Imaging in Oncology
Cancer Treatment and Research Steven T. Rosen, M.D., Series Editor Bergan, R. C. (ed.): Cancer Chemoprevention. 2001. ISBN 0-7923-7259-X. Raza, A., Mundle, S.D. (eds): Myelodysplastic Syndromes & Secondary Acute Myelogenous Leukemia 2001. ISBN: 0-7923-7396. Talamonti, M. S. (ed.): Liver Directed Therapy for Primary and Metastatic Liver Tumors. 2001. ISBN 0-7923-7523-8. Stack, M.S., Fishman, D.A. (eds): Ovarian Cancer. 2001. ISBN 0-7923-7530-0. Bashey, A., Ball, E.D. (eds): Non-Myeloablative Allogeneic Transplantation. 2002. ISBN 0-7923-7646-3. Leong, S. P.L. (ed.): Atlas of Selective Sentinel Lymphadenectomy for Melanoma, Breast Cancer and Colon Cancer. 2002. ISBN 1-4020-7013-6. Andersson, B., Murray D. (eds): Clinically Relevant Resistance in Cancer Chemotherapy. 2002. ISBN 1-4020-7200-7. Beam, C. (ed.): Biostatistical Applications in Cancer Research. 2002. ISBN 1-4020-7226-0. Brockstein, B., Masters, G. (eds): Head and Neck Cancer. 2003. ISBN 1-4020-7336-4. Frank, D.A. (ed.): Signal Transduction in Cancer. 2003. ISBN 1-4020-7340-2. Figlin, R. A. (ed.): Kidney Cancer. 2003. ISBN 1-4020-7457-3. Kirsch, M.; Black, P. McL. (ed.): Angiogenesis in Brain Tumors. 2003. ISBN 1-4020-7704-1. Keller, E.T., Chung, L.W.K. (eds): The Biology of Skeletal Metastases. 2004. ISBN 1-4020-7749-1. Kumar, R. (ed.): Molecular Targeting and Signal Transduction. 2004. ISBN 1-4020-7822-6. Verweij, J., Pinedo, H.M. (eds): Targeting Treatment of Soft Tissue Sarcomas. 2004. ISBN 1-4020-7808-0. Finn, W.G., Peterson, L.C. (eds.): Hematopathology in Oncology. 2004. ISBN 1-4020-7919-2. Farid, N. (ed.): Molecular Basis of Thyroid Cancer. 2004. ISBN 1-4020-8106-5. Khleif, S. (ed.): Tumor Immunology and Cancer Vaccines. 2004. ISBN 1-4020-8119-7. Balducci, L., Extermann, M. (eds): Biological Basis of Geriatric Oncology. 2004. ISBN Abrey, L.E., Chamberlain, M.C., Engelhard, H.H. (eds): Leptomeningeal Metastases. 2005. ISBN 0-387-24198-1 Platanias, L.C. (ed.): Cytokines and Cancer. 2005. ISBN 0-387-24360-7. Leong, S.P.L., Kitagawa, Y., Kitajima, M. (eds): Selective Sentinel Lymphadenectomy for Human Solid Cancer. 2005. ISBN 0-387-23603-1. Small, Jr. W., Woloschak, G. (eds): Radiation Toxicity: A Practical Guide. 2005. ISBN 1-4020-8053-0. Haefner, B., Dalgleish, A. (eds): The Link Between Inflammation and Cancer. 2006. ISBN 0-387-26282-2. Leonard, J.P., Coleman, M. (eds): Hodgkin’s and Non-Hodgkin’s Lymphoma. 2006. ISBN 0-387-29345. Leong, S.P.L. (ed): Cancer Clinical Trials: Proactive Strategies. 2006. ISBN 0-387-33224-3. Meyers, C. (ed): Aids-Associated Viral Oncogenesis. 2007. ISBN 978-0-387-46804-4. Ceelen, W.P. (ed): Peritoneal Carcinomatosis: A Multidisciplinary Approach. 2007. ISBN 978-0-387-48991-9. Leong, S.P.L. (ed): Cancer Metastasis and the Lymphovascular System: Basis for rational therapy. 2007. ISBN 978-0-387-69218-0. Raizer, J., Abrey, L.E. (eds): Brain Metastases. 2007. ISBN 978-0-387-69221-0. Woodruff, T., Snyder, K.A. (eds): Oncofertility. 2007. ISBN 978-0-387-72292-4. Angelos, P. (ed): Ethical Issues in Cancer Patient Care, Second Edition. 2008. ISBN 978-0-387-73638-9. Ansell, S. (ed): Rare Hematological Malignancies. 2008. ISBN 978-0-387-73743-0. Gradishar, W.J., Wood,W.C. (eds): Advances in Breast Care Management, Second Edition. 2008. 978-0-387-73160-5. Blake, M.A., Kalra, M.K. (eds): Imaging in Oncology. 2008. ISBN 978-0-387-75586-1.
Michael A. Blake, MRCPI, FFR (RCSI), FRCR Mannudeep K. Kalra, MD Editors
Imaging in Oncology
Michael A. Blake Department of Radiology Massachusetts General Hospital Boston, Massachusetts, United States
Mannudeep K. Kalra Department of Radiology Massachusetts General Hospital Boston, Massachusetts, United States
Series Editor: Steven T. Rosen Robert H. Lurie Comprehensive Cancer Center Northwestern University Chicago, IL United States Imaging in Oncology
ISBN-13: 978-0-387-75586-1
e-ISBN-13: 978-0-387-75587-8
Library of Congress Control Number: 2007936302 © 2008 Springer Science+Business Media, LLC 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. Printed on acid-free paper 9 8 7 6 5 4 3 2 1 springer.com
Preface
During the past two decades, and even more so in the last five years, radiology has evolved at a tremendous pace, and imaging technology continues to make great advances into morphological, as well as functional, aspects of oncologic diseases. Developments in computed tomography (CT) have led to the introduction to ultrafast, high-resolution single-source and dual-source multislice scanners. Positron emission tomography (PET) has stepped into the clinical limelight with the availability of vastly improved structural co-registration and overall improved diagnostic performance from recently developed PET-CT hybrid scanners. Magnetic resonance imaging (MRI) has become faster and more versatile with high magnetic strength systems, MR spectroscopy, diffusion weighted MRI, and flow mapping. Oncologic imaging guided interventional techniques such as radio frequency ablation, microwave and cryoablation procedures have also progressed immensely. From an oncologic point of view, these developments have improved patient care. Today, the role of imaging extends beyond traditional detection, localization, characterization, staging, follow-up and treatment of patients with cancer. CT is currently being investigated as a screening tool for colon and lung cancer. MRI has emerged as a modality of choice for imaging many cancers including hepatic, adrenal and most musculoskeletal cancers. Hybrid PET-CT scanners provide combined morphologic and functional information for tumor detection, and assessment of early tumor response to treatment. The growing, dynamic collaboration between the radiologic and oncologic communities is important to foster to ensure cancer patients receive optimal care. This book, “Imaging in Oncology,” describes the current status of imaging techniques in oncology, with the help of specialized contributions from world-renowned oncologic imaging experts.
v
Summary Statement
Scholarly overview of cancer imaging, incorporating the most recent research and clinical advances in oncologic radiology. Aims: ●
● ●
●
● ●
To serve as an up-to-date, attractive, broad overview book of oncologic imaging for radiologists and all involved in oncologic care, particularly oncologists and radiation therapists. To demonstrate the importance of oncologic imaging in medicine and surgery. To provide pertinent clinical and research information that underpins accurate interpretation and sensible use of cancer imaging. To review established oncologic imaging findings, algorithms and techniques in plain radiography, Ultrasound, CT, MR, Nuclear Medicine, PET and PET/CT, as well as image guided intervention. To highlight new developments and advances in oncologic imaging. To appeal to physicians in practice and in training, and to all interested in oncologic imaging.
Scope: ●
●
●
Scholarly, pertinently illustrated, well-referenced text with chapter contributions from world-renowned cancer imaging experts. Guidelines for each chapter provided by editors to provide for a coordinated, integrated text. Summary sections included in each chapter illuminating its key points
vii
Contents
Preface. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
v
Summary Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
vii
Section I 1
Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sanjeev Chawla, Harish Poptani, and Elias R. Melhem
3
2 Imaging of Spinal Tumors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Izlem Izbudak, Aylin Tekes, Juan Carlos Baez, and Kieran Murphy
43
3 PET Imaging of Brain Tumors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alan J. Fischman
67
4 Extracranial Head and Neck Neoplasms: Role of Imaging . . . . . . . . . . Myria Petrou and Suresh K. Mukherji
93
Section II 5
Imaging of Thoracic Malignancies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 Subba R. Digumarthy and Suzanne L. Aquino
6
Imaging of Mediastinal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Scott Moore, Hetal Dave-Verma, and Ajay Singh
7
Imaging Cardiac Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 Mannudeep K. Kalra and Suhny Abbara
ix
x
Contents
Section III 8
Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 Unni Udayasankar, Abbas Chamsuddin, Pardeep Mittal, and William C. Small
9
Recent Advances in Imaging of Pancreatic Neoplasms . . . . . . . . . . . . 229 Chad B. Rabinowitz, Hima B. Prabhakar, and Dushyant V. Sahani
10
Imaging of Colorectal Carcinoma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255 Jorge A. Soto
11 Imaging of Primary Malignant Tumors of Peritoneal and Retroperitoneal Origin . . . . . . . . . . . . . . . . . . . . . . . 281 J. Louis Hinshaw and Perry J. Pickhardt 12
Imaging of Urinary Tract Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 Michael A. Blake and Mannudeep K. Kalra
13 Current Status of Imaging for Adrenal Malignant Involvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 Michael A. Blake and Mannudeep K. Kalra 14
Recent Advances in Imaging of Male Reproductive Tract Malignancies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331 Jurgen J. Fütterer and J. Roan Spermon
Section IV 15
Imaging of Malignant Skeletal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . 367 Jay Pahade, Aarti Sekhar, and Sanjay K. Shetty
16
Radiology of Soft Tissue Tumors Including Melanoma. . . . . . . . . . . . 423 M.J. Shelly, P.J. MacMahon, and S. Eustace
Section V 17 Reticuloendothelium Malignancy: Current Role of Imaging . . . . . . . 455 Sunit Sebastian and Brian C. Lucey
Contents
xi
18
Pediatric Malignancies: Synopsis of Current Imaging Techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 469 Sabah Servaes, Monica Epelman, Avrum Pollock, and Karuna Shekdar
19
Interventional Radiology in Oncology . . . . . . . . . . . . . . . . . . . . . . . . . . 493 O.J. O’Connor, J. M. Buckley, and M. M. Maher
Section VI 20
Breast Tumor Imaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 515 Deirdre Coll
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 547
1
Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology Sanjeev Chawla, Harish Poptani, and Elias R. Melhem
1
Introduction
Primary brain tumors arise from various cell types of the brain, including glial cells, neurons, neuroglial precursor cells, pinealocytes, pericytes of the vessels, cells of the hypophysis, lymphocytes and the meninges [1, 2]. The incidence of primary brain tumors varies between subtypes, with the most common primary brain tumors in adults being gliomas and meningiomas. Gliomas can be histologically classified into astrocytomas, oligodendrogliomas, mixed oligoastrocytomas, ependymal tumors and tumors of the choroid plexus. Tumor malignancy or grade is generally assessed according to the World Health Organization (WHO) criteria, taking into account the presence of nuclear changes, mitotic activity, endothelial proliferation and necrosis [1, 3]. The most fatal and common primary brain neoplasm is the glioblastoma multiforme (GBM), which corresponds to WHO grade IV. Despite aggressive multimodal treatment strategy (surgery, radiation and chemotherapy), median survival of patients with GBM is limited to less than 14 months. A complex series of molecular events occur during tumor growth resulting in dysregulation of the cell cycle, alterations in apoptosis and cell differentiation, neo-vascularization as well as tumor cell migration and invasion into the normal brain parenchyma. Genetic alterations also play an important role in the development of glioma, including a loss, mutation or hypermethylation of the tumor suppressor gene, such as p53 or other genes involved in the regulation of the cell cycle. During progression from low-grade to high-grade, step-wise accumulation of genetic alterations occurs. Growth of certain tumors seems to be related to the presence of viruses and familial diseases that accelerate the progression of molecular alterations, or exposure to environmental chemicals, pesticides, herbicides and fertilizers [4-6]. A better understanding of tumorgenesis is crucial for the development of specific molecular therapies that specifically target the tumor and reduce patient morbidity and mortality. Positron emission tomography (PET), computed tomography (CT) and magnetic resonance imaging (MRI) are generally used for non-invasive diagnosis and Department of Radiology, University of Pennsylvania, Philadelphia, Pennsylvania, United States
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
3
4
S. Chawla et al.
understanding of tumor growth mechanism. Cranial CT and MRI, with and without contrast media, are widely used for primary diagnosis of brain tumors. CT is used for detection of calcifications in oligodendrogliomas, meningiomas or craniopharyngiomas, and for tumors that are located at the base of the skull. However, the discrimination of tumor boundaries from normal tissue or vasogenic edema, as well as the evaluation of tissue heterogeneity and tumor grading are often a challenge and are not adequately reflected on CT. Furthermore, the use of ionizing radiation and image acquisition only in the axial plane, limits its applicability. PET uses various radioactive agents to detect differences in metabolic and chemical activity in the body. PET measures a wide range of physiologic processes critical in understanding the pathophysiology of brain tumors with high sensitivity. It allows for detection of metabolic changes that occur prior to structural changes visible on CT and conventional MR images. However, the major limitation of PET is its relatively poor spatial resolution and a high incidence of false positives. Continuous developments in MRI provide new insights into the diagnosis, classification and understanding of the biology of brain tumors. MRI offers several advantages compared to CT and PET. MRI offers excellent spatial resolution (1 × 1× 1 mm3 in humans), very high gray-white matter contrast and acquisition of multiplanar images. MRI is particularly accurate in establishing the intra- or extra-axial origin of tumors. The use of three-dimensional (3-D) image acquisition and reconstruction with MRI is not only limited to diagnosis, but is also useful for pre-surgical planning, stereotactic procedures and radiotherapy. Despite optimization of sequences and protocols, the classification and grading of gliomas with conventional MRI is sometimes unreliable, with the sensitivity for glioma grading ranging from 55.1 percent to 83.3 percent [7]. Integration of diagnostic information from advanced MRI techniques like proton magnetic resonance spectroscopy (1H MRS), diffusion and perfusion-weighted imaging and functional MRI (fMRI) can further improve the classification accuracy of conventional anatomical MRI [8]. Advanced MRI techniques are also being used to gain additional information on metabolic and molecular tumor markers [9, 10]. In selected patients, MRI and PET are being used in conjunction to define the real extent of the tumor [11].
2
Magnetic Resonance Imaging
2.1
Diagnosis and Grading of Brain Tumors
2.1.1
Conventional MRI
General Features of Brain Tumors Due to the excellent soft tissue contrast and high spatial resolution, MRI provides exquisite anatomical details that aid in diagnosis, classification and understanding the biology of brain tumors. A routine MRI examination of patients with brain
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
5
tumors includes long TR/long TE (T2-weighted), short TR/short TE (T1-weighted), fluid-attenuated inversion recovery (FLAIR) and post-contrast T1 sequences. Detection of a tumor is based primarily on the presence of mass effect and signal alteration on these imaging sequences. The three main variables that differentiate tumors from normal tissue are: water content, regressive events and vascular architecture. Most brain tumors exhibit increased water content and, thus, appear hyperintense on T2-weighted and FLAIR images, and hypointense on T1-weighted images (Fig. 1.1 a,b, c and Fig. 1.2a,b, c). This hyperintensity is more pronounced in masses having a low nucleus/cytoplasm ratio (e.g., astrocytoma), than in masses with a high nucleus/cytoplasm ratio (e.g., medulloblastoma). The peritumoral hyperintensity on T2-weighted images is generally nonspecific and is thought to be due to tumor infiltration, vasogenic edema, or both.
Fig. 1.1 High-grade glioma. Axial T2-weighted image (a) demonstrates an ill-defined, hyperintense (compared to gray matter), heterogeneous mass in the left parietal lobe along with vasogenic edema along the white matter tracts. Note the presence of necrotic foci (arrow) within the tumor. This mass appears as iso to hypointense on T1-weighted image (b) and hyperintense on FLAIR image (c). There is a heterogeneous contrast enhancement within the mass on the corresponding post contrast T1-weighted image (d)
6
S. Chawla et al.
Fig. 1.2 Low-grade glioma. Axial T2-weighted image (a) demonstrates a homogenously hyperintense mass in the insular region extending into the right frontal lobe. This mass is well circumscribed with minimal mass effect and edema that appears hypointense on T1-weighted image (b) and hyperintense on FLAIR image (c). There is no evidence of abnormal contrast enhancement on the post contrast T1-weighted image (d)
Regressive events such as cyst formation, necrosis and hemorrhage, calcifications and fatty degenerative areas modulate the MRI appearance of brain tumors. Intratumoral cysts are secondary to focal mucoid degeneration and fluid transudation from cyst walls. Cysts can be filled with water, or contain considerable amounts of protein or other debris from prior hemorrhage. If the cyst contains water only, it has the same signal intensity as cerebrospinal fluid (CSF) on T2- and T1-weighted images. When the protein content increases, protons become bound in a hydration layer adjacent to the protein, significantly decreasing the T1 relaxation time of water, leading to an increase in the signal intensity on FLAIR and T1-weighted images. Necrotic areas result from ischaemic cell damage or intratumoral hemorrhagic events that result in the formation of pseudocystic areas. These areas typically appear hyperintense on T2 and hypointense on T1-weighted images, compared to normal brain parenchyma. Certain primary intracranial neoplasms and metastatic tumors demonstrate hemorrhage and calcification [12]. Both chronic hemorrhage and calcifications appear hypointense on T2 and T2-weighted images, due to the induction of paramagnetic susceptibilities [13]. Recently, corrected gradient echo phase imaging has been used to differentiate hemorrhage and calcification [14, 15]. An abnormal vascular architecture is a feature that is generally observed in tumors. Stimulation of the formation of new capillaries (neo-vasculature) within the tumor tissue is facilitated
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
7
by hypoxia and endothelial growth factor receptors (EGFR). In malignant gliomas, formation of capillaries with fenestrated endothelia is stimulated, which leads to disruption of the blood-brain barrier (BBB) and contrast enhancement [16], as shown on Fig. 1-1d. On the other hand, in some tumors with a functioning BBB, these capillaries exhibit near-normal features, hence, these tumors do not enhance on contrast-enhanced T1-weighted images [16] as shown on Fig. 1.2d. Metastatic tumors are characterized by the presence of typically leaky, non-central nervous system capillaries similar to their tissue of origin and, hence, exhibit intense enhancement. Extra-axial tumors, like meningiomas, arise from tissue whose capillaries lack tight junctions and, consequently, these tumors also exhibit contrast enhancement [16]. While the extent of a tumor in the brain can be evaluated by contrast enhancement, it is known that invasive tumor cells are also present beyond the enhancing portion of the tumor, particularly in gliomas. Since contrast enhancement on conventional MRI indicates disruption of BBB and not underlying regional vascularity, it cannot be used to predict histological grade [17]. However, Fayed, et al. [18] have reported a significant difference in the contrast-to-noise ratio (CNR) of gadolinium-enhancement between low- and high-grade gliomas. Using a CNR threshold of 35.86, these authors reported a sensitivity of 82.6 percent and a specificity of 91.7 percent for the prediction of malignancy. Besides primary information on the size and location of the tumor, conventional MRI (T1, T2 and post-contrast T1 images) provides additional information about secondary phenomena such as mass effect, edema, hemorrhage, necrosis and signs of increased intracranial pressure. General Features that Differentiate Intra-axial from Extra-axial Tumors Differentiation between intra-axial and extra-axial masses is crucial as clinical management of these tumors is different [19]. This distinction has been made easier by multiplanar capabilities of MRI. Key features that help in identifying an intra-axial mass include gyral expansion, thinning or effacement of the adjacent extra-axial subarachnoid space and peripheral displacement of blood vessels along the pial surface of the brain (best seen on contrast-enhanced images) [19]. Imaging features more characteristic of extra-axial intradural masses include local bony changes such as hyperostosis, or widening of pre-existing foramina or canals; displacement of brain surface vessels away from bone and dura; white matter buckling, and widening of the subarachnoid space adjacent to the mass; central displacement of both the gray-white junction and presence of blood vessels along the pial surface. Extradural masses show similar behavior, but they usually displace the dural sheet centrally [19]. Common Brain Tumors Occurring in Adults Intra-axial Tumors The most common tumors of intra-axial location are gliomas and metastases. Gliomas derived from brain cells can, thus, be classified as true brain tumors,
8
S. Chawla et al.
whereas metastases are deposits from extradural malignancies. Gliomas tend to be poorly demarcated from the normal brain parenchyma, whereas metastases are generally demarcated sharply. In cerebral hemispheric masses, imaging in the axial or coronal plane is usually best, whereas for lesions at or near the midline, sagittal imaging is often most informative. Typically, low-grade astrocytomas demonstrate increased signal intensity on T2-weighted images and are well circumscribed with no evidence of hemorrhage or necrosis [20]. On the other hand, GBMs – the most aggressive of the gliomas – are characterized by the presence of necrosis within the tumor, coupled with extensive peritumor fingerlike edema along the white matter tracts. Oligodendrogliomas may preferentially involve the cortical gray matter and foci of cystic degeneration are relatively common [12]. Many imaging characteristics of oligodendroglioma are similar to those of astrocytomas. Gyriform or ribbon-like pattern of calcification is frequently seen in oligodendrogliomas. It has been reported that FLAIR images tend to be superior to T2-weighted images in rendering greater tumor conspicuity and in depicting margins of oligodendrogliomas [21]. Contrast enhancement is helpful for differential diagnosis, as well as tumor grading. Low-grade fibrillary astrocytomas (WHO grade II) typically do not enhance, whereas anaplastic astrocytomas (grade III) either do not enhance or enhance focally. Pilocytic astrocytomas, though grade I tumors, often enhance markedly, at least in part. High-grade gliomas like GBMs nearly always enhance, however, the degree of enhancement depends on the relative proportions of viable tumor and tumor necrosis [22]. Intracranial metastases account for up to 40 percent of all adult brain neoplasms [23]. Metastases typically involve the cerebral or cerebellar hemispheres at the corticomedullary junction. These tumors usually appear as relatively well-defined masses in peripheral locations that demonstrate moderate edema and contrast enhancement (Fig. 1.3), however the enhancement of these tumors may be variable, such as solid, ring- like, irregular, homogenous or heterogeneous. Since ring enhancement is frequently observed in brain metastases, a single lesion may be mistaken for a GBM [12]. Contrast-enhanced T1-weighted images are most sensitive in detecting brain metastases, particularly lesions in the posterior fossa or multiple punctate metastases [24]. Some studies have reported that a triple dose of gadolinium is significantly better than a single dose for demonstration of metastases [25, 26]. Metastatic lesions of 1 cm or greater in diameter can be easily detected with a standard dose of contrast while a triple dose of gadolinium is necessary to detect lesions that are smaller than 5 mm in diameter [25]. Recently, it has been reported that the inclusion of FLAIR images along with pre- and post-contrast T1-weighted images aids in differentiating glioma from metastasis in patients with a solitary enhancing lesion, as gliomas frequently exhibit FLAIR signal abnormality in the non-enhancing adjacent cortex of the tumor [27]. Extra-axial Tumors The most common extra-axial tumors are meningiomas and schwannomas. Meningiomas develop from meningothelial cells while schwannomas arise from
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
9
Fig. 1.3 Solitary cerebral metastatic tumor from lung carcinoma. Axial T2-weighted image (a) shows a hyperintense mass along with surrounding fingerlike vasogenic edema involving the posterior right frontal lobe that appears hypointense on T1-weighted image (b) and hyperintense on FLAIR (c). The mass demonstrates partial rim enhancement along with centrally placed enhancing nodules on post-contrast T1-weighted image (d)
the nerve sheath, most commonly that of the vestibular portion of the eighth cranial nerve. Meningiomas are typically globular, sometimes lobulated masses attached to the dura that indent the brain and show comparatively little or no edema. Typically meningiomas appear iso- to hyperintense relative to gray matter on T2-weighted images, and iso- to slightly hypointense on T1-weighted images. They exhibit sharp margins, and the enhancement is often intense and homogeneous. At the point of attachment, meningiomas may induce hyperostosis or dural thickening. Abnormal dural enhancement may extend beyond the site of attachment [28, 29]. Schwannomas most commonly occur in the cerebello-pontine (CP) angle cistern, but may also arise from the oculomotor, facial and trigeminal nerves. Schwannomas usually differ from meningiomas in that they typically have a higher signal on T2-weighted images [30].
10
S. Chawla et al.
Less Common Brain Tumors Intra-axial Tumors Lymphomas, hemangioblastomas and gliomatosis cerebri are less common intra-axial tumors. Lymphomas most often involve the cerebral hemispheres; typically the white matter and basal ganglia, and approximately half are multifocal. Tumor extension to the ependymal or subarachnoid surfaces is common. The dense packing of tumor cells results in relative hypointensity on T2-weighted images, which is accentuated by the often-pronounced edema. Contrast enhancement is intense and generally solid; ring enhancement, secondary to central necrosis, is more common in immunocompromised than in immunocompetent patients [31]. Hemangioblastomas are highly vascular tumors of the cerebellum that often contain cysts. The MR signal of larger cysts usually differs from that of CSF due to a higher protein content of the cyst fluid. The solid tumor nodule varies in signal intensity, but typically shows strong, intense contrast enhancement. Gliomatosis cerebri is defined as a diffuse neoplastic glial cell infiltration of the brain involving several cerebral lobes. The abnormal cells grow along fiber tracts without destroying the brain parenchyma. Histologically, gliomatosis cerebri varies from grade II to IV, but contrast enhancement is rare. T2-weighted images show widespread hyperintensity of enlarged white matter structures, which cross the borders of lobes, the midline via the corpus callosum, or the level of the tentorium cerebelli [32]. Extra-axial Tumors Epidermoid cysts are lined by squamous cells and contain cholesterol and keratinized epithelial debris. Common locations of these benign lesions, which slowly expand and distort the CSF spaces, are the midline, the lateral ventricles, and the CP angle (Fig. 1.4). One may have difficulty distinguishing them from widened cisterns or
Fig. 1.4 Epidermoid cyst. Axial T2-weighted image (a) demonstrating a homogenously hyperintense mass in the left frontal lobe (a) that appears hyperintense on corresponding FLAIR image (b). The mass does not show any enhancement on the post contrast T1-weighted image (c)
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
11
Fig. 1.5 Ruptured Dermoid cyst. Sagittal T1-weighted image (a) and corresponding axial T1weighted image (b) demonstrating a hyperintense mass in the right cerebellopontine angle (arrows). Floating fat deposits within the lateral ventricles are visible on FLAIR image (c)
arachnoid cysts on standard spin echo sequences. The contrast between lesion and CSF may be improved by using a constructive interference in the steady state [33]. Dermoid cysts are generally midline lesions that contain oily material, which has a conspicuously bright signal on T1-weighted images. Since they are prone to rupturing into the CSF spaces, fat deposits floating on top of CSF may be seen in ventricles, cisterns, or sulci (Fig. 1.5). The keratinized components have a low signal on all imaging sequences. Colloid cysts are roundish masses that arise near the foramen of Monro and project into the third ventricle. If large enough, they cause obstructive hydrocephalus, which in pedunculated lesions can develop acutely. The MR appearance varies, depending on cyst content (hemosiderin, calcium, metal ions, lipids, cholesterol) [34].
2.1.2
Perfusion-Weighted Imaging
In brain tumors, perfusion-weighted imaging (PWI) measures the degree of tumor angiogenesis and capillary permeability, both of which are important biological markers of malignancy, grading and prognosis. Brain tumor vasculature plays a critical role in supplying nutrients and oxygen to tumor cells, and also provides a road map for tumor infiltration [35, 36]. Most widely used PWI methods include dynamic susceptibility contrast (DSC), dynamic contrast-enhanced (DCE) and arterial spin labeling (ASL).
Basic Principle of Perfusion Techniques The most robust and widely used quantitative variable derived from DSC imaging is the relative cerebral blood volume (rCBV). It has been shown that, in the absence of re-circulation and contrast material leakage, rCBV is proportional to the area under the contrast agent concentration time curve. In general, the assumptions of
12
S. Chawla et al.
negligible re-circulation and contrast material leakage are violated in the presence of a tumor. The effects of this assumption can be reduced by fitting a gamma variate function to the measured signal intensity time curve [37]. Both spin echo (SE) and gradient echo (GRE) sequences can be used for measuring rCBV. SE techniques have been shown to be sensitive to small vessels, whereas GRE images incorporate signals from large vessels, including normal veins as well as tumor microvascularity [38]. A strong correlation between tumor grade and blood volume has been observed with GRE technique [39]. DCE-PWI consists of rapid and repeated T1-weighted images after a bolus injection of contrast agent (e.g., Gd-DTPA). The signal intensity on these sequential images is converted into contrast concentration using calculated or assumed T1 relaxation times. Using mathematical modeling and the signal intensity from the artery (arterial input function) the size of the extravascular extracellular space (EES) and endothelial transfer coefficient, Ktrans, can be measured. The most widely used method is based on the pharmacokinetic model of Tofts and Kermode [40]. The estimated parameters are the permeability surface area product of the endothelium (PS), the fractional size of EES and the time course of blood plasma Gd-DTPA concentration or the arterial input function. While Ktrans is widely used in DCE studies, it is affected by several hemodynamic factors such as blood flow, blood volume, endothelial permeability and endothelial permeability surface area. In contrast to DCE imaging, endogenously labeled arterial water is used as a contrast agent in arterial spin labeling (ASL) techniques to measure tissue perfusion. The labeling of water protons is performed by application of a powerful magnetic field gradient to inflowing blood to invert its magnetization. The labeling can be implemented either with pseudo-continuous saturation or by flow-driven adiabatic inversion pulses [41, 42]. In an ASL experiment a pair of images are acquired, one in which blood and tissue water magnetization are different (spin labeled image), and another in which the two magnetic states are similar (control image). The perfusion parameters are then estimated from the subtracted image. Since water is used as an endogenous contrast (which being a small molecule is freely diffusible), the ASL methods are not confounded with the permeability issues faced when using the DSC or DCE methods [43] and, as such, provide a quantitative measure of perfusion. However, these methods suffer from poor signal to noise and relatively higher specific absorption rate.
Grading of Brain Tumors Astrocytomas and rCBV Gliomas are the most common type among primary tumors of the brain, with astrocytomas being the most common subtype. In astrocytomas vascular morphology is a critical parameter in determining malignant potential and survival. Several studies have shown a strong correlation between rCBV and astrocytoma grading [17, 44]. It is generally observed that as the grade of fibrillary astrocytoma increases, the
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
13
maximal rCBV tends to increase. However, it is still not clear as to what is the histological correlation of this increased rCBV. High-grade astrocytomas are characterized by a high level of histological variability, resulting in heterogeneous rCBV maps. On the other hand, low-grade astrocytomas tend to exhibit homogenous rCBV maps [45]. Non-astrocytic Gliomas and rCBV Some low-grade non-astrocytic gliomas occasionally present with high rCBV. Oligodendrogliomas are well-known for their delicate neo-angiogenic vessels, which have been classically described as having a dense network of branching capillaries resembling a “chicken wire” pattern [46]. As such, low-grade oligodendrogliomas may demonstrate elevated rCBV that can be as high as that of GBM [47]. Choroid plexus tumors are gliomas that arise from the choroid plexus within the ventricular system of the brain. These are highly vascular tumors composed of capillaries derived from the choroid plexus, which does not contain a BBB. This results in very leaky capillaries, which causes avid enhancement on post-contrast T1-weighted images. On DSC- PWI, choroid plexus tumors demonstrate marked leakage of Gd-DTPA at the start of the bolus, and the signal intensity curve does not return to baseline levels during image acquisition. Thus, rCBV measurements of choroid plexus tumors tend to be markedly under or over-estimated [48]. A recent DCE study has also shown a strong correlation of rCBV with glioma grade [49]. Brain Tumors Other Than Gliomas and rCBV Most meningiomas are biologically benign, WHO grade I tumors. The less common atypical meningiomas (WHO grade II) tend to be more clinically aggressive and are likely to recur even after complete resection. Regardless of grade, meningiomas are highly vascular tumors that usually derive their blood supply from dural vessels of the external carotid artery, though pial supply is not uncommon. Results from a recent preliminary study suggest that the type of vascular supply, dural or pial may affect the characteristics of the susceptibility-weighted signal intensity time curve. It was shown that a profound contrast leakage occurred during the bolus phase for meningiomas that derive their blood supply from dural vessels, compared with those tumors supplied by pial vessels. Although the validity of this observation has not been established, the concept of detecting the type of vascular supply to meningiomas by using DSC PWI has profound implications in selection of patients for preoperative embolization, which is limited to dural vessels compared to surgical planning for pial-supplied meninigiomas, which tend to bleed more during surgery [47]. Primary cerebral lymphoma (PCL) is a highly malignant brain tumor, usually of B-cell lymphocyte origin. PCLs can mimic malignant primary gliomas, metastatic brain tumors or even infection on anatomic MRI [50]. On DSC PWI, lymphomas tend to show elevated rCBV, but not to the same degree as GBM. This is probably due to the fact that florid angiogenesis is not a typical feature of PCL. Rather, PCL is well-known for its angiocentric histologic feature, in which the lymphoma cells
14
S. Chawla et al.
tend to center around preexistent brain vessels [51]. DSC PWI alone, however, may not provide accurate diagnostic differentiation between PCL and other types of malignant brain tumors. Metastatic brain tumors usually do not pose a diagnostic dilemma on MRI because they tend to be multiple, are located near the gray matter-white matter junction or the subarachnoid space, and often a known history of systemic malignancy is present. However, approximately 30 percent or more of all metastatic brain tumors can manifest as a single mass in the brain [52]. Common to all metastatic brain tumors is that their tumor capillaries do not resemble those of the brain, but of the organ where the systemic cancer arose [53]. Metastatic brain tumors contain capillaries that are highly leaky because the capillaries outside the brain do not possess the unique barrier function of the brain capillaries. This is reflected in the susceptibility-weighted signal intensity time curve where profound leakage of Gd-DTPA is noted in the early bolus phase. DSC PWI reportedly can be useful in differentiating primary high-grade gliomas from solitary metastatic brain tumors, where rCBV values within the vasogenic edema of metastases were significantly lower than those within the infiltrative edema of gliomas [54]. Glioma and Ktrans Endothelial permeability of vessels in brain tumors provides valuable information about BBB integrity, vascular morphology and the nature of neo-vascularization, as well as tumor pathophysiology and prognosis [55, 56]. Several recent studies have shown that quantitative estimates of microvascular permeability correlate with brain tumor grade [56-58]. Ktrans from DCE imaging have been assessed in many clinical settings and have been shown to be useful in determining the glioma grade [58] and treatment response [59]. Ktrans is a quantitative measure of the degree of increase in T1 due to accumulation of Gd-DTPA in tissue. Because higher-grade gliomas tend to demonstrate T1 enhancement after administration of Gd-DTPA, Ktrans correlates strongly with glioma grade and histologic proliferative marker, MIB-1 index [49, 60, 61]. With increasing glioma grade, there is a higher likelihood of T1-weighted contrast enhancement of the tumor and, hence, increasing Ktrans. Although Ktrans derived from DCE MRI is commonly used at several institutions, an alternative method of perfusion imaging using the dynamic susceptibility contrast has also been proposed [62]. In brain tumors the blood flow to the tumor tissue is often hampered by an abnormal vasculature comprised of immature or defective endothelium, tortuosity and thrombosis. Hence, the uptake of Gd-DTPA by the tumor is mainly limited by blood flow and not by permeability. However, in inflammatory lesions such as a multiple sclerosis plaque, the limiting factor for uptake of Gd-DTPA is permeability and not blood flow. In addition to the complexity of pharmacokinetic modeling, the inherent heterogeneity of a brain tumor and its vasculature poses a significant challenge in permeability measurements and accurate interpretation of the data.
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
15
Glioma and Absolute CBF Absolute or quantitative CBF values using ASL techniques have been used to characterize tumor grade [63, 64]. Wolf, et al. [63] reported accurate differentiation of high from low-grade gliomas in a study of 26 patients when excluding oligodendrogliomas. A recent study by Warmuth, et al. [64] compared the DSC method with ASL in primary and secondary CNS neoplasms, and reported that both are suitable for tumor grading. CBF measured from ASL provided similar information as rCBV measures generated with DSC-based perfusion MR methods (Fig. 1.6).
