Optical Imaging of Cancer: Clinical Applications

Optical Imaging of Cancer

Eben Rosenthal · Kurt R. Zinn Editors

Optical Imaging of Cancer Clinical Applications

123

Editors Eben Rosenthal Division of Otolaryngology University of Alabama Birmingham 1808 7th Avenue South Birmingham AL 35294-0012 USA [email protected]

Kurt R. Zinn Department of Radiology University of Alabama Birmingham 1530 3rd Avenue South Birmingham AL 35294-0012 Boshell Bldg. USA [email protected]

ISBN 978-0-387-93873-8 e-ISBN 978-0-387-93874-5 DOI 10.1007/978-0-387-93874-5 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2009928502 © Springer Science+Business Media, LLC 2009 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)

‘For Mary, Walker and Sarah’ Eben Rosenthal ‘For Tandra, and her dedication to imaging.’ Kurt R. Zinn

Contents

Part I

Optical Imaging Principles

Optical Imaging of Cancer: Enhancing Detection and Resection . . . . Kent T. Keyser and Christianne E. Strang

3

Endoscopic Techniques for Optical Imaging . . . . . . . . . . . . . . . . E. Namati, M.J. Suter, and G. McLennan

25

Design and Use of the Surgical Microscope in Fluorescence-Guided Surgery . . . . . . . . . . . . . . . . . . . . . . Max Sturgis Fluorophores for Optical Imaging . . . . . . . . . . . . . . . . . . . . . Iain Johnson Part II

49 59

Cancer Targeting Strategies

Overview of Cancer Detection and Monitoring Strategies . . . . . . . . Kurt R. Zinn

81

The Application of Tissue Autofluorescence in Detection and Management of Oral Cancer and Premalignant Lesions . . . . . . C.F. Poh, P. Lane, C. MacAulay, L. Zhang, and M.P. Rosin

101

Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy . . . . . . . . . . . . . . . . . . . . . . . . . . J. Oliver McIntyre and Lynn M. Matrisian

119

Illustrating Molecular Events with Light: A Perspective on Optical Reporter Genes . . . . . . . . . . . . . . . . . . . . . . . . . Pritha Ray

139

vii

viii

Contents

Part III Preclinical and Clinical Investigations Optical Imaging of Primary Tumors . . . . . . . . . . . . . . . . . . . . J. Robert Newman and Eben L. Rosenthal Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E.M. Sevick-Muraca

163

185

Optical Coherence Tomography for Cancer Detection . . . . . . . . . . Steven G. Adie and Stephen A. Boppart

209

Optical Imaging of Cancer: Neuro-oncologic Applications . . . . . . . . Stephen Yip and Khalid Shah

251

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

265

Contributors

Kent T. Keyser, PhD Department of Vision Sciences, University of Alabama at Birmingham, WORB 626, 1530 Third Avenue South, Birmingham, AL 35294-4390, USA, [email protected] Christianne E. Strang, PhD Department of Vision Sciences, University of Alabama at Birmingham, 1530 Third Avenue South, Birmingham, AL 35294-4390, USA, [email protected] Eman Namati, PhD Department of Internal Medicine, Carver College of Medicine, University of Iowa, Iowa City, IA, USA Melissa J. Suter, PhD Harvard Medical School, Wellman for Photomedicine, Massachusetts General Hospital, Boston, MA, USA

Center

Geoffrey McLennan, MD, PhD Departments of Internal Medicine, Biomedical Engineering and Radiology, Carver College of Medicine, University of Iowa, Iowa City, IA, USA, [email protected] Max Sturgis R&D Engineering – Surgical Consultant, Business Unit Surgical Operating Microscope, Leica Microsystems AG, Max Schmidheiny-Strasse 201, CH-9435 Heerbrugg, Switzerland, [email protected] Iain Johnson, PhD Life Technologies Corporation, 29851 Willow Creek Road, Eugene, OR 97402, USA, [email protected] Kurt R. Zinn, PhD Department of Radiology, 1530 3rd Avenue South, Boshell Building, Birmingham, AL 35294-0012, USA, [email protected] Catherine F. Poh, DDS, PhD Faculty of Dentistry, University of British Columbia, Cancer Control Research and Cancer Imaging, BC Cancer Agency/Cancer Research Centre, Rm: JBM 322, 2199 Wesbrook Mall, Vancouver, BC Canada V6T 1Z3, [email protected] Pierre M. Lane, PhD Cancer Imaging, BC Cancer Agency/Cancer Research Centre, 675 West 10th Avenue, Vancouver, BC, Canada V5Z 1L3, [email protected] Lewei Zhang, DMD, PhD Faculty of Dentistry, University of British Columbia, Rm: JBM 322, 2199 Wesbrook Mall, Vancouver, BC, Canada, V6T 1Z3, [email protected]

ix

x

Contributors

Calum E. MacAulay, PhD Cancer Imaging, BC Cancer Agency/Cancer Research Centre, 675 West 10th Avenue, Vancouver, BC, Canada V5Z 1L3, [email protected] Miriam P. Rosin, PhD Cancer Control Research, BC Cancer Agency/Cancer Research Centre, Biomedical Physiology & Kinesiology, Simon Fraser University, BC Cancer Agency Research Centre, 675 West 10th Avenue, Vancouver, BC, Canada V5Z 1L3, [email protected] J. Oliver McIntyre, PhD Department of Cancer Biology, Vanderbilt-Ingram Cancer Center and Vanderbilt University Institute of Imaging Science, Vanderbilt University, Medical Center, Nashville, TN 37232-6840, USA, [email protected] Lynn M. Matrisian, PhD Department of Cancer Biology, and VanderbiltIngram Cancer Center, Vanderbilt University Medical Center, Nashville, TN 37232-6840, USA Pritha Ray, PhD Advanced Center for Treatment, Research and Education in Cancer (ACTREC), Tata Memorial Center, Kharghar, Navi Mumbai, Maharastra, 410210, India, [email protected] Eben L. Rosenthal, MD Division of Otolaryngology – Head and Neck Surgery, Department of Surgery, University of Alabama at Birmingham, Birmingham, AL, USA, [email protected] J. Robert Newman, MD Division of Otolaryngology – Head and Neck Surgery, Department of Surgery, University of Alabama at Birmingham, Birmingham, AL, USA, [email protected] E.M. Sevick-Muraca, PhD The University of Texas Health Science Center, The Brown Foundation for Molecular Medicine, Center of Molecular Imaging, SRB 330A, Houston, TX 77030, USA, [email protected] Steven G. Adie, PhD Biophotonics Imaging Laboratory, Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA Stephen A. Boppart, MD, PhD Biophotonics Imaging Laboratory, Beckman Institute for Advanced Science and Technology; Departments of Electrical and Computer Engineering, Bioengineering, and Medicine, Colleges of Engineering and Medicine, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA, [email protected]

Contributors

xi

Stephen Yip, PhD Molecular Neurotherapy and Imaging Laboratory, Massachusetts General Hospital, Harvard Medical School, Bldg 149, 13th Street, Boston, MA 02129, USA Khalid Shah, PhD Molecular Neurotherapy and Imaging Laboratory, Massachusetts General Hospital, Harvard Medical School, Boston MA 02129, USA, [email protected]

Introduction

Optical detection for cancer represents the next great horizon for translational imaging. As molecularly targeted therapeutic agents are delivered to the clinic in increasing numbers, there is a parallel opportunity to advance optical imaging techniques. Because of its limited toxicity and potential for real-time information, optical imaging represents an underdeveloped modality in clinical medicine. This book explores the preclinical and clinical data to support the use of these techniques in cancer imaging. Fluorescent delivery techniques from monoclonal antibodies to vector delivery are explored in this book. Indications for use include monitoring vector delivery, early detection, and operative removal techniques in all stages of clinical development. This book represents the best and most clinically relevant techniques currently available. Birmingham, Alabama Birmingham, Alabama

Eben Rosenthal Kurt Zinn

xiii

Optical Imaging of Cancer: Enhancing Detection and Resection Kent T. Keyser and Christianne E. Strang

The purpose of this chapter is to provide a brief review of selected principles of microscopy and an overview of contemporary imaging methods, with an emphasis on fluorescence techniques used in biomedical research in general and cancer research in particular. For additional information, the reader is referred to the end of the chapter for a short list of excellent references and web sites that provide useful information. Light has both particle and wave characteristics and as a wave has a specific frequency and wavelength that we perceive as color. Light is energy and one can calculate the amount of energy carried by specific wavelengths of light with Planck’s law which states the energy in light, E = hv, with E given in ergs, h Planck’s constant, and v the frequency of light. The human eye is sensitive to light between 400 (indigo) and 700 nm (near infrared). The number and variety of tissues that are involved in seeing illustrate how critical vision is for people. For example, there are all of the structures of the eye that house and support the retina and ensure, at least in many people, that a focused image falls on the retina. The output neurons of the retina, the retinal ganglion cells, convey visual information to several target nuclei in the thalamus. These structures in turn project to regions of the cerebral cortex where higher level processing of visual information occurs. Remarkably, physiological studies suggest that at least 50% of the cortex of the human brain is devoted to processing visual information.

Limits of Resolution: Numerical Aperture and Conventional Microscopy Microscopes allow the visualization and study of objects that are much smaller than can be resolved by the naked eye. There are three important terms to bear in mind: 1. Magnification: the apparent enlargement of an object by an optical instrument (or computer program).

K.T. Keyser (B) Department of Vision Sciences, University of Alabama at Birmingham, AL 35294-4390, USA e-mail: [email protected]

E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_1, 

3

4

K.T. Keyser and C.E. Strang

2. Resolution: the smallest distance between two objects that allows them still to be distinguished as two separate things. This is the limiting factor in light microscopy. 3. Numerical aperture: this term refers to the efficiency of a light collecting element, usually a microscope objective. Efficiency refers to how well the objective collects light from a specimen. In general terms, the more efficiently the element collects the light, the higher the numerical aperture, and the better the resolution. Figure 1 will help to illustrate the concept of numerical aperture (NA). Light from an object radiates in all directions, and we can characterize the angle of the light given off from the object relative to the angle of the incident light. It is clear that objectives that collect light over a larger angle will yield a brighter image. The ability of an objective to collect light is indicated by the NA such that the larger the NA the more light will be collected by the objective. The NA of a microscope objective is a function of the angle over which light collected and of the refractive index, η, of the medium between the specimen and the objective such that NA = ηsinθ . At the interface between the cover glass and air, or between the air and the objective, light is reflected or refracted. The resultant loss of light from the specimen affects resolution and the refraction of light introduces aberration. One way to improve this situation is to introduce a material that has a refractive index very similar to that of the glass, such as immersion oil, between the cover glass and the lens.

Fig. 1 Dry objective vs. oil immersion. Light leaves the specimen at an angle relative to the angle of incident light. The larger the angle through which the objective collects light, the brighter the image. However, differences in the refractive index, η, of the medium between the specimen and the objective can cause diffraction, reflection, and light scatter, which decrease the effective numerical aperture and the resolution of the objective. The refractive index of immersion oil and glass is very similar and decreases aberration and light scatter and increases resolution

This increases the light collecting ability of the lens and reduces aberration. In summary, the NA of the light collecting element is the major factor that determines the resolution of a microscope objective such that the higher the NA, the better the resolution.

Fluorescence Imaging Fluorescence refers to a property displayed by some molecules in which absorption of light of a particular wavelength is followed by emission of light that is of longer wavelength than the light that was absorbed. Photon absorption occurs very rapidly,

Optical Imaging of Cancer

5

Fig. 2 Energy diagram of a fluorophore. Fluorescence is a process by which some molecules absorb light of a specific wavelength and after a very short period of time emit light at a longer wavelength. Photon absorption results in the movement of the molecule into an excited state. There is energy loss to the environment, after which the molecule relaxes to the lowest excited singlet state. Then after a few nanoseconds, the molecule relaxes to the ground state with the emission of a photon, i.e., fluorescence

on the order of 10–15 s, and the molecule moves from the ground state to an excited state (Fig. 2). Shortly thereafter energy is lost to the environment (10–11 s) through dipole interactions, and the molecule relaxes to the lowest excited singlet state and remains there for a few nanoseconds. Relaxation from this state is accompanied by the emission of a photon and, since energy was lost to the environment before this relaxation, the light that is emitted is of lower energy and hence longer wavelength. Remember that the energy in light is given by E = hV and so high-frequency (short λ) light has more energy than lower frequency (longer λ) light. The difference in frequency (or wavelength) between the absorbed light and the emitted light is called the Stokes shift and it is this difference that makes fluorescence-based imaging possible. The Stokes shift is named after Sir George G. Stokes who, in 1852, published a paper on fluorescence and the change in the wavelength of light. The experiment that was the basis for this paper involved the sun, a piece of blue glass from a church window, a solution of quinine, and a glass of wine that was used as an emission filter. Since the experiment was a success, perhaps we can assume the wine was not wasted after the experiment. The absorption and emission spectra of a hypothetical fluorophore are shown in Fig. 3 and, as depicted in the figure, in the absence of other interactions the emission spectrum is very similar to absorption spectrum. This is because an electron that moved from the ground state to the excited state upon photon absorption has a high probability of returning to the same vibrational ground

6

K.T. Keyser and C.E. Strang

Fig. 3 Absorption–emission spectra. The absorption spectra of this hypothetical dye peaks at 492 nm, while the emission spectra peaks at 520 nm. The energy emitted is less than the energy absorbed due to the Stokes shift

state from which it started. The symmetry between the absorption and the excitation spectra is termed the “mirror image rule.” Each fluorophore also has a characteristic quantum yield that is a measure of the emission efficiency of the fluorophore. The quantum yield, Q, of a fluorophore varies from 0 to 1 and is given by Q = emitted photons/absorbed photons. The phenomenon of fluorescence is the basis for exceptionally powerful imaging modalities. The advantages include high spatial and temporal resolution, high sensitivity, and quantifiability. In addition, fluorescence is extraordinarily sensitive to the environment surrounding the fluorophore in terms of pH, ion, and oxygen concentrations. A summary of the applications and limitations of each of the following methods is shown in Table 1. Fluorescence microscopy was developed in the early 1940 s. The deceptively simple idea is to deliver light of a specific frequency to the sample and then to separate the weak emitted light from the intense excitation light. The most common illumination system for fluorescence microscopy today is generally termed epi-illumination and a simplified diagram of a conventional epi-illumination microscope is shown in Fig. 4. In microscopes of this type, the excitation light is focused onto the specimen by the objective so the objective also serves as the condenser. Johan S. Ploem (born 1927) invented the epi-illumination system used in most fluorescence microscopes today. Ploem s design includes an excitation filter, a dichroic mirror or beamsplitter, and a barrier (or emission) filter, all gathered together in a small cube. This cube allows filter combinations to be changed by rotating a knob or moving a lever. In the context of this book it is significant that because of his contributions to the practice of microscopy, Ploem was elected as a fellow of the Papanicolaou Cancer Research Institute in 1977 and received the C. E. Alken Foundation award in 1982.

Optical Imaging of Cancer

7

Table 1 Overview of fluorescence methods and their applications Method

Applications

Caveats and limitations

Conventional widefield fluorescence microscopy

Localization of fluorophores within tissue, wide field of view with low magnification objectives. Global evaluation of, and image collection from, larger tissue samples. Identification of potential regions of interest to be used in higher resolution studies Localization of fluorophores within tissue. Collection of optical sections for 3D morphological reconstruction of structures. Suitable for FRAP, FRET, and FLIM studies with controlled intensity of laser excitation. Reasonable penetration of thick tissue. Limits of x–y axis resolution are 200–300 nM (depending on NA of objective) Collection of optical sections for 3D morphological reconstruction of structures. Suitable for FRAP, FRET, and FLIM studies. Decreased potential for damaging live cells. Good penetration of thick tissue. Limits of x–y-axis resolution are 200–300 nM (depending on NA of objective). Can also be used for uncaging or photoactivation Used for localization of protein(s) within cells or tissue by means of fluorescently labeled antibodies. Often used in fixed tissue, but labeling can be done in live cells

Photobleaching and limited resolution in x–y axis due to light scatter. Axial resolution limited. Limited penetration of thick tissue. Can cause photodamage to live cells

Confocal microscopy

Multiphoton microscopy

Fluorescence immunohistochemistry

Limit of z-axis resolution is usually in the range of 500 nM. Photobleaching of fluorophore and potential damage to live tissue with high intensity excitation light

Limits of z-axis resolution depend on the NA of the objective and in most stances are not better than 500 nm

Visualizing multiple proteins requires good separation of fluorophore absorption and emission spectra. Controls for antibody specificity are required. Vivid labeling does not prove that the antibody recognizes the protein of interest

8

K.T. Keyser and C.E. Strang Table 1 (continued) Method

Applications

FRAP

Potential damage to tissue Used to study the diffusion with high energy excitation properties of fluorescent light. The intensity of the molecules in living cells. excitation light must be Can provide information on calibrated to bleach the diffusion constant, mobile fluorophore without fraction, damaging the tissue binding/dissociation constants Often requires labeled protein Requires overlap (≥30%) between donor emission expressed by living cells or and acceptor absorption tissues. Can be used to spectra. The spectral determine conformation overlap of the fluorophore changes within a single requires multiple protein or interactions calibration images to rule between two proteins. For out false positives FRET interactions to occur, the fluorophores must usually be within 50 Å of one another Can be used to measure local Requires pulsed laser or lasers with variable intensity and ion or oxygen fast detectors concentration, pH, or other characteristics of the microenvironment. Can also be used to discriminate between the emissions of different fluorophores with similar spectral properties or to more easily quantify FRET interactions

FRET

FLIM

Caveats and limitations

Limits of Fluorescence Imaging: How Close Can Two Objects Be and Still Be Identifiable as Two Objects? The laws of diffraction are the limiting factor in the spatial resolution of a light microscope. As an example, consider the light emitted from a point source. In the objective, this point source is represented by the diffraction image of the point. The diffraction pattern that results from a uniformly illuminated circular aperture has a bright region in the center, known as the Airy disk (Fig. 5), which is named after George Biddell Airy, who served as the Astronomer Royal from 1835 to 1881. The central bright region, together with the accompanying concentric rings of diminishing brightness, is called the Airy pattern. The radius, r, of the Airy disc = 0. 61(λ/NA) with λ = the wavelength of the light and NA=numerical aperture of the objective. As two small objects are brought closer and closer together they can still be distinguished as two objects even though their images begin to overlap, as long

Optical Imaging of Cancer

9

Fig. 4 Fluorescence microscope. Fluorescence excitation is provided by a full-spectrum light source. An excitation filter is used to allow only selected wavelengths to pass. In this case, blue excitation light is passed, while other wavelengths are blocked. The excitation light is reflected down to the specimen, through the microscope objective, by a dichroic filter that reflects most of the blue light, but allows green light to pass through. Fluorescent molecules in the specimen absorb the blue light spectrum and, due to the Stokes shift, emit lower energy green light. The emitted green light passes through the dichroic filter. A final barrier filter blocks any remaining blue light so that only the green spectrum reaches the eyepiece

Fig. 5 Fluorescence is emitted from point sources in the specimen. In the microscope optics, the light from a source diffracts so that there is a central bright area surrounded by concentric rings (A). The central bright area is the Airy disc, the radius of which depends on the numerical aperture of the objective and the wavelength of light. As two objects move closer together, they are resolvable up to the point at which they are separated by r. This is the Rayleigh criterion and is the limit of resolution for fluorescence microscopes (B). As the objects move closer, at a separation of (0.78)r, they can no longer be resolved (C)

10

K.T. Keyser and C.E. Strang

as the distance separating them is equal to or greater than the radius of the Airy disk (Fig. 5). With this separation a small decrease in the intensity of the two diffraction images remains, which is referred to as the Rayleigh criterion. The Rayleigh criterion is named, confusingly enough, after John William Strutt, the third Baron Rayleigh who won a Nobel Prize in 1904 for his studies of gases and the discovery of argon. If the diffraction images of the objects overlap more than this (the distance separating them becomes less than the Rayleigh criterion) the two objects are no longer distinguishable as two objects. Depending on the wavelength of light that is used, conventional fluorescence microscopes are theoretically capable of resolving structures that are within about 0.2 μm of each other. However, the lateral resolution in many and perhaps even most imaging studies is usually between 400 and 600 nm. The failure to realize higher resolution in many studies is due to factors that include misalignment of the microscope, refractive index differences, optical aberration, and improper specimen preparation. An additional problem with conventional epi-fluorescence microscopes is that the entire specimen is exposed to the excitation light and the light that is emitted comes from throughout the specimen including regions above and below the focal plane. All of this light, including that from outside the focal plane, is collected by the microscope objective and the result is a loss of contrast and resolution. One way to overcome the loss of contrast and resolution is through computerbased image processing. Various computer software packages that rely on different algorithms can be used to generate an accurate description of the degradation of the image of a point source of light as it passes through the optical elements of a microscope. This description is called the point-spread function and the point-spread functions of each feature in a field of view make up the image. The point-spread function is unique for each microscope objective and must be determined for each objective used for image acquisition. Once determined, the point-spread function can be used to remove out-of-focus light from each image. This process is called deconvolution. However, high-fidelity image capture is also dependent on the sampling rate during image acquisition. Essentially all contemporary imaging systems use computerbased image acquisition and digitize the light collected from the specimen. That is, the analogue image that arrives at the detector after it passes through the optical path of the microscope is sampled and converted. How the information is sampled determines whether or not all of the information available in the image is collected. The Nyquist–Shannon sampling theorem establishes that when converting an analog signal to a digital signal, the sampling frequency must be such that the light from the smallest feature of interest should be sampled more than twice in order to be able to reconstruct the original perfectly from the sampled version. In other words, the sampling interval must be smaller than half of the desired resolution. If the sampling distance is larger than the Nyquist distance, information about the specimen is lost. This is called under-sampling, and in addition to the loss of information from the sample, under-sampling gives rise to aliasing artifacts that may show up as jagged edges or fringes. This is why cameras or other image acquisition

Optical Imaging of Cancer

11

devices with more megapixels, that is, a higher sampling frequency, yield higher fidelity images. Lastly, the resolution of the display monitor will affect the interpretation of the image. The value of a high NA objective can be compromised by data collection at low sampling rate or by viewing images on a computer monitor with low resolution, as in both of these cases data from a larger area are collapsed into a single pixel.

Confocal Laser Scanning Microscopy Confocal microscopes, which were introduced in the 1950 s, improve on conventional widefield fluorescence microscopes. The most common type of confocal microscope is the confocal laser scanning microscope (CLSM; Fig. 6). In these systems the excitation light from a laser is focused to a diffraction-limited spot at the

Fig. 6 Confocal microscope. In a point scanning confocal laser scanning microscope (CLSM), the excitation light is provided by a laser. Laser light sources provide excitation light of specific wavelengths, obviating the need for excitation filters. A pinhole aperture is used to set the amount of laser light that is reflected by the dichroic mirror onto the specimen. The laser moves line by line across the specimen in a given optical plane. Fluorescence emission passes the dichroic filter and is collected by a PMT connected to a computer which stores and integrates the photon counts from the PMT. A detector pinhole aperture is used to block emitted light that is not in focus. The focal plane can be changed, allowing multiple optical sections to be taken through the depth of the tissue

12

K.T. Keyser and C.E. Strang

focal point of the objective. This means that the intensity of the excitation light drops off dramatically above and below the plane of focus so that fluorophores in the focal plane are efficiently excited while those outside of the focal plane are not. The beam is swept back and forth in a raster pattern across the specimen by mirrors, typically driven by galvanometers or acousto-optical beam deflectors. In this way each point in the sample is illuminated. In addition, most laser scanning confocal microscopes position a small aperture, a pinhole, in the emitted light pathway. This arrangement means that only the fluorescence emitted from the focal volume of the objective is in focus and hence is able to pass through the aperture to the detector, normally a photomultiplier tube (PMT). The light originating from above or below the focal volume of the objective is not focused on the pinhole and does not pass through to the detector. Each image represents the light collected from a single optical section which, depending on the NA of the objective and other considerations, is rarely less than 600 nm thick. The microscope can then focus at a different level in the z-axis, that is to say at a different focal plane, and collect another image, and so forth. This yields a series of images in the z-axis (Fig. 7). These can be recombined and played back so that it appears that the observer is moving through the tissue, or the image “stack” can be rotated and the reconstructed image viewed in 3D. Alternatively, the

Fig. 7 Confocal optical sectioning. A retinal ganglion cell injected with fluorescent dye is shown as the maximum projection image (A) of stacked optical sections (B). Each optical section (C–F) is taken at a different depth along the z-axis and shows the dendritic structure at each depth. Scale bar = 80 μm

A

B

C

D

E

F

Optical Imaging of Cancer

13

stack of x–y images can be collapsed, yielding what is termed a maximum projection, which allows one to visualize the fluorescence in all of the optical sections at once.

Multiphoton Imaging As powerful as confocal imaging can be, it is still limited in its application to living cells and tissues by factors such as phototoxicity and photobleaching because the entire sample in the field of view is illuminated with the excitation light. Another approach that minimizes these problems is multiphoton excitation and imaging. Multiphoton absorption occurs when more than one photon (typically two or three) is absorbed within 10–18 s. The simultaneous absorption and transfer of energy from the two photons yield excitation equivalent to that resulting from absorption of a single shorter wavelength photon. For example, the absorption of two 800 nm photons provides the same excitation energy as one 400 nm photon. In order for this to occur, there must be a very high instantaneous intensity of excitation light and this is achieved through very short (100–150 fs) pulses from an ultrafast mode-locked laser. The excitation light is sufficiently intense to cause multiphoton absorption only at the geometric focus of the objective (Fig. 8). This is because the probability of two photon absorption outside the focal volume decreases as the fourth power of the distance along the z-axis. As a result, nearly all of the light emitted will come only from the fluorophores in the focal volume of the sample and image degradation caused by light emitted from outside the focal plane is eliminated. This also obviates the need for the pinhole that is used to reject out-of-focus photons in a conventional confocal microscope. In addition, the excitation light used in multiphoton imaging is typically long wavelength light and penetrates deeper into the sample. This is because fewer photons are absorbed by fluorophores out of the plane of focus and the longer wavelengths used for multiphoton excitation are scattered less by tissue constituents.

The Problem of the z-Axis As stated above, critical issues in fluorescence imaging relate to the size and thickness of the sample, the lateral (x–y) resolution of the microscope, the type of excitation, as well as the resolution of the collected image data and of the computer screen. The lateral resolution of the microscope will determine whether emissions from the two fluorophores will be captured in the same pixel if the proteins are side by side in reasonable proximity. This is often referred to as colocalization. More importantly, the z-axis resolution of the microscope will affect the interpretation of whether fluorescence originates from within, above, or below the focal plane. Determination of z-axis resolution is more complex than x–y resolution, but the resolution in the z-axis is always less than that in the x–y axis. For example, an oil immersion

14

K.T. Keyser and C.E. Strang

Fig. 8 Multiphoton microscope. In multiphoton microscopy, as in CLSM, a laser light source is used to excite fluorescent molecules in the specimen, and the resulting emissions are collected by a photomultiplier tube and integrated into an image via computer. However, the laser light used for confocal excitation can result in damage to tissue, so CLSM is not optimal for live cell or tissue imaging. Multiphoton microscopy allows for the use of lower energy excitation by using wavelength doubling. In this case, the laser yields conditions where two photons, each of which has 50% of the energy of the photon required for single photon excitation, are absorbed. These two photons together provide the same energy as the single photon. Efficient two-photon excitation and emission occurs only in the focal volume, so no pinholes are required

objective with a 1.3 NA will have an x–y resolution of perhaps 200 nm under nearly perfect conditions, while the axial, or z-axis, resolution will be no better than about 500 nm. This z-axis resolution may be sufficient to determine colocalization at a cellular level, but not at a subcellular level. The limitations of z-axis resolution are especially troubling in the case of conventional widefield fluorescent microscopy. For specimens that are less than 5 μm in thickness such as a monolayer of adherent cells or very thin tissue sections, colocalization analysis may be possible with a conventional fluorescence microscope. However, for thicker specimens it is difficult to determine whether or not two proteins are colocalized because widefield excitation light excites fluorophores at all depths simultaneously, and conventional fluorescence microscopes collect emission light from regions above and below the focal plane. Although the two fluorophores

Optical Imaging of Cancer

15

may appear to be in a single focal plane, they could in fact be above or below one another. For thicker specimens, confocal laser scanning microscopy offsets this problem to some extent. Images should be recorded as optical sections of limited axial dimension to determine if the fluorophores are indeed in the same x–y focal plane or whether they are vertically superimposed over each other along the microscope zaxis. The thickness of the optical section is determined by the numerical aperture of the objective and the use of a high NA lens can sometimes provide resolution sufficient to determine whether two or more proteins are expressed by a single cell.

Fluorescence Applications Immunohistochemistry and Immunofluorescence Fluorescence imaging is frequently used to study the expression pattern of proteins and their subcellular distribution. The detection of proteins in cells and tissues often relies on immunohistochemistry. This method relies on antibodies to detect proteins present in cells or tissues. Antibodies have binding sites for specific antigens and are produced by the body to attack foreign invaders such as bacteria, viruses, or pollen. Antibodies have distinct functional regions and are specific, binding tightly to a particular immunogenic region of the foreign object. The Fab (fragment antigen binding) region of the antibody is the recognition domain that binds to the antigenic region, or epitope, of the target protein. It has variable domains of heavy and light chains. The Fc region of the antibody is the effector region and mediates interactions with the other components of the immune system. In 1944 Albert Coons showed that a fluorescent molecule could be covalently bonded directly to the Fc region of an antibody made against a protein of interest. This binding does not affect antibody specificity so that labeled antibodies can be used to visualize the location and distribution of proteins of interest. The use of an antibody that is directly conjugated to a fluorescent molecule to detect proteins in cells or tissues is known as direct immunohistochemistry. In the case of indirect immunohistochemistry, the primary antibody is unlabeled. A secondary antibody, most commonly a fluorophore-labeled anti-isotype antibody, which recognizes the primary antibody, is used. Indirect immunohistochemistry has two major advantages over the direct method. First, there is nearly always a significant loss of antibody associated with the process of conjugating the fluorophore to the antibody, and antibodies may be available in limited quantities and are often expensive to purchase. Second, the indirect method offers higher sensitivity because multiple molecules of the fluorophore-bound secondary antibody can bind to each primary antibody molecule so that more light is emitted when the specimen is illuminated with light of the correct wavelength. Immunohistochemistry with two or more different fluorophore-bound antibodies is often used to determine whether the expression patterns of two or more proteins

16

K.T. Keyser and C.E. Strang

are similar or different. That is, are the proteins expressed by the same cells or tissue and what is the spatial relationship between the two proteins within a cell of interest? In experiments of this type, the emission from two or more fluorophores may overlap in the final image due to their close proximity within the specimen. This is called colocalization, and colocalization is operationally defined by whether the emissions from two or more fluorophores in close proximity will be captured within a single pixel. The availability of very specific fluorophores together with the optical sectioning capability of confocal and multiphoton microscopy has vastly improved our ability to detect colocalization in biological specimens. In contemporary fluorescence microscopy images are typically captured digitally and displayed on a computer monitor. In the case of green emitted light and red emitted light, mixing of green and red light in a computer monitor results in yellow, or shades of yellow–green through yellow–orange pixels. To address the question of whether or not two proteins are colocalized with conventional fluorescence microscopy, a number of issues need to be considered. First, the excitation and emission spectra of the fluorophores must not overlap. Fluorophores with narrow excitation and emission spectra are generally more suited to immunohistochemistry than are fluorophores with broad spectra. An example is shown in Fig. 9. Panels A, B, and C are single optical sections of tissue in which cells had been injected with Alexa 488 fluorescent dye (green; A), and subsequently labeled with an antibody against acetylcholine receptors (red; B). Panel C is an overlay image of the two labels and many of the cell bodies appear yellow–orange, suggesting colocalization.

Fig. 9 Colocalization. Panels A–C are single optical sections of neurons containing Alexa 488 fluorescent dye (green; A) labeled with an antibody against acetylcholine receptors (red; B) and an overlay image of the two labels (C). In these sections most of the cells containing Alexa 488 appear to express acetylcholine receptors. The images were taken with a 40×, 1.25 NA oil immersion objective. Scale bar = 40 μm

Fluorescence Recovery After Photobleaching (FRAP) This method was developed in the mid-1970 s in the laboratory of Watt W. Webb. Photobleaching refers to the reduction in fluorescence intensity (fading) with continued exposure to light of the correct wavelength to be absorbed by the fluorophore. For example, the commonly used fluorophore fluorescein can go through 30,000– 35,000 excitation–emission cycles before it loses the ability to fluoresce. Depending on a number of factors including, for example, the intensity of the excitation light and the concentration of molecular oxygen, this can take as little as 200–300 ms.

Optical Imaging of Cancer

17

This fluorescence does not recover. However, in many cases other unbleached fluorescently tagged molecules can move into the bleached region, and FRAP can be used to measure the ability of fluorescently tagged molecules to move over time. The recovery that occurs as fluorescent molecules move into the bleached region is monitored with low-intensity laser light (Fig. 10). Analysis of the fluorescence recovery can be used to determine important characteristics of a protein, including its diffusion constant, mobile fraction, transport rate, or binding/dissociation rate from other proteins. In practice, fluorescence in the region of interest is measured just before, during, and after photobleaching. A characteristic recovery curve showing the fluorescence intensity in the region of interest is shown in Fig. 10. Depending on the complexity of the interactions of the protein of interest with other molecules in the environment, a given curve may have differences in the slope, the plateau level, or it may display multiple plateaus and curves. If the curve is a single exponential, then the function Y = (F∞ – F0 )/(Fi – F0 ) can be used to calculate the mobile fraction (Y). In Fig. 10, Fi is the initial fluorescence level, t1/2 is the time that is necessary for the fluorescence to recover halfway between the fluorescence level after bleaching (F0 ) to the plateau level (F∞ ).

Fig. 10 Fluorescence recovery after photobleaching (FRAP). High-intensity light in the absorption spectrum of a given fluorescent dye will result in loss of fluorescence due to photobleaching. Low-intensity light can be used to measure the ability of fluorescently labeled molecules to move into the bleached region. Initial fluorescence (Fi ) in the region of interest is measured, fluorescence is photobleached (F0 ), and the region of interest is monitored until recovery reaches a plateau (F∞ ). In the simplest case, the recovery curve is a single exponential described by the function Y = (F∞ – F0 )/(Fi – F0 ). The t1/2 is the time that is necessary for the fluorescence to recover halfway under different conditions

Fluorescence Resonance Energy Transfer (FRET) More properly referred to as Förster resonance energy transfer after the German physical chemist Theodor Förster, FRET is a process whereby a fluorophore in the excited state can transfer its excitation energy to a neighboring fluorophore through dipole–dipole interactions without light emission. FRET is a powerful tool

18

K.T. Keyser and C.E. Strang

for the study of protein–protein interactions, protein–DNA interactions, and protein conformational changes. In order for FRET to occur, the emission spectrum of one fluorophore, the donor, must overlap the absorption spectrum of another fluorophore, the acceptor. The efficiency of energy transfer from one fluorophore to another is exquisitely dependent on the distance that separates the two fluorophores and varies as the inverse of the sixth power of the distance separating the two fluorophores. In practice, this requires that the distance separating the two be in the range of 10–50 Å(Å= 0.1 nm), although in rare instances FRET may occur when the separation is as much as 100 Å. When the specimen is exposed to excitation wavelengths appropriate for the donor and the two fluorophores are in close proximity, FRET is manifested by a decrease in the intensity of the donor emission (quenching) in the presence of the acceptor and increase (sensitization) in the emission intensity of the acceptor (Fig. 11).

Fig. 11 Förster resonance energy transfer (FRET). Two fluorophores, a donor and an acceptor, are required for FRET interactions. Laser (488 nm, light blue line) excitation of the donor (492 nm absorption peak) results in emitted energy with a peak at 520 nm (dotted green line). If the second fluorophore is within 50 Å of the donor, and there is spectral overlap (shaded region) between the emission spectra of the donor and the absorption spectra of the acceptor (bright green line, 575 nm), some of the energy is transferred non-radiatively to the acceptor. This results in the quenching of the donor fluorescence (solid green line, 520 nm) and sensitized emission by the acceptor (solid red line, 590 nm). Changes in the emission of the donor and acceptor allow the measurement of the separation of the fluorophores relative to one another

The most widely used donor and acceptor fluorophores for FRET studies are green fluorescent protein (GFP) and GFP variants, although other commercially available fluorophore pairs can be used successfully. The properties that must be considered in selecting fluorophores for FRET experiments include sufficient separation of the excitation spectra for selective excitation of the donor, sufficient overlap (≥30%) between the emission spectrum of the donor and the absorption spectrum of the acceptor to yield efficient energy transfer, and separation of the emission spectra between the donor and the acceptor sufficient to allow independent measurement of the fluorescence of each fluorophore.

Optical Imaging of Cancer

19

FRET Can Be Used to Determine the Distance Between Two Molecules In order to monitor interactions between two molecules, one of them is labeled with the donor fluorophore and the other with the acceptor fluorophore. When the molecules of interest are not in close proximity, excitation with light appropriate for the absorption spectrum of the donor yields the emission characteristic of the donor. When the two molecules are in close proximity (usually 10–50 Å), FRET occurs and under this condition the donor emission intensity is decreased and the acceptor emission is increased. Protein conformational changes can be detected as well. In this case the target protein is labeled with the donor and the acceptor at two different loci. If the conformation of the protein changes such that the distance or relative orientation of the donor and acceptor is affected, FRET may be observed. If a molecular interaction or a protein conformational change is dependent on ligand binding, the FRET technique is applicable for ligand detection.

Fluorescence Lifetime Imaging Microscopy (FLIM) As shown in Fig. 2, light absorption results in a fluorescent molecule moving from the ground state to an excited state, followed by thermal decay and relaxation to its lowest excited singlet state. The average time that a molecule remains in the excited singlet state prior to photon emission is the lifetime of that fluorophore. In fluorescence lifetime imaging microscopy (FLIM), the lifetime of the fluorescence, not its intensity, is used to create an image. Fluorescence lifetime imaging is an extraordinarily sensitive imaging method because the lifetime of a fluorophore can be altered in response to changes in the conformational state of the fluorophore or in response to interactions with the local environment. For example, interactions with ions or oxygen in the local environment can cause fluorophores to release the excited state energy faster, resulting in a decreased lifetime. Thus, by measuring changes in decay times for a single fluorophore, FLIM can be used to indirectly measure physiological changes in ion concentration, oxygen concentration, pH (the lifetime of a protonated fluorophore typically differs from that of the unprotonated fluorophore), or other changes in the microenvironment surrounding the fluorophore. Measurement of lifetime, rather than fluorescence intensity, means that FLIM is not dependent on local intensity variations, such as those observed when imaging thick tissue samples. Fluorescence lifetime should not be confused with fluorophore bleaching. Bleaching renders a fluorophore unable to move into the excited state, while the lifetime refers to the amount of time that the molecule remains in the excited state. Each fluorescent molecule has a characteristic lifetime, but the lifetime is affected by constituents of the local microenvironment. In a uniform environment, the decay of the fluorescent lifetime of a fluorophore usually can be described by a single exponential function. In more complex situations, such as those found in cells or tissues, multiple solvent environments exist

20

K.T. Keyser and C.E. Strang

and the decay of the fluorescent lifetime of a fluorophore is more often multiexponential. The formula [S1] = [S1]0 e–t describes the single exponential kinetics of fluorescence decay in a uniform environment. [S1] is the concentration of excited state molecules at a given time (t), [S1]0 is the initial concentration and  is the decay rate. The decay rate, also known as quenching, is the inverse of fluorescence lifetime. For commonly used fluorescent compounds, typical excited state decay times are within the range of 0.5–20 ns. Because the number of molecules in the excited state is affected by both radiative processes such as photon emission or heat transfer and non-radiative processes such as the energy transfer from a FRET donor to a FRET acceptor, the total decay rate is the sum of the decay of both types of processes,  tot =  rad +  nrad . If either the radiative or non-radiative decays are fast, the lifetime for that fluorophore is short. Lifetimes are calculated for each fluorophore and the pixels representing each are pseudocolored to create contrast or lifetime maps. FLIM can be used to discriminate between different fluorophores with similar absorption and emission spectra because each fluorophore has a characteristic lifetime. This is useful in cases in which there is a high level of autofluorescence in the sample being imaged or for quantification of FRET interactions. The possibility for direct excitation of the acceptor is a potentially confounding factor in quantification of FRET interactions. Multiple reference images or an alternate method of discriminating between the emissions of overlapping fluorophores, such as FLIM, are required.

Fluorescence Lifetime Measurement Fluorescence lifetimes are calculated using either the time domain method or the frequency domain method. The time domain method requires fast excitation pulses and fast detection circuits. Very brief pulses of light are used to sequentially illuminate each point in the sample and emissions are collected in discrete segments of time measured after the excitation flashes. The relative intensity as measured by photon counts in each of the time segments is used to generate a histogram of photon counts which are then used to calculate the lifetime (Fig. 12). The peak photon count of a fluorophore with a short lifetime would fall into the earlier time segment, while the peak photon count of the fluorophore with a longer lifetime would be collected in a later time segment. The frequency domain method requires the use of a light source, usually a laser, with circuits that allow the intensity to be modulated. An image intensifier such as a CCD camera is also required to collect the emissions. The wavelength of the excitation light is held constant, but the intensity is modulated with a frequency of 10–100 MHz. This results in modulation of the intensity of the fluorescence emission. However, due to the decay constant of the fluorescent molecule and the modulation frequency, the emitted light will be phase-shifted and will display less intensity modulation (Fig. 13). That is, because of the temporal characteristics of

Optical Imaging of Cancer

21

Fig. 12 Fluorescence lifetime imaging (FLIM), time domain method. In the time domain method, each point in the sample is excited sequentially by short laser pulses. The emitted fluorescence from each point is used to generate a histogram of photon counts over many time epochs. The differing decay characteristics that are extracted from the photon counts are used to distinguish the emissions of two fluorophores

the excitation–emission cycle of the fluorophore, the emission will be delayed in time compared to the excitation light, and the emission intensity will be less modulated than the excitation light (Fig. 13). Lifetime is calculated from the amount of phase-shift and the amount of modulation of emission intensity. In general, a fluorophore with a short fluorescence lifetime will be less phase-shifted and the emission

Fig. 13 FLIM frequency domain method. Frequency domain lifetime measurement uses an excitation light that is modulated at frequencies of 10–100 MHz (blue line). Modulation of the excitation light results in modulation of the intensity of the emitted light. The emitted light is phase-shifted and is less modulated. Because the amount of phase-shift and the amount of decrease in modulation are dependent on the decay constants of the fluorophore and the modulation frequency, fluorophores with similar emission spectra but shorter (dotted green line) or longer (solid green line) emission lifetimes can be distinguished from one another

22

K.T. Keyser and C.E. Strang

intensity modulation will more closely follow the excitation light modulation, while fluorophores with longer lifetimes will display greater phase-shifts and less modulation. If a reference of known lifetime is used, the amount of phase change and modulation of the sample fluorophore can also be calculated with respect to the standard.

Controls and Image Processing Fluorescence microscopy in all of its different guises is a powerful and flexible tool. Understanding of the principles, limitations, and strengths of each imaging modality will help in the selection of the appropriate method for addressing specific questions in biomedical research. As with any method, the data derived from fluorescence imaging studies are meaningless without proper control experiments. A simple example is that of controls for immunohistochemical studies. Often, investigators run controls in which the primary antibody is omitted, without realizing that this is only a control for secondary antibody specificity. Appropriate controls for the primary antibody include the use of preimmune or normal immunoglobulin with the protein concentration matched to that of the primary antibody, followed by secondary antibody processing. In addition, it is important that investigators, their laboratory staff, and students understand the limits of acceptable post-acquisition processing and the critically important issue of reporting how images were acquired and processed in publications. A simple example might be that of a 2-color experiment and subsequent brightness and contrast processing. In most situations both channels should be adjusted equally, and if other operations are carried out such as filtering or dynamic range adjustment, these manipulations should be described in the methods section. In addition, information including the make and model of imaging system, the type, magnification, and numerical aperture of objective lenses, the fluorophores, the image acquisition software, and processing software should be recorded for each image. Many journals, such as the American Journal of Physiology, now have specific requirements for authors who include digital image files as part of a manuscript submitted for publication.

The Future The diffraction-limited resolution limit for fluorescence imaging is under assault from various forms of “super-resolution” microscopy. Examples include photoactivated localization microscopy (PALM) and stimulated emission depletion microscopy (STED), both of which offer lateral resolution of 50 nm or better for fixed samples. To put this number into context, an individual ribosome is about 30 nm across. Furthermore, recent publications (e.g., Shroff et al., 2008) describe modifications of PALM, coupled with the use of a photostable photoactivatable

Optical Imaging of Cancer

23

probe, to allow super resolution imaging of live cells. While these imaging systems are not widely available currently, they have the potential to have an enormous impact on biomedical research, and cancer research in particular, in the very near future.

References and Recommended Readings Becker, W. and Bergman, A. (2003) Lifetime imaging for optical microscopy, http://www.beckerhickl.de/abstracts/Lifetime%20Imaging%20Techniques%20for%20Optical%20 Microscopy.htm Breusegem, S.Y., Levi, M., Barry, N.P. (2006) Fluorescence correlation and fluorescence lifetime imaging microscopy. Nephron Experimental Nephrology, 103:e41–49. Denk W. and Svoboda K., (1997) Photon upmanship: why multiphoton imaging is more than a gimmick. Neuron 18:351–357. Gerritsen, H.C., Ruonala, M.O., van den Heuval, D.J., van Bergen en Henegouwan, P.M.P. (2004) Fluorescence imaging of high affinity EGF receptor location, www.biophysics.org/ discussions/2004/gerritsen.pdf Herman, B. (1998) Fluorescence Microscopy, Second Edition, Springer-Verlag, New York. Michalet, X., Kapanidis, A.N., Laurence, T., Pinaud, F., Doose, S., Pflughoefft, M., Weiss, S. (2003) The power and prospects of fluorescence microscopies and spectroscopies. Annual Review of Biophysics and Biomolecular Structure 32:161–182, doi:10.1146/annurev. biophys.32.110601.142525. Pawley, J.B. (Ed.) (2006) The Handbook of Biological Confocal Microscopy, Third edition, Springer Science + Business Media, LLC, New York. Shroff, H., et al. (2008) Live-cell photoactivated localization microscopy of nanoscale adhesion dynamics. Nature Methods 5:5, doi:10.1038/NMETH.1202 417-423. Sir George Gabriel Stokes. (1852) On the optical properties of a recently discovered salt of Quinine. British Association Reports, pp. 15–16. So, P.T. C., Dong, C.Y., Masters, B.R., Berland, K.M. (2000) Two-photon excitation fluorescence microscopy. Annual Review of Biomedical Engineering 2:399–429, doi:10.1146/annurev. bioeng.2.1.399. Stubbs, C.D., et al. (2005) The use of time-resolved fluorescence imaging in the study of protein kinase C localisation in cells. BMC Cell Biology 6:22, doi:10.1186/1471-2121-6-22. Wang, J., Shyy, J. Y-J., Chien S. (2008) Fluorescence proteins, live-cell imaging, and mechanobiology: seeing is believing. Annual Review of Biomedical Engineering 10, doi:10.1146/annurev. bioeng.010308.161731.

Review Articles Michalet, X., Kapanidis, A.N., Laurence, T., Pinaud, F., Doose, S., Pflughoefft, M., Weiss, S. (2003) The power and prospects of fluorescence microscopies and spectroscopies. Annual Review of Biophysics and Biomolecular Structure 32:161–182, doi:10.1146/annurev. biophys.32.110601.142525. Wang, J., Shyy, J. Y-J., Chien S. (2008) Fluorescence proteins, live-cell imaging, and mechanobiology: seeing is believing. Annual Review of Biomedical Engineering 10, doi:10.1146/annurev. bioeng.010308.161731. So, P.T. C., Dong, C.Y., Masters, B.R., Berland, K.M. (2000) Two-photon excitation fluorescence microscopy. Annual Review of Biomedical Engineering 2:399–429, doi:10.1146/annurev. bioeng.2.1.399.

24

K.T. Keyser and C.E. Strang

Useful Webpages Please note that the following web sites contain useful information, but inclusion on this list does not imply product endorsement. American Journal of Physiology http://www.the-aps.org/publications/i4a/figures/fig_manip_2007.htm http://www.lambert-instruments.com/technologies/1_english/4_technologies/1_flim,_ fluorescence_lifetime_imaging_microscopy http://www.olympusconfocal.com/applications/flimintro.html http://www.olympusfluoview.com/applications/flipandfrap.html http://www.jobinyvon.com/SiteResources/Data/MediaArchive/files/Fluorescence/applications/ F-10.pdf http://las.perkinelmer.com/Content/ApplicationNotes/APP_Fluorescencerecovery.pdf

Endoscopic Techniques for Optical Imaging E. Namati, M.J. Suter, and G. McLennan

Introduction Lung cancer is the leading cause of cancer death in the industrialized world (Landis et al., 1998). Despite recent advancements to reduce the mortality associated with this disease, patient prognosis remains poor, with the current 5-year survival rate under 16%, a rate that has seen no dramatic change over the past 30 years (Jemal et al., 2007; Proctor 2001). It is set to become a worldwide epidemic with the World Health Organization estimating 10 million deaths per year worldwide by the year 2030 (Proctor 2001). While fatal in most cases, early stage lung cancer can be cured. At the time of presentation, less than 15% of patients have localized disease that may be amenable to surgical resection and potential cure; of these patients, the 5-year survival is still a low 60–70%, indicating that even earlier diagnosis is important.

Multidetector Computed Tomography We now have the ability to detect early, small lung cancer through multidetector computed tomography (MDCT) scanning, presenting as lung nodules less than 10 mm diameter (and often less than 5 mm diameter), but 60% of smokers have at least one such lung nodule, and less than 1% are cancer. Current recommendations are to follow these lung nodules over time to assess growth – thereby potentially missing the opportunity for cure. Hence a recently introduced term known as the ‘lung cancer paradox’ – early detection is possible, but early diagnosis is not. In addition, while MDCT scanning can now detect small peripheral lesions, central lesions are still often radiographically occult in the early stages of disease progression. Clearly, new methods for evaluating suspect areas in the human airways

G. McLennan (B) Department of Internal Medicine, Carver College of Medicine, University of Iowa, Iowa City, IA, USA e-mail: [email protected]

E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_2, 

25

26

E. Namati et al.

at a cellular and molecular level are needed and would provide important information for early diagnosis. The ability to characterize the cellular and molecular compositions of a tumor in vivo will allow diagnosis of the tissue, but also of equal importance, track the progress of proceeding therapies. This feedback mechanism will provide faster evaluation of therapeutic techniques and enable cellular characterization of tumors, which is also currently lacking. Eventually this may lead to a better understanding of lung cancer and in particular its path at an early stage.

Conventional Pulmonary Imaging Conventional endoscopic imaging or macro-optical imaging techniques are routinely employed in pulmonary disease management. The most common is the white light color or monochromatic bronchoscope; here a flexible fiberscope or chargecoupled device (CCD) endoscope is utilized to image the inside of the airway lumen with a white light illumination source. Assessment of the airway is performed based on subjective color, texture, and structural visualization of the reflectance image obtained through the bronchoscope. Many common lung pathologies can be distinguished using this form of bronchoscopy through assessment of such alterations as airway wall thickening, mucosal topography, mucosal color changes, and vascular changes. Bronchoscopy also provides a means for biopsy sampling using forceps biopsy, brush cytology, or needle aspiration through an auxiliary channel.

Autofluorescent Bronchoscopy Autofluorescent bronchoscopy, a form of bronchoscopy where endogenous fluorescence of the airway and tissues are visualized either individually or simultaneously alongside white light images, has recently become available. This form of bronchoscopy reveals autofluorescence principally from connective tissue such as collagen, but also from cellular chromophores such as flavins, NADH, and porphyrins (Richards-Kortum and Sevick-Muraca 1996; Feller-Kopman et al., 2005). As the bronchial mucosa becomes abnormal with cell proliferation, it becomes thicker, and it has been suggested that the autofluorescence signal becomes less detectable. The combined use of white light and autofluorescence bronchoscopy has been shown to increase the overall sensitivity of identifying lung pathologies in the airways when compared to white light bronchoscopy alone (Beamis et al., 2004; Feller-Kopman et al., 2005; Herth et al., 2006; Lam et al., 1993; 2000; Haussinger et al., 2005), although this increase came at the cost of specificity (Kennedy et al., 2001). The utility of autofluorescence bronchoscopy is increasingly seen as an important tool for the management of patients with early lung cancer.

Endoscopic Techniques for Optical Imaging

27

Recently, several other bronchoscopy techniques have also emerged, including computer-aided color and textural analysis of the airway lumen, fluorescence microvascular imaging using exogenous sodium fluorescein (Suter et al., 2005b, c, 2004b, 2008b), and narrow-band imaging (Herth et al., 2006).

Micro-optical Imaging Techniques Micro-optical imaging techniques have also generated interest from within the pulmonary field, for evaluation of lung pathology in vivo (McWilliams et al., 2002; Sutedja 2003). Micro-optical imaging devices that seem to have promise in diagnosing lung cancer based on their ability to visualize cellular and molecular markers and have the potential for endoscopic miniaturization for use within the airways are confocal microscopy (Aziz and Gmitro 1993; Delaney et al., 1993; Namati et al., 2008b; Sung et al., 2002; Kiesslich and Neurath 2005; Rouse et al., 2004; Flusberg et al., 2005a), optical coherence tomography (Fujimoto et al., 1995; 2000; Yang et al., 2004; Yun et al., 2006), various types of spectroscopy (Weissleder and Pittet 2008; Bard et al., 2005), and fluorescence or luminescence detection systems that rely on an administered compound.

Confocal Endo-microscopy Confocal endo-microscopy systems are now commercially available for both the gastro-intestinal and pulmonary systems, with promising early results. Development of these techniques, including technical advancements in the imaging hardware and development of fluorescent biomarker are currently ongoing. These endeavors may ultimately enable the diagnosis of suspect lesions in vivo.

Combination Strategies With early detection of suspicious nodules now possible with MDCT scanning, utilizing the three-dimensional (3D) data inherent to MDCT scans to aid bronchoscopy procedures has become possible. Several research and now commercial systems exist for identification of suspect lesions, visualization of the airways in three dimensions (virtual bronchoscopy), creation of a path to the suspect lesion (Kiraly et al., 2004; Ferguson and McLennan 2005; Tschirren et al., 2005; Negahdar et al., 2006; Baker et al., 2007; McLennan et al., 2007a; b) electromagnetic tracking of the bronchoscope for direct feedback of location with respect to the three-dimensional virtual airway image (Schwarz et al., 2003; Gildea et al., 2006; Schwarz et al., 2006; Eberhardt et al., 2007a, b; Seijo et al., 2007; Makris and Gourgoulianis 2008), and electromagnetic active guidance strategies (Riker et al., 2007).

28

E. Namati et al.

It is envisioned that a combination of MDCT and macro-optical imaging systems will be used for localization of suspect lesions through virtual pathway finding, electromagnetic navigation, and electromagnetic active guidance, while micro-optical imaging systems will be utilized for tissue analysis and possible diagnosis.

Conventional Optical (Macro-optical) Imaging Rigid bronchoscopy was first reported in 1897 by Gustav Killian (1898), where a rigid esophagoscope was used to remove a foreign body from the right main stem bronchus. Over the next century bronchoscopy continually advanced and slowly became adopted until the late 1960s where Ikeda et al. pioneered the use of the flexible fiber-optic bronchoscope opening an array of new possibilities (Ikeda et al., 1968). Flexible white light fiber-optic and digital CCD bronchoscopes are now a common and important tool for interventional pulmonology. These systems provide the ability to assess the central airways and enable tissue sampling through forceps biopsy, brush cytology, or needle aspiration. In addition, laser therapy, cryotherapy, electrocautery, and stenting are now possible and routinely performed using bronchoscopy (Herth et al., 2006). Bronchoscopy systems can be either fiber-optic or CCD based, monochromatic or color, but in any case they are referred to as white light bronchoscopy (WLB), as the illumination source is white. WLB systems provide reflectance-based imaging of the airways, as light reflected from the surface of the lumen is detected and viewed either through the bronchoscope or on a video monitor.

Autofluorescence Bronchoscopy In contrast to white light bronchoscopy, fluorescent or autofluorescent bronchoscopy, a developing form of bronchoscopy since the early 1990’s, detects fluorescent emission of tissue using a high intensity light source. These systems generally excite the tissue within the 400–490 nm spectral range and detect light in the greater than 500 nm range. Several commercial autofluorescent bronchoscopy systems have been developed (LIFE, SAFE-1000, D-light) and their utility has been investigated. Autofluorescence in the airways as discussed already is a weak signal produced by chromophores (Feller-Kopman et al., 2005). The detection of its presence and/or absence can be difficult, and it is not surprising that there is observer and technology-induced variance in the assessment of the test results. The color changes reported by the bronchoscopist are false colors indicating a presence and/or absence of the autofluorescence signal likely due to the mucosal thickness and have no other biological specificity for cancer. These tests are not standardized by any measurable output such as fluorescence intensity or quantifiable spectral change. Therefore, they are not particularly reproducible, resulting in limited adoption of this technology. From several different autofluorescence studies, an increase in the sensitivity of detecting pre-invasive bronchial lesions with respect to white light bronchoscopy was found (Feller-Kopman et al., 2005; Haussinger et al., 2005; Herth et al., 2006;

Endoscopic Techniques for Optical Imaging

29

Lam et al., 1993; 2000). However, autofluorescence bronchoscopy of pre-invasive lesions is still problematic as the specificity is very low, and up to half of the abnormal fluorescence regions are false-positives (Feller-Kopman et al., 2005; Haussinger et al., 2005). In light of these results, further controlled clinical evaluation is necessary including development of techniques to differentiate between pre-invasive lesions and bronchitis, a major obstacle in autofluorescence bronchoscopy.

Narrrow-Band Imaging An alternative bronchoscopy technique that has recently become available is narrow-band imaging (NBI). NBI proposes to increase the contrast of blood vessels by filtering out the illumination source with the exception of two narrow bands centered at 415 and 540 nm. Incidentally these are the peak absorption spectra of oxyhemoglobin, resulting in pronounced blood vessel contrast in the narrow-band images. It is known that dysplastic airway lesions have abnormal capillary formation, both from gross observation and histopathology. It is proposed that with the increase in vessel contrast obtained in NBI, such abnormalities are more apparent than when using white light bronchoscopy alone. Recent studies using NBI have shown that it is more specific for detecting dysplasia than white light bronchoscopy alone and in addition, more specific than white light bronchoscopy and autofluorescent bronchoscopy combined (Herth et al., 2006). However, these studies carry several limitations including observer bias and uncontrolled patient history and management (Vincent et al., 2007).

Qualitative Color Analysis White light bronchoscopy provides a visual link to bronchial mucosa topography and color; however, the subtle visual changes that indicate early stages in disease development may often be missed as a result of this highly subjective assessment. This is particularly apparent by the inexperienced bronchoscopist. The sensitivity of white light bronchoscopy for the detection of class III lesions has been reported to be as low as 10.6% with a corresponding specificity of 72.7% (Ernst et al., 2005). Recently, research focused on increasing the diagnostic yield of white light bronchoscopy while reducing user subjectivity has been performed using quantitative color analysis of the acquired white light bronchoscopy images (Gopalakrishnan 2003; Suter 2005; Suter et al., 2005d, 2004a, b). While the concept of color analysis of biological tissues is not new and has been used for a number of years in the field of dermatology, the utility in the pulmonary airways was limited until the introduction of color CCD chip bronchoscope (Aleva et al., 1998; Knyrim et al., 1987). A color analysis bronchoscope system with the primary objective of detecting color mucosal abnormalities was recently developed (Gopalakrishnan 2003; Suter 2005; Suter et al., 2005d, 2004a, b). This system was designed to provide real-time feedback to the treating physician by highlighting abnormal regions of interest on the WLB images adjacent to the live digital bronchoscope feed. Abnormal mucosal

30

E. Namati et al.

regions of interest were identified based on comparative color analysis between the live digital bronchoscope feed and a library of mucosal colors generated from healthy non-smoking volunteers. Ideally this form of assessment will serve to guide the treating physician during the clinical bronchoscopy exam to regions of interest that may be consequently investigated further through optical or conventional biopsy. Figure 1 depicts an example of a WLB image that has undergone quantitative color analysis. Regions of interest identified as being outside the determined range of healthy mucosal colors have been highlighted (Suter 2005).

Fig. 1 This figure demonstrates the result of classifying the pulmonary mucosal colors based on comparison with a developed healthy database. (a) The original image of a pulmonary airway with known papillomatosis. (b) The result of comparing the hue and saturation values in the image to the developed normative database. This image was classified correctly as abnormal (Suter 2005)

In addition to the intrinsic utility as a biopsy guidance tool, quantitative color analysis of the bronchial mucosa may prove useful as a non-invasive diagnostic tool. A preliminary study on the use of color analysis together with automated neural network classification to categorize pulmonary pathology demonstrated that it was possible to accurately distinguish between healthy mucosa, carcinomas, granulation tissue, papillomatosis, and airway mucosa changes associated with idiopathic stenosis of the airways (Suter 2005).

Fluorescein Bronchoscopy Fluorescein angiography is commonly used to qualitatively evaluate the circulation of the retina in order to detect and diagnose diseases including diabetes (Bjarnhall et al., 2002; Browning 1999). The process of fluorescein angiography involves injecting a bolus of fluorescein into the patient’s circulatory system and observing

Endoscopic Techniques for Optical Imaging

31

the time-based response in the retina (Flower 1973). Fluorescein is a highly fluorescent chemical compound that spontaneously fluoresces upon excitation with blue light. A large portion of the fluorescein injected into the blood stream binds to serum protein; however, it is the unbound fluorescein molecules that are responsible for the observed light emission (Boguta and Wrobel 2001). Recently bronchoscope-based fluorescence detection systems have been used in conjunction with fluorescein to assess the bronchial mucosal microvasculature (Suter et al., 2005a; Suter 2005). A 488 nm excitation source is delivered to the bronchial mucosa through a diffusing transmission fiber, and the fluorescein emission is detected by a fiber-optic bronchoscope connected to a CCD camera. A preliminary study conducted on consenting volunteers from both a healthy non-smoking population and from patients with a significant smoking history revealed a statistically significant (p-value < 0.05) difference in the detected fluorescein emission (Suter 2005). The smoking population was found to have reduced fluorescein emission intensity when compared to the non-smoking population. While it is unclear whether the decrease in detected fluorescein emission was a direct result in a change in the bronchial mucosal microvascularity, a result of airway wall thickening, or some other mucosal alteration, future research in the assessment of the detectable superficial bronchial microvasculature may work toward enhancing our current knowledge and understanding of the airway microvasculature and its role in tissue remodeling in response to injury or disease. Figure 2 (a) and (b) represents an example of a white light and fluorescein image acquired from a smoking individual with a significant obstructive cancer mass. In this example it is clear that the detected fluorescence emission is reduced over the identified mass.

Fig. 2 Image of an airway with a significant obstructive mass (top right quadrant) from a smoking individual with signs of chronic bronchitis. (a) White light illumination bronchoscopy image and (b) a detected fluorescence emission image of the airway shown in (a) during laser excitation. The fluorescence image is pseudocolored with blue to red representing low to high intensity

32

E. Namati et al.

Micro-optical Imaging Micro-optical imaging techniques have recently generated great interest in the medical field, for their potential to translate into clinical in vivo biopsy systems (McWilliams et al., 2002; Sutedja 2003). Promising techniques include optical coherent tomography (OCT) (Fujimoto et al., 1995; 2000; Yang et al., 2004; Yun et al., 2006), fiber-optic spectroscopy (Weissleder and Pittet 2008; Bard et al., 2005), confocal microscopy (Aziz and Gmitro 1993; Delaney et al., 1993; Namati et al., 2008; Sung et al., 2002; Kiesslich and Neurath 2005; Rouse et al., 2004; Flusberg et al., 2005a), and two-photon microscopy (Jung and Schnitzer 2003; Kim et al., 2008; Gobel et al., 2004; Bird and Gu 2002).

Optical Coherence Tomography Optical coherence tomography (OCT) is a non-contact optical imaging modality that provides tomographic images of tissue at resolutions comparable with architectural histology (<10 μm) (Bouma et al., 2000; Sergeev et al., 1997; Sivak et al., 2000; Tearney et al., 1997; Hsiung et al., 2005; Chen et al., 2007). The underlying concept of OCT is parallel to that of ultrasound, where measuring the delay of the source, as it is reflected off subsurface structures in biological tissues, generates depth information. Unlike ultrasound, however, a broadband light source is used in OCT, and due to the high speed of light propagation in tissue, optical reflectance is measured using low-coherence interferometry. Several preliminary ex vivo studies have been conducted in recent years on the use of optical coherence tomography for the assessment of bronchial mucosa (Yang et al., 2004; Ikeda et al., 2007; Tsuboi et al., 2005; Whiteman et al., 2006). These studies have demonstrated that optical coherence tomography can indeed be used to visualize and evaluate the pulmonary tissue demonstrating good correspondence with histopathological sections. Endoscopic optical coherence tomography has also been used to interrogate the bronchial mucosa in a limited in vivo human proof of principle study (Tsuboi et al., 2005). Recently a large 138-patient in vivo study was conducted demonstrating that epithelial thickness measurements of invasive carcinoma, carcinoma in situ, and dysplasia could be used to differentiate between metaplasia and hyperplasia (Lam et al., 2008). To date these in vivo studies have generally been limited to point sampling techniques due to the inherent limitations of the optical coherence tomography imaging technique. As lung cancer and its precursors are often multifocal and can arise anywhere within the airway tree (Kerr 2001), a diagnostic tool for evaluating this disease must be able to comprehensively investigate large areas in multiple bronchial segments during a clinical viable procedure time (1–5 min). Recently, a second-generation optical coherence tomography technology, termed optical frequency domain imaging (OFDI) or swept source optical coherence tomography, has been developed (Yun et al., 2003). While optical coherence tomography uses a broadband light source and

Endoscopic Techniques for Optical Imaging

33

detects the optical reflectance using low coherence interferometry, OFDI is based on frequency domain interferometry and utilizes a rapidly tuned wavelength swept source (Brinkmeyer and Ulrich 1990; Chinn et al., 1997; Golubovic et al., 1997; Lexer et al., 1997; Yun et al., 2003; 2003). The key advantage of OFDI over optical coherence tomography is that it provides images at rates that are 100× faster than conventional optical coherence tomography. Therefore OFDI, along with appropriate catheter designs, can be utilized to systematically image or screen the bronchial tree in a manner compatible with the temporal requirements of the bronchoscopy procedure. Figures 3 and 4 illustrate cross-sectional and volume-rendered images obtained in a recent study from ex vivo swine lungs in the central airways (Suter et al., 2008a). Layers of the bronchial wall including the epithelia, lamina propria, smooth muscle, perichondrium, and cartilage are clearly distinguishable.

Fig. 3 (a) Transverse and (b) longitudinal cross-sectional OFDI images of the bronchial wall with the corresponding histology (c). E, epithelia; LP, lamina propria; SM, smooth muscle; P, perichondrium; and C, cartilage

Confocal Endo-microscopy Confocal endo-microscopy, defined as confocal microscopy designed for endoscopic applications, has in particular gained attention for their miniaturization and high-resolution imaging potential. Their ability to non-invasively resolve cellular structures down to the micron level provides a high sensitivity and specificity

34

E. Namati et al.

Fig. 4 This figure shows a volume rendering of the comprehensive OFDI data set depicted in Fig. 3. Rings of cartilage are clearly identifiable in the three-dimensional rendering

for assessment of disease states in vivo. This has led to the development and commercialization of such systems. Confocal microscopy requires the scanning of a focused point source within the medium of interest, in this case a suspicious region inside the human body, with a high-powered light source and detection of fluorescent emission or reflectance via a sensitive photodetector. An image is obtained by reconstructing the rasterscanned detected light onto a grid (Minsky 1961, 1988). An array of confocal endomicroscopy systems currently exist. They generally fall under two groups: one that performs the laser scanning at the distal tip and a second that performs the scanning at the proximal tip (Flusberg et al., 2005a). The most common technique includes scanning a flexible fiber bundle, containing thousands of coherently placed fibers, at the proximal end that relays the laser scan to the distal tip (Aziz and Gmitro 1993, 1994; Collier et al., 2002; Drezek et al., 2000; MacAulay et al., 2004; Rouse and Gmitro 2000; Rouse et al., 2004; Sabharwal et al., 1999; Carlson et al., 2007a, b; Collier et al., 2007; Maitland et al., 2008; Smithpeter et al., 1998). This is technically the simplest approach; however, the resolution and field of view are ultimately limited by the fiber-to-fiber distance and number of fibers in the bundle. An increase in the resolution can be obtained by mounting an objective lens to the distal tip; however, any increase in resolution also results in a decrease in the field of view (Carlson et al., 2005; Chidley et al., 2006; Sung et al., 2002; Rouse et al., 2004; Jung and Schnitzer 2003). Several techniques have been proposed and built based on the second approach where scanning is performed at the distal tip. Most of these techniques utilize one or two single fibers for transmission and detection of light. Many different scanning apparatus mounted to the tip of the endoscope have been proposed and presented including a piezoelectric scanner (Rector et al., 2003; Myaing et al., 2006), a Microelectromechanical Systems (MEMS) single (Flusberg et al., 2005b; Dickensheets and Kino 1996) or dual axis scanner (Liu et al., 2006; 2007; Wang et al., 2003a,

Endoscopic Techniques for Optical Imaging

35

b, 2004), or a resonant tuning fork scanner (Delaney et al., 1993; 1994; King and Delaney 1994; Polglase et al., 2006). In either case, the resolution and field of view of the laser scanning can be adjusted or designed with varying specifications, and is not limited to the constraints discussed with the fiber bundle approach. Also, these techniques only require one or two single fibers to the distal tip, enabling highly flexible probe designs. Confocal endo-microscopy systems are now commercially available for the gastro-intestinal (GI) tract, skin, and pulmonary systems (CellVizio Inc, Optiscan Pty. Ltd., Lucid Inc.). Recently an increasing number of research studies have been undertaken using confocal endo-microscopy systems in the GI tract with some early studies in the cervical and pulmonary regions. The initial results from these studies indicate a significant increase in sensitivity and specificity of disease assessment either directly with confocal microscopy imaging or as an aid for accurate biopsy sampling (Anandasabapathy 2008; Becker et al., 2008; Collier et al., 2002; Deinert et al., 2007; Drezek et al., 2000; Goetz et al., 2006; Goetz and Kiesslich 2008; Hendee 2002; Hoffman et al., 2006; Hurlstone et al., 2008; Hurlstone and Sanders 2006; Hurlstone et al., 2008; Kakeji et al., 2006; Kiesslich et al., 2007a, b, 2005, 2006a, b; Kiesslich and Neurath 2005, 2006, 2007; Kitabatake et al., 2006; Liu et al., 2008; MacAulay et al., 2004; Miehlke et al., 2007; Pech et al., 2008; Polglase et al., 2005; Tan et al., 2007; Thong et al., 2007; Zheng et al., 2004). White light endoscopes for the GI tract are generally larger (>6 mm) than that used in the pulmonary system (<6 mm), hence deployment of confocal endo-microscopy systems for the lung has been limited. However, recently fabrication of small catheter-based systems that can be fed through the standard auxiliary port of a bronchoscope has been achieved.

Catheter-Based Confocal Microscopy (CBCM) Catheter-based confocal microscopy (CBCM), defined as systems small enough to be fed through small catheters (<3 mm), has recently been investigated in the pulmonary field, for characterization of normal and diseased lung tissue in mouse models (Namati et al., 2007a; b; Namati et al., 2008a; b, c), pigs (Rogers et al., 2008), and humans (Thiberville et al., 2007). Many of the developed catheter-based confocal microscopy systems have been designed to fit through the auxiliary channel of a standard bronchoscope, and therefore can be used to investigate lung tissue from the airway epithelium via the subtending airways (Namati et al., 2007a; b; Namati et al., 2008; Thiberville et al., 2007). In some cases the catheter-based confocal microscopy probes are small enough to reach the peripheral alveoli. Figure 5 (a)–(c) represents three images acquired using catheter-based confocal microscopy systems on a mouse, pig, and human lung, respectively. The mouse in this example was administered with sodium fluorescein (Namati et al., 2008), while the pig and human lung images represent the endogenous autofluorescence of the collagen and elastin network. A recent study using a commercial catheter-based confocal microscopy system (CellVizio, France) successfully demonstrated specific basement membrane

36

E. Namati et al.

Fig. 5 Catheter-based confocal microscopy image from a (a) mouse, (b) pig, and (c) human lung

alterations associated with pre-malignant bronchial legions. In this study, five distinct microscopic patterns were initially identified in normal regions from the proximal airways down to distal respiratory bronchi. In 29 patients with a high risk of lung cancer, disorganization of the elastin network associated with pre-invasive lesions, altered autofluorescent microstructure representing metaplasia or dysplasia, carcinoma in situ and invasive lesions were identified. In addition, spectral analysis was also simultaneously performed using the same device, and provided a quantitative adjunct for analysis of the connective tissue microstructure of the airways (Thiberville et al., 2007). This study provides positive initial results for assessment of lung cancer using a catheter-based confocal microscopy system in humans. A further series of alveolar-based studies have also been recently described, with alveoli from non-smoking and smoking patients being imaged. Clear depiction of the alveolar elastin matrix and alveolar macrophages has been shown using autofluorescence. The results of these studies indicate a significant increase in the number and motility of autofluorescent macrophages in smoking patients versus non-smoking patients (Heckly et al., 2008; Thiberville et al., 2006). In addition, clinical trials are now underway using this commercial catheterbased confocal microscopy system for assessment of alveolar structure and function in normal and diseased patients, as well as evaluation of lung pathologies such as lung cancer (MaunaKeaTechnologies 2008). Taking advantage of the fluorescence approach, where light of one wavelength is used to excite a specimen emitting light at a higher wavelength, highly sensitive biomarker tagging strategies can also be deployed. Two distinct optical biopsy strategies may unfold, one that uses specific fluorescent biomarker tagging of important genetic, molecular, and cellular structures and a second that assesses general information regarding the tumor micro-environment, such as pH, oxygen, glucose, and specific amino acids. An example of catheter-based confocal microscopy imaging on a mouse model of lung cancer is shown in Fig. 6 (a–c). Here three regions have been imaged in a living mouse lung: (a) ‘normal’ parenchyma, (b) suspicious parenchyma, and (c) tumor region. In this example, sodium fluorescein has been systemically administered for visualization of general tissue structure (green) while acridine orange has been applied topically over the region of interest for identification of nuclei (yellow-blue).

Endoscopic Techniques for Optical Imaging

37

Fig. 6 Catheter-based confocal microscopy image of a mouse model of lung cancer, (a) ‘normal’ parenchyma, (b) suspicious region, and (c) tumor region

Catheter Navigation and Guidance In the case of lung cancer, the majority of suspicious nodules identified in MDCT scans are in the peripheral region of the lung. However, using standard bronchoscopy systems only 10 out of the 26 generations of airways can be navigated due to size restrictions. The benefit of utilizing such micro-optical imaging systems discussed above for characterization and diagnosis of early peripheral lung nodules thus becomes important and requires new approaches. Accurate identification and biopsy of suspicious regions of interest (nodules and lymph nodes) in the central lung is a difficult task for even the well-trained bronchoscopist. The utility of a pathway finding system with the ability to track the tip of the bronchoscope and relate back to a virtual bronchoscopy image generated from a MDCT scan has recently become apparent. Virtual bronchoscopy systems have been investigated since the mid-1990 s (Ferretti et al., 1995; Summers et al., 1996; Vining et al., 1996; Ferretti et al., 1997; Higgins et al., 1998); however, only recently has their integration with electromagnetic guidance systems provided a practical utility during clinical procedures (Schwarz et al., 2003; Gildea et al., 2006; Schwarz et al., 2006; Eberhardt et al., 2007a; b; Seijo et al., 2007; Makris and Gourgoulianis 2008). Commercial systems are now available that provide virtual bronchoscopy and electromagnetic navigation for real-time tracking of the bronchoscope during a clinical bronchoscopy procedure. Virtual bronchoscopy systems can be used to create a path to the region of interest using automated or manual routines as shown in Fig. 7 (Kiraly et al., 2004; Ferguson and McLennan 2005; Tschirren et al., 2005; Negahdar et al., 2006; Baker et al., 2007; McLennan et al., 2007a, b). This path can then be followed during the live bronchoscopy procedure. However, there is an inherent problem associated with such a technique; virtual bronchoscopy is based on the MDCT scan, and hence the quality and depth of the airway segmentation is limited by the resolution of MDCT technology (Higgins et al., 1998). Currently with advanced airway segmentation algorithms up to 6–8 generations can be accurately identified as seen in Fig. 7 (Kiraly et al., 2002; 2004;

38

E. Namati et al.

Fig. 7 (a–c) Automated virtual path finding to a nodule of interest. The path is indicated in blue and the nodule is indicated by the yellow sphere. (a) Represents the automated segmented airways and direction path (blue), with the original CT data color coded for individual lobes. (b) Represents a volume reconstruction of the airways and the lung into its individual lobes. (c) Represents the Carina, 3rd, 5th, and 8th bifurcation leading toward the nodule of interest indicated by the yellow sphere. Images courtesy of VIDA diagnostics

Tschirren et al., 2005a, b). Several strategies exist for extending the path of interest, including manual segmentation of a single path (Graham et al., 2008), and further development in this area is ongoing. Navigation of catheter-based micro-optical imaging systems can now be implemented using such electromagnetic navigation systems. However, as indicated above, with limited visual feedback from the virtual bronchoscopy view in the peripheral regions, and the lack of steerability in current micro-optical imaging systems, new approaches for guidance are also needed. Guidance of catheters can be performed using 360◦ steerable catheter sheaths, as implemented by SuperDimension Inc.; however, such systems are again limited by the diameter of the sheath. A technique recently proposed (Riker et al., 2007) is the use of two large external magnets to steer a ferromagnetic-tipped catheter. A commercial system currently exists for steering catheters into coronary arteries (Grady et al., 1989, 1990; Ernst et al., 2004; McPherson and Warnick 2004; Chun et al., 2007; Kiemeneij et al., 2008; Mehta et al., 2008), and development of pulmonary applications is currently

Endoscopic Techniques for Optical Imaging

39

Fig. 8 (a) Bronchoscopy procedure utilizing a Niobe Stereotaxis magnetic guidance system; (b– d) a series of chest X-rays depicting a ferromagnetic guide wire steered toward a peripheral nodule; (e–g) magnified series of X-rays shown in (b–d) with the guide wire tip indicated by a arrow and the boundary of the peripheral nodule circled

underway (Di Biase et al., 2007) (Riker et al., 2007). Figure 8 represents a recent study, where a Niobe magnetic guidance system (Stereotaxis Inc, St. Louis) was used to direct a ferromagnetic-tipped guide wire to small peripheral lung nodules. This was part of a proof of concept study investigating the ability of such a technique for steering micro guide wires to peripheral nodules that would otherwise be very difficult to reach using standard bronchoscopy.

Summary and Future Direction In future, the combination of macro-optical imaging techniques such as white light, autofluorescent, and fluorescein bronchoscopy will be utilized for the initial identification of suspicious regions, while micro-optical techniques such as the optical coherence tomography or catheter-based confocal microscopy system will analyze and classify the regions of interest, enabling live pathologic assessment. In addition, with the integration of navigation and guidance systems, micro-optical catheters will be directed to regions of interest well past the limited bronchoscope range, enabling diagnosis of distant peripheral nodules. The digital imaging revolution has greatly advanced methods such as macroand micro-optical endoscopy and X-ray imaging techniques such as CT scanning. By merging these technologies together, with CT providing detection, some characterization, and pathway guidance (together with electromagnetic strategies) for the macro- and micro-optical bronchoscopic techniques, almost any lung nodule can be sampled within the lung. Nodules that are small and may be pre-cancerous can in the future be ablated at the same bronchoscopic procedure. This process can solve

40

E. Namati et al.

the lung cancer paradox, in much the same way as a dermatologist will ablate small lesions on sun-exposed skin at risk for skin cancers.

References Aleva, R. M., J. Kraan, M. Smith, N. H. T. Ten Hacken, D. S. Postma, and W. Timens. 1998. Techniques in human airway inflammation: quantity and morphology of bronchial biopsy specimens taken by forceps of three sizes. Chest 113 (1):182–185. Anandasabapathy, S. 2008. Endoscopic imaging: emerging optical techniques for the detection of colorectal neoplasia. Curr Opin Gastroenterol 24 (1):64–69. Aziz, D. J., and A. F. Gmitro. 1993. Confocal microscopy via a fiber optic imaging bundle. Paper read at Optical Fibers in Medicine VIII, at Los Angeles, CA, USA. Aziz, D. J., and A. F. Gmitro. 1994. Imaging performance of the fiber optic image-bundle confocal microscope. Paper read at Microscopy, Holography, and Interferometry in Biomedicine, at Budapest, Hungary. Baker, K. M., E. Namati, E. Hoffman, E. J. R. Van Beek, A. Ross, and G. McLennan. 2007. Virtual bronchoscopy: new applications for MDCT in pulmonology. Advances in MDCT 3 (4): 10–18. Bard, M. P., A. Amelink, M. Skurichina, M. den Bakker, S. A. Burgers, J. P. van Meerbeeck, R. P. Duin, J. G. Aerts, H. C. Hoogsteden, and H. J. Sterenborg. 2005. Improving the specificity of fluorescence bronchoscopy for the analysis of neoplastic lesions of the bronchial tree by combination with optical spectroscopy: preliminary communication. Lung Cancer 47 (1): 41–47. Beamis, J. F., Jr., A. Ernst, M. Simoff, R. Yung, and P. Mathur. 2004. A multicenter study comparing autofluorescence bronchoscopy to white light bronchoscopy using a non-laser light stimulation system. Chest 125 (5 Suppl):148S–149S. Becker, V., S. von Delius, M. Bajbouj, A. Karagianni, R. M. Schmid, and A. Meining. 2008. Intravenous application of fluorescein for confocal laser scanning microscopy: evaluation of contrast dynamics and image quality with increasing injection-to-imaging time. Gastrointest Endosc 68(2):319–323. Bird, D., and M. Gu. 2002. Fibre-optic two-photon scanning fluorescence microscopy. J Microsc 208 (Pt 1):35–48. Bjarnhall, G., O. Maepea, G. O. Sperber, C. Linden, E. Monestam, and A. Alm. 2002. Analysis of mean retinal transit time from fluorescein angiography in human eyes: normal values and reproducibility. Acata Ophthalmol Scand 80 (6):652–655. Boguta, A., and D. Wrobel. 2001. Fluorescein and phenolphthalein – correlation of fluorescence and photoelectric properties. J Fluorescence 11 (2):129–137. Bouma, B. E., G. J. Tearney, C. C. Compton, and N. S. Nishioka. 2000. High-resolution imaging of the human esophagus and stomach in vivo using optical coherence tomography. Gastrointest Endosc 51:467–474. Brinkmeyer, E., and R. Ulrich. 1990. High-resolution OCDR in dispersive waveguide. Electron Lett 26:413–414. Browning, D. J. 1999. Fundus photographic, fluorescein angiographic, and indocyanine green angiographic signs in successful laser chorioretinal venous anastomosis for central retinal vein occlusion. Ophthalmology 106 (12):2261–2268. Carlson, K., M. Chidley, K. B. Sung, M. Descour, A. Gillenwater, M. Follen, and R. RichardsKortum. 2005. In vivo fiber-optic confocal reflectance microscope with an injection-molded plastic miniature objective lens. Appl Opt 44 (10):1792–1797. Carlson, A. L., L. G. Coghlan, A. M. Gillenwater, and R. R. Richards-Kortum. 2007a. Dualmode reflectance and fluorescence near-video-rate confocal microscope for architectural, morphological and molecular imaging of tissue. J Microsc 228 (Pt 1):11–24.

