7 through 10 illustrate the depth dose comparisons among various isotopes. Published data on many radioisotope’s dosimetry are at different distances. And, in most cases, it is inappropriate to extrapolate depth doses from one distance to another due to many complex factors affecting dosimetry. Therefore, it is necessary to present four tables here instead of consolidating our results into a single data table. Table 7 shows depth dose comparison of three beta-emitting sources with 192Ir gamma linear source. These depth dose data are mostly measured and normalized at 1.5 mm from the source center. Lumen diameter of a human coronary artery varies depending upon location and disease. However, it may be estimated to be about 3 mm. Therefore, the dose normalizing point in Table 7 refers to the lumen surface, if the source is centered within it. At 2 mm deep within the arterial wall, i.e., 3.5 mm from the source, the depth dose of linear 192Ir source is 37% compared to 10% to 20% for beta sources. This seemingly indifferent depth dose may result into quite a different dose delivery under certain circumstances. Consider, for example, a dose of 10 Gy to be delivered at 2 mm depth in the arterial wall. For 192Ir linear source, a dose of 27 Gy would result at 1.5 mm from source. For a low-energy beta source like 32P, the lumen surface at 1.5 mm would receive 100 Gy. This example illustrates a major and most significant difference between Table 8. Depth Doses Around 32P Coated Palmaz-Schatz Stent (3.5 mm Diameter, 6.5 mm Length) Relative Dose Along Transverse Axis Distance from Stent Surface (mm)
32
P Coated P-S Stent (Duggan et al. 1998)
192 Ir Wire (7 mm length) (Kline et al. 1985)
0 0.5 1.0 1.5 2.0
100 32 18 10 6
100 [@ 1.75 mm from wire] 70 [@ 2.25 mm from wire] 55 [@ 2.75 mm from wire] 42 [@ 3.25 mm from wire] 35 [@ 3.75 mm from wire]
Table 9. Depth Doses Around 48V Stent (3.0 mm diameter, 15 mm length) Relative Dose Along Transverse Axis Distance from Stent Surface (mm)
192
Ir Wire
48
(Li et al. 1998)
V Stent (15 mm length) (Kline et al. 1985)
0 0.1 0.2 0.3 0.4
100 60 38 24 15
100 [@ 1.5 mm from wire] 91 [@ 1.6 mm from wire] 85 [@ 1.7 mm from wire] 79 [@ 1.8 mm from wire] 75 [@ 1.9 mm from wire]
0.5
10
70 [@ 2.0 mm from wire]
Intravascular Brachytherapy
883
Table 10. Depth Doses Around 90Y-Filled Balloon (3.0 mm diameter, 20 mm length) Relative Dose Along Transverse Axis Distance from Stent Surface (mm) 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
90
(Teirstein et al. 1997)
90 Y Wire (20 mm length) (Kline et al. 1985)
100 62 35 24 18 12 9 5
100 [@ 1.5 mm from wire] 56 [@ 2.0 mm from wire] 31 [@ 2.5 mm from wire] 20 [@ 3.0 mm from wire] 13 [@ 3.5 mm from wire] 8 [@ 4.0 mm from wire] 5 [@ 4.5 mm from wire] —
Y Stent
gamma and beta sources. Very steep dose gradients that exist around beta sources lead to large variations in dose within the arterial wall, even for a perfectly centered source. It can be concluded that source centering is crucial for beta sources. Depth doses around a 32P coated stent are shown in Table 8. They are compared to a 192Ir line source. Again, a radioactive stent exhibits very high dose gradient and therefore results in a large variation in dose within the arterial wall. Table 9 shows depth doses on a nitinol stent, which has a very shallow tissue penetration. The concept of such a radioactive stent activated in a cyclotron is an interesting physics exercise; however, its clinical utility would be limited. Finally, as shown in Table 10, we have compared depth doses of a 90Y-filled balloon with a 90Y wire. In the balloon geometry where radioactive 90Y liquid is distributed evenly throughout the expanded balloon, the depth doses are slightly (but not significantly) greater compared to wire geometry. An important advantage of beta sources over gamma isotopes is their high source strength. Laws of basic physics allow beta sources to accumulate enough radioactivity to offer dose rates of 2 to 5 Gy/min at 2 mm distance. In contrast, currently available source strength of 192Ir would exhibit dose rates 5 to 10 times smaller than 90Sr values. In other words, beta sources offer dose rates that suit ideally for coronary irradiation. Beta sources are easy to handle in terms or radiation protection. A very small amount of shielding will stop most beta radiation. In comparison, gamma sources such as 192Ir may produce significant levels of stray radiation, thereby requiring appropriate protection measures (Jani 2000). Radionuclides with a long half-life, such as 90Sr (28 years), are preferred for vascular brachytherapy. They offer constant dose rates over a long period of time and allow repeat use of the source without frequent calibrations. Isotopes such as 90Y, with a half-life of 64 hours, have a short useful lifespan and may not be used for multiple procedures unless done on the same day. These sources may have to be produced frequently and on site, each requiring source calibration and leak tests. If vascular brachytherapy is coupled with stenting to optimize the results of angioplasty, then it is important to consider the effect of a metallic stent on dose distribution. For beta radiation, the dose perturbation due to the presence of a stent may be significant and may not be ignored (Amols et al. 1996b). In addition, any effect of calcified plaque on the ability of beta radiation to penetrate may also need to be investigated.
884
Shirish K. Jani et al.
Figure 7. The Cordis Checkmate™ system.
Delivery Devices There were three devices approved by the Food and Drug Administration (FDA) for intracoronary brachytherapy in the United States. The Checkmate (Cordis Corporation, Miami, FL) is shown in Figure 7. It utilized an array of 192Ir seed sources manufactured by Best Medical International. The dose prescription was 14 Gy at 2 mm distance from the center of the source. The source ribbon was not centered within the delivery catheter’s lumen. Further details of this device can be found elsewhere (Jani et al. 1999). The Beta Cath device utilized an array of 90Sr seeds in a compact hand-held form as shown in Figure 8 (Novoste Corporation, Norcross, GA). Because of high dose rates around these beta-emitting sources, a dose of 20 to 25 Gy at 2 mm distance was typically delivered in a few minutes. The device utilized a specially designed catheter with three lumens, one of which delivered the source from the housing to its destination. This lumen was off-centered and so the dose around the Beta Cath catheter was radially asymmetrical. The Galileo device utilized a linear wire source of 32P imbedded at the end of a steel cable (Guidant Corporation, Houston, TX). As shown in Figure 9, the system was essentially a remote afterloader manufactured by Nucletron Corporation. A specially designed catheter was employed to center the source within the lumen. A dose of 20 Gy at 1 mm depth within the arterial wall was delivered in a few minutes. The microSelectron (Nucletron Corporation, Columbia, MD), which is well known to us in HDR remote afterloading treatments in cancer, was approved by the FDA for intravascular application in peripheral artery. The PARIS trial sponsored by Nucletron had utilized this device to deliver 14 Gy dose at 2 mm depth of the arterial wall when the patients were transferred to radiation oncology departments from interventional radiology after percutaneous transluminal coronary angioplasty (PTCA). Unfortunately, the Peripheral Artery Radiation Investigational Study (PARIS) trial was found to be negative.
Current Status of Intravascular Brachytherapy Some trials concluded in the early 2000s showed radiation coupled with anti-platelet therapy reduced the rate of restenosis without significant risk of late thrombosis (Waksman et al. 2001, 2003). The long-term
Intravascular Brachytherapy
885
Figure 8. The Novoste Beta Cath™ system.
Figure 9. The Guidant’s Galileo™ system.
follow-up of the patients undergoing brachytherapy for coronary restenosis were being reported as well (Waksman et al. 2004, Leon et al. 2004). But then came along an effective alternative therapy for controlling restenosis in patients with coronary disease, namely drug-coated stent (Moses et al. 2003, Stone et al. 2004). This newly designed stent delivers mechanical means as well as the drug therapy to the site of
886
Shirish K. Jani et al.
stenosis. It is easy to use and does not require the presence of radiotherapy staff in the cardiac catheterization laboratories. Its effectiveness in controlling the disease seems to be better than radiation. All these factors are leading to a quick disappearance of radiation as an agent in treating restenosis. By the end of 2003, the Cordis Corporation decided to discontinue its 192Ir-based Checkmate as a product. In September 2004, the Guidant Corporation also pulled their 32P-based Galileo from the market. The only remaining device for performing vascular brachytherapy is Beta Cath by Novoste Corporation. It faces challenges in the world of drug-coated stent.
Summary Vascular brachytherapy has offered many challenges to medical physicists. First, it required that we measure the dose at very close distances (1 to 2 mm and less) from a small source—something we had ignored all along in brachytherapy dosimetry. Second, it challenged us to develop the methods of calibration for beta-emitting stents. In addition, the seemingly small perturbations such as effect of a metallic stent (or contrast media in the lumen) on dose distribution of beta sources turned out to be not negligible. The IVBT also offered the physicists a new look at how cardiologists have utilized ultrasound technology in reconstructing the three-dimensional anatomy of an artery. In brief, it turned out to be an exciting and challenging field for physicists. It is hoped that the knowledge we have gained in near-field dosimetry of small radioactive sources will some day be utilized in other applications of brachytherapy in treating human diseases.
References Amols, H. I., L. E. Reinstein, and J. Weinberger. (1996a). “Dosimetry of a radioactive coronary balloon dilatation catheter for treatment of neointimal hyperplasia.” Med Phys 23:1783–1788. Amols, H. I., S. Mirzadeh, F. F. Knapp Jr., and J. Weinberger. (1996b). “Beta irradiation for restenosis: Considerations for stent implantation.” Circulation 94:I-210. Böttcher, H. D., B. Schopohl, D. Liermann, J. Kollath, and I. A. Asamietz. (1994). “Endovascular irradiation—a new method to avoid recurrent stenosis after stent implantation in peripheral arteries: Technique and preliminary results. Int J Radiat Oncol Biol Phys 29(1):183–186. Duggan, D. M., C. W. Coffey II, and S. Levi. (1998). “Dose distribution for a 32P-impregnated coronary stent: Comparison of theoretical calculations and measurements with radiochromic film.” Int J Radiat Oncol Biol Phys 40:713–720. Fischell, T. A., A. J. Carter, D. R. Fischell, M. T. Foster, and R. E. Fischell. “Radioisotope Stents” in Handbook of Vascular Brachytherapy. R. Waksman and P. W. Serruys. London: Martin Dunitz, Ltd., pp. 59–68, 1998. Hillstead, R. A., C. R. Johnson, and T. D. Weldon. “The Beta-Cath™ System” in Handbook of Vascular Brachytherapy. R. Waksman and P. W. Serruys. London: Martin Dunitz, Ltd.. pp. 41–52, 1998. Jani, S. K. Handbook of Dosimetry Data for Radiotherapy. Boca Raton, FL: CRC Press, pp. 137–167, 1993. Jani, S. K. (1999). “Physics of vascular brachytherapy.” J Invasive Cardiol 11:517–523. Jani, S. K. (2000). “Radiation safety of personnel during catheter-based Ir-192 coronary brachytherapy. J Invasive Cardiol 12:286–290. Jani, S. K., V. Massullo, and P. Teirstein. “The 192Ir Radioactive Seed Ribbon” in Handbook of Vascular Brachytherapy. R. Waksman and P. W. Serruys. London: Martin Dunitz, Ltd., pp. 27–32, 1998. Jani, S. K., V. Massullo, S. Steuterman, P. Tripuraneni, and P. S. Teirstein. (1997). “Physics and safety aspects of a coronary irradiation pilot study to inhibit restenosis using manually loaded 192Ir ribbons.” Semin Intervent Cardiol 2:119–123. Jani, S. K., G. B. Huppe, V. Massullo, P. Tripuraneni, and P. Teirstein. “Best: Manually Loaded Iridium 192 Ribbon” in Vascular Brachytherapy, 2nd Edition. R. Waksman (ed). Armonk, NY: Futura Publishing Company, Inc., pp. 485–488, 1999.
Intravascular Brachytherapy
887
Johns, H. E., and J. R. Cunningham. The Physics of Radiology. Springfield, IL: Charles C Thomas Publishers, pp. 71–99, 1983. Kline, R. W., M. T. Gillin, D. F. Grimm, and A. Niroomand-Rad. (1985). “Computer dosimetry of 192Ir wire.” Med Phys 12:634–638. Leon, M. B., P. S. Teirstein, J. W. Moses, P. Tripuraneni, A. J. Lansky, S. Jani, S. C. Wong, D. Fish, S. Ellis, D. R. Holmes, D. Kerieakes, and R. E. Kuntz. (2001). “Localized intracoronary gamma-radiation therapy to inhibit the recurrence of restenosis after stenting.” N Engl J Med 344:250–256. Leon, M. B., P. S. Teirstein, J. W. Moses, S. C. Wong, P. Tripuraneni, S. K. Jani, D. R. Holmes, A. Lanksy, and R. E. Kuntz. “Declining Long-Term Efficacy of Vascular Brachytherapy for In-Stent Restenosis: 5-Year FollowUp from the Gamma 1 Randomized Trial.” American Heart Association (AHA) 2004. Abstract. Li, A. N., N. L. Eigler, F. Litvack, and J. S. Whiting. (1998). “Characterization of a positron emitting V48 nitinol stent for intracoronary brachytherapy.” Med Phys 25:20–28. Minar, E., B. Pokrajac, R. Ahmadi, T. Maca, W. Seitz, A. Stumpflen, R. Potter, and H. Ehringer. (1998). “Brachytherapy for prophylaxis of restenosis after long-segment femoropopliteal angioplasty: Pilot study.” Radiology 208(1):173–179. Erratum in Radiology 208(3):834. Moses, J. W., M. B. Leon, J. J. Popma, P. J. Fitzgerald, D. R. Holmes, C. O’Shaughnessy, R. P. Caputo, D. J. Kereiakes, D. O. Williams, P. S. Teirstein, J. L. Jaeger, and R. E. Kuntz. (2003). “Sirolimus-eluting stents versus standard stents in patients with stenosis in a native coronary artery.” N Engl J Med 349:1315–1323. Mourtada, F. A., C. G. Soares, S. M. Seltzer, and S. H. Lott. (2000). “Dosimetry characterization of P-32 catheterbased vascular brachytherapy source wire.” Med Phys 27:1770–1776. Nath, R. “Physical Properties and Clinical Uses of Brachytherapy Radionuclide” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, R. Nath (eds.). Madison, WI: Medical Physics Publishing, pp. 7–38, 1995. Popma, J. J., M. Suntharalingam, A. J. Lansky, R. R. Heuser, B. Speiser, P. S. Teirstein, V. Massullo, T. Bass, R. Henderson, S. Silber, P. von Rottkay, R. Bonan, K. K. Ho, A. Osattin, and R. E. Kuntz. (2002). “Randomized trial of 90Sr/90Y ß-radiation versus placebo control for treatment of in-stent restenosis.” Circulation 106:1090–1096. Popowski, Y., V. Verin, I. Papirov, P. Nouet, M. Rouzaud, M. Schwager, P Urban, W. Rutishauser, and J. M. Kurtz. (1995). “Intra-arterial 90Y brachytherapy: Preliminary dosimetric study using a specially modified angioplasty balloon.” Int J Radiat Oncol Biol Phys 33:713–717. Raizner, R. A., and R. V. Calfee. “The Guidant Intravascular Brachytherapy System” in Handbook of Vascular Brachytherapy. R. Waksman and P. W. Serruys. London: Martin Dunitz, Ltd., pp. 53–58, 1998. Raizner, A. E., S. N. Oesterle, R. Waksman, P. W. Serruys, A. Colombo,Y. L. Lim, A. C. Yeung, W. J. van der Giessen, L. Vandertie, J. K. Chiu, L. R. White, P. J. Fitzgerald, G. L. Kaluza, and N. M. Ali. (2000). “Inhibition of restenosis with beta-emitting radiotherapy: Report of the Proliferation Reduction With Vascular Energy Trial (PREVENT).” Circulation 102:951–958. Soares, C. G., D. G. Halpern, and C.-K. Wang. (1998). “Calibration and characterization of beta-particle sources for intravascular brachytherapy.” Med Phys 25:339–346. Stone, G. W., S. G. Ellis, D. A. Cox, J. Hermiller, C. O’Shaughnessy, J. T. Mann, M. Turco, R. Caputo, P. Bergin, J. Greenberg, J. J. Popma, M. E. Russell, the TAXUS-IV Investigators. (2004). “A polymer-based, paclitaxeleluting stent in patients with coronary artery disease.” N Engl J Med 350:221–231. Teirstein, P. S., V. Massullo, S. Jani, J. J. Popma, G. S. Mintz, R. J. Russo, R. A. Schatz, E. M. Guarneri, S. Steuterman, N. B. Morris, M. B. Leon, and P. Tripuraneni. (1997). “Catheter-based radiotherapy to inhibit restenosis after coronary stenting.” N Engl J Med 336:1697–1703. Venselaar, J. L. M., P. H. van der Giessen, and W. J. F. Dries. (1996). “Measurement and calculation of the dose at large distances from brachytherapy sources: Cs-137, Ir-192 and Co-60.” Med Phys 23:537–543. Verin, V. A., and Y. G. Popowski. “Schneider—Sauerwein Intravascular Radiation System” in Handbook of Vascular Brachytherapy. R. Waksman and P. W. Serruys. London: Martin Dunitz, Ltd., pp. 95–102, 1998. Waksman, R. (1999). “Intracoronary radiation therapy for restenosis prevention: status of the clinical trials.” Cardiovasc Radiat Med 1:2059. Waksman, R., E. Cheneau, and A. E. Ajani. (2003). “Intracoronary radiation therapy improves the clinical and angiographic outcomes of diffuse in-stent restenosis lesions: Results of the Washington Radiation for In-Stent Restenosis Trial for Long Lesions (Long WRIST) Studies.” Circulation 107(13):1744–1749.
888
Shirish K. Jani et al.
Waksman, R., R. L. White, R. C. Chan, B. G. Bass, L. Geirlach, G. S. Mintz, L. F. Satler, R. Mehran, P. W. Serruys, A. J. Lansky, P. Fitzgerald, B. Bhargava, K. M. Kent, A. D. Pichard, and M. B. Leon. (2000). “Intracoronary gamma-radiation therapy after angioplasty inhibits recurrence in patients with in-stent restenosis.” Circulation 101:2165–2171. Waksman, R., A. E. Ajani, R. L. White, E. Pinnow, R. Dieble, T. B. Bui, M. Taaffe, L. Gruberg, G. S. Mintz, L. F. Satler, A. D. Pichard, K. M. Kent, and J. Lindsay. (2001). “Prolonged antiplatelet therapy to prevent late thrombosis after intracoronary gamma-radiation in patients with in-stent restenosis: Washington Radiation for In-Stent Restenosis Trial plus 6 months of clopidogrel (WRIST PLUS).” Circulation 103:2332–2335. Waksman, R., A. E. Raizner, A. C. Yeung, A. J. Lansky, and L. Vandertie. (2002). “Use of localised intracoronary beta radiation in treatment of in-stent restenosis: The INHIBIT randomised controlled trial. Lancet 359:551–557. Erratum in Lancet 359:1950. Waksman, R., A. E. Ajani, R. L. White, C. Chan, B. Bass, A. D. Pichard, L. F. Satler, K. M. Kent, R. Torguson, R. Deible, E. Pinnow, and J. Lindsay. (2004). “Five-year follow up after intracoronary gamma radiation therapy for in-stent restenosis.” Circulation 109(3):340–344.
Chapter 51
Brachytherapy with Miniature Electronic X-ray Sources Mark J. Rivard, Ph.D.,1 Larry A. DeWerd, Ph.D.,2 and Heather D. Zinkin, M.D.1 1 Department of Radiation Oncology, Tufts University School of Medicine Boston, Massachusetts 2 Department of Radiation Oncology, University of Wisconsin Madison, Wisconsin Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 889 History and Rationale for Electronic Brachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 889 Electronic Brachytherapy Source Manufacturers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 889 Advanced X-ray Technologies, Inc. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 890 Carl Zeiss Surgical GmbH. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 892 Xoft, Inc. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 894 Clinical Implementation Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 896 Regulatory Approvals and Licensing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 896 Dosimetric Characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 897 Treatment Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 897 Quality Assurance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 897 Clinical Trials for eBx Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 898
Introduction History and Rationale for Electronic Brachytherapy Over the past 100 years, the field of brachytherapy using radionuclide-based sources has become an integral component in the armamentarium of radiation therapy. Related to the discovery of radioactivity, brachytherapy was initiated using 226Ra, and later evolved to the current mix of sources available today, including 103Pd, 125I, 137Cs, and 192Ir as commonly used brachytherapy sources in the United States. However, utilization of radionuclide-based sources has unfortunately become more onerous given increased regulatory oversight and national security concerns. Thus, significant research over the past two decades has been performed to develop x-ray emitting, electronic brachytherapy (eBx) sources having similar properties to established radionuclide-based sources. Table 1 presents some of the potential advantages and disadvantages of eBx. Differences in cost between eBx and radionuclide-based brachytherapy programs have not yet been quantified. For specific comparisons, such as for a high dose rate (HDR) 50 kVp source and an HDR 192Ir source, the additional advantage of eBx is the ability to readily shield the source and to permit the presence of hospital personnel or loved ones to accompany the patient during treatment. The potential benefits of eBx need to be assessed for their practical utility in the forthcoming years.
Electronic Brachytherapy Source Manufacturers Research on eBx sources has been conducted in the past by many enterprises such as X-ray Therapeutics (formerly Interventional Innovations) purchased by Medtronic, X-Technologies (an Israeli firm), and RADI Medical Systems in Sweden. The sources manufactured by Pulse Sciences Inc. (Albuquerque, NM) have been studied by Karnas and colleagues (Karnas et al. 2001). Here, the operating voltage was 62.5 kV, and electrons were guided down a 4-mm diameter steel needle 20 cm long. Dose rates up to 3.84 Gy/hour were measured at 1 cm. Lack of electron focusing, subsequent diffuse x-ray production, and the relatively low dose rate diminished enthusiasm for the prototype studied.
889
890
Mark J. Rivard et al. Table 1. Advantages and Disadvantages of Electronic Brachytherapy (eBx) Compared to Radionuclide-based Brachytherapy Source
Advantages
Disadvantages
Radionuclide-based brachytherapy (Bx)
Well-established therapeutic use Well-established calibration procedures Fixed photon spectrum and half-life
Fixed dosimetry properties Radioactive waste concerns Need for regular source shipments due to decay
Electronic brachytherapy (eBx)
High specific activity, relatively small size User-adjustable dose rate (on/off) User-adjustable dosimetric properties Lessened radiological exposure to staff
Unproven clinical application No NIST calibration protocol Output variability among sources Typically larger in size
The three groups listed alphabetically below are currently working on development, and will be the only ones included in this review: 1. Advanced X-ray Technologies, Inc. (Birmingham, MI). 2. Carl Zeiss Surgical GmbH (Oberkochen, Germany). 3. Xoft Inc. (Fremont, CA). Further discussion follows on the specifics of what each company offers, and their positions in respect to dosimetric characterization and clinical implementation. Advanced X-ray Technologies, Inc. George Gutman leads the effort at Advanced X-ray Technologies (AXT) Inc., and has a number of patents on the process to generate x-rays and deliver them down a narrow, needlelike in vivo applicator onto a “pseudo” target for subsequent production of secondary x-rays (Gutman 2003a,b). Unlike the other systems described in this chapter, the AXT source works via indirect action. A 1.5 kW SEIFERT x-ray tube operated at 90 kV is used for primary x-ray production. Either silver (Ag) or tungsten (W) is used as the primary target. Secondary x-rays are produced with a tiny pseudotarget within the in vivo needle tube. This secondary target may be composed of molybdenum (Mo), neodymium (Nd), or erbium (Er). Figures 1a and 1b are diagrams of the source assembly and method of x-ray propagation, and Figure 2 shows the translation jig (Gutman et al. 2004; Lee et al. 2004). While the system has not yet received U.S. Food and Drug Administration (FDA) approval for humanuse, extensive measurements (in-water using thermoluminescent dosimeters (TLDs), and in-air using a scintillation counter) and calculations (MCNP5) are underway to characterize the spatial output of the device. Some of these properties have already been reported (Gutman et al., 2004; Lee et al., 2004). With the Ag:Mo target combination, AAPM Task Group 43 (TG-43) (Nath et al. 1995; Rivard et al., 2004) radial dose function values of g(0.5 cm) and g(3 cm) are 2.52 and 0.17, respectively (Lee et al., 2005). The most penetrating target combination is W:Nd, with g(0.5cm) ~1 and g(3cm) ~0.9. With the Ag:Mo target combination and an incident power of 0.32 mW on the secondary target, an in-water dose rate of 5.7 Gy/h may be achieved at 1 cm with the AXT source. In-water dose rates using the W:Nd target combination are about an order of magnitude less. However, these results are highly dependent (by a factor of 3) on the secondary target angle (see Figure 1b).
51–Brachytherapy with Miniature Electronic X-ray Sources
891
Figure 1a. The general schema for application of the Advanced X-Ray Technologies, Inc. source. [Reprinted from Phys Med Biol. “A novel needle-based miniature x-ray generating system.”G. Gutman, E. Sozontov, E. Strumban, F.-F. Yin, S.-W. Lee, and J. H. Kim. Vol. 49, pp. 4677–4688. © (2004), with permission of IOP Publishing, Bristol.]
Figure 1b. Schematic view of the Advanced X-Ray Technologies, Inc. source illustrating propagation of x-rays for in vivo application. [Reprinted from Phys Med Biol. “A novel needle-based miniature x-ray generating system.”G. Gutman, E. Sozontov, E. Strumban, F.-F. Yin, S.-W. Lee, and J. H. Kim. Vol. 49, pp. 4677–4688. © (2004), with permission of IOP Publishing, Bristol.]
892
Mark J. Rivard et al.
Figure 2. Photograph of the experimental setup and translation jig for the Advanced X-Ray Technologies, Inc., source. The primary x-ray source, guide tube, and secondary target are illustrated. [Reprinted from “Monte Carlo Dose Calculation and Experimental Result Comparison of a Novel Intensity-Modulated X-Ray Brachytherapy,” S.-W. Lee, F-F. Yin, G. Gutman, E. Strumban, E. Sozontov, and J. H. Kim, in proceedings of the 14th International Conference on the Use of Computers in Radiation Therapy (ICCR), Seoul, Korea, pp. 252–254. © (2004), with permission from Byong Yong Yi.]
Whereas HDR 192Ir sources have symmetric two-dimensional anisotropy function results about the longitudinal plane, this is not the case for the AXT source due to its construction geometry. For polar angles ranging from 180° to 45°, with 180° indicating the source distal direction, the AXT source output varied by approximately a factor of 3 for a wide variety of clinically relevant distances. Also due to source construction geometry, the dependence of source output on azimuthal angle varied by a factor of 6, which is much, more variable than for HDR 192Ir sources (Lee et al., 2005). To use this spatial dependence on output advantageously, the manufacturer is trying to apply the distribution as intensity-modulated brachytherapy. Carl Zeiss Surgical GmbH. Alan P. Sliski led the effort at Photoelectron Corporation for developing an electronic brachytherapy source, the INTRABEAM™ photon radiosurgery system (model PRS400), and has many U.S. patents for this technology (see, for example, Sliski et al., 1994). The product line was acquired by Carl Zeiss (Oberkochen, Germany) and is presently available from them. From a clinical perspective, this work is somewhat more evolved than the AXT source because patient treatments have been performed in a randomized clinical trial [TARGIT (targeted intraoperative radiotherapy)]1, and the device has been used intraoperatively for the irradiation of intracranial metastases and irradiation of the tumor bed following breast-conserving surgery (Douglas et al., 1996; Vaidya et al., 2001, 2002). A diagram of the INTRABEAM, as offered by Carl Zeiss Surgical GmbH, is shown in Figures 3 and 4. 1
TARGIT. “Targeted intraoperative radiotherapy.” http://www.ctg.ucl.ac.uk/TargitTest/index.htm, last accessed May 1, 2005.
51–Brachytherapy with Miniature Electronic X-ray Sources
893
Figure 3. Photograph of the INTRABEAM™ photon radiosurgery system (model PRS400), initially developed by the Photoelectron Corporation (now Boston Science Corporation), now offered by Carl Zeiss Surgical GmbH. The compact, battery powered generator and x-ray tube with a spherical intraoperative applicator are at the end of the manipulator arm. [Courtesy of Carl Zeiss Surgical GmbH.]
Figure 4. Clinical application of the INTRABEAM™ system in an interstitial neurosurgery setting. [Courtesy of Carl Zeiss Surgical GmbH.]
894
Mark J. Rivard et al.
X-rays are generated by the INTRABEAM using an electron beam focused from a conventional miniature electron gun onto the tip of a 3.2 mm diameter rigid drift tube (Biggs et al., 1993; Dinsmore et al., 1994, 1996; Yanch and Zervas 1995; Yanch and Harte 1996; Beatty et al., 1996; Hakim et al., 1997). The tip of the drift tube comprises a beryllium tube with hemispherical x-ray window coated on the inside with a thin gold target, used to stop most of the incident electrons, and coated on the outside with thin nickel and titanium nitride films for biocompatibility. The INTRABEAM can be used at 30, 40, and 50 kVp, and can run at 5, 10, 20, 40 µA. The 50 kVp electron beam is dithered around the center of the hemispherical beryllium window using a dynamic magnetic deflection system. The INTRABEAM system has a dose distribution that is approximately spherically symmetric but drops off in the proximal direction. Depth-dose data for a 25-min treatment time using the INTRABEAM system were reported as 20, 5, and 1 Gy at 0.1, 1.0, and 2.7 cm, respectively, from the outer surface of a 3.5-cm diameter spherical applicator (Vaidya et al., 2001). At a beam current of 40 µA in an uncooled catheter, dose rates at the surfaces of 1, 2, 3, and 4 cm diameter applicators are 780, 168, 60, and 30 Gy/h, respectively. Development of this technology and characterization of the dose distribution is underway using TLDs and ionization chambers (Soares and Sliski 2002; Soares et al., 2005). The development of the INTRABEAM is now finished, and it has been released for the U. S. market since 1999 by the FDA (as well as in Europe and China). Already more than 500 patients have received intraoperative radiotherapy following breast-conserving surgery with the INTRABEAM. Xoft, Inc. In 1998, Xoft, Inc. (formerly Xoft microTube, Inc.) intended to make eBx sources for intravascular brachytherapy to treat coronary artery disease. Like the other companies mentioned, Xoft has accrued intellectual property, and has U.S. patents approved and pending (see, for example, Forman, Lovoi, and Rusch 2000). With a dwindling market share due to introduction of drug-eluting stents, Xoft switched in 2001 to pursuing brachytherapy treatment for accelerated partial breast irradiation and other anatomic sites. Recently, small x-ray tubes have been developed that offer the prospect of electronic brachytherapy. Xoft Inc., has developed the AXXENT™ miniature x-ray brachytherapy source, and its dosimetric properties have been characterized (Rusch and Rivard 2004; Chiu-Tsao et al., 2004; Rivard et al., 2005). The source consists of a disposable, microminiature x-ray tube (Figure 5) integrated into a cooled, flexible, disposable sheath, which is directly attached to a treatment control console (Figure 6). Water circulating within the cooling sheath has intimate contact with the anode, allowing a higher power dissipation and higher dose-rate, without thermal damage to the source, surrounding probe structures, or the patient. The source is capable of operating voltages ranging from 20 kV to 50 kV. This provides photons with maximum energy less than 50 keV, so that shielding concerns are negligible in comparison to shielding needed for HDR 192Ir sources. Since the voltage (dose falloff) and current (beam intensity) may be modulated, an increased level of dose conformity may be possible than for a source having a fixed spectrum such as from a radionuclide. More research is needed examining these potential benefits. At an operating voltage of 50 kV, the source can produce an air kerma strength of 1,400 Gy/h at 1 cm (approximately three times that of a 10 Ci HDR 192Ir source) with a tube current of 300 µA. Operation down to 1 µA is feasible. Like conventional HDR remote afterloading brachytherapy sources using radionuclides attached to delivery drive wires, the source can be positioned within the patient at multiple dwell positions for providing highly conformal radiotherapy delivery. At 50 kV, the AAPM TG-43 (Nath et al., 1995; Rivard et al., 2004) radial dose function values (Figure 7) of g(0.5 cm) and g(3 cm) are 1.4 and 0.47, respectively (Rusch and Rivard 2004; Rusch et al., 2004). To date, no humans have been treated with the device, and an animal trial using Nubian milk goats has been completed towards seeking FDA approval in 2005. Patient treatments will commence with a phase IV trial for APBI using the AXXENT source at various institutions in North America. Initial clinical applications will operate at of 40, 45, and 50 kV and a fixed current of 300 µA.
51–Brachytherapy with Miniature Electronic X-ray Sources
895
Figure 5. Photographs and schematic diagram of the AXXENT™ x-ray source by Xoft, Inc. The LHS image shows the Source in operation. The bottom illustration shows the Source enclosed by a gray water-cooling sheath. The sheath outer diameter is 5.3 mm, and is quite flexible except for the distal 15 mm.
Figure 6. Schematic diagram of the entire AXXENT™ source probe illustrating the relationship of the source to the water coolant inlet/outlet and HV connector. Though depicted as a straight, needle-like device, in reality the probe is quite flexible and can negotiate turns as small as 5 cm in diameter. The RHS image shows the treatment control console adjacent to a radiotherapy stretcher bed.
896
Mark J. Rivard et al.
Figure 7. Comparison of point-wise radial dose functions, gP(r), for 103Pd, 125I, 192Ir, and the AXXENT™ eBx source operated at 50 kV.
Clinical Implementation Requirements With the aforementioned and future eBx sources, certain aspects should be in place preceding clinical implementation. Having FDA approval and an institutional license for clinical use are obvious. Proper clinical implementations with safe and consistent use among institutions are an additional hurdle, which can be overcome with appropriate protocols. In the absence of an AAPM task group addressing the specifics for eBx sources, some of the aspects needed for proper clinical implementation are described below.
Regulatory Approvals and Licensing The FDA has published a set of recommendations appropriate to radionuclide-based brachytherapy sources (U.S. FDA 1998, 2000, 2005). With the advent of eBx sources, these recommendations have been recently applied in a somewhat haphazard and inconsistent fashion toward protecting the public and providing full disclosure. Since these standards do not always directly apply to eBx, other standards need development. Should clinical use of eBx sources become significant in the United States, it is hoped that the FDA Center for Devices and Radiological Health (CRDH) will eventually develop recommendations specific to eBx sources and manufacturers. Regulation of x-ray tubes is often quite different from regulation of brachytherapy sources containing radionuclides, and no radioactive materials license is needed. X-ray tubes are generally under the purview of state regulations. The U.S. Nuclear Regulatory Commission (NRC) or agreement states issue approvals for radionuclide-based sources through the National Sealed Source and Device Registry (U.S. NRC 2005a). This database is now indefinitely closed to public access (U.S. NRC 2005b) due to Homeland Security concerns and maintenance.
51–Brachytherapy with Miniature Electronic X-ray Sources
897
Dosimetric Characterization Clearly one cannot perform brachytherapy using any type of source without first characterizing the dose distribution. However, it is surprising how many manufacturers of brachytherapy sources fail to satisfy the specific needs of many end-users when the product is deemed ready for clinical applications. In order that this shortcoming is avoided for eBx sources, it is worth mentioning the scope of data needed and particulars that must be addressed preceding clinical use of eBx sources. While the majority of eBx sources in this chapter are meant to serve as replacements for HDR 192Ir sources, the authors recommend that manufacturers and users follow the AAPM guidelines for low-energy (E < 50 keV), photon-emitting, radionuclide-based brachytherapy sources (Williamson et al., 1998) based on the average energy of the in-air photon energy spectrum of the eBx sources examined herein and their dose falloff or g(r). This also includes efforts towards inclusion on the Joint AAPM/ Radiological Physics Center (RPC) Source Registry (Radiological Physics Center 2005). Only after peer-review and scientific society consensus analysis of the requisite dosimetry parameters should the source be ready for clinical use. The authors do not recommend applying the low-energy, photon-emitting, teletherapy dosimetry protocol to these sources (Ma et al. 2001), and support use of the low-energy, photon-emitting, brachytherapy dosimetry protocol (Nath et al., 1995; Rivard et al., 2004).
Treatment Planning Treatment planning for eBx sources should be no different from conventional brachytherapy sources, and the reader is advised to consider customizing the recommendations of Fraass and colleagues (Fraass et al., 1998). There will be some subtleties like accounting for the inconsistency of coordinate system orientation, as is the case for Nucletron- and Varian-based HDR 192Ir sources, but clinical implementation and ongoing quality assurance (QA) efforts from a treatment-planning perspective should be straightforward if eBx source manufacturers can fabricate sources with relatively constant outputs. If this were not the case, the user would need to enter source-specific, as opposed to source model-specific, dosimetry parameters to essentially account for unique dosimetric properties. An onerous task at best! The assumption of source constancy as a function of time throughout the treatment course could also be in jeopardy. Clearly, product quality from the perspective of eBx source output consistency is a critical aspect to clinical implementation and subsequent treatment planning.
Quality Assurance It is the authors’ view that QA for the eBx sources examined in this chapter should initially be based on that for an HDR remote afterloader such as HDR 192Ir (Glasgow et al., 1993; Kubo et al., 1998). Nath and colleagues specify a code of practice and general QA practices for brachytherapy sources (Nath et al. 1997). From these guidelines, QA tasks appropriate to eBx sources should be adopted, such as new source calibration procedures and the frequency of subsequent radiation output checks. Additional QA tasks must be performed based on the specifics of the eBx and related hardware/software. For instance, the consistency of eBx source output as a function of radial distance, polar angle, and azimuthal angle is important. Manufacturer geometric tolerances are stringent for radionuclide-based sources due to intrinsic design feature. However, eBx sources may have significant variations in these three parameters from source to source. Furthermore, there are no national standards currently in place for calibrations of eBx sources as is the case for conventional, radionuclide-based sources (DeWerd et al., 2004). Clearly, understanding of manufacturer tolerances and subsequent QA efforts by medical physicists need to be well-coordinated preceding clinical implementation.
898
Mark J. Rivard et al.
Clinical Trials for eBx Sources Some of the motivating reasons for using eBx sources are given in Table 1. Based on equivalence of dose rates and photon energy spectra with conventional sources, HDR 192Ir and 125I respectively, it seems logical to suppose that clinical results using these new eBx sources would be similar, and that the main driving force behind this selection would be based on features and ease of use. However, one cannot truly be certain until randomized, unbiased clinical trials are performed to compare these novel sources with the conventional, radionuclide-based sources for which we have over 100 years of clinical experience. If eBx sources are included in the upcoming NSABP/RTOG2 Phase III clinical trial, perhaps a multivariate analysis would glean significant differences in toxicity or outcome? With a rigorous outlook and careful planning, the merits of new technology should be able to be assessed properly from the onset.
References Beatty, J., P. J. Biggs, K. Gall, P. Okunieff, F. S. Pardo, K. J. Harte, M. J. Dalterio, and A. P. Sliski. (1996). “A new miniature x-ray source for interstitial radiosurgery: Dosimetry.” Med Phys 23:53–62. Biggs, P., J. Beatty, K. Gall, K. Harte, and A. Sliski. (1993). “Absolute dosimetry for a new 40 kV x-ray device used for stereotactic radiation therapy.” (Abstract). Med Phys 20:925. Chiu-Tsao, S.-T., T. Rusch, S. Axelrod, H. Tsao, and L. Harrison. (2004). “Radiochromic film dosimetry for a new electronic brachytherapy source.” (Abstract). Med Phys 31:1913. DeWerd, L. A., M. S. Huq, I. J. Das, G. S. Ibbott, W. F. Hanson, T. W. Slowey, J. F. Williamson, and B. M. Coursey. (2004). “Procedures for establishing and maintaining consistent air-kerma strength standards for low-energy, photon-emitting brachytherapy sources: Recommendations of the Calibration Laboratory Accreditation Subcommittee of the American Association of Physicists in Medicine.” Med Phys 31:675–681. Dinsmore, M., J. C. Yanch, A. P. Sliski, and K. J. Harte. (1994). “New x-ray generator for interstitial radiotherapy.” Trans Am Nucl Soc 70:24–25. Dinsmore, M., K. J. Harte, A. P. Sliski, D. O. Smith, P. M. Nomikos, M. J. Dalterio, A. J. Boom, W. F. Leonard, P. E. Oettinger, and J. C. Yanch. (1996). “A new miniature x-ray source for interstitial radiosurgery: Device description.” Med Phys 23:45–52. Douglas, R. M., J. Beatty, K. Gall, R. F. Valenzuela, P. Biggs, P. Okunieff, and F. S. Pardo. (1996). “Dosimetric results from a feasibility study of a novel radiosurgical source for irradiation of intracranial metastases.” Int J Radiat Oncol Biol Phys 36:443–450. Forman, M. R., P. A. Lovoi, T. W. Rusch. (2002). Radiation for Inhibiting Hyperplasia After Intravascular Intervention. U.S. Patent 6,390,967. Fraass, B., K. Doppke, M. Hunt, G. Kutcher, G. Starkschall, R. Stern, and J. Van Dyk. (1998). “American Association of Physicists in Medicine Radiation Therapy Committee Task Group 53: Quality assurance for clinical radiotherapy treatment planning.” Med Phys 25:1773–1829. Also available as AAPM Report No. 62. Glasgow, G. P., J. D. Bourland, P. W. Grisby, J. A. Meli, and K. A. Weaver. “Remote Afterloading Technology.” AAPM Report No. 41. New York: American Institute of Physics, 1993. Gutman, G. (2003a). Apparatus and Method for Generating a High Intensity X-Ray Beam with a Selectable Shape and Wavelength. U.S. Patent 6,493,421. Gutman, G. (2003b). X-Ray System with Implantable Needle for Treatment of Cancer. U.S. Patent 6,580,940. Gutman, G., F.-F. Yin, E. Strumban, P. Petrashen, and J. H. Kim. (2003). “A device for intensity-modulated x-ray brachytherapy (IMXBT).” (Abstract). Med Phys 30:1469. Gutman, G., E. Sozontov, E. Strumban, F.-F. Yin, S.-W. Lee, and J. H. Kim. (2004). “A novel needle-based miniature x-ray generating system.” Phys Med Biol 49:4677–4688. Hakim, R., N. T. Zervas, F. Hakim, W. E. Butler, J. Beatty, J. C. Yanch, P. J. Biggs, K. P. Gall, A. P. Sliski. (1997). “Initial characterization of the dosimetry and radiology of a device for administering interstitial stereotactic radiosurgery.” Neurosurg 40:510–516. 2
National Surgical Adjuvant Breast and Bowel Project (NSABP). http://www.nsabp.pitt.edu/, last accessed May 1, 2005.
51–Brachytherapy with Miniature Electronic X-ray Sources
899
Karnas, S. J., N. Avvakumov, E. Yu, and J. J. Battista. (2001). “Operational characteristics of a prototype x-ray needle device.” Phys Med Biol 46:97–106. Kubo, H. D., G. P. Glasgow, T. D. Pethel, B. R. Thomadsen, and J. F. Williamson. (1998). “High dose-rate brachytherapy treatment delivery: Report of the AAPM Radiation Therapy Committee Task Group No. 59.” Med Phys 25(4):375–403. Also available as AAPM Report No. 61. Lee, S.-W., F-F. Yin, G. Gutman, E. Strumban, E. Sozontov, and J. H. Kim. (2004). “Monte Carlo Dose Calculation and Experimental Result Comparison of a Novel Intensity-Modulated X-Ray Brachytherapy.” 14th International Conference on the Use of Computers in Radiation Therapy (ICCR), Seoul, Korea, pp. 252–254. http://www.iccr.info/conference_3.asp last accessed May 1, 2005. Lee, S.-W., F.-F. Yin, G. Gutman, E. Strumban, E. Sozontov, and J. H. Kim. (2005). “Monte Carlo dose calculation and experimental result comparison of a novel intensity-modulated x-ray brachytherapy.” Med Phys (In press). Ma, C.-M., C. W. Coffey, L. A. DeWerd, C. Liu, R. Nath, S. M. Seltzer, and J. P. Seuntjens. (2001). “AAPM protocol for 40-300 kV x-ray beam dosimetry in radiotherapy and radiobiology.” Med Phys 28:868-893. Also available as AAPM Report No. 76. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee, Task Group No. 43.” Med Phys 22:209–234. Also available as AAPM Report No. 51. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24(10):1557–1598. Also available as AAPM Report No. 59. Radiological Physics Center (RPC). “Joint AAPM/RPC Registry of Low-energy Brachytherapy Seeds Meeting the AAPM dosimetric Prerequisites.” http://rpc.mdanderson.org/rpc/htm/Home_htm/Low-energy.htm last accessed May 1, 2005. Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. (2004). “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Also available as AAPM Report No. 84. Rivard, M. J, S. D. Davis, L. A. DeWerd, T. W. Rusch, and S. Axelrod. (2005). “Calculated and measured brachytherapy dosimetry parameters for the Xoft AXXENT X-Ray source: An electronic brachytherapy source.” Med Phys (submitted). Rusch, T. W., and M. J. Rivard. (2004). “Application of the TG-43 dosimetry protocol to electronic brachytherapy sources.” (Abstract). Radiother Oncol 71(S2):S84. Rusch, T. W., S. D. Davis, L. A. DeWerd, R. Burnside, S. Axelrod, and M. J. Rivard. (2004). “Characterization of a new miniature x-ray source for electronic brachytherapy.” (Abstract). Med Phys 31:1807. Sliski, A. P., M. T. Dinsmore, A. J. Boom, and N. T. Zervas. (1994). Low Power X-Ray Source with Implantable Probe for Treatment of Brain Tumors. U.S. Patent 5,369,679. Soares, C. G., and A. P. Sliski. (2002). “Absorbed dose measurements of a handheld 50-kV x-ray source with ionization chambers and plastic scintillators.” (Abstract). Med Phys 29:1351. Soares, C., C. Drupieski, B. Wingert, G. Pritchett, V. Pagonis, M. O’Brien, A. Sliski, P. Bilski, and P. Olko. (2005). “Absorbed dose measurements of a handheld 50 kVp X-ray source in water with thermoluminescence dosimeters.” Rad Prot Dosim 114:(In press). U.S. Food and Drug Administration (FDA). (1998) “Guidance for the Submission of Premarket Notifications for Radionuclide Dose Calibrators.” U.S. Dept. of Health and Human Services, Food and Drug Administration, Center for Devices and Radiological Health, Radiological Devices Branch, http://www.fda.gov/cdrh/ode/ radcalibrators.pdf, last accessed May 1, 2005. U.S. Food and Drug Administration (FDA). (2000) “Guidance for the Submission of Premarket Notifications for Photon-Emitting Brachytherapy Sources.” U.S. Dept. of Health and Human Services, Food and Drug Administration, Center for Devices and Radiological Health, Radiological Devices Branch, http://www.fda.gov/cdrh/ ode/guidance/1777.pdf, last accessed May 1, 2005. U.S. Food and Drug Administration (FDA). (2005) “Guidance Documents.” U.S. Dept. of Health and Human Services, Food and Drug Administration, Center for Devices and Radiological Health, Radiological Devices Branch, http://www.fda.gov/cdrh/guidance.html, last accessed May 1, 2005.
900
Mark J. Rivard et al.
U.S. Nuclear Regulatory Commissions (NRC). (2005a) “Sealed Sources and Devices.” http://www.nrc.gov/ materials/miau/sealed-source.html, last accessed May 1, 2005. U.S. Nuclear Regulatory Commissions (NRC). (2005b) “Sealed Source and Device Registry.” http://www.hsrd.ornl. gov/nrc/ssdinfo.htm, last accessed May 1, 2005. Vaidya, J. S., M. Baum, J. S. Tobias, D. P. D’Souza, S. Morgan, S. V. Naidu, M. Metaxas, K. J. Harte, A. P. Sliski, and E. Thomson. (2001). “Targeted intraoperative radiotherapy (Targit): An innovative method of treatment for early breast cancer.” Ann Oncol 12:1075–1080. Vaidya, J. S., M. Baum, J. S. Tobias, S. Morgan, and D. D’Souza. (2002). “The novel technique of delivering targeted intraoperative radiotherapy (Targit) for early breast cancer.” Eur J Surg Oncol 28:447–454. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, and R. Nath. (1998). “Dosimetric prerequisites for clinical use of new low energy photon interstitial brachytherapy sources.” Med Phys 25:2269–2270. Yanch, J. C., and N. T. Zervas. (1995). “The photon radiosurgery system.” Science Med 2:38–46. Yanch, J. C., and K. J. Harte. (1996). “Monte Carlo simulation of a miniature, radiosurgery x-ray tube using the ITS 3.0 coupled electron-photon transport code.” Med Phys 23:1551–1558.
Chapter 52
Thermal Therapy for Cancer Paul R. Stauffer, C.J. Diederich, and J. Pouliot Radiation Oncology Department University of California San Francisco, CA 94143 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901 Thermal Therapy Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 901 Tissue Heating Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 904 Mechanisms of Heating . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 905 Thermal Conduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 905 Electromagnetic (EM) Power Deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 905 Ultrasound (US) Power Deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 907 Methods of Heating Tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 908 External Electromagnetic Superficial Heating Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 909 External Superficial Ultrasound Heating Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 914 External Deep-Regional Heating Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 916 Interstitial and Intracavitary Deep Local Heating Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 919 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 923 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 923
Introduction This overview provides a basic outline of thermal therapy options for treatment of cancer. The report reviews the rationale for heating tissue in cancer therapy and briefly describes the devices and techniques that have proven useful in clinical applications to date. Due to highly variable tissue properties and the wide range of anatomic sites for tumors of different size, shape and depth in the body, an equally wide range of technology is required. We will illustrate basic heating principles with examples of performance from the most common applicator types and describe several examples of current workhorse devices and evolving technology that should be available in the near future.
Thermal Therapy Approaches The realm of thermal medicine includes all therapeutic treatments based on the transfer of heat energy into or out of body tissues. In this sense, thermal therapy may be performed either with cryotherapy (<50°C for >10 min) or with at least three overlapping but characteristically different protocols of elevated temperature: low temperature warming of tissue at approximately 39-41°C for a cumulative thermal dose of generally less than 5 cumulative equivalent minutes at 43°C (< 5 CEM43°C); moderate temperature hyperthermia (42-45°C for 15-60 min for a thermal dose of 15-240 CEM43°C, usually repeated for multiple fractions to achieve higher cumulative thermal dose); and high temperature thermal ablation therapy (usually > 50°C for > 4-6 min, or thermal doses > 512 CEM43°C). Cryotherapy ablation is an established branch of thermal medicine that has seen use in a number of clinical applications where thermal surgery is allowable due to adequate biological reserve and/or physical separation of the treatment volume from surrounding critical normal tissue structures. Low temperature thermal therapy has been demonstrated to augment the delivery and effectiveness of chemotherapy, even at temperatures as low as 39-41°C. By increasing blood perfusion, permeability of the tumor microvasculature, and cellular metabolic rate, local tissue hyperthermia has been shown to increase drug uptake and activity (Huang et al. 1994; Dewhirst et al. 2003), enhance DNA damage through accelerated cellular metabolism, and potentially decrease multiple drug resistance (van
901
902
Jean Pouliot et al.
der Zee and Gonzalez 2002). When these low temperatures are maintained for long time periods up to 72 hours or more and combined with long duration brachytherapy, dramatic radiosensitizing effects have also been demonstrated in animal models and in vitro (Xu et al. 2002; Armour 2004; Armour et al. In press). Historically, this protocol has generally been considered together with moderate temperature hyperthermia described below, and labeled “Hyperthermia” therapy. Recent biological and clinical investigations have identified sufficiently different mechanistic effects for long-term low temperature heating that clinical applications of this protocol should now be considered separately. Even so, heating considerations for this therapeutic approach are nearly identical to those described below, with a more modest thermal dose goal that is readily achievable clinically and is showing great promise for some clinical applications. Moderate temperature hyperthermia (HT) has been demonstrated in numerous biological and clinical investigations over the past three decades to significantly enhance clinical responses to radiation therapy and chemotherapy (van der Zee and Gonzalez 2002; Wust et al. 2002; Hehr et al. 2003). In general, the biological rationale for combining HT with RT is as follows (Overgaard 1989; Dewey 1994): i) hyperthermia alone is cytotoxic when delivered at temperatures > ~45°C for 60 min, temperatures that are typical in portions of the tumor that are most distant from the cooling effects of blood perfusion; ii) moderate hyperthermia (40-44°C) is a radiosensitizer which increases radiation damage and prevents subsequent cellular repair of sublethal radiation damage, especially in a nutrient deprived micro-environment with low blood supply, hypoxia, and low pH—such as that often found in large portions of a malignant tumor; and iii) hyperthermia increases tumor perfusion as blood vessels dilate to carry away excess heat; this generally increases tumor oxygenation and pH, thereby increasing sensitivity of tumor cells to radiation damage (Oleson 1995; Vujaskovic et al. 2000). These effects are complementary and partially selective to tumor cells where the hypoxic nutrient deprived cells are typically more radioresistant and susceptible to thermal damage, leading to thermal enhancement of radiation cell kill. These effects can be further
52–Thermal Therapy for Cancer
903
enhanced when heat and radiation treatments are applied simultaneously for maximum synergism (Overgaard 1989; Dewey 1994). Despite the very encouraging results of numerous Phase I and Phase II studies of moderate temperature hyperthermia as an adjuvant to radiation in the treatment of cancer (Sneed and Phillips 1991; Seegenschmiedt et al. 1996; Sneed et al. 1998a; Dewhirst et al. 2003), the results of early Phase III randomized trials performed in the US for superficial (Perez et al. 1991) and interstitial (Emami et al. 1996) treatments were not significantly positive. The lackluster results of these first trials were determined to be due to poor selection of heating technology and inadequate heating of the complete tumor volume. Due to the “failure” of these studies, significant efforts were undertaken to establish quality assurance procedures and guidelines for future trials (Dewhirst et al. 1990; Emami et al. 1991; Sapozink et al. 1991; Waterman et al. 1991; Lagendijk et al. 1998). Thus, the clinical rationale for combining heat with radiation therapy has been re-defined recently by more carefully controlled randomized clinical trials (Overgaard et al. 1995; Vernon et al. 1996; Sneed et al. 1998b; van der Zee et al. 2000; van der Zee and Gonzalez 2002) that demonstrate dramatic improvements in tumor response and control. For example: the complete response (CR) rate was increased from 41% to 59% in recurrent breast cancer (Vernon et al. 1996) and from 43% to 73% in previously irradiated recurrent breast cancer (Sherar et al. 1997); the CR rate for head and neck tumors was increased from 41% to 83%, with an increase in 5 year local control from 24% to 69% (Valdagni et al. 1988; Valdagni and Amichetti 1994); the two-year survival from glioblastoma multiforme was increased from 15% to 31% (Sneed et al. 1998b); the two-year local control for recurrent or metastatic malignant melanoma was increased from 28% to 46% (Overgaard et al. 1995); and the CR was increased from 39% to 55% in advanced pelvic disease (van der Zee et al. 2000). Moreover as demonstrated in numerous studies over the past two decades, the response rate improves dramatically with increasing thermal dose (Dewhirst and Sim 1984; Dewhirst et al. 1984; Oleson et al. 1989; Kapp and Cox 1993; Leopold et al. 1993; Oleson et al. 1993; Kapp and Cox 1995; Sherar et al. 1997), with recent data suggesting a minimum thermal dose of >10 CEM43°C is adequate for many sites. More extensive reviews of recently completed randomized trials of hyperthermia ± radiation therapy are given by van der Zee 2002 (van der Zee and Gonzalez 2002), and Dewhirst (Dewhirst et al. 2003). Thermal ablation therapy or high temperature hyperthermia is being used for direct tissue destruction of cancerous tumors as well as benign disease (e.g., prostatic BPH) when the treatment volume is located in anatomic sites having adequate biological reserve and/or separation from surrounding critical normal tissues. In this form of therapy, high temperatures are applied to completely destroy tumor vasculature and most importantly to disrupt and coagulate all structural proteins in the tumor cells. These effects are irreversible and lead to immediate cellular death in the high thermal dose region. In areas of lower dose near the borders of a thermally coagulated lesion, the cells may survive the initial thermal insult but go on to die within 2-3 days due to non-lethal hyperthermic effects as described above. It is within these border regions that adjunctive therapies such as radiation or chemotherapy can be most helpful in accentuating the radial penetration of tissue destruction. Although the effect of thermal ablation therapy is similar to cryotherapy in terms of immediate direct cell kill, thermal ablation requires a temperature differential of only 13°C (37-50°C) whereas tissue must be cooled almost –90°C (37 → –50°C) to ensure cryogenic death. With such a large gradient required for cryotherapy and only thermal conduction processes available to remove heat from the tissue, there is increasing interest in thermal ablation therapy as a more spatially and temporally controllable procedure for thermal surgery applications, even though cryotherapy is currently an established procedure for many sites. Clinical applications of thermal ablation therapy are expanding very rapidly in recent years for both benign and cancerous tissues (Stauffer and Goldberg 2004). This high temperature therapy works well in sites with well-localized tumors such as liver (Ahmed and Goldberg 2002; Erce and Parks 2003; Garcea et al. 2003), kidney (Murphy and Gill 2001), brain (Anzai et al. 1995; Leonardi and Lumenta 2002), lung (Dupuy et al. 2002), breast (Hayashi et al. 2003), and even bone (Rosenthal et al. 2003). Recent studies
904
Jean Pouliot et al.
have demonstrated the feasibility and efficacy of minimally invasive ablation therapy for treating recurrent prostate cancer using interstitial microwave antennas (Sherar et al. 2001; Sherar et al. 2003; Sherar et al. 2004), interstitial RF needles (Zlotta et al. 1998), interstitial ferromagnetic seeds (Tucker et al. 2002), and transrectal focused ultrasound combined with ultrasound image guidance (Uchida et al. 2002; Thuroff et al. 2003). One of the most precise technologies uses high-intensity focused ultrasound in combination with magnetic resonance (MR) imaging and MR thermal monitoring, providing for real-time control and assessment of the treatment for sites such as breast (Hynynen et al. 2001; Gianfelice et al. 2003) and uterine fibroids (Stewart et al. 2003; Tempany et al. 2003). Non-invasive treatment monitoring with MRI is now being utilized to guide thermal therapy in cases where real time precision control of treatment temperature and/or treatment effect is required (Quesson et al. 2000). Examples of thermal ablation therapy for non-oncologic applications include treatment of prostatic hyperplasia (Larson 2002; Zlotta and Djavan 2002), cardiac arrhythmias (Lin and Marchlinski 2003; Yee et al. 2003), uterine fibroids (Hayashi et al. 2003; Hindley et al. 2003), low back pain (Saal and Saal 2002), reduction of upper palate and turbinates to reduce snoring (Boudewyns and Van De Heyning 2000; Fischer et al. 2000), stabilization of skeletal joints (Enad et al. 2003; Noonan et al. 2003), and cosmesis such as dermal tightening (Goldberg 2003; Ruiz-Esparza and Gomez 2003). There are numerous examples of clinical thermal therapy techniques that provide a minimally-invasive alternative to surgical transurethral resection (TURP) for treating BPH, including transurethral microwave delivery (D’Ancona et al. 1997; Djavan et al. 1999; Ramsey and Dahlstrand 2000), transurethral RF energy (Bruskewitz et al. 1998), and transrectal high-intensity focused ultrasound (HIFU) (Sanghvi et al. 1999).