Differentiation of Gliomas from Metastases Intracranial metastases and primary high-grade gliomas are two common brain tumors encountered in adults. The management of these two tumors is different and can potentially affect clinical outcome. Several reports have demonstrated significantly higher rCBV in peritumoral region of high-grade gliomas than that of metastases [54, 65]. This is possibly due to the fact that, in primary high-grade gliomas, peritumoral areas contain altered capillary morphology. These tumors also exhibit scattered tumor cells infiltrating along newly formed or preexisting, but dilated, vascular channels. In metastases, on the other hand, the peritumoral region contains no infiltrating tumor cells [66].
Fig. 1.6 Parametric maps from a high-grade glioma. Relative cerebral blood volume map (a) and cerebral blood flow map (b) demonstrate foci of high blood volume and blood flow respectively (arrows) within the mass in the right parieto-occipital lobe consistent with high-grade glioma
16
2.1.3
S. Chawla et al.
Diffusion-weighted Imaging
Diffusion-weighted imaging (DWI) is an imaging technique in which random microscopic water motion is responsible for the contrast generated within the image. The diffusion of water molecules in the brain is characterized by its apparent diffusion coefficient (ADC) and the extent of directionality by fractional anisotropy (FA). DWI has been established as a reliable non-invasive method for the early detection of cerebral ischaemic stroke [67]. In brain tumors, ADC values have been used to distinguish normal brain tissue from necrosis, cysts, edema and solid tumor. These differences are thought to result from changes in the balance between intracellular and extracellular water, and due to changes in the structure of the two compartments [68].
Basic Principles of Diffusion Technique The DWI image is based on the principle that water molecules in any living tissue routinely undergo random motion. These images are typically obtained by measuring the loss of signal after a pulse that consists of a pair of diffusion gradients added on either side of the 180° pulse of a SE sequence. This sequence can be combined with different readout strategies, like echo planar imaging or spiral imaging. The degree of MR signal attenuation after application of the diffusion gradients depends upon the duration and strength of the magnetic field gradients and the diffusion coefficient of the tissue (D). The degree of signal loss can be represented by the term S/S0, in which signal after the application of diffusion gradients is represented by S, and signal before by S0. The ratio of S/S0 is proportional to the exponential of the diffusion coefficient (D) and the degree of diffusion weighting (b), and is represented as: S/S0 ∝ exp (-b*D) [69].
Grading of Gliomas Among the histological features used for glioma grading, cellularity has been the target of quantitative assessment with DWI. Translational movement of water molecules occurs in the extracellular space and any increase in swelling or cellularity causes a drop in ADC values. Hence, the higher the glioma’s grade, the lower the mean tumor ADC values [70-74] (Fig. 1.7a). Various attempts have been made to use ADC values for predicting the glioma grade, however, the results have been conflicting [75-78]. In a recent study, Fan, et al. [78] reported a significantly lower ADC value in solid portions of high-grade gliomas (0.52 ± 0.11 × 10−3 mm2/s), compared to low-grade gliomas (1.15 ± 0.16 × 10−3 mm2/s). However, Yang, et al. [75] reported that ADC values in low-grade tumors were lower than high-grade. Other studies have failed to observe any difference between the two grades of tumors [76, 77]. The limited role of ADC in glioma grading is likely due to the inherent heterogeneity associated with gliomas across different grades, within the same grade, and even within a single tumor.
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
17
Fig. 1.7 Diffusion parametric maps from a high-grade glioma. Axial ADC map (a) demonstrates restricted diffusion within the mass in the left parietal region extending into the lateral ventricular system (arrow). Corresponding FA map (b) demonstrates reduced anisotropy. Directionally encoded color map (c) showing displacement and infiltration of superior longitudinal fasciculus (arrow). Color indicates directions as follows: red, left-right; green, anteriorposterior; blue, superior-inferior. Fiber tracking image (d) shows diminished superior longitudinal fascicular fiber in the cerebral hemisphere involving the tumor compared to the contralateral side (arrow) (top view)
Another application of DWI in brain tumors is the use of fractional anisotropy (FA) maps derived from diffusion tensor imaging (DTI) in determining the integrity of white matter tracts in the vicinity of the tumor (Fig. 1.7b). Although there is a lack of direct histological correlation between FA maps and the status of white matter tracts near the tumor, a variation of FA matrices, such as the tumor infiltration index [79] or fiber attenuation index [80], may provide more specific information on the status of peritumoral edema in brain tumors. FA has been suggested as a good predictor of cell density and proliferation of GBMs [81]. However, the biological correlates for the DTI-derived matrices remain unclear and await much needed histological validation.
18
S. Chawla et al.
Differentiation of Gliomas from Metastases Krabbe, et al. [82] reported ADC values for metastases (1.2-2.73 × 10−3 mm2/s) and high-grade gliomas (0.72-2.61 × 10−3 mm2/s) which were not significantly different. Similarly, Kono, et al. [74] did not report any differences in ADC values between the two types of tumors using a larger patient population. These studies show that ADC values from solid enhancing portions of metastases are not useful in the differentiation from markedly enhancing high-grade gliomas. However, several studies have demonstrated the utility of DWI in differentiating these two entities when ADC values were measured from peritumoral regions, with some studies reporting significantly higher peritumoral ADC in metastases than that of gliomas [77, 83]. Tumor Types and ADC In a study of 48 patients with contrast-enhancing malignant tumors, significant differences in the ADCmin values were noted between lymphomas and GBM, and between lymphomas and anaplastic astrocytomas (AA). However, there were no differences between lymphomas and metastases, and between GBM, AAs and metastases [84]. On the other hand, Krabbe, et al. [82] did not observe any significant differences in the mean ADC values in patients with meningiomas, compared to highgrade gliomas or brain metastases. ADC values have also been correlated with cellularity of non-glial brain tumors. When compared with gliomas, lymphomas and medulloblastomas were shown to have lower ADC values because of densely packed cells in these tumors [85]. In a small sample of meningiomas, malignant or atypical meningiomas were found to have lower ADC values when compared with typical meningiomas [86]. 2.1.4
White Matter Tractography
Diffusion tensor imaging (DTI) is distinguished from DWI by its sensitivity to anisotropic or directionally dependent diffusion of water molecules. The anisotropic diffusion in the brain is largely attributed to the cytoarchitectural compositions of myelin and axons. DTI requires collection of diffusion data in at least six non-collinear directions. Anisotropic diffusion is characterized by a tensor, which fully describes the mobility of the water molecules in each direction and the correlation between them. The tensor can be diagonalized such that only the three non-zero elements, known as the eigen values, remain along the diagonal of the tensor. Each eigen value is associated with an eigenvector where the largest of the three eigen values corresponds to the principle eigenvector and describes the principal direction of diffusion at that point. As shown in Fig. 1.7c, by choosing the eigenvector associated with the largest eigen value, the principal diffusion direction of the brain structure to be examined can be color-coded, resulting in color-coded maps or directionally encoded FA maps [69]. In fiber tractography with DTI, white matter tract directions are mapped on the assumption that in each voxel a measure of the local fiber orientation is obtained.
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
19
Fiber tracking provides a 3-D depiction of white matter connectivity, which allows studying brain cytoarchitecture at a microscopic level. DTI-based fiber tractography necessitates definition of a seed region of interest (ROI) that is located at the path of the investigated fiber network system to initiate the fiber tracking process [87]. Selecting seed ROIs based on known anatomical landmarks has led to the identification of a large number of fiber bundles in healthy subjects [88, 89]. In patients with brain tumors, tractography enables visualization of specific fiber bundles that are either in proximity to a tumor or that are influenced by it. This information helps in intra-operative guidance for tumor resection [90]. Tractography maps of desired white matter tracts can be overlaid onto high-resolution anatomic images, and provide information on alterations in fiber tract directionality and integrity due to the presence of a brain tumor (Fig. 1.7d). Tractography of the corticospinal tract (CST) is beneficial for pre-surgical planning as it helps the surgeon to avoid injuring the CST during tumor resection. The role of DTI in brain tumors was recently evaluated in nine patients with eight gliomas and one metastatic adenocarcinoma [91]. The white matter tracts were characterized as being displaced, edematous, infiltrated or disrupted. Nine large white matter pathways in five patients were displaced; two patients with frontal oligodendrogliomas showed infiltration (confirmed pathologically). The investigators concluded that DTI is beneficial in pre-surgical planning, but it was not clear as to whether resection of anatomically intact fibers in abnormal-appearing areas of the brain would lead to postoperative deficits. In a separate study, DTI was used to determine abnormalities beyond those seen in T2-weighted scans on patients with low- and high-grade gliomas and metastases [92]. Abnormalities on DTI were larger than those seen on T2-weighted images in 10 out of 13 patients with high-grade gliomas, but not in metastases or lowgrade gliomas. Furthermore, four out of 13 of these cases showed new abnormalities in the contralateral hemisphere, suggesting the possibility of tumor spread across the corpus callosum. The implication of this study is that DTI can improve the targeting of radiation therapy to visible tumor volume, as well as encompassing ‘invisible’ tumor infiltrating the white matter pathways. Another potential application of this technique is in distinguishing between normal white matter, edematous brain and enhancing tumors, as reported in a study of nine patients with GBMs [93]. This technique may also be useful in tracking the distortion of white matter pathways by lowgrade gliomas, and anticipating the development of neurological impairment due to the presence of tumors.
2.1.5
Proton Magnetic Resonance Spectroscopy
Basic Principle of MRS Similar to the basic principles of MRI, MRS is based on the spin properties of atomic nuclei (e.g. 1H, 31P, 13C, 19F) when present in a strong magnetic field, that allows the nuclei to absorb and re-emit energy in response to a radio frequency pulse at the resonance frequency of that particular nuclei. The effective magnetic field, sensed by a particular nucleus, is affected by neighboring electrons. The separation of resonance
20
S. Chawla et al.
frequencies of different protons of a molecule due to the dissimilar chemical environment is described as the chemical shift (δ), and is expressed in parts per million (ppm). The height (maximum peak intensity) or the area under the peak yields relative measurements of the concentration of protons. The spectral information from a particular region of the brain is generally obtained by spatial localization, which is achieved by applying static and/or pulsed gradients. Localization methods commonly used in clinical 1H MRS include: pointresolved spectroscopy (PRESS), spatially resolved spectroscopy (SPARS) and the stimulated echo method (STEAM). As abundant water protons (70 M) impose limitations to observe intracellular metabolites (1-10 mM), the signal from water needs to be suppressed. The most frequently used method for suppressing the signal from water is chemical shift selective excitation (CHESS), which reduces the water signal by a factor of 1,000 [94]. Biochemical Features of a Normal Human Brain Spectra The most prominent resonances that are seen from normal human brain on in vivo 1 H MRS include N-acetyl aspartate (NAA), together with intense signals from creatine (tCr), choline-containing compounds (tCho), myo-inositol and multiple peaks from glutamate and glutamine (Glx) (Fig. 1.8). Biochemical Features of Tumor Spectra NAA (singlet at 2.02 ppm) is mainly distributed in intact neurons and neuronal processes such as axons. Studies on cultures of separated brain cells have revealed the presence of NAA in un-differentiated oligodendrocyte cells as well [95]. However, presence of NAA in mature oligodendrocytes and astrocytes is generally
Fig. 1.8 Proton magnetic resonance spectrum from a normal brain. Axial T1-weighted image (a) demonstrating voxel position at the centrum semiovale. 1H MRS spectrum (b) acquired with PRESS sequence displaying characteristic resonances [NAA (2.02 ppm); tCr (3.03 ppm); tCho (3.22 ppm); Glx (2.35 ppm) and mI (3.56 ppm)] from the voxel shown in (a)
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
21
not observed. The acetyl group of NAA has been suggested to have a role in biosynthesis of lipids, while the aspartyl group is involved in metabolism of several neurotransmitters [96]. NAA has generally been found to be either absent or reduced in brain tumors [97]. Total choline (tCho, single peak at 3.22 ppm comprising signals from choline, phosphocholine and glycerophosphocholine) is elevated in tumors, compared to normal brain tissue [97, 98]. Increased tCho is thought to be present due to accelerated membrane synthesis of rapidly dividing cancer cells. In vivo tCho levels have been shown to correlate with proliferative potential of the tumor as determined by immunohistochemical analysis of tumor biopsies using the Ki-67 labeling in gliomas [99] and meningiomas [100]. However, meningiomas, which are mostly benign and slow growing, can have tCho concentrations comparable to grade III astrocytomas [101]. Creatine-containing compounds (tCr; peak at 3.03 comprising signals from creatine and phosphocreatine) is reduced in astrocytomas, compared to normal brain tissue and is almost absent in meningiomas [98], schwannomas [102] and metastases [103]. Creatine-containing compounds are thought to have a role in maintaining energy-dependent systems in brain cells by serving as a reserve for high-energy phosphates. Creatine-containing compounds also serve as a buffer in adenosine triphosphate and adenosine diphosphate reservoirs as it is increased in hypometabolic states and decreased in hyper-metabolic states [104]. Myo-Inositol (mI; singlet at 3.56 ppm) has a short T2 and as such can only be observed with sequences using a short echo time (TE). Myo-Inositol is mostly found in astrocytes [97] and is high in low-grade gliomas [98, 105], but low or absent in non-glial tumors such as schwannomas [102] and meningiomas [98]. Higher levels of mI have been reported to distinguish hemangiopericytomas from meningiomas [106]. This metabolite is also a precursor for lipid metabolism and, hence, may be elevated due to increased cellular proliferation. Some studies suggest its role as an osmolyte and have attributed its increase in inflammatory processes [107]. The broad multiplet signals (between 2 and 2.4 ppm, and 3.76 ppm) from glutamate and glutamine (Glx) are most readily observed at short TE. Glx is prominently observed in meningiomas, possibly reflecting altered energy metabolism involving partial oxidation of glutamine leading to alanine (Ala, doublet at 1.47 ppm), which is also elevated in meningiomas [98, 101]. Glutathione (GSH; multiplets at 2.9 and 3.8 ppm) is part of the same metabolic pathway, and has been recently identified by in vivo 1H MRS and is reported to be higher in meningiomas, compared to astrocytomas [108]. Lactate (doublet at 1.33 ppm) is frequently observed in tumors, probably due to increased anaerobic glycolysis [109], and is often most prominent in high-grade tumors. However, studies have reported that lactate levels do not correlate with tumor grade [110] or metabolic rate [111]. Lactate is present in both intracellular and extracellular spaces, and its overall concentration is dependant upon metabolic rate of cells and clearance from the cell and the interstitium. A reduced clearance rate in necrotic or cystic regions leads to increased lactate levels, independent of any increase in glycolysis associated with high-grade tumors [110].
22
S. Chawla et al.
Lipids (0.9 and 1.3 ppm) are also characteristic of high-grade tumors at short TE [98, 112], but are only observed in about 41 percent of high-grade tumors [7]. Biopsy studies indicate that lipids correlate with necrosis [113], which is a histological characteristic of high-grade tumors. Increased lipid signals in tumors could be the result of membrane phospholipids released during cell breakdown, and may thus relate to the necrotic fraction. Studies in animal models suggest that triglycerides are stored as droplets in the cytosol as a result of hypoxic stress [114], and the presence of lipid resonances at 2.8 and 5.4 ppm have also been proposed as markers of apoptosis [115].
Grading of Gliomas As shown in Fig. 1.9, both single voxel 1H MRS and multivoxel proton magnetic resonance spectroscopic imaging (1H MRSI) have been used to evaluate the degree of malignancy of brain tumors [75, 116-118] (Fig. 1.9). The most common observation in grading of glioma is that the tCho/tCr and tCho/NAA ratios increase with grade; however, there is a significant overlap of these indices between grades, most likely due to the heterogeneity of tumors. Kaminogo M, et al. [119] reported that tCho/NAA was less than 1.0 in healthy tissues and greater than 1.0 in all but one of the gliomas studied. The tCho/tCr ratio tended to be higher with higher histological grades. A significant difference was observed between grades II and IV, but not between grades II and III or III and IV. Recently short TE data have shown a decrease in mI/tCr and mI/tCho with grade [98, 105]. Meyerland, et al. [9] reported that Lac/Water is useful in grading of gliomas. Some studies have also indicated that lipids play a role in grading of high-grade tumors [98, 112]. Since most tumors are heterogeneous, their spectra are likely to have contributions from multiple tissue compartments [120]. Along with viable tumor cells, there
Fig. 1.9 Proton magnetic resonance spectrum from a high-grade glioma. Axial post-contrast T1weighted image (a) showing homogenously enhancing mass in the right medial temporal lobe demonstrating the position of a voxel. Proton magnetic resonance spectrum (b) from the voxel shown in (a) exhibiting typical spectral features of a high-grade glioma. Note the reduced NAA and abnormally elevated peaks of tCho and Lip+Lac
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
23
may be necrotic and cystic regions, and in the case of highly infiltrative tumors there may also be contributions from normal brain tissue. Additionally, tumor growth is not well regulated and as such variations in cellular metabolism and cell density will occur, and as a tumor progresses it may become composed of cells of different grades [121]. Because of heterogeneous characteristics of tumors, accurate grading of gliomas has not been completely achieved by 1H MRS [7, 106, 122-124]. Differentiation of Gliomas from Metastases 1
H MRS has shown a great potential in differentiating high-grade gliomas from metastases, particularly when the lesion is solitary and conventional MRI is inconclusive. Ishimaru, et al. [103] reported elevated tCho from enhancing portions of all the high-grade gliomas and metastases studied. They also reported an absence of tCr and NAA in most of the metastases, indicating that the absence of these resonances could be used for differentiating them from gliomas. The absence of tCr in metastases may be the result of exhaustion of energy reserves due to a rapid cell proliferation, compared to gliomas. A study with a larger population – 51 patients – with a solitary brain tumor (33 gliomas, 18 metastases) has reported significantly elevated tCho/tCr in the peritumoral region of gliomas (2.28 ± 1.24), compared to metastases (0.76 ± 0.23) [54]. However, there was no significant difference in peritumoral NAA/tCr between the two groups, as there is no neuronal replacement or destruction in the peritumoral regions in either pathologic condition. It has been shown that in high-grade gliomas, tumor cells infiltrate along vascular channels but do not destroy the preexisting cytoarchitecture [17]. Vasogenic edema associated with metastases is also a passive process that does not necessarily destroy the underlying structure or neuronal tissue [125]. 2.1.6
Multiparametric Analysis
Histologically, gliomas often demonstrate considerable heterogeneity, with focal areas of malignant features dispersed over several regions. Due to this inherent heterogeneity, MRI techniques such as PWI, DWI and 1H MRS, when used independently, often lead to ambiguous results in grading of gliomas, specifically oligodendrogliomas [7, 106, 122-124]. The preexisting heterogeneity may also be the reason for overlapping rCBV and metabolite ratios between different glioma grades [126]. Limited studies have been performed to combine advanced MR techniques to more accurately characterize and predict survival of brain tumors. This approach offers detailed and complementary information on the complicated intra- and peritumoral architecture that reflects tumor vascularity, cellularity and metabolic information. Combined measurement of rCBV, ADC and 1H MRS indices has been shown to improve the sensitivity and specificity of glioma grading [116], and for the characterization of brain metastases [65] Figs. 1.10 through 1.12 demonstrate multiparametric analysis from representative cases of high-grade glioma, low-grade glioma and metastasis, respectively. Multiple MRI and 1H MRS techniques may
24
S. Chawla et al.
Fig. 1.10 Multiparametric analysis from representative case of a high-grade glioma. T1-weighted image (a) demonstrating a diffusely hypointense mass with surrounding edema in the left parietooccipital region. Co-registered FA map (b) showing regions of reduced anisotropy. Corresponding ADC map (c) and DWI image (d) demonstrating heterogeneous diffusion within the mass. Increased perfusion with a ring of high blood flow on CBF map (e) indicative of a high-grade glioma. Twodimensional CSI grid overlaid on T1-weighted image (lower panel) showing a representative voxel (black) and corresponding spectrum exhibiting various metabolites. Resonances of high Lip+Lac and tCho are consistent with findings of high-grade glioma
provide a more accurate assessment of brain tumors, providing useful information for guiding stereotactic biopsies, surgical resection and radiation treatment. Composite information from all relevant techniques is desirable as it is unlikely that all parameters will be useful in every patient or at every time-point. Multiparametric analysis also provides quantitative and correlative measurements that are closely related to the biological properties of the tumor, and reflect changes in tumor vascularity, cellularity and proliferation that are associated with tumor progression. Gupta, et al. [71] have previously reported a significant inverse correlation between tCho and ADC in gliomas. This study suggested that tumor cell density plays a major role in defining the level of tCho signal. Some studies have shown that there is a strong correlation between high vascularity (high rCBV), increased cellularity (low ADC) and increased membrane turnover (high tCho) in gliomas [65, 75].
2.1.7
Functional Magnetic Resonance Imaging
One of the principal surgical goals for the treatment of brain tumors is to minimize neurological deficits and to maximize resection of the pathological mass. To achieve this goal, functional eloquent brain areas must be identified. The standard methods used for identifying these areas are intra-operative mapping in the conscious patient, implantations of a subdural grid with extraoperative stimulation mapping
Fig. 1.11 Multiparametric analysis from representative case of a low-grade glioma. T1-weighted image (a) demonstrating a well-defined hypointense mass in the right parietal lobe with little peritumoral edema extending into the cortex. Co-registered FA map (b) showing low anisotropy. Corresponding ADC map (c) and DWI image (d) demonstrating high diffusivity indicative of lowgrade glioma, which is also confirmed by low blood flow on the CBF map (e) (arrow). Two-dimensional CSI grid overlaid on T1-weighted image (Lower panel) showing a representative voxel (black) and corresponding spectrum exhibiting various metabolites. Note lower tCho/tCr ratio and lower level of Lip+Lac compared to high-grade glioma (Fig. 1.10)
Fig. 1.12 Multiparametric analysis from a representative case of metastasis. Post-contrast T1weighted image (a) demonstrating an enhancing mass in the posterior right frontal lobe. Coregistered FA map (b) showing low anisotropy, ADC map (c) and DWI image (d) showing high diffusivity from the core of the mass and lower diffusivity from the rim. Markedly elevated blood volume is visible only from the central core of the mass on rCBV map (e) whereas in the peritumoral regions, volume is low. Lower panel depicting 2D CSI grid overlaid on T1-weighted image showing representative voxels from tumoral (1) as well as peritumoral (2) region. Corresponding spectra exhibiting various metabolites. Note elevated Lip+Lac and tCho peaks only from the tumor region suggesting a case of metastasis
26
S. Chawla et al.
or operative sensory-evoked potential recordings. However, these techniques are invasive, induce extreme stress on the conscious patient, and often require a craniotomy larger than what is necessary for tumor resection. fMRI is a non-invasive method that can be used pre-operatively and helps in establishing the relationship between the margins of the tumor and the functionally viable brain tissue. It is based upon the blood oxygenation level-dependent (BOLD) effect [127, 128]. It is believed that when the patient is asked to perform a certain language or motor task, the activation of neurons leads to an increase in oxygen consumption of these neuronal cells, which in turn induces a concomitant exaggerated increase of the local blood flow [129]. The decreased concentration of deoxygenated hemoglobin induces a higher signal on T2-weighted images, highlighting the functional areas of the brain relative to the task performed by the patient [130]. As the BOLD signal change following stimulation is relatively small, detection of a reliable change in activation is challenging. A number of post-processing steps like motion correction and smoothing of the data are required to accurately discern the functionally activated voxels. Functional maps are usually generated through a voxel-by- voxel statistical analysis of the time series MRI data. However, statistical procedures for preoperative mapping have not been standardized, as there is no consensus on the best statistical model for such studies or on the estimation of significance levels. Some authors apply a Bonferroni correction [16, 131], some put minimal thresholds on cluster sizes [132-134], while others perform permutation testing to estimate significance levels [135].
Influence of Tumors on BOLD Signal Evidence indicates that the BOLD response in the vicinity of certain tumors does not reflect the electrical neuronal activity as accurately as it does in healthy brain tissue [136-138]. Recent data indicate that cortical BOLD activation can be reduced near glial tumors, both at the edge of the tumor and in normal vascular territories, somewhat removed from the tumor. Loss of regional cerebral vasoactivity near these tumors has been suggested to be a contributing factor [139]. At the interface of tumors and normal brain, astrocytes and macrophages can continuously release nitric oxide that leads to a regionally increased cerebral blood flow and decreased oxygen extraction fraction during basal metabolism. These processes may result in a decreased BOLD signal intensity difference following activation [139]. Holodny, et al. [137] found that the number of activated voxels was 35 percent less at the tumor site, compared to the contralateral site. The authors suggested that this is possibly due to loss of autoregulation and a changed venous response, due to compression of the neighboring vasculature by the presence of a tumor. The presence of a tumor in the brain also leads to changes in regional tissue pH, glucose, lactate and adenosine triphosphate levels, although such effects on BOLD neuronal coupling have not been completely understood [140]. Glial tumors can induce abnormal vessel proliferation in adjacent normal brain tissue, altering regional CBF, rCBV, vasoactivity and, potentially, BOLD contrast. Other factors,
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
27
including vasogenic edema and hemorrhage, may also contribute to the observed decrease in the near-lesion BOLD contrast. Evidence for a substantial impact of vasogenic edema on BOLD contrast is lacking due to the small number of patients included in several of these studies. Micro-hemorrhages associated with intraparenchymal tumors could hinder the detection of change in susceptibility gradients that provide BOLD contrast. Various pharmacological agents used in the treatment of tumors may also influence the BOLD signal. There are indications that antihistamines reduce and that caffeine boosts the BOLD response [141]. It has also been reported that levodopa modifies the BOLD response [142]. Presurgical fMRI in patients with brain tumors is a promising clinical application as it allows risk assessment of therapeutic interventions, selection of patients for intra-operative mapping and guides brain surgery. Unfortunately, no randomized trials or outcome studies have been performed to evaluate the role of pre-surgical fMRI in determining the final outcome of the patient. Therefore, fMRI has not yet reached the status of routine clinical acceptance. Combining pre-surgical fMRI with other techniques such as DTI may aid in greater use of fMRI in clinical brain tumor applications.
Combined Use of fMRI and Tractography BOLD-fMRI and diffusion tensor tractography (DTT) may help in establishing the relationship between brain motor cortex, pyramidal tracts and gliomas, which might aid in optimizing surgical planning and guide microsurgery [143, 144]. Knowledge of the structural integrity and location of eloquent white matter tracts relevant to brain tumors is crucial in neurosurgical planning because damage to these clinically eloquent pathways can result in postoperative neurological deficits as devastating as damage to functional cortical areas. In many brain tumor cases partial or no loss in functional activity is observed leading to the assumption that the fibers, though deviated, are still partially functionally intact. In such cases white matter mapping using seed ROI based on known normal anatomical landmarks might be misleading, since the white matter is deviated from its normal location. This task becomes even more complicated when edema, tissue compression and degeneration are present. These changes deform the architecture of the white matter and, in some cases, prevent accurate selection of the seed ROI from which fiber tracking begins. Recently, it has been shown that selection of the seed points based on fMRI activations, which constrain the subjective seed ROI selection, enabled a more comprehensive mapping of fiber systems [145].
3 Differentiation of Recurrent Tumor from Radiation Necrosis Surgical resection of brain tumors is generally followed by chemo-radiotherapy that leads to radiation-induced necrosis. Despite aggressive and combined therapeutic regimes, brain tumors generally recur at or near the site of initial resection.
28
S. Chawla et al.
Differentiating radiation necrosis from recurrent tumor has important implications for the patient’s management since recurrent tumors may benefit from repeat surgery with adjunct chemotherapy, whereas radiation necrosis may be treated with steroids. However, the distinction of delayed radiation-induced necrosis from tumor recurrence by using conventional MRI has been difficult [146]. Both entities manifest mass effect with surrounding edema. Both processes can cause varying degree of BBB disruption that results in abnormal contrast enhancement. Early radiation necrosis is characterized by fibroid necrosis of blood vessel walls, followed by necrosis of surrounding parenchyma. Late vascular changes include wall thickening, hyalinization and telangiectasia. Extensive reactive gliosis, dystrophic calcification and cyst formation are commonly observed adjacent to the necrotic foci [147]. On the other hand, a recurrent tumor is characterized by angiogenesis [17]. Several advanced MRI techniques such as 1H MRS [148-151], PWI [17, 152] and DWI [153] have been used independently or in combination to differentiate tumor recurrence from radiation necrosis (Fig. 1.13). The typical change that occurs on 1H MRS of a tumor after radiation therapy is a reduction of tCho with a possible increase in lactate and/or lipids indicating the transformation of viable tumor cells towards necrosis [100]. Many groups have investigated whether the 1H MRS pattern is sufficiently different to distinguish between the two groups [148-151]. Although these studies concluded that 1H MRS is useful in accessing potential recurrence, there is a disagreement about a characteristic spectral pattern for either entity, since both decreases and increases in tCho have been reported post treatment. Some groups have reported that NAA and lactate can be used as reliable indicators of radiation change [150]. The difficulty lies in the fact that irradiated tumor beds usually contain a mixture of both tumor and radiation effects. In many studies, the voxel size used was too large, which led to partial volume effect from normal tissue. Using PWI, it has been reported that rCBV, a surrogate marker of angiogenesis, allows differentiation between the two lesions. Tumor recurrence is generally associated with elevation of rCBV, compared to patients with radiation necrosis [17, 152]. Moreover, DWI detects therapy-induced water diffusion changes and has been useful in differentiating between these two conditions [153]. Asao, et al. [154] have found significantly lower ADC values in recurring tumors than in tumors with radiation necrosis.
3.1
MRI in Stereotactic Biopsy
Stereotactic biopsy (SB) has evolved as a powerful and safe tool to provide tissue diagnoses with minimal disruption of normal brain function. Sampling error in biopsies of high-grade gliomas is well-known and is partly attributable to heterogeneity within a single tumor [155, 156]. The rate of conclusive histopathologic diagnosis by SB is highly variable, ranging from 60 percent to 98 percent. Furthermore, it has been recognized that the tiny amounts of tissue obtained by stereotactic biopsy may not be sufficient for a correct diagnosis, and may lead to errors that can have an impact on therapeutic management [157].
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
29
Fig. 1.13 Multiparametric analysis to differentiate tumor recurrence from radiation necrosis. Post-contrast T1-weighted images (a) showing contrast-enhanced masses in patients with radiation necrosis (upper panel) and tumor recurrence (lower panel). Co-registered FA (b), ADC maps (c) showing reduced anisotropy and elevated diffusivity within the mass in both patients. However, co-registered rCBV maps (d) demonstrating lower perfusion in patient with radiation necrosis compared to tumor recurrent patient (arrows). Proton spectra from the voxels shown on contrast enhanced T1-weighted images (a) showing higher tCho and lower Lip+Lac in patient with tumor recurrence compared to radiation necrosis
A few reports have shown that MRI-guided stereotactic biopsy specimen accurately represents the grade of glioma. The grading of gliomas is based on histologic evaluation of specimens from the most malignant region of the tumor. Most biopsies are guided by contrast-enhanced T1- weighted images [158], which depict areas of BBB breakdown and may not indicate the most malignant or vascular portion of the tumor. On the other hand, rCBV maps depict foci of greatest vascularity, which corresponds to the regions of maximum malignancy and can aid in directing precise stereotactic biopsy, particularly in non-enhancing tumors [159].
30
S. Chawla et al.
1 H MRSI has also been reported to be useful for stereotactic procedures and has demonstrated a potential to overcome the limitations of a single voxel study by encompassing the tumor as well as the normal brain. Pathological specimen taken from areas of increased tCho/tCr ratios and decreased NAA/tCr ratios can facilitate diagnosis by demonstrating increased cellularity and mitoses [160].
3.2
MRI in Experimental Brain Tumor Models
To understand brain tumor biology and to develop new treatment strategies, several experimental animal models (especially rodents) of glial tumors have been developed over the past several years [161, 162]. These experimental brain tumor models are aimed towards development of new radiotracers for cellular proliferation and protein synthesis, characterization of tracers and detection of early responses to therapeutic interventions [115]. Strategies for imaging transcriptional regulation and migration of tumor cells, and imaging expression of exogenous genes that carry a marker or therapeutic function, for the purpose of developing improved gene therapeutic vectors, have become possible with the help of MRI of animal models [163, 164]. In recent years stem cell therapy has proven to be a promising means to improve neurological function in brain tumor pathologies and has excellent potential to be clinically effective on a large scale [165]. With cell-specific MRI, the distribution and survival of magnetically labeled stem cells has been monitored [166]. In vivo detection of tumor cell migration, establishment of in vivo assays for tumor-specific signal transduction pathways, assessment of tumor-specific antigens and of labeled bone marrow-derived endothelial precursor cells has been greatly facilitated by combining MRI methods with animal models [167-171]. The combination of MRI techniques and specific animal tumor models has provided opportunities to characterize the tumor microenvironment, and physiology, and to understand their impact on tumor growth [172, 173]. Using rodent models, it has been shown that novel imaging techniques like T1 rho-weighted imaging has better demarcation potential for tumor borders than proton-density or T2-weighted imaging, which could be useful in treatment planning when combined with other imaging sequences [174]. Imaging of animal models has great implications especially when molecular, diagnostic and treatment modalities have to be translated from the bench to bedside.
4
Positron Emission Tomography
PET is an imaging technique that provides concentrations of trace amounts of compounds labeled with positron-emitting isotopes introduced into the body either by inhalation or intravenous administration [175]. PET techniques have a high sensitivity, such that very low levels of specific tracer accumulation can be detected,
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
31
but have an inherently limited spatial resolution [176]. In neuro-oncology the role of PET has been primarily limited to revealing highly specific quantitative information on the metabolic state of gliomas (Fig. 1.14). PET allows the quantitative localization of expression of endogenous or exogenous genes coding for enzymes
Fig. 1.14 Differentiation between low and high-grade glioma using PET. In low-grade glioma (grade II), glucose metabolism is similar to white matter (arrows) and amino acid uptake is only moderately increased. In high-grade gliomas, both glucose metabolism and amino acid uptake are increased (Printed with permission from Jacobs AH et al., NeuroRX. 2005; 2:333–347.)
32
S. Chawla et al.
or receptors by measuring the accumulation or binding of the respective enzyme substrates or receptor binding compounds [177]. Depending on the radiotracer, various molecular processes can be visualized by PET, most of them relating to an increased cell proliferation within gliomas. 2-[18F] fluoro-2-deoxy-D-glucose (FDG) is the most commonly used tracer for PET oncological studies. FDG’s relatively simple synthesis and long half-life, along with extensive knowledge of the mechanisms determining its uptake and retention, have made it quite popular in neuro-oncology [178]. It is well established that brain neoplasms present changes in glucose utilization, compared to normal brain tissue. FDG detects tumor glycolysis and has been used to detect the metabolic differences between normal brain, low-grade and high-grade gliomas [179]. Glucose and FDG share the same saturable carriers between blood and tissue, and FDG competes with glucose for hexokinase. FDG-6-P is trapped in cells in proportion to the glucose metabolic rate, and PET can detect its accumulation. With this tracer, changes in the oxidative metabolism were first demonstrated in vivo in brain neoplasms [180]. Activation of the gene coding for the synthesis of glucose transporter GLUT1 is a major early marker of malignant transformation. An over-expression of GLUT1 and GLUT3 has been observed in brain tumors [181], and this may explain the increased level of glucose extraction demonstrated with PET [182]. The FDG uptake into malignant cells is a consequence of increased expression of glucose transporter and glycolysis. PET-FDG has also been used to differentiate recurrent brain tumors from necrosis after radiation and/or chemotherapy [183]. The areas of necrosis indicate significantly reduced metabolism, while recurrent tumors are identified as having increased metabolism. Kim, et al. [184] evaluated PET imaging in 33 patients with brain tumors after radiation therapy, and found a sensitivity of 80 percent and a specificity of 94 percent for detection of tumor recurrence by FDG-PET. Oxygen-15 is a short-lived positron-emitting isotope that can be used to measure hemodynamic parameters. Using mathematical modeling, functional images of cerebral blood flow (CBF), oxygen extraction (OER), cerebral oxygen metabolic rate (CMRO2) and blood volume (CBV) can be derived from the combination of sequential studies with 15O2, C15O2 and C15O. Although blood flow in tumors is variable, oxygen metabolism is generally reduced in gliomas, in line with the relatively anaerobic metabolism of tumors. The low OER implies that the tumor is not ischaemic and that perfusion is sufficient to meet the metabolic need for oxygen in tumors before initiation of therapy [185]. Imaging with radiolabeled amino acids visualizes protein synthesis and amino acid transport phenomena, which are accelerated in tumors [186]. The radiolabeled amino acids methyl- [11C]-L-methionine ([11C] MET), [11C]-tyrosine, [18F] fluorotyrosine and O- (2-[18F]-fluoroethyl)-L-tyrosine have been reported to be more specific for brain tumor detection, compared to FDG [187]. [11C] MET-uptake correlates with cell proliferation In Vitro, the expression of Ki-67 and proliferating cell nuclear antigen, as well as to microvessel density, making it a potential biomarker for active tumor proliferation [188]. The intensity of [11C] MET uptake differentiates between grade II and grade III/ IV gliomas [189]. Increased [11C] MET uptake also depends on tumor type, with oligodendrogliomas accumulating more radiotracer
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
33
than astrocytomas from the same histological grade [189]. [11C] MET has also been used for guiding stereotactic biopsy of brain tumors [190]. Studies for assessing gene therapy in recurrent gliomas have also been performed. Transduction of the herpes simplex virus type-1 thymidine kinase, followed by subsequent activation of the prodrug ganciclovir, may be beneficial as adjuvant therapy [191]. In future, specific markers of tumor cell proliferation and gene expression may allow the application of PET, not only for diagnostic imaging but also for non-invasive biological characterization of malignant tumors and early monitoring of therapeutic interventions.