Endoscopic Techniques for Optical Imaging

41

Carlson, A. L., A. M. Gillenwater, M. D. Williams, A. K. El-Naggar, and R. R. Richards-Kortum. 2007b. Confocal microscopy and molecular-specific optical contrast agents for the detection of oral neoplasia. Technol Cancer Res Treat 6 (5):361–374. Chen, Y., A. D. Aguirre, P-L. Hsiung, S. Desai, P. R. Herz, M. Pedrosa, Q. Huang, M. Figueiredo, S-W. Huang, A. Koski, J. M. Schmitt, J. G. Fujimoto, and H. Mashimo. 2007. Ultrahigh resolution optical coherence tomography of Barrett s esophagus: preliminary descriptive clinical study correlating images with histology. Endoscopy 39 (7):599–605. Chidley, M. D., K. D. Carlson, R. R. Richards-Kortum, and M. R. Descour. 2006. Design, assembly, and optical bench testing of a high-numerical-aperture miniature injection-molded objective for fiber-optic confocal reflectance microscopy. Appl Opt 45 (11):2545–2554. Chinn, S. R., E. Swanson, and J. G. Fujimoto. 1997. Optical coherence tomography using a frequency-tunable optical source. Opt Lett 22:340–242. Chun, J. K., S. Ernst, S. Matthews, B. Schmidt, D. Bansch, S. Boczor, A. Ujeyl, M. Antz, F. Ouyang, and K. H. Kuck. 2007. Remote-controlled catheter ablation of accessory pathways: results from the magnetic laboratory. Eur Heart J 28 (2):190–195. Collier, T., M. Guillaud, M. Follen, A. Malpica, and R. Richards-Kortum. 2007. Realtime reflectance confocal microscopy: comparison of two-dimensional images and threedimensional image stacks for detection of cervical precancer. J Biomed Opt 12 (2):024021. Collier, T., A. Lacy, R. Richards-Kortum, A. Malpica, and M. Follen. 2002. Near real-time confocal microscopy of amelanotic tissue: detection of dysplasia in ex vivo cervical tissue. Acad Radiol 9 (5):504–512. Deinert, K., R. Kiesslich, M. Vieth, M. F. Neurath, and H. Neuhaus. 2007. In-vivo microvascular imaging of early squamous-cell cancer of the esophagus by confocal laser endomicroscopy. Endoscopy 39 (4):366–368. Delaney, P. M., M. R. Harris, and R. G. King. 1993. Novel microscopy using fibre optic confocal imaging and its suitability for subsurface blood vessel imaging in vivo. Clin Exp Pharmacol Physiol 20 (3):197–198. Delaney, P. M., R. G. King, J. R. Lambert, and M. R. Harris. 1994. Fibre optic confocal imaging (FOCI) for subsurface microscopy of the colon in vivo. J Anat 184 ( Pt 1):157–160. Di Biase, L., T. S. Fahmy, D. Patel, R. Bai, K. Civello, O. M. Wazni, M. Kanj, C. S. Elayi, C. K. Ching, M. Khan, L. Popova, R. A. Schweikert, J. E. Cummings, J. D. Burkhardt, D. O. Martin, M. Bhargava, T. Dresing, W. Saliba, M. Arruda, and A. Natale. 2007. Remote magnetic navigation: human experience in pulmonary vein ablation. J Am Coll Cardiol 50 (9):868–874. Dickensheets, D. L., and G. S. Kino. 1996. Micromachined scaning confocal optical microscope. 21(10):764–766. Drezek, R. A., T. Collier, C. K. Brookner, A. Malpica, R. Lotan, R. R. Richards-Kortum, and M. Follen. 2000. Laser scanning confocal microscopy of cervical tissue before and after application of acetic acid. Am J Obstet Gynecol 182 (5):1135–1139. Eberhardt, R., D. Anantham, A. Ernst, D. Feller-Kopman, and F. Herth. 2007a. Multimodality bronchoscopic diagnosis of peripheral lung lesions: a randomized controlled trial. Am J Respir Crit Care Med 176 (1):36–41. Eberhardt, R., D. Anantham, F. Herth, D. Feller-Kopman, and A. Ernst. 2007b. Electromagnetic navigation diagnostic bronchoscopy in peripheral lung lesions. Chest 131 (6): 1800–1805. Ernst, S., F. Ouyang, C. Linder, K. Hertting, F. Stahl, J. Chun, H. Hachiya, D. Bansch, M. Antz, and K. H. Kuck. 2004. Initial experience with remote catheter ablation using a novel magnetic navigation system: magnetic remote catheter ablation. Circulation 109 (12): 1472–1475. Ernst, A., M. J. Simoff, P. N. Mathur, R. C. Yung, and J. F. Beamis. 2005. D-light autofluorescence in the detection of premalignant airway changes: a multicenter trial. J Bronchology 12 (3): 133–138. Feller-Kopman, D., W. Lunn, and A. Ernst. 2005. Autofluorescence bronchoscopy and endobronchial ultrasound: a practical review. Ann Thorac Surg 80 (6):2395–2401.

42

E. Namati et al.

Ferguson, J. S., and G. McLennan. 2005. Virtual bronchoscopy. Proc Am Thorac Soc 2 (6):488–491, 504–505. Ferretti, G. R., J. Knoplioch, I. Bricault, C. Brambilla, and M. Coulomb. 1997. Central airway stenoses: preliminary results of spiral-CT-generated virtual bronchoscopy simulations in 29 patients. Eur Radiol 7 (6):854–859. Ferretti, G., J. Knoplioch, M. Coulomb, C. Brambilla, and P. Cinquin. 1995. [Endoluminal 3D reconstruction of the tracheo-bronchial tree (virtual bronchoscopy)]. J Radiol 76 (8):531–534. Flower, R. W. 1973. Injection technique for indocyanine green and sodium fluorescein dye angiography of the eye. Invest Ophthalmol 12 (12):881–903. Flusberg, B. A., E. D. Cocker, W. Piyawattanametha, J. C. Jung, E. L. Cheung, and M. J. Schnitzer. 2005a. Fiber-optic fluorescence imaging. Nat Methods 2 (12):941–950. Flusberg, B. A., J. C. Jung, E. D. Cocker, E. P. Anderson, and M. J. Schnitzer. 2005b. In vivo brain imaging using a portable 3.9 gram two-photon fluorescence microendoscope. Opt Lett 30 (17):2272–2274. Fujimoto, J. G., M. E. Brezinski, G. J. Tearney, S. A. Boppart, B. Bouma, M. R. Hee, J. F. Southern, and E. A. Swanson. 1995. Optical biopsy and imaging using optical coherence tomography. Nat Med 1 (9):970–972. Fujimoto, J. G., C. Pitris, S. A. Boppart, and M. E. Brezinski. 2000. Optical coherence tomography: an emerging technology for biomedical imaging and optical biopsy. Neoplasia 2 (1–2):9–25. Gildea, T. R., P. J. Mazzone, D. Karnak, M. Meziane, and A. C. Mehta. 2006. Electromagnetic navigation diagnostic bronchoscopy: a prospective study. Am J Respir Crit Care Med 174 (9):982–989. Gobel, W., J. N. Kerr, A. Nimmerjahn, and F. Helmchen. 2004. Miniaturized two-photon microscope based on a flexible coherent fiber bundle and a gradient-index lens objective. Opt Lett 29 (21):2521–2523. Goetz, M., A. Hoffman, P. R. Galle, M. F. Neurath, and R. Kiesslich. 2006. Confocal laser endoscopy: new approach to the early diagnosis of tumors of the esophagus and stomach. Future Oncol 2 (4):469–476. Goetz, M., and R. Kiesslich. 2008. Confocal endomicroscopy: in vivo diagnosis of neoplastic lesions of the gastrointestinal tract. Anticancer Res 28 (1B):353–360. Golubovic, B., B. E. Bouma, G. J. Tearney, and J. G. Fujimoto. 1997. Optical frequency-domain reflectometry using rapid wavelength tuning of a Cr4+:forsterite laser. Optics Letters 22: 1704–1706. Gopalakrishnan, D. 2003. Color Analysis of the Human Airway Wall. Master of Science, Dept of Biomedical Engineering, University of Iowa, Iowa City. Grady, M. S., M. A. Howard, 3rd, J. A. Molloy, R. C. Ritter, E. G. Quate, and G. T. Gillies. 1989. Preliminary experimental investigation of in vivo magnetic manipulation: results and potential application in hyperthermia. Med Phys 16 (2):263–272. Grady, M. S., M. A. Howard, 3rd, J. A. Molloy, R. C. Ritter, E. G. Quate, and G. T. Gillies. 1990. Nonlinear magnetic stereotaxis: three-dimensional, in vivo remote magnetic manipulation of a small object in canine brain. Med Phys 17 (3):405–415. Graham, M. W., J. D. Gibbs, and W. E. Higgins. 2008. Robust system for human airway-tree segmentation. Paper read at Medical Imaging, at San Diego. Haussinger, K., H. Becker, F. Stanzel, A. Kreuzer, B. Schmidt, J. Strausz, S. Cavaliere, F. Herth, M. Kohlhaufl, K. M. Muller, R. M. Huber, U. Pichlmeier, and T. Bolliger Ch. 2005. Autofluorescence bronchoscopy with white light bronchoscopy compared with white light bronchoscopy alone for the detection of precancerous lesions: a European randomised controlled multicentre trial. Thorax 60 (6):496–503. Heckly, B. G., L. Thiberville, C. Vever-Bizet, and B. Viellerobe. 2008. In vivo endoscopic autofluorescence microspectro-imaging of bronchi and alveoli. Paper read at SPIE: Endoscopic Microscopy III, Hendee, W. R. 2002. Confocal microscopy: a new technique for in vivo imaging. Acad Radiol 9 (5):503.

Endoscopic Techniques for Optical Imaging

43

Herth, F. J., R. Eberhardt, and A. Ernst. 2006. The future of bronchoscopy in diagnosing, staging and treatment of lung cancer. Respiration 73 (4):399–409. Higgins, W. E., K. Ramaswamy, R. D. Swift, G. McLennan, and E. A. Hoffman. 1998. Virtual bronchoscopy for three–dimensional pulmonary image assessment: state of the art and future needs. Radiographics 18 (3):761–778. Hoffman, A., M. Goetz, M. Vieth, P. R. Galle, M. F. Neurath, and R. Kiesslich. 2006. Confocal laser endomicroscopy: technical status and current indications. Endoscopy 38 (12):1275–1283. Hsiung, P-L., L. Pantanowitz, A. D. Aguirre, Y. Chen, D. Phatak, T. H. Ko, S. Bourquin, S. J. Schnitt, S. Raza, J. L. Connolly, H. Mashimo, and J. G. Fujimoto. 2005. Ultrahigh-resolution and 3-dimensional optical coherence tomography ex vivo imaging of the large and small intestines. Gastrointest Endosc 62 (4):561–574. Hurlstone, D. P., W. Baraza, S. Brown, M. Thomson, N. Tiffin, and S. S. Cross. 2008. In vivo real-time confocal laser scanning endomicroscopic colonoscopy for the detection and characterization of colorectal neoplasia. Br J Surg 95 (5):636–645. Hurlstone, D. P., and D. S. Sanders. 2006. Recent advances in chromoscopic colonoscopy and endomicroscopy. Curr Gastroenterol Rep 8 (5):409–415. Hurlstone, D. P., N. Tiffin, S. R. Brown, W. Baraza, M. Thomson, and S. S. Cross. 2008. In vivo confocal laser scanning chromo-endomicroscopy of colorectal neoplasia: changing the technological paradigm. Histopathology 52 (4):417–426. Ikeda, N., A. Hayashi, K. Iwasaki, M. Tsuboi, J. Usuda, and H. Kato. 2007. Comprehensive diagnostic bronchoscopy of central type early stage lung cancer. Lung Cancer 56 (3):295–302. Ikeda, S., N. Yanai, and S. Ishikawa. 1968. Flexible bronchofiberscope. Keio J Med 17 (1):1–16. Jemal, A., R. Siegel, E. Ward, T. Murray, J. Xu, and M. J. Thun. 2007. Cancer statistics, 2007. CA Cancer J Clin 57 (1):43–66. Jung, J. C., and M. J. Schnitzer. 2003. Multiphoton endoscopy. Opt Lett 28 (11):902–904. Kakeji, Y., S. Yamaguchi, D. Yoshida, K. Tanoue, M. Ueda, A. Masunari, T. Utsunomiya, M. Imamura, H. Honda, Y. Maehara, and M. Hashizume. 2006. Development and assessment of morphologic criteria for diagnosing gastric cancer using confocal endomicroscopy: an ex vivo and in vivo study. Endoscopy 38 (9):886–890. Kennedy, T. C., S. Lam, and F. R. Hirsch. 2001. Review of recent advances in fluorescence bronchoscopy in early localization of central airway lung cancer. Oncologist 6 (3):257–262. Kerr, K. M. 2001. Pulmonary preinvasive neplasia. J Clin Pathol 54:257–271. Kiemeneij, F., M. S. Patterson, G. Amoroso, G. Laarman, and T. Slagboom. 2008. Use of the Stereotaxis Niobe magnetic navigation system for percutaneous coronary intervention: results from 350 consecutive patients. Catheter Cardiovasc Interv 71 (4):510–516. Kiesslich, R., M. Goetz, K. Lammersdorf, C. Schneider, J. Burg, M. Stolte, M. Vieth, B. Nafe, P. R. Galle, and M. F. Neurath. 2007a. Chromoscopy-guided endomicroscopy increases the diagnostic yield of intraepithelial neoplasia in ulcerative colitis. Gastroenterology 132 (3): 874–882. Kiesslich, R., M. Goetz, M. Vieth, P. R. Galle, and M. F. Neurath. 2005. Confocal laser endomicroscopy. Gastrointest Endosc Clin N Am 15 (4):715–731. Kiesslich, R., M. Goetz, M. Vieth, P. R. Galle, and M. F. Neurath. 2007b. Technology insight: confocal laser endoscopy for in vivo diagnosis of colorectal cancer. Nat Clin Pract Oncol 4 (8):480–490. Kiesslich, R., L. Gossner, M. Goetz, A. Dahlmann, M. Vieth, M. Stolte, A. Hoffman, M. Jung, B. Nafe, P. R. Galle, and M. F. Neurath. 2006a. In vivo histology of Barrett s esophagus and associated neoplasia by confocal laser endomicroscopy. Clin Gastroenterol Hepatol 4 (8): 979–887. Kiesslich, R., A. Hoffman, M. Goetz, S. Biesterfeld, M. Vieth, P. R. Galle, and M. F. Neurath. 2006b. In vivo diagnosis of collagenous colitis by confocal endomicroscopy. Gut 55 (4): 591–592. Kiesslich, R., and M. F. Neurath. 2005. Endoscopic confocal imaging. Clin Gastroenterol Hepatol 3 (7 Suppl 1):S58–S60.

44

E. Namati et al.

Kiesslich, R., and M. F. Neurath. 2006. Chromoendoscopy and other novel imaging techniques. Gastroenterol Clin North Am 35 (3):605–619. Kiesslich, R., and M. F. Neurath. 2007. Endomicroscopy is born–do we still need the pathologist? Gastrointest Endosc 66 (1):150–153. Killian, G. 1898. Úber direkte Bronchoscopie. MMW 45 (27):844–847. Kim, P., M. Puoris haag, D. Cote, C. P. Lin, and S. H. Yun. 2008. In vivo confocal and multiphoton microendoscopy. J Biomed Opt 13 (1):010501. King, R. G., and P. M. Delaney. 1994. Confocal microscopy in pharmacological research. Trends Pharmacol Sci 15 (8):275–279. Kiraly, A. P., J. P. Helferty, E. A. Hoffman, G. McLennan, and W. E. Higgins. 2004. Threedimensional path planning for virtual bronchoscopy. IEEE Trans Med Imaging 23 (11): 1365–1379. Kiraly, A. P., W. E. Higgins, G. McLennan, E. A. Hoffman, and J. M. Reinhardt. 2002. Threedimensional human airway segmentation methods for clinical virtual bronchoscopy. Acad Radiol 9 (10):1153–1168. Kitabatake, S., Y. Niwa, R. Miyahara, A. Ohashi, T. Matsuura, Y. Iguchi, Y. Shimoyama, T. Nagasaka, O. Maeda, T. Ando, N. Ohmiya, A. Itoh, Y. Hirooka, and H. Goto. 2006. Confocal endomicroscopy for the diagnosis of gastric cancer in vivo. Endoscopy 38 (11):1110–1114. Knyrim, K., H. K. Seidlitz, F. Hagenmuller, and M. Classen. 1987. Color performance of video endoscopes: quantitative measurement of color reproduction. Endoscopy 19 (6):233–236. Lam, S., C. MacAulay, J. Hung, J. LeRiche, A. E. Profio, and B. Palcic. 1993. Detection of dysplasia and carcinoma in situ with a lung imaging fluorescence endoscope device. J Thorac Cardiovasc Surg 105 (6):1035–1040. Lam, S., C. MacAulay, J. C. leRiche, and B. Palcic. 2000. Detection and localization of early lung cancer by fluorescence bronchoscopy. Cancer 89 (11 Suppl):2468–2473. Lam, S., B. Standish, C. Baldwin, A. McWilliams, J. leRiche, A. Gazdar, A. I. Vitkin, V. Yang, N. Ikeda, and C. MacAulay. 2008. In vivo optical coherence tomography imaging of preinvasive bronchial lesions. Clin Cancer Res 14 (7):2006–2011. Landis, S. H., T. Murray, S. Bolden, and P. A. Wingo. 1998. Cancer statistics, 1998. CA Cancer J Clin 48 (1):6–29. Lexer, F., C. K. Hitzenberger, A. F. Fercher, and M. Kulhavy. 1997. Wavelength-tuning interferometry of intraocular distances. Appl Opt 36:6548–6553. Liu, H., Y. Q. Li, T. Yu, Y. A. Zhao, J. P. Zhang, J. N. Zhang, Y. T. Guo, X. J. Xie, T. G. Zhang, and P. V. Desmond. 2008. Confocal endomicroscopy for in vivo detection of microvascular architecture in normal and malignant lesions of upper gastrointestinal tract. J Gastroenterol Hepatol 23 (1):56–61. Liu, J. T., M. J. Mandella, S. Friedland, R. Soetikno, J. M. Crawford, C. H. Contag, G. S. Kino, and T. D. Wang. 2006. Dual-axes confocal reflectance microscope for distinguishing colonic neoplasia. J Biomed Opt 11 (5):054019. Liu, J. T., M. J. Mandella, H. Ra, L. K. Wong, O. Solgaard, G. S. Kino, W. Piyawattanametha, C. H. Contag, and T. D. Wang. 2007. Miniature near-infrared dual-axes confocal microscope utilizing a two-dimensional microelectromechanical systems scanner. Opt Lett 32 (3):256–258. MacAulay, C., P. Lane, and R. Richards-Kortum. 2004. In vivo pathology: microendoscopy as a new endoscopic imaging modality. Gastrointest Endosc Clin N Am 14 (3):595–620, xi. Maitland, K. C., A. M. Gillenwater, M. D. Williams, A. K. El-Naggar, M. R. Descour, and R. R. Richards-Kortum. 2008. In vivo imaging of oral neoplasia using a miniaturized fiber optic confocal reflectance microscope. Oral Oncol 44(11):1059–1066. Makris, D., and K. I. Gourgoulianis. 2008. Electromagnetic navigation diagnostic bronchoscopy and transbronchial biopsy. Chest 133 (3):829–830. MaunaKeaTechnologies 2008. Mauna Kea Technologies 2008 [cited. Available from www.maunakeatech.com. McLennan, G., J. S. Ferguson, K. Thomas, A. S. Delsing, J. Cook-Granroth, and E. A. Hoffman. 2007a. The use of MDCT-based computer-aided pathway finding for mediastinal and

Endoscopic Techniques for Optical Imaging

45

perihilar lymph node biopsy: a randomized controlled prospective trial. Respiration 74 (4): 423–431. McLennan, G., E. Namati, J. Ganatra, M. Suter, E. E. O Brien, K. Lecamwasam, E. J. R. Van Beek, and E. Hoffman. 2007b. Virtual Bronchoscopy. MRI 11 (1):10–20. McPherson, C. M., and R. E. Warnick. 2004. Results of contemporary surgical management of radiation necrosis using frameless stereotaxis and intraoperative magnetic resonance imaging. J Neurooncol 68 (1):41–47. McWilliams, A., C. MacAulay, A. F. Gazdar, and S. Lam. 2002. Innovative molecular and imaging approaches for the detection of lung cancer and its precursor lesions. Oncogene 21 (45): 6949–6959. Mehta, R., D. T. Hart, B. S. Nagra, Z. Liu, and B. K. Kantharia. 2008. Successful ablation of focal left atrial tachycardia using Stereotaxis Niobe remote magnetic navigation system. Europace 10 (3):280–283. Miehlke, S., A. Morgner, D. Aust, A. Madisch, M. Vieth, and G. Baretton. 2007. Combined use of narrow-band imaging magnification endoscopy and miniprobe confocal laser microscopy in neoplastic Barrett’s esophagus. Endoscopy 39 Suppl 1:E316. Minsky, M. 1961. Microscopy apparatus, edited by U. P. Office, USA. Minsky, M. 1988. Memoir on inventing the confocal microscope. Scanning 10:128–138. Myaing, M. T., D. J. MacDonald, and X. Li. 2006. Fiber-optic scanning two-photon fluorescence endoscope. Opt Lett 31 (8):1076–1078. Namati, E., J. Thiesse, J. de Ryk, and G. McLennan. 2007a. Dynamic in vivo alveolar morphology using a novel laser scanning confocal microscope. Paper read at 9th Biennial Conference of the Australian Pattern Recognition Society on Digital Image Computing Techniques and Applications (DICTA 2007), 05/12/2007, at Adelaide, Australia. Namati, E., J. Thiesse, J. de Ryk, and G. McLennan. 2007b. In vivo imaging of alveoli. Paper read at American Thoracic Society, 05/21/2007, at San Francisco, CA. Namati, E., J. Thiesse, J. de Ryk, and G. McLennan. 2008a. Alveolar Dynamics during respiration: are the pores of Kohn a pathway to recruitment? Am J Respir Cell Mol Biol 38 (5):572–578. Namati, E., J. Thiesse, J. de Ryk, and G. McLennan. 2008b. In vivo catheter-based histopathology of lung cancer in mice. Paper read at American Thoracic Society, 05/20/2008, at Toronto, Canada. Namati, E., J. Thiesse, J. de Ryk, and G. McLennan. 2008c. Longitudinal multi-modality threedimensional imaging of lung cancer in mice. Paper read at American Thoracic Society, 05/19/2008, at Toronto, Canada. Negahdar, M., A. Ahmadian, N. Navab, and K. Firouznia. 2006. Path planning for virtual bronchoscopy. Conf Proc IEEE Eng Med Biol Soc 1:156–159. Pech, O., T. Rabenstein, H. Manner, M. C. Petrone, J. Pohl, M. Vieth, M. Stolte, and C. Ell. 2008. Confocal laser endomicroscopy for in vivo diagnosis of early squamous cell carcinoma in the esophagus. Clin Gastroenterol Hepatol 6 (1):89–94. Polglase, A. L., W. J. McLaren, and P. M. Delaney. 2006. Pentax confocal endomicroscope: a novel imaging device for in vivo histology of the upper and lower gastrointestinal tract. Expert Rev Med Devices 3 (5):549–556. Polglase, A. L., W. J. McLaren, S. A. Skinner, R. Kiesslich, M. F. Neurath, and P. M. Delaney. 2005. A fluorescence confocal endomicroscope for in vivo microscopy of the upper- and the lower-GI tract. Gastrointest Endosc 62 (5):686–695. Proctor, R. N. 2001. Tobacco and the global lung cancer epidemic. Nat Rev Cancer 1 (1):82–86. Rector, D. M., D. M. Ranken, and J. S. George. 2003. High-performance confocal system for microscopic or endoscopic applications. Methods 30 (1):16–27. Richards-Kortum, R., and E. Sevick-Muraca. 1996. Quantitative optical spectroscopy for tissue diagnosis. Annu Rev Phys Chem 47:555–606. Riker, D. R., J. Scott Ferguson, K. J. Sprenger, S. Nibaur, K. Baker, J. Tschirren, P. H. Burmeister, and G. McLennan. 2007. Magnetic Navigation Guided Videobronchoscopy. Paper read at American Thoracic Society, 05/22/2007, at San Diego.

46

E. Namati et al.

Rogers, C. S., W. M. Abraham, K. A. Brogden, J. F. Engelhardt, J. T. Fisher, P. B. McCray Jr, G. McLennan, D. K. Meyerholz, E. Namati, L. S. Ostedgaard, R. S. Prather, J. R. Sabater, D. A. Stoltz, J. Zabner, and M. J. Welsh. 2008. The porcine lung as a potential model for cystic fibrosis. Am J Physiol Lung Cell Mol Physiol 295(2):L240–L263. Rouse, A. R., and A. F. Gmitro. 2000. Multispectral imaging with a confocal microendoscope. Opt Lett 25 (23):1708–1710. Rouse, A. R., A. Kano, J. A. Udovich, S. M. Kroto, and A. F. Gmitro. 2004. Design and demonstration of a miniature catheter for a confocal microendoscope. Appl Opt 43 (31):5763–571. Sabharwal, Y. S., A. R. Rouse, L. Donaldson, M. F. Hopkins, and A. F. Gmitro. 1999. Slit-scanning confocal microendoscope for high-resolution in vivo imaging. Appl Opt 38 (34):7133–7144. Schwarz, Y., J. Greif, H. D. Becker, A. Ernst, and A. Mehta. 2006. Real-time electromagnetic navigation bronchoscopy to peripheral lung lesions using overlaid CT images: the first human study. Chest 129 (4):988–994. Schwarz, Y., A. C. Mehta, A. Ernst, F. Herth, A. Engel, D. Besser, and H. D. Becker. 2003. Electromagnetic navigation during flexible bronchoscopy. Respiration 70 (5):516–522. Seijo, L. M., G. Bastarrika, M. D. Lozano, and J. J. Zulueta. 2007. [Preliminary experience with the use of electromagnetic navigation for the diagnosis of peripheral pulmonary nodules and enlarged mediastinal lymph nodes]. Arch Bronconeumol 43 (8):460–463. Sergeev, A. M., V. M. Gelikonov, G. V. Gelikonov, F. L. Feldchtein, R. V. Kuranov, N. D. Gladkova, N. M. Shakhova, L. B. Suopova, A. V. Shakhov, I. A. Kuznetzova, A. N. Denisenko, V. V. Pochinki, Y. P. Chumakoc, and O. S. Streltzova. 1997. In vivo endoscopic OCT imaging of precancer and cancer states of human mucosa. Opt Express 1:432–1440. Sivak, M. V., Jr., K. Kobayashi, J. A. Izatt, A. M. Rollins, R. Ung-Runyawee, A. Chak, R. C. Wong, G. A. Isenberg, and J. Willis. 2000. High-resolution endoscopic imaging of the GI tract using optical coherence tomography. Gastrointest Endosc 51:474–479. Smithpeter, C. L., A. K. Dunn, A. J. Welch, and R. Richards-Kortum. 1998. Penetration depth limits of in vivo confocal reflectance imaging. Appl Opt 37 (13):2749–2754. Summers, R. M., D. H. Feng, S. M. Holland, M. C. Sneller, and J. H. Shelhamer. 1996. Virtual bronchoscopy: segmentation method for real-time display. Radiology 200 (3):857–862. Sung, K. B., C. Liang, M. Descour, T. Collier, M. Follen, and R. Richards-Kortum. 2002. Fiberoptic confocal reflectance microscope with miniature objective for in vivo imaging of human tissues. IEEE Trans Biomed Eng 49 (10):1168–1172. Sutedja, G. 2003. New techniques for early detection of lung cancer. Eur Respir J Suppl 39: 57 s–66 s. Suter, M. J. 2005. Multi-modal Structural and Functional Description and Analyses of the Pulmonary Airway Mucosa, Department of Biomedical Engineering, University of Iowa, Iowa City. Suter, M. J., D. K. Kang, B. E. Bouma, and G. J. Tearney. 2008a. Multi-modal optical imaigng of the swine airways ex vivo. Paper read at American Thoracic Society, at Toronto, Canada. Suter, M. J., G. McLennan, J. M. Reinhardt, P. R. Montaue, P. Taft, J. Lee, and J. Zabner. 2005a. Bronchoscopic imaging of pulmonary mucosal vasculature responses to inflammatory mediators. J Biomed Opt 10 (3):034013. Suter, M., G. McLennan, J. M. Reinhardt, D. Riker, and E. A. Hoffman. 2005b. Macro-optical color assessment of the pulmonary airways with subsequent three-dimensional multidetectorx-ray-computed-tomography assisted display. J Biomed Opt 10 (5):051703. Suter, M. J., J. M. Reinhardt, and G. McLennan. 2008b. Integrated CT/bronchoscopy in the central airways preliminary results. Acad Radiol 15 (6):786–798. Suter, M., J. Reinhardt, P. Montague, P. Taft, J. Lee, J. Zabner, and G. McLennan. 2005c. Bronchoscopic imaging of pulmonary mucosal vasculature responses to inflammatory mediators. J Biomed Opt 10 (3):034013. Suter, M., J. M. Reinhardt, D. Riker, W. E. Higgins, E. A. Hoffman, and G. McLennan. 2005d. Macro-optical color assessment of the pulmonary airways with subsequent 3D MDCT assisted display. J Biomed Opt 10 (5):051703.

Endoscopic Techniques for Optical Imaging

47

Suter, M. J., J. M. Reinhardt, M. Sonka, W. E. Higgins, E. A. Hoffman, and G. McLennan. 2004a. Three dimensional true color topographical analysis of the pulmonary airways. Paper read at SPIE Conference on Medical Imaging, at San Diego. Suter, M. J., J. Tschirren, J. M. Reinhardt, M. Sonka, E. A. Hoffman, W. E. Higgins, and G. McLennan. 2004b. Evaluation of the human airway with multidetector X-ray computed tomography and optical imaging. Physiol Meas 25 (4):837–847. Tan, J., P. Delaney, and W. J. McLaren. 2007. Confocal endomicroscopy: a novel imaging technique for in vivo histology of cervical intraepithelial neoplasia. Expert Rev Med Devices 4 (6):863–871. Tearney, G. J., M. E. Brezinski, B. E. Bouma, S. A. Boppart, C. Pitris, J. F. Southem, and J. G. Fujimoto. 1997. In vivo endoscopic optical biopsy with optical coherence tomography. Science 276:2037–2039. Thiberville, L., G. B. Heckly, E. Peltier, and C. Cave. 2006. Perendoscopic alveolar imaging using fluorescent confocal fibered microscopy. Paper read at European Respiratory Society 2006. Thiberville, L., S. Moreno-Swirc, T. Vercauteren, E. Peltier, C. Cave, and G. B. Heckly. 2007. In vivo imaging of the bronchial wall microstructure using fibered confocal fluorescence microscopy. Am J Respir Crit Care Med 175 (1):22–31. Thong, P. S., M. Olivo, K. W. Kho, W. Zheng, K. Mancer, M. Harris, and K. C. Soo. 2007. Laser confocal endomicroscopy as a novel technique for fluorescence diagnostic imaging of the oral cavity. J Biomed Opt 12 (1):014007. Tschirren, J., E. A. Hoffman, G. McLennan, and M. Sonka. 2005a. Intrathoracic airway trees: segmentation and airway morphology analysis from low-dose CT scans. IEEE Trans Med Imaging 24 (12):1529–1539. Tschirren, J., E. A. Hoffman, G. McLennan, and M. Sonka. 2005b. Segmentation and quantitative analysis of intrathoracic airway trees from computed tomography images. Proc Am Thorac Soc 2 (6):484–487, 503–504. Tschirren, J., G. McLennan, K. Palagyi, E. A. Hoffman, and M. Sonka. 2005. Matching and anatomical labeling of human airway tree. IEEE Trans Med Imaging 24 (12):1540–1547. Tsuboi, M., A. Hayashi, N. Ikeda, H. Honda, Y. Kato, S. Ichinose, and H. Kato. 2005. Optical coherence tomography in the diagnosis of bronchial lesions. Lung Cancer 49:387–394. Vincent, B. D., M. Fraig, and G. A. Silvestri. 2007. A pilot study of narrow-band imaging compared to white light bronchoscopy for evaluation of normal airways and premalignant and malignant airways disease. Chest 131 (6):1794–1799. Vining, D. J., K. Liu, R. H. Choplin, and E. F. Haponik. 1996. Virtual bronchoscopy. Relationships of virtual reality endobronchial simulations to actual bronchoscopic findings. Chest 109 (2):549–553. Wang, T. D., C. H. Contag, M. J. Mandella, N. Y. Chan, and G. S. Kino. 2003a. Dual-axes confocal microscopy with post-objective scanning and low-coherence heterodyne detection. Opt Lett 28 (20):1915–1917. Wang, T. D., C. H. Contag, M. J. Mandella, N. Y. Chan, and G. S. Kino. 2004. Confocal fluorescence microscope with dual-axis architecture and biaxial postobjective scanning. J Biomed Opt 9 (4):735–742. Wang, T. D., M. J. Mandella, C. H. Contag, and G. S. Kino. 2003b. Dual-axis confocal microscope for high-resolution in vivo imaging. Opt Lett 28 (6):414–416. Weissleder, R., and M. J. Pittet. 2008. Imaging in the era of molecular oncology. Nature 452 (7187):580–589. Whiteman, S. C., Y. Yang, D. G. van Pittius, M. Stephens, J. Parmer, and M. A. Spiteri. 2006. Optical coherence tomography: real-time imaging of bronchial airways microstructure and detection of inflammatory/neoplastic morphologic changes. Clin Cancer Res 12 (3): 813–818. Yang, Y., S. C. Whiteman, D. G. van Pittius, Y. He, R. K. Wang, and M. A. Spiteri. 2004. Use of optical coherence tomography in delineating airways microstructure: comparison of OCT images to histopathological sections. Phys Med Biol 49 (7):1247–1255.

48

E. Namati et al.

Yun, S. H., C. Boudoux, G. J. Tearney, and B. E. Bouma. 2003. High-speed wavelength-swept semiconductor laser with a polygon-scanner-based wavelength filter. Opt Lett 28:1981–1983. Yun, S. H., G. J. Tearney, J. F. de Boer, N. Iftimia, and B. E. Bouma. 2003. High-speed optical frequency-domain imaging. Opt Express 11 (22):2953–2963. Yun, S. H., G. J. Tearney, B. J. Vakoc, M. Shishkov, W. Y. Oh, A. E. Desjardins, M. J. Suter, R. C. Chan, J. A. Evans, I. K. Jang, N. S. Nishioka, J. F. de Boer, and B. E. Bouma. 2006. Comprehensive volumetric optical microscopy in vivo. Nat Med 12 (12):1429–1433. Zheng, W., M. Harris, K. W. Kho, P. S. Thong, A. Hibbs, M. Olivo, and K. C. Soo. 2004. Confocal endomicroscopic imaging of normal and neoplastic human tongue tissue using ALA-inducedPPIX fluorescence: a preliminary study. Oncol Rep 12 (2):397–401.

Design and Use of the Surgical Microscope in Fluorescence-Guided Surgery Max Sturgis

Introduction This chapter will discuss the various issues surrounding surgical microscopes and the use of FGS (fluorescence-guided surgery). Fluorescence in surgery is presently used to delineate tumors or visualize blood flow so the surgeon can be guided as to where tumor margins may be or to visualize the changes in blood flow based on surgical intervention. This section will describe two intra-operative uses of optical imaging: (1) visualization of blood flow using ICG (indocyanine green) as the fluorescence agent and fluorescence in the NIR (near-infrared) range; and (2) use of 5-ALA (5-aminolevulinic acid) as the metabolic precursor of heme and inducing the synthesis of protoporphyrin IX (PpIX), with the PpIX becoming the fluorescent agent. Here the excitation light is in the UV range and the emission is in the red range. Each of these devices will be more fully described below, followed by a discussion on OR implementation and a discussion of clinical investigation for those who wish to design their own studies using a surgical microscope. Two operating microscopes in specific will be mentioned by name that the author is familiar with and that have been introduced by Leica Surgical of Heerbrugg, Switzerland. One, the Leica FL800, is used for visualization of blood flow. It uses ICG (indocyanine green) as the fluorescence agent and fluorescence in the NIR (near-infrared) range. The second device is the Leica FL400, which uses 5-ALA (5-aminolevulinic acid) as the metabolic precursor of heme and induces the synthesis of protoporphyrin IX (PpIX), with the PpIX becoming the fluorescent agent.

The Operating Microscope for Visualization of Blood Flow Compared to the standard operating microscope, there are three alterations which allow the fluorescent operating microscope (for example, the Leica FL800) to function: M. Sturgis (B) R&D Engineering – Surgical Consultant, Business Unit Surgical Operating Microscope, Leica Microsystems AG, Max Schmidheiny-Strasse 201, CH-9435, Heerbrugg, Switzerland e-mail: [email protected]

E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_3, 

49

50

M. Sturgis

1. Excitation light – The surgical microscope light source must be altered from normally filtered light to allow excitation light. Figure 1 is the normal microscope light, which blocks all energy which is not visible to the surgeon. When the fluorescent operating microscope is activated, the filters are changed on a motorized wheel inside the Xe light source to expand the light from 700 to 800 nm as shown in Fig. 2. This additional NIR energy allows the injected ICG combined with albumin to be excited and fluoresce.

Fig. 1 Normal surgical microscope light filtered below 400 nm and above 700 nm

Fig. 2 Vascular fluorescence light which in addition to normal light allows 700–800 nm light into field for excitation of ICG

Design and Use of the Surgical Microscope in Fluorescence-Guided Surgery

51

2. Seeing ICG fluorescence – A special black and white high-sensitive NIR camera and special 820–860 nm filter is now engaged to see the fluorescence which is invisible to the surgeon’s eye (Fig. 3).

Fig. 3 Special NIR camera adapter reflecting 820–860 nm energy to NIR camera

3. Recording software – As blood flow is a dynamic event and the ICG flow will last only~15 s, it is important to be able to easily play back images and play back in slow motion. Figure 4 is an illustration of the HDMD system with a FL800 loop being recorded. The vascular fluorescence loop can easily be played back with a single touch of loop button.

Fig. 4 Special recording hardware that allows surgeon to quickly play back vascular fluorescence loops at normal or slower speeds

Summary of fluorescence operating microscope – Figure 5 is an illustration of the components as they would appear in a vascular fluorescence surgical microscope. Vascular fluorescence surgical microscope images – Clinical applications can include assessment of flow through tumors or other structures (Fig. 6). Using the FL800 Leica vascular fluorescence microscope, ICG can be imaged in three stages as it flows through a human brain. Without pseudo-coloring, the images will appear in black and white.

52

M. Sturgis

Fluorescence Image OR Monitor

NIR-CCD

Emission sensor ICG-Filter

VIS-CCD

Emission filter Excitation Light

Leica FL800 Dual Video Adapter

ICG Fluorescence

Leica OH3 light Source

Object

Fig. 5 Components of the FL800 vascular fluorescence surgical microscope

Fig. 6 Serial flow through the brain over the course of 2–9 s post-injection bolus

Bolus passage after 2 seconds: arterial view

Bolus passage after 5 seconds: capilary view

Bolus passage after 9 seconds: venous view

Design and Use of the Surgical Microscope in Fluorescence-Guided Surgery

53

Technical considerations related to the OR implementation of the fluorescence stereomicroscopy– Initial use of these techniques in the operating room can be limited by the following technical factors: ICG on hand – It is critical that the OR keep at least four vials of ICG on hand. Although most surgical cases can be handled with one or two 25 mg vials, it is imperative to have more on hand as it is the cases which do not go as planned where ICG fluorescence works best. Correct injection of a bolus – It is common for a surgeon’s first experience with ICG fluorescence that the fluorescence is weak and lasts a long time. This is a sign that a correct bolus was not injected. For example, a correct bolus for a 25 mg dose (one vial) injected in the antecubital vein within 5 s will fluoresce in the brain for about 15 s. To get the same sort of response in a tibial vein, the injection will need to be cut to 2 s, as the ICG dilutes with the distance it travels and the quantity of blood it mixes with. Faster injections always produce stronger fluorescence with a shorter period of fluorescence. Understanding grey blood – As ICG is a hepatic uptake drug with a half-life of approximately 10 min, some neurovascular surgeons have successfully cut the dose in half and corresponding doubling of the speed of the injection. This half dose allows them do a high number of doses with only several minutes between doses as often needed for AVM (arteriovenous malformation) surgery. Understanding the effect of working distance – Light loses energy at a rate of 1/WD2 (WD = working distance). In practical terms, a microscope set at 400 mm receives half the fluorescent signal as a microscope set at 300 mm. Going to 200 mm doubles the light from 300 mm. With the typical surgical microscope, surgeons should be cautious at exceeding 300 mm when using ICG fluorescence. Understanding the limits of fluorescent-guided surgery – ICG vascular fluorescence-guided surgery has understandable limits. The NIR light will penetrate through tissue, but the thicker tissue or the greater density of the tissue may not allow it to properly penetrate. Some carotid arteries will fluoresce and others, due to plaque, will not.

Oncology Fluorescence-Guided Surgery of GBM (Glioblastoma Multiforme) Tumors This technique is based on the patient taking an oral dose of 5-ALA. The 5-ALA is metastasized into protoporphyrin IX (PpIX) selectively at the tumor site. When excited by the blue light, the PpIX will fluoresce at various shades of red. Pink and violet colors may show that tumor cells are mixed with healthy brain cells. The modifications to the standard surgical microscope are as follows: Excitation light – The fluorescence of the PpIX requires a heavy dose of blue light. Two 300 W Xe light sources are used on the FL400, which increase

54

M. Sturgis

Fig. 7 Effect of additional excitation light

the fluorescence emission brightness. As the fluorescence signal is critical to the resection, a stronger fluorescence signal clearly improves the overall situation. See Fig. 7. Emission spectrum – When PpIX is excited with the blue light as illustrated in Fig. 8, it will emit fluorescence between 480 and 730 nm. As blue light has a relatively short wavelength, it will penetrate the brain or tumor tissue ≤1.0 mm. Seeing the emission – As seen in Fig. 8, the red tumor fluorescence consists of much less energy than the blue excitation light needed to create the fluorescence. For the surgeon to see the fluorescence, a filter must be used to block the blue so the red is easy to see. However, unlike laboratory fluorescence,

Fig. 8 Emission of 5-ALA compared to normal vision

Design and Use of the Surgical Microscope in Fluorescence-Guided Surgery

55

Fig. 9 Optical emission filter allowing some blue and all red to reach surgeon

the surgeon still needs some light to be able to see his/her surgical instruments for the tumor resection. Therefore, the filter shown in Fig. 9 blocks roughly 90% of the blue light from reaching the surgeon but allows 100% of the red tumor fluorescence to reach the surgeon. This little bit of blue will appear bright in the surgeon’s final image. These excitation/emission filters were designed by Karl Storz of Germany. Documentation – Unlike the invisible vascular fluorescence, the fluorescence created by 5-ALA is best seen by the human eye. The human eye has excellent vision in this range. A special camera that alters internal video settings when the microscope is placed in the blue fluorescence mode is required, as normal white light settings do not provide good video. Summary – Figure 10 shows all the components as they would appear in a 5-ALA fluorescence surgical microscope. 5-ALA Fluorescence result – The fluorescence image shown in Fig. 11a shows a clear delineation between the tumor and the surrounding tissue. The nonfluorescence image shown in Fig. 11b illustrates how it is considerably more difficult to determine tumor from healthy brain tissue. Need for intra-operative pathology – The color differentiation among red, violet, and blue will lead the surgeon to understand the differences between tumor cells and brain cells. However, there are substantial differences in color presentation among different patients. The grade of tumor, 5-ALA uptake, dose, and a host of other factors lead the colors not to be perfect indicators when it comes to identifying healthy cells from tumor cells in different patients. Therefore, when first exposing a tumor to fluorescence, it is desirable to first take three to five biopsies and mark them according to color. In this way, when the biopsies return, the surgeon will have a good indication of tumor percents compared to the fluorescence emission color.

56

M. Sturgis

Fig. 10 Components of the FL400 oncology fluorescence surgical microscope

Fig. 11 (a) 5-ALA emission; (b) normal white light of tumor

Technical considerations related to the OR implementation of the oncology fluorescence stereomicroscopy– Initial use of these techniques in the operating room can be limited by the following technical factors: Timing of oral drug delivery – The drug will need to be given approximately 2 h before surgery, with adjustment of this time made to accommodate the time required to expose the GBM and your protocol. This can prove to be difficult in some situations.

Design and Use of the Surgical Microscope in Fluorescence-Guided Surgery

57

Intra-operative pathology – The OR will need to inform pathology to be ready for the biopsies so the surgeon receives the information before his/her resection is completed. Understanding the effect of working distance – Light loses energy at a rate of 1/WD2 (WD = working distance). In practical terms, a microscope set at 400 mm receives half the fluorescent signal of a microscope set at 300 mm. For example, going to 200 mm doubles the light from 300 mm. With the oncology fluorescence surgical microscope, surgeons should be cautious at varying the WD too much. Increasing WD by a 100 mm will cut the fluorescence emission roughly in half. Xe lamps – Xe lamps should be replaced after 150 h of use. This is in spite of the fact that manufacturers recommend Xe lamps be changed at 500 h. However, at that time frame, the Xe lamp has lost roughly half of its brightness. For normal white light vision, this may be acceptable, but the oncology fluorescence microscope will produce better results if lamps are kept below the 150 h time frame.

Design Considerations for Surgeons Interested in Designing Surgical Fluorescence Microscope Procedures Device regulatory issues – Fluorescence equipment is at the minimum a FDA Class 2 diagnostic device even when added to a surgical microscope. Those parts added to a microscope for fluorescence will require a 510(k) approval. If the microscope is used for a higher level care such as treatment, then the required approval level will be higher. Device approvals are generally not difficult to obtain since proving their function is not difficult. Pharmacological regulatory issues – Obtaining FDA approval for fluorescence agents that are presently used for other purposes should not be difficult. However, obtaining FDA approval for drugs presently not approved for other uses can be quite difficult and costly. Chromatic aberrations – Surgical microscopes are designed to bring light energy ranging from 400 to 720 nm (visible range) into focus at a single point. If a wavelength is shorter or longer than that range, it may require special optical corrections. For example, the FL800 ICG vascular fluorescence requires a different focal point for the camera and a set spot in the zoom for 830 nm light to come into perfect focus with the white light. Wavelength and tissue penetration – The 400 nm shorter excitation light described in this chapter and used in the Fl400 means that fluorescence cannot occur more than a short distance into tissue (≤1 mm). However, the 800 nm excitation light used on the FL800 vascular fluorescence makes it easy to penetrate some tissue up to 10 mm. Sensors – The world of CCD-type sensors is changing dramatically. Where we once hoped that a great CCD would display what we can see with our eyes,

58

M. Sturgis

we can now expect them to do more than our eyes can do. At present, NIR, IR, and selective wavelengths are easily within the reach of the clinical investigator. The newest black silicone technology could soon offer the scientific world sensors with one hundred times the energy sensitivity, which could greatly expand the usefulness of surgical fluorescence. Excitation light – Most fluorescence surgical microscopes, including the two mentioned here, get their excitation light from a filtered Xe standard light source. Although this is efficient from a manufacturing point of view, the photonics world is quickly changing to allow researchers to use laser diodes with very specific wavelength and energy control. Filter technology – Filter manufactures have also shown great improvement in their products during the last 10 years. Not only have narrow band filters become available, but very sharp cuts among different wavelengths present greater fluorescence detection opportunities. Image analysis computers – These computers have wide applications in pathology and in the medical industry, but have not made their way into the operating room. Their ability to quickly break apart image intensities and wavelengths make them an ideal addition to the surgical fluorescence microscope world. Photo dynamic therapy (PDT) – For PDT to become widespread in the surgical field, improvements in fluorescence technology must be improved. Over the years, a number of PDT studies have been done with variable results. However, they suffered from relatively poor light sources, imprecise filters, imprecise light delivery systems, and poor sensors. Today these obstacles are mostly overcome, but studies still need to be redesigned by medical researchers using the clearly better technology that the device industry can now provide.

Fluorophores for Optical Imaging Iain Johnson

Introduction Optical imaging using fluorescent probes, along with related biomedical imaging technologies such as PET, SPECT, MRI and bioluminescence, represents a return to in situ observational methods of biological discovery in contrast to the reductionist molecular level genomic and proteomic methods that have dominated biomedical research for the past 50 years. This chapter will delineate and compare the properties of fluorescent probes, which are the molecular constructs responsible for contrast generation in most fluorescence imaging applications. For this reason, fluorescence microscopy and other imaging technologies that can deliver this essential contextual information are of ever-increasing importance in biomedical research. This chapter will delineate and compare the properties of fluorescent probes, which are the molecular constructs responsible for contrast generation in most fluorescence imaging applications. To begin with, abbreviations are defined (Table 1) and some foundational terminology must be defined. Table 1 Abbreviations FITC = fluorescein isothiocyanate GFP = green fluorescent protein (usually refers to Aequoria victoria gene product GenBank accession number M62653, but sometimes also used more generically to refer to structurally and functionally homologous proteins from other species) ICG = indocyanine green MRI = magnetic resonance imaging NHS = N-hydroxysuccinimidyl PET = positron emission tomography R-PE = R-phycoerythrin SPECT = single photon emission computed tomography

I. Johnson (B) Life Technologies Corporation, 29851 Willow Creek Road, Eugene, OR 97402, USA e-mail: [email protected]

E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_4, 

59

60

I. Johnson

Definitions Fluorophore. A fluorophore is a molecule or an indivisible molecular assembly1 that emits light at specific wavelength in response to excitation by an external light source such as a laser or an incandescent lamp. In some cases, fluorophores exhibit usefully different spectroscopic properties (qv) at collective and individual levels. Fluorophores by themselves do not usually exhibit highly specific associations with molecular targets (qv fluorescent probe). For presentational purposes, we will subdivide fluorophores into three structurally distinct groups as follows: 1. Organic dyes (e.g., FITC, ICG) 2. Fluorescent proteins (e.g., GFP, R-PE) 3. Quantum dot nanocrystals (Qdots) Fluorescent probes. Fluorescent probes are exogenous molecules introduced into biological specimens for the purpose of generating image contrast. Structurally, fluorescent probes can be considered as binary molecular constructs consisting of targeting groups and fluorophores. Targeting groups are many and varied in terms of composition,2 but serve a single purpose from a functional standpoint. This purpose is to create a specific association between the probe and a molecular target within a complex and heterogeneous specimen. This specificity of association is the most crucial single aspect of the design and application of fluorescent probes. Without specificity of association, the information content of images in terms of fluorescence representing molecular distributions within the specimen is either degraded or entirely lost. Labeling. Labeling is the process of addition of the fluorescent probe to the specimen. The goal of the process is to create an unambiguous association between a detectable fluorescence signal and a molecular target within the specimen. Specifics of the labeling process are determined by the nature of the target tissue and will not be discussed in detail here. However, two general requirements must be fulfilled in all cases. (1) The probe must be delivered to the region of interest within the specimen and retained there for long enough to collect the experimental data. (2) The amount of probe delivered must be sufficient to achieve detection above background but not excessive to the extent of inducing perturbation of the structural and functional properties of the specimen. Autofluorescence. Fluorescence generated by molecules endogeneously present in the tissue is termed autofluorescence. Autofluorescence is addressed in a separate chapter within this text. In most cases, autofluorescence of cells and extracellular matrix emanates from an unresolvable heterogenous mixture of endogenous fluorophores.3 Because the concentration and spatial distribution of endogenous fluorophores cannot be experimentally controlled, exogenously introduced probes are used in the majority of fluorescence imaging applications. Autofluorescence is most commonly encountered as a background signal from which exogenous probe signals must be resolved to obtain high signal:background contrast.

Fluorophores for Optical Imaging

61

Mutiplex detection. Multiplex detection refers to the process of simultaneously labeling a specimen with two or more fluorescent probes to allow differentiation of multiple structural or functional features. As well as specific association with their targets, the probes must have distinctive spectroscopic properties that can be discriminated by the detection instrument. Most commonly, differences in the fluorescence emission spectrum are used for discrimination. Spectroscopic properties. Spectroscopic properties characterize the fluorescence excitation and emission of fluorophores. Spectroscopic properties relevant to practical fluorescence imaging that will be considered in this chapter are summarized in Table 2.

Operational Characteristics of Fluorescent Probes Having defined the necessary terminology, the remainder of this chapter will compare the three principal classes of fluorophore (qv) with respect to the operational characteristics listed in Table 3.

Excitation Spectrum and Extinction Coefficient Two primary spectroscopic considerations have to be made when selecting fluorescent probes, the first relating to the specimen and the second to the detection instrumentation. First, the excitation and emission spectra of the probe should be selected to coincide with the wavelength region in which the optical transmission of the specimen is maximal and its background autofluorescence is minimal. Based on this consideration, probes with excitation and emission spectra in the near infrared region (700–900 nm) are preferred for in vivo imaging applications. ICG is preferred over fluorescein as a tracer for retinal angiography primarily for this reason. Secondly, the instrument excitation wavelength output should be as close as possible to the peak of the fluorophore excitation spectrum. The fluorescence excitation spectra of organic dyes and fluorescent proteins consist of discontinuous peaks (e.g., Fig. 1a), requiring considerable attention be paid to this matching process in order to obtain efficient excitation. Excitation spectra of organic dye cohorts such as the Alexa Fluor series span the near ultraviolet, visible, and near infra-red regions (300–800 nm), whereas fluorescent proteins currently offer a more restricted range (∼400–600 nm). Quantum dot nanocrystals exhibit more continuous excitation profiles (Fig. 1b), facilitating multiplex detection schemes in which several probes with spectrally distinct emission profiles are excited at the same wavelength (Yezhelyev et al., 2007; Lin et al., 2007; Kobayashi et al., 2007). However, the general trend of decreasing optical transmission and increasing autofluorescence of tissues with decreasing wavelength to some extent counteracts the advantages gained by exciting quantum dot nanocrystals in the region <500 nm where their absorption is strongest

62

I. Johnson Table 2 Spectroscopic properties of fluorescent probes

Property

Definition

Significance

Fluorescence excitation spectrum

An X,Y plot of excitation wavelength versus number of fluorescence photons generated by a fluorophore

Absorption spectrum

An X,Y plot of wavelength versus absorbance of a chromophore or fluorophore

Fluorescence emission spectrum

An X,Y plot of emission wavelength versus number of fluorescence photons generated by a fluorophore

Extinction coefficient

Capacity for light absorption at specific wavelength∗

Fluorescence quantum yield

Number of fluorescence photons emitted per excitation photon absorbed Loss of fluorescence signal due to short-range interactions between the fluorophore and the local molecular environment, including other fluorophores (“self-quenching”) Destruction of the excited fluorophore due to photosensitized generation of reactive oxygen species (ROS), particularly singlet oxygen (1 O2 ∗ )

Optimum instrument setup should deliver excitation light as close to the peak of the excitation spectrum of the fluorophore as possible To a first approximation∗∗ , the absorption spectrum of a fluorophore is functionally equivalent to the fluorescence excitation spectrum Fluorescence emission spectral discrimination is the most straightforward basis for multiplex detection (qv) and for resolving probe fluorescence from background autofluorescence Fluorescence output per fluorophore (“brightness”) is proportional to the product of the extinction coefficient (at the operational excitation wavelength) and the fluorescence quantum yield See “extinction coefficient”

Quenching

Photobleaching

∗

Loss of fluorescence is reversible to the extent that separation of fluorophores can be altered in vivo

Results in irreversible loss of fluorescence signal from excessive exposure to excitation light

Extinction coefficient (ε; units: M–1 cm–1 ) is defined by the Beer–Lambert law A=ε.c.l where, A = absorbance, c = molar concentration, l = optical path length. From a functional point of view, the extinction coefficient of a fluorophore describes its capacity for light absorption in the same way that a specific activity describes the capacity of an enzyme to convert substrate to product. ∗∗ Generally true for single fluorophore species in homogenous solutions but not in more complex heterogenous samples.

Fluorophores for Optical Imaging

63

Table 3 Operational characteristics of fluorescent probes Excitation spectrum and extinction coefficient Emission spectrum Environment sensitivity and quenching Size Physical and chemical stability Conjugation and targeting Distribution and pharmacokinetics Toxicity Photostability

(Lim et al., 2003). Peak extinction coefficients of organic dyes and fluorescent proteins are generally in the range of 104 –105 M–1 cm–1 . The main exceptions to this stipulation are phycobiliproteins such as R-phycoerythrin (R-PE) that have multiple fluorescent prosthetic groups per protein monomer resulting in cumulative extinction coefficients of >106 M–1 cm–1 . Quantum dot nanocrystals exhibit extinction coefficients >106 M–1 cm–1 , approximately 100 times larger than the peak values of organic dyes (Fig. 1). This massively increased capacity for light absorption is the principle factor underlying the increased fluorescence output (brightness) of quantum dot nanocrystals.

Emission Spectrum For organic dyes and fluorescent proteins, the excitation and emission spectra are tightly coupled. In other words, the peak wavelength separation of the two spectra, referred to as the fluorescence Stokes shift, exhibits little variability (typically ∼20– 40 nm; Fig. 1a) among different fluorophores. From the user’s point of view, this means that if a fluorophore is selected based on matching the excitation spectrum to the instrument source wavelength, the emission wavelength window for fluorescence detection is then essentially selected by default. For quantum dot nanocrystals, the latitude in excitation wavelength selection discussed above means that excitation/emission wavelength pairings are less constrained. When combinations of fluorophores are being used for multiplex detection (defined above), the overlap between their emission spectra should be as small as possible to minimize the possibility of crosstalk and consequent misidentification of the signal origin. The three principle fluorophore families, organic dyes, fluorescent proteins (Snapp, 2005), and quantum dot nanocrystals all include variants with different emission spectra. Only organic dyes and quantum dot nanocrystals currently include fluorophores with emission in the near infrared region. The emission spectral bandshapes of quantum dot nanocrystals are narrower and more symmetrical than those of fluorescent proteins and organic dyes, providing better resolution (i.e., less inter-channel crosstalk) in multiplex detection applications (Fig. 1).

64

I. Johnson

Fig. 1 Absorption and fluorescence emission spectra of an organic dye (Alexa Fluor 594; a) and a quantum dot nanocrystal (Qdot 625; b). Absorption spectra are plotted in terms of extinction coefficient to illustrate the 100-fold larger values of this parameter typically exhibited by quantum dot nanocrystals relative to organic dyes. Comparison of the emission spectra illustrates the narrower bandwidths typical of quantum dot nanocrystals. In particular, Alexa Fluor 594 exhibits significant emission intensity at wavelengths >700 nm, whereas Qdot 625 exhibits essentially none

Fluorophores for Optical Imaging

65

Environment Sensitivity and Quenching Fluorescence output per fluorophore (“brightness”) is proportional to the product of the extinction coefficient (at the operational excitation wavelength) and the fluorescence quantum yield. Unlike the extinction coefficient, the quantum yield of a fluorophore is not wavelength dependent. Instead, the major operative variables are interactions of the fluorophore with the surrounding molecular environment.4 Fluorescence quenching (defined as loss of fluorescence signal due to short-range interactions between the fluorophore and the local molecular environment) is caused by an assortment of molecular mechanisms. The loss of fluorescence persists only for as long as the causative molecular interaction persists; and there is no permanent chemical or physical transformation of the fluorophore. By way of example, two rather ubiquitous quenching mechanisms will be briefly described. Photo-induced electron transfer (PET) is a process in which an electron is transferred from an oxidizable donor to an excited fluorophore, quenching the fluorescence of the latter. Examples of efficient electron donors are guanosine bases of nucleic acids and tryptophan residues of proteins. Although PET has a very short effective range (<1 nm), covalent coupling of fluorophores to protein or nucleic acid-targeting groups (see “Conjugation and Targeting”, below) can readily create the required proximity. Probes consisting of fluorophores attached close to guanosine bases or tryptophan residues will therefore generally exhibit lower fluorescence quantum yields than those with other nucleotide or amino acid contexts. Self-quenching is caused by the formation of fluorophore aggregates that have wavelength-shifted absorption spectra and drastically reduced fluorescence quantum yields relative to their constituent monomers. Self-quenching produces the counterintuitive result of fluorescence output decreasing as fluorophore concentration increases. Such concentration increases need only be at the microscopic level. Proteins and other biopolymers to which multiple fluorophores are attached provide scaffolds on which assembly of fluorophore aggregates can readily occur. Alexa Fluor dyes and Cy dyes (Fig. 2b) are designed to avoid self-quenching by incorporating negatively charged sulfonic acid substituent groups that produce mutual electrostatic repulsion and increased water solubility. However, in the case of indocyanine green (ICG),5 two sulfonic acid substituents (Fig. 2c) do not produce a sufficient increase in solubility to alleviate self-quenching of its fluorescence in concentrated aqueous solutions (Mordon et al., 1998). In addition to interactions with other molecules, various intramolecular mechanisms also produce environment-sensitive quantum yield variations. Protonation of the 3 -hydroxyl substituent of fluorescein (Fig. 2a) at pH ∼6.5 results in a 60% decrease of its quantum yield. Consequently, fluorescein is poorly suited for use in imaging probes that are targeted for internalization in acidic intracellular compartments (Koyama et al., 2007). In the case of cyanine dyes such as ICG and Cy5.5, intramolecular rotation around the central polymethine bridge (Fig. 2b, c) diminishes the fluorescence quantum yield. These dyes therefore tend to exhibit increased quantum yields in viscous solvents and when conjugated to biomolecular scaffolds that impede intramolecular flexibility.

66

I. Johnson

A HO

B

SO3

O

SO3

O

H3C

1

CH3

O 3S C O

OH

H3C

SO3

CH CH CH CH CH N

3

N

4

CH2

3

CH2

CH3

C O

CH2

O

CH2

N

O

CH3

O

CH2 CH2

2

C O

C

CH3

H3C

H3C

O

CH3 O

N

O

CH CH CH CH CH CH CH N

N

4

CH2 CH2

CH2

CH2

CH2

CH2 SO3

2

CH2

CH2

3

SO3

3

Fig. 2 Structural features of fluorescent dyes. a. 5-carboxyfluorescein, N-hydroxysuccinimidyl (NHS) ester; b. Cy5.5, NHS ester; c. indocyanine green (ICG). 1. Ionization of the 3 -hydroxyl substituent of fluorescein is responsible for the pH dependence of its spectroscopic properties in the physiological pH range. Cy5.5 and ICG lack analogous substituents and therefore do not exhibit comparable pH dependence. 2. N-hydroxysuccinimidyl ester functionality for coupling to amine-containing biomolecules (see Fig. 4a). 3. Sulfonic acid substituent conferring increased water solubility. 4. Conjugated polymethine bridge. The extension of this bridge from five carbons in Cy5.5 to seven carbons in ICG entirely accounts for the longer absorption/emission wavelength characteristics of ICG (absorption/emission spectral peaks are 675/694 nm and 780/815 nm for Cy5.5 and ICG, respectively)

Sensitivity to conditions results in a degree of unpredictability that is not conducive to functional reliability. Thus it is not surprising that in instances where fluorescent molecules have evolved in nature, fluorophores are usually shielded within a structurally defined protein matrix. GFP and phycobiliproteins provide excellent examples of this structural paradigm. This feature is emulated by quantum dot nanocrystals—their fluorescence quantum yield and other spectroscopic properties are generally insensitive to environmental conditions to the extent that the integrity of the passivation shell and polymer surface coating layers (Fig. 3) is maintained (Gao et al., 2004). Although the preceding discussion may give the impression that environmental sensitivity of fluorescent probes is an invariably negative trait, it is in fact a double-edged sword. On one hand, these effects are complicating factors in the relationship between the fluorescence signal intensity

Fluorophores for Optical Imaging

67

Fig. 3 Schematic representation of quantum dot nanocrystal structure comprising a semiconductor core (typically CdSe), a passivating shell (typically ZnS), a functionalized polymer coating, and multiple targeting groups such as streptavidin or antibodies

and probe concentration. On the other hand, calibrated and controlled environmentsensitive fluorescence responses can provide sensitive and selective readouts of biomolecular structure and function at the submicroscopic level.

Size The physical size of fluorescent probes impacts their solubility, pharmacokinetics (penetration into tissues and excretion), and biological activity. From all points of view, smaller is generally better. Organic dyes are generally planar molecules with longest dimensions of 1–2 nm6 and molecular weights in the order of 500–1500. As can be seen from a comparison of the structures of Cy5.5 and ICG in Fig. 2, increasing the excitation and emission wavelength ranges of a fluorophore is achieved by increasing its size.7 Thus a penalty in molecular size has to be paid to obtain the desirable objective of excitation/emission in the near infrared region. GFP is a small protein (molecular weight ∼ 27 kDa) consisting of 11 beta-strands forming a hollow cylinder approximately 4.2 nm in length by 2.4 nm in diameter that encapsulate the p-hydroxybenzylideneimidazolidinone fluorophore. Quantum dots typically consist of a spherical core of the semiconductor cadmium selenide (CdSe) surrounded by a zinc sulfide (ZnS) shell which is in turn surrounded by a hydrophilic polymer surface coating. As well as stabilizing the electronic excited state of the semiconductor, the ZnS shell also prevents the release of cytotoxic cadmium from the core. The hydrophilic coating confers water solubility and incorporates functional groups for crosslinking to antibodies, streptavidin, and other targeting groups. Overall, the trilaminar particles are about 4–10 nm in diameter, comparable in size to the phycobiliproteins. Attachment of targeting groups such as streptavidin or an antibody produces a total particle diameter of 10–20 nm (Fig. 3).

68

I. Johnson

Chemical and Physical Stability The physical and chemical stability of organic dyes, fluorescent proteins, and quantum dot nanocrystals is not a limiting factor in performing longitudinal imaging studies over periods of several days. The one major exception to this stipulation is instability induced by the imaging process itself (see “Photostability” below). Organic dyes in current use maintain their fluorescence under such harsh environmental conditions as lysosomal acidity and phagosomal oxidation. Green fluorescent protein is also a very stable molecule capable of withstanding extraction in organic solvents. Perhaps the major obstacle to long-term persistence of GFP fluorescence in vivo is a tendency for epigenetic silencing of transgene expression (Toth et al., 2007). Quantum dot nanocrystals remain intact, as assessed by size-exclusion chromatography, after intravenous administration and subsequent excretion in rats and mice (Choi et al., 2007). In view of their use in primarily an aqueous context, hydrolytic stability and aqueous solubility of fluorophores are of considerable practical importance. Most organic dye fluorophores are chemically stable in aqueous solutions of moderate pH for several days at minimum. However, aqueous solutions of activated fluorophore derivatives used for bioconjugation (Fig. 4) must be used within a few hours due to rapid hydrolysis of their NHS ester and maleimide functional groups. Once formed, the amide and thioether linkages (Fig. 4) are typically stable in aqueous solutions for several months.8

Fig. 4a. Schematic of NHS ester reaction with protein amine groups yielding a carboxamidelinked conjugate. The dye NHS ester derivative represented schematically here is typified by 5carboxyfluorescein, NHS ester shown in Fig. 2a. b. Schematic of maleimide reaction with protein sulfhydryl groups yielding a thioether-linked conjugate

Fluorophores for Optical Imaging

69

Physical stability of aqueous solutions of fluorophores usually presents more difficulties than chemical stability. In general, organic dyes (which are largely built on hydrocarbon sub-structures; Fig. 2) and quantum dot nanocrystals are poorly soluble in water.9 Derivatization with anionic sulfonic acid (Romieu et al., 2008) (Fig. 2) or carboxylate10 groups, inert polysaccharides (e.g., dextrans) or polyethylene glycol (PEG) substituents (Gao et al., 2004; Bentzen et al., 2005) is the most reliable synthetic strategy for improving aqueous solubility. Although increased aqueous solubility generally results in lower levels of nonspecific protein binding, these modifications occasionally facilitate additional binding modes such as electrostatic binding to basic proteins (e.g., nuclear histones) mediated by sulfonic acid substituents. Complexation with carriers provides either an alternative or adjunct to fluorophore derivatization for mitigating poor aqueous solubility. Materials that have been successfully employed for this purpose include cyclodextrins, nonionic surfactants (e.g., Cremophor EL), and serum proteins. A variety of solubility enhancing additives have been used to improve the stability of aqueous solutions of ICG used for intravenous injection11 (Alam et al., 2005; Maarek et al., 2001; Rajagopalan et al., 2000).

Conjugation and Targeting Coupling organic dyes to protein targeting groups such as antibodies is usually accomplished by reaction of succinimidyl (NHS) ester or isothiocyanate derivatives of the dye with the -amino groups of lysine side chains or N-terminal α-amino groups. Alternatively, maleimide or iodoactamide dye derivatives are reacted with free thiol groups of cysteine residues (Fig. 4). These labeling protocols are straightforward, requiring reaction of the dye with an aqueous (pH 7–9) protein solution for approximately 30 min, followed by purification of the dye–protein conjugate by gel filtration. Some degree of protein-specific optimization with respect to the dye:protein ratio is generally necessary. This is merely a reflection of the fact that proteins vary widely in size and in the number and distribution of reactive targets. Typically, no more than about 3–6 dyes can be attached per IgG antibody without self-quenching of fluorescence and/or inactivating the protein (e.g., binding site of the antibody). Since cysteine has a much lower incidence than lysine, thiol derivatization is more likely to afford opportunities for labeling at a single site (Backer et al., 2007; Lee et al., 2008). Fluorescent derivatization of peptides and other small molecule targeting groups presents a greater challenge because the fluorophore represents a larger fraction of the conjugate structure and therefore exerts more influence on its properties. Maintaining the target-binding affinity as close as possible to that of the unmodified ligand is usually the overriding objective. Factors relating to the fluorophore (size and electrostatic charge), the targeting group (position of fluorophore attachment), and the option of inserting a separating linker between them all impact the outcome in largely unpredictable ways. Optimal probe development therefore usually follows the path of synthesizing a panel of candidates with variations of these

70

I. Johnson

factors that are then screened for binding to the target either in vivo or in a surrogate in vitro model system (Achilefu et al., 2002). Due to the ubiquitous presence of cysteine and lysine residues, labeling of a specific protein can only be attained using purified samples and in situ applications are largely precluded. This limitation has spurred the recent development of “click” chemistry-coupling methods using bioorthogonal azide and alkyne functional groups (Laughlin et al., 2008). Compared to post-synthetic modification with organic dyes, GFP tagging of proteins is more precise because the point of attachment to the fusion partner and fluorophore:protein ratio are genetically prescribed. GFP tagging of proteins utilizes standard molecular biology techniques for construction, transfection, and expression of vectors (Snapp, 2005). To avoid structural perturbation, GFP is almost always ligated to the N- or C-terminus of the fusion partner. The p-hydroxybenzylideneimidazolidinone fluorophore of GFP is formed by a post-translational oxidation reaction. Consequently, in hypoxic conditions, GFP fluorescence may underrepresent the amount of protein expressed (Coralli et al., 2001). Current methods for coupling quantum dot nanocrystals to antibodies use a combination of the maleimide and NHS ester chemistries described above, assembled around the heterobifunctional crosslinker SMCC (Fig. 5). The resulting conjugate is multivalent, comprised of multiple antibodies per fluorophore rather than the multiple fluorophores per antibody stoichiometry of organic dye conjugates. The functional consequences of multivalency potentially include increased targetbinding affinity (avidity effect) and target-crosslinking artifacts (Clarke et al., 2008; Howarth et al., 2008).

Distribution and Pharmacokinetics Fluorophores. Optimizing the distribution and pharmacokinetics of fluorescent probes entails striking a balance between competing operational requirements. For probes delivered by intravenous injection (the most common method of administration for in vivo imaging), the probe must remain in circulation for long enough to allow detectable accumulation at the target (e.g., a tumor). How long is long enough depends primarily on the probe–target-binding affinity and the rate of probe transport from the vasculature to the target. On the other hand, eventual excretion of the probe limits the duration of toxic side effects and reduces off-target background signals. From a design standpoint, the distribution of a fluorescent probe within the specimen should be controlled by the properties of the targeting group rather than the fluorophore itself. This expected outcome can be compromised in cases where the size of the fluorophore is similar to that of the targeting group or multiple fluorophores are attached to a large targeting group. For example, the dye:protein ratio of antibody conjugates significantly influences their in vivo distribution (Schellenberger et al., 2004; Pèlegrin et al., 1991; Tadatsu et al., 2006). Instances of GFP tagging causing aberrant localization of fusion partner proteins

Fluorophores for Optical Imaging

71 O N

NH2-

-NH2

N O C

O N

-N H

C

N

N O

N H

2-

2

NH

-NH2

NH 2

2-

-NH 2 O

-NH 2

2-

O C

-N H

-

H2

H2 -N H2 -N

-NH2

-NH 2

N H

2

NH

O NH 2

O

2

-NH 2

NH

2

NH

-N H

NH 2

H2 -N H2 -N

2 NH

H2 -N

2-

O

-NH 2

NH 2-

-NH2

NH

CH2 N

O

O

NH 2

-NH 2

NH

2-

2-

NH 2

O

-NH 2 -N H

2

-N H

O

H2 -N

H2 N

NH

2

-NH2

O C

SMCC

NH 2

NH 2-

H2 N

NH 2

-NH 2 -N H

O

+ H HS -S S -S

-S S

-

DTT

-

-SS-

HS HS-

-SH

HS-

-SH

-SS-

=

-S H

CH2

S -S

-S S

-

-

-SS-

S

HS-

O N

2

NH

NH 2

NH 2

2-

-

-

-SS-

S

-SH

N O

H2 -N H2 -N

S -S

-S S

-

2

-NH2

S-

2 NH

-NH 2

-S

O C

-NH 2 O -N H

H 2O N

N H

O

-SS-

NH

NH

2-

S

HS-

C

NH

N

S -S

H2 -N -NH 2

2-

O

2

-N H

H2 N

NH 2

NH

-NH2

O C

-NH 2 -N H

O

Fig. 5 Schematic of antibody–quantum dot nanocrystal conjugation via hinge reduction chemistry. Quantum dot nanocrystals with amine-functionalized surfaces are first coupled to the crosslinker succinimidyl trans-4-(maleimidylmethyl) cyclohexane-1-carboxylate (SMCC) through the same reactive chemistry as shown in Fig. 4a. Reduction of antibody disulfide bridges by treatment with dithiothreitol (DTT) generates free sulfhydryl groups which are then reacted with the maleimides introduced by SMCC through the same reactive chemistry as shown in Fig. 4b

72

I. Johnson

are generally sporadic. When they do occur, such problems are often the result of overexpressing fusion proteins in order to increase fluorescence signal levels. Quantum Dots. Because quantum dot nanocrystals are much larger than organic dyes and GFP, their tendency to exert a dominant effect on probe distribution properties is greater. This has not prevented the successful application of quantum dot nanocrystal–antibody conjugates in in vivo imaging applications (Tada et al., 2007). Intravenous injection of quantum dot nanocrystals without targeting groups results in rapid and persistent accumulation in the kidney, liver, and spleen (Schipper et al., 2007; Yang et al., 2007). The lack of clearance is at least in part attributable to the size of current generation quantum dot nanocrystals (Fig. 3) which is above the threshold for renal filtration and urinary excretion (Choi et al., 2007), providing a major impetus for ongoing efforts to engineer smaller particles (Howarth et al., 2008). Non-Targeted Dyes. The transport and metabolism pathways of non-targeted organic dye tracers such as ICG and fluorescein are reminiscent of circulating drugs. The fluorescence/concentration relationships of both dyes are impacted. Intravenously injected ICG rapidly becomes almost entirely bound to plasma proteins. This prolongs its retention in the vasculature compared to fluorescein, which exhibits rapid extravasation with accompanying tissue necrosis. Binding of ICG to plasma proteins results in a 30-nm shift of its absorption spectrum (peak shift 780– 810 nm). ICG is eliminated by excretion into the bile, with a circulating half-life of about 3 min in human subjects. Intravenously injected fluorescein is rapidly converted to fluorescein monoglucuronide in the liver. Glucuronidation has essentially the same effect on the spectroscopic properties of fluorescein as protonation – the absorption peak shifts from 490 to 454 nm and the peak extinction coefficient and fluorescence quantum yield both decrease by ∼70%.

Toxicity Fluorescent probes can exert both chemical and photochemical toxic effects. The systemic toxicity of fluorescent dyes is rather low (Lutty, 1978). The intravenous LD50 value (in mice) is 60–80 mg/kg for ICG and 300 mg/kg for fluorescein, compared to typical administered doses for in vivo imaging procedures of 0.05–0.5 mg/kg. However, it is not uncommon for organic dyes to become intracellularly compartmentalized, resulting in microscopic concentrations >0.1 mM with accompanying localized toxic effects. For example, cationic cyanine dyes readily accumulate in mitochondria and cause respiratory inhibition (Anderson et al., 1993). The addition of sulfonic acid substituents (Fig. 2b, c) counteracts this tendency. This basic paradigm of minimal systemic impacts but context-dependent functional and structural perturbations at cellular level, particularly where selective uptake results in elevated local concentrations of probes, is recapitulated by both fluorescent protein and quantum dot nanocrystal probes. Incidences of pathological abnormalities directly attributable to GFP expression (Guo et al., 2007)

Fluorophores for Optical Imaging

73

are sporadic and vastly outnumbered by viable and phenotypically normal GFP transgenic animals (Megason et al., 2006). Levels of EGFP in mouse retinal cells exceeding 0.1 mM (determined by quantitative confocal microscopy) have been found to have no discernible effect on cell morphology and function (Rex et al., 2005). In terms of in vivo imaging applications, immune reactions evoked by GFP appear to be more prevalent than toxic effects (Stripecke et al., 1999; Steinbauer et al., 2003). In vivo assessments of quantum dot nanocrystal toxicity are still at the preliminary stage as refinements and modifications of the materials themselves continue to be developed (Maysinger et al., 2007). Particle size, electrostatic charge, concentration, surface coating integrity, and oxidative decomposition are among the many potential contributing factors (Hardman 2006). Release of cadmium from the core12 is inhibited by the ZnS passivation shell (Fig. 3) and further restricted by surface coating layers (Derfus et al., 2004). Reported perturbation effects of quantum dot nanocrystals in vitro include induction of autophagy in human mesenchymal stem cells (Selevesterov et al., 2006) and proinflammatory cytokine release in primary human epidermal keratinocytes (Ryman-Rasmussen et al., 2007). Phototoxicity is a more pervasive problem than chemical toxicity because its causative mechanism involves not only the fluorophore but also excitation light and oxygen (Fig. 6). Furthermore, endogenous fluorophores as well as exogenous

1

1. hνEX 2. ISC

1Dye

3Dye*

2

CH

3O

1Dye

2

NH3

Cys

+

1O

2*

NH3

CH CH2S SCH2 CH

1O* 2

NH3

+

C O

C O

O

O

CH2SH

C O

NH3

H2O

O

CH CH2SO3H C O O

O

O

3

CH

1O * 2

O

O

CH CH CH O

CH CH

+O

N

N

N

N

CH3

CH3

CH3

CH3

Fig. 6 Mechanistic commonality of phototoxicity and photobleaching. Both phenomena are initiated by photosensitized generation of highly reactive singlet oxygen (1 O2 ∗ ). The controlling factors in this process (1) are the dye (fluorophore), oxygen (3 O2 ), and excitation light (hνEX ). Phototoxicity (2) results from indiscriminate reaction of 1 O2 ∗ with proteins, nucleic acids, and lipids. In the example shown here, cysteine is converted to cystine or cysteine sulfonic acid. Photobleaching (3) results from reaction of 1 O2 ∗ with a fluorophore resulting in disruption of the conjugated πelectron system that is essential for absorption and fluorescence emission (Byers et al., 1976). The example shown here represents the carbocyanine fluorophore DiOC1 (3)

74

I. Johnson

fluorescent probes can contribute to photosensitized generation of singlet oxygen (1 O2 ∗ ), Thus fluorescence excitation in the near infra-red region, where endogenous fluorophores are less abundant, is beneficial from the point of view of minimizing phototoxicity as well as autofluorescence.

Photostability Photobleaching is a complex process in which a fluorophore can participate twice – once as a catalyst and once as substrate (Fig. 6). Furthermore, photobleaching rates are also modulated by environmental factors such as aggregation in the case of cyanine dyes (Byers et al., 1976). Photobleaching is an irreversible process.13 The only way to restore signal to a specimen that has incurred photobleaching is to supply fresh fluorophore. Although some organic dyes and fluorescent proteins are intrinsically more resistant to photobleaching than others, the differences are generally not large enough to justify interchanging fluorophores for this reason alone. Quantum dot nanocrystals are the principle exception to this stipulation. Their inorganic core composition and polymer coatings render them largely immune to the reactions shown in Fig. 6 and thus clearly superior to organic dyes in experiments involving long periods of continuous excitation (Le Gac et al., 2006). In general, the best course for controlling photobleaching is to minimize the intensity and duration of excitation, particularly in vivo where the remedy of oxygen scavenging treatments often employed in immunofluorescence microscopy is not applicable.

Conclusions and Prospectus Even from the brief glimpse afforded by this chapter, it can be seen that optimizing the performance of fluorescent probes in vivo involves consideration of a large number of variables. This complexity can appear daunting to a newcomer to the field. On the other hand, once mastered, this same apparent complexity becomes enabling, providing capacities for multidimensional interrogation of biological specimens. These capacities go far beyond those introduced here. Applications of other fluorescence attributes such excited state lifetimes, fluorescence resonance energy transfer (FRET), and multiphoton excitation and the use of light as a spatial or temporal selection agent (photoactivation) are well documented elsewhere (Akers et al., 2007; Jares-Erijman and Jovin, 2003; Dunn and Sutton, 2008; Patterson and LippincottSchwartz, 2008). Despite its multidimensional capacities, fluorescence detection does not encompass the full range of molecular and cellular identification and temporal and spatial resolution required for in vivo imaging of biological processes. Consequently there is an impetus toward development of probes combining fluorophores with PET or MRI contrast agents and integrated multimode instrument platforms to support them (Cherry, 2006; Culver et al., 2008).

Fluorophores for Optical Imaging

75

Notes 1. That is, if the molecular assembly is dismantled, the property of fluorescence is lost. 2. Examples of targeting groups include antibodies for recognition of specific protein targets, oligonucleotides (for recognition of specific DNA sequences via in situ hybridization), and proteins expressed in tandem fusions with GFP and other fluorescent proteins. 3. Autofluorescence background may also contain contributions from compounds supplied in cell culture media (e.g., riboflavin) or animal diets (e.g., chlorophyll). 4. Under optimal environmental conditions, the fluorescence quantum yields of almost all fluorophores in current use as biological probes, including organic dyes, fluorescent proteins, and quantum dot nanocrystals fall within a range spanning only one order of magnitude (0.l–1.0). 5. ICG is FDA approved for injection in humans. Diagnostic applications include retinal angiography, assessment of cardiac output and liver function, and estimation of plasma volume. 6. The fluorescein molecule is about 1 nm across the long (horizontal) axis depicted in Fig. 2a. 7. Specifically, increases in wavelength are the result of increasing the length of the conjugated π-electron system of the fluorophore. 8. Commercial preparations of dye-labeled antibodies are typically provided as 1 mg/mL solutions in phosphate-buffered saline and are stable for at least 3 months when stored refrigerated in this condition. 9. For the purposes of this discussion, “poorly soluble” means a limiting solubility of <0.1 mM (or 0.1 mg/ml for a compound with molecular weight =1000). 10. Carboxylates are only fully anionic at pH>5. 11. Typical formulations are 5 mg/ml (∼6 mM) ICG in sterile water. 12. Cadmium release is of particular concern in view of the propensity of quantum dots to accumulate in the kidney and liver, which are primary targets of cadmium toxicity. 13. Although reversible photobleaching is a known process for some fluorophores (e.g., cyanine dyes), it is usually only observable under single-molecule detection conditions.

References Achilefu S, Jimenez HN, Dorshow RB, Bugaj JE, Webb EG, Wilhelm RR, Rajagopalan R, Johler J, Erion JL. (2002) J Med Chem. 45:2003–2015. Akers W, Lesage F, Holten D, Achilefu S. (2007) Mol Imaging. 6:237–246. Alam A, Chavan AJ, Flower RW. (2005) US Patent 6,944,493. Anderson WM, Wood JM, Anderson AC. (1993) Biochem Pharmacol. 45:2115–2122. Backer MV, Levashova Z, Patel V, Jehning BT, Claffey K, Blankenberg FG, Backer JM. (2007) Nat Med. 13:504–509.