Tissue Heating Considerations Human tissues have heterogeneous properties that effect the resultant temperature distributions during heating. These include the electrical (dielectric constant and electrical conductivity), acoustic (acoustic velocity and attenuation), and thermal properties (thermal conductivity, heat capacity, and blood perfusion rate). These properties have a dramatic impact on how energy is radiated and/or conducted from a given heat applicator, how that energy is absorbed within the tissue, and how it is thermally redistributed within the volume. However, the most influential and difficult to compensate for physiologic variable in delivering thermal therapy is blood perfusion, which varies spatially due to the heterogeneous vasculature and temporally due to dynamic changes in blood vessel size as a function of increased temperature. Blood flow is very effective at cooling or redistributing thermal energy, making ideal uniformly high temperature distributions impossible in tissue perfused at varying rates with 37°C blood. Furthermore, the effects of blood flow change dynamically during treatment, further exacerbating efforts to deliver a uniform prescribed thermal dose. In most cases, the minimum temperatures are limited by blood perfusion and maximum temperatures are limited by patient discomfort and subsequent complications (burns). Thus, the most useful heating techniques can dynamically adjust their heating patterns to compensate for heterogeneous and dynamic blood perfusion and thereby produce more uniform tissue temperature distributions. The thermal dosimetry of a hyperthermia treatment is normally quantified by temperature parameters Tmax, T90, Tmin, T50, and Tave as a function of time (Dewhirst et al. 2003). Tissue damage at both low and high temperature exposures has been shown to be linearly dependent on exposure time and non-linearly dependent on temperature elevation. In order to quantify thermal damage, a comprehensive Arrhenius analysis was applied to many different cell lines and in vivo tissues. While minor uncertainties remain in the exact calculation of thermal dose over the wide range of temperatures from body temperature to thermally ablative temperatures at times exceeding 70 °C, the thermal isoeffect dose relationship proposed by Sapareto and Dewey (Sapareto and Dewey 1984) remains the most accurate and commonly used method of quantifying thermal damage to tissue. The data demonstrate that cell killing doubles for every
52–Thermal Therapy for Cancer
905
1 °C rise in temperature above 43 °C, while cell killing is reduced by one quarter for every 1 °C below 43 °C. Subsequent efforts to quantify heterogeneous thermal dose distributions in tumors refined the method of dose calculation to include not only the thermal isoeffect dose relationship but also the concept of minimum (Tmin) and 90th percentile (T90) temperatures within the tumor (Dewhirst et al. 1984; Dewey 1994). With this approach, thermal isoeffect dose is accumulated over time and expressed in equivalent minutes at 43 °C (EM43°C) at each measured temperature point. To describe thermal dose within an entire tumor, the dose is generally calculated every minute as a function of the current T90 (the temperature exceeded by 90% of the measured temperature points) and thermal doses summed together for each minute of treatment to obtain the total dose in Cumulative Equivalent Minutes at 43°C for the T90 temperature (CEM43°CT90) (Dewey 1994; Dewhirst et al. 2003). In general, EM43(C thermal doses of around 120240 min generate tissue necrosis, but the sensitivity of living tissue is highly variable. Brain and rectum are most sensitive (EM43°C = 30-50 min threshold) while muscle is more heat tolerant (EM43°C = 240 min). For moderate temperature hyperthermia potentiation of radiation and chemotherapy treatments, a total cumulative thermal dose of CEM43°CT90 > 6-10 minutes dramatically improves the clinical effect (Sherar et al. 1997; Jones et al. 2003; Jones et al. 2005).
Mechanisms of Heating There are three primary mechanisms for heating tissue: i) thermal conduction of heat energy driven by a temperature gradient; ii) resistive (ohmic) or dielectric losses in tissue from an applied electromagnetic (EM) field; and iii) mechanical losses from molecular oscillations caused by an applied ultrasound (US) pressure wave. A few comments regarding these basic mechanisms are given below. The underlying physical principles are described in the primary articles referenced below, in several excellent review articles, (Guy et al. 1974; Christensen and Durney 1981; Cheung and Neyzari 1984; Fessenden and Hand 1995; Stauffer 2000) and in books covering the field of hyperthermia in general (Nussbaum 1982; Field and Hand 1990; Gautherie 1990; Seegenschmiedt et al. 1995).
Thermal Conduction Thermal conduction is the oldest and simplest technique for heating tissue, normally applied either by total immersion of an arm, leg, or body in heated air, water, or other fluid, by circulation of externallyheated blood through an isolated tissue volume (e.g. leg, kidney), or by direct tissue contact with a heated surface. Due to efficient thermoregulation of the skin which exhibits up to 15 fold increase in blood flow in response to excess heat (Song et al. 1995), thermal convection cooling from 37 °C blood is capable of maintaining very high thermal gradients and it is generally not possible to heat effectively more than 23 mm distance from a hot surface if the applied temperature is below pain and skin necrosing temperature thresholds (Orenberg et al. 1986; Field and Hand 1990; Lagendijk 1990). Thus, thermal conduction is more useful for smoothing and regulating non-uniform temperature distributions induced by external power sources than as a sole mechanism of heating sub-surface disease.
Electromagnetic (EM) Power Deposition Over the frequency range of interest for depositing power in tissue with non-ionizing radiation, all tissues may be considered as lossy dielectrics that are both poor electrical conductors and poor insulators. The electrical properties of human tissues vary considerably from each other and as a function of frequency, and may be characterized by the relative dielectric constant (er) and electrical conductivity (s). The primary mechanism of EM heating of tissue also varies with frequency. At radiofrequencies below approximately 10 MHz, the alternating field induces a net movement of free electrons and power deposition results from
906
Jean Pouliot et al.
“joule heating” or resistive losses from the induced current. At microwave frequencies above about 100 MHz, the radiative mode of electromagnetic propagation and dielectric losses in tissue predominate over conduction current losses (Figure 1). Under these conditions, heating results primarily from friction caused by mechanical interactions between adjacent polar water molecules, which oscillate in the time varying electric field. For both radiofrequency (RF) and microwave (MW) radiation, absorbed power density decreases exponentially with depth in tissue as shown in Figure 2a. In order to select the optimum frequency of EM field for depositing energy in a tumor, the critical factors are tumor size and depth below the tissue surface relative to EM wavelength, and proximity to adjacent critical normal tissue structures. A convenient compilation of parameters such as electrical conductivity, dielectric constant, and the wavelengths and penetration depths of electromagnetic waves in air, muscle and fat tissues is given in Fessenden and Hand (1995). For the practical range of frequencies used in hyperthermia from 1-1000 MHz, the wavelengths in soft tissue vary from about 4.5 cm at 1000 MHz up to 2 m at the lower RF frequencies. The maximum spatial resolution of power deposition (focal spot size) is approximately one half this wavelength. As applicator size decreases relative to wavelength, geometric divergence of the wavefront as it exits the small aperture increasingly restricts penetration, as seen in Figure 2b (Turner and Kumar 1982; Hand 1990). The effective heating depth may decrease further as a result of power deposition peaks in the spatially complex antenna near field, and heterogeneities of tissue properties which increase reflection and refraction perturbations of the EM field at tissue interfaces. These later problems may be accommodated to some extent by appropriate applicator design and methods of coupling energy from the applicator into the tissue volume. This has been the subject of significant research over the past two decades and has produced a variable degree of success at improving the localization and penetration of EM energy in the body, as described in the following sections. Clearly, the upper microwave frequencies may be expected to provide localized heating of skin and surface tissues while the lower RF frequencies will heat larger and deeper regions of the body.
Figure 1. Electromagnetic scale showing fixed relationship between frequency, free space wavelength, and energy level of electromagnetic wave. Wavelength in tissue is lo /(er )1/2.
52–Thermal Therapy for Cancer
(a)
907
(b)
Figure 2. (a) 1/e penetration depth of plane wave electromagnetic energy into human soft tissue. b) 1/e penetration depth of 434 MHz radiation in soft tissue from various size waveguide apertures compared to plane wave, normalized to 100% at 5 mm depth in tissue.
Ultrasound (US) Power Deposition Ultrasound energy propagates through tissue as a traveling pressure wave. Variations in pressure from alternating compressive and expansive forces produce an oscillating physical displacement of tissue molecules in the direction of wave propagation. Collisions between vibrating molecules produce heat from mechanical losses, which attenuate energy from the traveling wave. Similar to EM radiation, ultrasound intensity decreases exponentially with distance into tissue, with a tradeoff between effective localization of power superficially at the higher frequencies and deeper penetration due to decreased attenuation at the lower frequencies. Because the speed of sound (cs) is orders of magnitude lower than the speed of light, the wavelength of ultrasound in tissue over the frequency range of interest (f = 0.5-10 MHz) is between λ = 0.1 and 3 mm. With wavelengths much shorter than the dimensions of both tumors and applicators, dispersion of the beam is minimal and well-collimated beams may be directed into very small volumes, unlike much longer EM waves. Because of the combination of short wavelength and lower attenuation, US sources can be used to penetrate deep in the body while still focusing into small tumors (Figure 3). Tabulations of the acoustical properties of mammalian tissues and more detailed coverage of the physical interactions of ultrasound with tissue are available in a number of excellent reviews (Hunt 1990; Fessenden and Hand 1995; Hynynen 1995). The primary parameters affecting the coupling of ultrasound energy into a given tumor volume at depth are the ultrasound source frequency, beam geometry, and tissue properties. In clinical practice, the tumor volume almost always appears in the transducer near field, where there are spatially close fluctuations from minimum to maximum beam intensity. Even with their spatially complex fields, ultrasound frequencies between 1.0 and 3.5 MHz are generally used for external ultrasound hyperthermia applications due to their penetration depth characteristics. Although intensity peaks and nulls are only a few mm apart, secondary mechanisms such as reflections, scattering, and thermal conduction within the tissue as well as perfusion related convection effects smooth the resulting temperature distributions. For unfocused applicators with no focal gain, effective heating may be accomplished to a maximum depth of approximately 3-6 cm, usually using frequencies around 3.5 MHz (Samulski et al., 1990). Use of lower frequencies with non-focused US applicators to heat larger and deeper tissue volumes generally produces increased pain due to unavoidable absorption in underlying bone. Deeper penetration is possible by focus-
908
Jean Pouliot et al.
Figure 3. Penetration of planewave ultrasound in human soft tissue.
ing the US energy from a large surface area into a small focal region at depth and using geometric gain and directional control to minimize irradiation of all bone and air interfaces. Since the wavelengths are small, it is possible to build phased arrays with an overall small cross-section that can be electronically scanned or illuminated to develop complex beams. Treatment of larger tumors may be accomplished by mechanically or electrically scanning a small focal volume around at depth to spread out the time averaged power deposition. Another major factor affecting the coupling of ultrasound into tissue is the presence of reflecting interfaces between applicator and tumor. Minimizing perturbation of the US beam begins with appropriate preparation of the tissue surface to eliminate all air gaps or bubbles. Equally important is a continuous path of temperature-controlled non-compressible fluid media (e.g., degassed water, US gel) between applicator and skin. Appropriate bolusing techniques are described in the literature (Samulski et al., 1990; Diederich et al., 1994). Fortunately, the acoustic impedances of most soft tissues are similar so that once inside the skin there is little reflective loss between tissue layers. The impedances of bone, air, and lung are considerably different than those of soft tissues however, causing significant reflection and refraction of ultrasound energy at these interfaces. Almost complete reflection at soft tissue-gas interfaces and rapid absorption of the transmitted portion of the wave at soft tissue-bone interfaces often cause the most difficulty in using ultrasound clinically, especially for unfocused beams aimed into tissue overlying bone or air.
Methods of Heating Tissue Over the past three decades, hundreds of RF, microwave and ultrasound devices have been investigated for improving the spatial and temporal control of power deposition in human tissue. Many devices have been developed at independent university research laboratories and are not yet available to the hyperthermia community. Other devices have been commercialized but are still limited in availability to patient clinics. While ongoing development efforts continue to produce net gains in applicator functionality, there has already been significant improvement in the hardware and software available for clinical use today as compared to the equipment used in early clinical trials of the 1980s and 1990s. Design principles and general performance capabilities of most of the available technologies have been reviewed previously (Stauffer 1990; Handl-Zeller 1992; Hynynen 1995; Lee 1995; Seegenschmiedt et al. 1995; Stauffer et al. 1995; Seegenschmiedt et al. 1996; Diederich and Hynynen 1999; Stauffer 2000). Given the dramatic
52–Thermal Therapy for Cancer
909
changes in heating technology over the years, we will briefly review the underlying technology, illustrate with examples of representative heating systems in current clinical use, and end with a look towards evolving technology for the clinic of tomorrow.
External Electromagnetic Superficial Heating Technology The most basic EM applicator used for heating superficial tissue is the microwave waveguide with single linearly polarized monopole feed, as shown in Figure 4a (Guy 1971; Sandhu 1982; Lee 1995). Aperture size is generally designed with one side at least a half wavelength long in order to radiate a well formed TE10 mode electric field oriented almost entirely tangential to the skin-fat-muscle tissue interfaces to minimize overheating of high resistivity fat. In order to keep waveguide dimensions small to fit tumors on contoured anatomy, the interior is often filled or lined with high dielectric constant material to reduce the effective wavelength in the waveguide structure. The Figure 4b plot of power deposition pattern or specific absorption rate (SAR) from such an applicator demonstrates the central gaussian peak typical of TE10 mode radiation emanating from a metal aperture, with power deposition falling off to less than 50% of its peak before reaching the antenna perimeter laterally, or tissue depths greater than 1-1.5 cm (Chou et al., 1990; straube et al., 1990).
Figure 4. Linearly polarized single aperture waveguide applicators and corresponding SAR pattern 1 cm deep in muscle equivalent medium. Note the 50% SARmax contour covers about 50% of the aperture face.
910
Jean Pouliot et al.
In order to heat larger and deeper disease, the logical approach is to construct multi-aperture arrays. Fenn et al. (Fenn et al. 1999) describe a 915 MHz dual opposed waveguide pair for heating breast with the addition of implanted electric field probes and an adaptive phased array feedback control routine to optimize phase focusing of heat at depth in breast. Unfortunately as seen in Figure 2, the attenuation of 915 MHz radiation in soft tissue is such that only a minor gain in penetration depth is possible even with aggressive phase control of multiple apertures (Gross et al. 1990; Fenn et al. 1993). Hand et al (Hand et al. 1986) investigate the theoretical improvement in lateral uniformity of SAR from a 4 × 4 array of 4 cm square (waveguide type) tangential electric field apertures as compared to a single 19 cm square aperture. Diederich and Stauffer (1993) evaluated the first commercial device of this type, a 915 MHz 16 element planar waveguide array shown in Figure 5. They found this Microtherm 1000 (Labthermics Technologies Corp., Champaign IL) applicator suitable for treating superficial tissue regions up to 13 × 13 × 1.5 cm. Stauffer et al. (1994a, 1994b) reported a tight distribution of T90 and T50 values (40.9 and 42.6 °C respectively) averaged over 60 min treatment intervals in the first 27 patients from stationary sensors at 961 intratumoral points 0.5-1.5 cm deep, and 41.3 and 42.7 °C, respectively, for 4039 additional points mapped across the tumor target surface. The authors concluded that this water bag coupled planar array provides 50% iso-SAR coverage extending almost to the outer perimeter of the array for uniform heating of up to 169 cm2 areas, while offering impressive adaptability of heating patterns to accommodate irregular tumor shape (Diederich and Stauffer 1993). Another array heating approach makes use of inductive loop coupled Current Sheet Applicators (CSA) which are smaller (7.3 × 5.9 × 3.3 cm) and lighter in weight than typical waveguide applicators and can be connected together in hinged flexible arrays for contoured surfaces (Bach Anderson et al. 1984; Johnson et al. 1987; Gopal et al. 1992; Gopal and Cetas 1993; Prior et al. 1995). Initial clinical results using 433 MHz four element CSA arrays for superficial chestwall disease have demonstrated more uniform and higher overall temperature distributions than possible with earlier devices, with Tmin and Tave tumor temperatures of 41.0 ± 1.5°C and 42.2 ±1.4°C, respectively obtained in the first patient series (Hand et al. 1993; Leigh et al. 1994). Electromagnetic horn applicators are a close variant of the waveguide, with tapered openings to control the divergence of the radiated field. In general, horns provide somewhat larger effective field size than equivalent size waveguides, with 50% SAR contours often extending to the aperture boundary in one direction. As with other applicators, the difficult challenge has been to obtain sufficient SAR between adjacent horns of an array. Van Rhoon et al. (Rietveld et al., 1998; Van Roon et al., 1998) describe a novel flared horn with the two sides parallel to the electric field replaced with low εr Lucite to expand the SAR
Figure 5. 16 element microwave waveguide planar array applicator used for heating superficial disease <1.5 cm deep, shown treating multiple field diffuse chestwall disease spreading across the side and back of patient.
52–Thermal Therapy for Cancer
911
distribution in the H-plane. In addition, a cone shaped wedge of low εr PVC material is mounted in the center of the waveguide to further reduce SAR centrally. The heating effectiveness of up to 6 element arrays of Lucite Cone Applicators (LCA) has been compared to that of conventional horn arrays in matched clinical treatments (Rietveld et al. 1999). The data demonstrate superior uniformity of heating from the new applicators with invasively measured temperatures averaging 0.28°C higher overall and 0.43°C higher peripherally (Figure 6). Due to improved heating around the applicator periphery, the LCA applicators are currently in routine use in Rotterdam in up to 3 × 2 arrays, treating surface areas up to 600 cm2. Printed circuit board (PCB) based microstrip antenna technology is receiving increasing attention in recent years due to the ability to form almost arbitrarily large arrays from relatively low cost, lightweight and flexible PCB material. Several investigators have reported theoretical power deposition patterns possible with microstrip patches (Underwood et al. 1992), slot apertures (Montecchia 1992), and various configuration spiral microstrip antennas (Lee et al. 1992; Ryan et al. 1995). Even with substantial coupling bolus thickness, these microstrip structures radiate complex E fields with a significant normal field component that tends to reduce the penetration depth of effective heating. Thus microstrip applicators appear best suited to tumors that extend up to and include the tissue surface rather than those located beneath a layer of high resistivity fat. The largest single aperture microstrip applicator reported in the literature measures about 20 × 30 cm and can produce an effective field size (≥ 50% SARmax) of roughly 12.5 × 24 cm when driven at 434 MHz (Lamaitre et al. 1996). These Contact Flexible Microstrip Applicators (CFMA) are available in several different sizes that can be used at frequencies ranging from 40 MHz to 915 MHz. The applicators have been shown to produce large effective field sizes up to 400 cm2 with relatively uniform SAR patterns under the aperture perimeter in homogenous tissue phantoms (Gelvich et al. 1996; Lamaitre et al. 1996). The applicators have been used in clinical studies outside the US, though clinical thermal dosimetry and efficacy results have not been reported yet. While the applicators provide a large effective heating area, concerns remain about the lack of adjustability of heating under a single feed applicator when placed over heterogeneous tissue. Thus the development of CFMA devices has continued to multi-aperture arrays, which should become a useful clinical device for large area shallow depth heating in the near future. Stauffer et al. (1994a,b) described the concept of a Conformal Microwave Array (CMA) applicator consisting of an array of square radiating apertures etched from a single layer of flexible copper foil and driven non-coherently at 915 MHz. Subsequently, radiation patterns from the square annular slot Dual Concentric Conductor (DCC) apertures were analyzed theoretically with Finite Difference Time Domain (FDTD) simulations for a variety of aperture sizes and design configurations, and the simulations verified with measurements of SAR in muscle equivalent phantoms (Stauffer et al., 1995, 1998; Rossetto et
Figure 6. 10 × 10 cm flared horn and Lucite cone applicators, with corresponding SAR patterns that demonstrate superior coverage of the LCA aperture face with >50% SARmax. Courtesy of Rietveld and Van Rhoon (Van Rhoon et al. 1998; Rietveld et al. 1999).
912
Jean Pouliot et al.
al., 1998). In an effort to predict SAR patterns from large conformal arrays in heterogeneous non-perfect loads, Rossetto and Stauffer (1999) characterized SAR patterns for DCC arrays placed over variable thickness water bolus (0.25-0.75 cm thickness), over rib sized bones lying 0.5 or 1 cm deep in muscle beneath 0 or 0.5 cm layers of fat, and for the array heating through a 0.5 cm water bolus with various size pockets of air trapped between the water bolus and skin surface. Several important conclusions were reached which may have significance for other microwave applicators as well. First, neither the presence of ribs close to the skin surface nor variations of up to a factor of three in water bolus thickness produced unacceptable patterns of power deposition from the 915 MHz microwave applicators. In other words, these realistic considerations for clinical use of microwave array applicators did not significantly alter the uniformity of surface heating from the array applicators. However, the presence of even small air bubbles between the microwave radiators and tissue surface was found to alter power deposition locally around the air discontinuity, with the degree of perturbation depending on bubble size and location under the DCC slot aperture. This later conclusion is not unexpected in light of other investigators who have shown as high as 20% perturbation from low er temperature probe catheters in microwave fields (Chan et al., 1988). Although thermal conduction and convection effects in tissue will tend to smooth out the temperature distribution which results from a perturbed SAR pattern, this study re-emphasizes the importance of minimizing air in the path of high dielectric (water) coupled microwave apertures regardless of design. Figure 7 shows the back side microstrip feedline network of a 4 × 5 aperture CMA applicator placed over a 6 mm thick water bolus through which temperature controlled degassed deionized water is circulated to maintain control of skin surface temperature. Elastic and Velcro straps are used to produce a snug but comfortable fit of the microwave array and bolus to the upper torso, and an elastic overgarment may be worn to secure the applicator in place during patient movement. In one configuration, the patient is free to stand and move around the room close to the microwave power equipment during hyperthermia treatment. Preliminary clinical evaluations of the CMA applicators such as those seen in Figure 7 have reported improved patient comfort and effective heating of surface disease to a depth of ≥10 mm (Stauffer 2000; Stauffer et al., 2001). Following the treatment protocol of most hyperthermia applicators to date, the CMA applicator has generally been used to heat tissue as soon as possible before or after the radiotherapy treatment it is intended to enhance. Due to the time required for the patient to move between the linear accelerator and hyperthermia room and the time to set up the applicator on the patient, heat and
Figure 7. CMA applicators placed over chestwall disease and secured in place with elastic fabric. Note the white coaxial cable feedlines to microwave connectors at the edge of the PCB, and fiber optic sensors that can be scanned in an out of catheters which cross the tissue surface under the array.
52–Thermal Therapy for Cancer
913
radiation treatments are usually separated by at least 30-60 min. As shown by Overgaard (1989), significantly higher thermal enhancement of the radiation dose may be possible by combining heat and radiation doses simultaneously for maximum synergism. In order to facilitate simultaneous treatments, a “Combination Applicator” is currently under development (Stauffer et al., 2005, 2005; Taschereau et al., 2004), with an early prototype pictured in Figure 8. Like the CMA heat applicator, the basic structure diagrammed in Figure 9 includes a thin temperature regulated water bolus chamber to couple microwaves from a flexible printed circuit board (PCB) microwave antenna array into tissue and to provide surface cooling for maintaining skin temperatures uniformly below 45 °C. A porous foam layer is placed inside the bolus chamber to maintain appropriate thickness of the water layer as required for good power deposition patterns (Rossetto and Stauffer 1999; Rossetto et al., 2000; Neuman et al, 2002) and to provide good conformity of the bolus to slightly irregular skin surfaces. An array of brachytherapy catheters spaced 1 cm apart is included on the back of the applicator to provide uniform irradiation of the tissue surface by a scanning HDR source. These catheters are integrated into the CMA assembly on the back side of the PCB antenna array behind a second foam spacer that ensures optimal spacing to the skin for most uniform radiation dose considerations. Thermal mapping catheters on the bolus front surface accommodate scanning fiber-optic probes for monitoring the temperature distribution of the skin surface during the simultaneously applied heat and brachytherapy. Figure 10 shows a “Rando” phantom model of the human torso with 15 × 15 × 1 cm chestwall recurrence target covered by a conformal Combination Applicator and partially inflated air bladder with elastic overgarment to ensure conformity of the applicator to the complex anatomic contour. Existing HDR treatment planning software (IPSA) with slightly modified dose optimization constraints to fit a “planar” surface rather than the interstitial implant approach more commonly used with HDR equipment was used
Figure 8. “L-shaped” Combination Applicator on human torso model.
914
Jean Pouliot et al.
Figure 9. Exploded view of Combination Applicator showing catheters for brachytherapy and temperature monitoring, and separate layers for the circulating temperature controlled water bolus, flexible PCB microwave antenna array, dielectric spacer for optimizing brachytherapy dose distribution, and outer air bladder.
Figure 10. Model of human torso with 15×15×1 cm chestwall recurrence target covered by conformal Combination Applicator, and corresponding radiation dose distribution superimposed on the pre-treatment CT scan as calculated by IPSA, showing excellent coverage of the 1 cm deep tissue target with prescribed 100 cGy dose.
to preplan source dwell times for most uniform coverage of the target volume with the prescribed dose. Based on the treatment planning CT scans of the torso model, Figure 10 also shows the IPSA program generated 3-D surface rendering of the treatment setup with all catheter and possible source locations over the 15 × 15 cm target region, and the corresponding calculated radiation dose distribution. Note there is good conformity of the applicator to the complex curvature of the torso model, and good conformity of the 100 cGy prescribed dose to the entire 1 cm deep “tumor” target (in purple).
External Superficial Ultrasound Heating Equipment The simplest approach for superficial ultrasound heating is a single piston style transducer operating at a frequency in the range of 1 to 3.5 MHz where the effective heating depth of an unfocused ultrasound beam is about 2 to 6 cm. Figure 11 shows a diagram and early clinical treatment configuration of a piston style transducer. The ultrasound must include a short column of degassed water to transmit the pressure wave into tissue The water is temperature regulated to cool the skin surface, since power deposition is maximum at the skin surface (Corry and Barlogie 1982). Due to lack of control over heating patterns from single
52–Thermal Therapy for Cancer
915
Figure 11. Simple piston type ultrasound applicator with crystal plate that vibrates sending pressure waves through temperature controlled degassed water and into the tissue. This outdated approach shown for illustration.
transducer systems, most clinical facilities have replaced these simple devices with more controllable multiple transducer array applicators. A considerably more adjustable heating device is the Sonotherm 1000 (Labthermics Technologies Inc., Champaign IL) ultrasound planar array applicator which has a 4×4 array of 3.6 cm square transducers mounted in a 15 × 15 cm housing that couples to tissue through an extendable floppy bolus of temperature regulated degassed water. Because of its well collimated beam profile, this non-focused array has been used successfully for precisely directing power deposition into superficial tumors up to 4 to 6 cm deep in the neck, thorax, and extremities (Samulski et al., 1990). Figure 12 shows a typical clinical setup for the Sonotherm applicator and a measured intensity beam plot in homogeneous water. Note the extremely well collimated beam produced by the 12 outer transducers, with no power deposited centrally for this case of the center 4 transducers turned off. To take advantage of the precise focusing and deep penetration capabilities of ultrasound pressure waves in tissue, other ultrasound treatment approaches have been investigated that use more sophisticated transducers and especially arrays of focused transducers aimed to a common focal spot at depth. Early ultrasound array systems used a small number of low frequency transducers (e.g., 4-8 transducers at 1 MHz) producing a small focal spot which was mechanically scanned rapidly around the tumor to paint a larger temporally averaged heating pattern (Hynynen et al., 1987, 1990). In principle, when scanning a small high-intensity focal region, the ratio of heating energy deposited near the focal spot compared to that in surface tissues within the entrance portal is diminished directly in proportion to the desired treatment volume. For a given surface window, increasing the scan area to treat larger tumors at depth requires correspondingly higher SAR in surface tissues to maintain adequate power delivery at depth, providing a practical upper limit on treatment volume. In practice, this has not been a serious limitation as Scanned Focused Ultrasound (SFUS) has been used successfully in the clinic for a variety of superficial and moderate depth tumors including those in the breast, lower abdomen, pelvis, and head and neck tumors including brain (via craniectomy). Continued development, including the addition of higher frequency transducers (4 MHz) with reduced penetration and a patient pain feedback button, have reduced patient pain during treatment and extended the usefulness of this scanning ultrasound technique to skin and larger targets up to 20 × 20 × 3-4 cm depth. Improved spatial control of power deposition has been reported due to computer control of operating frequency and applied power as a function of scan position, taking into account tissue geometry and patient pain. These software enhancements improved the average minimum
916
Jean Pouliot et al.
Figure 12. Degassed water coupled 16-transducer planar array ultrasound applicator placed on recurrent breast disease. Acoustic pressure squared pattern (proportional to SAR) in degassed water obtained by turning off the center 4 transducers and leaving on the peripheral 12 transducers to demonstrate directional control of power deposition. Note the intensity peaks, related to acoustic wavelength, which are smoothed by thermal conduction in tissue to produce a more uniform temperature distribution.
and maximum tumor temperatures achieved with this SFUS system in the clinic to 41.1±1.1°C and 45.9±2.7°C (Anhalt et al., 1993). Another novel approach combines the positive attributes of a stationary transducer array with the potential of improved heating uniformity from computer controlled real-time adjusted power deposition. Moros et al. (1994, 1995) describe a Scanning Ultrasound Reflector Linear Array System (SURLAS) for heating superficial tissue targets by moving a triangular or wedge shaped acoustic reflector back and forth across the target tissue within a thin water filled rectangular bolus structure that is sealed to the body surface. Heating is provided by a 1 × 4 element planar array of ultrasound transducers at one or both ends of the scan box aimed across the target surface toward the moving reflector. A current version includes a high frequency (4.9 MHz) transducer array at one end for heating shallow depth tissue and a matching geometry lower frequency (1.9 MHz) array at the opposite end for heating tumors that extend more deeply below the surface (Moros et al., 1999). The diagram of Figure 13 shows the general concept of ultrasound energy directed laterally across the tumor surface within a thin water bolus layer and then redirected down into the tumor by the scanning reflector. By dynamic computer control of both the power and percentage of high and low frequency insonication as a function of reflector position over the target, the penetration depth of effective heating can be varied from superficial to deep during the scan to accommodate tumor dimensions and heterogeneity (Moros et al., 1999; Novak et al., 2005). Currently the system can heat up to about 200 cm2 areas of superficial tissue when the bolus compartment is coupled to gently contoured anatomy. A particularly notable feature of this design is that the water bolus and scanning reflector are designed from non-radiation perturbing materials so that external beam radiation may be delivered simultaneously with the ultrasound heating energy through the same treatment window for maximum synergism of treatments.
External Deep-Regional Heating Technology In order to treat deeper in the human body, a number of tradeoffs are necessary regarding the size, depth, and localization of the target volume. The most common heating techniques involve the use of radiated or conducted electromagnetic fields at frequencies between 10-100 MHz where wavelengths are long
52–Thermal Therapy for Cancer
917
Figure 13. Scanning ultrasound reflector linear array system (SURLAS) using either one or two different frequency linear arrays of transducers directed laterally across the water bolus to a scanning reflector that redirects beam energy into tissue. The system modulates frequency and power during the scan to match the heating pattern to tumor depth. Bolus and scanning reflector construction provides a radiation translucent window for unimpeded simultaneous irradiation from linear accelerator. With permission from Eduardo Moros (Washington University).
compared with tumor and body dimensions (as seen from Figures 1 and 2a). This longer wavelength radiation generally produces heat in a relatively large region of the body that includes the tumor and relies on different tumor characteristics (e.g., low perfusion, high electrical conductivity) to produce localized tumor heating rather than precisely focused power deposition. Perhaps the most straightforward approach to depositing energy within the entire body cross section is the radiofrequency capacitive heating device which has seen extensive clinical application in Asia. The Thermotron RF-8 is one commercial implementation of this approach, with two opposing arms that support variable size saline bolus-coupled metal plate electrodes in direct contact with and on opposite sides of the patient (Figure 14). With this configuration, conduction current at 8 MHz is coupled capacitively through the electrode bolus and into the patient. With the resistivity of fat tissue approximately 60 times higher than underlying muscle in the same current path between opposing plates, resistive heating is highest in the fat layer near the surface, which should ideally be thin compared to the tumor target. Adjustment of the heating pattern is accomplished by changing the size and position of electrodes relative to the tumor, with a smaller diameter electrode on the tumor side to concentrate the current. For patients with a sufficiently thin fat layer overlying the tumor, current that has been coupled through the fat may preferentially flow through low resistivity tumor that is intermingled and lies in parallel within higher resistance normal tissues at depth. The system is intended to counter preferential heating of superficial fat with aggressive skin cooling using bolus temperatures as low as 10°C (van Rhoon et al., 1992). T50 tumor temperatures ranging from 41.6 to 42.3 have been reported for a variety of sites in several multi-institution clinical studies in Asia (Kaheki et al., 1990; Hiraoka et al. 1995). Another EM device that can deposit energy deep in the body with just a single external power source is the Coaxial TEM applicator (De Leeuw et al., 1991). This device consists of a coaxial cable like structure that is large enough to place the entire patient inside a hollow 60 cm diameter “inner conductor” chamber that is filled with coupling water. With only a single 70 MHz power generator to produce an axially directed electric field, partial steering of the SAR around the body cross section is possible by shifting patient position within the 60 cm aperture cross section. Initial clinical investigation of the device has demonstrated success at producing a clinically useful moderate temperature rise in deep pelvic tumor masses with good patient tolerance and acceptable toxicity. For the first 64 patients, the reported average
918
Jean Pouliot et al.
Figure 14. RF capacitive heating device used to generate electric current between two saline coupled metal plate electrodes of different size.
T90 and T50 tumor temperatures were 39.9+1.0°C and 40.7+1.0°C with an accompanying systemic temperature rise of 1.9 +0.7°C (Van Es et al., 1995). Since it is desirable to both penetrate deep in the body and restrict heating to specific tumor containing regions at depth, there have been a number of electromagnetic radiating array applicators developed which are designed to steer energy deposition within the body. One such device is the four waveguide array or Matched Phased Array (MPA) system which allows custom positioning of waveguide sources on the patient surface accompanied by phase and amplitude adjustments of all four sources to steer power deposition at depth. This device has demonstrated the ability to generate low to moderate temperatures (40-41°C) in deep tumors in clinical studies since 1987 (Gonzalez Gonzalez et al., 1992). Other developments have investigated novel approaches that superimpose EM waves from two more inductively or capacitively coupled apertures to focus and steer energy at depth (Franconi et al., 1991, 1993; Kato et al. 1991, 1995; Fujita et al., 1993) but clinical use of these devices has not been reported yet. Probably the most widespread and successful deep regional heating device to date is based on the 60120 MHz dipole phased array technology developed by BSD Medical Corp. (Salt Lake City, UT) and distributed worldwide in several developmental stages. The original deep heating Annular Phased Array System (APAS) had 4 amplifiers driving 8 dipole antennas positioned in a fixed pattern around the patient circumference and coupled with deionized water bolus (Turner 1984). This first design was tested in a large clinical trial in the 1980s and found to be lacking in adjustability since it did not include software control of relative power and phase for the four amplifiers (Sapozink et al., 1988). Subsequent development produced the Sigma-60 applicator operating over the same frequency range but providing increased flexibility of control from four independent phase and amplitude controls for the 8 dipoles as well as a more patient friendly interface, as seen in Figure 15. Improved localization has been reported for this device as well as clinical utility in a number of deep tissue sites (Saulski et al., 1987; Myerson et al., 1991), though some investigators were still frustrated with the resolution of power steering for other sites (Anscher et al., 1992). The company has recently extended its product line with a new series of Sigma-Eye applicators, which provide axial as well as lateral steering of power deposition with three rings of 8 dipole antennas each (Figure 15). This later configuration has sufficient adjustability that the company markets an MRI-compatible heat applicator system along with an MRI magnet that facilitates pre-treatment planning scans of the patient in the intended treatment configuration. Moreover, associated software is available to non-invasively monitor deep tissue temperature and physiologic changes during heat treatment (Gellermann et al. 2000, 2005; Wust et al., 2004). This system is currently in clinical use with ongoing optimization of techniques to integrate the capabilities of the two complimentary systems.
52–Thermal Therapy for Cancer
919
Figure 15. Deep heating electromagnetic system with patient in position for heating deep-seated tumors in lower abdomen or pelvis. (Photos with permission from BSD Medical Corp.) An MRI compatible version of the system may be used with an MR unit for pre-treatment planning and real-time non-invasive temperature monitoring.
To take advantage of the improved adjustability of an RF deep heating system with 12 independently controllable antennas, current systems include more sophisticated 3-D treatment planning capabilities. Patient and tumor-specific treatment planning is time consuming and not necessarily required for significant improvement in steering power deposition into a tumor target at depth. Currently, treatment planning is most often accomplished starting from a library of standard patient models that has already been prepared from CT scans for use in pre-treatment planning of “matching patients”. Tissue dielectric and blood perfusion properties have been identified for all tetrahedral cells in the standard models to enable rapid calculation of 3-D SAR and temperature patterns for patients whose anatomy matches that of a standard model. Figure 16 demonstrates three steps in pre-treatment planning possible with the commercial Hyperplan software included with the BSD Medical Corp (Salt Lake City, UT) phased array radiofrequency heating system.
Interstitial and Intracavitary Deep Local Heating Technology For improved localization of heat in smaller target volumes at depth, hyperthermia may be applied using interstitial heat sources implanted directly within the tumor or by placing heat sources within natural body cavities adjacent to the tumor tissue. Although invasive, these approaches provide a higher degree of localization than is possible with externally applied systems since heat energy is disbursed from directly within or immediately around the target tissue. Typically, interstitial heat is applied in conjunction with interstitial brachytherapy, utilizing the same implanted needles or catheters. Complete reviews of the numerous technologies used for invasive heating are described elsewhere (Handl-Zeller 1992; Seegenschmiedt and Sauer 1993; Seegenschmiedt et al. 1995; Stauffer et al. 1995). Additional examples of recent invasive heating approaches have recently been highlighted in a special issue of the International Journal of Hyperthermia covering Thermal Ablation Therapy applications (Stauffer and Goldberg 2004). Three recent advances in interstitial heating technology, which provide improved means for delivering deep local therapy with better spatial and temporal control of the temperature distributions than has been possible in the past, are reviewed as examples below. Multi-electrode RF capacitively coupled devices consist of linear arrays of small tubular electrodes that are implanted in tissue inside one or more plastic brachytherapy catheters. RF (27 MHz) current flow between electrodes is coupled capacitively through the plastic catheter wall to deposit energy in the
920
Jean Pouliot et al.
Figure16. HyperPlan model of body phantom allows pre-treatment planning of phase and amplitude shifts required to steer power deposition peaks into a tumor. Steps in the pre-planning show the tetrahedral standard model with tumor location in pink, the HyperPlan calculated SAR pattern, and calculated 41ºC temperature contour superimposed on underlying structure within a human pelvis. (With permission from BSD Medical Corp.)
surrounding tissue by resistive losses (Crezee et al., 1999). The relative RF current from each electrode of an array can be adjusted to custom tailor the heating pattern to the heterogeneous tumor, and can be varied during treatment to accommodate dynamically changing blood perfusion cooling. Very sophisticated thermal treatment planning systems have been developed in support of this technology, which include accurate representation of thermally significant vasculature (De Bree et al., 1998). These devices have been used for highly adjustable hyperthermia therapy in sites such as brain (Hulshof et al., 2004) and prostate (van Vulpen et al., 2002). Although offering somewhat more limited 3-D spatial control of the heating distribution, a multi-needle RF Localized Current Field (LCF) system has evolved from earlier implementations (Doss and McCabe 1976; Astrahan and Norman 1982) into a computer controlled system capable of long duration low temperature heating protocols (up to 72 hours, <41°C) (Martinez et al, 1993). The heating current to each needle is controlled separately in response to temperature sensors within the needles, and can be controlled remotely from outside the treatment suite using standard computer networking (Figure 17). The enhanced computer control features make operation of long duration treatments more manageable for the operator and patient. With this system, the electrical current (typically 0.5-1.0 MHz similar to electrosurgical instruments) distributes along the length of exposed metal needle and traverses the intervening tissue between active electrodes. The length of the heating tip can be modified by placing insulating sleeves over portions of the needles, particularly to minimize current near the needle entrance sites through normal skin. Currently active electrodes are selected under computer control to vary the current pathways to produce most uniform time-averaged heating within the implant volume. For both RF techniques mentioned here, the penetration of thermal energy falls off steeply (1/r2) with distance from each source, necessitating multiple heating electrodes be placed no more than about 1.25 cm apart for sufficiently uniform heating within the narrow therapeutic window of low to moderate temperature hyperthermia (Strohbehn 1987; Hand et al., 1991; Stauffer et al., 1995). Wider spacing of electrodes may be possible in thermal ablation applications where a larger therapeutic window of acceptable temperatures is more forgiving of temperature variability within the target volume (Haemmerich and Lee 2005; Haemmerich et al., 2005). Over the past three decades, extensive effort has been devoted to the development of implantable microwave antennas from miniature coaxial cable using different design strategies such as dipoles, helical coils, and modified sleeve-, choke-, and helical-dipole configurations to produce variable heating lengths in tissue. Most often the antennas are driven at 915 MHz, though 433 MHz and 2.45 GHz anten-
52–Thermal Therapy for Cancer
921
Figure 17. Multi-needle LCF hyperthermia system designed for long duration, mild temperature hyperthermia combined with brachytherapy. Advances include quick RF connections to multiple needle implant templates, real-time temperature control, and remote treatment monitoring/control capabilities. (Photos courtesy of Peter Corry, William Beaumont Hospital).
nas are also common for some applications. The radial penetration of power deposition around microwave antennas is greater than from RF electrodes. Depending on the specific antenna and tissue load configuration, attenuated energy falls off as 1/rn, where n ranges from 1 to 3 for practical heating configurations (Zhu et al., 1998). Thus microwave antennas are typically placed 1.25 to 2 cm apart for hyperthermia applications (Strohbehn 1987; Stauffer et al., 1995). Dipole antennas have the greatest radial penetration near at the axial center of the radiating antenna, but the heating pattern is gaussian shaped rather than uniform along the dipole length and not ideal for many applications. Helical coil antennas can produce more uniform power deposition along the coil at the distal tip of antenna (~1.5 to 4 cm), but somewhat less penetration. Miniature (~1 mm diameter) helical coil antennas of various heating lengths are pre-selected to fit the heating pattern to the tumor dimensions and have been used in many clinical hyperthermia studies (Seegenschniedt et al., 1996), including sites such as brain (Sneed et al. 1998) and prostate (Sherar et al., 2003, 2004). An example of a commercial system is shown in Figure 18, using helical coil microwave antennas that are compatible with HDR brachytherapy catheters. In recent years, a number of transurethral thermotherapy devices have been developed to treat prostate BPH with helical coil or helical-dipole microwave antennas. When used for thermal ablation with a large therapeutic temperature window, these single antenna devices exhibit heating lengths up to 4 cm and ~15 mm or more effective radial penetration of heating (Larson 2002). Compared to other interstitial and intracavitary techniques, ultrasound technology offers the highest degree of spatial and temporal control over 3-D power deposition distributions during hyperthermia and thermal ablation therapy treatments (Diederich and Hynynen 1990). Endocavitary ultrasound applicators constructed from linear arrays of sectioned tubular ultrasound transducers (Diederich and Hynynen 1990) have been implemented in the clinic for transrectal heating of prostate cancer (Fosmire et al., 1993; Hurwitz et al., 2001). An expandable balloon with circulating temperature regulated cooling water couples the heating power from semi-cylindrical transducers into tissue and helps to protect the rectum from thermal damage. Applicators are available now with sectored transducer elements and individually controllable output power (at ~1.6 MHz) to provide both axial and angular control of heating. Effective penetration from the large diameter rectal applicators is 2-3 cm from the rectal wall, extending treatment throughout the prostate gland and seminal vesicles. Alternatively, transrectal high-intensity focused ultrasound devices have also been used for local thermal ablation of prostate cancer and BPH, relying on
922
Jean Pouliot et al.
Figure 18. Interstitial microwave antennas suitable for hyperthermia in combination with interstitial brachytherapy (with permission from BSD Medical Corp.). Associated treatment planning program screen shot shows calculated SAR contours for a multi-catheter implant and control parameters of power and phase for multiple antenna array.
mechanical steering of a small high intensity focal region to multiple overlapping positions within the target volume. Interstitial ultrasound applicators utilize arrays of small tubular ultrasound radiators, designed to be inserted within 13-14 gauge plastic implant catheters which are also used for interstitial HDR brachytherapy (Diederich 1996; Nau et al., 2001). Water-flow is used during power application to couple the ultrasound into tissue and improve thermal penetration by cooling the catheter-tissue interface. The applicators are fabricated with multiple tubular segments and separate power connections, so that the power deposition pattern can be adjusted in real time along the applicator axis. Ultrasound energy emanating from each transducer section is highly collimated within the borders of each segment so that the axial length of the therapeutic temperature zone remains well defined by the number of active elements over a large range of applied power levels and treatment durations (Diederich et al. 1996; Nau et al., 2000; Deardorff et al. 2001). Using ultrasound applicators with a variable number of radiating sections, the length of heating within each catheter is easily tailored to fit the clinical target using the same catheter implant required for HDR brachytherapy. Unlike RF microwave and thermal conduction type interstitial sources, the angular heating pattern of tubular ultrasound radiators can be modified by sectoring the cylindrical transducer surface. In this fashion, active zones can be selected (i.e., 90°, 180°, or 360°) to produce angularly selective heating patterns that significantly expand the flexibility of use of these interstitial applicators (Deardorff and Diederich 1999; Nau et al., 2000). With operator control over the orientation of directional heating, high power levels may be directed into tumor tissue in one direction while protecting critical normal tissues nearby. Alternatively, the directional applicators may be rotated dynamically during treatment with real-time power adjustments to precisely tailor the heating pattern around one or more applicators. Interstitial ultrasound devices have proven especially effective for delivering selective thermal ablation under real-time MR thermal monitoring (Kangasniemi et al., 2002; Nau et al., 2005). For interstitial ultrasound devices with tubular sources, the radial penetration of energy falls off as 1/r. The enhanced radial penetration of ultrasound allows larger applicator separation than microwave antennas, with a preferred separation of 1.5 to 2.0 cm between applicators to maintain therapeutic hyperthermia temperatures. An example of an interstitial ultrasound hyperthermia treatment in conjunction with HDR brachytherapy is shown in Figure 19.
52–Thermal Therapy for Cancer
923
(a) (b)
(c)
Figure 19. (a) Multi-element interstitial ultrasound applicators suitable for heating in combination with HDR brachytherapy; (b) CT image of a brachytherapy implant pattern in the prostate, with directional (210°) ultrasound applicators placed posterior with energy directed anterior to protect the rectum. Temperature sensors were placed in other catheters of the brachytherapy implant array. (c) Measured temperature profiles indicating therapeutic heating along the desired 2 cm active heating zone.
Conclusion There are numerous applications for therapeutic heat and cold which require a large variety of clinical approaches. This review describes the basic thermal therapy approaches in terms of mechanisms of interaction with tissue and the technology available for heating tissue. Basic heating principles are illustrated with examples of performance from devices that have been used for clinical hyperthermia in the past. Typical heating systems currently being used in the clinic are described and contrasted with a look at evolving technology to be available in the near future. A list of references to more extensive reviews as well as primary literature covering these heating systems is included for the interested reader.
References Ahmed, M. and S. N. Goldberg (2002). “Thermal ablation therapy for hepatocellular carcinoma.” J Vasc Interv Radiol 13(9 Pt 2): S231-44. Anhalt, D. P., K. Hynynen, R. B. Roemer, et al. (1993). Scanned ultrasound hyperthermia for treating superficial disease. Hyperthermic oncology 1992, vol. 2. Gerner, E. and T. Cetas. Tucson, Arizona Board of Regents. 2: 191-192.
924
Jean Pouliot et al.
Anscher, M. S., T. V. Samulski, K. A. Leopold, et al. (1992). “Phase i/ii study of external radio frequency phased array hyperthermia and external beam radiotherapy in the treatment of prostate cancer: Technique and results of intraprostatic temperature measurements.” International Journal of Radiation Oncology Biology Physics 24(3): 489-495. Anzai, Y., R. Lufkin, A. DeSalles, et al. (1995). “Preliminary experience with MR-guided thermal ablation of brain tumors.” American Journal of Neuroradiology 16(1): 39-48; discussion 49-52. Armour, E. P. (2004). “Long duration mild temperature hyperthermia and brachytherapy.” Int. J. Hyperthermia 20(2): 175-89. Armour, E. P., Z. Wang, P. M. Corry, et al. ((In press)). “The role of thermotolerance in long duration 41¯c sensitization of rat 9l cells to HDR brachytherapy.” International Journal of Radiation Oncology Biology Physics. Astrahan, M. A. and A. Norman (1982). “A localized current field hyperthermia system for use with 192-iridium interstitial implants.” Med. Phys. 9(3): 419-424. Bach Anderson, J., A. Baurn, K. Harmark, et al. (1984). “A hyperthermia system utilizing a new type of inductive applicator.” IEEE Trans. Biomed. Eng. 31: 21-7. Boudewyns, A. and P. Van De Heyning (2000). “Temperature-controlled radiofrequency tissue volume reduction of the soft palate (somnoplasty) in the treatment of habitual snoring: Results of a European multicenter trial.” Acta Otolaryngol 120(8): 981-5. Bruskewitz, R., M. M. Issa, C. G. Roehrborn, et al. (1998). “A prospective, randomized 1-year clinical trial comparing transurethral needle ablation to transurethral resection of the prostate for the treatment of symptomatic benign prostatic hyperplasia.” J. Urology 159(5): 1588-93; discussion 1593-. Chan, K. W., C. K. Chou, J. A. McDougall, et al. (1988). “Perturbations due to the use of catheters with non-perturbing probes.” Int. J. Hyperthermia 4(6): 699-702. Cheung, A. Y. and A. Neyzari (1984). “Deep local hyperthermia for cancer therapy: External electromagnetic and ultrasound techniques.” Cancer Research (Supplement) 44: 4736s-4744s. Chou, C. K., J. A. McDougall, K. W. Chan, et al. (1990). “Evaluation of captive bolus applicators.” Med. Phys. 17: 705-709. Christensen, D. A. and C. H. Durney (1981). “Hyperthermia production for cancer therapy: A review of fundamentals and methods.” J. Microwave Power 16: 89-105. Corry, P. M. and B. Barlogie (1982). Clinical application of high frequency methods for local hyperthermia. Physical aspects of hyperthermia. Nussbaum, G. H. New York, American Institute of Physics: 307-328. Crezee, J., R. S. Kaatee, J. F. van der Koijk, et al. (1999). “Spatial steering with quadruple electrodes in 27 MHz capacitively coupled interstitial hyperthermia.” Int. J. Hyperthermia 15(2): 145-56. D’Ancona, F. C., E. A. Francisca, W. P. Witjes, et al. (1997). “High energy thermotherapy versus transurethral resection in the treatment of benign prostatic hyperplasia: Results of a prospective randomized study with 1 year of followup.” J. Urology 158(1): 120-5. De Bree, J., J. J. Lagendijk, B. W. Raaymakers, et al. (1998). “Treatment planning of brain implants using vascular information and a new template technique.” IEEE Trans Med Imaging 17(5): 729-36. De Leeuw, A. A., J. Mooibroek and J. J. Lagendijk (1991). “Specific absorption rate steering by patient positioning in the “coaxial tem” system: Phantom investigation.” Int. J. Hyperthermia 7(4): 605-611. Deardorff, D. L. and C. J. Diederich (1999). “Angular directivity of thermal coagulation using air-cooled directcoupled interstitial ultrasound applicators.” Ultrasound Med. Biol. 25(4): 609-22. Deardorff, D. L., C. J. Diederich and W. H. Nau (2001). “Control of interstitial thermal coagulation: Comparative evaluation of microwave and ultrasound applicators.” Med. Phys. 28(1): 104-17. Dewey, W. C. (1994). “Arrhenius relationships from the molecule and cell to the clinic.” Int. J. Hyperthermia 10(4): 457-483. Dewhirst, M. W., E. Jones, T. V. Samulski, et al. (2003). Hyperthermia. Cancer medicine. Kufe, D., R. Pollock, R. Weischelbaumet al. Hamilton BC, Decker. 1: 623-36. Dewhirst, M. W., T. L. Phillips, T. V. Samulski, et al. (1990). “RTOG quality assurance guidelines for clinical trials using hyperthermia.” International Journal of Radiation Oncology Biology Physics 18: 1249-1259. Dewhirst, M. W. and D. A. Sim (1984). “The utility of thermal dose as a predictor of tumor and normal tissue responses to combined radiation and hyperthermia.” Cancer Research (Suppl) 44: 4772s-4780s.