5 ●
●
●
●
●
Key Points Neuroimaging of brain tumors has evolved from a purely anatomy-based discipline to one that incorporates morphologic abnormality with physiologic alterations in cellular metabolism and hemodynamics. MRI and PET have become an essential part of the diagnostic protocol to diagnose, guide surgery, monitor therapy response and predict prognosis of patients with brain tumors. The incorporation of advanced MRI – such as DWI, 1H MRS and PWI – as part of the clinical imaging protocol has empowered neuro-radiologists to begin the process of combining radiology with biology to provide meaningful and clinically relevant end points and biomarkers for clinical trials and assessment of malignancy. BOLD-fMRI and diffusion tensor tractography can non-invasively localize the relationship between motor cortex, pyramidal tracts and gliomas to optimize surgical planning with preservation of eloquent areas and subcortical white matters tracts. Molecular imaging is a rapidly growing area that should enable evaluation of physiological, biochemical and genetic processes that occur in brain tumors.
References 1. Kleihues P, Soylemezoglu F, Schauble B, Scheithauer BW, Burger PC. Histopathology, classification, and grading of gliomas. Glia 1995; 15:211-221. 2. Daumas-Duport C, Beuvon F, Varlet P, Fallet-Bianco C. [Gliomas: WHO and Sainte-Anne Hospital classifications]. Ann Pathol 2000; 20:413-428. 3. Louis DN, Holland EC, Cairncross JG. Glioma classification: a molecular reappraisal. Am J Pathol 2001; 159:779-786. 4. Furnari FB, Huang HJ, Cavenee WK. Genetics and malignant progression of human brain tumors. Cancer Surv 1995; 25:233-275. 5. Ichimura K, Bolin MB, Goike HM, Schmidt EE, Moshref A, Collins VP. Deregulation of the p14ARF/MDM2/p53 pathway is a prerequisite for human astrocytic gliomas with G1-S transition control gene abnormalities. Cancer Res 2000; 60:417-424.
34
S. Chawla et al.
6. Musicco M, Filippini G, Bordo BM, Melotto A, Morello G, Berrino F. Gliomas and occupational exposure to carcinogens: case-control study. Am J Epidemiol 1982; 116:782-790. 7. Negendank WG, Sauter R, Brown TR, et al. Proton magnetic resonance spectroscopy in patients with glial tumors: a multicenter study. J Neurosurg 1996; 84:449-458. 8. Al-Okaili RN, Krejza J, Wang S, Woo JH, Melhem ER. Advanced MRI Techniques in the Diagnosis of Intra-axial Brain Tumors in Adults. Radiographics 2006; 26 Suppl 1:S173-189. 9. Meyerand ME, Pipas JM, Mamourian A, Tosteson TD, Dunn JF. Classification of biopsyconfirmed brain tumors using single-voxel MR spectroscopy. AJNR Am J Neuroradiol 1999; 20:117-123. 10. Haberg A, Kvistad KA, Unsgard G, Haraldseth O. Preoperative blood oxygen level-dependent functional magnetic resonance imaging in patients with primary brain tumors: clinical application and outcome. Neurosurgery 2004; 54:902-914; discussion 914-905. 11. Pirzkall A, Li X, Oh J, et al. 3D MRSI for resected high-grade gliomas before RT: tumor extent according to metabolic activity in relation to MRI. Int J Radiat Oncol Biol Phys 2004; 59:126-137. 12. Dean BL, Drayer BP, Bird CR, et al. Gliomas: classification with MRI. Radiology 1990; 174:411-415. 13. Atlas SW, Thulborn KR. MR detection of hyperacute parenchymal hemorrhage of the brain. AJNR Am J Neuroradiol 1998; 19:1471-1477. 14. Gupta RK, Rao SB, Jain R, et al. Differentiation of calcification from chronic hemorrhage with corrected gradient echo phase imaging. J Comput Assist Tomogr 2001; 25:698-704. 15. Yamada N, Imakita S, Sakuma T, Takamiya M. Intracranial calcification on gradient-echo phase image: depiction of diamagnetic susceptibility. Radiology 1996; 198:171-178. 16. Atlas SW, Howard RS, 2nd, Maldjian J, et al. Functional magnetic resonance imaging of regional brain activity in patients with intracerebral gliomas: findings and implications for clinical management. Neurosurgery 1996; 38:329-338. 17. Knopp EA, Cha S, Johnson G, et al. Glial neoplasms: dynamic contrast-enhanced T2-weighted MRI. Radiology 1999; 211:791-798. 18. Fayed N, Morales H, Modrego PJ, Pina MA. Contrast/Noise ratio on conventional MRI and choline/creatine ratio on proton MRI spectroscopy accurately discriminate low-grade from high-grade cerebral gliomas. Acad Radiol 2006; 13:728-737. 19. Ricci PE. Imaging of adult brain tumors. Neuroimaging Clin N Am 1999; 9:651-669. 20. Cozad SC, Townsend P, Morantz RA, Jenny AB, Kepes JJ, Smalley SR. Gliomatosis cerebri. Results with radiation therapy. Cancer 1996; 78:1789-1793. 21. White ML, Zhang Y, Smoker WR, et al. Fluid-attenuated inversion-recovery MRI in assessment of intracranial oligodendrogliomas. Comput Med Imaging Graph 2005; 29:279-285. 22. Andreula CF, Recchia-Luciani AN. Rationale for the use of contrast media in MRI. Neuroimaging Clin N Am 1997; 7:461-498. 23. Posner JB, Chernik NL. Intracranial metastases from systemic cancer. Adv Neurol 1978; 19:579-592. 24. Sze G, Milano E, Johnson C, Heier L. Detection of brain metastases: comparison of contrastenhanced MR with unenhanced MR and enhanced CT. AJNR Am J Neuroradiol 1990; 11:785-791. 25. Yuh WT, Tali ET, Nguyen HD, Simonson TM, Mayr NA, Fisher DJ. The effect of contrast dose, imaging time, and lesion size in the MR detection of intracerebral metastasis. AJNR Am J Neuroradiol 1995; 16:373-380. 26. Sze G, Johnson C, Kawamura Y, et al. Comparison of single- and triple-dose contrast material in the MR screening of brain metastases. AJNR Am J Neuroradiol 1998; 19:821-828. 27. Tang YM, Ngai S, Stuckey S. The solitary enhancing cerebral lesion: can FLAIR aid the differentiation between glioma and metastasis? AJNR Am J Neuroradiol 2006; 27:609-611. 28. Tokumaru A, O’Uchi T, Eguchi T, et al. Prominent meningeal enhancement adjacent to meningioma on Gd-DTPA-enhanced MR images: histopathologic correlation. Radiology 1990; 175:431-433.
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
35
29. Nagele T, Petersen D, Klose U, Grodd W, Opitz H, Voigt K. The “dural tail” adjacent to meningiomas studied by dynamic contrast-enhanced MRI: a comparison with histopathology. Neuroradiology 1994; 36:303-307. 30. Kremer P, Forsting M, Hamer J, Sartor K. MR enhancement of the internal auditory canal induced by tissue implant after resection of acoustic neurinoma. AJNR Am J Neuroradiol 1998; 19:115-118. 31. Johnson BA, Fram EK, Johnson PC, Jacobowitz R. The variable MR appearance of primary lymphoma of the central nervous system: comparison with histopathologic features. AJNR Am J Neuroradiol 1997; 18:563-572. 32. Kleihues P, Burger PC, Scheithauer BW. The new WHO classification of brain tumors. Brain Pathol 1993; 3:255-268. 33. Ikushima I, Korogi Y, Hirai T, et al. MR of epidermoids with a variety of pulse sequences. AJNR Am J Neuroradiol 1997; 18:1359-1363. 34. Ernemann U, Rieger J, Tatagiba M, Weller M. An MRI view of a ruptured dermoid cyst. Neurology 2006; 66:270. 35. Plate KH, Risau W. Angiogenesis in malignant gliomas. Glia 1995; 15:339-347. 36. Zagzag D, Zhong H, Scalzitti JM, Laughner E, Simons JW, Semenza GL. Expression of hypoxia-inducible factor 1alpha in brain tumors: association with angiogenesis, invasion, and progression. Cancer 2000; 88:2606-2618. 37. Boxerman JL, Schmainda KM, Weisskoff RM. Relative cerebral blood volume maps corrected for contrast agent extravasation significantly correlate with glioma tumor grade, whereas uncorrected maps do not. AJNR Am J Neuroradiol 2006; 27:859-867. 38. Aronen HJ, Perkio J. Dynamic susceptibility contrast MRI of gliomas. Neuroimaging Clin N Am 2002; 12:501-523. 39. Donahue KM, Krouwer HG, Rand SD, et al. Utility of simultaneously acquired gradient-echo and spin echo cerebral blood volume and morphology maps in brain tumor patients. Magn Reson Med 2000; 43:845-853. 40. Tofts PS, Kermode AG. Measurement of the blood-brain barrier permeability and leakage space using dynamic MRI. 1. Fundamental concepts. Magn Reson Med 1991; 17:357-367. 41. Dixon WT, Du LN, Faul DD, Gado M, Rossnick S. Projection angiograms of blood labeled by adiabatic fast passage. Magn Reson Med 1986; 3:454-462. 42. Sardashti M, Schwartzberg DG, Stomp GP, Dixon WT. Spin-labeling angiography of the carotids by presaturation and simplified adiabatic inversion. Magn Reson Med 1990; 15:192-200. 43. St Lawrence KS, Frank JA, McLaughlin AC. Effect of restricted water exchange on cerebral blood flow values calculated with arterial spin tagging: a theoretical investigation. Magn Reson Med 2000; 44:440-449. 44. Aronen HJ, Gazit IE, Louis DN, et al. Cerebral blood volume maps of gliomas: comparison with tumor grade and histologic findings. Radiology 1994; 191:41-51. 45. Sugahara T, Korogi Y, Kochi M, et al. Correlation of MRI-determined cerebral blood volume maps with histologic and angiographic determination of vascularity of gliomas. AJR Am J Roentgenol 1998; 171:1479-1486. 46. Daumas-Duport C, Tucker ML, Kolles H, et al. Oligodendrogliomas. Part II: A new grading system based on morphological and imaging criteria. J Neurooncol 1997; 34:61-78. 47. Cha S, Tihan T, Crawford F, et al. Differentiation of low-grade oligodendrogliomas from lowgrade astrocytomas by using quantitative blood-volume measurements derived from dynamic susceptibility contrast-enhanced MRI. AJNR Am J Neuroradiol 2005; 26:266-273. 48. Gelabert-Gonzalez M, Fernandez-Villa JM, Lopez-Garcia E, Gonzalez-Garcia J, Garcia-Allut A. [Choroid plexus tumors]. Rev Neurol 2001; 33:177-183. 49. Patankar TF, Haroon HA, Mills SJ, et al. Is volume transfer coefficient (K(trans) ) related to histologic grade in human gliomas? AJNR Am J Neuroradiol 2005; 26:2455-2465. 50. Pollack IF, Lunsford LD, Flickinger JC, Dameshek HL. Prognostic factors in the diagnosis and treatment of primary central nervous system lymphoma. Cancer 1989; 63:939-947. 51. Cotton F, Ongolo-Zogo P, Louis-Tisserand G, et al. [Diffusion and perfusion MRI in cerebral lymphomas]. J Neuroradiol 2006; 33:220-228.
36
S. Chawla et al.
52. Posner JB. Management of brain metastases. Rev Neurol (Paris) 1992; 148:477-487. 53. Long DM. Capillary ultrastructure in human metastatic brain tumors. J Neurosurg 1979; 51:53-58. 54. Law M, Cha S, Knopp EA, Johnson G, Arnett J, Litt AW. High-grade gliomas and solitary metastases: differentiation by using perfusion and proton spectroscopic MRI. Radiology 2002; 222:715-721. 55. Stewart PA, Hayakawa K, Farrell CL, Del Maestro RF. Quantitative study of microvessel ultrastructure in human peritumoral brain tissue. Evidence for a blood-brain barrier defect. J Neurosurg 1987; 67:697-705. 56. Provenzale JM, Wang GR, Brenner T, Petrella JR, Sorensen AG. Comparison of permeability in high-grade and low-grade brain tumors using dynamic susceptibility contrast MRI. AJR Am J Roentgenol 2002; 178:711-716. 57. Law M, Yang S, Babb JS, et al. Comparison of cerebral blood volume and vascular permeability from dynamic susceptibility contrast-enhanced perfusion MRI with glioma grade. AJNR Am J Neuroradiol 2004; 25:746-755. 58. Roberts HC, Roberts TP, Brasch RC, Dillon WP. Quantitative measurement of microvascular permeability in human brain tumors achieved using dynamic contrast-enhanced MRI: correlation with histologic grade. AJNR Am J Neuroradiol 2000; 21:891-899. 59. George ML, Dzik-Jurasz AS, Padhani AR, et al. Non-invasive methods of assessing angiogenesis and their value in predicting response to treatment in colorectal cancer. Br J Surg 2001; 88:1628-1636. 60. Roberts HC, Roberts TP, Bollen AW, Ley S, Brasch RC, Dillon WP. Correlation of microvascular permeability derived from dynamic contrast-enhanced MRI with histologic grade and tumor labeling index: a study in human brain tumors. Acad Radiol 2001; 8:384-391. 61. Ludemann L, Grieger W, Wurm R, Budzisch M, Hamm B, Zimmer C. Comparison of dynamic contrast-enhanced MRI with WHO tumor grading for gliomas. Eur Radiol 2001; 11:1231-1241. 62. Johnson G, Wetzel SG, Cha S, Babb J, Tofts PS. Measuring blood volume and vascular transfer constant from dynamic, T(2)*-weighted contrast-enhanced MRI. Magn Reson Med 2004; 51:961-968. 63. Wolf RL, Wang J, Wang S, et al. Grading of CNS neoplasms using continuous arterial spin labeled perfusion MRI at 3 Tesla. J Magn Reson Imaging 2005; 22:475-482. 64. Warmuth C, Gunther M, Zimmer C. Quantification of blood flow in brain tumors: comparison of arterial spin labeling and dynamic susceptibility-weighted contrast-enhanced MRI. Radiology 2003; 228:523-532. 65. Chiang IC, Kuo YT, Lu CY, et al. Distinction between high-grade gliomas and solitary metastases using peritumoral 3-T magnetic resonance spectroscopy, diffusion, and perfusion imagings. Neuroradiology 2004; 46:619-627. 66. Burger PC. Morphologic correlates in gliomas: where do we stand? Monogr Pathol 1990:16-29. 67. Muir KW, Buchan A, von Kummer R, Rother J, Baron JC. Imaging of acute stroke. Lancet Neurol 2006; 5:755-768. 68. Bulakbasi N, Kocaoglu M, Ors F, Tayfun C, Ucoz T. Combination of single-voxel proton MR spectroscopy and apparent diffusion coefficient calculation in the evaluation of common brain tumors. AJNR Am J Neuroradiol 2003; 24:225-233. 69. Sundgren PC, Dong Q, Gomez-Hassan D, Mukherji SK, Maly P, Welsh R. Diffusion tensor imaging of the brain: review of clinical applications. Neuroradiology 2004; 46:339-350. 70. Bulakbasi N, Guvenc I, Onguru O, Erdogan E, Tayfun C, Ucoz T. The added value of the apparent diffusion coefficient calculation to magnetic resonance imaging in the differentiation and grading of malignant brain tumors. J Comput Assist Tomogr 2004; 28:735-746. 71. Gupta RK, Sinha U, Cloughesy TF, Alger JR. Inverse correlation between choline magnetic resonance spectroscopy signal intensity and the apparent diffusion coefficient in human glioma. Magn Reson Med 1999; 41:2-7.
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
37
72. Sugahara T, Korogi Y, Kochi M, et al. Usefulness of diffusion-weighted MRI with echo-planar technique in the evaluation of cellularity in gliomas. J Magn Reson Imaging 1999; 9:53-60. 73. Kitis O, Altay H, Calli C, Yunten N, Akalin T, Yurtseven T. Minimum apparent diffusion coefficients in the evaluation of brain tumors. Eur J Radiol 2005; 55:393-400. 74. Kono K, Inoue Y, Nakayama K, et al. The role of diffusion-weighted imaging in patients with brain tumors. AJNR Am J Neuroradiol 2001; 22:1081-1088. 75. Yang D, Korogi Y, Sugahara T, et al. Cerebral gliomas: prospective comparison of multivoxel 2D chemical-shift imaging proton MR spectroscopy, echoplanar perfusion and diffusionweighted MRI. Neuroradiology 2002; 44:656-666. 76. Lam WW, Poon WS, Metreweli C. Diffusion MRI in glioma: does it have any role in the pre-operation determination of grading of glioma? Clin Radiol 2002; 57:219-225. 77. Rollin N, Guyotat J, Streichenberger N, Honnorat J, Tran Minh VA, Cotton F. Clinical relevance of diffusion and perfusion magnetic resonance imaging in assessing intra-axial brain tumors. Neuroradiology 2006; 48:150-159. 78. Fan GG, Deng QL, Wu ZH, Guo QY. Usefulness of diffusion/perfusion-weighted MRI in patients with non-enhancing supratentorial brain gliomas: a valuable tool to predict tumor grading? Br J Radiol 2006; 79:652-658. 79. Lu S, Ahn D, Johnson G, Law M, Zagzag D, Grossman RI. Diffusion-tensor MRI of intracranial neoplasia and associated peritumoral edema: introduction of the tumor infiltration index. Radiology 2004; 232:221-228. 80. Roberts TP, Liu F, Kassner A, Mori S, Guha A. Fiber density index correlates with reduced fractional anisotropy in white matter of patients with glioblastoma. AJNR Am J Neuroradiol 2005; 26:2183-2186. 81. Beppu T, Inoue T, Shibata Y, et al. Fractional anisotropy value by diffusion tensor magnetic resonance imaging as a predictor of cell density and proliferation activity of glioblastomas. Surg Neurol 2005; 63:56-61; discussion 61. 82. Krabbe K, Gideon P, Wagn P, Hansen U, Thomsen C, Madsen F. MR diffusion imaging of human intracranial tumors. Neuroradiology 1997; 39:483-489. 83. Lu S, Ahn D, Johnson G, Cha S. Peritumoral diffusion tensor imaging of high-grade gliomas and metastatic brain tumors. AJNR Am J Neuroradiol 2003; 24:937-941. 84. Yamasaki F, Kurisu K, Satoh K, et al. Apparent diffusion coefficient of human brain tumors at MRI. Radiology 2005; 235:985-991. 85. Guo AC, Cummings TJ, Dash RC, Provenzale JM. Lymphomas and high-grade astrocytomas: comparison of water diffusibility and histologic characteristics. Radiology 2002; 224:177-183. 86. Filippi CG, Edgar MA, Ulug AM, Prowda JC, Heier LA, Zimmerman RD. Appearance of meningiomas on diffusion-weighted images: correlating diffusion constants with histopathologic findings. AJNR Am J Neuroradiol 2001; 22:65-72. 87. Mori S, van Zijl PC. Fiber tracking: principles and strategies - a technical review. NMR Biomed 2002; 15:468-480. 88. Wakana S, Jiang H, Nagae-Poetscher LM, van Zijl PC, Mori S. Fiber tract-based atlas of human white matter anatomy. Radiology 2004; 230:77-87. 89. Catani M, Howard RJ, Pajevic S, Jones DK. Virtual in vivo interactive dissection of white matter fasciculi in the human brain. Neuroimage 2002; 17:77-94. 90. Yu CS, Li KC, Xuan Y, Ji XM, Qin W. Diffusion tensor tractography in patients with cerebral tumors: a helpful technique for neurosurgical planning and postoperative assessment. Eur J Radiol 2005; 56:197-204. 91. Witwer BP, Moftakhar R, Hasan KM, et al. Diffusion-tensor imaging of white matter tracts in patients with cerebral neoplasm. J Neurosurg 2002; 97:568-575. 92. Price SJ, Burnet NG, Donovan T, et al. Diffusion tensor imaging of brain tumors at 3T: a potential tool for assessing white matter tract invasion? Clin Radiol 2003; 58:455-462. 93. Sinha S, Bastin ME, Whittle IR, Wardlaw JM. Diffusion tensor MRI of high-grade cerebral gliomas. AJNR Am J Neuroradiol 2002; 23:520-527.
38
S. Chawla et al.
94. Burlina AP, Aureli T, Bracco F, Conti F, Battistin L. MR spectroscopy: a powerful tool for investigating brain function and neurological diseases. Neurochem Res 2000; 25:1365-1372. 95. Bhakoo KK, Williams IT, Williams SR, Gadian DG, Noble MD. Proton nuclear magnetic resonance spectroscopy of primary cells derived from nervous tissue. J Neurochem 1996; 66:1254-1263. 96. Birken DL, Oldendorf WH. N-acetyl-L-aspartic acid: a literature review of a compound prominent in 1H-NMR spectroscopic studies of brain. Neurosci Biobehav Rev 1989; 13:23-31. 97. Ross B, Michaelis T. Clinical applications of magnetic resonance spectroscopy. Magn Reson Q 1994; 10:191-247. 98. Howe FA, Barton SJ, Cudlip SA, et al. Metabolic profiles of human brain tumors using quantitative in vivo 1H magnetic resonance spectroscopy. Magn Reson Med 2003; 49:223-232. 99. Shimizu H, Kumabe T, Shirane R, Yoshimoto T. Correlation between choline level measured by proton MR spectroscopy and Ki-67 labeling index in gliomas. AJNR Am J Neuroradiol 2000; 21:659-665. 100. Shino A, Nakasu S, Matsuda M, Handa J, Morikawa S, Inubushi T. Noninvasive evaluation of the malignant potential of intracranial meningiomas performed using proton magnetic resonance spectroscopy. J Neurosurg 1999; 91:928-934. 101. Manton DJ, Lowry M, Blackband SJ, Horsman A. Determination of proton metabolite concentrations and relaxation parameters in normal human brain and intracranial tumors. NMR Biomed 1995; 8:104-112. 102. Murphy M, Loosemore A, Clifton AG, et al. The contribution of proton magnetic resonance spectroscopy (1HMRS) to clinical brain tumor diagnosis. Br J Neurosurg 2002; 16:329-334. 103. Ishimaru H, Morikawa M, Iwanaga S, Kaminogo M, Ochi M, Hayashi K. Differentiation between high-grade glioma and metastatic brain tumor using single-voxel proton MR spectroscopy. Eur Radiol 2001; 11:1784-1791. 104. Kreis R, Ernst T, Ross BD. Development of the human brain: In vivo quantification of metabolite and water content with proton magnetic resonance spectroscopy. Magn Reson Med 1993; 30:424-437. 105. Castillo M, Smith JK, Kwock L. Correlation of myo-inositol levels and grading of cerebral astrocytomas. AJNR Am J Neuroradiol 2000; 21:1645-1649. 106. Barba I, Moreno A, Martinez-Perez I, et al. Magnetic resonance spectroscopy of brain hemangiopericytomas: high myoinositol concentrations and discrimination from meningiomas. J Neurosurg 2001; 94:55-60. 107. Norfray JF, Tomita T, Byrd SE, Ross BD, Berger PA, Miller RS. Clinical impact of MR spectroscopy when MRI is indeterminate for pediatric brain tumors. AJR Am J Roentgenol 1999; 173:119-125. 108. Opstad KS, Provencher SW, Bell BA, Griffiths JR, Howe FA. Detection of elevated glutathione in meningiomas by quantitative in vivo 1H MRS. Magn Reson Med 2003; 49:632-637. 109. Stubbs M, Veech RL, Griffiths JR. Tumor metabolism: the lessons of magnetic resonance spectroscopy. Adv Enzyme Regul 1995; 35:101-115. 110. Kugel H, Heindel W, Ernestus RI, Bunke J, du Mesnil R, Friedmann G. Human brain tumors: spectral patterns detected with localized H-1 MR spectroscopy. Radiology 1992; 183:701-709. 111. Alger JR, Frank JA, Bizzi A, et al. Metabolism of human gliomas: assessment with H-1 MR spectroscopy and F-18 fluorodeoxyglucose PET. Radiology 1990; 177:633-641. 112. Auer DP, Gossl C, Schirmer T, Czisch M. Improved analysis of 1H-MR spectra in the presence of mobile lipids. Magn Reson Med 2001; 46:615-618. 113. Kuesel AC, Sutherland GR, Halliday W, Smith IC. 1H MRS of high-grade astrocytomas: mobile lipid accumulation in necrotic tissue. NMR Biomed 1994; 7:149-155. 114. Barba I, Cabanas ME, Arus C. The relationship between nuclear magnetic resonance-visible lipids, lipid droplets, and cell proliferation in cultured C6 cells. Cancer Res 1999; 59:1861-1868. 115. Hakumaki JM, Poptani H, Sandmair AM, Yla-Herttuala S, Kauppinen RA. 1H MRS detects polyunsaturated fatty acid accumulation during gene therapy of glioma: implications for the in vivo detection of apoptosis. Nat Med 1999; 5:1323-1327.
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
39
116. Law M, Yang S, Wang H, et al. Glioma grading: sensitivity, specificity, and predictive values of perfusion MRI and proton MR spectroscopic imaging compared with conventional MRI. AJNR Am J Neuroradiol 2003; 24:1989-1998. 117. Herminghaus S, Dierks T, Pilatus U, et al. Determination of histopathological tumor grade in neuroepithelial brain tumors by using spectral pattern analysis of in vivo spectroscopic data. J Neurosurg 2003; 98:74-81. 118. Hsu YY, Chang CN, Wie KJ, Lim KE, Hsu WC, Jung SM. Proton magnetic resonance spectroscopic imaging of cerebral gliomas: correlation of metabolite ratios with histopathologic grading. Chang Gung Med J 2004; 27:399-407. 119. Kaminogo M, Ishimaru H, Morikawa M, et al. Diagnostic potential of short echo time MR spectroscopy of gliomas with single-voxel and point-resolved spatially localised proton spectroscopy of brain. Neuroradiology 2001; 43:353-363. 120. Ott D, Hennig J, Ernst T. Human brain tumors: assessment with in vivo proton MR spectroscopy. Radiology 1993; 186:745-752. 121. Paulus W, Peiffer J. Intratumoral histologic heterogeneity of gliomas. A quantitative study. Cancer 1989; 64:442-447. 122. Cheng LL, Chang IW, Louis DN, Gonzalez RG. Correlation of high-resolution magic angle spinning proton magnetic resonance spectroscopy with histopathology of intact human brain tumor specimens. Cancer Res 1998; 58:1825-1832. 123. Sabatier J, Ibarrola D, Malet-Martino M, Berry I. [Brain tumors: interest of magnetic resonance spectroscopy for the diagnosis and the prognosis]. Rev Neurol (Paris) 2001; 157:858-862. 124. McKnight TR, von dem Bussche MH, Vigneron DB, et al. Histopathological validation of a three-dimensional magnetic resonance spectroscopy index as a predictor of tumor presence. J Neurosurg 2002; 97:794-802. 125. Zhang M, Olsson Y. Hematogenous metastases of the human brain–characteristics of peritumoral brain changes: a review. J Neurooncol 1997; 35:81-89. 126. Fayed N, Modrego PJ. The contribution of magnetic resonance spectroscopy and echoplanar perfusion-weighted MRI in the initial assessment of brain tumors. J Neurooncol 2005; 72:261-265. 127. Beaulieu C, Allen PS. Determinants of anisotropic water diffusion in nerves. Magn Reson Med 1994; 31:394-400. 128. Ogawa S, Menon RS, Tank DW, et al. Functional brain mapping by blood oxygenation leveldependent contrast magnetic resonance imaging. A comparison of signal characteristics with a biophysical model. Biophys J 1993; 64:803-812. 129. Heeger DJ, Ress D. What does fMRI tell us about neuronal activity? Nat Rev Neurosci 2002; 3:142-151. 130. Toronov V, Walker S, Gupta R, et al. The roles of changes in deoxyhemoglobin concentration and regional cerebral blood volume in the fMRI BOLD signal. Neuroimage 2003; 19:1521-1531. 131. Wilkinson ID, Romanowski CA, Jellinek DA, Morris J, Griffiths PD. Motor functional MRI for pre-operative and intraoperative neurosurgical guidance. Br J Radiol 2003; 76:98-103. 132. Pujol J, Conesa G, Deus J, et al. Presurgical identification of the primary sensorimotor cortex by functional magnetic resonance imaging. J Neurosurg 1996; 84:7-13. 133. Pujol J, Conesa G, Deus J, Lopez-Obarrio L, Isamat F, Capdevila A. Clinical application of functional magnetic resonance imaging in presurgical identification of the central sulcus. J Neurosurg 1998; 88:863-869. 134. Lehericy S, Duffau H, Cornu P, et al. Correspondence between functional magnetic resonance imaging somatotopy and individual brain anatomy of the central region: comparison with intraoperative stimulation in patients with brain tumors. J Neurosurg 2000; 92:589-598. 135. Lee CC, Ward HA, Sharbrough FW, et al. Assessment of functional MRI in neurosurgical planning. AJNR Am J Neuroradiol 1999; 20:1511-1519. 136. Fandino J, Kollias SS, Wieser HG, Valavanis A, Yonekawa Y. Intraoperative validation of functional magnetic resonance imaging and cortical reorganization patterns in patients with brain tumors involving the primary motor cortex. J Neurosurg 1999; 91:238-250.
40
S. Chawla et al.
137. Holodny AI, Schulder M, Liu WC, Wolko J, Maldjian JA, Kalnin AJ. The effect of brain tumors on BOLD functional MRI activation in the adjacent motor cortex: implications for image-guided neurosurgery. AJNR Am J Neuroradiol 2000; 21:1415-1422. 138. Ulmer JL, Hacein-Bey L, Mathews VP, et al. Lesion-induced pseudo-dominance at functional magnetic resonance imaging: implications for preoperative assessments. Neurosurgery 2004; 55:569-579; discussion 580-561. 139. Schreiber A, Hubbe U, Ziyeh S, Hennig J. The influence of gliomas and nonglial spaceoccupying lesions on blood-oxygen-level-dependent contrast enhancement. AJNR Am J Neuroradiol 2000; 21:1055-1063. 140. Hossmann KA, Linn F, Okada Y. Bioluminescence and fluoroscopic imaging of tissue pH and metabolites in experimental brain tumors of cat. NMR Biomed 1992; 5:259-264. 141. Laurienti PJ, Field AS, Burdette JH, Maldjian JA, Yen YF, Moody DM. Dietary caffeine consumption modulates fMRI measures. Neuroimage 2002; 17:751-757. 142. Buhmann C, Glauche V, Sturenburg HJ, Oechsner M, Weiller C, Buchel C. Pharmacologically modulated fMRI–cortical responsiveness to levodopa in drug-naive hemiparkinsonian patients. Brain 2003; 126:451-461. 143. Kamada K, Todo T, Masutani Y, et al. Combined use of tractography-integrated functional neuronavigation and direct fiber stimulation. J Neurosurg 2005; 102:664-672. 144. Li ZX, Dai JP, Jiang T, et al. [Function magnetic resonance imaging and diffusion tensor tractography in patients with brain gliomas involving motor areas: clinical application and outcome]. Zhonghua Wai Ke Za Zhi 2006; 44:1275-1279. 145. Schonberg T, Pianka P, Hendler T, Pasternak O, Assaf Y. Characterization of displaced white matter by brain tumors using combined DTI and fMRI. Neuroimage 2006; 30:1100-1111. 146. Bonavita S, Di Salle F, Tedeschi G. Proton MRS in neurological disorders. Eur J Radiol 1999; 30:125-131. 147. Burger PC, Fuller GN. Pathology–trends and pitfalls in histologic diagnosis, immunopathology, and applications of oncogene research. Neurol Clin 1991; 9:249-271. 148. Fulham MJ, Bizzi A, Dietz MJ, et al. Mapping of brain tumor metabolites with proton MR spectroscopic imaging: clinical relevance. Radiology 1992; 185:675-686. 149. Sijens PE, Vecht CJ, Levendag PC, van Dijk P, Oudkerk M. Hydrogen magnetic resonance spectroscopy follow-up after radiation therapy of human brain cancer. Unexpected inverse correlation between the changes in tumor choline level and post-gadolinium magnetic resonance imaging contrast. Invest Radiol 1995; 30:738-744. 150. Usenius T, Usenius JP, Tenhunen M, et al. Radiation-induced changes in human brain metabolites as studied by 1H nuclear magnetic resonance spectroscopy in vivo. Int J Radiat Oncol Biol Phys 1995; 33:719-724. 151. Taylor JS, Langston JW, Reddick WE, et al. Clinical value of proton magnetic resonance spectroscopy for differentiating recurrent or residual brain tumor from delayed cerebral necrosis. Int J Radiat Oncol Biol Phys 1996; 36:1251-1261. 152. Sugahara T, Korogi Y, Tomiguchi S, et al. Posttherapeutic intraaxial brain tumor: the value of perfusion-sensitive contrast-enhanced MRI for differentiating tumor recurrence from nonneoplastic contrast-enhancing tissue. AJNR Am J Neuroradiol 2000; 21:901-909. 153. Hein PA, Eskey CJ, Dunn JF, Hug EB. Diffusion-weighted imaging in the follow-up of treated high-grade gliomas: tumor recurrence versus radiation injury. AJNR Am J Neuroradiol 2004; 25:201-209. 154. Asao C, Korogi Y, Kitajima M, et al. Diffusion-weighted imaging of radiation-induced brain injury for differentiation from tumor recurrence. AJNR Am J Neuroradiol 2005; 26:1455-1460. 155. Willems JG, Alva-Willems JM. Accuracy of cytologic diagnosis of central nervous system neoplasms in sterotactic biopsies. Acta Cytol 1984; 28:243-249. 156. Fratkin JD, Ward MM, Roberts DW, Sullivan MM. CT-guided stereotactic biopsy of intracranial lesions: correlation between core biopsy and aspiration smear. Diagn Cytopathol 1986; 2:126-132. 157. Krieger MD, Chandrasoma PT, Zee CS, Apuzzo ML. Role of stereotactic biopsy in the diagnosis and management of brain tumors. Semin Surg Oncol 1998; 14:13-25.