76

I. Johnson

Bentzen EL, Tomlinson ID, Mason J, Gresch P, Warnement MR, Wright D, Sanders-Bush E, Blakely R, Rosenthal SJ. (2005) Bioconjug Chem. 16:1488–1494. Byers GW, Gross S, Henrichs PM. (1976) Photochem Photobiol. 23:37–43. Cherry SR. (2006) Annu. Rev. Biomed. Eng. 8:35–62. Choi HS, Liu W, Misra P, Tanaka E, Zimmer JP, Itty Ipe B, Bawendi MG, Frangioni JV. (2007) Nat Biotechnol. 25:1165–1170. Clarke SJ, Hollmann CA, Aldaye FA, Nadeau JL. (2008) Bioconjug Chem. 19:562–568. Coralli C, Cemazar M, Kanthou C, Tozer GM, Dachs GU. (2001) Cancer Res. 61:4784–4790. Culver J, Akers W, Achilefu S. (2008) J Nucl Med. 49:169–172. Derfus AM, Chan WC, Bhatia SN. (2004) Nano Lett. 4:11–18. Dunn KW, Sutton TA. (2008) ILAR J.49:66–77. Gao X, Cui Y, Levenson RM, Chung LW, Nie S. (2004) Nat Biotechnol. 22:969–976. Guo JK, Cheng EC, Wang L, Swenson ES, Ardito TA, Kashgarian M, Cantley LG, Krause DS. (2007) Transgenic Res. 16:829–834. Hardman R. (2006) Environ Health Perspect. 114:165–172. Howarth M, Liu W, Puthenveetil S, Zheng Y, Marshall LF, Schmidt MM, Wittrup KD, Bawendi MG, Ting AY. (2008) Nat Methods 5:397–399. Jares-Erijman EA, Jovin TM. (2003) Nat Biotechnol. 21:1387–1395. Kobayashi H, Hama Y, Koyama Y, Barrett T, Regino CA, Urano Y, Choyke PL. (2007) Nano Lett. 7:1711–1716. Koyama Y, Hama Y, Urano Y, Nguyen, DM, Choyke PL, Kobayashi H. (2007) Clin Cancer Res. 13:2936–2945. Laughlin ST, Baskin JM, Amacher SL, Bertozzi CR. (2008) Science 320:664–667. Le Gac S, Vermes I, van den Berg A. (2006) Nano Lett. 6:1863–1869. Lee SB, Hassan M, Fisher R, Chertov O, Chernomordik V, Kramer-Marek G, Gandjbakhche A, Capala J. (2008) Clin Cancer Res. 14:3840–3849. Lim YT, Kim S, Nakayama A, Stott NE, Bawendi MG, Frangioni JV. (2003) Mol Imaging 2:50–64. Lin S, Xie X, Patel MR, Yang YH, Li Z, Cao F, Gheysens O, Zhang Y, Gambhir SS, Rao JH, Wu JC. (2007) BMC Biotechnol 7:67. Lutty GA. (1978) Toxicol Appl Pharmacol. 44:225–249. Maarek JM, Holschneider DP, Harimoto J. (2001) J Photochem Photobiol B. 65:157–164. Maysinger D, Behrendt M, Lalancette-Hébert M, Kriz J. (2007) Nano Lett. 7:2513–2520. Megason S, Amsterdam A, Hopkins N, Lin S. (2006) Green Fluorescent Protein: Properties, Applications and Protocols, 2nd Edn (Chalfie M, Kain SR, Eds.) Wiley-Liss, Wilmington, DE, pp. 285–303. Mordon S, Devoisselle JM, Soulie-Begu S, Desmettre T. (1998) Microvasc Res. 55:146–152. Patterson GH, Lippincott-Schwartz J. (2008) Methods Cell Biol. 85:45–61. Pèlegrin A, Folli S, Buchegger F, Mach JP, Wagnières G, van den Bergh H. (1991) Cancer 67:2529–2537. Rajagopalan R, Uetrecht P, Bugaj JE, Achilefu SA, Dorshow RB. (2000) Photochem Photobiol. 71:347–350. Rex TS, Peet JA, Surace EM, Calvert PD, Nikonov SS, Lyubarsky AL, Bendo E, Hughes T, Pugh EN Jr, Bennett J. (2005) Mol Vis. 11:1236–1245. Romieu A, Brossard D, Hamon M, Outaabout H, Portal C, Renard PY. (2008) Bioconjug Chem. 19:279–289. Ryman-Rasmussen JP, Riviere JE, Monteiro-Riviere NA. (2007) J Invest Dermatol. 127:143–153. Schellenberger EA, Weissleder R, Josephson L. (2004) Chembiochem. 5:271–274. Schipper ML, Cheng Z, Lee SW, Bentolila LA, Iyer G, Rao J, Chen X, Wu AM, Weiss S, Gambhir SS. (2007) J Nucl Med. 48:1511–1518. Seleverstov O, Zabirnyk O, Zscharnack M, Bulavina L, Nowicki M, Heinrich JM, Yezhelyev M, Emmrich F, O’Regan R, Bader A. (2006) Nano Lett. 6:2826–2832.

Fluorophores for Optical Imaging

77

Snapp E. (2005) Current Protocols in Cell Biology, John Wiley & Sons, Inc., New York, pp. 21.4.1– 21.4.13. Steinbauer M, Guba M, Cernaianu G, Köhl G, Cetto M, Kunz-Schughart LA, Geissler EK, Falk W, Jauch KW. (2003) Clin Exp Metastasis. 20:135–141. Stripecke R, Carmen Villacres M, Skelton D, Satake N, Halene S, Kohn D. (1999) Gene Ther. 6:1305–1312. Tada H, Higuchi H, Wanatabe TM, Ohuchi N. (2007) Cancer Res. 67:1138–1144. Tadatsu Y, Muguruma N, Ito S, Tadatsu M, Kusaka Y, Okamoto K, Imoto Y, Taue H, Sano S, Nagao Y. (2006) J Med Invest. 53:52–60. Toth ZE, Shahar T, Leker R, Szalayova I, Bratincsák A, Key S, Lonyai A, Németh K, Mezey E. (2007) Exp Cell Res. 313:1943–1950. Yang RS, Chang LW, Wu JP, Tsai MH, Wang HJ, Kuo YC, Yeh TK, Yang CS, Lin P. (2007) Environ Health Perspect. 115:1339–1343 Yezhelyev MV, Al-Hajj A, Morris C, Marcus AI, Liu T, Lewis M, Cohen C, Zrazhevskiy P, Simons JW, Rogatko A, Nie S, Gao X, O’Regan RM. (2007) Adv Mater. 19:3146–3151.

Overview of Cancer Detection and Monitoring Strategies Kurt R. Zinn

Introduction There is tremendous potential for patients to benefit from recent advances in optical technologies and molecular imaging approaches. While these clinical applications are just beginning, the data from preclinical studies indicate significant improvements in early detection of cancerous growth and metastasis by various optical-based techniques. In addition, it is now possible to monitor biological processes such as apoptosis activation or protease activity (e.g., cathepsin D) in cancerous tissue in real time and to repeatedly monitor changes of these processes as an index of effective therapy. All optical imaging requires contrast, either endogenous to the tissue or via administration of an exogenous agent. The latter may be a chemical or genetic contrast agent that leads to specific signal in the target tissue. The exogenous contrast agents leading to specific signal include small molecules, peptides, antibodies, engineered antibody or receptor fragments, viral vectors, and nanoparticles. An optical contrast agent is often referred to as a fluorescent probe or fluorophore, while additional names are being coined for more specific and later imaging approaches, as presented in Table 1. The term “imaging biomarker” refers to imaging surrogates that may replace traditional endpoints for clinical trials and is a general reference that applies to all imaging modalities, not just optical imaging. Research to improve the sensitivity and specificity of optical imaging for cancer detection and monitoring includes the following areas: improvements in image acquisition and processing, probe targeting, probe activation, and use of genetic reporters that are more specific for cancer. The success of tumor targeting requires providing contrast to image tumor location for detection or monitoring of disease. The goal is to increase sensitivity (fluorescence intensity) over background. Target identification: All optical imaging applications require contrast to discriminate cancerous from normal tissue. Endogenous targets from within the cancer can be probed by autofluorescence methods and thereby produce contrast. Differences

K.R. Zinn (B) Department of Radiology, 1530 3rd Avenue South, Boshell Building, Birmingham, AL 35294-0012, USA

E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_5, 

81

Proteins or other scaffolds labeled with a fluorophore and quencher (or two fluorophores) that are in close proximity Chemical compound

Proteolytic beacon

Nanoprobe double labeled for dual modality imaging

DNA or RNA, alone, or with virus, or within cell

GFP = green fluorescent protein, RFP = red fluorescent protein, mCherry = improved RFP USPIO labeled with Cy5.5

Genetic reporter, genetic contrast agent

Dual labeled probe

USPIO = ultra-small iron particle, may also be composed of gold, carbon, or other element

USPIO

Nanoprobe, nanoparticle, nanorod

5-aminolevulinic acid

Stem-loop oligonucleotide labeled with a fluorophore and quencher

Molecular beacon

Optical contrast agent requiring activation

Excitation light for fluorophore absorbs is not visible on the emission side, there are many fluorophores that have different wavelengths of emission, including the near infrared Activation when fluorophore is separated from quencher due to specific DNA or RNA binding and elimination of stem-loop structure Activation when the fluorophore and quencher (or two fluorophores) are separated due to a specific proteolytic cleavage Metabolism leads to incorporation of fluorescent porphyrin in cancer cells Elemental composition may provide MR contrast; fluorophore for optical imaging may be attached to nanoparticle, additional coating or composition may allow for increased circulation in vivo, or targeting potential Expression of fluorescent reporter protein after delivery of the DNA or RNA to the target cells MR contrast with optical fluorophore

Chemical compound or structure that absorbs at one wavelength and emits at another

Cy5.5; indocyanine green (ICG), FITC; Quantum dot (semiconductor, Cd, Se) broad excitation but narrow emissions at selectable wavelengths

Fluorescent probe, fluorophore

Basis of contrast

Probe composition

Example

Term

Table 1 Contrast terminology for fluorescence imaging

82 K.R. Zinn

Overview of Cancer Detection and Monitoring Strategies

83

in autofluorescence in cancerous tissue relative to normal tissue can be explained by structural and metabolic changes associated with cancer. The chapter in this book entitled “The Application of Tissue Autofluorescence in Detection and Management of Oral Cancer and Premalignant Lesions” (Poh et al.,) provides details on the use of endogenous contrast for optical imaging of cancer, including clinical studies. With respect to exogenous contrast for optical imaging, there are various strategies to identify targets and screen probes that bind to them. The target might be identified following analyses of cancer specimens by genomic, proteonomic, or histological analyses. The selected target can be used in high throughput in vitro screening assays, where thousands of chemical molecules can be screened to identify binding or inhibitory properties. A similar process is applied to develop a therapeutic biological agent. In the past, drug companies and/or academic scientists were responsible for target identification out of motivations to develop therapeutic agents. Certain therapeutic drugs or biological agents can be exploited as imaging probes, but often modifications are necessary as the ideal properties for therapeutically targeted molecules are not the same as those designed for imaging. Regardless of the selection method, the desired properties of a target for optical imaging include a high concentration in cancerous tissue relative to normal tissue and a capacity to interact with the imaging probe in a manner that leads to signal amplification. The target might be an enzyme that can either metabolically trap the probe or activate the probe. In this manner one target protein could interact with numerous probe molecules. Similarly, the target might be a cancer-specific receptor that could bind and internalize the probe, while new receptors are continuously produced and returned to the surface for binding more probe. The target and imaging probe must have a capacity to interact.

Contrast Mechanisms and Properties of an Ideal Imaging Probe Optical imaging probes are diagnostic tools that enable detection of tumors or their metabolic processes from normal tissue. There are four mechanisms to generate fluorescent contrast in tumor, namely (1) by blood or lymph flow, (2) binding and retention, (3) activation, and (4) genetic. The first three mechanisms require administration of a probe, while the fourth mechanism requires delivery of DNA or RNA that encodes for production of a fluorescent protein reporter. Figure 1 summarizes how contrast is achieved for each of these basic mechanisms and also provides an example for each mechanism. Blood or Lymph Flow. Visualization of blood flow or lymphatic drainage allows for cancer detection or nodal staging. This method capitalizes on the high blood flow of tumors, similar to the method that computed tomography used iodinated contrast to detect tumors. This accomplished when the contrast agent is delivered in a controlled manner and cancer is identified because of differences that are detectable from the normal anatomy. The use of fluorescence imaging to detect tumors via lymphatic drainage is reviewed in the chapter entitled “Nodal Staging of

84

K.R. Zinn

a

Intravenous Injection of ICG Intratumor Injection of ICG

Tumor

Lymph Node

Tumor Lymph Node

Excitation

Emission Time

Fig. 1 Mechanisms to achieve contrast for fluorescence imaging. (a) Blood flow or lymph flow after injection of fluorophore. The example shows injection of the ICG fluorophore either intravenously or intratumor. (b) Binding and retention. The example shows injection of a tumortargeting antibody with attached fluorophore. (c) Activation. The example shows activation of a quenched fluorophore by cathepsin D. (d) Genetic. The example shows expression of red fluorescent protein (RFP) following delivery with an adenoviral (Ad) vector

Overview of Cancer Detection and Monitoring Strategies

85

b Antibody Targeting Tumor Receptor Fluorophore

Receptors with increased expression on cancer cells

Excitation

Emission

Fig. 1 (continued)

Cancer Using Diagnostic, NIR Fluorescence Imaging Techniques,” by Eva SevickMuraca. Binding and Retention. If the probe achieves contrast by selectively binding to tumor followed by retention, then rapid clearance from blood is ideal, since it leads to improved signal-to-noise ratio, especially if tumor-bound probe was not cleared.

86

K.R. Zinn

c

Quenched Fluorophore

Cathepsin-D released from cancer cells

Excitation

No Emission Due to Quenching

Cleavage Excitation

Emission

Fig. 1 (continued)

A high concentration of unbound probe in blood could prevent small lesions from becoming visible, especially for surface reflectance imaging. Retention in tumor following binding is an important requirement for imaging, as a fast metabolic halflife of the fluorescent probe leading to excretion would lead to decreased sensitivity for tumor detection.

Overview of Cancer Detection and Monitoring Strategies

87

d Intravenous Injection

Ad-CMV-RFP

infection

Intratumor Injection

Excitation

RFP = red fluorescent protein

CMV

RFP

transcription

translation Emission

Fig. 1 (continued)

Activation. Probe activation to achieve fluorescence might be related to pH, or release of Ca2+ , or it may include a metabolic step, whereby an enzymatic process converts a non-fluorescent probe into a fluorescent one. For example, 5aminolevulinic acid is non-fluorescent, but administration of the compound leads to accumulation of fluorescent porphyrins in tumors. This technique has the benefit of using enzymatic amplification to increase imaging signal. More commonly, the activation is related to the release of a quenched fluorophore. McIntire et al., include a detailed discussion of this mechanism in a chapter of this book entitled

88

K.R. Zinn

“Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy.” Another mechanism of activation is by binding of the fluorophore to a specific protein sequence, for example, binding of biarsenical dyes to genetically encoded tetracysteine motifs, as described by Martin et al. (2005). Genetic. Genetic fluorescence is accomplished by using a genetically encoded reporter that leads to expression of a fluorescent protein or other optical reporter such as luciferase. Fluorescent proteins span the visible spectrum, from cyan to red (reviewed by Shaner et al. (2005) and Hoffman (2005)) and are constantly being improved in terms of brightness and shift to the near infrared for improved tissue penetration (Shaner et al., 2008; Shcherbo et al., 2007). The genetic code for the fluorescent protein is typically encoded in a plasmid or virus (e.g., adenovirus, adeno-associated virus or AAV, or lentivirus) or other vector as described in Fig. 2. Plasmids and lentiviral vectors are commonly employed to transfect cells to produce various reporter proteins for in vitro studies. However, implementation into clinical practice requires significantly more work and is an endeavor of researchers in the gene therapy field.

Agent AAV, Lentivirus

Reverse transcription

Integration Or Episomal

Imaging reporter

Expression of Reporter

Cell DNA

Bacteria

Plasmid Adenovirus

nanoparticle

Fig. 2 A summary of methods to deliver genetic reporters to tumor cells. Figure is a modification with permission, originally prepared by Christine Herrmann, PhD, Baylor College of Medicine, from “The Reproductive Life Cycle of a Retrovirus-HIV” for “Introduction to Viruses” on BioEd Online 2008

It is possible to use a viral vector to induce the expression of fluorescent reporters in cancer in vivo. An example is presented in Fig. 3, showing a red fluorescent protein (RFP) induced by an adenoviral (Ad) vector in a green fluorescent tumor. The green fluorescence resulted from transfection of the cell line with a plasmidencoding green fluorescent protein (GFP).

Overview of Cancer Detection and Monitoring Strategies

A.Bright Field

B. RFP

C. GFP

D. Fused

89

Fig. 3 Imaging expression of RFP in a GFP-positive prostate tumor xenograft in a mouse. The Ad-RFP was injected intratumor

In this example, the vector (Ad-RFP) was directly injected into the tumor, and reporter fluorescence was induced when the cancer cells became infected. This particular Ad vector was replication competent and was therefore able to spread within the tumor causing the tumor to be killed. The final chapter of this section entitled “Illustrating Molecular Events with Light: A Perspective on Optical Reporter Genes” by Pritha Ray provides extensive details on how genetic reporters are used to generate contrast. Imaging probes using the first three mechanisms described in the preceding paragraphs must have access to the target and appropriate circulating half-life (if by blood delivery) for effective imaging contrast. The probe should be non-toxic or at least minimally toxic, with appropriate clearance for the body. The probe should not induce a strong immune response. Additional requirements include low cost, easy preparation, and purification; a kit formulation is also desirable. Achieving contrast by the fourth genetic mechanism requires administration of a deliver agent that is analogous to the imaging probe. The agent might be a virus or plasmid contained within an immune or stem cell. One advantage of the genetic mechanism is that amplification can be achieved when the viral vector spreads or cell divides, as the genetic component would be replicated during that process. Further, the control element (promoter) leading to expression of the genetic reporter may be one

90

K.R. Zinn

that is specific for the cancer. Conditional replication of adenoviral vectors has been reported, including with viral particles that are fluorescent (Le et al., 2005). Finally, the viral vector or cell delivery may also provide a tumor-specific targeting capability (Stoff-Khalili et al., 2008; Takayama et al., 2007; Stoff-Khalili et al., 2007; Rocconi et al., 2007).

Clinical Implementation of Optical Imaging The appearance, size, or palpable differences of cancerous lesions may render them detectable. However, these characteristics are often inadequate for detection of small lesions and may not show differences following successful therapeutic intervention. Further, even if a small lesion is detectable by conventional methods, there may be residual undetected disease that remains following surgical intervention, even if surgical borders are evaluated by current practice. The tumor may have spread locally or to draining lymph nodes. Routine screenings and evaluations including oral cancer screenings, colonoscopy, and pulmonary bronchoscopy frequently fail to identify malignant lesions at an early stage when they are the most treatable. It would be desirable to have new diagnostic tools to build on existing infrastructure and methods. Therein lies a primary advantage of optical imaging, since optical instruments are already in clinical practice, for example in the operating room (stereomicroscopes), with surgical robots, and in the clinic where endoscopes of all varieties are used for procedures including colposcopy, laryngoscopy, arthroscopy, colposcopy, bronchoscopy, and various other endoscopies, which are being implemented to improve minimally invasive surgical techniques. These instruments can be easily equipped for application of fluorescence-based imaging, especially for planar reflectance-based imaging of superficial mucosal surfaces, including surgical procedures. As an additional example, currently gynecologists apply acetic acid and iodine solutions to cervical mucosa to visualize differences in color for normal versus dysplastic or cancerous tissue; these differences are visible by colposcopy. In a similar manner, these types of procedures and others could be extended using fluorescent probes that are activated by precancerous or cancerous tissues. A recent report of lymph node mapping with ICG demonstrated the ease with which dynamic fluorescence imaging could be applied in the clinical setting (Sevick-Muraca et al., 2008). This study illustrates the first mechanism by which contrast can be generated in tumors by imaging lymph node drainage from breast tumors. It established that a dose as low as 10 μg of ICG fluorophore was sufficient for dynamic imaging studies. An earlier report described a quantitative light-based method for non-invasive imaging of human breast cancer (Ntziachristos et al., 2000). In this study, the images were obtained using diffuse optical tomography, with contrast enhancement of blood flow provided by ICG that was administered to the patient. MRI was performed concurrently on the same patient and showed that ICGenhanced optical images co-registered accurately with gadolinium-enhanced MRI, thereby validating the ability of DOT to image breast tumors. Other reports documented ICG and modified photodynamic agents in combination with sophisticated

Overview of Cancer Detection and Monitoring Strategies

91

optical techniques (frequency-domain photon migration) to detect spontaneous cancer in the canine mammary chain (Gurfinkel et al., 2000; Reynolds et al., 1999). These methods have been significantly improved (Joshi et al., 2006). Fluorescence-based probes for imaging of tumor-specific receptors use the “binding and retention” mechanism for achieving contrast. Ballou et al. (1998, 1997) first reported retention of antibodies conjugated with near infrared dyes in tumors. Jackson and his group used light-based imaging to demonstrate the accumulation of a tumor receptor-specific, single chain Fv fragment labeled with Cy5 fluorescent dye in mice bearing melanoma xenografts (Ramjiawan et al., 2000). Optical imaging also detected tumor accumulation of a somatostatin-avid peptide conjugated with a near infrared fluorescent dye (Becker et al., 2000; 2001; Licha et al., 2001). More recently a library-derived peptidomimetic targeting alpha4beta1 integrin was reported for fluorescence imaging with Cy5.5 (Peng et al., 2008). Clinically approved antibodies for therapeutic applications were also conjugated with near infrared fluorophores and successfully applied for tumor imaging in preclinical imaging studies, including Her2-binding trastuzumab (Gee et al., 2008), CEAbinding arcitumomab (Lisy et al., 2008), and EGFR-binding cetuximab (Gleysteen et al., 2008; Withrow et al., 2007; Rosenthal et al., 2007). Using EGFR-targeted therapies; it was demonstrated that this technique could image microscope disease in preclinical surgical resection models (Gleysteen et al., 2008; Withrow et al., 2008). Eben Rosenthal reviews these data in his chapter of this book. One clinically relevant example of the “activation” mechanism for contrast generation used 5-aminolevulinic acid administration in patients prior to glioma surgery to enable surgeons to better identify glioma margins (Stummer et al., 2006). The 5aminolevulinic acid is non-fluorescent, but the compound leads to accumulation of fluorescent porphyrins in the glioma that are visible by fluorescence during surgery, allowing surgeons to better identify margins. Progression-free survival in patients where this procedure was applied was 41%, significantly higher than the standard practice, with survival at 21%. Additional examples of the activation mechanism include oxidation–reduction sensitive probes, one for sensing nucleic acids (Abe et al., 2008) and another fluorogenic probe for 3α-hydroxysteroid dehydrogenase (Yee et al., 2004). A biological thiol sensor was reported (Shibata et al., 2008), as was a glucosamine-bound fluorescent probe targeting lysosomes for breast tumor imaging (Li et al., 2008). The first examples of the “activation” mechanism of contrast via cancer-specific proteases were described by Mahmood and Weissleder using Cy5.5 probes that were inactive (autoquenched) when injected in the mice, but became specifically activated by proteases expressed in breast xenograft tumors (Mahmood et al., 1999; Weissleder et al., 1999). Since that time other investigators have expanded this approach (Fingleton et al., 2004; McIntyre et al., 2004; McIntyre and Matrisian, 2003; Scherer et al., 2008; Mahmood and Weissleder, 2003; Kozloff et al., 2006). A more detailed discussion of this mechanism for contrast is included in the chapter by McIntyre and Matrisian. Fluorescent genetic reporters contained within viral vectors have allowed for detection of extremely small tumors and metastasis in preclinical studies. Various

92

K.R. Zinn

xenografted tumors less than 1 mm in diameter in the chest and abdomen were detected with a replication-competent viral vector-encoding GFP (Adusumilli et al., 2006; Eisenberg et al., 2006). Chaudhuri et al. described using an adenoviral vectorencoding GFP to detect clusters of 50–100 ovarian cancer cells in the abdomen of mice (Chaudhuri et al., 2001a; 2002; 2001b). This work was extended to include a blood-based reporter (secreted embryonic alkaline phosphatase) so that in vivo imaging followed blood assays that indicated the presence of cancer (Chaudhuri et al., 2003). GFP has been routinely applied as a light-based reporter after stable integration of the GFP gene in cancer cells prior to implantation in mice (Bennett et al., 1997; Chishima et al., 1997a, b; c, d; Li et al., 2000; Yang et al., 2000a; Dardalhon et al., 1999; Pfeifer et al., 2001; Yang et al., 2000b; 1998; Yang et al., 1999). Chishima et al., demonstrated that GFP-expressing tumor cells were visualized after tumor-bearing mice were dissected, and the metastasis of cancer was detected in many different organs (Chishima et al., 1997a, b, c, d; Chirmule et al., 2000). This work was later extended by Yang et al. (1999) to include noninvasive imaging of GFP-positive melanoma metastasis in mice. Hoffman and his group reported imaging results from a study where an adenoviral vector-encoding enhanced GFP was injected into different organs of nude mice (Yang et al., 2000b). Light-based in vivo imaging showed GFP expression in different organs. Enhanced GFP and other red-shifted reporters are not cytotoxic and have stable fluorescence signal that can be readily detected. A chapter in this book entitled “Illustrating Molecular Events with Light: A Perspective on Optical Reporter Genes” by Pritha Ray provides additional details on the use of genetic reporters to generate contrast in tumors.

Optical Imaging in Context: Comparison to Conventional Modalities Optical imaging is one of several modalities with the potential to impact patient care. While not the subject of the current chapter or this book, the other modalities will be discussed briefly in order for comparisons to be made with optical imaging. Positron emission tomography (PET) has expanded in recent years, especially since the emergence of the dual modality PET/CT instrument. The latter advance allowed precise anatomical registration of the PET probe and facilitated the more widespread application of clinical PET. PET is a sensitive, whole-body and three-dimensional imaging but requires administration of a probe that is radiolabeled with a positron-emitting radionuclide. The most used probe is F-18-FDG that can detect the higher metabolic rate of cancer and can identify residual disease (5–7 mm) or tumors responding to therapy. Other probes in development target cancer hypoxia or various receptors or transporters present in cancer or in specific locations in the brain. Several comprehensive reviews have been published (Wester, 2007; Cai and Chen, 2008; Mankoff et al., 2008; Bading and Shields, 2008; Plathow and Weber, 2008). A related technology that also requires administration of a radiolabeled probe is single photon emission computed tomography (SPECT)

Overview of Cancer Detection and Monitoring Strategies

93

and SPECT/CT. This technology uses various forms of Tc-99m-labeled (primarily, 140 keV emission), In-111, Ga-67, I-123, or I-131-labeled probes. With respect to clinically approved cancer imaging, the bone agent (Tc-99m-MDP) is excellent for imaging bone metastasis, while In-111-octreoscan detects somatostatin receptors in neuroendocrine tumors. Of interest in molecular imaging is the potential for SPECT to detect multiple probes, with different gamma ray emissions, simultaneously. Detection of two probes is not difficult; and with careful planning, up to three probes can be administered. A similar capability does not exist with PET, where only one probe can be given at one time. The PET/CT and SPECT/CT instruments are large and expensive. The patient imaging sessions are at least 30 min and must be carefully linked with the timing of the administered probe. Post-processing of the collected data is necessary for interpretation by a radiologist. Magnetic resonance imaging (MRI) methods are routinely applied for cancer detection and monitoring of response to therapy. These instruments are also quite large and expensive and require post-processing to generate images, making them more difficult to apply in a surgical setting. MRI modalities include diffusion-weighted MRI (DWI), dynamic contrast-enhanced MRI, and magnetic resonance spectroscopy. These techniques have been applied to evaluate cancer treatment efficacy, including in neoadjuvant therapy of breast cancer (Pickles et al., 2006; Su et al., 2006; Yankeelov et al., 2007). DWI measures the water diffusion, which is increased with therapy-induced cell necrosis or apoptosis in tumors within 2–4 days, especially in responding patients (Galons et al., 1999; Chenevert et al., 2000). Dynamic contrast-enhanced MRI measures early changes (within 2–7 days) in the tumor microvasculature, as quantified by changes in volume transfer constant (Ktrans ), fractional vascular plasma volume (vp ), and fractional extravascular–extracellular volume (ve ) (Wilmes et al., 2007; Liu et al., 2005). Magnetic resonance spectroscopy can quantify the substantial decrease of choline or phosphomonoester (PME) in tumors after successful treatment within a few days after therapy is started (Su et al., 2006). Existing non-optical imaging technologies do not meet the current needs for the early detection and monitoring of cancer. Imaging technologies have improved in their sensitivity to image ovarian, breast, and other cancers non-invasively by PET, CT, MRI, SPECT, and ultrasonography. However, these methods fall short in fulfilling the need for early and accurate diagnosis of neoplastic disease or to monitor response. In 1998 PET imaging with F-18-FDG was able to detect primary breast lesions over 1 cm in diameter (Wahl, 1998), and PET–CT has improved detection down to 5 mm. Grab et al., concluded that a negative finding on PET or MRI would not exclude early ovarian neoplasia (Grab et al., 2000). Kubik-Huch et al. (2000) reported in a comparative study that PET, CT, and MRI were not a replacement for surgery in the detection of microscopic peritoneal disease. While PET imaging offered less accurate spatial assignment of small lesions compared with CT and MRI, the latter two modalities were less specific than F-18-FDG PET. In a separate report, Tempany et al. (2000) reported that CT and MR were equivocal for imaging advanced ovarian cancer. Kurjak et al. (2000) reported that transvaginal color Doppler and three-dimensional power Doppler ultrasound imaging improved the

94

K.R. Zinn

ability to differentiate benign from malignant ovarian masses. Taken together, nonoptical imaging techniques are not satisfactory for the early and accurate detection of tumors smaller than 5 mm in diameter.

Advantages and Distinctions of Fluorescence Imaging The greatest advantage of optical technologies is the ease and relatively low cost with which it can be applied to expand capabilities within the current clinical practice. Because optical imaging can be used in real time, is easily adaptable to existing equipment, and allow anatomic visualization, the technology allows the surgeon to have an improved tool to detect and resect tumors. The high spatial resolution and sensitivity of fluorescence are additional characteristics that distinguish the modality from others. There is no other imaging modality that can detect cancer less than 1 mm in size in real time in clinical settings. Additional advantages include a capacity to image several fluorophores simultaneously, a feature to detect fluorescence only when “activated” by a cancer-specific process, and the potential to apply genetic reporter to achieve cancer-specific fluorescence. There are no radioactivity concerns using fluorescence. Finally, contrast can be produced by additional mechanisms such as changes in fluorescence lifetime that may depend on local environment within cancer.

Challenges for Clinical development The current emphasis by federal, industrial, and academic institutions to promote molecular imaging stems from a real desire to enable personalized medicine and also to save money for both drug development and patient care. The FDA as part of its Critical Path Initiative summarized its position in 2005 by stating that “Imaging is a key technology for assessing, accelerating the development of, and guiding the use of new therapeutic options. The Agency believes that synergy between current drug development programs and current imaging techniques can be created for drug development to work in a more cost effective manner.” Similarly, the National Cancer Institute established the In vivo Imaging Workspace in 2005 with the following intentions: “Initial efforts will involve enlisting the widest possible representation from cancer centers, industry, organizations, and standards-setting groups. Among the earliest of the cooperative workspace tasks will be the identification of overall aims and the most urgent challenges in cancer imaging and sharing of data. The workspace will define the needs for and participate in creating, optimizing, and validating tools and methods to extract meaning from in vivo imaging data.” Both the FDA and the NIH appreciate the need to develop standardized methods to apply molecular imaging methods in clinical practice. They are joined by a number of scientific organizations (Academy of Molecular Imaging, Society of Molecular, Society of Nuclear Medicine) as well as academic organizations in this challenging endeavor.

Overview of Cancer Detection and Monitoring Strategies

95

Economics and Regulatory Issues for Imaging Probe Development The previous sections of this chapter discussed the mechanisms to generate fluorescent contrast for clinical applications and provided examples of how fluorescence imaging has been applied in preclinical and clinical cancer studies. A logical basis for using a fluorescent probe to generate contrast in a tumor and subsequent demonstration in a preclinical model are only the initial steps in probe development. A probe must undergo rigorous testing with review and final approval by the FDA before it can be more widely applied. Aside from the regulatory approval issues, a private enterprise must pay for the development in the hopes that a commercial product will be developed that can generate reimbursements to cover capital investment. Finally, when the FDA-approved commercial product is ready, the company must then convince the Center for Medicare and Medicaid Services and other insurance groups to allow reimbursement, in order to pay for the probe, instrument time, and physicians’ interpretation services. The regulatory challenges for development of imaging probes were reviewed recently (Hoffman et al., 2007). The cost for a company to develop a new optical imaging probe to commercialization is estimated to be between 100 and 200 million U.S. dollars over 8–10 years, based on prior experience with costs of contrast agents for other imaging modalities (nuclear medicine, CT, MRI, etc.) (Nunn, 2006; Nunn, 2007). In 2004 the total worldwide market for all contrast agents was estimated to be around 8.5 billion U.S. dollars or about 1–2% of the total drug market of ˜850 billion U.S. dollars. Because the contrast agent or imaging probe market is significantly less than the therapeutic drug market, the pharma has limited enthusiasm for development of imaging agents. In 2003 the top two imaging agents (Omnipaque, Cardiolite) each had worldwide sales of approximately 400 million U.S. dollars. By comparison the number 10 drug in therapeutic sales was ∼3.4 billion U.S. dollars for the same year. Another issue is that current contrast agents typically have multiple indications or are used in very common diseases where many studies are done. New optical agents may be even more targeted which improve detection and the toxicity profile, but therefore have less of a market. Based on all this information, there are significant challenges for commercialization of optical imaging agents, as they are pushed forward for use in personalized medicine. However, clinicians and imaging scientists agree that the patient benefits using this approach are worth the effort.

Conclusions There is great potential for development of fluorescence imaging probes that can improve currently applied optical technologies for cancer detection and monitoring. The enthusiasm for these probes is balanced by significant challenges in obtaining regulatory approvals and reimbursement, which will require adequate demonstration that their impact not only saves lives but reduces overall health-care costs.

96

K.R. Zinn

References Abe, H., J. Wang, K. Furukawa, K. Oki, M. Uda, S. Tsuneda, and Y. Ito, A reductiontriggered fluorescence probe for sensing nucleic acids. Bioconjug Chem, 2008. 19(6): 1219–26. Adusumilli, P.S., B.M. Stiles, M.K. Chan, D.P. Eisenberg, Z. Yu, S.F. Stanziale, R. Huq, R.J. Wong, V.W. Rusch, and Y. Fong, Real-time diagnostic imaging of tumors and metastases by use of a replication-competent herpes vector to facilitate minimally invasive oncological surgery. Faseb J, 2006. 20(6): 726–8. Bading, J.R. and A.F. Shields, Imaging of cell proliferation: status and prospects. J Nucl Med, 2008. 49 Suppl 2: 64S–80S. Ballou, B., G.W. Fisher, J.S. Deng, T.R. Hakala, M. Srivastava, and D.L. Farkas, Cyanine fluorochrome-labeled antibodies in vivo: assessment of tumor imaging using Cy3, Cy5, Cy5.5, and Cy7. Cancer Detect Prev, 1998. 22(3): 251–7. Ballou, B., G.W. Fisher, T.R. Hakala, and D.L. Farkas, Tumor detection and visualization using cyanine fluorochrome-labeled antibodies. Biotechnol Prog, 1997. 13(5): 649–58. Becker, A., C. Hessenius, S. Bhargava, C. Grotzinger, K. Licha, J. Schneider-Mergener, B. Wiedenmann, and W. Semmler, Cyanine dye labeled vasoactive intestinal peptide and somatostatin analog for optical detection of gastroenteropancreatic tumors. Ann N Y Acad Sci, 2000. 921: 275–8. Becker, A., C. Hessenius, K. Licha, B. Ebert, U. Sukowski, W. Semmler, B. Wiedenmann, and C. Grotzinger, Receptor-targeted optical imaging of tumors with near-infrared fluorescent ligands. Nat Biotechnol, 2001. 19(4): 327–31. Bennett, J., D. Duan, J.F. Engelhardt, and A.M. Maguire, Real-time, noninvasive in vivo assessment of adeno-associated virus-mediated retinal transduction. Invest Ophthalmol Vis Sci, 1997. 38(13): 2857–63. Cai, W. and X. Chen, Multimodality molecular imaging of tumor angiogenesis. J Nucl Med, 2008. 49 Suppl 2: 113S–28S. Chaudhuri, T.R., Z. Cao, V.N. Krasnykh, A.V. Stargel, N. Belousova, E.E. Partridge, and K.R. Zinn, Blood-based screening and light based imaging for the early detection and monitoring of ovarian cancer xenografts. Technol Cancer Res Treat, 2003. 2(2): 171–80. Chaudhuri, T.R., Z. Cao, B.E. Rogers, E.E. Partridge, and K.R. Zinn, Non-invasive light-based imaging of GFP-positive ovarian xenografts. Abstract Book p. 70, Annual Investigators Workshop, 2001a. Chaudhuri, T.R., Z. Cao, E.E. Partridge, and K.R. Zinn, A comparative analysis of optical and gamma camera imaging for non-invasive detection of ovarian cancer. Eur J Nucl Med Mol I, 2002. 29(1):s229. Chaudhuri, T.R., J.M. Mountz, B.E. Rogers, E.E. Partridge, and K.R. Zinn, Light-based imaging of green fluorescent protein-positive ovarian cancer xenografts during therapy. Gynecol Oncol, 2001b. 82(3): 581–9. Chenevert, T.L., L.D. Stegman, J.M. Taylor, P.L. Robertson, H.S. Greenberg, A. Rehemtulla, and B.D. Ross, Diffusion magnetic resonance imaging: an early surrogate marker of therapeutic efficacy in brain tumors. J Natl Cancer Inst, 2000. 92(24): 2029–36. Chirmule, N., S.E. Raper, L. Burkly, D. Thomas, J. Tazelaar, J.V. Hughes, and J.M. Wilson, Readministration of adenovirus vector in nonhuman primate lungs by blockade of CD40-CD40 ligand interactions. J Virol, 2000. 74(7): 3345–52. Chishima, T., Y. Miyagi, L. Li, Y. Tan, E. Baranov, M. Yang, H. Shimada, A.R. Moossa, and R.M. Hoffman, Use of histoculture and green fluorescent protein to visualize tumor cell host interaction. In Vitro Cell Dev Biol Anim, 1997a. 33(10): 745–7. Chishima, T., Y. Miyagi, X. Wang, Y. Tan, H. Shimada, A. Moossa, and R.M. Hoffman, Visualization of the metastatic process by green fluorescent protein expression. Anticancer Res, 1997b. 17(4A): 2377–84.

Overview of Cancer Detection and Monitoring Strategies

97

Chishima, T., Y. Miyagi, X. Wang, H. Yamaoka, H. Shimada, A.R. Moossa, and R.M. Hoffman, Cancer invasion and micrometastasis visualized in live tissue by green fluorescent protein expression. Cancer Res, 1997c. 57(10): 2042–7. Chishima, T., M. Yang, Y. Miyagi, L. Li, Y. Tan, E. Baranov, H. Shimada, A.R. Moossa, S. Penman, and R.M. Hoffman, Governing step of metastasis visualized in vitro. Proc Natl Acad Sci USA, 1997d. 94(21): 11573–6. Dardalhon, V., N. Noraz, K. Pollok, C. Rebouissou, M. Boyer, A.Q. Bakker, H. Spits, and N. Taylor, Green fluorescent protein as a selectable marker of fibronectin- facilitated retroviral gene transfer in primary human T lymphocytes. Hum Gene Ther, 1999. 10(1): 5–14. Eisenberg, D.P., P.S. Adusumilli, K.J. Hendershott, S. Chung, Z. Yu, M.K. Chan, M. Hezel, R.J. Wong, and Y. Fong, Real-time intraoperative detection of breast cancer axillary lymph node metastases using a green fluorescent protein-expressing herpes virus. Ann Surg, 2006. 243(6): 824–30; discussion 830–2. Fingleton, B., R. Menon, K.J. Carter, P.D. Overstreet, D.L. Hachey, L.M. Matrisian, and J.O. McIntyre, Proteinase activity in human and murine saliva as a biomarker for proteinase inhibitor efficacy. Clin Cancer Res, 2004. 10(23): 7865–74. Galons, J.P., M.I. Altbach, G.D. Paine-Murrieta, C.W. Taylor, and R.J. Gillies, Early increases in breast tumor xenograft water mobility in response to paclitaxel therapy detected by noninvasive diffusion magnetic resonance imaging. Neoplasia, 1999. 1(2): 113–7. Gee, M.S., R. Upadhyay, H. Bergquist, H. Alencar, F. Reynolds, M. Maricevich, R. Weissleder, L. Josephson, and U. Mahmood, Human breast cancer tumor models: molecular imaging of drug susceptibility and dosing during HER2/neu-targeted therapy. Radiology, 2008. 248(3): 925–35. Gleysteen, J.P., J.R. Newman, D. Chhieng, A. Frost, K.R. Zinn, and E.L. Rosenthal, Fluorescent labeled anti-EGFR antibody for identification of regional and distant metastasis in a preclinical xenograft model. Head Neck, 2008. 30(6): 782–9. Grab, D., F. Flock, I. Stohr, K. Nussle, A. Rieber, S. Fenchel, H.J. Brambs, S.N. Reske, and R. Kreienberg, Classification of asymptomatic adnexal masses by ultrasound, magnetic resonance imaging, and positron emission tomography. Gynecol Oncol, 2000. 77(3): 454–9. Gurfinkel, M., A.B. Thompson, W. Ralston, T.L. Troy, A.L. Moore, T.A. Moore, J.D. Gust, D. Tatman, J.S. Reynolds, B. Muggenburg, K. Nikula, R. Pandey, R.H. Mayer, D.J. Hawrysz, and E.M. Sevick-Muraca, Pharmacokinetics of ICG and HPPH-car for the detection of normal and tumor tissue using fluorescence, near-infrared reflectance imaging: a case study. Photochem Photobiol, 2000. 72(1): 94–102. Hoffman, R.M., The multiple uses of fluorescent proteins to visualize cancer in vivo. Nat Rev Cancer, 2005. 5(10): 796–806. Hoffman, J.M., S.S. Gambhir, and G.J. Kelloff, Regulatory and reimbursement challenges for molecular imaging. Radiology, 2007. 245(3): 645–60. Joshi, A., W. Bangerth, K. Hwang, J.C. Rasmussen, and E.M. Sevick-Muraca, Fully adaptive FEM based fluorescence optical tomography from time-dependent measurements with area illumination and detection. Med Phys, 2006. 33(5): 1299–310. Kozloff, K.M., L. Quinti, C. Tung, R. Weissleder, and U. Mahmood, Non-invasive imaging of osteoclast activity via near-infrared cathepsin-K activatable optical probe. J Musculoskelet Neuronal Interact, 2006. 6(4): 353. Kubik-Huch, R.A., W. Dorffler, G.K. von Schulthess, B. Marincek, O.R. Kochli, B. Seifert, U. Haller, and H.C. Steinert, Value of (18F)-FDG positron emission tomography, computed tomography, and magnetic resonance imaging in diagnosing primary and recurrent ovarian carcinoma. Eur Radiol, 2000. 10(5): 761–7. Kurjak, A., S. Kupesic, V. Sparac, and D. Kosuta, Three-dimensional ultrasonographic and power Doppler characterization of ovarian lesions. Ultrasound Obstet Gynecol, 2000. 16(4): 365–71. Le, L.P., J. Li, V.V. Ternovoi, G.P. Siegal, and D.T. Curiel, Fluorescently tagged canine adenovirus via modification with protein IX-enhanced green fluorescent protein. J Gen Virol, 2005. 86(Pt 12): p. 3201–8.

98

K.R. Zinn

Li, C., T.R. Greenwood, and K. Glunde, Glucosamine-bound near-infrared fluorescent probes with lysosomal specificity for breast tumor imaging. Neoplasia, 2008. 10(4): 389–98. Li, C.Y., S. Shan, Q. Huang, R.D. Braun, J. Lanzen, K. Hu, P. Lin, and M.W. Dewhirst, Initial stages of tumor cell-induced angiogenesis: evaluation via skin window chambers in rodent models. J Natl Cancer Inst, 2000. 92(2): 143–7. Licha, K., C. Hessenius, A. Becker, P. Henklein, M. Bauer, S. Wisniewski, B. Wiedenmann, and W. Semmler, Synthesis, characterization, and biological properties of cyanine- labeled somatostatin analogues as receptor-targeted fluorescent probes. Bioconjug Chem, 2001. 12(1): 44–50. Lisy, M.R., A. Goermar, C. Thomas, J. Pauli, U. Resch-Genger, W.A. Kaiser, and I. Hilger, In vivo near-infrared fluorescence imaging of carcinoembryonic antigen-expressing tumor cells in mice. Radiology, 2008. 247(3): 779–87. Liu, G., H.S. Rugo, G. Wilding, T.M. McShane, J.L. Evelhoch, C. Ng, E. Jackson, F. Kelcz, B.M. Yeh, F.T. Lee, Jr., C. Charnsangavej, J.W. Park, E.A. Ashton, H.M. Steinfeldt, Y.K. Pithavala, S.D. Reich, and R.S. Herbst, Dynamic contrast-enhanced magnetic resonance imaging as a pharmacodynamic measure of response after acute dosing of AG-013736, an oral angiogenesis inhibitor, in patients with advanced solid tumors: results from a phase I study. J Clin Oncol, 2005. 23(24): 5464–73. Mahmood, U., C.H. Tung, A. Bogdanov, Jr., and R. Weissleder, Near-infrared optical imaging of protease activity for tumor detection. Radiology, 1999. 213(3): 866–70. Mahmood, U. and R. Weissleder, Near-infrared optical imaging of proteases in cancer. Mol Cancer Ther, 2003. 2(5): 489–96. Mankoff, D.A., J.M. Link, H.M. Linden, L. Sundararajan, and K.A. Krohn, Tumor receptor imaging. J Nucl Med, 2008. 49 Suppl 2: 149S–63S. Martin, B.R., B.N. Giepmans, S.R. Adams, and R.Y. Tsien, Mammalian cell-based optimization of the biarsenical-binding tetracysteine motif for improved fluorescence and affinity. Nat Biotechnol, 2005. 23(10): 1308–14. McIntyre, J.O., B. Fingleton, K.S. Wells, D.W. Piston, C.C. Lynch, S. Gautam, and L.M. Matrisian, Development of a novel fluorogenic proteolytic beacon for in vivo detection and imaging of tumour-associated matrix metalloproteinase-7 activity. Biochem J, 2004. 377(Pt 3): 617–28. McIntyre, J.O. and L.M. Matrisian, Molecular imaging of proteolytic activity in cancer. J Cell Biochem, 2003. 90(6): 1087–97. Ntziachristos, V., A.G. Yodh, M. Schnall, and B. Chance, Concurrent MRI and diffuse optical tomography of breast after indocyanine green enhancement. Proc Natl Acad Sci U S A, 2000. 97(6): 2767–72. Nunn, A.D., The cost of developing imaging agents for routine clinical use. Invest Radiol, 2006. 41(3): 206–12. Nunn, A.D., The cost of bringing a radiopharmaceutical to the patient s bedside. J Nucl Med, 2007. 48(2): 169. Peng, L., R. Liu, M. Andrei, W. Xiao, and K.S. Lam, In vivo optical imaging of human lymphoma xenograft using a library-derived peptidomimetic against alpha4beta1 integrin. Mol Cancer Ther, 2008. 7(2): 432–7. Pfeifer, A., T. Kessler, M. Yang, E. Baranov, N. Kootstra, D.A. Cheresh, R.M. Hoffman, and I.M. Verma, Transduction of liver cells by lentiviral vectors: analysis in living animals by fluorescence imaging. Mol Ther, 2001. 3(3): 319–22. Pickles, M.D., P. Gibbs, M. Lowry, and L.W. Turnbull, Diffusion changes precede size reduction in neoadjuvant treatment of breast cancer. Magn Reson Imaging, 2006. 24(7): 843–7. Plathow, C. and W.A. Weber, Tumor cell metabolism imaging. J Nucl Med, 2008. 49 Suppl 2: 43S–63S. Ramjiawan, B., P. Maiti, A. Aftanas, H. Kaplan, D. Fast, H.H. Mantsch, and M. Jackson, Noninvasive localization of tumors by immunofluorescence imaging using a single chain Fv fragment of a human monoclonal antibody with broad cancer specificity. Cancer, 2000. 89(5): 1134–44.

Overview of Cancer Detection and Monitoring Strategies

99

Reynolds, J.S., T.L. Troy, R.H. Mayer, A.B. Thompson, D.J. Waters, K.K. Cornell, P.W. Snyder, and E.M. Sevick-Muraca, Imaging of spontaneous canine mammary tumors using fluorescent contrast agents. Photochem Photobiol, 1999. 70(1): 87–94. Rocconi, R.P., Z.B. Zhu, M. Stoff-Khalili, A.A. Rivera, B. Lu, M. Wang, R.D. Alvarez, D.T. Curiel, and S.K. Makhija, Treatment of ovarian cancer with a novel dual targeted conditionally replicative adenovirus (CRAd). Gynecol Oncol, 2007. 105(1): 113–21. Rosenthal, E.L., B.D. Kulbersh, T. King, T.R. Chaudhuri, and K.R. Zinn, Use of fluorescent labeled anti-epidermal growth factor receptor antibody to image head and neck squamous cell carcinoma xenografts. Mol Cancer Ther, 2007. 6(4): 1230–8. Scherer, R.L., J.O. McIntyre, and L.M. Matrisian, Imaging matrix metalloproteinases in cancer. Cancer Metastasis Rev, 2008. 27(4): 679–90. Sevick-Muraca, E.M., R. Sharma, J.C. Rasmussen, M.V. Marshall, J.A. Wendt, H.Q. Pham, E. Bonefas, J.P. Houston, L. Sampath, K.E. Adams, D.K. Blanchard, R.E. Fisher, S.B. Chiang, R. Elledge, and M.E. Mawad, Imaging of lymph flow in breast cancer patients after microdose administration of a near-infrared fluorophore: feasibility study. Radiology, 2008. 246(3): 734–41. Shaner, N.C., M.Z. Lin, M.R. McKeown, P.A. Steinbach, K.L. Hazelwood, M.W. Davidson, and R.Y. Tsien, Improving the photostability of bright monomeric orange and red fluorescent proteins. Nat Methods, 2008. 5(6): 545–51. Shaner, N.C., P.A. Steinbach, and R.Y. Tsien, A guide to choosing fluorescent proteins. Nat Methods, 2005. 2(12): 905–9. Shcherbo, D., E.M. Merzlyak, T.V. Chepurnykh, A.F. Fradkov, G.V. Ermakova, E.A. Solovieva, K.A. Lukyanov, E.A. Bogdanova, A.G. Zaraisky, S. Lukyanov, and D.M. Chudakov, Bright far-red fluorescent protein for whole-body imaging. Nat Methods, 2007. 4(9): 741–6. Shibata, A., K. Furukawa, H. Abe, S. Tsuneda, and Y. Ito, Rhodamine-based fluorogenic probe for imaging biological thiol. Bioorg Med Chem Lett, 2008. 18(7): 2246–9. Stoff-Khalili, M.A., A.A. Rivera, J.M. Mathis, N.S. Banerjee, A.S. Moon, A. Hess, R.P. Rocconi, T.M. Numnum, M. Everts, L.T. Chow, J.T. Douglas, G.P. Siegal, Z.B. Zhu, H.G. Bender, P. Dall, A. Stoff, L. Pereboeva, and D.T. Curiel, Mesenchymal stem cells as a vehicle for targeted delivery of CRAds to lung metastases of breast carcinoma. Breast Cancer Res Treat, 2007. 105(2): 157–67. Stoff-Khalili, M.A., A.A. Rivera, A. Nedeljkovic-Kurepa, A. DeBenedetti, X.L. Li, Y. Odaka, J. Podduturi, D.A. Sibley, G.P. Siegal, A. Stoff, S. Young, Z.B. Zhu, D.T. Curiel, and J.M. Mathis, Cancer-specific targeting of a conditionally replicative adenovirus using mRNA translational control. Breast Cancer Res Treat, 2008. 108(1): 43–55. Stummer, W., U. Pichlmeier, T. Meinel, O.D. Wiestler, F. Zanella, and H.J. Reulen, Fluorescenceguided surgery with 5-aminolevulinic acid for resection of malignant glioma: a randomised controlled multicentre phase III trial. Lancet Oncol, 2006. 7(5): 392–401. Su, M.Y., H.M. Baik, H.J. Yu, J.H. Chen, R.S. Mehta, and O. Nalcioglu, Comparison of choline and pharmacokinetic parameters in breast cancer measured by MR spectroscopic imaging and dynamic contrast enhanced MRI. Technol Cancer Res Treat, 2006. 5(4): 401–10. Takayama, K., P.N. Reynolds, Y. Adachi, L. Kaliberova, J. Uchino, Y. Nakanishi, and D.T. Curiel, Vascular endothelial growth factor promoter-based conditionally replicative adenoviruses for pan-carcinoma application. Cancer Gene Ther, 2007. 14(1): 105–16. Tempany, C.M., K.H. Zou, S.G. Silverman, D.L. Brown, A.B. Kurtz, and B.J. McNeil, Staging of advanced ovarian cancer: comparison of imaging modalities– report from the Radiological Diagnostic Oncology Group. Radiology, 2000. 215(3): 761–7. Wahl, R.L., Overview of the current status of PET in breast cancer imaging. Q J Nucl Med, 1998. 42(1): 1–7. Weissleder, R., C.H. Tung, U. Mahmood, and A. Bogdanov, Jr., In vivo imaging of tumors with protease-activated near-infrared fluorescent probes. Nat Biotechnol, 1999. 17(4): 375–8. Wester, H.J., Nuclear imaging probes: from bench to bedside. Clin Cancer Res, 2007. 13(12): 3470–81.

100

K.R. Zinn

Wilmes, L.J., M.G. Pallavicini, L.M. Fleming, J. Gibbs, D. Wang, K.L. Li, S.C. Partridge, R.G. Henry, D.R. Shalinsky, D. Hu-Lowe, J.W. Park, T.M. McShane, Y. Lu, R.C. Brasch, and N.M. Hylton, AG-013736, a novel inhibitor of VEGF receptor tyrosine kinases, inhibits breast cancer growth and decreases vascular permeability as detected by dynamic contrast-enhanced magnetic resonance imaging. Magn Reson Imaging, 2007. 25(3): 319–27. Withrow, K.P., J.P. Gleysteen, A. Safavy, J. Skipper, R.A. Desmond, K. Zinn, and E.L. Rosenthal, Assessment of indocyanine green-labeled cetuximab to detect xenografted head and neck cancer cell lines. Otolaryngol Head Neck Surg, 2007. 137(5): 729–34. Withrow, K.P., J.R. Newman, J.B. Skipper, J.P. Gleysteen, J.S. Magnuson, K. Zinn, and E.L. Rosenthal, Assessment of bevacizumab conjugated to Cy5.5 for detection of head and neck cancer xenografts. Technol Cancer Res Treat, 2008. 7(1): 61–6. Yang, M., E. Baranov, P. Jiang, F.X. Sun, X.M. Li, L. Li, S. Hasegawa, M. Bouvet, M. Al-Tuwaijri, T. Chishima, H. Shimada, A.R. Moossa, S. Penman, and R.M. Hoffman, Whole-body optical imaging of green fluorescent protein-expressing tumors and metastases. Proc Natl Acad Sci USA, 2000a. 97(3): 1206–11. Yang, M., E. Baranov, A.R. Moossa, S. Penman, and R.M. Hoffman, Visualizing gene expression by whole-body fluorescence imaging. Proc Natl Acad Sci USA, 2000. 97(22): 12278–82. Yang, M., S. Hasegawa, P. Jiang, X. Wang, Y. Tan, T. Chishima, H. Shimada, A.R. Moossa, and R.M. Hoffman, Widespread skeletal metastatic potential of human lung cancer revealed by green fluorescent protein expression. Cancer Res, 1998. 58(19): 4217–21. Yang, M., P. Jiang, Z. An, E. Baranov, L. Li, S. Hasegawa, M. Al-Tuwaijri, T. Chishima, H. Shimada, A.R. Moossa, and R.M. Hoffman, Genetically fluorescent melanoma bone and organ metastasis models. Clin Cancer Res, 1999. 5(11): 3549–59. Yankeelov, T.E., M. Lepage, A. Chakravarthy, E.E. Broome, K.J. Niermann, M.C. Kelley, I. Meszoely, I.A. Mayer, C.R. Herman, K. McManus, R.R. Price, and J.C. Gore, Integration of quantitative DCE-MRI and ADC mapping to monitor treatment response in human breast cancer: initial results. Magn Reson Imaging, 2007. 25(1): 1–13. Yee, D.J., V. Balsanek, and D. Sames, New tools for molecular imaging of redox metabolism: development of a fluorogenic probe for 3 alpha-hydroxysteroid dehydrogenases. J Am Chem Soc, 2004. 126(8): 2282–3.

The Application of Tissue Autofluorescence in Detection and Management of Oral Cancer and Premalignant Lesions C.F. Poh, P. Lane, C. MacAulay, L. Zhang, and M.P. Rosin

Introduction There is a wealth of literature that supports the use of tissue autofluorescence in the screening and diagnosis of precancers in the lung, uterine cervix, skin, and oral cavity. This approach is already in clinical use in the lung, and the mechanism of action of tissue autofluorescence has been well described in the cervix. Data are now emerging supporting its clinical usage in the detection and management of oral cancer and premalignant lesions. In this chapter, we will describe the biology underlying tissue autofluorescence, briefly review its current application in the management of lung and cervical cancers, and finally focus on its potential clinical utility in the detection and management of oral cancer and premalignant lesions. Since first proposed by Slaughter et al. (1953), “field cancerization” has become an overarching concept that has been widely applied in the management of cancers, including those in the oral cavity. With the advent of molecular technology, this concept has broadened, moving from a definition that focused primarily on the clinical description of the multiple lesions that can occur in a tissue over time to encompass the underlying molecular change envisaged as driving these clinical changes. It is becoming increasingly apparent that genetically altered cells are often widespread across the epithelium of patients’ mucosal surfaces, extending into clinically and histologically normal tissue. These clinical lesions can be ill-defined, patchy in appearance with intermittent normal tissue, complicating decisions on when and where to biopsy for histological and molecular assessment. Genetically altered cells can spread quite widely and may be at least partly responsible for the formation of multiple lesions at different anatomical sites in the oral cavity, which can become apparent clinically over time. Even within a single lesion, the subclinical extension

C.F. Poh (B) Faculty of Dentistry, University of British Columbia, Cancer Control Research and Cancer Imaging, BC Cancer Agency/Cancer Research Centre, Rm: JBM 322, 2199 Wesbrook Mall, Vancouver, BC, Canada, V6T 1Z3 e-mail: [email protected]

E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_6, 

101

102

C.F. Poh et al.

of altered cells can result in incomplete treatment with ensuing risk of local recurrence and formation of second primary malignancy. There is an urgent need for new approaches that can be applied in clinical practice to improve the detection, risk assessment, and management of field alterations in high-risk tissue, particularly approaches that can be used in real-time settings to identify regions of risk requiring assessment. Advances in optical technology are providing us with a new view of the dynamic fields underlying cancer development. Among such technologies, assessment of tissue autofluorescence represents a particularly promising approach. Changes in fluorescence reflect a complex interplay of alterations to fluorophores in the tissue and structural changes in tissue morphology, occurring with progression of the disease (Pavlova et al., 2003; Richards-Kortum and Sevick-Muraca, 1996). Tools have been developed that provide the clinician with an ability to directly visualize such change. There is a growing body of literature supporting the ability of these tools to facilitate the detection and diagnosis of cancers and precancers in the lung, uterine cervix, skin, and oral cavity (Pavlova et al., 2003; Gillenwater et al., 1998; Heintzelman et al., 2000; Ingrams et al., 1997; Lam et al., 1993; Muller et al., 2003; Ramanujam et al., 1994; Zeng et al., 2000; Chang et al., 2002; Collier et al., 2003). Due to the potential significant impact on disease management, these tools can be used to “shed new light on an old problem” (Califano et al., 1996). In this chapter, we will describe the biology underlying tissue autofluorescence, briefly review the application of tissue autofluorescence in the management of lung, cervical, and skin cancers, and finally focus on the more recent application of this technology in management of oral cancer and premalignant lesions.

Biology Underlying Tissue Autofluorescence Not unlike many scientific discoveries, the association of dysplastic progression with the loss of tissue autofluorescence was observed serendipitously during a study designed to show something quite different, well before its biological plausibility was hypothesized. Instruments designed to detect loss of tissue fluorescence for the detection of cancerous and precancerous lesions were developed and commercialized well before the underlying biological mechanisms were understood. Various researchers, in particular the group led by Richards-Kortum, established biological plausibility by examining the optical properties of freshly excised tissue specimens using confocal fluorescence microscopy and spectroscopy and correlating this with specimen histology. The biological mechanisms involved in tissue autofluorescence and their correlation with dysplastic progression are better understood now; however, there is still more to learn. In general, native fluorescence detected at the tissue surface is a function of tissue morphology and biochemistry. Intrinsic tissue fluorescence, due to naturally occurring fluorophores in the epithelium and stroma, is modified by

The Application of Tissue Autofluorescence in Detection and Management

103

local tissue morphology through absorption and scattering, first during the application of the excitation light and then during the collection of the emission light. In general, absorption and scattering modify the intensity and spectral distribution of the detected fluorescence. Fluorophores that have been found to be useful in the optical screening and diagnosis of precancers are those that excite in the violet–blue part of the visible spectrum (400–450 nm) through the ultraviolet A (UV-A, 315–400 nm) and have properties that have been spectroscopically correlated with dysplastic progression (Pavlova et al., 2003; Richards-Kortum and Sevick-Muraca, 1996; Drezek et al., 2001a, b). Endogenous fluorophores relevant to screening and diagnosis of precancers are illustrated in Fig. 1. The crosshairs (“+”) indicate excitation–emission maxima for each fluorophore, while the ellipses represent the approximate range of excitation and emission.

Fig. 1 Excitation and emission wavelengths of endogenous fluorophores. The figure shows fluorescence intensity as a function of excitation (Ex) and emission (Em) wavelength for endogenous fluorophores relevant to the visualization of autofluorescence for screening and diagnosis of precancers. The crosshairs (“+”) indicate excitation–emission maxima for each fluorophore, while the ellipses represents their approximate range of excitation and emission

Collagen fluorescence originates from the cross-links that bind collagen fibrils together to form fibers. Maximum fluorescence is observed at 340-nm excitation (420-nm emission) and there is significant fluorescence when excited between 410 and 470 nm. In this range of excitation, the emission maximum continuously shifts to the red from 475 nm at 410-nm excitation to ∼540 nm at 470-nm excitation (Sokolov et al., 2002). The reduced form of nicotinamide adenine dinucleotide (NADH) and the oxidized form of flavin adenine dinucleotide (FAD) are important fluorophores that also excite within the UV-A through violet–blue wavelengths. Maximum NADH fluorescence occurs at 340-nm excitation and 450-nm emission,

104

C.F. Poh et al.

while that of FAD occurs at 450-nm excitation and 515-nm emission (RichardsKortum and Sevick-Muraca, 1996). The excitation and emission bands of collagen are much broader than those of NADH/FAD, probably due to the contribution of several different fluorophores to the overall spectrum. Porphyrin fluorescence has a broad excitation band (maximum at 436 nm) and a narrow emission band (maximum at 630 nm). It is generally associated with bacterial and fungal infections and is not diagnostic for early cancer. Fluorescence from the collagen cross-links originates in the stroma (collagen matrix) while that of FAD and NADH originate in the cells of the epithelium. Experiments on freshly excised surgical samples have shown that the majority of fluorescence originates from the collagen and only a small fraction comes from the epithelium (Pavlova et al., 2008). Figure 2 illustrates this distribution of fluorescence between the epithelium and stroma and summarizes the biological mechanisms that influence tissue fluorescence during dysplastic progression.

Fig. 2 Biological mechanisms responsible for tissue autofluorescence in oral lesions. Blue excitation light causes cellular fluorescence due to reduced form of nicotinamide adenine dinucleotide (NADH) and the oxidized form of flavin adenine dinucleotide (FAD) in the epithelium and collagen fluorescence due to cross-links in the stroma. Overall fluorescence intensity decreases with dysplastic progression as indicated by the biological mechanisms labeled 1 through 5

Tissue and nuclear morphology have a significant impact on fluorescence via scattering of the excitation and emission lights. The morphological changes that

The Application of Tissue Autofluorescence in Detection and Management

105

accompany dysplastic progression impact the scattering properties of the epithelium which in turn modifies the observed fluorescence. Epithelial thickness increases with dysplastic progression due to increased cell division and decreased apoptosis. In addition, nuclear changes observed during dysplastic progression increase nuclear scattering (Collier et al., 2003; Mourant et al., 2000). A thicker epithelium composed of more cells with increased nuclear scattering means that less excitation light will reach the collagen cross-links in the stroma where the bulk of the fluorescence is generated. The same holds true for the fluorescent emission propagating out of the tissue. The increased epithelial thickness and increased nuclear scattering associated with dysplastic progression therefore tend to reduce the intensity of autofluorescence (Fig. 2, callouts 1 and 2). Autofluorescence from collagen cross-links has been shown to decrease in the immediate vicinity of dysplasia (Drezek et al., 2001a). This loss of florescence is generally attributed to changes in collagen biochemistry, possibly due to the breakdown of the extracellular matrix induced by the dysplastic cells. One hypothesis is that matrix metalloproteinase (MMP) expression in host stromal cells and the consequent remodeling of the extracellular matrix is induced by altered signaling from dysplastic epithelial cells (Thomas et al., 1999; Heppner et al., 1996). The collagen remodeling associated with dysplastic progression therefore causes a decrease in observed tissue autofluorescence (Fig. 2, callout 3). Carcinogenesis leads to increased micro-vasculature in the stroma as the colony of dysplastic cells in the epithelium recruits increased blood supply. This increases the concentration of hemoglobin in the stroma. Hemoglobin absorbs strongly in the violet–blue (maximum absorption at 420 nm) and competes with the collagen crosslinks to absorb excitation light. Autofluorescence intensity therefore decreases with the micro-visualization associated with carcinogenesis (Fig. 2, callout 4). The cofactors NADH and FAD associated with respiration and electron transport are good indicators of cellular metabolism and increased cell division. It has been shown, using a confocal fluorescence microscope to observe fluorescence from fresh cervical specimens, that fluorescence intensity due to NADH increases with dysplastic progression and that of FAD decreases (Pavlova et al., 2003; Drezek et al., 2001a). The increased metabolic activity associated with dysplastic progression therefore tends to decrease the autofluorescence due to FAD (Fig. 2, callout 5). Changes in the optical properties of tissue due to inflammation can also lead to a loss of autofluorescence. Increased hemoglobin concentration due to microvascularization and increased blood volume at the site of inflammation causes the absorption of blue excitation light before it can induce collagen fluorescence. Pavlova et al. (2008) have speculated that the reduction in stromal fluorescence in the presence of chronic inflammation could also be linked to the displacement of structural fibers by the infiltrating lymphocytes which are much less fluorescent and also promote the expression of matrix-degrading proteases leading to the breakdown of collagen cross-links. These authors also suggest that inflammation may be discriminated from dysplasia by using UV light to excite NADH fluorescence in addition to that produced by FAD and collagen.

106

C.F. Poh et al.

The five biological mechanisms outlined in this section all involve a loss of collagen or FAD fluorescence with increasing dysplastic progression. Based on the present knowledge of the origins of fluorescence and its change with dysplastic progression, it is thought that the loss of fluorescence is primarily due to breakdown of the collagen matrix and increased hemoglobin absorption. Secondary to these effects is increased scattering in the epithelium, epithelial thickening, and a decrease in FAD concentration.

Current Applications of Tissue Autofluorescence in the Management of Lung, Cervical, and Skin Cancers The visualization of tissue autofluorescence has an interesting history. One of the first reported uses of fluorescence was in dermatology for the identification of skin lesions and infection. In this application, UV light from a Wood’s lamp, a source of UV light at 365 nm invented by Robert Wood in 1903, excited fluorophores on the skin associated with fungal or bacterial infections. The UV light reflected from the skin was invisible, however, the presence of fluorescent light, usually red porphyrin fluorescence, indicated infection and aided in the localization of the lesion. The visualization of exogenous porphyrin or 5-aminolevulinic acid (5-ALA) induced protoporphyrin IX (PPIX) florescence followed for the detection and localization of malignancy. Visualization of infected lesions due to endogenous porphyrins produced by bacteria has also been described in the literature. Visualization of porphyrin fluorescence in the lung was first reported in 1985 (Kato and Cortese, 1985). During a study by Lam and colleagues at the BC Cancer Agency (Vancouver, BC) aimed at porphyrin dose reduction and determination of the optimal excitation wavelength for distinguishing normal from premalignant and malignant bronchial mucosa (Palcic et al., 1991), they discovered the best discrimination resulted from autofluorescence alone (zero porphyrin dose) (Kennedy et al., 2001). Autofluorescence bronchoscopy is now an established clinical technique that is used around the world to address the limitations of white light bronchoscopy, specifically, its limited ability to detect small intraepithelial and microinvasive lesions in the central airway. The first autofluorescence bronchoscopes were developed at the British Columbia Cancer Research Centre (Lam et al., 1993; Palcic et al., 1991) and commercialized in 1998 by Xillix Technologies (Vancouver, BC) as the light-induced fluorescence endoscopy (LIFE) device. The LIFE system used 442-nm excitation light and the red–green fluorescence ratio captured by an intensified camera to image precancers in the lung. Dysplastic and cancerous tissue shows an increase in the red–green fluorescence ratio compared to that of normal tissue. A multi-institutional trial of a total of 700 lesions showed that this device provided a significant 6.2-time increase in relative sensitivity compared to white light bronchoscopy alone for localization of lesions with moderate dysplasia, severe dysplasia, and carcinoma in situ (Lam et al., 1998). The LIFE device has been approved for use in the United States,

The Application of Tissue Autofluorescence in Detection and Management

107

Canada, Europe, and Japan; and other vendors including Karl Storz (D-Light AF), Pentax (Safe 1000), and Richard Wolf (DAFE) have produced similar instruments. The application of tissue autofluorescence to the management of cervical cancer has also induced significant research effort and device commercialization. Several companies are developing optical devices as an adjunct to colposcopy to improve the detection of precancerous cervical abnormalities that have the potential of becoming invasive cancer. In contrast to the lung devices described above, these devices acquire florescence and reflectance imagery of the entire cervix under computer control and then trained algorithms determine disease state using spectral and contextual features extracted from the fluorescence and reflectance imageries. Some devices also employ a point probe to improve specificity at operator-selected points on the cervix. This indirect method of imaging, where spectral data are captured and interpreted under computer control, is in contrast to direct fluorescence visualization where no image is acquired and the fluorescent is interpreted by the operator directly. The LightTouch cervical cancer test, developed by Guided Therapeutics (Norcross, GA), formerly SpectRx, is undergoing pivotal (phase II) clinical trials in anticipation of a premarket approval application to the FDA. Published data from one of these clinical trials (DeSantis et al., 2007) found that the device is capable of detecting more than 95% of CIN 2+, with a corresponding specificity for benign cervices of 55% in a population of 572 women scheduled for colposcopy. The LUMA cervical imaging system for the early detection of cancer, being developed by MediSectra (Lexington, MA), received FDA approval as an aid to clinicians examining women with abnormal Pap tests. A multicenter clinical study (Alvarez et al., 2007) of 193 women who underwent colposcopy followed by LUMA was used to achieve FDA approval. Of the 50 cases of precancer (CIN 2+) detected in the study, colposcopy identified 41 cases, while LUMA detected an additional 9 cases that colposcopy had missed. The hyperspectral diagnostic imaging (HSDI) system, developed by STI Medical Systems (Honolulu, HI), has also undergone FDA clinical trials. One study (Parker et al., 2002) claimed 98% sensitivity and 99% sensitivity for detecting CIN 1 in a population of 33 women with abnormal Pap smears. Remicalm LLC (Houston, TX) is developing the RemiScope for fluorescence imaging and the multi-optical wand (MOW) for point spectroscopy. In pre-clinical trials, the MOW, now in phase II clinical trials, achieved a sensitivity/specificity similar to that of other spectroscopy devices (Nath et al., 2004). A pilot study using the RemiScope reported a sensitivity of 79% and a specificity of 88% for differentiating HGSIL from LGSIL or normal (Ferlay et al., 2004). The lung and cervical devices discussed above employ indirect visualization of tissue autofluorescence because they use a sensitive camera to record the fluorescence which is then presented to the operator as a processed image on a monitor. In contrast, current devices for the oral cavity and skin employ direct visualization of tissue autofluorescence where the emitted fluorescence is optically filtered and relayed directly to the operator’s eye. This straightforward approach reduces device cost, simplifies its application and interpretation, and makes it more suitable as a screening tool.

108

C.F. Poh et al.

The application of direct FV (fluorescence visualization) to dermatology has focused primarily on efforts to better delineate tumor margins for treatment. Lui et al., utilized a diffuse blue light source and separate goggles (glasses with special optical filters) worn by the operator to block the reflected excitation light and pass only the tissue autofluorescence (Zeng et al., 2000; Lui et al., 2001). The goggles employed dual band-pass filters (490–560 nm and >620 nm) to optimally visualize skin autofluorescence. In a study of 41 patients with basal cell carcinoma (BCC), tumor margins were delineated first under white light and then again using direct florescence visualization. Tumors were excised and a two-dimensional image delineating the true histopathologic tumor margin was constructed by systematic sectioning of the entire tumor. The histopathologic margin was compared to the margins delineated under white light and direct FV. Over the entire range of margins tested, fluorescence visualization more accurately estimated the histologic margins of the BCC as compared to standard white light examination.

Potential Clinical Utilities of Tissue Autofluorescence in Oral Cancer and Premalignant Lesions The oral cavity is an ideal site to model the clinical applicability of tissue optics for detection, risk assessment and management of high-risk fields and clinical lesions. This is largely due to its ease of access and the presence of an established histological progression pattern from premalignancy to cancer. Moreover, oral cancer is often identified at a late stage with poor prognosis and treatment is frequently disfiguring and debilitating. There is a pressing clinical need to develop new tools for early detection and management of this disease. In 1999, we established a large longitudinal cohort in British Columbia, Canada, to follow patients with oral cancer and dysplasia over time with a goal toward developing and validating new tools for better control of the disease. Currently the oral cancer prediction longitudinal (OCPL) study has accrued over 700 patients, roughly equal proportions of cancers (at risk of recurrence) and dysplasia (at risk of progression). Three platforms are attached to it: genomic, computer microscopy imaging, and visualization. Among the tools developed within the OCPL study has been a simple handheld field-of-view device for direct visualization of tissue autofluorescence in the R (LED Med. Inc., oral cavity which has since been commercialized as VELScope White Rock, BC). The device uses a band-pass filter centered at 425 nm to filter light from a metal halide lamp (providing an excitation spectrum composed primarily of the 405- and 436-nm peaks), a long-pass filter centered at 475 nm to observe the emission of tissue autofluorescence, and a proprietary image enhancement filter to improve contrast between normal and abnormal tissue (Lane et al., 2006). Under direct FV, the normal oral mucosa emits pale green autofluorescence. Clinical lesions that retained the normal green autofluorescence under direct FV were classified as lesions with FV retained (FVR). Tissue that showed a distinct reduction

The Application of Tissue Autofluorescence in Detection and Management

109

in the normal pale green and appeared as dark brown to black was classified as FV loss (FVL). In the following sections, we will present lessons learned with this device during the follow-up of these high-risk patients and the initialization of studies in community settings. More specifically we will describe the use of direct FV: (1) to detect oral cancers; (2) to detect primary dysplasia; (3) to guide treatment of cancers and high-grade lesions, through delineation of surgical margins; and (4) to monitor patient in follow-up after treatment for recurrent and/or new disease. Finally, we will discuss future directions of research in our group that are aimed at validation of FV within community settings. Detection of oral cancer. Currently the decision of where and when to biopsy an oral mucosal lesion in order to determine whether it is a cancer is based on a set of clinical features: appearance of the lesion, size, color, and location. Oral changes can be ill-defined, diffuse, or patchy; alternatively, they may be clinically occult. The first validation of the use of direct FV was reported for patients in the cohort in 2006 (Lane et al., 2006). This study involved a small subset of 33 patients with invasive squamous cell carcinoma; 6 normal cases were used as controls. All were biopsy confirmed. All of these cancers were FVL (i.e., showed loss of fluorescence). In contrast, there was no change in fluorescence for any of the normal samples. Since that time we have examined ∼120 cancers. For some, we have known diagnosis at initial assessment, since they were being assessed in a referral clinic. Others were identified as they developed in patients in the OCPL study during follow-up. We estimate that ∼95–100% of cancers are detected with this system, showing FVL. Detection of primary dysplasia. It is critical that a visualization device detects not only cancers but also those other lesions that have a high probability of malignant transformation. This would include high-grade lesions (HGLs, severe dysplasia/carcinoma in situ) and a proportion of low-grade lesions (LGLs, mild and moderate dysplasia). We have been assessing the ability of FV to identify such lesions to ensure that they are biopsied for risk assessment by histology and, more recently, with genetic markers. The literature reports an elevated rate of progression for HGLs (Crissman and Zarbo, 1989; Fresko and Lazarus, 1981; Hayward and Regezi, 1977; Summerlin, 1996). In a recent analysis of the OCPL database, we examined 124 HGLs. Eighty (65%) of these lesions had been treated by surgical excision and the remaining lesions were left for follow-up. For patients not receiving treatment, progression rates to cancer were 42%, 56%, and 70% in 2, 3, and 5 years of follow-up, respectively. In addition, data from our molecular (Baldwin et al., 2005; Garnis et al., 2004) and computer microscopy imaging platforms (Rosin et al., 2007) also show a shift to characteristics associated with progression risk for HGLs. A published abstract covering approximately 2 years worth of data in the OCPL database showed that the majority of HGLs were FVL (Zhang et al., 2008). Of the HGLs present at first clinic visit with a diagnosis or which developed into HGL during follow-up, approximately 82 of 83 such cases are FVLs.

110

C.F. Poh et al.

The same abstract showed that 59 of 76 LGLs were FVL. The significance of fluorescence loss in these lesions has yet to be established. To date, we have shown a strong association between FVL and the presence of a loss of heterozygosity profile that is associated with a high risk of progression to oral cancer (Rosin et al., 2000). We are only beginning to see progression in this group for LGLs with FV status with an average follow-up time of only 24 months. This follow-up time is very short since our experience with progression rates for LGLs based on molecular profiles shows approximately 50% of the lesions progressing within 5 years. Delineation of surgical margins of cancers and HGLs. Histological assessment of margin positivity is one of the better predictors of local recurrence after surgery (McMahon et al., 2003; Sutton et al., 2003). More recently, molecular predictors have been added to this assessment of margins (Brennan et al., 1995; Goldenberg et al., 2004). One of the challenges to the surgeon is how to set the margin in the operating room to ensure capture of all high-risk tissue. Current practice is to arbitrarily remove a 10-mm normal-looking oral mucosa border around the clinical cancerous lesion if anatomically possible. However, the frequent positive surgical margins and the high rate of recurrence of carcinomas at the primary site (10–30% of cases) (Brennan et al., 1995; Leemans et al., 1994; Tabor et al., 2004; Tabor et al., 2001; Hittelman et al., 1996; Partridge et al., 2000) indicate the inadequacy of this approach. There is an urgent need to develop new approaches in examining cancerous field for a better management. To date, there are only two articles that have used FV to examine margins around oral cancers. Svistun et al. (2004) examined autofluorescence in tumor margins from freshly resected oral tissue ex vivo with several different excitation and emission wavelengths and illumination at 400 nm, and observation at 530 nm was correlated with histopathologic diagnosis. In November 2006, we published the first study that used direct FV in vivo real time in surgical settings. The objective was to study FVL changes in tumor margins and determine whether such field changes were of high risk, i.e., associated with histological and molecular high-risk patterns (Poh et al., 2006). All cancers showed FV loss (FVL) with this loss extending beyond the clinically visible tumor boundary in at least one direction in 19 of the 20 cancers examined. Of note, this subclinical lateral extension was uneven around the clinical lesion. Also the extension beyond the clinically apparent perimeter varied from 4 to 25 mm. Hence, setting an arbitrary uniform margin would not necessarily control a lesion, since as detailed below this FVL extension involved high-risk change. Strikingly, 32 of 36 FVL margin biopsies (clinically normal) showed histological change, 7 SCC, 10 HGLs, and 15 LGLs. Three of the 20 cancer cases (5 margins) had high-grade histological change present beyond the 10-mm conventional margin. Current management would require salvage surgery to remove such change. In addition, molecular analysis of the FVL margins with low-grade or no dysplasia showed a significant association with loss of heterozygosity (LOH) patterns previously associated with a 26.3-fold increase in local tumour recurrence (Rosin et al., 2002). This suggests that FVL also identifies histologically low-grade margins that contain high-risk molecular clones.

The Application of Tissue Autofluorescence in Detection and Management

111

As an extension of this study, we have also examined FV fields around HGLs going through surgery. There is currently no agreement on the management of such lesions. Using the same protocol, we examined 22 HGLs in surgery with FV status. The data were strikingly similar to those observed for invasive cancers. All lesions showed FVL. The region of loss extended beyond clinical boundaries in 20 (91%) of the cases. As with the invasive cancers, the lateral extension around the clinically apparent lesion was uneven, ranging from 1 to 25 mm (mean, 7.1 ± 5.4 mm). In four cases (five margins), histological assessment of the FVL margins demonstrated the presence of high-grade dysplasia beyond 10 mm. These results support the presence of significant lateral extension of cancer field in both invasive cancer and pre-invasive HGLs. Direct tissue autofluorescence might provide a new way to examine the field alteration around the clinically apparent high-risk lesions in the oral cavity for lateral extension. In British Columbia, we established a consensus among surgeons in 2005 to set the surgical margin at 10 mm beyond the FVL perimeter for cancers and HGLs. Lesions so treated are being followed over time. A demonstration of impact on recurrence for such lesions will be analyzed by comparing outcome historically observed prior to this change in surgical practice. Follow-up of cancer patients after treatment for recurrence or new disease. Patients with a history of oral cancer have a high risk of developing a local recurrence of the tumor or the development of a second primary, especially within the first 2 years, despite intensive follow-up (1–3) (Dhooge et al., 1998; Mashberg and Samit, 1989; Silverman and Gorsky, 1990). Treatment often induces reactive white and red lesions at the previous cancer site that are not readily differentiated from (pre)-malignant changes, hence complicating follow-up. Repeated comparative biopsies for such patients are impractical. We have started to gain experience in using direct FV to examine the former cancer site on a regular basis. We have been focusing on two scenarios: the reappearance of a clinical lesion that may or may not show FV and the appearance of FVL without clinical change (i.e., clinically occult FVL). In the latter case, important determinants are the persistence of FVL over time and subsequent development of clinical lesions and changes to the size and intensity of FVL fields. An interesting case that exemplifies some of the data we are now accruing is shown in Fig. 3. This is a patient with oral cancer treated with surgery (before FV was used to guide surgical margin determinations), although there is some reactive change to the clinical scar immediately after treatment which tends to diminish over time. In the case shown, this did not occur; instead, the FVL became larger in size and increased in intensity, eventually developing into a clinically apparent lesion. The lesion on biopsy was shown to be a carcinoma in situ. This progressive alteration in FVL is an important attribute in evaluating such field change after treatment. A small subset analysis has been completed of 41 FVL sites that persisted after surgical treatment. Thirty-four of these sites later developed clinically visible lesions during follow-up. To date, 21 have been biopsied, with 16 (76%) showing high-grade and 3 (14%) low-grade dysplasia and 2 being

112

C.F. Poh et al.