52–Thermal Therapy for Cancer
925
Dewhirst, M. W., D. A. Sim, S. Sapareto, et al. (1984). “Importance of minimum tumor temperature in determining early and long-term responses of spontaneous canine and feline tumors to heat and radiation.” Cancer Res. 44: 43-50. Diederich, C. J. (1996). “Ultrasound applicators with integrated catheter-cooling for interstitial hyperthermia: Theory and preliminary experiments.” Int. J. Hyperthermia 12(2): 279-97. Diederich, C. J. and K. Hynynen (1999). “Ultrasound technology for hyperthermia.” Ultrasound Med. Biol. 25(6): 871-87. Diederich, C. J. and K. H. Hynynen (1990). “The development of intracavitary ultrasonic applicators for hyperthermia: A design and experimental study.” Med. Phys. 17: 626-634. Diederich, C. J., I. S. Khalil, P. R. Stauffer, et al. (1996). “Direct coupled interstitial ultrasound applicators for simultaneous thermobrachytherapy: A feasibility study.” Int. J. Hyperthermia 12(3): 401-19. Diederich, C. J. and P. R. Stauffer (1993). “Pre-clinical evaluation of a microwave planar array applicator for superficial hyperthermia.” Int. J. Hyperthermia 9: 227-246. Diederich, C. J., P. R. Stauffer and D. Bozzo (1994). “An improved bolus configuration for commercial multielement ultrasound and microwave hyperthermia systems.” Med. Phys. 21(9): 1401-3. Djavan, B., C. Seitz, K. Ghawidel, et al. (1999). “High-energy transurethral microwave thermotherapy in patients with acute urinary retention due to benign prostatic hyperplasia.” Urology 54(1): 18-22. Doss, J. D. and C. W. McCabe (1976). “A technique for localized heating in tissue: An adjunct to tumor therapy.” Med. Instrum. 10: 16-21. Dupuy, D. E., W. W. Mayo-Smith, G. F. Abbott, et al. (2002). “Clinical applications of radio-frequency tumor ablation in the thorax.” Radiographics 22 Spec No: S259-69. Emami, B., C. Scott, C. A. Perez, et al. (1996). “Phase iii study of interstitial thermoradiotherapy compared with interstitial radiotherapy alone in the treatment of recurrent or persistent human tumors. A prospectively controlled randomized study by the radiation therapy group.” Int. J. Radiat. Oncol. Biol. Phys. 34(5): 1097-104. Emami, B., P. Stauffer, M. W. Dewhirst, et al. (1991). “RTOG quality assurance guidelines for interstitial hyperthermia.” International Journal of Radiation Oncology Biology Physics 20: 1117-1124. Enad, J. G., F. D. Kharrazi, N. S. ElAttrache, et al. (2003). “Electrothermal capsulorrhaphy in glenohumeral instability without bankart tear.” Arthroscopy 19(7): 740-5. Erce, C. and R. W. Parks (2003). “Interstitial ablative techniques for hepatic tumours.” Br J Surg 90(3): 272-89. Fenn, A. J., C. J. Diederich and P. R. Stauffer (1993). “An adaptive-focusing algorithm for a microwave planar phasedarray hyperthermia system.” Linc. Lab. J. (USA) 6(2): 269-288. Fenn, A. J., G. L. Wolf and R. M. Fogle (1999). “An adaptive microwave phased array for targeted heating of deep tumours in intact breast: Animal study results.” Int. J. Hyperthermia 15(1): 45-61. Fessenden, P. and J. W. Hand (1995). Hyperthermia therapy physics. Medical radiology: Radiation therapy physics. Smith, A. R. Berlin, Heidelberg, Springer-Verlag: 315-363. Field, S. B. and J. W. Hand, Eds. (1990). An introduction to the practical aspects of clinical hyperthermia. London, New York, Taylor & Francis. Fischer, Y., J. Gosepath, R. G. Amedee, et al. (2000). “Radiofrequency volumetric tissue reduction (RFVTR) of inferior turbinates: A new method in the treatment of chronic nasal obstruction.” Am J Rhinol 14(6): 355-60. Fosmire, H., K. Hynynen, G. W. Drach, et al. (1993). “Feasibility and toxicity of transrectal ultrasound hyperthermia in the treatment of locally advanced adenocarcinoma of the prostate.” International Journal of Radiation Oncology Biology Physics 26(2): 253-259. Franconi, C., L. Raganella, C. A. Tiberio, et al. (1991). “Low-frequency RF hyperthermia: Iv—a 27 MHz hybrid applicator for localized deep tumor heating.” IEEE Trans. Biomed. Eng. 38(3): 287-293. Franconi, C., J. Vrba, Jr. and F. Montecchia (1993). “27 MHz hybrid evanescent-mode applicators (HEMA) with flexible heating field for deep and safe subcutaneous hyperthermia.” Int. J. Hyperthermia 9(5): 655-673. Fujita, Y., H. Kato and T. Ishida (1993). “An RF concentrating method using inductive aperture-type applicators.” IEEE Trans. Biomed. Eng. 40(1): 110-113. Garcea, G., T. D. Lloyd, C. Aylott, et al. (2003). “The emergent role of focal liver ablation techniques in the treatment of primary and secondary liver tumours.” Euro J Cancer 39(15): 2150-64. Gautherie, M., Ed. (1990). Methods of external hyperthermia heating. Clinical Thermology: Subseries thermotherapy. Berlin, Heidelberg, Springer-Verlag.
926
Jean Pouliot et al.
Gellermann, J., W. Wlodarczyk, H. Ganter, et al. (2005). “A practical approach to thermography in a hyperthermia/magnetic resonance hybrid system: Validation in a heterogeneous phantom.” Int. J. Radiat. Oncol. Biol. Phys. 61(1): 267-77. Gellermann, J., P. Wust, D. Stalling, et al. (2000). “Clinical evaluation and verification of the hyperthermia treatment planning system HyperPlan.” Int. J. Radiat. Oncol. Biol. Phys. 47(4): 1145-56. Gelvich, E. A., V. N. Mazokhin and I. I. Troshin (1996). “An attempt at quantitative specification of SAR distribution homogeneity.” Int. J. Hyperthermia 12(3): 431-6. Gianfelice, D., A. Khiat, Y. Boulanger, et al. (2003). “Feasibility of magnetic resonance imaging-guided focused ultrasound surgery as an adjunct to Tamoxifen therapy in high-risk surgical patients with breast carcinoma.” J Vasc Interv Radiol 14(10): 1275-82. Goldberg, D. J. (2003). “Lasers for facial rejuvenation.” Am J Clin Dermatol 4(4): 225-34. Gonzalez Gonzalez, D., J. D. P. Van Dijk, F. Oldenburger, et al. (1992). Results of combined treatment with radiation and hyperthermia in 111 patients with large or deep seated tumors. Hyperthermic oncology 1992, vol. 1. Gerner, E. W. Tucson, Arizona Board of Regents: 415b. Gopal, M. K. and T. C. Cetas (1993). “Current sheet applicators for clinical microwave hyperthermia.” IEEE Trans. Mic. Theory Tech. 41(3): 431-437. Gopal, M. K., J. W. Hand, M. L. D. Lumori, et al. (1992). “Current sheet applicator arrays for superficial hyperthermia of chestwall lesions.” Int. J. Hyperthermia 8(2): 227-240. Gross, E. J., T. C. Cetas, P. R. Stauffer, et al. (1990). “Experimental assessment of phased-array heating of neck tumours.” Int. J. Hyperthermia 6(2): 453-474. Guy, A. W. (1971). “Electromagnetic fields and relative heating patterns due to a rectangular aperture source in direct contact with bilayered biological tissue.” IEEE Trans. Mic. Theory Tech. 19: 214-223. Guy, A. W., J. Lehmann and J. B. Stonebridge (1974). “Therapeutic applications of electromagnetic power.” IEEE Proceedings 62(1): 55-75. Haemmerich, D. and F. T. Lee, Jr. (2005). “Multiple applicator approaches for radiofrequency and microwave ablation.” Int J Hyperthermia 21(2): 93-106. Haemmerich, D., F. T. Lee, Jr., D. J. Schutt, et al. (2005). “Large-volume radiofrequency ablation of ex vivo bovine liver with multiple cooled cluster electrodes.” Radiology (USA) 234(2): 563-8. Hand, J. W. (1990). Biophysics and technology of electromagnetic hyperthermia. Methods of external hyperthermia heating. Gautherie, M. Berlin, Heidelberg, Springer-Verlag: 1-60. Hand, J. W., J. L. Cheetham and A. J. Hind (1986). “Absorbed power distributions from coherent microwave arrays for localized hyperthermia.” IEEE Trans. Mic. Theory Tech. 34(5): 484-489. Hand, J. W., B. S. Trembly and M. V. Prior (1991). Physics of interstitial hyperthermia: Radiofrequency and hot water tube techniques. Hyperthermia and oncology, vol. 3. Urano, M. and E. Douple. Zeist, VSP: 99-134. Hand, J. W., C. C. Vernon, M. V. Prior, et al. (1993). Current sheet applicator arrays for superficial hyperthermia. Hyperthermic oncology 1992, vol. 2. Gerner, E. and T. Cetas. Tucson, Arizona Board of Regents. 2: 193-7. Handl-Zeller, L., Ed. (1992). Interstitial hyperthermia. Wien, New York, Springer-Verlag. Hayashi, A. H., S. F. Silver, N. G. van der Westhuizen, et al. (2003). “Treatment of invasive breast carcinoma with ultrasound-guided radiofrequency ablation.” Am J Surg 185(5): 429-35. Hehr, T., P. Wust, M. Bamberg, et al. (2003). “Current and potential role of thermoradiotherapy for solid tumours.” Onkologie 26(3): 295-302. Hindley, J. T., W. M. Gedroyc and L. Regan (2003). “Interstitial laser coagulation for uterine myomas.” Am J Obstet Gynecol 188(3): 859; author reply 860. Hiraoka, M., Y. Nishimura, Y. Nagata, et al. (1995). “Clinical results of thermoradiotherapy for soft tissue tumours.” Int. J. Hyperthermia 11(3): 365-377. Huang, S. K., P. R. Stauffer, K. Hong, et al. (1994). “Liposomes and hyperthermia in mice: Increased tumor uptake and therapeutic efficacy of doxorubicin in sterically stabilized liposomes.” Cancer Res. 54(8): 2186-91. Hulshof, M. C., B. W. Raaymakers, J. J. Lagendijk, et al. (2004). “A feasibility study of interstitial hyperthermia plus external beam radiotherapy in glioblastoma multiforme using the multi electrode current source (MECS) system.” Int J Hyperthermia 20(5): 451-63. Hunt, J. W. (1990). Principles of ultrasound used for generating localized hyperthermia. An introduction to the practical aspects of clinical hyperthermia. Field, S. B. and J. W. Hand. London, New York, Taylor & Francis: 371-422.
52–Thermal Therapy for Cancer
927
Hurwitz, M. D., I. D. Kaplan, G. K. Svensson, et al. (2001). “Feasibility and patient tolerance of a novel transrectal ultrasound hyperthermia system for treatment of prostate cancer.” Int J Hyperthermia 17(1): 31-7. Hynynen, K. (1995). Ultrasound heating technology. Thermoradiotherapy and thermochemotherapy. Seegenschmiedt, M. H., P. Fessenden and C. C. Vernon. Berlin, Heidelberg, Springer-Verlag. 1, Chapter 12: 253-77. Hynynen, K., O. Pomeroy, D. N. Smith, et al. (2001). “MR imaging-guided focused ultrasound surgery of fibroadenomas in the breast: A feasibility study.” Radiology (USA) 219(1): 176-85. Hynynen, K., R. Roemer, D. Anhalt, et al. (1987). “A scanned, focused, multiple transducer ultrasonic system for localized hyperthermia treatments.” Int. J. Hyperthermia 3(1): 21-35. Hynynen, K., D. Shimm, D. Anhalt, et al. (1990). “Temperature distributions during clinical scanned, focused ultrasound hyperthermia treatments.” Int. J. Hyperthermia 6(5): 891-908. Johnson, R. H., A. W. Preece, J. W. Hand, et al. (1987). “A new type of lightweight low-frequency electromagnetic hyperthermia applicator.” IEEE Trans. Mic. Theory Tech. 35(12): 1317-1321. Jones, E., J. Oleson, L. Prosnitz, et al. (2005). “A randomized trial of hyperthermia and radiation for superficial tumors.” JCO(May In Press). Jones, E. L., T. V. Samulski, M. W. Dewhirst, et al. (2003). “A pilot phase ii trial of concurrent radiotherapy, chemotherapy, and hyperthermia for locally advanced cervical carcinoma.” Cancer 98(2): 277-82. Kakehi, M., K. Ueda, T. Mukojima, et al. (1990). “Multi-institutional clinical studies on hyperthermia combined with radiotherapy or chemotherapy in advanced cancer of deep-seated organs.” Int. J. Hyperthermia 6(4): 719-740. Kangasniemi, M., C. J. Diederich, R. E. Price, et al. (2002). “Multiplanar MR temperature-sensitive imaging of cerebral thermal treatment using interstitial ultrasound applicators in a canine model.” J Magn Reson Imaging 16(5): 522-31. Kapp, D. S. and R. S. Cox (1993). “Cumulative minutes of isoeffective hyperthermia with t90=43¯c is the treatment parameter most predictive for outcome in patients with metastatic adenocarcinoma of the breast and single tumor nodules per treatment field.” International Journal of Radiation Oncology Biology Physics 27 (Suppl. 1): 315-316. Kapp, D. S. and R. S. Cox (1995). “Thermal treatment parameters are most predictive of outcome in patients with single tumor nodules per treatment field in recurrent adenocarcinoma of the breast.” Int. J. Radiat. Oncol. Biol. Phys. 33: 887-99. Kato, H., J. W. Hand, M. V. Michael, et al. (1991). “Control of specific absorption rate distribution using capacitive electrodes and inductive aperture-type applicators: Implications for radiofrequency hyperthermia.” IEEE Trans. Biomed. Eng. 38(7): 644. Kato, H., N. Uchida, T. Kasai, et al. (1995). “A new applicator utilizing distributed electrodes for hyperthermia: A theoretical approach.” Int. J. Hyperthermia 11(2): 287-294. Lagendijk, J. J. W. (1990). Thermal models: Principles and implementation. Practical aspects of clinical hyperthermia. Field, S. B. and J. W. Hand. London, Taylor and Francis: 478-512. Lagendijk, J. J. W., G. C. van Rhoon, S. N. Hornsleth, et al. (1998). “ESHO quality assurance guidelines for regional hyperthermia.” Int. J. Hyperthermia 14(2): 125-33. Lamaitre, G., J. D. P. Van Dijk, E. A. Gelvich, et al. (1996). “SAR characteristics of three types of contact flexible microstrip applicators for superficial hyperthermia.” Int. J. Hyperthermia 12(2): 255-69. Larson, T. R. (2002). “Rationale and assessment of minimally invasive approaches to benign prostatic hyperplasia therapy.” Urology 59(2 Suppl 1): 12-6. Lee, E. R. (1995). Electromagnetic superficial heating technology. Thermoradiotherapy and thermochemotherapy. Seegenschmiedt, M. H., P. Fessenden and C. C. Vernon. Berlin, Heidelberg, Springer-Verlag. 1, Chapter 10: 193-217. Lee, E. R., T. R. Wilsey, P. Tarczy-Hornoch, et al. (1992). “Body conformable 915 MHz microstrip array applicators for large surface area hyperthermia.” IEEE Trans. Biomed. Eng. 39(5): 470-483. Leigh, B. R., B. Stea, J. R. Cassady, et al. (1994). “Clinical hyperthermia with a new device: The current sheet applicator.” International Journal of Radiation Oncology Biology Physics 30(4): 945-951. Leonardi, M. A. and C. B. Lumenta (2002). “Stereotactic guided laser-induced interstitial thermotherapy (SLITT) in gliomas with intraoperative morphologic monitoring in an open MR: Clinical experience.” Minim Invasive Neurosurg 45(4): 201-7.
928
Jean Pouliot et al.
Leopold, K. A., M. W. Dewhirst, T. V. Samulski, et al. (1993). “Cumulative minutes with T90 greater than Tempindex is predictive of response of superficial malignancies to hyperthermia and radiation.” International Journal of Radiation Oncology Biology Physics 25: 841-847. Lin, D. and F. E. Marchlinski (2003). “Advances in ablation therapy for complex arrhythmias: Atrial fibrillation and ventricular tachycardia.” Curr Cardiol Rep 5(5): 407-14. Martinez, A. A., D. Gersten and P. M. Corry (1993). Interstitial thermo-brachytherapy with continuous mild hyperthermia for the treatment of locally advanced or recurrent cancer of the female tract. Diagnostic imaging and radiation oncology. Seegenschmiedt, H. M. and R. Sauer. Bangalore, India, Macmillan India Ltd. Montecchia, F. (1992). “Microstrip-antenna design for hyperthermia treatment of superficial tumors.” IEEE Trans. Biomed. Eng. 39(6): 580-588. Moros, E. G., X. Fan and W. L. Straube (1999). “Experimental assessment of power and temperature penetration depth control with a dual frequency ultrasonic system.” Med. Phys. 26(5): 810-17. Moros, E. G., W. L. Straube, E. E. Klein, et al. (1995). “Simultaneous delivery of electron beam therapy and ultrasound hyperthermia utilizing scanning reflectors: A feasibility study.” International Journal of Radiation Oncology Biology Physics 31(4): 893-904. Moros, E. G., W. L. Straube and R. J. Myerson (1994). “Devices and techniques for the clinical application of concomitant heat and ionizing radiation by external means.” Biomed. Eng. Appl. Basis Comm. 6: 328-39. Murphy, D. P. and I. S. Gill (2001). “Energy-based renal tumor ablation: A review.” Semin Urol Oncol 19(2): 133-40. Myerson, R. J., L. Leybovich, B. Emami, et al. (1991). “Phantom studies and preliminary clinical experience with the BSD 2000.” Int. J. Hyperthermia 7(6): 937-951. Nau, W. H., C. J. Diederich and E. C. Burdette (2001). “Evaluation of multielement catheter-cooled interstitial ultrasound applicators for high-temperature thermal therapy.” Med. Phys. 28(7): 1525-34. Nau, W. H., C. J. Diederich, A. B. Ross, et al. (2005). “MRI-guided interstitial ultrasound thermal therapy of the prostate: A feasibility study in the canine model.” Med. Phys. 32(3): 733-43. Nau, W. H., C. J. Diederich and P. R. Stauffer (2000). “Directional power deposition from direct-coupled and cathetercooled interstitial ultrasound applicators.” Int. J. Hyperthermia 16(2): 129-44. Neuman, D. G., P. R. Stauffer, S. Jacobsen, et al. (2002). “SAR pattern perturbations from resonance effects in water bolus layers used with superficial microwave hyperthermia applicators.” International Journal of Hyperthermia 18(3): 180-93. Noonan, T. J., J. M. Tokish, K. K. Briggs, et al. (2003). “Laser-assisted thermal capsulorrhaphy.” Arthroscopy 19(8): 815-9. Novak, P., E. G. Moros, W. L. Straube, et al. (2005). “Surlas: A new clinical grade ultrasound system for sequential or concomitant thermoradiotherapy of superficial tumors: Applicator description.” Med. Phys. 32(1): 230-40. Nussbaum, G. H., Ed. (1982). Physical aspects of hyperthermia. New York, American Association of Physicists in Medicine. Oleson, J. R., M. W. Dewhirst, J. M. Harrelson, et al. (1989). “Tumor temperature distributions predict hyperthermia effect.” International Journal of Radiation Oncology Biology Physics 16: 559-570. Oleson, J. R., T. V. Samulski, K. A. Leopold, et al. (1993). “Sensitivity of hyperthermia trial outcomes to temperature and time: Implications for thermal goals of treatment.” International Journal of Radiation Oncology Biology Physics 25: 289-297. Oleson, J. R. (1995). “Hyperthermia from the clinic to the laboratory: A hypothesis.” Int. J. Hyperthermia 11: 315-322. Orenberg, E. K., F. R. Noodleman, J. A. Koperski, et al. (1986). “Comparison of heat delivery systems for hyperthermia treatment of psoriasis.” Int. J. Hyperthermia 2(3): 231-241. Overgaard, J. (1989). “The current and potential role of hyperthermia in radiotherapy.” International Journal of Radiation Oncology Biology Physics 16: 535-549. Overgaard, J., D. Gonzalez Gonzalez, M. C. C. M. Hulshof, et al. (1995). “Randomised trial of hyperthermia as adjuvant to radiotherapy for recurrent or metastatic malignant melanoma.” Lancet 345: 540-543. Perez, C. A., T. Pajak, B. Emami, et al. (1991). “Randomized phase iii study comparing irradiation and hyperthermia with irradiation alone in superficial measurable tumors. Final report by the radiation therapy oncology group.” Am. J. Clin. Oncol. 14(2): 133-41.
52–Thermal Therapy for Cancer
929
Prior, M. V., M. L. D. Lumori, J. W. Hand, et al. (1995). “The use of a current sheet applicator array for superficial hyperthermia: Incoherent versus coherent operation.” IEEE Trans. Biomed. Eng. 42(7): 694-698. Quesson, B., J. A. de Zwart and C. T. Moonen (2000). “Magnetic resonance temperature imaging for guidance of thermotherapy.” J Magn Reson Imaging 12(4): 525-33. Ramsey, E. W. and C. Dahlstrand (2000). “Durability of results obtained with transurethral microwave thermotherapy in the treatment of men with symptomatic benign prostatic hyperplasia.” Journal of Endourology 14(8): 671-5. Rietveld, P. J. M., M. L. D. Lumori, J. Van Der Zee, et al. (1998). “Quantitative evaluation of 2*2 arrays of Lucite cone applicators in flat layered phantoms using gaussian-beam-predicted and thermographically measured SAR distributions.” Phys. Med. Biol. 43(8): 2207-20. Rietveld, P. J. M., W. L. J. Van Putten, J. Van Der Zee, et al. (1999). “Comparison of the clinical effectiveness of the 433 MHz Lucite cone applicator with that of a conventional waveguide applicator in applications of superficial hyperthermia.” International Journal of Radiation Oncology Biology Physics 43(3): 681-7. Rosenthal, D. I., F. J. Hornicek, M. Torriani, et al. (2003). “Osteoid osteoma: Percutaneous treatment with radiofrequency energy.” Radiology (USA) 229(1): 171-5. Rossetto, F., C. J. Diederich and P. R. Stauffer (2000). “Thermal and SAR characterization of multielement dual concentric conductor microwave applicators for hyperthermia, a theoretical investigation.” Med. Phys. 27(4): 745-53. Rossetto, F. and P. R. Stauffer (1999). “Effect of complex bolus-tissue load configurations on SAR distributions from dual concentric conductor applicators.” IEEE Trans. Biomed. Eng. 46(11): 1310-19. Rossetto, F., P. R. Stauffer, V. Manfrini, et al. (1998). “Effect of practical layered dielectric loads on SAR patterns from dual concentric conductor microstrip antennas.” Int. J. Hyperthermia 14(6): 513-34. Ruiz-Esparza, J. and J. B. Gomez (2003). “The medical face lift: A noninvasive, nonsurgical approach to tissue tightening in facial skin using nonablative radiofrequency.” Dermatol Surg 29(4): 325-32; discussion 332. Ryan, T. P., V. L. Backus and C. T. Coughlin (1995). “Large stationary microstrip arrays for superficial microwave hyperthermia at 433 MHz: SAR analysis and clinical data.” Int. J. Hyperthermia 11(2): 187-209. Saal, J. A. and J. S. Saal (2002). “Intradiscal electrothermal therapy for the treatment of chronic discogenic low back pain.” Clin Sports Med 21(1): 167-87. Samulski, T. V., W. J. Grant, J. R. Oleson, et al. (1990). “Clinical experience with a multi-element ultrasonic hyperthermia system: Analysis of treatment temperatures.” Int. J. Hyperthermia 6(5): 909-922. Samulski, T. V., D. S. Kapp, P. Fessenden, et al. (1987). “Heating deep seated eccentrically located tumors with an annular phased array system: A comparative clinical study using two annular array operating configurations.” International Journal of Radiation Oncology Biology Physics 13: 907-916. Sandhu, T. S. (1982). Clinical hyperthermia with microwaves. Physical aspects of hyperthermia. Nussbaum, G. H. NY, American Institute of Physics Inc.: 329-356. Sanghvi, N. T., R. S. Foster, R. Bihrle, et al. (1999). “Noninvasive surgery of prostate tissue by high intensity focused ultrasound: An updated report.” Euro J Ultrasound 9(1): 19-29. Sapareto, S. A. and W. C. Dewey (1984). “Thermal dose determination in cancer therapy.” International Journal of Radiation Oncology Biology Physics 10: 787-800. Sapozink, M. D., T. C. Cetas, P. M. Corry, et al. (1988). “Final report of the NCI hyperthermia equipment evaluation contractor’s group.” Int. J. Hyperthermia 4(1): 1-132. Sapozink, M. D., P. M. Corry, D. S. Kapp, et al. (1991). “RTOG quality assurance guidelines for clinical trials using hyperthermia for deep-seated malignancy.” International Journal of Radiation Oncology Biology Physics 20(5): 1109-1115. Seegenschmiedt, M. H., P. Fessenden and C. C. Vernon, Eds. (1995). Thermoradiotherapy and thermochemotherapy: Volume 1 biology, physiology, and physics. Berlin, Heidelberg, Springer-Verlag. Seegenschmiedt, M. H., P. Fessenden and C. C. Vernon (1996). Thermoradiotherapy and thermochemotherapy: Volume 2, clinical applications. Berlin, New York, Springer-Verlag. Seegenschmiedt, M. H., G. Klautke, R. Seidel, et al. (1996). Clinical practice of interstitial thermoradiotherapy. Thermoradiotherapy and thermochemotherapy: Volume 2, clinical applications. Seegenschmiedt, M. H., P. Fessenden and C. C. Vernon. Berlin, New York, Springer-Verlag: 207-320.
930
Jean Pouliot et al.
Seegenschmiedt, M. H. and R. Sauer, Eds. (1993). Interstitial and intracavitary thermoradiotherapy. Medical radiology: Diagnostic imaging and radiation oncology. Berlin, Heidelberg, Springer-Verlag. Sherar, M., F. F. Liu, M. Pintilie, et al. (1997). “Relationship between thermal dose and outcome in thermoradiotherapy treatments for superficial recurrences of breast cancer: Data from a phase iii trial.” International Journal of Radiation Oncology, Biology, Physics 39(2): 371-80. Sherar, M. D., M. R. Gertner, C. K. Yue, et al. (2001). “Interstitial microwave thermal therapy for prostate cancer: Method of treatment and results of a phase I/II trial.” J. Urology 166(5): 1707-14. Sherar, M. D., J. Trachtenberg, S. R. Davidson, et al. (2004). “Interstitial microwave thermal therapy and its application to the treatment of recurrent prostate cancer.” Int J Hyperthermia 20(7): 757-68. Sherar, M. D., J. Trachtenberg, S. R. Davidson, et al. (2003). “Interstitial microwave thermal therapy for prostate cancer.” J Endourol 17(8): 617-25. Sneed, P. K. and T. L. Phillips (1991). “Combining hyperthermia and radiation: How beneficial?” Oncology 5: 99-108. Sneed, P. K., P. R. Stauffer, G. C. Li, et al. (1998a). Hyperthermia. Textbook of radiation oncology. Leibel, S. A. and T. L. Phillips. Philadelphia, W B Saunders Co.: 1241-62. Sneed, P. K., P. R. Stauffer, M. W. McDermott, et al. (1998b). “Survival benefit of hyperthermia in a prospective randomized trial of brachytherapy boost +/- hyperthermia for glioblastoma multiforme.” Int. J. Radiat. Oncol. Biol. Phys. 40(2): 287-95. Song, C. W., I. B. Choi, B. S. Nah, et al. (1995). Microvasculature and perfusion in normal tissues and tumors. Thermoradiotherapy and thermochemotherapy: Volume 1, biology, physiology and physics. Seegenschmiedt, M. H., P. Fessenden and C. C. Vernon. Berlin, New York, Springer-Verlag: 139-156. Stauffer, P., J. Schlorff, R. Taschereau, et al. (2004). Combination applicator for simultaneous heat and radiation. Intl. Conf. IEEE Eng. Med. Biol. Society, San Francisco, IEEE Press. Stauffer, P. R. (1990). Techniques for interstitial hyperthermia. An introduction to the practical aspects of clinical hyperthermia. Field, S. B. and J. W. Hand. London, New York, Taylor & Francis: 344-370. Stauffer, P. R. (2000). Thermal therapy techniques for skin and superficial tissue disease. A critical review, matching the energy source to the clinical need. Ryan, T. P. Bellingham WA, SPIE Optical Engineering Press. CR75: 327-67. Stauffer, P. R., C. J. Diederich and D. Bozzo (1994a). Conformal array microwave applicator for superficial hyperthermia of large contoured surfaces. IEEE MTT-S Intl. Mic. Symp. Digest. Stauffer, P. R., C. J. Diederich and M. H. Seegenschmiedt (1995). Interstitial heating technologies. Thermoradiotherapy and thermochemotherapy: Volume 1, biology, physiology and physics. Seegenschmiedt, M. H., P. Fessenden and C. C. Vernon. Berlin, New York, Springer-Verlag: 279-320. Stauffer, P. R., C. J. Diederich, P. K. Sneed, et al. (1994b). Preliminary clinical experience with planar and conformal microwave array applicators for hyperthermia. Fourteenth Annual Meeting of the North American Hyperthermia Society, Nashville, Tennessee. Stauffer, P. R. and S. N. Goldberg (2004). “Introduction: Thermal ablation therapy.” Int J Hyperthermia 20(7): 671-7. Stauffer, P. R., S. Jacobsen and D. Neuman (2001). Microwave array applicator for radiometry controlled superficial hyperthermia. Thermal treatment of tissue: Energy delivery and assessment. Ryan, T. P. San Jose, Proceedings of SPIE. 4247: 19-29. Stauffer, P. R., M. Leoncini, V. Manfrini, et al. (1995). “Dual concentric conductor radiator for microwave hyperthermia with improved field uniformity to periphery of aperture.” IEICE Trans. on Communicat. E78-B(6): 826-35. Stauffer, P. R., F. Rossetto, M. Leoncini, et al. (1998). “Radiation patterns of dual concentric conductor microstrip antennas for superficial hyperthermia.” IEEE Trans. Biomed. Eng. 45(5): 605-13. Stauffer, P. R., J. L. Schlorff, T. Juang, et al. (2005). Progress on system for applying simultaneous heat and brachytherapy to large-area surface disease. SPIE BIOS 2005, San Jose, SPIE Press. Stewart, E. A., W. M. Gedroyc, C. M. Tempany, et al. (2003). “Focused ultrasound treatment of uterine fibroid tumors: Safety and feasibility of a noninvasive thermoablative technique.” Am J Obstet Gynecol 189(1): 48-54. Straube, W. L., R. J. Myerson, B. Emami, et al. (1990). “SAR patterns of external 915 MHz microwave applicators.” Int. J. Hyperthermia 6(3): 665-670.
52–Thermal Therapy for Cancer
931
Strohbehn, J. W. (1987). Interstitial techniques for hyperthermia. Physics and technology of hyperthermia. Field, S. B. and C. Franconi. Dordrecht, Boston, Lancaster, Martinus Nijhoff Publishers: 211-240. Taschereau, R., P. R. Stauffer, I. C. Hsu, et al. (2004). “Radiation dosimetry of a conformal heat-brachytherapy applicator.” Technology in Cancer Research and Treatment 3(4): 347-58. Tempany, C. M., E. A. Stewart, N. McDannold, et al. (2003). “MR imaging-guided focused ultrasound surgery of uterine leiomyomas: A feasibility study.” Radiology (USA) 226(3): 897-905. Thuroff, S., C. Chaussy, G. Vallancien, et al. (2003). “High-intensity focused ultrasound and localized prostate cancer: Efficacy results from the European multicentric study.” J Endourol 17(8): 673-7. Tucker, R. D., C. E. Platz, C. Huidobro, et al. (2002). “Interstitial thermal therapy in patients with localized prostate cancer: Histologic analysis.” Urology 60(1): 166-9. Turner, P. F. and L. Kumar (1982). “Computer solution for applicator heating patterns.” National Cancer Institute Monograph 61: 521-523. Turner, P. F. (1984). “Regional hyperthermia with an annular phased array.” IEEE Trans. Biomed. Eng. 31: 106-114. Uchida, T., N. T. Sanghvi, T. A. Gardner, et al. (2002). “Transrectal high-intensity focused ultrasound for treatment of patients with stage t1b-2n0m0 localized prostate cancer: A preliminary report.” Urology 59(3): 394-8; discussion 398-9. Underwood, H. R., A. F. Peterson and R. L. Magin (1992). “Electric-field distribution near rectangular microstrip radiators for hyperthermia heating: Theory versus experiment in water.” IEEE Trans. Biomed. Eng. 39(2): 146-153. Valdagni, R. and M. Amichetti (1994). “Report of long-term follow-up in a randomized trial comparing radiation therapy and radiation therapy plus hyperthermia to metastatic lymph nodes in stage iv head and neck patients.” International Journal of Radiation Oncology Biology Physics 28: 163-169. Valdagni, R., M. Amichetti and G. Pani (1988). “Radical radiation alone versus radical radiation plus microwave hyperthermia for n3 (tnm-uicc) neck nodes: A prospective randomized clinical trial.” International Journal of Radiation Oncology Biology Physics 15: 13-24. van der Zee, J. and G. D. Gonzalez (2002). “The Dutch deep hyperthermia trial: Results in cervical cancer.” Int J Hyperthermia 18(1): 1-12. van der Zee, J., D. Gonzalez-Gonzalez, G. C. van Rhoon, et al. (2000). “Comparison of radiotherapy alone with radiotherapy plus hyperthermia in locally advanced pelvic tumours: A prospective, randomised, multicentre trial.” Lancet 355: 1119-25. Van Es, C. A., H. K. Wyrdeman, A. A. de Leeuw, et al. (1995). “Regional hyperthermia of pelvic tumours using the Utrecht “coaxial tem” system: A feasibility study.” Int. J. Hyperthermia 11(2): 173-186. Van Rhoon, G. C., P. J. M. Rietveld and J. Van Der Zee (1998). “A 433 MHz Lucite cone waveguide applicator for superficial hyperthermia.” Int. J. Hyperthermia 14(1): 13-27. van Rhoon, G. C., J. van der Zee, M. P. Broekmeyer-Reurink, et al. (1992). “Radiofrequency capacitive heating of deep-seated tumours using pre-cooling of the subcutaneous tissues: Results on thermometry in Dutch patients.” Int. J. Hyperthermia 8(6): 843-854. van Vulpen, M., B. W. Raaymakers, J. J. Lagendijk, et al. (2002). “Three-dimensional controlled interstitial hyperthermia combined with radiotherapy for locally advanced prostate carcinoma—a feasibility study.” Int. J. Radiat. Oncol. Biol. Phys. 53(1): 116-26. Vernon, C. C., J. W. Hand, S. B. Field, et al. (1996). “Radiotherapy with or without hyperthermia in the treatment of superficial localized breast cancer: Results from five randomized controlled trials.” Int. J. Radiat. Oncol. Biol. Phys. 35(4): 731-44. Vujaskovic, Z., J. M. Poulson, A. A. Gaskin, et al. (2000). “Temperature-dependent changes in physiologic parameters of spontaneous canine soft tissue sarcomas after combined radiotherapy and hyperthermia treatment.” International Journal of Radiation Oncology, Biology, Physics 46(1): 179-85. Waterman, F. M., M. W. Dewhirst, P. Fessenden, et al. (1991). “RTOG quality assurance guidelines for clinical trials using hyperthermia administered by ultrasound.” International Journal of Radiation Oncology Biology Physics 20(5): 1099-1107. Wust, P., J. Gellermann, M. Seebass, et al. (2004). “[part-body hyperthermia with a radiofrequency multiantenna applicator under online control in a 1.5 T MR-tomograph].” Rofo 176(3): 363-74. Wust, P., B. Hildebrandt, G. Sreenivasa, et al. (2002). “Hyperthermia in combined treatment of cancer.” Lancet Oncol 3(8): 487-97.
932
Jean Pouliot et al.
Xu, M., R. J. Meyerson, W. L. Straube, et al. (2002). “Radiosensitization of heat resistant human tumour cells by 1 hour at 41.1c and its effect on DNA repair.” Int. J. Hyperthermia 18(5): 385-403. Yee, R., S. Connolly and H. Noorani (2003). “Clinical review of radiofrequency catheter ablation for cardiac arrhythmias.” Can J Cardiol 19(11): 1273-84. Zhu, L., L. X. Xu and N. Chencinski (1998). “Quantification of the 3-d electromagnetic power absorption rate in tissue during transurethral prostatic microwave thermotherapy using heat transfer model.” IEEE Trans. Biomed. Eng. 45(9): 1163-72. Zlotta, A. R. and B. Djavan (2002). “Minimally invasive therapies for benign prostatic hyperplasia in the new millennium: Long-term data.” Curr Opin Urol 12(1): 7-14. Zlotta, A. R., B. Djavan, C. Matos, et al. (1998). “Percutaneous transperineal radiofrequency ablation of prostate tumour: Safety, feasibility and pathological effects on human prostate cancer.” Br J Urol 81(2): 265-75.
Chapter 53
Radiofrequency Ablation Christopher L. Brace, Ph.D.1, and Paul F. Laeseke, M.D.2 1 Department of Electrical and Computer Engineering 2 Department of Radiology University of Wisconsin, Madison, Wisconsin Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 933 Role of Radiofrequency Ablation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 933 Historical Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 934 Standard Terminology and Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 934 Applications of RF Ablation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935 Clinical Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935 Liver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 935 Lung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 938 Kidney . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 938 Bone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 938 Clinical Research Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 939 Principles of RF Ablation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940 RF Heating in Tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940 Joule Heating by RF in Tissue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 940 Tissue Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 941 Zones of Ablation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 941 Drawbacks to RF Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 942 RF Ablation Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 944 Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 944 Single Electrode Designs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 944 Expandable Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945 Multiple Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945 Grounding Pads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 946 RF Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947 Feedback and Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947 Advanced Power Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947 Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948 Imaging and Navigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948 Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948 Monitoring and Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949
Introduction Role of Radiofrequency Ablation Approximately one in four deaths in the United States is caused by cancer, with 246,880 deaths expected in 2005 from colon, kidney, liver, and lung tumors (Jemal at al., 2005). Unfortunately, while surgical resection improves survival in many of these patients, only a minority of patients are eligible for surgery due to advanced disease, co-morbidities, prior resection, or limited organ reserve. This has led to the development of minimally invasive procedures for the local control of tumors. Radiofrequency (RF) ablation is a member of this diverse group of image-guided oncologic interventions. It falls within the heat-based
933
934
Christopher L. Brace and Paul F. Laeseke
thermal ablation therapies, which attempt to destroy tumors by subjecting the tumor and surrounding tissue to extremely high, tumoricidal temperatures in a minimally invasive fashion.
Historical Background The use of RF energy to make lesions was first described in 1976 by Organ (1976). However, it was not until the early 1990s that the modern era of RF ablation was born, when McGahan and colleagues reported its use in an animal model McGahan et al., 1990; 1992). RF ablation rapidly received Food and Drug Administration (FDA) approval as a method of treatment for focal malignancies because of its similarities to surgical electrocautery. Since that time, RF ablation has experienced an explosion in popularity, as evidenced by the exponential rise in the number of articles in the literature and clinical procedures being performed (Figure 1).
Standard Terminology and Procedure In 2003 a proposal for standardization of terms and reporting criteria for image-guided tumor ablation was proposed and has since been adopted (Goldberg et a., 2003). Here we briefly describe the terms relevant to RF ablation that will be used in this chapter, and summarize how the procedure is performed. Ablation zone or zone of ablation refers to the radiological region of induced treatment effect. Likewise, zone of coagulation is used to describe the pathologic region of treatment. Finally, the RF ablation system is composed of an active electrode (or more generally, applicator), generator, and dispersive electrodes or ground pads used to complete the circuit (Figure 2). RF ablation can be performed as an open, laparoscopic, or percutaneous procedure. Percutaneous ablation is performed under imaging guidance [ultrasound, computed tomography (CT), or magnetic resonance imaging (MRI)] and is associated with minimal morbidity and a rapid return to baseline status compared to traditional surgical approaches. Prior to starting the procedure, ground pads are placed on the patient. Four pads are generally placed on the patient’s thighs. An attempt is made to keep the pads equidistant from the electrode to ensure homogeneous distribution of current flow to the pads and to prevent superficial tissue burns from high current densities underneath one or more pads. After the electrode is placed
Figure 1. Number of publications for RF ablation (RF), microwave ablation (MW), and cryoablation (Cryo) in the liver since 1990.
53–Radiofrequency Ablation
935
into the tumor, RF energy is delivered by the generator for a period of time. Imaging is performed immediately after the procedure to verify complete ablation of the tumor or to check for complications. These images also serve as a baseline when checking follow-up scans for local recurrence. Follow-up imaging is performed at approximately 1, 3, and 6 months and approximately every 6 months thereafter until 2 years, by which time the majority of local recurrences have occurred.
Applications of RF Ablation Clinical Applications Currently, RF ablation is being performed clinically for the treatment of malignancies in the neck, chest, abdomen, and musculoskeletal system with several other tumor types under investigation (Farrel et al., 2003; Davis, Choi, and Blankenbaker 2004; Decadt and Siriwardena 2004; Lencioni et al., 2004). While most commonly used for the treatment of malignant liver tumors, interest is burgeoning for ablation of renal cell carcinoma (Zagoria 2004), inoperable lung tumors (Lencioni et al., 2004), osteoid osteomas and bone metastases (Cioni et al., 2004; Davis, Choi, and Blankenbaker 2004), and breast cancer (Hayashi et al. 2003; Fornage et al., 2004). As a new and developing field, long-term reports of successes, side effects, and complications from RF ablation have only recently begun to emerge. As with any new procedure, there will inevitably be a learning curve that will depend on the user and the resources available to effect progress. Because different disease sites often have unique characteristics, the lessons learned from one application may not be transferable to another. Liver RF ablation is now considered standard therapy for a variety of conditions in the liver including: hepatocellular carcinoma in cirrhotics, hepatic metastatic colorectal cancer in non-operative patients, and debulking of symptomatic liver tumors (Figure 3). As is the case with hepatic resection, presence of extrahepatic metastases is considered a contraindication for performing RF ablation. Currently, there is no
Figure 2. Overview of an RF ablation system. The generator provides current to the electrode, which heats tissue near its tip. The body tissues and ground pads complete the electrical circuit.
936
Christopher L. Brace and Paul F. Laeseke
Figure 3. Successful percutaneous RF ablation in a patient with hepatic carcinoma (HCC) and limited hepatic reserve due to cirrhosis. (a) Preoperative CT scan demonstrates an enhancing HCC. (b) Immediate post-ablation scan demonstrates a successful ablation without residual tumor. (c) Two years later, CT demonstrates no residual enhancement and shrinking of the ablation zone. The increased ascites is due to progressive hepatic failure secondary to cirrhosis. This patient would have been at high risk for liver failure if a hepatic resection had been performed.
consensus regarding the maximum size of tumors that can be treated. Greater experience and technological advancements are allowing a more aggressive approach to the size and number of tumors that can be treated; however, tumor size and number are still important factors affecting local recurrence rates following the procedure and must be considered during patient selection. Several studies have shown a higher local recurrence rate for tumors greater than 4 cm in diameter (Machi et al., 2001; Kosari et al., 2002; Kuvshinoff and Ota 2002). Hepatic RF ablation is considered a relatively safe procedure. The overall mortality rate is approximately 0.3%, and major and minor complication rates are approximately 2% and 5%, respectively (Livraghi et al., 2003; Curley et al., 2004). Complications associated with hepatic RF ablation include hemorrhage, intra-abdominal abscess, neoplastic seeding, bile duct stricture, bowel perforation, and pain from diaphragm or body wall injury. The success of clinical RF ablation therapy has been mixed. A recent systematic review of the clinical RF ablation literature between 1995 and 2002 reported local failure rates after treatment of malignant liver tumors ranging from 0 to 34% (Decadt and Siriwardena 2004). Contributing factors to local recurrence after RF ablation include the heat sink effect of local blood flow (Figure 4) (Rossi et al., 1999; Chinn et al., 2001; Lu et al., 2002), difficulty in ultrasound imaging of RF ablation zones (Figure 5) (Cha et al., 2000; Liu et al., 2001), and an inadequately tumoricidal environment with surviving tumor cells even within RF ablation zones (Solbiati et al., 1997). Recent developments in RF ablation technology that have enhanced the ability to create larger and more reliable zones of coagulation will be discussed later in this chapter. Because the field of tumor ablation is relatively new, long-term survival data has not been available to date. Recent data presented at the 2004 Radiological Society of North America (RSNA) Annual Meeting strongly suggests a long-term survival advantage for patients with hepatic colorectal metastases treated with RF ablation. One series reported 1, 2, 3, and 5-year survival rates of 96.2%, 64.2%, 45.7%, and 22.1%, respectively, with a median survival of 33 months (Solbiati et al., 2004b). The subgroup consisting of patients treated after January 2000 had survival rates at 1, 2, and 3 years of 94.5%, 79.5%, and 59.7%, respectively. This 3-year survival rate appears to approach the survival rates reported in the surgical literature after hepatic resection (Fong et al., 1999). A second investigator at the same meeting reported similar results with a 5-year survival rate of 24.1% after RF ablation of hepatic colorectal metastases (Lencioni 2004a,b).
53–Radiofrequency Ablation
937
Figure 4. Perivascular RF treatment failure. (a) Pre-procedure scan demonstrates metastasis (block arrow) adjacent to the inferior vena cava (IVC, arrow). (b) Intraprocedural scan demonstrates RF electrode (block arrow) several centimeters from the IVC, but precisely centered within the tumor. (c) At 1-month follow-up, the ablation zone (block arrow) appears to adequately cover the entire tumor. (d) Three-month follow-up scan shows local recurrence (block arrow) on medial aspect of ablation zone next to the IVC (arrow). Blood flow in the IVC was enough to cool and preserve perivascular tumor despite successful ablation of the remainder of the metastasis.
Figure 5. Intraprocedural ultrasound shows ablation zone being formed with three electrodes (small arrows) in place. The hyperechoic region (block arrow) corresponds to microbubbles being formed. This area only roughly correlates with the true size of the ablation zone and obscures visualization of the tumor during subsequent placement of electrodes making sequential, overlapping ablations difficult. This ablation was performed with a multiple-electrode system developed to overcome this limitation.
938
Christopher L. Brace and Paul F. Laeseke
Lung Lung cancer is the most common cause of cancer death in both sexes worldwide, and current trends in smoking indicate that it will remain so in the coming decades. Furthermore, lung metastases have been found at autopsy in almost one-third of all people with malignant disease. Surgery has become an accepted treatment for lung tumors in selected patients with an approximately 25% 10-year survival rate. However, a large number of patients have unresectable tumors or comorbidities precluding surgery, or refuse surgery. RF ablation is a promising therapy for this group of people. In theory, lung tumors are well suited to RF ablation because the surrounding air in adjacent normal parenchyma provides an insulating effect facilitating energy concentration within the tumor tissue (the oven effect). On the other hand, air increases circuit impedance levels, which reduces the amount of power that can be deposited into aerated lung, and precludes ultrasound guidance. To date, over 500 patients have been treated for primary lung tumors (e.g., non-small-cell lung cancer) or lung metastases (Steinke et al., 2004). While no long-term data are available to determine what impact RF lung ablation will have on the survival of patients, early results are promising, given that experience is limited and the technology has not been optimized for the unique environment of the lung. Full tumor destruction has been achieved in the vast majority of procedures, and the rate of minor complications (small pneumothorax, pleural effusion, or hemorrhage) is less than 30%. The rate of pneumothoraces requiring a chest tube is 10% to 30%, and the rate of pleural effusions requiring drainage is less than 10%. Major complications are rare but several have been reported, including a massive pulmonary hemorrhage and several deaths. Kidney A minimally invasive treatment option for small renal cell cancer (RCC) is desirable because of the increased number of these tumors being diagnosed by cross-sectional imaging (CT, ultrasound, MRI). Many of these tumors are asymptomatic and are incidentally detected on scans that were performed for a different complaint. Many patients are elderly, or have other underlying conditions, which put them at increased risk from complications of conventional open surgery. Interest in a minimally invasive treatment has been increasing now that partial nephrectomy has proven as effective as radical nephrectomy for treatment of small RCC (Novick 2004). Partial nephrectomy, like ablation, has the advantage of preserving much of the functioning renal tissue, lessening the chances of post-procedure renal failure. When treating kidneys with percutaneous ablation, the retroperitoneal location minimizes the risk of major bleeding, while the exophytic location of many renal tumors decreases the chances of major complications from injury to the central collecting system. Central tumors are more difficult to treat due to the heat sink effect of large vessels within the renal hilum. Indications for renal tumor ablation are: a prior partial or total nephrectomy, pre-existing renal insufficiency, various co-morbidities making the patient a high surgical risk, or syndromes with multiple tumors (e.g., von Hippel-Lindau). Renal tumor ablation is widely considered a safe procedure. Hemorrhage is the most common complication, but it is most often minor and self-limiting. Major complications are quite rare (<2%) and include significant hemorrhage, ureteral stricture, and anesthesia-related adverse effects. Several more years of patient follow-up are needed to determine the impact on patient survival, since renal cell carcinomas have low biologic activity. Bone The original application of RF ablation in the musculoskeletal system was for the treatment of facet joint osteoarthritis. More recently, it has become a common treatment for osteoid osteomas (small benign tumors of bone that are often painful and usually occur in the extremities of children and young adults) and, in many centers, it has now become the treatment of choice (Cioni et al., 2004). It is also used to relieve pain from bone metastases (Poggi et al., 2003; Goetz et al., 2004). The therapeutic effect of RF
53–Radiofrequency Ablation
939
ablation is thought to be related to debulking viable tumor and inhibiting the release of local inflammatory modulators. Symptoms from painful metastases often resolve immediately after RF treatment and post-treatment pain can usually be controlled with oral analgesics and nonsteroidal anti-inflammatory medications. In addition to local control of the tumor mass, RF ablation may achieve an improvement in tumor markers and paraneoplastic syndromes. Experience has shown improved pain and quality of life scores after ablation. Major complications are rare, and the addition of methylmethacrylate cement can help stabilize vertebral bodies and other lytic tumors in the pelvis. Orthopedic stabilization may be needed for tumors involving the femur and tibia. RF ablation is also an established method for treating certain conditions outside of the field of oncology. It is used as a minimally invasive way to treat cardiac arrhythmias such as atrial fibrillation, atrial flutter, and supraventricular tachycardia. The cardiac ablation catheter is inserted into a blood vessel and threaded to the heart where it is used to ablate tissue causing the abnormal rhythms. Furthermore, RF ablation has also been used in the symptomatic treatment of patients with Parkinson’s disease. Thalamotomies and pallidotomies can provide significant improvement in the disabling symptoms of tremor, rigidity, and bradykinesia.
Clinical Research Directions RF ablation is currently being investigated in several other organ systems. It is a promising new technology for the treatment of prostatic carcinoma that has proven to induce extensive necrosis. However, follow-up is too short to determine its place in the treatment of prostate cancer. RF has been used to treat tumors in the head and neck, including medullary carcinoma of the thyroid. Preliminary work in which RF ablation was followed by standard surgical resection has shown this technique is effective for ablation of small (≤2 cm) primary breast tumors. Finally, early experiences suggest that RF ablation is effective for local control of adrenal neoplasms; however, it has been associated with hypertensive crisis and will require further evaluation to determine both the efficacy and safety of the procedure (Chini et al., 2004). Greater experience with RF ablation and more sophisticated RF systems are leading to a more aggressive approach to the tumors being treated. While RF ablation is considered a relatively safe procedure, it does have associated complications especially when treating peripheral tumors. Performing ablations close to perihepatic structures can result in thermal damage to them. Injury to the body wall or diaphragm can result in prolonged, severe pain while injury to the bowel can lead to perforation and life-threatening sepsis or abscess formation. Several different methods have been investigated to separate the liver from surrounding structures prior to ablation, thereby minimizing risk of thermal injury. These include balloons, CO2, and fluids [sterile water, saline, and 5% dextrose (D5W) in water]. Fluids are easier to place than balloons and are sonolucent, unlike CO2, allowing for ultrasound guidance. Of the fluids, D5W is advantageous in that it is iso-osmolar (will not cause fluid shifts, unlike sterile water) and is nonionic, thereby limiting conductance of the RF current (unlike saline). Preliminary data shows a greater protective effect with intraperitoneal D5W compared to saline (Laeseke et al., 2005). Furthermore, intraperitoneal D5W has been shown to decrease post-procedural pain scores and patient-controlled analgesia (Hinshaw et al., 2005). Figure 6 illustrates the use of D5W in separating the liver and colon. Efforts to decrease local recurrence rates and to increase survival rates have focused on increasing ablation zone size and combining RF ablation with adjuvant therapies. Increasing ablation zone size has largely involved modifications to electrode design. In addition to these advancements, infusion of hypertonic saline has been used to significantly increase ablation zone size (Ahmed et al., 2002; Lee et al., 2004. However infusion of hypertonic saline may limit the predictability of the ablation zone. Coagulation and intratumoral drug accumulation have also been increased by combining RF ablation and IV liposomal doxorubicin (Goldberg et al., 2001; Ahmed and Goldberg 2004). More research is necessary to determine the increased effectiveness of RF ablation in combination with chemotherapy or radiation.