1 Anatomic, Physiologic and Metabolic Imaging in Neuro-Oncology
41
158. Kelly PJ, Daumas-Duport C, Kispert DB, Kall BA, Scheithauer BW, Illig JJ. Imaging-based stereotaxic serial biopsies in untreated intracranial glial neoplasms. J Neurosurg 1987; 66:865-874. 159. Cha S, Knopp EA, Johnson G, Wetzel SG, Litt AW, Zagzag D. Intracranial mass lesions: dynamic contrast-enhanced susceptibility-weighted echo-planar perfusion MRI. Radiology 2002; 223:11-29. 160. Baik HM, Choe BY, Son BC, et al. Feasibility of proton chemical shift imaging with a stereotactic headframe. Magn Reson Imaging 2003; 21:55-59. 161. Kondziolka D, Somaza S, Comey C, et al. Radiosurgery and fractionated radiation therapy: comparison of different techniques in an in vivo rat glioma model. J Neurosurg 1996; 84:1033-1038. 162. Hormigo A, Friedlander DR, Brittis PA, Zagzag D, Grumet M. Reduced tumorigenicity of rat glioma cells in the brain when mediated by hygromycin phosphotransferase. J Neurosurg 2001; 94:596-604. 163. Hakumaki JM, Poptani H, Puumalainen AM, et al. Quantitative 1H nuclear magnetic resonance diffusion spectroscopy of BT4C rat glioma during thymidine kinase-mediated gene therapy in vivo: identification of apoptotic response. Cancer Res 1998; 58:3791-3799. 164. Jacobs AH, Dittmar C, Winkeler A, Garlip G, Heiss WD. Molecular imaging of gliomas. Mol Imaging 2002; 1:309-335. 165. Zhang Z, Jiang Q, Jiang F, et al. in vivo magnetic resonance imaging tracks adult neural progenitor cell targeting of brain tumor. Neuroimage 2004; 23:281-287. 166. Magnitsky S, Watson DJ, Walton RM, et al. in vivo and Ex Vivo MRI detection of localized and disseminated neural stem cell grafts in the mouse brain. Neuroimage 2005; 26:744-754. 167. Su H, Forbes A, Gambhir SS, Braun J. Quantitation of cell number by a positron emission tomography reporter gene strategy. Mol Imaging Biol 2004; 6:139-148. 168. Doubrovin M, Ponomarev V, Beresten T, et al. Imaging transcriptional regulation of p53dependent genes with positron emission tomography in vivo. Proc Natl Acad Sci U S A 2001; 98:9300-9305. 169. Serganova I, Doubrovin M, Vider J, et al. Molecular imaging of temporal dynamics and spatial heterogeneity of hypoxia-inducible factor-1 signal transduction activity in tumors in living mice. Cancer Res 2004; 64:6101-6108. 170. Uhrbom L, Nerio E, Holland EC. Dissecting tumor maintenance requirements using bioluminescence imaging of cell proliferation in a mouse glioma model. Nat Med 2004; 10:1257-1260. 171. Wen B, Burgman P, Zanzonico P, et al. A preclinical model for noninvasive imaging of hypoxia-induced gene expression; comparison with an exogenous marker of tumor hypoxia. Eur J Nucl Med Mol Imaging 2004; 31:1530-1538. 172. Gillies RJ, Raghunand N, Karczmar GS, Bhujwalla ZM. MRI of the tumor microenvironment. J Magn Reson Imaging 2002; 16:430-450. 173. Garcia-Martin ML, Martinez GV, Raghunand N, Sherry AD, Zhang S, Gillies RJ. High resolution pH(e) imaging of rat glioma using pH-dependent relaxivity. Magn Reson Med 2006; 55:309-315. 174. Poptani H, Duvvuri U, Miller CG, et al. T1rho imaging of murine brain tumors at 4 T. Acad Radiol 2001; 8:42-47. 175. Phelps ME, Mazziotta JC. Positron emission tomography: human brain function and biochemistry. Science 1985; 228:799-809. 176. Phelps ME. PET: the merging of biology and imaging into molecular imaging. J Nucl Med 2000; 41:661-681. 177. Price P. PET as a potential tool for imaging molecular mechanisms of oncology in man. Trends Mol Med 2001; 7:442-446.
42
S. Chawla et al.
178. Heiss WD, Heindel W, Herholz K, et al. Positron emission tomography of fluorine-18-deoxyglucose and image-guided phosphorus-31 magnetic resonance spectroscopy in brain tumors. J Nucl Med 1990; 31:302-310. 179. Herholz K, Heindel W, Luyten PR, et al. in vivo imaging of glucose consumption and lactate concentration in human gliomas. Ann Neurol 1992; 31:319-327. 180. Mineura K, Yasuda T, Kowada M, Shishido F, Ogawa T, Uemura K. Positron emission tomographic evaluation of histological malignancy in gliomas using oxygen-15 and fluorine18-fluorodeoxyglucose. Neurol Res 1986; 8:164-168. 181. Nishioka T, Oda Y, Seino Y, et al. Distribution of the glucose transporters in human brain tumors. Cancer Res 1992; 52:3972-3979. 182. Brooks DJ, Beaney RP, Lammertsma AA, et al. Glucose transport across the blood-brain barrier in normal human subjects and patients with cerebral tumors studied using [11C]3-O-methyl-D-glucose and positron emission tomography. J Cereb Blood Flow Metab 1986; 6:230-239. 183. Glantz MJ, Hoffman JM, Coleman RE, et al. Identification of early recurrence of primary central nervous system tumors by [18F]fluorodeoxyglucose positron emission tomography. Ann Neurol 1991; 29:347-355. 184. Kim EE, Chung SK, Haynie TP, et al. Differentiation of residual or recurrent tumors from post-treatment changes with F-18 FDG PET. Radiographics 1992; 12:269-279. 185. Tyler JL, Diksic M, Villemure JG, et al. Metabolic and hemodynamic evaluation of gliomas using positron emission tomography. J Nucl Med 1987; 28:1123-1133. 186. Isselbacher KJ. Sugar and amino acid transport by cells in culture–differences between normal and malignant cells. N Engl J Med 1972; 286:929-933. 187. Jager PL, Vaalburg W, Pruim J, de Vries EG, Langen KJ, Piers DA. Radiolabeled amino acids: basic aspects and clinical applications in oncology. J Nucl Med 2001; 42:432-445. 188. Sato N, Suzuki M, Kuwata N, et al. Evaluation of the malignancy of glioma using 11Cmethionine positron emission tomography and proliferating cell nuclear antigen staining. Neurosurg Rev 1999; 22:210-214. 189. Herholz K, Holzer T, Bauer B, et al. 11C-methionine PET for differential diagnosis of low-grade gliomas. Neurology 1998; 50:1316-1322. 190. Goldman S, Levivier M, Pirotte B, et al. Regional methionine and glucose uptake in highgrade gliomas: a comparative study on PET-guided stereotactic biopsy. J Nucl Med 1997; 38:1459-1462. 191. Jacobs A, Tjuvajev JG, Dubrovin M, et al. Positron emission tomography-based imaging of transgene expression mediated by replication-conditional, oncolytic herpes simplex virus type 1 mutant vectors in vivo. Cancer Res 2001; 61:2983-2995.
2
Imaging of Spinal Tumors Izlem Izbudak1, Aylin Tekes1, Juan Carlos Baez2, and Kieran Murphy3
Key Points ●
●
●
●
●
●
●
●
Primary spinal tumors are rare, metastasis to spine is common and comprise 90 percent of spinal column tumors. The advent of MRI revolutionized the characterization of spinal tumors by providing detailed direct visualization of the bone marrow and spinal cord in multiple planes, allowing earlier detection and treatment for both intradural and extradural tumors. Sagittal T1-weighted and STIR images are the most sensitive sequences for the bone marrow lesions, even in the early phase. Contrast-enhanced T1-weighted images are more sensitive when fat-saturation is applied. Whole body MRI with STIR sequence is currently possible in a short scan time revealing metastatic disease in both the spine and in solid organs. Differentiating osteoporotic acute compression fractures of the vertebra from malignant compression fractures is challenging even with MRI. MR techniques, such as DWI or chemical shift imaging, have been studied and quantitative techniques might be helpful. Early signs of cord or cauda equina compression is progressive sharp nerve root pain aggravated by bending or coughing; a limited sagittal T2-weighted MRI of the spine would be enough to evaluate compression. Most (90 percent to 95 percent) of the intramedullary tumors are malignant and predominantly composed of glial components. The most common types are ependymomas in adults and astrocytomas in children. Advanced imaging techniques such as MR spectroscopy, DWI, MT, and functional studies are currently limited by the strong magnetic field inhomogeneities
1
Johns Hopkins University, Department of Radiology, Neuroradiology Division.
2
Johns Hopkins University, School of Medicine.
3
Johns Hopkins University, Department of Radiology, Neurointerventional Division
Corresponding author: Izlem Izbudak, MD, 1211 Asquith Pines Place, Saverna Park, Arnold, MD 21012 e-mail:
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
43
44
●
I. Izbudak et al.
present in the spinal cord region, respiratory and cardiac movements, and the small size of the spinal cord. We recommend the use of vertebroplasty for painful destructive vertebral lesions. The few complications reported have mostly been related to excessive cement injection, underlining the need of excellent imaging conditions to control the cement injection.
Spinal tumors can be grouped into 2 main categories: extradural (bone) and intradural. Intradural tumors are further grouped into intradural/intramedullary and intradural/extramedullary components. Oncologists more commonly treat extradural malignant spinal tumors. Therefore, we will concentrate more on this group of tumors, complications such as compression fractures or cord compression and the percutaneous therapy with vertebroplasty. Relatively common benign and malignant primary bone tumors of the vertebrae are also briefly discussed, as well as the common intradural extramedullary and intradural intramedullary tumors.
1
Extradural Tumors
1.1
Malignant
1.1.1
Metastasis
The spine is a common site for metastatic disease of the breast, prostate, lung, kidney, thyroid, uterine carcinoma and melanoma. The lumbar spine is the most affected, followed by thoracic, cervical spine and sacrum [1]. Once established in an osseous location, metastatic tumor cells activate osteoclasts, which ultimately lead to bone resorption. [2, 3]. Direct tumor cell bone lyses also ensues. Conventional radiographs show metastatic bone lesions only after the loss of more than 50 percent of the bone mineral content at the site of the disease. However, they are helpful for characterization of the lesion as lytic, blastic or mixed. Additionally the fracture risk is traditionally determined on plain radiographs [4]. Computed tomography (CT) is valuable as an adjunct in detailing osseous anatomy, character and extent of the specific lesion. Also CT is used for guiding biopsies for previously detected vertebral lesions. MRI is a sensitive modality for the detection of metastatic disease and sometimes it provides improved specificity in characterization of the lesion. MRI can evaluate the lesion, its intramedullary and extramedullary extent, the degree of cortical involvement, the absence or presence of periosteal involvement and the extent of the soft tissue mass. Another advantage of MRI is to detect compressive myelopathy. The vertebral metastasis may be focal or diffuse, and diffuse metastasis may show a homogenous or heterogenous signal pattern on MRI. Diffuse inhomogenous metastasis can be differentiated from normal inhomogenous fatty marrow in elderly
2 Imaging of Spinal Tumors
45
Fig. 2.1 Diffuse prostate cancer metastasis. (a) Contrast-enhanced sagittal T1-weighted MR image does not reveal a definite enhancing lesion. (b) Sagittal STIR sequence shows multi-level bright tumor deposits (arrows) on the background of suppressed fatty marrow signal
individuals by using a short tau inversion recovery (STIR) sequence which shows multiple bright metastatic deposits within the background of dark patchy fatty marrow (Fig. 2.1). Lytic lesions may be seen in almost all tumor types. Bone metastases of bladder, kidney and thyroid cancer are invariably lytic. The lytic lesions usually show avid contrast enhancement on fat-saturated T1-weighted images (Fig. 2.2). Blastic lesions are frequently seen in prostate and breast cancer, occasionally in lung, stomach, pancreas and cervix carcinomas, and infrequently in colorectal cancer [5]. MRI shows focal areas of low signal intensity on both T1- and T2 sequences, and high signal intensity on STIR, though less conspicuous than the lytic pattern. With the recent development of turbo STIR sequences, it is possible to image the whole body in 30 to 40 minutes by using MRI, which also reveals solid organ
46
I. Izbudak et al.
Fig. 2.2 Diffuse breast cancer metastasis with soft tissue component. (a) Sagittal T1-weighted MR image shows diffuse decreased signal of the vertebrae that is almost isointense with the intervertebral discs. There is posterior soft tissue mass at T12 level (arrow). (b) Contrast-enhanced fat-saturated sagittal T1-weighted image demonstrates heterogenous enhancement of the vertebrae and the metastatic soft tissue mass (arrow)
metastasis such as liver, lung or brain, in addition to axial and peripheral skeleton metastasis [6, 7, 8]. Eustace, et al. [6] compared scintigraphy to whole body turbo STIR MRI in 25 patients with known or suspected skeletal metastasis and found that MRI is 96.5 percent sensitive and 100 percent specific with a positive predictive value (PPV) of 100 percent, whereas scintigraphy is 72 percent sensitive and 98 percent specific with a PPV of 95 percent.
1.1.2
Lymphoma
Primary bone involvement occurs in 3 percent to 5 percent of the patients with Non-Hodgkin’s Lymphoma, and 25 percent of them have secondary bone involvement. Primary bone involvement is rare in Hodgkin’s disease. Secondary bone involvement occurs in 5 percent to 20 percent of patients with Hodgkin’s disease during the course of the disease, but in only 1 percent to 4 percent at presentation. The radiographic and CT findings are nonspecific and represent late manifestations, more commonly osteolytic; ranging from a permeative moth-eaten pattern to a more geographic area of osteolytic destruction [9]. Patchy sclerosis, mixed osteolytic-sclerotic pattern and, rarely, “ivory vertebrae” are seen.
2 Imaging of Spinal Tumors
47
Fig. 2.3 Metastatic B-cell lymphoma. (a) Sagittal T2-weighted MR image shows slight heterogenous signal in L5 vertebra, soft tissue mass in the anterior epidural space (thin arrow) and a small hypointense nodule within the thecal sac adjacent to cauda equina roots at L4 level (arrow). (b) Contrast-enhanced fat-saturated T1-weighted sagittal image reveals bone marrow metastasis in L5 showing diffuse heterogenous enhancement with an enhancing soft tissue component in the anterior epidural space (thin arrow) and intradural extramedullary metastatic nodule also showing enhancement (arrow)
Lymphoma may produce diffuse infiltration of the bone marrow, usually in the low-grade Non-Hodgkin’s type, and it can only be detected on MRI. T1weighted images demonstrate diffuse hypointensity of the vertebrae and bright signal of the intervertebral discs. The finding of high signal intensity marrow on T2-weighted fat-suppressed images or an obvious contrast enhancement, particularly on fat-saturated contrast-enhanced T1-weighted images, help to differentiate it from normal hypercellular marrow. In Hodgkin’s lymphoma, intermediate and high-grade non-Hodgkin’s lymphoma, the bone involvement is usually focal. In Non-Hodgkin’s lymphoma epidural soft tissue mass can occur alone or as a component of vertebral or paraspinal tumors, and this might be present either at diagnosis or during the disease course (Fig. 2.3).
1.1.3
Leukemia
Leukemia usually shows diffuse bone marrow infiltration, rather than focal disease, and results in a decreased signal on T1-weighted images, presenting diagnostic challenges similar to diffuse lymphoma. In both diseases the decrease in bone marrow signal intensity
48
I. Izbudak et al.
on T1-weighted images can be homogenous or heterogenous yielding a “salt and pepper” pattern [10]. In patients with myelogenous leukemia, paraspinal soft tissue masses, e.g., chloromas, can occur with similar signal characteristics to the adjacent soft tissues.
1.1.4
Multiple Myeloma
In multiple myeloma (MM) there is replacement of the bone marrow with abnormal plasma cells which trigger excessive bone resorption and inhibition of bone formation [11]. The skeletal radiography is the primary imaging modality for detecting bone changes, and it is included in the Durie-Salmon clinical staging criteria of newly diagnosed multiple myeloma. CT is sensitive for punched-out lytic lesions, expansile lesions with soft tissue masses, diffuse osteopenia and fractures (Fig. 2.4). Mahnken, et al. [12] compared multi-detector CT of the spine to MRI and radiography in 18 patients with MM and found that, compared to conventional radiography, an additional 24 affected vertebrae, 15 additional vertebral fractures and six vertebrae at further risk of fracture were detected on CT. In patients with newly diagnosed MM, low tumor burden is normally associated with a normal MRI pattern; whereas, high tumor burden is suspected when marrow is diffusely hypointense on T1-weighted images, hyperintense on STIR images and enhancing on gadolinium-enhanced images. Overall marrow signal intensity may be homogeneous or heterogenous. Spinal compression fractures occur in 55 percent to 70 percent of patients with MM [13] (Fig. 2.5).
1.1.5
Differentiating Metastatic from Acute Benign Compression Fractures by Using MRI
Compression fractures due to metastatic malignancy are frequently seen in the same age group of osteoporotic compression fractures, and differentiation often affects appropriate clinical staging, treatment planning and prognostic determination in patients with known nonosseous malignancies [14]. Chronic benign compression fractures can be easily detected due to absence of abnormal signal intensity in a compressed vertebra [14,15]; however, acute osteoporotic compression fractures can be difficult to differentiate from malignant compression fractures. Jung, et al. [16] reported that distinction between metastatic and acute osteoporotic compression fractures could be made on the basis of MRI findings. The sensitivity, specificity and accuracy for metastatic compression fractures were 100 percent, 93 percent and 95 percent, respectively. MRI findings suggestive of metastatic compression fractures were as follows: convex posterior border of the vertebral body, abnormal signal intensity of the pedicle or posterior element, epidural mass, encasing epidural mass, focal paraspinal mass and other spinal metastasis. MRI findings suggestive of acute osteoporotic compression fracture were as follows: low-signal-intensity band on T1- and T2-weighted images, spared normal
2 Imaging of Spinal Tumors
49
Fig. 2.4 Multiple myeloma metastasis. (a) Axial CT scan at C1-2 level showing lytic expansile destruction of C1 vertebra. (b) Coronal 2D reformat CT image demonstrates additional multiple punched out lytic lesions (arrows) in whole C-spine secondary to multiple myeloma
bone marrow signal intensity of the vertebral body, retropulsion of a posterior bone fragment and multiple compression fractures [16] (Fig. 2.6). Chemical shift MRI (in-phase, opposed-phase imaging) of the spine was hypothesized to be sensitive and specific for differentiating pathologic from acute compression fractures, based on the fact that the presence of fat and water in normal marrow results in suppression of signal intensity on the opposed-phase images [17,18]. Whereas, in pathologic fractures, normal fat-containing marrow is replaced
50
I. Izbudak et al.
Fig. 2.5 Multiple myeloma spinal metastasis with compression fracture. (a) Sagittal STIR image shows more than 50 percent compression of the T1 vertebra with slight compression on the ventral cord (arrow). Heterogenous signal increase in T1, C7, C6 vertebrae (thin arrows) and less conspicuous foci of signal increase in C2, C3 and C4. (b) Sagittal contrast-enhanced fat-saturated T1-weighted MRI better demonstrates the diffuse bone marrow metastatic foci in the vertebrae
with a tumor resulting in lack of suppression on the opposed-phase images. Zajick, et al. [18] showed a substantial decrease in signal intensity for all normal vertebrae and for benign lesions, compared to a minimal decrease or an increase in signal intensity for metastasis (Fig. 2.7). Diffusion-weighted MRI (DWI) was also recently studied for differentiating acute benign compression fractures from malignant compression fractures, based on the assumption that interstitial water increase in bone marrow edema due to benign compression fractures can be differentiated from restricted motion of water molecules in tumor cell infiltration. Baur, et al. [19] reported that DWI provides excellent distinction between malignant and benign vertebral compression fracture, but Castillo, et al. failed to demonstrate the advantage of diffusion-weighted MRI of the spine over conventional MRI [19,20]. Zhou, et al. [21] and Maeda, et al. [22] used quantitative DWI techniques to improve the specificity.
1.1.6
Cord Compression
Cord compression is the worst complication of metastatic bone disease and should be recognized early to avoid irreversible neurological damage, progressing to paraplegia. Studies have consistently shown that the malignant cord or cauda equina
2 Imaging of Spinal Tumors
51
Fig. 2.6 Osteoporotic compression fractures. T1-weighted sagittal image shows an old compression fracture in T12 vertebra (star) which has normal signal intensity revealing its chronic nature. At L2 vertebra a linear hypointense line (arrows) is seen with slight vertebral compression, suggesting a relatively acute fracture
compression is diagnosed late in the evolution of a compressive lesion and that ability to walk after treatment is directly associated with ability to walk at the time of diagnosis [23]. The most widely recognized features of cord compression (weakness, sensory loss, bowel and bladder problems) occur late in the natural history of malignant cord compression. The clinical features of early compression, according to a prospective observational study by Levack, et al. [23] is: progressive severe nerve root pain, described as sharp and precipitated by coughing or bending, irrespective of whether the pain was thoracic or lumbosacral. In their study the site of pain correlated poorly with site of compression; therefore, they recommend MRI of the whole spine for patients with known malignancy presenting with severe back or nerve root pain. A limited sagittal T2-weighted MRI of the spine would be sufficient for that purpose and would require only a short scanning time. Contrast administration is not necessary to diagnose cord or nerve root compression. The mass effect on the
52
I. Izbudak et al.
Lesion Type
Metastasis Fracture Hemangioma Schmorl Node Endplate Degeneration
−40
−20 0 20 40 60 80 100 Proportional Signal Intensity Decrease on Out-of-Phase Images Compared with In-Phase Images
Fig. 2.7 Box plot demonstrates proportional decrease in signal intensity on out-of-phase images compared with in-phase images for five lesion types, as determined at chemical shift MRI. There is statistically significant difference in between metastasis and benign lesions. (Reprinted from Zajick DC, et al. Radiology 2005; 237:590-6.)
cord may be associated with abnormal signal intensity on T2-weighted images representing acute myelopathy. DWI-DTI with fiber tracking and computation of fractional anisotropy (FA) and apparent diffusion coefficient (ADC) show promising results for early signs of cord compression in the absence of signal abnormality in T2-weighted images [24].
1.1.7
Chordoma
Chordomas are uncommon aggressive extradural lesions of the bony spine arising from remnants of the primitive notochord and representing approximately 3 percent to 5 percent of primary bone tumors [9]. Chordomas are found predominantly in the sacrococcygeal area (50 percent), with 30 percent to 40 percent arising from the basisphenoid region and the remainder from the vertebral bodies. The intervertebral disc and two or more adjacent vertebrae are commonly affected, and there is often a paraspinal soft tissue mass that may possess a calcified matrix [9]. Plain films show bony destruction with areas of amorphous calcification in a high percentage of cases. CT additionally shows paravertebral soft tissue masses, including the epidural component. MRI demonstrates total destruction of the vertebrae, initially without collapse, and spread to adjacent vertebral bodies across the disc space. T1-weighted images show isointense (75 percent) or hypointense (25 percent) lesions with increased T2 signal intensity in all cases, with the majority possessing low signal septa or calcification. The pattern of enhancement varies.
2 Imaging of Spinal Tumors
1.1.8
53
Chondrosarcoma
Chondrosarcomas are malignant cartilaginous neoplasms that rarely affect the spine, although they account for between 11 percent and 22 percent of all primary bone tumors [9, 25]. The peak incidence is between 20 and 60 years of age, peaking in the fifth and sixth decades. Seventy-five percent of chondrosarcomas are primary; the remainder result from malignant transformation of a preexisting cartilaginous lesion (enchondromatosis or osteochondroma). On plain films they cause lytic destruction and calcified matrix in the form of radiodense swirls, rings or arcs [9]. Frequently there is an associated soft tissue mass, which may be better demonstrated by CT scan. On MRI the signal intensity of chondrosarcomas is heterogenous: focal areas of decreased signal intensity on T2-weighted images due to prominent calcifications, or high signal intensity representing cartilage. Areas of hemorrhage can also occur, producing paramagnetic or susceptibility signal changes within the lesions. With gadolinium, enhancement is usually avid in a ring and arc septal pattern. Dynamic contrast-enhanced MRI has the potential for differentiating benign cartilaginous tumors from chondrosarcomas. Geirnaerdat, et al. [26] investigated the role of fast contrast-enhanced MRI in differentiating eight enchondromas and 11 osteochondromas from 18 chondrosarcomas. The sensitivity was 89 percent, specificity 84 percent, positive predictive value 84 percent, and negative predictive value 89 percent.
1.2
Benign
1.2.1
Vertebral Hemangioma
Vertebral hemangiomas are common, incidentally discovered, asymptomatic lesions in the vertebrae with a 60 percent occurence in the thoracic region and 29 percent in the lumbar region, and fewer in the cervical spine and sacrum. They are multiple in about one-third of cases [9]. On plain film they appear as vertically oriented thick trabeculae, and on axial CT these trabeculae are often surrounded by low-attenuation fat, producing a spotted appearance [9]. On T1-weighted MR images they appear mottled and of increased signal intensity, although sometimes they cannot be visible on T1-weighted images. On T2weighted images they are consistently bright. Presence of fat cells and dilated vessels with interstitial edema most likely accounts for the high signal intensity on T1- and T2-weighted images [27]. Contrast enhancement of the vertebral hemangioma is variable, depending on its appearance on T1-weighted images and the type of sequence used after contrast injection (Fig. 2.8). They usually show enhancement with gadolinium, either homogenous or peripheral. Occasionally, asymptomatic vertebral hemangioma shows low signal intensity on T1-weighted images, with marked enhancement on post-contrast T1-weighted SE images. This might be
54
I. Izbudak et al.
Fig. 2.8 Multiple typical hemangiomas. (a) Sagittal T1-weighted MR image shows round, bright lesions in T11, T12, L4, L5 and S1 vertebrae (arrows). (b) These lesions appear bright on this sagittal T2-weighted image as well (arrows)
confused with metastasis, however, CT still shows a rather specific hemangioma pattern with thickened trabeculae.
1.2.2
Giant Cell Tumor
Giant cell tumors comprise 4 percent to 5 percent of all primary bone tumors and are almost always seen in skeletally mature bone, after epiphyseal closure, usually between 20 and 40 years of age [9,25]. In Dahlin and Unni’s series they were the second most common benign spinal tumor, after vertebral hemangiomas [28]. Only 5 percent of giant cell tumors occur in the spine. Additionally, they are the most common benign neoplasm involving the sacrum [9]. Plain films show a geographically expansile lytic lesion, rarely with a sclerotic border. CT better shows bone destruction and may be useful in demonstrating soft tissue masses (Fig. 2.9). MRI reveals a destructive mass arising within the vertebral body, which demonstrates low to intermediate signal intensity on T1-weighted images and high signal intensity on T2-weighted images. Hemorrhage may be seen and fluid-fluid levels
2 Imaging of Spinal Tumors
55
Fig. 2.8 (continued) (c) Contrast-enhanced fat-saturated sagittal T1-weighted image demonstrates no enhancement in these lesions except for minimal partial enhancement of the one in T11 vertebra (arrow).
are occasionally occur but are more typical for secondary aneurysmal bone cysts arising from giant cell tumors [25].
1.2.3
Aneurysmal Bone Cyst
Aneurysmal bone cysts are expansile, blood-filled lesions that most frequently occur in adolescents (80 percent are found in persons younger than 20 years). They most commonly involve the posterior elements of the spine and frequently (up to 32 percent) arise in association with a preexisting lesion, such as a giant cell tumor, osteoblastoma, chondroblastoma or fibrous dysplasia [25]. Thirty percent of aneurysmal bone cysts involve the spine, most commonly the lumbar region, followed by the cervical spine (22 percent). Aneurysmal bone cysts may spread to the vertebral body (40 percent) or involve an adjacent vertebra or rib [25]. Plain films of the spine demonstrate an expanding, radiolucent, or lytic lesion usually involving the posterior elements with marked thinning of adjacent cortical bone [9]. CT can confirm the geographic expansion of lesions, delineate multicystic components with fluid-fluid levels, and define soft tissue extension.
56
I. Izbudak et al.
Fig. 2.9 Giant cell tumor. (a) Sagittal 2D-reformat CT image demonstrates lytic destructive mass in the sacral vertebrae. (b) Axial contrast-enhanced CT image with soft tissue window shows diffuse enhancing tumor within the sacral region. This patient was 13 years old, a very unusual age for giant cell tumor
MRI typically demonstrates numerous well-defined cystic cavities that are surrounded by a rim of low signal intensity that contains areas of low and high signal intensity on both T1- and T2-weighted images [9,25]. Lesions involving the vertebral body may be destructive and produce collapse. Fluid-fluid levels occur, but the signal characteristics of the layers are variable and may differ from cavity to cavity in the same lesion, depending on the chronicity of the hemorrhage.
2 2.1
Intradural Tumors Intradural/Intramedullary Tumors
Intramedullary spinal cord tumors account for approximately 25 percent of all spine tumors. Most (90 percent to 95 percent) of the intramedullary tumors are malignant and predominantly composed of glial components. Ependymomas are the most common type in adults and astrocytomas in children. Symptoms are usually insidious. Conventional radiography and CT often failed to reveal the true extent of intramedullary spinal neoplasms until gross expansion, erosion or scalloping of the spinal canal or scoliosis had occurred. Myelography, either with conventional
2 Imaging of Spinal Tumors
57
radiography or CT, revealed an intramedullary mass as a complete or partial block in the flow of intrathecal contrast material [29]. The advent of MRI revolutionized the characterization of intramedullary spinal cord lesions by providing detailed direct visualization of the spinal cord in multiple planes. MRI has become a key tool for the differentiation of operative tumors from non-operative lesions, such as multiple sclerosis (MS) plaques, transverse myelitis or cord infarction. Within the tumor subpopulation, MRI has allowed for detailed preoperative planning through understanding the precise limits of the tumor and surrounding edema, the presence of cysts or evidence of preexisting hemorrhage. MRI is the current imaging modality of choice in the evaluation of spinal cord masses [30].
2.1.1
MRI Technique of the Spinal Cord
Baseline MRI of the spinal cord should include thin slice T1, T1 with gadolinium and fast spin echo T2 sequences obtained in the axial and sagittal planes. T1-weighted sequences are ideal for identifying regions of cord enlargement or other anomalies in the contour of the cord, and provide a baseline for comparison with the post-gadolinium images. The post-gadolinium images allow for the differentiation of tumor from cyst and, in many cases, demonstrate the viable tumor boundary [30]. There are three important MRI features: cord expansion, enhancement following contrast administration and cyst formation [31, 32]. Contrast-enhanced images are important to define the extent of disease and are particularly useful in distinguishing associated “benign” cysts and syrinx from neoplastic involvement. Rostral or caudal cysts are commonly associated with intramedullary neoplasms, are considered to be reactive and do not enhance. Fluid attenuated inversion recovery (FLAIR) and STIR imaging have been compared to spin echo T2-weighted images mostly for MS of the spinal cord and have revealed controversial and superior lesion detection, respectively. Imaging of the spinal cord using MR spectroscopy (1H-MRS), diffusion-weighted and tensor imaging (DWI-DTI), magnetization transfer (MT) pulse sequence and functional MRI has lagged behind their use in brain imaging, mainly due to strong magnetic field inhomogeneities present in the spinal cord region, respiratory and cardiac movements, and the small size of the spinal cord. These techniques have been previously studied mostly in non-invasively treated diseases such as demyelinating, ischemic and traumatic lesions of the spinal cord [33, 34, 35]. Ducreux, et al. [36] reported DTI imaging of five spinal cord astrocytomas revealing difficulties with fiber tracking due to the presence of surrounding edema and cyst formations.
2.1.2
Ependymoma
Ependymomas represent approximately 60 percent of all glial-based tumors of the spinal cord and filum terminale. Spinal ependymomas are slow-growing tumors that tend to manifest in young adulthood. Most of the spinal cord ependymomas are multi-segment lesions occurring most commonly in the cervical region with or without involvement of the thoracic cord. Only 6.5 percent involve either the distal thoracic cord or the
58
I. Izbudak et al.
conus medullaris (Fig. 2.10) [29, 37]. Cyst formation and hemorrhage is common, especially at the tumor margins. Hemorrhage and calcification are more common than in astrocytomas. The myxopapillary subtype tends to occur in the filum terminale as a lesion in the cauda equina, and they have a greater tendency to bleed (Fig. 2.11).
Fig. 2.10 Intramedullary cervical ependymoma. (a) Sagittal T1-weighted MR image shows an expanded spinal cord from C5 through T1, with an isointense oval lesion at the level of C5-6 (arrow). (b) Sagittal T2-weighted MR image confirms syringohydromelia in the cranial and caudal aspect (arrows) of this isointense mass. (c) Contrast-enhanced sagittal T1-weighted MR image shows homogeneous contrast enhancement in this well circumscribed mass lesion (arrow)
Fig. 2.11 (continued)
2 Imaging of Spinal Tumors
59
Fig. 2.11 Myxopapillary ependymoma of the filum terminale. (a) Sagittal T1-weighted MR image shows isointense oval mass extending from the conus medullaris to L2 (arrow). (b) Sagittal T2weighted MR image reveals hypo-isointense signal to the cord in the caudal aspect of this lesion suggestive of solid component (arrow) and hyperintensity in the caudal aspect suggestive of cyst formation (arrow head). (c) Contrast-enhanced sagittal T1-weighted MR image demonstrates intense enhancement of the solid portion (arrow). Lack of contrast enhancement in the periphery of the cyst suggests that this is a polar cyst
Intramedullary ependymomas tend to be centrally located with sharp margins. Homogenous or heterogenous enhancement is noted after contrast enhancement. A cyst with an enhancing nodule may be evident. Syringohydromyelia may be associated especially with cervical ependymomas. There is an association between neurofibromatosis type 2 (NF-2). Complete surgical resection usually results in cure.
2.1.3
Astrocytoma
About one-third of all spinal cord gliomas are astrocytomas. More than half of all astrocytomas are seen in the thoracic region, usually in the upper thoracic cord. They usually involve multiple segments and are low-grade. There is a greater incidence of spinal cord astrocytomas in patients with NF-1 [38]. Astrocytomas are typically eccentric within the posterior spinal cord and are more infiltrative than ependymomas. Tumoral cysts are often eccentric. MRI of astrocytomas may be indistinguishable from ependymomas. Full diameter, ill-defined, diffuse and fusiform enlargement
60
I. Izbudak et al.
Fig. 2.12 Intramedullary cervical astrocytoma. (a) Sagittal T2-weighted MR image shows slightly hyperintense expansion of the cervical spinal cord with cysts at the level of odontoid process of C2 and C7- T2 (arrows). (b) Contrast-enhanced sagittal T1-weighted MR image reveals a multisegment heterogeneous enhancing intramedullary lesion with indistinct borders showing the infiltrative nature of this tumor. Note that the cyst in the caudal aspect of this lesion shows peripheral enhancement revealing a tumoral cyst (arrow), whereas the cysts in the inferior aspect are polar cysts without peripheral contrast enhancement
of the cord may be seen (Fig. 2.12). Although low-grade, nearly all astrocytomas enhance after contrast administration with a uniform or heterogeneous enhancement pattern. The actual tumor margins may extend beyond the enhancing margins. Total resection is not necessarily a goal in treating these tumors, and radiation therapy is administered in an attempt to eradicate residual disease [39]. Gangliogliomas are a rare subtype of spinal cord astrocytomas characterized by slow growth and a good prognosis, however there is a tendency for local reoccurrence and metastasis. They are more common in children than in adults. Scoliosis and bony remodeling are more common than with astrocytomas or ependymomas. Other differentiating features include holocord and long cord segment involvement, mixed signal on T1W image, lack of edema and absence of hemosiderin [40].
2.1.4
Hemangioblastoma
Hemangioblastomas constitute 1.0 percent to 7.2 percent of all spinal cord tumors. Although most of these tumors (75 percent) are intramedullary, they may involve intradural space or even extradural. [41]. Multiple lesions indicate the manifestation of von Hippel Lindau syndrome. These tumors are usually discrete, small and are
2 Imaging of Spinal Tumors
61
very vascular. The typical MR appearance is that of a large intramedullary cyst with a mural nodule. The spinal cord is often diffusely enlarged, out of proportion to the solid component of the tumor nidus. They have the largest syrinx formation compared with the other intramedullary tumors. Angiography and preoperative embolization may be useful in selected cases to limit intraoperative blood loss [42].
2.1.5
Lymphoma
Although most cases of spinal lymphoma involve the epidural compartment and bony vertebra, lymphoma may also be confined to the spinal cord. Spinal cord involvement is usually metastatic. Lymphoma demonstrates solid enhancement with adjacent high T2 signal intensity consistent with edema. Cord enlargement is not as severe as with other intramedullary neoplasms [29].
2.2
Intradural / Extramedullary Tumors
Extramedullary spinal cord tumors account for more than 70 percent of intradural spinal cord tumors in adults. The most common primary tumors are derived from sheath cells covering the spinal-nerve roots (schwannomas and neurofibromas), or meningial cells located along the spinal cord surface (meningiomas). Other tumor types, such as hemangiopericytomas, lipomas, paragangliomas, epidermoid cysts and dermoid cysts, are less common [43].
2.2.1
Nerve Sheath Tumors
Nerve sheath tumors (NFT) comprise schwannomas and neurofibromas and represent 30 percent intradural/extramedullary tumors. The peak incidence is in the fourth and fifth decade. They may be indistinguishable by imaging standards. Multiple nerve sheath tumors in the spine frequently are associated with NF-1 and NF-2. 70 percent are intradural/extramedullary; 15 percent are extradural or both intradural extramedullary in a dumbbell configuration, whereas less than 1 percent are intramedullary. NFTs are most frequent in the thoracic spine (40 percent) with an equal incidence in the cervical and lumbar spine [44]. Neurofibromas and schwannomas are heterogeneously hyperintense on T2-weighted images, and enhance heterogeneously.