SCC

D2

1998/06

2002/5

D2

A

C

B

D

CIS

2005/8 2006/2 2006/12

Scar 2008/01

Fig. 3 The use of FV during follow-up for detection of cancer recurrence. A 45-year-old female non-smoker presented with an invasive squamous cell carcinoma on the right lateral tongue in June 1998 and was treated with surgery. The follow-up biopsies at the fourth and the seventh year showed moderate dysplasia (D2). At the eighth year of follow-up (February, 2006) after the initial surgery, (a) white light image showing a scar without clinically visible lesion on the right tongue (arrow); (b) FV image showing some FVL area (arrow). Ten months after, a 7-mm red lesion appeared at the center of the scar (c, arrow) with a significant increase in the size and intensity of the area of FV loss (d, arrow). The biopsy showed carcinoma in situ and the site was treated with FV-guided surgical excision. At the most recent 1-year post-surgery follow-up visit, a wellhealed scar with no sign of local recurrence was seen. Lower panel: Timeline to demonstrate the date (below the line, yyyy/mm) of the follow-up visit with biopsy (color circles) and the diagnosis (above the line, SCC: squamous cell carcinoma; D2: moderate dysplasia; CIS: carcinoma in situ) and the follow-up visit with no biopsy (white circle). The red thunderbolt figures indicate a surgical procedure

nondysplastic. For those (seven cases) without clinical lesion development, all were biopsied and showed three high-grade and three low-grade dysplasia and one being nondysplastic. We are also focusing our efforts on looking at the entire oral mucosa of cancer patients after treatment, to determine whether FV will detect new anatomically separate lesions, possibly second primaries. We have previously reported one such case in 2007 (Poh et al., 2007b). Figure 4 shows another such case in which a patient had both multiple recurrences at the former cancer site and the development of new lesions more than 10 cm from the original site. FVL detected both recurrent and new lesions supporting its usage in individuals with very unstable, widespread disease. Of specific interest in this case was the appearance of two distinct patches of clinical change in the patient’s new lesion with a distinct FVL “strand” linking the two patches (Fig. 4f). This case demonstrates the use of FV in follow-up for mapping and detection of lesions at multiple sites. Thus, following field changes with fluorescence may generate interesting and clinically important information on the biology underlying field cancerization.

The Application of Tissue Autofluorescence in Detection and Management

113

A

B

C

D

E

F

CIS Left Anterior floor of mouth 2004/10 Right posterior Soft palate & retromolar trigone

CIS

CIS

2005/6

2006/4

CIS

2007/11

2007/2

Fig. 4 The use of FV during follow-up for recurrence and new lesions developed at a distant anatomical site. A 58-year-old female smoker developed a carcinoma in situ (CIS) on the anterior floor of mouth in October 2004. This lesion was excised with two recurrences as carcinoma in situ (CIS) over a 2-year period. At the 18-month follow-up visit (November, 2007) after the last surgery, (a) a diagram to map the locations of all three clinical lesions: right posterior soft palate (red), right retromolar trigone (yellow), and left anterior floor of mouth (purple); (b–f) clinical images taken at this time. (b) A white light image showing a discrete white and red lesion (arrow) on the left floor of mouth, recurring at the left lateral margin of the previous surgical site for CIS and it showed FVL (c; arrow). (d) Two new red lesions developed at distant sites: right posterior soft palate (arrowhead, diagnosed as CIS) and right retromolar trigone/ pad extending to lingual aspect (arrow). Both lesions showed FVL (e, arrowhead and arrow) with a distinct FVL “strand” linking the two new patches. (f) Mark the FVL “strand” (dash line) linking the two red patches on the white light image. This case demonstrates the use of FV in follow-up for mapping and detection of lesions at multiple sites. Due to the patient’s medical conditions, further management of these lesions is pending. Lower panel: Timelines to demonstrate the date (below the line, yyyy/mm) of follow-up visits with biopsy (color circles) and the diagnosis (above the line, CIS: Carcinoma in situ) and follow-up visits with no biopsy (white circle). The red thunderbolt figures indicate a surgical procedure

Future Directions As we have indicated above, the full impact of FV on our understanding of the biology of oral cancer and premalignant disease is yet to be determined and its clinical impact is still being explored to a large degree. This technology may provide a mechanism for monitoring lesions over time, before and after treatment, and for looking at clinically non-apparent spread of abnormal cells. It can also be used to quickly monitor the entire oral field for temporal change and the development of

114

C.F. Poh et al.

new regions of concern. This is of major importance for those individuals with wide oral mucosal abnormality who are at high risk for second lesion development. All of the research described above has taken place within high-risk referral clinics. We have also begun to explore the role of direct FV in community settings. This technology is being used in a high-risk complex community in Vancouver’s Downtown Eastside (Poh et al., 2007a). A primary focus of this study is to observe the impact of common oral conditions (e.g., infection or inflammation) which could potentially confound an FV assessment. This is critical to establishing the utility of the device for detection of cancers and high-risk lesions in the community. A second study involves a series of 25 dental practices, with clinicians trained in the use of FV integrated into the conventional oral mucosal examination (Laronde et al., 2008). That study will also look for confounders and is aimed at establishing an educational strategy for community dental practitioners. The latter study will integrate computer-based technology into the assessment of community dentists to help them minimize false positivities especially during the evolution of screening behavior in such clinics. In summary, the data presented in this article indicate that FV is a valuable tool, with potential for expanding yet again the concept of field. It has already demonstrated clinical impact both in the management of high-risk patients and the need in community settings. Acknowledgements The authors would like to acknowledge the funding support from the National Institute of Dental and Craniofacial Research (R01DE13124 and R01DE17013), from the Canadian Institutes of Health Research (MOP-77663), and from the Canadian Cancer Society (CSS-20336) and a Clinician Scientist Award from the Canadian Institutes of Health Research and Scholar Award from Michael Smith Foundation for Health Research (C.F. Poh).

References R. D. Alvarez, T. C. Wright, and O. D. Grp, “Increased detection of high-grade cervical intraepithelial neoplasia utilizing an optical detection system as an adjunct to colposcopy,” Gynecologic Oncology, vol. 106, pp. 23–28, 2007. C. Baldwin, C. Garnis, L. Zhang, M. P. Rosin, and W. L. Lam, “Multiple microalterations detected at high frequency in oral cancer,” Cancer Research, vol. 65, pp. 7561–7, 2005. J. A. Brennan, L. Mao, R. H. Hruban, J. O. Boyle, Y. J. Eby, W. M. Koch, S. N. Goodman, and D. Sidransky, “Molecular assessment of histopathological staging in squamous-cell carcinoma of the head and neck,” New England Journal of Medicine, vol. 332, pp. 429–35, 1995. J. Califano, P. van der Riet, W. Westra, H. Nawroz, G. Clayman, S. Piantadosi, R. Corio, D. Lee, B. Greenberg, W. Koch, and D. Sidransky, “Genetic progression model for head and neck cancer: implications for field cancerization,” Cancer Research, vol. 56, pp. 2488–92, 1996. S. K. Chang, M. Follen, A. Malpica, U. Utzinger, G. Staerkel, D. Cox, E. N. Atkinson, C. MacAulay, and R. Richards-Kortum, “Optimal excitation wavelengths for discrimination of cervical neoplasia,” IEEE Transactions on Biomedical Engineering, vol. 49, pp. 1102–111, 2002. T. Collier, D. Arifler, A. Malpica, M. Follen, and R. Richards-Kortum, “Determination of epithelial tissue scattering coefficient using confocal microscopy,” IEEE Journal of Selected Topics in Quantum Electronics, vol. 9, pp. 307–313, 2003.

The Application of Tissue Autofluorescence in Detection and Management

115

J. D. Crissman and R. J. Zarbo, “Dysplasia, in situ carcinoma, and progression to invasive squamous cell carcinoma of the upper aerodigestive tract,” American Journal of Surgical Pathology, vol. 13 Suppl 1, pp. 5–16, 1989. T. DeSantis, N. Chakhtoura, L. Twiggs, D. Ferris, M. Lashgari, L. Flowers, M. Faupel, S. Bambot, S. Raab, and E. Wilkinson, “Spectroscopic imaging as a triage test for cervical disease: a prospective multicenter clinical trial,” Journal of Lower Genital Tract Disease, vol. 11, pp. 18–24, 2007. I. J. Dhooge, M. De Vos, and P. B. Van Cauwenberge, “Multiple primary malignant tumors in patients with head and neck cancer: results of a prospective study and future perspectives,” Laryngoscope, vol. 108, pp. 250–6, 1998. R. Drezek, C. Brookner, I. Pavlova, I. Boiko, A. Malpica, R. Lotan, M. Follen, and R. RichardsKortum, “Autofluorescence microscopy of fresh cervical-tissue sections reveals alterations in tissue biochemistry with dysplasia,” Photochemistry and Photobiology, vol. 73, pp. 636–41, 2001a. R. Drezek, K. Sokolov, U. Utzinger, I. Boiko, A. Malpica, M. Follen, and R. Richards-Kortum, “Understanding the contributions of NADH and collagen to cervical tissue fluorescence spectra: modeling, measurements, and implications,” Journal of Biomedical Optics, vol. 6, pp. 385–96, 2001b. J. Ferlay, F. Bray, P. Pisani, and D. M. Parkin, “GLOBOCAN 2002: cancer incidence, mortality and prevalence worldwide,” IARC CancerBase No. 5. version 2.0 ed. Lyon: IARC Press, 2004. D. Fresko and S. S. Lazarus, “Oral carcinoma in situ. Its progression to squamous, basosquamous, and basal-cell carcinoma,” Archives of Pathology & Laboratory Medicine, vol. 105, pp. 15–9, 1981. C. Garnis, B. P. Coe, A. Ishkanian, L. Zhang, M. P. Rosin, and W. L. Lam, “Novel regions of amplification on 8q distinct from the MYC locus and frequently altered in oral dysplasia and cancer,” Genes Chromosomes Cancer, vol. 39, pp. 93–8, 2004. A. Gillenwater, R. Jacob, R. Ganeshappa, B. Kemp, A. K. El-Naggar, J. L. Palmer, G. Clayman, M. F. Mitchell, and R. Richards-Kortum, “Noninvasive diagnosis of oral neoplasia based on fluorescence spectroscopy and native tissue autofluorescence,” Archives of OtolaryngologyHead & Neck Surgery, vol. 124, pp. 1251–8, 1998. D. Goldenberg, S. Harden, B. G. Masayesva, P. Ha, N. Benoit, W. H. Westra, W. M. Koch, D. Sidransky, and J. A. Califano, “Intraoperative molecular margin analysis in head and neck cancer,” Archives of Otolaryngology–Head & Neck Surgery, vol. 130, pp. 39–44, 2004. J. R. Hayward and J. A. Regezi, “Oral dysplasia and in situ carcinoma: clinicopathologic correlations of eight patients,” Journal of Oral Surgery, vol. 35, pp. 756–62, 1977. D. L. Heintzelman, U. Utzinger, H. Fuchs, A. Zuluaga, K. Gossage, A. M. Gillenwater, R. Jacob, B. Kemp, and R. R. Richards-Kortum, “Optimal excitation wavelengths for in vivo detection of oral neoplasia using fluorescence spectroscopy,” Photochem Photobiol, vol. 72, pp. 103–13, 2000. K. J. Heppner, L. M. Matrisian, R. A. Jensen, and W. H. Rodgers, “Expression of most matrix metalloproteinase family members in breast cancer represents a tumor-induced host response,” American Journal of Pathology, vol. 149, pp. 273–82, 1996. W. N. Hittelman, H. J. Kim, J. S. Lee, D. M. Shin, S. M. Lippman, J. Kim, J. Y. Ro, and W. K. Hong, “Detection of chromosome instability of tissue fields at risk: in situ hybridization,” Journal of Cellular Biochemistry. Supplement, vol. 25, pp. 57–62, 1996. D. R. Ingrams, J. K. Dhingra, K. Roy, D. F. Perrault, I. D. Bottrill, S. Kabani, E. E. Rebeiz, M. M. Pankratov, S. M. Shapshay, R. Manoharan, I. Itzkan, and M. S. Feld, “Autofluorescence characteristics of oral mucosa,” Head and Neck-Journal for the Sciences and Specialties of the Head and Neck, vol. 19, pp. 27–32, 1997. H. Kato and D. A. Cortese, “Early detection of lung-cancer by means of hematoporphyrin derivative fluorescence and laser photoradiation,” Clinics in Chest Medicine, vol. 6, pp. 237–53, 1985.

116

C.F. Poh et al.

T. C. Kennedy, S. Lam, and F. R. Hirsch, “Review of recent advances in fluorescence bronchoscopy in early localization of central airway lung cancer,” Oncologist, vol. 6, pp. 257–62, 2001. S. Lam, T. Kennedy, M. Unger, Y. E. Miller, D. Gelmont, V. Rusch, B. Gipe, D. Howard, J. C. LeRiche, A. Coldman, and A. F. Gazdar, “Localization of bronchial intraepithelial neoplastic lesions by fluorescence bronchoscopy,” Chest, vol. 113, pp. 696–702, 1998. S. Lam, C. MacAulay, J. Hung, J. LeRiche, A. E. Profio, and B. Palcic, “Detection of dysplasia and carcinoma in situ with a lung imaging fluorescence endoscope device,” Journal of Thoracic and Cardiovascular Surgery, vol. 105, pp. 1035–40, 1993. P. M. Lane, T. Gilhuly, P. Whitehead, H. Zeng, C. F. Poh, S. Ng, P. M. Williams, L. Zhang, M. P. Rosin, and C. E. MacAulay, “Simple device for the direct visualization of oral-cavity tissue fluorescence,” Journal of Biomedical Optics, vol. 11, pp. 024006, 2006. D. M. Laronde, J. L. Bottorff, T. G. Hislop, C. Y. Poh, B. Currie, P. M. Williams, and M. P. Rosin, “Voices from the community–experiences from the dental office: initiating oral cancer screening,” Journal of the Canadian Dental Association, vol. 74, pp. 239–41, 2008. C. R. Leemans, R. Tiwari, J. J. Nauta, I. van der Waal, and G. B. Snow, “Recurrence at the primary site in head and neck cancer and the significance of neck lymph node metastases as a prognostic factor,” Cancer, vol. 73, pp. 187–90, 1994. H. Lui, S. Said, L. Warshawski, D. Zloty, D. McLean, C. MacAulay, and H. Zeng, “Fluorescence visualization with blue light more accurately estimates the histopathologic margins of basal cell carcinoma as compared to clinical examination alone,” presented at European Conferences on Biomedical Optics, Munich, Germany, 2001. A. Mashberg and A. M. Samit, “Early detection, diagnosis, and management of oral and oropharyngeal cancer,” CA: A Cancer Journal for Clinicians, vol. 39, pp. 67–88, 1989. J. McMahon, C. J. O Brien, I. Pathak, R. Hamill, E. McNeil, N. Hammersley, S. Gardiner, and E. Junor, “Influence of condition of surgical margins on local recurrence and disease-specific survival in oral and oropharyngeal cancer,” British Journal of Oral & Maxillofacial Surgery, vol. 41, pp. 224–31, 2003. J. R. Mourant, M. Canpolat, C. Brocker, O. Esponda-Ramos, T. M. Johnson, A. Matanock, K. Stetter, and J. P. Freyer, “Light scattering from cells: the contribution of the nucleus and the effects of proliferative status,” Journal of Biomedical Optics, vol. 5, pp. 131–7, 2000. M. G. Muller, T. A. Valdez, I. Georgakoudi, V. Backman, C. Fuentes, S. Kabani, N. Laver, Z. Wang, C. W. Boone, R. R. Dasari, S. M. Shapshay, and M. S. Feld, “Spectroscopic detection and evaluation of morphologic and biochemical changes in early human oral carcinoma,” Cancer, vol. 97, pp. 1681–92, 2003. A. Nath, K. Rivoire, S. Chang, L. West, S. B. Cantor, K. Basen-Engquist, K. Adler-Storthz, D. D. Cox, E. N. Atkinson, G. Staerkel, C. Macaulay, R. Richards-Kortum, and M. Follen, “A pilot study for a screening trial of cervical fluorescence spectroscopy,” International Journal of Gynecological Cancer, vol. 14, pp. 1097–107, 2004. B. Palcic, S. Lam, J. Hung, and C. Macaulay, “Detection and localization of early lung-cancer by imaging techniques,” Chest, vol. 99, pp. 742–3, 1991. M. F. Parker, G. C. Mooradian, G. S. Okimoto, D. M. O Connor, K. Miyazawa, and S. J. Saggese, “Initial neural net construction for the detection of cervical intraepithelial neoplasia by fluorescence imaging,” American Journal of Obstetrics and Gynecology, vol. 187, pp. 398–402, 2002. M. Partridge, S. Pateromichelakis, E. Phillips, G. G. Emilion, R. P. A’Hern, and J. D. Langdon, “A case-control study confirms that microsatellite assay can identify patients at risk of developing oral squamous cell carcinoma within a field of cancerization,” Cancer Research, vol. 60, pp. 3893–8, 2000. I. Pavlova, K. Sokolov, R. Drezek, A. Malpica, M. Follen, and R. Richards-Kortum, “Microanatomical and biochemical origins of normal and precancerous cervical autofluorescence using laser-scanning fluorescence confocal microscopy,” Photochemistry and Photobiology, vol. 77, pp. 550–5, 2003.

The Application of Tissue Autofluorescence in Detection and Management

117

I. Pavlova, M. Williams, A. El-Naggar, R. Richards-Kortum, and A. Gillenwater, “Understanding the biological basis of autofluorescence imaging for oral cancer detection: high-resolution fluorescence microscopy in viable tissue,” Clinical Cancer Research, vol. 14, pp. 2396–404, 2008. C. F. Poh, G. Hislop, B. Currie, R. Lee, S. Sikorski, C. Zed, L. Zhang, C. Macaulay, and M. P. Rosin, “Oral cancer screening in a high-risk underserved community–Vancouver Downtown Eastside,” Journal of Health Care for the Poor and Underserved, vol. 18, pp. 767–78, 2007a. C. F. Poh, S. P. Ng, P. M. Williams, L. Zhang, D. M. Laronde, P. Lane, C. Macaulay, and M. P. Rosin, “Direct fluorescence visualization of clinically occult high-risk oral premalignant disease using a simple hand-held device,” Head Neck, vol. 29, pp. 71–6, 2007b. C. F. Poh, L. Zhang, D. W. Anderson, J. S. Durham, P. M. Williams, R. W. Priddy, K. W. Berean, S. Ng, O. L. Tseng, C. MacAulay, and M. P. Rosin, “Fluorescence visualization detection of field alterations in tumor margins of oral cancer patients,” Clinical Cancer Research, vol. 12, pp. 6716–22, 2006. N. Ramanujam, M. F. Mitchell, A. Mahadevan, S. Warren, S. Thomsen, E. Silva, and R. Richardskortum, “In-vivo diagnosis of cervical intraepithelial neoplasia using 337-nm-excited laser-induced fluorescence,” Proceedings of the National Academy of Sciences of the United States of America, vol. 91, pp. 10193–7, 1994. R. Richards-Kortum and E. Sevick-Muraca, “Quantitative optical spectroscopy for tissue diagnosis,” Annual Review of Physical Chemistry, vol. 47, pp. 555–606, 1996. M. P. Rosin, X. Cheng, C. Poh, W. L. Lam, Y. Huang, J. Lovas, K. Berean, J. B. Epstein, R. Priddy, N. D. Le, and L. Zhang, “Use of allelic loss to predict malignant risk for low-grade oral epithelial dysplasia,” Clinical Cancer Research, vol. 6, pp. 357–62, 2000. M. P. Rosin, W. L. Lam, C. Poh, N. D. Le, R. J. Li, T. Zeng, R. Priddy, and L. Zhang, “3p14 and 9p21 loss is a simple tool for predicting second oral malignancy at previously treated oral cancer sites,” Cancer Research, vol. 62, pp. 6447–50, 2002. M. P. Rosin, C. F. Poh, M. Guillard, P. M. Williams, L. Zhang, and C. MacAulay, “Visualization and other emerging technologies as change makers for oral cancer prevention,” Annals of the New York Academy of Sciences, vol. 1098, pp. 1–17, 2007. S. Silverman, Jr. and M. Gorsky, “Epidemiologic and demographic update in oral cancer: California and national data–1973 to 1985,” Journal of the American Dental Association, vol. 120, pp. 495–9, 1990. D. P. Slaughter, H. W. Southwick, and W. Smejkal, “Field cancerization in oral stratified squamous epithelium; clinical implications of multicentric origin,” Cancer, vol. 6, pp. 963–8, 1953. K. Sokolov, J. Galvan, A. Myakov, A. Lacy, R. Lotan, and R. Richards-Kortum, “Realistic threedimensional epithelial tissue phantoms for biomedical optics,” Journal of Biomedical Optics, vol. 7, pp. 148–56, 2002. D. J. Summerlin, “Precancerous and cancerous lesions of the oral cavity,” Dermatologic Clinics, vol. 14, pp. 205–23, 1996. D. N. Sutton, J. S. Brown, S. N. Rogers, E. D. Vaughan, and J. A. Woolgar, “The prognostic implications of the surgical margin in oral squamous cell carcinoma,” International Journal of Oral and Maxillofacial Surgery, vol. 32, pp. 30–4, 2003. E. Svistun, R. Alizadeh-Naderi, A. El-Naggar, R. Jacob, A. Gillenwater, and R. Richards-Kortum, “Vision enhancement system for detection of oral cavity neoplasia based on autofluorescence,” Head and Neck, vol. 26, pp. 205–15, 2004. M. P. Tabor, R. H. Brakenhoff, H. J. Ruijter-Schippers, J. A. Kummer, C. R. Leemans, and B. J. Braakhuis, “Genetically altered fields as origin of locally recurrent head and neck cancer: a retrospective study,” Clinical Cancer Research, vol. 10, pp. 3607–13, 2004. M. P. Tabor, R. H. Brakenhoff, V. M. van Houten, J. A. Kummer, M. H. Snel, P. J. Snijders, G. B. Snow, C. R. Leemans, and B. J. Braakhuis, “Persistence of genetically altered fields in head and neck cancer patients: biological and clinical implications,” Clinical Cancer Research, vol. 7, pp. 1523–32, 2001.

118

C.F. Poh et al.

G. T. Thomas, M. P. Lewis, and P. M. Speight, “Matrix metalloproteinases and oral cancer,” Oral Oncology, vol. 35, pp. 227–33, 1999. H. Zeng, D. I. McLean, C. MacAulay, and H. Lui, “Autofluorescence properties of skin and applications in dermatology,” Proceedings of the SPIE – The International Society for Optical Engineering Biomedical Photonics and Optoelectronic Imaging, 8–10 Nov. 2000, vol. 4224, pp. 366–73, 2000. L. Zhang, S. Ng, C. F. Poh, P. M. Williams, D. M. Laronde, C. MacAulay, and M. P. Rosin, “Fluorescence visualization identifies primary oral premalignant lesions (OPL) with high-risk molecular patterns,” presented at the 99th Annual Meeting of the American Association for Cancer Research, San Diego, 2008.

Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy J. Oliver McIntyre and Lynn M. Matrisian

Introduction: Proteases and Cancer A variety of physiological processes such as wound healing and tissue remodeling are mediated by a plethora of proteinases – enzymes that can hydrolyze peptide bonds – of which as many as 622 have been identified in the human genome. These proteinases are classified into five of the seven clans of peptidases with known catalytic type: S (serine), C (cysteine), A (aspartyl), M (metallo), and T (threonine) [MEROPS, http://merops.sanger.ac.uk; (Rawlings et al. 2008) (Fig. 1)]. In many physiological processes, the proteinases mediate and/or regulate both intercellular signaling, such as in the release and/or processing of chemokines, and intracellular pathways, such as in the apoptotic pathways leading to programmed cell death. Dysregulation of the temporal and/or spatial co-ordination of these intracellular and/or intercellular pathways disrupts the normal physiology and rhythm of life that can be manifest in unregulated growth such as occurs in tumors and their metastatic progeny. The evolving revelation of the diverse range of biological functions of the proteinases in normal growth and development of multicellular organisms has been accompanied by recognition of the significance of a variety of proteinases and proteolytic cascades in the pathophysiology of cancers. Both intracellular and extracellular proteinases from a broad range of the 72 families of proteases in the human degradome [(Barrett et al. 1998; Rawlings et al. 2008) (Fig. 1)] are now known to either contribute to or are implicated in various aspects of tumor growth, invasion, and metastasis. In addition, more recent work, based primarily on loss-of-function animal models, has identified a number of proteinases, including members of the cysteine, serine, and metallo clans, that have roles in tumor suppression [reviewed in (Lopez-Otin and Matrisian 2007)]. Cancers that progress to metastatic disease have a poor prognosis and are life threatening. A critical step in this pathological pathway to metastasis is the breaching of the basement membrane that permits escape of cells from a primary tumor J. Oliver McIntyre (B) Department of Cancer Biology, Vanderbilt–Ingram Cancer Center and Vanderbilt University Institute of Imaging Science, Vanderbilt University, Medical Center, Nashville, TN, USA e-mail: [email protected] E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_7, 

119

120

J.O. McIntyre and L.M. Matrisian

Fig. 1 Protease classification of the human degradome. The human proteolytic enzymes are classified into cysteine, metallo, serine, threonine (T), and aspartyl (A) clans as enumerated in each segment. The enzymes are further subdivided into families as indicated by the number in parentheses. The number closest to the center of the chart indicates the number of human enzymes in the clan. Based on data from MEROPS (http://merops.sanger.ac.uk )

into the circulation and/or lymphatic system. In this regard, the metastatic process can be viewed as a dysregulation of the complex interplay between the cellular components of tissues and their surrounding matrix. In multicellular organisms, such interactions between cells and their environment, including proteins and other components of the matrix, remain poorly understood though it is these kinds of interactions that define not only the composition but also the size and shape of tissues, organs, and whole organisms. A significant advance in understanding the biochemical and physiological processes involved in such cell–matrix interactions was the discovery more than 40 years ago of collagenase (now called collagenase-1 or MMP-1), a proteinase that was involved in the resorption of the tadpole tail during morphogenesis (Gross and Lapiere 1962). Collagenase is now recognized as matrix metalloproteinase (MMP), a family of extracellular, zinc-dependent M-proteinases capable of degrading all components of the extracellular matrix [(Brinckerhoff and Matrisian 2002; Woessner and Nagase 2000) for reviews] that, in humans, constitute 24 distinct gene products. Lance Liotta’s pioneering work in the late 1970s (Liotta et al. 1980) indicated that the degradation of collagen in the basement membrane is an important component of tumor invasion and metastasis. It is now recognized that the MMPs are just one of several proteinase families to participate in matrix degradation. Virtually all MMP family members have been associated with tumor growth or progression, but particularly well known are the gelatinases MMP-2 and MMP-9 (Egeblad and Werb 2002). The ADAMS (a disintegrin and metalloproteinase) and ADAMTS (ADAM with thrombospondin motif) M-proteases, with a total of more than 50 members, participate in extracellular proteolysis of the polymeric/fibrillar and non-fibrillar matrix proteins as well as non-matrix proteins (Rocks et al. 2008). The cathepsins, lysosomal enzymes of the C- or A-protease clans, have also been implicated in matrix degradation and tumor progression (Herszenyi et al. 2000; Koblinski et al. 2000; Lecaille et al. 2002). The C-protease cathepsin K is produced by osteoclasts and degrades bone matrix components in specialized extracellular

Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy

121

acidic compartments (Bromme and Kaleta 2002; Lecaille et al. 2002). Cathepsin D, an A-protease, has long been associated with breast cancer progression (Duffy 1996; Rochefort et al. 2000). The C-protease cathepsin B is expressed in a number of steps in malignant progression being implicated in tumor–stromal interactions and matrix degradation as well as neovascularization and angiogenesis (Koblinski et al. 2000). Likewise, the S-proteases, particularly those of the S1 or trypsin-like family that contribute to normal homeostasis, have been implicated in pathological processes including cancer (Netzel-Arnett et al. 2003). Notably, the plasminogen/plasmin system participates in tissue remodeling and extracellular matrix degradation and is one of the main proteolytic cascades involved in tumor cell invasion and metastasis (Berger 2002). For example, the urokinase-type plasminogen activator (UPA), an S-protease in the plasminogen/plasmin pathway, has been implicated in gastrointestinal neoplastic disease (Herszenyi et al. 2000). Other S-proteases implicated in cancer include the more recently identified membrane-anchored proteinases that appear to perform complex regulatory cellular signaling functions both at the plasma membrane and within the extracellular matrix but exhibit dysregulation in tumors (Netzel-Arnett et al. 2003; Caughey 2007). In particular, a group of type II integral membrane proteinases, including seprase and hepsin, that interact with a variety of membrane-associated molecules and substrates, appear to localize at cell surface protrusions called invadopodia and play a prominent role in cell migration and matrix invasion, processes that are essential for tumor invasion, angiogenesis, and metastasis (Chen and Kelly 2003; Kelly et al. 2008). For many of these proteases, the activation process is mediated by a proteolytic cascade, e.g., plasmin and stromelysin-1 (MMP-3) cooperate to produce fully activated collagenase from procollagenase (Brinckerhoff and Matrisian 2002). Based on the co-localization of the serine proteinases, MMPs and cathepsin B, it has been postulated that these kinds of tumor-associated extracellular proteinases participate in proteolytic cascades on the tumor cell surface (Overall and Dean 2006) that may also contribute to the pathophysiology of disease progression in cancer. The role of the tumor microenvironment in the establishment of metastasis, and the determination of site-specific metastasis, was recognized in the late 1800s in Paget’s “seed and soil” hypothesis. While the molecular determinants of the “seed” are much better delineated than those of the “soil” for either primary or metastatic lesions, work over the last decade has begun to identify key players, including proteases, in the tumor microenvironment that contribute to tumor progression and/or organ-specific metastatic disease. There is growing recognition of the importance of an inflammatory microenvironment in promoting cancer initiation and progression (Coussens and Werb 2002). The recruitment of tumor-infiltrating bone-marrow-derived cells that contribute to the inflammatory response such as macrophage, neutrophils, mast cells, and eosinophils, appears to promote malignancy by releasing a variety of factors including extracellular proteases (Melnikova and Bar-Eli 2007). Indeed, tumor-associated macrophages that secrete a variety of factors including matrix proteinases appear to be key regulators in the link between inflammation and cancer (Sica et al. 2008)). Thus proteinases play key roles in a number of steps in cancer growth, progression, and metastasis. These enzymes, as

122

J.O. McIntyre and L.M. Matrisian

expressed within the milieu of the tumor microenvironment, present as prime targets not only for in vivo detection and imaging of tumors but also for assessing their response to therapeutic intervention. Although the current state of knowledge predominantly associates the enhanced expression/activity of proteases with cancer progression and poor prognosis, more than 30 proteolytic enzymes negatively regulate some aspect of cancer and thus appear to function, at least in some settings, as tumor suppressors (Lopez-Otin and Matrisian 2007). These include both intracellular proteases, such as caspases, deubiquitylases, and autophagins, and extracellular proteases, particularly a number of M-proteases (MMPs, ADAMTSs), C-proteases (cathepsins), and S-proteases (kallikreins). For many of these proteases, however, tumor suppression may be dependent on the biological context, e.g., MMP12 expressed by tumor cells is associated with poor outcome whereas its expression by host macrophages confers a good prognosis (Kerkela et al. 2002; Lopez-Otin and Matrisian 2007). Proteases have been validated as drug targets, with many examples of highly effective protease inhibitors in the pharmaceutical armamentarium, including angiotensin-converting enzyme inhibitors and HIV protease inhibitors. For the treatR ) has been approved for the treatment of ment of cancer, bortezomib (Velcade multiple myeloma and mantle cell lymphoma (www.velcade.com). Bortezomib binds slowly and reversibly inhibits the chymotrypsin-like activity of the 26S proteosome, and with lower affinity also targets the caspase-like activity (Orlowski and Kuhn 2008). However, clinical trials with small molecule inhibitors of the MMPs failed to demonstrate efficacy and, in some cases, patients treated with the MMP inhibitor fared worse than placebo-controlled patients (Coussens et al. 2002). These results may be explained, at least in part, by the complexity of the biological activities of MMP substrates. Despite the limitations on therapeutic targeting of tumor proteases, the expression of proteases by cancer cells and within the immediate tumor microenvironment presents an opportunity for visualization of the tumor, prognostic information, assessment of response to therapy, and providing assistance in the identification of tumor margins in surgical resection. This chapter focuses on advances in optical visualization of proteinases in malignant and pre-malignant tissues.

Imaging Proteinases A variety of imaging modalities are utilized for the clinical detection and imaging of either primary tumors or metastatic disease for both staging cancer and selecting appropriate therapeutic intervention. While response to therapy is generally assessed by measuring a reduction in tumor size, it is now recognized that noninvasive imaging of tissue functions can provide an early indication of response, particularly with some of the molecular-targeted and/or anti-angiogenic therapies (Brindle 2008). Utilizing proteinases as targets for in vivo imaging has been developing over the past decade with several probes now having been designed for optical imaging, a modality that is particularly useful for imaging tumors in small animal

Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy

123

models. In addition, as summarized in a recent review (Scherer et al. 2008a), a few probes have been reported that were designed to image proteases by bioluminescence, positron emission tomography (PET), single photon emission computed tomography (SPECT), or magnetic resonance imaging (MRI), including an MMPtargeted novel magnetic resonance imaging contrast agent with a solubility switch (Lepage et al. 2007). These developments demonstrate the potential for multi-modal imaging of tumor-associated proteases using a variety of well-established clinical imaging methods. The strategies for imaging proteases have been guided by the biological regulation of their catalytic activity. Most proteases are synthesized as inactive zymogens with activity being regulated predominantly by activation induced by a variety of stimuli rather than at the level of transcription and/or translation. In addition, catalytic activities are often further modulated by interaction with specific proteinbased inhibitors. Thus, significant efforts have been made to develop tools that allow direct monitoring of protease activity in the context of their native biological environment. This includes substrates whose processing by a protease can be easily monitored to assess substrate turnover, binding specificity, and enzyme efficiency (Dive et al. 2008). Modifications of simple fluorogenic substrates have yielded a variety of substrate-based probes that have been used to visualize proteolytic activity in living cells, tissues, and in vivo in small animals, taking advantage of the signal amplification afforded by the enzyme catalysis (Scherer et al. 2008a). An alternative strategy aimed at enhancing the imaging selectivity has been the development of activity-based probes (ABPs) based on small molecule protease inhibitors making use of highly selective reactive functional groups to limit the complexity of protease targeted by a single probe (Kato et al. 2005; Sadaghiani et al. 2007a; Fonovic and Bogyo 2007). A number of new probes and techniques are being developed for molecular imaging of proteases. These can be applied to give both temporal and spatial resolutions in the context of disease progression and have potential for assessing response to therapy.

Optical Imaging Optical imaging is capable of evaluating a number of in vivo processes with the mechanism of contrast generally requiring the accumulation of a fluorescent reagent at the target site. A multitude of fluorescently labeled probes have been developed that target cell surface receptors, enzyme biodistribution, protein function, and gene regulation (Van de Wiele and Oltenfreiter 2006). The main objective of optical imaging is to accumulate fluorophores at a targeted region that upon excitation emit photons. Since the detection of response is obtained by probing the organism or target tissue with light, optical methods provide minimally invasive detection of cancer. The method is made practical by the development of fluorescent probes that are selective and specific for particular biological targets. Sensitivity of optical fluorescence imaging is further improved by using chromophores that both absorb and fluoresce in the NIR region of the electromagnetic spectrum where tissue has both

124

J.O. McIntyre and L.M. Matrisian

low absorption and reduced scattering (Weissleder and Ntziachristos 2003). Two main strategies have been developed for imaging proteases: the selective cleavage of a substrate (substrate-based imaging agents) and the use of small molecule inhibitors designed with high selectivity for specific proteases (activity-based probes, ABPs).

Substrate-Based Imaging Agents Fluorogenically labeled substrates have been designed that are quenched due to the proximity of the fluorophores or that utilize Förster Resonance Energy Transfer (FRET) to quench the fluorescent signal that is then enhanced upon proteolytic cleavage (McIntyre et al. 2004; McIntyre and Matrisian 2003; Tsien 2005). The in vivo optical detection and imaging of protease activity was first demonstrated less than a decade ago by Weissleder and colleagues in mouse xenograft tumors (Weissleder et al. 1999). The optical contrast agents developed utilized near-infrared (NIR) fluorophores as optical sensors attached to a linear polylysine–polyethylene glycol copolymer. The proximity of the fluorophores on the polymer substrate quenched the fluorescence (homo-FRET) with fluorescence being enhanced following proteolytic cleavage of the poly-lysine peptide linker, producing an optically detected near-infrared fluorescence (NIRF) signal associated with the tumor. Optimizing substrates specific for MMPs, with a similar concept of off-toon fluorescence, has been used to develop self-quenched and FRET-pair proteolytic beacons to specifically image tumor-associated MMPs. Bremer et al. (2001) developed the first probes capable of imaging MMP activity using a NIR FRET substrate-based probe containing the peptide sequence GPLGVRGK designed to detect MMP-2 activity. Although the substrate could also be cleaved by other MMPs and proteases, enhanced fluorescence was demonstrated in HT1080 human fibrosarcoma xenografts, with response being inhibited by treatment with a synthetic MMP inhibitor (Bremer et al. 2001). This probe served as the first reporter probe to detect MMP activity demonstrating the ability to sense and image MMP response directly in vivo. Interestingly, a peptide-based NIRF probe, quenched by heterotransfer to a NIR absorber and designed to detect MMP7 activity (Pham et al. 2004), appears to also provide detection of tumor-associated MMP activity without the use of a polymer delivery vehicle (Wellington Pham, personal communication). In more recent studies, the polymer-based protease substrates developed by the Weissleder group have been used to assess protease activities in murine arthritis (Izmailova et al. 2007), in cardiovascular disease (Jaffer et al. 2007), and for early detection of colonic adenocarcinomas in mice using NIR microcatheter imaging (Alencar et al. 2007). The first MMP probes developed lacked the ability to quantitatively assess and efficiently determine specific MMP activity. McIntyre et al. have generated optical proteolytic beacons (PBs) built on a dendrimeric polymer core and designed initially for MMP detection (McIntyre and Matrisian 2003; McIntyre et al. 2004). The prototype visible-range proteolytic beacon for MMP-7, PBvisM7, consisted of R PAMAM dendrimer (Generations 4, nominal MW, 14,215) coupled to Starburst a substrate peptide labeled at its N-terminal with fluorescein (FL) (McIntyre et al.

Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy

125

2004). This fluorogenic substrate was designed to image MMP7 activity using a FRET pair as sensor and also included an internal reference fluorophore (McIntyre et al. 2004). The internal reference of the beacon allows the quantification of both the cleaved and the uncleaved substrates providing a means to directly correlate MMP activity with fluorescent signal. The substrate used in the beacon, FL(AHX)RPLA∗LWRS(AHX)C-COOH (asterisk denotes MMP7 cleavage site), was shown to be more selective for MMP7 than other MMPs that are often active within the tumor microenvironment. A caproyl linker (AHX) was included adjacent to the N-terminal FL so as to diminish the solubility of the FL-RPLA peptide produced by proteolysis. The PB demonstrated differential response between MMP7positive and MMP7-negative xenograft tumors in vivo (McIntyre et al. 2004). The more recently reported NIR version of the MMP7-selective PB, PBnirM7, uses AF700 or Cy5.5 as the sensor in place of FL and AF750 instead of TMR for the reference (Fig. 2). This beacon thus has a NIR FRET pair that provides for reduced tissue absorption and scattering of both excitation and emission photons

Fig. 2 Diagrammatic structure (not to scale) of a proteolytic beacon (PB). PB-M7NIR is constructed on a PEGylated-PAMAM generation-4 dendrimer (ethylenediamine core) with R Alexafluor 750 (AF750, internal reference) linked to surface amines (not shown) and cyanine5.5 (Cy5.5)-labeled peptide proteinase sensors, Cy5.5(Ahx)RPLA∗LWRS(Ahx)C-. Ahx is aminohexanoic acid and the peptide linker is a substrate for MMP7

126

J.O. McIntyre and L.M. Matrisian

Fig. 3 Fluorescence imaging of tumor-associated MMP7 activity in mouse tumors using a proteolytic beacon (PB). Images (Xenogen IVIS200) show white light photographs overlaid with fluorescence images of the Cy5.5 sensor channel imaged at either ∼4 h (panel A) or ∼1 h (panel B and scale bar) after i.v. injection of MMP7-selective PB-M7NIR. Panel A. Subcutaneous SW480neo (C, control) or SW480mat (M, expressing MMP7) xenograft colon tumors in a nude mouse (dorsal, caudal view) imaged in vivo. Panel B. A segment of explanted APC Min mouse intestine (jejunum and ileum) with spontaneous polyps (red circles) imaged ex vivo

as compared versus the prototype PB with visible chromophores, resulting in an improved response of the PB (Fig. 3) (Scherer et al. 2008b). The selectivity for MMP7 and enhanced sensitivity of the NIR-PB was assessed by ex vivo imaging of benign tumors in the intestines of adenomatous polyposis coli multiple intestinal neoplasia (APCMin ) mice injected systemically with the PB; the images revealed enhanced sensor fluorescence over the polyps (see Fig. 3). There is some variation in the sensor fluorescence observed in each polyp that may reflect heterogeneity in the expression of MMP7 in the polyps as observed previously by immunohistochemistry following fixation (Wilson et al. 1997). Quantification of both the sensor and the reference fluorescence over a large number of polyps shows a significant increase in Sensor/Reference ratios in the adenomas of APCMin mice compared to control intestinal tissue or adenomas from MMP7-null Min mice (Scherer et al. 2008b). This approach was highly sensitive, as benign tumors with diameters as small as 1 mm could be detected. Preliminary studies in our laboratory indicate that some additional enhancement in the sensitivity for detection of MMPs and/or other proteases in the tumor microenvironment can be achieved by the use of PBs incorporating a peptide cleavable by more than one protease, i.e., trading specificity of the PB for enhanced sensitivity in detecting proteases in the tumor microenvironment (McIntyre and Matrisian, unpublished results). For optical imaging, Achilefu and colleagues have prepared a number of NIR optical contrast agents designed to either bind to or be metabolized by tumors

Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy

127

and, together with Britton Chance, have demonstrated the feasibility of detecting 2 cm-deep subsurface tumors using a metabolism-enhanced NIR fluorescent contrast agent and NIRF in vivo imaging (Achilefu et al. 2002; Chen et al. 2003; Achilefu 2004). A number of fluorogenic triple-helical peptide substrates have been developed as collagen-mimetic substrates for MMPs (Lauer-Fields and Fields 2002), though these have primarily found application for distinguishing substrate specificity for MMPs or as tools for high-throughput screening for MMP inhibitors (Minond et al. 2007; Lauer-Fields et al. 2009). Recently, the Achilefu group reported development of a self-assembling homotrimeric triple-helical peptide (THP), incorporating segments of type V collagen, with high specificity to MMP2 and MMP9. Preliminary results indicated that this kind of triple-helical peptide, containing homo-FRET quenched fluorophores, could be cleaved by MMP-2 yielding enhanced fluorescence (Barry Edwards, personal communication). Alternative strategies for optical imaging of substrate-based imaging probes include an approach described by the Tsien group to use activatable cell-penetrating peptides (ACPPs), consisting of a polyarginine membrane-translocating motif linked via an MMP-cleavable peptide (PLG∗LAG) to an appropriate masking polyanionic domain (a cleavable peptide hairpin), to deliver fluorescent labels to within tumor cells both in vitro and in vivo after cleavage by tumor-associated proteases (Jiang et al. 2004). Such ACPPs offer a general strategy toward both imaging and delivery of therapeutics in a variety of diseases in which extracellular proteases have been implicated. A different kind of strategy for optical imaging of proteases makes use of fluorescent-labeled peptides that bind with high affinity to the target protease, e.g., using a labeled peptide selected from a phage-display library to bind with high affinity to hepsin, a transmembrane serine protease that is generally upregulated in prostate cancer (Kelly et al. 2008). However, such peptide-based ligands do not necessarily discriminate active from inactive protease and thus are distinct from either substrate or activity-based probes that are specific for the active proteases. The proteolytic cleavage strategy for FRET substrates has also been adapted to achieve MMP7-activated photodynamic therapy with a molecular beacon designed to produce singlet oxygen upon irradiation but only after proteolytic cleavage (Zheng et al. 2007). Such innovative modifications of imaging beacons to also provide targeted delivery of therapy portend anticipated developments in this area.

Small Molecule Activity-Based Probes (ABPs) for Proteases Activity-based proteomics uses small molecules referred to as “activity-based probes” (ABPs) to identify distinct sets of proteins within a complex proteome. Such ABPs consist of three parts: a reactive group for covalent attachment to the enzyme, a linker region that can modulate reactivity and specificity, and a “tag” for identification and/or purification of modified enzymes. These activity-dependent tags provide analysis of changes in enzyme activity rather than simple protein abundance. The ABPs are designed for covalent modification of target proteins and are

128

J.O. McIntyre and L.M. Matrisian

often based on protease inhibitors developed on the basis of subtle differences in reaction mechanisms for the major protease clans and families [reviewed in Powers et al. (2002)]. However, of the major protease clans, only proteases in the Sand C-clans use a catalytic mechanism that can be co-opted to give direct covalent modification of the primary active site nucleophile. For enzymes that lack a direct acyl-enzyme intermediate in the catalytic hydrolysis reaction, such as the metalloproteases (M-clan), ABPs are based on ligands that bind to the catalytic metal in the active site. The reactive component of the ABPs are often based on suitable modifications of suicide substrates [for reviews see Evans and Cravatt (2006); Jeffery and Bogyo (2003); Powers et al. (2002); Sadaghiani et al. (2007b); Speers and Cravatt (2004)]. The linker region of the ABPs not only connects the reactive group to the tag, used for identification and/or purification of the target, but can also incorporate specificity elements for targeting the probe to a specific enzyme, family, or clan, e.g., using peptide or peptide-like structures as linkers in ABPs for targeting proteases. The tags of ABPs include biotin, fluorescent, and radioactive moieties [for review see Sadaghiani et al. (2007b)] that provide rapid identification and/or purification of probe-modified proteins. Over the last decade, ABPs have been developed for profiling S-, T-, and C-clans of proteases within the degradome by exploiting the conserved active site nucleophiles in these protease families (Sadaghiani et al. 2007a; Dive et al. 2008). Since the M-clan of proteases lack a corresponding conserved nucleophile in the active site, ABPs for this clan have been based on high-affinity metal ligands and a photo-reactive group such as in the ABPs for zinc-dependent proteases including the MMPs (Chan et al. 2004; Saghatelian et al. 2004). However, the MMP-targeted ABPs have had limited success in identifying active MMPs in biological samples due, in part, to low levels of the active forms of these enzymes (David et al. 2007).

ABPs for C-Clan Proteases ABPs for C-clan proteases have been developed using a variety of covalently reactive functional groups and selective linker sequences designed to target small sets of related C-protease sub-families with overlapping substrate specificity. A number of these have been developed to target various deubiquitinating proteases, one of the largest families of proteases (>60 members in the human genome) (LopezOtin and Overall 2002) that function to regulate the removal of ubiquitin from target proteins and thereby control degradation by the proteasome. ABPs designed to target ubiquitin-specific proteases (USPs) make use of the full-length ubiquitin chain (76 amino acids) as the linker in the probe since this family of C-proteases recognize folded ubiquitin as substrate. ABPs targeting USPs have been used to identify new protease families and to monitor changes in activities of USPs in cancer cells. A number of ABPs have been developed for two other significant sub-families of C-proteases: the papain family (clan CA/CB) that are primarily lysosomal and the cytosolic caspases (clan CD) involved in the regulation of apoptosis (programmed cell death). ABPs for these two families of proteases are generally

Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy

129

short, tri- or tetra-peptides carrying reactive functional groups such as epoxides, acyloxymethyl ketones (AOMK), or vinyl sulfones and have been used both for biomarker discovery and for imaging in cancer.

ABPs for S-Clan Proteases ABPs targeting the large S-clan proteases include fluorophosphonate probes that also target the larger family of serine hydrolase enzymes as well as probes with the less reactive diphenyl phosphonate containing peptide-based scaffolds more selective for S-proteases. Since the S-protease family comprises a number of often small and highly specialized sub-families, less progress has been made on S-clan ABPs. Such probes have been designed to target diverse enzymes in trypsin, chymotrypsin, and granzyme families but have yet to find specific applications to cancer.

ABPs for M-Clan Proteases ABPs designed to covalently modify the active site of zinc M-class proteases are generally based on either a broad-spectrum or a selective synthetic inhibitor that is integrated with a suitable photolabile group incorporated into the inhibitor structure. With these ABPs, efficient labeling occurs when, in the enzyme-ABP complex, the reactive group is directed at the enzyme active site rather than being exposed to aqueous solvent. Results from studies with a series of hydroxamate-based ABPs designed to target MMPs also showed that the reactivity varied considerably with different MMPs with some probes also reacting with a broad range of the zinc metalloprotease family (David et al. 2007). Studies with cells expressing MMPs and using these kinds of ABPs suggest that MMP active forms are present in extremely low amounts, a result in agreement with gelatin zymography data and which may explain the failure to detect MMP active forms in previous reports. These results support the notion that MMPs are mostly present in their zymogen form and in complex with TIMPs, a situation that could be very specific to the MMP family, as compared to other classes of zinc metalloproteases. In this respect, many zinc metalloproteases identified by the ABP profiling approach are expressed directly as active forms, for which no natural inhibitors have been reported (Sieber et al. 2006).

Imaging Protease Activity in Tumors One of the major challenges in cancer diagnosis is the early detection of small primary tumors (Weissleder et al. 1999). Since many enzyme activities are upregulated in tumor cells, probes that report on enzymatic activity provide valuable tools for early diagnostic imaging (Weissleder et al. 1999; Mahmood and Weissleder 2003; Sloane et al. 2006). In general, optical imaging techniques are used to image protease activity in vivo. The cost, space, and time involved in optical imaging are less demanding compared to other imaging modalities. Furthermore, the advantage of

130

J.O. McIntyre and L.M. Matrisian

optical imaging methods include the use of non-ionizing low-energy radiation, high sensitivity with the possibility of detecting micron-sized objects, and continuous data acquisition in real time and in an intact environment. Optical imaging in the NIR region between 700 and 900 nm has a low absorption by intrinsic photoactive biomolecules and allows light to penetrate several centimeters into the tissue, a depth that is sufficient to image practically all small animals (Zuzak et al. 2002). Imaging in the NIR region has less tissue autofluorescence, markedly improving the target/background ratio as compared with the visible region of the spectrum (Rudin and Weissleder 2003). The detection sensitivity depends on both selection of the fluorescent probe and optimization of imaging geometry for detection with a highly sensitive CCD camera. These kinds of optical imaging systems are capable of detecting a small number of photons that are transmitted through living tissues permitting real-time images to be collected within a few seconds. A fast and relatively easy imaging procedure makes this modality attractive for potential clinical use. Fluorescence-mediated tomography (FMT) has recently been shown to three-dimensionally localize and quantify fluorescent probes in deep tissues at high sensitivity (Ntziachristos et al. 2002; Montet et al. 2007). Current methods for imaging enzymes mainly rely on antibody labeling or on substrates that become fluorescent after enzyme cleavage (Baruch et al. 2004; Sloane et al. 2006). Although antibodies are specific for their enzyme targets, they are not cell permeable and do not give information about enzyme activity. Fluorescent substrates are useful for the activity-based imaging of proteases; however, these compounds often suffer from a lack of specificity, leading to cleavage by multiple classes of proteases (Baruch et al. 2004; Sloane et al. 2006). Furthermore, with fluorescent substrate reporters, it is not possible to determine which protease is responsible for substrate processing in vivo. By contrast, ABPs covalently bind to active enzymes, thus permitting assignment of imaging signals to specific enzymes with a number of ABPs that target C-clan proteases having been used to image enzyme activity in tumor cells both in vitro and in vivo (Joyce et al. 2004; Blum et al. 2005; 2007).

Imaging Cancer with ABP-Based Imaging Agents Small molecule ABPs have been used for monitoring the level of active proteases in cells and tissues and applied particularly to assess samples that differ in stage or types of disease pathology. ABP-detected proteases that show altered levels during disease progression may serve as useful biomarkers and/or as therapeutic targets. Since ABPs form direct covalent bonds with their targets, active proteases can also be localized either in whole cells or in animals by fluorescence studies in vitro or in vivo. ABPs designed to target various clans of proteases such as the M-clan and Cclan ABPs have been used to profile human tumors and tumor cell lines and identify novel enzyme activities for the diagnosis and treatment of cancer (Evans and Cravatt

Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy

131

2006; Fonovic and Bogyo 2007; Schmidinger et al. 2006). For example, profiling a set of human breast cancer cell lines with FP-rhodamine, an ABP targeting the serine hydrolase superfamily, showed upregulation of a distinct set of these enzymes, revealing potential biomarkers (Jessani et al. 2002). FP-rhodamine has been used to show that a number of serine hydrolase activities, such as uPA and tissue plasminogen activator (tPA), are highly elevated in MDA-MB-231 cell line variants and correlate with increased tumor growth and metastasis upon reintroduction into mice (Jessani et al. 2004). A library of M-clan ABPs based on peptide hydroxamate scaffold with a photocrosslinker showed elevated neprilysin, alanyl aminopeptidase, and ADAM10 activities in both breast carcinoma and melanoma cell lines (Saghatelian et al. 2004; Sieber et al. 2006). ABPs for the C-clan ubiquitinspecific proteases (USPs) have been used to identify unique and tumor-specific activities in human tumor cell lines (Ovaa et al. 2004) and in human cervical cancer biopsies A number of ABPs have also been applied to functionally characterize enzyme activities in mouse models of cancer. For example, a C-clan ABP targeting the papain family revealed upregulation of cathepsin X activity partially compensating for a deficiency in cathepsin B in a mammary tumor line (Vasiljeva et al. 2006). These kinds of data suggest that proteases can dynamically compensate for each other in genetically modified animals. Using the same kind of C-clan ABP, it was shown in a mouse model that develops pancreatic beta-cell tumors that the levels of multiple cysteine cathepsins were highly upregulated and linked to both angiogenesis and tumor invasiveness (Joyce et al. 2004). An additional benefit from using these kinds of ABPs is that the targeted cathepsins could subsequently be identified and quantified by biochemical analysis (Joyce et al. 2004). Fluorescenttagged ABPs are constitutively fluorescent and give a high nonspecific fluorescent background when used in living cells. To ameliorate this problem, a cathepsinspecific ABP was developed that becomes fluorescent only upon binding to its enzyme target and has been used to image cathepsin activity levels in both fibroblast and breast cancer cell lines (Blum et al. 2005). Importantly, the fluorescent signals could be specifically blocked by pre-treatment of cells with a general cysteine protease inhibitor. More recently, near-infrared-labeled versions of the C-clan cathepsin probes produced spatially resolvable fluorescence in the tumor tissues of live mice that correlated with the levels of active cathepsins in those tissues; ex vivo analysis of tumor tissues further confirmed that the fluorescence observed in the live animals was due to specific probe labeling of active cathepsins (Blum et al. 2007).

Substrate-Based Imaging Agents in Cancer The substrate-based proteolytic beacons selective for MMP-7, PBvisM7, and PBnirM7 have been used to detect MMP-7 activity in xenograft tumors in mice. In these studies, pairs of xenograft tumors were established on the rear flanks of each animal: one tumor with human colorectal tumor cells that express several

132

J.O. McIntyre and L.M. Matrisian

MMP family members but do not express detectable amounts of endogenous MMP7 and a second with the same cells transfected with an MMP-7 expression vector (Witty et al. 1994). Imaging is achieved following i.v. injection of a single bolus of either PBvisM7 or PBnirM7 (McIntyre et al. 2004; Scherer et al. 2008b). Approximately 2–4 h following PB injection, the reference (R) signal is low both in the control and in the MMP7-transfected tumors, while the sensor (S) channel shows an approximately 10-fold difference between the control and the MMP7-expressing tumors with a comparable difference in S/R ratio (Scherer et al. 2008b). The secondgeneration PBnirM7 has been used to detect intestinal adenomas in the multiple intestinal neoplasia (Min) mouse model of familial polyposis (Scherer et al. 2008b). In those studies, the animals were sacrificed post i.v. administration of PBnirM7 revealing enhanced fluorescence of the MMP7 sensor associated with a number of adenomas in the intestinal tract examined ex vivo (Fig. 3). The S/R ratio was consistently and significantly higher in Min adenomas compared to normal intestinal tissue of mice lacking the Min mutation, and in Min adenomas from MMP7-null mice. Similar studies by the Weissleder group also demonstrated the detection of adenoma-associated protease activity in the intestines of Min mice (Marten et al. 2002). Taken together, the fluorescence imaging studies of living mice indicate that PB-M7s can be used to detect and selectively image MMP-7 activity in vivo via the enhanced fluorescence of the sensor in the proteolyzed reagent that results in an increase in sensor/reference ratio. The optical imaging approach using new optical reporters has potential for highly sensitive, non-invasive, in vivo detection, and imaging of tumor-associated proteolytic activity.

Conclusion and Future Directions Over the past several years, the field of activity-based proteomics has produced a variety of technologies for the direct study of enzymes in their biological context. Protease probes that monitor the activity of numerous diverse enzyme classes have been synthesized, and these probes have been applied to a number of biologically and pathologically relevant fields. Additionally, a number of new tools, including gel-free screening systems and quenched probes, have been developed that allow rapid identification and visualization of enzyme activity in vitro and in vivo. Both substrate-based imaging probes and ABPs have been applied to the identification and evaluation of potential enzyme inhibitors in the physiologically relevant environments of a complex proteome, cell, or even whole animal. However, challenges in the field of activity-based proteomics still remain to be addressed. In order to identify new probe scaffolds that allow for greater proteomic coverage by ABPs, structurally diverse probe libraries need to be developed. Perhaps the most important challenge facing activity-based proteomics is the need to combine the data from activity-based assays with relevant biological experiments to gain a more complete understanding of enzyme function in cancer and other biological processes and diseases.

Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy

133

The continued use of optical imaging of proteolytic activity has exciting potential both for the understanding of cancer and in applications to cancer detection, diagnosis, and treatment. For preclinical studies, extension into the use of multi-photon fluorescence microscopy for intra-vital imaging of protease activities should facilitate the further delineation of specific roles of proteases in processes critical to tumor progression. New developments in NIR optical tomography (Nioka and Chance 2005; Montet et al. 2007) are yielding promising optical approaches for imaging in clinical practice, particularly as a complementary modality for breast cancer detection (Zhu et al. 2005; Chance et al. 2005). The advent of such tomographic fluorescence imaging as well as advances in quantitative real-time catheter-based fluorescence molecular imaging technology (Upadhyay et al. 2007) portend new applications of imaging in the era of molecular oncology (Weissleder and Pittet 2008). The development of new kinds of targeted optical reagents, including those providing for both imaging and therapy (Chen et al. 2005; Zheng et al. 2007), will likely provide new paradigms for the clinician. Non-invasive imaging techniques for proteolytic activity provide an extraordinary opportunity to increase the sensitivity of detecting early-stage tumors and to identify tumors that require particularly aggressive therapy. The significant potential of protease-sensing optical imaging beacons and probes for delineating tumor margins and the tumor–stromal interface as well as for imaging tumor response to treatment is likely to be realized with further improvements in sensitivity and selectivity of these versatile tools. With time and the rapid advance in technology, we are likely to see a sharp increase in the number and types of applications of protease probes to oncology. Acknowledgements We thank Randy Scherer for the images of tumors and adenomas with proteolytic beacons. This work was supported by National Institutes of Health grant CA60867 to L.M.M., grant P30 068485 to the Vanderbilt-Ingram Cancer Center, and grant U24 CA126588, the Southeastern Center for Imaging Animal Models, Vanderbilt University Institute of Imaging Science.

References Achilefu, S. 2004. Lighting up tumors with receptor-specific optical molecular probes. Technology in Cancer Research and Treatment 3 (4):393–409. Achilefu, S., H. N. Jimenez, R. B. Dorshow, J. E. Bugaj, E. G. Webb, R. R. Wilhelm, R. Rajagopalan, J. Johler, and J. L. Erion. 2002. Synthesis, in vitro receptor binding, and in vivo evaluation of fluorescein and carbocyanine peptide-based optical contrast agents. Journal of Medicinal Chemistry 45 (10):2003–15. Alencar, H., M. A. Funovics, J. Figueiredo, H. Sawaya, R. Weissleder, and U. Mahmood. 2007. Colonic adenocarcinomas: near-infrared microcatheter imaging of smart probes for early detection–study in mice. Radiology 244 (1):232–8. Barrett, A. J., N. D. Rawlings, and J. F. Woessner, Jr. 1998. Handbook of Proteolytic Enzymes. London: Academic Press. Baruch, A., D. A. Jeffery, and M. Bogyo. 2004. Enzyme activity–it’s all about image. Trends in Cell Biology 14 (1):29–35. Berger, D. H. 2002. Plasmin/plasminogen system in colorectal cancer. World Journal of Surgery 26 (7):767–771.

134

J.O. McIntyre and L.M. Matrisian

Blum, G., S. R. Mullins, K. Keren, M. Fonovic, C. Jedeszko, M. J. Rice, B. F. Sloane, and M. Bogyo. 2005. Dynamic imaging of protease activity with fluorescently quenched activity-based probes. Nature Chemical Biology 1 (4):203–9. Blum, G., G. von Degenfeld, M. J. Merchant, H. M. Blau, and M. Bogyo. 2007. Noninvasive optical imaging of cysteine protease activity using fluorescently quenched activity-based probes. Nature Chemical Biology 3 (10):668–77. Bremer, C., C. H. Tung, and R. Weissleder. 2001. In vivo molecular target assessment of matrix metalloproteinase inhibition. Nature Medicine 7 (6):743–8. Brinckerhoff, C. E., and L. M. Matrisian. 2002. Matrix metalloproteinases: a tail of a frog that became a prince. Nature Reviews Molecular Cell Biology 3 (3):207–14. Brindle, K. 2008. New approaches for imaging tumour responses to treatment. Nature Reviews Cancer 8 (2):94–107. Bromme, D., and J. Kaleta. 2002. Thiol-dependent cathepsins: pathophysiological implications and recent advances in inhibitor design. Current Pharmaceutical Biotechnology 8 (18): 1639–58. Caughey, G. H. 2007. Mast cell tryptases and chymases in inflammation and host defense. Immunological Reviews 217:141–54. Chan, E. W., S. Chattopadhaya, R. C. Panicker, X. Huang, and S. Q. Yao. 2004. Developing photoactive affinity probes for proteomic profiling: hydroxamate-based probes for metalloproteases. Journal of the American Chemical Society 126 (44):14435–46. Chance, B., S. Nioka, J. Zhang, E. F. Conant, E. Hwang, S. Briest, S. G. Orel, M. D. Schnall, and B. J. Czerniecki. 2005. Breast cancer detection based on incremental biochemical and physiological properties of breast cancers: a six-year, two-site study. Academic Radiology 12 (8):925–33. Chen, Y., A. Gryshuk, S. Achilefu, T. Ohulchansky, W. Potter, T. Zhong, J. Morgan, B. Chance, P. N. Prasad, B. W. Henderson, A. Oseroff, and R. K. Pandey. 2005. A novel approach to a bifunctional photosensitizer for tumor imaging and phototherapy. Bioconjugate Chemistry 16 (5):1264–74. Chen, W. T., and T. Kelly. 2003. Seprase complexes in cellular invasiveness. Cancer Metastasis Review 22 (2–3):259–69. Chen, Y., G. Zheng, Z. H. Zhang, D. Blessington, M. Zhang, H. Li, Q. Liu, L. Zhou, X. Intes, S. Achilefu, and B. Chance. 2003. Metabolism-enhanced tumor localization by fluorescence imaging: in vivo animal studies. Optics Letters 28 (21):2070–72. Coussens, L. M., B. Fingleton, and L. M. Matrisian. 2002. Matrix metalloproteinase inhibitors and cancer: trials and tribulations. Science 295 (5564):2387–92. Coussens, L. M., and Z. Werb. 2002. Inflammation and cancer. Nature 420 (6917):860–7. David, A., D. Steer, S. Bregant, L. Devel, A. Makaritis, F. Beau, A. Yiotakis, and V. Dive. 2007. Cross-linking yield variation of a potent matrix metalloproteinase photoaffinity probe and consequences for functional proteomics. Angewandte Chemie (International Ed. in English) 46 (18):3275–7. Dive, V., Paulick, M., McIntyre, J.O., Matrisian, L.M., and Bogyo, M. 2008. Activity-based imaging and biochemical profiling tools for analysis of the cancer degradome. Edited by D. Edwards, The Cancer Degradome, New York: Springer Science. Duffy, M. J. 1996. Proteases as prognostic markers in cancer. Clinical Cancer Research 2 (4): 613–18. Egeblad, M., and Z. Werb. 2002. New functions for the matrix metalloproteinases in cancer progression. Nature Reviews Cancer 2 (3):161–74. Evans, M. J., and B. F. Cravatt. 2006. Mechanism-based profiling of enzyme families. Chemical Reviews 106 (8):3279–301. Fonovic, M., and M. Bogyo. 2007. Activity based probes for proteases: applications to biomarker discovery, molecular imaging and drug screening. Current Pharmaceutical Design 13 (3): 253–61.

Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy

135

Gross, J., and C. M. Lapiere. 1962. Collagenolytic activity in amphibian tissues: a tissue culture assay. Proceedings of the National Academy of Sciences of the United States of America 48:1014–1022. Herszenyi, L., M. Plebani, P. Carraro, M. De Paoli, G. Roveroni, R. Cardin, F. Foschia, Z. Tulassay, R. Naccarato, and F. Farinati. 2000. Proteases in gastrointestinal neoplastic diseases. Clinica Chimica Acta 291 (2):171–87. Izmailova, E. S., N. Paz, H. Alencar, M. Chun, L. Schopf, M. Hepperle, J. H. Lane, G. Harriman, Y. Xu, T. Ocain, R. Weissleder, U. Mahmood, A. M. Healy, and B. Jaffee. 2007. Use of molecular imaging to quantify response to IKK-2 inhibitor treatment in murine arthritis. Arthritis and Rheumatism 56 (1):117–28. Jaffer, F. A., P. Libby, and R. Weissleder. 2007. Molecular imaging of cardiovascular disease. Circulation 116 (9):1052–61. Jeffery, D. A., and M. Bogyo. 2003. Chemical proteomics and its application to drug discovery. Current Opinion in Biotechnology 14 (1):87–95. Jessani, N., M. Humphrey, W. H. McDonald, S. Niessen, K. Masuda, B. Gangadharan, J. R. Yates, 3rd, B. M. Mueller, and B. F. Cravatt. 2004. Carcinoma and stromal enzyme activity profiles associated with breast tumor growth in vivo. Proceedings of the National Academy of Sciences of the United States of America 101 (38):13756–61. Jessani, N., Y. Liu, M. Humphrey, and B. F. Cravatt. 2002. Enzyme activity profiles of the secreted and membrane proteome that depict cancer cell invasiveness. Proceedings of the National Academy of Sciences of the United States of America 99 (16):10335–40. Jiang, T., E. S. Olson, Q. T. Nguyen, M. Roy, P. A. Jennings, and R. Y. Tsien. 2004. Tumor imaging by means of proteolytic activation of cell-penetrating peptides. Proceedings of the National Academy of Sciences of the United States of America 101 (51):17867–72. Joyce, J. A., A. Baruch, K. Chehade, N. Meyer-Morse, E. Giraudo, F. Y. Tsai, D. C. Greenbaum, J. H. Hager, M. Bogyo, and D. Hanahan. 2004. Cathepsin cysteine proteases are effectors of invasive growth and angiogenesis during multistage tumorigenesis. Cancer Cell 5 (5): 443–53. Kato, D., K. M. Boatright, A. B. Berger, T. Nazif, G. Blum, C. Ryan, K. A. Chehade, G. S. Salvesen, and M. Bogyo. 2005. Activity-based probes that target diverse cysteine protease families. Nature Chemical Biology 1 (1):33–8. Kelly, K. A., S. R. Setlur, R. Ross, R. Anbazhagan, P. Waterman, M. A. Rubin, and R. Weissleder. 2008. Detection of early prostate cancer using a hepsin-targeted imaging agent. Cancer Research 68 (7):2286–91. Kerkela, E., R. Ala-aho, P. Klemi, S. Grenman, S. D. Shapiro, V. M. Kahari, and U. Saarialho-Kere. 2002. Metalloelastase (MMP-12) expression by tumour cells in squamous cell carcinoma of the vulva correlates with invasiveness, while that by macrophages predicts better outcome. Journal of Pathology 198 (2):258–69. Koblinski, J. E., M. Ahram, and B. F. Sloane. 2000. Unraveling the role of proteases in cancer. Clinica Chimica Acta 291 (2):113–35. Lauer-Fields, J. L., and G. B. Fields. 2002. Triple-helical peptide analysis of collagenolytic protease activity. Biological Chemistry 383 (7–8):1095–105. Lauer-Fields, J. L., D. Minond, P. S. Chase, P. E. Baillargeon, S. A. Saldanha, R. Stawikowska, P. Hodder, and G. B. Fields. 2009. High throughput screening of potentially selective MMP-13 exosite inhibitors utilizing a triple-helical FRET substrate. Bioorganic & Medicinal Chemistry 17(3):990–1005. Lecaille, F., J. Kaleta, and D. Bromme. 2002. Human and parasitic papain-like cysteine proteases: their role in physiology and pathology and recent developments in inhibitor design. Chemical Reviews 102 (12):4459–88. Lepage, M., W. C. Dow, M. Melchior, Y. You, B. Fingleton, C. C. Quarles, C. Pepin, J. C. Gore, L. M. Matrisian, and J. O. McIntyre. 2007. Noninvasive detection of matrix metalloproteinase activity in vivo using a novel magnetic resonance imaging contrast agent with a solubility switch. Molecular Imaging 6 (6):393–403.

136

J.O. McIntyre and L.M. Matrisian

Liotta, L. A., K. Tryggvason, S. Garbisa, I. Hart, C. M. Foltz, and S. Shafie. 1980. Metastatic potential correlates with enzymatic degradation of basement membrane collagen. Nature 284:67–8. Lopez-Otin, C., and L. M. Matrisian. 2007. Emerging roles of proteases in tumour suppression. Nature Reviews Cancer 7 (10):800–8. Lopez-Otin, C., and C. M. Overall. 2002. Protease degradomics: a new challenge for proteomics. Nature Reviews Molecular Cell Biology 3 (7):509–19. Mahmood, U., and R. Weissleder. 2003. Near-infrared optical imaging of proteases in cancer. Molecular Cancer Therapeutics 2 (5):489–96. Marten, K., C. Bremer, K. Khazaie, M. Sameni, B. Sloane, C. H. Tung, and R. Weissleder. 2002. Detection of dysplastic intestinal adenomas using enzyme-sensing molecular beacons in mice. Gastroenterology 122 (2):406–14. McIntyre, J. O., B. Fingleton, K. S. Wells, D. W. Piston, C. C. Lynch, S. Gautam, and L. M. Matrisian. 2004. Development of a novel fluorogenic proteolytic beacon for in vivo detection and imaging of tumour-associated matrix metalloproteinase-7 activity. Biochemical Journal 377 (Pt 3):617–28. McIntyre, J. O., and L. M. Matrisian. 2003. Molecular imaging of proteolytic activity in cancer. Journal of Cellular Biochemistry 90 (6):1087–97. Melnikova, V., and M. Bar-Eli. 2007. Inflammation and melanoma growth and metastasis: the role of platelet-activating factor (PAF) and its receptor. Cancer Metastasis Reviews 26 (3–4): 359–71. Minond, D., J. L. Lauer-Fields, M. Cudic, C. M. Overall, D. Pei, K. Brew, M. L. Moss, and G. B. Fields. 2007. Differentiation of secreted and membrane-type matrix metalloproteinase activities based on substitutions and interruptions of triple-helical sequences. Biochemistry 46 (12):3724–33. Montet, X., J. L. Figueiredo, H. Alencar, V. Ntziachristos, U. Mahmood, and R. Weissleder. 2007. Tomographic fluorescence imaging of tumor vascular volume in mice. Radiology 242 (3): 751–8. Netzel-Arnett, S., J. D. Hooper, R. Szabo, E. L. Madison, J. P. Quigley, T. H. Bugge, and T. M. Antalis. 2003. Membrane anchored serine proteases: a rapidly expanding group of cell surface proteolytic enzymes with potential roles in cancer. Cancer Metastasis Review 22 (2–3): 237–58. Nioka, S., and B. Chance. 2005. NIR spectroscopic detection of breast cancer. Technology in Cancer Research and Treatment 4 (5):497–512. Ntziachristos, V., C. H. Tung, C. Bremer, and R. Weissleder. 2002. Fluorescence molecular tomography resolves protease activity in vivo. Nature Medicine 8 (7):757–60. Orlowski, R. Z., and D. J. Kuhn. 2008. Proteasome inhibitors in cancer therapy: lessons from the first decade. Clinical Cancer Research 14 (6):1649–57. Ovaa, H., B. M. Kessler, U. Rolen, P. J. Galardy, H. L. Ploegh, and M. G. Masucci. 2004. Activitybased ubiquitin-specific protease (USP) profiling of virus-infected and malignant human cells. Proceedings of the National Academy of Sciences of the United States of America 101 (8):2253–8. Overall, C. M., and R. A. Dean. 2006. Degradomics: systems biology of the protease web. Pleiotropic roles of MMPs in cancer. Cancer Metastasis Reviews 25 (1):69–75. Pham, W., Y. Choi, R. Weissleder, and C. H. Tung. 2004. Developing a peptide-based near-infrared molecular probe for protease sensing. Bioconjugate Chemistry 15 (6):1403–7. Powers, J. C., J. L. Asgian, O. D. Ekici, and K. E. James. 2002. Irreversible inhibitors of serine, cysteine, and threonine proteases. Chemical Reviews 102 (12):4639–750. Rawlings, N. D., F. R. Morton, C. Y. Kok, J. Kong, and A. J. Barrett. 2008. MEROPS: the peptidase database. Nucleic Acids Research 36 (Database issue):D320–5. Rochefort, H., M. Garcia, M. Glondu, V. Laurent, E. Liaudet, J. M. Rey, and P. Roger. 2000. Cathepsin D in breast cancer: mechanisms and clinical applications, a 1999 overview. Clinica Chimica Acta 291 (2):157–70.

Proteinase Optical Imaging Tools for Cancer Detection and Response to Therapy

137

Rocks, N., G. Paulissen, M. El Hour, F. Quesada, C. Crahay, M. Gueders, J. M. Foidart, A. Noel, and D. Cataldo. 2008. Emerging roles of ADAM and ADAMTS metalloproteinases in cancer. Biochimie 90 (2):369–79. Rudin, M., and R. Weissleder. 2003. Molecular imaging in drug discovery and development. Nature Reviews. Drug Discovery 2 (2):123–31. Sadaghiani, A. M., S. H. Verhelst, and M. Bogyo. 2007a. Tagging and detection strategies for activity-based proteomics. Current Opinion in Chemical Biology 11 (1):20–8. Sadaghiani, A. M., S. H. Verhelst, V. Gocheva, K. Hill, E. Majerova, S. Stinson, J. A. Joyce, and M. Bogyo. 2007b. Design, synthesis, and evaluation of in vivo potency and selectivity of epoxysuccinyl-based inhibitors of papain-family cysteine proteases. Chemical Biology 14 (5):499–511. Saghatelian, A., N. Jessani, A. Joseph, M. Humphrey, and B. F. Cravatt. 2004. Activity-based probes for the proteomic profiling of metalloproteases. Proceedings of the National Academy of Sciences of the United States of America 101 (27):10000–5. Scherer, R. L., J. O. McIntyre, and L. M. Matrisian. 2008a. Imaging matrix metalloproteinases in cancer. Cancer Metastasis Reviews 27(4):679–690. Scherer, R. L., M. N. VanSaun, J. O. McIntyre, and L. M. Matrisian. 2008b. Optical imaging of matrix metalloproteinase-7 activity in vivo using a proteolytic nanobeacon. Molecular Imaging 7(3):118–131. Schmidinger, H., A. Hermetter, and R. Birner-Gruenberger. 2006. Activity-based proteomics: enzymatic activity profiling in complex proteomes. Amino Acids 30 (4):333–50. Sica, A., P. Allavena, and A. Mantovani. 2008. Cancer related inflammation: the macrophage connection. Cancer Letters 267(2):204–215. Sieber, S. A., S. Niessen, H. S. Hoover, and B. F. Cravatt. 2006. Proteomic profiling of metalloprotease activities with cocktails of active-site probes. Nature Chemical Biology 2 (5):274–81. Sloane, B. F., M. Sameni, I. Podgorski, D. Cavallo-Medved, and K. Moin. 2006. Functional imaging of tumor proteolysis. Annual Review of Pharmacology and Toxicology 46:301–15. Speers, A. E., and B. F. Cravatt. 2004. Chemical strategies for activity-based proteomics. Chembiochem: A European Journal of Chemical Biology 5 (1):41–7. Tsien, R. Y. 2005. Building and breeding molecules to spy on cells and tumors. FEBS Letters 579 (4):927–32. Upadhyay, R., R. A. Sheth, R. Weissleder, and U. Mahmood. 2007. Quantitative real-time catheterbased fluorescence molecular imaging in mice. Radiology 245 (2):523–31. Van de Wiele, C., and R. Oltenfreiter. 2006. Imaging probes targeting matrix metalloproteinases. Cancer Biotherapy & Radiopharmaceuticals 21 (5):409–17. Vasiljeva, O., A. Papazoglou, A. Kruger, H. Brodoefel, M. Korovin, J. Deussing, N. Augustin, B. S. Nielsen, K. Almholt, M. Bogyo, C. Peters, and T. Reinheckel. 2006. Tumor cell-derived and macrophage-derived cathepsin B promotes progression and lung metastasis of mammary cancer. Cancer Research 66 (10):5242–50. Weissleder, R., C. H. Tung, U. Mahmood, and A. Bogdanov, Jr. 1999. In vivo imaging of tumors with protease-activated near-infrared fluorescent probes. Nature Biotechnology 17 (4): 375–8. Weissleder, R., and V. Ntziachristos. 2003. Shedding light onto live molecular targets. Nature Medicine 9 (1):123–8. Weissleder, R., and M. J. Pittet. 2008. Imaging in the era of molecular oncology. Nature 452 (7187):580–9. Wilson, C. L., K. J. Heppner, P. A. Labosky, B. L. M. Hogan, and L. M. Matrisian. 1997. Intestinal tumorigenesis is suppressed in mice lacking the metalloproteinase matrilysin. Proceedings of the National Academy of Sciences of the United States of America 94 (4):1402–407. Witty, J. P., S. McDonnell, K. Newell, P. Cannon, M. Navre, R. Tressler, and L. M. Matrisian. 1994. Modulation of matrilysin levels in colon carcinoma cell lines affects tumorigenicity in vivo. Cancer Research 54:4805–12.

138

J.O. McIntyre and L.M. Matrisian

Woessner, J. F., and H. Nagase. 2000. Matrix Metalloproteinases and TIMPs. New York: Oxford University Press Inc. Zheng, G., J. Chen, K. Stefflova, M. Jarvi, H. Li, and B. C. Wilson. 2007. Photodynamic molecular beacon as an activatable photosensitizer based on protease-controlled singlet oxygen quenching and activation. Proceedings of the National Academy of Sciences of the United States of America 104 (21):8989–94. Zhu, Q., E. B. Cronin, A. A. Currier, H. S. Vine, M. Huang, N. Chen, and C. Xu. 2005. Benign versus malignant breast masses: optical differentiation with US-guided optical imaging reconstruction. Radiology 237 (1):57–66. Zuzak, K. J., M. D. Schaeberle, E. N. Lewis, and I. W. Levin. 2002. Visible reflectance hyperspectral imaging: characterization of a noninvasive, in vivo system for determining tissue perfusion. Analytical Chemistry 74 (9):2021–8.

Illustrating Molecular Events with Light: A Perspective on Optical Reporter Genes Pritha Ray

Introduction A major medical success achieved in the last century was the decline of the overall mortality rate of cancer patients by providing high-quality medical care. Many groundbreaking discoveries along with imaging formed the base of this achievement in modern medicine. Noninvasive molecular imaging using small animals is a relatively recent field as compared to clinical imaging in biomedical science. Rapid development in imaging strategies and imaging instrumentations quickly brought this field to the forefront of modern medicine and is now contributing to the development of the very basic steps of clinical imaging. Noninvasive molecular imaging can be categorized into direct and indirect imaging. Theoretically, indirect imaging involves reporter genes that after introduction in cells or animals indirectly measure the expression of an endogenous gene or promoter, while direct imaging engages those molecular probes that attempt to monitor a specific molecular process (e.g., receptor/ligand binding) in living cells or animals. While direct imaging is more relevant for clinical studies, it is more difficult to achieve. For example, to image any receptor directly we need a labeled ligand, and it is practically impossible to develop a labeled ligand for every receptor present in our body. In this scenario, reporter genes play important role to form a general integrated platform for many different applications. Reporter genes are not new, identified long back and in research for few decades. As more and more reporters have been identified, careful and intelligent molecular manipulations were attempted resulting in smarter reporter molecules. These reporter molecules have better transcriptional and translational efficiency in mammalian cells and can be adapted for in vivo imaging from living animals. Since the main theme of this book is “Clinical translation of optical imaging,” this chapter will focus primarily on the optical reporters and their applications in preclinical and clinical studies. Descriptions of other reporter genes useful P. Ray (B) Advanced Center for Treatment, Research and Education in Cancer (ACTREC), Tata Memorial Center, Kharghar, Navi Mumbai, Maharastra, 410210, India e-mail: [email protected]

E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_8, 

139

140

P. Ray

for imaging such as PET reporter genes, SPECT reporter genes, or MR reporter genes can be found in some of the excellent reviews (Massoud and Gambhir 2003; Gilad et al. 2007; Massoud and Gambhir 2007).

Optical Reporter Genes As the name suggests, all optical reporter proteins are able to emit light at different wavelengths due to their bioluminescent, fluorescent, or chemiluminescent properties. Chemiluminescence is the emission of light because of a chemical reaction. The enthalpy of the reaction gives rise to an atom or a molecule in a vibronically excited state; when the electron decays back to ground state, a photon is emitted. Bioluminescence is a form of chemiluminescence where visible light emission occurs from luminous organisms. In fluorescence, absorption of a specific wavelength of light causes an excited state of the chromophore (a core region of the protein) resulting in emission of lower energy photons (fluorescence) of higher wavelength. In nature, marine organisms are the main sources of these optical reporters.