940
Christopher L. Brace and Paul F. Laeseke
Figure 6. Separating liver and colon with 5% dextrose in water (D5W). (a) Pre-procedure CT scan shows esophageal metastasis adjacent to the colon. (b) Intraprocedure CT scan shows D5W (arrow) being instilled into the peritoneal cavity via a needle between the tumor and colon. A separation of several millimeters was achieved. (c) Post-procedure scan shows successful ablation with no enhancement. The procedure was performed without complications or injury to the bowel.
Principles of RF Ablation RF Heating in Tissues Joule Heating by RF in Tissue For RF ablation, cell death is achieved from irreversible thermal damage. Such damage occurs for tissue temperatures above 50 °C when the temperature is maintained for at least 4 to 6 minutes, and nearly instantaneously for temperatures above 60 °C. Heat is generated inside tissues through joule heating, a process by which electrical current passing through a conductive medium is converted to thermal energy. In RF ablation, current is conducted from the electrode, through tissues, to the ground pads (Figure 2). Thermal energy transfer in RF ablation is governed by the transient heat-transfer equation:
ρC
∂T ∂t
= ∇ ⋅ k∇T + J ⋅ E − Q p ,
(1)
where ρ is material density (kg/m3), C is specific heat (J/kg·K), T is temperature (K), t is time (s), k is thermal conductivity (W/kg·m3), J is the current density vector (A/m2), E is the electric field vector (V/m), and Qp represents additional heat transfer, or Q p = wbl Cbl ρbl ρ (T − Tbl ) ,
(2)
where wbl is local perfusion rate (mL/min/kg), Cbl is the specific heat of blood (J/kg·K), rbl is blood density (kg/m3), and Tbl is blood temperature (K). Upon examination of equation (1), it is clear that the heat generation term is J·E. Recalling that J = σE, where a is electrical conductivity (S/m), the heat generation term can be written simply as
53–Radiofrequency Ablation
941
Q p = wbl Cbl ρbl ρ (T − Tbl ) .
(3)
Active heating (J·E) causes a large temperature gradient (∇T) to build near the electrode, which moves radially outward provided that (1) energy is continually deposited into the tissue and (2) thermal conduction is greater than perfusion losses (∇ · k∇T > Qp). Active heating only occurs within a few millimeters of the electrode surface because current density and power decrease as 1/r2 and 1/r4, respectively (Organ 1976). The majority of the ablation zone is heated by thermal conduction (∇ · k∇T). A computed timeprogression of heating shows that the most active heating occurs near the distal and proximal tips of a single electrode (due to high current densities) and that heat gradually transfers into the tissue over the ablation time (Figure 7). Tissue Properties The highly temperature- and frequency-dependent electrical and thermal properties of tissue have been explored in the literature (Duck 1990). In current RF ablation systems, the frequency of operation is held constant so that when modeling or designing components, only the temperature-dependent properties are considered. Of particular importance is the change in electrical conductivity with temperature. An electrical path is provided in tissues by the electrolytic extracellular fluid inherent to tissue. The rapid decrease in electrical conductivity during a heating cycle can be attributed to dehydration of, and physical changes within, tissues. Tissue conductivity actually increases while heating from 37 °C to 78 °C when heated in a water bath (Pop et al., 2003). As tissue temperature advances beyond 95 °C, however, liquid water is quickly converted to steam. The expulsion of liquid begins to reduce the conductivity of the tissue. When water is completely removed, charring can occur. The eschar formed effectively eliminates conduction unless rehydrated. The effect of this sudden drop in conductivity is that electrode temperature or circuit impedance (~1/σ) must be monitored during a procedure to avoid charring. Temperature and impedance feedback will be discussed in the “RF Ablation Equipment” section. Other tissue properties and their temperature coefficients are given in Table 1. Zones of Ablation The zone of ablation is the volume of tissue around the electrode that is heated to lethal temperatures. In practice, the ablation zone has two major components: an area of complete necrosis and an area of partial necrosis (Figure 8). All cells inside the area of complete necrosis are assumed to be necrotic but viable
Figure 7. Current density and time-progression of a single-electrode ablation. Current density is greatest at the distal and proximal ends of the electrode. The final zone of ablation is a result of thermal conduction from the electrode.
942
Christopher L. Brace and Paul F. Laeseke Table 1. Electrical and thermal properties (when known, the temperature coefficient is given in %/ °C) of tissues at 460 kHz. (Data from Duck 1990 and Gabriel, Lau, and Gabriel 1996). Electrical conductivity σ (S/m)
Relative permissivity εr
Specific heat C (J/kg·K)
Thermal conductivity k (W/kg·m3)
Density ρ (kg/m3)
Bone
0.022
175.6
440
0.233 (+ 0.26)
1900
Heart
0.278 (+ 20.30)
3356 (– 0.10)
3700
0.585 (+ 0.24)
1060
Kidney
0.226 (+ 1.3)
3521 (– 0.50)
3900
0.543 (+ 0.26)
1050
Liver
0.146 (+ 1.3)
2861 (– 0.20)
3300
0.564 (+ 0.25)
1050
Lung
0.122
1045
3000
0.407 (+ 0.95)
240
0.443 (+ 0.302)
3772 (– 0.10)
3700
0.518 (+ 0.24)a
1030
Tissue
Muscle a
Porcine estimate
cells remain in the outlying area of partial necrosis, hemorrhage, and congestion. This has been confirmed by examination of histopathology (thomsen 1991; Shock et al., 2004). In general, the most necrotic areas will be nearest to the electrode, since this is the area where most heat is generated. For a single electrode, the zone of ablation resembles an ellipsoid: nearly circular in the transverse plane, but elliptical in the longitudinal plane (Figure 7). State-of-the-art single electrodes are able to ablate areas ~2.0 cm in diameter and ~3.5 cm long in vivo. Efforts have been made to make ablation zones as spherical as possible because tumors are generally spherical, but ablation zone shape depends largely on the electrode design and tissue environment. Perfusion is always the limiting factor for ablation zone size and shape. The largest zones of ablation are often created with surgical techniques that cut off blood flow during heating (Rossi et al., 1999; Chinn et al., 2001; Lu et al., 2002). These techniques are not consistent with minimally invasive procedures; therefore, research efforts are focused on increasing the zone of ablation to treat large tumors without surgical intervention. Several overlapping ablations are most commonly used to treat large tumors but this technique can be time-consuming and technically difficult (Figure 9). The number of overlapping ablations needed to treat a volume larger than that which can be ablated with a single device increases dramatically with tumor size (Khajanchee et al., 2004). The “RF Ablation Equipment” section will discuss advances in electrode and system design that have led to clinically useful RF ablation systems.
Drawbacks to RF Treatment Despite the successes of RF ablation in recent years, there are lingering drawbacks inherent to the joule heating mechanism. One is the possibility of superficial burns beneath the ground pads, which must be placed on the patient to provide a closed electrical circuit for current flow. From equation (3), it is clear that if the current density, J, becomes too large beneath the ground pad, enough heat may be generated to cause mild to severe burns. Ground pads are designed to disperse the current over a wide area, avoiding sharp corners where current density may stack up, to combat superficial heating. Multiple pads may be used to increase the total pad area, but proper placement is critical. Pads placed nearest to the electrode generally create the lowest-resistance path for current flow, collect a greater amount of current, and are more likely to cause burns. Another obstacle in RF ablation is the rapid decrease in tissue conductivity with increased temperature and dehydration. If the tissue becomes sufficiently dehydrated (a hallmark of rapid heating), the
53–Radiofrequency Ablation
Figure 8. Zone of ablation with well-defined areas of (a) complete necrosis and (b) partial necrosis.
Figure 9. Overlapping multiple ablations are required to treat many tumors.
943
944
Christopher L. Brace and Paul F. Laeseke
conductivity drops to zero, current cannot flow and, from equation (3), heat is no longer generated. Generally, power is turned off for some time while the tissue cools and re-hydrates, but this tends to increase treatment time. This downtime provided the incentive for multiple-electrode systems, to be discussed later. Infusing saline through the electrode has been investigated to reduce charring effects, but this technique has its own pitfalls, including reduced control of the ablation zone and increased complication rates (Livraghi et al., 1997). Finally, current RF ablation systems are plagued by relatively high local recurrence rates, particularly in perivascular locations due to the temperature-limiting “heat-sink effect” that large vessels provide (Bilchik et al., 1999; McGahan and Dodd 2001; Lu et al., 2002). This effect may be overcome by faster heat generation (J·E >> Qp), but has yet to be realized or proven in RF ablation. Higher-power generators and innovations in electrode design may help reduce local recurrence rate (see the next section).
RF Ablation Equipment Four companies currently have FDA 510k approval for ablation devices: Berchtold (Tuttlingen, Germany), Boston Scientific (Natick, MA), RITA Medical Systems (Mountain View, CA), and Valleylab (Boulder, CO). Each of these companies offers an RF generator operating at either 375 kHz (Berchtold) or 460 kHz (Boston Scientific, RITA, Valleylab) with variable output powers and their own selections of electrodes and ground pads, which will be discussed in this section.
Electrodes Electrode design plays a key role in the speed, size, and quality of an RF ablation. A general survey of electrode types, their properties, and common traits will be discussed and references will be provided for more advanced information. Single Electrode Designs The simplest electrode design is that of a single, current-carrying conductor (electrode) inserted into the liver via an electrically insulated shaft (Figure 10a). When an electric potential is applied, current flows from the electrode to the ground pads. The highest current densities are near the distal tip and proximal base of the electrode and heat is generated within 1 to 2 mm of the electrode surface. Electrode surface area is inversely proportional to the impedance of the RF circuit and is determined from the length and diameter of the electrode. Thus, for a given diameter (17 gauge [1.5 mm] is most common for single electrodes) and input current, a longer electrode provides a lower impedance path but reduces the current density around the electrode. Commercially available generators with single electrodes are able to create ~1.6 × 4 cm elliptical zones of ablation in vivo (Goldberg et al., 1995). Larger generators would likely be able to power slightly longer (or larger diameter) electrodes. Charring is very problematic for these electrodes. Cooled electrode: The long, narrow ablation zone created by a single electrode is not ideal for treating tumors, which are usually spherical. A cooled electrode and shaft were introduced to use the available energy more efficiently (Goldberg et al., 1996; Lorentzen 1996). Cooling the electrode moderates the amount of unwanted shaft heating and reduces the amount of charring near the electrode, thereby increasing the total amount of RF energy deposited during the ablation. Thus, cooled electrodes are able to ablate more spherical, ~2.0 × 3.5 cm volumes in vivo. Ground pads are required for these electrodes to operate. The Valleylab Cool-tip™ series is based on this concept. Infusion electrode: Infusing hypertonic saline through the electrode into the tissue can also reduce the likelihood of charring and increase the zone of ablation (Mittleman et al., 1995; Shake et al., 1997). The infused electrolytic solution keeps a reliable conduction path between the electrode and surrounding
53–Radiofrequency Ablation
945
Figure 10. Commercially available electrode designs: (a) single Cool-tip™; (b) Cool-tip™ cluster; (c) RITA Starburst™; (d) Boston Scientific LeVeen™.
tissue open. By keeping the tissue hydrated, charring is nearly eliminated. Zones of ablation ~4 cm in diameter have been reported with a single infusion electrode, but the shape of the ablation zone is often uncontrollable (Livraghi et al., 1997; Pereira et al., 2004). Saline-infused electrodes may be characterized by higher complication rates than other needle electrode designs, but further study is needed (Boehm et al., 2002). RITA Starburst™ Xli and Berchtold® products employ saline infusion. Expandable Electrodes Both RITA and Boston Scientific offer expandable electrode systems (Figures 10b and 10c), though each system operates in a slightly different manner. A trocar with several retracted prongs is guided to the tumor site, at which point the prongs are protracted into the tissue. In general, prongs are operated in an equipotential (same input voltage) fashion with ground pads used to collect current from all prongs simultaneously. The shape of the expandable arrays and amount of expansion determines the total zone of ablation shape and size. Procedure times are generally longer (~30 minutes) than single electrode designs, but have more reproducible results and are less susceptible to blood flow effects. Since most expandable electrodes are 14 gauge (2.1 mm) in diameter, they are more popular for open surgical use. Some minor complications related to tissue charring and poor retraction with expandable systems have been reported in the lung, but have not affected performance or use in other organ systems (Steinke et al., 2003). Multiple Electrodes Bipolar: In all previous electrode designs, current flows from the electrode to the ground pads. However, an interstitial ground return can be used to eliminate the need for pads. When a voltage is applied to one electrode while another electrode is grounded, current flows directly between the electrodes; the electrodes are operated in bipolar mode (McGahan et al., 1996). The grounded electrode may be on the same applicator for a single insertion or on a second applicator altogether. Many electrodes may be operated in this
946
Christopher L. Brace and Paul F. Laeseke
manner by switching the active and grounded electrodes during ablation, but at a given time only two electrodes are active. Bipolar operation has the advantage of direct current flow and heating in a specific area. Heat can be generated faster and higher temperatures can be achieved within the ablation zone (Figure 11). Such benefits may make this mode well suited for ablation of vascular areas. Depending on how the electrodes are inserted, this mode can be more invasive than single-electrode ablation. Currently, if a single insertion is used, then control of the individual current levels and feedback from each electrode is not possible. If multiple insertion points are used, then the total invasiveness and procedure complexity is increased. These limitations may be acceptable for certain procedures (Haemmerich et al., 2003). Equipotential: It is also possible to operate several electrodes at the same electrical potential, provided the inter-electrode spacing is small enough to allow heat generated by each electrode to overlap with the adjoining electrodes. This is the principle behind the expandable electrodes discussed previously and the Cool-tip™ cluster electrode (Figure 10d). It has been shown that an array of equipotential single electrodes creates a repulsive force in the center of the array, forcing current to flow to the periphery of the array (Haemmerich et al., 2005). Thus, equipotential electrodes are undesirable unless the spacing is small enough to allow overlap of thermal gradients, or large enough to operate independently. Zones of ablation ~3.0 cm have been reported with clustered electrodes (Goldberg et al., 1998; Pereira et al., 2004). Switched: As mentioned earlier, single electrodes are likely to cause impedance (or temperature) spikes that require power shutdown and a period of cooling before ablation can resume. This provided an impetus for switched, multiple-electrode RF ablation (Lee et al., 2002). Rather than shutting power off completely, power is switched to another electrode. Several electrodes may be operated in this manner to ablate large tumors or multiple tumors simultaneously. Each electrode is electrically independent, so the repulsive force of equipotential electrodes is eliminated. Ablation zone diameters approaching 5 cm have been realized using switched RF ablation. RF switching is currently an area of active and fruitful ablation research and has recently become commercially available with the Valleylab Switching Controller™ (Lee et al., 2003; Haemmerich et al., 2005).
Grounding Pads To complete the RF ablation circuit and allow current flow, a path to ground must be provided. The most common method is to place grounding pads (a.k.a., dispersive electrodes) onto the skin of the patient. The
Figure 11. Expandable-type electrodes operated in (a) monopolar and (b) bipolar modes. Note the greater amount of heating between the bipolar electrodes and resulting large zone of ablation.
53–Radiofrequency Ablation
947
pads contain a conducting adhesive that keeps them in physical and electrical contact with the body. They are connected to the ground side of the generator and close the RF ablation circuit loop. Ground pads are problematic for several reasons; they (1) add additional complexity to the ablation procedure, (2) require additional preparation (shaving) of the application site, and (3) must collect current in a limited area, which can lead to high enough current densities to heat the underlying skin and cause burns. For these reasons, some manufacturers offer impedance monitoring or thermal sensing of the ground pads to ensure proper placement and to avoid skin burns. Certain electrode designs discussed previously do not require ground pads and may be better suited to use in tissues that are poor electrical conductors (e.g., bone, lung).
RF Generator The role of the RF generator is to supply energy to the ablation circuit. In commercial generators, feedback, control, and pulsed power techniques are used to couple energy into the tissue efficiently and without charring. The engineering behind RF generators is beyond the scope of this chapter but advanced control and application of power will be covered. Feedback and Control Each commercial vendor uses a slightly different feedback method to avoid overheating the tissue. Valleylab and Boston Scientific systems continuously measure the impedance of the RF circuit. In the Valleylab system, power is turned off when the impedance level of an electrode increases significantly (e.g., ~3 Ω) above some baseline value. Such impedance spikes are characteristic of tissue on the verge of charring. Power is removed from the electrode to allow the surrounding tissue to cool and rehydrate, then is reapplied a short time later. This pulsed-application continues for the duration of the procedure (typically 12 to 15 minutes). Boston Scientific systems use impedance feedback to define the endpoint of the ablation. Power is applied continuously until Roll-off®—when the impedance begins to spike. The procedure is either terminated at this point or power automatically decreased for the remainder of the ablation, which is typically ~30 minutes total. RITA systems use temperature feedback to control power application. Each prong of the expandable array has a thermocouple at the tip. Temperatures are continuously monitored for each prong, and when the average temperature of the prongs reaches a user-defined level (e.g., 100 °C), power adjusted for the remainder of the procedure (typically ~30 minutes). The user may also elect to control for power or treatment time. Advanced Power Application The size of an ablation zone is directly related to the amount of RF energy delivered to the tissue and how that energy is distributed. Recall that power is the time-derivative of energy; higher power means energy is deposited in less time. Maximum commercial generator output power is 200 to 250 W, depending on manufacturer and application. To maximize the zone of ablation and minimize the risk of bleeding after a procedure, advanced power application techniques have been developed. Pulsing: Using feedback and control algorithms or fixed time intervals, it has been shown that turning power on and off during the ablation uses the available power more efficiently and creates larger zones of ablation. This is because overheating of tissue causes the conduction path and active heating to be eliminated. Since temperatures above 60 °C do not kill cells much faster than at 60 °C (near-instantaneous), the tissue may be allowed to cool from 100 °C without affecting the efficacy of the procedure. The cooling process regenerates the necessary conduction path required for heating. Cautery: When extracting the electrode after ablation, clinicians use a cautery technique to reduce the likelihood of bleeding in the needle tract or seeding of viable tumor cells from the electrode itself. A
948
Christopher L. Brace and Paul F. Laeseke
simple procedure of manually ablating the entire tract to 90 °C is generally considered sufficient. Some generators offer a separate cauterization mode of operation.
Future Directions The future of RF ablation is being drawn up in research laboratories today. Many recent developments (e.g., switched electrodes, adjuvant therapies) have already been discussed that will allow RF ablation to be faster and more effective, treat larger tumors, and be applied for new applications. New advances in imaging, contrast agents, and computer modeling will also expand the horizons for RF ablation. This section deals with other areas of ongoing research.
Imaging and Navigation The feasibility of tracking the RF electrode, before and during ablation, has recently been demonstrated (Varro, Locklin, and Wood 2004; Viswanathan et al., 2005). In the technique proposed by Varro et al., a laser-positioning device attached to the CT gantry is used to align the electrode correctly before insertion based on pre-ablation images and planning. The laser allows a precise insertion angle to be maintained, while the insertion depth is tracked on the electrode shaft. Thus, the electrode may be correctly placed without additional imaging guidance. Viswanathan et al., describe a similar concept, but make use of coiled RF markers inside the electrode to communicate position to a GPS-like system of radiotransducers placed at landmarks on the body. These landmarks are correlated with pre-ablation imaging scans so that the electrode’s anatomical position is indicated by the RF. This technique also allows the tumor to be landmarked and the ablation procedure planned and carried out without additional imaging guidance. Finally, virtual sonography—a technique that combines CT and ultrasound to create real-time volumetric images—is currently being explored for use with ablative technologies.
Equipment In addition to the technologies discussed already, high-power generators and optimization of current systems are being actively researched. The primary goals of recent device research have been to create larger zones of ablation in less time to broaden the scope of RF ablation. Optimization of multipolar and switched RF ablation may achieve both of these goals (Lee et al., 2003). The higher heat generated by each method may also reduce the local recurrence rate, particularly in perivascular regions subject to the “heat-sink effect.” New ground pad designs have been proposed to shoulder the burden of increased energy deposition by spreading out the return current, thereby reducing the current density (Tungjitkusolmun et al., 2000). Finally, all electrode designs, particularly those that utilize the full generator duty cycle (expandable, multiple-electrode) may benefit from higher input powers. Whether large input powers can increase the ablation zone size, decrease treatment time, or operate more than three electrodes simultaneously remains to be seen.
Monitoring and Evaluation Ablation monitoring and post-ablation imaging are both under continual development. Contrast-enhanced CT and ultrasound may decrease the local recurrence rate by allowing physicians to see and completely treat viable tumor immediately post-ablation (Meloni et al., 2001). In Meloni et al., contrast-enhanced ultrasound was found to have 83.3% power at detecting viable tumor post-ablation, with the advantages of imaging without ionizing radiation and using the same imaging system as used for electrode guidance. Contrast-enhanced CT was found to be the most sensitive (100%), but requires the use of x-rays, longer
53–Radiofrequency Ablation
949
imaging times, and is more cumbersome to operate. Solbiati et al., found that using contrast-enhanced ultrasound reduced the local recurrence rate from 16.1% to 5.9% (Solbiati et al., 2004a). Ultrasound elastography and MR thermometry are two techniques under investigation to monitor progress during ablation. Ultrasound elastography is a technique by which the elasticity parameters of tissue are imaged with ultrasound. These parameters are correlated to the temperature of tissue and may be used to determine the zone of ablation. Ultrasound elastography has recently been shown to make more accurate determinations of ablation zone volume than conventional CT (Lui et al., 2004). A similar MR elastography technique has also been explored but requires greater cost and imaging time (Wu et al., 2001). Monitoring of ablation temperature is possible with MR thermometry. Temperature changes may be mapped based on relaxation time (T1), resonance frequency shift of water protons, or diffusion coefficient. Temperature-sensitive contrast agents may also be used. Such techniques allow 1 °C changes in temperature to be imaged with 1 to 2 mm spatial resolution and 1 to 2 s temporal resolution per slice (Quessson, de Zwart, and Moonen 2000). Imaging time and a requirement for MR-compatible equipment are the major drawbacks.
Conclusions While RF ablation technology is relatively new, its role in cancer treatment has already been established. Advances in medical imaging and RF ablation technology will improve the outcomes for existing indications. Moreover, its role in other organ systems will continue to expand. The use of adjuvant therapies will also facilitate the advancement of RF ablation as a treatment option for tumors beyond the capabilities of current systems. There is little doubt that with the current cancer incidence rates and the aging population, advancements in RF ablation will play a key role in reducing the significant burden cancer places on both patients and society.
References Ahmed, M., and S. N. Goldberg. (2004). “Combination radiofrequency thermal ablation and adjuvant IV liposomal doxorubicin increases tissue coagulation and intratumoural drug accumulation.” Int J Hyperthermia 20(7):781–802. Ahmed, M., S. M. Lobo, J. Weinstein, J. B. Kruskal, G. S. Gazelle, E. F. Halpern, S. K. Afzal, R. E. Lenkinski, and S. N. Goldberg. (2002). “Improved coagulation with saline solution pretreatment during radiofrequency tumor ablation in a canine model.” J Vasc Interv Radiol 13(7):717–724. Bilchik, A. J., D. M. Rose, D. P. Allegra, P. J. Bostick, E. Hsueh, and D. L. Morton. (1999). “Radiofrequency ablation: A minimally invasive technique with multiple applications.” Cancer J Sci Am 5(6):356–361. Boehm, T., A. Malich, S. N. Goldberg, J. R. Reichenbach, I. Hilger, P. Hauff, M. Reinhardt, M. Fleck, and W. A. Kaiser. (2002). “Radio-frequency tumor ablation: Internally cooled electrode versus saline-enhanced technique in an aggressive rabbit tumor model.” Radiology 222(3):805–813. Cha, C. H., F. T. Lee Jr., J. M. Gurney, B. K. Markhardt, T. F. Warner, F. Kelcz, and D. M. Mahvi. (2000). “CT versus sonography for monitoring radiofrequency ablation in a porcine liver.” Am J Roentgenol 175(3):705–711. Chini, E. N., M. J. Brown, M. A. Farrell, and J. W. Charboneau. (2004). “Hypertensive crisis in a patient undergoing percutaneous radiofrequency ablation of an adrenal mass under general anesthesia.” Anesth Analg 99(6):1867–1869, table of contents. Chinn, S. B., F. T. Lee Jr., G. D. Kennedy, C. Chinn, C. D. Johnson, T. C. Winter 3rd, T. F. Warner, and D. M. Mahvi. (2001). “Effect of vascular occlusion on radiofrequency ablation of the liver: results in a porcine model.” Am J Roentgenol 176(3):789–795. Cioni, R., N. Armillotta, I. Bargellini, V. Zampa, C. Cappelli, P. Vagli, G. Boni, S. Marchetti, V. Consoli, and C. Bartolozzi. (2004). “CT-guided radiofrequency ablation of osteoid osteoma: Long-term results.” Eur Radiol 14(7):1203–1208.
950
Christopher L. Brace and Paul F. Laeseke
Curley, S., P. Marra, K. Beaty, L. Ellis, J. Vauthey, E. Abdalla, C. Scaife, C. Raut, R. Wolff, H. Choi, E. Loyer, P. Vallone, F. Fiore, F. Scordino, V. De Rosa, R. Orlando, S. Pignata, B. Daniele, and F. Izzo. (2004). “Early and late complications after radiofrequency ablation of malignant liver tumors in 608 patients.” Ann Surg 239:450–458. Davis, K. W., J. J. Choi, and D. G. Blankenbaker. (2004). “Radiofrequency ablation in the musculoskeletal system.” Semin Roentgenol 39(1):129–144. Decadt, B., and A. K. Siriwardena. (2004). “Radiofrequency ablation of liver tumours: Systematic review.” Lancet Oncol 5(9):550–560. Duck, F. A. Physical Properties of Tissue: A Comprehensive Reference Book. London: Academic Press, 1990. Farrell, M. A., W. J. Charboneau, D. S. DiMarco, G. K. Chow, H. Zincke, M. R. Callstrom, B. D. Lewis, R. A. Lee, and C. C. Reading. (2003). “Imaging-guided radiofrequency ablation of solid renal tumors.” Am J Roentgenol 180(6):1509–1513. Fong, Y., J. Fortner, R. L. Sun, M. F. Brennan, and L. H. Blumgart. (1999). “Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: Analysis of 1001 consecutive cases.” Ann Surg 230(3):309–318; discussion 318–321. Fornage, B. D., N. Sneige, M. I. Ross, A. N. Mirza, H. M. Kuerer, B. S. Edeiken, F. C. Ames, L. A. Newman, G. V. Babiera, and S. E. Singletary. (2004). “Small (< or = 2-cm) breast cancer treated with US-guided radiofrequency ablation: Feasibility study.” Radiology 231(1):215–224. Gabriel, S., R. W. Lau, and C. Gabriel. (1996). “The dielectric properties of biological tissues: II. Measurements in the frequency range 10 Hz to 20 GHz.” Phys Med Biol 41(11):2251–2269. Goetz, M., M. Callstrom, J. Charboneau, M. Farrell, T. Maus, T. Welch, G. Wong, J. Sloan, P. Novotny, I. Petersen, R. Beres, D. Regge, R. Capanna, M. Saker, D. Gronemeyer, A. Gevargez, K. Ahrar, M. Choti, T. de Baere, and J. Rubin. (2004). “Percutaneous image-guided radiofrequency ablation of painful metastases involving bone: A multicenter study.” J Clin Oncol 22:300–306. Goldberg, S. N., G. S. Gazelle, S. L. Dawson, W. J. Rittman, P. R. Mueller, and D. I. Rosenthal (1995). “Tissue ablation with radiofrequency: Effect of probe size, gauge, duration, and temperature on lesion volume.” Acad Radiol 2(5):399–404. Goldberg, S. N., G. S. Gazelle, L. Solbiati, W. J. Rittman, and P. R. Mueller. (1996). “Radiofrequency tissue ablation: Increased lesion diameter with a perfusion electrode.” Acad Radiol 3(8):636–644. Goldberg, S. N., L. Solbiati, P. F. Hahn, E. Cosman, J. E. Conrad, R. Fogle, and G. S. Gazelle. (1998). “Large-volume tissue ablation with radio frequency by using a clustered, internally cooled electrode technique: Laboratory and clinical experience in liver metastases.” Radiology 209(2):371–379. Goldberg, S. N., P. F. Saldinger, G. S. Gazelle, J. C. Huertas, K. E. Stuart, T. Jacobs, and J. B. Kruskal. (2001). “Percutaneous tumor ablation: increased necrosis with combined radio-frequency ablation and intratumoral doxorubicin injection in a rat breast tumor model.” Radiology 220(2):420–427. Goldberg, S. N., J. W. Charboneau, G. D. Dodd 3rd, D. E. Dupuy, D. A. Gervais, A. R. Gillams, R. A. Kane, F. T. Lee Jr., T. Livraghi, J. P. McGahan, H. Rhim, S. G. Silverman, L. Solbiati, T. J. Vogl, and B. J. Wood. (2003). “Image-guided tumor ablation: Proposal for standardization of terms and reporting criteria.” Radiology 228(2):335–345. Haemmerich, D., A. W. Wright, D. M. Mahvi, F. T. Lee Jr., and J. G. Webster. (2003). “Hepatic bipolar radiofrequency ablation creates coagulation zones close to blood vessels: A finite element study.” Med Biol Eng Comput 41(3):317–323. Haemmerich, D., F. T. Lee Jr., D. J. Schutt, L. A. Sampson, J. G. Webster, J. P. Fine, and D. M. Mahvi. (2005). “Largevolume radiofrequency ablation of ex vivo bovine liver with multiple cooled cluster electrodes.” Radiology 234(2):563–568. Hayashi, A. H., S. F. Silver, N. G. van der Westhuizen, J. C. Donald, C. Parker, S. Fraser, A. C. Ross, and I. A. Olivotto. (2003). “Treatment of invasive breast carcinoma with ultrasound-guided radiofrequency ablation.” Am J Surg 185(5):429–435. Hinshaw, J. L., D. L. Chicks, P. F. Laeseke, T. C. I. Winter, and F. T. Lee Jr. (2005). “RF Ablation of Peripheral Liver Tumors: Intraperitoneal 5% Dextrose in Water (D5W) Decreases Post-procedure Pain.” Annual Meeting of the Society of Gastrointestinal Radiology, San Antonio, TX.
53–Radiofrequency Ablation
951
Jemal, A., T. Murray, E. Ward, A. Samuels, R. C. Tiwari, A. Ghafoor, E. J. Feuer, and M. J. Thun. (2005). “Cancer statistics, 2005.” CA Cancer J Clin 55(1):10–30. Khajanchee, Y., D. Streeter, L. Swanstrom, and P. Hansen. (2004). “A mathematical model for preoperative planning of radiofrequency ablation of hepatic tumors.” Surg Endosc 18:696–701. Kosari, K., M. Gomes, D. Hunter, D. J. Hess, E. Greeno, and T. D. Sielaff. (2002). “Local, intrahepatic, and systemic recurrence patterns after radiofrequency ablation of hepatic malignancies.” J Gastrointest Surg 6(2):255–263. Kuvshinoff, B. W., and D. M. Ota. (2002). “Radiofrequency ablation of liver tumors: Influence of technique and tumor size.” Surgery 132(4):605–611; discussion 611–612. Laeseke, P. F., L. A. Sampson, C. L. Brace, T. C. Winter 3rd, J. P. Fine, and F. T. Lee Jr. (2005). “Unintended thermal injuries from radiofrequency ablation: Protection with 5% dextrose in water.” Am J Roentgenol (in press). Lee, F. T., D. Haemmerich, J. G. Webster, A. S. Wright, C. D. Johnson, and D. M. Mahvi. (2002). Radio-frequency Ablation System Using Multiple Electrodes. Wisconsin Alumni Research Foundation (WARF) Technology License WARF-P01339US. Patent pending. Lee Jr., F. T., D. Haemmerich, A. S. Wright, D. M. Mahvi, L. A. Sampson, and J. G. Webster. (2003). “Multiple probe radiofrequency ablation: Pilot study in an animal model.” J Vasc Interv Radiol 14(11):1437–1442. Lee, J. M., J. K. Han, S. H. Kim, K. S. Shin, J. Y. Lee, H. S. Park, H. Hur, and B. I. Choi. (2004). “Comparison of wet radiofrequency ablation with dry radiofrequency ablation and radiofrequency ablation using hypertonic saline preinjection: ex vivo bovine liver.” Korean J Radiol 5(4):258–265. Lencioni, R. (2004a). “Tumor Radiofrequency Ablation Italian Network (TRAIN): Long-term Results in Epatic Colorectal Cancer Metastases.” Radiological Society of North America Scientific Assembly and Annual Meeting, Chicago. Lencioni, R. (2004b). “Tumor Radiofrequency Ablation Italian Network (TRAIN): Long-term Survival Outcomes in Patients with Stage 0 Hepatocellular Carcinoma.” Radiological Society of North America Scientific Assembly and Annual Meeting, Chicago. Lencioni, R., L. Crocetti, R. Cioni, A. Mussi, G. Fontanini, M. Ambrogi, C. Franchini, D. Cioni, O. Fanucchi, R. Gemignani, R. Baldassarri, C. A. Angeletti, and C. Bartolozzi. (2004). “Radiofrequency ablation of lung malignancies: Where do we stand?” Cardiovasc Intervent Radiol 27(6):581–590. Liu, J. B., B. B. Goldberg, D. A. Merton, N. M. Rawool, L. Needleman, and F. Forsberg. (2001). “The role of contrastenhanced sonography for radiofrequency ablation of liver tumors.” J Ultrasound Med 20(5):517–523. Liu, W., U. Techavipoo, T. Varghese, J. A. Zagzebski, Q. Chen, and F. T. Lee Jr. (2004). “Elastographic versus xray CT imaging of radio frequency ablation coagulations: An in vitro study.” Med Phys 31(6):1322–1332. Livraghi, T., S. N. Goldberg, F. Monti, A. Bizzini, S. Lazzaroni, F. Meloni, S. Pellicano, L. Solbiati, and G. S. Gazelle. (1997). “Saline-enhanced radio-frequency tissue ablation in the treatment of liver metastases.” Radiology 202(1):205–210. Livraghi, T., L. Solbiati, M. F. Meloni, G. S. Gazelle, E. F. Halpern, and S. N. Goldberg. (2003). “Treatment of focal liver tumors with percutaneous radio-frequency ablation: Complications encountered in a multicenter study.” Radiology 226(2):441–451. Lorentzen, T. (1996). “A cooled needle electrode for radiofrequency tissue ablation: Thermodynamic aspects of improved performance compared with conventional needle design.” Acad Radiol 3(7):556–563. Lu, D. S., S. S. Raman, D. J. Vodopich, M. Wang, J. Sayre, and C. Lassman. (2002). “Effect of vessel size on creation of hepatic radiofrequency lesions in pigs: Assessment of the ‘heat sink’ effect.” Am J Roentgenol 178(1):47–51. Machi, J., S. Uchida, K. Sumida, W. M. Limm, S. A. Hundahl, A. J. Oishi, N. L. Furumoto, and R. H. Oishi. (2001). “Ultrasound-guided radiofrequency thermal ablation of liver tumors: Percutaneous, laparoscopic, and open surgical approaches.” J Gastrointest Surg 5(5):477–489. McGahan, J. P. (sic: McGhana, J. P.), and G. D. Dodd 3rd (2001). “Radiofrequency ablation of the liver: Current status.” Am J Roentgenol 176(1):3–16. McGahan, J. P., P. D. Browning, J. M. Brock, and H. Tesluk. (1990). “Hepatic ablation using radiofrequency electrocautery.” Invest Radiol 25(3):267–270. McGahan, J. P., J. M. Brock, H. Tesluk, W. Z. Gu, P. Schneider, and P. D. Browning. (1992). “Hepatic ablation with use of radio-frequency electrocautery in the animal model.” J Vasc Interv Radiol 3(2):291–297. McGahan, J. P., W. Z. Gu, J. M. Brock, H. Tesluk, and C. D. Jones. (1996). “Hepatic ablation using bipolar radiofrequency electrocautery.” Acad Radiol 3(5):418–422.
952
Christopher L. Brace and Paul F. Laeseke
Meloni, M. F., S. N. Goldberg, T. Livraghi, F. Calliada, P. Ricci, M. Rossi, D. Pallavicini, and R. Campani. (2001). “Hepatocellular carcinoma treated with radiofrequency ablation: Comparison of pulse inversion contrastenhanced harmonic sonography, contrast-enhanced power Doppler sonography, and helical CT.” Am J Roentgenol 177(2):375–380. Mittleman, R. S., S. K. Huang, W. T. de Guzman, H. Cuenoud, A. B. Wagshal, and L. A. Pires. (1995). “Use of the saline infusion electrode catheter for improved energy delivery and increased lesion size in radiofrequency catheter ablation.” Pacing Clin Electrophysiol 18(5 Pt 1):1022–1027. Novick, A. C. (2004). “Laparoscopic and partial nephrectomy.” Clin Cancer Res 10(18 Pt 2):6322S–6327S. Organ, L. W. (1976). “Electrophysiologic principles of radiofrequency lesion making.” Appl Neurophysiol 39(2):69–76. Pereira, P. L., J. Trubenbach, M. Schenk, J. Subke, S. Kroeber, I. Schaefer, C. T. Remy, D. Schmidt, J. Brieger, and C. D. Claussen. (2004). “Radiofrequency ablation: in vivo comparison of four commercially available devices in pig livers.” Radiology 232(2):482–490. Poggi, G., C. Gatti, M. Melazzini, G. Bernardo, M. Strada, C. Teragni, A. Delmonte, C. Tagliaferri, C. Bonezzi, M. Barbieri, A. Bernardo, and P. Fratino. (2003). “Percutaneous ultrasound-guided radiofrequency thermal ablation of malignant osteolyses.” Anticancer Res 23(6D):4977–4983. Pop, M., A. Molckovsky, L. Chin, M. C. Kolios, M. A. Jewett, and M. D. Sherar. (2003). “Changes in dielectric properties at 460 kHz of kidney and fat during heating: Importance for radio-frequency thermal therapy.” Phys Med Biol 48(15):2509–2525. Quesson, B., J. A. de Zwart, and C. T. Moonen. (2000). “Magnetic resonance temperature imaging for guidance of thermotherapy.” J Magn Reson Imaging 12(4):525–533. Rossi, S., F. Garbagnati, I. De Francesco, F. Accocella, L. Leonardi, P. Quaretti, A. Zangrandi, C. Paties, and R. Lencioni. (1999). “Relationship between the shape and size of radiofrequency induced thermal lesions and hepatic vascularization.” Tumori 85(2):128–132. Shake, J. G., D. W. Larson, C. T. Salerno, R. W. Bianco, and R. M. Bolman 3rd. (1997). “The role of electrolyte in lesion size using an irrigated radiofrequency electrode.” J Invest Surg 10(6):339–346; discussion 346–348. Shock, S. A., K. Meredith, T. F. Warner, L. A. Sampson, A. S. Wright, T. C. Winter 3rd, D. M. Mahvi, J. P. Fine, and F. T. Lee Jr. (2004). “Microwave ablation with loop antenna: in vivo porcine liver model.” Radiology 231(1):143–149. Solbiati, L., T. Ierace, S. N. Goldberg, S. Sironi, T. Livraghi, R. Fiocca, G. Servadio, G. Rizzatto, P. R. Mueller, A. Del Maschio, and G. S. Gazelle. (1997). “Percutaneous US-guided radio-frequency tissue ablation of liver metastases: treatment and follow-up in 16 patients.” Radiology 202(1):195–203. Solbiati, L., T. Ierace, M. Tonolini, and L. Cova. (2004a). “Guidance and monitoring of radiofrequency liver tumor ablation with contrast-enhanced ultrasound.” Eur J Radiol 51:S19–S23. Solbiati, L., T. Livraghi, T. Ierace, F. Meloni, L. Cova, and S. Goldberg. (2004b). “Radiofrequency Ablation for Liver Colorectal Metastases: Is It Possible to Equal the 5-year Survival Rates of Surgery?” Radiological Society of North America Scientific Assembly and Annual Meeting, Chicago. Steinke, K., J. King, D. Glenn, and D. Morris. (2003). “Percutaneous radiofrequency ablation of lung tumors: difficulty withdrawing the hooks resulting in a split needle.” Cardiovasc Intervent Radiol 26:583–585. Steinke, K., P. E. Sewell, D. Dupuy, R. Lencioni, T. Helmberger, S. T. Kee, A. L. Jacob, D. W. Glenn, J. King, and D. L. Morris. (2004). “Pulmonary radiofrequency ablation—an international study survey.” Anticancer Res 24(1):339–343. Thomsen, S. (1991). “Pathologic analysis of photothermal and photomechanical effects of laser-tissue interactions.” Photochem Photobiol 53(6):825–835. Tungjitkusolmun, S., E. J. Woo, H. Cao, J. Z. Tsai, V. R. Vorperian, and J. G. Webster. (2000). “Finite element analyses of uniform current density electrodes for radio-frequency cardiac ablation.” IEEE Trans Biomed Engr 47(1):32–40. Varro, Z., J. K. Locklin, and B. J. Wood. (2004). “Laser navigation for radiofrequency ablation.” Cardiovasc Intervent Radiol 27(5):512–515. Viswanathan, A., B. Wood, N. Glossop, F. Banovac, K. Cleary, and J. Kruecker. (2005). “Multimodality Navigation for Radiofrequency Ablation (RFA) with Tracked Needles.” Annual Meeting of the Radiological Society of North America (RSNA), SSE03-06, Chicago, IL.
53–Radiofrequency Ablation
953
Wu, T., J. P. Felmlee, J. F. Greenleaf, S. J. Riederer, and R. L. Ehman. (2001). “Assessment of thermal tissue ablation with MR elastography.” Magn Reson Med 45(1):80–87. Zagoria, R. J. (2004). “Imaging-guided radiofrequency ablation of renal masses.” Radiographics 24 Suppl 1:S59–S71.
Chapter 54
Microspheres as Microbrachytherapy Bruce R. Thomadsen, Ph.D., James S. Welsh, Ph.D., and Richard J. Hammes, Ph.D. University of Wisconsin Madison, Wisconsin Definition of Microbrachytherapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 955 Description of Radiotherapeutic Microspheres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 955 Intra-arterial Microsphere Delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 956 SIR-Spheres® . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 956 Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 956 Delivery Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 957 TheraSpheres™ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 959 Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 959 Delivery Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 961 Prescription . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 961 Doses to Neighboring Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 962 Dose from Shunting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 962 Dose from Proximity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963 Dose from Leaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963 Post-procedural Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963 Removal of the Delivery System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963 Survey for Contamination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 963 Measure the Exposure Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 964 Measure the Residual Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965 Verify the Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 965
Definition of Microbrachytherapy Microbrachytherapy consists of delivery of radiation using microscopic carriers. While forceps easily pick up conventional brachytherapy sources, microbrachytherapy sources require magnification even to distinguish them. Microbrachytherapy falls between conventional brachytherapy and systemic treatment using radiopharmaceuticals, such as 131I for thyroid cancer or 153Sm EDTMP1 for metastatic bone pain. Often, the dividing line to either side blurs. Two exciting and currently growing examples of this modality are microspheres used for treatment of liver metastases and radioimmunotherapy. At the present, the physics of radioimmunotherapy tends to differ little from that for other systemically administered radionuclide therapy, and the discussion of that falls outside of the scope of this text. Thus, this chapter focuses on radiotherapeutic microspheres.
Description of Radiotherapeutic Microspheres Brachytherapy microspheres consists of very small beads, on an average 20 to 40 µm in diameter, that carry a radionuclide. As of this writing, the commercial products use 90Y as the radionuclide engine, but experimental work has been reported using 166Ho (Nijsen et al., 1999; Seppenwoolde et al., 2004) and 164Dy for neutron activation (Adnani 2004). Various materials have been used for the microspheres proper: glass (Herba et al., 1988), micropolymer, resin, starch, and polylactic acid among many other materials. The
1
Ethylene Diamine Tetramethylene Phosphonate
955
956
Bruce R. Thomadsen et al.
method of incorporation of the radionuclide in the microspheres depends on the particular microsphere material and varies widely. Of the two commercially available products: SIR-Spheres® (SIRTex, Lane Cove, New South Wales, Australia) consist of resin micropolymers with a diameter of 32 µm ± 10 µm. With an estimated activity per microsphere of 50 Bq, a typical patient administration of 2.5 GBq contains approximately 50 million microspheres. TheraSpheres™ (Nordion, Kanata, Canada) consists of glass microspheres with a diameter of 32 µm ± 10 µm. With a higher activity per sphere, 2500 Bq, a typical dose of 5 GBq consists of around 4 million microspheres.
Intra-arterial Microsphere Delivery In concept, the most common application of radiolabeled microspheres is in therapy for liver metastases. Most of the normal liver parenchyma receives its blood supply from the hepatic portal vein, while the blood supply for most liver tumors comes through the hepatic artery. This difference in the source of blood opens the possibility of using microspheres to selectively irradiate tumor tissue and preserve functioning liver. For this application, the microspheres are injected intra-arterially, through a catheter originating by insertion through the femoral artery. The microspheres are suspended in a slurry, and as they follow the arterial flow, to a capillary. The diameter of the microspheres exceeds that of the capillary, so the microspheres lodge at the mouth of the capillary. Since 90Y decays with a 64.2-hour half-life, the treatment takes approximately a fortnight (two weeks) for 97% completion (5 half-lives). The microspheres do not dissolve or erode within this time, but remain in place. Getting the microspheres into the artery is a bit more complicated than simply injecting them in the manner of a shot. To do so could cause a clog in the artery (or the plastic tubing leading to the artery), and lead to high doses in unintended locations. Thus, the delivery slowly instills a dilute slurry. The actual details differ somewhat between the two commercially available products, with some differences due to the material of the spheres themselves.
SIR-Spheres® SIR-Spheres have been approved by the U. S. Food and Drug Administration (FDA) for use in the United States for treatment of colorectal cancer metastases to the liver. Other uses require either an approved clinical trial or humanitarian use approval. Preparation SIR-Spheres come in a shipping vial with a manufacturer’s assay in activity. Verification of the stated activity can be quite challenging. The vials exhibit marked variations in thickness (as shown in Figure 1). The variation in the construction of the shipping vials produces inconsistent readings in well ionization chambers (dose calibrators) in terms of indicated activity. Compared with the manufacturer’s assay, measured activity using a constant calibration factor varied by ±11% in one series (Thomadsen 2003). The inconsistency makes quality assurance of the received activity very difficult. The slurry of microspheres is mixed well and, using a syringe, a volume taken from the shipping vial in the approximate proportion of the activity prescribed, that is
Vsyringe = Vshipping
APrescribed Ashipping
,
(1)
55–Microspheres as Microbrachytherapy
957
Figure 1. Shipping vials showing individual variations, noted by the arrows.
where V indicates volume, A the activity, and the subscripts “shipping” that quantity for the shipping vial, and “syringe,” obviously for the syringe. The process actually entails iterative measurement of the shipping vial to verify that the correct activity has been withdrawn. After pulling up the prescribed amount plus about 10% (which is likely to remain in the delivery system after the installation), the material is moved to the delivery vial (see Figure 2) and the fluid level in the vial brought to approximately 5 mm from the top with water. The vial is placed in an acrylic cylinder on the bottom of the delivery box. Two needles are placed inserted through the diaphragm of the vial (see Figure 3): one that will take the slurry from the vial to the catheter in the patient and one that brings water into the vial pushing the slurry out the first needle. As noted above, delivery of the microspheres in too concentrated a slurry could clog the catheter line or result in bunching of the microspheres in a single location in the tumor bed. Thus, to begin with the needle leading to the patient takes the fluid from the top of the vial. The polymer microspheres will settle to the bottom of the water slurry, but would float in a normal saline solution (or a solution containing iodine contrast agent). The tip of the needle bringing the water into the vial sits at the bottom of the vee in the bottom of the vial. All tubing is primed (filled with water) before connection. In the procedure room, all staff wears surgical caps, foot coverings, masks, and lead aprons, just as with any interventional procedure. Additional precautions include covering the floor under the work area and tables used for the radionuclide with absorbent pads. Delivery Technique The person actually delivering the microspheres (usually the interventional radiologist but occasionally the radiation oncologist or nuclear medicine specialist depending on experience and local institutional policy) pushes water into the vial in small pulses that stir the microspheres into the slurry. As the vial remains closed, the hydraulic pressure of the incoming water forces the slurry out the needle at the top of the vial, into the catheter leading to the hepatic artery.
958
Bruce R. Thomadsen et al.
Figure 2. The SIR-Spheres delivery box. The plastic box provides shielding from the beta particles emitted by the 90Y. In the figure the arrows indicate: (a) the delivery vial, and (b) the rotating knob to switch the flow to the patient between the microsphere slurry and the flush syringe.
Figure 3. A close-up showing the needles entering the delivery vial: on the right leading to the patient and on the left coming from the water-filled syringe.
55–Microspheres as Microbrachytherapy
959
Any leaks in the vial would allow the fluid to escape, causing a radioactive contamination problem, and prevent the pressure from moving the slurry into the treatment catheter. Aside from the inflowing and outflowing needles, a needle penetrates the diaphragm at least one other time, and often more, to instill the slurry in the vial from the shipping container. Each of these penetrations carries the potential to cause a pressure leak. To prevent this, after insertion of the needles and before the procedure, a light coat of superglue is spread across the surface of the diaphragm. Additionally, it has been recommended that needles no larger than 18 gauge be used to puncture the rubber diaphragm to introduce the slurry. As the water pushes the slurry out, even in the dilute form at the top of the vial, the concentration is fairly high. To prevent clogging of the line, the injection is paused after filling about one-third of the tubing leading from the box to the catheter. At that point, the radiologist turns the stopcock from MICROSPHERES to FLUSH, and injects plain water through the tubing, clearing the line. After clearing the line, the position of the catheter within the patient can be confirmed though fluoroscopy or cine using contrast. Periodic imaging with contrast also serves to verify that the material from of the catheter still flows into the target artery. However, the tubing and catheter must be cleared of the contrast medium before reinitiating the microsphere infusion because the presence of contrast can result in bunching of the microspheres. As the microspheres fill the mouths of the capillaries, the forward flow into the hepatic artery can slow or even stop, resulting in retrograde flow into the right gastric, gastroduodenal, or other arteries and an inadvertent dose to structures receiving blood from those arteries. A second infusion just like the first follows clearing the tubing and catheter. This procedure continues until the fluid in the vial becomes relatively dilute and clear, about five repetitions. At that time, the needle leading to the patient is pushed to the bottom of the vee in the vial, and the syringe that was pushing water into the vial is replaced with one filled with air. The radiologist then injects air into the vial, forcing the liquid out and into the tubing (stopping when the air reaches the stopcock leading directly to the patient catheter), thus completing the delivery of the radionuclide. In the experience at the University of Wisconsin, the fraction of the activity actually delivered to the patient comprises over 99% of the activity initially in the delivery vial.
TheraSpheres™ TheraSpheres have been approved by the U. S. FDA for use in the United States for treatment of primary hepatocelluar cancer of the liver. Other uses require either an approved clinical trial or humanitarian use approval. Preparation TheraSpheres arrive in the dose delivery vial with a manufacturer’s assay in activity. Surrounding the vial is a plastic cylinder. The manufacturer recommends verification of the activity by placing the whole vial into the well chamber, and thus measuring the bremsstrahlung. For a CRC 12 or 15, Nordion suggests a setting of 47.5. Verification of this setting value remains the responsibility of the institution’s medical physicist. At the moment, traceability relies on sources available from the manufacturer. Unlike SIR-Spheres, instead of removing the desired amount from the shipping container, the intent is to deliver the entire contents of the vial. Thus, an approximate amount is ordered (at this writing, 3, 5, 7, 10, 15, or 20 GBq) and the timing for the delivery determined based on decay in order to instill the correct activity. The connections of the components of the delivery system, while differing slightly from the description for SIR-Spheres, generally follow the same concepts. Figure 4 shows the overall schematic while Figure 5 shows a photograph of the equipment.
960
Bruce R. Thomadsen et al.
Figure 4. Schematic of the delivery process for TheraSpheres. (Figure courtesy of Nordion, Kanata, Canada.)
Figure 5. The physical setup for TheraSpheres delivery as shown in Figure 4. (Figure courtesy of Nordion, Kanata, Canada.)
55–Microspheres as Microbrachytherapy
961
Delivery Technique The delivery technique also generally follows that discussed above. Because TheraSpheres consist of glass, the use of saline, instead of plain water, poses no problem, as unlike resin microspheres, glass microspheres are unaffected by the buoyancy difference.
Prescription The U. S. Nuclear Regulatory Commission (NRC) approved the use of tagged microspheres as a brachytherapy procedure (NRC http://www.nrc.gov/materials/miau/med-use-toolkit/microsphere.html). Thus, microsphere treatments may only be prescribed by authorized users that satisfy the training and experience requirements for brachytherapy, i.e., radiation oncologists. Screening for treatment always includes a study using 99mTc-labeled macroaggregated albumin (MAA) to determine the degree of shunting to the lung or gastrointestinal tract. The fraction shunt either enters into the dosimetry calculations or, if greater than 20%, can be reason to abort the treatment. The NRC Licensing Guidance for 90Y microspheres states that the written directive (prescription) should be in terms of the dose (in rad, Gy, rem, or Sv) to the primary treatment site (target) and specify the maximum dose allowed to sites outside the target. While both manufacturers give procedures for dose calculation in their product material, the calculations often have little bearing on the true dose delivered. The calculation of dose for TheraSpheres assumes that the entire liver is the target. The calculation formalism then becomes Activity [GBq ] =
Dose[Gy ] ⋅ massliver [ kg ] 50[
Gy ⋅ kg GBq
] ⋅ (1 − Fshunt )
,
(2)
where the liver mass is estimated by computed tomography (CT) and the fractional shunt to the lung, Fshunt is based on the relative counts in the lung compared to the liver as determined on planar images of the 99m Tc-labeled MAA study. The constant for 90Y in the denominator, 50, comes from the product of: Equilibrium dose constant = 0.54 kg • Gy • GBq–1 • h–1, and Mean life = 1.44 • half-life = 1.44 • 64.2 h = 92.4 h rounded to a whole number from 49.8. As noted above, only certain activities can be ordered and the time for treatment administration is then determined so the calculated activity will be delivered. A typical prescribed dose is 150 Gy. The same equation holds for dose to the tumor, but the assessment becomes more difficult. Sarfaraz discusses the dosimetry techniques (Sarfaraz et al., 2003). The assessment comes from partitioning the activity observed in the tumor relative to that in the normal liver on single photon emission computed tomography (SPECT) images, and relating those regions to the mass as determined by CT. The difficulty often comes in defining the tumor volumes, particularly when there are multiple small tumors throughout the liver. Thus, while determining the average dose to the whole volume delimited by the liver is fairly straightforward, calculation of the actual dose to regions of tumor and, by exclusion, to the normal liver can become very challenging. Calculation for SIR-Spheres could take the same route, and the information is contained in the manufacturer’s data sheets. However, recognizing the difficulty (to impossibility) of determining the dose to tumors in most clinical settings, and in spite of the licensing guidance, the NRC has allowed prescription
962
Bruce R. Thomadsen et al.
of treatment for SIR-Spheres to be based on a model that was used in clinical trials. Three factors determine the dose for delivery: 1. Extent of disease (i.e., fraction of liver involved with tumor). 2. Fractional shunting to the lung. 3. The amount of the liver treated (i.e. right, left or both liver lobes). Table 1 summarizes the values assigned for each of these factors. The primary determination of the activity to use considers the extent of tumor involvement in the liver. The fractional shunting to the lung reduces the amount of radionuclide that can be injected based on lung tolerance. The prescribed activity is further modified based on whether the right, left, or entire liver is to be treated. Currently, most SIR-Spheres patients undergo treatment only to one lobe of the liver at a time, since that seems to be better tolerated.