2.2.2
Meningiomas
Meningiomas are slow-growing tumors comprising approximately 25 percent of primary intraspinal tumors. There is a female predominance of 4 to 10:1. They usually
62
I. Izbudak et al.
present after the fourth decade. Most (90 percent) of the meningiomas are intradural extramedullary, whereas only 5 percent are extradural [38]. A total of 5 percent are intradural extramedullary and extradural in a dumbbell fashion. The most common location is the thoracic region (80 percent), and approximately two-thirds of thoracic meningiomas occur in the dorsal spinal canal. The vast majority are solitary and benign. Foraminal widening and pedicular erosion may occasionally be present. The tumor may be calcified. Meningiomas are usually discrete lesions, with a broad dural base and typically isointense to spinal cord on T1W images and T2W images, and enhance homogenously. NF-2 is associated with multiple meningiomas [45].
2.2.3
Metastasis
Intradural/extramedullary leptomeningeal carcinomatosis is rare and intramedullary metastasis is even rarer. Primary neuroectodermal tumor (PNET) is the most common primary central nervous system malignancy; systemic cancers that produce leptomeningeal metastases are breast (36 percent), lymphoma (28 percent), lung (16 percent), and melanoma (10 percent) [46]. Contrast-enhanced MRI is excellent for evaluation of leptomeningeal carcinomatosis, and it may be thick and linear or nodular.
3
Vertebroplasty
The percutaneous injection of bone cement into a vertebral body is called vertebroplasty and was first performed in 1987 in France [47]. It is a simple procedure with profound patient benefits [48,49] and a success rate approaching 80 percent. The procedure is performed to treat pain from benign osteoporotic vertebral fractures, primary malignant disease such as multiple myeloma or lymphoma, metastatic disease and benign tumors of the bone, particularly hemangioma. Conventional therapy for malignant disease consists of bed rest, bracing and anti-inflammatory or commonly opiate medications, and radiation therapy. The surgical option is corpectomy cage placement and stabilization above and below the fracture with pedicle screws, and internal fixation. These surgical procedures require significant post procedural recovery and have associated morbidity and mortality in patients who often have limited life expectancies. By comparison vertebroplasty leads to a durable partial or complete pain reduction in 80 percent to 90 percent of patients [50,51,52]. Pain relief is usually observed within the first 72 hours after treatment [51,53]. This procedure involves the injection of bone cement into a vertebral body via a percutaneous route under X-ray guidance. The cement is injected as a semi-liquid substance via a needle that is 11- to 14-gauge in size that has been passed into the vertebral body via a transpedicular, or less often, a paraspinous approach (Fig. 2.13).
2 Imaging of Spinal Tumors
63
Fig. 2.13 Vertebroplasty for L4 lesion. (a) Lateral spot image shows an 11-gauge needle approaching percutaneously through pedicle into the vertebral body. (b) The opacified cement is seen in the vertebral body after injection of the cement and removal of the needle
For treatments involving S1 levels, most of the time a trans-illiac bone access is performed to treat the center of the vertebral body. In a case in which previous surgical stabilization with transpedicular surgical fixation screws has been done, a lateral approach is chosen to perform complementary vertebroplasty. An alternative in the future will be the use of a curved needle from the level above through the disc space into the adjacent vertebral body. Destruction of the posterior vertebral wall, with or without compression of the spinal canal, complete loss of vertebral body height and the presence of osteoblastic metastatic lesions are considered relative contraindications [54, 55]. A critical issue is the relationship between the degree of vertebral body filling and the likelihood of achieving pain relief in patients with malignant disease. It has been clearly shown that pain relief is not related to the amount of cement injected. Instead, pain relief is related to the distribution of cement in the vertebral body and, more particularly, its distribution in fracture plains. The complication rate, though low, is definitely related to the amount of cement injected.
Conclusion In summary, modern imaging techniques, particularly MRI, provide invaluable information on imaging of spinal tumors. Acknowledgments We would like to thank Dr. Nafi Aygun for the images he provided.
64
I. Izbudak et al.
References 1. Algra PR, Bloem JL, Tissing H, et al. Detection of vertebral metastases: comprasion between MRI and bone scintigraphy. Radiographics 1991; 11:219-32. 2. Mundy Gr, Yoneda T. Facilitation and suppression of bone metastasis. Clin Orthop 1995;312:34-44. 3. Clohisy DR, perkins SL, Ramnaraine ML. Review of cellular mechanisms of tumor osteolysis. Clin Orthop 2000;373:104-14. 4. Papagelopoulos PJ, Savvidou OD, Galanis EC, et al. Advances and challenges in diagnosis and management of skeletal metastases. Orthopedics 2006;29[7];608-20. 5. Padhani A, Husband J. Bone metastases. In: Husband JES, Reznek RH, eds. Imaging in Oncology Oxford, U.K.: Isis Medical Media Ltd.; 1998:765–787 6. Eustace S, Tello R, DeCarvalho V, et al. A comparison of whole-body TurboSTIR MRI and Planar 99 mTc-methylene Diphosphonate scintigraphy in the examination of patients with suspected skeletal metastases. AJR 1997; 169:1655-61. 7. Walker R, Kessar P, Blanchard R, et al. Turbo STIR magnetic resonance imaging as a wholebody screening tool for metastases in patients with breast carcinoma: preliminary clinical experience. J Magn Reson Imaging. 2000 Apr;11(4):343-50. 8. Frat A, Agildere M, Gencoglu A, et al. Value of whole-body turbo short tau inversion recovery magnetic resonance imaging with panoramic table for detecting bone metastases: comparison with 99 MTcmethylene diphosphonate scintigraphy. J Comput Assist Tomogr. 2006 Jan-Feb;30(1):151-6. 9. Masaryk TJ. Spinal Tumors. In: Modic MT, Masaryk TJ, Ross JS, eds. Magnetic Resonance Imaging of the Spine. Ed. 2. St. Louis, Missouri. Mosby-Year Book Inc. 1994; 249-314. 10. Vande Berg BC, Lecouvet FE, Michaux L et al. Magnetic resonance imaging of the bone marrow in hematological malignancies. Eur Radiol 1998;8:1335-44. 11. Jantunen E, Laakso M. Biophosphonates in multiple myeloma:current status; future perspectives. Br J Haematol 1996;93:501-6. 12. Mahnken AH, Wildberger JE, Gehbauer G. et al. Multidetector CT of the spine in multiple myeloma: Comparison with MRI and radiography. Am J Roentgenol 2002;178:1429-36. 13. Lecouvet FE, Vande Berg BC, Maldague BE, et al. Vertebral compression fractures in multiple myeloma: Distribution and appearance at MRI. Radiology 1997;204:195-9. 14. Baker LL, Goodman SB, Perkash I, Lane B, Enzmann DR. Benign versus pathologic compression fractures of vertebral bodies: assessment with conventional spin-echo, chemical shift, and STIR MRI. Radiology 1990;174:495-502. 15. An HS, Andreshak TG, Nguyen C, Williams A, Daniels D. Can we distinguish between benign versus malignant compression fractures of the spine by magnetic resonance imaging? Spine 1995; 20:1776-82. 16. Jung H, Jee W, McCauley TR, Ha K, Choi K. Discrimination of metastatic from acute osteoporotic compression spinal fractures with MRI. Radiographics 2003;23:179-87. 17. Erly WK, Oh ES, Outwater EK. The utility of in-phase/opposed-phase imaging in differentiating malignancy from acute benign compression fractures of the spine. Am J neuroradiol 2006; 27;1183-88. 18. Zajick DC, Morrison WB, Schweitzer ME, Parellada JA, Carrino JA. Benign and Malignant Processes: Normal values and differentiation with chemical shift MRI in vertebral marrow. Radiology 2005;237:590-6. 19. Baur A, Huber A, Ertl-Wagner B, et al. Diagnostic value of increased diffusion weighting of a steady-state free precession sequence for differentiating acute benign osteoporotic fractures from pathologic vertebral compression fractures. AJNR Am J Neuroradiol 2001;22:366-72. 20. Castillo M, Arbelaez A, Smith JK, Fisher LL. Diffusion-weighted MRI offers no advantage over routine noncontrast MRI in the detection of vertebral metastasis. AJNR Am J Neuroradiol 2002;21:948-53.
2 Imaging of Spinal Tumors
65
21. Zhou XJ, Leeds NE, McKinnon GC, Kumar AJ. Characterization of benign and metastatic compression fractures with quantitative diffusion MRI. AJNR Am J Neuroradiol 2002;23:165-70. 22. Maeda M, Sakuma H, Maier SE, Takeda K. Quantitative assessment of diffusion abnormalities in benign and malignant vertebral compression fractures by line scan diffusion-weighted imaging. AJR 2003; 181:1203-9. 23. Levack P, Graham J, Collie D, et al. Don’t wait for a sensory level-Listen to the symptoms: a prospective audit of the delays in diagnosis of malignant cord compression. Clinical Oncology 2002;14:472-480. 24. Facon D, Ozanne A, Fillard P, et al. AJNR AM J Neuroradiol 2005 26:1587-94. 25. Keogh C, Bergin D, Brennan D, and Eustace S. MRI of bone tumors of the cervical spine. MR Clinics of North America 2000;8(3):513-27. 26. Geirnaerdt MJA, Hogendoorn PCW, Bloem JL, Taminiau AHM, and van der Woude HJ. Cartilaginous Tumors: Fast Contrast-enhanced MRI. Radiology. 2000;214:539-46. 27. Baudrez V, Gallant C, Lecouvet FE, et al. Vertebral hemangioma: MR-histological correlation in autopsy specimens. Radiology 1999;213(P):245. 28. Dahlin DC, Unni KK. Bone tumors; general aspects and data on 8,542 cases. Springfield III, Thomas, 1986, 62-9. 29. Koeller KK, Rosenblum RS, Morrison AL. Neoplasms of the spinal cord and filum terminale: radiologic-pathologic correlation. RadioGraphics 2000;20:1721–1749. 30. Waldron JS, Cha S. Radiographic Features of Intramedullary Spinal Cord Tumors. Neurosurgery Clinics of North America 2006;17:13-19. 31. Takemoto K, Matsumura Y, Hashimoto H, et al. MRI of intraspinal tumors: capability in histological differentiation and compartmentalization of extramedullary tumors. Neuroradiology 1988;30:303-309. 32. Marliani AF, Clementi V, Albini-Riccioli L. Quantitative proton magnetic resonance spectroscopy of the human cervical spinal cord at 3 Tesla. Magn Reson Med 2006;57:160-163. 33. Dillon WP, Norman D, Newton TH, et al. Intradural spinal cord lesions: Gd-DTPA enhanced MRI. Radiology 1989;170:229-237. 34. Rossi C, Boss A, Linding TM et al. Diffusion tensor imaging of the spinal cord at 1.5 and 3.0 tesla. Rofo. 2007 Mar; 179(3):219-24. 35. Kendi AT, Tan FU, Kendi M, et al. MR spectroscopy of cervical spinal cord in patients with multiple sclerosis. Neuroradiology. 2004 Sep; 46(9) 764-9. 36. Ducreux D, Lepeintre JF, Fillard P, et al. MR diffusion tensor imaging and fiber tracking in 5 spinal cord astrocytomas. AJNR Am J Neuroradiol. 2006 Jan; 27(1):214-6. 37. Ferrante L, Mastronardi L, Celli P, Lunardi P, Acqui M, Fortuna A. Intramedullary spinal cord ependymomas: a study of 45 cases with long-term follow-up. Acta Neurochir 1992;119:74-79. 38. Dickman CA, Fehlings MG, Gokaslan ZL. In: Dickman CA, Fehlings MG, Gokaslan ZL, eds. Spinal Cord and Spinal Column Tumors: Principles and Practice. New York, NY: Thieme; 2006:145-176. 39. Cooper P. Outcome after operative treatment of intramedullary spinal cord tumors in adults: intermediate and long-term results in 51 patients. Neurosurgery 1989;25:855-859. 40. Patel U, Pinto RS, Miller DC, et al. MR of spinal cord ganglioglioma. AJNR Am J Neuroradiol. 1998;19:879-887. 41. Osborn AG. Tumors, cysts, and tumorlike lesions of the spine and spinal cord. In: Osborn A, eds. Diagnostic neuroradiology. St Louis, Mo: MosbyYear Book, 1994;895-916. 42. Bloomer CW, Ackerman A, Bhatia RG. Imaging for spine tumors and new applications.Top Magn Reson Imaging. 2006 Apr; 17(2):69-87. 43. Traul DE, Shaffrey ME, Schiff D. Part I: Spinal-cord neoplasms-intradural neoplasms. Lancet Oncol. 2007 Jan;8(1):35-45. 44. Levy WJ, Latchaw J, Hahn JF, et al. Spinal neurofibromas: a report of 66 cases and a comparison with meningiomas. Neurosurgery. 1986;18:331-334. 45. DeVerdelhan O, Haegelen C, Carsin-Nicol B, et al. MRI features of spinal schwannomas and meningiomas. J Neuroradiol. 2005;32:42-49.
66
I. Izbudak et al.
46. Perrin RG, Livingston KE, Aarabi B. Intradural extramedullary spinal metastasis. A report of 10 cases. J Neurosurg. 1982;56:835-837. 47. Galibert P, Deramond H. Percutaneous acrylic vertebroplasty as a treatment of vertebral angioma as well as painful and debilitating diseases. Chirurgie 1990;116:326-34 48. Cotten A., Boutry N., Cortet B., Assaker R., Demondion X., Leblond D., Chastanet P., Duquesnoy B., Deramond H. Percutaneous vertebroplasty: state of the art. Radiographics 1998;18:311-20. 49. Dufresne AC, Brunet E, Sola-Martinez MT, Rose M, Chiras J. Percutaneous vertebroplasty of the cervico-thoracic junction using an anterior route. Technique and results. Report of nine cases. J Neuroradiol 1998 Jul; 25(2):123-8. 50. Weill A, Chiras J, Simon JM, Rose M, Sola-Martinez T, Enkaoua E. Spinal metastases: indications for and results of percutaneous injection of acrylic surgical cement. Radiology 1996; 199:241-7. 51. Barr M., Barr J. Invited commentary. RadioGraphics 1998;18:320-322. 52. Cotten A, Deramond H, Cortet B, Lejeune JP, Leclerc X, Chastanet P, Clarisse J Preoperative percutaneous injection of methyl methacrylate and N-butyl cyanoacrylate intervertebral hemangiomas. AJNR Am J Neuroradiol 1996 Jan;17(1):137-42. 53. Jensen ME, Evans AJ, Mathis JM, Kallmes DF, Cloft HJ, Dion JE. Percutaneous polymethylmethacrylate vertebroplasty in the treatment of osteoporotic vertebral body compression fractures: technical aspects. AJNR Am J Neuroradiol 1997 Nov-Dec18 (10):1897-904. 54. Cortet B, Cotten A, Boutry N, Dewatre F, Flipo RM, Duquesnoy B, Chastanet P, Delcambre B. Percutaneous vertebroplasty in patients with osteolytic metastases or multiple myeloma. Rev Rhum Engl Ed 1997 Mar;64(3):177-83. 55. Martin JB, Sugiu K, San Millian R, Murphy K J, Piotin M, Rufenacht DA. Vertebroplasty: clinical experience and follow-up results. Bone Vol. 25, No. 2, Supplement, 11S-15S, 1999.
3
PET Imaging of Brain Tumors Alan J. Fischman MD, PhD
1
Introduction
The incidence of primary brain tumors is ∼11:100,000 of the population. In the year 2006, ∼18,820 new cases of brain and other nervous system tumors were diagnosed in the United States [1] and these tumors were the cause of death in ∼12,820 patients. Despite advances in diagnosis and therapy, the prognosis for patients with primary brain tumors remains very poor; age-adjusted five-year survival is 30.8 percent. Primary brain tumor is the most prevalent solid tumor in children, and patients 19 years old or younger have a five-year survival of 65 percent. Patients aged 44 or younger have a five-year survival of 58.7 percent. In the elderly, prognosis is extremely poor with a five-year survival of less than 6.5 percent in patients aged 65 and older [1]. The epidemiology of primary brain tumors is extremely complex and includes lesions with both benign and malignant histologies. Between 1985 and 1992, over 60,000 patients diagnosed with primary brain tumors were reported to the National Cancer Data Base (NCDB) [2], and in this group the most frequent tumors were glioblastoma multiforme (GBM) and astrocytoma. The World Health Organization (WHO) has established a four-level classification system (Grade I to IV) with Grade I being most benign, and Grade IV most malignant. The most malignant tumors – astrocytomas and GBM (Grade III to IV) – had overall 30 percent and two percent five-year survival in the NCDB series [2].
2
Basics of Positron Emission Tomography
Positron emission tomography (PET) obtains in vivo regional biochemical and physiologic information about healthy and diseased human brain tissue while the patient is comfortable, conscious and alert. This capability is the result of four
Director of Nuclear Medicine, Massachusetts General Hospital, Professor, Harvard Medical School
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
67
68
A.J. Fischman
major technological developments: 1) “user friendly” cyclotrons for producing positron emitting isotopes, 2) techniques for the rapid synthesis of radiopharaceuticals necessary for biochemical and physiologic studies, 3) mathematical models and practical algorithms to obtain critical information from the data, and 4) the PET instrumentation to safely detect the radiopharmaceuticals in vivo in a regional and quantitative manner. With PET, molecular and physiological processes involved in the working of healthy or diseased human brains can be studied at a level of anatomic and quantitative detail that was previously impossible. Before the introduction of the PET scanner, we could only infer what went on within the brain from postmortems or animal studies. Data obtained from PET studies have a greater level of quantitative reliability than the results that can be obtained with any other imaging modality. In this context, the concentration of a radiopharmaceutical in a volume of tissue measured with a PET camera is identical to that of a sample of excised tissue measured with a well counter. Compared with all other Nuclear Medicine procedures, PET has higher resolution, sensitivity and quantitative fidelity. Since the radionuclides used in PET can be incorporated into almost any drug or natural biological molecule, the number of potential PET radiopharmaceuticals is unlimited. This gives PET the capability of providing quantitative images of a variety of physiological and biochemical processes, including: blood flow, blood volume, blood-brain barrier permeability, oxygen utilization, glucose utilization, amino acid transport, protein synthesis, cell proliferation and tissue hypoxia. A partial list of PET tracers that have been used for these purposes is presented in Table 3.1.
3
General Consideration for PET Imaging of the Brain
Since the brain is immobilized in the cranium, it is the ideal organ for most methods of imaging (PET, SPECT, CT, MRI). With the addition of external immobilization devices and motion correction algorithms, the brain is the only organ of the body for which the high resolution of modern PET cameras (< 5 mm) can be realized. In addition, cross modality co-registration (e.g.,PET with MRI or CT), is simpler and more accurate than in any other area of the body. The introduction and proliferation of hybrid PET-CT devices has had major impact on PET imaging for most applications in oncology. With these cameras imaging time is reduced by using the CT for attenuation correction, and the CT data provides an anatomic reference for interpreting the PET images. In contrast, PET-CT offers almost no advantage for imaging brain tumors [3]; the imaging time is already short and the optimal procedure for correlating functional data with anatomy is co-registration with MRI.
15
15
O-Water, C O2 O-Oxygen 15 O (11C)-Carbon monoxide 11 C-Aminoisobuteric acid, 82Rb-Chloride, 68Ga-EDTA 2-18F-Fluoro-2-deoxy-D-glucose (FDG) 11 C-Methionine, Leucine, Tyrosine, 3,4 dihydroxy-618 F-fluoro-L-phenylalanine (FDOPA), 18F-α-Methyl tyrosine 18 F-Fluoromisonidazole, 18F Fluoroazomycin Arabinoside (FAZA) 11 C- Thymidine, 2′deoxy-3′-18F-Fluorothymidine (FLT), 18 F-9-[4-fluoro-3-(hydroxymethyl) butyl]guanine (FHBG), 124 I- 2′-fluoro-2′-deoxy-1-ß-D-arabinofuranosyl-5-iodouracil (FIAU)
15
Table 3.1 PET Tracers for Studying Brain Tumors Radiopharmaceutical Perfusion Oxygen utilization Blood volume Blood-brain barrier permeability Glucose utilization Amino acid transport and protein synthesis Hypoxia Cell proliferation Monitoring of gene therapy
Parameter Evaluated
66, 67 15 76, 77
4, 5 6 6 74 75 7, 11, 13, 14
References
3 PET Imaging of Brain Tumors 69
70
A.J. Fischman
4 Positron Emission Tomography (PET) Radiopharmaceuticals for Imaging Primary Brain Tumors 2-[18F]-fluoro-2-deoxy-D-glucose (FDG) is currently the most commonly used radiopharmaceutical for imaging brain tumors. This is due both to the wide availability of this tracer and the intimate relationship between glucose metabolism and malignancy. Other tracers labeled with 15O, 11C and 18F have also been employed for imaging brain tumors. Tracers labeled with 15O include: bolus injected H2150 or inhaled C1502 for studying tumor perfusion, 15O for studying oxygenation and inhaled C150 for studying tumor blood volume [4,5,6]. Tracers labeled with 11C, include methionine, leucine and tyrosine for studying amino acid transport and protein synthesis in tumors [7]. Imaging with 11C methionine has been shown to be of great value for imaging low-grade gliomas [8,9,10]. Tritiated thymidine is the “gold standard” for studying cell proliferation in vitro and this pyrimidine nucleoside, when labeled with 11C, is an excellent tracer for studying cell proliferation in brain tumors by PET [11]. 11C labeled methionine and thymidine appear to have better specificity for tumor proliferation, compared with inflammation, and may therefore have advantages over FDG for differentiating recurrent tumors from radiation necrosis [12]. Unfortunately, due to the requirement of an on-site cyclotron and radiochemistry facilities and the short physical half-lives of 15O (2 minutes) and 11C (20 minutes), tracers labeled with these radionuclides are not practical for routine clinical application. In recent years additional 18F labeled radiopharmaceuticals have been developed and applied for imaging brain tumors. These tracers include: 3,4dihydroxy-6-18F-fluoro-l-phenylalanine (FDOPA, [13]) and O-(2-[18F] fluoroethyl)L-tyrosine (FET, [14]) for studying amino acid transport and protein synthesis, and 3′-Deoxy-3′-18F-fluorothymidine (FLT, [15]) for studying cell proliferation. Although these agents are not currently available though regional radiopharmacies on a routine basis, this problem is being addressed by several venders of PET radiopharmaceuticals.
5
FDG PET for the Evaluation of Brain Tumors
FDG is a glucose analogue that is transported and phosphorylated in normal tissues and tumors by the same transporters and enzymes as glucose. However, after phosphorylation to FDG-6-PO4 at the first step in glucose utilization, the lack of an OH group at the 2-position of FDG prevents further metabolism. Thus, FDG acts as a trapped tracer that provides a snapshot of glucose utilization at the time of injection. For FDG-PET studies patients are required to have no caloric intake for at least four hours prior to injection. Following IV injection of the tracer (∼ 5 mCi for 3-D acquisitions) the patients rest quietly in a dimly lit room for ∼45 minutes during tracer uptake. During this “uptake period” visual and auditory stimuli are minimized
3 PET Imaging of Brain Tumors
71
to avoid cortical activation that could confound image interpretation. In general, blindfolds and ear plugs are not employed. Images are acquired in a single bed position (usually ∼ 15 cm) for six to eight minutes, and reconstructed by filtered back-projection or iterative algorithms (OSEM). Attenuation correction of the images is performed using transmission images, analytical algorithms or CT data. The images are formatted into trans-axial, sagittal and coronal projections. In many cases PET images are co-registered with MRI data (by computer methods such as the mutual information algorithm [16]) for precise correlation of functional and anatomic information. In addition, in our laboratory, the PET images are transformed to the Talarack coordinate system, co-registered with PET images of age-matched cohorts of normal subjects, and the difference in FDG uptake between the patients and controls are presented as Z-scores. The regional metabolic data that is provided by FDG-PET in brain tumor patients provides important information about tumor grade, prognosis and recurrence [17]. However, in contrast to other tissues, glucose is almost the exclusive energy substrate for brain metabolism. Thus, normal brain tissue has very high background accumulation of FDG, particularly in gray matter structures. Although malignant tumors in the brain avidly accumulate FDG, the level of FDG utilization in even high-grade tumors is similar to (and in some cases less than) normal gray matter structures. The physiological accumulation of FDG in normal structures reduces the conspicuity of lesions and limits the ability of FDG-PET to detect and characterize small lesions. In addition, the fact that areas of inflammation and metastatic disease frequently occur at the interface of gray and white matter further complicate analysis of PET data. Clearly, image interpretation can be greatly facilitated by co-registration with MRI and/or databases of FDG uptake in normal brains.
6
FDG-PET in the Initial Diagnosis of Brain Tumors
In the initial diagnosis of brain tumors FDG-PET has been extremely useful for assessing tumor grade, identifying optimal sites for biopsy and assessment of prognosis (Table 3.2).
Table 3.2 Indications for FDG-PET in Brain Tumor Patients Initial Evaluation of Tumors Determination of grade/degree of malignancy Determination of an optimal site (s) for stereotactic biopsy Assessment of prognosis Post-therapy Re-evaluation Detection of recurrent tumor Detection of residual tumor after surgery Monitoring of tumor progression Grading malignancy Differentiation between recurrent tumor and necrosis
72
7
A.J. Fischman
Assessment of Tumor Grade
In early studies by Patronas, et al. [18] it was demonstrated that the rate of glycolysis in brain tumors as measured by FDG-PET is a more accurate index of tumor grade than contrast enhancement. Accumulation of FDG in normal gray and white matter structures provide a convenient reference for comparisons with uptake in tumors. Using this qualitative method of image analysis, low-grade (Grade I and II) have levels of FDG uptake that are less than or similar to normal white matter, whereas high-grade (Grade III and IV) have levels of FDG accumulation that approach or exceed gray matter. These findings are consistent with the results of semi-quantitative image analysis. In a study by Delbeke, et al. [19] a series of patients with histologically proven high- (n=32) and low- (n=26) grade brain tumors were evaluated in terms of tumor-towhite-matter (T/WM) and tumor-to-gray-matter (T/GM) ratios to determine cut-off values for differentiating between low- (Grade I and II) and high-grade (Grade III and IV) tumors. These investigators demonstrated that T/WM ratios of more than 1.5 and T/GM ratios more than 0.6 established the diagnosis of high-grade tumors with a sensitivity of 94 percent and a specificity of 77 percent. An example of a patient with a high-grade glioma of the right thalamus is shown in Fig. 3.1A. After therapy there is reduced FDG uptake in the original site, but evidence of a new area of hypermetabolism in the contralateral thalamus (B). With further therapy this area of hypermetabolism decreased in intensity (C). Approximately one year later the tumor recurred in the right thalamus (D).
8
Identification of Optimal Sites for Stereotactic Biopsy
Primary brain tumors have poorly defined borders and frequently there is significant variation in grade within a tumor’s volume. In a study by Paulus and Peiffer, histological features of multiple samples from 50 brain tumors were studied [20]. These investigators found varying grades in 82 percent of the tumors with 62 percent of the tumors containing both high- and low-grade features. These findings illustrated the potential for sampling error and explained the growth pattern of the tumors, e.g., irregular contours of tumors may be due to faster growth of high-grade components. In addition, high-grade tumors such as GBM often arise as focal areas of malignant degeneration in lower grade lesions. Since these changes are not easily detected by anatomic imaging, evaluation of these lesions by CT- or MRI-guided stereotactic biopsy can have significant sampling error, which may result in understaging. By identifying foci of maximal hypermetabolism in these heterogeneous lesions, FDG-PET can guide stereotactic biopsy to sample the highest grade areas of the tumor [21, 22, 23, 24]. Fig. 3.2, shows FDG-PET images of a patient with a recurrent tumor in the region of the right caudate nucleus. Previously, the patient had surgical resection of
3 PET Imaging of Brain Tumors
73
Fig. 3.1 Serial FDG-PET studies of a patient with thalamic glioma. (A) The initial study at the time of diagnosis demonstrates intense hypermetabolism in the right thalamus. (B) After therapy there is reduced FDG uptake in the original site, but evidence of a new area of hypermetabolism in the contralateral thalamus. (C) With further therapy this area of hypermetabolism decreased in intensity. (D) Approximately one year later the tumor recurred in the right thalamus
Fig. 3.2 Co-registered FDG-PET (A) and MRI (B) images of a patient with recurrent tumor in the region of the right caudate nucleus. Previously, the patient had surgical resection of an anaplastic oligodendroblastoma in this region and presented with heterogeneous enhancement on MRI. The PET study clearly demonstrates areas of FDG accumulation that are greater than normal gray matter; consistent with high-grade tumor recurrence
74
A.J. Fischman
Fig. 3.3 Co-registered FDG-PET (A) and MRI (B) images of a low-grade astrocytoma. This tumor was enhancing on MRI and the co-registered PET image shows a level of FDG metabolism that is similar to normal white matter; consistent with a low-grade lesion
an anaplastic oligodendroblastoma in this region and presented with heterogeneous enhancement on MRI. The PET study clearly demonstrates areas of FDG accumulation that are greater than normal gray matter, consistent with high-grade tumor recurrence. Fig. 3.3 illustrates PET images of a low-grade astrocytoma. This tumor was enhancing on MRI and the co-registered PET image shows a level of FDG metabolism that is similar to normal white matter; consistent with a low-grade lesion.
9
Assessment of Prognosis
The level of FDG accumulation in a primary brain tumor can yield important information about prognosis. In an early study of 29 patients with untreated and treated brain tumors, Alavi, et al. [25] demonstrated that patients with hypermetabolic lesions had significantly shorter survival compared with patients that had hypometabolic tumors. In the cases with high-grade tumors, patients with hypometabolic tumors had a one-year survival of 78 percent, whereas those with hypermetabolic tumors had a one-year survival of only 29 percent. Similar results were demonstrated in a study of 45 patients with high-grade tumors by Patronas, et al. [26]. In this investigation, patients with hypermetabolic tumors had an average survival of five months, while patients with eu- or hypometabolic tumors had an average survival of 19 months. In studies of patients with low-grade tumors [27, 28], the development of focal areas of hypermetabolism has been associated with poorer prognosis. These investigations suggested that, independent of prior therapy for both low- and high-grade tumors, there is a relationship between FDG uptake and tumor aggressiveness.
3 PET Imaging of Brain Tumors
75
In another study [29] it was demonstrated that the prognostic value of FDG-PET can be enhanced by performing serial studies.
10
FDG-PET in the Post-therapy Re-evaluation of Brain Tumor Patients Detection of Recurrent Tumor
As in the initial diagnosis of primary brain tumors, FDG-PET can also be useful in the evaluation of recurrent tumors. Fig. 3.4 shows serial FDG-PET images of a patient who had previous resection of a GBM in the left temporal region and presented with a nodular area of enhancement on MRI. The first two images (A and B) show hypometabolism in this region, consistent with a good response to therapy. The third (C) and fourth (D) images clearly demonstrate FDG accumulation that is greater than normal gray matter, consistent with high-grade tumor recurrence. Fig. 3.5 shows serial FDG-PET studies of a patient with a left frontal glioma. The first three images (A, B and C) show hypometabolism at the site of the tumor, consistent with a good response to therapy. In the fourth image there is increased FDG uptake in the posterior aspect of the tumor, consistent with recurrence.
Fig. 3.4 Serial co-registered FDG-PET images of a patient who had previous resection of a GBM in the left temporal region and presented with a nodular area of enhancement on MRI. The first two images (A and B) show hypometabolism in this region; consistent with a good response to therapy. The third (C) and fourth (D) images clearly demonstrate FDG accumulation that is greater than normal gray matter; consistent with high-grade tumor recurrence
76
A.J. Fischman
Fig. 3.5 Serial co-registered FDG-PET studies of a patient with a left frontal glioma. The first three images (A, B and C) show hypometabolism at the site of the tumor; consistent with a good response to therapy. In the fourth image (D) there is increased FDG uptake in the posterior aspect of the tumor; consistent with recurrence
11
Detection of Residual Tumor after Surgery
Following tumor resection, both residual tumors and post-surgical changes can show abnormal enhancement on MRI and, thus, cannot be distinguished by this technique. In contrast, post-surgical changes do not result in increased FDG uptake [30]. Thus, when a rim of contrast enhancement is observed surrounding a resection cavity, but hypermetabolism is not detected by FDG-PET, recurrent tumors can be excluded with a relatively high degree of confidence. In contrast abnormal hypermetabolic activity after surgery suggests a recurrent high-grade tumor. In the situation of early recurrence, FDG-PET can be used to define the region of highest grade tumor for stereotactic biopsy [31].
12
Monitoring of Tumor Progression and Grading of Malignancy
High-grade tumors such as GBM often arise as focal areas of malignant degeneration in lower grade lesions, and these changes are not easily detected by anatomic imaging. In contrast, the detection of a new focal area hypermetabolism with FDG-PET
3 PET Imaging of Brain Tumors
77
is straightforward and provides important information about tumor progression and change in grade.
13
Differentiation Between Recurrent Tumor and Necrosis
Following radiation therapy, radiation necrosis is usually associated with reduced FDG accumulation in the treatment field [32]. However, in some cases, increased FDG activity after high-dose radiation therapy can be caused by tracer uptake in metabolically active macrophages that accumulate at the therapy site. Although the level of uptake is usually moderate (between white and gray matter) and relatively uniform, in some cases accumulation can be equal to or greater than gray matter and may even have nodular characteristics (Fig. 3.6). In these situations, radiation necrosis cannot be differentiated from a recurrent tumor. In an investigation by Barker, et al. [33], 55 patients with high-grade brain tumors that were treated by surgery and radiation and had enlarging areas of enhancement on MRI suggesting tumor recurrence or radiation necrosis were studied. The results of this investigation demonstrated that high FDG accumulation (equal to or greater than gray matter) was associated with poorer prognosis, compared to patients with low FDG accumulation (less than gray matter). In a study by Chao, et al. [34] it was demonstrated that FDG-PET had a sensitivity of 75 percent and a specificity of 81 percent for differentiating recurrent tumors from radiation necrosis in 47 patients with primary and metastatic brain tumors who underwent stereotactic radiosurgery. In patients with brain metastasis, co-registration of PET with MRI increased sensitivity from 65 percent to 86 percent [34].
Fig. 3.6 Co-registered MRI (A) and FDG-PET (B) images of a patient with a left temporal tumor. The PET study clearly demonstrates an area of FDG accumulation that is greater than normal gray matter; consistent with high-grade tumor recurrence. However, biopsy demonstrated post radiation necrosis
78
A.J. Fischman
Although these studies suggest that FDG-PET is useful for differentiation of tumor recurrence from radiation necrosis, other studies have been less encouraging. In a study by Ricci, et al. [35], MRI and FDG-PET images, and medical records of 84 consecutive patients with a history of a treated brain tumor were evaluated retrospectively. In all patients recurrent tumor or radiation necrosis was suggested by clinical or MRI findings. When contralateral white matter was used as the reference, the sensitivity and specificity of FDG-PET were 86 percent and 22 percent. With contralateral gray matter as the reference, sensitivity and specificity were 73 percent and 56 percent. Overall, nearly one-third of the patients would have been treated inappropriately in either scheme if the PET scan had been the sole determinant of therapy. Thus, although FDG-PET is a useful technique for differentiation tumor recurrence from radiation necrosis, sensitivity and specificity are less than ideal. In a recent review by Hustinx, et al. [36], it was suggested that only the combination of FDG with a radiolabeled amino acid analogue (MET or one of the more recently developed 18F labeled tracers) can provide a comprehensive characterization of suspected brain tumor recurrence.
14
Metastatic Brain Lesions
As in the rest of the body, tumor metastasis to the brain usually has high levels of glucose utilization and is detectable by FDG-PET. When the level of FDG accumulation in these lesions is greater than normal gray matter, or when they are located in white matter, CNS metastasis from peripheral tumors are easily identified by FDG-PET. In contrast, due to the high level of glucose utilization in normal brain tissue, lesions that are only mildly or moderately hypermetabolic are much less conspicuous. In addition, since brain metastases result from hematogenous seeding and frequently occur at the cortical gray-white junction, small lesions in these regions can be difficult, if not impossible, to detect by FDG-PET. Moreover, the edema that frequently surrounds metastatic lesions has low FDG accumulation and partial-volume averaging can reduce detectability. In many situations, lesion detectability can be greatly augmented by co-registration of FDG-PET images with MRI. In a study by Larcos and Maisey [37] whole-body FDG-PET studies of 273 patients with various primary tumors were reviewed to determine the utility of additional brain imaging to screen for CNS metastases. The results of this study demonstrated brain lesions in only 2 percent of the cases, and unsuspected metastases in only 0.7 percent. Thus, the authors concluded that the addition of brain imaging to wholebody FDG-PET is of very limited value. Overall, detection of one or more hypermetabolic foci in brains of patients undergoing whole-body plus brain imaging for the evaluation of metastatic disease from peripheral tumors can be quite specific. However, sensitivity for detecting small lesions, particularly those with low to moderate levels of FDG uptake, is much lower than MRI and is likely to yield numerous false negative results. Thus, unless a patient
3 PET Imaging of Brain Tumors
79
has specific CNS symptoms, the addition of brain imaging to whole-body PET is of very limited clinical utility and is generally not recommended.