Bioluminescence Reporters Luciferase Luciferase is a generic name because none of the major luciferases share sequence homology with each other. Luciferases can be found in bacteria, fungi, dinoflagellates, radiolarians and about 17 metazoan phyla and 700 genera, mostly marine (Greer and Szalay 2002). Phylogenetic analyses suggest that luciferin-luciferase systems have had more than 30 independent origins. To date, only a handful of them have been cloned, modified for better expression, and in use for in vivo imaging. Among them, the most popular luciferases were isolated from beetles (firefly and click beetle-coleoptera), jellyfish and sea pansies (cnidaria), and bacteria (Vibrio spp. and Photorhabdus luminescens). Each of these luciferases has their unique sequence, structure, and substrate requirements that can be and had been optimized for imaging living animals. Beetle Luciferases The firefly (Photinus pyralis) is one of the most familiar bioluminescent creatures in nature. The gene encoding firefly luciferase was first cloned by de Wet et al. (1985) and further engineered for higher expression in mammalian cells (de Wet et al. 1987). In cells, this enzyme catalyzes its substrate (a benzothiazole luciferin) in the presence of Mg-ATP and emits a yellow–green light with an emission peak at ∼ 560 nm. The photons emitted at this range can efficiently travel through the tissues and suffer less absorption by hemoglobin and oxy-hemoglobin present in the blood of living organism (Day et al. 2004). Later few more beetle luciferases especially

Illustrating Molecular Events with Light

141

from click beetles (Wood et al. 1989) drew attention due to their light emission at further red-shifted wavelength (580–625 nm) which has the obvious advantage of better tissue penetration required for in vivo imaging (Doyle et al. 2004; Massoud and Gambhir 2003). All these luciferases are now being extensively exploited for various repetitive and noninvasive imaging applications. Marine Luciferases In the ocean, bioluminescence is found across nearly all taxa, from bacteria to fish. These luciferases do not share much homology at the nucleotide or amino acid level among different organisms. They can be classified into four groups based on the types of substrates they use for the bioluminescence reaction. The existence of such a large variety of luciferases and luciferins has immense significance for capturing food, preying, defense, reproduction, and many other vital activities. The details will be found in a review by Hastings (1983). Among all these marine luciferases, the luciferase from anothozoan sea pansy Renilla has been developed as an excellent bioluminescence reporter protein with a substrate called coelenterazine for in vivo imaging (Bhaumik and Gambhir 2002). In contrast to beetle luciferase-luciferin reaction, the renilla luciferase–coelenterazine reaction does not require ATP and emits light at blue–green (emission peak – 480 nm) region. Luciferin and coelenterazine do not cross-react and thus are advantageous for simultaneous imaging of two different molecular events (Bhaumik and Gambhir 2002). Bacterial Luciferases Luminous bacteria are very abundant and widely distributed in nature. They belong to three genera: Vibrio, Photobacterium, and the freshwater or soil species Xenorabdus (Meighen 1993). Bacterial luciferase is a 77 kDa chimeric protein of two nonidentical subunits (A and B) coded by two adjacent genes LuxA and LuxB that form the regulated lux operon. The other cistrons (C & D) situated in this operon codes for the substrate and thus bacterial luciferase has distinct advantage over firefly or renilla luciferase by making its own substrate. This potentially eliminates the injection of substrate and overcomes the high variability often noticed in in vivo imaging. However, in certain studies (e.g., monitoring time Kinetics of a molecular process) a constitutively on luciferase-luciferin system poses serious constraint. The LuxA and B genes and other members of the Lux operon have been cloned from different species of bacteria to human pathogenic bacteria, and their route of infection has been monitored in living mice by imaging (Yu et al. 2003; Doyle et al. 2004; Piwnica-Worms et al. 2004). Gaussia Luciferase Gaussia luciferase is a secreted luciferase isolated from a marine copepod (Gaussia princes) that generates 200-fold higher signal than renilla luciferase after addition/injection of coelenterazine in cell culture and from living animals

142

P. Ray

(Tannous et al. 2005). This unique secretary nature allows this luciferase to report from the cells as well as from the environment. Since evolution created a large number of naturally occurring luciferases, many new ones are being and will be identified with unique properties and will be adapted for molecular imaging.

In Vivo Bioluminescence Imaging Bioluminescence reporters have one potential drawback of generating very low level of light (5–10 photons/cell) that limits imaging of living cells with an ordinary or simple optical microscope. However, recent development of specialized and highly sensitive photon detectors (such as cooled, or intensified, charge-coupled device (CCD) cameras) has enabled external detection of very low levels of light emitted from living cells and animals as models of human biology and diseases (Thorne and Contag 2005) (Fig. 1). This requires that the imaging system incorporates an extremely light-tight enclosure, a very sensitive camera capable of long exposure times (occasionally as long as 20 min) and collection optics with a low f-number to collect as many photons as possible (Troy et al. 2004). After addition or injection of substrate, the cells or anaesthetized animals are placed in the light tight chamber and then light signals are captured for minutes. Visible light is dramatically absorbed and scattered by mammalian tissues, and thus the utility of this region of the spectrum for imaging is largely in small animals or superficial tissue sites of larger animals and humans. The use of light in the near-infrared (NIR) region of the spectrum, where tissue is less absorbing, allows interrogation of biological processes that occur several centimeters deep in tissue. Generations of red-shifted luciferases are therefore more efficient to produce better signals (Loening et al. 2007; Zhao et al. 2005). However, absence of auto-luminescence in mammalian tissues results in very high signal to noise ratio (SNR) yielding very sensitive measurements (Troy et al. 2004).

Fig. 1 Real-time bioluminescence imaging of a living mouse bearing tumors expressing N2atk20 rl fusion protein. A total of 2 x 106 of N2a cells stably expressing the tk20 rl fusion gene, and control N2a cells, were implanted on the left and right shoulders, respectively, of a single nude mouse and imaged daily using the optical CCD camera after injection of coelenterazine. A gradual increase in bioluminescence was observed in the tumor expressing tk20 rl fusion over time but not in the control tumor. Reproduced from Ray et al. (2003)

Illustrating Molecular Events with Light

143

To date, bioluminescence imaging is the most sensitive imaging technique for small animals and probably can detect as little as 10–15 moles of probe/liter of tissues (Massoud and Gambhir 2003). Fluorescent Proteins The second class of optical reporters or the fluorescent proteins are found exclusively in marine animals and had been used in research for the last three decades. Currently, a palette of these proteins are available that emit light at every wavelength of the visible spectrum ranging from blue to far-red region. Green Fluorescent Proteins The first fluorescent protein (green fluorescent protein or GFP) was cloned from a jellyfish called Aquaria victoria by Prasher et al. (1992), which emits light at 529 nm (green) after being excited at 480 nm (blue) light (Prasher et al. 1992). The power of this protein was soon realized when few groups showed that it can be functionally expressed in other organisms and as fusion partner with other proteins. Subsequently, many GFP mutants were created for better stabilization, monomerization, better expression in mammalian cells and different emission peaks (Wahlfors et al. 2001). All these reporters contain an 11-stranded β barrel, with both ends capped between β strands. This structure is conserved in most of the fluorescent proteins with minimal sequence homology at DNA level. The fluorophore is buried in an α helix and mainly formed by an intramolecular cyclization of the core amino acids Ser65, Tyr66, Gly67 producing a p-hydroxybenzylideneimidazolinone in the center of the interior α helix. This structure is required for both absorption and fluorescence; however, the rate of this process is also determined by the surrounding sequence and the external environment (pH, presence of oxygen) (Remington 2006). Red Fluorescent Proteins The other widely used fluorescent protein was drFP583 (Matz et al. 1999) or DsRed isolated from coral that has excitation and emission peak at 558 and 583 nm. This red-shifted emission peak generated considerable expectation for its application for in vivo imaging. However, the native protein, DsRed, exists as an obligate tetramer and often form insoluble aggregates in cells. Several groups have then tried mutagenesis and generated a family of mutant proteins with monomeric or dimeric structure and with much higher emission spectra (farthest is the mCherry mutant at 615 nm) (Shaner et al. 2004, 2005; Campbell et al. 2002). In DsRed, and other red-shifted fluorescent proteins, the peptide bond before the Ser65/Tyr66/Gly67 fluorophore is oxidized, leading to delocalized electron density over the polypeptide bond and longer wavelength excitation and emission. Variations on this structure produce long-range emissions from blue to far red.

144

P. Ray

In vivo fluorescence imaging: Parallel to the molecular development of fluorescent proteins, advancement in sensitivity of light detectors (photodiodes and charge coupled detectors or CCD cameras) leads to revolution in fluorescence microscopy (confocal microscopy, total internal reflection fluorescence or TIRF, two-photon microscopy) (Weissleder and Pittet 2008; Levenson and Mansfield 2006; Hoffman 2005). It is now possible to detect a single fluorescent molecule over time from a single live cell. This improvement in technology also leads to the development of in vivo fluorescence imaging systems such as Xenogen optical imaging system, Maestro Spectral imaging systems that enable us to visualize expression of fluorescent proteins from a whole animal (Troy et al. 2004). In fluorescent imaging, an external light of appropriate wavelength is used to excite a target fluorescent molecule present inside the body, followed almost immediately by release of longerwavelength, lower-energy light for imaging. Other than the fluorescent reporter genes, optical contrast agents, quantum dots, nanoparticles, or fluorescent dyes are commonly used for cellular and molecular imaging in small animals, and these molecules have better potential to be translated from preclinical research to patient care. Unlike bioluminescence, imaging of fluorescence reporters does not need a chemical substrate. However, the use of fluorescence proteins can be constrained by the combination of absorption and scatter of both excitation and emission light and the natural autofluorescence of mammalian tissue. In vivo fluorescence imaging deeply suffers from tissue attenuation of light signals (in contrast to bioluminescence which is one-way trafficking of light, fluorescent imaging requires two-way trafficking of light) and autofluorescence properties of the biomolecules. This leads to a significant drop in signal to noise ratio and depth-dependent attenuation of signal. Autofluorescence is strongest when excited in the blue range of the spectrum and weaker when excited in the red or NIR (700–800 nm) range. Autofluorescence in tissue is primarily from components in skin (collagen, which fluoresces in the green) and food (chlorophyll breakdown products, which fluoresce in the red), although muscle, organs, and tumors are also autofluorescent, mostly in the visible, but also out into the near-infrared region. There are two primary methodologies for removing the interference from tissue autofluorescence in in vivo imaging. The first involves utilizing fluorescence lifetime information – tissue has a different fluorescence lifetime than the fluorophores used in biomedicine, and on this basis, the signals can be separated. However, the pulsed lasers typically used for this type of imaging can be expensive. The second means of removing interference from tissue autofluorescence is by using multispectral imaging and unmixing to separate and isolate the contributions from each of the fluorophores in a sample. With appropriate software tools for generating accurate spectral characterizations of the target and autofluorescent species, at least five spectrally distinct fluorescent sources can be unmixed in a single animal subject (Levenson and Mansfield 2006). In vivo fluorescent imaging encompasses a wide range of resolution and imaging depths, including <400 μM with intravital microscopy and whole-body imaging with 1–3 mm spatial resolution at <10 cm with fluorescence molecular tomography (FMT) (Montet et al. 2007;

Illustrating Molecular Events with Light

145

Ntziachristos et al. 2005). Development of filter sets and algorithms for unmixing spectra of several different fluorescent signals results in greater sensitivity for multiple fluorescent reporters in vivo. With all these advanced instrumentations utilizing the optical properties of light and highly sensitive improved mutants of fluorescent reporters, in vivo fluorescence imaging is slowly becoming an essential option for noninvasive optical imaging (Fig. 2).

Fig. 2 In vivo fluorescence imaging of brain tumor expressing green fluorescent protein. Realtime whole-body imaging of a U87-GFP human glioma growing in the brain of a nude mouse at (a) 1 week, (b) 3 weeks, and (c) 5 weeks after surgical orthotopic implantation. Reproduced from Hoffman and Yang (2006)

Development of Specialized Reporters After completion of the sequencing of human genome, a major thrust in biomedical research is to understand the interaction between different cellular proteins. Proteins are known as “molecular work horse,” and interaction between different proteins at different time points govern multiple cellular process. The same protein can interact with different proteins at different time or in different cell types resulting in completely different outcome. Protein–protein interaction is thus vital for life. Many human diseases occur as a result of aberrant protein–protein interactions due to either a missing protein partner or the presence of a wrong protein partner. Only over the last 4–5 years, specialized methods were developed that enabled us to detect protein–protein interactions from living animals using reporter genes (Massoud and Gambhir 2007).

Modified Mammalian Two-Hybrid Reporter System Yeast two-hybrid system is a classical method primarily established in yeast system to identify new cellular protein partners. Briefly this system contains two parts: (1) a protein (x) fused with a transactivator binding domain and (2) a known or assumed interacting protein (y) fused with DNA-binding domain and a separate plasmid with

146

P. Ray

a reporter gene under the transactivator responsible regulatory region. Interaction of x and y brings the DNA-binding domain and the transactivator domain in physical proximity thereby activating the transactivator. The activated transcription factor will then bind to the regulatory region and drive the transcription of the reporter gene. Presence of the reporter gene expression thus confirms the specific interaction of two proteins (x & y). Several groups have used this modified mammalian two-hybrid system to detect the interaction of two proteins by bioluminescence and micro-PET imaging noninvasively in living animals (Ray et al. 2002; Luker et al. 2002, 2003) (Fig. 5.1). The main drawback of this system is that it identifies the interaction occurring only in the nucleus.

Fluorescence Resonance Energy Transfer Protein–protein interaction often critically depends on the distance and close association of the protein partners. Fluorescence resonance energy transfer (FRET) can efficiently measure the distance-dependent interaction (>100 Å) of two fluorescence proteins. If the two fluorescence proteins (acceptor and donor) come in close proximity either as a fusion protein or through interactions of another two proteins, photon energy from the acceptor (excited by a specific wavelength) can excite the donor and release of emission light at another (higher) wavelength. FRET is a very popular method for measuring protein interactions from living cells (Galperin et al. 2004). However, measuring protein–protein interactions through FRET has not been achieved from living animals due to the inherent limitations of whole-body fluorescence imaging as described in the Section: In vivo fluorescence imaging.

Bioluminescence Resonance Energy Transfer In recent years, researchers have developed another distance-dependent assay to measure protein–protein interaction from living cells that can be applied to living animals as well. This bioluminescence resonance energy transfer (BRET) applies the similar phenomena like FRET but uses a bioluminescent protein as donor and a fluorescence protein as an acceptor. Here the energy transfer that excites the fluorescent donor comes from the emitted light of the bioluminescent acceptor after addition of proper substrate. BRET was easily adapted for whole-body imaging by eliminating the need of an excitation light (requires for FRET) and thereby reducing the tissue autofluorescence and tissue attenuation of the excited light (Branchini et al. 2005; De et al. 2007).

Spilt Reporter System The cytoplasm is the main site of the cellular milieu where most proteins interact with each other. Many of the cellular proteins, including important enzymes, form their quaternary functionally active structure by assembling one or more

Illustrating Molecular Events with Light

147

polypeptide fragments. Adaptation of this strategy to the “split reporter protein” system leads to development of a very sophisticated and efficient method to measure real-time protein–protein interactions. In the “split protein” strategy, a single reporter protein/enzyme is cleaved into N-terminal and C-terminal segments, and each segment is fused to one of two interacting proteins (X and Y). Physical interactions between the two proteins X and Y reconstitutes the functional reporter protein leading to signal generation that can be measured. An appropriate split point in the reporter should lead to two fragments that do not have significant affinity for each other and yet when brought together (through interaction of the two proteins being studied for their mutual affinity) lead to detectable signal. This split protein strategy can work through either protein-fragment complementation assays (PCA) or intein-mediated reconstitution assays. To date, several reporter proteins (e.g., β-lactamase, β-galactosidase, ubiquitin, dihydrofolate reductase, firefly luciferase, renilla luciferase, green fluorescent protein) have been adapted for splitprotein strategies by finding various split sites for each reporter protein (Galarneau et al. 2002; Stagljar et al. 1998; Ozawa et al. 2001; Paulmurugan and Gambhir 2003; Paulmurugan et al. 2002). The split luciferases and split GFP proteins are being extensively used for imaging many vital protein–protein or receptor–ligand interactions (Fig. 5.3).

Multimodality Reporters The last but not the least specialized reporter molecule involves a combination of one bioluminescent and one fluorescent gene molecularly stitched together either using an IRES (internal ribosome entry site) or two promoters or by direct fusing with each other. Scientists have developed several multimodality vectors and applied them to various applications ranging from detection of metastasis, immune cell trafficking, bacterial infection, and so forth (Ray et al. 2003; Gafni et al. 2004; Cao 2006; Kim et al. 2004) (Fig. 5.4). All these specialized reporters have extensively been used in preclinical setting with small animal optical imaging but yet to reach the clinic. However, they bring in very important information of normal and deregulated cells from small animal models and accelerate a systems-wide understanding of biological complexity.

Molecular Vectors Cancer biology faced a revolution as molecular vectors were designed from naturally occurring biomolecules that can propagate and insert foreign DNA in nonnative cells. This opened up both therapeutic and monitoring (imaging) potential for modern medicine. Reporter genes emerged as important tools from their early application in cell and molecular biology to the current in vivo imaging research. Imaging gene expression with reporter genes does not reflect the endogenous physiological

148

P. Ray

situation, but they offer a big advantage as one does not need a separate probe for imaging each and every single molecular event. The process of delivery of reporter genes is still being improved with new vectors, and hybrid vectors are being constantly constructed and validated in living animals.

Plasmids A plasmid is an extrachromosomal DNA molecule found in bacteria that can replicate and propagate its genome inside another cell. Plasmids serve as important tools for cell and molecular biology as they can carry and express a foreign gene in various types of cells. However, plasmids do not serve as a good vehicle for long-term studies since they cannot transfect primary and immune cells efficiently. Commercially available plasmid vectors often carry an antibiotic-resistance gene (either bacterial or mammalian) that helps them selectively grow in medium containing that particular antibiotic. Plasmids are commonly used for transient experimental studies and for creating stable cell lines. Several commercially available plasmid vectors with different reporter genes (e.g., firefly luciferase, renilla luciferase, gfp) are routinely used for different purposes (Pike et al. 2006).

Viral Vectors The second class of molecular vectors was isolated from different viruses (DNA or adeno and RNA or retroviruses) which are more efficient in gene delivery, sustained expression, and clinically relevant than the plasmid vectors (Fig. 3). Though all these virus particles were isolated from pathogenic or harmful natural virus, they were modified into safe vehicles by deleting the harmful genes. Based on their usage for gene therapy and imaging, there are four major groups of viral vectors: adenovirus, retrovirus, lentivirus, and adeno-associated virus. Another class of DNA-virus known as herpes simplex virus (HSV) genome is also under development as a vehicle for imaging gene therapy but needs to undergo substantial improvement (such as, complete abolishment of the lytic cycle, higher expression level) and therefore is not discussed in this review.

Adenoviruses Adenoviruses are the most common human pathogen that infects the eye, upper respiratory tract, and gastrointestinal epithelium. These are nonenveloped viruses containing a double-stranded DNA genome of about 35 kb. For developing an inactivated virus, two regulatory genes the E1 and E3 are deleted and then supplied in trans (gene elements provided in another plasmid/complementary strand), either by a helper virus or by a plasmid, when packaging the virus. Graham et al. (1977) first developed a cell line carrying the E1 gene that enabled the production of recombinant adenovirus in a helper-free environment . Since then, adenoviral vectors have

Illustrating Molecular Events with Light

149

Fig. 3 Optical imaging of molecular events with specialized reporters from living mouse. 1. Imaging protein–protein interaction in living mice with a modified yeast two-hybrid strategy. (a) Schematic diagram of the system for imaging the interaction of proteins X and Y. The first step involves the vectors pA-gal4-x and pB-vp16-y, which are used to drive transcription of gal4-x and vp16-y through use of promoters A and B. In the second step, the two fusion proteins GAL4X and VP16-Y interact because of the specificity of protein X for protein Y. Subsequently, the GAL4-X-Y-VP16 binds to GAL4-binding sites [five GAL4-binding sites (bs) are available] on a reporter template. This leads to VP16-mediated transactivation of firefly luciferase reporter gene expression under the control of GAL4 response elements in a minimal promoter. Transcription of the firefly luciferase reporter gene leads to firefly luciferase protein, which, in turn, leads to a detectable visible light signal in the presence of the appropriate substrate (D-luciferin), ATP, Mg2+ , and oxygen. The NFκB promoter was used for either pA or pB and TNFα-mediated induction. (b) In vivo optical CCD imaging of mice carrying transiently transfected 293T cells for induction of the yeast two-hybrid system. All images shown are the visible light image superimposed on the optical CCD bioluminescence image with a scale in photons/sec/cm2 /steridian (sr). Mice in top row were imaged after injection of D-luciferin but with no TNFα-mediated induction. Mice in bottom row were imaged after injection of D-luciferin after TNFα -mediated induction, showing marked gain in signal from the peritoneum over 30 h. Reproduced with permission from Ray et al. (2002). 2. Imaging protein–protein interaction with bioluminescence resonance energy transfer (BRET). (a) Schematic showing small molecule-mediated protein–protein interaction leading to bioluminescence resonance energy transfer (BRET). FKBP12 is fused to the N-terminus of RLUC donor protein, and a FRB is fused to the C-terminus of GFP2 acceptor protein. When the genes encoding for both of these two fusion proteins are expressed inside cells and rapamycin is

150

P. Ray



Fig. 3 (continued) present to mediate FRB–FKBP12 interaction, then resonance energy transfer occurs. This BRET signal can be detected by using the deap blue coelenterazine (DBC) substrate for RLUC. (b) Detection of in vivo BRET2 signal from specific protein–protein interaction. Dorsal view of a nude mouse implanted s.c. with 5×106 293T cells transiently transfected either with pBRET2 (L) or with pFKBP12-hRluc (LL) alone or cotransfected with pFKBP12-hRluc and pGFP2 -FRB (LR) in presence (right panel) or absence (left panel) of rapamycin. Mice that received the small molecule mediator drug, rapamycin (5 mg/kg) were injected i.p. immediately after cell implantation. The scan was performed 7 h after drug administration. Mice were scanned for 5 min integration time using either GFP2 or DBC filters in succession by injecting with 25 μg DBC intravenously. Reproduced with permission from De and Gambhir (2005). 3. Imaging protein–protein interaction with split-reporter-based complementation strategy. (a) Schematic diagram of two strategies for using split reporters to monitor protein–protein interactions. (A) Complementationmediated restoration of firefly luciferase activity. N-terminal half of firefly luciferase is attached to protein X through a short peptide FFAGYC, and the C-terminal half of firefly luciferase is connected to protein Y through the peptide CLKS. Interaction of protein X and Y recovers Fluc activity through protein complementation. (B) Split Intein (DnaE)-mediated protein splicing leads to firefly luciferase reconstitution. The N-terminal half of firefly luciferase is connected to the N-terminal half of DnaE (DnaE-n) with peptide FFAGYC. The N-terminal half of DnaE in turn is connected to protein X. Similarly, the C-terminal half of firefly luciferase is connected to the C-terminal half of DnaE (DnaE-c) with peptide CLKS, and the C-terminal half of intein is in turn connected to protein Y. The interaction of proteins X and Y mediates reconstitution through splicing of the N and C halves of DnaE. (b) In vivo optical CCD imaging of mice carrying transiently transfected 293T cells for induction of the complementation-based split luciferase system. All images shown are the visible light image superimposed on the optical CCD bioluminescence image with a scale in photons/sec/cm2 /steridian (sr). Mice were imaged in a supine position after i.p. injection of D-luciferin. (A) A set of nude mice were repetitively imaged after s.c. implantation of 293T cells transiently transfected with plasmids PD (site B), PG (site C), PD plus PG (site D), and mock transfected cells (site A). One group of mice was induced with TNF-α and the other group was not induced. Images are from one representative mouse from each group immediately after implanting cells (0 h) and 18 and 36 h after TNF- induction. The induced mouse showed higher Fluc signal at site D when compared with the mouse not receiving TNF-. The Fluc signal significantly increases after receiving TNF-. Reproduced with permission from Paulmurugan et al. (2002). 4. Multimodality imaging of reporter genes in living subjects. (a) Schematic diagram of a multimodality triple fusion vector carrying a fluorescence (monmeric red fluorescent protein or enhanced green fluorescent protein), a bioluminescence (firefly luciferase or renilla luciferase), and a PET (HSV1-sr39 thymidine kinase or HSV1-thymidine kinase) reporter genes connected by two small peptide linkers is shown in the middle. Transcription of this fusion vector yields a single mRNA and subsequent translation leads to a single polypeptide that is capable of retaining partial if not full activities of the three proteins fused. (b) Fluorescence, bioluminescence, and microPET imaging of triple fusion reporter vector expression in the same living nude mouse. Ten million 293T cells transiently expressing the CMV-hrl-mrfp-ttk, CMV-ttk, CMV-mrfp1, and CMVhrl plasmids were implanted s.c. at four sites on ventral side of a nude mouse and imaged the next day for fluorescence/bioluminescence and PET, using a cooled CCD camera and microPET respectively. Fluorescence imaging was performed by placing the mouse in a CCD camera for 1 s, and a fluorescence image was acquired with an excitation filter at 500–550 nm and an emission filter at 575–650 nm. Cells expressing the fusion (2.1a) and mrfp1 (2.1c) genes showed fluorescence, and the signal is recorded as maximum photons/sec/cm2 /sr (2.1). The same mouse was then scanned in the CCD camera for bioluminescence following injection of coelenterazine via tail vein, and bioluminescence signal is found in cells expressing the fusion (2.2a) and hrl (2.2d) and is recorded as maximum photons/sec/cm2 /sr (2.2). Following the optical scan, the same mouse was imaged by microPET using FHBG. Cells expressing the fusion reporter gene (2.3a and 2.4a) and ttk gene (2.3b and 2.4b) showed FHBG accumulation in coronal section (2.3) and trans-axial section (2.4). Nonspecific accumulation of tracer was found in the gastrointestinal tracts (GI) and bladder (attributable to clearance of FHBG). Reproduced with permission from Ray et al. (2004)

Illustrating Molecular Events with Light

151

received much attention as gene transfer agents and widely used in gene therapy and imaging applications (Honigman et al. 2001; Wu et al. 2002; Rehemtulla et al. 2002). The exact mechanism by which an adenovirus targets a host cell is poorly understood. Most likely the virus attaches with the cell membrane through a receptor and then enters through endocytosis. Inside the cell, the virus replicates as an independent episomal molecule but does not integrate into genome. Adenoviral vectors have high transduction efficiency, are capable of containing DNA inserts of 8 kb size, and can infect both replicating and differentiated cells. Rehemtulla et al. (2002) first monitored the efficacy of an adenoviral-mediated brain tumor gene therapy treated with 5-fluorocytosine using bioluminescence imaging (Rehemtulla et al. 2002). Subsequently, many studies were conducted on cardiac gene therapy, pancreatic islet transplantation, application of oncolytic adenovirus on antivascular and antitumor therapy, etc., using bioluminescence imaging of luciferase gene expression by adenoviral vectors (Chang et al. 2003; Wu et al. 2002; Afanasieva et al. 2003; Jounaidi et al. 2007; Sato et al. 2008; Serganova and Blasberg 2005).

Retroviruses Retroviruses are enveloped viruses that contain a single-stranded RNA which is reverse transcribed into a double-stranded DNA molecule after infecting a host cell. This DNA can then integrate into the host genome and express the viral genes and other genes of interest following successive generation. The basal retroviral genome can be divided into three transcriptional units: gag (codes for genes responsible for viral capsule proteins), pol (codes reverse transcriptase and integrase proteins), and env (codes for protein needed for receptor recognition and envelop anchoring). Presence of long-terminal repeat (LTR) is a very important feature of retroviral genome that initiates the viral DNA synthesis and regulates the transcription of viral genomes (De and Gambhir 2003). Retroviruses have been a popular tool for gene therapy and molecular imaging due to long-term and relatively stable expression of the transferred genes into the cells they transduce (Serganova et al. 2007). Moreover, their integration into the host cell usually does not alter the normal physiology. Subsequently, replication-deficient retroviral vectors and packaging cell lines were created for safety purpose. A major disadvantage of retroviral vector is that they can only transduce cells that divide shortly after viral infection. Thus, this vector has limited applications to important target organs like the liver, skeletal muscle, haematopoetic stem cells, and neuronal cells (Fig. 4).

Lentiviruses Lentiviruses are a subclass of retroviruses of which HIV (human immunodeficiency virus) is the most recently discovered member. The major advantage of

152

P. Ray

Fig. 4 Schematic diagrams showing basic differences of two major viral gene delivery approaches. (a) Retrovirus, containing a RNA genome, attaches to a host cell by binding to a specific cell surface receptor, and a double-stranded DNA is synthesized from the RNA genome by a process called reverse transcription. This DNA integrates into the host genome ensuring sustained expression of the transgene. (b) Adenovirus is comprised of a double-stranded genome. After entering the cell by a receptor-mediated endocytosis process, the viral DNA is released within the cell, which then leads to a synthesis of proteins using host cell machinery. Unlike retro or lentivirus, DNA of adenovirus does not integrate into the host genome. Reproduced with permission from De and Gambhir (2003)

using lentiviral vector is that they can infect proliferating and nonproliferating cells and integrate the required genes into host genome thereby exerting a stable expression. Several generations of lentiviral vectors have been introduced to eliminate the potential risk of remaking an infectious viral particle inside the cells. These safe viral vectors are now being evaluated as delivery vectors for therapeutic as well as imaging studies (De et al. 2003; Gafni et al. 2004; Bradbury et al. 2007).

Adeno-associated Viruses Adeno-associated viruses are nonpathogenic satellite viruses of other human viruses and require coinfection with either adenovirus or herpes simplex virus for their replications. AAV vectors drew attention due to their very site-specific integration capacity in Chromosome 19 with the help of their inverted terminal repeats.

Illustrating Molecular Events with Light

153

However, AAV vector can only incorporate 5 kb of foreign DNA that impose limitation for using a larger gene fragment (Gafni et al. 2004).

Applications Compared with other imaging modalities, there are many issues to resolve in order for optical reporter gene imaging to reach clinical applications. Several major issues include (1) insertion of the reporter genes in patients (e.g., gene therapy vector required, many approvals needed), (2) high tissue attenuation of light from depth more than 5 mm for most fluorescent genetic reporters, (3) logistical problems to image bioluminescent reporter genes (equipment, substrate injections, etc.). However, there remains interest in this approach primarily due to diversity of success for preclinical applications in small animals. Modern biomedical science has benefited greatly from what has been possible due to the advances in optical imaging instruments and the genetic reporter technology.

Preclinical Applications It is difficult to entirely encompass the current field of preclinical applications involving optical imaging due to the enormous number of existing and growing studies. Broadly, the field can be categorized in few major sections such as tumor/cancer biology, gene therapy/adaptive therapy, and cell trafficking. All these sections are interrelated and the abovementioned reporters and specialized reporters are employed to translate the molecular interaction in normal physiological milieu.

Tumor/Cancer Biology Application of bioluminescence imaging to follow bacterial infections or tumor cell proliferation was pioneered in 1990s by the laboratory of Dr. Chris Contag at Stanford University (Contag et al. 1998). They along with a company called Xenogen first described a highly sensitive CCD camera that can collect very low-level light from living animals (Contag et al. 1998). The first bioluminescence imaging of tumor cells was performed by Edinger et al. (1999) in Stanford University where they had implanted Hela cells stably expressing the firefly luciferase in living mice and imaged over time. They showed it was possible to image as low as 103 cells and follow metastasis with time (Edinger et al. 1999). Later many groups have used similar strategies to monitor drug therapy, micrometastasis, route of progression of different forms of cancer, and imaging cell death from treated tumors (Cao et al. 2006; Sahai et al. 2005; Sato et al. 2008). Transgenic animals especially transgenic mice have evolved as very useful tool to understand and identify critical molecular pathways associated with disease. Vooijs et al. (2002) described a conditional mouse model for retinoblastoma-dependent sporadic cancer and monitored the pituitary tumor development by bioluminescence imaging. Several groups used

154

P. Ray

luciferase to image tissue-specific promoter activity relevant to different cancers (Iyer et al. 2001). Understanding the direct and drug-modulated protein interactions is long-time focus for biotech industries. Recent development of in vivo bioluminescence imaging with yeast two-hybrid system, BRET technology, split reporters are immensely helping to translate the knowledge of cell-based assays of protein interactions to living animals (Massoud and Gambhir 2003, 2007) (Fig. 3(1)–(3)). Finally, multimodality fusion vectors are able to follow the same molecular events from single cell to multicellular organisms using these cell-based and animal-based imaging modalities (Fig. 5.4). In vivo whole-body fluorescence imaging of tumor metastases has lower sensitivity than bioluminescence imaging mostly due to autofluorescence of the tissues. However, a few groups have successfully imaged tumor proliferation and angiogenesis by creating special surgical procedure called “skin flap” technique in living mice and using a whole-body mouse imaging system (Hoffman 2005). They had also taken advantage of dual color produced by green and red fluorescence proteins (simultaneously excited by a blue LED flashlight) for simultaneous imaging of different molecular events in real time. Other in vivo fluorescence imaging systems such as IVIS-200 (Caliper, Alameda) (Winnard et al. 2006), MaestroTM in-vivo spectral imaging system (CRI, Cambridge) (Ray et al. 2007) were also used for imaging tumors expressing fluorescence reporters from living animals. Intravital multiphoton microscopy imaging is an alternative to whole-body imaging that can achieve depth resolution of up to 800 μm on a localized area using fluorescence (Brown et al. 2001). These images provide a great detail of the vasculature, migrations of tumor cells with much less interference from autofluorescence, and therefore are of interest for specialized studies.

Gene Therapy/Adoptive Therapy An alternative method of curing diseases specifically genetic diseases where few drugs are effective is gene therapy. Though envisioned back in 1960, gene therapy made very slow progress in reality for curing human diseases. This is primarily due to longer validation period where noninvasive imaging played a major role. The field of gene therapy/cell therapy or adaptive immune therapy has been extensively validated by reporter gene imaging, using different viral vectors described in the section: Viral vectors. Adenoviral (Ad) vectors are promising gene therapy vehicles due to their in vivo stability and efficiency, but their potential utility is compromised by their restricted tropism. Researchers modified the receptors present on the adenovirus to retarget them different tissues and imaged the efficiency of targeted delivery of the adenoviral vectors (Mocanu et al. 2007; Adams et al. 2002). Monitoring the efficacy of gene therapy either using a combination of reporter and therapeutic gene/drug (such as yeast cytosine deaminase/5 fluoro-cytosine) combination or by introducing an apoptosis-inducing gene (such as p53-VP22, tumor necrosis factor-related apoptosis-inducing ligand or TRAIL) in luciferase-expressing tumors,

Illustrating Molecular Events with Light

155

or by grafting pancreatic islets transduced with an adenovirus-expressing luciferase extracted many clinically relevant information (Rehemtulla et al. 2002; Zender et al. 2002; Lu et al. 2004; Iyer et al. 2005). Lentiviral vectors are able to integrate in the mammalian genome exerting a sustained expression and are preferred choice for gene therapy. Researchers have validated the use of lentiviral vectors for gene therapy in different models (Cao et al. 2007; Gafni et al. 2004).

Cell Trafficking Cell trafficking is a natural phenomenon of development which is also associated with various immunological diseases and cancer. Better understanding of trafficking patterns of cells is very crucial for optimizing immune cell therapies. The classical procedure of imaging cell trafficking involves ex vivo labeling cells either with a paramagnetic particle (SPIO) or with a radiolabeled tracer (such as [18F]FDG, indium-111 oxine), injecting the cells in living animals and then monitoring either with MR or SPECT/PET modalities over time. However, over time, cells tend to lose the tracer due to dilution through cell division, tracer release, and radioactive decay; hence prelabeling is not a suitable approach for long-term cell trafficking monitoring. This potential drawback can be overcome by labeling the cells with reporter genes and watching the trafficking in real time in living animals by imaging. Cell trafficking with reporter genes has been explored with MR, PET, and optical imaging with various levels of sensitivity and resolution. Edinger et al. (2003) have evaluated the migration of two different types of tumor cells (lymphoma and leukemia) labeled with luciferase-gfp fusion reporter in control mice, irradiated mice, and mice undergoing chemotherapy by bioluminescence imaging. They also followed the trafficking of the CIK (cytokine-induced killer cells) adoptively transferred with the same luciferase-gfp fusion reporter to a lymphoma tumor and successive tumor regression in real time (Edinger et al. 1999). However, optical imaging suffers from signal attenuation and therefore probably not the ideal modality for tracking cells for clinical applications; however, a multimodality approach (optical with PET or optical with MR) is more appropriate (Shu et al. 2005; Yaghoubi et al. 2007).

Clinical Application The scope of applying optical imaging with reporter genes in clinic is still in infancy. While near-infrared fluorescence endoscopy in combination with fluorescent substrates or fluorescently labeled peptides, compact optical breast imaging system, hand-held fluorescence imaging system for sentinel lymph-node mapping, and few other techniques have already entered clinical practice, the reporter gene-based optical imaging have several disadvantages for direct clinical application. The main obstructions of implementing reporter gene imaging in clinic are (i) introduction of

156

P. Ray

a foreign gene/probe and (ii) attenuation of signal with depth of tissue in human body. However, a combination of optical and nonoptical (radionuclide-based PET or SPECT imaging or MR imaging) techniques will have the possibility to bring reporter genes in clinic. We can envision that in case of human cell therapy studies where patient’s own T cells will be injected back after ex vivo transduction with a therapeutic and a reporter gene to produce the therapeutic effect, visualization and assessment with imaging will have tremendous clinical impact. Such clinical studies are already ongoing using HSV1-thymidine kinase, a PET reporter gene in glioblastoma patients from City of Hope in collaboration with UCLA and Stanford University (Yaghoubi 2008). It is only a matter of time that creation of further sophisticated optical imaging instruments and optical probes would allow optical reporter genes play a similar role.

Challenges and Future Directions In the past decades, enormous advancements have been made in the clinical and preclinical imaging sciences by developing new technologies and integrating the knowledge of biology, chemistry, physics, and engineering in interdisciplinary centers across the world. Imaging has become an indispensable tool in cancer research, clinical trials, and medical practice. Perhaps the biggest growth has occurred in the area of fluorescence imaging with various technologies being adapted for in vivo analysis. However, several inherent limitations have restricted the fast translation of reporter gene imaging into the clinic. Reporter gene-based bioluminescence and fluorescence imaging is and will remain complementary to other imaging modalities, such as positron emission tomography, magnetic resonance imaging, and computed tomography but provides a number of unique capabilities from its intrinsic molecular sensitivity, repeatability and relatively low-cost instrumentation and probes (material). The limited penetration of light into tissues means its main role will be in surface imaging including the deep tissue imaging with fiber optics/endoscopy. However, in preclinical setting optical imaging will probably win over all other existing imaging modalities because of its high sensitivity, low-cost instrumentation, and by the wealth of specialized reporters (described in section: Development of Specialized Reporters) to detect the interactions of cellular proteins. In summary, reporter gene imaging has already exhibited its value across a wide range of biomedical applications, and its progress will continue to accelerate the fundamental understanding of cancer thus allowing earlier detection, better treatment, and new therapies. Acknowledgements The author sincerely acknowledges the mentorship, guidance, and encouragement from Prof. Sanjiv Sam Gambhir throughout her research in this exciting field of molecular imaging. The author further wishes to express her gratitude to Drs. Abhijit De and Drs. Shariar Yaghoubi for their help and for the critical review of the manuscript and thanks all of her colleagues at Stanford and UCLA for their help and support.

Illustrating Molecular Events with Light

157

References Adams, J. Y., M. Johnson, M. Sato, F. Berger, S. S. Gambhir, M. Carey, M. L. Iruela-Arispe, and L. Wu. 2002. Visualization of advanced human prostate cancer lesions in living mice by a targeted gene transfer vector and optical imaging. Nat Med 8 (8):891–7. Afanasieva, T. A., M. Wittmer, A. Vitaliti, M. Ajmo, D. Neri, and R. Klemenz. 2003. Single-chain antibody and its derivatives directed against vascular endothelial growth factor: application for antiangiogenic gene therapy. Gene Ther 10 (21):1850–9. Bhaumik, S., and S. S. Gambhir. 2002. Optical imaging of renilla luciferase reporter gene expression in living mice. Proc Natl Acad Sci U S A 99 (1):377–82. Bradbury, M. S., G. Panagiotakos, B. K. Chan, M. Tomishima, P. Zanzonico, J. Vider, V. Ponomarev, L. Studer, and V. Tabar. 2007. Optical bioluminescence imaging of human ES cell progeny in the rodent CNS. J Neurochem 102 (6):2029–39. Branchini, B. R., M. H. Murtiashaw, T. L. Southworth, S. R. Wilkinson, N. F. KHattak, J. C. Rosenberg, and M. Zimmer. 2005. Mutagenesis evidence that the partial reactions of firefly bioluminescence are catalyzed by different conformations of the luciferase C-terminal domain. Biochemistry 44:1385–93. Brown, E. B., R. B. Campbell, Y. Tsuzuki, L. Xu, P. Carmeliet, D. Fukumura, and R. K. Jain. 2001. In vivo measurement of gene expression, angiogenesis and physiological function in tumors using multiphoton laser scanning microscopy. Nat Med 7 (7):864–8. Campbell, R. E., A. E. Palmer, O. Tour, P. A. Steinbach, G. S. Baird, D. A. Zacharias, and R. Y. Tsien. 2002. A monomeric red fluorescent protein. Proc Natl Acad Sci U S A 99:7877–82. Cao, F., M. Drukker, S. Lin, A. Y. Sheikh, X. Xie, Z. Li, A. J. Connolly, I. L. Weissman, and J. C. Wu. 2007. Molecular imaging of embryonic stem cell misbehavior and suicide gene ablation. Cloning Stem Cells 9 (1):107–17. Cao, F., S. Lin, X. Xie, P. Ray, M. Patel, X. Zhang, M. Drukker, S. J. Dylla, A. J. Connolly, X. Chen, I. L. Weissman, S. S. Gambhir, and J. C. Wu. 2006. In vivo visualization of embryonic stem cell survival, proliferation, and migration after cardiac delivery. Circulation 113 (7):1005–14. Chang, S. C., H. L. Chuang, Y. R. Chen, J. K. Chen, H. Y. Chung, Y. L. Lu, H. Y. Lin, C. L. Tai, and J. Lou. 2003. Ex vivo gene therapy in autologous bone marrow stromal stem cells for tissue-engineered maxillofacial bone regeneration. Gene Ther 10 (24):2013–9. Contag, P. R., I. N. Olomu, D. K. Stevenson, and C. H. Contag. 1998. Bioluminescent indicators in living mammals. Nat Med 4 (2):245–7. Day, J. C., L. C. Tisi, and M. J. Bailey. 2004. Evolution of beetle bioluminescence: the origin of beetle luciferin. Luminescence 19 (1):8–20. De, A., and S. S. Gambhir. 2003. PET in imaging gene expression and therapy. Positron Emission Tomography: Basic Science and Clinical practice, Springer Publication, New York, 845–68. De, A., and S. S. Gambhir. 2005. Noninvasive imaging of protein–protein interactions from live cells and living subjects using bioluminescence resonance energy transfer. Faseb J 19 (14):2017–9. De, A., X. Z. Lewis, and S. S. Gambhir. 2003. Noninvasive imaging of lentiviral-mediated reporter gene expression in living mice. Mol Ther 7 (5 Pt 1):681–91. De, A., A. M. Loening, and S. S. Gambhir. 2007. An improved bioluminescence resonance energy transfer strategy for imaging intracellular events in single cells and living subjects. Cancer Res 67 (15):7175–83. de Wet, J. R., K. V. Wood, M. DeLuca, D. R. Helinski, and S. Subramani. 1987. Firefly luciferase gene: structure and expression in mammalian cells. Mol Cell Biol 7 (2):725–37. de Wet, J. R., K. V. Wood, D. R. Helinski, and M. DeLuca. 1985. Cloning of firefly luciferase cDNA and the expression of active luciferase in Escherichia coli. Proc Natl Acad Sci U S A 82 (23):7870–3. Doyle, T. C., S. M. Burns, and C. H. Contag. 2004. In vivo bioluminescence imaging for integrated studies of infection. Cell Microbiol 6 (4):303–17. Edinger, M., Y.A. Cao, M.R. Verneris, M.H. Bachmann, C.H. Contag, and R.S. Negrin. 2003. Blood 101 (2): 640–8.

158

P. Ray

Edinger, M., T. J. Sweeney, A. A. Tucker, A. B. Olomu, R. S. Negrin, and C. H. Contag. 1999. Noninvasive assessment of tumor cell proliferation in animal models. Neoplasia 1 (4):303–10. Gafni, Y., G. Pelled, Y. Zilberman, G. Turgeman, F. Apparailly, H. Yotvat, E. Galun, Z. Gazit, C. Jorgensen, and D. Gazit. 2004. Gene therapy platform for bone regeneration using an exogenously regulated, AAV-2-based gene expression system. Mol Ther 9 (4):587–95. Galarneau, A., M. Primeau, L. E. Trudeau, and S. W. Michnick. 2002. Beta-lactamase protein fragment complementation assays as in vivo and in vitro sensors of protein protein interactions. Nat Biotechnol 20 (6):619–22. Galperin, E., V. V. Verkhusha, and A. Sorkin. 2004. Three-chromophore FRET microscopy to analyze multiprotein interactions in living cells. Nat Methods 1 (3):209–17. Gilad, A. A., P. T. Winnard, Jr., P. C. van Zijl, and J. W. Bulte. 2007. Developing MR reporter genes: promises and pitfalls. NMR Biomed 20 (3):275–90. Graham, F. L., J. Smiley, W. C. Russell, and R. Nairn. 1977. Characteristics of a human cell line transformed by DNA from human adenovirus type 5. J Gen Virol 36 (1):59–74. Greer, L. F., 3rd, and A. A. Szalay. 2002. Imaging of light emission from the expression of luciferases in living cells and organisms: a review. Luminescence 17 (1):43–74. Hastings, J. W. 1983. Biological diversity, chemical mechanisms, and the evolutionary origins of bioluminescent systems. J Mol Evol 19 (5):309–21. Hoffman, R. M. 2005. Advantages of multi-color fluorescent proteins for whole-body and in vivo cellular imaging. J Biomed Opt 10 (4):41202. Hoffman, R. M., and M. Yang. 2006. Whole-body imaging with fluorescent proteins. Nat Protoc 1 (3):1429–38. Honigman, A., E. Zeira, P. Ohana, R. Abramovitz, E. Tavor, I. Bar, Y. Zilberman, R. Rabinovsky, D. Gazit, A. Joseph, A. Panet, E. Shai, A. Palmon, M. Laster, and E. Galun. 2001. Imaging transgene expression in live animals. Mol Ther 4 (3):239–49. Iyer, M., M. Sato, M. Johnson, S. S. Gambhir, and L. Wu. 2005. Applications of molecular imaging in cancer gene therapy. Curr Gene Ther 5 (6):607–18. Iyer, M., L. Wu, M. Carey, Y. Wang, A. Smallwood, and S. S. Gambhir. 2001. Two-step transcriptional amplification as a method for imaging reporter gene expression using weak promoters. Proc Natl Acad Sci U S A 98 (25):14595–600. Jounaidi, Y., J. C. Doloff, and D. J. Waxman. 2007. Conditionally replicating adenoviruses for cancer treatment. Curr Cancer Drug Targets 7 (3):285–301. Kim, Y. J., P. Ray, P. Dubey, S. S. Gambhir, and O. N. Witte. 2004. Multimodality imaging of lymphocytic migration using lentiviral-based transduction of a tri-fusion reporter gene. Mol Imaging Biol 6 (5):331–40. Levenson, R. M., and J. R. Mansfield. 2006. Multispectral imaging in biology and medicine: slices of life. Cytometry A 69 (8):748–58. Loening, A. M., A. M. Wu, and S. S. Gambhir. 2007. Red-shifted renilla reniformis luciferase variants for imaging in living subjects. Nat Methods 4 (8):641–3. Lu, Y., H. Dang, B. Middleton, Z. Zhang, L. Washburn, M. Campbell-Thompson, M. A. Atkinson, S. S. Gambhir, J. Tian, and D. L. Kaufman. 2004. Bioluminescent monitoring of islet graft survival after transplantation. Mol Ther 9 (3):428–35. Luker, G. D., V. Sharma, C. M. Pica, J. L. Dahlheimer, W. Li, J. Ochesky, C. E. Ryan, H. PiwnicaWorms, and D. Piwnica-Worms. 2002. Noninvasive imaging of protein–protein interactions in living animals. Proc Natl Acad Sci U S A 99 (10):6961–6. Luker, G. D., V. Sharma, C. M. Pica, J. L. Prior, W. Li, and D. Piwnica-Worms. 2003. Molecular imaging of protein–protein interactions: controlled expression of p53 and large T-antigen fusion proteins in vivo. Cancer Res 63 (8):1780–8. Massoud, T. F., and S. S. Gambhir. 2003. Molecular imaging in living subjects: seeing fundamental biological processes in a new light. Genes Develop 17 (5):545–80. Massoud, T. F., and S. S. Gambhir. 2007. Integrating noninvasive molecular imaging into molecular medicine: an evolving paradigm. Trends Mol Med 13 (5):183–91.

Illustrating Molecular Events with Light

159

Matz, M. V., A. F. Fradkov, Y. A. Labas, A. P. Savitsky, A. G. Zaraisky, M. L. Markelov, and S. A. Lukyanov. 1999. Fluorescent proteins from nonbioluminescent Anthozoa species. Nat Biotechnol 17 (10):969–73. Meighen, E. A. 1993. Bacterial bioluminescence: organization, regulation, and application of the lux genes. Faseb J 7 (11):1016–22. Mocanu, J. D., K. W. Yip, N. M. Alajez, W. Shi, J. H. Li, S. J. Lunt, E. H. Moriyama, B. C. Wilson, M. Milosevic, K. W. Lo, N. van Rooijen, P. Busson, C. Bastianutto, and F. F. Liu. 2007. Imaging the modulation of adenoviral kinetics and biodistribution for cancer gene therapy. Mol Ther 15 (5):921–9. Montet, X., J. L. Figueiredo, H. Alencar, V. Ntziachristos, U. Mahmood, and R. Weissleder. 2007. Tomographic fluorescence imaging of tumor vascular volume in mice. Radiology 242 (3):751–8. Ntziachristos, V., J. Ripoll, L. V. Wang, and R. Weissleder. 2005. Looking and listening to light: the evolution of whole-body photonic imaging. Nat Biotechnol 23 (3):313–20. Ozawa, T., A. Kaihara, M. Sato, K. Tachihara, and Y. Umezawa. 2001. Split luciferase as an optical probe for detecting protein–protein interactions in mammalian cells based on protein splicing. Anal Chem 73 (11):2516–21. Paulmurugan, R., and S. S. Gambhir. 2003. Monitoring protein–protein interactions using split synthetic renilla luciferase protein-fragment-assisted complementation. Anal Chem 75 (7):1584–9. Paulmurugan, R., Y. Umezawa, and S. S. Gambhir. 2002. Noninvasive imaging of protein–protein interactions in living subjects by using reporter protein complementation and reconstitution strategies. Proc Natl Acad Sci U S A 99 (24):15608–13. Pike, L., J. Petravicz, and S. Wang. 2006. Bioluminescence imaging after HSV amplicon vector delivery into brain. J Gene Med 8 (7):804–13. Piwnica-Worms, D., D. P. Schuster, and J. R. Garbow. 2004. Molecular imaging of host-pathogen interactions in intact small animals. Cell Microbiol 6 (4):319–31. Prasher, D. C., V. K. Eckenrode, W. W. Ward, F. G. Prendergast, and M. J. Cormier. 1992. Primary structure of the Aequorea victoria green-fluorescent protein. Gene 111 (2):229–33. Ray, P., A. De, J. J. Min, R. Y. Tsien, and S. S. Gambhir. 2004. Imaging tri-fusion multimodality reporter gene expression in living subjects. Cancer Res 64 (4):1323–30. Ray, P., H. Pimenta, R. Paulmurugan, F. Berger, M. E. Phelps, M. Iyer, and S. S. Gambhir. 2002. Noninvasive quantitative imaging of protein–protein interactions in living subjects. Proc Natl Acad Sci U S A 99 (5):3105–10. Ray, P., R. Tsien, and S. S. Gambhir. 2007. Construction and validation of improved triple fusion reporter gene vectors for molecular imaging of living subjects. Cancer Res 67 (7):3085–93. Ray, P., A. M. Wu, and S. S. Gambhir. 2003. Optical bioluminescence and positron emission tomography imaging of a novel fusion reporter gene in tumor xenografts of living mice. Cancer Res 63 (6):1160–5. Rehemtulla, A., D. E. Hall, L. D. Stegman, U. Prasad, G. Chen, M. S. Bhojani, T. L. Chenevert, and B. D. Ross. 2002. Molecular imaging of gene expression and efficacy following adenoviralmediated brain tumor gene therapy. Mol Imaging 1 (1):43–55. Remington, S. J. 2006. Fluorescent proteins: maturation, photochemistry and photophysics. Curr Opin Struct Biol 16 (6):714–21. Sahai, E., J. Wyckoff, U. Philippar, J. E. Segall, F. Gertler, and J. Condeelis. 2005. Simultaneous imaging of GFP, CFP and collagen in tumors in vivo using multiphoton microscopy. BMC Biotechnol 5:14. Sato, M., M. L. Figueiredo, J. B. Burton, M. Johnson, M. Chen, R. Powell, S. S. Gambhir, M. Carey, and L. Wu. 2008. Configurations of a two-tiered amplified gene expression system in adenoviral vectors designed to improve the specificity of in vivo prostate cancer imaging. Gene Ther 15 (8):583–93. Serganova, I., and R. Blasberg. 2005. Reporter gene imaging: potential impact on therapy. Nucl Med Biol 32 (7):763–80. Serganova, I., V. Ponomarev, and R. Blasberg. 2007. Human reporter genes: potential use in clinical studies. Nucl Med Biol 34 (7):791–807.

160

P. Ray

Shaner, N. C., R. E. Campbell, P. A. Steinbach, B. N. Giepmans, A. E. Palmer, and R. Y. Tsien. 2004. Improved monomeric red, orange and yellow fluorescent proteins derived from Discosoma sp. red fluorescent protein. Nat Biotechnol 22 (12):1567–72. Shaner, N. C., P. A. Steinbach, and R. Y. Tsien. 2005. A guide to choosing fluorescent proteins. Nat Methods 2 (12):905–9. Shu, C. J., S. Guo, Y. J. Kim, S. M. Shelly, A. Nijagal, P. Ray, S. S. Gambhir, C. G. Radu, and O. N. Witte. 2005. Visualization of a primary anti-tumor immune response by positron emission tomography. Proc Natl Acad Sci U S A 102 (48):17412–7. Stagljar, I., C. Korostensky, N. Johnsson, and S. Te Heesen. 1998. A genetic system based on splitubiquitin for the analysis of interactions between membrane proteins in vivo. Proc Natl Acad Sci U S A 95 (9):5187–92. Tannous, B. A., D. E. Kim, J. L. Fernandez, R. Weissleder, and X. O. Breakefield. 2005. Codonoptimized Gaussia luciferase cDNA for mammalian gene expression in culture and in vivo. Mol Ther 11 (3):435–43. Thorne, S. H., C. H. Contag. 2005. Using in vivo bioluminescence imaging to shed light on cancer biology. Proc IEEE 93 (4):750–62. Troy, T., D. Jekic-McMullen, L. Sambucetti, and B. Rice. 2004. Quantitative comparison of the sensitivity of detection of fluorescent and bioluminescent reporters in animal models. Mol Imaging 3 (1):9–23. Vooijs, M., J. Jonkers, S. Lyons, and A. Berns. 2002. Noninvasive imaging of spontaneous retinoblastoma pathway-dependent tumors in mice. Cancer Res 62 (6):1862–7. Wahlfors, J., S. Loimas, T. Pasanen, and T. Hakkarainen. 2001. Green fluorescent protein (GFP) fusion constructs in gene therapy research. Histochem Cell Biol 115 (1):59–65. Weissleder, R., and M. J. Pittet. 2008. Imaging in the era of molecular oncology. Nature 452 (7187):580–9. Winnard, P. T., Jr., J. B. Kluth, and V. Raman. 2006. Noninvasive optical tracking of red fluorescent protein-expressing cancer cells in a model of metastatic breast cancer. Neoplasia 8 (10):796–806. Wood, K. V., Y. A. Lam, H. H. Seliger, and W. D. McElroy. 1989. Complementary DNA coding click beetle luciferases can elicit bioluminescence of different colors. Science 244 (4905):700–2. Wu, J. C., M. Inubushi, G. Sundaresan, H. R. Schelbert, and S. S. Gambhir. 2002. Optical imaging of cardiac reporter gene expression in living rats. Circulation 105 (14):1631–4. Yaghoubi, S. S., R. J. Creusot, P. Ray, C. G. Fathman, and S. S. Gambhir. 2007. Multimodality imaging of T-cell hybridoma trafficking in collagen-induced arthritic mice: image-based estimation of the number of cells accumulating in mouse paws. J Biomed Opt 12 (6):064025. Yaghoubi, S. S., M. C. Jensen, N. Satyamurthy, S. Budhiraja, D. Paik, J. Czernin, and S. S. Gambhir. 2009. Non-invasive detection of therapeutic cytolytic T cells with 18F-FHBG PET in a patient with glioma. Nat. Clin. Pract. Oncol. 6 (1):53–8. Nat Clin Pract Oncol. Yu, Y. A., T. Timiryasova, Q. Zhang, R. Beltz, and A. A. Szalay. 2003. Optical imaging: bacteria, viruses, and mammalian cells encoding light-emitting proteins reveal the locations of primary tumors and metastases in animals. Anal Bioanal Chem 377 (6):964–72. Zender, L., R. Kock, M. Eckhard, B. Frericks, T. Gosling, T. Gebhardt, S. Drobek, M. Galanski, F. Kuhnel, M. Manns, and S. Kubicka. 2002. Gene therapy by intrahepatic and intratumoral trafficking of p53-VP22 induces regression of liver tumors. Gastroenterology 123 (2):608–18. Zhao, H., T. C. Doyle, O. Coquoz, F. Kalish, B. W. Rice, and C. H. Contag. 2005. Emission spectra of bioluminescent reporters and interaction with mammalian tissue determine the sensitivity of detection in vivo. J Biomed Opt 10 (4):41210.

Optical Imaging of Primary Tumors J. Robert Newman and Eben L. Rosenthal

Introduction The rapidly emerging field of optical imaging of tumors is an exciting area of translational research. A search in PubMed for “clinical optical imaging in cancer” reveals that the first article published was in 1964 and almost half of all the articles have been published in the last 5 years. Fluorescent probes and mechanisms to target them to cancer cells were developed for in vitro experiments and rodent models of cancer. However, with improvements in hardware, targeting agents, and fluorescent agents, optical imaging is rapidly gaining new and expanded applications in preclinical and clinical research. At the forefront of this field is the discovery of monoclonal antibodies as targeting agents and fluorescent markers that can reliably be conjugated to proteins, while minimizing interference from background fluorescence. Further sophistication and flexibility of imaging systems have and will allow these techniques to move further into the clinical realm. In this chapter, we will review the basic aspects of cancer optical imaging, including therapeutic, diagnostic, and surgical applications.

Photodynamic Detection Cancer detection and treatment modalities have been developed from photoactive compounds that can enhance visual demarcation between normal and neoplastic tissues. Photodynamic diagnosis techniques exploit the fluorescent properties of certain compounds or their biologic derivatives in combination with unique changes the compounds undergo in cancer tissue. Because these compounds will fluoresce in the visible spectrum when illuminated with light of the appropriate wavelength, visualization of malignant or pre-malignant tissue occurs without J.R. Newman (B) Division of Otolaryngology – Head and Neck Surgery, Department of Surgery, University of Alabama at Birmingham, Birmingham, AL, USA e-mail: [email protected]

E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_9, 

163

164

J.R. Newman and E.L. Rosenthal

the need for a complex image processing apparatus required for near-infrared imaging.

Aminolevulinic Acid and Hexaminolevulinate There are several photoactive agents currently in clinical use which non-specifically accumulate within malignant tissues at a significantly higher ratio compared to normal tissue. The most commonly used compounds are 5-aminolevulinic acid (ALA) and hexaminolevulinate (HAL). HAL contains a hexyl ester group that improves its solubility, facilitating distribution of the compound across tissues layers and in the delivery vehicle. When given orally, HAL also is better retained within the tumor and has higher cell cytotoxicity compared to ALA (Casas et al., 2001; Wu et al., 2003). These agents can be applied topically to mucosa within natural orifices or given orally. ALA is a natural precursor of the photoactive intermittent protoporphyrin IX. When exogenous ALA or related compounds are administered, the hemoglobin synthesis pathway becomes overloaded at the rate-limiting enzyme ferrochelatase (Fig. 1), which is responsible for the last step in the synthesis of hemoglobin. This results in the accumulation of the reactant protoporphyrin IX, a photosensitizing compound (Kennedy and Pottier, 1992). Protoporphyrin IX has been found to have an excitation frequency from 380 to 440 nm, which can be reproduced by a xenon light source. When protoporphyrin IX relaxes from a triplet state to a singlet state, it releases energy in the form of 635 nm light. Because the concentration of mitochondria (where hemoglobin biosynthesis occurs) is higher in malignant tissue, these cells quickly accumulate protoporphyrin IX and therefore preferentially emit more light than surrounding tissue (Regula et al., 1995). Compounds that exploit this pathway that are currently in clinical trials or approved for use include Hexvix (HAL; manufactured by GE Healthcare, London, United Kingdom), Medac (5-aminolevulinic acid, Hamburg Germany), and Photofrin (porfimer sodium, Axcan Pharma, Birmingham, AL). These compounds can be used for photodynamic therapy (PDT) or photodynamic diagnosis (also referred to more broadly as photodetection). Photodynamic diagnosis techniques are most commonly applied to detect precancerous lesions, carcinoma in situ, or other early lesions. Typically the photoactive compound can be administered topically within a biologic cavity such as the bladder or colon and then retained there for a limited amount of time prior to visualization by light with a wavelength between 380 and 440 nm. Biopsies of suspicious tissue are obtained for pathologic examination to confirm the presence of pre-malignant or malignant disease. Topical application in this manner has been used in the clinical setting for colon cancer, cervical intraepithelial neoplasm (CIN), bladder cancer, and oral cavity malignancies (Mayinger et al., 2008; Hillemanns et al., 2008; Fradet et al., 2007; Leunig et al., 2000). In non-epithelial cancers not accessible by a natural orifice, systemic administration allows specific uptake in tumors and tissues with high metabolic activity; ALA has been used clinically for detection of parathyroid glands and glioblastomas (Prosst et al., 2006; Stepp et al., 2007; Stummer et al., 2006) by oral administration.

Optical Imaging of Primary Tumors

165

Fig. 1 Hemoglobin biosynthesis pathway. Addition of exogenous ALA leads to accumulation of the photoactive compound protoporphyrin IX in the mitochondria. Excitation of protoporphyrin IX with a xenon light source causes tissues with a high mitochondrial content, such as parathyroid glands or malignant tissue, to fluoresce. Uro = uroporphyrinogen, UroD = uroporphyrinogen decarboxlyase, CPG = coproporphyrinogen

Clinical Studies Using Photodynamic Detection Topical Administration Most studies using HAL and ALA have been performed in Europe where these agents are available commercially for use in patients. HAL has been approved for early detection of bladder cancer in Europe and is currently in clinical trials

166

J.R. Newman and E.L. Rosenthal

in the United States. The US trial, sponsored by GE Healthcare (the maker of Hexvix), is an observational study to assess the impact of blue light cystoscopy after administration of HAL in the diagnosis and treatment of patients with noninvasive bladder cancer. Clinical studies have found that sensitivity is significantly improved over white light cystoscopy for detection of carcinoma in situ (Fradet et al., 2007). Furthermore, the 4-year follow-up data have demonstrated that overall there is a 10% increase in recurrence-free survival in patients who received photodetection treatment as opposed to the white light group. The two differences were found to be statistically significant and independent of each other, even when controlling for associated risk factors. Assessment of bladder cancer most commonly uses bladder rinses for several hours prior to endoscopic examination. There is heterogeneous distribution of the compound across the mucosa which results in variation of imaging characteristics between the patients. Comparison of ALA to HAL has demonstrated that HAL has twice the fluorescence of ALA at 40 times lower concentration because HAL is better distributed through the urothelial layers (Marti et al., 1999; Witjes and Douglass, 2007). Improved uptake by tissues is thought to result from improved aqueous solubility and excellent protoporphyrin IX formation at lower doses. Current practice is to use 1 hour of sensitization of bladder epithelium prior to cystoscopy, although shorter retention times are reported. Other studies have used transrectal administration of HAL with endoscopic examination of the colon using blue light fluorescence compared to white light (Mayinger et al., 2008). Photodynamic diagnosis in this case revealed detection of 99% of pre-malignant or malignant lesions compared to 71% of lesions detected with white light alone. Photodynamic diagnosis was better at demonstrating polyps than white light alone. Similar to bladder studies, fluorescent characteristics of precancerous lesions identified on colonoscopy improve with longer exposure times to HAL (retention times). HAL has also been used for photodynamic diagnosis in cervical intraepithelial neoplasia (CIN). A clinical series of 24 women with CIN grades 1 through 3 received topical HAL. Fluorescence was found to have improved diagnosis of CIN in this setting (Hillemanns et al., 2008). Photodynamic therapy in this setting can also be used for treatment since the lesions are very superficial, although studies thus far have produced equivocal outcomes (Hillemanns et al., 1999; Barnett et al., 2003). Similar studies have been performed for oral cavity malignancies, but have generated limited enthusiasm. Topical administration of ALA to oral cavity mucosa has been assessed, but it is somewhat more difficult to apply in a consistent manner as compared to the colon and bladder. Topically applied bioadhesive patches have been developed for delivery of ALA to the lip, but this is primarily for photodynamic therapy as opposed to diagnostic assessment (McCarron et al., 2006). Significant variation in the histological composition of the mucosa throughout the oral cavity and oropharynx makes this more difficult as well. Furthermore, oral cavity cancers tend to be associated with field cancerization (Braakhuis et al., 2002) which can cause diffuse and heterogeneous changes throughout the mucosa.

Optical Imaging of Primary Tumors

167

Oral Administration Systematic administration of ALA has been used clinically to detect brain tumors and parathyroid adenomas (Stepp et al., 2007; Stummer et al., 2006; Utsuki et al., 2007). Brain tumors are notoriously difficult to distinguish from normal tissue during surgical resection. Facilitating translation of optical imaging techniques is the reliance of neurosurgeons on microscopes during ablative procedures. A large randomized phase III trial of 322 patients with glioblastomas was analyzed (Stummer et al., 2006). This trial evaluated 6-month progression-free survival and found that those with white light had a 21% survival as opposed to those with 5 ALA that had a 41% survival, these differences were statistically significant. This trial has been criticized for not using surgical navigation techniques, and for not giving ALA to patients in the control group (but not using fluorescent guided surgery). Other studies using ALA to assess intraoperative tumor margins have found a high rate of false-positive results, most commonly associated with reactive or inflammatory tissue surrounding the tumor or related to concurrent infection (Utsuki et al., 2007). Parathyroid glands are often less than a centimeter in size and are removed for a variety of clinical indications. Because parathyroid gland identification in some patients is challenging due to aberrant locations or anatomic variations, the use of an imaging modality to positively identify parathyroid glands in real time has been investigated. Mice do not have readily identifiable parathyroid tissue, therefore studies on photodynamic detection of parathyroid tissues have been performed in rats. ALA has been found to be effective in preclinical and clinical assessment of parathyroid tissue; preclinical studies have found that parathyroid tissue fluorescence is 3.2 times higher than thyroid tissue on which it often lays (Fig. 2) (Prosst et al., 2004). These studies and others have identified optimal timing and dose in a preclinical model (Prosst et al., 2004, 2005). ALA preferentially accumulates in parathyroid tissue because of its high mitochondrial content. The use of optical imaging may facilitate endoscopic parathyroid surgery, which generally improves cosmetic and functional outcomes of these surgical procedures, and has been successful in this setting (Prosst et al., 2006).

Fig. 2 ALA-induced parathyroid fluorescence. Digital images using blue light illumination taken before (A) and 2 h after (B) a 500 mg/kg intraperitoneal injection of 5-ALA. (B) Arrows indicate the fluorescent bilateral parathyroid glands. (C) Histology of fluorescence guided biopsy confirms parathyroid fluorescence (magnification = 20×). Bar = 5 mm

168

J.R. Newman and E.L. Rosenthal

Photodynamic Therapy (PDT) Although not the focus of this text, ALA and related compounds have been used for therapeutic oncology. Administration of these photosensitizers followed by application of the appropriate wavelength light will result in the generation of reactive oxygen species-mediated cell death. Compared to diagnostic uses, PDT uses higher wavelengths of light between 630 and 850 nm to excite oxygen to a high energy state. Photofrin (porfirmer sodium) is the most widely used photosensitizer and has been widely studied (Berg et al., 2005; Chang and Wilder-Smith, 2005). Photofrin is FDA approved for use in treatment of precancerous lesions (Barrett’s esophagus (Overholt et al., 2007)) or in palliative management of obstructing bronchial cancers. It is a partially purified derivative of hematoporphyrin and can be activated by light at 630 nm. Although this treatment has been used successfully for some very superficial cancers, this technique has not been employed for larger malignancies because thickness of the tumor cannot be accurately assessed. Importantly, photodynamic therapy does not allow tumor margins to be assessed on final pathology since the tissue is not removed as part of the procedure. ALA has also been investigated for use in PDT (Pech et al., 2005). In fact, ALA has significantly less toxicity when uninvolved tissue is exposed to light compared to Photofrin and therefore is safer to use and has less complications, such as esophageal stenosis. Logically, the use of fluorescence for operative guidance and PDT have been united in the treatment of glioblastoma in the clinical setting (Eljamel et al., 2007). ALA-mediated PDT has been used alone for treatment of gliomas (Patrice et al., 2006; Hirschberg et al., 2006, 2004), but has limited application since a craniotomy is required and results do not compare favorably with conventional external beam radiation therapy. However, using a single agent for both imaging and treatment has significant appeal, as tumor can be maximally ablated by conventional surgical techniques and then the margins treated by photodynamic therapy to clear the residual microscopic disease (Stepp et al., 2007; Eljamel et al., 2007)).

Molecularly Targeted Contrast Agents Mechanisms to Molecularly Target Tumors There have been multiple mechanisms proposed to target tumor cells in vivo which can be considered in several broad categories: protease dependent, receptor ligand, and antibodies (Table 1). Each of these has been assessed in vivo for potential application in the optical detection of tumors. Protease-dependent agents exploit high levels of proteolytic enzymes found within tumors. Typically a substrate is administered that becomes fluorescent upon cleavage by tumor-specific proteases. These agents achieve high levels of fluorescence quickly, since the tumor-derived enzyme can cleave multiple fluorescent molecules. This topic is covered in more depth by McIntire et al. in the chapter of this book entitled “Proteinase optical imaging tools for cancer detection and response to therapy”. Receptor ligands have

Y

Y Y

Y

Y

Y

Trastuzumab (Herceptin) bevacizumab (Avastin) Panitumumab Colon

Colon, HNSCC

Breast

Glioblastoma, human embryonic kidney cells Glioma, medulloblastoma Breast, HNSCC

Bladder, esophageal, oral

Bladder, CIN, colon

Bladder, colon, glioblastoma, oral, PTG

Tumor types

Giusti et al., (2008) and Giusti et al., (2007)

Withrow et al., (2008), Stollman et al., (2008)

Stepp et al., (2007), Stummer et al., (2006), Marti et al., (1999), Witjes and Douglass (2007), McCarron et al., (2006), Utsuki et al., (2007), Prosst et al., (2004), (2005), Prosst et al., (2006) Mayinger et al., (2008), Hillemanns et al., (2008), Fradet et al., (2007), Marti et al., (1999), Witjes and Douglass (2007) Berg et al., (2005), Chang and Wilder-Smith (2005), Overholt et al., (2007) Garanger et al., (2007), Hsu et al., (2006), Jin et al., (2007), Wu et al., (2006) Mamelak et al., (2006), Hockaday et al., (2005), Veiseh et al., (2007) Rosenthal et al., (2007), Kulbersh et al., (2007), Gleysteen et al., (2007), Barrett et al., (2007) Koyama et al., (2007a), (2007b), Mume et al., (2005)

Citations

Abbreviations: ALA, aminolevulinic acid; HAL, hexaminolevulinate; RGD, Arg–Gly–Asp; CTX, chlorotoxin; PTG, parathyroid gland; CIN, cervical intraepithelial neoplasia; HNSCC, head and neck squamous cell carcinoma.

Y

Y

Cetuximab (Erbitux)

Y

N

Antibodies

N

Y

N

Y

Photofrin

Y

RGD base peptides CTX

N

HAL

Y

Clinical trial

Tumor specific ligands

N

ALA

Photodynamic agents

FDA approval

Agent

Class

Table 1 Targeting agents for optical imaging

Optical Imaging of Primary Tumors 169

170

J.R. Newman and E.L. Rosenthal

commonly been used to bind to integrins or molecular complexes on the cell surface of tumor cells. A significant amount of information for ligands as well as antibody tumor targeting has been derived from single photon emission computed tomography (SPECT) imaging and positron emission tomography (PET) techniques. Also commonly used for molecular imaging are antibodies and related constructs targeted against cell surface-expressed molecules. Antibodies most commonly employed for imaging purposes have been developed for therapeutic uses and then subsequently applied for imaging purposes. When evaluating targeting agents in cancer for optical imaging, experiments must consider the in vivo model, the targeting molecule, and the fluorophore. The nature of fluorophores is addressed by Ian Johnson in separate chapter, and animal models to evaluate surgically guided optical fluorescence will be discussed at the end of this chapter. Development of an effective agent for targeting tumors requires many characteristics (Table 2). First, the ideal agent for optical imaging of cancer should be non-toxic and specifically accumulate in the tumor, but not in normal tissues. Second, the agent should be specific for a variety of different tumor types. Although efficacy for different tumor types would need to be evaluated, establishing safety data for a single compound would significantly reduce the overall costs associated with development of these agents. Third, the tumor should consistently and homogeneously take up the molecule. If the tumor takes up the molecule in a heterogeneous fashion and a mosaic fluorescent appearance is present, then there may be small non-expressing areas of tumor that would go undetected (which reduces the sensitivity of this technique). Fourth, in order to allow for screening of early tumors, the targeted molecule should be expressed at high levels early in tumor progression in order to allow for detection of small amounts of early disease. Clinically, screening is perhaps one of the most important potential applications of this technique. Screening would allow the detection of sub-clinical cancer, especially if the specificity is sufficient to preclude histopathological confirmation of disease. Fifth, it is clinically important that the molecules have a rapid uptake within a short time (6 h) between the administration of the drug and imaging. Unless the agent can be used shortly after administration, the patient will need two visits: one to administer the agent and a second to perform the imaging. Proteins that are gradually taken up by tumors perhaps have the longest delay of up to 48–72 h. Protease-dependent techniques tend to have a shorter time frame due to the expediential nature of the enzymatic activation. Sixth, the target should be

Table 2 Characteristics of an ideal imaging agent

FDA approved for therapeutic indication Non-toxic Accumulates in tumor and not normal tissues Targets variety of tumor types Homogenously taken up by tumor Target expressed early in tumor progression Rapidly taken up by tumor Targeting molecule is retained by the tumor

Optical Imaging of Primary Tumors

171

tightly bound to the tumor and not diffuse into the circulation. This is more commonly associated with protease-dependent detection methods and less with agents that target cell surface receptors. It is commonly considered an advantage if the agent is rapidly cleared from the circulation because the background fluorescence derived from the unbound fluorescent agent in the circulating blood will be minimized. However, the longer a targeted agent circulates, the more of the agent is sequestered by the tumor, particularly if the agent is incorporated into the tumor cell after cell surface binding. A rapidly cleared receptor targeting agent (such as an antibody) may not accumulate within the tumor and would provide inadequate fluorescence.

Tumor-Specific Ligands Tumor-specific ligands bound to optical probes have been used to target growth factor receptors, integrins, or molecular complexes located on tumors. Typically these ligands are specific to certain tumor types. Perhaps the RGD peptides (Arg-Gly-Asp) are the most frequently used for tumor targeting and have been utilized for selective drug delivery (Garanger et al., 2007). They have also been used for tumor imaging in preclinical models for optical imaging and some clinical trials for SPECT (single photon emission computed tomography)-based imaging (Hsu et al., 2006; Jin et al., 2007; Wu et al., 2006). Various configurations of this peptide sequence have been deployed for imaging purposes and are largely specific to certain epithelial tumors. Others have explored the use of chlorotoxin for molecular imaging. Chlorotoxin is a 36 amino acid peptide that binds the MT1-MMP/MMP-2/TIMP2 complex. Chlorotoxin is known to specifically target the neuro system and therefore has been particularly useful with optical imaging of brain tumors. Chlorotoxin has found to be non-toxic in humans and has been radiolabeled for anti-neoplastic treatment for patients with gliomas (Mamelak et al., 2006; Hockaday et al., 2005). It has been demonstrated that conjugating chlorotoxin with Cy5.5 leads to biophotonic imaging of multiple tumor types including gliomas and medulloblastomas (Veiseh et al., 2007). The clinical application of this remains in question, however, because the studies have not provided stereomicroscopic imaging required for operative intervention, which is critical for demonstrating the clinical value. As noted earlier, identification of normal and tumor junction in invasive brain tumors is particularly challenging for the surgeon. Although fluorescence has been demonstrated using receptor ligands for biophotonic imaging and confocal microscopy, it is unclear if it has a role in the clinical detection of these tumors.

Antibodies Perhaps the most intuitive method for molecular imaging of tumors is the use of antibodies that have been previously developed for therapeutic reasons. This strategy has multiple advantages, including the fact that the targeted agent would already be FDA approved and therefore would minimize development costs. Furthermore,

172

J.R. Newman and E.L. Rosenthal

antibodies have the potential to molecularly characterize the tumor. A diagnostic dose of antibodies could be used to determine antigen expression level and quantitative optical imaging could be used to measure uptake of the antibody into the tumor and retention of the antibody in vivo which may vary significantly from patient to patient. Currently, levels of EGFR expression measured by immunohistochemistry have not been found to correlate with tumor response to targeted anti-epidermal growth factor receptor (anti-EGFR) antibody (Bonner et al., 2006; Vallbohmer et al., 2005). Alternatively, optical imaging could be used to predict tumor response to targeted therapy by multiple parameters such as fluorescence intensity, loss of fluorescence over time, or lifetime fluorescence. Because some antibodies are directed against proteins and not cell surface receptors, these are less advantageous for molecular imaging. A theoretical disadvantage of using antibodies for molecular targeting is that a sub-therapeutic dose is given to the patient for diagnostic purposes which may desensitize the tumor to subsequent therapeutic use of the antibody and promote resistance within the tumor to that agent. These theoretical risks would need to be weighed against the benefits after in vivo and clinical assessment. Anti-HER2-directed antibody (trastuzumab or Herceptin) has been used to image metastatic cells in vivo and molecularly characterize breast tumors in several studies of interest (Koyama et al., 2007a; b; Mume et al., 2005). Trastuzumab conjugated to a fluorescent probe selectively targeted pulmonary nodules formed by systematically injected embryonic fibroblasts that were engineered to express HER2 (Koyama et al., 2007b). This suggests the possible utility of this bioconjugate for imaging of breast cancer cells in vivo. However, less than a third of breast cancers overexpress the HER2 gene (Slamon et al., 1987), which limits the clinical use of this antibody for tumor detection. Probably the best application of optical imaging using Herceptin will be to identify those tumors that overexpress HER2 to guide treatment strategy. Herceptin-labeled antibodies have been used to molecularly characterize the receptor expression profile of breast tumors in vivo using a cocktail of therapeutic antibodies labeled with different fluorescent probes (Koyama et al., 2007a). Herceptin has also been conjugated to nanoparticles to improve ultrasound identification of cells that overexpress the HER2 gene (Liu et al., 2007). Avastin (bevacizumab) is targeted against vascular endothelial growth factor (VEGF) protein which circulates in the blood system and stimulates the VEGF receptor. It is highly expressed within most tumor types (Berger et al., 1995). Because this protein diffuses out of the tumor, the antibody antigen complex will diffuse out, theoretically limiting its potential for molecular imaging. The sensitivity and specificity of this antibody to detect tumors are low compared to therapeutic antibodies that bind to cell surface bound molecules (Rosenthal et al., 2007; Withrow et al., 2008). Others have used radiolabeled bevacizumab to image VEGF expression in murine xenografts using planar scintigraphy (Stollman et al., 2008). Fluorescently labeled bevacizumab may have limited uses for tumor detection, but may be used for determining potential response to VEGF treatment or predicting outcomes based on the optical imaging parameters suggested above.

Optical Imaging of Primary Tumors

173

Cetuximab (anti-epidermal growth factor receptor antibody, Erbitux) has been conjugated to Cy5.5 for molecular imaging and it has also been shown to have significant potential to detect small collections of less than 500 tumor cells (Rosenthal et al., 2007; Kulbersh et al., 2007). Cetuximab fits many of the characteristics of an ideal imaging agent, as it is preferentially expressed in tumor over normal tissues (Spaulding and Spaulding, 2002; Bonner et al., 2002), has already been FDA approved for patient use, has a relatively low toxicity profile (Bonner et al., 2006), its receptor has been shown to be overexpressed in a variety of tumors (Rivera et al., 2008), and is expressed on the surface of tumor cells. In fact, xenografted tumors are reliably detected in murine models of primary and metastatic human tumors, but xenografted normal human skin is not (Rosenthal et al., 2007; Kulbersh et al., 2007; Gleysteen et al., 2007). Interestingly, the cetuximab–Cy5.5 bioconjugate could be used to monitor disease response to cetuximab treatment using common optical imaging techniques. Administration of cetuximab has not been shown to minimize receptor uptake of fluorescence intensity (Gleysteen et al., 2007), which would be a requisite for monitoring disease response. Authors have also investigated the use of this agent to characterize receptor expression by tumors (Barrett et al., 2007). In addition to cetuximab, panitumumab has been approved for use in metastatic colon cancer (Giusti et al., 2008, 2007), but there is no imaging data with this agent.

Potential Clinical Applications Clinical Significance Molecular targeting in oncology has been largely considered in the therapeutic context; however, these same targeting mechanisms used to block function can provide a novel method of detecting or characterizing tumors by targeting specific enzymes, receptors, or proteins present on tumor cells but not on normal surrounding tissue. In addition to tumor detection, it is possible that these technologies could be used for sub-typing tumors based on receptor expression. The ability to molecularly characterize tumors in this non-invasive, biological manner could have significant implications for treatment strategies, staging, and prognosis. As noted above, it may be possible to measure the molecular characteristics of a tumor using a cocktail of targeting agents. For example, using a combination of three or four fluorescently labeled antibodies that target several growth factor receptors, it may be possible to molecularly characterize biomarker profiles of tumors that would guide treatment and provide prognostic information. For example, a fluorescent profile could be obtained prior to initiating therapy. However, this would require the ability to compare fluorescence intensity between patients. Quantification of fluorescence measurements remains a difficult task due to several factors that significantly effect fluorescence intensity including (1) the angle of the light that is shown on the tumor, (2) distance of the light from the tumor, and (3) variability in the thickness and size of the tumor. Furthermore, any overlying tissue of variable

174

J.R. Newman and E.L. Rosenthal

thickness (e.g., skin) would result in significant scatter that would be unpredictable. Recently, there has been identification of a potential correction using normalization values of white light reflection (Upadhyay et al., 2007). This helps to reduce variations in intensity associated with angulations of light on the surface, but does less for potential scatter for overlying tissue. However, it may be possible to measure other parameters such as lifetime fluorescence, change in fluorescence over time, or other characteristics that may guide quantitation of fluorescence with tumor variables (Upadhyay et al., 2007). Optical imaging for detection of cancer can be thought of broadly in three clinical applications: (1) screening, (2) staging, and (3) guiding procedures. When used for screening, molecularly targeted agents could confirm the diagnosis of sub-clinical malignancy and identify malignancies earlier and with greater confidence (avoiding or reducing the number of invasive biopsies). Accurate assessment of tumor size is difficult when clinicians rely on manual palpation or white light visualization alone, which explains why positive margins are as high as 40% in some tumor resections (McMahon et al., 2003; Woolgar and Triantafyllou, 2005). Classically tumor detection is thought of as a staging modality, such as in the use of PET scanning. It can also be an interactive process where optical imaging is used to identify and then resect tumor. After resection, margins can be assessed for sub-clinical disease. Screening for very early malignancies can be difficult due to the small size of the lesion, or the tissue changes can be so subtle that histopathological assessment is required for diagnosis. In the screening setting, the molecular agent would be administered and then followed by minimally invasive techniques (such as endoscopy) using optical imaging techniques. In this setting, small, sub-clinical tumors would be detected earlier than when using conventional techniques such as white light or metabolic imaging techniques. Because early tumors tend to have low metabolic rates and are small, PET or anatomic imaging has limited potential to detect them. Another advantage of optical imaging is that the technique would be performed and interpreted by the treating physician, not by a radiologist. Although potentially very sensitive, optical imaging for screening is only feasible in certain body cavities where endoscopic techniques are feasible. Because it requires absence of ambient light and superficial or epithelial disease, these techniques are ideal for use in colonoscopy (colon cancer), broncoscopy (lung cancer), laryngoscopy (laryngeal cancer), and colposcopy (cervical cancer) for screening. It would be less beneficial for examination of the lung which would require more invasive procedures such as video assisted thorascopy (VAT). Although near-infrared probes have a depth of penetration of 1–2 cm, early lesions would likely be very small with low-expression levels and therefore detection techniques would be limited to surface epithelium.