Doses to Neighboring Structures Other organs receive dose from microsphere treatments in three ways.
Dose from Shunting As noted above, the planning phase uses scans with 99mTc-labeled MAA to quantitate shunting to the lungs and also to other abdominal organs. Such shunting can result from natural anatomical connections between the arterial flow or poor placement of the treatment catheter. Placement of the catheter requires periodic verification during the treatment delivery as the tip of the catheter can work itself free of the targeted artery. In addition, as the capillary bed fills with microspheres, the flow from the catheter tip into the target artery slows. If the flow into the capillary bed becomes slower than the microsphere delivery, or if the capillary flow virtually stops, the injected microspheres can spill out of the targeted artery and flow into neighboring arteries, such as branches of the common hepatic artery including the right gastric or gastroduodenal, delivering doses to organs they feed. Table 1. Factors that Determine the Activity To Prescribe for SIR-Spheres Extent of Disease
Lung Shunting
Target
Fraction of liver involvement
Base Activity in GBq
>50% 25%–50% < 25%
3 2.5 2
Fraction of counts in the lung
Dosage Modifier
< 10% 10%–15% 15%–20% > 20%
1.0 0.8 0.6 0.0 (DO NOT PROCEED)
Part of Liver
Dosage Modifier
Whole Liver Right Lobe Only Left Lobe Only
1.0 0.7 0.3
The prescribed activity = Base Activity × Lung-shunt Modifier × Fraction-of-liver Modifier.
55–Microspheres as Microbrachytherapy
963
Dose from Proximity Even though the radionuclide engine, 90Y, only emits beta particles, microspheres in tumors near the surface of the liver deliver doses to any tissue lying in contact. One organ commonly receiving dose in this manner is the duodenum, and the stomach often receives some dose as well. Without better treatment planning, screening for cases where such a dose may become significant remains elusive. Of course, the beta particles also produce bremsstrahlung (which is used sometimes to image the source distribution after instillation), but the bremsstrahlung dose is insufficient to produce injuries in neighboring structures.
Dose from Leaching Finally, some dose could result were the radionuclide to detach from the microspheres. To date, no study has shown significant leaching of the yttrium from the carriers.
Post-procedural Procedures After the instillation of the microspheres, the treatment team has four more tasks to perform.
Removal of the Delivery System The delivery vial and tubing will all be radioactive waste after the procedure and must be removed before looking for unexpected contamination. Usually, these would be placed in a plastic box for transport to the storage facility.
Survey for Contamination A survey for radioactive contamination certainly is routine following any instillation of unsealed sources into a patient. However, radioactive microspheres are not considered unsealed sources! The manufacturers have gone to great effort to ensure that the radionuclide remains in the microspheres. That being noted, because of the small size of the microspheres, they do act similarly to spilled solutions, but not identically. The initial steps are identical to those for radioactive solutions: using an appropriate survey meter (a GM with a pancake probe works well), check for activity above background on all the staff in the procedure room (particularly hands—front and back—the bottoms of the feet as in Figure 6, and their body front and back); and the floor, tables and carts; and linens used in the procedure. Were contamination found, the remedial steps depend on the location. 1. Contamination on cloth or paper. Contaminated cloth can be removed, bagged, and placed in the plastic box used for the tubing and vial. Storing the article in a safe, shielded location, the activity can be allowed to decay to background and the article returned to service. Contaminated paper should be bagged and disposed in the same manner as any other solid, radioactive waste. 2. Contamination on hands or other places on the body. Assuming that the staff was appropriately covered and wearing gloves, the likelihood of contamination on skin remains very low. However, should that occur and the contamination level is low, the affected person should wash the area with copious amounts of water and mild soap (or, not having any choice, whatever soap is available). After washing, the area is monitored again, and if still above background, washed again. After three cycles of washing and monitoring, if the readings are not falling, the washing should stop, and the physicist should consult an organization such as REAC/TS (Radiation Assistance Training Center/Training Site) (865-576-1005).
964
Bruce R. Thomadsen et al.
Figure 6. Surveying the bottoms of feet.
3. Contamination on surfaces. Microsphere contamination of surfaces presents two distinct hazards. The first is that the microspheres may be difficult to get out of cervices. While solutions follow capillary action, a slurry of microspheres will do so only if unimpeded. Cracks and imperfections in surfaces may hold the microspheres back from following the fluid, particularly if they have settled out of solution. Thus, decontamination may be more difficult than with a straight solution. Furthermore, any spill may allow the microspheres to aerosolize or spread by rolling about as the carrier solution dries. Thus, such a spill should be kept moist and covered until removed.
Measure the Exposure Rate As with any permanent brachytherapy, to release the patient the exposure at a meter must remain less than 50 µSv/h (5 mR/h). For these cases, the regulation never prevents release but does require documentation of the readings, which typically run between 5 and 30 µSv/h.
55–Microspheres as Microbrachytherapy
965
Measure the Residual Activity Part of determining the activity instilled requires assay of the activity remaining in the delivery vial and associated tubing. Neither measurement is entirely straightforward. Such measurements may use either the dose calibrator or a survey meter (not a GM counter). With the dose calibrator, the vial is simply measured before and after the treatment, and the residual in the vial taken as the ratio of the readings. A similar procedure is followed when using a survey meter. The results must be seen as only approximate in either case. The delivery vial narrows to a vee over the last approximately 12 mm from the bottom. For the initial measurement a significant proportion of the microspheres sit above this vee, while for the residual measurement, any activity will reside in the bottom. Thus, the attenuation of the wall of the vial will be different for the two measurement conditions. The difficulty is even more pronounced for the tubing. With no activity measurement before to calibrate the tubing geometry, simply placing the tubing in a dose calibrator may give erroneously very high readings if the holder fails to stop all the beta particles from entering the sensitive volume of the chamber, or even impinging on metallic walls producing more bremsstrahlung than the original vial would have. Even enclosing the tubing in a plastic container of similar thickness to the vial fails to produce similar enough geometries to make such measurements very useful. Making similar measurements with a survey meter may improve the reliability, particularly if the readings are performed at a great distance, but only if the original, preprocedure reading is performed in the same geometry. In summary, the post-procedural verification readings may be more useful at satisfying regulatory requirements than assessing patient treatments.
Verify the Distribution Bremsstrahlung planar images can give a rough qualitative impression of the relative activity distribution between the liver and neighboring structures. Quantitatively, the images can confirm the installation of a normal amount of radionuclide, although the actual calculated concentration should not be taken as a serious index. Unfortunately, the bremsstrahlung signal is too low for a meaningful SPECT study.
References Adnani, N. (2004). “Dosimetry of in situ activated dysprosium microspheres.” Phys Med Biol 49:733–746. Herba, M. J., F. F. Illescas, M. P. Thirlwell, G. J. Boos, L. Rosenthall, M. Atri, and P.M. Bret. (1988). “Hepatic malignancies: Improved treatment with intraarterial Y-90.” Radiology 169:311–314. Nijsen, J. F. W., B. A. Zonnenberg, J. R. W. Woittiez, D. W. Rook, I. A. Swildens-van Woudenberg, P. P. van Rijk, and A. D. van Het Schip. (1999). “Holmium-166 poly lactic acid microspheres applicable for intra-arterial radionuclide therapy of hepatic malignancies: effects of preparation and neutron activation techniques.” Euro J Nucl Med 26:699–704. Sarfaraz M., A. Kennedy, Z. Cao, G. Sackett, C. Yu, M. Lodge, R. Murthy, B. Line, and D. Van Echo. (2003). “Physical aspects of yttrium-90 microsphere therapy for nonresectable hepatic tumors.” Med Phys 30:199–203. Seppenwoolde, J.-H., J. F. Nijsen, L. W. Bartels, S. Zielhuis, A. D. van Het Schip, and C. J. Bakker. (2004). “Internal radiation therapy of liver tumors: Qualitative and quantitative magnetic resonance imaging of biodistribution of holmium-loaded microspheres animal models.” Magn Reson Med 53:76–84. Thomadsen, B. R, (2003). Dosimetric Criteria for Microsphere Brachytherapy. Annual meeting of Council on Ionizing Radiation Standards and Measurement, Gaithersburg, MD, October 27, 2003.
Brachytherapy Physics, Second Edition Compiled References A–L Aarts, E. H. L., and J. H. M. Korst. Simulated Annealing and Boltzmann Machine. New York: John Wiley & Sons, 1989. Ahmed, M., and S. N. Goldberg. (2004). “Combination radiofrequency thermal ablation and adjuvant IV liposomal doxorubicin increases tissue coagulation and intratumoural drug accumulation.” Int J Hyperthermia 20(7):781–802. Ahmed, M., S. M. Lobo, J. Weinstein, J. B. Kruskal, G. S. Gazelle, E. F. Halpern, S. K. Afzal, R. E. Lenkinski, and S. N. Goldberg. (2002). “Improved coagulation with saline solution pretreatment during radiofrequency tumor ablation in a canine model.” J Vasc Interv Radiol 13(7):717–724. Aizman, A., P. T. Finger, U. Shabto, A. Szechter, and A. Berson. (2004). “Palladium 103 (103Pd) plaque radiation therapy for circumscribed choroidal hemangioma with retinal detachment.” Arch Ophthalmol 122:1652–1656. Akabani, G., I. Cokgor, R. E. Coleman, D. González Trotter, T. Wong, H. S. Friedman, A. Garcia-Turner, J. E. Herndon II, D. DeLong, R. E. McLendon, X.-G. Zhao, C.N. Pegram, D.D. Bigner, and M. R. Zalutsky. (2000). “Dosimetry and dose-response relationships in newly diagnosed patients treated with iodine-131-labeled anti-tenacin monoclonal antibody therapy.” Int J Radiat Oncol Biol Phys 46:947–958. Alberti, W., B. Pothmann, P. Tabor, K. Muskalla, K. P. Hermann, and D. Harder. (1991). “Dosimetry and physical treatment planning for iodine eye plaque therapy.” Int J Radiat Oncol Biol Phys 20:1087–1092. Altschuler, M. D., P. A. Findlay, and R. D. Epperson. (1983). “Rapid, accurate 3D location of multiple seeds in implant radiation treatment planning.” Phys Med Biol 28:1305–1318. American Association of Physicists in Medicine (AAPM). (1994). “Comprehensive QA for radiation oncology. Report of AAPM Radiation Therapy Committee Task Group 40.” Med Phys 21:581–618. Also available as AAPM Report No. 46. American Association of Physicists in Medicine (AAPM). (1997). “Code of practice for brachytherapy physics: AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24:1557–1598. Also available as AAPM Report No. 60. American Association of Physicists in Medicine (AAPM). (1998). “High dose-rate brachytherapy treatment delivery: AAPM Radiation Therapy Committee Task Group No. 59.” Med Phys 25:375–403. Also available as AAPM Report No. 61. American Association of Physicists in Medicine (AAPM). Report No. 21. “Specification of Brachytherapy Source Strength.” New York: American Institute of Physics, New York, 1987. American Association of Physicists in Medicine (AAPM). Report No. 41. “Remote Afterloading Technology.” New York: American Institute of Physics, New York, 1993. American Cancer Society (ACS). Cancer Facts and Figures 2005. Atlanta, GA: ACS, 2005. Amols, H. I., and I. I. Rosen. (1981). “A three-film technique for reconstruction of radioactive seed implants.” Med Phys 8:210. Amols, H. I., L. E. Reinstein, and J. Weinberger. (1996a). “Dosimetry of a radioactive coronary balloon dilatation catheter for treatment of neointimal hyperplasia.” Med Phys 23:1783–1788. Amols, H. I., S. Mirzadeh, F. F. Knapp Jr., and J. Weinberger. (1996b). “Beta irradiation for restenosis: Considerations for stent implantation.” Circulation 94:I-210. Anagnostopoulos, G., D. Baltas, A. Geretschlaeger, T. Martin, P. Papagiannis, N. Tselis, and N. Zamboglou. (2003). “In vivo thermoluminescence dosimetry dose verification of transperineal 192Ir high-dose-rate brachytherapy using CT-based planning for the treatment of prostate cancer.” Int J Radiat Oncol Biol Phys 57(4):1183–1191. Anderson, L. L. (1974). “Characteristics of 252Cf neutron sources used for radiotherapy.” Eur J Cancer 10:203–205. Anderson, L. L. (1976). “Spacing nomograph for interstitial implants of 125I seeds.” Med Phys 3:48–51. Anderson, L. L. (1986). “A ‘natural’ volume-dose histogram for brachytherapy.” Med Phys 13:898–903.
1
Anderson, L. L. (1995). “Dose Specification and Quantification of Implant Quality” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (ed.). Madison, WI: Medical Physics Publishing, 1995. Anderson, L. L., and A. D. Osian. (1986). “Brachytherapy optimization and evaluation.” Endocuriether Hypertherm Oncol 2:S25–S32. Anderson, L. L., and J. L. Presser. “Classical Systems I for Temporary Interstitial Implants: Manchester and Quimby Systems” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds). Madison, WI: Medical Physics Publishing, pp. 301–322, 1995. Anderson, L. L., B. S. Hilaris, and L. K. Wagner. (1985). “A nomograph for planar implant planning.” Endocuriether Hypertherm Oncol 1:9–15. Anderson, L. L., F. W. Mick, K. Zabrouski, and Y. Watanabe. (1995). “Photoelectrons facilitate autoradiography for 192Ir remote afterloaders.” Med Phys 22(11):1759–1761. Anderson, L. L., J. V. Moni, and L. B. Harrison. (1993). “A nomograph for permanent implants of palladium-103 seeds.” Int J Radiat Oncol Biol Phys 27:129–136. Anderson, L. L., M. R. Hoffman, P. J. Harrington, and G. Starkschall. (1997). “Atlas generation for intraoperative high dose rate brachytherapy.” J Brachyther Int 13:333–340. Anderson, L. L., R. Nath, K. A. Weaver, D. Nori, T. L. Phillips, Y. H. Son, S. Chiu-Tsao, A. S. Meigooni, J. A. Meli, and V. Smith. Interstitial Brachytherapy, Physical, Biological and Clinical Considerations. New York: Raven Press, 1990. Anderson, L., and S. Chiu-Tsao. “Physical Aspects of Eye Plaque Brachytherapy Using Photon Emitters” in Radiotherapy of Intraocular and Orbital Tumors. R. Sagerman and W. Alberti (eds.). New York: Springer-Verlag, pp. 347–361, 1993. Anderson, L., R. Nath, A. J. Olch, and J. Roy. (1991). “American Endocurietherapy Society recommendations for dose specification in brachytherapy.” Endocuriether/Hypertherm Oncol 7:1–12. Angelopoulos, A., P. Baras, L. Sakelliou, P. Karaiskos, and P. Sandilos. (2000). “Monte Carlo dosimetry of a new 192Ir high dose rate brachytherapy source.” Med Phys 27(11):2521–2527. Ankem, M. K., V. S. DeCarvalho, A. M. Harandgozo, V. H. Hartano, M. Perrotti, K. Han, W. J. Shih, E. Malka, E. C. White, R. Maggio, R. Ioffreda, J. W. Goldsmith, and R. E. Weiss. (2002). “Implications of radioactive seed migration to the lungs after prostate brachytherapy.” Urology 59:555–559. Armour, E. P., J. R. White, A. Armin, P. M. Corry, M. Coffey, C. DeWitt, and A. Martinez. (1997). “Pulsed low dose rate brachytherapy in a rat model: Dependence of late rectal injury on radiation pulse size.” Int J Radiat Oncol Biol Phys 38:825–834. Armpilia, C. I., R. G. Dale, I. P. Coles, B. Jones, and V. Antipas. (2003). “The determination of radiobiologically optimized half-lives for radionuclides used in permanent brachytherapy implants.” Int J Radiat Oncol Biol Phys 55:378–385. Arthur, D. W., D. E. Wazer, D. Koo, L. Berle, L. Cuttino, M. Yunes, M. Rivard, D. Todor, S. Tong, T. Tenenholz, and T. DiPetrillo. (2003b). “The importance of dose volume histogram evaluation in partial breast brachytherapy: A study of dosimetric parameters.” Int J Radiat Oncol Biol Phys 57(2):S361–S362. Arthur, D. W., F. A. Vicini, R. R. Kuske, D. E. Wazer, and S. Nag. (2002). “Accelerated partial breast irradiation: An updated report from the American Brachytherapy Society.” Brachytherapy 1:184–190. Arthur, D. W., F. A. Vicini, R. R. Kuske, D. E. Wazer, and S. Nag. (2003a). “Accelerated partial breast irradiation: An updated report from the American Brachytherapy Society.” Brachytherapy 2(2):124–130. Arthur, D.W., D. Koo, R. D. Zwicker, S. Tong, H. D. Bear, B. J. Kaplan, B. D. Kavanagh. L. A. Warwicke, D. Holdford, C. Amir, K. J. Archer, and R. K. Schmidt-Ullrich. (2003a). “Partial breast brachytherapy after lumpectomy: Low dose-rate and high dose-rate experience.” Int J Radiat Oncol Biol Phys 56:681–689. Ash, D., A. Flynn, J. Battermann, T. de Reijke, P. Lavagnini, and L. Blank. (2000). “ESTRO/EAU/EORTC recommendations on permanent seed implantation for localized prostate cancer.” Radiother Oncol 57:315–321. Astrahan, M. A. (2003). “A patch source model for treatment planning of ruthenium ophthalmic
2
applicators.” Med Phys 30:1219-1228. Astrahan, M. A. (2005). “Improved treatment planning for COMS eye plaques.” Int J Radiat Oncol Biol Phys 61(4):1227–1242. Astrahan, M. A., G. Luxton, G. Jozsef, P. E. Liggett, and A. Petrovich. (1990a). “Optimization of 125I ophthalmic plaque brachytherapy.” Med Phys 17:1053–1057. Astrahan, M. A., G. Luxton, G. Jozsef, T. D. Kampp, P. E. Liggett, M. D. Sapozink, and Z. Petrovich. (1990b). “An interactive treatment planning system for ophthalmic plaque radiotherapy.” Int J Radiat Oncol Biol Phys 18:679–687. Astrahan, M. A., G. Luxton, Q. Pu, and Z. Petrovich. (1997). “Conformal episcleral plaque therapy.” Int J Radiat Oncol Biol Phys 39(2):505–519. Attix, F. H. Introduction to Radiological Physics and Radiation Dosimetry. New York: John Wiley and Sons, 1986. Attix, F. H., and V. H. Ritz. (1957). “A determination of the gamma-ray emission of radium.” J Res Natl Bureau Standards 59:293–305. Augsburger, J. J., J. W.Gamel, V. F. Sardi et al. (1986). “Enucleation vs. cobalt plaque radiotherapy for malignant melanomas of the choroid and ciliary body.” Arch Ophthalmol 104:655–661. Augsburger, J. J., R. G. Peyster, A. M. Markoe, E. G. Guillet, J. A. Shields, and M. E. Haskin. (1987). “Computed tomography of posterior uveal melanoma.” Arch Ophthalmol 105:1512–1516. Baglan, K. A., A. A. Martinez, R. C. Frazier, V. R. Kini, L. L. Kestin, P. Y. Chen, G. Edmundson, E. Mele, D. Jaffray, and F. A. Vicini. (2001). “The use of high-dose rate brachytherapy alone after lumpectomy in patients with early-stage breast cancer treated with breast-conserving therapy.” Int J Radiat Oncol Biol Phys 50:1003–1011. Ballester, F., C. Hernandez, J. Perez-Calatayud, and F. Lliso. (1997). “Monte Carlo calculation of dose rate distributions around 192Ir wires.” Med Phys 24:1221–1228. Ballester, F., D. Granero, J. Perez-Calatayud, E. Casal, and V. Puchades. (2004). “Monte carlo dosimetric study of best industries and Alpha Omega Ir-192 brachytherapy seeds.” Med Phys 31(12):3298–3305. Ballester, F., J. L. Lluch, Y. Limami, M. A. Serrano, E. Casal, J. Perez-Calatayud, and F. Lliso. (2000). “A Monte Carlo investigation of the dosimetric characteristics of the CSM11 137Cs source from CIS.” Med Phys 27(9):2182–2189. Ballester, F., V. Puchades, I. L. Lluch, M. A. Serrano-Andres, Y. Limami, J. Perez-Calatayud, and E. Casal. (2001). “Monte-Carlo dosimetry of the HDR 12i and Plus 192Ir sources.” Med Phys 28(12):2586–2591. Baltas, D. S. Giannouli, A. Garbi, F. Diakonos, K. Gramani, G. T. Ioannidis, A. Tsalpatouros, N. Uzunoglu, C. Kolotas, and N. Zmboglou. (1998). “Application of the Monte Carlo integration (MCI) method for calculation of the anisotropy of 192Ir brachytherapy sources.” Phys Med Biol 43(6):1783–1801. Baltas, D., C. Kolotas, K. Geramani, R. F. Mould, G. Ioannidis, M. Kekchidi, and N. Zamboglou. (1998). “A conformal index (COIN) to evaluate implant quality and dose specification in brachytherapy.” Int J Radiat Oncol Biol Phys 40(2):515–524. Batho, H. R., and N. E. T. Young. (1964). “Tissue absorption corrections for linear radium source.” Br J Radiol 37:689. Beatty, J., P. J. Biggs, K. Gall, P. Okunieff, F. S. Pardo, K. J. Harte, M. J. Dalterio, and A. P. Sliski. (1996). “A new miniature x-ray source for interstitial radiosurgery: Dosimetry.” Med Phys 23:53–62. Beaulieu, L., L. Archambault, S. Aubin, E. Oral, R. Taschereau, and J. Pouliot. (2004a). “The robustness of dose distributions to displacement and migration of 125I permanent seed implants over a wide range of seed number, activity, and designs.” Int J Radiat Oncol Biol Phys 58:1298–1308. Beddoe, A. H. (1975). “Isoexposure curves for Co-60 ophthalmic applicators.” Aust Radiol 19:145-151 Beitler, J., B. McCormick, and R. Ellsworth. (1990). “Ocular melanoma: Total dose and dose rate effects with Co-60 plaque therapy.” Radiology 176:275–278.
http://www.redjournal.org/medline/record/MDLN.2353101 Bell, A. G. (1903). “The uses of radium.” Ann Med 6:261. Benedetto, A. R. (1995). “The Brachytherapy Regulatory Environment: Organization of a Radiation Safety Program” and its appendix “Compilation of Selected Rules for Brachytherapy Use of USNRC
3
Licensed Materials” by J. O. Eichling in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (Eds.). AAPM 1994 Summer School Proceedings. Madison, WI: Medical Physics Publishing, pp. 163–173 and 174–183, 1995. Berger, M. J. (1968). “Energy deposition in water by photons from point isotropic sources.” J Nucl Med Suppl 1:17–25. Bernard, M., B. Guille, and G. Duvalet. (1975). “Mesure du debit d’exposition linéique nominal des sources a une dimension, utilisées en curiethérapie.” J Radiol Electrol 56:785–790. Beyer, D. C. (1991). “Pterygia: single-fraction postoperative beta irradiation.” Radiology 178:569–571. Beyer, D., R. Nath, W. Butler, G. Merrick, J. Blasko, S. Nag, and C. Orton. (2000). “American brachytherapy society recommendations for clinical implementation of NIST-1999 standards for (103)palladium brachytherapy.” Int J Radiat Oncol Biol Phys 47(2):273–275. Bice, W. S., B. R. Prestidge, J. J. Prete, and D. F. DuBois. (1998). “Clinical impact of implementing the recommendations of the AAPM Task Group 43 on permanent prostate brachytherapy using 125I.” Int J Radiat Oncol Biol Phys 40:1237–1241. Bice, W. S., Jr., B. R. Prestidge, and M. F. Sarosdy. (2001). “Sector analysis of prostate implants.” Med Phys 28(12):2561–2567. Bice, W. S., Jr., B. R. Prestidge, P. D. Grimm, J. L. Friedland, V. Feygelman, M. Roach 3rd, J. J. Prete, D. F. Dubois, and J. C. Blasko. (1998). “Centralized multiinstitutional postimplant analysis for interstitial prostate brachytherapy.” Int J Radiat Oncol Biol Phys 41(4): 921–927. Bice, W. S., Jr., D. F. Dubois, J. J. Prete, and B. R. Prestidge. (1999). “Source localization from axial image sets by iterative relaxation of the nearest neighbor criterion.” Med Phys 26(9):1919–1924. Biggs, P. J., and D. M. Kelley. (1983). “Geometric reconstruction of seed implants using a three-film technique.” Med Phys 10:701–704. Biggs, P., J. Beatty, K. Gall, K. Harte, and A. Sliski. (1993). “Absolute dosimetry for a new 40 kV x-ray device used for stereotactic radiation therapy.” (Abstract). Med Phys 20:925. Bilaniuk, L. T., J. F. Schenck, R. A. Zimmerman, H. R. Hart, T. H. Foster, W. A. Edelstein, H. I. Goldberg, and R. I. Grossmam. (1985). “Ocular and orbital lesions: Surface coil MR imaging.” Radiology 156:669–674. Bilchik, A. J., D. M. Rose, D. P. Allegra, P. J. Bostick, E. Hsueh, and D. L. Morton. (1999). “Radiofrequency ablation: A minimally invasive technique with multiple applications.” Cancer J Sci Am 5(6):356–361. BIR/IPSM. “Recommendations for Brachytherapy Dosimetry: Report of a Joint Working Party. London: British Institute of Radiology and Institute of Physical Sciences in Medicine, 1993. Blank, L. Overview of LDR-PDR Differentiators. Presented at Nucletron Workshop: From Traditional to Modern Brachytherapy. Transferring from LDR to PDR. Amsterdam, The Netherlands, 2004. Blasko, J. C., H. Ragde, and D. Schumacher. (1987). “Transperineal percutaneous iodine-125 implantation for prostatic carcinoma using transrectal ultrasound and template guidance.” Endocuriether/Hypertherm Oncol 3:131–139. Blasko, J. C., H. Ragde, and P. D. Grimm. (1991). “Transperineal ultrasound-guided implantation of the prostate: morbidity and complications.” Scand J Urol Nephrol Suppl 137:113–118. Blichert-Toft, M., M. Brincker, J. A. Andersen, K. W. Andersen, C. K. Axelsson, H. T. Mouridsen, P. Dombernowsky, M. Overgaard, C. Gadeberg, G. Knudsen, et al. (1988). “A Danish randomized trial comparing breast-preserving therapy with mastectomy in mammary carcinoma.” Acta Oncol 27:671–677. Boehm, T., A. Malich, S. N. Goldberg, J. R. Reichenbach, I. Hilger, P. Hauff, M. Reinhardt, M. Fleck, and W. A. Kaiser. (2002). “Radio-frequency tumor ablation: Internally cooled electrode versus saline-enhanced technique in an aggressive rabbit tumor model.” Radiology 222(3):805–813. Bohm, T. D., S. L Griffin, P. M. DeLuca, and L. A. DeWerd. (2005). “The effect of ambient pressure on well chamber response: Monte Carlo calculated results for the HDR 1000 Plus.” Med Phys 32:1103–1114. Bond, W. H., D. Richards, and E. Turner. (1965). “Experience with radioactive gold in the treatment of craniopharyngioma.” J Neurol Neurosurg Psychiat 28:30–38. Bosworth, J. L., S. Packer, M. Rotman, T. Ho, and P. T. Finger. (1988). “Choroidal melanoma: I-125 plaque therapy.” Radiology 169:249-251.
4
Böttcher, H. D., B. Schopohl, D. Liermann, J. Kollath, and I. A. Asamietz. (1994). “Endovascular irradiation—a new method to avoid recurrent stenosis after stent implantation in peripheral arteries: Technique and preliminary results. Int J Radiat Oncol Biol Phys 29(1):183–186. Boutillon, M. “Values of g for Photon Energies” in CCEMRI Report (1)185-18. Paris: Consultative Committee for Standards of Measuring Ionizing Radiations, 1985. Boutilon, M., and A. M. Perroche-Rous. (1987). “Re-evaluation of the W value for electrons in dry air.” Phys Med Biol 32:213–291. Brahme, A., J. Nilsson, and D. Belkic. (2001). “Biologically optimized radiation therapy.” Acta Oncol 40:725–734. Brenner, D. J. (1993). “Accelerated repopulation during radiotherapy–Evidence for delayed onset.” Radiat Oncol Invest 1:167–172. Brenner, D. J. (1997). “Radiation biology in brachytherapy.” J Surg Oncol 65:66–70. Brenner, D. J., A. A. Martinez, G. K. Edmundson, C. Mitchell, H. D. Thames, and E. P. Armour. (2002). “Direct evidence that prostate tumors show high sensitivity to fractionation (low _/_ ratio) similar to late responding normal tissue.” Int J Radiat Oncol Biol Phys 52:6–13. Brenner, D. J., and E. J. Hall. (1991). “Conditions for the equivalence of continuous to pulsed low dose rate brachytherapy.” Int J Radiat Oncol Biol Phys 20:181–190. Brenner, D. J., E. J. Hall, G. Randers-Pehrson, Y. Huang, G. W. Johnson, R. W. Miller, B. Wu, M. E. Vazquez, C. Medvedosky, and B. V. Worgul. (1996). “Quantitative comparisons of continuous and pulsed low dose rate regimens in a model late-effect system.” Int J Radiat Oncol Biol Phys 34:905–910. Bridal, S. L., J. M. Correas, A. Saied, and P. Laugier. (2003) “Milestones on the road to hier resolution, quantitative, functional ultrasonic imaging.” IEEE Proceedings 91:1543–1561. Bridier, A., H. Kafrouni, J. P. Houlard, and A. Dutreix. “Comparison des distributions de dose en curietherapie interstitielle autour des sources continues et discontinues.” Tnt. Symp. on Dosimetry in Radiotherapie. IAEA SM-298/23. Vienna: IAEA, 1988. Briesmeister, J. F. (2000). “MCNP—A general Monte Carlo N-particle transport code—Version 4C.” Los Alamos National Laboratory Report LA-13709-M. March 2000. Brinkmann, D., and R. W. Kline. (1998). “Automated seed localization from CT datasets of the prostate.” Med Phys 25(9):1667–1672. British Committee on Radiation Units and Measurements (BCRU). (1984). “Specification of brachytherapy sources, memorandum from the British Committee on Radiation Units and Measurements.” Br J Radiol 57:941–942. British Institute of Radiology (BIR). “Recommendations for Brachytherapy Dosimetry.” Report of a joint working party of the BIR and the Institute of Physical Sciences in Medicine (IPSM). London: British Institute of Radiology, 1993. Burghardt, E., H. M. Hofmann, F. Ebner, J. Haas, K. Tamussino, and E. Justich. (1989). “Magnetic resonance imaging in cervical cancer: A basis for objective classification.” Gynecol Oncol 33:61–67. Burlin, T. E. (1966). “A general theory of cavity ionization.” Br J Radiol 39:361. Butler, W. M. (2005). Personal communication. Butler, W. M., A. T. Dorsey, K. R. Nelson, and G. M. Merrick. (1998). “Quality assurance calibration of I-125 Rapid Strand in a sterile environment.” Int J Radiat Oncol Biol Phys 41:217–222. Butler, W. M., and G. S. Merrick. (1996). “I-125 Rapid Strand™ loading technique.” Radiat Oncol Invest 4:48–49. Butler, W. M., G. S. Merrick, J. H. Lief, and A. T. Dorsey. (2000). “Comparison of seed loading approaches in prostate brachytherapy.” Med Phys 27:381–392. Cance, M., and J. P. Simoen. (1983). “Etalonnage des sources de curiethérapie” Compte-Rendus du Congrès de la Société Française des Physiciens d’Hépitaux. Paris: French Society of Hospital Physicists (SFPH), pp. 87–104, 1983. Carlsson, A. K., and A. Ahnesjo. (2000). “The collapsed cone superposition algorithm applied to scatter dose calculations in brachytherapy.” Med Phys 27(10):2320–2332. Casal, E., F. Ballester, J. L. Lluch, J. Perez-Calatayud, and F. Lliso. (2000). “Monte Carlo calculations of dose rate distributions around the Amersham CDCS-M-type 137Cs source. Med Phys 27(1):132–140.
5
Cavallo, L. M., B. M. Coursey, S. B. Garfinkel et al., (1973). “Needs for radioactivity standards and measurements in different fields.” Nucl Instr Methods 112:5–18. Cember. H. Introduction to Health Physics, 3rd ed. New York: McGraw-Hill, 1996. Cha, C. H., F. T. Lee Jr., J. M. Gurney, B. K. Markhardt, T. F. Warner, F. Kelcz, and D. M. Mahvi. (2000). “CT versus sonography for monitoring radiofrequency ablation in a porcine liver.” Am J Roentgenol 175(3):705–711. Charra-Brunaud, C., I.-C. Hsu, V. Weinberg, and J. Pouliot. (2003). “Analysis of the interaction between number of implant catheters and dose-volume histograms in prostate high-dose-rate brachytherapy using a computer model.” Int J Radiat Oncol Biol Phys 56(2):586–591. Charyulu, K. K. N. (1980). “Transperineal interstitial implantation of prostate cancer: A new method.” Int J Radiat Oncol Biol Phys 6:1261–1266. Chase, N.E., H. L. Atkins, and J.W. Correll. (1961). “Interstitial irradiation of brain tumors with iridium 192.” Radiology 77:842–843. Chassagne, D., and J. C. Horiot. (1977). “Proposals for common definitions of reference points in gynecological brachytherapy.” J Radiol Electrol Med Nucl 58:371–373. (French.) Chavaudra, J. (1995). “Pierre and Marie Curie-Sklodowska.” Med Phys 22:1877–1887. Chen, Y., A. L. Boyer, and L. Xing. (2000). “A dose-volume histogram based optimization algorithm for ultrasound guided prostate implants.” Med Phys 27:2286–2292. Chini, E. N., M. J. Brown, M. A. Farrell, and J. W. Charboneau. (2004). “Hypertensive crisis in a patient undergoing percutaneous radiofrequency ablation of an adrenal mass under general anesthesia.” Anesth Analg 99(6):1867–1869, table of contents. Chinn, S. B., F. T. Lee Jr., G. D. Kennedy, C. Chinn, C. D. Johnson, T. C. Winter 3rd, T. F. Warner, and D. M. Mahvi. (2001). “Effect of vascular occlusion on radiofrequency ablation of the liver: results in a porcine model.” Am J Roentgenol 176(3):789–795. Chiu-Tsao, S.-T., T. Duckworth, C. Zhang, N. S. Patel, C. Y. Hsiung, L. Wang, J. A. Shih, and L. B. Harrison. (2004b). “Dose response characteristics of new models of GAFCHROMIC films: dependence on densitometer light source and radiation energy.” Med Phys 31(9):2501–2508. Chiu-Tsao, S.-T., T. L. Duckworth, N. S. Patel, J. Pisch, and L. B. Harrison. (2004a). “Verification of Ir192 near source dosimetry using GAFCHROMIC film.” Med Phys 31(2):201–207. Chiu-Tsao, S.-T., T. Rusch, S. Axelrod, H. Tsao, and L. Harrison. (2004). “Radiochromic film dosimetry for a new electronic brachytherapy source.” (Abstract). Med Phys 31:1913. Cho, P. S., R. H. Johnson, and T. W. Griffin. (1995). “Cone-beam CT for radiotherapy applications.” Phys Med Biol 40(11):1863–1883. Cho, S. H., and R. Muller-Runkel. (1997). “Validity of the interval method for the determination of the anisotropy factor of high dose rate 192Ir sources.” Int J Radiat Oncol Biol Phys 37(2):483–487. Cioni, R., N. Armillotta, I. Bargellini, V. Zampa, C. Cappelli, P. Vagli, G. Boni, S. Marchetti, V. Consoli, and C. Bartolozzi. (2004). “CT-guided radiofrequency ablation of osteoid osteoma: Long-term results.” Eur Radiol 14(7):1203–1208. Cionini, L., P. Pacini, S. Marzano, et al. (1993). “Exclusive brachytherapy after conservative surgery in cancer of the breast.” (Abstract). Lyon Chir 89:128. Clark, R. M., P. B. McCulloch, M. N. Levine, M. Lipa, R. H. Wilkinson, L. J. Mahoney, V. R. Basrur, B. Nair, R. S. McDermot, and C. S. Wong. (1992). “Randomized clinical trial to assess the effectiveness of breast irradiation following lumpectomy and axillary dissection for node-negative breast cancer.” J Natl Cancer Inst 84:683–689. Clark, R. M., T. Whelan, M. Levine, R. Roberts, A. Willan, P. McCulloch, M. Lipa, R. H. Wilkinson, and L. J. Mahoney. (1996). “Randomized clinical trial of breast irradiation following lumpectomy and axillary dissection for node-negative breast cancer: An update. Ontario Clinical Oncology Group.” J Natl Cancer Inst 88:1659–1664. Clarke, D. H., F. A. Vicini, and H. Jacobs. “High Dose Rate Brachytherapy for Breast Cancer” in High Dose Rate Brachytherapy: A Textbook. S. Nag (ed.). Armonk, NY: Futura Publishing, pp. 321–329, 1994. Clarke, D., J. L. Curtis, A. Martinez, L. Fajardo, and D. Goffinet. (1983). “Fat necrosis of the breast simulating recurrent carcinoma after primary radiotherapy in the management of early stage breast carcinoma.” Cancer 52:442–445.
6
Cleary, K., V. Watson, D. Lindisch, D. Stoianovici, A. Patriciu, D. Mazilu, R. H. Taylor, G. Fichtinger, S. Xu, C. S. White, J. Donlon, and M. Taylor. (2005). “Precision placement of instruments for minimally invasive procedures using a needle driver robot.” Int J Med Robot Comput Assist Surg 1:40–47. Cohen, G. N., H. I. Amols, and M. Zaider. (2000). “An independednt dose-to-point calculation program for the verification of high-dose-rate brachytherapy treatment planning.” Int J Radiat Oncol Biol Phys 48(4):1251–1258. Cohen, G. N., H. I. Amols, M. J. Zelefsky, and M. Zaider. (2002). “The Anderson nomograms for permanent interstitial prostate implants: A briefing for practitioners.” Int J Radiat Oncol Biol Phys 53(2):504-511. Cohen, L. (1955). “Protracted interstitial irradiation of tumors using 182Ta.” Br J Radiol 28:338. Coleman, D. J., R. H. Silverman, A. Chabi, M. J. Rondeau, K. K. Shung, J. Cannata, and H. Lincoff. (2004b). “High-resolution ultrasonic imaging of the posterior segment.” Ophthalmology 111(7):1344–1351. Coleman, D. J., R. H. Silverman, M. J. Rondeau, H. C. Boldt, H. O. Lloyd, F. L. Lizzi, T. A. Weingeist, X. Chen, S. Vangveeravong, and R. Folberg. (2004a). “Noninvasive in vivo detection of prognostic indicators for high-risk uveal melanoma: Ultrasound parameter imaging.” Ophthalmology 111(3):558–564. Collaborative Ocular Melanoma Study Group. (1999). “Echography (ultrasound) procedures for the collaborative ocular melanoma study. COMS Report No. 12.” J Ophth Nurs Technol Part I 18(4):143–149, Part II 18(5):219–232. Collaborative Ocular Melanoma Study Group. (2001a). “Collaborative Ocular Melanoma Study (COMS) randomized trial of I-125 Brachytherapy for medium choroidal melanoma: I. Visual acuity after 3 years. COMS Report No. 16.” Ophthalmology 108:348–366. Collaborative ocular melanoma study group. (2001b). “the coms Randomized Trial Of Iodine 125 brachytherapy for choroidal melanoma, II: Characteristics of patients enrolled and not enrolled: COMS Report No. 17.” Arch Ophthalmol 119:951–965. Collaborative Ocular Melanoma Study Group. (2001c). “The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma, III: Initial mortality findings; COMS Report No. 18.” Arch Ophthalmol 119:969–982. Collé, R. (1999). “Chemical digestion and radionuclide assay of TiNi-encapsulated 32P intravascular brachytherapy sources.” Appl Radiat Isot 50:811–833. Comité Francais de Mesure des Rayonnements Ionisants (CMFRI). “Recommendations pour la determination des doses absorbees en curiethérapie. Rapport du Comite Francais ‘Mesure des Rayonnements Ionisants’ No. 1.” Paris: Bureau National de Metrologie, 1983. Cooper, J. S. (1978). “Postoperative irradiation of pterygia: Ten more years of experience.” Radiology 128:753–756. Cooper, J. S., and I. A. Lerch. “Postoperative irradiation of pterygia: An unexpected effect of the time/dose relationship.” Radiology 135(3):743–745. Corbett, J. F., J. J. Jezioranski, J. Crook, T. Tran, and I. W. Yeung. (2001). “The effect of seed orientation deviations on the quality of 125I prostate implants.” Phys Med Biol 46(11):2785–2800. Correll, J. W., N. E. Chase, and H. L. Atkins. (1961). “New technique for interstitial irradiation of brain tumors.” J Neurosurg 18:800–803. Coursey, B. M., L. J. Goodman, D. D. Hoppes et al. (1992). “The needs for brachytherapy source calibration in the united States.” Nucl Instr Methods Phys Res 312:246-250. Coutard, H. (1932). “Roentgentherapy of epitheliomas of the tonsillar region, hypopharynx, and larynx, from 1920 to 1926.” Am J Roentgenol 28:313–331, 343–348. Cox, J.D. (ed.). Moss’ Radiation Oncology: Rationale, Technique, Results, 7th edition. St. Louis: MosbyYear Book, 1994. Crile Jr., G., and C. B. Esselstyn Jr. (1990). “Factors influencing local recurrence of cancer after partial mastectomy.” Cleveland Clin J Med 57:143–146. Crook, J. M., B. A. Esche, G. Chaplain, J. Isturiz, I. Sentenac, and J. C. Horiot. (1987). “Dose-volume analysis and the prevention of radiation sequelae in cervical cancer.” Radiother Oncol 8:321–332.
7
Cross, W. G., J. Hokkanen, H. Järvinen, F. Mourtada, P. Sipilä, C. G. Soares, and S. Vynckier. (2001). “Calculation of beta-ray dose distributions from ophthalmic applicators and comparison with measurements in a model eye.” Med Phys 28:1385. Curley, S., P. Marra, K. Beaty, L. Ellis, J. Vauthey, E. Abdalla, C. Scaife, C. Raut, R. Wolff, H. Choi, E. Loyer, P. Vallone, F. Fiore, F. Scordino, V. De Rosa, R. Orlando, S. Pignata, B. Daniele, and F. Izzo. (2004). “Early and late complications after radiofrequency ablation of malignant liver tumors in 608 patients.” Ann Surg 239:450–458. Cygler, J., J. Szanto, M. Soubra, and D. Rogers. (1990). “Effects of gold and silver backings on the dose rate around an 125I seed.” Med Phys 17:172–178. D’Souza, W. D., and R. R. Meyer. (2001). “Dose homogeneity as a function of source activity in optimized I-125 prostate implant treatment plans.” Int J Radiat Oncol Biol Phys 51:1120–1130. D’Souza, W. D., R. R. Meyer, B. R. Thomadsen, and M. C. Ferris. (2001). “An iterative sequential mixed-integer approach to automated prostate brachytherapy treatment plan optimization.” Phys Med Biol 46(2):297–322. Daftari, I., D. Barash, S. Lin, and J. O’Brien. (2001). “Use of high-frequency ultrasound imaging to improve delineation of anterior uveal melanoma for proton irradiation.” Phys Med Biol 46(2):579–590. Dale, R. G. (1983). “Some theoretical deviations relating to the tissue dosimetry of brachytherapy nuclides, with particular reference to iodine-125.” Med Phys 10:176-183. Dale, R. G. (1985). “The application of the linear-quadratic doseeffect equation to fractionated and protracted radiotherapy.” Br J Radiol 58:515–528. Damore, S. J., A. M. N. Syed, A. Puthawala, and A. Sharma. (2000). “Needle displacement during HDR brachytherapy in the treatment of prostate cancer.” Int J Radiat Oncol Biol Phys 46(5):1205–1211. Das, R. K., R. Patel, H. Shah, H. Odau, and R. R. Kuske. (2004). “3D CT-based high-dose-rate breast brachytherapy implants: treatment planning and quality assurance.” Int J Radiat Oncol Biol Phys 59:1224–1228. Daskalov, G. M., E. Loffler, and J. F. Williamson. (1998). “Monte Carlo-aided dosimetry of a new high dose-rate brachytherapy source.” Med Phys 25(11):2200–2208. Daskalov, G. M., R. S. Baker, D. W. Rogers, and J. F. Williamson. (2000). “Dosimetric modeling of the microselectron high-dose rate 192Ir source by the multigroup discrete ordinates method.” Med Phys 27(10):2307–2319. Daskalov, G. M., R. S. Baker, D. W. Rogers, and J. F. Williamson. (2002). “Multigroup discrete ordinates modeling of 125I 6702 seed dose distributions using a broad energy-group cross section representation.” Med Phys 29:113–124. Dattoli, M., and K. A. Walker. (1997). “Simple method to stabilize the prostate during transperineal prostate brachytherapy.” Int J Radiat Oncol Biol Phys 388:341–342. Davis, B. J., Brinkmann, D. H., Kruse, J. J., Herman, M. G., LaJoie, W. N., Schwart, D. J., Pisansky, T. M., and R. W. Kline. (2004). “Selective identification of different brachytherapy sources, ferromagnetic seeds, and fiducials in the prostate using an automated seed sorting algorithm.” Brachytherapy 3:106–112. Davis, B. J., R. R. Kinnick, M. Fatemi, E. P. Lief, R. A. Robb, and J. F. Greenleaf. (2003). “Measurement of the ultrasound backscatter signal from three seed types as a function of incidence angle: application to permanent prostate brachytherapy.” Int J Radiat Oncol Biol Phys 57(4):1174–1182. Davis, K. W., J. J. Choi, and D. G. Blankenbaker. (2004). “Radiofrequency ablation in the musculoskeletal system.” Semin Roentgenol 39(1):129–144. de la Zerda, A., S. Chiu-Tsao, H. S. Tsao, J. Lin, and J. H. Kim. (1992). “Effect of a COMS eye plaque on 125 I dose distribution, in particular reference to the penumbra characteristics.” (Abstract). Med Phys 19:776. de la Zerda, A., S. Chiu-Tsao, H. S. Tsao, J. Lin, L. Boulay, I. Kanna, and J. H. Kim. (1996). “125I eye plaque dose distribution including penumbra characteristics.” Med Phys 23:407–418. Deasy, J. O. (1997). “Multiple local minima in radiotherapy optimization problems with dose-volume constraints.” Med Phys 24:1157–1161.
8
Deasy, J. O., and C. G. Soares. (1994). “Extrapolation chamber measurements of 90Sr+90Y beta-particle ophthalmic applicator dose rates.” Med Phys 21:91. Decadt, B., and A. K. Siriwardena. (2004). “Radiofrequency ablation of liver tumours: Systematic review.” Lancet Oncol 5(9):550–560. DeJong, K. A. “Genetic Algorithms Are NOT Function Optimizers” in Foundations of Genetic Algorithms: 2. E. D. Whitley (ed.). FOGA Proceedings July 24–29, 1992, Vail, Colorado. San Mateo, CA: Morgan Kaufman, pp. 5–18, 1993. Delclos, L., G. H. Fletcher, E. B. Moore, and V. A. Sampiere. (1980). “Minicolpostats, dome cylinders, other additions and improvements of the Fletcher-Suit afterloadable system: Indications and limitations of their use.” Int J Radiat Oncol Biol Phys 6:195–1206. Delclos, L., G. H. Fletcher, V. Sampiere, and W. H. Grant. (1978). “Can the Fletcher gamma ray colpostat system be extrapolated to other systems?” Cancer 41:970–979. DeMarco, J. J., G. Hugo, and T. D. Solberg. (2002). “Dosimetric parameters for three low-energy brachytherapy sources using the Monte Carlo N-particle code.” Med Phys 29:662–668. Dempsey, J. F., J. A. Williams, J. B. Stubbs, T. J. Patrick, and J. F. Williamson. (1998). “Dosimetric properties of a novel brachytherapy balloon applicator for the treatment of malignant brain-tumor resection-cavity margins.” Int J Radiat Oncol Biol Phys 42:421–429. Devine, M. (2004). Private communication. Isotope Products Laboratories, Valencia, CA. DeWerd, L. A. “Source Standardization and Calibration for Intravascular Brachytherapy” in Intravascular Brachytherapy/Fluoroscopically Guided Interventions. S. Balter, R. C. Chan, and T. B. Shope, Jr. (eds.). American Association of Physicists in Medicine 2002 Summer School proceedings. Medical Physics Monograph 28. Madison, WI: Medical Physics Publishing, pp. 423–443, 2002. DeWerd, L. A. Brachytherapy Dosimetric Assessment: Source Calibration, Radiological Society of North America, Categorical Course in Brachytherapy Physics. B. Thomadsen (ed.). Oak Brook, IL: RSNA, 1997. DeWerd, L. A., and B. R. Thomadsen. “Source Strength Standards and Calibration of HDR/PDR Sources” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.). American Association of Physicists in Medicine (AAPM) 1994 Summer School. Madison, WI: Medical Physics Publishing, 1995. DeWerd, L. A., M. S. Huq, I. J. Das, G. S. Ibbott, W. F. Hanson, T. W. Slowey, J. F. Williamson, and B. M. Coursey. (2004). “Procedures for establishing and maintaining consistent air-kerma strength standards for low-energy, photon-emitting brachytherapy sources: Recommendations of the Calibration Laboratory Accreditation Subcommittee of the American Association of Physicists in Medicine.” Med Phys 31:675–681. DeWerd, L. A., S. M. DeWerd, and F. H. Attix. (1993). Solution to Inverse Square Equations Involving Distance Error and Scatter Correction. Wisconsin Medical Physics Report #197 Madison; WI. (Available from the University of Wisconsin.) DeWitt, K. D., I.-C. Hsu, V. K. Weinberg, É. Lessard, and J. Pouliot. (2004). “3-D inverse treatment planning for the tandem and ovoid applicator in cervical cancer.” Int J Radiat Oncol Biol Phys, In press. Deye, James A. “Codes and Regulations, Radiation” in Encyclopedia of Medical Devices and Instruments. John G. Webster (Ed.). New York: John Wiley & Sons, 1988. DiBiase, S. J., K. Wallner, K. Tralins, and S. Sutlief. (2000). “Brachytherapy radiation doses to the neurovascular bundles.” Int J Radiat Oncol Biol Phys 46(5):1301–1307. Dickler, A., M. Kirk, J. Choo, W. C. Hsi, J. Chu, K. Dowlatshahi, D. Francescatti, and C. Nguyen. (2004). “Treatment volume and dose optimization of MammoSite breast brachytherapy applicator.” Int J Radiat Oncol Biol Phys 59:469–474. Diffey, B. L., and S. C. Klevenhagen. (1975). “An experimental and calculated dose distribution in water around CDC-K type cesium-137 sources.” Med Phys 20:446. Dinsmore, M., J. C. Yanch, A. P. Sliski, and K. J. Harte. (1994). “New x-ray generator for interstitial radiotherapy.” Trans Am Nucl Soc 70:24–25.
9
Dinsmore, M., K. J. Harte, A. P. Sliski, D. O. Smith, P. M. Nomikos, M. J. Dalterio, A. J. Boom, W. F. Leonard, P. E. Oettinger, and J. C. Yanch. (1996). “A new miniature x-ray source for interstitial radiosurgery: Device description.” Med Phys 23:45–52. Dominici, H. (1907). “Modifications histologiques déterminées par le rayonnement du radium.” Arch D’elect Méd 15:835. Douglas, R. M., J. Beatty, K. Gall, R. F. Valenzuela, P. Biggs, P. Okunieff, and F. S. Pardo. (1996). “Dosimetric results from a feasibility study of a novel radiosurgical source for irradiation of intracranial metastases.” Int J Radiat Oncol Biol Phys 36:443–450. Douysset, G., J. Gouriou, F. Delaunay, L. DeWerd, K. Stump, and J. Micka. (2005). “Comparison of dosimetric standards of USA and France for HDR brachytherapy.” Phys Med Biol 50:1961–1978. Dove, D. B. (1959). “Effect of dosemeter size on measurements close to a radioactive source.” Br J Radiol 62:202–204. Dubois, D. F., B. R. Prestidge, L. A. Hotchkiss, W. S. Bice, and J. J. Prete. (1997). “Source localization following permanent transperineal prostate interstitial brachytherapy using magnetic resonance imaging.” Int J Radiat Oncol Biol Phys 39(5):1037–1041. Dubois, D. F., W. S. Bice, and B. R. Prestidge. (2001). “CT and MRI derived source localization error in a custom prostate phantom using automated image coregistration.” Med Phys 28(11):2280–2284. Duck, F. A. Physical Properties of Tissue: A Comprehensive Reference Book. London: Academic Press, 1990. Duggan, D. M., C. W. Coffey II, and S. Levi. (1998). “Dose distribution for a 32P-impregnated coronary stent: Comparison of theoretical calculations and measurements with radiochromic film.” Int J Radiat Oncol Biol Phys 40:713–720. Durrance, F. Y., and G. H. Fletcher. (1968). “Computer calculation of dose contribution to regional lymphatics from gynecological radium insertions.” Radiology 91:140–148. Dutreix, A., and A. Wambersie. (1968). “Étude de la repartition des doses autour de sources ponctu elles alignées.”Acta Radiol 7:389–400. Dutreix, A., and A. Wambersie. (1975). “Specification of gamma-ray brachytherapy sources.” Br J Radiol 48:1034. Dutreix, A., and G. Marinello. “The Paris System.” Chapter 5 in Modern Brachytherapy. Chicago, IL: Year Book Medical Publishers, 1987. Dutreix, A., G. Marinello, and A. Wambersie. Dosimetrie en Curiethérapie. Paris: Masson, 1982. Edmundson, G. K. “Geometry-based Optimization for Stepping Source Implants” in Brachytherapy HDR and LDR. A. A. Martinez, C. G. Orton, and R. F. Mould, (eds.). Columbia, MD: Nucletron, pp. 184–192, 1990. Edmundson, G. K. “Volume Optimization: An American Viewpoint” in Brachytherapy: From Radium to Optimization. R. J. Mould and J. Batterman (eds.). Columbia, MD: Nucletron, pp. 314–318, 1994. Edmundson, G. K., D. Yan, and A. Martinez. (1995). “Intraoperative optimization of needle placement and dwell times for conformal prostate brachytherapy. Int J Radiat Oncol Biol Phys 33:1257–1263. Edmundson, G. K., F. A. Vicini, P. Y. Chen, C. Mitchell, and A. A. Martinez. (2002). “Dosimetric characteristics of the MammoSite RTS, a new breast brachytherapy applicator.” Int J Radiat Oncol Biol Phys 52:1132–1139. Edmundson, G. K., N. R. Rizzo, M. Teahan, D. Brabbins, F. A. Vicini, and A. Martinez. (1993). “Concurrent treatment planning for outpatient high dose rate prostate template implants.” Int J Radiat Oncol Biol Phys 27(5):1215–1223. Erickson, B., R. Jones, J. Rownd, K. Albano, and M. Gillan. (2000). “Is the tandem and ring applicator a suitable alternative to the high dose rate Selectron tandem and ovoid applicator?” J Brachytherapy Int 16:131–144. Ezzell, G. A. (1994). “Quality assurance of treatment plans for optimized high dose rate brachytherapy—planar implants.” Med Phys 21:659–661. Ezzell, G. A. (2000a). “Limitations of the straight-line assumption for endobronchial HDR brachytherapy treatments.” Med Phys 27(1):151–153. Ezzell, G. A. (2000b). “A manual algorithm for computing dwell times for two-catheter endobronchial treatments using HDR brachytherapy.” Med Phys 27(5):1030–1033.