15
Other Brain Tumors
In addition to CNS metastasis from peripheral tumors, other tumors such as GBM and lymphoma can result in multifocal CNS involvement and lesions that cross the midline of the brain. In the case of lymphoma, both Hodgkin’s and Non-Hodgkin’s disease can affect the CNS, both as primary lesions and as metastases from peripheral disease. In addition, in rare cases primary lymphomas can present with metastases to peripheral tissues. FDG-PET is very sensitive for evaluating extracerbral lymphoma and, in general, CNS lymphoma is also hypermetabolic. Fig. 3.7
Fig. 3.7 FDG-PET study of a patient with CNS lymphoma (A) that crossed the midline in frontal lobes and metastasized to the periphery; producing hypermetabolic lesions in the paratrachael regions (B)
80
A.J. Fischman
illustrates an FDG-PET study of a patient with a CNS lymphoma (A) that metastasized to the periphery and produced hypermetabolic lesions in the paratrachael regions (B). In contrast to CNS lymphoma, nonmalignant lesions such as toxoplasmosis have lower metabolic activity. Several studies have demonstrated that FDGPET can effectively differentiate CNS lymphoma from infection in patients with non-specific abnormalities on CT and/or MRI [38, 39, 40]. In contrast to lower grade gliomas, GBM can produce multifocal CNS involvement which, in some cases, can cross the midline. In addition, in rare cases, GBM can metastasize to extraneural sites [41].
16
Other PET Tracers for Imaging Brain Tumors
Although FDG has been, and in the near term will continue to be, the most widely used PET tracer for the clinical evaluation of brain tumors, other agents have been shown to yield critical physiological, molecular and metabolic data about tumor pathophysiology. Studies with these agents have provided important information about tumor blood flow, oxygen utilization, hypoxia, amino acid metabolism, lipid synthesis and cell proliferation [42].
17
Amino Acid Transport and Metabolism
Numerous amino acids have been labeled with positron emitters for imaging brain tumors [43]. The mechanism(s) by which radiolabeled amino acids accumulate in tumors are related to alterations in amino acid transport, protein synthesis or blood-brain barrier permeability. Although compared with FDG-PET, the effect of inflammation on tracer uptake is less important and tumor specificity is not absolute with amino acid imaging. 11C methionine has been applied extensively for imaging brain tumors; however, other tracers such as: 11C Tyrosine [44], 2-18F Tyrosine [45],O-[2-18Ffluorethyl)-L-tyrosine [FET, [46]) and most recently 3,4-dihydroxy-618 F-fluoro-l-phenylalanine (FDOPA, (13]) show great promise for this application.
18 11
11
C Methionine
C Methionine (MET) accumulates in brain tumors by multiple mechanisms, including: (1) Increased transport, (2) Increased protein synthesis, (3) Breakdown of the blood-brain barrier, (4) Methyl group transfer in lipid synthesis, and (5) The fact that methionine is the first amino acid incorporated in the synthesis of all proteins. Due to these multiple mechanisms for increased tracer accumulation in tumors, MET-PET is extremely sensitive for tumor detection and defining tumor borders, but is of limited value for studying specific aspects of tumor biology.
3 PET Imaging of Brain Tumors
81
Several studies have demonstrated that MET-PET is superior to FDG-PET for defining tumor margins and for distinguishing radiation necrosis from recurrent tumors [47, 48]. In a study by Chung, et al. [49], the utility of MET-PET in patients with brain lesions that were eu- or hypometabolic on FDG-PET was studied. The results of this investigation demonstrated that 33 out of 35 patients with tumors that were eu- or hypometabolic on FDG-PET showed increased uptake of MET (80 percent sensitivity). In contrast, 10 of 10 benign lesions had normal or decreased uptake of MET (100 percent specificity). MET-PET can also be useful for target region selection for stereotactic brain biopsies. In a study by Pirotte, et al. (50] it was demonstrated that 23 tumors had increased MET uptake, whereas two benign lesions had normal uptake. Recently, Nariai, et al. [51] performed a retrospective analysis of MET-PET studies in 194 patients with known or suspected glioma to determine how MET uptake correlated with tumor pathological features and prognosis. Tumor uptake was quantified as a ratio to contralateral healthy brain tissue (T/N). The results of this study demonstrated significant differences in T/N ratio between nonneoplastic lesions, low-grade gliomas and malignant gliomas. In patients with malignant gliomas, a significant difference in survival was observed between cases with and without postoperative tumor remnant, based on elevated MET uptake. In a study by Pirotte, et al. [52], the roles of MET-PET and FDG-PET for selecting sites for CT- or MRI- guided stereotactic biopsy were compared in 45 patients. Histologically based diagnoses were obtained in all patients (39 tumors, six benign lesions), and biopsies were performed in all tumors with the aid of PET guidance. The histological and imaging data demonstrated that all tumors had an area of abnormal MET uptake, and 33 had abnormal FDG uptake. All six benign lesions had no MET uptake. These authors concluded that: (1) When FDG shows limitations in target selection, MET is a good alternative because of its high specificity, and (2) As a single-tracer procedure and regardless of FDG uptake, MET is a better choice for PET guidance in neurosurgical procedures. In another study, Kim, et al. [53] compared the prognostic value of MET- and FDG-PET in glioma patients. In this investigation, MET-PET and FDG-PET were performed within a time interval of two weeks in 47 brain tumor patients (19 GBM, 28 others). Univariate and multivariate analyses were performed to determine significant prognostic factors. The findings of this study demonstrated that tumor pathology (glioblastoma or not), age, Karnofsky performance status (KPS) and MET uptake were significant predictors of prognosis by univariate analysis. Multivariate analysis demonstrated that tumor pathology, KPS and MET-PET were significant independent predictors. The Ki-67 proliferation index was significantly correlated with MET uptake, but not with FDG uptake. Fig. 3.8 illustrates intense MET uptake in an area of enhancement on MRI in a patient with left frontal GBM. Fig. 3.9 shows FDG- and MET-PET images of a patient with a high-grade glioma of the splenium of the corpus callosum. Although the FDG-PET images show only moderate uptake in this region, the MET-PET images show intense uptake.
82
A.J. Fischman
Fig. 3.8 Co-registered MET-PET (A) and MRI (B) images of a patient with a left frontal GBM. The PET study clearly demonstrates an area of MET accumulation that is greater than normal gray matter; consistent with high-grade tumor recurrence
Fig. 3.9 FDG- and MET-PET images of a patient with a high-grade glioma of the splenium of the corpus callosum. Although the FDG-PET images show only moderate uptake in this region, the MET-PET images show intense uptake
MET-PET has also been shown to be useful for differentiating recurrent tumors from radiation necrosis. In a study by Tsuyuguchi, et al. [54] MET-PET was performed in 21 patients with suspected recurrent brain tumors or radiation injury after stereotactic radiosurgery. The findings of the study demonstrated that the sensitivity and specificity of MET-PET for detecting tumor recurrence were 77.8 percent and 100 percent, respectively. In another study, Van Laere, et al. [55] performed a direct
3 PET Imaging of Brain Tumors
83
comparison of MET- and FDG-PET for detecting glioma recurrence. In this investigation, uptake of FDG and MET was determined on the same day in 30 patients after therapy for a primary brain tumor (23 Grade II to IV astrocytomas, four oligodendrogliomas and three mixed oligo-astrocytomas). The findings of this investigation demonstrated pathologically increased tracer uptake in 28 of 30 MET-PET scans, but only in 17 of 30 FDG-PET scans. The inter-observer agreement was 100 percent for MET and 73 percent for FDG. Kaplan-Meier survival analysis demonstrated significant differences for both FDG-PET (p=0.007) and MET-PET (p=0.014). The combination of FDG- and MET-PET resulted in the highest prognostic accuracy (p=0.003); however, MET-PET alone was the best prognostic predictor in the subgroup of patients with primary astrocytoma. The authors concluded that: (1) FDG- and MET-PET provide complementary prognostic information in patients with suspected brain tumor recurrence or progression after therapy, and (2) Due to its sensitivity and clearer delineation of the suspected recurrence, MET is the single tracer of choice for evaluating these patients. Overall, these studies indicate that PET with a combination of tracers may be useful for providing a metabolic profile for a specific tumor [36]. High uptake on FDG-PET would confirm the presence of a high-grade tumor, whereas low uptake could represent a low-grade tumor, post-therapy changes, infarct, a benign lesion or, in rare cases, a high-grade lesion (Fig. 3.9). This study could then be followed by MET-PET to differentiate low- or intermediate-grade tumors from post-therapy changes, infarct or benign lesions.
19
O-(2-[18F]fluoroethyl)- L-tyrosine
O-(2-[18F]fluoroethyl)- L-tyrosine (FET) has been shown to be a useful tracer for diagnosing recurrent glioma. In a study by Popperl, et al. [14], FET-PET was performed in 52 patients with glioma (primary grading: 27=WHO Grade IV, 16=Grade III, 9=Grade II, 1=Grade I), and clinically suspected recurrence at four to 180 months after different treatments. The results of this investigation demonstrated that all patients had FET uptake, of varying intensity, in the area of the primary tumor after initial therapy. In 42 patients with confirmed recurrence there was additional distinct focal FET uptake with significantly higher values, compared with 11 patients without clinical signs of recurrence (these patients had only low and homogeneous FET uptake at the margins of the resection cavity). With respect to tumor grading, there was a slight, but not statistically significant, increase from WHO II to WHO III and WHO IV recurrence. The authors concluded that FET-PET reliably distinguishes between post- therapeutic benign lesions and tumor recurrence after initial treatment of low- and high-grade gliomas. In another study by Weckesser, et al. [56], FET-PET was performed in 44 patients referred for the evaluation of a suspected brain tumor. The results of this investigation demonstrated increased FET uptake in 35 of 44 lesions. All histologically confirmed gliomas and many other lesions showed a variable degree of FET. No uptake was observed in nine lesions
84
A.J. Fischman
(one inflammatory lesion, one dysembryoplastic neuroepithelial tumor, one mature teratoma and six lesions without histological confirmation).
20
3,4-dihydroxy-6-18F-fluoro-l-phenylalanine (FDOPA)
In addition to its primary application for evaluation of patients with movement disorders [57], as an amino acid analog, FDOPA, is also an important tracer for studying amino acid metabolism in tumors. A recent study by Chen, et al. [13] examined 81 patients undergoing evaluation for brain tumors. Tracer kinetics in normal brain tissue and tumors were estimated. PET uptake was quantified by SUV and the ratio of tumor to normal hemispheric tissue uptake (T/N). In addition, PET uptake of FDOPA was quantified as ratios of tumor to striatal uptake (T/S) and of tumor to white matter uptake (T/W). The accuracies of FDOPA-PET and FDG-PET were determined by comparing imaging data with histologic findings and clinical follow-up. The results of this study demonstrated that tumor uptake of FDOPA was rapid and peaked at ∼15 minutes after intravenous injection. Tumor vs. striatal uptake could be distinguished by the difference in peak times. Both high- and low-grade tumors were well visualized with FDOPA. The sensitivity for identifying tumors was significantly higher with FDOPA, compared with FDG at comparable specificities, as determined by visual inspection, especially for low-grade tumors. Receiveroperating-characteristic curve analysis demonstrated that the optimal threshold for FDOPA-PET was a T/S of greater than 1.0 (sensitivity, 96 percent; specificity, 100 percent), or an T/N of greater than 1.3 (sensitivity, 96 percent; specificity, 86 percent). For all of the patients studied, the diagnostic accuracy of FDOPA PET at these thresholds was: sensitivity, 98 percent; specificity, 86 percent; positive predictive value, 95 percent, and negative predictive value, 95 percent. No significant difference in tumor uptake on FDOPA-PET was seen between low- and high-grade tumors, or between contrast-enhancing and non-enhancing tumors. In general radiation necrosis was distinguishable from tumor recurrence on FDOPA-PET (P < 0.00001). Overall, it was demonstrated that FDOPA-PET was more accurate than FDG-PET for imaging of low-grade tumors, evaluating recurrent tumors and for distinguishing tumor recurrence from radiation necrosis.
21
Cellular Proliferation
Tritiated thymidine is the “gold standard” for studying cell proliferation in vitro and this pyrimidine nucleoside, when labeled with 11C, is an excellent tracer for studying cell proliferation in brain tumors by PET. Unlike other PET radiopharmaceuticals, uptake of 11C thymidine (dThr) by tumors provides a direct measure of nucleotide metabolism, DNA synthesis and cell proliferation in vivo. In a study by Eary, et al. [58] a series of 13 patients underwent closely spaced dThd-PET, FDG-
3 PET Imaging of Brain Tumors
85
PET and MRI procedures, and the results were compared by standardized visual analysis. The dThd images were qualitatively different from the other two studies in ∼50 percent of the cases, suggesting that dThd provides different information from FDG-PET and MRI. In two cases recurrent tumor was more apparent on dThd-PET, compared with FDG-PET. In two other patients uptake of dThd in the tumor was less than FDG uptake, and these patients had slower tumor progression than the three patients with high uptake of both tracers. Since DThd-PET can demonstrate the presence of viable tumors in areas where FDG uptake is normal, assessment of tumor viability by imaging cell proliferation may be superior to FDG-PET for monitoring disease activity and response to therapy. Recently 3′-Deoxy-3′-18F-fluorothymidine (FLT) was developed as a PET tracer for imaging tumor cell proliferation [15]. This agent is a substrate for thymidine kinase, and its kinetic behavior in tumors can be described by a model that is very similar to the one used for quantitative analysis of FDG-PET data. In a recent study by Chen, et al. [59] the accuracy of FLT and FDG images were compared in a series of 25 patients with newly diagnosed or previously treated gliomas. More than half of the patients underwent resection after the PET studies, and correlations between tracer uptake and the Ki-67 proliferation index were examined. The predictive power of FLT and FDG for tumor progression and survival was analyzed using Kaplan-Meier statistics. The results of this study demonstrated that tumor uptake of FLT is rapid with peak uptake at five to 10 minutes after injection, which remained stable for up to 75 minutes. This indicated that a 30-minute scan beginning at five minutes after injection was adequate for imaging. FLT visualized all high-grade (Grade III or IV) tumors; however, Grade II tumors did not show appreciable uptake. Absolute uptake of FLT was much lower compared with FDG; however, image contrast was much better (T/N ratio, 3.85 vs. 1.49). These findings were not surprising since (1) The rate of DNA synthesis in tumors is much lower than the rate of glucose utilization and (2) In contrast to FDG, FLT does not cross the intact blood-brain barrier. SUVmax for FLT was more highly correlated with Ki-67 index (r = 0.84; p < 0.0001), than SUVmax for FDG (r = 0.51; P = 0.07). In addition FLT uptake had more significant predictive power for tumor progression and survival (P = 0.0005 and P = 0.001, respectively). Thus, FLT appears to be a promising tracer as a marker of proliferation, particularly in high-grade tumors. In another study [60], 10 patients with recurrent GBM were studied by FLT-PET and Gd-DTPA MRI. All tumors had increased FLT uptake and showed Gd-DTPA enhancement. In addition, there was a significant correlation between the volume of the metabolically active part of the tumor determined by PET and the volume of enhancement; however, there were differences in the areas of Gd-DTPA enhancement and FLT uptake. In a study by Muzi, et al. [61] 12 patients with primary brain tumors were imaged by dynamic FLT-PET with arterial blood sampling and the data were analyzed by kinetic modeling. The results of this investigation demonstrated that, in tumors that show breakdown of the blood-brain barrier, transport dominates FLT uptake. Transport across the blood-brain barrier and modest rates of FLT phosphorylation appear to limit the assessment of cellular proliferation using FLT to highly
86
A.J. Fischman
proliferative tumors with breakdown of the blood-brain barrier. These findings are consistent with the results described above [59, 60].
22
Lipid Synthesis
A variety of tumors show increased synthesis of membrane lipids, which is associated with increased choline metabolism. In a study by Hara, et al. [62] it was shown that PET with 11C choline can be used for imaging brain tumors. Subsequently, DeGrado, et al. [63] prepared 18F labeled choline (FCH) and demonstrated its utility for tumor imaging. Like MET, FCH accumulates in brain tumors and has low background accumulation in normal tissue. In a study by Hara, et al. [64] PET studies with 18F- and 11C-choline were performed in 12 patients with suspected glioma by CT and MRI criteria. Uptake of both tracers was always low in low-grade gliomas, and high in high-grade glioma. In general, 18F choline yielded slightly superior results, compared with 11C-choline. Radiolabeled acetate (ACE) could also be a useful tracer for studying lipid metabolism in brain tumors. In a study by Liu, et al. [65] PET with11C ACE and FDG was performed sequentially in 26 patients with primary astrocytomas. The results of this investigation demonstrated that the sensitivity and specificity of ACE, for discriminating high- from low-grade astrocytomas, were 42 percent and 86 percent, respectively. This was lower than the results obtained with FDG, which showed sensitivity and specificity of 79 percent and 100 percent. The authors concluded that ACE is a promising tracer for detecting primary astrocytomas, but is of limited value in the differentiation of high- and low-grade tumors.
23
Tissue Hypoxia
Since regions of tumor hypoxia are associated with relative resistance to radiation therapy, identification of these areas is of great importance for effective treatment planning. Tracers such as 18F Fluoromisonidazole (FMISO) and 1-(5-fluoro-5deoxy-alpha-D-arabinofuranosyl)-2-nitroimidazole (FAZA,) have been developed as PET tracers for imaging hypoxia [66, 67]. Although these tracers have been useful for identifying areas of hypoxia in peripheral tumors, applications to brain tumors have been limited. However in a study by Bruehlmeier, et al. [68], PET with FMISO and 15O-H2O was used to measure in vivo hypoxia and perfusion in 11 patients with various brain tumors. The results of this investigation demonstrated that FMISO-PET can be used to define the spatial description of hypoxia in brain tumors that is independent of blood-brain barrier disruption and tumor perfusion. In a study by Cher, et al. [69] 17 patients with newly primary gliomas were evaluated for tumor hypoxia by FMISO-PET, and correlations were made with FDG-PET and tumor markers of angiogenesis and hypoxia. The results of this investigation
3 PET Imaging of Brain Tumors
87
demonstrated FMISO uptake in all high-grade gliomas, but not in low-grade gliomas. In addition a significant relationship was found between FDG or FMISO uptake and expression of VEGF-R1 and Ki 67.
24
15
O Labeled Tracers
Tracers labeled with 15O include: bolus injected H2150 or inhaled C1502 for studying tumor perfusion, 15O for studying oxygenation and inhaled C150 for studying tumor blood volume [70, 71. 72]. Although none of these tracers has the general applicability of the radiopharmaceuticals described above, they can have considerable value in specific applications. For example, because perfusion to the cerebral cortex is changed by cortical activation, perfusion studies with injected H2150 or inhaled C1502 can be employed to localize specific areas of eloquent cortex [73]. In these types of studies images of cerebral perfusion are acquired at rest and during the performance of a specific task. By subtracting the images and co-registration with MRI, the brain area that mediates the task can be identified. This type of study is of particular value in planning surgery or radiation therapy in patients with distorted cerebral anatomy after previous surgery.
25 ●
●
●
●
●
●
●
●
Key Points
The unique metabolic information provided by FDG-PET can be extremely valuable in both the initial diagnosis and post-therapy re-evaluation of brain tumors. The major advantages of FDG as a PET tracer include: its wide availibility, the relatively long physical half-life of 18F, compared with other positron emitters and, most importantly, the correlation between glucose utilization rate and tumor grade. These factors have important practical, diagnostic and prognostic implications for the medical and surgical management of patients with brain tumors. The major drawbacks of FDG-PET are limited spatial resolution and high background activity in gray matter. In contrast, although CT and MRI provide exquisite anatomic detail, they do not yield metabolic information and are of limited value in assessing tumor grade. Thus, when FDG-PET images are co-registered with CT or MRI data, fusion images of metabolic and anatomic information provides a powerful tool for tumor evaluation. In addition, co-registration of serial PET studies performed at initial diagnosis at various times after treatment is very helpful for studying treatment effects and tumor recurrence. FDG-PET can also be useful in detecting malignant degeneration in low-grade tumors, which can be of great value for guiding stereotactic biopsy and/or surgery.
88 ●
●
A.J. Fischman
The low level of glucose metabolism in low-grade tumors reduces the sensitivity of FDG-PET for detection of these lesions. In treated low-grade tumors, PET studies with other tracers such as MET, FDOPA, FET or FLT appear to be superior to FDG-PET for detection of recurrent or residual tumors.
References 1. Jemal A, Siegel R, Ward E, Murray T, Xu J, Smigal C, Thun MJ. Cancer statistics, 2006. CA Cancer J Clin 2006;56(2):106-130. 2. Surawicz TS, Davis F, Freels S, Laws ER, Menck HR, Laws ER Jr. Brain tumor survival: results from the National Cancer Data Base. J Neurooncol 1998;40(2):151-160. 3. Alavi A, Mavi A, Basu S, Fischman AJ. Is PET-CT the only option? Eur J Nucl Med (In Press). 4. Herscovitch P, Markham J, Raichle ME. Brain blood flow measured with intravenous H215O. I. Theory and error analysis. J Nucl Med 1983;24(9):782-789. 5. Raichle ME, Martin WR, Herscovitch P, Mintun MA, Markham. J. Brain blood flow measured with intravenous H215O. II. Implementation and validation. J Nucl Med 1983;24(9): 790-798. 6. Mintun MA, Raichle ME, Martin WR, Herscovitch P. Brain oxygen utilization measured with O-15 radiotracers and positron emission tomography. J Nucl Med 1984;25(2):177-187. 7. Jager PL, Vaalburg W, Pruim J, de Vries EG, Langen KJ, Piers DA Radiolabeled amino acids: basic aspects and clinical applications in oncology. J Nucl Med 2001;42(3):432-445. 8. Derlon JM, Chapon F, Noel MH, Khouri S, Benali K, Petit-Taboue MC Houtteville JP, Chajari MH, Bouvard G. Non-invasive grading of oligodendrogliomas: correlation between in vivo metabolic pattern and histopathology. Eur J Nucl Med 2000;27(7):778-787. 9. Ribom D, Eriksson A, Hartman M, Engler H, Nilsson A, Langstrom B, Bolander H Bergstrom M, Smits A. Positron emission tomography 11C-methionine and survival in patients with lowgrade gliomas. Cancer 2001;92(6):1541-1549. 10. Chung JK, Kim K, Kim Sk SK, Lee J, Paek S, Yeo S, Jeong M, Lee S, Jung W, Lee C Usefulness of 11C-methionine PET in the evaluation of brain lesions that are hypo- or isometabolic on (18)F-FDG PET. Eur J Nucl Med 2002;29(2):176-182. 11. Eary JF, Mankoff DA, Spence AM, Berger MS, Olshen A, Link JM, O’Sullivan F Krohn KA. 2-[C-11]thymidine imaging of malignant brain tumors. Cancer Res 1999;59(3):615-621. 12. Reinhardt MJ, Kubota K, Yamada S, Iwata R, Yaegashi H. Assessment of cancer recurrence in residual tumors after fractionated radiotherapy: a comparison of fluorodeoxyglucose, L-methionine and thymidine. J Nucl Med 1997;38(2):280-287. 13. Chen W, Silverman DH, Delaloye S, Czernin J, Kamdar N, Pope W, Satyamurthy N, Schiepers C, Cloughesy T. 18F-FDOPA PET imaging of brain tumors: comparison study with 18F-FDG PET and evaluation of diagnostic accuracy. J Nucl Med 2006;47(6):904-911. 14. Popperl G, Gotz C, Rachinger W, Gildehaus FJ, Tonn JC, Tatsch K. Value of O-(2[18F]fluoroethyl)- L-tyrosine PET for the diagnosis of recurrent glioma. Eur J Nucl Med Mol Imaging 2004; 31(11):1464-1470. 15. Shields AF, Grierson JR, Dohmen BM, Machulla HJ, Stayanoff JC Lawhorn-Crews JM, Obradovich JE, Muzik O, Mangner TJ Imaging proliferation in vivo with [F-18]FLT and positron emission tomography. Nat Med 1998;4(11):1334-1336. 16. Skalsk J, Wahl RL, Meyer CR. Comparison of Mutual Information-Based Warping Accuracy for Fusing Body CT and PET by 2 Methods: CT Mapped onto PET Emission Scan Versus CT Mapped onto PET Transmission Scan. Journal of Nuclear Medicine 2002; 43(9): 1184-1187.
3 PET Imaging of Brain Tumors
89
17. Di Chiro G. Positron emission tomography using [18F] fluorodeoxyglucose in brain tumors. A powerful diagnostic and prognostic tool. Invest Radiol 1987;22(5):360-371. 18. Patronas NJ, Brooks RA, DeLaPaz RL, Smith BH, Kornblith PL, Di Chiro G Glycolytic rate (PET) and contrast enhancement (CT) in human cerebral gliomas. AJNR Am J Neuroradiol 1983;4(3):533-535. 19. Delbeke D, Meyerowitz C, Lapidus RL, Maciunas RJ, Jennings MT, Moots PL Kessler RM. Optimal cutoff levels of F-18 fluorodeoxyglucose uptake in the differentiation of low-grade from high-grade brain tumors with PET. Radiology 1995;195(1):47-52. 20. Paulus W, Peiffer J. Intratumoral histologic heterogeneity of gliomas. A quantitative study. Cancer 1989;64(2):442-447. 21.Hanson MW, Glantz MJ, Hoffman JM, Friedman AH, Burger PC, Schold SC Coleman RE FDGPET in the selection of brain lesions for biopsy. J Comput Assist Tomogr 1991;15(5):796-801. 22. Pirotte B, Goldman S, Brucher JM, Zomosa G, Baleriaux D, Brotchi J Levivier M. PET in stereotactic conditions increases the diagnostic yield of brain biopsy. Stereotact Funct Neurosurg 1994;63(1-4):144-149. 23. Pirotte B, Goldman S, David P, Wikler D, Damhaut P, Vandesteene A, Salmon L Brotchi J, Levivier M. Stereotactic brain biopsy guided by positron emission tomography (PET) with [F-18]fluorodeoxyglucose and [C-11]methionine. Acta Neurochir Suppl (1997) 68:133-138. 24. Pirotte B, Goldman S, David P, Wikler D, Damhaut P, Vandesteene A, Salmon L Brotchi J, Levivier M Stereotactic brain biopsy guided by positron emission tomography (PET) with [F18]fluorodeoxyglucose and [C-11]methionine. Acta Neurochir Suppl 1997;68:133-138. 25. Alavi JB, Alavi A, Chawluk J, Kushner M, Powe J, Hickey W, Reivich M Positron emission tomography in patients with glioma. A predictor of prognosis. Cancer 1988;62(6):1074-1078 26. Patronas NJ, Di Chiro G, Kufta C, Bairamian D, Kornblith PL, Simon R Larson SM Prediction of survival in glioma patients by means of positron emission tomography. J Neurosurg 1985;62(6):816-822. 27. De Witte O, Levivier M, Violon P, Salmon I, Damhaut P, Wikler D Hildebrand J, Brotchi J, Goldman S, Wikler D Jr. Prognostic value positron emission tomography with [18F]fluoro-2deoxy-D-glucose in the low-grade glioma. Neurosurgery 1996;39(3):470-6; discussion 476-477. 28. Francavilla TL, Miletich RS, Di Chiro G, Patronas NJ, Rizzoli HV, Wright DC. Positron emission tomography in the detection of malignant degeneration of low-grade gliomas. Neurosurgery (1989 Jan) 24(1):1-5 29. Schifter T, Hoffman JM, Hanson MW, Boyko OB, Beam C, Paine S, Schold SC Burger PC, Coleman RE. Serial FDG-PET studies in the prediction of survival in patients with primary brain tumors. Comput Assist Tomogr 1993; 17(4):509-561. 30. Hanson MW, Hoffman JM, Glantz MJ, et al. FDG PET in the early postoperative evaluation of patients with brain tumor. 1990;31:799p. 31. Glantz MJ, Hoffman JM, Coleman RE, Friedman AH, Hanson MW, Burger PC Herndon JE, Meisler WJ, Schold SC, Herndon JE, Schold SC. Identification of early recurrence of primary central nervous system tumors by [18F]fluorodeoxyglucose positron emission tomography. Ann Neurol 1991;29(4):347-355. 32. Di Chiro G, Oldfield E, Wright DC, De Michele D, Katz DA, Patronas NJ Doppman JL, Larson SM, Ito M, Kufta CV. Cerebral necrosis after radiotherapy and/or intraarterial chemotherapy for brain tumors: PET and neuropathologic studies. AJR Am J Roentgenol 1988;150(1):189-197. 33. Barker FG, Chang SM, Valk PE, Pounds TR, Prados MD, Barker FG. 18-Fluorodeoxyglucose uptake and survival of patients with suspected recurrent malignant glioma. Cancer 1997;79(1):115-126. 34. Chao ST, Suh JH, Raja S, Lee SY, Barnett G. The sensitivity and specificity of FDG PET in distinguishing recurrent brain tumor from radionecrosis in patients treated with stereotactic radiosurgery. Int J Cancer 2001;96(3):191-197. 35. Ricci PE, Karis JP, Heiserman JE, Fram EK, Bice AN, Drayer BP Differentiating recurrent tumor from radiation necrosis: time for re- evaluation of positron emission tomography? AJNR Am J Neuroradiol 1998;19(3):407-413.
90
A.J. Fischman
36. Hustinx R, Pourdehnad M, Kaschten B, Alavi A. PET imaging for differentiating recurrent brain tumor from radiation necrosis. Radiol Clin North Am 2005;43(1):35-47. 37. Larcos G, Maisey MN. FDG-PET screening for cerebral metastases in patients with suspected malignancy. Nucl Med Commun 1996;17(3):197-198. 38. Hoffman JM, Waskin HA, Schifter T, Hanson MW, Gray L, Rosenfeld S Coleman RE. FDGPET in differentiating lymphoma from nonmalignant central nervous system lesions in patients with AIDS. J Nucl Med 1993;34(4):567-575. 39. Heald AE, Hoffman JM, Bartlett JA, Waskin HA Differentiation of central nervous system lesions in AIDS patients using positron emission tomography (PET). Int J STD AIDS 1996;7(5):337-346. 40. Roelcke U, Leenders KL. Positron emission tomography in patients with primary CNS lymphomas. J Neurooncol 1999;43(3):231-236. 41. Astner ST, Pihusch R, Nieder C, Rachinger W, Lohner H, Tonn JC, Molls M Grosu AL. Extensive local and systemic therapy in extraneural metastasized glioblastoma multiforme. Anticancer Res 2006; 26(6C):4917-4920. 42. Wong TZ, van der Westuizen GJ, Coleman RE. Positron emission tomography imaging of brain tumors. Neuroimag Clin North Am 2002;12:615-626. 43. Jager PL, Vaalburg W, Pruim J, de Vries EG, Langen KJ, Piers DA. Radiolabeled amino acids: basic aspects and clinical applications in oncology. J Nucl Med 2001;42(3):432-445. 44. Pruim J, Willemsen AT, Molenaar WM, van Waarde A, Paans AM, Heesters MA Go KG, Visser GM, Franssen EJ, Vaalburg W. Brain tumors: L-[1-C-11]tyrosine PET for visualization and quantification of protein synthesis rate. Radiology 1995;197(1):221-226. 45. Wienhard K, Herholz K, Coenen HH, Rudolf J, Kling P, Stocklin G, Heiss WD Increased amino acid transport into brain tumors measured by PET of L- (2-18F) fluorotyrosine. J Nucl Med 1991;32(7):1338-1346. 46. Weckesser M, Langen KJ, Rickert CH, Kloska S, Straeter R, Hamacher K Kurlemann G, Wassmann H, Coenen HH, Schober O. O-(2-[18F]fluorethyl)-L-tyrosine PET in the clinical evaluation of primary brain tumours. Eur J Nucl Med Mol Imaging 2005;32(4):422-429. 47. Conti PS. Introduction to imaging brain tumor metabolism with positron emission tomography (PET). Cancer Invest 1995;13(2):244-259. 48. Ericson K, Lilja A, Bergstrom M, Collins VP, Eriksson L, Ehrin E von Holst H, Lundqvist H, Langsrom B B, Mosskin M Positron emission tomography with ([11C]methyl)-L-methionine, [11C] D- glucose, and [68Ga] EDTA in supratentorial tumors. J Comput Assist Tomogr 1985;9(4):683-689. 49. Chung JK, Kim YK, Kim SK, Lee YJ, Paek S, Yeo JS, Jeong JM, Lee DS, Jung HW Lee MC Usefulness of 11C-methionine PET in the evaluation of brain lesions that are hypo- or isometabolic on 18F-FDG PET. Eur J Nucl Med Mol Imaging 2002;29(2):176-182. 50. Pirotte B, Goldman S, David P, Wikler D, Damhaut P, Vandesteene A, Salmon L Brotchi J, Levivier M. Stereotactic brain biopsy guided by positron emission tomography (PET) with [F-18]fluorodeoxyglucose and [C-11]methionine. Acta Neurochir Suppl 1997;68:133-138. 51. Nariai T, Tanaka Y, Wakimoto H, Aoyagi M, Tamaki M, Ishiwata K, Senda M Ishii K, Hirakawa K, Ohno K. Usefulness of L-[methyl-11C] methionine-positron emission tomography as a biological monitoring tool in the treatment of glioma. J Neurosur 2005;103(3): 498-507. 52. Pirotte B, Goldman S, Massager N, David P, Wikler D, Lipszyc M, Salmon I Brotchi J, Levivier M. Combined use of 18F-fluorodeoxyglucose and 11C-methionine in 45 positron emission tomography-guided stereotactic brain biopsies. J Neurosurg 2004;101(3):476-483. 53. Kim S, Chung JK, Im SH, Jeong JM, Lee DS, Kim DG, Jung HW, Lee MC. 11C-methionine PET as a prognostic marker in patients with glioma: comparison with 18F-FDG PET. Eur J Nucl Med Mol Imaging 2005;32(1):52-50. 54. Tsuyuguchi N, Sunada I, Iwai Y, Yamanaka K, Tanaka K, Takami T, Otsuka Y Sakamoto S, Ohata K, Goto T, Hara M. Methionine positron emission tomography of recurrent metastatic brain tumor and radiation necrosis after stereotactic radiosurgery: is a differential diagnosis possible? J Neurosurg 2003;98(5):1056-1064.