Intraoperative Guidance Tumor resections are currently performed in the same manner they were 100 years ago using gross palpation and clinician-based understanding of tumor margins. For accurate surgical resections using optical imaging, the molecular probe would need

Optical Imaging of Primary Tumors

175

to have a very homogeneous uptake within the tumor to allow detection of small islands of tumor cells. Adoption of this technique in the operating room could be conducted by marking the tumor edges by optical imaging and then resecting the tumor using white light. After gross tumor removal, the wound bed could be reexamined under near-infrared light and small areas of potential residual disease can be detected and re-resected or biopsied for frozen section to assess the need for further resection. The advantage of using optical imaging over conventional techniques is that it would minimize the morbidities associated with removal of uninvolved, normal tissue. Furthermore, injection of the fluorescent agent could be used to detect sub-clinical metastasis that occurs regionally. For example, in head and neck cancer accumulation of fluorescence could be detected during nodal basin dissection which is routinely performed. This would minimize the length of the procedure as well as patient morbidities associated with removal of a large number of uninvolved lymph nodes.

Head and Neck Squamous Cell Carcinoma Most primary oral cavity cancers and previously irradiated HNSCCs require surgical resection for definitive treatment. Indeed, the incidence of involved or close surgical margins in head and neck cancer patients approaches 40% on histopathological review (McMahon et al., 2003; Ravasz et al., 1991). Failure to obtain a complete tumor resection results in significantly worse outcome in head and neck cancer patients (Ravasz et al., 1991; Cook et al., 1993). Despite the importance of obtaining negative margins at the time of surgical resection, there have been minimal advances in the intraoperative detection of tumors. As a result, a significant portion of uninvolved structures are removed with the cancer in order to obtain a negative margin. Studies have been performed demonstrating the value of optical imaging in many cancer types. However, significant preclinical data exist for head and neck squamous cell carcinoma and glioblastomas. Currently, tumor margins are determined intraoperatively by a combination of palpation, visual inspection, and microscopic assessment of frozen tissue sections. With the exception of previously untreated tongue cancers, assessment of tumor margins by intraoperative palpation is limited because tumors are adjacent to bone, adjacent to cartilage (larynx), in deep tissues (pterygoid muscles), or in previously irradiated tissues. More accurate assessment of tumor extent could limit the ablative defect size. Our lab has shown in preclinical, xenograft models of head and neck cancer that cetuximab labeled with Cy5.5 for molecular imaging has significant potential to detect small collections of tumor of less than 500 cells (Kulbersh et al., 2007). Interestingly, cetuximab (anti-EGFR antibody, Erbitux) conjugated to Cy5.5 may be used to monitor disease response with common endoscopy techniques. Administration of cetuximab–Cy5.5 with therapeutic doses of cetuximab has not been shown to minimize receptor uptake of fluorescence intensity (Gleysteen et al., 2007). In our initial experiments we evaluated utility of cetuximab conjugated to Cy5.5 to detect sub-clinical primary (Rosenthal et al., 2006) and metastatic (Gleysteen et al., 2008)

176

J.R. Newman and E.L. Rosenthal

tumors in orthotopic and flank murine models. Furthermore, we evaluated the potential of bevacizumab (Withrow et al., 2008) and anti-CD147 antibody (Newman et al., 2008) bio-conjugated to Cy5.5 to detect tumors in flank xenograft murine models.

In Vivo Model to Determine Sensitivity and Specificity Modeling detection of sub-clinical tumor detection in a mouse model has many caveats, but remains an excellent method for comparing the effectiveness of different targeting agents and fluorophores. Although antibody clearance and the absence of endogenous human tissues make it difficult to predict how these agents will perform in clinical trials, these models represent a preclinical modality to optimize fluorescent probes and targeting agents. Our model has been used effectively to evaluate different antibodies (Withrow et al., 2008; Kulbersh et al., 2007) and fluorophores (Rosenthal et al., 2006; Withrow et al., 2007) and has been designed to closely recapitulate the clinical setting. For imaging, we used a custom-built Leica fluorescent stereomicroscope fitted with a Cy5.5 filter and an ORCA ER charge-coupled device camera (Hamamatsu, Bridgewater, NJ). Mock surgical resections are performed on mice bearing HNSCC xenografts to determine if Cy5.5-labeled cetuximab could be used to aid in tumor removal (Fig. 3). The cetuximab–Cy5.5 conjugate is systemically administered in sub-therapeutic doses (50 μg) to mice bearing subcutaneously xenografted HNSCC cell lines. After 48 or 72 h, the tumor underwent near-total resection, 2-mm cupped forceps were used to biopsy areas of negative and positive fluorescence (Fig. 4). These small biopsies were then paraffin embedded and serially sectioned for analysis. Fluorescent stereomicroscopic findings were then correlated with histological findings. Best results have been obtained with therapeutic (rather than diagnostic) doses. Specificity usually approaches 100%, meaning that there is minimal background and fluorescence signifies the presence of tumor. Sensitivity is usually lower because small islands of tumor cells are biopsied from the non-fluorescent wound bed. It is unknown whether this model can accurately translate to the clinic, however, it helps to compare various antibodies, doses, time points, and fluorophores in vivo. Models used for metastatic disease have been developed by our laboratory (Gleysteen et al., 2008). For detection of regional (cervical) lymphadenopathy, mice were given tongue xenografts with HNSCC tumor cell lines (Fig. 5, A–B). After 14 days, mice were injected systemically with 50 μg of cetuximab–Cy5.5. After 72 h of the injection of the labeled antibody, each mouse was sacrificed and placed on its back with forelegs outstretched and pinned down. A skin incision was made from the rib cage to the chin and the cervical skin was removed. Bright field and fluorescent images (at 800- and 200-ms exposure) of the neck were then taken (Fig. 5, C–E). Fluorescent regions were excised until no fluorescence could be detected at 800 ms exposure time. Each sample was then fixed, hematoxylin and eosin stained, sectioned, and placed on slides. Biopsies of the tongue (primary tumor) were also collected to pathologically confirm the presence of tumor. Fluorescent

Optical Imaging of Primary Tumors

177

Fig. 3 Fluorescence-guided tumor removal. Gross images (A, D, G), bright field stereomicroscopy (B, E, H), and red psuedocolored near-infrared fluorescent images (C, F, I) obtained during a mock resection of a SCC-1 flank xenograft. SCID mice bearing SCC-1 xenografts were systemically injected with cetuximab–Cy5.5. After 72 h, the skin was removed (A–C) and then the tumor partially resected (D–E). After the tumor was completely excised the wound bed was again imaged (G–I). Negative biopsies were taken from non-fluorescent areas (white arrow) and positive biopsies taken from fluorescent regions (green arrow). Bar = 1 mm A. FLANK XENOGRAFT MODEL RESULTS Results

50 µg dose

250 µg dose

Positive predictive value

100%

100%

Negative predictive value

100%

100%

Sensitivity

86%

91%

Specificity

100%

100%

Fig. 4 Results of fluorescence-guided biopsies. Biopsies taken from the tumor bed using 2-mm cup forceps demonstrate foci of tumor (arrows) that were not detected by fluorescence (B, false negative) and those that were detected by fluorescence (C, true positive). Bar = 100 μm

regions removed from the neck were found to contain tumor by histopathologic review. Pulmonary metastases were created by systemically injecting mice with HNSCC cells by tail vein. After 11 days, mice were injected with 50 μg of cetuximab–Cy5.5. On day 14, the lungs were removed from the chest to minimize background fluorescence and were placed in a dish on a black background. Brightfield and fluorescent images were obtained for each lung individually (Fig. 6). A speckled pattern of fluorescence was seen when lungs bearing metastatic tumors were imaged for Cy5.5 fluorescence. The lungs were then paraffin embedded, hematoxylin–eosin stained, and placed on slides for pathologic examination. Histopathology verified that the

178

J.R. Newman and E.L. Rosenthal

Fig. 5 Lymph node resection guided by cetuximab–Cy5.5 fluorescence. Orthotopic tumor model demonstrates the ability to visualize primary (B) and regional (C) tumor foci using antibodyguided fluorescence, 72 h after systemic injection of cetuximab–Cy5.5. Fluorescent areas were excised until fluorescence disappeared (D). Hematoxylin and eosin (H&E) stain (E) demonstrates the presence of tumor (arrow) within the lymph node

speckling pattern visualized during fluorescent imaging was due to the presence of metastatic tumor. To accurately assess the potential of these imaging strategies for the detection of tumors it is important to use a clinically relevant model and equipment. Imaging using small animal photon-based imagers (IVIS, Xenogen Corporation, California) cannot be used clinically and therefore assessment of optical imaging

Fig. 6 Cetuximab–Cy5.5 detection of pulmonary metastasis. Eleven days after systemic injection with HNSCC tumor cells, mice bearing pulmonary metastasis were injected with cetuximab– Cy5.5 bioconjugate. After 72 h, the lungs were excised and placed on a black background (A). Near-infrared fluorescent images revealed a distinctive milliary pattern of fluorescence (B). Tumor micrometastases were confirmed by pathology (C)

Optical Imaging of Primary Tumors

179

for clinical application should be limited to stereomicroscopy. We have recapitulated the surgical setting to facilitate the translation of this technology into the clinical setting. However, there are several pitfalls in these models. Although the targeting agent (cetuximab) is FDA approved, the fluorophore Cy5.5 is not. As mentioned previously, a translation to the clinic is more difficult with the time delay of 48–72 h between injection and maximal fluorescence. Our imaging system, while very similar to clinical microscopes, has yet to be used successfully with actual surgical operating microscopes. Differences in antibody clearance and expression of the antigen in normal tissues will significantly affect background fluorescence. Furthermore, the immunodeficient murine models do not have a desmoplastic or inflammatory reaction that is typical for human epithelial tumors. The infiltration of other cell types will certainly alter the fluorescence intensity at the borders and throughout the tumor. Finally, in visualizing the metastatic tumor in the lung, our best results were obtained by removing the lung and placing it on a black background. This would obviously be problematic in clinical use. Although it remains unclear if these models can translate to the clinic, they remain helpful to compare the potential of various agents.

Glioblastomas Attempts to accurately determine tumor margins using aminolevulinic acid and small proteins have been discussed above. Because the blood brain barrier remains an obstacle to protein transport, the use of antibodies and other large molecules has not been considered ideal for treatment or imaging of gliomas. However, this remains controversial since the tumor significantly disrupts the blood–brain barrier. Some recent work suggests that labeled chlorotoxin can effectively target intracranial tumors with intact blood–brain barriers (Veiseh et al., 2007). These authors demonstrated the ability to visualize tumor through intact skull and scalp and demonstrated that the blood–brain barrier was not disturbed.

Barriers to Clinical Practice Fluorescent agents that can be easily conjugated to proteins are not approved for clinical use. Although indocyanine green is an FDA approved fluorescent agent with spectral properties at high wavelengths, it is not possible to conjugate this molecule to proteins with any consistency. There are several reports of indocyanine green being conjugated to various ligands or antibodies, but these have not been duplicated (Withrow et al., 2008; De Grand and Frangioni, 2003; Ke et al., 2003). The most commonly used agent for conjugation to other molecules has been the Cy5.5 and Cy5 molecules, commercially available through GE Healthcare. Although easily bio-conjugated, they are not clinically approved. Cy5.5 and related agents are chemically very similar to indocyanine green which is used with minimal toxicity

180

J.R. Newman and E.L. Rosenthal

in patients for cardiac studies and lymph node mapping studies (Speich et al., 1988; Benya et al., 1989), however, the safety of Cy5.5 has not been established. FDA approval of these and other agents that can be easily bio-conjugated is of paramount importance in order to put these agents into early clinical trials. Surgical equipment and endoscopy hardware for near-infrared detection requires additional adaptation. As near-infrared probes cannot be detected directly by the human eye, some type of camera and monitor for viewing the image needs to be attached to the endoscope or microscope. There are multiple techniques that have been proposed for this application, including two side-by-side cameras with alternating images (De Grand and Frangioni, 2003) or a toggle switch for near-infrared and white light imaging. Anticipated barriers also include hardware limitations – particularly in surgery which requires significant interaction with the tissue. There are multiple techniques that can be used for this including the use of instrumentation that can be seen in infrared light. The technology exists for capturing these images, as indocyanine green is currently used in the clinic, but their use is experimental in nature and has not been commercially developed. Also limiting clinical use of optical imaging agents is the lack of quantitative detection methods. In the absence of a method for normalizing fluorescence intensity, lesions can be detected but could be difficult to characterize. Diagnostic agents have a significantly smaller market compared to therapeutic agents, and therefore the economics of diagnostic agents makes them less appealing to the pharmaceutical industry. Development costs of diagnostic and therapeutic agents are very similar, but because therapeutic agents require multiple doses and larger doses, these agents are relatively lucrative for the industry. Given that the requirements for FDA approval are similar, it is much more difficult to solicit industry partnership for the development of diagnostic agents. Therefore, it is particularly important for agents that are being developed to target multiple tumor types and have several potential clinical applications such as screening or surgical resection.

Summary Optical imaging to guide therapy has significant clinical advantages. It puts the imaging technology in the hands of clinicians and allows for real-time interactive imaging. However, because the best fluorophores for optical imaging are not currently approved for human use and hardware has not become commercially available for use, the clinical application of optical imaging remains several years in the future.

References Barnett AA, Haller JC, Cairnduff F, Lane G, Brown SB, Roberts DJ: A randomised, doubleblind, placebo-controlled trial of photodynamic therapy using 5-aminolaevulinic acid for the treatment of cervical intraepithelial neoplasia. Int J Cancer 2003, 103:829–832.

Optical Imaging of Primary Tumors

181

Barrett T, Koyama Y, Hama Y, Ravizzini G, Shin IS, Jang BS, Paik CH, Urano Y, Choyke PL, Kobayashi H: In vivo diagnosis of epidermal growth factor receptor expression using molecular imaging with a cocktail of optically labeled monoclonal antibodies. Clin Cancer Res 2007, 13:6639–6648. Benya R, Quintana J, Brundage B: Adverse reactions to indocyanine green: a case report and a review of the literature. Cathet Cardiovasc Diagn 1989, 17:231–233. Berg K, Selbo PK, Weyergang A, Dietze A, Prasmickaite L, Bonsted A, Engesaeter BO, AngellPetersen E, Warloe T, Frandsen N, Hogset A: Porphyrin-related photosensitizers for cancer imaging and therapeutic applications. J Microsc 2005, 218:133–147. Berger DP, Herbstritt L, Dengler WA, Marme D, Mertelsmann R, Fiebig HH: Vascular endothelial growth factor (VEGF) mRNA expression in human tumor models of different histologies. Ann Oncol 1995, 6:817–825. Bonner JA, De Los Santos J, Waksal HW, Needle MN, Trummel HQ, Raisch KP: Epidermal growth factor receptor as a therapeutic target in head and neck cancer. Semin Radiat Oncol 2002, 12:11–20. Bonner JA, Harari PM, Giralt J, Azarnia N, Shin DM, Cohen RB, Jones CU, Sur R, Raben D, Jassem J, et al.: Radiotherapy plus cetuximab for squamous-cell carcinoma of the head and neck. N Engl J Med 2006, 354:567–578. Braakhuis BJ, Tabor MP, Leemans CR, van der Waal I, Snow GB, Brakenhoff RH: Second primary tumors and field cancerization in oral and oropharyngeal cancer: molecular techniques provide new insights and definitions. Head Neck 2002, 24:198–206. Casas A, Perotti C, Fukuda H, Rogers L, Butler AR, Batlle A: ALA and ALA hexyl ester-induced porphyrin synthesis in chemically induced skin tumours: the role of different vehicles on improving photosensitization. Br J Cancer 2001, 85:1794–1800. Chang CJ, Wilder-Smith P: Topical application of photofrin for photodynamic diagnosis of oral neoplasms. Plast Reconstr Surg 2005, 115:1877–1886. Cook JA, Jones AS, Phillips DE, Soler Lluch E: Implications of tumour in resection margins following surgical treatment of squamous cell carcinoma of the head and neck. Clin Otolaryngol Allied Sci 1993, 18:37–41. Fradet Y, Grossman HB, Gomella L, Lerner S, Cookson M, Albala D, Droller MJ: A comparison of hexaminolevulinate fluorescence cystoscopy and white light cystoscopy for the detection of carcinoma in situ in patients with bladder cancer: a phase III, multicenter study. J Urol 2007, 178 :68–73; discussion 73. Garanger E, Boturyn D, Dumy P: Tumor targeting with RGD peptide ligands-design of new molecular conjugates for imaging and therapy of cancers. Anticancer Agents Med Chem 2007, 7:552–558. Giusti RM, Shastri KA, Cohen MH, Keegan P, Pazdur R: FDA drug approval summary: panitumumab (Vectibix). Oncologist 2007, 12:577–583. Giusti RM, Shastri K, Pilaro AM, Fuchs C, Cordoba-Rodriguez R, Koti K, Rothmann M, Men AY, Zhao H, Hughes M, et al.: U.S. Food and drug administration approval: panitumumab for epidermal growth factor receptor-expressing metastatic colorectal carcinoma with progression following fluoropyrimidine-, oxaliplatin-, and irinotecan-containing chemotherapy regimens. Clin Cancer Res 2008, 14:1296–1302. Gleysteen JP, Duncan RD, Magnuson JS, Skipper JB, Zinn K, Rosenthal EL: Fluorescently labeled cetuximab to evaluate head and neck cancer response to treatment. Cancer Biol Ther 2007, 6: 1181–1185. Gleysteen JP, Newman JR, Chhieng D, Frost A, Zinn KR, Rosenthal EL: Fluorescent labeled antiEGFR antibody for identification of regional and distant metastasis in a preclinical xenograft model. Head Neck 2008, 30(6):782–789. De Grand AM, Frangioni JV: An operational near-infrared fluorescence imaging system prototype for large animal surgery. Technol Cancer Res Treat 2003, 2:553–562. Eljamel MS, Goodman C, Moseley H: ALA and Photofrin(R) Fluorescence-guided resection and repetitive PDT in glioblastoma multiforme: a single centre Phase III randomised controlled trial. Lasers Med Sci 2007, 23(4):361–367.

182

J.R. Newman and E.L. Rosenthal

Hillemanns P, Korell M, Schmitt-Sody M, Baumgartner R, Beyer W, Kimmig R, Untch M, Hepp H: Photodynamic therapy in women with cervical intraepithelial neoplasia using topically applied 5-aminolevulinic acid. Int J Cancer 1999, 81:34–38. Hillemanns P, Wang X, Hertel H, Andikyan V, Hillemanns M, Stepp H, Soergel P: Pharmacokinetics and selectivity of porphyrin synthesis after topical application of hexaminolevulinate in patients with cervical intraepithelial neoplasia. Am J Obstet Gynecol 2008, 198:300.e1–300.e7. Hirschberg H, Sorensen DR, Angell-Petersen E, Peng Q, Tromberg B, Sun CH, Spetalen S, Madsen S: Repetitive photodynamic therapy of malignant brain tumors. J Environ Pathol Toxicol Oncol 2006, 25:261–279. Hirschberg H, Sun CH, Tromberg BJ, Yeh AT, Madsen SJ: Enhanced cytotoxic effects of 5aminolevulinic acid-mediated photodynamic therapy by concurrent hyperthermia in glioma spheroids. J Neurooncol 2004, 70:289–299. Hockaday DC, Shen S, Fiveash J, Raubitschek A, Colcher D, Liu A, Alvarez V, Mamelak AN: Imaging glioma extent with 131I-TM-601. J Nucl Med 2005, 46:580–586. Hsu AR, Hou LC, Veeravagu A, Greve JM, Vogel H, Tse V, Chen X: In vivo near-infrared fluorescence imaging of integrin alphavbeta3 in an orthotopic glioblastoma model. Mol Imaging Biol 2006, 8:315–323. Jin ZH, Josserand V, Foillard S, Boturyn D, Dumy P, Favrot MC, Coll JL: In vivo optical imaging of integrin alphaV-beta3 in mice using multivalent or monovalent cRGD targeting vectors. Mol Cancer 2007, 6:41. Ke S, Wen X, Gurfinkel M, Charnsangavej C, Wallace S, Sevick-Muraca EM, Li C: Near-infrared optical imaging of epidermal growth factor receptor in breast cancer xenografts. Cancer Res 2003, 63:7870–7875. Kennedy JC, Pottier RH: Endogenous protoporphyrin IX, a clinically useful photosensitizer for photodynamic therapy. J Photochem Photobiol B 1992, 14:275–292. Koyama Y, Barrett T, Hama Y, Ravizzini G, Choyke PL, Kobayashi H: In vivo molecular imaging to diagnose and subtype tumors through receptor-targeted optically labeled monoclonal antibodies. Neoplasia 2007a, 9:1021–1029. Koyama Y, Hama Y, Urano Y, Nguyen DM, Choyke PL, Kobayashi H: Spectral fluorescence molecular imaging of lung metastases targeting HER2/neu. Clin Cancer Res 2007b, 13:2936–2945. Kulbersh BD, Duncan RD, Magnuson JS, Skipper JB, Zinn K, Rosenthal EL: Sensitivity and specificity of fluorescent immunoguided neoplasm detection in head and neck cancer xenografts. Arch Otolaryngol Head Neck Surg 2007, 133:511–515. Leunig A, Betz CS, Mehlmann M, Stepp H, Arbogast S, Grevers G, Baumgartner R: Detection of squamous cell carcinoma of the oral cavity by imaging 5-aminolevulinic acid-induced protoporphyrin IX fluorescence. Laryngoscope 2000, 110:78–83. Liu J, Li J, Rosol TJ, Pan X, Voorhees JL: Biodegradable nanoparticles for targeted ultrasound imaging of breast cancer cells in vitro. Phys Med Biol 2007, 52:4739–4747. Mamelak AN, Rosenfeld S, Bucholz R, Raubitschek A, Nabors LB, Fiveash JB, Shen S, Khazaeli MB, Colcher D, Liu A, et al.: Phase I single-dose study of intracavitary-administered iodine131-TM-601 in adults with recurrent high-grade glioma. J Clin Oncol 2006, 24:3644–3650. Marti A, Lange N, van den Bergh H, Sedmera D, Jichlinski P, Kucera P: Optimisation of the formation and distribution of protoporphyrin IX in the urothelium: an in vitro approach. J Urol 1999, 162:546–552. Mayinger B, Neumann F, Kastner C, Degitz K, Hahn EG, Schwab D: Early detection of premalignant conditions in the colon by fluorescence endoscopy using local sensitization with hexaminolevulinate. Endoscopy 2008, 40:106–109. McCarron PA, Ma LW, Juzenas P, Lani V, Woolfson A, Zawislak AA, Moan J: Facilitated delivery of ALA to inaccessible regions via bioadhesive patch systems. J Environ Pathol Toxicol Oncol 2006, 25:389–402. McMahon J, O Brien CJ, Pathak I, Hamill R, McNeil E, Hammersley N, Gardiner S, Junor E: Influence of condition of surgical margins on local recurrence and disease-specific survival in oral and oropharyngeal cancer. Br J Oral Maxillofac Surg 2003, 41:224–231.

Optical Imaging of Primary Tumors

183

Mume E, Orlova A, Malmstrom PU, Lundqvist H, Sjoberg S, Tolmachev V: Radiobromination of humanized anti-HER2 monoclonal antibody trastuzumab using N-succinimidyl 5-bromo-3-pyridinecarboxylate, a potential label for immunoPET: Nucl Med Biol 2005, 32: 613–622. Newman JR, Gleysteen JP, Baranano CF, Bremser JR, Zhang W, Zinn KR, Rosenthal EL: Stereomicroscopic fluorescence imaging of head and neck cancer xenografts targeting CD147. Cancer Biol Ther 2008, 7:1063–1070. Overholt BF, Wang KK, Burdick JS, Lightdale CJ, Kimmey M, Nava HR, Sivak MV, Jr., Nishioka N, Barr H, Marcon N, et al.: Five-year efficacy and safety of photodynamic therapy with Photofrin in Barrett s high-grade dysplasia. Gastrointest Endosc 2007, 66:460–468. Patrice T, Olivier D, Bourre L: PDT in clinics: indications, results, and markets. J Environ Pathol Toxicol Oncol 2006, 25:467–485. Pech O, Gossner L, May A, Rabenstein T, Vieth M, Stolte M, Berres M, Ell C: Long-term results of photodynamic therapy with 5-aminolevulinic acid for superficial Barrett s cancer and highgrade intraepithelial neoplasia. Gastrointest Endosc 2005, 62:24–30. Prosst RL, Gahlen J, Schnuelle P, Post S, Willeke F: Fluorescence-guided minimally invasive parathyroidectomy: a novel surgical therapy for secondary hyperparathyroidism. Am J Kidney Dis 2006, 48:327–331. Prosst RL, Schroeter L, Gahlen J: Kinetics of intraoperative fluorescence diagnosis of parathyroid glands. Eur J Endocrinol 2004, 150:743–747. Prosst RL, Schroeter L, Gahlen J: Enhanced ALA-induced fluorescence in hyperparathyroidism. J Photochem Photobiol B 2005, 79:79–82. Prosst RL, Willeke F, Schroeter L, Post S, Gahlen J: Fluorescence-guided minimally invasive parathyroidectomy: a novel detection technique for parathyroid glands. Surg Endosc 2006, 20:1488–1492. Ravasz LA, Slootweg PJ, Hordijk GJ, Smit F, van der Tweel I: The status of the resection margin as a prognostic factor in the treatment of head and neck carcinoma. J Craniomaxillofac Surg 1991, 19:314–318. Regula J, MacRobert AJ, Gorchein A, Buonaccorsi GA, Thorpe SM, Spencer GM, Hatfield AR, Bown SG: Photosensitisation and photodynamic therapy of oesophageal, duodenal, and colorectal tumours using 5 aminolaevulinic acid induced protoporphyrin IX–a pilot study. Gut 1995, 36:67–75. Rivera F, Vega-Villegas ME, Lopez-Brea MF: Cetuximab, its clinical use and future perspectives. Anticancer Drugs 2008, 19:99–113. Rosenthal EL, Kulbersh BD, Duncan RD, Zhang W, Magnuson JS, Carroll WR, Zinn K: In vivo detection of head and neck cancer orthotopic xenografts by immunofluorescence. Laryngoscope 2006, 116:1636–1641. Rosenthal EL, Kulbersh BD, King T, Chaudhuri TR, Zinn KR: Use of fluorescent labeled antiepidermal growth factor receptor antibody to image head and neck squamous cell carcinoma xenografts. Mol Cancer Ther 2007, 6:1230–1238. Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A, McGuire WL: Human breast cancer: correlation of relapse and survival with amplification of the HER-2/neu oncogene. Science 1987, 235:177–182. Spaulding DC, Spaulding BO: Epidermal growth factor receptor expression and measurement in solid tumors. Semin Oncol 2002, 29:45–54. Speich R, Saesseli B, Hoffmann U, Neftel KA, Reichen J: Anaphylactoid reactions after indocyanine-green administration. Ann Intern Med 1988, 109:345–346. Stepp H, Beck T, Pongratz T, Meinel T, Kreth FW, Tonn J, Stummer W: ALA and malignant glioma: fluorescence-guided resection and photodynamic treatment. J Environ Pathol Toxicol Oncol 2007, 26:157–164. Stollman TH, Scheer MG, Leenders WP, Verrijp KC, Soede AC, Oyen WJ, Ruers TJ, Boerman OC: Specific imaging of VEGF-A expression with radiolabeled anti-VEGF monoclonal antibody. Int J Cancer 2008, 122:2310–2314.

184

J.R. Newman and E.L. Rosenthal

Stummer W, Pichlmeier U, Meinel T, Wiestler OD, Zanella F, Reulen HJ: Fluorescence-guided surgery with 5-aminolevulinic acid for resection of malignant glioma: a randomised controlled multicentre phase III trial. Lancet Oncol 2006, 7:392–401. Upadhyay R, Sheth RA, Weissleder R, Mahmood U: Quantitative real-time catheter-based fluorescence molecular imaging in mice. Radiology 2007, 245:523–531. Utsuki S, Oka H, Sato S, Shimizu S, Suzuki S, Tanizaki Y, Kondo K, Miyajima Y, Fujii K: Histological examination of false positive tissue resection using 5-aminolevulinic acidinduced fluorescence guidance. Neurol Med Chir (Tokyo) 2007, 47:210–213; discussion 213-214. Vallbohmer D, Zhang W, Gordon M, Yang DY, Yun J, Press OA, Rhodes KE, Sherrod AE, Iqbal S, Danenberg KD, et al.: Molecular determinants of cetuximab efficacy. J Clin Oncol 2005, 23:3536–3544. Veiseh M, Gabikian P, Bahrami SB, Veiseh O, Zhang M, Hackman RC, Ravanpay AC, Stroud MR, Kusuma Y, Hansen SJ, et al.: Tumor paint: a chlorotoxin:Cy5.5 bioconjugate for intraoperative visualization of cancer foci. Cancer Res 2007, 67:6882–6888. Withrow KP, Gleysteen JP, Safavy A, Skipper J, Desmond RA, Zinn K, Rosenthal EL: Assessment of indocyanine green-labeled cetuximab to detect xenografted head and neck cancer cell lines. Otolaryngol Head Neck Surg 2007, 137:729–734. Withrow KP, Newman JR, Skipper JB, Gleysteen JP, Magnuson JS, Zinn K, Rosenthal EL: Assessment of bevacizumab conjugated to Cy5.5 for detection of head and neck cancer xenografts. Technol Cancer Res Treat 2008, 7:61–66. Witjes JA, Douglass J: The role of hexaminolevulinate fluorescence cystoscopy in bladder cancer. Nat Clin Pract Urol 2007, 4:542–549. Woolgar JA, Triantafyllou A: A histopathological appraisal of surgical margins in oral and oropharyngeal cancer resection specimens. Oral Oncol 2005, 41:1034–1043. Wu Y, Cai W, Chen X: Near-infrared fluorescence imaging of tumor integrin alpha v beta 3 expression with Cy7-labeled RGD multimers. Mol Imaging Biol 2006, 8:226–236. Wu SM, Ren QG, Zhou MO, Wei Y, Chen JY: Photodynamic effects of 5-aminolevulinic acid and its hexylester on several cell lines. Sheng Wu Hua Xue Yu Sheng Wu Wu Li Xue Bao (Shanghai) 2003, 35:655–660.

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques E.M. Sevick-Muraca

Introduction In 2007, approximately 1.5 million Americans were diagnosed with invasive carcinomas (American Cancer Society Cancer Facts & Figures, 2007). For the majority of these patients, their disease was staged using the TNM disease classification approved by the American Joint Committee on Cancer whereby Tumor extent, regional lymph Node involvement, and presence or absence of distant Metastases are determined. Anatomical imaging with computed tomography (CT), magnetic resonance (MR), or ultrasound (US) typically provides a measurement of the size and extent of the locoregional disease while distant metastases in advanced disease are detected through whole-body nuclear bone scanning or positron emission tomography (PET). Yet, even though lymph node (LN) status is recognized as the single most important prognostic factor for the majority of cancer patients (Greene et al., 2002), there is a paucity of techniques to diagnostically image the lymphatic vascular compartment to assess disease. Indeed, there is ample evidence to suggest that the lymphatic system provides the first route of cancer dissemination throughout the body. For example, Su et al. (2006) have shown that the metastatic potential of colon, head and neck, breast, prostate, and lung cancers is positively correlated with vascular endothelial growth factor receptor-3 and vascular endothelial growth factor-C, which constitute the receptor and ligand system responsible for lymphangiogenesis. The implications of this and other reports are that primary tumors “grow” their own lymphatic or “highway” systems for cancer cell dissemination throughout the body. Another implication may be that lymphangiogenesis is a potentially important therapeutic/diagnostic target for prevention and/or management of metastatic cancer (Harrell et al., 2007; Bono et al., 2004; Skobe et al., 2001; Hirakawa et al., 2007). E.M. Sevick-Muraca (B) The University of Texas Health Science Center, The Brown Foundation for Molecular Medicine, Center of Molecular Imaging, Houston, Texas 77030, USA e-mail: [email protected]

E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_10, 

185

186

E.M. Sevick-Muraca

Limitations of Conventional Techniques Current methods to non-invasively image the cancer status of nodal basins require the LN to be greater than 1 cm in size, therefore TNM staging commonly relies on surgical resection of LN basins that drain tumor sites with ex vivo pathological evaluation of resected tissues to determine cancer involvement. LN biopsies are routinely performed in breast, melanoma, head and neck, colon, lung, and other cancers. However, several problems exist with the current surgical practices of LN staging. • Current status of staging occult LN disease: The “sampling problem.” Assessment of nodal status is performed by surgical “sampling” of discrete LNs to provide an indication of the cancer status of the entire tumor draining lymphatic basin. As a result, a propensity exists for inaccurate under-staging of nodal involvement (Kingsmore et al., 2005). In breast cancers, for example, studies have shown that the greater the number of lymph nodes resected, the greater the chance for detecting diseased LNs (Bland et al., 1999). Although an increased number of resected LNs translates into a decreased chance for misdiagnosing node-positive disease and for undertreatment of cancer patients, it also translates into an increased risk for development of lymphedema (Williams et al., 2005) (see below). As a result, the concept of minimal disruption of the lymphatics via the “sentinel lymph node biopsy” (SLNB) procedure has become standard practice in breast cancer staging and is increasingly employed in other cancers. In SLNB, the “sentinel” or first tumor draining lymph node is resected and if found to be cancer positive, then additional LNs within the draining basin are resected to provide more accurate staging. • Lack of intraoperative, molecular pathology or the accuracy problem: Once LNs are resected, accurate pathological examination to determine the presence of cancer is generally conducted. Yet, intraoperative molecular pathology for accurate staging is not always possible. In many cases, SLNB requires rapid surgical pathological evaluation of resected sentinel LNs at the time of the surgical procedure for the surgeon to determine whether additional LNs need to be resected. Unfortunately, the level of ex vivo, molecular pathological examination required for accuracy of nodal staging requires too lengthy a time to perform SLNB and additional LN dissection within the same surgical setting. For example, in breast cancer SLNB, cytological touch preparations are typically performed to provide staging information for rapid non-molecular assessment of the SLN so that additional LN dissections can take place during the same surgical procedure. In retrospective studies, reverse transcription polymerase chain reaction (RT-PCR) or immunohsitochemistry (IHC) analyses of resected nodal tissues have detected micrometastases in breast cancer patients who have gone on to experience metastatic relapse after being diagnosed as sentinel node negative using standard hematoxylin and eosin (H&E) staining (Querzoli, et al., 2006; Cummings et al., 2002). Whereas developments in intraoperative molecular pathology using new technologies seek to reduce analysis time to a reported

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques

187

40 min (Hughes et al., 2006; Blumencranz et al., 2007), the need for immediate molecular analysis to determine if cancer cells are present in LNs remains unmet. As a result, the lack of intraoperative molecular pathology combined with the minimal number of LNs resected in SLB contribute to (i) a reduced accuracy of nodal staging that biases for undertreatment (Tjan-Heijnen et al., 2001) and (ii) controversial recommendations for more invasive LN dissections independent of SLN status (Zavagno et al., 2005). • Lymphedema as a complication of lymph node dissections: Acquired lymphedema is a surgical complication that can present itself after extensive lymph node dissection for TNM staging. Acquired lymphedema is an incurable disease in which excessive edema occurs in the limbs normally drained by the lymphatic basin subjected to nodal dissection. It has been estimated that as many as 62% of all breast cancer survivors who undergo axillary lymph node dissection encounter the debilitating form of breast cancer-related lymphedema (BRCL) with initial symptoms appearing weeks to years after the surgical trauma (Armer, 2005; Ridner, 2006; Mansel et al., 2006). The etiology and manifestation of lymphedema remain largely unknown due in part to our lack of knowledge of lymphatic repair mechanisms (de Vries et al., 2006). No method exists to predict if patients will develop lymphedema after nodal dissection.

Sentinel Lymph Node Assessment Most of our information regarding lymphatic system architecture arises from human cadaver studies, and both lymphvascular architecture and drainage patterns are highly variable among individuals (Foeldi and Foeldi, 2006). Consequently, tumor drainage to sentinel LNs must be detected by one or a combination of two methods that employ radiotracers or isosulfan blue: • After intradermal (ID) or subcutaneous administration of 99m Tc sulfur colloid near a tumor, gamma camera imaging is used to detect drainage patterns to the sentinel LNs, and a hand-held gamma probe is currently used intraoperatively to identify draining LNs for resection. • Following intraoperative peritumoral, ID, or subcutaneous injection of isosulfan blue dye, lymph channels and draining LNs are visually identified for resection. Despite these localization techniques that successfully identify sentinel LNs, not all LNs are surgically accessible for resection and evaluation. Even though inner upper and lower quadrant breast tumors drain supraclavicular and internal mammary nodes and even though nodal disease in these basins are associated with the poorest prognoses (Yao et al., 2007; Singletary et al., 2002), surgical morbidity prevents their resection and nodal staging is not performed on these most pertinent LN drainage basins. The possible tissue and nerve damage involved with resection of cubital LNs in the arm and popliteal LNs in the legs of melanoma patients also often

188

E.M. Sevick-Muraca

obviates their dissection SLNB. Similarly in prostate and bladder cancer, virtually no nodal staging is performed due to the inaccessibility and morbidity of surgically resecting prostate-draining lymph nodes in the pelvis simply for diagnostic purposes. Controversies in nodal staging. There are two controversies in the literature associated with nodal staging: (1) Is aggressive lymph node dissection without diagnostic confirmation of disease beneficial to cancer patients? and (2) What level of molecular pathology provides an accurate indicator of nodal disease? • Dissection of non-occult or “macroscopic and clinically apparent” lymph node metastases are typically performed without pathological examination, and the therapeutic advantage of surgery is not debated. Depending on tumor type, LN dissections are performed therapeutically on a suspicion of occult lymph node metastases. For example, in locally advanced breast cancer patients, complete ALND after adjuvant chemotherapy is typically performed, even though nodal staging is not performed to suggest that microscopic or macroscopic LN disease exists after a regimen of adjuvant chemotherapy (Shen et al., 2007). In prostate cancer, some surgeons perform aggressive, extended pelvic lymph node dissections (PLNDs) at the time of prostatectomy to remove possible “occult” metastases or as a prophylactic to remove dissemination routes for extracapsular remaining disease at prostatectomy (Lerner et al., 1993; Stein et al., 2001; Weingartner et al., 1996; Allaf et al., 2004) – similarly to the previous practice of complete ALND in breast cancer patients that has been replaced by SLNB. Similar clinical care is reported for bladder cancer. Badgwell and colleagues (Badgwell et al., 2007) report controversial, but favorable improvement in survival rates of patients with sentinel node-positive melanoma in the lower limbs when aggressive PLND is performed, despite the lack of evidence for occult or non-occult LN disease in the LN basins of these patients. In sentinel node-positive melanoma patients, current National Comprehensive Cancer Network practice guidelines recommend therapeutic regional LN dissections (NCCN Guideline, 2006). • There is also controversy over what constitutes clinically significant nodal disease. Under current TNM guidelines for breast cancer, resected LNs are typically deemed cancer positive from H&E micrometastases (<2 cm and >0.2 mm) whether or not stromal changes indicative of malignancy are present (Singletary et al., 2002). While “nanometastases” detected by molecular IHC (with anti-cytokeratin in the case of epithelial cancers) have been shown in some studies to be clinically significant predictors of metastatic relapse in some patients (Querzoli, et al., 2006), it remains controversial as to whether the benefits of systemic chemotherapy (or LN dissections) due to the observation of isolated tumor cells in the lymphatic space warrant the risks associated with the treatment itself. Nonetheless, nodes deemed cancer positive by IHC, yet are histologically negative for micrometastases by H&E are designated pN0(i+) but classified as negative in the current TNM staging system. While pathologists, surgeons, and

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques

189

oncologists currently debate the size of micrometastases that constitutes clinically relevant LN disease, the presence of isolated tumor cells and other markers of malignancy, such as stromal reorganization, may provide a better indicator of minimal, yet clinically relevant, occult LN disease. The opportunities of near-infrared (NIR) fluorescence imaging with agents that molecularly target cancer cells and stromal reorganization within the lymphatic space could enable a non-invasive approach to perform nodal staging in many cancers. With a non-invasive method to monitor cancer-positive lymph nodes and provide rapid intraoperative diagnosis of tissues, the difficulties and controversies regarding TNM staging could be resolved.

Practice and Research of Technique to Image Nodal Disease To differentially image diseased LNs, contrast agents must be employed. As described pictorially in Fig. 1, four distinct routes exist to introduce molecular imaging agents into the lymphatic space: (i) ID for delivery into the lymphatic vessels via the lymph plexus in the dermis and subcutaneously into the interstitium for permeation into lymphatic capillaries and vessels; (ii) direct introduction into the lymphatic space by lymphatic vessel cannulation; and (iii) intravenously (IV) for

Fig. 1 Injection routes to introduce contrast agents into the lymphatics are through interstitial (intradermal or subcutaneous) administration, direct administration into a cannulated lymphatic vessel, or intravenous injection. Reproduced from Sharma et al. (2008)

190

E.M. Sevick-Muraca

transit from the vascular space to the interstitium for lymphatic permeation or for direct deposition via the microcirculation within lymph nodes (Sharma et al., 2008). Particles less than ∼11 nm in diameter can pool in LNs via blood circulation, but can also diffuse into lymphatic vessels via gap junctions between lymph endothelial cells (LECs) under a hydrostatic pressure gradient (Blacker et al., 2007). Particles as large as 100 nm in diameter can extravasate into the interstitial space where they are phagocytized by macrophages or they can be taken up through the lymph plexus and convected through lymph flow into lymphatic vessels and draining lymph nodes. Particles >100 nm typically remain trapped in the interstitium (Moghimi and RajabiSiahboomi, 1996).

Lymphoscintigraphy and Isosulfan Blue Dye This technique is commonly used in clinical application to identify sentinel lymph nodes. As previously described, current LN mapping procedures use lymphoscintigraphy with 99m Tc sulfur colloid (∼100 nm diameter) and/or visual intraoperative identification of lymphatic drainage with isosulfan blue. Both 99m Tc and isosulfan blue are administered intraparenchymally or ID for uptake and convection into the lymphatics to detect SLNs. Unfortunately, neither agent can differentiate between diseased and normal LNs, and macroscopic LN disease can obstruct lymph flow that prevents LN mapping.

Superparamagnetic MR Contrast Agents Superparamagnetic MR contrast agents have also been proposed as lymphotropic agents. These agents consist of 4–6 nm diameter particles with a magnetic core surrounded by a layer of dextran or starch derivative. These particles act to shorten proton relaxation times, giving a negative signal that is distinct from non-contrasted tissues. Ferumoxtran, a suspension of dextran-coated ultrasmall iron oxide particles (USIOP) <50 nm in diameter, has been injected IV (Bellin et al., 1998) and detects LN disease based on size. The negative contrast provided by accumulated USIOPs, the large dose required for systemic administration, and their differentiation of diseased LN based on size rather than the occult presence of cancer cells makes them unlikely contrast agent candidates for tumor nodal staging.

US, MR, CT Lymphangiography Techniques US, MR, CT lymphangiography techniques that employ iodinated or gadolinium complexes (Partsch et al., 1984, 1983; Suga et al., 2003; Suga et al., 2006; Rabin et al., 2006; Lohrmann et al., 2006a, 2007, 2006b; Matsushima et al., 2007; Ruehm

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques

191

et al., 2001) or microbubbles (Goldberg et al., 2004; Hauff et al., 2004) administered ID or through cannulated lymph vessels to detect enlarged LNs are likewise unlikely lymph imaging agents. Besides the difficulty of lymph vessel cannulation, administration of large milliliter volumes of agent required to map lymphatic drainage is prohibitive in the ID space (see below), technically difficult and painful for the patient. Even if conjugated to molecular targeting entities specific for cancer, the femto- and pico-molar sensitivities required for molecular imaging of diseased lymph nodes are unattainable in US-, MR-, or CT-based imaging, which at best provide an assessment of enlarged LNs. Molecular imaging approaches to target in vivo extracellular disease markers specific to cancer depend on exquisite sensitivities to detect exogenous contrast agents. Nuclear medicine offers the only clinically established imaging approaches capable of femto- to picomolar sensitivities required for molecular imaging of cancerpositive lymph nodes. Unfortunately, 18 FDG PET via systemic administration offers low specificity for occult LN disease. On the other hand, radioimmunoscintigraphy and radiolabeled small molecular/peptide agent may be combined with PET, and to a lesser extent, single photon emission computed tomography (SPECT), for non-invasive nodal staging.

Radioimmunoscintigraphy Techniques Radioimmunoscintigraphy techniques using ID or direct lymphatic delivery of radiolabeled antibodies was first proposed by Weinstein and colleagues for nodal staging of cancer in the early 1980 s (Weinstein et al., 1982; Weinstein et al., 1983). Advantages include (i) high lymph node uptake of imaging agent that enables significantly lower doses of labeled antibodies than that required for systemic IV administration, (ii) rapid lymph node uptake enabling minimal time between ID antibody administration and imaging, and (iii) increased TBR from reduced non-specific binding that would otherwise occur with IV administration. This technique depends on clearance of labeled antibodies from normal LNs and retention in diseased LNs, enabling noninvasive staging. Nuclear applications of intralymphatic and ID administration have since been explored by several investigators, mostly using radiolabeled monoclonal antibodies in intact or fragmented forms to detect tumor draining LNs in a spectrum of cancer patients ranging from melanoma, T-cell lymphoma, breast, and prostate cancers (Deland et al., 1980 ; Tjandra et al., 1989 ; Kairemo, 1990 ; Keenan et al., 1987a; Keenan et al., 1987b ; Abdelnabi et al., 1990 ; Lotze et al., 1986 ; Nelp et al., 1987 ; Engelstad et al., 1986). The results are mixed owing probably to a number of factors, including the following: • Poor resolution. Radioimmunoscintigraphy results to date employ gamma emitters for gamma scintigraphy or SPECT. Scintigrams are associated with notoriously low resolution (Fig. 2), while tomographic SPECT provides a low photon count rate owing to its pinhole aperture configuration.

192

E.M. Sevick-Muraca

Fig. 2 Example of lymphoscintigraphy showing channels and SLNs demarked by non-specific 99mc Tc

• Radiotracers have finite half-lives and reduced photon counts with increased time after administration. As a result, the long clearance times of large antibodies from normal LNs may not enable differentiation between diseased and normal LNs. • Immune reactions. Although small molecules may offer a better opportunity for rapid clearance from normal LNs and retention in diseased LNs, results to date have focused upon large antibodies that are not fully humanized and thereby can elicit an immune reaction, which itself impacts clearance through normal LNs.

NIR Fluorescence Imaging In contrast to radionuclides used in nuclear medicine, fluorophores have no physical finite half-life and can be reactivated with propagating light to produce multiple emissions. If one considers a fluorophore with a nanosecond fluorescent lifetime (i.e., the mean time between absorption of an excitation photon and subsequent radiative relaxation to the ground state) and a quantum efficiency of one-tenth, the theoretical limit of the number of imaging photon events per fluorophore could be 100,000,000 photons per second. When one compares a fluorescent decay processes to the decay of a radionuclide which at most emits one photon event ever, the advantages of NIR fluorescence imaging becomes obvious. Three caveats challenge the realization of a nearly billionth-fold greater sensitivity of NIR fluorescence over “gold-standard” nuclear imaging: autofluorescence, agent photostability, penetration depth, and availability of instrumentation. Autofluorescence: To avoid confounding effects of autofluorescence, successful fluorescent imaging agents must be excited and emit in the NIR wavelength range, typically ≥750 nm (Adams and Alitalo, 2007; Adams et al., 2007).

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques

193

Agent photostability: Only one NIR fluorophore is available for human use for indications other than fluorescent imaging. The dye indocyanine green (IC-Green, Akorn Pharmaceuticals, Buffalo Grove, IL), is approved for IV administration, at 5 mg/kg up to a total maximum of 25 mg, for use in hepatic clearance, cardiovascular function, and in assessing retinopathy. However, IC-Green does not have a functional group for conjugation to molecularly targeting entities such as antibodies or their fragments, aptamers, or peptides. Instead, IC-Green associates with albumin in the blood and as described below provides a unique opportunity to translate NIR fluorescence instrumentation for lymphatic imaging. Recently NIR excitable fluorophores with functional groups for conjugation have become commercially available, expanding opportunities for NIR imaging, but none have been approved for human use, and their stability in conjugates remains to be tested. Penetration depth: Recently, we demonstrated imaging with dual-labeled RGD peptides against αv β3 integrin expression and dual-labeled trastuzumab R , Genentech, Inc) against human epidermal growth factor (Herceptin receptor-2 (HER2) expression in subcutaneous xenograft models to show comparable optical planar imaging, scintigraphy, and tomography, with enhanced SNR in NIR fluorescence compared to nuclear imaging (Houston et al., 2005; Sampath et al., 2007). Because subcutaneous xenograft models offer shallow penetration depths, the results may be expected and similar to what might be encountered for intraoperative guidance. Below, we demonstrate the ability to collect NIR fluorescence from deep-tissue regions in humans. Success is dependent on (i) eliminating autofluorescence by choice of NIR rather than red-excitable fluorophores, (ii) reducing the overwhelming excitation light leakage through filters designed to pass only the weak fluorescence signals and reject backscattered excitation light (Hwang et al., 2005; Hwang et al., 2006); and (iii) employing an integrating camera similar in design to nuclear medicine systems, but outfitted with time-of-flight capabilities to enable tomography (Sevick-Muraca and Paithankar, 1999). Availability of instrumentation: The translation of molecular imaging agents for nodal staging requires use of a device with demonstrated capability to detect investigational imaging agents. Unfortunately until recently, NIR fluorophores were unavailable, and the large proportion of standard laboratory instrumentation is not capable of measuring fluorescence in the NIR wavelength range. Most importantly, no instrumentation is commercially available to sensitively measure NIR fluorescence from microdose administration. Recently, we have demonstrated non-invasive NIR fluorescence imaging in humans using microdose administration of NIR fluorescent agents to justify the use of optical molecular imaging agents in humans (Sevick-Muraca et al., 2008; SevickMuraca and Rasmussen, 2008) and summarize our results herein. The development of “first-in-human” NIR optical imaging agents for molecularly based nodal staging requires the demonstration of an imaging device capable of collecting images

194

E.M. Sevick-Muraca

following microdose administration of NIR fluorescent agents in humans. The Food and Drug Administration defines “microdose” as 1/100th of the dose that yields pharmacological dose for imaging agents, 100 μg for a peptide or 30 nanomoles for protein-based imaging agents. Guidance for Industry, Investigators, and Reviewers: Exploratory IND Studies, CDER, January 2006.

Approach and Methods In our studies to demonstrate the feasibility of NIR fluorescence imaging, we capitalized upon the history of safe use of non-specific NIR fluorescent dye, indocyanine green (or IC-Green, Akorn Inc., Buffalo Grove, IL), in order to (i) demonstrate NIR lymph imaging in breast cancer patients who undergo standard of care SLN mapping procedures using 99m technetium sulfur colloid and/or isosulfan dye to intraoperatively locate LNs within the breast and axilla and (ii) assess lymphatic function in normal subjects and in those with lymphedema. In a dose escalation study, we demonstrated non-invasive NIR optical imaging of lymph function and architecture following as little as 10 μg ID or subcutaneous delivery of IC-Green. We hypothesize that if human fluorescence images can be collected with microdose administration of IC-Green, then the device and approach can be used as a platform for future development of molecularly targeting, protein and peptide-based NIR fluorescent imaging agents. Most importantly, upon demonstrating NIR fluorescence imaging following microdose of the comparatively dim IC-Green, we demonstrate the ability to perform NIR fluorescence imaging with “first-in-humans” molecular imaging agents for nodal staging. Non-invasive imaging of the fluorescent IC-Green was performed with an intensified charged coupled device (CCD) (Reynolds et al., 1997; Sevick-Muraca et al., 2008). Briefly, the device has three principal components: (i) a NIR sensitive image intensifier; (ii) a 16 bit dynamic range, frame transfer charge coupled device (CCD) camera; and (iii) a 80 mW 785 nm laser diode to provide the excitation light for activating the IC-Green. The laser diode beam was expanded using a plano-convex lens and a holographic optical diffuser such that approximately 0.02 m2 of the breast and axilla surfaces were illuminated with surface fluence less than 1.9 mW/cm2 . A 785-nm holographic notch band rejection filter and an 830-nm image quality bandpass filter were placed before the 28 mm lens to selectively reject the excitation light and pass the emitted 830 nm fluorescence (Hwang et al., 2005). A collection of 50–400 images with 512 × 512 resolution and 200–800 ms exposure time were acquired enabling near-real-time visualization of IC-Green trafficking. For image registration, white light images were sometimes acquired with a neutral density filter as the camera captured focused images of the skin surface. In other cases the excitation light leakage through the filters provided sufficient images for registration. Subjects were imaged immediately following intradermal injections of IC-Green. The protocols used for this Phase I feasibility study were conducted under combinational investigational new drug applications for off-label use of IC-Green. The HIPAA compliant studies were approved by the Institutional Review Board (IRB) as

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques

195

well as hospital-affiliated clinical sites. Details of the studies can be found elsewhere (Sevick-Muraca et al., 2008; Sevick-Muraca and Rasmussen, 2008; Rasmussen et al. xxxx).

Comparison of Scintigraphy and NIR Fluorescence Imaging of Lymph Nodes in Breast Cancer Patients Figure 3(a) and (b) depicts the white light photograph and corresponding fluorescent image depicting an example of lymph trafficking after four periareolar intradermal injections of 20 μg of IC-Green in 100 μL in the right breast of a breast cancer patient (Sevick-Muraca et al., 2008). Figure 3(c) illustrates the corresponding lymphoscintigraphy in the same subject who was also injected with radiotracer at the

Fig. 3 White light view (a), NIR fluorescence view (b), and frontal scintigram (c) of the right breast of a breast cancer patient who received a total of 100 μg IC-Green intradermally surrounding the areola. The stick diagram illustrates the breast view in (a) and (b). Lymphatic propulsion occurred in this subject. Reproduced from Sevick-Muraca and Paithankar (1999)

196

E.M. Sevick-Muraca

same time as IC-Green. In Fig. 3(b), two crossing lymph channels are shown to propel IC-Green “packets” into the axilla, at a velocity of ∼0.2 cm/s and at time intervals of ∼20 s. Three fluorescent, negative LNs were resected 5 h after agent administration, all of which were radioactive, but none were found to be blue from the isosulfan dye used to locate the tumor draining nodes. Not all subjects exhibited lymph propulsion as illustrated in the case described above. Figure 4(a) and (b) shows the white light photograph and corresponding fluorescent image which shows two uniformly fluorescent lymph channels draining

Fig. 4 White light view (a), NIR fluorescence view (b), and frontal scintigram (c) of the left breast of breast cancer patient who received a total of 100 μg IC-Green intradermally surrounding the areola. The stick diagram illustrates the breast view in (a) and (b). The two lymphatic channels present on the fluorescence image were also seen on the scintigram. There was no lymphatic propulsion seen in this subject. Reproduced from Sevick-Muraca and Paithankar (1999)

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques

197

Fig. 5 An example white light (left) and fluorescent (right) image of a resected fat containing a sentinel node from a breast cancer patient. The images show co-localization of the blue and NIR fluorescent dyes. Reproduced from Sevick-Muraca and Paithankar (1999)

to separate locations within the axilla in the left breast of a breast cancer patient following four intradermal administration of 25 μg in 100 μL. The scintigram confirms the drainage to two different LNs within the axilla. One radioactive, blue, and fluorescent, negative LN was resected 25 h after agent administration. These studies show the ability to non-invasively detect tumor-draining lymph nodes from NIR fluorescence imaging. Using the intensified camera system, we were able to non-invasively detect fluorescence in tissues after as little as 10 μg administration of IC-Green (Sevick-Muraca et al., 2008). Figure 5 provides a typical example of a LN immediately after resection that is both blue (from the isosulfan blue) and fluorescent (from the IC-Green). The overlay of white light and fluorescent images demonstrate the co-localization of the blue and NIR fluorescent dyes and demonstrate that the absorption of excitation and fluorescent light does not impair the collection of fluorescence from the sample. These results show the ability for rapid ex vivo or intraoperative in vivo molecular pathology to potentially improve TNM staging with NIR-based molecular imaging agents.

Functional Lymph Imaging in Health and Disease With as little as 100 μg of IC-Green, we have been able to quantitatively image propulsive lymph velocity in major lymph bundles and deep lymph vessels that drain into lymph node basins. In the hands, 100 μg of ICG was administered interdigitally in four portions using conventional 27 g needles. Almost immediately after administration, trafficking of lymph fluid “packets” from the site of injections through the radial and ulnar lymphatic bundles could be seen with fluorescent imaging.

198

E.M. Sevick-Muraca

In the foot, trafficking from ID injections of 25 μg IC-Green were made in the first two interdigital spaces, heel, posterior medial calf, and anterior medial thigh resulted in visualization of propulsive lymph flow within the ventromedial lymph bundle of the legs. Deep lymph vessels such as the dorsolateral bundle feeding the popliteal lymph nodes as well as the deep femoral vessels were imaged, but images were particularly diffuse owing to tissue scattering. Nonetheless, in either arm or leg, lymph propulsion occurred at apparent velocities ∼1 cm/s and appearing at intervals ranging from 13 to 120 s. No apparent correlation was found between blood pressure or heart rate and frequencies or velocities of propelled lymph “packets.” From dynamic imaging of normal subjects, we found propelled fluorescent lymph frequently accumulated in localized regions of the lymphatics, before being released to the next localized region for transit to the axilla (in the arms) or to the inguinal region (in the legs). We hypothesize that these localized lymphatic regions represent lymphangions (depicted at apparent mean intervals of 2.8 cm in the foot as illustrated in Fig. 6). In normal subjects, draining lymph nodes in the limbs and the major draining lymph node basin in the axilla and groin area were also imaged. In addition to normal subjects, we have also imaged subjects with acquired lymphedema, particularly breast cancer related lymphedema. Of particular interest is the asymptomatic left arm of a lymphedema subject with a history of bilateral mastectomy. The subject underwent radical mastectomy of the right breast and later developed swelling of the right arm. Nine years after her right mastectomy, the subject underwent modified radical mastectomy of left breast but remained free of symptoms of lymphedema at the time of imaging in the left arm (Fig. 7). Although the edematous right arm was observed to have diffused IC-Green and few lymphatics vessels to our surprise, the left asymptomatic arm exhibited sites with apparent hyperplasia of thin, small cutaneous lymphatics, and bidirectional flow in a lymphatic trunk (Fig. 8) – an observation not seen in normal subjects and hypothesized to result from genetic mutations in the transcriptional factor FOXC2 (Fang et al., 2000; Petrova et al., 2004; Mellor et al., 2007). These results suggest that the opportunity to rapidly collect images from photon-plentiful NIR fluorescence imaging enable new methods to diagnose disease based on lymph architecture and function.

Radio- and Optical-Immunoscintigraphy for Cancer Nodal Staging Given the ability to detect lymph nodes following trace administration of NIR fluorophore and modest success of radioimmunoscintigraphy, we previously combined both nuclear and optical agents to label an antibody targeting HER2 which is often overexpressed in metastatic breast cancer. Recently we have synthesized a dual-labeled antibody-(111 In-DTPA)n -trastuzumab-(IRDye800)m and demonstrated imaging specificity for HER2 in vitro and in vivo within a preclinical model (Engelstad et al., 1986; Houston et al., 2005).

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques

199

Fig. 6 NIR fluorescence images as a function of time of a forearm after intradermal administration of 100 mg of ICG was administered interdigitally

Figure 9 shows a series of whole-body fluorescence (A), planar scintigraphy (B), and SPECT/CT (C) images from athymic nude mice with HER2 overexpressing SKBr3 human breast cancer xenografts after i.v. injection of (111 In-DTPA)n trastuzumab-(IRDye800)m . These and other results on dual labeling show that the target to background ratio computed from scintigraphy and fluorescence images is comparable, consistent with the single and stable, dual-labeled agent. However,

200

E.M. Sevick-Muraca

Fig. 7 Fluorescence images of nodal regions in normal subjects: (a) three median lymphatic bundles that pool the fluorescent dye into three lymph nodes in the axilla; (b) afferent and efferent lymphatic vessels feeding and draining the fluorescent cubital lymph node in the medial forearm and elbow; (c) the popliteal lymph node in the back of the right knee; and (d) fluorescent signals demarking up to six superficial inguinal lymph nodes

the signal-to-noise ratio of the NIR fluorescence is significantly greater than that from the scintigraphy, consistent with the increased photon count rate afforded by NIR fluorescence imaging. To assess applicability for nodal staging, we injected the dual-labeled agent intradermally for transit to the axillary lymph nodes of tumorfree mice. In order for the agent to detect lymph metastasis, clearance from normal lymph nodes needs to occur well before the half-life of the radiotracer employed. Figure 10 shows the rate of clearance of (In-DTPA)n-trastuzumab-(IRDye800)m in the tumor-free axillary lymph nodes of mice averaged over time in five animals. We

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques

201

Fig. 8 Fluorescent images of (b) symptomatic and (c) asymptomatic arms of a breast cancer patient with unilateral lymphedema and with a history of bilateral mastectomy. In the symptomatic right arm, IC-Green appears to be diffused within the arm with a few lymphatic vessels apparent on the medial side of elbow and upper arm and that feed a network of hyperplastic cutaneous vessels in the upper arm. The asymptomatic arm depicts a site of increased number of small lymphatics and a discrete vessel in the medial side of the forearm Reproduced from Rasmussen et al., 2009

found variability in the uptake of the imaging agent into the axillary node among the mice, which is represented by a large error bar at t = 0 h (Fig. 10c), but all mice attained maximum fluorescence in nodes within t = 1 h. Half the labeled antibody cleared from the lymph nodes within 10–12 h, suggesting that radioisotopes with longer half-lives might not provide sufficient signal. This result is consistent with previous reports using radiolabeled antibodies and suggests the advantages of using an NIR fluorescent agent for imaging cancer-positive lymph nodes rather than a radiotracer. Nonetheless, using a smaller targeting moiety, such as a peptide or affibody rather than an antibody, could improve clearance rates from normal lymph nodes and possibly improve specificity for detecting cancer-positive lymph nodes.

Conclusion and Summary Given the opportunities (1) of NIR fluorescence for image-guided resection of cancer-positive lymph nodes and rapid molecular pathology of resected LNs and (2) of nuclear imaging for certain, deep-tissue, non-invasively imaging, the combination of nuclear and optical imaging for nodal staging of several cancers presurgically, intraoperatively, and within the surgical pathology suite may have the most impact on accuracy of TNM. The opportunity to detect NIR fluorescence following microdose administration of fluorophore bodes well for the translation of NIR fluorescence-based molecular imaging agents developed for nodal staging of cancers.

202

E.M. Sevick-Muraca

Fig. 9 (a) and (b) are representative white light and whole-body fluorescence images of an athymic nude mouse inoculated with SKBr3-luc xenografts, taken 48 h after i.v. administration of (111 In-DTPA)n -trastuzumab-(IRDye800)m . High agent uptake is visible in the left flank which is consistent with tumor region. Similar uptake is observed in nuclear imaging obtained from 111 In as represented by planar scintigraphy (c) and SPECT/CT (d). Reproduced from Houston et al. (2005)

Nomenclature ALND BCRL CCD CT HER2 ID IV IHC H&E

axillary lymph node dissection breast cancer-related lymphedema charge coupled device computed tomography human epidermal growth factor receptor-2 intradermal intravenous immunohistochemistry hematoxylin and eosin

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques

203

Fig. 10 Representative white light (a) and optical imaging (b) of accumulation and clearance of (In-DTPA)n -trastuzumab-(IRDye800)m in the axillary node after administration into the forepaw foot pad. Quantitative analysis of lymph node trafficking in five mice is represented in (c). Reproduced from Sampath et al. (2008)

LN MR NIR PET PLND RT-PCR SPECT SLNB SLN USIPO US

lymph nodes magnetic resonance near-infrared positron emission tomography pelvic lymph node dissection reverse transcription polymerase chain reaction single photon emission computed tomography lymph node biopsy sentinel lymph node ultrasmall iron oxide particles and ultrasound

Acknowledgements This work was supported in part by R01 CA112679 and R01 CA136404.

204

E.M. Sevick-Muraca

References Abdelnabi, H.H., et al., Clinical-experience with intra lymphatic administration of in-111 labeled monoclonal-antibody pay-276 for the detection of pelvic nodal metastases in prostaticcarcinoma. European Journal of Nuclear Medicine, 1990. 16(3): 149–156. Adams, R.H. and K. Alitalo, Molecular regulation of angiogenesis and lymphangiogenesis. Nature Reviews. Molecular Cell Biology, 2007. 8(6): 464–478. Adams, K.E., et al., Comparison of visible and near-infrared wavelength-excitable fluorescent dyes for molecular imaging of cancer. Journal of Biomedical Optics, 2007. 12(2): 020417–1–024017–9. Allaf, M.E., et al., Anatomical extent of lymph node dissection: Impact on men with clinically localized prostate cancer. Journal of Urology, 2004. 172(5): 1840–1844. American Cancer Society Cancer Facts & Figures. 2007. Armer, J.M., The problem of post-breast cancer lymphedema: impact and measurement issues. Cancer Investigation, 2005. 23(1): 76–83. Badgwell, B., et al., Pelvic lymph node dissection is beneficial in subsets of patients with nodepositive melanoma. Annals of Surgical Oncology, 2007. 14(10): 2867–2875. Bellin, M.F., et al., Lymph node metastases: safety and effectiveness of MR imaging with ultrasmall superparamagnetic iron oxide particles – initial clinical experience. Radiology, 1998. 207(3): 799–808. Blacker, H.A., et al., How regenerating lymphatics function: lessons from lizard tails. Anatomical Record-Advances in Integrative Anatomy and Evolutionary Biology, 2007. 290(1): 108–114. Bland, K.I., et al., Axillary dissection in breast-conserving surgery for stage I and II breast cancer: a national cancer data base study of patterns of omission and implications for survival. Journal of the American College of Surgeons, 1999. 188(6): 586–595. Blumencranz, P., et al., Sentinel node staging for breast cancer: intraoperative molecular pathology overcomes conventional histologic sampling errors. American Journal of Surgery, 2007. 194(4): 426–432. Bono, P., et al., High LYVE-1-positive lymphatic vessel numbers are associated with poor outcome in breast cancer. Clinical Cancer Research, 2004. 10(21): 7144–7149. Cummings, M.C., et al., Occult axillary lymph node Metastases in breast cancer do matter results of 10-year survival analysis. American Journal of Surgical Pathology, 2002. 26(10): 1286–1295. Deland, F.H., E.E. Kim, and D.M. Goldenberg, Lymphoscintigraphy with radionuclide-labeled antibodies to carcinoembryonic antigen. Cancer Research, 1980. 40(8): 2997–3000. Engelstad, B.L., et al., Phase-1 immunolymphoscintigraphy with an in-111 labeled antimelanoma monoclonal-antibody. Radiology, 1986. 161(2): 419–422. Fang, J.M., et al., Mutations in FOXC2 (MFH-1) a forkhead family transcription factor, are responsible for the hereditary lymphedema-distichiasis syndrome. American Journal of Human Genetics, 2000. 67(6): 1382–1388. Foeldi, M. and E. Foeldi, Foeldi s Textbook of Lymphology, ed. L. Biotext, 2006, San Francisco, CA. Goldberg, B.B., et al., Sentinel lymph nodes in a swine model with melanoma: contrast-enhanced lymphatic US. Radiology, 2004. 230(3): 727–734. Greene, F.L., et al., AJCC Cancer Staging Manual. 6th ed. 2002, Springer, New York. Harrell, M.I., B.M. Iritani, and A. Ruddell, Tumor-induced sentinel lymph node lymphangiogenesis and increased lymph flow precede melanoma metastasis. American Journal of Pathology, 2007. 170(2): 774–786. Hauff, P., et al., Molecular targeting of lymph nodes with L-selectin ligand-specific US contrast agent: a feasibility study in mice and dogs. Radiology, 2004. 231(3): 667–673. Hirakawa, S., et al., VEGF-C-induced lymphangiogenesis in sentinel lymph nodes promotes tumor metastasis to distant sites. Blood, 2007. 109(3): 1010–1017.

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques

205

Houston, J.P., et al., Quality analysis of in vivo near-infrared fluorescence and conventional gamma images acquired using a dual-labeled tumor-targeting probe. Journal of Biomedical Optics, 2005. 10(5): 054010. Hughes, S.J., et al., A rapid, fully automated, molecular-based assay accurately analyzes sentinel lymph nodes for the presence of metastatic breast cancer. Annals of Surgery, 2006. 243(3): 389–398. Hwang, K., et al., Improved excitation light rejection enhances small-animal fluorescent optical imaging. Molecular Imaging, 2005. 4(3): 194–204. Hwang, K., et al., Influence of excitation light rejection on forward model mismatch in optical tomography. Physics in Medicine and Biology, 2006. 51(22): 5889–902. Kairemo, K.J.A., Immunolymphoscintigraphy with Tc-99m-labeled monoclonal-antibody (Bw431-26) reacting with carcinoembryonic antigen in breast-cancer. Cancer Research, 1990. 50(3): S949–S954. Keenan, A.M., et al., Immunolymphoscintigraphy in patients with lymphoma after subcutaneous injection of indium-111-labeled T101 monoclonal-antibody. Journal of Nuclear Medicine, 1987a. 28(1): 42–46. Keenan, A.M., et al., Immunolymphoscintigraphy and the dose dependence of in-111 labeled T101 monoclonal-antibody in patients with cutaneous T-cell lymphoma. Cancer Research, 1987b. 47(22): 6093–6099. Kingsmore, D.B., et al., Axillary recurrence in breast cancer. European Journal of Surgical Oncology, 2005. 31(3): 226–231. Lerner, S.P., et al., The rationale for en-bloc pelvic lymph-node dissection for bladder-cancer patients with nodal metastases - long-term results. Journal of Urology, 1993. 149(4): 758–765. Lohrmann, C., E. Foeldi, and M. Langer, Indirect magnetic resonance lymphangiography in patients with lymphedema – preliminary results in humans. European Journal of Radiology, 2006a. 59(3): 401–406. Lohrmann, C., et al., High-resolution MR lymphangiography in patients with primary and secondary lymphedema. American Journal of Roentgenology, 2006b. 187(2): 556–561. Lohrmann, C., et al., Gaclotericlol for MR imaging of lymphatic vessels in lymphoedematous patients: initial experience after intracutaneous injection. British Journal of Radiology, 2007. 80(955): 569–573. Lotze, M.T., et al., Monoclonal-antibody imaging of human-melanoma – radioimmunodetection by subcutaneous or systemic injection. Annals of Surgery, 1986. 204(3): 223–235. Mansel, R.E., et al., Randomized multicenter trial of sentinel node biopsy versus standard axillary treatment in operable breast cancer: the ALMANAC trial. Journal of National Cancer Institute, 2006. 98(9): 599–609. Matsushima, S., et al., Nonenhanced magnetic resonance lymphoductography: visualization of lymphatic system of the trunk on 3-dimensional heavily T2-weighted image with 2-dimensional prospective acquisition and correction. Journal of Computer Assisted Tomography, 2007. 31(2): 299–302. Mellor, R.H., et al., Mutations in FOXC2 are strongly associated with primary valve failure in veins of the lower limb. Circulation, 2007. 115(14): 1912–1920. Moghimi, S.M. and A.R. RajabiSiahboomi, Advanced colloid-based systems for efficient delivery of drugs and diagnostic agents to the lymphatic tissues. Progress in Biophysics and Molecular Biology, 1996. 65(3): 221–249. NCCN Guideline The NCCN GUIDELINE: Clinical Practice Guidelines in Oncology (Version 3.2005). (C) 2006 National Comprehensive Cancer Network, Inc., Available at: http://www.nccn.org . Nelp, W.B., et al., Preliminary studies of monoclonal-antibody lymphoscintigraphy in malignantmelanoma. Journal of Nuclear Medicine, 1987. 28(1): 34–41. Partsch, H., A. Urbanek, and B. Wenzelhora, The dermal lymphatics in lymphoedema visualized by indirect lymphography. British Journal of Dermatology, 1984. 110(4): 431–438. Partsch, H., B.I. Wenzelhora, and A. Urbanek, Differential-diagnosis of lymphedema after indirect lymphography with lotasul. Lymphology, 1983. 16(1): 12–18.

206

E.M. Sevick-Muraca

Petrova, T.V., et al., Defective valves and abnormal mural cell recruitment underlie lymphatic vascular failure in lymphedema distichiasis. Nature Medicine, 2004. 10(9): 974–981. Querzoli, P., et al., Axillary lymph node nanometastases are prognostic factors for disease-free survival and metastatic relapse in breast cancer patients. Clinical Cancer Research, 2006. 12(22): 6696–6701. Rabin, O., et al., An X-ray computed tomography imaging agent based on long-circulating bismuth sulphide nanoparticles. Nature Materials, 2006. 5(2): 118–122. Rasmussen, J.C., I-C. Tan, M.V. Marshal, C.E. Fife, and E.M. Sevick-Muraca, Lymphatic imaging in humans with near-infrared fluorescence. Current Opinion in Biotechnology, 2009. 20:1–9. Reynolds, J.S., T.L. Troy, and E.M. Sevick-Muraca, Multipixel techniques for frequency-domain photon migration imaging. Biotechnology Progress, 1997. 13(5): 669–680. Ridner, S.H., Pretreatment lymphedema education and identified educational resources in breast cancer patients. Patient Education and Counseling, 2006. 61: 72–79. Ruehm, S.G., T. Schroeder, and J.F. Debatin, Interstitial MR lymphography with gadoterate meglumine: initial experience in humans. Radiology, 2001. 220(3): 816–821. Sampath, L., et al., Dual-labeled trastuzumab-based imaging agent for the detection of Human Epidermal Growth Factor Receptor-2 (HER2) overexpression in breast cancer. Journal of Nuclear Medicine, 2007 (cover illustration). 48: 1501–1510. Sampath, L., W. Wang, and E.M. Sevick-Muraca, Near infrared optical for nodal staging. Journal of Biomedical Optics, 2008. 13:(4) 041312. Sevick-Muraca, E.M., et al., Imaging of lymph flow in breast cancer patients after microdose administration of a near-infrared fluorophore: feasibility study. Radiology, 2008. 246(3): 734–741. Sevick-Muraca, E.M. and D.Y. Paithankar, Fluorescence imaging system and measurement, 1999. Us patent 5,865,754. Sevick-Muraca, E.M. and J.C. Rasmussen, Molecular imaging with optics: primer case for nearinfrared fluorescence techniques in personalized medicine. Journal of Biomedical Optics, 2008. 13(4): 041303–1–16. Sharma, R., et al., New horizons for imaging lymphatic function. The New York Academy of Sciences, 2008. 1131: 13–36. Shen, J., et al., Feasibility and accuracy of sentinel lymph node biopsy after preoperative chemotherapy in breast cancer patients with documented axillary metastases. Cancer, 2007. 109(7): 1255–1263. Singletary, S.E., et al., Revision of the American Joint Committee on Cancer staging system for breast cancer. Journal of Clinical Oncology, 2002. 20(17): 3628–3636. Skobe, M., et al., Induction of tumor lymphangiogenesis by VEGF-C promotes breast cancer metastasis. Nature Medicine, 2001. 7(2): 192–198. Stein, J.P., et al., Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. Journal of Clinical Oncology, 2001. 19(3): 666–675. Su, J.L., et al., The VEGF-C/Flt-4 axis promotes invasion and metastasis of cancer cells. Cancer Cell, 2006. 9(3): 209–223. Suga, K., et al., Visualization of breast lymphatic pathways with an indirect computed tomography lymphography using a nonionic monometric contrast medium iopamidol – preliminary results. Investigative Radiology, 2003. 38(2): 73–84. Suga, K., et al., Lymphatic drainage from esophagogastric tract: feasibility of endoscopic CT lymphography for direct visualization of pathways. Radiology, 2006. 238(1): 952–960. Tjandra, J.J., et al., Immunolymphoscintigraphy for the detection of lymph-node metastases from breast-cancer. Cancer Research, 1989. 49(6): 1600–1608. Tjan-Heijnen, V.C.G., et al., Micro-metastases in axillary lymph nodes: an increasing classification and treatment dilemma in breast cancer due to the introduction of the sentinel lymph node procedure. Breast Cancer Research and Treatment, 2001. 70(2): 81–88. Weingartner, K., et al., Anatomical basis for pelvic lymphadenectomy in prostate cancer: results of an autopsy study and implications for the clinic. Journal of Urology, 1996. 156(6): 1969–1971.

Nodal Staging of Cancer Using Diagnostic Optical Imaging Techniques

207

de Vries, M., et al., Morbidity after inguinal sentinel lymph node biopsy and completion lymph node dissection in patients with cutaneous melanoma. European Journal of Surgical Oncology, 2006. 32(7): 785–789. Weinstein, J.N., et al., Monoclonal anitbodies in the lymphatics: toward the diagnosis and therapy of tumor metastases. Science, 1982. 218(4579): 1334–1337. Weinstein, J.N., et al., Monoclonal antibodies in the lymphatics: selective delivery to lymph node metastases of a solid tumor. Science, 1983. 222(4622): 423–426. Williams, A.F., P.J. Franks, and C.J. Moffatt, Lymphoedema: estimating the size of the problem. Palliative Medicine, 2005. 19(4): 300–313. Yao, M.S., et al., Internal mammary nodal chain drainage is a prognostic indicator in axillary node-positive breast cancer. Annals of Surgical Oncology, 2007. 14(10): 2985–2993. Zavagno, G., et al., Axillary recurrence after negative sentinel lymph node biopsy without axillary dissection: a study on 479 breast cancer patients. European Journal of Surgical Oncology, 2005. 31(7): 715–720.

Optical Coherence Tomography for Cancer Detection Steven G. Adie and Stephen A. Boppart

Introduction Optical Coherence Tomography (OCT) Optical coherence tomography (OCT) is an emerging high-resolution medical and biological imaging technology that is currently making the transition from the research lab into clinical practice. OCT is analogous to ultrasound B-mode imaging except that reflections of light are detected rather than sound. This technique is attractive for medical imaging because it permits real-time in situ imaging of tissue microstructure with resolution approaching that of conventional histology, but without the need for excision and histological processing. Although OCT penetration depth is on the order of 1–2 mm (Schmitt 1999), its fiber-optic implementation enables the use of compact endoscopic probes that facilitate internal access within the body, including epithelial layers, where 85% of all cancers originate (Gurjar et al. 2001). The diagnostic potential of OCT can be enhanced through various contrast mechanisms directly or indirectly relating to the physiological function of tissue, such as blood flow, oxygenation, or tissue birefringence, or with the use of targeted site-specific exogenous contrast agents. It is useful to compare OCT to other medical imaging modalities in terms of resolution and penetration depth. As can be seen in Fig. 1, the general trend is that higher resolution is accompanied by reduced penetration depth. OCT bridges the gap between confocal microscopy and high-frequency ultrasound. Imaging in nontransparent tissues is facilitated by the existence of a “biological window,” where absorption of near-infrared wavelengths is minimal and light can penetrate relatively deep into tissue. In most scattering tissues, imaging depths of 1–2 mm can be achieved (Schmitt 1999). Typical axial (or depth) resolutions are on the order of 5–10 μm, while ultrahigh resolution systems have achieved axial resolutions under 1 μm (Povazay et al. 2002). S.A. Boppart (B) Biophotonics Imaging Laboratory, Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA e-mail: [email protected]

E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_11, 

209

210

S.G. Adie and S.A. Boppart

Fig. 1 Comparison of OCT with standard clinical imaging modalities. Resolution diminishes with increasing tissue depth

Clinical Applications OCT has found the most widespread application in ophthalmology, due largely to the relative transparency of ocular tissues. The superior resolving capabilities of OCT over confocal scanning laser ophthalmoscopes, the technology which OCT has largely displaced, has made it the modality of choice for clinical imaging of many retinal diseases (Schuman et al. 2004). OCT has also been applied to the detection of vulnerable plaques in vivo with good success (Bouma et al. 2003) and to the long-range in vivo monitoring of upper airway profiles in the study of sleep apnea (Armstrong et al. 2006). The other main application areas focus on cancer detection in various parts of the body including the breast (Boppart et al. 2004), gastrointestinal tract (Pitris et al. 2000), bladder (Zagaynova et al. 2004), skin (Gambichler et al. 2007b), oral cavity (Kawakami-Wong et al. 2007), cervix (Escobar et al. 2006), lung (Whiteman et al. 2006), and brain (Bohringer et al. 2006). The ability of OCT to image tissue in situ with near-histological resolution permits visualization of morphological differences between normal and neoplastic tissues, suggesting the possibilities of performing in vivo optical biopsies and guiding surgical procedures. Several OCT beam delivery systems have been developed (Schmitt 1999; Bouma and Tearney 2002), providing the flexibility for imaging in a variety of clinical settings. Fixed or handheld probes facilitate open-field noninvasive imaging, while forward-imaging endoscopes (Boppart et al. 1997) and fiber-optic-based catheters enable minimally invasive internal body imaging (Zara and Lingley-Papadopoulos

Optical Coherence Tomography for Cancer Detection

211

2008). Needle-based imaging provides deep access to suspicious lesions and has recently been achieved by incorporating an OCT fiber probe into a 23-gauge (∼330 μm) needle for fine-needle aspiration, (Iftimia et al. 2005).

Translation of OCT Technology Translation of OCT technology from the research lab into clinical use can be described with reference to the development process for pharmaceutical agents (Zysk et al. 2007c). The cycle begins with technology development (step 0), through to its acceptance as the standard of care (step 4). Step 1 involves pilot studies to determine whether the technology is safe and whether it is feasible to differentiate between normal and diseased tissues for a specific imaging application, e.g., breast cancer. In step 2, a limited trial is conducted to determine the usefulness of the technique for differentiating normal and diseased tissues or for grading disease states. It should be noted that during steps 0–2, the technology may not yet be clearly delineated, and technology developments can relatively easily be incorporated into the trials. Evaluation of imaging technology developments is conducted via the logical progression from cell line and tissue phantom measurements to animal (pre-clinical) and human tissue measurements. Step 3 expands validation to large-scale multicenter trials, during which technology standardization (controlling for OCT system performance) becomes important. Validation of OCT imaging in animal or human tissue is commonly conducted by comparison to light microscopy of histological sections, the medical gold standard. Sensitivity and specificity of the new imaging technique is determined in steps 2 and 3 via double-blinded evaluation of co-located measurements Adoption of technology in to the standard of care (step 4) is accompained by further commercialization and widespread establishment of the relevant standard operating procedures for management of the disease. Chapter outline in this chapter we present a brief background of the principles of OCT and then discuss major translational efforts toward clinical oncological applications. We discuss in detail aspects of OCT translational research on breast cancer and provide an overview of translational research progress for other main application areas such as gastrointestinal, bladder, skin, and oral cancer. For each application area, the scope for OCT intervention is described, followed by a discussion of feasibility studies that demonstrate the potential contribution of OCT imaging to the area. Specific technological developments underpinning the clinical diagnostic capabilities for each specific application area are highlighted. Finally, the findings (and methods) of clinical or animal studies testing the efficacy of OCTbased diagnostics are discussed. Emphasis is placed on validation of OCT imaging by comparison to the medical gold standard of histology. Section “Computer-Aided Tissue Classification, Multimodal Imaging and Contrast Enhancement” includes promising future directions for further improvement to the diagnostic capabilities of OCT, while the Section “Barriers to Clinical Application” discusses barriers to clinical adoption of OCT technology. We conclude in Section “Summary and Outlook on Clinical Adoption” with an outlook on the clinical adoption of OCT technology for oncological applications.