10
Ezzell, G. A., and R. W. Luthmann. “Clinical Implementation of Dwell Time Optimization Techniques for Single Stepping-Source Remote Applicators” in Brachytherapy Physics, J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.). AAPM Summer School 1994. Madison, WI: Medical Physics Publishing, pp. 617–639, 1995. Farrell, M. A., W. J. Charboneau, D. S. DiMarco, G. K. Chow, H. Zincke, M. R. Callstrom, B. D. Lewis, R. A. Lee, and C. C. Reading. (2003). “Imaging-guided radiofrequency ablation of solid renal tumors.” Am J Roentgenol 180(6):1509–1513. Farrow, D. C., W. C. Hunt, and J. M. Samet. (1992). “Geographic variation in the treatment of localized breast cancer.” N Engl J Med 326:1097–1101. Faverly, D. R., L. Burgers, P. Bult, and R. Holland. (1994). “Three dimensional imaging of mammary ductal carcinoma in situ: Clinical implications.” Semin Diagn Pathol 11:193–198. Fentiman, I. S., C. Poole, D. Tong, P. J. Wintr, H. M. Mayles, P. Turner, M. A. Chaudary, and R. D. Rubens. (1991). “Iridium implant treatment without external radiotherapy for operable breast cancer; a pilot study.” Eur J Cancer 27:447–450. Feygelman, V., J. L. Friedland, R. M. Sanders, B. K. Noriega, and J. M. Pow-Sang. (1996a). “Improvement in dosimetry of ultrasound-guided prostate implants with the use of multiple stabilization needles.” Med Dosim 21(2):109–112. Feygelman. V., B. K. Noriega, R. M. Sanders, and J. L. Friedland. (1996). “A simple method for verifying activity of iodine-125 seeds in rigid absorbable suture.” Med Dosim 21: 261–262. Fichtinger, G., A. Deguet, G. Fischer, E. Balogh, K. Masamune, R. H. Taylor, L. M. Fayad, and S. J. Zinreich. (2005a). “CT image overlay for percutaneous needle insertions.” Comput Assist Surg (in review). Fichtinger, G., A. Krieger, R. C. Susil, A. Tanacs, L. L. Whitcomb, and E. Atalar. “Transrectal Prostate Biopsy Inside Closed MRI Scanner with Remote Actuation, Under Real-Time Image Guidance.” Proceedings to Medical Image Computing and Computer-Assisted Interventions 2002. Lecture Notes in Computer Science, Vol. 2488, Part 1. New York: Springer-Verlag, pp. 91–98, 2002. Fichtinger, G., E. C. Burdette, A. Tanacs, A. Patriciu, D. Mazilu, L. L. Whitcomb, and D. Stoianovici. (2005b). “Robotically assisted prostate brachytherapy with transrectal ultrasound guidance – preliminary experiments.” Brachytherapy (in review). Fichtinger, G., T. L. DeWeese, A. Patriciu, A. Tanacs, A. Mazilu, J. H. Anderson, K. Masamune, R. H. Taylor, and D. Stoianovici. (2004). “Robotically assisted prostate biopsy and therapy with intraoperative CT guidance.” Acad Radiol 9:60–74. Fisch, B. M., B. Pickett, V. Weinberg, and M. Roach. (2001). “Dose of radiation received by the bulb of the penis correlates with risk of impotence after three-dimensional conformal radiotherapy for prostate cancer.” Urology 57(5):955–959. Fischell, T. A., A. J. Carter, D. R. Fischell, M. T. Foster, and R. E. Fischell. “Radioisotope Stents” in Handbook of Vascular Brachytherapy. R. Waksman and P. W. Serruys. London: Martin Dunitz, Ltd., pp. 59–68, 1998. Fisher, B., C. Redmond, R. Poisson, R. Margolese, N. Wolmark, L. Wickerham, E. Fisher, M. Deutsch, R. Caplan, Y. Pilch, et al. (1989). “Eight-year results of a randomized clinical trial comparing total mastectomy and lumpectomy with or without radiation in the treatment of breast cancer.” N Engl J Med 320:822–828. Fleming, P., A. M. Nisar Syed, D. Neblett, A. Puthawala, F. W. George 3rd, and D. Townsend. (1980). “Description of an afterloading 192Ir interstitial-intracavitary technique in the treatment of carcinoma of vagina.” Obstet Gynecol 55:525–535. Fletcher, G. H. (1953). “Cervical radium applicators with screening in the direction of bladder and rectum.” Radiology 60:77–84. Fletcher, G. H. Textbook of Radiotherapy. Third Edition. Philadelphia, PA: Lea & Febiger, 1980. Flocks, R. H., H. D. Kerr, H. B. Elkins, and D. Culp. (1952). “Treatment of carcinoma of the prostate by interstitial radiation with radioactive gold (Au-198): A preliminary report.” J Urol 68:510–522. Fong, Y., J. Fortner, R. L. Sun, M. F. Brennan, and L. H. Blumgart. (1999). “Clinical score for predicting recurrence after hepatic resection for metastatic colorectal cancer: Analysis of 1001 consecutive cases.” Ann Surg 230(3):309–318; discussion 318–321.
11
Forman, M. R., P. A. Lovoi, T. W. Rusch. (2002). Radiation for Inhibiting Hyperplasia After Intravascular Intervention. U.S. Patent 6,390,967. Fornage, B. D., N. Sneige, M. I. Ross, A. N. Mirza, H. M. Kuerer, B. S. Edeiken, F. C. Ames, L. A. Newman, G. V. Babiera, and S. E. Singletary. (2004). “Small (< or = 2-cm) breast cancer treated with US-guided radiofrequency ablation: Feasibility study.” Radiology 231(1):215–224. Forrest, A. P., H. J. Stewart, D. Everington, R. J. Prescott, C. S. McArdle, A. N. Harnett, D. C. Smith, and W. D. George. (1996). “Randomised controlled trial of conservation therapy for breast cancer: 6-year analysis of the Scottish trial. Scottish Cancer Trials Breast Group.” Lancet 348:708–713. Forrest, S. (1993). “Genetic algorithms: Principles of natural selection applied to computation.” Science 261:872–878. Fowler, J. E., W. Barzell, B. S. Hilaris, and W. F. Whitmore. (1979). “Complications of 125iodine implantation and pelvic lymphadenectomy in the treatment of prostatic cancer.” J Urol 121:447–451. Fowler, J. F. (1989). “The linear-quadratic formula and progress in fractionated radiotherapy.” Br J Radiol 62:679–694. Fowler, J., and M. Mount. (1992). “Pulsed brachytherapy: The conditions for no significant loss of therapeutic ratio compared with traditional low dose rate brachytherapy.” Int J Radiat Oncol Biol Phys 23:661–669. Fraass, B., K. Doppke, M. Hunt, G. Kutcher, G. Starkschall, R. Stern, and J. Van Dyk. (1998). “American Association of Physicists in Medicine Radiation Therapy Committee Task Group 53: Quality assurance for clinical radiotherapy treatment planning.” Med Phys 25:1773–1829. Also available as AAPM Report No. 62. Fujita, M., Y. Hirokawa, K. Kashiwado, Y. Akagi, K. Kashimoto, H. Kiriu, K. Ohtani, and T. Wada. (1996). “An analysis of mandibular bone complications in radiotherapy for T1 and T2 carcinoma of the oral tongue.” Int J Radiat Oncol Biol Phys 24:333–339. Fukushima, S., T. Inoue, T. Inoue, and S. Ozeki. (1999). “Postoperative irradiation of pterygium with 90Sr eye applicator.” Int J Radiat Oncol Biol Phys 43:597–600. Gabor, J., O. E. Streeter, and M. A. Astrahan. (2003). “The use of linear programming in optimization of HDR implant dose distributions.” Med Phys 30(5):751–760. Gabriel, S., R. W. Lau, and C. Gabriel. (1996). “The dielectric properties of biological tissues: II. Measurements in the frequency range 10 Hz to 20 GHz.” Phys Med Biol 41(11):2251–2269. Gaspar, L. E., S. Nag, A. Herskovic, R. Mantravadi, and B. Speiser. (1997). “American Brachytherapy Society (ABS) consensus guidelines for brachytherapy of esophageal cancer.” Int J Radiat Oncol Biol Phys 38(1):127–132. Geman, S., and D. Geman. (1984). “Stochastic relaxation, Gibbs distributions, and Bayesian restoration of images.” IEEE Trans Pattern Anal Mach Intell (PAMI) 6:721–741. Gerbaulet, A. and A. Bridier. “Institut Gustave Roussy Method” in A Practical Manual of Brachytherapy by B. Pierquin and G. Marinello. F. Wilson, B. Erickson, and J. Cunningham (trans.). Madison, WI: Medical Physics Publishing, pp. 176–181, 1997. Giannouli, S., N. Milickovic, D. Baltas, M. Lahanas, N. Uzunoglu, C. Kolotas, and N. Zamboglou. (2000). “Autoactivation of source dwell positions for HDR brachytherapy treatment planning.” Med Phys 27(11):2517–2520. Gibb, R., and J. B. Massey. (1980). “Radium dosage: SI units and the Manchester system.” Br J Radiol 53:1100–1101. Gifford, K. A 3-D CT Assisted Monte Carlo Evaluation of Intracavitary Brachytherapy Implants. Ph.D. dissertation. The University of Texas Graduate School of Biomedical Sciences. Houston, TX, 2004. Gifford, K., J. Horton, E. Jackson, T. Steger, M. Heard, F. Mourtada, A. Lawyer, and G. Ibbott. (2005).”Verification of Monte Carlo calculations around a Fletcher Suit Delclos ovoid with radiochromic film and normoxic polymer gel dosimetry.” Med Phys (accepted for publication). Gillin, M. T., R. W. Kline, J. F. Wilson, and J. D. Cox. (1984). “Single and double plane implants: A comparison of the Manchester System with the Paris System.” Int J Radiat Oncol Biol Phys 10:921–925. Glasgow, G. P. (2002a). “Nuclear Regulatory Commission regulatory status of approved intravascular brachytherapy systems.” Cardiovasc Radiat Med 3:1–11.
12
Glasgow, G. P. (2002b). “Is an Ir-192 permanent seed implant feasible for prostate brachytherapy?” Brachytherapy 1:195–203. Glasgow, G. P., J. D. Bourland, P. W. Grisby, J. A. Meli, and K. A. Weaver. “Remote Afterloading Technology.” AAPM Report No. 41. New York: American Institute of Physics, 1993. Godden, T. J. The Physical Aspects of Brachytherapy. Medical Physics Handbook 19. Bristol: Adam Hilger, 1988. Godden, T. L. Physical Aspects of Brachytherapy. Philadelphia: Adam Hilger, 1988. Goetsch, S. J., and K. A. Sunderland. (1991). “Surface dose rate calibration of Sr-90 plane ophthalmic applicators.” Med Phys 18:161. Goetsch, S. J., F. H. Attix, D. W. Pearson, and B. R. Thomadsen. (1991). “Calibration of 192Ir high doserate afterloading systems.” Med Phys 18:462–467. Goetsch, S. J., F. H. Attix, L. A. DeWerd, and B. R. Thomadsen. (1992). “A new well ionization chamber for the calibration of iridium-192 high dose rate sources.” Int J Radiat Oncol Biol Phys 24:167–170. Goetz, M., M. Callstrom, J. Charboneau, M. Farrell, T. Maus, T. Welch, G. Wong, J. Sloan, P. Novotny, I. Petersen, R. Beres, D. Regge, R. Capanna, M. Saker, D. Gronemeyer, A. Gevargez, K. Ahrar, M. Choti, T. de Baere, and J. Rubin. (2004). “Percutaneous image-guided radiofrequency ablation of painful metastases involving bone: A multicenter study.” J Clin Oncol 22:300–306. Goldberg, S. N., G. S. Gazelle, L. Solbiati, W. J. Rittman, and P. R. Mueller. (1996). “Radiofrequency tissue ablation: Increased lesion diameter with a perfusion electrode.” Acad Radiol 3(8):636–644. Goldberg, S. N., G. S. Gazelle, S. L. Dawson, W. J. Rittman, P. R. Mueller, and D. I. Rosenthal (1995). “Tissue ablation with radiofrequency: Effect of probe size, gauge, duration, and temperature on lesion volume.” Acad Radiol 2(5):399–404. Goldberg, S. N., J. W. Charboneau, G. D. Dodd 3rd, D. E. Dupuy, D. A. Gervais, A. R. Gillams, R. A. Kane, F. T. Lee Jr., T. Livraghi, J. P. McGahan, H. Rhim, S. G. Silverman, L. Solbiati, T. J. Vogl, and B. J. Wood. (2003). “Image-guided tumor ablation: Proposal for standardization of terms and reporting criteria.” Radiology 228(2):335–345. Goldberg, S. N., L. Solbiati, P. F. Hahn, E. Cosman, J. E. Conrad, R. Fogle, and G. S. Gazelle. (1998). “Large-volume tissue ablation with radio frequency by using a clustered, internally cooled electrode technique: Laboratory and clinical experience in liver metastases.” Radiology 209(2):371–379. Goldberg, S. N., P. F. Saldinger, G. S. Gazelle, J. C. Huertas, K. E. Stuart, T. Jacobs, and J. B. Kruskal. (2001). “Percutaneous tumor ablation: increased necrosis with combined radio-frequency ablation and intratumoral doxorubicin injection in a rat breast tumor model.” Radiology 220(2):420–427. Gong, L., P. S. Cho, B. H. Han, K. E. Wallner, S. G. Sutlief, S. D. Pathak, D. R. Haynor, and Y. Kim. (2002). “Ultrasonography and fluoroscopic fusion for prostate brachytherapy dosimetry.” Int J Radiat Oncol Biol Phys 54(5):1322–1330. Gray, L. H. (1936). “An ionization method for the absolute measurement of gamma-ray energy.” Proc Roy Soc A156:578-596. Green, S. B., W. R. Shapiro, P. C. Burger, R. G. Selker, J. C. VanGilder, S. Saris, M. G. Malkin, J. Mealy, J. Neal, J. T. Robertson, and J. Olson. “A Randomized Trial of Interstitial Radiotherapy (RT) Boost for Newly Diagnosed Malignant Glioma: Brain Tumor Cooperative Group (BTCG) Trial 8701.” Proceedings of the American Society of Clinical Oncology, Dallas, TX, p. 174, 1994. Griffin, S. L., L. A. DeWerd, J. A. Micka, and T. D. Bohm. (2005). “The effect of ambient pressure on well chamber response: Experimental results with empirical correction factors.” Med Phys 32:700–709. Grigsby, P, B. Siegel, and F. Dehdashti. (2001). “Lymph node staging by positron emission tomography in patients with carcinoma of the cervix.” J Clinical Oncol 19:3745–3749. Grigsby, P. W., J. F. Williamson, and C. A. Perez. (1992). “Source configuration and dose rates for the Selectron afterloading equipment for gynecologic applicators.” Int J Radiat Oncol Biol Phys 24(2):321–327. Grigsby, P.W., A. Georgiou, J. F. Williamson, and C. A. Perez. (1993). “Anatomic variation of gynecological brachytherapy prescription points.” Int J Radiat Oncol Biol Phys 27:725–729. Guinot, J. L., L. Arribas, M. L. Chust, J. L. Mengual, E. Garcia Miragall, M. Carrascosa, P. P. Escolar, V. Crispin, and C. Guardino. (2003). “Lip cancer treatment with high dose rate brachytherapy.” Radiother Oncol 69(1):113–115.
13
Gupta, S., S. Nag, and J. Gupta. (1993). “Pulmonary embolism of permanently implanted radioactive iodine-125 seeds.” Endocuriether/Hypertherm Oncol 9:27–31. Gutin, P. H., and R. H. Dormandy. (1982). “A coaxial catheter system for afterloading radioactive sources for interstitial irradiation of brain tumors.” J Neurosurg 56:734–735. Gutin, P. H., T. L. Phillips, and Y. Hosobuchi. (1981). “Permanent and removable implants for the brachytherapy of brain tumors.” Int J Radiat Oncol Biol Phys 7:1371–1381. Gutman, G. (2003a). Apparatus and Method for Generating a High Intensity X-Ray Beam with a Selectable Shape and Wavelength. U.S. Patent 6,493,421. Gutman, G. (2003b). X-Ray System with Implantable Needle for Treatment of Cancer. U.S. Patent 6,580,940. Gutman, G., E. Sozontov, E. Strumban, F.-F. Yin, S.-W. Lee, and J. H. Kim. (2004). “A novel needlebased miniature x-ray generating system.” Phys Med Biol 49:4677–4688. Gutman, G., F.-F. Yin, E. Strumban, P. Petrashen, and J. H. Kim. (2003). “A device for intensitymodulated x-ray brachytherapy (IMXBT).” (Abstract). Med Phys 30:1469. Haas, J. S., R. D. Dean, and C. M. Mansfield. (1985). “Dosimetric comparison of the Fletcher family of gynecologic colpostats 1950–1980.” Int J Radiat Oncol Biol Phys 11:1317–1321. Haemmerich, D., A. W. Wright, D. M. Mahvi, F. T. Lee Jr., and J. G. Webster. (2003). “Hepatic bipolar radiofrequency ablation creates coagulation zones close to blood vessels: A finite element study.” Med Biol Eng Comput 41(3):317–323. Haemmerich, D., F. T. Lee Jr., D. J. Schutt, L. A. Sampson, J. G. Webster, J. P. Fine, and D. M. Mahvi. (2005). “Large-volume radiofrequency ablation of ex vivo bovine liver with multiple cooled cluster electrodes.” Radiology 234(2):563–568. Hakim, R., N. T. Zervas, F. Hakim, W. E. Butler, J. Beatty, J. C. Yanch, P. J. Biggs, K. P. Gall, A. P. Sliski. (1997). “Initial characterization of the dosimetry and radiology of a device for administering interstitial stereotactic radiosurgery.” Neurosurg 40:510–516. Hall, E. J. Radiobiology for the Radiologist. New York: Harper and Row Publishers, 1994. Hall, E. J., R. Oliver, and B. J. Shepstone. (1966). “Routine dosimetry with tantalum-182 and iridium-192 wires.” Acta Radiol 4:155. Hanson, W. F., and M. Graves. “ICRU Recommendations on Dose Specification in Brachytherapy” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds). Madison, WI: Medical Physics Publishing, pp. 361–378, 1995. Harrison, L.B., B. D. Minsky, W. E. Enker, B. Mychalczak, J. Guillem, P. B. Paty, L. L. Anderson, C. White, and A. M. Cohen. (1998). “High dose rate intraoperative radiation therapy (HDR-IORT) as part of the management strategy for locally advanced primary and recurrent rectal cancer.” Int J Radiat Oncol Biol Phys 42(2):325–330. Hawighorst, H., S. O. Schoenber, P. G. Knapstein, M. V. Knopp, U. Schaeffer, M. Essig, and G. van Kaick. (1998). “Staging of invasive cervical carcinoma and of pelvic lymph nodes by high resolution MRI with a phased-array coil in comparison with pathological findings.” J Comput Assist Tomogr 22:75–81. Hayashi, A. H., S. F. Silver, N. G. van der Westhuizen, J. C. Donald, C. Parker, S. Fraser, A. C. Ross, and I. A. Olivotto. (2003). “Treatment of invasive breast carcinoma with ultrasound-guided radiofrequency ablation.” Am J Surg 185(5):429–435. Health Physics News and Notices (2002). “IAEA action plan to combat nuclear terrorism.” Health Physics 82:908–909. Health Physics News and Notices (2003). “IAEA and UPU join forces to protect mail.” Health Physics 84:129–130. Heintz, B. H., R. E. Wallace, and J. M. Hevezi. (2001). “Comparison of I-125 sources used for permanent interstitial implants.” Med Phys 28:671–682. Hendee, W. R., and G. S. Ibbott. Radiation Therapy Physics, Second Edition. St. Louis, MO: Mosby, 1996. Henschke, U. K. (1960). “‘Afterloading’ applicator for radiation therapy of carcinoma of the uterus.” Radiology 74:834. Henschke, U. K., and D. C. Lawrence. (1965). “Cesium-131 seeds for permanent implants.” Radiology 85:1117–1119.
14
Heyman, J. (1924). “Technique and results in the treatment of carcinoma of the uterine cervix at “Radiumhemmet” Stockholm.” J Obstet Gynecol Brit Emp 31:1–19. Higson, G. R. Medical Device Safety: The Regulation of Medical Devices for Public Health and Safety. Bristol, England: Institute of Physics Publishing, 2001. Hilaris, B., D. Nori, and L. L. Anderson. Atlas of Brachytherapy. New York: Macmillan Publishing Company, 1988. Hillstead, R. A., C. R. Johnson, and T. D. Weldon. “The Beta-Cath™ System” in Handbook of Vascular Brachytherapy. R. Waksman and P. W. Serruys. London: Martin Dunitz, Ltd.. pp. 41–52, 1998. Hinshaw, J. L., D. L. Chicks, P. F. Laeseke, T. C. I. Winter, and F. T. Lee Jr. (2005). “RF Ablation of Peripheral Liver Tumors: Intraperitoneal 5% Dextrose in Water (D5W) Decreases Post-procedure Pain.” Annual Meeting of the Society of Gastrointestinal Radiology, San Antonio, TX. Hitchon, P. W., J. C. VanGilder, B. C. Wen, and S. K. Jani. (1992). “Brachytherapy for malignant recurrent and untreated gliomas.” Stereotact Funct Neurosurg 59:174–178. Hitchon, P. W., S. K. Jani, J. C. Van Gilder, J. C. Godersky, and J. F. Doornbos. (1987). “Interstitial radiation in recurrent gliomas.” Appl Neurophysiol 50:292–294. Holland, J. H. Adaptation in Natural and Artificial Systems. University of Michigan, Ann Arbor. 1975. Holland, R., J. L. Connolly, R. Gelman, M. Mravunac, J. H. Hendriks, A. L., Verbeek, S. J. Schnitt, B. Silver, J. Boyages, and J. R. Harris. (1990). “The presence of an extensive intraductal component following a limited excision correlates with prominent residual disease in the remainder of the breast.” J Clin Oncol 8:113–118. Holm, H. H., N. Juul, J. F. Pedersen, H. Hansen, and I Stroyer. (1983). “Transperineal 125-iodine seed implantation in prostatic cancer guided by transrectal ultrasonography.” J Urol 130:283–286. (2000) J Urol 167:985–988. Discussion 988–989. Holupka, E. J., P. M. Meskell, E. C. Burdette, and I. D. Kaplan. (2004). “An automatic seed finder for brachytherapy CT postplans based on the Hough transform.” Med Phys 31(9):2672–2679. Horiuchi, J., M. Takeda, H. Shibuya, S. Matsumoto, M. Hoshina, and S. Suzuki. (1991). “Usefulness of 198 Au grain implants in the treatment of oral and oropharyngeal cancer.” Radiother Oncol 21(1):29–38. Horwitz, E. M., A. J. Frazier, F. A. Vicini, D. H. Clarke, G. K. Edmundson, R. D. Keidan, G. S. Gustafson, C. F. Dmuchowski, and A. A. Martinez. (1997). “The impact of temporary iodine-125 interstitial implant boost in the primary management of squamous cell carcinoma of the oropharynx.” Head Neck 19(3):219–226. Hosobuchi, Y., T. L. Philips, T. A. Stupar, and P. H. Gutin. (1980). “Interstitial brachytherapy of primary brain tumors. preliminary report.” J Neurosurg 53:613–617. Hricak, H., and K. Yu. (1996). “Radiology in invasive cervical cancer.” Am J Roentgenol 167:1101–1108. Hricak, H., B. Hamm, R. C. Semelka, C. E. Cann, T. Nauert, E. Secaf, J. L. Stern, and K. J. Wolf. (1991). “Carcinoma of the uterus: Use of gadopentetate dimeglumine in MR imaging.” Radiology 181:95–106. Hsu, I.-C., É. Lessard, V. Weinberg, and J. Pouliot. (2004). ”Comparison of inverse planning simulated annealing and geometrical optimization for prostate high-dose-rate brachytherapy.” Brachytherapy 3:147–152. Hsu, I.-C., J. Speight, J. Hai, É. Vigneault, T. Phillips, and J. Pouliot. (2002). “A comparison between tandem and ovoids and interstitial gynecologic template brachytherapy dosimetry using a hypothetical computer model.” Int J Radiat Oncol Biol Phys 52(2):538–543. Huang, K., P. K. Sneed, M. S. Berger, S. Kunwar, D. A. Larson, W. M. Wara, and M. W. McDermott. (2002). “Permanent iodine-125 brachytherapy for brain metastasis resection cavities.” Int J Radiat Oncol Biol Phys 50:80 (Poster presentation). IEC 60601-2-17 http://www.astro.org/publications/astronews/2005/Jan/IECReport.htm last accessed May 1, 2005. Imamura, H., S. Haga, T. Shimizu, O. Watanabe, J. Kinoshita, H. Nagumo, T. Kajiwara, and M. Aiba. (2000). “Relationship between the morphological and biological characteristics of intraductal components accompanying invasive ductal breast carcinoma and patient age.” Breast Cancer Res Treat 62:177–184.
15
Ingber, L. (1996). “Adaptive simulated annealing (ASA): Lessons learned.” Control and Cybernetics 25(1):33–54. International Atomic Energy Agency (IAEA) IAEA-TECDOC-1079. “Calibration of Brachytherapy Sources.” Vienna, Austria: IAEA, 1999. International Atomic Energy Agency (IAEA). TECDOC-1274. “Dosimetry and Medical Radiation Physics Section, IAEA, Calibration of photon and beta ray sources used in brachytherapy, Guidelines on Standardized Procedures at Secondary Standards Dosimetry Laboratories (SSDLs) and Hospitals.” Vienna: IAEA, 2002. International Commission on Radiation Units and Measurements (ICRU). Report No. 62. “Prescribing, Recording and Reporting Photon Beam Therapy (Supplement to ICRU Report 50).” Bethesda, MD: ICRU, 1999. International Commission on Radiation Units and Measurements (ICRU). Report No. 33. “Radiation Quantities and Units.” Bethesda, MD: ICRU, 1980. International Commission on Radiation Units and Measurements (ICRU). Report No. 38. “Dose and Volume Specification for Reporting Intracavitary Therapy in Gynecology.” Bethesda, MD: ICRU, 1985. International Commission on Radiation Units and Measurements (ICRU). Report No. 42. “Use of Computers in External Beam Radiotherapy Procedures with High-Energy Photons and Electrons.” Bethesda, MD: ICRU, 1987. International Commission on Radiation Units and Measurements (ICRU). Report No. 50. “Prescribing, Recording and Reporting Photon Beam Therapy.” Bethesda, MD: ICRU, 1993. International Commission on Radiation Units and Measurements (ICRU). Report No. 58. “Dose and Volume Specification for Reporting Interstitial Therapy.” Bethesda, MD: ICRU, 1997. International Commission on Radiation Units and Measurements (ICRU). Report No. 24. “Determination of Absorbed Dose in a Patient Irradiated by Beams of X or Gamma Rays in Radiotherapy Procedures.” Bethesda, MD: ICRU, 1976. International Commission on Radiation Units and Measurements (ICRU). Report No. 29. “Dose Specification for Reporting External Beam Therapy with Photons and Electrons.” Bethesda, MD: ICRU, 1978. International Commission on Radiation Units and Measurements (ICRU). Report No. 35. “Radiation Dosimetry: Electron Beams with Energies Between 1 and 50 MeV.” Bethesda, MD: ICRU, 1984. International Electrotechnical Commission (IEC), Geneva, Switzerland, http://www.iecee.org/cbscheme/Standard/med.htm last accessed May 1, 2005. Interstitial Collaborative Working Group (ICWG). Interstitial Brachytherapy, Physical, Biological and Clinical Considerations. New York: Raven Press, 1990. Jain, A., Y. Zhou, T. Mustufa, E. C. Burdette, G. S. Chirikjian, and G. Fichtinger. (2005). “Matching and Reconstruction of Brachytherapy Seeds using the Hungarian Algorithm (MARSHAL).” Med Phys (in review). Jani, S. K. (1999). “Physics of vascular brachytherapy.” J Invasive Cardiol 11:517–523. Jani, S. K. (2000). “Radiation safety of personnel during catheter-based Ir-192 coronary brachytherapy. J Invasive Cardiol 12:286–290. Jani, S. K. Handbook of Dosimetry Data for Radiotherapy. Boca Raton, FL: CRC Press, 1993. Jani, S. K., G. B. Huppe, V. Massullo, P. Tripuraneni, and P. Teirstein. “Best: Manually Loaded Iridium 192 Ribbon” in Vascular Brachytherapy, 2nd Edition. R. Waksman (ed). Armonk, NY: Futura Publishing Company, Inc., pp. 485–488, 1999. Jani, S. K., J. E. Wacha, P. W. Hitchon, and J. C. VanGilder. (1987b). “Dosimetric comparison between stereotaxic and volumetric brain implants.” Med Dosimetry 12:23–26. Jani, S. K., P. W. Hitchon, J. C. Van Gilder, and B. C. Wen. (1987). “Normal brain irradiation during stereotactic brain implants using radioactive iodine-125.” Appl Neurophysiol 50:295–301. Jani, S. K., P. W. Hitchon, J. C. Van Gilder, E. C. Pennington, and D. H. Hussey. (1987a). “Choice of radioisotope in stereotactic interstitial radiotherapy of small brain tumors.” Appl Neurophysiol 50:295–301. Jani, S. K., V. Massullo, and P. Teirstein. “The 192Ir Radioactive Seed Ribbon” in Handbook of Vascular Brachytherapy. R. Waksman and P. W. Serruys. London: Martin Dunitz, Ltd., pp. 27–32, 1998.
16
Jani, S. K., V. Massullo, S. Steuterman, P. Tripuraneni, and P. S. Teirstein. (1997). “Physics and safety aspects of a coronary irradiation pilot study to inhibit restenosis using manually loaded 192Ir ribbons.” Semin Intervent Cardiol 2:119–123. Jemal, A., T. Murray, E. Ward, A. Samuels, R. C. Tiwari, A. Ghafoor, E. J. Feuer, and M. J. Thun. (2005). “Cancer statistics, 2005.” CA Cancer J Clin 55(1):10–30. Jereczek-Fossa, B. A., and R. Orecchia. (2002). “Radiotherapy-induced mandibular bone complications.” Cancer Treat Rev 28(1):65–74. Jhingran, A., and P. J. Eifel. (2000). “Perioperative and postoperative complications of intracavitary radiation for FIGO stage I-III carcinoma of the cervix.” Int J Radiat Oncol Biol Phys. 46(5):1177–1183. Johns, H. E., and J. R. Cunningham. The Physics of Radiology, 4th Edition. Springfield, IL: Charles C Thomas Publishers, 1983. Johns, H. E., and J. R. Cunningham. The Physics of Radiology. Springfield, IL: Charles C Thomas Publishers, pp. 71–99, 1983. Jones, C. H. “MicroSelectron HDR Source Localisation Techniques.” Proceedings of the 5th International SELECTRON Users’ Meeting 1988. The Hague – The Netherlands. R. F. Mould (ed.). Leersum, The Netherlands: Nucletron International B. V., 1989. Judy, P.F., S. Balter, D. Bassano, E. C. Mccullough, J. T. Payne, and L. Rothenberg. (1977). “Phantoms for Performance Evaluation and Quality Assurance of CT Scanners.” AAPM Report No. 1. Chicago: AAPM, 1977. Jürgenliemk-Schulz, I. M., L. J. Hartman, J. M. Roesink, R. J. Tersteeg, I. van Der Tweel, H. B. Kal, M. P. Mourits, and H. K. Wyrdeman. (2004). “Prevention of pterygium recurrence by postoperative single-dose beta-irradiation: A prospective randomized clinical double-blind trial.” Int J Radiat Oncol Biol Phys 59:1138–1147. Kaplan, I. D., E. J. Holupka, P. Meskell, S. J. Soon, B. Saltzman, P. Church, and G. P. Kearney. (2000). “Intraoperative treatment planning for radioactive seed implant therapy for prostate cancer.” Urology 56:492–495. Karaiskos, P., A. Angelopoulos, E. Pantelis, P. Papagiannis, L. Sakelliou, E. Kouwenhoven, and D. Baltas. (2003). “Monte Carlo dosimetry of a new 192Ir pulsed dose rate brachytherapy source.” Med Phys 30(1):9–16. Karaiskos, P., A. Angelopoulos, P. Baras, H. Rozaki-Mavrouli, P. Sandilos, L. Vlachos, and L. Sakelliou. (2000). “Dose rate calculations around 192Ir brachytherapy sources using a Sievert integration model.” Phys Med Biol 45(2):383-98. Karaiskos, P., P. Papagiannis, A. Angelopoulos, L. Sakelliou, D. Baltas, P. Sandilos, and L. Vlachos. (2001). “Dosimetry of 192Ir wires for LDR interstitial brachytherapy following the AAPM TG-43 dosimetric formalism.” Med Phys 28:156–166. Karaiskos, P., P. Papagiannis, L. Sakelliou, G. Anagnostopoulos, and D. Baltas. (2001). “Monte Carlo dosimetry of the selectSeed 125I interstitial brachytherapy seed.” Med Phys 28:1753–1760. Karnas, S. J., N. Avvakumov, E. Yu, and J. J. Battista. (2001). “Operational characteristics of a prototype x-ray needle device.” Phys Med Biol 46:97–106. Kassa, B., F. Mourtada, J. L. Horton, and R. G. Lane. (2004). “Contrast effects on dosimetry of a partial breast irradiation system.” Med Phys 31:1976–1979. Katz, A., and P. J. Eifel. (2000). “Quantification of intracavitary brachytherapy parameters and correlation with outcome in patients with carcinoma of the cervix.” Int J Radiat Oncol Biol Phys 48:1417–1425. Keisch, M., F. A. Vicini, R. R. Kuske, M. Herbert, J. White, C. Quiet, D. Arthur, T. Scroggins, and O. Streerer. (2003). “Initial clinical experience with the MammoSite breast brachytherapy applicator in women with early stage breast cancer treated with breast-conserving therapy.” Int J Radiat Oncol Biol Phys 55:289–293. Kellerer, A. M., and H. H. Rossi. (1978). “Generalized formulation of dual radiation action.” Radiat Res 75:471–488. Kestin, L. L., D. A. Jaffray, G. K. Edmundson, A. A. Martinez, J. W. Wong, V. R. Kini, P. Y. Chen, and F. A. Vicini. (2000). “Improving the dosimetric coverage of interstitial high-dose rate breast implants.” Int J Radiat Oncol Biol Phys 46:35–43.
17
Keynes, G. (1929). “The treatment of primary carcinoma of the breast with radium.” Acta Radiol 10:393–402. Keyser, G. M. (1951). “Absorption corrections for radium standardization.” Can J Phys 29:301. Khajanchee, Y., D. Streeter, L. Swanstrom, and P. Hansen. (2004). “A mathematical model for preoperative planning of radiofrequency ablation of hepatic tumors.” Surg Endosc 18:696–701. Khan, F. The Physics of Radiation Therapy. Baltimore, MD: Williams and Wilkins, pp. 380–381, 385–387, 2003. Khoo, V. S., J. L. Bedford, S. Webb, and D. P. Dearnaley. (2003). “Class solutions for conformal external beam prostate radiotherapy.” Int J Radiat Oncol Biol Phys 55(4):1109–1120. Kiefer, J. Biological Radiation Effects. New York: Springer-Verlag, 1990. King, T.A., J. S. Bolton, R. R. Kuske, G. M. Fuhrman, T. G. Scroggins, and X. Z. Jiang. (2000). “Longterm results of wide-field brachytherapy as the sole method of radiation therapy after segmental mastectomy for Tis,1,2 breast cancer.” Am J Surg 180:299–304. Kirk, M. C., W. C. Hsi, J. C. H. Chu, H. Niu, Z. Hu, D. Bernard, A. Dickler, and C. Nguyen. (2004). “Dose perturbation induced by radiographic contrast inside brachytherapy balloon applicators.” Med Phys 31:1219–1224. Kirov, A. S., and J. F. Williamson. (1997). “Two-dimensional scatter integration method for brachytherapy dose calculations in 3D geometry.” Phys Med Biol 42(11):2119–2135. Kirov, A. S., and J. F. Williamson. (2001). “Monte Carlo-aided dosimetry of the Source Tech Medical Model STM125I 125I interstitial brachytherapy source.” Med Phys 28:764–772. Kline, R. W., M. T. Gillin, D. F. Grimm, and A. Niroomand-Rad. (1985). “Computer dosimetry of 192Ir wire.” Med Phys 12:634–638. Kolkman-Deurloo, I. K., A. G. Visser, C. G. Niel, N. Driver, and P. C. Levendag. (1994). “Optimization of interstitial volume implants.” Radiother Oncol 31:229–239. Kondo, V. S., and M. L. Randolph. (1960). “Effect of finite size of ionization chambers on measurements of small photon sources.” Radiat Res 13:37–60, Kooy, H. M., R. A. Cormack, G. Mathiowitz, C. Tempany, and A. V. D’Amico. (2000). “A software system for interventional magnetic resonance image-guided prostate brachytherapy.” Comput Aided Surg 5:401–413. Kosari, K., M. Gomes, D. Hunter, D. J. Hess, E. Greeno, and T. D. Sielaff. (2002). “Local, intrahepatic, and systemic recurrence patterns after radiofrequency ablation of hepatic malignancies.” J Gastrointest Surg 6(2):255–263. Krieger, A., R. C. Susil, C. Menard, J. A. Coleman, G. Fichtinger, E. Atalar, and L. L. Whitcomb. (2005). “Design of a novel MRI compatible manipulator for image guided prostate intervention.” IEEE Trans Biomed Engr 52:306–313. Krishnan, L., W. R. Jewell, E. C. Krishnan, R. Cherian, and F. Lin. (1992). “Breast cancer with extensive intraductal component: Treatment with immediate interstitial boost irradiation.” Radiology 183:273–276. Krishnaswamy, V. (1971). “252Cf dosimetry.” Br J Radiol 527:903. Krishnaswamy, V. (1972). “Dose distribution about 137Cs sources in tissue.” Radiology 105:181–184. Krishnaswamy, V. (1978). “Dose distribution around an I-125 seed source in tissue.” Radiology 126:489491. Krishnaswamy, V. (1979). “Dose tables for 125I seed implants.” Radiology 132:727–730. Kubo, H. (1985). “Exposure contribution from Ti K x rays produces in the titanium capsule of the clinical I-125 seed.” Med Phys 12:215–220. Kubo, H. D., B. M. Coursey, W. F. Hanson, R. W. Kline, S. M. Seltzer, R. E. Shuping, and J. F. Williamson. (1998). “Report of the ad hoc committee of the AAPM Radiation Therapy Committee on 125I sealed source dosimetry.” Int J Radiat Oncol Biol Phys 40:697–702. Kubo, H. D., G. P. Glasgow, T. D. Pethel, B. R. Thomadsen, and J. F. Williamson. (1998). “High doserate brachytherapy treatment delivery: Report of the AAPM Radiation Therapy Committee Task Group No. 59.” Med Phys 25(4):375–403. Also available as AAPM Report No. 61. Kubo, H., and R. B. Chin. (1992). “Simple mathematical formulas for quick checking of single catheter high dose rate brachytherapy treatment plans.” Endocuriether/Hypertherm Oncol 8:165–169.
18
Kumar, P. P., and F. F. Bartone. (1981). “Transperineal percutaneous I-125 implant of prostate.” Urology 130:283–286. Kumar, P. P., and R. R. Good. (1986). “Vicryl carrier for I-125 seeds: Percutaneous transperineal insertion.” Radiology 159:276. Kurhanewicz, J., D. Vigneron, and S. J. Nelson. (2000). “Three-dimensional magnetic resonance spectroscopic imaging of brain and prostate cancer.” Neoplasia 2(1-2):166–189. Kuske, R. R. American Brachytherapy Society. School of breast brachytherapy. January 17–19, 2003, Las Vegas, Nevada. Kuske, R. R., J. S. Bolton, and W. Hanson. (1998). RTOG 95-17: A Phase I/II Trial To Evaluate Brachytherapy as the Sole Method of Radiation Therapy for Stage I and II Breast Carcinoma. Philadelphia: Radiation Therapy Oncology Group 1998:1–34. Kuske, R. R., J. S. Bolton, R. M. Wilenzick, W. M. P. McKinnon, B. Pullen, T. G. Scroggins, E. L. Zakris, and B. B. Fineberg. (1994). “Brachytherapy as the sole method of breast irradiation in Tis, T1, T2, N0,1 breast cancer.” ASTRO, San Francisco, CA, Oct 2–7, 1994. (Abstr.) Int J Radiat Oncol Biol Phys 30(S1):245. Kutcher, G. J., L. Coia, M. Gillin, W. F. Hanson, S. Leibel, R. J. Morton, J. R. Palta, J. A. Purdy, L. R. Reinstein, G. K. Svensson, M. Weller, and L. Wingfield. (1994). “Comprehensive QA for radiation oncology: Report of the AAPM Radiation Therapy Committee Task Group No. 40.” Med Phys 21:581–615. Also available as AAPM Report No. 46. Kuvshinoff, B. W., and D. M. Ota. (2002). “Radiofrequency ablation of liver tumors: Influence of technique and tumor size.” Surgery 132(4):605–611; discussion 611–612. Kwan, D. K., A. R. Kagan, A. J. Olch, P. Y. M. Chan, B. Hintz, and M. Wollin. (1983). “Single and double-plane iridium-192 interstitial implants: Implantation guidelines and dosimetry.” Med Phys 10:456–461. Lachaine, M. E., J. C. Gorman, and M. G. Palisca. (2003). “A fast, independent dose check of HDR plans.” J Appl Clin Med Phys 4(2):149–155. Lachance, B., D. Beliveau-Nadeau, É. Lessard, M. Chretien, I.-C. Hsu, J. Pouliot, L. Beaulieu, and É. Vigneault. (2002). “Early clinical experience with anatomy-based inverse planning dose optimization for high-dose-rate boost of the prostate.” Int J Radiat Oncol Biol Phys 54:86–100. Laeseke, P. F., L. A. Sampson, C. L. Brace, T. C. Winter 3rd, J. P. Fine, and F. T. Lee Jr. (2005). “Unintended thermal injuries from radiofrequency ablation: Protection with 5% dextrose in water.” Am J Roentgenol (in press). Lahanas, M., D. Baltas, and N. Zamboglou. (1999). “Anatomy-based three-dimensional dose optimization in brachytherapy using multiobjective genetic algorithms.” Med Phys 26(9):1904–1918. Lahanas, M., D. Baltas, and N. Zamboglou. (2003). “A hybrid evolutionary algorithm for multi-objective anatomy-based dose optimization in high-dose-rate brachytherapy.” Phys Med Biol 48:399–415. Lahanas, M., D. Baltas, and S. Giannouli. (2003a). “Global convergence analysis of fast multiobjective gradient-based dose optimization algorithms for high-dose-rate brachytherapy.” Phys Med Biol 48(5):599–617. Lam, S. T., P. S. Cho, R. J. Marks II, and S. Narayanan. (2004). “Three-dimensional seed reconstruction for prostate brachytherapy using Hough trajectories.” Phys Med Biol 49(4):557–569. Laperriere, N. J., P. M. Leung, S. McKenzie, M. Milosevic, S. Wong, J. Glen, M. Pintilie, and M. Bernstein. (2002). “Randomized study of brachytherapy in the initial management of patients with malignant astrocytoma.” Int J Oncol 21(4):817–23. Laughlin, J. S., W. M. Silver, E. I. Holodny, and F. W. Ritter. (1963). “A dose description system for interstitial radiation therapy.” Am J Roentgenol 89:470–490. Laurence, G. C. (1935). “Intensity and dosage near radium needles.” Radiology 25:166–181. Lazarescu, G., and J. J. Battista. (1997). “Analysis of the radiobiology of ytterbium-169 and iodine-125 permanent brachytherapy implants.” Phys Med Biol 42:1727–1736. Lea, D. E., and D. G. Catcheside. (1942). “The mechanism of the induction by radiation of chromosome aberrations in Tradescantia.” J Genet 44:216–245. Lee Jr., F. T., D. Haemmerich, A. S. Wright, D. M. Mahvi, L. A. Sampson, and J. G. Webster. (2003). “Multiple probe radiofrequency ablation: Pilot study in an animal model.” J Vasc Interv Radiol 14(11):1437–1442.
19
Lee, E. K., and M. Zaider. (2003). “Intraoperative dynamic dose optimization in permanent prostate implants.” Int J Radiat Oncol Biol Phys 56(3):854–861. Lee, E. K., R. J. Gallagher, D. Silvern, C. S. Wuu, and M. Zaider. (1999). “Treatment planning for brachytherapy: An integer programming model, two computational approaches and experiments with permanent prostate implant planning.” Phys Med Biol 44(1):145–165. Lee, F. T., D. Haemmerich, J. G. Webster, A. S. Wright, C. D. Johnson, and D. M. Mahvi. (2002). Radiofrequency Ablation System Using Multiple Electrodes. Wisconsin Alumni Research Foundation (WARF) Technology License WARF-P01339US. Patent pending. Lee, J. M., J. K. Han, S. H. Kim, K. S. Shin, J. Y. Lee, H. S. Park, H. Hur, and B. I. Choi. (2004). “Comparison of wet radiofrequency ablation with dry radiofrequency ablation and radiofrequency ablation using hypertonic saline preinjection: ex vivo bovine liver.” Korean J Radiol 5(4):258–265. Lee, P. C., S. J. Starr, K. Zuhlke, and B. J. Moran. (1999). “Comparisons of a proposed five-seed assay method with the single-seed and batch assay methods for I-125 seeds in ultrasound-guided prostate implants.” Radiat Oncol Invest 7:374–381. Lee, S.-W., F.-F. Yin, G. Gutman, E. Strumban, E. Sozontov, and J. H. Kim. (2005). “Monte Carlo dose calculation and experimental result comparison of a novel intensity-modulated x-ray brachytherapy.” Med Phys (In press). Lee, S.-W., F-F. Yin, G. Gutman, E. Strumban, E. Sozontov, and J. H. Kim. (2004). “Monte Carlo Dose Calculation and Experimental Result Comparison of a Novel Intensity-Modulated X-Ray Brachytherapy.” 14th International Conference on the Use of Computers in Radiation Therapy (ICCR), Seoul, Korea, pp. 252–254. http://www.iccr.info/conference_3.asp last accessed May 1, 2005. Lee, W. R., A. F. deGuzman, S. K. Tomlinson, and D. L. McCullough. (2002). “Radioactive sources embedded in suture are associated with improved postimplant dosimetry in men treated with prostate brachytherapy.” Radiother Oncol 65:123–127. Lencioni, R. (2004a). “Tumor Radiofrequency Ablation Italian Network (TRAIN): Long-term Results in Epatic Colorectal Cancer Metastases.” Radiological Society of North America Scientific Assembly and Annual Meeting, Chicago. Lencioni, R. (2004b). “Tumor Radiofrequency Ablation Italian Network (TRAIN): Long-term Survival Outcomes in Patients with Stage 0 Hepatocellular Carcinoma.” Radiological Society of North America Scientific Assembly and Annual Meeting, Chicago. Lencioni, R., L. Crocetti, R. Cioni, A. Mussi, G. Fontanini, M. Ambrogi, C. Franchini, D. Cioni, O. Fanucchi, R. Gemignani, R. Baldassarri, C. A. Angeletti, and C. Bartolozzi. (2004). “Radiofrequency ablation of lung malignancies: Where do we stand?” Cardiovasc Intervent Radiol 27(6):581–590. Lenz, M. (1927). “Radiotherapy of cancer of cervix at the radium institute, Paris, France.” Am J Roentegenol Radiat Ther 17:335–342. Leon, M. B., P. S. Teirstein, J. W. Moses, P. Tripuraneni, A. J. Lansky, S. Jani, S. C. Wong, D. Fish, S. Ellis, D. R. Holmes, D. Kerieakes, and R. E. Kuntz. (2001). “Localized intracoronary gammaradiation therapy to inhibit the recurrence of restenosis after stenting.” N Engl J Med 344:250–256. Leon, M. B., P. S. Teirstein, J. W. Moses, S. C. Wong, P. Tripuraneni, S. K. Jani, D. R. Holmes, A. Lanksy, and R. E. Kuntz. “Declining Long-Term Efficacy of Vascular Brachytherapy for In-Stent Restenosis: 5-Year Follow-Up from the Gamma 1 Randomized Trial.” American Heart Association (AHA) 2004. Abstract. Lerma, F. A., and J. F. Williamson. (2002). “Accurate localization of intracavitary brachytherapy applicators from 3D CT imaging studies.” Med Phys 28:325–333. Lessard, É. (2004). “Development and clinical introduction of an inverse planning dose optimization by simulated annealing (IPSA) for high dose rate brachytherapy.” Thesis abstract. Med Phys 31(10):2935. Lessard, É., and J. Pouliot. (2001). “Inverse planning anatomy-based dose optimization for HDRbrachytherapy of the prostate using fast simulated annealing algorithm and dedicated objective function.” Med Phys 28:773–779. Lessard, É., I.-C. Hsu, and J. Pouliot. (2002). “Inverse planning for interstitial gynecological template brachytherapy: Truly anatomy-based planning.” Int J Radiat Oncol Biol Phys 54(5): 1243–1250.
20
Levendag, P. C., R. Peters, C. A. Meeuwis, L. L. Visch, D. Sipkema, C. de Pan, and P. I. Schmitz. (1997). “A new applicator design for endocavitary brachytherapy of cancer in the nasopharynx.” Radiother Oncol 45(1):95–98. Li, A. N., N. L. Eigler, F. Litvack, and J. S. Whiting. (1998). “Characterization of a positron emitting V48 nitinol stent for intracoronary brachytherapy.” Med Phys 25:20–28. Li, Z. (2002). “Monte Carlo calculations dosimetry parameters of the Urocor Prostaseed 125I source.” Med Phys 29:1029–1034. Li, Z., I. A. Nalcacioglu, S. Ranka, S. K. Sahni, J. R. Palta, W. Tome, and S. Kim. (2001). “An algorithm for automatic, computed-tomography-based source localization after prostate implant.” Med Phys 28(7):1410–1415. Li, Z., J. Palta, and J. Fan. (2000). “Monte Carlo calculations and experimental measurements of dosimetry parameters of a new Pd-103 source.” Med Phys 27:1108–1112. Lief, E. P., B. J. Davis, and K. E. Wallner. (1998). “CT scanning for post-operation seed localization in permanent prostate implants.” Med Phys 25:A172 (Abstract). Lien, H. H., V. Blomlie, T. Iversen, C. Trope, K. Sundfor, and V. M. Abeler. (1993). “Clinical stage I carcinoma of the cervix: value of MR imaging in determining invasion into the parametrium.” Acta Radiol 34:130–132. Liljegren, G., L. Holmberg, H. O. Adami, G. Westman, S. Graffman, and J. Bergh. (1994). “Sector resection with or without postoperative radiotherapy for stage I breast cancer: Five-year results of a randomized trial. Uppsala-Orebro Breast Cancer Study Group.” J Natl Cancer Inst 86:717–722. Lin, P.-J. P., T. J. Beck, C. Borras. G. Cohen, R. A. Jucius, R. J. Kriz, E. Nickoloff, L. Rothenberg, K. J. Strauss, and T. Villafana. “Specification and Acceptance Testing of Computed Tomography Scanners.” AAPM Report No. 39. New York: American Institute of Physics, 1993. Lindsay, P., J. Battista, and J. Van Dyk (2001). “The effect of seed anisotrophy on brachytherapy dose distributions using 125I and 103Pd,” Med Phys 28(3):336–345. Ling, C. C., and Z. C. Gromadzki. (1981). “Activity uniformity of 192Ir seeds.” Int J Radiat Oncol Biol Phys 7:665–669. Ling, C. C., E. D. Yorke, I. J. Spiro, D. Kubiatowicz, and D. Bennett. (1983). “Physical dosimetry of I125 seeds of a new design for interstitial implant.” Int J Radiat Oncol Biol Phys 9:1747–1752. Ling, C. C., L. L. Anderson, and W. U. Shipley. (1979). “Dose inhomogeneity in interstitial implants using 125I seeds.” Int J Radiat Oncol Biol Phys 5:419–425. Ling, C., M. C. Schell, K. R. Working, K. Jentzch, L. Harisiadis, S. Carabell, and C. Rogers. (1987). “CT-assisted assessment of bladder and rectum dose in gynecological implants.” Int J Radiat Oncol Biol Phys 13:1577–1582. Liu, H., G. Cheng, Y. Yu, R. Brasacchio, D. Rubens, J. Strang, L. Liao, and E. Messing. (2003). “Automatic localization of implanted seeds from postimplant CT images.” Phys Med Biol 48:1191–1203. Liu, J. B., B. B. Goldberg, D. A. Merton, N. M. Rawool, L. Needleman, and F. Forsberg. (2001). “The role of contrast-enhanced sonography for radiofrequency ablation of liver tumors.” J Ultrasound Med 20(5):517–523. Liu, L., S. C. Prasad, and D. A. Bassano. (2004). “Determination of 137Cs dosimetry parameters according to the AAPM TG-43 formalism.” Med Phys 31(3):477–483. Liu, W., U. Techavipoo, T. Varghese, J. A. Zagzebski, Q. Chen, and F. T. Lee Jr. (2004). “Elastographic versus x-ray CT imaging of radio frequency ablation coagulations: An in vitro study.” Med Phys 31(6):1322–1332. Livraghi, T., L. Solbiati, M. F. Meloni, G. S. Gazelle, E. F. Halpern, and S. N. Goldberg. (2003). “Treatment of focal liver tumors with percutaneous radio-frequency ablation: Complications encountered in a multicenter study.” Radiology 226(2):441–451. Livraghi, T., S. N. Goldberg, F. Monti, A. Bizzini, S. Lazzaroni, F. Meloni, S. Pellicano, L. Solbiati, and G. S. Gazelle. (1997). “Saline-enhanced radio-frequency tissue ablation in the treatment of liver metastases.” Radiology 202(1):205–210. Loevinger, R. (1993). “Wide-angle free-air chamber for calibration of low-energy brachytherapy sources (abstract).” Med Phys 20:907.