3 PET Imaging of Brain Tumors
91
55. Van Laere K, Ceyssens S, Van Calenbergh F, de Groot T, Menten J, Flamen P Bormans G, Mortelmans L. Direct comparison of 18F-FDG and 11C-methionine PET in suspected recurrence of glioma: sensitivity, inter-observer variability and prognostic value. Eur J Nucl Med Mol Imaging 2005;32(1):39-51. 56. Weckesser M, Langen KJ, Rickert CH, Kloska S, Straeter R, Hamacher K Kurlemann G, Wassmann H, Coenen HH, Schober O O-(2-[18F]fluorethyl)-L-tyrosine PET in the clinical evaluation of primary brain tumours. Eur J Nucl Med Mol Imaging 2005;32(4):422-429. 57. Fischman AJ. Role of [18F] DOPA PET imaging in assessing movement disorders. Radio Clin North Amer 2005;43: 93-106. 58. Eary JF, Mankoff DA, Spence AM, Berger MS, Olshen A, Link JM, O’Sullivan F Krohn KA 2-[C-11]thymidine imaging of malignant brain tumors. Cancer Res 1999;59(3):615-621. 59. Chen W, Cloughesy T, Kamdar N, Satyamurthy N, Bergsneider M, Liau L Mischel P, Czernin J, Phelps ME, Silverman DH Imaging proliferation in brain tumors with 18F-FLT PET: comparison with 18F-FDG. J Nucl Med 2005;46(6):945-952. 60. Yamamoto Y, Wong TZ, Turkington TG, Hawk TC, Reardon DA, Coleman RE 3′-Deoxy-3′[F-18]fluorothymidine positron emission tomography in patients with recurrent glioblastoma multiforme: comparison with Gd- DTPA enhanced magnetic resonance imaging. Mol Imaging Biol 2006;8(6):340-347. 61. Muzi M, Spence AM, O’Sullivan F, Mankoff DA, Wells JM, Grierson JR, Link JM Krohn KA. Kinetic analysis of 3′-deoxy-3′-18F-fluorothymidine in patients with gliomas. J Nucl Med 2006;47(10):1612-1621. 62. Hara T, Kosaka N, Shinoura N, Kondo T. PET imaging of brain tumor with [methyl11 C]choline. J Nucl Med 1997;38(6):842-847. 63. DeGrado TR, Baldwin SW, Wang S, Orr MD, Liao RP, Friedman HS, Reiman R Price DT, Coleman RE. Synthesis and evaluation of 18F-labeled choline analogs as oncologic PET tracers. J Nucl Med 2001;42(12):1805-1814. 64. Hara T, Kondo T, Hara T, Kosaka N. Use of 18F-choline and 11C-choline as contrast agents in positron emission tomography imaging-guided stereotactic biopsy sampling of gliomas. J Neurosurg 2003; 99(3):474-479. 65. Liu RS, Chang CP, Chu LS, Chu YK, Hsieh HJ, Chang CW, Yang BH, Yen SH Huang MC, Liao SQ, Yeh SH. PET imaging of brain astrocytoma with 1-11C-acetate. Eur J Nucl Med Mol Imaging 2006;33(4):420-427. 66. Dubois L, Landuyt W, Haustermans K, Dupont P, Bormans G, Vermaelen P Flamen P, Verbeken E, Mortelmans L, Evaluation of hypoxia in an experimental rat tumour model by 18F fluoromisonidazole PET and immunohistochemistry. Br J Cancer 2004;91(11):1947-1954. 67. Reischl G, Ehrlichmann W, Bieg C, Solbach C, Kumar P, Wiebe LI, Machulla HJ. Preparation of the hypoxia imaging PET tracer [18F] FAZA: reaction parameters and automation. Appl Radiat Isot 2005;62(6):897-901. 68. Bruehlmeier M, Roelcke U, Schubiger PA, Ametamey SM. Assessment of hypoxia and perfusion in human brain tumors using PET with 18F-fluoromisonidazole and 15O-H2O. J Nucl Med 2004;45(11):1851-1859. 69. Cher LM, Murone C, Lawrentschuk N, Ramdave S, Papenfuss A, Hannah A O’Keefe GJ, Sachinidis JI, Berlangieri SU, Fabinyi G, Scott AM. Correlation of hypoxic cell fraction and angiogenesis with glucose metabolic rate in gliomas using 18F-fluoromisonidazole, 18F-FDG PET, and immunohistochemical studies. J Nucl Med 2006;47(3):410-418. 70. Herscovitch P, Markham J, Raichle ME. Brain blood flow measured with intravenous H215O. I. Theory and error analysis. J Nucl Med 1983;24(9):782-789. 71. Raichle ME, Martin WR, Herscovitch P, Mintun MA, Markham J. Brain blood flow measured with intravenous H215O. II. Implementation and validation. J Nucl Med 1983;24(9):790-798. 72. Mintun MA, Raichle ME, Martin WR, Herscovitch P. Brain oxygen utilization measured with O-15 radiotracers and positron emission tomography. J Nucl Med 1984;25(2):177-187. 73. Duncan JD, Moss SD, Bandy DJ, Manwaring K, Kaplan AM, Reiman EM, Chen K Lawson MA, Wodrich DL. Use of positron emission tomography for presurgical localization of eloquent brain areas in children with seizures. Pediatr Neurosurg 1997;26(3):144-156.
92
A.J. Fischman
74. Valk PE, Budinger TF, Levin VA, Silver P, Gutin PH, Doyle WK. PET of malignant cerebral tumors after interstitial brachytherapy. Demonstration of metabolic activity and correlation with clinical outcome. J Neurosurg (1988 Dec) 69(6):830-838. 75. Di Chiro G, DeLaPaz RL, Brooks RA, Sokoloff L, Kornblith PL, Smith BH Patronas NJ, Kufta CV, Kessler RM, Johnston GS, Manning RG, Wolf AP Glucose utilization of cerebral gliomas measured by [18F] fluorodeoxyglucose and positron emission tomography. Neurology 1982;32(12):1323-1329. 76. Min JJ, Iyer M, Gambhir SS. Comparison of [18F]FHBG and [14C]FIAU for imaging of HSV1-tk reporter gene expression: adenoviral infection vs stable transfection. Eur J Nucl Med Mol Imaging 2003.30(11):1547-1560. 77. Tjuvajev JG, Doubrovin M, Akhurst T, Cai S, Balatoni J, Alauddin MM, Finn R Bornmann W, Thaler H, Conti PS, Blasberg RG. Comparison of radiolabeled nucleoside probes (FIAU, FHBG, and FHPG) for PET imaging of HSV1-tk gene expression. J Nucl Med 2002;43(8):1072-1083.
4
Extracranial Head and Neck Neoplasms: Role of Imaging Myria Petrou, MA, MBChB and Suresh K. Mukherji, MD
Key Points ●
●
●
● ●
1
Imaging plays a critical part in the staging and subsequent clinical management of patients with head and neck cancers. CT tends to be the study of choice for neoplasms below the level of the soft palate. MRI is advantageous in imaging the nasopharynx and soft palate, and can be useful when assessing for skull base invasion and perineural spread. CT and MRI can provide complementary information that can guide treatment. PET-CT is assuming an evolving role in the clinical decision making process.
Introduction
The development of modern imaging techniques has had a substantial impact on the management of head and neck neoplasms. Important decisions once made at surgery are now made in advance, using information from a number of imaging modalities such as computed tomography (CT), magnetic resonance imaging (MRI) and positron emission tomography (PET) [1-9]. Imaging is not only an important component of the initial diagnosis and treatment phase, but is also the primary method of post-treatment follow-up. The intent of this chapter is to provide an overview of the role of imaging in the diagnosis and management of extracranial head and neck neoplasms. We will specifically emphasize information derived from imaging studies that can alter clinical management.
Department of Radiology, Division of Neuroradiology, University of Michigan Health System, Ann Arbor, Michigan Corresponding author: Myria Petrou, MA, MBChB. Department of Radiology, University of Michigan Hospitals, 1500 E. Medical Center Drive, Ann Arbor, MI 48109-0030 e-mail:
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
93
94
2
M. Petrou and S.K. Mukherji
Techniques
There has been considerable debate regarding the use of CT and MRI for imaging head and neck neoplasms. The evolution of multidetector CT has enabled fast and detailed evaluation of anatomy and pathology in the neck, and CT tends to be the study of choice for neoplasms below the level of the soft palate. MRI is advantageous in imaging the nasopharynx and soft palate, and can be useful when assessing for skull base invasion and perineural spread [5, 10-12]. Positron Emission Tomography CT (PET-CT) is gaining greater acceptance in imaging of head and neck neoplasms, with its major strength being the combination of anatomic and metabolic information [7-9, 13]. It can be used for detection of metastatic cervical lymph nodes with a negative predictive value of 90 percent [4, 7-9, 13-15]. It can also detect the primary tumor in patients presenting with metastatic squamous cell carcinoma of the head and neck with an unknown primary in 30-50% of cases [16-18]. The role of PET-CT is expected to increase with further technology advances that will permit optimal diagnostic quality CT imaging.
3
General Goals of Imaging:
– Define extent of tumor at initial diagnosis and help determine resectability – Assess for nodal metastases – Post-treatment surveillance Our discussion will first address the role of imaging in determining tumor resectability. We will then go on to discuss the role of imaging in the initial management of a number of head and neck neoplasms by location. Some information on nodal involvement will be presented; a detailed discussion of lymph node pathology is, however, beyond the scope of this review.
4
Imaging and Assessment of Tumor Resectability
The 2002 American Joint Commission on Cancer (AJCC) revised the T-stage classification of head and neck cancers [19]. Advanced (T4) stage cancers were subdivided into T4a and T4b categories. T4b tumors were generally categorized as surgically unresectable; T4a lesions, although requiring extensive surgery, could still be classified as resectable. The AJCC identifies three repetitive criteria for T4b cancers for most aerodigestive system locations: (1) Vascular encasement and invasion, (2) Prevertebral space invasion and (3) Invasion of mediastinal structures [19, 20].
4 Extracranial Head and Neck Neoplasms: Role of Imaging
95
MRI can be particularly helpful in assessing carotid encasement. If the vessel shows greater than 270 degrees of circumferential encasement, it usually cannot be resected from the artery at surgery. Tumors that surround the artery by less than 180 degrees can be readily resected; most tumors showing 180- to 270degree encasement can also be resected [21]. If there is concern for carotid encasement in a patient that is otherwise considered a surgical candidate, a temporary balloon occlusion can be performed under angiographic guidance. Intraluminal tumors, although specific for infiltration and unresectability, are infrequently seen and lack sensitivity. Regarding pre-vertebral muscle infiltration, Hsu, et al. [22] have demonstrated that preservation of a high T1 signal intensity retropharyngeal fat stripe is an excellent indicator that tumors have not become fixed to the prevertebral fascia or musculature. The actual presence of prevertebral invasion cannot be accurately ascertained by imaging characteristics, and is best determined at surgery or endoscopy [23]. The imaging criteria for mediastinal invasion have not been studied in as much detail. CT and MRI features that should raise suspicion for mediastinal involvement are mediastinal fat stranding, encasement/abutment of the trachea, esophagus and mediastinal vessels, esophageal wall thickening or soft tissue within the tracheal cartilages. Presence of two or more of these imaging findings results in greater specificity and sensitivity [24, 25].
5 5.1
Specific Tumor Locations: Nasopharynx
A wide variety of malignant neoplasms can theoretically arise from the nasopharyngeal mucosa; undifferentiated carcinoma is the most common form, accounting for up to 98 percent of all nasopharyngeal malignancies in the Asian population [26]. Radiation therapy is the mainstay of treatment [27]. Chemotherapy usually is tried for patients with recurrent or metastatic disease, and a combination of radiation and chemotherapy has been tried for treatment of locally advanced disease [28-31]. Surgery plays a minor role and is limited to resection of residual or recurrent disease in the nasopharynx and lymph nodes. Most tumors originate in the fossa of Rosenmuller [32]. Tumors tend to spread submucosally with early infiltration of the palatal muscles, particularly the levator veli palatini. Because the muscle is responsible for opening the eustacian tube orifice during swallowing, dysfunction leads to disequilibrium of air pressure in the middle ear and the nasopharynx [33]. The tumor itself may also obstruct the eustacian tube orifice. These factors commonly result in serous otitis media [34]. Endoscopy can grossly underestimate the extent of nasopharyngeal carcinoma (NPC) and imaging, MRI in particular, can be very helpful in determining the full extent of disease (Fig. 4.1).
96
M. Petrou and S.K. Mukherji
Fig. 4.1 (a) Extensive enhancing soft tissue mass involving the entire nasopharynx and extending posterolaterally to involve the carotid space bilaterally; the left carotid artery is completely encased by the mass. There is frank erosion of the clivus. (b) Asymmetric abnormal enhancement also noted along the left cavernous sinus and Meckel’s cave (arrows), suggestive of perineural spread of neoplasm
NPC spreads along well defined routes [5,10, 35-39]: ●
●
●
●
●
Anterior Spread: Tumors spread anteriorly into the nasal fossa and can erode into the maxillary sinus. From the nasal fossa there may be infiltration into the pterygopalatine fossa via the sphenopalatine foramen. If the tumor spreads to the pterygopalatine fossa it can extend along the maxillary nerve and into the intracranial compartment via the foramen rotundum. Tumor in the pterygopalatine fossa can also spread superiorly to the orbital apex; from here the tumor can extend to the intracranial compartment via the superior orbital fissure. Lateral Spread: This is the most common direction of spread. It can be recognized on imaging by the infiltration of the normally fat-filled parapharyngeal space (PPS). Further lateral spread can involve the masticator space and there can be subsequent perineural spread along the mandibular division of the trigeminal nerve. Posterior Spread: NPC can infiltrate the retropharyngeal space and prevertebral muscles. In advanced cases one can see destruction of the vertebral body and involvement of the spinal canal. Posterolateral extension to the carotid sheath and posterosuperior extension to the jugular foramen and hypoglossal canal can also be seen. Inferior Spread: Some tumors spread inferiorly along the posterior pharyngeal wall. Inferior spread can be more lateral, along the PPS. In advanced cases, one can see tumors in the oropharynx that can spread to the soft palate. Superior Spread: This can result in erosion of the clivus, sphenoid sinus floor, petrous apex and foramen lacerum. Skull base erosion is detected in up to onethird of patients.
4 Extracranial Head and Neck Neoplasms: Role of Imaging
97
Cervical lymph node metastases are common in NPC. Often it is enlarged lymph nodes that lead the patients to seek medical attention. Seventy-five percent of patients have enlarged lymph nodes at initial presentation and bilateral cervical lymphadenopathy can be seen in up to 80 percent [40]. Nodal metastases show an orderly inferior spread, and affected nodes are larger in the upper neck. Lateral retropharyngeal lymphadenopathy can be seen in 65 percent of patients with cervical lymph node metastases. Although the lateral retropharyngeal lymph nodes are considered first-echelon lymph nodes, 35 percent of metastases bypass these and reach the cervical lymph nodes [41, 42]. CT and MRI are complementary modalities for the evaluation of nasopharyngeal carcinoma. MRI is of greater utility in defining soft tissue extension, obliteration of fat planes and perineural spread [5, 10, 39]. The TNM staging system for NPC and a summary of clinically relevant information that can be obtained from imaging, is shown in Table 4.1.
5.2
Oral Cavity
The oral cavity is separated from the oropharynx by a plane formed by the soft palate, anterior tonsillar pillars and circumvallate papilla. The contents of the oral cavity include the oral tongue, floor of mouth, gingival, gingivobuccal and buccomasseteric regions, hard palate and mandible [43, 44]. Squamous cell carcinoma (SCCA) accounts for 90 percent of all malignant tumors involving the oral cavity [45]. Other malignancies in this location include lymphomas, sarcomas and minor salivary gland tumors. The lips are the most common site of SCCA (44.9 percent), followed by the oral tongue (16.5 percent), floor of mouth (12.1 percent), lower gingival (12.1 percent), palate and upper gingival (4.7 percent) and buccal mucosa (9.7 percent) [46].
Table 4.1 TNM Staging System for Nasopharyngeal Carcinoma and important information that can be derived from imaging TNM Staging: T1 Tumor confined to the nasopharynx T2 Tumor extends to the soft tissues of the oropharynx and/or nasal fossa T2a- without parapharyngeal extension T2b- with parapharyngeal extension T3 Tumor invades bony structures and/or paranasal sinuses T4 Tumor with intracranial extension and/or involvement of cranial nerves, infratemporal fossa, hypopharynx, orbit or masticator space. Imaging Checklist: ● ● ● ●
Detailed assessment of spread pattern, with emphasis on deep and superior extension Skull base erosion Involvement of the mandibular division of the fifth cranial nerve. Cavernous sinus involvement
98
5.3
M. Petrou and S.K. Mukherji
Oral Tongue
The majority of SCCAs in the oral tongue arise from its lateral borders and undersurface [46, 47]. These tumors remain localized to the oral tongue until they become very large. Tumors arising from the anterior and middle third of the tongue tend to spread to the floor of the mouth and to the root of the tongue. Tumors that arise from the posterior one-third of the oral tongue can extend to the tongue base and show spread patterns similar to tongue base cancers. Surgical options for carcinomas arising from the oral tongue are wide local excision, partial glossectomy or total glossectomy. Low-volume, superficial lesions with no evidence for perineural spread or vascular invasion may be adequately treated with wide local excision or localized radiation therapy. Advanced lesions that do not cross the midline may be adequately treated with a partial glossectomy. If tumors cross the midline and show evidence of invasion of the contralateral lingual neurovascular bundle, then total glossectomy is needed for curative resection, versus nonsurgical organ preservation therapy [46-48], as shown in Fig. 4.2. Bone erosion is seen only with very advanced tongue carcinomas. Lateral tumors that extend inferiorly into the floor of the mouth acquire spread patterns of floor of mouth tumors. Infiltration of the floor of the mouth would require removal of involved structures, in addition to a partial glossectomy. The extent of bone erosion identified on CT would determine the type of mandibulectomy [49].
Fig. 4.2 Large enhancing mass (*) involving the left anterior aspect of the oral, tongue and extending over the midline to involve the contralateral aspect of the tongue. Patient was treated with a near total glossectomy
4 Extracranial Head and Neck Neoplasms: Role of Imaging
5.4
99
Floor of Mouth
The surgical approach for these cancers is dependent upon the extent of tumor determined by physical exam and imaging. The crucial information that imaging can provide is the presence of bone erosion and the degree of submucosal extension [50, 51]. Low-volume superficial lesions are often not seen on imaging studies. These lesions are resected by wide local excision through a transoral approach. Tumors that extend along the mylohyoid or hyoglossus muscles require a combined transoral and cervical approach. Tumors can also extend posteriorly to the tongue base and may occasionally spread over the free margin of the mylohyoid to the soft tissues of the neck. These advanced tumors require a mandibulotomy, in addition to a combined transoral and cervical approach [44]. The relationship of a floor of mouth cancer with the midline lingual septum and the contralateral neurovascular bundle has to be determined prior to surgery. Obliteration of the fat planes surrounding the lingual vessels suggests perineural and perivascular spread [52]. Tumors that show invasion of the ipsilateral and contralateral neurovascular bundles require a total glossectomy. However, given the morbidity associated with this procedure, patients commonly opt for organ sparing treatment alternatives. Bone erosion upstages lesions to T4. The extent of bone erosion, as defined by CT, can determine the type of surgery [Fig. 4.3]. If erosion is limited to the lingual cortex, a marginal mandibulectomy can be performed. If there is extension into the marrow, segmental mandibulectomy followed by reconstruction would be needed [53].
Fig. 4.3 (a) and (b) Enhancing mass involving the anterior aspect of the floor of mouth (arrow), effacing the adjacent fat plane and abutting the mandibular cortex. Despite the proximity to the mandibular cortex, the cortex appears intact. Lack of osseous involvement was confirmed at surgery
100
5.5
M. Petrou and S.K. Mukherji
Lip Carcinoma
Early lesions are difficult to differentiate from the normal orbicularis oris muscle. The lesions that require imaging are advanced, infiltrative tumors with margins that cannot be determined on clinical inspection. Imaging can help define the full extent of these lesions, as well as identify bone erosion. Subtle bone erosion can occur along the buccal surface of the mandibular or maxillary alveolar ridge and is best detected with CT. The presence of bone erosion upgrades the cancer stage to T4, thereby contraindicating treatment with wide local excision and necessitating resection with either a partial mandibulectomy or maxillectomy [5, 54].
5.6
Buccal Carcinoma
The buccal mucosa covers the lips and cheeks. It is continuous with the gingiva of the buccal surface of the maxillary and mandibular alveolar ridges and with the retromolar trigone. Buccal carcinomas most commonly originate along the lateral walls and are usually low-grade. Early buccal carcinomas are difficult to visualize and may be indistinguishable from the orbicularis oris muscle. The most common pattern of spread is lateral submucosal extension along the buccinator muscle to the pterygomandibular raphe, and erosion of the underlying lesion. Low-volume lesions limited to the mucosa can be resected with a wide local excision. Tumors that extend submucosally to the pterygomandibular raphe need more extensive resection. Erosion of the underlying bone requires a partial maxillectomy or segmental mandibulectomy [53, 54].
5.7 Gingiva and Hard Palate Carcinoma The gingiva is the mucous membrane covering the floor of the mouth, the mandible and the maxilla. The upper gingiva is continuous with the hard palate mucosa. Primary SCCA of the hard palate is rare; SCCA of the hard palate usually represent spread of primary gingival SCCA. Primary malignant tumors of the hard palate are usually minor salivary gland carcinomas, with the most common being adenoid cystic and mucoepidermoid types [55]. Cross-sectional imaging often underestimates the extent of gingival and hard palate tumors, which can be better assessed with endoscopy. CT is particularly helpful in evaluating bone erosion. Again, the degree of bone erosion determines the surgical approach. Low-volume, superficial tumors that do not erode the underlying bone may be excised with an intraoral wide local excision. Slow-growing lesions arising from the lower gingiva may result in saucerization of the mandibular cortex and preservation of the medullary cavity. Tumors that demonstrate this type of erosion require a marginal mandibulectomy. Invasion of the medullary cavity
4 Extracranial Head and Neck Neoplasms: Role of Imaging
101
necessitates a segmental mandibulectomy. This type of bone erosion usually occurs from involvement of the occlusal surface and is more common in edentulous patients. Bone erosion of the maxillary alveolar ridge or hard palate by upper gingival carcinomas requires a partial maxillectomy, rather than a wide local excision. Thin section coronal CT (1 to 2 mm contiguous sections) is an ideal method for identifying hard palate erosion [54]. Patients with adenoid cystic carcinomas of the hard palate can commonly exhibit perineural spread along the greater and lesser palatine nerves to the pterygopalatine fossa [45]. MR imaging should be performed to evaluate for perineural spread along the pterygopalatine fossa, foramen rotundum and the cavernous sinus [11,12, 56]. Cavernous sinus involvement by adenoid cystic carcinoma or SCC precludes surgical resection of the primary tumor at many institutions.
5.8
Retromolar Trigone
The retromolar trigone is a small triangle-shaped area posterior to the last molars. Because of their unique location, SCCs that arise in the retromolar trigone have complex spread patterns [5]. The pterygomandibular raphe is a band of connective tissue situated beneath the mucosal surface of the retromolar trigone. It forms a junction between the oral cavity, nasopharynx and oropharynx and serves as a common insertion point for the buccinator, orbicularis oris and superior constrictor muscles. Therefore, tumors arising in the retromolar trigone may grow anteriorly along the orbicularis oris or buccinator into the buccal region; they may also grow posteriorly along the superior constrictor into the tonsil. Alternatively, superior growth along the pterygomandibular raphe allows access to the skull base and nasopharynx, and inferior growth results in invasion of the floor of the mouth. Imaging plays an important role in defining the full extent of deep spread of retromolar trigone carcinomas, and prevents wide local excision of apparently superficial lesions on clinical exam that demonstrate deep extension on imaging. Early retromolar trigone tumors can erode bone, given the proximity of the maxillary tuberosity and the anterior aspect of the ascending ramus of the mandible (Fig. 4.4). Bone erosion is usually clinically occult and can be only identified on imaging. Bone erosion renders these lesions T4 and requires a partial mandibulectomy or maxillectomy [45, 54]. T-staging details and pertinent imaging findings for oral cavity cancers are given in Table 4.2.
5.9
Oropharynx
The oropharynx consists of the pharyngeal wall between the nasopharynx and the pharyngoepiglottic fold, the soft palate, the tonsillar region and the tongue base.
102
M. Petrou and S.K. Mukherji
Fig. 4.4 (a) and (b) Extensive enhancing mass centered in the retromolar trigone (*), extending to and resulting in erosion of the left anterior mandibular ramus (arrow)
Table 4.2 TNM Staging System for Oral Cavity Carcinomas and important information that can be derived from imaging TNM Staging: T1 Tumor 2 cm or less in greatest dimension T2 Tumor >2 cm but <4 cm in greatest dimension T3 Tumor > 4 cm in greatest dimension T4a (Lip) Tumor invades through cortical bone, inferior alveolar nerve, floor of mouth, or skin of face. T4a (Oral cavity) Tumor invades through cortical bone, into deep (extrinsic) muscle of tongue, maxillary sinus or skin of face T4b Tumor involves masticator space, pterygoid plates or skull base and/or encases internal carotid artery Imaging Checklist: Lip Carcinomas: Bone erosion Soft tissue invasion Floor of mouth carcinomas: Extent of bone erosion Deep invasion along the mylohyoid and hyoglossus muscles Relationship to ipsilateral lingual neurovascular bundle Extension across midline and relationship to contralateral neurovascular bundle Tongue base invasion Extension into soft tissues of the neck Oral Tongue: Invasion of ipsilateral lingual neurovascular bundle Extension across midline and relationship to contralateral neurovascular bundle Invasion of floor of mouth and associated bone erosion (continued)
4 Extracranial Head and Neck Neoplasms: Role of Imaging
103
Table 4.2 (continued) Buccal Mucosa: Submucosal extension Bone erosion Gingiva and hard palate: Bone erosion Perineural invasion of the incisive canal, greater and lesser palatine foramina Retromolar Trigone: Bone erosion Submucosal spread Perineural invasion
SCCA and its variants account for more than 90 percent of oropharyngeal malignancies. Lymphomas are the second most common malignancy and typically arise from Waldeyer’s ring. Other malignancies that may occur in the oropharynx include minor salivary gland tumors, mucosal melanomas, sarcomas and lymphoepitheliomas [54].
5.10
Tonsil and Soft Palate Carcinoma
Treatment of tonsillar and soft palate carcinomas depends on lesion size and involvement of surrounding structures (Fig. 4.5). In these tumors, bone erosion is an unusual finding and occurs only with advanced soft palate carcinomas that have spread to the hard palate. Detailed imaging of the soft tissue extent of tonsillar and soft palate carcinomas provides information that may affect their surgical management. Although CT is commonly used to evaluate the primary lesion, the extent of these tumors can be optimally assessed with MRI. Early tumors, localized in the tonsillar fossa or the soft palate, can be treated with wide local excision through an intraoral approach or can be treated with definitive radiation therapy [57, 58]. Advanced tumors that extend submucosally through the superior constrictor and invade the parapharyngeal and masticator spaces require resection of parts of the tongue base, mandible and maxilla vs. organ preservation therapy [59, 60]. Progressive growth of these tumors may result in encasement of the internal carotid artery or extension superiorly along the fascial and muscle planes into the skull base.
5.11
Tongue Base Carcinoma
The most commonly performed surgical procedure for treatment of tongue base carcinomas is a partial glossectomy. This procedure requires preservation of one lingual artery and one hypoglossal nerve. CT or MRI may be used preoperatively to evaluate patients with a tongue base carcinoma. Important information that directly impacts surgical management can be derived from imaging. Specifically,
104
M. Petrou and S.K. Mukherji
Fig. 4.5 Bilateral tonsillar masses (*), left greater than right. Right-sided mass is confined to the tonsillar fossa. Left-sided mass extends anteriorly to involve the posterior aspect of the tongue (arrow heads). Mass also extends posterolaterally to involve the left parapharyngeal space
involvement of the ipsilateral neurovascular bundle, submucosal involvement in adjacent areas such as the floor of the mouth and tumor extension across the midline is information that can be derived from imaging. If the tumor extends across the midline, its relationship to the contralateral neurovascular bundle has to be determined. Spread across the midline and proximity to the opposite neurovascular bundle requires a total glossectomy, which is often considered an unacceptable surgical option [54]. T-staging details and key imaging findings in the evaluation of oropharyngeal carcinomas are given in Table 4.3.
5.12
Hypopharynx
The hypopharynx is the portion of the upper aerodigetive tract that extends from the hyoid bone superiorly and extends inferiorly to the inferior aspect of the cricoid
4 Extracranial Head and Neck Neoplasms: Role of Imaging
105
Table 4.3 TNM Staging System for Oropharyngeal Carcinoma and important information that can be derived from imaging TNM Staging: T1 Tumor 2 cm or less in greatest dimension T2 Tumor >2 cm but <4 cm in greatest dimension T3 Tumor > 4 cm in greatest dimension T4a Tumor invades through cortical bone, inferior alveolar nerve, floor of mouth, or skin of face T4a Tumor invades the larynx, deep/extrinsic muscles of the tongue, medial pterygoid, hard palate or mandible T4b Tumor invades the lateral pterygoid muscle, pterygoid plates, lateral nasopharynx or skull base or encases carotid artery Imaging Checklist: Tonsil, soft palate and posterior pharyngeal wall tumors: ● ● ● ● ●
Detailed evaluation of submucosal extension into the soft tissues of the neck Tongue base invasion Encasement of carotid artery Bone erosion Prevertebral muscle invasion
Tongue base carcinomas: ● ● ●
Extension to floor of mouth and surrounding structures Relationship to ipsilateral lingual neurovascular bundle Extension across midline and relationship to contralateral neurovascular bundle
cartilage. Above the hyoid is the oropharynx; below the cricoid cartilage the hypopharynx becomes the cervical esophagus. Most investigators divide the hypopharynx into three regions: the posterior pharyngeal wall, the piriform sinuses and the postcricoid region [61]. More than 95 percent of all hypopharyngeal tumors are SCCAs. Hypopharyngeal tumors can remain asymptomatic for a long time. At the time of diagnosis, up to 75 percent of patients have cervical lymph node metastases [57]. Hypopharyngeal tumors can be treated surgically or with definitive radiation treatment. The extent of disease is often underestimated at endoscopy because of the tendency of these tumors to spread submucosally [62]; this is best assessed with imaging. The specific issues that need to be addressed in patients with hypopharyngeal carcinomas are whether the tumor crosses the midline, extension of tumor into the apex of the piriform sinus, presence of cartilage erosion, inferior tumor extension and invasion of the prevertebral muscles [5, 54, 63-65]. Low-volume superficial lesions located in the superior aspect of the hypopharynx may be resected with a transoral wide local excision. Small lesions located more inferiorly may require a midline mandibulolabial approach and a transhyoid or lateral pharyngotomy. Advanced lesions involving the lateral or posterior pharyngeal walls, but not crossing the midline, may be treated with a partial laryngopharyngectomy. Advanced tumors that cross the midline require a total laryngopharyngectomy.
106
M. Petrou and S.K. Mukherji
Tumor involvement of the apex of the piriform sinus is an important factor in carcinomas arising in this location that are to be treated surgically. The apex of the piriform sinus is located at the level of the true vocal cords. Because of the proximity of the apex to the cricoarytenoid joint, there is a high likelihood of laryngeal invasion and cartilage erosion. Therefore, piriform sinus carcinomas involving the apex are treated with laryngopharyngectomy; tumors that spare the apex may be treated with partial laryngopharyngectomy [5, 54]. The inferior margin of the tumor and its relation to the esophageal inlet should also be addressed. Invasion of the postcricoid region with extension to the cervical esophagus requires esophagectomy, in addition to total laryngopharyngectomy [5, 54, 63]. Overall tumor volume has been shown to be an important outcome predictor, for both tumors treated surgically as well as for those treated with radiation [66, 67]. The rate of second primary tumors in patients with hypopharyngeal cancer is higher than for the remainder of head and neck cancers, with a reported frequency of up to 15 percent [68], and one should carefully look for them on cross-sectional imaging studies. T-staging details and key imaging information for hypopharyngeal carcinomas is summarized in Table 4.4.
5.13
Larynx
The larynx is divided into the supraglottis, glottis and subglottis. The supraglottic larynx is located above the true vocal cord (TVC) and extends from the tongue base and valleculae to the laryngeal ventricle. The contents of the supraglottic larynx are the epiglottis, aryepiglottic folds, false vocal cords, laryngeal ventricle and the arytenoid processes of the arytenoid cartilages [61]. Table 4.4 TNM Staging System for Hypopharyngeal Carcinoma and important information that can be derived from imaging TNM Staging: T1 Tumor limited to one subsite of the hypopharynx and 2 cm or less in greatest dimension T2 Tumor invades more than one subsite of hypopharynx or an adjacent site or measures > 2 cm but < 4 cm in greatest diameter without fixation of the hemilarynx. T3 Tumor measures > 4 cm or with fixation of the hemilarynx T4a Tumor invades thyroid/cricoid cartilage, hyoid bone, thyroid gland or central compartment soft tissue T4b Tumor invades prevertebral space, encases carotid artery or invades mediastinal structures Imaging Checklist: ● ● ● ●
Detailed assessment of spread pattern, with emphasis on deep and superior extension Skull base erosion Involvement of the mandibular division of the fifth cranial nerve Cavernous sinus involvement
4 Extracranial Head and Neck Neoplasms: Role of Imaging
107
The glottic larynx consists of the TVC and anterior and posterior commissure. The subglottic larynx extends from the undersurface of the TVC to the base of the cricoid cartilage. Treatment of laryngeal carcinomas depends on the location and extent of the primary tumor [69, 70].
5.14
Supraglottis
Imaging plays an important role in the staging and treatment of patients with SCCA of the supraglottic larynx. The majority of supraglottic carcinomas are detected by endoscopy. Although mucosal abnormality is usually adequately assessed on direct visualization, deep extent and submucosal involvement are difficult to identify by endoscopy and can be better assessed by cross-sectional imaging [54, 69]. The most commonly performed laryngeal conservation surgery for supraglottic carcinoma is the supraglottic laryngectomy. The surgery entails resection of the false vocal cords, aryepiglottic folds, epiglottis, pre-epiglottic space, superior half of the thyroid cartilage and hyoid bone. The inferior margin of a standard supraglottic laryngectomy is the laryngeal ventricle. Although the arytenoid cartilages are usually spared, one of them may be removed if involved with a tumor. Imaging studies in patients with a supraglottic malignancy should attempt to determine whether the patient would be a candidate for a supraglottic laryngectomy. Contraindications to supraglottic laryngectomy include transglottic extension, cartilage invasion and exolaryngeal tumor spread [Fig. 4.6]. There are some modifications of the surgical approach that allow resection of portions of the thyroid and arytenoids cartilages, but cartilage invasion generally precludes its use. Other factors that contraindicate supraglottic laryngectomy, such as mucosal involvement of the piriform sinus, vocal cord fixation and proximity of the tumor to the circumvallate papillae, are difficult to assess with imaging and can be better evaluated with endoscopy [71]. Supracricoid laryngectomy with cricohyoidopexy is currently performed to resect advanced supraglottic carcinomas that extend to the ventricle, invade the glottis and, in selected cases, invade small areas of the thyroid cartilage. Contraindications to this operation include subglottic spread with invasion of the cricoid cartilage [5, 54, 69, 70]. Other voice-preserving laryngectomy procedures include nearly total laryngectomy, which may be indicated in transglottic lesions with subglottic invasion. This technique allows partial removal of the cricoid cartilage. Total laryngectomy may be necessary as primary treatment for extensive SCCA of the larynx, with invasion of the laryngeal cartilages and the subglottis, for salvage after failed radiotherapy, for focal recurrence after partial laryngectomy and for osteochondroradionecrosis that can occur as a complication of radiation therapy [72]. A growing number of patients with supraglottic carcinoma are now being treated nonsurgically. The results of primary radiation therapy for supraglottic carcinomas have been variable; primary control rates range between 70-90% for T1 and T2
108
M. Petrou and S.K. Mukherji
Fig. 4.6 Large enhancing laryngeal mass noted at the level of the right false vocal cord (a) and extending inferiorly to the level of the true vocal cords (b). Mass extends over the anterior commissure to involve the left side. There is suggestion of exolaryngeal spread through the thyroid cartilage (arrow). A thin rind of abnormal soft tissue extends inferiorly along the right lateral aspect of the subglottic region at the level of the cricoid cartilage, consistent with transglottic extension of neoplasm (c). There is also asymmetric sclerosis of the right aspect of the cricoid cartilage (d)
lesions, and are almost 40 percent for T3 and T4 lesions [72]. Pretreatment CT imaging can be predictive of local control in these patients. Specifically tumor volume, pre-epiglottic spread and cartilage invasion are important issues to address in planning definitive radiotherapy. Tumors with volumes less than 6 ml have an 89 percent chance of local control with radiation treatment alone, as opposed to a 52 percent chance of local control in tumors greater than 6 ml. There does not appear to be a relationship between invasion of the pre-epiglottic space and local control by radiation treatment. There is, however, a decreased likelihood of preserving a functional larynx post radiation with pre-epiglottic space involvement. The presence of cartilage invasion on pretreatment imaging does correlate with an increased incidence of treatment failure [73]. T-staging and important imaging information regarding supraglottic carcinomas is given in Table 4.5.