212

S.G. Adie and S.A. Boppart

Principles of OCT Low-Coherence Interferometry The origins of OCT lie in white-light interferometry that led to the fiber-based optical coherence-domain reflectometry (Youngquist et al. 1987). Initially the technique was proposed for locating faults or breaks in long optical fibers in telecommunication networks, but later was used to obtain one-dimensional depth resolved measurements in biological tissues (Swanson et al. 1992). The addition of lateral beam scanning, as outlined in the landmark paper published in Science (Huang et al. 1991), enabled noninvasive cross-sectional tomographic imaging of biological tissues. Since the speed of light is too fast to permit direct ultra-precise measurement of the time of flight, low-coherence interferometry is employed to indirectly sense the depth of scatterers in a sample. Figure 2 presents a simplified schematic of an OCT system based on a fiber-optic Michelson interferometer. Light from a broadband source is split into the sample and reference arms. Light reflected from each arm recombines at the coupler and a detector measures the resulting interference. With time-domain detection, the broadband optical spectrum is detected simultaneously, often utilizing a pair of photodiodes in a balanced configuration to reduce common-mode noise. This produces observable interference fringes only when the path lengths of the reference and sample arms are equal to within the source coherence length, typically on the order of 5–10 μm.

Fig. 2 Schematic of a typical fiber-optic OCT system showing both time-domain and spectraldomain detection

Optical Coherence Tomography for Cancer Detection

213

Scanning of the reference arm then effectively scans the sensing/probing depth and point-wise recording of the interference generates an axial scan, known as an “Ascan” (following ultrasound terminology). A galvanometer mounted mirror in the sample arm laterally scans the beam on the sample to generate a two-dimensional cross-sectional (B-mode) OCT image comprised of a series of laterally displaced A-scans. Spectral-domain OCT (SD-OCT) offers an alternative detection scheme where the broad optical bandwidth is dispersed onto a multi-element line-scan camera. Each pixel, or spectral bin, has a relatively small bandwidth (long coherence length) and therefore simultaneously detects the resultant interference from all depths without the need to scan the reference mirror. An inverse Fourier transform of the spectral interference fringes recorded by the line-scan camera produces the A-scan (Fercher et al. 2003). A similar scheme requiring an inverse Fourier transform of spectral interference fringes is swept-source acquisition – achieved by scanning (or sweeping) the wavelength of a laser source over its optical bandwidth to acquire an A-scan (Choma et al. 2003; Yasuno et al. 2005). Since signal from all depths are acquired in parallel, spectral-domain (SD-) and swept-source (SS-) OCT offer a 20-30 dB improvement in sensitivity (see Section “OCT System Performance” for definition), when compared to a similar time-domain system of the same A-scan rate (de Boer et al. 2003; Choma et al. 2003). Alternatively, as implied by Equation 4, spectral-domain detection can acquire images of the same sensitivity 100–1000 times faster than a similar time-domain system.

OCT System Performance The performance of an OCT system is described in terms of its resolution, sensitivity, and imaging speed. The full-width half-maximum (FWHM) axial resolution z is governed by the coherence length, which depends primarily on the source bandwidth according to (Schmitt 1999) 2 ln 2 λ2 · , (1) π λ where λ is the FWHM optical bandwidth (assumed Gaussian) and λ is the center wavelength. Lateral resolution is governed by the numerical aperture (NA) of the focused beam and is typically given as the focused Gaussian beam spot size. The lateral resolution x is given by (Saleh and Teich 1991) z =

4λ f · , (2) π d where f is the focal length of the objective lens and d is the beam width incident upon it. Traditionally, improvement of lateral resolution in optical imaging comes at the expense of depth-of-field. In OCT the depth-of-field is defined as twice the Rayleigh range, zR , of the focused Gaussian beam. Known as the confocal parameter, the depth-of-field is proportional to the square of the lateral resolution and is given by (Saleh and Teich 1991) x =

214

S.G. Adie and S.A. Boppart

2zR =

π x2 . 2λ

(3)

The term optical coherence microscopy (OCM) has been used to describe high lateral resolution imaging achieved with the use of objective lenses with high numerical apertures. Depth-of-field limitations in OCM can be overcome in time-domain detection by dynamic focusing (Lexer et al. 1999). With spectral-domain detection, the superior phase stability (due to a static reference arm and higher speed data acquisition) is utilized by a computational technique known as interferometric synthetic aperture microscopy (ISAM) (Ralston et al. 2007) to overcome such depth-of-field limitations. The sensitivity S of an OCT system describes its minimum detectable reflectivity or the weakest optical signal detectable. The sensitivity of an OCT system in the optical shot noise limit is proportional to the optical power incident on the sample and inversely proportional to the A-scan rate and is given by (Fercher et al. 2003) S=

1 ηλ Ps , 4 hc B

(4)

where η is the quantum efficiency of the detector, h the Plank’s constant, c the speed of light, Ps the optical power incident on the sample, and B the electronic detection bandwidth. In time-domain OCT, B is related to the A-scan frequency through the scan velocity, v, according to B = (2 ln 4)(4/π )(v/λ) (Swanson et al. 1992), and in spectral-domain detection it is related to the A-scan acquisition rate according to B = 1/(2Ti ) (de Boer et al. 2003), where Ti is the camera integration time. (In spectral-domain detection, S in Equation 4 is reduced by a factor of 2, since only the real part of the complex cross-spectral density is measured (de Boer et al. 2003).) A further factor affecting the imaging performance of the various acquisition schemes is the dynamic range of reflectivity (or scattering signal strength) that the OCT system can measure. In practice, the bit depth at the digitization stage of the signal acquisition can place limits on the achievable dynamic range (Liu and Brezinski 2007). In time-domain and swept-source acquisition, AC-coupled detection allows the full bit depth of the analog-to-digital converter to be applied to the interference part of the signal. However, the bit depth of line-scan cameras used in SD-OCT is spread over a relatively large incoherent background containing interference fringes superimposed on top, thus reducing the effective bit depth spanning the interference signal. This limits the ability of the system to detect both strong signals near the surface and weak signals from deep within a turbid sample.

Oncological Tissue Optics and OCT Imaging Several characteristics of OCT images can be used to differentiate between cancerous and normal tissues. In highly scattering tissues, OCT typically does not resolve individual cells, but is sensitive to the disruption of normal tissue architecture or

Optical Coherence Tomography for Cancer Detection

215

morphology (e.g., layer thickness or scattering regularity), which is often associated with carcinomas (Fujimoto 2003; Pan et al. 2001). The relative strength of the optical scattering (Isenberg et al. 2005; Qi et al. 2006) and attenuation (Goldberg et al. 2008) in OCT images provides another parameter to differentiate normal from cancerous tissues. Joint spatial frequency and textural image analysis, which is sensitive to speckle patterns as well as tissue architectural features, can also provide distinction between cancerous and normal tissues (Zysk and Boppart 2006; Qi et al. 2006; Goldberg et al. 2008). The physical mechanism giving rise to information in OCT images is optical scattering from microscopic refractive index variations within the tissue (Schmitt 1999). The nuclei of cells are understood to be one of the main sources of scattering detected by OCT (Xie et al. 2002; Mourant et al. 2000). Fundamental studies demonstrate differences in light scattering from cells at various stages of neoplastic progression (Drezek et al. 2003; Ramachandran et al. 2007). These light–tissue interactions underpin the oncological diagnostic capabilities of OCT. Traditional assessment of neoplastic changes based on cellular features such as atypia of cell nuclei, accelerated rate of growth, and local invasion are motivating the pursuit of cellular imaging. New developments in optical sources for ultrahigh resolution (Drexler 2004) and real-time methods for extending depth-of-field in OCM (Ralston et al. 2008) suggest that turbid-tissue diagnostics based on cellular resolution is possible. These improvements are expected to further improve the sensitivity and specificity of OCT for detection of cancer (Fujimoto 2003). The penetration depth in a turbid biological sample is limited by multiple scattering (Pan et al. 1997; Yadlowsky et al. 1995; Adie et al. 2007). Dominant tissue chromophores, such as hemoglobin, water, and melanin, have relatively low absorption over the 700–1300 nm biological optical window. Below 700 nm, hemoglobin absorption dominates, while water absorption becomes significant above 1400 nm. Studies on the effect of wavelength on OCT penetration depth have shown that penetration depth is optimized with longer 1300 nm optical sources (Brezinski and Fujimoto 1999; Aguirre et al. 2006), since scattering decreases nearly monotonically with increasing wavelength (Fercher et al. 2003).

Clinical Oncology Applications Breast Cancer Breast Cancer Management and Scope for OCT Intervention The clinical management of breast cancer offers several opportunities where current practice could potentially benefit from OCT imaging. Figure 3 summarizes the management of breast cancer, and potential intervention opportunities for OCT, ranging from early stage screening to the analysis and/or guidance of surgical tissue resection. This potential derives from the high-resolution and high-acquisition

216

S.G. Adie and S.A. Boppart

Fig. 3 Screening and surgical treatment of breast cancer (rectangular boxes) and potential areas for OCT intervention (ovals). The current clinical practice for each stage of care is summarized by the text at the bottom

rate of OCT, its inherent compatibility with the use of compact fiber-based probes, and the ability to construct portable systems. Limitations of optical scattering on penetration depth can be somewhat circumvented through the use of compact optical fiber probes permitting access within the ductal structure of the breast or to a suspicious lesion via the tip of a biopsy needle. The feasibility of in vivo imaging with compact fiber-optic probes has been demonstrated via fiber-optic ductoscopy (white light, noninterferometric imaging) of women with nipple discharge (Shen et al. 2000; Liu et al. 2008). The high-resolution and acquisition speed of OCT could also be utilized to image margins of resected tissue or directly image the open surgical field to a depth of 1–2 mm, providing an imaging capability not currently available to surgeons. OCT research efforts on breast cancer have to date focused on development of a needle biopsy guidance capability and on intraoperative imaging applications, and thus will be the main subject of this section. The motivation for OCT-guided needle biopsy is underpinned by the relatively high nondiagnostic sampling rate for needle biopsy procedures. Pijnappel et al. reported an 8–12% “miss rate” for large-core needle biopsy (LCNB) obtained under ultrasound guidance (Pijnappel et al. 2004). Thus localized optical imaging of tissues at the needle tip, along with real-time feedback, has the potential to improve

Optical Coherence Tomography for Cancer Detection

217

the nondiagnostic sampling rate. The penetration depth, high resolution, and fiberoptic compatibility of OCT are well suited to the requirements for needle biopsy guidance. Open surgery can be classified into breast conserving procedures (lumpectomy) or complete breast removal (mastectomy). The aim of lumpectomy procedures is to remove all cancerous tissue, while preserving normal breast tissue. When combined with adjuvant radiation therapy, it provides an effective alternative to mastectomy (Fisher et al. 2002). The significance of margin status was studied for 276 patients undergoing re-excision or mastectomy subsequent to initial tumor excision (Cellini et al. 2004). Close margins in the initial excision procedure resulted in a residual cancer rate of 53% in the re-excised specimen, while positive or undetermined margins increased this rate to 68 and 67%, respectively. Intraoperative feedback on margin status of resected tissue could enable the surgeon to remove additional remaining cancerous tissue within the same procedure and reduce the rate of residual cancer. Current methods of feedback available to surgeons during open breast surgery are X-ray imaging or histological analysis of resected masses. These operate on distinctly different spatial resolutions, and both are time-consuming and suffer from inadequate sampling. Histological analysis enables microscopic imaging of tissue slices, but the time frame required for intraoperative feedback places severe constraints on the number of slices that can be analyzed – typically only 6–10 cryosections may be analyzed, potentially leading to sampling errors. More rapid feedback than cryosectioning is available by the touch-prep technique, where a microscope slide is touched to excised tissue margins intraoperatively, stained and viewed under a microscope. However, the method does not provide high sensitivity and specificity. X-ray imaging (of tumors displaying X-ray contrast) can provide macroscale detail to detect the presence of a primary tumor, but only provides a 2D projection and is not suited to imaging small-scale structure of irregular borders or metastasizing cells. X-ray contrast may also miss lesions, ducts, and structures that are not calcified. Thus, the high resolution and real-time 3D imaging capabilities of OCT could provide an important advance for the surgical removal of tumors. Sentinel lymph node biopsy has become the standard of care for assessing the involvement of axillary lymph nodes (Veronesi et al. 2003; McMasters et al. 2000) and staging the progression of breast cancer. The sentinel node is the first node in the lymphatic drainage network form a given cancerous region of the breast, and where micro-metastases are known to sequentially infiltrate the chain of lymph nodes. Thus patients with a negative sentinel node can be spared an axillary node dissection and associated morbidity such as lymphedema, arm numbness, impaired shoulder mobility, arm weakness, and infections in the breast, chest, or arm (Swenson et al. 2002). The sentinel node is identified intraoperatively by detecting the levels of radioactive tracer or blue dye injected into the peri-tumoral region of the breast (McMasters et al. 2000). In practice, however, this method is imprecise since radioactive tracer and/or blue dye levels are detected in multiple “sentinel” nodes (Zakaria et al. 2007). The high-resolution, real-time imaging capability of OCT

218

S.G. Adie and S.A. Boppart

could potentially be useful for intraoperative screening and detection of micrometastasis in these “sentinel” lymph nodes and potentially even do so in situ, without having to disrupt the normal lymphatic drainage of the tissue. OCT Atlas of Breast Cancer OCT imaging of normal and cancerous breast tissues began in the optics lab to investigate the feasibility for tissue classification and to generate an atlas of breast OCT images with co-registered histology. This pursuit guides the interpretation of OCT images of breast tissue, and when coupled with blinded studies on a sufficiently large sample size, enables benchmarking of the sensitivity and specificity of OCT-based tissue classification. In this section, we describe the major OCT image features, with supporting images of early stage and advanced breast carcinoma measurements taken in the optics lab. OCT images were selected to present both cancerous and normal breast tissue side-by-side to minimize the effects of system-specific parameters such as wavelength and resolution on the main findings. Sections “Intraoperative Tumor Margin Detection” and “Intraoperative Lymph Node Imaging” extend the translational research efforts to encompass clinical intraoperative measurements. Figure 4 presents images of ductal carcinoma in situ (DCIS) from a recent study by Hsiung et al. of 119 freshly excised human breast tissue specimens from 35 women aged 29–81 (Hsiung et al. 2007). Lobules and surrounding fibrous stroma identifiable in the histology image correlate to regions in the OCT image of relatively low and high scattering, respectively. Interestingly, the authors found that tumor cells within lobules appear to be lower scattering than the surrounding fibrous stroma. They also note dilatation of the lobules caused by the tumor cells, accompanied by architectural distortion of the surrounding stroma. Similarly, although

Fig. 4 (a) OCT image of DCIS lesions in lobules. Tumor cells within lobules appear uniformly low scattering. Dilatation and architectural distortion of the lobules are visible. A microcalcification (C, circled area) within the lobules appears highly scattered with pronounced shadowing. (b) Histologic specimen corresponding to OCT image. (Hematoxylin–eosin stain; original magnification, 40×.) OCT measurement taken with a 1.1 μm wavelength system with 3.5 μm axial resolution and 6 μm transverse resolution. Figure reprinted with permission from Hsiung et al. (2007)

Optical Coherence Tomography for Cancer Detection

219

lumina were not observable for ductal hyperplasia or DCIS, dilation and distortion were observed in ducts containing DCIS lesions. Evidence of microcalcification was also observable as a focal point of higher scattering within a lobule (not present in given plane of the histology image). Figure 5 presents the measurement of invasive ductal carcinoma (left) bordering normal adipose tissue (right) along with the corresponding histology. Low-scattering lipid-filled adipocytes, scattering membranes, and highly scattering cell-dense tumor regions are visible in the OCT and histology images. The boundary between tumor and adipose tissues as observed in the OCT image correlates well with the histology image. It is interesting to compare the scattering response between invasive ductal carcinoma and invasive lobular carcinoma. In the study by Hsiung et al., the scattering associated with ductal carcinoma was found to be greater than the surrounding stroma, but scattering from invasive lobular carcinoma was lower than the surrounding stroma (Hsiung et al. 2007). Figure 6 presents measurements of infiltrating lobular carcinoma. The presence of entrapped adipose cells is visible, and isolated bands of increased scattering were attributed to intervening fibrotic stroma. Due to the lower scattering observed for tumor cells in DCIS and invasive lobular carcinoma, the authors suggest that malignant cellular change may not be the primary source of increased scattering for invasive ductal carcinoma. Instead they suggest that this is due to a desmoplastic reaction to infiltrating cancer. Thus, while ductal carcinoma is likely to be distinguishable from normal stroma, infiltrating lobular carcinoma may be more difficult to differentiate from normal fibrous stroma.

Fig. 5 (a) OCT image of invasive ductal carcinoma (left) and its boundary with adipose (right) and (b) corresponding histology. The arrows highlight selected details of correspondence between the images. Data taken with an 800 nm Ti:Sapphire OCT system with 4 μm axial resolution and 15 μm transverse resolution. Figure reprinted with permission from Zysk and Boppart (2006)

220

S.G. Adie and S.A. Boppart

Fig. 6 (a) OCT image of a solid variant infiltrative lobular carcinoma. Regions with densely infiltrating tumor cells appear low scattering and homogeneous, with isolated regions of entrapped fat. (b) Histologic specimen corresponding to OCT image. (Hematoxylin–eosin stain; original magnification, 40×.) OCT measurement taken with a 1.1 μm wavelength system with 3.5 μm axial resolution and 6 μm transverse resolution. Figure reprinted with permission from Hsiung et al. (2007)

Intraoperative Tumor Margin Detection Two main categories of real-time intraoperative imaging are conceivable – ex vivo imaging of resected masses or direct in situ (in vivo) imaging of tissue within the surgical field. Ex vivo imaging of resected masses before any form of tissue processing provides the first step to investigate the intraoperative imaging capabilities of OCT. After imaging, tissue can be marked with ink to facilitate subsequent co-localized comparison of OCT measurements with histopathology. Intraoperative ex vivo studies are facilitated by portable OCT systems located within the surgical suite. Important technical requirements for such a system are acquisition speed, resolution, and penetration depth. Other desirable characteristics are the acquisition of 3D volumes and the incorporation of needle-based probes to facilitate deeper tissue imaging. Spectral-domain or swept-source acquisition provides important advantages over time-domain acquisition, since it enables 100–1000 times higher acquisition speed without sacrificing signal-to-noise ratio. The use of 1300 nm optical sources maximizes penetration depth in highly scattering tissue. Commercially available portable sources at 1300 nm have recently obtained bandwidths in excess of 100 nm, providing for axial resolutions better than 6 μm in tissue. The feasibility of intraoperative imaging has been investigated in a study of 18 resected specimens (Nguyen et al. 2007, 2008a, 2008b). Figure 7 presents an OCT image and corresponding histology section immunohistochemically stained for the estrogen receptor (ER), demonstrating the detection of a positive margin. The central area of high scattering in the OCT image correlated to the tumor as highlighted by the ER+ image and is flanked by normal adipose tissue. Tissue architecture and scattering from tumor cells were significantly different from that of normal adipose tissue or cells and could be readily differentiated from residual blood on the specimen or from coagulation artifacts from the surgical

Optical Coherence Tomography for Cancer Detection

221

Fig. 7 Intraoperative detection of positive margins in invasive papillary cancer with OCT (top) and corresponding ER+ immunohistochemically stained section (bottom). Arrows highlight correlation of tumor architecture between the images

resection. However, differentiation of tumor from stroma presents a greater challenge and may benefit from polarization-sensitive OCT imaging (discussed further in Section “Enhancing Endogenous Contrast”). Further work is underway to assess the sensitivity and specificity of intraoperative OCT-based tissue classification. Intraoperative Lymph Node Imaging An in vitro feasibility study with rat and human lymph nodes investigated the potential for OCT-based diagnostics and to provided a baseline for intraoperative imaging studies (Luo et al. 2005). Three-dimensional OCT imaging of lymph nodes from a carcinogen-induced rat mammary tumor model enabled visualization of anatomical features such as the capsule, lymphoid follicles, cortex, and medullary sinuses, and good correlation with histology was obtained. Measurement of a human lymph node with metastatic squamous cell carcinoma showed OCT-based visualization of microvasculature with good correspondence to histology. Figure 8 presents measurement of human lymph nodes with late stage metastatic and necrotic squamous cell carcinoma. The structure of blood vessels, regions of squamous cell carcinoma, and advanced necrosis are well correlated between OCT and H&E-stained histology images. Despite a penetration depth limitation 1–2 mm with OCT imaging, relevant morphological features were accessible from the surface. This study suggests that OCT has the potential to provide intraoperative guidance on which nodes are resected for the purposes of breast cancer staging. The feasibility of intraoperative sentinel lymph node assessment was investigated in a study of 15 patients that imaged 17 normal, 1 reactive, and 2 metastatic nodes (Nguyen et al. 2008a, b; Nguyen et al. 2007). Figure 9 presents results from one

222

S.G. Adie and S.A. Boppart

Fig. 8 Three-dimensional ex vivo OCT images and corresponding H&E histology of (a) human lymph node with metastatic squamous cell carcinoma and (b) human lymph node with metastatic and necrotic squamous cell carcinoma. (Scale bars = 100 μm.) Published with permission from Luo et al. (2005)

Fig. 9 (a) Intraoperative OCT imaging of a reactive sentinel lymph node with (b) corresponding histology and (c, d, e) intraoperative OCT imaging of normal lymph nodes

Optical Coherence Tomography for Cancer Detection

223

reactive and several normal nodes. The capsule is distinguishable from the cortex in the images of the normal nodes by the relatively higher scattering from the capsule layer. However, the reactive sentinel lymph node has uniformly strong scattering across the capsule-to-cortex boundary, which is maintained deeper within the cortex. Needle-Biopsy Guidance Needle-based biopsy guidance with OCT is underpinned by the inherent compatibility of OCT with miniature fiber-optic probes. These probes may be incorporated into traditional needles as small as 31-gauge (inner diameter, 127 μm) (Reed et al. 2002) or potentially into core biopsy devices to provide optical guidance at the tip. Imaging with such compact fiber-based probes is currently restricted to the acquisition of A-scans. Due to the lack of lateral beam scanning and the use of single fiber probes, the method is perhaps more accurately described as low-coherence interferometry (LCI) (Youngquist et al. 1987; Swanson et al. 1992), a technique that is the historical precursor to scanned image-based OCT. Breast biopsies routinely performed in clinical practice fall into two categories – fine-needle aspiration biopsy (FNAB) or large-core needle biopsy (LCNB). FNAB aspirates fluid and cells from the location of a tumor, while LCNB extracts a solid tissue sample for subsequent histological analysis. In order to minimize the acquisition of nondiagnostic normal tissue, five core biopsy samples are commonly acquired, with more recommended to reliably diagnose certain calcified lesions (Liberman et al. 1994). FNAB is typically performed with 22-gauge needles, in various lengths from 5 to 20 cm, with external diameters of 0.6–1.0 mm (Frable 1983), while LCNB typically uses 18- to 14-guage needles (Reynolds 2000), with outer diameters up to 2.1 mm. Two mechanisms of contrast have been investigated for LCI-guided needle biopsy. The first is optical scattering as manifest in the A-scan (Zysk and Boppart 2006; Goldberg et al. 2008) and the second is the group refractive index (Zysk et al. 2007a, b). The LCI scattering profile of adipose tissue produces a periodic response with large signal peaks corresponding to the boundary of the adipose cells; stroma produced a denser scattering response with higher frequency oscillations than adipose tissue; while tumor also produced a dense response, but with greater attenuation than either stroma or adipose (Zysk and Boppart 2006). The group refractive index of adipose tissue has been shown in a rat mammary tumor model, which closely resembles ductal carcinoma in humans, to be significantly higher than that of tumor (Zysk et al. 2006). The combination of scattering and refractive index information was found to enhance the diagnostic value over separate independent use of the parameters (Zysk et al. 2008). A relatively compact needle-based device for measuring the scattering response of tissue has been developed for LCI-guided FNAB procedures (Iftimia et al. 2005) comprising a 250 μm single-mode optical fiber housed without a focusing lens within a 23-gauge FNA needle (with inner and outer diameters of ∼330 and ∼570 μm, respectively). The needle was attached to a regular syringe through a hub to enable tissue aspiration while allowing the fiber free movement at the tip of

224

S.G. Adie and S.A. Boppart

the needle. A similar device was used in a subsequent study to develop automated algorithms to test the feasibility of differentiating fibroglandular tissue from adipose (Goldberg et al. 2008). Two needle designs have been developed for the measurement of refractive index. In the first design, the optical path length through a sample residing in a fixedwidth channel is measured (Zysk et al. 2007a). This allows calculation of the group refractive index since optical path length is the product of the sample refractive index and thickness. The second design is based on operating principle of a refractometer, but with the addition of LCI gating of the reflection from the needle tip in contact with a scattering sample (Zysk et al. 2007b). The phase refractive index of the sample is obtained since it governs the Fresnel reflection coefficients at the needle tip. Both designs require further work on miniaturization and incorporation into a needle biopsy device. The incorporation of computer-aided diagnosis (CAD) algorithms (see Section “Computer-Aided Diagnosis”) could further enhance the diagnostic value.

Gastrointestinal Cancers Gastrointestinal endoscopic OCT imaging is enabled by the development of endoscopic sample arm probes, which can be inserted via the biopsy channel of a standard gastro-endoscope. The feasibility of in vivo imaging of the gastrointestinal tract has demonstrated the capability to distinguish between normal and abnormal mucosae (Pitris et al. 2000; Jackle et al. 2000), suggesting a potential role for optical biopsy. Recent OCT research suggests its capability to image large surface areas in real time, and thereby potentially also overcome the sampling constraints of current excisional biopsy (Adler et al. 2007). Figure 10 provides an example of ex vivo

Fig. 10 In vitro OCT images and corresponding histology of (a) normal colon showing columnar epithelial morphology with crypt structures and (b) carcinoma showing disruption of normal epithelial structure and disorganization of the crypts, with corresponding histology. Tissue layer labels: m, mucosal; mm, muscularis mucosal; and sm, submucosal. Figure reprinted with permission from Fujimoto (2003)

Optical Coherence Tomography for Cancer Detection

225

OCT imaging of colon tissue, comparing the morphology of normal and cancerous tissues. Two application areas of interest stand out as the most promising for endoscopic OCT imaging in the gastrointestinal tract – Barrett’s esophagus and the colon. We present an overview of the major translational research studies and the impact of recent technological improvements for each application.

Barrett’s Esophagus The accuracy of OCT-based detection of dysplasia in Barrett’s esophagus was investigated in a double-blinded study comparing OCT with histological analysis (Isenberg et al. 2005). A total of 314 biopsy sites from 33 patients were imaged using an OCT system with axial and lateral resolutions of 10 and 25 μm, respectively. Four endoscopists interpreted the OCT images, using two criteria as indicative of dysplasia-reduced light scattering and loss of tissue architecture. The researchers reported a sensitivity and specificity of 68 and 82%, respectively, and with an overall accuracy of 78%, but with significant variability between the endoscopists who analyzed OCT images. Despite only moderate sensitivity and specificity, the study demonstrated a high negative predictive value of 89%, which could prove advantageous for biopsy guidance. Although the statistics of this study do not yet justify OCT-based optical biopsy, technological improvements to resolution and computer-aided diagnosis are expected to improve surveillance of Barrett’s esophagus. The impact of computer-aided diagnosis and resolution has been investigated by two recent studies (Qi et al. 2006; Chen et al. 2008). (Details of computer-aided detection strategies are discussed in Section “Computer-Aided Diagnosis.”) In the study by Qi et al. (2006) 106 OCT–histology paired images from 13 patients resulted in detection sensitivity and specificity of 82 and 74%, respectively, and an accuracy of 83%. This is promising for high-throughput screening applications and for avoiding inter-observer variability. The subsequent study by Chen et al. (2008) utilized similar computer-aided algorithms to compare co-located OCT data sets acquired at 13 and 5 μm axial resolutions. Ultrahigh axial resolution provided enhanced visualization of clinically relevant features such as glandular structures. Computer-aided discrimination of Barrett’s from normal esophagus was enhanced, although further investigation is required to establish the improvement in distinguishing dysplasia from Barrett’s esophagus. Further enhancements in axial resolution, and combination with OCM to obtain cellular-level resolution, are expected to bring substantial further improvements in tissue classification.

Colon Cancer The feasibility of utilizing OCT in vivo to detect dysplasia of the lower GI tract was investigated in a study examining 44 colon polyps from 24 patients (Pfau et al. 2003). Standard OCT resolution of 10–20 μm and a frame rate of 4 frames per

226

S.G. Adie and S.A. Boppart

second were employed. Hyperplastic polyps were similar to normal mucosa, but dysplastic tissue (adenomas) was differentiated by loss of tissue organization and reduced light scattering. The impact of resolution and 3D acquisition was conducted with freshly excised tissue from the large and small intestines of 23 patients (Hsiung et al. 2005). Several combinations of axial and transverse resolutions were investigated with resolutions ranging from 11.6 μm axial × 23 μm transverse to 3.5 μm axial × 6 μm transverse. The epithelial layer of the mucosa and individual villi, glands, and crypts could be visualized. Neoplasia and inflammation were manifest in the OCT images as architectural distortion of glands. The highest resolution provided enhanced visualization of fine structures, but the reduced depth-of-field from the high 6 μm lateral resolution did not permit clear visualization of both mucosa and submucosal layers within the same image. A recent study has greatly improved the real-time imaging capability of OCT, acquiring 3D data of rabbit colon in vivo at a frame rate of 50 frames per second (Adler et al. 2007). New swept-source technology called a Fourier-domain modelocked frequency-swept laser enabled unprecedented acquisition speeds, with axial resolutions of 5–7 μm. The development of high-speed swept laser sources significantly enhances the real-time capabilities of OCT not only for colon cancer but other applications as well.

Bladder Cancer The current clinical standard for early bladder cancer detection is cytoscopy, which provides en face imaging of the surface of the bladder wall. Its sensitivity is limited, especially for diagnosis of an occult high-grade malignancy known as carcinoma in situ, and therefore is often combined with multiple random biopsies (Wang et al. 2007b; Lingley-Papadopoulos et al. 2008). However, this procedure still misses 50% of early flat bladder cancers (Wang et al. 2007b). The diagnostic sensitivity and specificity of early bladder cancer diagnosis could therefore benefit from high-resolution imaging of subsurface tissue morphology. In vivo OCT imaging in the human bladder is made possible by interfacing endoscopic sample probes via the probe channel of a cytoscope. Diagnostic potential for bladder cancer has recently benefited from technology developments incorporating microelectromechanical systems (MEMS) scanners and SD-OCT to improve resolution and acquisition rate (Wang et al. 2007b), as well as fluorescence-guided OCT imaging (Pan et al. 2003; Wang et al. 2005). The first systematic OCT study measuring tissue morphological changes during tumorigenesis was performed in the rat model of bladder cancer (Pan et al. 2001). At each point during tumorigenesis, at least two Fisher rat bladder samples exposed to methyl-nitroso-urea (MNU) were studied ex vivo with a benchtop OCT system. Tumorigenesis was associated with increased urothelial thickness and backscattering, demonstrating the feasibility of detection and staging of early invasive but superficial carcinoma. Inadequate resolution was cited as limiting the

Optical Coherence Tomography for Cancer Detection

227

capability to distinguish neoplasia or malignancies from hyperplasia or dysplasia. Subsequent theoretical modeling of light–tissue interaction utilized histomorphometric evaluations of urothelial cellular morphology to simulate the impact of bladder tumorigenesis on OCT images (Xie et al. 2002). The theoretical modeling supported the finding of a previous animal study that OCT has the potential to differentiate inflammation, hyperplasia, and neoplasia by quantifying changes in urothelial thickening and backscattering. Human studies have demonstrated the feasibility of OCT in vivo and produced findings similar to the animal studies. A study of 680 OCT images from 66 patients found that squamous metaplasia appeared as thicker and less transparent epithelium with a jagged boundary, while a complete loss of layered structures of the bladder wall was associated with transitional cell carcinoma (Zagaynova et al. 2002). A subsequent study further investigated the capability of OCT for evaluating transitional cell carcinoma, with 261 OCT images of 87 areas from 24 patients (Manyak et al. 2005). OCT images were graded on integrity of layered structures and classified by an investigator independent from the histology images as benign, abnormal but not invasive, and abnormal and invasive. Architectural disruption within all three layers was considered indicative of abnormal and invasive. High sensitivities and specificities of 100 and 89% were obtained, and the overall accuracy of OCT diagnosis was 92%. An excellent negative predictive value of 100% suggests a significant role in biopsy guidance; however, extension of the lateral scan dimensions beyond the 2 mm used in this study was desired for large-scale applications (Fig. 11).

Fig. 11 In vivo cytoscope OCT imaging of normal human bladder (a) and human bladder cancer (c) with corresponding histology (b and d) has benefited from recent developments in endoscopic probe design incorporating MEMS-based technology. Labels: U, urothelium; LP, lamina propria; M, muscularis. Figure reprinted with permission from Wang et al. (2007b)

228

S.G. Adie and S.A. Boppart

Recent technology developments in sample probe design have enhanced the capabilities of OCT for in vivo detection of bladder cancers, enabling an endoscopic image quality comparable to that obtained with bench systems (Wang et al. 2007b). With incorporation of MEMS technology, lateral resolution was improved to ∼12 μm and lateral scan range was expanded beyond 4.5 mm. This enabled visualization of finer morphological details and the imaging of larger fields of view improved localization of cancer boundaries. The combination of fluorescence imaging with white-light cytoscopy has demonstrated promising results to improve the diagnostic value of white-light cytoscopy alone (Grossman et al. 2007). Fluorescence imaging has also been investigated for use in conjunction with photodynamic therapy (Chang et al. 1997). A multimodal approach incorporating fluorescence imaging has also been investigated for OCT (Wang et al. 2005). (See Section “Multimodal Optical Imaging” for further details.)

Skin Cancer The skin represents a prime opportunity for OCT imaging due to its accessibility to optical imaging. Translational oncological applications in dermatology have recently taken significant strides toward clinical use, largely due to technology refinement and commercial development. Two clinical conditions, basal cell carcinoma and malignant melanoma, have been the primary subject of these studies. Due to the poor survival rate associated with advanced malignant melanoma, early detection that permits differentiation between benign and malignant melanocytic lesions is one of the most important issues in dermatology. A feasibility study characterizing the features of basal cell carcinoma (BCC) was conducted using a commercial OCT system from LightLab Imaging that was modified for dermatological application (Olmedo et al. 2006). From the study of 49 lesions from 23 patients, morphological features of superficial, nodular, micronodular, and infiltrative islands of BCC that were observed in histological images correlated accurately with OCT. A follow-up study by the same group (Olmedo et al. 2007) compared the thickness of BCC measured with OCT to light microscopy measurements, obtaining excellent agreement. In a similar study, 43 BCCs were measured with a SkinDex 300 from ISIS Optronics with lateral and axial resolutions of 3 and 5 μm, respectively (Gambichler et al. 2007a). Loss of skin architecture and disarrangement of the epidermis and upper dermis were characteristic of BCC lesions, although no statistically significant difference was found between nodular, multifocal, superficial, and infiltrative BCCs. Two recent studies conducted with a SkinDex 300 system from ISIS Optronics investigated the feasibility of characterizing melanocytic lesions in vivo (Gambichler et al. 2007b; de Giorgi et al. 2005). The study by Gambichler et al. (2007b) measured 75 patients with 52 benign nevi and 40 malignant melanomas. The rete ridge architecture typically present at the dermoepidermal junction of

Optical Coherence Tomography for Cancer Detection

229

benign nevi was disrupted by malignant melanomas. More significantly, malignant melanomas were differentiable from benign nevi by the existence of “icicle-shaped” structures that corresponded to dense dermal infiltrates of melanocytes as confirmed by histology.

Oral Cancer The diagnosis of malignancy in the oral cavity is problematic due to the frequent lack of gross signs or obvious symptoms (Jung et al. 2005). Current diagnostic techniques require repeated surgical biopsy of benign lesions, with the documented number of advanced lesions outnumbering localized lesions by 2:1 (Jung et al. 2005; Wilder-Smith et al. 2004). Early detection of localized lesions results in a 5-year survival rate of 75% as compared to 16% for those with cancer metastasis (Jung et al. 2005; Wilder-Smith et al. 2004). Early-stage, noninvasive detection and monitoring of oral dysplasia could significantly improve the screening of high-risk populations and reduce morbidity and mortality. Translational OCT research into oral cancer has included both in vivo animal and in vivo human studies. The standard animal model for oral carcinogenesis is the hamster cheek pouch model. A study of 35 hamsters performed in vivo OCT imaging at all stages throughout carcinogenesis (Wilder-Smith et al. 2004). OCT images were classified for malignancy by two trained observers based on several architectural features such as changes in keratinization, epithelial thickening, epithelial proliferation and invasion, broadening of rete pegs, irregular epithelial stratification, and basal hyperplasia. Each observer was blinded to the other, but scored both OCT and histology images. (OCT and histology images were not matched to each other during the analysis to avoid bias.) Excellent sensitivity and specificity of 100 and 96%, respectively, for differentiating malignant from nonmalignant lesions was obtained. The overall agreement of OCT diagnosis with histopathological analysis was 80%. The incorporation of 3D acquisition and Doppler blood flow imaging enabled enhanced visualization and monitoring of vascular and perfusion changes (Jung et al. 2005; Wilder-Smith et al. 2005). The presence of small blood vessels was often seen in close proximity to tumor tissue (Wilder-Smith et al. 2005). Further improvements in OCT technology, in particular the attainment of cellular resolution, are expected to provide cytological level diagnostic capabilities. A recent study of 41 patients investigated the feasibility of distinguishing normal and pathological mucosa of the oral cavity and oropharynx (Ridgway et al. 2006). Imaging of normal mucosa revealed layers such as the epithelium, basement membrane, and lamina propria. Microstructures such as papillae, glands, ducts, and blood vessels were also detectable. Malignant lesions were associated with a disruption of tissue-layered architecture and elimination of the basement membrane. Imaging of lesion borders revealed transition zones between cancerous and normal layered mucosal architecture.

230

S.G. Adie and S.A. Boppart

The feasibility of OCT to evaluate laryngeal pathologies in vivo was investigated in a study involving 82 patients (Wong et al. 2005). The study provided baseline information on epithelial thickness, structure of lamina propria, and integrity of the basement membrane for a range of pathologies including microinvasive squamous cell carcinoma. The authors concluded that identifying the disruption of the basement membrane was the single greatest capability of OCT imaging, which could be utilized to characterize lesions, identify areas for biopsy, and perhaps aid earlier detection of malignant lesions by reducing biopsy sampling errors.

Computer-Aided Tissue Classification, Multimodal Imaging, and Contrast Enhancement Computer-Aided Diagnosis In order for OCT to reach widespread clinical use, low operator training requirements will be needed. As in other clinical imaging modalities (Gibbs and Turnbull 2003; Shankar et al. 2003), computer-aided tissue identification techniques are expected to assist in this transition. Additionally, given the data acquisition rates possible with state-of-the-art OCT systems, rigorous human interpretation of every image is not possible in real time. Thus computer algorithms become essential for real-time feedback during biopsy or surgical guidance procedures. Prior to the development of effective algorithms, it is necessary to have an understanding of the OCT image properties from each tissue type. Parameters that could be exploited for classification include image intensity and scattering coefficients, spatial frequency information, image texture, other known feature structures and contrasting mechanisms, for e.g., wavelength-dependent absorption or scattering, or tissue birefringence properties. This section discusses recent efforts on the development of computational algorithms for cancer detection with OCT and the various methods applied to validate their performance. A feasibility study demonstrated the potential of breast tissue classification based on spatial frequencies of OCT images (Zysk and Boppart 2006). Excised tissues from three patients were used as a training data set to isolate the OCT image characteristics of tumor (invasive ductal carcinoma), adipose, and stroma. The tissues were separated in the laboratory based on visual and tactile inspections and verified with H&E histology. Classification techniques for the identification of breast tumor tissue were applied to an OCT image (Fig. 5) from a fourth patient. A total of 1666 scan lines of adipose tissue, 1408 scan lines of tumor tissue, and 941 scan lines of stroma tissue were analyzed in this study. Figure 12 presents characteristic A-scans of breast tissues taken from the training data and their unique spatial frequency spectra. Two techniques were evaluated separately and in combination. The first technique took advantage of the unique Fourier-domain signature from each scan line, and the second exploited the periodic scattering response from adipose tissue. When applied to the task of tissue classification of A-scans

Optical Coherence Tomography for Cancer Detection

231

Fig. 12 Spatial (top) and Fourier spectral signatures (bottom) of different breast tissue types and representative histology (top insets)

within an OCT image, the combined technique yielded the optimal overall performance, with approximate tumor tissue (ductal carcinoma) sensitivity and specificity measurements of 97 and 68%, respectively. The estimated sensitivity surpasses that of conventional X-ray mammography and ultrasound, which have reported invasive ductal carcinoma detection sensitivities of approximately 81 and 94%, respectively (Berg et al. 2004). Due to the small sample size in this study, further work is required to evaluate the impact of biological variability on classification performance. A more extensive study of 260 excised breast tissue samples from 58 patients subsequently extended the number of classification parameters to include the A-scan slope and standard deviation (Goldberg et al. 2008). These parameters were applied to the differentiation of adipose from fibroglandular (stromal) tissue, for the purpose of reducing the number of nondiagnostic samples taken during fine-needle aspiration biopsy. Algorithms were trained on a set of 71 A-scans and validated on a set of 86 A-scans. A multivariate Gaussian model was used to calculate the probability that a test A-scan matched the class parameters of the training data set. The reported sensitivities and specificities of 98.1 and 82.5%, respectively, and an overall accuracy of 91.9% suggests that automated differentiation of adipose tissue from fibroglandular tissue is feasible.

232

S.G. Adie and S.A. Boppart

OCT-based computer-aided diagnosis has also been investigated for the detection of dysplasia in Barrett’s esophagus to enable real-time sampling of large surface areas (Qi et al. 2006). A total of 405 OCT images from 13 patients were acquired in vivo, from which 106 images satisfied the inclusion criteria for the study. Since loss of structure within OCT images was correlated to dysplasia (Isenberg et al. 2005), texture analysis methods (Gossage et al. 2003) were chosen to evaluate local intensity variations and homogeneity. Principle component analysis (PCA), utilizing six separate textural features, was employed to reduce the vector dimension while preserving information from each feature. The overall performance of the algorithm was evaluated by leave-one-out cross-validation and receiver-operating characteristic (ROC) curves. Although the sensitivity and specificity of 82 and 74%, respectively, and an accuracy of 83% do not indicate that OCT can replace standard surveillance, further technology refinements are expected to bring increased accuracy. The effects of improved axial resolution on tissue classification (based on textural analysis methods (Gossage et al. 2003) and PCA combined with linear discriminant analysis (LDA)) have been studied (Chen et al. 2008). Scatter plots of the first two linear discrimination functions showed significantly improved discrimination between Barrett’s and normal esophagus, which may also be promising for the detection of dysplasia. Significant further enhancement to computer-aided diagnosis is expected with additional resolution improvement that enables joint analysis of cellular and architectural features. Texture analysis methods have also been investigated for automated classification of bladder tissues (Lingley-Papadopoulos et al. 2008). A total of 182 OCT images from 68 different areas taken in vivo from 21 patients were compared to histology from co-located biopsies. Leave-one-out cross-validation, utilizing 18 textural features, was used to classify the OCT data sets as normal, exudative, dysplastic, or carcinoma in situ. The sensitivity and specificity of differentiating cancerous from noncancerous tissue was 92 and 62%, respectively. The high number of false positives was primarily due to the difficulty in distinguishing infiltrative inflammation from cancerous tissue, suggesting that further research is required to improve the specificity. With current OCT resolution and acquisition rates, computer-aided diagnosis has increasingly become important for screening large volumetric data sets for suspicious regions. Ongoing research will further develop and test CAD algorithms, as well as their implementation for real-time feedback.

Multimodal Optical Imaging Due to the constraints that real-time, high-resolution acquisition places on the OCT field of view, practical application of OCT to many oncological areas calls for localization and measurement of suspicious areas only. An efficient approach to this problem is a multimodal solution, using, for example, wide-field fluorescence imaging to guide high-resolution OCT measurements. In addition, the diagnostic

Optical Coherence Tomography for Cancer Detection

233

capability of OCT could potentially be improved by combining it with the biochemical sensitivity of fluorescence spectroscopy. Fluorescence signals measured from tissue can be endogenous (autofluorescence) or can be enhanced via the application of exogenous agents such as 5-aminolevulinic acid (5-ALA). Fluorescence imaging can be implemented in wide field to provide OCT guidance or in raster-scanned single-point acquisition to enhance the diagnostic capabilities of OCT. We give a brief account of studies pursuing these approaches for carcinoma of the cervix and vulva (Kuranov et al. 2002), bladder cancer (Pan et al. 2003; Wang et al. 2005), and colon cancer (Tumlinson et al. 2004; Hariri et al. 2006, 2007). A feasibility study of neoplasia in the cervix and vulva of six patients produced promising results for the determination of tumor boundaries (Kuranov et al. 2002). OCT and laser-induced fluorescence (LIF) imaging were performed under colposcopic control using separate imaging systems, 2.5–3 h after 5-ALA administration. Excitation was performed at 633 nm, and fluorescence spectra were measured over the range 500–900 nm. Two diagnostically significant bands at 675 and 700 nm were attributed to the presence of porphyrins and photoporphyrins, respectively. It was found that the tumor boundary detected by OCT and fluorescence imaging (and confirmed by histology) extended by about 2 mm beyond the colposcopically determined boundary. A subsequent paper by the same research group expanded the study to 21 patients (Sapozhnikova et al. 2005) and found that tumor boundaries as confirmed by histology were 1–4 mm farther than colposcopically determined boundaries. While lack of sufficient image information or false positives limited boundary detection probabilities to 77 and 79.5% for OCT and fluorescence spectroscopy, respectively, a combination of these optical methods increased the probability to 94.5%. The diagnostic capabilities of fluorescence-guided OCT (FGOCT) were investigated at several stages during the development of bladder transitional cell carcinomas in 54 Fisher rats (Wang et al. 2005). Large area ex vivo imaging (FOV=17×23 mm) of 5-ALA-induced fluorescence (local intravesical rather than systemic intravenous instillation) was used to select suspicious areas for separate OCT measurement. Bandwidths used for excitation and detection were 380–420 and 620–700 nm, respectively. The specificity of fluorescence detection was significantly improved from 53 to 93% with FGOCT, while the sensitivity was increased from 79% with fluorescence imaging to 100% with FGOCT. Development of an integrated probe (4 mm diameter) has demonstrated simultaneous OCT and fluorescence imaging, and further miniaturization (diameter of 2.8 mm) is expected to enable in vivo FGOCT of the bladder (Pan et al. 2003). The feasibility of combining OCT and laser-induced fluorescence (LIF) spectroscopy for distinguishing between normal, dysplastic colon, and inflammatory bowel disease (IBD) has been investigated in animal models (Hariri et al. 2007). Ex vivo imaging was performed on 30 normal mice and 10 mice each modeling colorectal cancer and IBD, respectively. Excitation wavelengths of 325 and 442 nm produced diagnostically significant endogeneous fluorescence peaks for IBD at 635 and 670 nm, which were attributed to increased porphyrin production by bacteria. This preliminary data suggest that the synergy of OCT and LIF could play a role

234

S.G. Adie and S.A. Boppart

in differentiating between the categories of normal, dysplasia, and IBD. In vivo application of OCT and LIF has been enabled by the development of a miniature endoscope (2 mm outer diameter) enabling simultaneous OCT and LIF measurements (Tumlinson et al. 2004). The LIF subsystem consisted of a single excitation fiber and two collection fibers alongside an OCT imaging fiber and GRIN lens. Both subsystems were cemented to a rod prism, and a pullback of the optical probe was performed within a thin-walled silica tube to generate images over a 6 mm field. This probe was utilized to investigate the feasibility of OCT–LIF imaging in mice under anesthesia, over several timepoints during the development of colon cancer (Hariri et al. 2006). In addition to morphological disruption of tissue boundary lines as measured with OCT, fluorescence excitation at 325 and 442 nm produced a decreased 405 nm intensity and a peak at 680 nm associated with progression of the disease. Further studies are warranted to fully assess the significance of these findings.

Enhancing Endogenous Contrast The diagnostic capability of OCT has been improved with the development of several extensions designed for functional imaging of tissues including refractive index, polarization, Doppler blood flow imaging, spectroscopy, and elastography. In addition to the native OCT contrast, these enhancement techniques enable imaging of tissue properties that are directly or indirectly modified by abnormal pathology. Development of such contrast enhancement mechanisms has paralleled that of other medical imaging modalities such as CT, MRI, and ultrasound. In this section, we briefly discuss these endogenous contrast enhancements poised for translation and how they could enhance cancer diagnosis. Refractive index variations and the subsequent scattering of incident light are the primary source of the OCT signal. On a cellular level, the refractive index has been shown to be sensitive to nuclear changes occurring as a result of cell mitosis (Boppart et al. 1998) and dysplasia (Gurjar et al. 2001), while on a bulk tissue scale it has been correlated with tumor malignancy (Das et al. 1997). A recent study in rat mammary tissues demonstrated the capability of distinguishing adipose from tumor or stromal tissue (Zysk et al. 2006). The group refractive index of a scattering sample can be measured by adapting the high-resolution imaging capabilities of OCT to perform two independent measurements allowing the optical and physical path lengths of the sample. Based on this concept, several techniques for measuring the refractive index have been proposed for OCT (Tearney et al. 1995; Zvyagin et al. 2003; Zysk et al. 2007a). Alternatively, the addition of LCI-based depth resolution has demonstrated refractive index measurements based on the Fresnel reflection coefficients from a needle tip in contact with a scattering sample (Zysk et al. 2007b). The development of needle-based designs for refractive index measurement in the breast (mainly consisting of adipose) could potentially reduce the number of nondiagnostic adipose samples taken during needle biopsy (Zysk et al. 2007a, b).

Optical Coherence Tomography for Cancer Detection

235

Polarization-sensitive OCT (PS-OCT) provides a contrast mechanism to potentially diagnose tissue composition or abnormalities, based on its polarization transforming properties (de Boer and Milner 2002). Birefringence in tissue emanates from regions of highly ordered macromolecular structures such as collagen fibers, cartilage, muscle, and the retinal nerve fiber layer and can be detected in a depthdependent manner with polarization-sensitive OCT (PS-OCT) systems. PS-OCT has been used to measure birefringence in the retinal nerve fiber layer (Cense et al. 2004), where it is believed that the increased pressure on the retina due to glaucoma reduces retinal birefringence. PS-OCT was used to determine the burn depth in rat skin, by sensing a reduction in birefringence caused by the denaturation of dermal collagen in response to thermal damage (Park et al. 2001). In oncology, PS-OCT has been investigated for the diagnosis of cervical intraepithelial neoplasms (Lee et al. 2008) and invasive basal cell carcinoma (Strasswimmer et al. 2004). The fibrous stroma (for example, of the breast) is composed largely of collagen fibers, which is expected to be distinct from scattering structures of tumors. The detection and quantification of polarization transformation with PS-OCT also has the potential to provide improved discrimination between tumor and stromal tissue types. Spectroscopic OCT utilizes the broadband property of the light source to perform spatially localized spectroscopy over the source optical bandwidth. Spectroscopic OCT provides the capability to image wavelength-dependent scattering or absorption, potentially enabling functional imaging and molecular contrast. Spectroscopic OCT was demonstrated over a 650–1000 nm wavelength range using an ultrabroadband Ti:Al2 O3 laser source, providing enhanced contrast of melanocytes in mesenchymal cells of the African frog tadpole, Xenopus laevis (Morgner et al. 2000). Optical sources centered about the 800 nm isosbestic point of hemoglobin have enabled spatially resolved, relative oxygenation saturation measurements in whole blood (Lu et al. 2008; Faber et al. 2005). An important property to note is that spectroscopic OCT imaging is governed by the time–frequency relation, effectively imposing a trade-off between spectral and spatial resolutions (Oldenburg et al. 2007; Graf and Wax 2007; Xu et al. 2005). This compromise can be avoided (at the expense of a narrow depth-of-field) by performing spectroscopic OCM (Xu et al. 2006). Molecular contrast imaging can be performed using a pump-probe technique, which has been used to enhance absorption-based contrast, where a target molecule pumped into a transient state has increased absorption for the OCT probe beam (Rao et al. 2003). The studies highlighted here provide the foundation and framework for application of spectroscopic OCT to the detection or monitoring of tumorigenesis. Doppler OCT measures flow in blood vessels via a Doppler shift in the interferogram fringe carrier frequency (Izatt et al. 1997; Zvyagin et al. 2000) or by processing sequential phase-resolved A-scans (Zhao et al. 2000). Doppler OCT has also been applied to measuring blood flow in the retina (Leitgeb et al. 2003) and in skin (Barton et al. 1999). Doppler OCT has also been investigated for monitoring microvasculature tissue response during photodynamic therapy in animal models (Standish et al. 2007b, a), including with a needle design to extend the depth coverage to interstitial spaces (Li et al. 2006). Preliminary experience has also been gained with endoscopic Doppler OCT of the gastrointestinal tract (Yang

236

S.G. Adie and S.A. Boppart

et al. 2005). In the future, Doppler OCT could play a greater role in monitoring tumor vasculature and angiogenesis in vivo. OCT elastography, originally introduced by Schmitt, measures local tissue deformations under applied mechanical pressure (Schmitt 1998). High-resolution maps of local tissue deformation obtained by OCT elastography essentially provide a “highresolution palpation” capability. Conceptually this follows the development of elastography techniques for magnetic resonance and ultrasound imaging (Greenleaf et al. 2003; Gao et al. 1996). Traditionally OCT elastography employs quasi-static mechanical compression and speckle tracking to obtain spatially resolved tissue displacements (Schmitt 1998; Kirkpatrick et al. 2006), but recent phase-sensitive SDOCT methods (Wang et al. 2007a) potentially enable the measurement of dynamic mechanical excitation. The initial application that OCT elastography has been investigated for is the detection of vulnerable arterial plaques (Rogowska et al. 2006; 2004; Chan et al. 2004). OCT elastography has been used to obtain strain maps in engineered tissues in vitro and the X. laevis tadpole in vivo (Ko et al. 2006). Tumors are often detected via increased stiffness relative to the surrounding tissues. As measured with other imaging modalities such as ultrasound and MRI-based elastography, Young’s elastic modulus (the ratio of stress to strain) may vary by up to 4 orders of magnitude in soft tissues; and breast tumors may vary from that of surrounding tissue by up to a factor of 90 (Greenleaf et al. 2003; Sarvazyan et al. 1998). A recent study demonstrated that optical coherence elastography can map dynamic elastic moduli of normal and neoplastic human breast tissue ex vivo (Liang et al. 2008).

Exogenous Contrast Agents for OCT When imaging biological tissues, it is often desirable to enhance the signals measured from specific structures. Contrast agents which produce a specific image signature have been utilized in virtually every imaging modality including ultrasound, CT, MRI, nuclear medicine, and optical microscopy, among many others. Recently, new engineered contrast agents and molecular contrast techniques specifically designed for OCT have been developed and characterized (Boppart et al. 2005; Yang 2005). Since OCT detects only coherent light, exogenous contrast is obtained using optical contrast agents that alter the local scattering or absorption properties or that are dynamically excited to produce changes in the amplitude or phase of the light. Submicron particles (>200 nm) introduced into the circulatory or lymphatic systems typically remain in circulation until cleared by the reticuloendothelial system, while nanoparticles and molecules (<100 nm) are able to extravasate and provide contrast enhancement outside the vasculature. Agents may be functionalized with antibodies or molecules to target them to specific molecules, cells, or tissue types and thus provide additional selectivity that can enhance the utility of OCT as an emerging diagnostic technique. Scattering-based contrast agents produce strong scattering due to significantly different refractive index from tissue. Perfluorocarbon-filled albumin microspheres and nanoparticles of gold, melanin, carbon, and iron oxide were found to produce

Optical Coherence Tomography for Cancer Detection

237

strong scattering in OCT (Lee et al. 2003; Agrawal et al. 2006; Barton et al. 2002). In vivo measurements of mouse liver after tail-vein injection of nontargeted goldcoated protein microspheres demonstrated increased signal scattering from liver, which highlighted low scattering from the liver vasculature (Lee et al. 2003). In other in vivo measurements using microspheres with encapsulated iron-oxide, scattering variations were observed in vascular regions of exposed mouse intestinal wall immediately after tail-vein injection, with minimal changes in an avascular region (Boppart et al. 2005). With further work, these agents could be used to highlight tumors that have increased vasculature due to angiogenesis. Absorption-based contrast agents are selected to be spectroscopically active over the optical bandwidth of the OCT light source. Spectroscopic OCT using agents with sharp, stable absorption peaks within the source optical bandwidth can utilize differential absorption between “active” and “control” regions of the spectrum to deduce relative dye concentrations. This is because the negative contrast provided by a spectroscopically neutral absorber that is spatially localized within a heterogeneous sample is not easily distinguishable from the inherent morphology of the sample. Candidates include near-infrared dyes with well-characterized absorption spectra, or plasmon-resonant nanoparticles or gold nanocages with tunable peak absorption can be utilized (Boppart et al. 2005; Oldenburg et al. 2006; Cang et al. 2005). In one study, the concentration of the FDA-approved near-infrared dye indocyanine green was mapped within a stage 54 X. laevis (Yang et al. 2004). In another study, spectroscopic OCT was used to localize the distribution of a NIR dye, with strong absorption over the shorter half of the source optical spectrum, in the vascular system of a botanical specimen (Xu et al. 2004). The use of plasmon-resonant gold nanoparticles could also serve as multifunctional therapeutic agents, utilizing their strong, tunable absorption spectra to induce local hyperthermia in cells and tissues. Modulating probes provide increased discrimination from stationary background signals. An example of this from fluorescence microscopy is magnetically controlled, blinking fluorescent probes (Anker and Kopelman 2003). This advantage can also be obtained in OCT by using novel dynamic contrast agents that are physically modulated in space using an external magnetic field (Oldenburg et al. 2005b, a). Iron oxide such as magnetite is suitable because of its high magnetic susceptibility, ferromagnetic property, and known biocompatibility after polymer coating. Magnetomotive contrast, manifest via scattering changes in the OCT image, was demonstrated in bulk three-dimensional cell scaffolds containing macrophages labeled with microparticles of iron oxide (Oldenburg et al. 2005). This work raises interesting possibilities of combining large field of view MRI molecular contrast imaging with small field of view, high-resolution magnetomotive OCT imaging. Modulated molecular contrast can also be obtained in OCT via a pump-probe technique (Rao et al. 2003). The location of a target molecule population, with an absorption spectrum temporarily modified by a separate optical pump beam, is obtained via increased absorption of the OCT probe beam. Toggling the power of the optical pump beam produces modulation of absorption contrast. The technique has been used with the FDA-approved indocyanine green (Yaqoob et al. 2006). It has

238

S.G. Adie and S.A. Boppart

been adapted to measure the transient absorption of fully allowed electronic transitions on very short timescales before a steady state is achieved (Applegate and Izatt 2006). Imaging this shorter ground state recovery time can be used to separate contributions from multiple chromophores. This was demonstrated on a mixture of human whole blood and rhodamine 6G.

Barriers to Clinical Application Successful translation of technologies for medical use typically follows phases ranging from system development and early feasibility studies, to regulatory approval and larger clinical trials, and eventually to adoption as the standard of care. During this process, progression to the next phase is largely dependent upon the demonstration of safety and efficacy in the previous phase. Past experiences can highlight certain barriers that can impede the timely evaluation of device performance during various phases of the translation. These barriers to the translation of OCT technology can be categorized into cultural and logistical (Zysk et al. 2007c) barriers. Logistical barriers can be further broken down into geographic and administrative, technology transfer and intellectual property management, and operational barriers associated with the clinical adoption of the new technology. Translation of technologies to the clinic requires a successful match between technological innovation and a clinical problem. This is achieved through cooperation between physical scientists or engineers, medical professionals, and industry. Cultural barriers can often impede progress through the stages of translation, due to the wide differences in the intellectual content between disciplines and differing career expectations and interests. Dialogue between physical scientists or engineers and physicians can be difficult (at least initially) because of the different technical language used by each discipline. Efforts required during the translational research cycle (particularly the middle phases) often do not align with the career focus of academic scientists and engineers who develop new technologies, and physicians who treat patients using proven methods and technologies. Cultural differences can be addressed at an early stage through multidisciplinary education, both at the undergraduate and graduate levels, to impart the value of interdisciplinary synergy. The fostering of relationships between medical departments at teaching hospitals, and physics or engineering departments can also help bridge gaps between these communities. Physician-scientists or physician-engineers are uniquely poised to address these cultural barriers. Institutions that have a significant research focus and ties to academic institutions (such as joint appointments) can provide the setting to overcome these cultural barriers. Further opportunities for cross-disciplinary interaction could be facilitated by more conferences specially designed to attract both engineers and clinicians. This will provide clinicians with a greater appreciation of stateof-the-art technology and engineers with better exposure to outstanding clinical problems. Logistical constraints can be geographic and administrative in nature, pertaining to the physical locations of technology development and deployment for clinical

Optical Coherence Tomography for Cancer Detection

239

feasibility studies. Medical campuses where patient care is delivered may not always be close to academic institutions where the technology is developed, and thus are not always ideal for cross-institutional collaboration. This can be addressed by the allocation of physical space at both locations to encourage interaction. Crossinstitutional centers or institutes with a focus on translational research, that are in close proximity to hospitals delivering patient care, can provide the ideal environment for patient recruiting, imaging studies, data analysis, and collaboration between interdisciplinary research team members. Administrative constraints can frequently present challenges to successful translation research activities as many existing institutions and administrative units were developed under the traditional concept of discrete academic departments. Translational research that spans multiple academic departments or colleges, or even different institutions, often must develop new paths to encourage and support translational activities. Recognizing the value and awarding successful investigators for this type of academic pursuit is also a current topic of discussion for many promotion and tenure committees at academic institutions. The increasing development of interdisciplinary institutes or centers has begun to reduce many of these barriers. Funding translational research activities can at times be problematic. Because this type of research does not always begin with hypothesis-driven questions to address basic science issues, and because translational research is not commonly at a stage of development for large randomized controlled clinical trials, researchers in this area are often confronted with the challenge to promote more technologydriven proposals. Fortunately, major federal funding agencies in the United States and elsewhere have recognized the significant potential benefit of translational research activities and have taken steps to directly target and support these types of studies. In the United States, intellectual property control over federally funded research is granted to universities by the Bayh–Dole act. However, since university research focuses on scientific discovery and development of new technologies, institutions may not always have the resources to devote to technology translation. Because the transfer of technology from the research laboratory into the private sector is such an important step in the translational research cycle, investigators must often closely align themselves with institutional technology transfer offices, and become familiar with the technology transfer process. Management of intellectual property involving multiple institutions can present complex legal implications. However, significant value can be added by industry input through prototype development for large-scale clinical trials, regulatory approval logistics, and subsequent commercialization for clinical deployment. Early involvement of all stakeholders in clinical trials can maximize the resources and funding available for translation. Dedicated research parks with venture capital funding opportunities can help expedite and guide university researchers through the process. Relationships between university departments and industry can be fostered through programs to train students to work with industry as part of their degree requirements. Operational barriers can arise from the new procedures and expertise requirements introduced into the clinical setting. This includes the training of medical

240

S.G. Adie and S.A. Boppart

staff and experts for operating and interpreting OCT images, and the establishment of standard operating procedures. Standardization of OCT system performance will become important to obtain reproducible images that are interpretable according to commonly established criteria. Revision of these criteria should ideally be conducted in the earlier stages of translation, where OCT system upgrades (e.g., improved resolution or faster acquisition) are more easily deployed for clinical studies. Standard operating procedures relate to the practical role that OCT will play and its impact have on the standard of care. In other words, how will the extra diagnostic information alter the current standard of care? Health insurance reimbursement will depend on the added diagnostic value of the technology, how it affects the standard of care, and how it improves the patient treatment and outcome, as well as the cost of OCT imaging procedures. Awareness of these existing barriers is important so that solutions can be implemented and adopted. Because translational research activities represent a major cultural shift in the research enterprise, it is likely that a generational change will have to occur. For now, this change can begin with exposing the next generation of physician-investigators to the merits of translational research, and the potential it has to significantly impact our quality of healthcare.

Summary and Outlook on Clinical Adoption Optical coherence tomography provides unique capabilities within the medical imaging landscape. Combining image resolutions approaching that of conventional histology, penetration depths of 1–3 mm, and various contrasting mechanisms, it has the ability to probe tissue morphology as well as function in a wide array of applications. A wealth of research demonstrating its potential for cancer detection suggests that it is well poised to make a significant impact in multiple branches of oncology. The array of ongoing OCT research supporting its translation to the clinic ranges from technology development, to the development of contrasting mechanisms, to large-scale clinical studies. The OCT technology has the flexibility to be applied to many branches in oncology including the gastrointestinal tract, dermatology, and open breast surgery, to name only a few. This flexibility also provides the potential to contribute to multiple stages of clinical care such as screening, biopsy guidance and diagnosis, endoscopic or open surgical guidance, and monitoring of treatments. In order for OCT to enter the mainstream standard of care, it must demonstrate added diagnostic value in large-scale clinical trials that is equal or better than current clinical methods. Regulatory approvals for the use of medical devices in the clinical standard of care require that the approved uses be well developed and validated. Therefore, OCT imaging will need to be validated for specific procedures, rather than a clinical field of application, and standard operating protocols will need to be developed. Technology translation can be a lengthy process, and it is useful to overview developments in the commercial OCT landscape to get a sense of the time frame

Optical Coherence Tomography for Cancer Detection

241

for its adoption into clinical practice. While the entry point for the engagement of industry reflects the status of the field, it also serves to accelerate the translation process itself. The first commercial ophthalmic OCT systems were introduced by Carl Zeiss Meditec in 1996. They gained FDA approval in 2002 for retinal imaging and, as of 2007, have sold in excess of 6000 units of its Stratus OCTTM systems worldwide (Zysk et al. 2007c). In 2006, Heidelberg Engineering received FDA clearance for anterior segment imaging. This suggests a time frame of 6–10 years to obtain regulatory approval after the appearance of the first commercial systems targeting specific procedures. Representative OCT companies with commercial systems targeting oncological applications include Imalux Corporation (Cleveland, Ohio), ISIS Optronics GmbH (Mannheim, Germany), LightLab Imaging (Westford, Massachusetts), Novacam Technologies, Incorporated (Pointe-Claire, Canada), OCT Medical Imaging, Incorporated (Irvine, California), and Lantis Laser Incorporated (Denville, New Jersey). Other representative companies marketing OCT systems for research include Bioptigen, Inc. (Research Triangle Park, North Carolina), offering a broad selection of application-specific scanners with their 840 and 1310 nm systems, and Thorlabs, Inc., with systems that also incorporate Doppler and polarization-sensitive imaging. In perspective, the development and clinical adoption of OCT is following a timeline not unlike that which occurred for ultrasound, CT, SPECT/PET, and MR imaging. OCT has already entered the clinical standard of care in ophthalmology, and applications in interventional cardiology are poised to be the next area for clinical adoption. Oncological applications will likely soon follow after ongoing clinical trials rigorously demonstrate and validate its diagnostic capabilities. The achievements reviewed in this chapter, the rate of progress of the field, and the rapid commercial developments suggest that OCT is poised to become a major medical imaging modality for clinical oncology over the next decade. Acknowledgments We wish to thank all the past and current members of the Biophotonics Imaging Laboratory at the Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, for their technical contributions and physical efforts toward the development and clinical translation of OCT and related optical biomedical imaging technologies. We also wish to thank all of our colleagues and collaborators that have advanced this field to the state of the art. We thank those that have contributed images and data and apologize to those who could not be represented due to limited publication space. The work from the Biophotonics Imaging Laboratory was supported in part by grants from the National Institutes of Health (Roadmap Initiative/NIBIB R21 EB005321, NIBIB R01 EB005221, NCI R21/R33 CA115536), the National Science Foundation (BES 03-47747, BES 05-19920), Carle Foundation Hospital, and the Grainger Foundation. Additional information can be found at http://biophotonics. illinois.edu.

References Adie, S. G., T. R. Hillman, and D. D. Sampson. 2007. Detection of multiple scattering in optical coherence tomography using the spatial distribution of Stokes vectors. Optics Express 15 (26):18033–18049.

242

S.G. Adie and S.A. Boppart

Adler, D. C., Y. Chen, R. Huber, J. Schmitt, J. Connolly, and J. G. Fujimoto. 2007. Threedimensional endomicroscopy using optical coherence tomography. Nature Photonics 1 (12): 709–716. Agrawal, A., S. Huang, A. W. H. Lin, M. H. Lee, J. K. Barton, R. A. Drezek, and T. J. Pfefer. 2006. Quantitative evaluation of optical coherence tomography signal enhancement with gold nanoshells. Journal of Biomedical Optics 11 (4):041121. Aguirre, A. D., N. Nishizawa, J. G. Fujimoto, W. Seitz, M. Lederer, and D. Kopf. 2006. Continuum generation in a novel photonic crystal fiber for ultrahigh resolution optical coherence tomography at 800 nm and 1300 nm. Optics Express 14 (3):1145–1160. Anker, J. N., and R. Kopelman. 2003. Magnetically modulated optical nanoprobes. Applied Physics Letters 82 (7):1102–1104. Applegate, B. E., and J. A. Izatt. 2006. Molecular imaging of endogenous and exogenous chromophores using ground state recovery pump-probe optical coherence tomography. Optics Express 14 (20):9142–9155. Armstrong, J. J., M. S. Leigh, D. D. Sampson, J. H. Walsh, D. R. Hillman, and P. R. Eastwood. 2006. Quantitative upper airway imaging with anatomic optical coherence tomography. American Journal of Respiratory and Critical Care Medicine 173 (2):226–233. Barton, J. K., J. B. Hoying, and C. J. Sullivan. 2002. Use of microbubbles as an optical coherence tomography contrast agent. Academic Radiology 9:S52–S55. Barton, J. K., J. A. Izatt, M. D. Kulkarni, S. Yazdanfar, and A. J. Welch. 1999. Three-dimensional reconstruction of blood vessels from in vivo color Doppler optical coherence tomography images. Dermatology 198 (4):355–361. Berg, W. A., L. Gutierrez, M. S. NessAiver, W. B. Carter, M. Bhargavan, R. S. Lewis, and O. B. Ioffe. 2004. Diagnostic accuracy of mammography, clinical examination, US, and MR imaging in preoperative assessment of breast cancer. Radiology 233 (3):830–849. Bohringer, H. J., D. Boller, J. Leppert, U. Knopp, E. Lankenau, E. Reusche, G. Huttmann, and A. Giese. 2006. Time-domain and spectral-domain optical coherence tomography in the analysis of brain tumor tissue. Lasers in Surgery and Medicine 38 (6):588–597. Boppart, S. A., B. E. Bouma, C. Pitris, J. F. Southern, M. E. Brezinski, and J. G. Fujimoto. 1998. In vivo cellular optical coherence tomography imaging. Nature Medicine 4 (7):861–865. Boppart, S. A., B. E. Bouma, C. Pitris, G. J. Tearney, J. G. Fujimoto, and M. E. Brezinski. 1997. Forward-imaging instruments for optical coherence tomography. Optics Letters 22 (21): 1618–1620. Boppart, S. A., W. Luo, D. L. Marks, and K. W. Singletary. 2004. Optical coherence tomography: feasibility for basic research and image-guided surgery of breast cancer. Breast Cancer Research and Treatment 84 (2):85–97. Boppart, S. A., A. L. Oldenburg, C. Y. Xu, and D. L. Marks. 2005. Optical probes and techniques for molecular contrast enhancement in coherence imaging. Journal of Biomedical Optics 10 (4):041208. Bouma, B. E., and G. J. Tearney. 2002. Handbook of Optical Coherence Tomography. New York: Marcel Dekker. Bouma, B. E., G. J. Tearney, H. Yabushita, M. Shishkov, C. R. Kauffman, D. D. Gauthier, B. D. MacNeill, S. L. Houser, H. T. Aretz, E. F. Halpern, and I. K. Jang. 2003. Evaluation of intracoronary stenting by intravascular optical coherence tomography. Heart 89 (3):317–320. Brezinski, M. E., and J. G. Fujimoto. 1999. Optical coherence tomography: High-resolution imaging in nontransparent tissue. IEEE Journal of Selected Topics in Quantum Electronics 5 (4): 1185–1192. Cang, H., T. Sun, Z. Y. Li, J. Y. Chen, B. J. Wiley, Y. N. Xia, and X. D. Li. 2005. Gold nanocages as contrast agents for spectroscopic optical coherence tomography. Optics Letters 30 (22): 3048–3050. Cellini, C., S. T. Hollenbeck, P. Christos, D. Martins, J. Carson, S. Kemper, E. LaVigne, E. Chan, and R. Simmons. 2004. Factors associated with residual breast cancer after re-excision for close or positive margins. Annals of Surgical Oncology 11 (10):915–920.

Optical Coherence Tomography for Cancer Detection

243

Cense, B., H. C. Chen, B. H. Park, M. C. Pierce, and J. F. de Boer. 2004. In vivo birefringence and thickness measurements of the human retinal nerve fiber layer using polarization-sensitive optical coherence tomography. Journal of Biomedical Optics 9 (1):121–125. Chan, R. C., A. H. Chau, W. C. Karl, S. Nadkarni, A. S. Khalil, N. Iftimia, M. Shishkov, G. J. Tearney, M. R. Kaazempur-Mofrad, and B. E. Bouma. 2004. OCT-based arterial elastography: robust estimation exploiting tissue biomechanics. Optics Express 12 (19): 4558–4572. Chang, S. C., G. Buonaccorsi, A. J. MacRobert, and S. G. Bown. 1997. 5-Aminolevulinic acid (ALA)-induced protoporphyrin IX fluorescence and photodynamic effects in the rat bladder: An in vivo study comparing oral and intravesical ALA administration. Lasers in Surgery and Medicine 20 (3):254–264. Chen, Y., A. D. Aguirre, P. L. Hsiung, S. Desai, P. R. Herz, M. Pedrosa, Q. Huang, M. Figueiredo, S. W. Huang, A. Koski, J. M. Schmitt, J. G. Fujimoto, and H. Mashimo. 2007. Ultrahigh resolution optical coherence tomography of Barrett’s esophagus: preliminary descriptive clinical study correlating images with histology. Endoscopy 39 (7):599–605. Chen, Y., A. D. Aguirre, P. L. Hsiung, S. W. Huang, H. Mashimo, J. M. Schmitt, and J. G. Fujimoto. 2008. Effects of axial resolution improvement on optical coherence tomography (OCT) imaging of gastrointestinal tissues. Optics Express 16 (4):2469–2485. Choma, M. A., M. V. Sarunic, C. H. Yang, and J. A. Izatt. 2003. Sensitivity advantage of swept source and Fourier domain optical coherence tomography. Optics Express 11 (18):2183–2189. Das, B. B., F. Liu, and R. R. Alfano. 1997. Time-resolved fluorescence and photon migration studies in biomedical and model random media. Reports on Progress in Physics 60 (2): 227–292. de Boer, J. F., B. Cense, B. H. Park, M. C. Pierce, G. J. Tearney, and B. E. Bouma. 2003. Improved signal-to-noise ratio in spectral-domain compared with time-domain optical coherence tomography. Optics Letters 28 (21):2067–2069. de Boer, J. F., and T. E. Milner. 2002. Review of polarization sensitive optical coherence tomography and Stokes vector determination. Journal of Biomedical Optics 7 (3):359–371. de Giorgi, V., M. Stante, D. Massi, L. Mavilia, P. Cappugi, and P. Carli. 2005. Possible histopathologic correlates of dermoscopic features in pigmented melanocytic lesions identified by means of optical coherence tomography. Experimental Dermatology 14 (1):56–59. Drexler, W. 2004. Ultrahigh-resolution optical coherence tomography. Journal of Biomedical Optics 9 (1):47–74. Drezek, R., M. Guillaud, T. Collier, I. Boiko, A. Malpica, C. Macaulay, M. Follen, and R. RichardsKortum. 2003. Light scattering from cervical cells throughout neoplastic progression: influence of nuclear morphology, DNA content, and chromatin texture. Journal of Biomedical Optics 8 (1):7–16. Escobar, P. F., L. Rojas-Espaillat, S. Tisci, C. Enerson, J. Brainard, J. Smith, N. J. Tresser, F. I. Feldchtein, L. B. Rojas, and J. L. Belinson. 2006. Optical coherence tomography as a diagnostic aid to visual inspection and colposcopy for preinvasive and invasive cancer of the uterine cervix. International Journal of Gynecological Cancer 16 (5):1815–1822. Faber, D. J., E. G. Mik, M. C. G. Aalders, and T. G. van Leeuwen. 2005. Toward assessment of blood oxygen saturation by spectroscopic optical coherence tomography. Optics Letters 30 (9):1015–1017. Fercher, A. F., W. Drexler, C. K. Hitzenberger, and T. Lasser. 2003. Optical coherence tomography – principles and applications. Reports on Progress in Physics 66 (2):239–303. Fisher, B., S. Anderson, J. Bryant, R. G. Margolese, M. Deutsch, E. R. Fisher, J-H. Jeong, and N. Wolmark. 2002. Twenty-year follow-up of a randomized trial comparing total mastectomy, lumpectomy, and lumpectomy plus irradiation for the treatment of invasive breast cancer. The New England Journal of Medicine 347 (16):1233–1241. Frable, W. J. 1983. Fine-needle aspiration biopsy – a review. Human Pathology 14 (1):9–28. Fujimoto, J. G. 2003. Optical coherence tomography for ultrahigh resolution in vivo imaging. Nature Biotechnology 21 (11):1361–1367.

244

S.G. Adie and S.A. Boppart

Gambichler, T., A. Orlikov, R. Vasa, G. Moussa, K. Hoffmann, M. Stucker, P. Altmeyer, and F. G. Bechara. 2007a. In vivo optical coherence tomography of basal cell carcinoma. Journal of Dermatological Science 45 (3):167–173. Gambichler, T., P. Regeniter, F. G. Bechara, A. Orlikov, R. Vasa, G. Moussa, M. Stucker, P. Altmeyer, and K. Hoffmann. 2007b. Characterization of benign and malignant melanocytic skin lesions using optical coherence tomography in vivo. Journal of the American Academy of Dermatology 57 (4):629–637. Gao, L., K. J. Parker, R. M. Lerner, and S. F. Levinson. 1996. Imaging of the elastic properties of tissue – a review. Ultrasound in Medicine and Biology 22 (8):959–977. Gibbs, P., and L. W. Turnbull. 2003. Textural analysis of contrast-enhanced MR images of the breast. Magnetic Resonance in Medicine 50 (1):92–98. Goldberg, B. D., N. V. Iftimia, J. E. Bressner, M. B. Pitman, E. Halpern, B. E. Bouma, and G. J. Tearney. 2008. Automated algorithm for differentiation of human breast tissue using low coherence interferometry for fine needle aspiration biopsy guidance. Journal of Biomedical Optics 13 (1):014014. Gossage, K. W., T. S. Tkaczyk, J. J. Rodriguez, and J. K. Barton. 2003. Texture, analysis of optical coherence tomography images: feasibility for tissue classification. Journal of Biomedical Optics 8 (3):570–575. Graf, R. N., and A. Wax. 2007. Temporal coherence and time-frequency distributions in spectroscopic optical coherence tomography. Journal of the Optical Society of America A-Optics Image Science and Vision 24 (8):2186–2195. Greenleaf, J. F., M. Fatemi, and M. Insana. 2003. Selected methods for imaging elastic properties of biological tissues. Annual Review of Biomedical Engineering 5:57–78. Grossman, H. B., L. Gomella, Y. Fradet, A. Morales, J. Presti, C. Ritenour, U. Nseyo, M. J. Droller, and Pc B30201 Study Grp. 2007. A phase III, multicenter comparison of hexaminolevulinate fluorescence cystoscopy and white light cystoscopy for the detection of superficial papillary lesions in patients with bladder cancer. Journal of Urology 178 (1):62–67. Gurjar, R. S., V. Backman, L. T. Perelman, I. Georgakoudi, K. Badizadegan, I. Itzkan, R. R. Dasari, and M. S. Feld. 2001. Imaging human epithelial properties with polarized light-scattering spectroscopy. Nature Medicine 7 (11):1245–1248. Hariri, L. P., A. R. Tomlinson, N. H. Wade, D. G. Besselsen, U. Utzinger, E. W. Gerner, and J. K. Barton. 2007. Ex vivo optical coherence tomography and laser-induced fluorescence spectroscopy imaging of murine gastrointestinal tract. Comparative Medicine 57 (2):175–185. Hariri, L. P., A. R. Tumlinson, D. G. Besselsen, U. Utzinger, E. W. Gerner, and J. K. Barton. 2006. Endoscopic optical coherence tomography and laser-induced fluorescence spectroscopy in a murine colon cancer model. Lasers in Surgery and Medicine 38 (4):305–313. Hsiung, P. L., L. Pantanowitz, A. D. Aguirre, Y. Chen, D. Phatak, T. H. Ko, S. Bourquin, S. J. Schnitt, S. Raza, J. L. Connolly, H. Mashimo, and J. G. Fujimoto. 2005. Ultrahighresolution and 3-dimensional optical coherence tomography ex vivo imaging of the large and small intestines. Gastrointestinal Endoscopy 62 (4):561–574. Hsiung, P. L., D. R. Phatak, Y. Chen, A. D. Aguirre, J. G. Fujimoto, and J. L. Connolly. 2007. Benign and malignant lesion in the human breast depicted with ultrahigh resolution and dimensional optical coherence tomography. Radiology 244 (3):865–874. Huang, D., E. A. Swanson, C. P. Lin, J. S. Schuman, W. G. Stinson, W. Chang, M. R. Hee, T. Flotte, K. Gregory, C. A. Puliafito, and J. G. Fujimoto. 1991. Optical Coherence Tomography. Science 254 (5035):1178–1181. Iftimia, N. V., B. E. Bouma, M. B. Pitman, B. Goldberg, J. Bressner, and G. J. Tearney. 2005. A portable, low coherence interferometry based instrument for fine needle aspiration biopsy guidance. Review of Scientific Instruments 76 (6):064301. Isenberg, G., M. V. Sivak, A. Chak, R. C. K. Wong, J. E. Willis, B. Wolf, D. Y. Rowland, A. Das, and A. Rollins. 2005. Accuracy of endoscopic optical coherence tomography in the detection of dysplasia in Barrett’s esophagus: a prospective, double-blinded study. Gastrointestinal Endoscopy 62 (6):825–831.