21
Loftus, T. P. (1970). “Standardization of 137Cs gamma-ray sources in terms of exposure units (roentgens).” J Res Nat Bur Stand 74A:1–6. Loftus, T. P. (1980). “Standardization of 192Ir gamma-ray sources in terms of exposure.” J Res Nat Bur Stand 85:19–25. Loftus, T. P. (1984). “Exposure standardization of 125I seeds used for brachytherapy.” J Res Nat Bur Stand 89:295–303. Loftus, T. P. (1988). “Calibration of gamma-ray emitting brachytherapy sources.” Nat Bur Stand Spec Pub 89:2250. Lorentzen, T. (1996). “A cooled needle electrode for radiofrequency tissue ablation: Thermodynamic aspects of improved performance compared with conventional needle design.” Acad Radiol 3(7):556–563. Lu, D. S., S. S. Raman, D. J. Vodopich, M. Wang, J. Sayre, and C. Lassman. (2002). “Effect of vessel size on creation of hepatic radiofrequency lesions in pigs: Assessment of the ‘heat sink’ effect.” Am J Roentgenol 178(1):47–51. Luse, R. W., J. Blasko, and P. Grimm. (1997). “A method for implementing the American Association of Physicists in Medicine Task Group-43 dosimetry recommendations for 125I transperineal prostate seed implants.” Int J Radiat Oncol Biol Phys 37:737–741. Luxton, G., and G. Jozsef. (1999). “Radial dose distribution, dose to water and dose rate constant for monoenergetic photon point sources from 10 keV to 2 MeV: EGS4 Monte Carlo model calculation.” Med Phys 26:2531–2538. Lymperopoulou, G., E. Pantelis, P. Papagiannis, P. Rozaki-Mavrouli, L. Sakelliou, D. Baltas, and P. Karaiskos. (2004). “A monte carlo dosimetry study of vaginal 192Ir brachytherapy applications with a shielded cylindrical applicator set.” Med Phys 31(11):3080–3086.
22
Brachytherapy Physics, Second Edition Compiled References M–Z Ma, C.-M., C. W. Coffey, L. A. DeWerd, C. Liu, R. Nath, S. M. Seltzer, and J. P. Seuntjens. (2001). “AAPM protocol for 40-300 kV x-ray beam dosimetry in radiotherapy and radiobiology.” Med Phys 28:868-893. Also available as AAPM Report No. 76. Machi, J., S. Uchida, K. Sumida, W. M. Limm, S. A. Hundahl, A. J. Oishi, N. L. Furumoto, and R. H. Oishi. (2001). “Ultrasound-guided radiofrequency thermal ablation of liver tumors: Percutaneous, laparoscopic, and open surgical approaches.” J Gastrointest Surg 5(5):477–489. MacPherson, M.S., and J. J. Battista. (1995). “Dose distributions and dose rate constants for new ytterbium-169 brachytherapy seeds.” Med Phys 22:89–96. Magee, B., R. Swindell, M. Harris, and S. S. Banerjee. (1996). “Prognostic factors for breast recurrence after conservative breast surgery and radiotherapy: Results from a randomised trial.” Radiother Oncol 39:223–227. Major, T., C. Polgár, J. Fodor, A. Somogyi, and G. Németh. (2002). “Conformality and homogeneity of dose distributions in interstitial implants at idealized target volumes: A comparison between the Paris and dose-point optimized systems.” Radiother Oncol 62(1):103–111. Marinello, G., M. Valero, S. Leung, and B. Pierquin. (1985). “Comparative dosimetry between iridium wires and seed ribbons.” Int J Radiat Oncol Biol Phys 11:1733–1739. Martinez, A, R. Cox, and G. Edmundson. (1984). “A multiple-site perineal applicator (MUPIT) for treatment of prostate, anorectal, and gynecological malignancies.” Int J Radiat Oncol Biol Phys 10:297–305. Martinez, A. A., I. Pataki, D. Edmundson, E. Sebastian, D. Brabbins, and G. Gustafson. (2001). “Phase II prospective study of the use of conformal high-dose-rate brachytherapy as monotherapy for the treatment of favorable stage prostate cancer: A feasibility report.” Int J Radiat Oncol Biol Phys 49(1):61–69. Martinez, A., J. Gonzalez, J. Stromberg, G. Edmundson, M. Plunkett, G. Gustafson, D. Brown, D. Yan, F. Vicini, and D. Brabbins. (1995). “Conformal prostate brachytherapy: Initial experience of a phase I/II dose-escalating trial.” Int J Radiat Oncol Biol Phys 33(5):1019–1027. Maruyama, Y., H. W. Chin, A. B. Young, J. Beach, P. Tibbs, S. Goldstein, and W. Markesbery. (1984). “CT and MR for brain tumor implant therapy using CF-252 neutrons.” J Neurooncol 2:349–360. Masamune, K., G. Fichtinger, A. Patriciu, R. C. Susil, R. H. Taylor, L. R. Kavoussi, J. H. Anderson, I. Sakuma, T. Dohi, and D. Stoianovici. (2001). “System for robotically assisted percutaneous procedures with computed tomography guidance.” Comput Assist Surg 6:370–383. Matsumoto, K., E. Tada, N. Tamesa, S. Tomita, and T. Ohmoto. (1998). “Stereotactic brachytherapy for a cystic metastatic brain tumor in the midbrain.” Case report. J Neurosurg 88:141–144. Mayo, C. S., and K. Ulin. (2001). “A method for checking high dose rate treatment times for vaginal applicators.” J Appl Clin Med Phys 2(4):184–190. Mayr, M. T., I. R. Crocker, E. K. Butker, H. Williams, G. A. Cotsonis, J. J. Olson. (2002). “Results of interstitial brachytherapy for malignant brain tumors.” Int J Oncol 21:817–823. Mazeron, J. J., A. Gerbaulet, J. M. Simon, and C. Hardiman. (1998). “How to optimize therapeutic ratio in brachytherapy of head and neck squamous cell carcinoma?” Acta Oncol 37:583–591. Mazeron, J. J., G. Noël, and J. M. Simon. (2002). “Head and Neck Brachytherapy.” Semin Radiat Oncol 12(1):95–108. Mazeron, J. J., J. M. Crook, G. Marinello, W. Walop, and B. Pierquin. (1990). “Prognostic factors of local outcome for T1, T2 carcinomas of oral tongue treated by iridium-192 implantation.” Int J Radiat Oncol Biol Phys 19(2):281–285. McCrae, D., J. Rogers, and A. Dritschilo. (1987). “Dose-volume and complication in interstitial implants for breast carcinoma.” Int J Radiat Oncol Biol Phys 13:525–529. McGahan, J. P. (sic: McGhana, J. P.), and G. D. Dodd 3rd (2001). “Radiofrequency ablation of the liver: Current status.” Am J Roentgenol 176(1):3–16.
23
McGahan, J. P., J. M. Brock, H. Tesluk, W. Z. Gu, P. Schneider, and P. D. Browning. (1992). “Hepatic ablation with use of radio-frequency electrocautery in the animal model.” J Vasc Interv Radiol 3(2):291–297. McGahan, J. P., P. D. Browning, J. M. Brock, and H. Tesluk. (1990). “Hepatic ablation using radiofrequency electrocautery.” Invest Radiol 25(3):267–270. McGahan, J. P., W. Z. Gu, J. M. Brock, H. Tesluk, and C. D. Jones. (1996). “Hepatic ablation using bipolar radiofrequency electrocautery.” Acad Radiol 3(5):418–422. McGinley, P. Shielding Techniques for Radiation Oncology Facilities, 2nd ed. Madison, WI: Medical Physics Publishing, 2002. Medich, D. C., and J. J. Munro III. (2005). “Monte Carlo dosimetric characterization of an ytterbium-169 high dose rate brachytherapy source.” Med Phys (in review). Meertens, H., and R. van der Laarse. (1985). “Screens in ovoids of a Selectron cervix applicator.” Radiother Oncol 3(1):69–80. Meigooni, A. S., H. Zhang, C. Perry, S. A. Dini, and R. A. Koona. (2003). “Theoretical and experimental determination of dosimetric characteristics for BrachySeed Pd-103 model Pd-1 source.” Appl Radiat Isot 58:533–541. Meigooni, A. S., H. Zhang, J. R. Clark, V. Rachabatthula, and R. A. Koona. (2004). “Dosimetric characteristics of the new RadioCoil™ 103Pd wire line source for use in permanent brachytherapy implants.” Med Phys 31:3095–3105. Meigooni, A. S., S. A. Dini, K. Sowards, J. L. Hayes, and A. Al-Otoom. (2002). “Experimental determination of the TG-43 dosimetric characteristics of EchoSeed™ model 6733 125I brachytherapy source.” Med Phys 29:939–942. Meigooni, A. S., Z. Bharucha, M. Yoe-Sein, and K. Sowards. (2001). “Dosimetric characteristics of the Best® double-wall 103Pd brachytherapy source.” Med Phys 28:2568–2575. Meigooni, A., C. B. Saw, and R. Nath. “Basic Physics of Brachytherapy” in Principles and Practice of Brachytherapy. Subir Nag (ed.). Armonk, NY: Futura Publishing Company, Inc., 1997. Meisberger, L. L., R. J. Keller, and R. J. Shalek. (1968) “The effective attenuation in water of the gamma rays of gold 198, iridium 192, cesium 137, radium 226, and cobalt 60.” Radiology 90:953–957. Mellenberg, D. E., and E. C. Pennington. (1999). “103Pd loaded cartridge air kerma strength verification.” Med Dosim 24:73–75. Meloni, M. F., S. N. Goldberg, T. Livraghi, F. Calliada, P. Ricci, M. Rossi, D. Pallavicini, and R. Campani. (2001). “Hepatocellular carcinoma treated with radiofrequency ablation: Comparison of pulse inversion contrast-enhanced harmonic sonography, contrast-enhanced power Doppler sonography, and helical CT.” Am J Roentgenol 177(2):375–380. Ménard, C., R. C. Susil, P. Choyke, G. S. Gustafson, W. Kammerer, H. Ning, R. W. Miller, K. L. Ullman, N. S. Crouse, S. Smith, É. Lessard, J. Pouliot, V. Wright, E. McVeigh, C. N. Coleman, and K. Camphausen. (2004). ”MRI-guided HDR prostate brachytherapy in a standard 1.5T scanner.” Int J Radiat Oncol Biol Phys 59(5):1414–1423. Ménard, C., R. C. Susil, P. Choyke, J. Coleman, R. Grubb, A. Gharib, A. Krieger, P. Guion, D. Thomasson, K. Ullman, S. Gupta, V. Espina, L. Liotta, E. Petricoin, L. L. Whitcomb, G. Fichtinger, E. Atalar, N. C. Coleman, and K. Camphausen. (2004). “An interventional MRI technique for the molecular characterization of intra-prostatic dynamic contrast enhancement.” Molecular Imag (in press). Mendenhall, W. M. (2004). “Mandibular osteoradionecrosis.” J Clin Oncol 22(24):4867–4868. Meredith, W. J. (ed.). Radium Dosage: The Manchester System. Edinburgh: Livingston, 1967. Meredith, W. J. “Dosage for Cancer of the Cervix Uteri.” Chapter 6 in Radium Dosage: The Manchester System, ed. 2. Edinburgh: E. & S. Livingstone, pp. 42–50, 1967a. Meredith, W. J. “Physical Aspects of the Interstitial Treatment System.” Chapter 10 in Radium Dosage: The Manchester System, ed. 2. Edinburgh: E. & S. Livingstone, pp. 83–107, 1967. Merrick, G. S., K. E. Wallner, and W. M. Butler. (2003). “Permanent interstitial brachytherapy for the management of carcinoma of the prostate gland.” J Urol 169:1643–1652. Merrick, G. S., K. Wallner, W. M. Butler, R. W. Galbreath, J. H. Lief, and M. L. Benson (2001). “A comparison of radiation dose to the bulb of the penis in men with and without prostate brachytherapy-induced erectile dysfunction.” Int J Radiat Oncol Biol Phys 50(3):597–604.
24
Merrick, G. S., W. M. Butler, A. T. Dorsey, and H. L. Walbert. (1997). “Prostatic conformal brachytherapy: 125I/103Pd postoperative dosimetric analysis.” Radiat Oncol Investig 5(6):305–313. Merrick, G. S., W. M. Butler, A. T. Dorsey, J. H. Lief, and M. L. Benson. (2000). “Seed fixity in the prostate/periprostatic region following brachytherapy.” Int J Radiat Oncol Biol Phys 46(1):215–220. Messing, E. M., J. B. Zhang, D. J. Rubens, R. A. Brasacchio, J. G. Strang, A. Soni, M. C. Schell, P. G. Okunieff, and Y. Yu. (1999). “Intraoperative optimized inverse planning for prostate brachytherapy: Early experience.” Int J Radiat Oncol Biol Phys 44(4):801–808. Metropolis, N., A. Rosenbluth, N. Rosenbluth, A. Teller, and E. Teller. (1953). “Equation of state calculations by fast computing machines.” J Chem Phys 21:1087–1092. Milickovic, N., M. Lahanas, M. Papagiannopoulou, N. Zamboglou, and D. Baltas. (2002). “Multiobjective anatomy-based dose optimization for HDR-brachytherapy with constraint free deterministic algorithms.” Phys Med Biol 47:2263–2280. Millar, W. T., J. H. Hendry, and P. A. Canney. (1996). “The influence of the number of fractions and biexponential repair kinetics on biological equivalence in pulsed brachytherapy.” Brit J Radiol 69:457–468. Miller, A. V., M. G. Davis, and J. L. Horton. (1996). “A method for verifying treatment times for simple high-dose-rate endobronchial brachytherapy procedures.” Med Phys 23:1903–1908. Miller, T., and P. Grigsby. (2002). “Measurement of tumor volume by PET to evaluate prognosis in patients with advanced cervical cancer treated by radiation therapy.” Int J Radiat Oncol Biol Phys 53:353–359. Million, R. R., N. J. Cassisi, and A. A. Mancuso. “Oral Cavity” in Management of Head and Neck Cancer: A Multidisciplinary Approach. R. R. Million and N. J. Cassisi (eds.). Philadelphia: Lippincott, 1994. Minar, E., B. Pokrajac, R. Ahmadi, T. Maca, W. Seitz, A. Stumpflen, R. Potter, and H. Ehringer. (1998). “Brachytherapy for prophylaxis of restenosis after long-segment femoropopliteal angioplasty: Pilot study.” Radiology 208(1):173–179. Erratum in Radiology 208(3):834. Mittleman, R. S., S. K. Huang, W. T. de Guzman, H. Cuenoud, A. B. Wagshal, and L. A. Pires. (1995). “Use of the saline infusion electrode catheter for improved energy delivery and increased lesion size in radiofrequency catheter ablation.” Pacing Clin Electrophysiol 18(5 Pt 1):1022–1027. Mizowaki, T., G. N. Cohen, A. Y. Fung, and M. Zaider. (2002). “Towards integrating functional imaging in the treatment of prostate cancer with radiation: The registrations of the MR spectroscopy imaging to ultrasound/CT images and its implementation in treatment planning.” Int J Radiat Oncol Biol Phys 54:1558–1564. Monteiro-Grillo, I., L. Gaspar, M. Monteiro-Grillo, F. Pires, and J. M. Ribeiro da Silva. (2000). “Postoperative irradiation of primary or recurrent pterygium: Results and sequelae.” Int J Radiat Oncol Biol Phys 48:865–869. Moses, J. W., M. B. Leon, J. J. Popma, P. J. Fitzgerald, D. R. Holmes, C. O’Shaughnessy, R. P. Caputo, D. J. Kereiakes, D. O. Williams, P. S. Teirstein, J. L. Jaeger, and R. E. Kuntz. (2003). “Sirolimuseluting stents versus standard stents in patients with stenosis in a native coronary artery.” N Engl J Med 349:1315–1323. Mossman, K. L. (2003). “Restructuring nuclear regulations.” Environ Health Perspect 111:13–17. Mott, J. H., J. E. Livsey, and J. P. Logue. (2004). “Development of a simultaneous boost IMRT class solution for a hypofractionated prostate cancer protocol.” Br J Radiol 77:377–386. Mourtada, F. A., C. G. Soares, S. M. Seltzer, and S. H. Lott. (2000). “Dosimetry characterization of P-32 catheter-based vascular brachytherapy source wire.” Med Phys 27:1770–1776. Mourtada, F., T. Wareing, J. Horton, J. McGhee, D. Barnett, K. Gifford, G. Failla, and R. A. Mohan. (2004). “A deterministic dose calculation method applied to the dosimetry of shielded intracavitary brachytherapy applicators.” Med Phys 31:1807. Mundinger, F. “Rationale and Methods of Interstitial Iridium 192 or Iodine 125 Protracted Long-term Irradiation” in Stereotactic Cerebral Irradiation. Inserm Symposium No. 12. G. Szikla (ed.). Amsterdam: Elsevier North Holland, 1979. Mundinger, F., W. Birg, and C. B. Ostertag. (1978). “Treatment of small cerebral gliomas with CT-aided stereotaxic curietherapy.” Neuroradiology 16:564–567. Munro, J. J. (2004). Private communication, Implant Sciences Corporation, Wakefield, MA.
25
Munro, J. J., III. (2004). “169Ytterbium low energy gamma source for high dose rate brachytherapy.” Implant Sciences Corporation Edition. Wakefield, MA. Murphy, M. K., R. K. Piper, L. R. Greenwood, M. G. Mitch, P. J. Lamperti, S. M. Seltzer, M. J. Bales, and M. H. Philips. (2004). “Evaluation of the new cesium-131 seed for use in low-energy x-ray brachytherapy.” Med Phys 6:1529–1538. Mutic, S., J.R. Palta, E. K., Butker, I. J. Das, M. S. Huq, L. D. Loo, B. J. Salter, C. H. McCollough, and J. Van Dyk. (2003). “Quality assurance for computed-tomography simulators and the computedtomography simulation process: Report of the AAPM Radiation Therapy Committee Task Group No. 66.” Med Phys 30:2762–2792. Also available as AAPM Report No. 83. Mutic, S., P. Grigsby, D. Low, J. Dempsey, W. Harms, R. Laforest, W. Bosch, and T. Miller. (2002) “PET-guided three-dimensional treatment planning of intracavitary gynecologic implants.” Int J Radiat Oncol Biol Phys 52:1104–1110. Nag, S. (1985). “Transperineal iodine-125 implantation of the prostate under transrectal ultrasound and fluoroscopic control.” Endocuriether/Hypertherm Oncol 1:207–211. Nag, S. (2000). “Brachytherapy for prostate cancer: summary of American Brachytherapy Society recommendations.” Semin Urol Oncol 18(2):133–136. Nag, S., A. A. Abitbol, L. L. Anderson, J. C. Blasko, A. Flores, L. B. Harrison, B. S. Hilaris, A. A. Martinez, M. P. Mehta, and D. Nori. (1993). “Consensus guidelines for high dose rate remote brachytherapy in cervical endometrial, and endobronchial tumors.” Int J Radiat Oncol Biol Phys. 27(5):1241–1244. Nag, S., B. Erickson, B. Thomadsen, C. Orton, J. D. Demanes, and D. Petereit. (2000). “The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for carcinoma of the cervix.” Int J Radiat Oncol Biol Phys 48(1):201–211. Nag, S., B. Erickson, S. Parikh, N. Gupta, M. Varia, and G. Glasgow. (2000). “The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for carcinoma of the endometrium.” Int J Radiat Oncol Biol Phys 48(3):779–790. Nag, S., C. Chao, B. Erickson, J. Fowler, N. Gupta, A. Martinez, B. Thomadsen; American Brachytherapy Society. (2002). “The American Brachytherapy Society recommendations for lowdose-rate brachytherapy for carcinoma of the cervix, Int J Radiat Oncol Biol Phys 52(1):33–48. Nag, S., D. Beyer, J. Friedland, P. Grimm, and R. Nath. (1999). “The American brachytherapy society (ABS) recommendations for transperineal permanent brachytherapy of prostate cancer.” Int J Radiat Oncol Biol Phys 44:789–799. Nag, S., D. Shasha, N. Janjan, I. Petersen, and M. Zaider, (2001b). “The American Brachytherapy Society recommendations for brachytherapy of soft tissue sarcomas.” Int J Radiat Oncol Biol Phys 49(4):1033–1043. Nag, S., E. R. Cano, D. J. Demanes, A. A. Puthawala, and B. Vikram. (2001). “The American Brachytherapy Society recommendations for high-dose-rate brachytherapy for head-and-neck carcinoma.” Int J Radiat Oncol Biol Phys 50(5):1190–1198. Nag, S., H. Cardenes, S. Chang, I. Das, B. Erickson, G. Ibbott, J. Lowenstein, J. Roll, B. Thomadsen, and M. Varia. (2004). “Proposed guidelines for image-based intracavitary brachytherapy for cervical carcinoma: Report from Image-Guided Brachytherapy Working Group.” Int J Radiat Oncol Biol Phys 60:1160–1172. Nag, S., J. M. Quivey, J. D. Earle, D. Followill, J. Fontanesi, and P. T. Finger. (2003a). “The American Brachytherapy Society recommendations for brachytherapy of uveal melanomas.” Int J Radiat Oncol Biol Phys 56(2):544–555. Nag, S., J. P. Ciezki, R. Cormack, S. Doggett, J. K. DeWyngaert, G. K. Edmundson, R. G. Stock, N. N. Stone, Y. Yu, and M. J. Zelefsky. (2001). “Intraoperative planning and evaluation of permanent prostate brachytherapy: Report of the American Brachytherapy Society.” Int J Radiat Oncol Biol Phys 51:1422–1430. Nag, S., P. E. Cole, I. Crocker, S. K. Jani, K. V. Kishnasastry, V. Massullo, R. Nath, D. Nori, S. Parikh, P. Rubin, B. Speiser, P. S. Teirstein, P. Tripuraneni, R. Waksman, and J. F. Williamson. (1999a). “The American Brachytherapy Society perspective on intravascular brachytherapy.” Cardiovasc Radiat Med 1(1):8–19.
26
Nag, S., P. Lukas, D. S. Thomas, and L. B. Harrison. “Intraoperative High Dose Rate Remote Brachytherapy” in High Dose Rate Brachytherapy: A Textbook. S. Nag. Armonk NY: Futura Publishing Company, 1994. Nag, S., P. S. Fernandes, and R. Bahnson. (1998). “Transperineal image-guided permanent brachytherapy for localized cancer of the prostate. Review article.” Urologic Oncol 4:191–202. Nag, S., R. Dobelbower, G. Glasgow, G. Gustafson, N. Syed, B. Thomadsen, and J. F. Williamson. (2003b). “Inter-society standards for the performance of brachytherapy: A joint report from ABS, ACMP and ACRO.” Crit Rev Oncol Hematol 48(1):1–17. Nag, S., R. J. Ellis, S. Merrick, R. Bahnson, K. Wallner, and R. Stock. (2002). “American Brachytherapy Society recommendations for reporting morbidity after prostate brachytherapy.” Int J Radiat Oncol Biol Phys 54(2):462–470. Nag, S., R. Martinez-Monge, H. Zhang, and N. Gupta. (1998). “Simplified non-looping functional loop technique for HDR brachytherapy.” Radiother Oncol 48:339–341. Nag, S., R. R. Kuske, F. A. Vicini, D. W. Arthur, and R. D. Zwicker. (2001a). “Brachytherapy in the treatment of breast cancer.” Oncology (Huntingt) 15(2):195–205. Nag, S., S. Vivekanandam, and R. Martinez-Monge. (1997). “Pulmonary embolization of permanently implanted radioactive palladium-103 seeds for carcinoma of the prostate.” Int J Radiat Oncol Biol Phys 39:667–670. Nag, S., W. Bice, K. DeWyngaert, B. Prestidge, R. Stock, R., and Y. Yu. (2000). “The American brachytherapy society recommendations for permanent prostate brachytherapy postimplant dosimetric analysis.” Int J Radiat Oncol Biol Phys 46:221–230. Nakamatsu, K., Y. Nishimura, Y. Yagyu, R. Koike, and S. Kanamori. (2004). “A randomized trial of 30Gy/3 fractions versus 40Gy/4 fractions in postoperative strontium-90 radiation therapy (RT) for pterygia.” Int J Radiat Oncol Biol Phys 60:S551. Nakamoto, Y., A. Eisbruch E. D. Achtyes, Y. Sugawara, K. R. Reynolds, C. M. Johnston, and R. L. Wahl. (2002). “Prognostic value of positron emission tomography using F-18-fluordeoxyglucose in patients with cervical cancer undergoing radiotherapy.” Gynecol Oncol 84:289–295. Narayan, K., R. J. Hicks, T. Jobling, T. Bernshaw, and A. F. McKenzie. (2001). “A comparison of MRI and PET scanning in surgically staged loco-regionally advanced cervical cancer: Potential impact on treatment.” Int J Gynecol Cancer 11:263–271. Narayanan, S., P. S. Cho, and R. J. Marks II. (2002). “Fast cross-projectional algorithm for reconstruction of seeds in prostate brachytherapy.” Med Phys 29:1572–1579. Narayanan, S., P. S. Cho, and R. J. Marks II. (2004). “Three-dimensional seed reconstruction from an incomplete data set for prostate brachytherapy.” Phys Med Biol 49(15):3483–3494. Nath, R. “Physical Characteristics and Clinical Uses of Brachytherapy Radionuclides” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath. Madison, WI: Medical Physics Publishing, pp. 7–37, 1995. Nath, R. “Physical Properties and Clinical Uses of Brachytherapy Radionuclide” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, R. Nath (eds.). Madison, WI: Medical Physics Publishing, pp. 7–38, 1995. Nath, R., H. Amols, C. Coffey, D. Duggan, S. Jani, Z. Li, M. Schell, C. Soares, J. Whiting, P. E. Cole, I Crocker, and R. Schwartz. (1999). “Intravascular brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 60.” Med Phys 26:119–152. Also available as AAPM Report No. 66. Nath, R., L. L. Anderson, G. Luxton, K. A. Weaver, J. F. Williamson, and A. S. Meigooni. (1995). “Dosimetry of interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee, Task Group No. 43.” Med Phys 22:209–234. Also available as AAPM Report No. 51. Nath, R., L. L. Anderson, J. A. Meli, A. J. Olch, J. A. Stitt, and J. F. Williamson. (1997). “Code of practice for brachytherapy physics: Report of the AAPM Radiation Therapy Committee Task Group No. 56.” Med Phys 24(10):1557–1598. Also available as AAPM Report No. 59. Nath, R., N. Yue, and L. Liu. (1999). “On the depth of penetration of photons and electrons for intravascular brachytherapy.” Cardiovasc Radiat Med 1:72–79.
27
Nath, R., N. Yue, K. Shahnazi, and P. J. Bongiorni. (2000). “Measurement of dose-rate constant for 103Pd seeds with air-kerma strength calibration based upon a primary national standard.” Med Phys 27:655–658. National Commission on Radiation Protection and Measurements (NCRP) Commentary No. 11, Dose Limits for Individuals Who Receive Exposure from Radionuclide Therapy Patients. Bethesda, MD: NCRP, 1995. National Council on Radiation Protection and Measurements (NCRP). NCRP Report No. 5. “A Handbook of Radioactivity Measurements Procedures.” Bethesda, MD: NCRP, 1985. National Council on Radiation Protection and Measurements (NCRP). NCRP Report No. 40. “Protection Against Radiation from Brachytherapy Sources.” Bethesda, MD: NCRP, 1972. National Council on Radiation Protection and Measurements (NCRP). NCRP Report No. 91. “Recommendations on Limits for Exposure to Ionizing Radiation.” Bethesda, MD: NCRP, 1987. National Council on Radiation Protection and Measurements (NCRP). Report No. 41. “Specification of Gamma-Ray Brachytherapy Sources.” Bethesda, MD: NCRP, 1974. Neblett, D. L. “Clinical Techniques and Applications Available for Interstitial Implantation” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds). Madison, WI: Medical Physics Publishing, pp. 281–300, 1995. Neblett, D. L., A. M. N. Syed, A. A. Puthawala, R. Harrop, H. S. Frey, and S. E. Hogan. (1985). “An interstitial implant technique evaluated by contiguous volume reference.” Endocuriether/Hypertherm 1:213–221. Netherlands Commission on Radiation Dosimetry (NCS). “Quality control of sealed beta sources in brachytherapy. Recommendations on detectors, measurement procedures and quality control of beta sources.” Report 14, 2004. This report is available through http://www.ncs-dos.org. Netherlands Commission on Radiation Dosimetry (NCS). Report No. 4. “Recommendations for Dosimetry and Quality Control of Radioactive Sources Used in Brachytherapy.” Bilthoven: (Netherlands Commission on Radiation Dosimetry, 1991. Nishimura, Y., A. Nakai, T. Yoshimatsu, Y. Yagyu, K. Nakamatsu, H. Shindo, and O. Ishida. (2000). “Long-term results of fractionated strontium-90 radiation therapy for pterygia.” Int J Radiat Oncol Biol Phys 46:137–141. Novick, A. C. (2004). “Laparoscopic and partial nephrectomy.” Clin Cancer Res 10(18 Pt 2):6322S–6327S. Ohtake, T., R. Abe, I. Kimijima, T. Fukushima, A. Tsuchiya, K. Hoshi, and H. Wakasa. (1995). “Intraductal extension of primary invasive breast carcinoma treated by breast-conservative surgery. Computer graphic three-dimensional reconstruction of the mammary duct-lobular systems.” Cancer 76:32–45. Oliver, G., and C. Wright. (1969). “Dosimetry of an implantable 252Cf source.” Radiology 92:143. Organ, L. W. (1976). “Electrophysiologic principles of radiofrequency lesion making.” Appl Neurophysiol 39(2):69–76. Orton, C. (1974). “Time-dose factors in brachytherapy.” Br J Radiol 47:603–607. Pajic, B., A. Pallas, D. Aebersold, G. Gruber, and R. H. Greiner. (2004). “Prospective study on exclusive, nonsurgical strontium-/yttrium-90 irradiation of pterygia.” Strahlenther Onkol 180:510–516. Pantelis, E., D. Baltas, K. Dardoufas, P. Karaiskos, P. Papagiannis, H. Rozaki-Mavrouli, and L. Sakelliou. (2002). “On the dosimetric accuracy of a Sievert integration model in the proximity of 192Ir HDR sources.” Int J Radiat Oncol Biol Phys 53(4):1071–1084. Parker, H. M., “A dosage system for interstitial radium therapy. II. Physical aspects.” Br J Radiol 11:252–266. Parsons, J. T. “The Effect of Radiation on Normal Tissues of the Head and Neck” in Management of Head and Neck Cancer: A Multidisciplinary Approach. R. R. Million and N. J. Cassisi (eds.). Philadelphia: Lippincott, 1994b. Parsons, J. T. “Time-Dose-Volume Relations in Radiation Therapy” in Management of Head and Neck Cancer: A Multidisciplinary Approach. R. R. Million and N. J. Cassisi (eds.). Philadelphia: Lippincott, 1994a.
28
Parsons, J. T., S. P. Stringer, A. A. Mancuso, and R. R. Million. “Nasal Vestibule, Nasal Cavity, and Paranasal Sinuses” in Management of Head and Neck Cancer: A Multidisciplinary Approach. R. R. Million and N. J. Cassisi (eds.). Philadelphia: Lippincott, 1994. Pasteau, O., and P. Degrais. (1914). “The radium treatment of cancer of the prostate.” Arch Roentgenol Ray 28:396–410. Patel, R. R., H. K. Shah, R. K. Das, A. Forouzannia, H. Odau, and R. R. Kuske. “Interstitial High Dose Rate Accelerated Partial Breast Irradiation Alone in Breast Conservation Therapy: The University of Wisconsin Experience.” Proceedings of the 89th Scientific Assembly and Annual Meeting of the Radiological Society of North America, #012-RO. p. 147, 2003. Paterson, R. (1952) “Studies in optimum dosage: the Mackenzie Davidson Memorial Lecture.” Br J Radiol 25:505–516. Paterson, R., and H. M. Parker in Radium Dosage: The Manchester System. W. J. Meredith (ed.). Baltimore, MD: Williams and Wilkins Company, pp. 6–38, 1949. Paterson, R., and H. M. Parker. (1934). “A dosage system for gamma-ray therapy.” Br J Radiol 7:592–632. Paterson, R., and H. M. Parker. (1938). “A dosage system for interstitial radium therapy.” Br J Radiol 11:252–266, 313–340. Paterson, R., and H. M. Parker. (1995). “A dosage system for gamma ray therapy. 1934.” Br J Radiol 68:H60–100. Paterson, R., H. M. Parker, F. W. Spiers, M. C. Tod, S. K. Stephenson, and W. J. Meredith. Radium Dosage, The Manchester System. W. J. Meredith (ed). Baltimore, MD: The Williams and Wilkins Company, 1947. Paul, J. M. R. F. Koch, P. C. Philip, and F. R. Khan. (1986). “Uniformity of dose distribution in interstitial implants.” Endocuriether Hypertherm Oncol 2:107–118. Paul, J. M., R. F. Koch, and P. C. Philip. (1988). “Uniform analysis of dose distribution in intersti tial brachytherapy dosimetry systems.” Radiother Oncol 13:105–125. Pedley, I. D. (2002). “Transperineal interstitial permanent prostate brachytherapy for carcinoma of the prostate.” Surg Oncol 11:25–34. Pelloski, C. E., M. Palmer, G. M. Chronowski, A. Jhingran, J. Horton, and P. J. Eifel. (2005). “Comparison between computed tomography-based volumetric calculations and ICRU reference point estimates of the doses of radiation delivered to the bladder and rectum during intracavitary radiation therapy for cervical cancer.” Int J Radiat Oncol Biol Phys (Accepted for publication). Pereira, P. L., J. Trubenbach, M. Schenk, J. Subke, S. Kroeber, I. Schaefer, C. T. Remy, D. Schmidt, J. Brieger, and C. D. Claussen. (2004). “Radiofrequency ablation: in vivo comparison of four commercially available devices in pig livers.” Radiology 232(2):482–490. Perera, F., F. Chisela, J. Engel, and V. Ventkatesan. (1995). “Method of localization and implantation of the lumpectomy site for high dose rate brachytherapy after conservative surgery for T1 and T2 breast cancer.” Int J Radiat Oncol Biol Phys 31:959–965. Perez, C. A., and L. W. Brady (eds.). Principles and Practice of Radiation Oncology, 3rd Edition. Philadelphia: Lippincott-Raven, 1998. Perez, C. A., and L.W. Brady. Principles and Practices of Radiation Oncology, Second Edition. Philadelphia: J. B. Lippincott Company, 1992. Perez, C. A., R. Zwicker, and J. Williamson. “Clinical Applications of Brachytherapy. I. LDR and PDR” in Principles and Practice of Radiation Oncology. C. A. Perez, L. W. Brady, E. C. Halperin, and R. K. Schmidt-Ullrich (eds.). Philadelphia: Lippincott, Williams and Wilkins, pp. 538–603, 2003. Perez-Calatayud, J., D. Granero, and F. Ballester (2004). “Phantom size in brachytherapy source dosimetric studies.” Med Phys 31(7):2075–2081. Perez-Calatayud, J., D. Granero, F. Ballester, V. Puchades, and E. Casal. (2004). “Monte Carlo dosimetric characterization of the Cs-137 selectron/LDR source: Evaluation of applicator attenuation and superposition approximation effects.” Med Phys 31(3):493–499. Perez-Calatayud, J., F. Ballester, J. L. Lluch, M. A. Serrano-Andres, E. Casal, V. Puchades, and Y. Limami. (2001a). “Monte Carlo calculation of dose rate distributions around the Walstam CDC.Ktype 137Cs sources.” Phys Med Biol 46(7):2029–2040.
29
Perez-Calatayud, J., F. Ballester, M. A. Serrano-Andres, J. L. Lluch, V. Puchades, Y. Limami, and E. Casal. (2002). “Dosimetric characteristics of the CDC-type miniature cylindrical 137Cs brachytherapy sources.” Med Phys 29(4):538–543. Perez-Calatayud, J., F. Ballester, M. A. Serrano-Andres, V. Puchades, J.L. Lluch, Y. Limami, and E. Casal. (2001c). “Dosimetry characteristics of the Plus and 12i GammaMed PDR 192Ir sources.” Med Phys 28(12):2576–2585. Perez-Calatayud, J., F. Lliso, F. Ballester, M. A. Serrano, J. L. Lluch, Y. Limami, V. Puchades, and E. Casal. (2001b). “A Monte Carlo study of dose rate distribution around the specially asymmetric CSM3-a 137Cs source.” Phys Med Biol 46(7):N169–174. Perez-Calatayud, J., F. Lliso, V. Carmona, F. Ballester, and C. Hernandez. (1999). “Monte Carlo calculation of dose rate distributions around 0.5 and 0.6 mm in diameter 192Ir wires.” Med Phys 26(3):395–401. Pernot, M., E. Luporsi, S. Hoffstetter, D. Peiffert, P. Aletti, C. Marchal, P. Kozminski, A. Noel, and P. Bey. (1997). “Complications following definitive irradiation for cancers of the oral cavity and the oropharynx (in a series of 1134 patients).” Int J Radiat Oncol Biol Phys 37(3):577–585. Pierquin, B., A. Dutreix, C. H. Paine, D. Chassagne, G. Marinello, and D. Ash. (1978). “The Paris system in interstitial radiation therapy.” Acta Radiol Oncol 17:33–48. Pierquin, B., and G. Marinello. A Practical Manual of Brachytherapy. Madison, WI: Medical Physics Publishing, 1997. Pierquin, B., D. Chassagne, F. Baillet, and C. Paine. (1973). “Clinical observations on the time factor in interstitial radiotherapy using iridium-192.” Clin Radiol 24:506–509. Pierquin, B., J. F. Wilson, and D. Chassagne. Modern Brachytherapy. New York: Masson, 1987. Podgorsak, M. B., L. A. DeWerd, B. R. Thomadsen, and B. R. Paliwal. (1992). “Thermal and scatter effects on the radiation sensitivity of well chambers used for high dose rate Ir-192 calibrations.” Med Phys 19:1311–1314. Poggi, G., C. Gatti, M. Melazzini, G. Bernardo, M. Strada, C. Teragni, A. Delmonte, C. Tagliaferri, C. Bonezzi, M. Barbieri, A. Bernardo, and P. Fratino. (2003). “Percutaneous ultrasound-guided radiofrequency thermal ablation of malignant osteolyses.” Anticancer Res 23(6D):4977–4983. Polgar, C., Z. Sulyok, J. Fodor, Z. Orosz, T. Major, Z. Takacsi-Nagy, L. C. Mangel, A. Somogyi, M. Kasler, and G. Nemeth. (2002). “Sole brachytherapy of the tumor bed after conservative surgery for T1 breast cancer: Five-year results of a phase I-II study and initial findings of a randomised phase III trial.” J Surg Oncol 80:121–128. Pop, M., A. Molckovsky, L. Chin, M. C. Kolios, M. A. Jewett, and M. D. Sherar. (2003). “Changes in dielectric properties at 460 kHz of kidney and fat during heating: Importance for radio-frequency thermal therapy.” Phys Med Biol 48(15):2509–2525. Popma, J. J., M. Suntharalingam, A. J. Lansky, R. R. Heuser, B. Speiser, P. S. Teirstein, V. Massullo, T. Bass, R. Henderson, S. Silber, P. von Rottkay, R. Bonan, K. K. Ho, A. Osattin, and R. E. Kuntz. (2002). “Randomized trial of 90Sr/90Y ß-radiation versus placebo control for treatment of in-stent restenosis.” Circulation 106:1090–1096. Popowski, Y., V. Verin, I. Papirov, P. Nouet, M. Rouzaud, M. Schwager, P Urban, W. Rutishauser, and J. M. Kurtz. (1995). “Intra-arterial 90Y brachytherapy: Preliminary dosimetric study using a specially modified angioplasty balloon.” Int J Radiat Oncol Biol Phys 33:713–717. Potish, R. A., and B. J. Gerbi. (1987). “Cervical cancer: Intracavitary dose specification and prescription.” Radiology 165:555–560. Pötter R, Kovacs G, Lenzen B, et al. (1991). “Technique of MRI assisted brachytherapy treatment planning.” Brachytherapy Journal 5:145–148. Potter, R., E. Van Limbergen, W. Dries, Y. Popowski, V. Coen, C. Fellner, D. Georg, C. Kirisits, P. Levendag, H. Marijnissen, H. Marsiglia, J. J. Mazeron, B. Pokrajac, P. Scalliet, and V. Tamburini. (2001). “Prescribing, recording, and reporting in endovascular brachytherapy. Quality assurance, equipment, personnel and education.” Radiother Oncol 59(3):339–360. Potters, L., E. Calugaru, K. B. Thornton, T. Jackson, and D. Huang. (2003). “Toward a dynamic real-time intraoperative permanent prostate brachytherapy methodology.” Brachytherapy 2:172–180. Pouliot, J., D. Tremblay, J. Roy, and S. Filice. (1996). “Optimization of permanent 125I prostate implants using fast simulated annealing.” Int J Radiat Oncol Biol Phys 36(3):711–720.
30
Pouliot, J., É. Lessard, and I.-C. Hsu. “Number of catheters in prostate high dose rate brachytherapy: The role of inverse planning.” ESTRO Joint Brachytherapy Meeting GEC/ESTRO-ABS-GLAC, Barcelona, Spain, May 13–15, 2004. Pouliot, J., R. Taschereau, C. Coté, J. Roy, and D. Tremblay. (1999). “Dosimetric aspects of permanents radioactive implants for the treatment of prostate cancer.” Physics in Canada 55(2):61–68. Pouliot, J., Y. Kim, É. Lessard, I.-C. Hsu, D. B. Vigneron, and J. Kurhanewicz. (2004). “Inverse planning for HDR prostate brachytherapy used to boost dominant intraprostatic lesions defined by magnetic resonance spectroscopy imaging.” Int J Radiat Oncol Biol Phys 59(4):1196–1207. Poupon, L., B. Castelain, E. Lartigau. (2003). “Pulse dose rate brachytherapy: Optimization and place of imaging.” Cancer Radiather 7:136–146 (French). Prados, M. D., P. H. Gutin, T. L. Phillips, W. M. Wara, P. K. Sneed, and D. A. Larson. (1992). “Interstitial brachytherapy for newly diagnosed patients with malignant gliomas: The UCSF experience.” Int J Radiat Oncol Biol Phys 24:593–597. Press, W. H., S. A. Teukolsky, W. T. Vettering, and B. P. Flannery. Numerical Recipes in C, Second Edition. New York: Cambridge University Press, 1988. Prestidge, B. P. (2004). “Early Experience with Cone-Beam CT for Brachytherapy.” Materials of 8th Annual International Conference and Workshop “New And Future Developments In Radiotherapy,” sponsored by Wayne State University, School of Medicine, Detroit, MI. November 12–14, 2004, San Diego, CA. Price, R. R., L. Axel, T. Morgan, R. Newman, W. Perman, N. Schneiders, M. Selikson, M. Wood, and S. R. Thomas. (1990). “Quality assurance methods and phantoms for magnetic resonance imaging.” Med Phys 17(2):287–295. Also available as AAPM Report No. 28. Proxima Therapeutics, Alpharetta, GA, www: http://www.proximatherapeutics.com, 2004. Puthawala, A. A., A.M.N. Syed, D. L. Eads, D. Neblett, L. Gillin, and T. C. Gates. (1985). “Limited external irradiation and interstitial 192-iridium implant in the treatment of squamous cell carcinoma of the tonsillar region.” Int J Radiat Oncol Biol Phys 11(9):1595–1602. Quesson, B., J. A. de Zwart, and C. T. Moonen. (2000). “Magnetic resonance temperature imaging for guidance of thermotherapy.” J Magn Reson Imaging 12(4):525–533. Quimby, E. H. (1932). “The grouping of radium tubes in packs and plaques to produce the desired distribution of radiation.” Am J Roentgenol Radium Ther 27:18–39. Quimby, E. H. (1935). “Physical factors in interstitial radium therapy.” Am J Roentgenol Radiat Ther 33:306–316. Quimby, E. H. (1941). “The specification of dosage in radium therapy.” Am J Roentgenol Radiat Ther 45:1–18. Quimby, E. H. (1944). “Dosage table for linear radium sources.” Radiology 43:572. Quimby, E. H., and V. Castro. (1953). “The calculation of dosage in interstitial radium therapy.” Am J Roentgenol 70:739–749. Raben, A., H. Chen, A. Grebler, J. Geltzeiler, M. Geltzeiler, I. Keselman, S. Litvin, S. Sim, A. Hanlon, and J. Yang. (2004). “Prostate seed implantation using 3D-computer assisted intraoperative planning vs. a standard look-up nomogram: Improved target conformality with reduction in urethral and rectal wall dose.” Int J Radiat Oncol Biol Phys 60:1631–1638. Radiological Physics Center (RPC). “Joint AAPM/RPC Registry of Low-energy Brachytherapy Seeds Meeting the AAPM dosimetric Prerequisites.” http://rpc.mdanderson.org/rpc/htm/Home_htm/Lowenergy.htm last accessed May 1, 2005. Raizner, A. E., S. N. Oesterle, R. Waksman, P. W. Serruys, A. Colombo, Y. L. Lim, A. C. Yeung, W. J. van der Giessen, L. Vandertie, J. K. Chiu, L. R. White, P. J. Fitzgerald, G. L. Kaluza, and N. M. Ali. (2000). “Inhibition of restenosis with beta-emitting radiotherapy: Report of the Proliferation Reduction With Vascular Energy Trial (PREVENT).” Circulation 102:951–958. Raizner, R. A., and R. V. Calfee. “The Guidant Intravascular Brachytherapy System” in Handbook of Vascular Brachytherapy. R. Waksman and P. W. Serruys. London: Martin Dunitz, Ltd., pp. 53–58, 1998. Redpath, A. T. (2002). “Automatic determination of needle and source positions for brachytherapy of the prostate using 125Iodine rapid strand.” Radiother Oncol 64:215–227. Regaud, C. (1924). “Some biological aspects of radiation therapy of cancer.” Am J Roentgenol 10:97.
31
Regaud, C. (1930). “Sur les principes radiophysiologiques de la radiothérapie des cancers.” Acta Radiol 11:455. Ribeiro, G. G., G. Dunn, R. Swindell, M. Harris, and S. S. Banerjee. (1990). “Conservation of the breast using two different radiotherapy techniques: Interim report of a clinical trial.” Clin Oncol 2:27–34. Ritz, V. H. (1960). “Standard free-air chamber for the measurement of low-energy x-rays (20 to 100k – constant potential).” J Res Nat Bur Stand 64:49. Rivard M. J., B. L. Kirk, and L. C. Leal. (2005). “Impact of radionuclide physical distribution on brachytherapy dosimetry parameters.” Nucl Sci Engr 149:101–106. Rivard, M. J, S. D. Davis, L. A. DeWerd, T. W. Rusch, and S. Axelrod. (2005). “Calculated and measured brachytherapy dosimetry parameters for the Xoft AXXENT X-Ray source: An electronic brachytherapy source.” Med Phys (submitted). Rivard, M. J. (2000). “Neutron dosimetry for a general 252Cf brachytherapy source.” Med Phys 27:2803–2815. Rivard, M. J. (2001). “Monte Carlo calculations of AAPM Task Group Report No. 43 dosimetry parameters for the MED3631-A/M 125I source.” Med Phys 28:629–637. Rivard, M. J. (2002). “Comprehensive Monte Carlo calculations of AAPM Task Group Report No. 43 dosimetry parameters for the Model 3500 I-Plant 125I brachytherapy source.” Appl Radiat Isotop 57:381–389. Rivard, M. J., B. M. Coursey, L. A. DeWerd, W. F. Hanson, M. S. Huq, G. S. Ibbott, M. G. Mitch, R. Nath, and J. F. Williamson. (2004). “Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.” Med Phys 31:633–674. Also available as AAPM Report No. 84. Rivard, M. J., C. S. Melhus, and B. L. Kirk. (2004). “Brachytherapy dosimetry parameters calculated for a new 103Pd source.” Med Phys 31:2466–2470. Rivard, M. J., D-A. R. Evans, and I. A. Kay. (2005). “A technical evaluation of the Nucletron FIRST® system: Conformance of a remote afterloading brachytherapy seed implantation system to manufacturer specifications and AAPM Task Group report recommendations.” J Appl Clin Med Phys 6:22–50. Rivard, M. J., J. Sganga, G. A. Cardarelli, B. A. Bornstein, M. J. Engler, J. Tsai, S. A. Kaufman, T. A. DiPetrillo, R. Graham, and D. E. Wazer. (2003). “A quantitative methodology for measuring conformance index of MammoSite catheter treatments.” Brachytherapy 2:43. Rivard, M. J., W. M. Butler, L. A. DeWerd, M. S. Huq, G. S. Ibbott, Z. Li, M. G. Mitch, R. Nath, and J. F. Williamson. (2004a). “Erratum: ‘Update of AAPM Task Group No. 43 Report: A revised AAPM protocol for brachytherapy dose calculations.’” [Med Phys 31:633–674 (2004)].” Med Phys 31:3532–3533. Rodriguez, R. R., J. D. Demanes, and G. A. Altieri. (1999). “High dose rate brachytherapy in the treatment of prostate cancer.” Brachytherapy 13(3):503–523. Roesch, W. C. (1958). “Dose for nonelectronic equilibrium conditions.” Radiat Res 9:399–410. Rogus, R. D., M. J. Smith, and H. D. Kubo. (1998). “An equation to QA check the total treatment time for single-catheter HDR brachytherapy.” Int J Radiat Oncol Biol Phys 40(1):245–248. Rose, P., L. Adler, M. Rodriquez, P. F. Faulhaber, F W. Abdul-Karim, and F. Miraldi. (1999). “Positron emission tomography for evaluating para-aortic nodal metastasis in locally advanced cervical cancer before surgical staging: A surgicopathologic study.” J Clin Oncol 17(1):41–45. Rosen I. I., K. M. Khan, R. G. Lane, and C. A. Kelsey. (1982). “The effect of geometric errors in the reconstruction of Iridium-92 seed implants.” Med Phys 9:220. Rosen, P. P., A. A. Fracchia, J. A. Urban, D. Schottenfeld, and G. F. Robbins. (1975). “‘Residual’ mammory carcinoma following simulated partial mastectomy.” Cancer 35:739–747. Rosenthal, M. S., and R. Nath. (1983). “An automatic seed identification technique for interstitial implants using three isocentric radiographs.” Med Phys 10(4):475–479. Rossi, H. H., and M. Zaider. Microdosimetry and Its Applications. Springer-Verlag Telos, 1996. Rossi, S., F. Garbagnati, I. De Francesco, F. Accocella, L. Leonardi, P. Quaretti, A. Zangrandi, C. Paties, and R. Lencioni. (1999). “Relationship between the shape and size of radiofrequency induced thermal lesions and hepatic vascularization.” Tumori 85(2):128–132.
32
Roston, A. Y., and M. E. El-Sayed. (1987). “Fat necrosis of the breast: An unusual complication of lumpectomy and radiotherapy in breast cancer.” Clin Radiol 38:31. Roy, J. N., K. E. Wallner, P. J. Harrington, C. C. Ling, and L. L. Anderson. (1993). “A CT-based evaluation method for permanent implants: application to prostate.” Int J Radiat Oncol Biol Phys 26(1):163–169. Roy, J. N., K. E. Wallner, S. T. Chiu-Tsao, L. L. Anderson, and C. C. Ling. (1991). “CT-based optimized planning for transperineal prostate implant with customized template.” Int J Radiat Oncol Biol Phys 21:483–489. Rozenfeld, M., and D. Jette. (1984). “Quality assurance of radiation dosage: Usefulness of redundancy.” Radiology 150:241–244. Rusch, T. W., and M. J. Rivard. (2004). “Application of the TG-43 dosimetry protocol to electronic brachytherapy sources.” (Abstract). Radiother Oncol 71(S2):S84. Rusch, T. W., S. D. Davis, L. A. DeWerd, R. Burnside, S. Axelrod, and M. J. Rivard. (2004). “Characterization of a new miniature x-ray source for electronic brachytherapy.” (Abstract). Med Phys 31:1807. Russell, K. J., and J. C. Blasko. (1993). “Recent advances in interstitial brachytherapy for localized prostate cancer.” Probl Urol 7:260–278. Sarrazin, D., M. G. Le, R. Arriagada, G. Contessa, F. Fontaine, M. Spielmann, F. Rochard, T. Le Chevalier, and J. Lacour. (1989). “Ten-year results of a randomized trial comparing a conservative treatment to mastectomy in early breast cancer.” Radiother Oncol 14:177–184. Saw, C .B., L. J. Korb, B. Darnell, K. V. Krishna, and D. Ulewicz. (1998). “Independent technique of verifying high-dose rate (HDR) brachytherapy treatment plans.” Int J Radiat Oncol Biol Phys 40(3):747–750. Saw, C. B., and N. Suntharalingam. (1988). “Reference dose rates for single- and double-plane Ir-192 implants.” Med Phys 15:391–396. Saw, C. B., and N. Suntharalingam. (1991). “Quantitative assessment of interstitial implants.” Int J Radiat Oncol Biol Phys 20:135–139. Saw, C. B., L. J. Korb, B. Darnell, K. V. Krishna, and D. Ulewicz. (1998). “Independent technique of verifying high-dose rate (HDR) brachytherapy treatment plans.” Int J Radiat Oncol Biol Phys 40:747–750. Saw, C. B., N. Suntharalingam, and A. Wu. (1993). “Concept of dose nonuniformity in interstitial brachytherapy.” Int J Radiat Oncol Biol Phys 26(3):519–527. Scheidler, J., A. F. Heuck, M. Steinborn, R. Kimmig, and M. F. Reiser. (1998). “Parametrial invasion in cervical carcinoma: Evaluation of detection at MR imaging with fat suppression.” Radiology 206:125–129. Scheidler, J., H. Hricak, D. B. Vigneron, K. K. Yu, D. L. Sokolov, L. R. Huang, C. J. Zaloudek, S. J. Nelson, P. R. Carroll, and J. Kurhanewicz. “Prostate cancer: Localization with three-dimensional proton MR spectroscopic imaging—clinicopathologic study.” Radiology 213(2):473–480. Schmidt-Ullrich, R., R. D. Zwicker, A. Wu, and K. Kelly. (1991). “Interstitial Ir-192 implants of the oral cavity: The planning and construction of volume implants.” Int J Radiat Oncol Biol Phys 20:1079–1085. Schnitt, S., J. Hayman, R. Gelman, T. J. Eberlein, S. M. Love, K. Mayzel, R. T. Osteen, A. J. Nixon, S. Pierce, J. L. Connolly, P. Cohen, L. Schneider, B. Silver, A. Recht, and J. R. Harris. (1996). “A prospective study of conservative surgery alone in the treatment of selected patients with stage I breast cancer.” Cancer 77:1094–1100. Schoeppel, S. L., B. A. Fraass, M. P. Hopkins, M. L. La Vigne, A. S. Lichter, D. L. McShan, S. Noffsinger, C. Perez-Tamayo, and J. A. Roberts. (1989). “A CT-compatible version of the Fletcher system intracavitary applicator: clinical application and 3-dimensional treatment planning.” Int J Radiat Oncol Biol Phys 17:1103–1109. Schoeppel, S. L., J. H. Ellis, M. L. LaVigne, R. A. Schea, and J. A. Roberts. (1992). “Magnetic resonance imaging during intracavitary gynecologic brachytherapy.” Int J Radiat Oncol Biol Phys 23(1):169–174.