Optical Coherence Tomography for Cancer Detection

245

Izatt, J. A., M. D. Kulkami, S. Yazdanfar, J. K. Barton, and A. J. Welch. 1997. In vivo bidirectional color Doppler flow imaging of picoliter blood volumes using optical coherence tomograghy. Optics Letters 22 (18):1439–1441. Jackle, S., N. Gladkova, F. Feldchtein, A. Terentieva, B. Brand, G. Gelikonov, V. Gelikonov, A. Sergeev, A. Fritscher-Ravens, J. Freund, U. Seitz, S. Schroder, and N. Soehendra. 2000. In vivo endoscopic optical coherence tomography of the human gastrointestinal tract – toward optical biopsy. Endoscopy 32 (10):743–749. Jung, W. G., J. Zhang, J. R. Chung, P. Wilder-Smith, M. Brenner, J. S. Nelson, and Z. P. Chen. 2005. Advances in oral cancer detection using optical coherence tomography. IEEE Journal of Selected Topics in Quantum Electronics 11 (4):811–817. Kawakami-Wong, H., S. G. Gu, M. J. Hammer-Wilson, J. B. Epstein, Z. P. Chen, and P. WilderSmith. 2007. In vivo optical coherence tomography-based scoring of oral mucositis in human subjects: a pilot study. Journal of Biomedical Optics 12 (5):051702. Kirkpatrick, S. J., R. K. Wang, and D. D. Duncan. 2006. OCT-based elastography for large and small deformations. Optics Express 14 (24):11585–11597. Ko, H. J., W. Tan, R. Stack, and S. A. Boppart. 2006. Optical coherence elastography of engineered and developing tissue. Tissue Engineering 12 (1):63–73. Kuranov, R. V., V. V. Sapozhnikova, H. M. Shakhova, V. M. Gelikonov, E. V. Zagainova, and S. A. Petrova. 2002. Combined application of optical methods to increase the information content of optical coherent tomography in diagnostics of neoplastic processes. Quantum Electronics 32 (11):993–998. Lee, T. M., A. L. Oldenburg, S. Sitafalwalla, D. L. Marks, W. Luo, F. J. J. Toublan, K. S. Suslick, and S. A. Boppart. 2003. Engineered microsphere contrast agents for optical coherence tomography. Optics Letters 28 (17):1546–1548. Lee, S. W., J. Y. Yoo, J. H. Kang, M. S. Kang, S. H. Jung, Y. Chong, D. S. Cha, K. H. Han, and B. M. Kim. 2008. Optical diagnosis of cervical intraepithelial neoplasm (CIN) using polarization-sensitive optical coherence tomography. Optics Express 16 (4): 2709–2719. Leitgeb, R. A., L. Schmetterer, W. Drexler, A. F. Fercher, R. J. Zawadzki, and T. Bajraszewski. 2003. Real-time assessment of retinal blood flow with ultrafast acquisition by color Doppler Fourier domain optical coherence tomography. Optics Express 11 (23):3116–3121. Lexer, F., C. K. Hitzenberger, W. Drexler, S. Molebny, H. Sattmann, M. Sticker, and A. F. Fercher. 1999. Dynamic coherent focus OCT with depth-independent transversal resolution. Journal of Modern Optics 46 (3):541–553. Li, H., B. A. Standish, A. Mariampillai, N. R. Munce, Y. X. Mao, S. Chiu, N. E. Alarcon, B. C. Wilson, A. Vitkin, and V. X. D. Yang. 2006. Feasibility of interstitial Doppler optical coherence tomography for in vivo detection of microvascular changes during photodynamic therapy. Lasers in Surgery and Medicine 38 (8):754–761. Liang, X., A. L. Oldenburg, V. Crecea, E. J. Chaney, and S. A. Boppart. 2008. Optical micro-scale mapping of dynamic biomechanical tissue properties. Optics Express 16: 11052–11065. Liberman, L., D. D. Dershaw, P. P. Rosen, A. F. Abramson, B. M. Deutch, and L. E. Hann. 1994. Stereotaxic 14-gauge breast biopsy – how many core biopsy specimens are needed. Radiology 192 (3):793–795. Lingley-Papadopoulos, C. A., M. H. Loew, M. J. Manyak, and J. M. Zara. 2008. Computer recognition of cancer in the urinary bladder using optical coherence tomography and texture analysis. Journal of Biomedical Optics 13:024002. Liu, B., and M. E. Brezinski. 2007. Theoretical and practical considerations on detection performance of time domain, Fourier domain, and swept source optical coherence tomography. Journal of Biomedical Optics 12 (4):044007. Liu, G. Y., J. S. Lu, K. W. Shen, J. Wu, C. M. Chen, Z. Hu, Z. Z. Shen, T. Q. Zhang, and Z. M. Shao. 2008. Fiberoptic ductoscopy combined with cytology testing in the patients of spontaneous nipple discharge. Breast Cancer Research and Treatment 108 (2):271–277.

246

S.G. Adie and S.A. Boppart

Lu, C. W., C. K. Lee, M. T. Tsai, Y. M. Wang, and C. C. Yang. 2008. Measurement of the hemoglobin oxygen saturation level with spectroscopic spectral-domain optical coherence tomography. Optics Letters 33 (5):416–418. Luo, W., F. T. Nguyen, A. M. Zysk, T. L. S. Ralston, J. Brockenbrough, D. L. Marks, A. L. Oldenburg, and S. A. Boppart. 2005. Optical biopsy of lymph node morphology using optical coherence tomography. Technology in Cancer Research & Treatment 4 (5):539–547. Manyak, M. J., N. D. Gladkova, J. H. Makari, A. M. Schwartz, E. V. Zagaynova, L. Zolfaghari, J. M. Zara, R. Iksanov, and F. I. Feldchtein. 2005. Evaluation of superficial bladder transitionalcell carcinoma by optical coherence tomography. Journal of Endourology 19 (5):570–574. McMasters, K. M., T. M. Tuttle, D. J. Carlson, C. M. Brown, R. D. Noyes, R. L. Glaser, D. J. Vennekotter, P. S. Turk, P. S. Tate, A. Sardi, P. B. Cerrito, and M. J. Edwards. 2000. Sentinel lymph node biopsy for breast cancer: a suitable alternative to routine axillary dissection in multi-institutional practice when optimal technique is used. Journal of Clinical Oncology: Official Journal of the American Society of Clinical Oncology 18 (13):2560–2566. Morgner, U., W. Drexler, F. X. Kartner, X. D. Li, C. Pitris, E. P. Ippen, and J. G. Fujimoto. 2000. Spectroscopic optical coherence tomography. Optics Letters 25 (2):111–113. Mourant, J. R., M. Canpolat, C. Brocker, O. Esponda-Ramos, T. M. Johnson, A. Matanock, K. Stetter, and J. P. Freyer. 2000. Light scattering from cells: the contribution of the nucleus and the effects of proliferative status. Journal of Biomedical Optics 5 (2):131–137. Nguyen, F. T., A. M. Zysk, E. J. Chaney, J. G. Kotynek, U. J. Oliphant, F. J. Bellafiore, K. M. Rowland, P. A. Johnson, and S. A. Boppart. 2008a. Intraoperative evaluation of breast tumor margins with optical coherence tomography. (unpublished). Nguyen, F. T., A. M. Zysk, E. J. Chaney, J. G. Kotynek, U. J. Oliphant, F. J. Bellafiore, K. M. Rowland, P. A. Johnson, S.G. Adie, and S. A. Boppart. 2008b. Evaluation of optical coherence tomography for the intraoperative assessment of lymph nodes in breast cancer. (unpublished). Nguyen, F. T., A. M. Zysk, J. G. Kotynek, F. J. Bellafiore, K. M. Rowland, P. A. Johnson, E. J. Chaney, and S. A. Boppart. 2007. Portable real-time optical coherence tomography system for intraoperative imaging and staging of breast cancer, at San Jose, CA, United States, Proc. SPIE Vol. 6430:64300. Oldenburg, A. L., J. R. Gunther, and S. A. Boppart. 2005a. Imaging magnetically labeled cells with magnetomotive optical coherence tomography. Optics Letters 30 (7):747–749. Oldenburg, A. L., M. N. Hansen, D. A. Zweifel, A. Wei, and S. A. Boppart. 2006. Plasmonresonant gold nanorods as low backscattering albedo contrast agents for optical coherence tomography. Optics Express 14 (15):6724–6738. Oldenburg, A. L., F. J. J. Toublan, K. S. Suslick, A. Wei, and S. A. Boppart. 2005b. Magnetomotive contrast for in vivo optical coherence tomography. Optics Express 13 (17):6597–6614. Oldenburg, A. L., C. Y. Xu, and S. A. Boppart. 2007. Spectroscopic optical coherence tomography and microscopy. IEEE Journal of Selected Topics in Quantum Electronics 13 (6):1629–1640. Olmedo, J. M., K. E. Warschaw, J. M. Schmitt, and D. L. Swanson. 2006. Optical coherence tomography for the characterization of basal cell carcinoma in vivo: a pilot study. Journal of the American Academy of Dermatology 55 (3):408–412. Olmedo, J. M., K. E. Warschaw, J. M. Schmitt, and D. L. Swanson. 2007. Correlation of thickness of basal cell carcinoma by optical coherence tomography in vivo and routine histologic findings: a pilot study. Dermatologic Surgery 33 (4):421–426. Pan, Y. T., R. Birngruber, and R. Engelhardt. 1997. Contrast limits of coherence-gated imaging in scattering media. Applied Optics 36 (13):2979–2983. Pan, Y. T., J. P. Lavelle, S. I. Bastacky, S. Meyers, G. Pirtskhalaishvili, M. L. Zeidel, and D. L. Farkas. 2001. Detection of tumorigenesis in rat bladders with optical coherence tomography. Medical Physics 28 (12):2432–2440. Pan, Y. T., T. Q. Xie, C. W. Du, S. Bastacky, S. Meyers, and M. L. Zeidel. 2003. Enhancing early bladder cancer detection with fluorescence-guided endoscopic optical coherence tomography. Optics Letters 28 (24):2485–2487.

Optical Coherence Tomography for Cancer Detection

247

Park, B. H., C. Saxer, S. M. Srinivas, J. S. Nelson, and J. F. de Boer. 2001. In vivo burn depth determination by high-speed fiber-based polarization sensitive optical coherence tomography. Journal of Biomedical Optics 6 (4):474–479. Pfau, P. R., M. V. Sivak, A. Chak, M. Kinnard, R. C. K. Wong, G. A. Isenberg, J. A. Izatt, A. Rollins, and V. Westphal. 2003. Criteria for the diagnosis of dysplasia by endoscopic optical coherence tomography. Gastrointestinal Endoscopy 58 (2):196–202. Pijnappel, R. M., M. van den Donk, R. Holland, W. P. Th M. Mali, J. L. Peterse, J. H. C. L. Hendriks, and P. H. M. Peeters. 2004. Diagnostic accuracy for different strategies of imageguided breast intervention in cases of nonpalpable breast lesions. British Journal of Cancer 90 (3):595–600. Pitris, C., C. Jesser, S. A. Boppart, D. Stamper, M. E. Brezinski, and J. G. Fujimoto. 2000. Feasibility of optical coherence tomography for high-resolution imaging of human gastrointestinal tract malignancies. Journal of Gastroenterology 35 (2):87–92. Povazay, B., K. Bizheva, A. Unterhuber, B. Hermann, H. Sattmann, A. F. Fercher, W. Drexler, A. Apolonski, W. J. Wadsworth, J. C. Knight, P. S. J. Russell, M. Vetterlein, and E. Scherzer. 2002. Submicrometer axial resolution optical coherence tomography. Optics Letters 27 (20): 1800–1802. Qi, X., M. V. Sivak, G. Isenberg, J. E. Willis, and A. M. Rollins. 2006. Computer-aided diagnosis of dysplasia in Barrett’s esophagus using endoscopic optical coherence tomography. Journal of Biomedical Optics 11 (4):044010. Ralston, T. S., D. L. Marks, P. S. Carney, and S. A. Boppart. 2007. Interferometric synthetic aperture microscopy. Nature Physics 3 (2):129–134. Ralston, T. S., D. L. Marks, P. S. Carney, and S. A. Boppart. 2008. Real-time interferometric synthetic aperture microscopy. Optics Express 16 (4):2555–2569. Ramachandran, J., T. M. Powers, S. Carpenter, A. Garcia-Lopez, J. P. Freyer, and J. R. Mourant. 2007. Light scattering and microarchitectural differences between tumorigenic and non-tumorigenic cell models of tissue. Optics Express 15 (7):4039–4053. Rao, K. D., M. A. Choma, S. Yazdanfar, A. M. Rollins, and J. A. Izatt. 2003. Molecular contrast in optical coherence tomography by use of a pump-probe technique. Optics Letters 28 (5): 340–342. Reed, W. A., M. F. Yan, and M. J. Schnitzer. 2002. Gradient-index fiber-optic microprobes for minimally invasive in vivo low-coherence interferometry. Optics Letters 27 (20):1794–1796. Reynolds, H. E. 2000. Core needle biopsy of challenging benign breast conditions: a comprehensive literature review. American Journal of Roentgenology 174 (5):1245–1250. Ridgway, J. M., W. B. Armstrong, S. Guo, U. Mahmood, J. P. Su, R. P. Jackson, T. Shibuya, R. L. Crumley, M. Gu, Z. P. Chen, and B. J. F. Wong. 2006. In vivo optical coherence tomography of the human oral cavity and oropharynx. Archives of Otolaryngology-Head & Neck Surgery 132 (10):1074–1081. Rogowska, J., N. A. Patel, J. G. Fujimoto, and M. E. Brezinski. 2004. Optical coherence tomographic elastography technique for measuring deformation and strain of atherosclerotic tissues. Heart 90 (5):556–562. Rogowska, J., N. Patel, S. Plummer, and M. E. Brezinski. 2006. Quantitative optical coherence tomographic elastography: method for assessing arterial mechanical properties. British Journal of Radiology 79 (945):707–711. Saleh, B. E. A., and M. C. Teich. 1991. Fundamentals of Photonics, Wiley Series in Pure and Applied Optics. New York: Wiley. Sapozhnikova, V. V., N. M. Shakhova, V. A. Kamensky, S. A. Petrova, L. B. Snopova, and R. V. Kuranov. 2005. Capabilities of fluorescence spectroscopy using 5-ALA and optical coherence tomography for diagnosis of neoplastic processes in the uterine cervix and vulva. Laser Physics 15 (12):1664–1673. Sarvazyan, A. P., O. V. Rudenko, S. D. Swanson, J. B. Fowlkes, and S. Y. Emelianov. 1998. Shear wave elasticity imaging: A new ultrasonic technology of medical diagnostics. Ultrasound in Medicine and Biology 24 (9):1419–1435.

248

S.G. Adie and S.A. Boppart

Schmitt, J. M. 1998. OCT elastography: imaging microscopic deformation and strain of tissue. Optics Express 3 (6):199–211. Schmitt, J. M. 1999. Optical coherence tomography (OCT): a review. IEEE Journal of Selected Topics in Quantum Electronics 5 (4):1205–1215. Schuman, J. S., C. A. Puliafito, and J. G. Fujimoto. 2004. Optical Coherence Tomography of Ocular Diseases: SLACK Inc., Thorofare, NJ. Shankar, P. M., V. A. Dumane, C. W. Piccoli, J. M. Reid, F. Forsberg, and B. B. Goldberg. 2003. Computer-aided classification of breast masses in ultrasonic B-scans using a multiparameter approach. IEEE Transactions on Ultrasonics Ferroelectrics and Frequency Control 50 (8):1002–1009. Shen, K. W., J. Wu, J. S. Lu, Q. X. Han, Z. Z. Shen, M. Nguyen, Z. M. Shao, and S. H. Barsky. 2000. Fiberoptic ductoscopy for patients with nipple discharge. Cancer 89 (7):1512–1519. Standish, B. A., X. Jin, J. Smolen, A. Mariampillai, N. R. Munce, B. C. Wilson, I. A. Vitkin, and V. X. D. Yang. 2007a. Interstitial Doppler optical coherence tomography monitors microvascular changes during photodynamic therapy in a Dunning prostate model under varying treatment conditions. Journal of Biomedical Optics 12 (3):034022. Standish, B. A., V. X. D. Yang, N. R. Munce, L. M. W. K. Song, G. Gardiner, A. Lin, Y. I. Mao, A. Vitkin, N. E. Marcon, and B. C. Wilson. 2007b. Doppler optical coherence tomography monitoring of microvascular tissue response during photodynamic therapy in an animal model of Barrett’s esophagus. Gastrointestinal Endoscopy 66 (2):326–333. Strasswimmer, J., M. C. Pierce, B. H. Park, V. Neel, and J. F. de Boer. 2004. Polarization-sensitive optical coherence tomography of invasive basal cell carcinoma. Journal of Biomedical Optics 9 (2):292–298. Swanson, E. A., D. Huang, M. R. Hee, J. G. Fujimoto, C. P. Lin, and C. A. Puliafito. 1992. Highspeed optical coherence domain reflectometry. Optics Letters 17 (2):151–153. Swenson, K. K., M. J. Nissen, C. Ceronsky, L. Swenson, M. W. Lee, and T. M. Tuttle. 2002. Comparison of side effects between sentinel lymph node and axillary lymph node dissection for breast cancer. Annals of Surgical Oncology 9 (8):745–753. Tearney, G. J., M. E. Brezinski, J. F. Southern, B. E. Bouma, M. R. Hee, and J. G. Fujimoto. 1995. Determination of the refractive-index of highly scattering human tissue by optical coherence tomography. Optics Letters 20 (21):2258–2260. Tumlinson, A. R., L. P. Hariri, U. Utzinger, and J. K. Barton. 2004. Miniature endoscope for simultaneous optical coherence tomography and laser-induced fluorescence measurement. Applied Optics 43 (1):113–121. Veronesi, U., G. Paganelli, G. Viale, A. Luini, S. Zurrida, V. Galimberti, M. Intra, P. Veronesi, C. Robertson, P. Maisonneuve, G. Renne, C. De Cicco, F. De Lucia, and R. Gennari. 2003. A randomized comparison of sentinel-node biopsy with routine axillary dissection in breast cancer. The New England Journal of Medicine 349 (6):546–553. Wang, Z. G., D. B. Durand, M. Schoenberg, and Y. T. Pan. 2005. Fluorescence guided optical coherence tomography for the diagnosis of early bladder cancer in a rat model. Journal of Urology 174 (6):2376–2381. Wang, R. K. K., S. Kirkpatrick, and M. Hinds. 2007a. Phase-sensitive optical coherence elastography for mapping tissue microstrains in real time. Applied Physics Letters 90 (16):164105. Wang, Z. G., C. S. D. Lee, W. C. Waltzer, J. X. Liu, H. K. Xie, Z. J. Yuan, and Y. T. Pan. 2007b. In vivo bladder imaging with microelectromechanical systems-based endoscopic spectral domain optical coherence tomography. Journal of Biomedical Optics 12 (3):034009. Whiteman, S. C., Y. Yang, D. G. van Pittius, M. Stephens, J. Parmer, and M. A. Spiteri. 2006. Optical coherence tomography: real-time imaging of bronchial airways microstructure and detection of inflammatory/neoplastic morphologic changes. Clinical Cancer Research 12 (3):813–818. Wilder-Smith, P., W. G. Jung, M. Brenner, K. Osann, H. Beydoun, D. Messadi, and Z. P. Chen. 2004. In vivo optical coherence tomography for the diagnosis of oral malignancy. Lasers in Surgery and Medicine 35 (4):269–275.

Optical Coherence Tomography for Cancer Detection

249

Wilder-Smith, P., T. Krasieva, W. G. Jung, J. Zhang, Z. P. Chen, K. Osann, and B. Tromberg. 2005. Noninvasive imaging of oral premalignancy and malignancy. Journal of Biomedical Optics 10 (5):051601. Wong, B. J. F., R. P. Jackson, S. G. Guo, J. M. Ridgway, U. Mahmood, J. P. Su, T. Y. Shibuya, R. L. Crumley, M. Gu, W. B. Armstrong, and Z. P. Chen. 2005. In vivo optical coherence tomography of the human larynx: normative and benign pathology in 82 patients. Laryngoscope 115 (11):1904–1911. Xie, T. Q., M. L. Zeidel, and Y. T. Pan. 2002. Detection of tumorigenesis in urinary bladder with optical coherence tomography: optical characterization of morphological changes. Optics Express 10 (24):1431–1443. Xu, C. Y., F. Kamalabadi, and S. A. Boppart. 2005. Comparative performance analysis of timefrequency distributions for spectroscopic optical coherence tomography. Applied Optics 44 (10):1813–1822. Xu, C. Y., C. Vinegoni, T. S. Ralston, W. Luo, W. Tan, and S. A. Boppart. 2006. Spectroscopic spectral-domain optical coherence microscopy. Optics Letters 31 (8):1079–1081. Xu, C. Y., J. Ye, D. L. Marks, and S. A. Boppart. 2004. Near-infrared dyes as contrast-enhancing agents for spectroscopic optical coherence tomography. Optics Letters 29 (14):1647–1649. Yadlowsky, M. J., J. M. Schmitt, and R. F. Bonner. 1995. Multiple-scattering in optical coherence microscopy. Applied Optics 34 (25):5699–5707. Yang, C. H. 2005. Molecular contrast optical coherence tomography: a review. Photochemistry and Photobiology 81 (2):215–237. Yang, C. H., L. E. L. McGuckin, J. D. Simon, M. A. Choma, B. E. Applegate, and J. A. Izatt. 2004. Spectral triangulation molecular contrast optical coherence tomography with indocyanine green as the contrast agent. Optics Letters 29 (17):2016–2018. Yang, V. X. D., S. J. Tang, M. L. Gordon, B. Qi, G. R. Gardiner, M. Cirocco, P. Kortan, G. B. Haber, G. Kandel, I. A. Vitkin, B. C. Wilson, and N. E. Marcon. 2005. Endoscopic Doppler optical coherence tomography in the human GI tract: initial experience. Gastrointestinal Endoscopy 61 (7):879–890. Yaqoob, Z., E. McDowell, J. G. Wu, X. Heng, J. Fingler, and C. H. Yang. 2006. Molecular contrast optical coherence tomography: a pump-probe scheme using indocyanine green as a contrast agent. Journal of Biomedical Optics 11 (5):054017. Yasuno, Y., V. D. Madjarova, S. Makita, M. Akiba, A. Morosawa, C. Chong, T. Sakai, K. P. Chan, M. Itoh, and T. Yatagai. 2005. Three-dimensional and high-speed swept-source optical coherence tomography for in vivo investigation of human anterior eye segments. Optics Express 13 (26):10652–10664. Youngquist, R. C., S. Carr, and D. E. N. Davies. 1987. Optical coherence-domain reflectometry – a new optical evaluation technique. Optics Letters 12 (3):158–160. Zagaynova, E., M. J. Manyak, O. Streltsova, N. Gladkova, F. Feldchtein, and V. Kamensky. 2004. A multicenter study of optical coherence tomography for diagnosis and guided surgery of bladder cancer. Journal of Clinical Oncology 22 (14):391S–391S. Zagaynova, E. V., O. S. Streltsova, N. D. Gladkova, L. B. Snopova, G. V. Gelikonov, F. I. Feldchtein, and A. N. Morozov. 2002. In vivo optical coherence tomography feasibility for bladder disease. Journal of Urology 167 (3):1492–1496. Zakaria, S., A. C. Degnim, C. G. Kleer, K. A. Diehl, V. M. Cimmino, A. E. Chang, L. A. Newman, and M. S. Sabel. 2007. Sentinel lymph node biopsy for breast cancer: how many nodes are enough? Journal of Surgical Oncology 96 (7):554–559. Zara, J. M., and C. A. Lingley-Papadopoulos. 2008. Endoscopic OCT approaches toward cancer diagnosis. IEEE Journal of Selected Topics in Quantum Electronics 14 (1):70–81. Zhao, Y. H., Z. P. Chen, C. Saxer, S. H. Xiang, J. F. de Boer, and J. S. Nelson. 2000. Phaseresolved optical coherence tomography and optical Doppler tomography for imaging blood flow in human skin with fast scanning speed and high velocity sensitivity. Optics Letters 25 (2):114–116.

250

S.G. Adie and S.A. Boppart

Zvyagin, A. V., J. B. FitzGerald, K. K. M. B. D. Silva, and D. D. Sampson. 2000. Realtime detection technique for Doppler optical coherence tomography. Optics Letters 25 (22): 1645–1647. Zvyagin, A. V., K. K. M. B. D. Silva, S. A. Alexandrov, T. R. Hillman, J. J. Armstrong, T. Tsuzuki, and D. D. Sampson. 2003. Refractive index tomography of turbid media by bifocal optical coherence refractometry. Optics Express 11 (25):3503–3517. Zysk, A. M., S. G. Adie, J. J. Armstrong, M. S. Leigh, A. Paduch, D. D. Sampson, F. T. Nguyen, and S. A. Boppart. 2007a. Needle-based refractive index measurement using low-coherence interferometry. Optics Letters 32 (4):385–387. Zysk, A. M., and S. A. Boppart. 2006. Computational methods for analysis of human breast tumor tissue in optical coherence tomography images. Journal of Biomedical Optics 11 (5):054015. Zysk, A. M., E. J. Chaney, and S. A. Boppart. 2006. Refractive index of carcinogen-induced rat mammary tumours. Physics in Medicine and Biology 51 (9):2165–2177. Zysk, A. M., D. L. Marks, D. Y. Liu, and S. A. Boppart. 2007b. Needle-based reflection refractometry of scattering samples using coherence-gated detection. Optics Express 15 (8):4787–4794. Zysk, A. M., F. T. Nguyen, E. J. Chaney, J. G. Kotynek, U. J. Oliphant, F. J. Bellafiore, P. A. Johnson, K. M. Rowland, and S. A. Boppart. 2008. Clinical feasibility of microcopicallyguided breast needle biopsy using a fiber-optic probe with computer-aided detection. (unpublished). Zysk, A. M., F. T. Nguyen, A. L. Oldenburg, D. L. Marks, and S. A. Boppart. 2007c. Optical coherence tomography: a review of clinical development from bench to bedside. Journal of Biomedical Optics 12 (5):051403.

Optical Imaging of Cancer: Neuro-oncologic Applications Stephen Yip and Khalid Shah

Introduction Clinical Significance Tumors of the central nervous system (CNS) present unique challenges to the clinician with respect to diagnosis and therapeutics. Location of the tumors within the brain confined by the calvaria and the complex boney architecture of the skull base also present difficulty in efficiently imaging lesions especially in computer tomographic (CT) imaging of lesions in the posterior fossa. The most common central nervous system (CNS) tumors are metastatic lesions that arise from tumors elsewhere in the body, of which lung and breast primaries constitute a large proportion. Primary brain tumors, those that arise from cellular constituents residing within the brain such as glial cells and neurons, are particularly vexing in both diagnosis and treatment improvements in imaging technology with the combined use of neurophysiological monitoring have lead to more aggressive surgical resections of tumors. The photonic based imaging will play significant role in the implementation of future experimental therapeutics and image-guided therapies in neuro-oncology. Glial tumors include neoplastic derivatives of astrocytes and oligodendrocytes which probably arise from maldevelopment of stem and progenitor cells or from de-differentiation of mature cells (Bachoo et al., 2002). Specific genetic and epigenetic abnormalities underlie many of the biological and behavioral dysfunctions of brain tumors (Louis, 2006). Regardless, glial tumors such as astrocytomas and oligodendrogliomas exhibit primitive and often dysregulated behavior including autonomous growth and the ability to extensively migrate and infiltrate along and through the adjacent neuropil and its supporting structures. This phenomenon was initially described by Scherer and accounts for the difficulty in the clinical management of primary brain tumors (Scherer, 1940). Extensive infiltration and insinuation

K. Shah (B) Molecular Neurotherapy and Imaging Laboratory, Massachusetts General Hospital, Harvard Medical School, Boston MA 02129, USA e-mail: [email protected]

E. Rosenthal, K.R. Zinn (eds.), Optical Imaging of Cancer, C Springer Science+Business Media, LLC 2009 DOI 10.1007/978-0-387-93874-5_12, 

251

252

S. Yip and K. Shah

of tumor cells with normal surrounding neuropil, migration along blood vessels, the subpial space, and white matter tracts, and the proximity to normal CNS structures preclude the effective surgical resection of tumor (Holland, 2000). In addition, doselimitation in whole brain irradiation and the restriction of the blood brain barrier (BBB) on the delivery of both traditional therapeutic agents as well as novel molecular targeting compounds place additional limitations on the effective management of primary brain tumors as well as metastatic or secondary brain tumors.

Advances in Imaging Improvements in imaging technology with the combined use of neurophysiological monitoring have lead to more aggressive surgical resections of tumors (Lyons and Vora, 2007). In addition, current imaging modalities play essential roles in all phases of management of the brain tumor patients. These technologies will also play significant roles in the implementation of future experimental therapeutics and image-guided therapies. This chapter will focus on photonic-based imaging in neuro-oncology and therefore will not discuss the more traditional neuro-imaging modalities such as CT and MRI.

Current Options in Clinical Diagnostic Imaging of CNS Malignancies The most commonly utilized forms of imaging techniques for the clinical evaluation of CNS tumors are CT and MRI with and without contrast agents. Furthermore, use of different acquisition sequences in MRI and innovations in the delivery of contrast agent also contribute to information in addition to size and localization of the tumor. More physiological and functionally based techniques such as magnetic resonance spectroscopy (MRS) and positron emission tomography (PET) using various metabolic tracers allow for better localization and functional profiling of tumors (Jacobs et al., 2005). These include tissue edema, hemorrhage, and information on vascular permeability of the tumor – all of which contribute to better characterization of the tumor and also give additional information as to the pathology of the lesion such as differentiation between recurrence of tumor versus radiation necrosis (Jacobs et al., 2005; Price, 2007). Nevertheless, shortcomings exist for the current generation of imaging technologies especially in achieving high enough resolution for the detection of individual and small numbers of tumor cells which are the major culprit in local and distant recurrences in gliomas. It is thought that identification and localization of these cells will aid in effecting a more surgically complete eradication of the tumor and also will give more advanced warning of disease recurrence. Optical imaging using various molecular probes permits the real-time visualization of molecular and cellular processes, which provide information pertaining to tumor behavior, delivery vehicles, and response to therapy. In recent years, cleverly designed genetically engineered fluorescent and bioluminescent probes have been incorporated into both

Optical Imaging of Cancer: Neuro-oncologic Applications

253

tumor cells and delivery vehicles (e.g., viruses and stem cells) which have significantly improved our understanding of the molecular processes that underlie malignant cell behavior as well as, the delivery and therapeutic efficacy of new agents. These probes interrogate individual cellular pathways and therefore offer considerably higher sensitivity and specificity in the detection of neoplastic processes than more traditional techniques that are more tissue based rather than cell or pathway based (Brindle, 2008).

In Vivo Optical Imaging in Neuro-oncology The various forms of photonic-based imaging, either in isolation or in combination with more traditional imaging techniques, permit multi-modality imaging of molecular events specific to the tumor and its microenvironment as well as biological responses to therapeutics are superior to traditional imaging in many aspects (see Table 1). The current generations of imaging technology such as CT, MRI, and PET are useful in providing some of this information; however, photonic-based detection using novel reporter probes (some with dual function as both diagnostics and therapeutic) offers great hope in revolutionizing the management of cancer patients (Brindle, 2008; Weissleder and Pittet, 2008). Nevertheless, despite promises offered by molecular optical imaging, currently there is little clinical realization of these imaging techniques in neuro-oncology. We will first focus on advances in preclinical applications of optical imaging in the realms of diagnostics and therapeutics with emphasis in neuro-oncology. Table 1 Characteristics of different optical imaging techniques Imaging methods

Molecular marker

Tissue depth

FRI-visible GFP, BFP, 1 cm CFP,YFP, RFP 1 cm FRI–NIR Proteases: Cathepsin-D and Cathepsin-H, MMP-2, Caspase-3 FMT No limit BLI

Fluc, Rluc, Gluc

* Abbreviations:

1–2 cm

Potential for human imaging

Advantages

Reporter/ process

Yes, under certain limits Yes, under certain limits

Sensitive, detects Fluorescent fluorochromes in proteins live and dead cells Activatable, detects Peptide-caged fluorochromes in activatable live and dead cells NIR fluorochromes

Yes

Morphological, tumor imaging Gene expression, cell tracking, quick and easy

Yes, under certain limits

NIR fluorochromes Luciferin, coelenterazine

FRI, fluoresence resonance imaging; NIR, near infrared imaging; FMT, fluresence mediated tomography; BLI, bioluminescence imaging; GFP, green fluorescent protein; CFP, cyan fluorescent protein; YFP, yellow fluorescent protein; RFP, red fluorescent protein; MMP, matrix mettaloprotease.

254

S. Yip and K. Shah

Fluorescence Imaging Fluorescence-based in vivo imaging of brain tumors using fluorescein was first reported in 1948 (Moore et al., 1948). Certain molecules, known as fluorochromes, emit light at a longer wavelength after excitation at shorter wavelength. Prior to that, there had been reports of endogenous porphyrins contributing to autofluorescence when exposed to ultraviolet light. Porphyrins and their derivatives have been used for the in vivo detection of neoplastic tissue based on the principle of preferential uptake of the compound directly or conjugated to antibodies that recognize tumor-specific antigens (Mannino et al., 2008). Fluorescence-based imaging has been in clinical use in many forms (Berg et al., 2005). Taking advantage of this phenomenon, fluorescence-guided neurosurgery has been in limited use for the past 25 years as both an intraoperative adjunct for the visualization of residual tumor and also as a mean of eradicating these tumor cells via photo-activation of the compound (see below) with generation of cytotoxic reactive oxygen species (ROS) such as singlet oxygen and superoxide anions (Stylli and Kaye, 2006; Utsuki et al., 2007). With special interests to clinical neuro-oncology, photodynamic therapy or PDT-using derivatives of porphyrin has a long history in both the diagnostics and therapeutic realms. The major impediment to the clinical implementation of fluorescence-based imaging is imposed by tissue barrier on the activation and detection of the fluorochromes. This is somewhat overcome when imaging is used intraoperatively during surgical resection of brain tumors since normal tissue barriers against light penetration imposed by skin, underlying subcutaneous soft tissues, and the calvarium are eliminated by craniotomy (Stylli and Kaye, 2006; Utsuki et al., 2007; Madsen and Hirschberg, 2006; Okuda et al., 2007) Non-neuro-oncological applications such as detection and eradication of pre-malignant and malignant lesions in the exposed skin and mucosal linings of organs and tracts accessible by endoscopes such as the oro-digestive tract, the airway, and the bladder are more amendable to this particular technology and are covered in a separate chapter in this book (Berg et al., 2005; Krammer and Plaetzer, 2008). The advent of endoscopic techniques in surgical neuro-oncology allows limited use of these photosensitizer compounds in minimal access neurosurgery (Tamura et al., 2007). This is essentially the earliest form of optical imaging in neuro-oncology and has the longest history with respect to clinical applications; however, disadvantages of the technology concern specificity of labeling, relative persistence of the photosensitizer in the patient and therefore photosensitivity, low penetrability of photo-activation secondary to short excitation wavelength, and inability of the photosensitizer to probe complex biological and molecular interactions. Also, tissue autofluorescence constitutes an additional disadvantage in the in vivo applications of fluorescence-based imaging. This is solved partially by the use of photosensitizers with unique biophysical properties. To fully realize the potential of photonicbased imaging in the in vivo setting – a new generation of fluorescence reporter probes coupled with development of advanced illumination and new and complex computational models have been assembled and evolved into fluorescence molecular tomography (FMT) (Ntziachristos et al., 2005). FMT – either alone or in

Optical Imaging of Cancer: Neuro-oncologic Applications

255

conjunction with more established imaging techniques in the forms of FMT–CT and FMT–MRI are more suitable for pre-clinical use at this juncture and again are restricted by biophysical constraints imposed by the activating wavelength (Ntziachristos et al., 2005; Zacharakis et al., 2005). There are some reports on the potential utility of optical spectroscopy but again they are best suited for the operating room after dural opening to permit adequate photo-excitation and detection (Lin et al., 2005; Majumder et al., 2007).

Photodynamic Therapy (PDT) The same photoactivable fluorescence compounds, mainly porphyrin based, used for tumor imaging also generate cytotoxic oxygen radicals which can be used to mediate destruction of tumor cells. Photodynamic therapy or PDT is sometimes used as an adjuvant therapy immediately post-resection in the operating theater (Beck et al., 2007; Stummer et al., 2008). The patient typically receives the photosensitizing agent intravenously or orally (e.g., 5-aminolevulinic acid) a day prior to surgery and the resection cavity is bathed in light of the appropriate excitation wavelength (usually in the presence of an optical diffuser) to activate the photosensitizer which is preferentially taken up by rapidly proliferating tumor cells (Madsen and Hirschberg, 2006). PDT appears to be a relatively non-toxic form of adjuvant therapy and might offer the surgeon a method of eradicating infiltrating glioma cells in the resection margin without significant damage to normal neural structures. Similar principle of using the same compound with properties suitable for both diagnostics and therapeutic applications is being expanded currently in the development of molecular beacons (Stefflova et al., 2007a, b).

Near Infrared Fluorescence Reflectance Imaging (NIRF-i) Tissue excitation at the near infrared region (650–900 nm) permits deep penetration of tissue in a non-invasive fashion. This is due to the low absorption by hemoglobin, water, and lipids in this range (Fig. 1). An added benefit is the elimination of autofluorescence from non-targeted tissue. Combination of photo-excitation within the NIRF spectra in conjunction with a specific fluorescent probe permit real-time interrogation of unique physiological processes in deep tissues. This is achieved via the use of NIRF fluorochromes conjugated to quenching peptides separated by protease recognition sites. For example, in vivo cancerous tissue with increased expression of proteases such as cathepsins and metalloproteases can be visualized by introduction of these designer probes followed by NIR excitation and detection of unique fluorescence signals (Mahmood and Weissleder, 2003). The same principle of incorporating unique protease recognition site into NIR fluorochromes for the detection of specific proteases also allows for detection and localization of regional apoptosis in the living animal by taking advantage of the unique caspases cleavage site (Messerli et al., 2004). Protease-dependent mechanisms for optical imaging are discussed in a separate chapter.

256

S. Yip and K. Shah

Fig. 1 (A) Interaction of light with tissue. The absorption coefficient of light in tissue is dependent on wavelength and results from absorbers such as hemoglobins, lipids, and water. Given the decreased absorption of light in the near infrared (NIR) region compared with visible light (∼400–650 nm) and infrared light (>900 nm), tissue penetration of NIR photons may be up to 10–15 cm. (B) Chemical structure of repeating graft copolymer segment indicating quenched and the activated state after the cleavage at the enzyme recognition sites indicated by arrows (Adapted from Shah and Weissleder (2005))

NIRF-i has been demonstrated in several animal glioma models (Chen et al., 2004; Hsu et al., 2006). One example utilized peptide-based probe to preferentially localize tumor-associated vasculature expressing αv β3 integrin. In addition, probes with customizable modules allow for NIRF-i-based detection of tumorspecific biological processes as well as functioning as a photosensitizing compound with therapeutic potential (Stefflova et al., 2007a, b). These dual function “molecular beacons” consist of improved photosensitizing molecules that allow for NIRF activation thus achieving satisfactory tissue penetration of activating light to effect sensitive detection of probes in a pathway-specific fashion and also the generation of cytotoxic molecules for the destruction of tumor cells.

Optical Imaging of Cancer: Neuro-oncologic Applications

257

Bioluminescence Imaging Bioluminescence imaging or BLI has several advantages over fluorescence-based imaging in that genetically engineered vectors expressing different forms of luciferase enzymes and exogenously administered substrates (luciferins) generate photons which can be sensitively detected using cooled charge-coupled device cameras in living animals. The absence of background noise and the ability to specifically engineer different versions of the enzymes with unique substrates and kinetics profiles allow for versatility in in vivo applications. The luciferases from Renilla (Rluc) and firefly (Fluc) have different substrates, coelenterazine and D-luciferin, respectively, and can be imaged in tumors in the same living mouse with kinetics of light production being separable in time by separate injections of these two substrates (Bhaumik and Gambhir, 2002; Shah et al., 2003). However, the same restriction in the form of limited tissue penetration of activating light as fluorescence-based imaging also applied to BLI. This limits the suitability of BLI for clinical use but this imaging modality has proven to be extremely useful in animal models for real-time multimodal monitoring of gene expression, cell tracking, and the monitoring of therapeutic response (Weissleder and Pittet, 2008; Shah and Weissleder, 2005). Expression of genetically engineered luciferase in implanted glioma xenografts allows for live, real-time, non-invasive, in vivo monitoring of tumor growth and response to therapy (Shah et al., 2003; Burgos et al., 2003; Soling et al., 2004; Kemper et al., 2006; Szentirmai et al., 2006; Dinca et al., 2007). In the work coordinated in our laboratory, glioma cells stably expressing Fluc were implanted subcutaneously into nude mice and tumor growth was monitored in vivo over time by luciferin administration and BLI. HSV amplicon vectors bearing the genes for TRAIL and Rluc were injected directly into these Fluc-positive gliomas allowing superimposition of gene delivery to the tumor by coelenterazine administration and BLI (Shah et al., 2003). This dual imaging approach has direct applications in studying the delivery of gene therapy vectors and simultaneously monitoring therapeutic effects in vivo (Wessels et al., 2007). Clever design of gene therapy vectors incorporating drug-inducible imaging markers whose expressions are proportional to those of the gene of interest permits in vitro detection with fluorescent microscopy and in vivo non-invasive monitoring using PET or BLI and gives information regarding expression level of the therapeutic cargo (Winkeler et al., 2007). Recent developments in advanced computational modeling coupled with imaging techniques have allowed three-dimensional realization of bioluminescence signal which permits more precise and quantitative determination (Wang et al., 2008).

Quantum Dots Advent of the use of fluorescent semiconductor nanocrystals or quantum dots (qdots) in biological applications has given researchers another tool in the advanced imaging of living tissues (Gao et al., 2004; Michalet et al., 2005). Qdots have unique properties including size-tunable light emission and improved signal-to-noise ratio

258

S. Yip and K. Shah

which permits multiplex analysis. In addition, development of qdots excitable with wavelength in the NIR allows for improved tissue penetration and so degree of in vivo imaging. One potential use of qdots in glioma management involves injection of qdots into the systemic circulation which are taken up by glioma-infiltrating macrophages thus labeling the tumor (Popescu and Toms, 2006). This indirect form of optical labeling appears to show some promise for identification of tumor deposits in pre-clinical animal models (Jackson et al., 2007; Muhammad et al., 2007; Wang et al., 2007). However, impact of qdots on neuro-oncology is more practically achievable at this point in in vitro diagnostics – all the characteristics mentioned above make qdots superior than organic dyes and fluorescent proteins for in vitro diagnostics (Xing et al., 2007). One can envision combinations of in vivo and in vitro use of qdots to enhance the sensitivity and specificity of brain tumor diagnosis.

Bimodal Fluorescence and Bioluminescence Imaging Transplantation of genetically manipulated stem cells of human origin to the central nervous system offers immense potential for the treatment of several neurological disorders. Monitoring the expression levels of transgenes and following cells at a cellular resolution in vivo is critical to the development of such therapies in vivo. Initial reports of neural stem cells (NSC) exhibiting gliomatropic behavior and the ability to deliver therapeutic molecules to glioma microdeposits had led to a rapid explosion of research in using NSC and more recently mesenchymal stem cells (MSC) as “smart” delivery vehicles for both diagnostic and therapeutic agents to glioma deposits (Benedetti et al., 2000; Aboody et al., 2000; Yip et al., 2006). Both NSC and MSC exhibit tremendous migratory potential which intriguingly mirrors that of glioma cells. Various groups have shown that NSC and MSC engineered to express tumor-specific therapeutic agents such as tumor necrosis factor apoptosis inducing ligand (TRAIL) and interleukins display exquisite tumor-tropic migratory behavior and the ability to express diagnostic and therapeutic genes in close proximity to “escaping” glioma cells and effecting their destruction (Shah et al., 2003; Ehtesham et al., 2002). Subsequently, several groups including our laboratory have demonstrated stable expression of BLI markers in stem cells which allow for the serial in vivo imaging of gliomatropic migration of these cells toward the implanted tumor even when implanted away from the tumor or even in the contralateral hemisphere. In an earlier study, we demonstrated tracking of fluorescence- and BLI-tagged mouse NSC engineered to express soluble TRAIL (s-TRAIL) in animals implanted with glioma cells tagged with a different BLI marker. This essentially allowed for the imaging of multiple events including migration of NSC and progression or regression of tumor burden in these animals (Tang et al., 2003; Shah et al., 2005). Recently, we have engineered lentiviral vectors bearing fusions between fluorescent and bioluminescent marker proteins and employed bioluminescence imaging and intravital microscopy to study the fate of human neural stem cells (hNSCs) in relation to gliomas in vivo. We used GFP–Rluc expressing malignant human glioma model and implanted hNSC-expressing Fluc–DsRed2 intracranially. Dual bioluminescence intravital microscopy and correlative neuropathology revealed that

Optical Imaging of Cancer: Neuro-oncologic Applications

259

transduced hNSCs migrate extensively toward and into glioma tumors and do not differentiate into terminal neural cell types within 2 weeks of implantation (Fig. 2). Similar studies, using primary NSC transduced with lentiviral vectors bearing both luciferase GFP, have been performed to non-invasively assess the survival and

Fig. 2 Imaging human neural stem cell fate in in vivo model of glioma. (A–D) Bioluminescence imaging of mice implanted with GFP–Fluc expressing hNSC in mice with established Rluc– DsRed2 gliomas. Fluc images of mice on day 3 (A), day 7 (B), day 10 (C), and Rluc image on day 10 (D) are shown. (E) Mice implanted with Gli36–GFP–Rluc glioma cells stereotactically into the right frontal lobe were implanted with Fluc–DsRed2 hNSC 2 days later. Mice were imaged by intravital microscopy (E), on day 10 after hNSC implantation, sacrificed and brains were sectioned and confocal microscopy was performed. (F) Fluorescent image showing hNSC (red) infiltrating the tumor (green); magnification (×40) (G–J). Immunohistochemistry on brain sections from Gli36–glioma bearing mice implanted with hNSC expressing GFP–Fluc, 10 days post-implantation. Representative images of brain sections immunostained for nestin (G), Ki67 (H), GFAP (I), and MAP-2 (J). (Green – GFP expression; Red – Ki67, GFAP, or MAP-2 expression; Yellow – co-expression of GFP and nestin) (Adapted from Shah et al., (2008))

260

S. Yip and K. Shah

residence time of transplanted NSC in the spinal cord injury models in living animals (Okada et al., 2005). These studies demonstrate the combined application of different imaging modalities in evaluating the fate of NSC in real time. Many altered pathways in cancer cells depend on growth factor receptors. The majority of glioblastomas demonstrate atypical genetic changes in the EGFR locus resulting in increased cell surface expression of the receptor protein and/or expression of a mutated form of the protein that is autonomous of ligand activation (Sugawa et al., 1990; Kesari et al., 2005). This ultimately results in the growth advantage of malignant glioma cells but may make them susceptible to targeted inhibition by kinase inhibitors. Therefore, it is useful to profile the EGFR expression and activation status in malignant gliomas which provides both molecularly prognostic as well as therapeutic information. In an effort to dissect the role of EGFR expression in glioma progression in vivo and evaluate targeted therapies for gliomas, we have genetically engineered glioma cells to visualize the dynamics of EGFR and targeted therapies in real time in vivo. Employing engineered lentiviral vectors bearing fusions between EGFR and its exon 2–7 deleted variant (EGFRvIII) with GFP and Renilla luciferase (Rluc), we show that there is a direct correlation between EGFR expression and glioma cell proliferation in the initial stages of glioma progression. In order to monitor and evaluate EGFR targeted therapies: (1) we have engineered short hairpin RNAs (shRNAs) and (2) anti-EGFR monoclonal antibody cetuximab which is approved for clinical use. Using EGFR–GFP–Rluc/Fluc–DsRed2 glioma model, we show that cetuximab results in a considerable reduction in glioma cell proliferation in culture and glioma burden in vivo that can be monitored at in real time at cellular resolution (Fig. 3). This study serves as a template to follow the role of growth factor receptor expression in tumor progression and to image therapeutic efficacy of targeted therapies in cancer.

Summary Optical or photonic-based imaging of brain tumors offers significant advantage over more traditional methods of imaging. These include the ability to design and apply pathway- and process-specific molecular probes which permit more sensitive and specific detection of the tumor. This is especially important in the management of primary brain tumors given that identification of residual tumor deposits assist in more effective surgical resection of the tumor while sparing normal brain structures. Also, these advanced imaging techniques, when coupled with molecular probes designed against tumor-specific pathways, permit the monitoring of novel molecularly targeted treatments. Pre-clinical evaluations of novel anti-brain tumor therapies in both in vitro and in vivo animal models have been greatly advanced by these novel imaging techniques via the application of fluorescenceand bioluminescence-tagged molecules and cells. Porphyrin-based fluorescence technology as both adjunct diagnostic and therapeutic tool has longest history in clinical neuro-oncology compared to all the other optical-based imaging technology. Its optimal use will likely be in the operating room where tissue planes are

Optical Imaging of Cancer: Neuro-oncologic Applications

261

Fig. 3 Monitoring the dynamics of EGFR targeted therapy in vivo: Mice with established EGFR– GFP–Rluc and Fluc–DsRed2 expressing gliomas were injected with cetuximab–cy5.5 and 3 days later were imaged by intravital microscopy. (A) Intravital fluorescent pictures of a day 4 EGFR– GFP–Rluc/Fluc–DsRed2 glioma: EGFR–GFP–Rluc (green); Fluc–DsRed2 (red) cetuximab–cy5.5 (blue). (B) Cetuximab–cy5.5 binding to the normal brain cells as compared to the glioma cells. (C, D, and E) Confocal images from histology on coronal brain sections sacrificed on day 6 after implantation: EGFR–GFP–Rluc (C) cetuximab–cy5.5 (D), and merged image (E). (F) Mice with established Fluc–DsRed2 expressing gliomas were treated with cetuximab–cy5.5 every 3 days for a period of 2 weeks and bioluminescence imaging was performed to quantify the effects of cetuximab on glioma proliferation in vivo. In panel B, data are mean±SD and ∗ P<0.05 vs. normal brain. In panel F, data are mean±SD and ∗ P<0.05 vs. Cetuximab–cy5.5 (Adapted from Arwert et al., (2007))

exposed due to the limited tissue penetration of the excitation light and in the detection of weak (relatively) emission by the photosensitizer. Nevertheless, continual development of novel pathway- and molecule-based optical probes in parallel with development of cancer pathway-specific therapeutic agents as well as ongoing advances in the engineering of optical imaging devices and computational models will definitely advance the field of neuro-oncological optical imaging.

262

S. Yip and K. Shah

References Aboody KS, Brown A, Rainov NG, Bower KA, Liu S, Yang W, et al. Neural stem cells display extensive tropism for pathology in adult brain: evidence from intracranial gliomas. Proc Natl Acad Sci U S A. 2000 Nov 7;97(23):12846–51. Arwert E, Hingtgen S, Figueiredo JL, Bergquist H, Mahmood U, Weissleder R, et al. Visualizing the dynamics of EGFR activity and antiglioma therapies in vivo. Cancer Res. 2007 Aug 1;67(15):7335–42. Bachoo RM, Maher EA, Ligon KL, Sharpless NE, Chan SS, You MJ, et al. Epidermal growth factor receptor and Ink4a/Arf: governing terminal differentiation and transformation stem cell to astrocyte axis. Cancer Cell. 2002;1(3):269–77. Beck TJ, Kreth FW, Beyer W, Mehrkens JH, Obermeier A, Stepp H, et al. Interstitial photodynamic therapy of nonresectable malignant glioma recurrences using 5-aminolevulinic acid induced protoporphyrin IX. Lasers Surg Med. 2007 Jun;39(5):386–93. Benedetti S, Pirola B, Pollo B, Magrassi L, Bruzzone MG, Rigamonti D, et al. Gene therapy of experimental brain tumors using neural progenitor. Nat Med. 2000;6(4): 447–50. Berg K, Selbo PK, Weyergang A, Dietze A, Prasmickaite L, Bonsted A, et al. Porphyrin-related photosensitizers for cancer imaging and therapeutic applications. J Microsc. 2005 May;218(Pt 2):133–47. Bhaumik S, Gambhir SS. Optical imaging of Renilla luciferase reporter gene expression in living mice. Proc Natl Acad Sci U S A. 2002 Jan 8;99(1):377–82. Brindle K. New approaches for imaging tumour responses to treatment. Nat Rev Cancer. 2008 Feb;8(2):94–107. Burgos JS, Rosol M, Moats RA, Khankaldyyan V, Kohn DB, Nelson MD, Jr., et al. Time course of bioluminescent signal in orthotopic and heterotopic brain tumors in nude mice. Biotechniques. 2003 Jun;34(6):1184–8. Chen X, Conti PS, Moats RA. In vivo near-infrared fluorescence imaging of integrin alphavbeta3 in brain tumor xenografts. Cancer Res. 2004 Nov 1;64(21):8009–14. Dinca EB, Sarkaria JN, Schroeder MA, Carlson BL, Voicu R, Gupta N, et al. Bioluminescence monitoring of intracranial glioblastoma xenograft: response to primary and salvage temozolomide therapy. J Neurosurg. 2007 Sep;107(3):610–6. Ehtesham M, Kabos P, Kabosova A, Neuman T, Black KL, Yu JS. The use of interleukin 12-secreting neural stem cells for the treatment of intracranial glioma. Cancer Res. 2002;62(20):5657–63. Gao X, Cui Y, Levenson RM, Chung LW, Nie S. In vivo cancer targeting and imaging with semiconductor quantum dots. Nat Biotechnol. 2004 Aug;22(8):969–76. Holland EC. Glioblastoma multiforme: the terminator. Proc Natl Acad Sci U S A. 2000;97(12):6242–4. Hsu AR, Hou LC, Veeravagu A, Greve JM, Vogel H, Tse V, et al. In vivo near-infrared fluorescence imaging of integrin alphavbeta3 in an orthotopic glioblastoma model. Mol Imaging Biol. 2006 Nov-Dec;8(6):315–23. Jackson H, Muhammad O, Daneshvar H, Nelms J, Popescu A, Vogelbaum MA, et al. Quantum dots are phagocytized by macrophages and colocalize with experimental gliomas. Neurosurgery. 2007 Mar;60(3):524–9; discussion 9-30. Jacobs AH, Kracht LW, Gossmann A, Ruger MA, Thomas AV, Thiel A, et al. Imaging in neurooncology. NeuroRx. 2005 Apr;2(2):333–47. Kemper EM, Leenders W, Kusters B, Lyons S, Buckle T, Heerschap A, et al. Development of luciferase tagged brain tumour models in mice for chemotherapy intervention studies. Eur J Cancer. 2006 Dec;42(18):3294–303. Kesari S, Ramakrishna N, Sauvageot C, Stiles CD, Wen PY. Targeted molecular therapy of malignant gliomas. Curr Neurol Neurosci Rep. 2005 May;5(3):186–97.

Optical Imaging of Cancer: Neuro-oncologic Applications

263

Krammer B, Plaetzer K. ALA and its clinical impact, from bench to bedside. Photochem Photobiol Sci. 2008 Mar;7(3):283–9. Lin WC, Mahadevan-Jansen A, Johnson MD, Weil RJ, Toms SA. In vivo optical spectroscopy detects radiation damage in brain tissue. Neurosurgery. 2005 Sep;57(3):518–25; discussion-25. Louis DN. Molecular pathology of malignant gliomas. Annu Rev Pathol. 2006;1:97–117. Lyons MK, Vora SA. Brain tumors: current issues in diagnosis and management. Semin Neurol. 2007 Sep;27(4):312–24. Madsen S, Hirschberg H. Photodynamic therapy and detection of high-grade gliomas. J Environ Pathol Toxicol Oncol. 2006;25(1–2):453–66. Mahmood U, Weissleder R. Near-infrared optical imaging of proteases in cancer. Mol Cancer Ther. 2003 May;2(5):489–96. Majumder SK, Gebhart S, Johnson MD, Thompson R, Lin WC, Mahadevan-Jansen A. A probability-based spectroscopic diagnostic algorithm for simultaneous discrimination of brain tumor and tumor margins from normal brain tissue. Appl Spectrosc. 2007 May;61(5): 548–57. Mannino S, Molinari A, Sabatino G, Ciafre SA, Colone M, Maira G, et al. Intratumoral vs systemic administration of meta-tetrahydroxyphenylchlorin for photodynamic therapy of malignant gliomas: assessment of uptake and spatial distribution in C6 rat glioma model. Int J Immunopathol Pharmacol. 2008 Jan–Mar;21(1):227–31. Messerli SM, Prabhakar S, Tang Y, Shah K, Cortes ML, Murthy V, et al. A novel method for imaging apoptosis using a caspase-1 near-infrared fluorescent probe. Neoplasia. 2004 MarApr;6(2):95–105. Michalet X, Pinaud FF, Bentolila LA, Tsay JM, Doose S, Li JJ, et al. Quantum dots for live cells, in vivo imaging, and diagnostics. Science. 2005 Jan 28;307(5709):538–44. Moore GE, Peyton WT, French LA. The clinical use of fluorescein in neurosurgery. The localization of brain tumors. J Neurosurg. 1948;5:392–8. Muhammad O, Popescu A, Toms SA. Macrophage-mediated colocalization of quantum dots in experimental glioma. Methods Mol Biol. 2007;374:161–71. Ntziachristos V, Ripoll J, Wang LV, Weissleder R. Looking and listening to light: the evolution of whole-body photonic imaging. Nat Biotechnol. 2005 Mar;23(3):313–20. Okada S, Ishii K, Yamane J, Iwanami A, Ikegami T, Katoh H, et al. In vivo imaging of engrafted neural stem cells: its application in evaluating the optimal timing of transplantation for spinal cord injury. FASEB J. 2005 Nov;19(13):1839–41. Okuda T, Kataoka K, Taneda M. Metastatic brain tumor surgery using fluorescein sodium: technical note. Minim Invasive Neurosurg. 2007 Dec;50(6):382–4. Popescu MA, Toms SA. In vivo optical imaging using quantum dots for the management of brain tumors. Expert Rev Mol Diagn. 2006 Nov;6(6):879–90. Price SJ. The role of advanced MR imaging in understanding brain tumour pathology. Br J Neurosurg. 2007 Dec;21(6):562–75. Scherer HJ. The forms of growth in gliomas and their practical significance. Brain. 1940;63:1–35. Shah K, Bureau E, Kim DE, Yang K, Tang Y, Weissleder R, et al. Glioma therapy and realtime imaging of neural precursor cell migration and tumor regression. Ann Neurol. 2005 Jan;57(1):34–41. Shah K, Hingtgen S, Kasmieh R, Figueiredo JL, Garcia-Garcia E, Martinez-Serrano A, et al. Bimodal viral vectors and in vivo imaging reveal the fate of human neural stem cells in experimental glioma model. J Neurosci. 2008 Apr 23;28(17):4406–13. Shah K, Tang Y, Breakefield X, Weissleder R. Real-time imaging of TRAIL-induced apoptosis of glioma tumors in vivo. Oncogene. 2003 Oct 9;22(44):6865–72. Shah K, Weissleder R. Molecular optical imaging: applications leading to the development of present day therapeutics. NeuroRx. 2005 Apr;2(2):215–25. Soling A, Theiss C, Jungmichel S, Rainov NG. A dual function fusion protein of Herpes simplex virus type 1 thymidine kinase and firefly luciferase for noninvasive in vivo imaging of gene therapy in malignant glioma. Genet Vaccines Ther. 2004 Aug 4;2(1):7.

264

S. Yip and K. Shah

Stefflova K, Chen J, Zheng G. Using molecular beacons for cancer imaging and treatment. Front Biosci. 2007a;12:4709–21. Stefflova K, Chen J, Zheng G. Killer beacons for combined cancer imaging and therapy. Curr Med Chem. 2007b;14(20):2110–25. Stummer W, Beck T, Beyer W, Mehrkens JH, Obermeier A, Etminan N, et al. Long-sustaining response in a patient with non-resectable, distant recurrence of glioblastoma multiforme treated by interstitial photodynamic therapy using 5-ALA: case report. J Neurooncol. 2008 Mar;87(1):103–9. Stylli SS, Kaye AH. Photodynamic therapy of cerebral glioma – a review. Part II – clinical studies. J Clin Neurosci. 2006 Aug;13(7):709–17. Sugawa N, Ekstrand AJ, James CD, Collins VP. Identical splicing of aberrant epidermal growth factor receptor transcripts from amplified rearranged genes in human glioblastomas. Proc Natl Acad Sci U S A. 1990 Nov;87(21):8602–6. Szentirmai O, Baker CH, Lin N, Szucs S, Takahashi M, Kiryu S, et al. Noninvasive bioluminescence imaging of luciferase expressing intracranial U87 xenografts: correlation with magnetic resonance imaging determined tumor volume and longitudinal use in assessing tumor growth and antiangiogenic treatment effect. Neurosurgery. 2006 Feb;58(2):365–72; discussion -72. Tamura Y, Kuroiwa T, Kajimoto Y, Miki Y, Miyatake S, Tsuji M. Endoscopic identification and biopsy sampling of an intraventricular malignant glioma using a 5-aminolevulinic acidinduced protoporphyrin IX fluorescence imaging system. Technical note. J Neurosurg. 2007 Mar;106(3):507–10. Tang Y, Shah K, Messerli SM, Snyder E, Breakefield X, Weissleder R. In vivo tracking of neural progenitor cell migration to glioblastomas. Hum Gene Ther. 2003 Sep 1;14(13):1247–54. Utsuki S, Miyoshi N, Oka H, Miyajima Y, Shimizu S, Suzuki S, et al. Fluorescence-guided resection of metastatic brain tumors using a 5-aminolevulinic acid-induced protoporphyrin IX: pathological study. Brain Tumor Pathol. 2007;24(2):53–5. Wang G, Cong W, Shen H, Qian X, Henry M, Wang Y. Overview of bioluminescence tomography – a new molecular imaging modality. Front Biosci. 2008;13:1281–93. Wang J, Yong WH, Sun Y, Vernier PT, Koeffler HP, Gundersen MA, et al. Receptortargeted quantum dots: fluorescent probes for brain tumor diagnosis. J Biomed Opt. 2007 Jul–Aug;12(4):044021. Weissleder R, Pittet MJ. Imaging in the era of molecular oncology. Nature. 2008 Apr 3;452(7187):580–9. Wessels JT, Busse AC, Mahrt J, Dullin C, Grabbe E, Mueller GA. In vivo imaging in experimental preclinical tumor research – a review. Cytometry A. 2007 Aug;71(8):542–9. Winkeler A, Sena-Esteves M, Paulis LE, Li H, Waerzeggers Y, Ruckriem B, et al. Switching on the lights for gene therapy. PLoS ONE. 2007;2(6):e528. Xing Y, Chaudry Q, Shen C, Kong KY, Zhau HE, Chung LW, et al. Bioconjugated quantum dots for multiplexed and quantitative immunohistochemistry. Nat Protoc. 2007;2(5):1152–65. Yip S, Sabetrasekh R, Sidman RL, Snyder EY. Neural stem cells as novel cancer therapeutic vehicles. Eur J Cancer. 2006 May 11. Zacharakis G, Kambara H, Shih H, Ripoll J, Grimm J, Saeki Y, et al. Volumetric tomography of fluorescent proteins through small animals in vivo. Proc Natl Acad Sci U S A. 2005 Dec 20;102(51):18252–7.

Index

A ACPPs, see Activatable cell-penetrating peptides Activatable cell-penetrating peptides, 127 Activity-based probes, 123, 127–128 for C-clan proteases, 128–129 for S-clan and M-clan proteases, 129 Acyloxymethyl ketones, 129 Adaptive immune therapy, reporter gene imaging, 154–155 Adeno-associated viruses, in clinical imaging, 148, 151–153 Adenomatous polyposis coli multiple intestinal neoplasia, 126 Adenoviral (Ad) vector, 84, 88, 148, 151, 154 Airy disk, defined, 8 5-ALA, see 5-aminolevulinic acid 5-Aminolevulinic acid, 49, 164 dose, 53 oral administration, 167 topical administration, 165–166 Antibodies, in molecular imaging of tumors, 171–173 Anti-EGFR, see Anti-epidermal growth factor receptor Anti-epidermal growth factor receptor, 172 Anti-HER2-directed antibody, application, 172 AOMK, see Acyloxymethyl ketones APC Min , see Adenomatous polyposis coli multiple intestinal neoplasia Aquaria victoria, 143 Autofluorescence definition, 60 in tissue, 101, 144 See also Fluorophores, for optical imaging Autofluorescent bronchoscopy in lung cancer, 26–29 role, 106 Avastin protein, 172

B Bacterial luciferases, 141 Barrett’s esophagus dysplasia, OCT imaging, 225 Basal cell carcinoma, 108, 228 BBB, see Blood brain barrier BCC, see Basal cell carcinoma Beetle luciferases, 140–141 Bimodal fluorescence, in brain tumor detection, 258–260 Bioluminescence imaging, 257 for brain tumor detection, 258–260 clinical applications, 153–156 defined, 140 limitations and future prospects, 156 molecular vectors, 147–153 in protein-protein interaction study, 145–147 reporters, 140–145 See also Optical imaging Bioluminescence resonance energy transfer, 146 Bladder cancer, OCT imaging, 226–228 See also Optical coherence tomography BLI, see Bioluminescence imaging Blood brain barrier, 252 Blood flow visualization, FGS, 49–53 See also Fluorescence-guided surgery (FGS), surgical microscope Bortezomib drug, in mantle cell lymphoma treatment, 122 Brain tumor detection, optical imaging bimodal fluorescence and bioluminescence imaging, 258–260 clinical evaluation, imaging techniques, 252–253 fluorescence imaging, 254–257 quantum dots, 257–258 in vivo fluorescence imaging of, 145

265

266 BRCL, see Breast cancer-related lymphedema Breast cancer, OCT imaging, 215–224 See also Optical coherence tomography Breast cancer-related lymphedema, 187 BRET, see Bioluminescence resonance energy transfer C CAD, see Computer-aided diagnosis Cadmium selenide (CdSe), 67 Cancer ABP-based imaging agents, 130–131 detection and monitoring strategies, 81–83 clinical development, 94 clinical implementation of optical imaging, 90–92 fluorescence imaging, 94 ideal imaging probe, 83–90 imaging probe development, 95 optical imaging, 92–94 detection, proteinase optical imaging tools for future prospects, 132–133 imaging proteinases, 122–129 proteases and cancer, 119–122 tumors, imaging protease activity in, 129–132 and proteases, 119–122 substrate-based imaging agents in, 131–132 treatment, endoscopic techniques for, 25–28 catheter navigation and guidance, 37–39 conventional optical imaging, 28–31 future prospects, 39–40 micro-optical imaging, 32–37 Cancer detection and treatment, optical imaging clinical applications, 173–180 photodynamic techniques, 163 aminolevulinic acid and hexaminolevulinate, 164–165 clinical studies, 165–168 target mechanisms for, 168, 170–171 tumor-specific ligands and antibodies, 171–173 Cancer detection, optical coherence tomography in bladder cancer, 226–228 in breast cancer, 215–224 clinical applications, 210–211, 238–240 and computer-aided tissue identification techniques, 230–232

Index future prospects, 240–241 in gastrointestinal cancers, 224–226 multimodal optical imaging, 232–234 in oral cancer, 229–230 principles low-coherence interferometry, 212–213 and oncological tissue optics, 214–215 system performance, 213–214 in skin cancer, 228–229 technology, translation, 211 Cancer nodal staging, optical imaging limitations, 186–187 near-infrared fluorescence imaging, 192–194 approaches and methodologies, 194–195 and scintigraphy, 195–196 radio and optical-immunoscintigraphy, 198–201 research technique for imaging, 189–192 sentinel lymph node assessment, 187–189 Cathepsin D and breast cancer, 121 Catheter-based confocal microscopy, 35–37 CBCM, see Catheter-based confocal microscopy CCD, see Charge-coupled device C-clan proteases, ABPs, 128–129 Cell trafficking, imaging, 155 Central nervous system (CNS), malignancies bimodal fluorescence and bioluminescence imaging, 258–260 clinical evaluation, imaging techniques for, 252–253 evaluation, imaging techniques for, 252–253 fluorescence imaging, 254–257 quantum dots, 257–258 Cervical cancer, tissue autofluorescence, 106–108 Cervical intraepithelial neoplasm, 164 Cetuximab, usage, 173 Charge-coupled device, 142, 194 Chemiluminescence, defined, 140 Chlorotoxin, for tumor molecular imaging, 171 CIK, see Cytokine-induced killer cells CIN, see Cervical intraepithelial neoplasm Clinical imaging, in biomedical science, 139 limitations and future prospects, 156 molecular vectors in, 147–153 optical reporter genes, 140–145 protein–protein interaction study, 145–147 See also Optical imaging

Index CLSM, see Confocal laser scanning microscopy Colon cancer, OCT imaging, 225–226 Computed tomography (CT), 185, 251 Computer-aided diagnosis, 224 Computer-aided tissue identification techniques, 230–232 Confocal endo-microscopy, role, 27, 33–35 Confocal laser scanning microscopy, 11–13 Conventional optical imaging, in cancer, 28–31 See also Cancer Conventional pulmonary imaging, 26 C-protease cathepsin B, expression, 121 C-protease cathepsin K, production, 120–121 Cytokine-induced killer cells, 155 D DCIS, see Ductal carcinoma in situ Diffusion-weighted MRI, 93 Direct imaging, 139 Doppler OCT, application, 236–237 DsRed fluorescent proteins, 143–144 Ductal carcinoma in situ, 218 DWI, see Diffusion-weighted MRI E Endoscopic fluorescence detection, of lung cancer, 25 Endoscopic techniques, for cancer treatment, 25–28 catheter navigation and guidance, 37–39 conventional optical imaging, 28–31 future prospects, 39–40 micro-optical imaging, 32–37 Enhancing endogenous contrast and OCT imaging, 234–236 Exogenous contrast and OCT, in biological tissue imaging, 236–238 F FAD, see Flavin adenine dinucleotide F-18-FDG PET imaging, role, 93 F-18-FDG, usage, 92 FGOCT, see Fluorescence-guided OCT Fine-needle aspiration, 223 Firefly (Photinus pyralis), 140 Flavin adenine dinucleotide, 103 FLIM, see Fluorescence Lifetime Imaging Microscopy FL400 oncology fluorescence surgical microscope, components, 56 Fluorescein bronchoscopy, in retina circulation, 30–31 Fluorescence-guided OCT, 233

267 Fluorescence-guided surgery (FGS), surgical microscope, 49 for blood flow visualization, 49–53 design, 57–58 GBM tumors, surgery, 53–57 Fluorescence imaging, 4–8 advantages and distinctions, 94 applications, 15–16 controls and image processing, 22 future prospects, 22–23 limitations, 8–11 Fluorescence Lifetime Imaging Microscopy, 19–20 Fluorescence lifetime, measurement, 20–22 Fluorescence-mediated tomography, 130 Fluorescence molecular tomography, 144, 254 Fluorescence quenching, causes, 65 See also Fluorophores, for optical imaging Fluorescence Recovery After Photobleaching, 16–17 Fluorescence Resonance Energy Transfer, 17–19, 146 Fluorescence visualization (FV), applications, 108 Fluorescent dyes, features, 66 Fluorescent operating microscope, functions, 49–53 Fluorescent probes, definition, 60 See also Fluorophores, for optical imaging Fluorophores, for optical imaging, 59 definitions, 60–61 fluorescent probes, characteristics chemical and physical stability, 68–69 conjugation and targeting, 69–70 distribution and pharmacokinetics, 70–72 emission spectrum, 63–64 environment sensitivity and quenching, 65–67 excitation spectrum and extinction coefficients, 61–63 photostability, 74 size, 67 toxicity, 72–74 Fluorophores, in cancer diagnosis, 103 See also Cancer FL800 vascular fluorescence surgical microscope, components, 52 FMT, see Fluorescence-mediated tomography; Fluorescence molecular tomography FNAB, see Fine-needle aspiration F¨orster Resonance Energy Transfer, 124

268 FRAP, see Fluorescence Recovery After Photobleaching FRET, see Fluorescence Resonance Energy Transfer; F¨orster Resonance Energy Transfer FV application, in tumor treatment, 108–113 See also Oral cancer detection FVL, see FV loss FV loss, 109 FVR, see FV retained FV retained, 108

Index

G Gastrointestinal cancers, OCT imaging for, 224–226 See also Optical coherence tomography Gaussia luciferase, 141–142 Gene therapy, reporter gene imaging, 154–155 GFP, see Green fluorescent protein Glial tumors, 251 Glioblastoma Multiforme (GBM), oncology fluorescence-guided surgery of, 53–57 Green fluorescent protein, 18, 88, 143

I IBD, see Inflammatory bowel disease ICG, see Indocyanine green ICG fluorophore, usage, 90 IC-Green, see Indocyanine green Imaging biomarker, definition, 81 Imaging probe development, in cancer studies, 95 See also Cancer, detection and monitoring strategies Immunohistochemistry (IHC), 186 Indirect imaging, 139 Indocyanine green, 49, 194 Inflammatory bowel disease, 233 Institutional Review Board, 194 Internal ribosome entry site, 147 Intraoperative lymph node imaging, 221–223 Intraoperative tumor margin detection, 220–221 Invadopodia, definition of, 121 Invasive ductal carcinoma, OCT image of, 219 IRB, see Institutional Review Board IRES, see Internal ribosome entry site Isosulfan blue dye, in sentinel lymph nodes identification, 190

H HAL, see Hexaminolevulinate Head and Neck Squamous Cell Carcinoma, 175–176 Hematoxylin and Eosin staining, 186 Hemoglobin biosynthesis pathway, 165 HER2, see Human epidermal growth factor receptor-2 Herceptin-labeled antibodies, usage, 172 Herpes simplex virus, 148 H&E staining, see Hematoxylin and Eosin staining Hexaminolevulinate, 164 oral administration, 167 topical administration, 165–166 HGLs, see High-grade lesions High-grade lesions, 109 HNSCCs, see Head and Neck Squamous Cell Carcinoma hNSCs, see Human neural stem cells HSDI, see Hyperspectral diagnostic imaging HSV, see Herpes simplex virus Human degradome, protease classification, 120 Human epidermal growth factor receptor-2, 193 Human neural stem cells, 258 Hyperspectral diagnostic imaging, 107

L Large-core needle biopsy, 223 Laser-induced fluorescence imaging, 233 LCI-guided needle biopsy, mechanisms for, 223 LCNB, see Large-core needle biopsy LDA, see Linear discriminant analysis LECs, see Lymph endothelial cells Leica fluorescent stereomicroscope, 176 Leica FL800, usage, 49 Lentiviruses, in clinical imaging, 151–152 LGLs, see Low-grade lesions LIFE, see Light-induced fluorescence endoscopy LIF imaging, see Laser-induced fluorescence imaging Light-induced fluorescence endoscopy, 106 LightTouch cervical cancer test, advantage, 107 Linear discriminant analysis, 232 LOH, see Loss of heterozygosity Long-terminal repeat, 151 Loss of heterozygosity, 110 Low-grade lesions, 109 LTR, see Long-terminal repeat Luciferase, 140 LUMA cervical imaging system, advantage, 107

Index Lung cancer autofluorescent bronchoscopy, 26–27 catheter navigation and guidance, 37–39 CBCM, 35–37 endoscopic fluorescence detection, 25 MDCT, 25–26 micro-optical imaging techniques, 27 tissue autofluorescence, 106–108 Lymphangiography techniques, for tumor nodal staging, 190–191 Lymph endothelial cells, 189–190 Lymph node (LN), 185 imaging, 195–201 scintigraphy and NIR fluorescence imaging, 195–197 See also Cancer nodal staging, optical imaging Lymphoscintigraphy, in sentinel lymph nodes identification, 190 M Maestro Spectral imaging systems, 144 Magnetic resonance imaging (MRI), 93 Magnetic resonance spectroscopy, 185, 252 Marine luciferases, 141 Matrix metalloproteinase, 105, 120 M-clan proteases, ABPs for, 129 MDCT, see Multi detector computed tomography MEMS, see Microelectromechanical systems Mesenchymal stem cells, 258 Methyl-nitroso-urea, 226 Microelectromechanical systems, 226 Micro-optical imaging in cancer, 32–37 techniques, in lung cancer, 27 See also Cancer Mirror image rule, definition, 6 MMP, see Matrix metalloproteinase MNU, see Methyl-nitroso-urea Modified mammalian two-hybrid reporter system, 145–146 MRS, see Magnetic resonance spectroscopy MSC, see Mesenchymal stem cells Multidetector computed tomography, 25–26 Multimodality reporters, 147 Multimodal optical imaging, for cancer detection, 232–234 Multi-optical wand (MOW), 107 Multiphoton imaging, 13

269 Multiplex detection, definition, 61 See also Fluorophores, for optical imaging N NADH, see Nicotinamide adenine dinucleotide Narrow-band imaging, 29 NBI, see Narrow-band imaging Near-infrared fluorescence, 124 Near Infrared Fluorescence Reflectance Imaging, 255–256 Needle-based biopsy guidance, of OCT, 223–224 Neural stem cells, 258 Nicotinamide adenine dinucleotide, 103 NIRF, see Near-infrared fluorescence NIRF-i, see Near Infrared Fluorescence Reflectance Imaging Noninvasive molecular imaging, 139 See also Optical imaging NSC, see Neural stem cells Numerical aperture (NA), 4 O OCPL, see Oral cancer prediction longitudinal OCT, see Optical coherence tomography OFDI, see Optical frequency domain imaging Optical coherence tomography, 32–33, 209 in bladder cancer, 226–228 in breast cancer, 215–224 clinical applications, 210–211, 238–240 and computer-aided tissue identification techniques, 230–232 future prospects, 240–241 in gastrointestinal cancers, 224–226 multimodal optical imaging, 232–234 in oral cancer, 229–230 principles low-coherence interferometry, 212–213 and oncological tissue optics, 214–215 system performance, 213–214 in skin cancer, 228–229 technology, translation, 211 See also Optical imaging Optical frequency domain imaging, 32–33 Optical imaging for brain tumor detection bimodal fluorescence and bioluminescence imaging, 258–260 clinical evaluation, imaging techniques for, 252–253 evaluation, imaging techniques for, 252–253 fluorescence imaging, 254–257 quantum dots, 257–258

270 Optical imaging (cont.) for cancer detection and treatment, 92–94 clinical applications, 173–180 photodynamic techniques, 163–168 target mechanisms, 168, 170–171 tumor-specific ligands and antibodies, 171–173 for cancer nodal staging limitations, 186–187 near-infrared fluorescence imaging, 192–196 radio and optical-immunoscintigraphy, 198–201 research technique for imaging, 189–192 sentinel lymph node assessment, 187–189 characteristics of techniques, 253 in neuro-oncology, in vivo, 253 preclinical applications, 153 probe, in tumors detection, 83–90 (see also Cancer, detection and monitoring strategies) targeting agents, 169 usage, 172 Optical imaging, fluorophores, 59 definitions, 60–61 fluorescent probes, characteristics chemical and physical stability, 68–69 conjugation and targeting, 69–70 distribution and pharmacokinetics, 70–72 emission spectrum, 63–64 environment sensitivity and quenching, 65–67 excitation spectrum and extinction coefficients, 61–63 photostability, 74 size of, 67 toxicity, 72–74 Optical-immunoscintigraphy, for cancer nodal staging, 198–201 Optical reporter genes, bioluminescence reporters, 140–145 clinical applications, 153–156 limitations and future prospects, 156 molecular vectors in, 147–153 in protein-protein interaction study, 145–147 Oral cancer detection OCT imaging for, 229–230 (see also Optical coherence tomography) tissue autofluorescence in, 101–102

Index applications, 106–108 biology, 102–106 clinical utilities, 108–113 future prospects, 113–114 Oral cancer prediction longitudinal, 108 P PALM, see Photo activated localization microscopy PCA, see Principle component analysis; Protein-fragment complementation assays PDT, see Photodynamic therapy PEG, see Polyethylene glycol Pelvic lymph node dissections, 188 PET, see Photo-induced electron transfer; Positron emission tomography Phosphomonoester (PME), 93 Photo activated localization microscopy, 22 Photobleaching and fluorescent probes, 74 See also Fluorophores, for optical imaging Photodynamic diagnosis techniques, for cancer detection, 163 aminolevulinic acid and hexaminolevulinate, 164–165 clinical studies oral administration, 167 photodynamic therapy, 168 topical administration, 165–166 See also Cancer detection and treatment, optical imaging Photodynamic therapy, 58, 164, 168, 255 Photofrin, application, 168 Photo-induced electron transfer, 65 Photomultiplier tube, 12 Plasmids, in clinical imaging, 148 PLNDs, see Pelvic lymph node dissections PMT, see Photomultiplier tube Point-spread function, definition, 10 See also Fluorescence imaging Polarization-sensitive OCT, 235 Polyethylene glycol, 69 Porphyrin fluorescence, role, 104 Positron emission tomography, 92, 170, 185, 252 PpIX, see Protoporphyrin IX Premalignant lesions detection tissue autofluorescence, 101–102 applications, 106–108 biology, 102–106 clinical utilities, 108–113 future prospects, 113–114 Primary brain tumors, 251

Index Primary tumors, optical imaging clinical applications for, 173–180 photodynamic diagnosis, clinical studies aminolevulinic acid and hexaminolevulinate, 164–165 oral administration, 167 photodynamic therapy, 168 topical administration, 165–166 target mechanisms for, 168, 170–171 tumor-specific ligands and antibodies, 171–173 Principle component analysis, 232 Protease-dependent agents, 168 Proteases and cancer, 119–122 Proteinase optical imaging, for cancer detection future prospects, 132–133 imaging proteinases, 122–129 proteases and cancer, 119–122 tumors, imaging protease activity in, 129–132 Protein-fragment complementation assays, 147 Proteolytic beacons (PBs), 124 Protoporphyrin IX, 49 PS-OCT, see Polarization-sensitive OCT Pulmonary disease management, macro-optical imaging techniques in, 26 Pulmonary metastasis, cetuximab–Cy5.5 detection of, 178 Q QDots, see Quantum dot nanocrystals Quantum dot nanocrystals, 60, 72 See also Fluorophores, for optical imaging Quantum dots (qdots), biological applications, 257–258 R Radioimmunoscintigraphy techniques, for tumor nodal staging, 191–192, 198–201 Rayleigh criterion, definition, 10 Red fluorescent protein, 88, 143–145 Remicalm LLC, usage, 107 Renilla luciferase, 260 Reporter gene imaging, disadvantages, 155–156 role of, 139–140 Retroviruses, in clinical imaging, 151 Reverse transcription polymerase chain reaction, 186 RFP, see Red fluorescent protein Rluc, see Renilla luciferase

271 R-PE, see R-phycoerythrin R-phycoerythrin, 63 RT-PCR, see Reverse transcription polymerase chain reaction S Scintigraphy and NIR fluorescence imaging, of lymph nodes, 195–196 S-clan proteases, ABPs, 129 Sentinel lymph node biopsy, 186–189, 217–218 Short hairpin RNAs, 260 shRNAs, see Short hairpin RNAs Signal to noise ratio, 142 Single photon emission computed tomography, 92, 170, 191 Skin cancer OCT imaging for, 228–229 (see also Optical coherence tomography) tissue autofluorescence in, 106–108 SLNB, see Sentinel lymph node biopsy SNR, see Signal to noise ratio SPECT, see Single photon emission computed tomography Split reporter system, 146–147 STED, see Stimulated emission depletion microscopy Stimulated emission depletion microscopy, 22 Stokes shift, definition, 5 Superparamagnetic MR contrast agents, for tumor nodal staging, 190–191 Surgical fluorescence microscope procedures, designing in, 57–58 See also Fluorescence-guided surgery (FGS), surgical microscope Swept source optical coherence tomography, see Optical frequency domain imaging T THP, see Triple-helical peptide Tissue autofluorescence in oral cancer and premalignant lesions detection, 101–102 applications, 106–108 biology in, 102–106 clinical utilities, 108–113 future prospects, 113–114 Tissue plasminogen activator, 131 TNM disease classification, 185 Toxicity and fluorescent probes, 72–74 See also Fluorophores, for optical imaging tPA, see Tissue plasminogen activator

272 TRAIL, see Tumor necrosis factor apoptosis inducing ligand Triple-helical peptide, 127 Tumor and cancer biology, bioluminescence imaging in, 153–154 Tumor cells target mechanisms for, 168, 170–171 tumor-specific ligands and antibodies, 171–173 Tumor necrosis factor apoptosis inducing ligand, 258 Tumor of central nervous system clinical significance and imaging for, 251–252 Tumors, imaging protease activity in, 129–132 U Ubiquitin-specific proteases, 128, 131 Ultrasmall iron oxide particles, 190–191 Ultrasound (US), 185 UPA, see Urokinase-type plasminogen activator Urokinase-type plasminogen activator, 121

Index USIOP, see Ultrasmall iron oxide particles USPs, see Ubiquitin-specific proteases V Vascular endothelial growth factor, 172 VAT, see Video assisted thorascopy VEGF, see Vascular endothelial growth factor R VELScope , role, 108 Video assisted thorascopy, 174 Viral vectors, in clinical imaging, 148 W White light bronchoscopy, 28–30 WLB, see White light bronchoscopy X X-ray imaging, role, 217 Y Yeast two-hybrid reporter system, 145–146 Z Z-axis resolution, 13–15 See also Fluorescence imaging