33
Schoeppel, S. L., M. L. LaVigne, M. K. Martel, D. L. McShan, B. A. Fraass, and J. A. Roberts. (1994). “Three-dimensional treatment planning of intracavitary gynecologic implants: analysis of ten cases and implications for dose specification.” Int J Radiat Oncol Biol Phys 28:277–283. Schulz, I. M., L. J. Hartman, J. M. Roesink, R. J. Tersteeg, I. van Der Tweel, H. B. Kal, M. P. Mourits, and H. K. Wyrdeman. (2003). “Prevention of pterygium recurrence by postoperative single dose beta-irradiation: A prospective double-blind trial.” Int J Radiat Oncol Biol Phys 57:S252. Schwartz, D. J., B. J. Davis, R. J. Vetter, T. M. Pizansky, M. G. Herman, T. M. Wilson, W. N. LaJoie, and A. L. Oberg. (2003). “Radiation exposure to operating room personnel during transperineal interstitial permanent prostate brachytherapy.” Brachytherapy 2:98–102. Seltzer, S. M., and P. J. Lamperti. (1999). Status of NIST Primary Standards for 125I and 103Pd Therapy Seeds Based on the Wide-Angle Free-Air Chamber (WAFAC): Proceedings of a CIRMS Workshop, National Institute of Standards and Technology. Seltzer, S. M., P. J. Lamperti, R. Loevinger, C. G. Soares, and J. T. Weaver. (1999). “New NIST airkerma strength standards for 125I and 103Pd brachytherapy seeds (abstr).” Med Phys 25:170. Seltzer, S. M., P. J. Lamperti, R. Loevinger, M. G. Mitch, J. T. Weaver, and B. M. Coursey. (2003). “New national air-kerma-strength standards for 125I and 103Pd brachytherapy seeds.” J Res Nat Bur Stand 108:337–358. Sethi, T., D. Ash, A. Flynn, and G. Workman. (1996). “Replacement of hairpin and loop implants by optimized straight line sources.” Radiother Oncol 39:117–121. Shah, N. M., T. Tenenholz, D. Arthur, T. DiPetrillo, B. Bornstein, G. Cardarelli, M. J. Zheng, M. J. Rivard, S. Kaufman, and D. E. Wazer. (2004). “MammoSite and interstitial brachytherapy for accelerated partial breast irradiation: Factors that affect toxicity and cosmesis.” Cancer 101:727–734. Shalek, R. J., and M. Stovall. “Dosimetry in Implant Therapy” in Radiation Dosimetry, vol III. F. H. Attix and E. Tochlin (eds.). New York: Academic Press, pp. 776–798, 1969. Shalek, R. J., and M. Stovall. Brachytherapy Dosimetry, The Dosimetry of Ionizing Radiation. Vol. III. K. R. Kase, B. E. Bjarngard, and F. H. Attix (eds). San Diego, CA: Academic Press, Inc., 1990. Shalek, R. J., and M. Stovall. Implant Dosimetry. The University of Texas M.D. Anderson Hospital and Tumor Institute at Houston, 1967. Shank, B. (1986. “Breast preservation and brachytherapy.” Endocuriether/Hypertherm Oncol 2:17–24. Sharma, A., A. M. N. Syed, A. Puthawala, and A. A. Farid. (1998). “HDR planning for prostate cancer implants using Varisource remote afterloader.” J Brachytherapy International 14:1–14. Sharma, S. H., J. F. Williamson, and E. Cytacki. (1982). “Dosimetric analysis of stereo and orthogonal reconstruction of interstitial implants.” Int J Radiat Oncol Biol Phys 8:1803. Sheline, G. E., W. M. Wara, and V. Smith. (1980). “Therapeutic irradiation and brain injury.” Int J Radiat Oncol Biol Phys 6:1215–1228. Shen, D., Z. Lao, J. Zeng, W. Zhang, I. Sesterhenn, L. Sun, J. W. Moul, E. H. Herskovits, G. Fichtinger, and C. Davatzikos. (2004). “Optimization of biopsy strategy by a statistical atlas of prostate cancer distribution.” Med Image Anal 8:139–150. Shewchuk, J. R. (1994). “An introduction to the conjugate gradient method without the agonizing pain.” http://www-2.cs.cmu.edu/~jrs/jrspapers.html. Shimizutani, K., Y. Koseki, T. Inoue, T. Teshima, S. Furukawa, K. Kubo, H. Fuchihata, N. Masaki, H. Ikeda, and Y. Tanaka. (1995). “Application of 198Au grains for carcinoma of oral cavity.” Strahlenther Onkol 171(1):29–34. Shock, S. A., K. Meredith, T. F. Warner, L. A. Sampson, A. S. Wright, T. C. Winter 3rd, D. M. Mahvi, J. P. Fine, and F. T. Lee Jr. (2004). “Microwave ablation with loop antenna: in vivo porcine liver model.” Radiology 231(1):143–149. Sievert, R. M. (1921). “Die Intensitätsverteilung der primären γ-Strahlung in der Nähe medizinischer Radiumpräparate.” Acta Radiol 1:89–128. Simon, J. M., J. J. Mazeron, S. Pohar, C. Le Pechoux, J. M. Crook, L. Grimard, P. Piedbois, J. P. Le Bourgeois, and B. Pierquin. (1993). “Effect of intersource spacing on local control and complications in brachytherapy of mobile tongue and floor of mouth.” Radiother Oncol 26(1):19–25.
34
Sironi, S., F. De Cobelli, G. Scarfone, E. Colombo, G. Bolis, A. Ferrari, and A. DelMaschio. (1993). “Carcinoma of the cervix: Value of plain and gadolinium-enhanced MR imaging in assessing degree of invasiveness.” Radiology 188:797–801. Slanina, J., M. Wannenmacher, K. Kuphal, H. Knufermann, C. Beck, and W. Schilli. (1982). “Interstitial radiotherapy with 198Au seeds in the primary management of carcinoma of the oral tongue: Results in Freiburg/Breisgau from January 1964 to July 1980.” Int J Radiat Oncol Biol Phys 8:1683–1689. Sliski, A. P., M. T. Dinsmore, A. J. Boom, and N. T. Zervas. (1994). Low Power X-Ray Source with Implantable Probe for Treatment of Brain Tumors. U.S. Patent 5,369,679. Sloboda, R. S. (1992). “Optimization of brachytherapy dose distribution by simulated annealing.” Med Phys 19:955–964. Sloboda, R., J. E. Pedersen, and R. M. Halperin. (2003). “Is there a preferred strength for regularly spaced 125 I seeds in inverse-planned prostate implants?” Int J Radiat Oncol Biol Phys 55:234–244. Sloboda, R., J. E. Pedersen, J. Hanson, and R. M. Halperin. (2003). “Dosimetric consequences of increased seed strength for I-125 prostate implants.” Radiother Oncol 68:295–297. Sminia, P., C. J. Schneider, K. Koedooer, G. van Tienhoven, L. E. Blank, and D. G. Gonzalez. (1998). “Pulse frequency in pulsed brachytherapy based on tissue repair kinetics.” Int J Radiat Oncol Biol Phys 41:139–150. Smith, A., P. Almond, and L. Delclos. (1974). “Evaluation of 252Cf neutron emitter for interstitial and intracavitary radiation therapy.” Eur J Cancer 10:369. Smith, V., A. S. Meigooni, and S.-T. Chiu-Tsao. “Orthogonal Techniques,” “Stereo-Shift Technique and Automatic Reconstruction Techniques,” and “Specialized Techniques” in Interstitial Brachytherapy: Physical, Biological, and Clinical Considerations. Interstitial Collaborative Working Group. New York: Raven Press, Ltd., pp. 239–252, 1990. Sneed, P. K., K. R. Lamborn, D. A. Larson, and M. D. Prados. (1996). “Demonstration of brachytherapy boost dose-response relationships in glioblastoma multiforme.” Int J Radiat Oncol Biol Phys 35:37–44. Snyder, K. M., R. G. Stock, S. M. Hong, Y. C. Lo, and N. N. Stone. (2001). “Defining the risk of developing grade 2 proctitis following 125I prostate brachytherapy using a rectal dose-volume histogram analysis.” Int J Radiat Oncol Biol Phys 50(2):335–341. Soares, C. G. (1991). “Calibration of ophthalmic applicators at NIST: A revised approach.” Med Phys 18:787–793. Soares, C. G. (1995). “Comparison of NIST and manufacturer calibrations of 90Sr+90Y ophthalmic applicators.” Med Phys 22:1487. Soares, C. G. (2004). “Calibration of Ophthalmic Applicators.” IRD-P-09 report. http://physics.nist.gov/Divisions/Div846/QualMan/Procdures/WebProcedure09v200.pdf. Soares, C. G., and A. P. Sliski. (2002). “Absorbed dose measurements of a handheld 50-kV x-ray source with ionization chambers and plastic scintillators.” (Abstract). Med Phys 29:1351. Soares, C. G., D. G. Halpern, and C.-K. Wang. (1998). “Calibration and characterization of beta-particle sources for intravascular brachytherapy.” Med Phys 25:339–346. Soares, C. G., S. Vynckier, H. Järvinen, W. G. Cross, P. Sipilä, D. Flühs, B. Schaeken, F. A. Mourtada, G. A. Bass, and T. T. Williams. (2001). “Dosimetry of beta-ray ophthalmic applicators: Comparison of different measurement methods.” Med Phys 28:1373-1384. Soares, C., C. Drupieski, B. Wingert, G. Pritchett, V. Pagonis, M. O’Brien, A. Sliski, P. Bilski, and P. Olko. (2005). “Absorbed dose measurements of a handheld 50 kVp X-ray source in water with thermoluminescence dosimeters.” Rad Prot Dosim 114:(In press). Solbiati, L., T. Ierace, M. Tonolini, and L. Cova. (2004a). “Guidance and monitoring of radiofrequency liver tumor ablation with contrast-enhanced ultrasound.” Eur J Radiol 51:S19–S23. Solbiati, L., T. Ierace, S. N. Goldberg, S. Sironi, T. Livraghi, R. Fiocca, G. Servadio, G. Rizzatto, P. R. Mueller, A. Del Maschio, and G. S. Gazelle. (1997). “Percutaneous US-guided radio-frequency tissue ablation of liver metastases: treatment and follow-up in 16 patients.” Radiology 202(1):195–203. Solbiati, L., T. Livraghi, T. Ierace, F. Meloni, L. Cova, and S. Goldberg. (2004b). “Radiofrequency Ablation for Liver Colorectal Metastases: Is It Possible to Equal the 5-year Survival Rates of
35
Surgery?” Radiological Society of North America Scientific Assembly and Annual Meeting, Chicago. Speight, J., K. Shinohara, B. Pickett, V. Weinberg, I.-C. Hsu, and M. Roach III. (2000). “Prostate volume change after radioactive seed implantation: Possible benefit of improved dose volume histogram with perioperative steroid.” Int J Radiat Oncol Biol Phys 48(5):1461–1467. Sralek, R. J., and M. Stovall. “Brachytherapy Dosimetry” in The Dosimetry of Ionizing Radiation. K. R. Kase, B. E. Bjarngard and F. H. Attix (eds.). New York: Academic Press, pp. 259–321, 1990. Steel, G. Gordon (ed.). Basic Clinical Radiobiology. Oxford: Oxford University Press, 1997. Steel, G. Gordon. Growth Kinetics of Tumors. Gloucestershire, UK: Clarendon Press, 1977. Steggerda, M. J., and B. J. Mijnheer. (1994). “Replacement corrections of a Farmer-type ionization chamber for the calibration of Cs-137 and Ir-192 sources in a solid phantom.” Radiother Oncol 31:76–84. Steinfeld, A. D., B. R. Donahue, and L. Plaine. (1991). “Pulmonary embolization of iodine-125 seeds following prostate implantation.” Urology 37:149–150. Steinke, K., J. King, D. Glenn, and D. Morris. (2003). “Percutaneous radiofrequency ablation of lung tumors: difficulty withdrawing the hooks resulting in a split needle.” Cardiovasc Intervent Radiol 26:583–585. Steinke, K., P. E. Sewell, D. Dupuy, R. Lencioni, T. Helmberger, S. T. Kee, A. L. Jacob, D. W. Glenn, J. King, and D. L. Morris. (2004). “Pulmonary radiofrequency ablation—an international study survey.” Anticancer Res 24(1):339–343. Stitt, J. A., J. F. Fowler, B. R. Thomadsen, D. A. Buchler, B. P. Paliwal, and T. J. Kinsella. (1992). “High dose rate intracavitary brachytherapy for carcinoma of the cervix: The Madison system: I. Clinical and radiobiological considerations.” Int J Radiat Oncol Biol Phys. 24(2):335–348. Stock, R. G., N. N. Stone, M. Dahlal, and Y. C. Lo. (2002). “What is the optimal dose for 125I prostate implants? A dose-response analysis of biochemical control, posttreatment prostate biopsies, and long-term urinary symptoms.” Brachytherapy 1(2):83–89. Stock, R. G., N. N. Stone, M. F. Wesson, and J. K. DeWyngaert. (1995). “A modified technique allowing interactive ultrasound-guided three-dimensional transperineal prostate implantation.” Int J Radiat Oncol Biol Phys 32(1):219–225. Stone, G. W., S. G. Ellis, D. A. Cox, J. Hermiller, C. O’Shaughnessy, J. T. Mann, M. Turco, R. Caputo, P. Bergin, J. Greenberg, J. J. Popma, M. E. Russell, the TAXUS-IV Investigators. (2004). “A polymer-based, paclitaxel-eluting stent in patients with coronary artery disease.” N Engl J Med 350:221–231. Stone, N. N., R. G. Stock, J. K. DeWyngaert, and A. Tabert. (1995). “Prostate brachytherapy: Improvements in prostate volume measurements and dose distribution using interactive ultrasound guided implantation and three-dimensional dosimetry.” Radiat Oncol Invest 3:185–195. Stone, N. N., S. Hong, Y.-C. Lo, V. Howard, and R. G. Stock. (2003). “Comparison of intraoperative dosimetric implant representation with postimplant dosimetry in patients receiving prostate brachytherapy.” Brachytherapy 2:17–25. Stovall, M., and R. J. Shalek. (1972). “A review of computer techniques for dosimetry of interstitial and intracavitary radiotherapy.” Comput Programs Biomed 1:125. Stump, K. E., L. A. DeWerd, J. A. Micka, and D. R. Anderson. (2002). “Calibration of new high dose rate 192Ir sources.” Med Phys 29:1483–1488. Su, Y., B. J. Davis, M. G. Herman, and R. A. Robb. (2004). “Prostate brachytherapy seed localization by analysis of multiple projections: Identifying and addressing the seed overlap problem.” Med Phys 31(5):1277–1287. Subak, L., H. Hricak, B. Powell, L. Azizi, and J. L. Storm. (1995). “Cervical carcinoma: Computed tomography and magnetic resonance imaging for preoperative staging.” Obstet Gynecol 86(1):43–50. Susil, R. C., A. Krieger, J. A. Derbyshire, A. Tanacs, L. L. Whitcomb, E. R. McVeigh, G. Fichtinger, and E. Atalar. (2003). “System for MR Image–guided prostate interventions: Canine study.” Radiology 228:886–894.
36
Susil, R. C., A. Krieger, J. A. Derbyshire, A. Tanacs, L. L. Whitcomb, E. R. McVeigh, G. Fichtinger, E. Atalar. (2005b). “A system for MRI-guided diagnostic and therapeutic prostate interventions.” Radiology (in press). Susil, R. C., C. Ménard, A. Krieger, J. A. Coleman, K. Camphausen, P. Choyke, K. Ullman, S. Smith, G. Fichtinger, L. L. Whitcomb, N. C. Coleman, and E. Atalar. (2005a). “Transrectal prostate biopsy and fiducial marker placement in a standard 1.5T MRI scanner.” Urology (in review). Syed, A. M. N., A. Puthawala, D. Neblett et al. (1986). “Transperineal interstitial/intracavitary ‘SyedNeblett’ applicator in the treatment of carcinoma of the uterine cervix.” Endocurither/Hypertherm 2:1–13. Syed, A. M. N., A. Puthawala, N. Barth, A. Sharma, and A. Londrc. (1997). “High dose rate brachytherapy in the treatment of carcinoma of prostate: Preliminary results.“ J Brachytherapy No vol given:1–14. Syed, A. M. N., A. Puthawala, P. Austin, J. Cherlow, J. Perley, L. Tansey, A. Shanberg, D. Sawyer, R. Baghdassarian, B. Wachs, et al. (1992). “Temporary iridium-192 implant in the management of carcinoma of the prostate.” Cancer 69:2515–2524. Syed, A. M. N., and A. A. Puthawala. Proceedings of the Workshop on HDR and LDR Brachytherapy Techniques for Prostate, Head and Neck and GYN Malignancies. Long Beach, CA: Long Beach Endocuritherapy Foundation, 1996. Syed, A. M. N., and A. Puthawala. “Interstitial Implant Techniques” in Brachytherapy into the 21st Century. Long Beach, CA: Endocurietherapy Research Foundation, 1996. Syed, A. M. N., N. Feder, F. George, et al. (1975). “Management of Extensive Residual Cancer with Interstitial Iridium Implant: A Preliminary Report” in Afterloading: 20 Years of Experience 19551975. New York, NY, pp. 119–124, 1975. Taasan, V., B. Shapiro, J. A. Taren, and W. H. Beierwaltes. (1985). “Phosphorous–32 therapy of cystic grade IV astrocytoma: technique and preliminary application.” J Nucl Med 26:1335. Tapen, E. M., J. C. Blasko, P. D. Grimm, H. Ragde, R. Luse, S. Clifford, J. Sylvester, and T. W. Griffin. (1998). “Reduction of radioactive seed embolization to the lung following prostate brachytherapy.” Int J Radiat Oncol Biol Phys 42:1063–1067. Taschereau, R., J. Roy, and J. Pouliot. (1999). “Monte Carlo simulation of prostate implants to improve dosimetry and compare planning methods.” Med Phys 26(9):1952–1959. Taschereau, R., J. Roy, J. Pouliot, and D. Tremblay. (2000). “Seed misplacement and stabilizing needles in transperineal permanent prostate implants.” Radiother Oncol 55:59–63. Taschereau, R., P. Stauffer, I.-C. Hsu, J. L. Schlorff, A. J. Milligan, and J. Pouliot. (2004). “Radiation dosimetry of a conformal heat-brachytherapy applicator.” Technol Cancer Res Treat 3(4):347–358. Taylor, B. N., and C. E. Kuyatt. (1994). “Guidelines for Evaluating and Expressing the Uncertainty of NIST Measurement Results.” NIST Technical Note 1297. Washington, DC: U.S. Dept. of Commerce, 1994. Tedgren, A. K., and A. Ahnesjo. (2003). “Accounting for high Z shields in brachytherapy using collapsed cone superposition for scatter dose calculation.” Med Phys 30(8):2206–2217. Teirstein, P. S., V. Massullo, S. Jani, J. J. Popma, G. S. Mintz, R. J. Russo, R. A. Schatz, E. M. Guarneri, S. Steuterman, N. B. Morris, M. B. Leon, and P. Tripuraneni. (1997). “Catheter-based radiotherapy to inhibit restenosis after coronary stenting.” N Engl J Med 336:1697–1703. Tewari, K., F. Cappuccini, A. M. N. Syed, A. Puthawala, P. J. DiSaia, M. L. Berman, A. Manetta, and B. J. Monk. (1999). “Interstitial brachytherapy in the treatment of advanced and recurrent vulvar cancer.” Am J Obstet Gynecol 181(1):91-98. Thames, H. D., and J. H. Hendry. Fractionation in Radiotherapy. Bristol, PA: Taylor and Francis, 1987. Thomadsen, B. R. “Clinical Implementation of Remote Afterloading Interstitial Brachytherapy” in Brachytherapy Physics. J. F. Williamson, B. R. Thomadsen, and R. Nath (eds.). Madison, WI: Medical Physics Publishing, pp. 679–698, 1995. Thomadsen, B. R. Achieving Quality in Brachytherapy. Bristol: Institute of Physics Publishing, 1999. Thomadsen, B. R. Calculation checks for volume implants (personal communication). University of Wisconsin, Madison, WI, 1993.
37
Thomadsen, B. R., P. V. Houdek, R. van der Laarse, G. Edmundson, I. K. Kolkman-Deurloo, and A. G. Visser. “Treatment Planning and Optimization” in High Dose Rate Brachytherapy: A Textbook. S. Nag (ed.). Armonk, NY: Futura Publishing, pp. 79–145, 1994. Thomadsen, B. R., S. Shahabi, D. Buchler, W. Giese, and M. Mehta. (1990). “Differential loadings of brachytherapy templates.” Endocuriether/Hypertherm Oncol 6:197–202. Thomadsen, B. R., S. Shahabi, J. A. Stitt, D. A. Buchler, J. F. Fowler, B. R. Paliwal, and T. J. Kinsella. (1992). “High dose rate intracavitary brachytherapy for carcinoma of the cervix: The Madison system: II. Procedural and physical considerations.” Int J Radiat Oncol Biol Phys 24:349–357. Thomadsen, B., L. DeWerd, T. McNutt, S. DeWerd, and D. Schmidt. (1999). “Assessment of the strength of individual 192Ir seeds in ribbons.” Med Phys 26:2471–2475. Thomason, C., T. R. Mackie, and M. J. Lindstrom. (1991). “Effect of source encapsulation on the energy spectra of 192Ir and 137Cs seed sources.” Phys Med Biol 36:496–506. Thomsen, S. (1991). “Pathologic analysis of photothermal and photomechanical effects of laser-tissue interactions.” Photochem Photobiol 53(6):825–835. Tod, M. C., and W. J. Meredith. (1938). “A dosage system for use in the treatment of cancer of the uterine cervix.” Br J Radiol 11:809–824. Tod, M. C., and W. J. Meredith. (1953). “Treatment of cancer of cervix uteri—a revised ‘Manchester method.’” Br J Radiol 26:252–257. Todor, D. A., G. N. Cohen, H. I. Amols, and M. Zaider. (2002). “Operator-free, film-based 3D seed reconstruction in brachytherapy.” Phys Med Biol 47:2031–2048. Todor, D. A., M. Zaider, G. N. Cohen, M. F. Worman, and M. J. Zelefsky. (2003). “Intraoperative dynamic dosimetry for prostate implant.” Phys Med Biol 48:1153–1171. Tornes, A., and M. Eriksen. (2001). “Improved visualization of prostate brachytherapy sources.” J Brachytherapy Int 17:126. Tubic, D., A. Zaccarin, J. Pouliot, and L. Beaulieu. (2001a). “Automated seed detection and threedimensional reconstruction. I. Seed localization from fluoroscopic images or radiographs.” Med Phys 28:2265–2271. Tubic, D., A. Zaccarin, J. Pouliot, and L. Beaulieu. (2001b). “Automated seed detection and threedimensional reconstruction. II. Reconstruction of permanent prostate implants using simulated annealing.” Med Phys 28:2272–2279. Tungjitkusolmun, S., E. J. Woo, H. Cao, J. Z. Tsai, V. R. Vorperian, and J. G. Webster. (2000). “Finite element analyses of uniform current density electrodes for radio-frequency cardiac ablation.” IEEE Trans Biomed Engr 47(1):32–40. U.S. Nuclear Regulatory Commission Enforcement Action 04-087: “St. Vincent Hospital & Health Care Center – Notice of Violation.” Washington, D.C., 2004. U.S. Nuclear Regulatory Commission. (2003). Code of Federal Regulations, title 10, chapter 1: Energy, Part 35: Medical Use of By-product Material. Washington D.C.: Nuclear Regulatory Commission. U.S. Food and Drug Administration (FDA). (1998) “Guidance for the Submission of Premarket Notifications for Radionuclide Dose Calibrators.” U.S. Dept. of Health and Human Services, Food and Drug Administration, Center for Devices and Radiological Health, Radiological Devices Branch, http://www.ctg.ucl.ac.uk/TargitTest/index.htm, last accessed May 1, 2005. U.S. Food and Drug Administration (FDA). (2000) “Guidance for the Submission of Premarket Notifications for Photon-Emitting Brachytherapy Sources.” U.S. Dept. of Health and Human Services, Food and Drug Administration, Center for Devices and Radiological Health, Radiological Devices Branch, http://www.ctg.ucl.ac.uk/TargitTest/index.htm, last accessed May 1, 2005. U.S. Food and Drug Administration (FDA). (2005) “Guidance Documents.” U.S. Dept. of Health and Human Services, Food and Drug Administration, Center for Devices and Radiological Health, Radiological Devices Branch, http://www.fda.gov/cdrh/guidance.html, last accessed May 1, 2005. U.S. Nuclear Regulatory Commission (1981). Code of Federal Regulations, title 10, chapter 1, part 19 (Notices, Instructions, and Reports to Workers; Inspections). [Online] ission (1983). FC83-20, Revision 2–Facility Interlocks and Safety Devices for High, Medium, and Pulsed Dose-Rate Afterloading Units. Washington, DC. U.S. Nuclear Regulatory Commission (1986). FC86-4, Revision 1–Information Required for Licensing Remote Afterloading Devices. Washington, DC.
38
U.S. Nuclear Regulatory Commission (2002a). Code of Federal Regulations, title 10, parts 20, 32, and 35 (Medical Use of Byproduct Material: Final Rule). Federal Register, vol. 67, no. 79 (April 24):20250–20397. [Online]. Washington, DC: Nuclear Regulatory Commission. U.S. Nuclear Regulatory Commission (2002b). NRC Regulatory Issue Summary 2002-20, “Clarification of Requirements under 10 CFR 35.432, “Calibration Measurements of Brachytherapy Sources” [Online] ulatory Commission (2003). NRC Information Notice 2003-21: High-dose rate remote afterloader equipment failure. November 24, 2003. [Online] atory Commission (2004a). Licensing guidance Novoste and Guidant Intravascular Brachytherapy (IVB) systems. May 21, 2004. [Online] Commission (2004b). Newsletter NUREG/BR-0117/04-2: Nuclear Material Safety and Safeguards [Online] U.S. Nuclear Regulatory Commission. (1991a). Code of Federal Regulations, title 10, Rules and Regulations. Washington, DC: Nuclear Regulatory Commission. U.S. Nuclear Regulatory Commission. (1991b) Code of Federal Regulations, title 10, part 35, para 35.32, Energy, Rules and Regulations. Washington, DC: Nuclear Regulatory Commission. U.S. Nuclear Regulatory Commission. (1991c). Regulatory Guide 8.33 (Task DG-8001), Quality Management Program, para 3.1.6 and 3.2.10. Washington, DC: Nuclear Regulatory Commission. U.S. Nuclear Regulatory Commissions (NRC). (2005a) “Sealed Sources and Devices.” http://www.fda.gov/cdrh/guidance.htmlegistry.” ssed May 1, 2005. UL 60601 .org/publications/astronews/2005/Jan/IECRMayeport.htm |Hoppes, and F. J. Schima. (1992). “New and revised half-life measurements results.” Nucl Instrum Meth Phys Res A312:349–352. tory of HDR afterloading in brachytherapy.” Strahlenther Onkol 82 (suppl):2–6. Vaidya, J. S., M. Baum, J. S. Tobias, D. P. D’Souza, S. Morgan, S. V. Naidu, M. Metaxas, K. J. Harte, A. P. Sliski, and E. Thomson. (2001). “Targeted intraoperative radiotherapy (Targit): An innovative method of treatment for early breast cancer.” Ann Oncol 12:1075–1080. Vaidya, J. S., M. Baum, J. S. Tobias, S. Morgan, and D. D’Souza. (2002). “The novel technique of delivering targeted intraoperative radiotherapy (Targit) for early breast cancer.” Eur J Surg Oncol 28:447–454. van der Laarse, R. (1994). “The stepping source dosimetry system as an extension of the Paris System.” Brachytherapy from Radium to Optimization 34:319–330. van der Laarse, R., and R. W. De Boer. “Computerized Highr Dose Rate Brachytherapy Treatment Planning” in Brachytherapy HDR and LDR. A. A. Martinez, C. G. Orton, and R. F. Mould (eds.). Columbia, MD: Nucletron Corporation, 1990. van Dijk, E., I. K. K. Kolkman-Deurloo, and P. M. G. Damen. (2004). “Determination of the reference air kerma rate for 192Ir brachytherapy sources and the related uncertainty.” Med Phys 31:2826–2833. van Dongen, J. A., H. Bartelink, I. S. Fentiman, T. Lerut, F. Mignolet, G. Olthuis, E. van der Schueren, R. Sylvester, J. Winter, K. van Zijl, et al. (1992). “Randomized clinical trial to assess the value of breast-conserving therapy in stage I and II breast cancer, EORTC 10801 trial.” J Natl Cancer Inst Monogr 11:l5–18. Van Kleffens, H. J., and W. M. Sta. (1979. “Application of stereo x-ray photogrammetry (SRM) in the determination of absorbed dose values during intracavitary radiation therapy.” Int J Radiat Oncol Biol Phys 5:557. van Laarhoven, P. J. M., and E. H. L. Aarts. Simulated Annealing: Theory and Applications. Dordrecht, The Netherlands: D. Reidel Publishing, 1987. Van Miert, P. J., and J. F. Fowler. (1956). “The use of tantalum-182 in the treatment of early bladder cancer.” Br J Radiol 29:508. van’t Riet, A., A. C. Mak, M. A. Moerland. L. H. Elders, and W. van der Zee. (1997). “A conformation number to quantify the degree of conformality in brachytherapy and external beam irradiation: Application to the prostate.” Int J Radiat Oncol Biol Phys 37(3):731–736. Varro, Z., J. K. Locklin, and B. J. Wood. (2004). “Laser navigation for radiofrequency ablation.” Cardiovasc Intervent Radiol 27(5):512–515. Venselaar, J. L. M., P. H. van der Giessen, and W. J. F. Dries. (1996). “Measurement and calculation of the dose at large distances from brachytherapy sources: Cs-137, Ir-192 and Co-60.” Med Phys 23:537–543. Venselaar, J. L., H. W. Bierhuizen, and R. Klop. (1995). “A method to check treatment time calculations in Ir-192 high-dose-rate volume implants.” Med Phys 22:1499–1500.
39
Venselaar, J. L., W. F. Brouwer, B. H. van Straaten, and A. H. Aalbers. (1994). “Intercomparison of calibration procedures for Ir-192 HDR sources in The Netherlands and Belgium.” Radiother Oncol 30:155–161. Verhaegen, F., E. van Dijk, H. Thierens, A. Aalbers, and J. Seuntjens. (1992). “Calibration of low activity 192 Ir brachytherapy sources in terms of reference air kerma rate with large volume spherical ionization chambers.” Phys Med Biol 37:2071–2082. Verin, V. A., and Y. G. Popowski. “Schneider—Sauerwein Intravascular Radiation System” in Handbook of Vascular Brachytherapy. R. Waksman and P. W. Serruys. London: Martin Dunitz, Ltd., pp. 95–102, 1998. Veronesi, U., A. Banfi, M. Del Vecchio, R. Saccozzi, C. Clemente, M. Greco, A. Luini, E. Marubini, G. Muscolino, F. Rilke, et al. (1986). “Comparison of Halsted mastectomy with quadrantectomy, axillary dissection, and radiotherapy in early breast cancer: Long-term results.” Eur J Cancer Clin Oncol 9:1085–1089. Veronesi, U., A. Luini, M. Del Vecchio, M. Greco, V. Galimberti, M. Merson, F. Rilke, V. Sacchini, R. Saccozzi, T. Savio, et al. (1993). “Radiotherapy after breast-preserving surgery in women with localized cancer of the breast.” N Engl J Med 328:1587–1591. Vicini, F. A., K. L. Baglan, L. L. Kestin, C. Mitchell, P. Y. Chen, R. C. Frazier, G. Edmundson, N. S. Goldstein, P. Benitez, and R. R. Huang. (2001). “Accelerated treatment of breast cancer.” J Clin Oncol 19(7):1993–2001. Vicini, F. A., P. Y. Chen, M. Fraile, G. S. Gustafson, G. K. Edmundson, D. A. Jaffray, P. Benitez, J. Pettinga, B. Madrazo, J. A. Ingold, N. S. Goldstein, R. C. Matter, and A. A. Martinez. (1997). “Lowdose-rate brachytherapy as the sole radiation modality in the management of patients with earlystage breast cancer treated with breas-conserving therapy: Preliminary results of a pilot trial.” Int J Radiat Oncol Biol Phys 38(2):301–310. Vicini, F. A., V. R. Kini, G. Edmundson, G. S. Gustafson, J. Stromberg, and A. Martinez. (1999). “A comprehensive review of prostate cancer brachytherapy: Defining an optimal technique.” Int J Radiat Oncol Biol Phys 44:483–491. Vicini, F. A., V. R. Kini, P. Y. Chen, E. Horwitz, G. Gustafson, P. Benitez, G. Edmundson, N. Goldstein, K. McCarthy, and A. Martinez. (1999). “Irradiation of tumor bed alone after lumpectomy in selected patients with early-stage breast cancer treated with breast-conserving therapy.” J Surg Oncol 70:33–40. Vikram, B., and B. S. Hilaris. (1981). “A non-looping afterloading technique for interstitial implants at the base of the tongue.” Int J Radiat Oncol Biol Phys 7:419–422. Vikram, B., and S. Mishra. (1994). “Permanent iodine-125 implants in postoperative radiotherapy for head and neck cancer with positive surgical margins.” Head Neck 16(2):155–157. Visser, A. G. (1989). “An intercomparison of the accuracy of computer planning systems for brachytherapy.” Radiother Oncol 15:245–258. Viswanathan, A., B. Wood, N. Glossop, F. Banovac, K. Cleary, and J. Kruecker. (2005). “Multimodality Navigation for Radiofrequency Ablation (RFA) with Tracked Needles.” Annual Meeting of the Radiological Society of North America (RSNA), SSE03-06, Chicago, IL. Waksman, R. (1999). “Intracoronary radiation therapy for restenosis prevention: status of the clinical trials.” Cardiovasc Radiat Med 1:2059. Waksman, R., A. E. Ajani, R. L. White, C. Chan, B. Bass, A. D. Pichard, L. F. Satler, K. M. Kent, R. Torguson, R. Deible, E. Pinnow, and J. Lindsay. (2004). “Five-year follow up after intracoronary gamma radiation therapy for in-stent restenosis.” Circulation 109(3):340–344. Waksman, R., A. E. Ajani, R. L. White, E. Pinnow, R. Dieble, T. B. Bui, M. Taaffe, L. Gruberg, G. S. Mintz, L. F. Satler, A. D. Pichard, K. M. Kent, and J. Lindsay. (2001). “Prolonged antiplatelet therapy to prevent late thrombosis after intracoronary gamma-radiation in patients with in-stent restenosis: Washington Radiation for In-Stent Restenosis Trial plus 6 months of clopidogrel (WRIST PLUS).” Circulation 103:2332–2335. Waksman, R., A. E. Raizner, A. C. Yeung, A. J. Lansky, and L. Vandertie. (2002). “Use of localised intracoronary beta radiation in treatment of in-stent restenosis: The INHIBIT randomised controlled trial. Lancet 359:551–557. Erratum in Lancet 359:1950.
40
Waksman, R., E. Cheneau, and A. E. Ajani. (2003). “Intracoronary radiation therapy improves the clinical and angiographic outcomes of diffuse in-stent restenosis lesions: Results of the Washington Radiation for In-Stent Restenosis Trial for Long Lesions (Long WRIST) Studies.” Circulation 107(13):1744–1749. Waksman, R., R. L. White, R. C. Chan, B. G. Bass, L. Geirlach, G. S. Mintz, L. F. Satler, R. Mehran, P. W. Serruys, A. J. Lansky, P. Fitzgerald, B. Bhargava, K. M. Kent, A. D. Pichard, and M. B. Leon. (2000). “Intracoronary gamma-radiation therapy after angioplasty inhibits recurrence in patients with in-stent restenosis.” Circulation 101:2165–2171. Wallace, R. E., and J. J. Fan. (1998). “Evaluation of a new brachytherapy iodine-125 source by AAPM TG43 formalism.” Med Phys 25:2190–2196. Wallace, R. E., and J. J. Fan. (1999). “Report on the dosimetry of a new design 125iodine brachytherapy source.” Med Phys 26:1925–1931. Wallace, R. E., and J. J. Fan. (1999a). “Dosimetric characterization of a new design 103palladium brachytherapy source.” Med Phys 26:2465–2470. Wang, R., and R. S. Sloboda. (1998a). “Influence of source geometry and materials on the transverse axis dosimetry of 192Ir brachytherapy sources.” Phys Med Biol 43(1):37–48. Wang, R., and R. S. Sloboda. (1998b). “Monte Carlo dosimetry of the VariSource high dose rate 192Ir source.” Med Phys 25(4):415–423. Wang, X. H., and L. Potters. (2001). “A theoretical derivation of the nomograms for permanent prostate brachytherapy.” Med Phys 28(4):683–687. Watanabe, Y., and L. L. Anderson. (1997). “A system for nonradiographic source localization and realtime planning of intraoperative high dose rate brachytherapy.” Med Phys 24(12):2014–2023. Wazer, D. E., L. Berle, R. Graham, M. Chung, J. Rothschold, T. Graves, B. Cady, K. Ulin, R. Ruthazer, and T. A. DiPetrillo. (2002). “Preliminary results of a phase I/II study of HDR brachytherapy alone for T1/T2 breast cancer.” Int J Radiat Oncol Phys 53(4):889–897. Weaver, J. T., T. P. Loftus, and R. Loevinger. “NBS Measurement Services: Calibration of Gamma-RayEmitting Brachytherapy Sources.” National Bureau of Standards (NIST) Special Publication 250-19, Washington, DC: NIST, 1988. Weaver, K. (1998). “Anisotropy functions for I-125 and Pd-103 sources.” Med Phys 25:2271–2278. Weaver. K. A. (1986). “The dosimetry of 125I seed eye plaques.” Med Phys 13:78–83. Weaver K. A., R. Peksens, and C. Barnett. “Special Brachytherapy Procedures and Dosimetry for Tumors of the Brain, Eye, Head and Neck, and Perineum” in Radiation Oncology Physics 1986. J. G. Kereiakes, H. R. Elson, and C. G. Born. (eds.). AAPM Medical Physics Monograph No. 15. New York: American Institute of Physics, pp. 677–699, 1987. Weaver, K. A., V. Smith, D. Huang, C. Barnett, M. Schell, and C. C. Ling. (1989). “Dose parameters of 125i and 192ir seed sources.” Med Phys 18:636–643. Weeks, K. J. (1998). “Monte Carlo dose calculations for a new ovoid shield system for carcinoma of the uterine cervix.” Med Phys 25(12):2288–2292. Weeks, K. J., and G. S. Montana. (1997). “Three-dimensional applicator system for carcinoma of the uterine cervix.” Int J Radiat Oncol Biol Phys 37:455–463. Weeks, K. J., and J. C. Dennett. (1990). “Dose calculation and measurements for a CT-compatible version of the Fletcher applicator.” Int J Radiat Oncol Biol Phys 18:1191–1198. Weeks, K. J., G. S. Montana, and G. C. Bentel. (1991). “Design of a plastic minicolpostat applicator with shields.” Int J Radiat Oncol Biol Phys 21(4):1045-1052. Weeks, K. J., S. L. Schoeppel, K. Pruss, M. Hopkins, and C. Perez-Tamayo. (1989). “Computer tomography-compatible afterloading Fletcher-Suit-Delclos colpostat with adjustable shielding.” Endocurie Hypertherm Oncol 5:169–174. Wefer, A. E., H. Hricak, D. B. Vigneron, F. V. Coakley, Y. Lu, J. Wefer, U. Mueller-Lisse, P. R. Carroll, and J. Kurhanewicz. (2000). “Sextant localization of prostate cancer: Comparison of sextant biopsy, magnetic resonance imaging and magnetic resonance spectroscopic imaging with step section histology.” J Urol 164:400–404. Whitmore, W. F., B. S. Hilaris, H. Grabstald, and M. Batata. (1974). “Implantation of 125I in prostatic cancer.” Surg Clin North Am 54:887–895.
41
Whitmore, W. F., Jr., B. Hilaris, and H. Grabstald, H. (1972). “Retropubic implantation of iodine-125 in the treatment of prostatic cancer.” J Urol 108:918–920. Whitmore, W. F., Jr., B. Hilaris, and H. Grabstald, H. (2002). “Retropubic implantation of iodine-125 in the treatment of prostatic cancer. 1972.” J Urol 167(2Pt2):981–983; discussion 984. Whyte, G. H. (1955). “Attenuation of g radiation in cylindrical geometry.” Br J Radiol 28:635. Wilder, R. B., J. M. Buatti, J. M. Kittelson, D. S. Shimm, P. M. Harari, E. E. Rogoff, and J. R. Cassady. (1992). “Pterygium treated with excision and postoperative beta irradiation.” Int J Radiat Oncol Biol Phys 23:533–537. Williamson, J. F. (1986). “The accuracy of the line and point dose approximation in Ir-192 dosimetry.” Int J Radiat Oncol Biol Phys 12:409. Williamson, J. F. (1988). “Monte Carlo and analytic calculation of absorbed dose near 137Cs intracavitary sources.” Int J Radiat Oncol Biol Phys 15:227–237. Williamson, J. F. (1988). “Theoretical evaluation of dose distributions in water about models 6711 and 6702 125I seeds.” Med Phys 15:891–897. Williamson, J. F. (1988). “Monte Carlo evaluation of specific dose constants in water for 125I seeds.” Med Phys 15:686–694. Williamson, J. F. (1990). “Dose calculations about shielded gynecological colpostats.” Int J Radiat Oncol Biol Phys 19:167–178. Williamson, J. F. (1991). “Comparison of measured and calculated dose rates in water near I-125 and Ir192 seeds.” Med Phys 28:776–786. Williamson, J. F. (1996). “The Sievert integral revisited: Evaluation and extension to low energy brachytherapy sources.” Int J Radiat Oncol Biol Phys 36:1239–1250. Williamson, J. F. (1998b). “Monte Carlo-based dose-rate tables for the Amersham CDCS.J and 3M model 3600 137Cs tubes.” Int J Radiat Oncol Biol Phys 41(4):959–970. Williamson, J. F. (2000). “Monte Carlo modeling of the transverse-axis dose distribution of the model 200 103Pd interstitial brachytherapy source.” Med Phys 27:643–654. Williamson, J. F. “Clinical Brachytherapy Physics” in Principles and Practice of Radiation Oncology, 4th ed. C. A. Perez, L. W. Brady, E. C. Halperin, and R. K. Schmidt-Ullrich (eds.). Philadelphia: Lippincott Williams & Wilkins, 2003. Williamson, J. F. “Monte Carlo Simulation of Photon Transport Phenomena” in Monte Carlo Simulation in the Radiological Sciences. R. L. Morin (ed.). Boca Raton, FL: CRC, pp. 53–102, 1988. Williamson, J. F. “Physics of Brachytherapy” in Principles and Practice of Radiation Oncology, 3rd Edition. C. A. Perez and L. W. Brady (eds.). Philadelphia: Lippincott-Raven, 1998a. Williamson, J. F., and D. A. Brenner. “Physics and Radiobiology of Brachytherapy” in Principles and Practice of Radiation Oncology, 4rd Edition. C. A. Perez, L. W. Brady, E. Halperin et al. (eds.). Philadelphia: J.B. Lippincott Company, pp. 472–537, 2003. Williamson, J. F., and R. Nath. (1991). “Clinical implementation of AAPM Task Group 32 recommendations on brachytherapy source strength specification.” Med Phys 18:439–448. Williamson, J. F., and T. Seminoff. (1987). “Template-guided interstitial implants: 137Cs-reusable sources as a substitute for 192Ir.” Radiology 165:265–269. Williamson, J. F., and Z. Li. (1995). “Monte Carlo aided dosimetry of the microselectron pulsed and high dose-rate 192Ir sources.” Med Phys 22(6):809–819. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, and R. Nath. (1998). “Dosimetric prerequisites for clinical use of new low energy photon interstitial brachytherapy sources.” Med Phys 25:2269–2270. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, G. Ibbott, R. Nath, and M. J. Rivard. (2000a). “Important notice for radiation therapy physicists using 125I or 103Pd brachytherapy sources (Letter).” Online: http://rpc.mdanderson.org/rpc/htm/Home_htm/Lowenergy%20documents/ImpNotice_EndUsers_v12.pdf. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, R. Nath, and G. Ibbott. (1999). “Guidance to users of Nycomed Amersham and North American Scientific, Inc. I-125 interstitial sources: Recommendations of the AAPM Radiation Therapy Committee ad hoc Subcommittee on low-energy seed dosimetry.” Med Phys 26:570–573.
42
Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, R. Nath, M. J. Rivard, and G. Ibbott. (1999a). “On the use of apparent activity (Aapp) for treatment planning of 125I and 103Pd interstitial brachytherapy sources: Recommendations of the AAPM Radiation Therapy Committee Subcommittee on low-energy brachytherapy source dosimetry.” Med Phys 26:2529–2530. Williamson, J. F., B. M. Coursey, L. A. DeWerd, W. F. Hanson, R. Nath, M. J. Rivard, and G. Ibbott. (2000). “Recommendations of the American Association of Physicists in Medicine on 103Pd interstitial source calibration and dosimetry: Implications for dose specification and prescription.” Med Phys 27:634–642. Also available as AAPM Report No. 69. Williamson, J. F., F. M. Khan, S. C. Sharma, and G. D. Fullerton. (1982). “Methods for the routine calibration of brachytherapy sources,” Radiology 142:511–516. Williamson, J. F., G. Ezzell, A. Olch, and B. R. Thomadsen. “Quality Assurance of High Dose Rate Brachytherapy” in High Dose Rate Brachytherapy: A Textbook. S. Nag (ed.), Armonk, NY: Futura Publishing, pp. 147–212, 1994. Williamson, J. F., H. Perera, Z. Li, and W. R. Lutz. (1993). “Comparison of calculated and measured heterogeneity correction factors for 125I, 137Cs and 192Ir brachytherapy sources near localized heterogeneities.” Med Phys 20:209–222. Williamson, J. F., L. L. Anderson, D. W. Grigsby, A. Martinez, R. Nath, D. Neblett, A. Olch, and K. Weaver. (1993). “American Endocurietherapy Society recommendations for specification of brachytherapy source strength.” Endocuriether/Hypertherm Oncol 9:1–7. Williamson, J. F., R. Baker, and Z. Li. (1991). “A convolution algorithm for brachytherapy dose computations in heterogeneous geometries.” Med Phys 18:1256–1265. Williamson, J. F., Z. Li, and J. W. Wong. (1993). “One-dimensional scatter-subtraction method for brachytherapy dose calculation near bounded heterogeneities.” Med Phys 20:233–244. Williamson, J., G. Daskalov, R. Baker, D. Rogers, and I. Kawrakow. (2001). “Accuracy and efficiency comparisons between three-dimensional multigroup discrete ordinates and voxel based Monte Carlo methods for dosimetric modeling of the Model 6702 125I seed.” (Abstract). Med Phys 28(6):1229. Withers, H. R., J. M. G. Taylor, and B. Maciejewski. (1988). “The hazard of accelerated tumor clonogen repopulation during radiotherapy.” Acta Oncol 27:131–146. Wu, A., E. S. Sternick, and D. J. Muise. (1988). “Effect of gold shielding on the dosimetry of an I-125 seed at close range.” Med Phys 15: 627–628. Wu, A., and F. Krasin. (1990). “Film dosimetry analyses on the effect of gold shielding for iodine-125 eye plaque therapy for choroidal melanoma.” Med Phys 17:843–846. Wu, A., K. Ulin, and E. S. Sternick. (1988). “A dose homogeneity index for evaluating 192Ir interstitial breast implants.” Med Phys 15:104–107. Wu, A., K. Ulin, and E. S. Sternick. (1988). “A dose uniformity index for evaluating Ir-192 interstitial breast implants.” Med Phys 15:104–107. Wu, T., J. P. Felmlee, J. F. Greenleaf, S. J. Riederer, and R. L. Ehman. (2001). “Assessment of thermal tissue ablation with MR elastography.” Magn Reson Med 45(1):80–87. Wuu, C. S., R. D. Ennis, P. B. Schiff, E. K. Lee, and M. Zaider. (2000). “Dosimetric and volumetric criteria for selecting a source activity and a source type (125I or 103Pd) in the presence of irregular seed placement in permanent prostate implants.” Int J Radiat Oncol Biol Phys 47:815–820. Wyckoff, H. O. “From ‘Quantify of Radiation’ and ‘Dose’ to ‘Exposure’ and ‘Absorbed Dose:' An Historical Review.” Report No. Lauriston Taylor Lecture 4, National Council on Radiation Protection and Measurements, Washington, DC, 1980. Xia, P., N. Lee, Y.-M. Liu, I. Poon, V. Weinberg, E. Shin, J. M. Quivey, and L. J. Verhey. (2004). “A study of planning dose constraints for treatment of nasopharyngeal carcinoma using a commercial inverse treatment planning system.” Int J Radiat Oncol Biol Phys 59(3):886–896. Yamada, Y., L. Potters, M. Zaider, G. Cohen, E. Venkatraman, and M. Zelefsky. (2003). “Impact of intraoperative edema during transperineal permanent prostate brachytherapy on computer-optimized and preimplant planning techniques.” Am J Clin Oncol 26(5):130–135. Yanch, J. C., and K. J. Harte. (1996). “Monte Carlo simulation of a miniature, radiosurgery x-ray tube using the ITS 3.0 coupled electron-photon transport code.” Med Phys 23:1551–1558. Yanch, J. C., and N. T. Zervas. (1995). “The photon radiosurgery system.” Science Med 2:38–46.
43
Yang, G., L. E. Reinstein, S. Pai, Z. Xu, and D. L. Carroll. (1998). “A new genetic algorithm technique in optimization of permanent 125I prostate implants.” Med Phys 25(12):2308–2315. Yannuzzi, L. A., M. D. Ober, J. S. Slakter, R. F. Spaide, Y. L. Fisher, R. W. Flower , and R. Rosen. (2004). “Ophthalmic fundus imaging: Today and beyond.” Am J Ophthalmol 137(3):511–524. Yoo, S., M. E. Kowalok, B. R. Thomadsen, and D. L. Henderson. (2003). “Treatment planning for prostate brachytherapy using region of interest adjoint functions and a greedy heuristic.” Phys Med Biol 48(24):4077–4090. Young, M. E. J., and H. F. Batho. (1964). “Dose tables for linear radium sources calculated by an electronic computer.” Br J Radiol 37:38–44. Yu, K. K., H. Hricak, L. Subak et al. (1995). “Preoperative staging of cervical cancer with MR imaging: Comparison of body versus phase array coil.” Radiology 197:321. Yu, Y., and M. C. Schell. (1996). “A genetic algorithm for the optimization of prostate implants.” Med Phys 23:2085–2091. Yu, Y., J. B. Zhang, G. Cheng, M. C. Schell, and P. Okunieff. (2000). “Multi-objective optimization in radiotherapy: Applications to stereotactic radiosurgery and prostate brachytherapy” Artif Intell Med 19:39–51. Yu, Y., J. B. Zhang, R. A. Brasacchio, P. G. Okunieff, D. J. Rubens, J. G. Strang, A. Soni, and E. M. Messing. (1999). “Automated treatment planning engine for prostate seed implant brachytherapy.” Int J Radiat Oncol Biol Phys 43(3):647–652. Yu, Y., L. L. Anderson, Z. Li, D. E. Mellenberg, R. Nath, M. C. Schell, F. M. Waterman, A. Wu, and J. C. Blasko. (1999). “Permanent prostate seed implant brachytherapy: Report of the American Association of Physicists in Medicine Task Group No. 64.” Med Phys 26(10):2054–2076. Also available as AAPM Report No. 68. Zagoria, R. J. (2004). “Imaging-guided radiofrequency ablation of renal masses.” Radiographics 24 Suppl 1:S59–S71. Zaider, M., and G. N. Minerbo. (2000). “Tumor control probability: A formulation applicable to any temporal protocol of dose delivery.” Phys Med Biol 45:279–293. Zaider, M., and H. H. Rossi. Radiation Science for Physicians and Public Health Workers. Norwell, MA: Kluwer Academic Publishers, 2001. Zaider, M., C.-S. Chui, G. Cohen, D. A. Silvern, and P. Gajiwala. (2000a). “Real-time optimization for prostate LDR brachytherapy.” Int J Radiat Oncol Biol Phys 48:601–608. Zaider, M., M. J. Zelefsky, E. K. Lee, K. L. Zakian, H. I. Amols, J. Dyke, G. Cohen, Y. C. Hu, A. K. Endi, C.-S. Chui, and J. A. Koutcher. (2000). “Treatment planning for prostate implants using magnetic-resonance spectroscopy imaging.” Int J Radiat Oncol Biol Phys 47(4):1085–1096. Zaider, M., M. J. Zelefsky, G. N. Cohen, C.-S. Chui, E. D. Yorke, L. Ben-Porat, and L. Happersett. (2005). “Methodology for biologically-based treatment planning for combined low-dose-rate (permanent implant) and high-dose-rate (fractionated) treatment of prostate cancer.” Int J Radiat Oncol Biol Phys 61:702–713. Zannis, V. J., L. C. Walker, B. Barclay-White, and C. A. Quiet. (2003). “Postoperative ultrasound-guided percutaneous placement of a new breast brachytherapy balloon catheter.” Am J Surg 186:383–385. Zelefsky, M, J., Y. Yamada, C. Marion, S. Sim, G. Cohen, L. Ben-Porat, D. Silvern, and M. Zaider. (2003). “Improved conformality and decreased toxicity with intraoperative computer-optimized transperineal ultrasound-guided prostate brachytherapy.” Int J Radiat Oncol Biol Phys 55(4):956–963. Zelefsky, M. J., Y. Yamada, G. Cohen, E. S. Venkatraman, A. Y. C. Fung, E. Furhang, D. Silvern, and M. Zaider. (2000). “Postimplantation dosimetric analysis of permanent transperineal prostate implantation: improved dose distributions with an intraoperative computer-optimized conformal planning technique.” Int J Radiat Oncol Biol Phys 48:601–608. Zhang, P., A. S. Beddar, and C. H. Sibata. (2004). “AAPM TG-43 formalism for brachytherapy dose calculation of a 137Cs tube source.” Med Phys 31(4):755–759. Zhou, C., and F. Inanc. (2003). “Integral-transport-based deterministic brachytherapy dose calculations.” Phys Med Biol 48(1):73–93. Zimmerman, R. A., and L. T. Bilaniuk. (1988). “Ocular MR imaging.” Radiology 168:875–876.
44
Zwicker, R. D., and R. Schmidt-Ullrich. (1995). “Dose uniformity in a planar interstitial implant system.” Int J Radiat Oncol Biol Phys 31:149–155. Zwicker, R. D., R. Schmidt-Ullrich, and B. Schiller. (1985). “Planning of Ir-192 seed implants for boost irradiation to the breast.” Int J Radiat Oncol Biol Phys 11:2163–2170.
45