4 Extracranial Head and Neck Neoplasms: Role of Imaging
109
Table 4.5 TNM Staging System for Supraglottic Laryngeal Carcinoma and important information that can be derived from imaging TNM Staging: T1 Tumor limited to one subsite of the supraglottis with normal vocal cord mobility T2 Tumor invades mucosa of more than one adjacent subsite of supraglottis or glottis or region outside the supraglottis without fixation of the larynx T3 Tumor limited to the larynx with vocal cord fixation and/or invades any of the following: postcricoid area, pre-epiglottic tissues, paraglottic space, and/or minor thyroid cartilage cartilage erosion (eg inner cortex) T4a Tumor invades through the thyroid cartilage, and/or invades tissues beyond the larynx (e.g., trachea, soft tissues of the neck including deep extrinsic muscles of the tongue, strap muscles, thyroid or esophagus) T4b Tumor invades prevertebral space, encases carotid artery or invades mediastinal structures Imaging Checklist: ● ● ● ● ● ●
Tumor Volume Cartilage Invasion Exolaryngeal spread Pre-epiglottic region invasion Transglottic extension Pyriform sinus invasion
5.15
Glottis
The true vocal cords are the most common site of laryngeal carcinomas with the ratio of glottic to supraglottic carcinomas being approximately 3:1. Low-volume tumors that are limited to the midportion of the true vocal cords and do not involve the anterior commissure or the ipsilateral cricoarytenoid joint may be treated with laser cordectomy. The most commonly performed partial laryngectomy procedure to treat tumors of the true vocal cord is the vertical hemilaryngectomy. The standard hemilaryngectomy involves resection of the affected true vocal cord, the ispilateral false cord and the adjacent thyroid cartilage. Modifications of the hemilaryngectomy allow for resection of more advanced tumors. A supracricoid laryngectomy may be used to remove more advanced glottic tumors [5, 54, 70]. Invasion of the anterior commissure, posterior commissure or paraglottic fat preclude treatment by cordectomy. Imaging findings that contraindicate vertical hemilaryngectomy are cartilage invasion, transglottic extension to involve the false vocal cord, extension across the anterior commissure with involvement of more than one-third of the contralateral true vocal cord, extension into the posterior commissure and subglottic extension that is >10 mm anteriorly or >5 mm posteriorly [70]. Subglottic extension contraindicates supracricoid laryngectomy and requires that total laryngectomy be performed. Radiation therapy has gained acceptance as an effective treatment for T1 and T2 carcinomas of the true vocal cords. Reported local control rates are lower for T3 lesions. Tumor volume and the status of the laryngeal cartilages is important information that can be derived from imaging and can contribute to risk assessment.
110
M. Petrou and S.K. Mukherji
Tumor volumes <3.5 ml with no cartilage sclerosis or sclerosis of a single laryngeal cartilage is associated with a 90 percent chance of local control with RT alone. Tumor volumes <3.5 ml with sclerosis of two laryngeal cartilages and tumor volumes >3.5 ml and no/single cartilage sclerosis are associated with a 50 percent chance of local control with RT. Furthermore, tumor volumes >3.5 ml and combined sclerosis of the ispilateral arytenoid and adjacent cricoid cartilage only have a 20 percent chance of successful local control with RT alone [74, 75].
5.16
Subglottis
Primary subglottic tumors are less common than glottic and supraglottic carcinomas. Early lesions are usually treated with RT. An advanced tumor involving the undersurface of the true vocal cords is usually treated with a total laryngectomy. Because the cricoid cartilage is the foundation of the larynx, a voice sparing laryngectomy with a partial cricoidectomy is not a feasible option for low-volume tumors. Imaging can help evaluate the presence of cartilage invasion, determine the inferior extent of the tumor and identify the extent of tracheal resection necessary for adequate margins [5, 54, 70]. T-staging details and important imaging findings in cases of glottic and subglottic carcinomas are shown in Table 4.6.
Table 4.6 TNM Staging System for Glottic and Subglottic Carcinomas and important information that can be derived from imaging Glottic Carcinomas- TNM Staging: T1 Tumor limited to the vocal cords (may involve anterior or posterior commissure) with normal mobility T2 Tumor extends to the supraglottis and/or the subglottis, or impaired vocal cord mobility T3 Tumor limited to the larynx with vocal cord fixation and/or invades paraglottic space and/or minor thyroid cartilage cartilage erosion (e.g., inner cortex) T4a Tumor invades through the thyroid cartilage, and/or invades tissues beyond the larynx. (e.g., trachea, soft tissues of the neck including deep extrinsic muscles of the tongue, strap muscles, thyroid or esophagus) T4b Tumor invades prevertebral space, encases carotid artery or invades mediastinal structures Imaging Checklist: ● ● ● ● ● ● ●
Tumor volume Transglottic extension Cartilage sclerosis/invasion Subglottic extension and relation to the cricoid cartilage Involvement of anterior and posterior commissure Involvement of cricoarytenoid joint Deep invasion of the paraglottic fat (continued)
4 Extracranial Head and Neck Neoplasms: Role of Imaging
111
Table 4.2 (continued) Subglottic Carcinomas-TNM Staging: T1 Tumor limited to the subglottis T2 Tumor extends to the vocal cords with normal or impaired mobility T3 Tumor limited to the larynx with vocal cord fixation T4a Tumor invades cricoid or thyroid cartilage, and/or invades tissues beyond the larynx. (e.g., trachea, soft tissues of the neck including deep extrinsic muscles of the tongue, strap muscles, thyroid or esophagus) T4b Tumor invades prevertebral space, encases carotid artery or invades mediastinal structures Imaging Checklist: ● ●
Inferior extent Cartilage Invasion
5.17
Sinonasal Carcinoma
Carcinomas of the sinonasal cavity constitute approximately 3-4% all head and neck malignancies [76]. Overall, they have a poor prognosis because they may be detected at an advanced stage. Squamous cell carcinoma accounts for approximately 80 percent of all sinonasal malignancies. Minor salivary gland malignancies such as adenocarcinoma and adenoid cystic carcinoma account for approximately 10 percent; these tend to arise in the palate and spread to the nasal cavity and paranasal sinuses. Melanomas can also occur in this location and 10-20% of these are amelanotic. Esthesioneuroblastomas also occur in the upper nasal cavity or ethmoid vault, arising from the olfactory nerves in this location. Other aggressive neoplasms involving the sinonasal cavity include ameloblastomas, sarcomas and lymphoma; these neoplasms are rare and only constitute a very small fraction of sinonasal malignancies [77]. Sinonasal malignancies usually spread by direct extension and/or perineural involvement [78-80]. Imaging can play an important role in defining the anatomic extent of disease and is an invaluable part of pre-surgical planning. The superior and posterior boundaries of the maxillary sinuses are the most important in terms of determining surgical management [80]. Direct extension superiorly into the orbit and intracranial spread via the ethmoid air cells makes obtaining tumor free margins difficult. Posterior extension by either direct invasion or perineural spread can result in involvement of the masticator space, infratemporal fossa, orbit and intracranial compartment, and can preclude curative surgery. The inferior and medial margins of the maxillary sinuses are, on the other hand, more readily resected en bloc [77]. Tumors involving the ethmoid air cells can cross the fovea ethmoidalis and cribriform plate superiorly into the intracranial compartment. In this case a craniofacial resection is required, usually involving both the expertise of a neurosurgeon as well as a head and neck surgeon [81, 82]. Lateral extension of ethmoid neoplasms across the thin lamina papyracea may result in intraorbital spread and usually requires orbital exenteration for clean margins [83-85]. Tumors arising in the sphenoid sinus can be difficult to completely resect because of the surrounding vital structures.
112
M. Petrou and S.K. Mukherji
CT and MRI are complementary in the assessment and staging of sinonasal neoplasms and both should preferably be acquired prior to any surgical intervention, including biopsy [77-79, 85, 86]. CT is more sensitive and accurate in assessing the osseous margins of the sinonasal cavity, the osseous floor of the anterior cranial fossa and the walls of the orbit [78]. Osseous destruction is seen most commonly with carcinomas. Pre- and post-contrast MRI can be particularly helpful in assessing tumor extension outside the sinonasal cavity [85-88]. Potential areas of tumor extension that must be assessed in all patients with sinonasal malignancies are intracranial spread (anterior and middle cranial fossae), and spread to the palate, orbits, pterygopalatine fossa and skull base. Lymph node metastases in adults are most often found in the internal jugular and submandibular nodes [77]. One of the challenges for the radiologist in assessing tumor extension is distinguishing tumors from co-existing inflammatory change. This is best accomplished with pre- and post-contrast MRI, with fat suppression. Inflammatory secretions frequently have high water content and, as a result, high T2 signal; they also demonstrate peripheral rim enhancement. Most sinonasal tumors are highly cellular and, as a result, have low to intermediate T2 signal and a more solid pattern of enhancement. However, the protein and water content, as well as viscosity characteristics, can lead to a complex appearance of inflammatory changes on MRI [89, 90]. Sinonasal masses that frequently invade the skull base and result in intracranial spread include carcinomas, esthesioneuroblastoma, lymphoma and sarcomas. Benign lesions such as inverted papillomas, polyps and mucoceles can also invade the skull base with very similar imaging findings of osseous destruction. CT may detect cortical erosion, but MRI is probably more sensitive in assessing skull base invasion. Low bone marrow signal is seen within the bone marrow on unenhanced T1-weighted images. Low T1 signal can also be seen with edema or hematopoietic marrow and should, therefore, be interpreted in conjunction with T2 and enhancement characteristics [91-94]. Intracranial spread is also best evaluated with MRI. Imaging findings that suggest malignant involvement of the dura include the presence of discontinuous dural enhancement, regions of thickening or nodularity greater than 5 mm and presence of high T2 signal within the adjacent brain parenchyma [85, 87]. CT and MRI can both be used in assessing orbital spread of tumors (Fig. 4.7). Osseous destruction with involvement of the orbital fat, manifesting as stranding on imaging, has been used as one of the features suggestive of orbital invasion. Although orbital fat stranding has a high positive predictive value regarding orbital invasion, its absence does not exclude orbital involvement. Orbital fat stranding is better appreciated on CT, as compared to MRI, which as a result tends to underestimate orbital involvement. Other criteria used to ascertain orbital involvement include the relation between the tumor and the periorbital structures (e.g., the periosteum of the bones comprising the orbit), the overall integrity of the osseous structures surrounding the tumor and the appearance of the extraocular muscles (enlargement, displacement, signal abnormality). None of these criteria are, however, particularly accurate, and an intra-operative assessment with histology on a frozen section is often used to make a definitive diagnosis [77, 87, 88].
4 Extracranial Head and Neck Neoplasms: Role of Imaging
113
Fig. 4.7 (a) and (b) Aggressive right ethmoidal soft tissue mass, destroying the ethmoid septa and extending to the right orbit. Patient was treated with extensive surgery including a right orbital exenteration
References 1. Katsantonis GP, Archer CR, Rosenblum BN et al. The degree to which accuracy of preoperative staging of laryngeal carcinoma has been enhanced by computed tomography. Otolaryngol Head Neck Surg.1986; 95:52-62. 2. Charlin B, Brazeau-Lamontagne L, Guerrier B, et al. Assessment of laryngeal cancer: CT scan versus endoscopy. J Otolaryngol.1989; 18:283-8. 3. Sulfaro S, Barzan L, Querin F et al. T staging of the laryngohypopharyngeal carcinoma. A 7year multidisciplinary experience. Arch. Otolaryngol Head Neck Surg. 1989; 115:613-20 4. Benchaou M, Lehmann W, Slosman DO et al. The role of FDG-PET in the preoperative assessment of N-staging in head and neck cancer. Acta Orolarygol 1996; 116:332-35. 5. Mukherji, SK, Pillsbury H, Castillo M. Imaging squamous cell carcinoma of the upper aerodigestive tract: What the clinicians need to know. Radiology 1997; 205:629-646. 6. Zbaren P, Becker M, Lang H. Pretherapeutic staging of hypopharyngeal carcinoma. Clinical findings, computed tomography, and magnetic resonance imaging compared with histopathologic evaluation. Arch Otolaryngol Head Neck Surg. 1997; 123:908-13. 7. Bruschini P, Georgetti A, Bruschini L, et al. Positron emission tomography (PET) in the staging of head and neck cancer: comparison between PET and CT. Acta Otorhinolaryngol Ital 2003; 23:446-53. 8. Branstetter BF 4th, Blodgett TM, Zimmer LA et al. Head and neck malignancy: is PET/CT more accurate than PET or CT alone? Radiology 2005; 234: 580-86. 9. Mukherji SK, Bradford CR. Controversies: is there a role for positron-emission tomographic CT in the initial staging of head and neck squamous cell carcinoma? AJNR Am J Neuroradiol. 2006; 27:243-5. 10. Ng SH, Chang TC, Ko SF, et al. Nasopharyngeal Carcinoma. CT and MR assessment. Neuroradiology 1997; 39:741-6. 11. Ginsberg LE. MR imaging of perineural tumor spread. Magn. Res Imaging Clin N Am 2002; 10:511-25. 12. Gandhi D, Gujar S, Mukherji SK. Magnetic resonance imaging of perineural spread of head and neck malignancies. Top Magn Reson Imaging 2004; 15:79-85. 13. Fakhry N, Lussato D, Jacob T. Comparison between PET and PET/CT in recurrent head and neck cancer and clinical implications. Eur Arch Otorhinolaryngol. 2007; 264:531-8.
114
M. Petrou and S.K. Mukherji
14. Di Martino E, Nowak B, Hassan HA et al. Diagnosis and staging of head and neck cancer: a comparison of modern imaging modalities (positron emission tomography, computed tomography, color-coded duplex sonography) with panendoscopic and histopathologic findings. Arch Otolaryngol Head Neck Surg. 2000; 126:1457-61. 15. Murukami R, Uozumi H, Hirai T et al. Impact of FDG-PET/CT Imaging on Nodal Staging for Head-And-Neck Squamous Cell Carcinoma. Int J Radiat Oncol Biol Phys. 2007; 68:377-82. 16. Sheikholeslam-Zadeh R, Choufani G, Goldman S, et al. Unknown primary detected by FDGPET. A review of the present indications of FDG-PET in head and neck cancers. Acta Otorhinolaryngol Belg. 2002; 56:77-82. 17. Delgado-Bolton RC, Fernandez-Perez C, Gonzalez-Mate A, et al. Meta-analysis of the performance of 18F-FDG PET in primary tumor detection in unknown primary tumors. J Nucl Med. 2003; 44:1301-14. 18. Gutzeit A, Antoch G, Kuhl H, et al. Unknown primary tumors: detection with dual-modality PET/CT–initial experience. Radiology 2005; 234:227-34. 19. Greene FL, Page DL, Felming ID, eds. AJCC Cancer Staging Manual. 6th ed. Philadelphia: Lippincott Raven; 2002. 20. Yousem DM, Gad K, Tufano RP. Resectability Issues with Head and Neck Cancer. AJNR Am J Neuroradiol. 2006; 27:2024-36. 21. Yousem DM, Hatabu H, Hurst RW et al. Carotid artery invasion by head and neck masses: prediction with MR imaging. Radiology 1995; 195:715-20. 22. Hsu WC, Loevner LA, Karpati R, et al. Accuracy of magnetic resonance imaging in predicting absence of fixation of head and neck cancers to the prevertebral space. Head Neck 2005; 27:95-100. 23. Loevner LA, Ott IL, Yousem DM et al. Neoplastic fixation to the prevertebral compartment by squamous cell carcinoma of the head and neck. AJR Am J Roentgenol 1998; 170:1389-94. 24. Wang JC, Takashima S, Takayama F, et al. Tracheal invasion by thyroid carcinoma: prediction using MR imaging. AJR Am J Roentgenol 2001; 177:929-36. 25. Roychowdhury S, Loevner LA, Yousem DM et al. MR imaging for predicting neoplastic infiltration of the cervical esophagus. AJNR Am J Neuroradiol 2000; 21:1681-87. 26. Shanmugaratnam K, Chan SH, de-The G: Histopathology of nasopharyngeal carcinoma: Correlations with epidemiology, survival rates and other biological characteristics. Cancer 1979; 44:1029-44 27. Chua DTT, Sham JST, Kwong DLW, et al. Treatment outcome after radiotherapy alone for patients with stage I-II nasopharyngeal carcinoma. Cancer 2003; 98:74-80. 28. Huncharek M, Kupelnick B. Combined chemoradiation versus radiation therapy alone in locally advanced nasopharyngeal carcinoma. Am J Clin Oncol 2002; 25:219-223. 29. Baujat B, Audry H, Jean Bourhis J et al. Chemotherapy in locally advanced nasopharyngeal carcinoma: an individual patient data meta-analysis of eight randomized trials and 1753 patients. Int J Radiation Oncology Biol Phys 2003; 57: 481-488. 30. Langendijk JA, Leemans R, Buter J et al. The additional value of chemotherapy to radiotherapy in locally advanced nasopharyngeal carcinoma: a meta-analysis of the published literature. J Clin Oncol 2004; 22: 4604-4612. 31. Ma BBY, Chan ATC. Recent perspectives in the role of chemotherapy in the management of advanced nasopharyngeal carcinoma. Cancer 2005; 103: 22-31. 32. Sham JS, Wei WI, Zong YS et al. Detection of subclinical nasopharyngeal carcinoma by fiberoptic endoscopy and multiple biopsies. Lancet 1990; 335:371-4. 33. Hsu MM, Young YH, Lin KL. Eustachian tube function of patients with nasopharyngeal carcinoma. Ann Otol Rhinol Laryngol 1995; 104: 453-5. 34. Sham JS, Wei WI, Lau SK et al. Serous otitis media and paranasopharyngeal extension of nasopharyngeal carcinoma. Head Neck 1992; 14:19-23. 35. Sham JS, Cheung YK, Choy D, et al. Nasopharyngeal carcinoma: CT evaluation of patterns of tumor spread. AJNR Am J Neuroradiol 1991;12:265-70.
4 Extracranial Head and Neck Neoplasms: Role of Imaging
115
36. Chong VF, Fan YF, Toh KH et al. MRI and CT features of nasopharyngeal carcinoma with maxillary sinus involvement. Australas Radiol 1995; 39:2-9 37. Chong VF, Fan YF. Maxillary nerve involvement in nasopharyngeal carcinoma. AJR Am J Roentgenol 1996; 167:1309-12 38. Chong VF, Fan YF, Mukherji SK. Carcinoma of the nasopharynx. Semin Ulrtrasound CT MR 1998; 19: 449-62. 39.Chong VF, Fan YF, Mukherji SK. Malignancies of the nasopharynx and the skull base. In Oncologic Imaging (2nd ed) Bragg DG, Rubin P, Hricak H (Eds). Philadelphia W.B. Saunders 2002; 202-231. 40. Chong VF, Fan YF, Khoo JB. MRI features of cervical nodal necrosis in metastatic disease. Clin Radiol 1996; 51:103-9. 41. Sham JS, Choy D, Wei WI. Nasopharyngeal carcinoma: Orderly neck node spread. Int J Radiat Oncol Biol Phys 1990; 19:929-33. 42. Chong VF, Fan YF, Khoo JB. Retropharyngeal lymphadenopathy in nasopharyngeal carcinoma. Eur J Radiol 1995; 21:100-5. 43. Hollinshead WH: Textbook of Anatomy. Hagerstown, MD, Harper & Row, 928-948, 1974. 44. Mukherji SK, Castillo M. Normal cross-sectional anatomy of the nasopharynx, oropharynx and oral cavity. Neuroimaging Clin N Am 1998; 8:211-8. 45. Stambuck HE, Karimi S, Lee N, et al. Oral cavity and oropharynx tumors. Radiol Clin North Am. 2007;45:1-20 46. Million RA, Cassisi NJ, Mancuso AA: Oral Cavity. In Million RR, Cassisi NJ (eds): Management of Head and Neck Cancer: A multidisciplinary approach. Philadelphia, JB Lippincott, 599-626, 1994. 47. Wang CC. Radiation Therapy for head and neck neoplasms. 3rd edition. New York: WileyLiss; 1997. 48. Sessions DG, Spector GJ, Lenox J, et al. Analysis of treatment results for oral tongue cancer. Laryngoscope 2002;112: 616-25. 49. O’Brien CJ, Adams JR, McNeil EB et al. Influence of bone invasion and extent of mandibular resection on local control of cancers of the oral cavity and oropharynx. Int J Oral Maxillofac Surg. 2003; 32: 492-7. 50. Spiro R.H., Huvos A.G., Wong G.Y., et al: Predictive value of tumor thickness in squamous carcinoma confined to the tongue and floor of the mouth. Am J Surg 1986; 152: 345-350. 51. Mukherji SK, Isaacs DL, Creager A et al. CT detection of mandibular invasion by squamous cell carcinoma of the oral cavity. AJR Am J Roentgenol. 2001; 177:237-43. 52. Mukherji SK, Weeks S, Castillo M. Squamous cell carcinomas that arise in the oral cavity and tongue base: Can CT help predict perineural or vascular invasion? Radiology 1996; 198:157-62. 53. Genden EM, Rinardo A, Jacobson A et al. Management of mandibular invasion: when is a marginal mandibulectomy appropriate? Oral Oncol 2005; 41:776-82. 54. Mukherji SK, Fatterpekar G, Chong VF. Malignancies of the oral cavity, oropharynx and hypopharynx. In Oncologic Imaging (2nd ed) Bragg DG, Rubin P, Hricak H (Eds). Philadelphia W.B. Saunders 2002; 202-231. 55. Weber AL, Romo L, Hashmi S. Malignant tumors of the oral cavity and oropharynx:clinical, pathologic and radiologic evaluation. Neuroimag Clin N Am 2003;13: 443-464. 56. Ginsberg LE, DeMonte F. Imaging of perineural tumor spread from palatal carcinoma. AJNR Am J Neuroradiol 1998;22: 1417-22. 57. Million RA, Cassisi NJ, Mancuso AA et al: Oropharynx. In Million RR, Cassisi NJ (eds): Management of Head and Neck Cancer: A multidisciplinary approach. Philadelphia, JB Lippincott, 599-626, 1994. 58. Mendenhall WM, Morris CG, Amdur RJ et al. Definitive radiotherapy for tonsillar squamous cell carcinoma.Am J Clin Oncol. 2006;29: 290-7. 59. Shirazi HA, Sivanandan R, Goode R, et al. Advanced-staged tonsillar squamous carcinoma: organ preservation versus surgical management of the primary site. Head Neck 2006; 28: 587-94.
116
M. Petrou and S.K. Mukherji
60. Poulsen M, Porceddu SV, Kingsley PA. Locally advanced tonsillar squamous cell carcinoma: Treatment approach revisited. Laryngoscope 2007; 117: 45-50. 61. Merati AL, Rieder AA. Normal endoscopic anatomy of the pharynx and larynx. Am J Med. 2003; 115 Suppl 3A:10S-14S. 62. Saleh E, Mancuso AA, Stringer S. Relative roles of computed tomography and endoscopy for determining the inferior extent of pyriform sinus carcinoma: correlative histopathologic study. Head Neck. 1993; 15: 44-52. 63. Wenig BL, Ziffra KL, Mafee MF, et al. MR imaging of squamous cell carcinoma of the larynx and hypopharynx. Otolaryngol Clin North Am. 1995;28: 609-619. 64. Becker M. Larynx and hypopharynx. Radiol Clin North Am. 1998; 36:891-920. 65. Schmalfuss IM. Imaging of the hypopharynx and cervical esophagus. Magn Reson Imaging Clin N Am. 2002; 10: 495-509. 66. Pameijer FA, Mancuso AA, Mendenhall WM, et al. Evaluation of pretreatment computed tomography as a predictor of local control in T1/T2 pyriform sinus carcinoma treated with definitive radiotherapy. Head Neck. 1998; 20:159-168. 67. Keberle M, Hoppe F, Dotzel S et al. Tumor volume as determined by computed tomography predicts local control in hypopharyngeal squamous cell carcinoma treated with primary surgery. Eur Radiol 2004;14:286-91. 68. Million RR. Pharyngeal walls, pyriform sinus, postcricoid pharynx. In: Million RR, ed. Management of head and neck cancer. Philadelphia: JB Lippincott 1994:502-532. 69. Yousem DM, Tufano RP. Laryngeal Imaging. Magn Reson Imaging Clin N Am. 2002; 10: 451-465. 70. Mukherji SK, Becker M. Carcinoma of the larynx. In Oncologic Imaging (2nd ed) Bragg DG, Rubin P, Hricak H (Eds). Philadelphia W.B. Saunders 2002; 233-61. 71. Zbaren P, Becker M, Laeng H. Staging of laryngeal cancer: Endoscopy, computed tomography and magnetic resonance imaging versus histopathology. Eur Arch Otolaryngol 1997; 254:117-122. 72. Gallo A, Mocetti O, De Vincentiis M, et al. Neoplastic infiltration of laryngeal cartilages: Histocytochemical study. Laryngoscope 1992; 102: 891-5. 73. Freeman DE, Mancuso AA, Parsons JT, et al. Irradiation alone for supraglottic larynx carcinoma: can CT findings predict treatment results?. Int J Radiat Oncol Biol Phys. 1990;19: 485-490. 74. Mukherji SK, Mancuso AA, Mendenhall Wm et al. Can pretreatment CT predict local control of T2 glottic carcinomas treated with radiation therapy alone? AJNR Am J Neuroradiol 1995; 16:655-662. 75. Pameijer FA, Mancuso AA, Mendenhall WM, et al. Can pretreatment computed tomography predict local control in T3 squamous cell carcinoma of the glottic larynx treated with definitive radiotherapy? Int J Radiat Oncol Biol Phys 1997; 37: 1011-1021. 76. Muir C, Weiland L. Upper aerodigestive tract cancers. Cancer 1995; 75(suppl1): 147-53. 77. Loevner LA. Paranasal Sinus Neoplasms. In Oncologic Imaging (2nd ed) Bragg DG, Rubin P, Hricak H (Eds). Philadelphia W.B. Saunders 2002; 160-81. 78. Curtin HD, Williams R, Johnson J. CT of perineural tumor extension: pterygopalatine fossa. AJNR Am J Neuroradiol. 1984; 5:731-737. 79. Kraus DH, Lanzieri CF, Wanamaker JR, et al. Complementary use of computed tomography and magnetic resonance imaging in assessing skull base lesions. Laryngoscope. 1992; 102:623-629. 80. Maroldi R, Farina D, Battaglia D et al. MR of malignant nasosinusal neoplasms: Frequently asked questions. Eur J Radiol 1997; 24:181-190. 81. Van Tuyl R, Gissack GS. Prognostic factors in craniofacial surgery. Laryngoscope 1991; 101:240-44. 82. Osguthorpe JD, Patel S. Craniofacial approaches to sinus malignancy. Otolaryngol Clin North Am 1995; 28:1239-1257. 83. Perry C, Levine PA, Williamson BR, et al. Preservation of the eye in paranasal sinus cancer surgery. Arch Otolaryngol Head Neck Surg. 1988; 114: 632-634.
4 Extracranial Head and Neck Neoplasms: Role of Imaging
117
84. McCary WS, Levine PA. Management of the eye in the treatment of sinonasal cancers. Otolaryngol Clin North Am. 1995; 28: 1231-1238 85. Hermans R, De Vuysere S, Marchal G. Squamous cell carcinoma of the sinonasal cavities [Review] Semin Ultrasound CT MR 1999; 20:15086. Loevner LA, Sonners AI. Imaging of neoplasms of paranasal sinuses. Magn Reson Imaging Clin N Am. 2002;10:467-93. 87. Eisen MD, Yousem DM, Montone KT, et al. Use of preoperative MR to predict dural, perineural, and venous sinus invasion of skull base tumors. AJNR Am J Neuroradiol 1996; 17:1937-45. 88. Eisen MD, Yousem DM, Loevner LA et al. Preoperative imaging to predict orbital invasion by tumor. Head Neck 2000; 22:456-62. 89. Som PM, Shapiro MD, Biller Hf et al. Sinonasal tumors and inflammatory tissues: Differentiation with MR imaging. Radiology 1988; 167:803-8. 90. Hasso AN, Lambert D. Magnetic resonance imaging of the paranasal sinuses and nasal cavities.Top Magn Reson Imaging 1994; 6:209-23. 91. Daffner RH, Lupetin AR, Dash N et al. MRI in the detection of malignant infiltration of bone marrow. AJR Am J Roentgenol 1986;146: 353-8. 92. Kimura F, Kim KS, Friedman H, et al. MR imaging of the normal and abnormal clivus. AJR Am J Roentgenol 1990; 11:1015-21 93. Ricci C, Cova M, Kang YS et al. Normal age-related patterns of cellular and fatty bone marrow distribution in the axial skeleton: MR imaging study. Radiology 1990; 177: 83-8. 94. Poulton TB, Murphy WD, Duek JL et al. Bone marrow reconversion in adults who are smokers: MR imaging findings. AJR Am J Roentgenol 1993; 161:1217-21.
5
Imaging of Thoracic Malignancies Subba R. Digumarthy1, MD and Suzanne L. Aquino2, MD
Technological advances have revolutionized the radiologic imaging of malignancy with tremendous improvements in disease localization and characterization. There has been a great emphasis on extensive research and its clinical application in oncologic imaging. Thoracic malignancies form an important group of malignancies involving the body. In this chapter, we review the current concepts in imaging of primary and metastatic thoracic malignancies.
1
Bronchogenic Carcinoma
Bronchogenic carcinoma is the most frequently occurring cancer in the world, with an incidence of 1.2 million cases in 2000. [1] In the United States lung cancer had an estimated incidence of 173,770 and a mortality of 160,440 in 2004. [1] Cigarette smoking is the most important risk factor in its development, but exposures including radon and asbestos, have also been associated with lung cancer. [2, 3] Patients with bronchogenic carcinoma may present with cough, hemoptysis, dyspnea or systemic manifestations such as fatigue, weight loss or fever. Ten percent of patients are asymptomatic and 40 percent of patients have advanced stage disease at initial presentation. [4] The primary tumor of bronchogenic carcinoma may manifest as a solitary pulmonary nodule, mass or an area of consolidation. Proximal tumors in the airway frequently lead to bronchial obstruction with atelectasis, which may be segmental, lobar or involve an entire lung. On occasion, a post-obstructive pneumonia results and is the cause for a patient’s initial symptoms. (Fig. 5.1)
1 Assistant Radiologist, Massachusetts General Hospital, Instructor in Radiology, Harvard Medical School 2 Associate Radiologist, Massachusetts General Hospital, Associate Professor of Radiology, Harvard Medical School
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
121
122
S.R. Digumarthy and S.L. Aquino
Table 5.1 Classification of thoracic malignancies Tumors of Lung & Bronchi Bronchogenic carcinoma Non-small cell carcinoma (Squamous cell carcinoma- including Pancoast tumor, Adenocarcinoma, its subtype of Bronchioloalveolar cell Carcinoma, Large cell undifferentiated carcinoma and Large cell neuroendocrine carcinoma) Small cell carcinoma Carcinoid Lymphoma- Hodgkin’s Disease and Non-Hodgkins Lymphoma Metastases- Endobronchial, Pulmonary, Lymphangitic Carcinomatosis, Nodal metastases Tumors of Pleura Malignant Mesothelioma Metastases
Fig. 5.1 Post obstructive changes. Axial CT scan of endobronchial squamous cell cancer of the left lower lobe causing distal collapse and lung abscess
A solitary pulmonary nodule (SPN) is an opacity less than 3 cm in diameter, surrounded by air-containing lung and without associated lymphadenopathy or atelectasis. An opacity that is 3 cm or greater in diameter is categorized as a mass. A nodule or mass may be round or oval, with smooth, lobulated, irregular or spiculated margins. Eighty-four to 90 percent of nodules with spiculated margins are malignant. [5, 6] With High Resolution CT (HRCT) imaging a SPN may be optimally classified as ground glass, solid or mixed. [7, 8] Mixed solid/ground glass and ground glass nodules have a higher rate of malignancy than solid nodules on screening CT [9]. Kim, et al. quantified the ground glass attenuation of pulmonary nodules on CT and found a greater extent of ground glass opacity in those with bronchioloalveolar cell carcinoma (BAC) histology. These tumors tend to have a more favorable prognosis. [10] They may be treated with wedge resection, rather than the usual lobectomy performed for solid nodules. [10, 11] Air-space consolidation, as a primary manifestation of lung cancer, is most characteristic of adenocarcinomas, particularly BAC cell type (Fig. 5.2). The tumor consolidation may be peripheral,
5 Imaging of Thoracic Malignancies
123
Fig. 5.2 Multifocal bronchoalveolar carcinoma. Axial CT of multifocal consolidation containing air bronchograms
segmental or lobar and contain an air bronchogram. On CT a pattern of peripheral consolidation, with stretching and narrowing of bronchi, widening of the branching angle of bronchi, bulging of the interlobar fissure, as well as co-existing pulmonary nodules, in the same or different lobe should suggest BAC. [12, 13] In a study by Jung, et al., mucus-filled bronchi were associated with pneumonia and not BAC. [13] The CT-angiogram sign of enhancing pulmonary vessels in an area of consolidation is non-specific and may occur in post-obstructive pneumonitis and pneumonia, as well as BAC. [14, 15] The overall prognosis is not significantly different from that for the other types of non-small cell lung cancer [16]. The consolidative form of BAC has a 26 percent five-year survival rate, compared to 39 percent of focal nodular BAC. [17] Squamous cell carcinoma frequently develops centrally near the hila and major airways. Common features on CT include central cavitation and/or necrosis. Adenocarcinoma of the lung tends to develop in the periphery of the lung and/or in preexisting scars (e.g., scar carcinoma). Ground glass nodules and uni- or multifocal consolidation are more often associated with the BAC subtype of adenocarcinoma. Pancoast tumor, also called superior sulcus tumor, constitutes less than 5 percent of all lung cancers. (Fig. 5.3) These tumors arise in the lung apex and tend to directly invade the parietal pleura, chest wall, lower trunks of brachial plexus and sympathetic chain, ribs and spine. Neurologic involvement leads to C8-T1 radiculopathy and/or Horner’s syndrome. MRI, including MR angiography, provides the most accurate radiologic assessment of tumor invasion into the chest wall, vessels, brachial plexus and spine [18-20].
124
S.R. Digumarthy and S.L. Aquino
b
a
c Fig. 5.3 Pancoast tumor. (a) Axial CT scan of right superior sulcus tumor invading the bracial plexus. (b) Axial CT scan of circumferential tumor at the right apex. (c) PET scan with increased 18 FDG uptake corresponding to tumor
Table 5.2 TNM Staging of Lung Cancer [22] TUMOR T1 T2 N0 IA IB N1 IIA IIB N2 IIIA IIIA N3 IIIB IIIB Highlighted Stages- Surgically resectable NODES
2
T3 IIB IIIA IIIA IIIB
T4 IIIB IIIB IIIB IIIB
IV= M1
Staging
The TNM (Tumor, Node, Metastasis) system for staging and classifying Non-Small cell Lung Carcinoma (NSLC) has been revised to provide greater specificity for identifying patient groups with similar prognosis, and has been adopted by the American Joint Committee on Cancer and the Union Internationale Contre le Cancer [21] [Tables 5.2 and 5.3]. Table 5.4 depicts the five-year survival rates for the different TNM Stages.
Table 5.3 TNM Descriptors [21] Primary tumor (T) TX Primary tumor cannot be assessed, or tumor proven by the presence of malignant cells in sputum or bronchial washings, but not visualized by imaging or bronchoscopy T0 No evidence of primary tumor Tis Carcinoma in situ T1 Tumor £ 3 cm in greatest dimension, surrounded by lung or visceral pleura, without bronchoscopic evidence of invasion more proximal than the lobar bronchus* (e.g., not in the main bronchus) T2 Tumor with any of the following features of size or extent: > 3 cm in greatest dimension Involves main bronchus, 2 cm distal to the carina Invades the visceral pleura Associated with atelectasis or obstructive pneumonitis that extends to the hilar region, but does not involve the entire lung T3 Tumor of any size that directly invades any of the following: chest wall (including superior sulcus tumors), diaphragm, mediastinal pleura, parietal pericardium; or tumor in the main bronchus, < 2 cm distal to the carina, but without involvement of the carina; or associated atelectasis or obstructive pneumonitis of the entire lung T4 Tumor of any size that invades any of the following: mediastinum, heart, great vessels, trachea, esophagus, vertebral body, carina; or tumor with a malignant pleural or pericardial effusion,† or with satellite tumor nodule(s) within the ipsilateral primary-tumor lobe of the lung Regional lymph nodes (N) NX Regional lymph nodes cannot be assessed N0 No regional lymph node metastasis N1 Metastasis to ipsilateral peribronchial and/or ipsilateral hilar lymph nodes, and intrapulmonary nodes involved by direct extension of the primary tumor N2 Metastasis to ipsilateral mediastinal and/or subcarinal lymph node(s) N3 Metastasis to contralateral mediastinal, contralateral hilar, ipsilateral or contralateral scalene or supraclavicular lymph node(s) Distant metastasis (M) MX Presence of distant metastasis cannot be assessed M0 No distant metastasis M1 Distant metastasis present‡ * The uncommon superficial tumor of any size with its invasive component limited to the bronchial wall, which may extend proximal to the main bronchus, is also classified T1. † Most pleural effusions associated with lung cancer are due to tumor. However, there are a few patients in whom multiple cytopathologic examinations of pleural fluid show no tumor. In these cases, the fluid is non-bloody and is not an exudate. When these elements and clinical judgment indictate that the effusion is not related to the tumor, the effusion should be excluded as a staging element and the patient’s disease should be staged T1, T2, or T3. Pericardial effusion is classified according to the same rules. ‡ Separate metastatic tumor nodule(s) in the ipsilateral non-primary-tumor lobe(s) of the lung also are classified M1. Table 5.4 Five-year survival rates [21] IA 61 percent IB 38 percent IIA 34 percent IIB 23 percent IIIA 9-13 percent IIIB 3-7 percent IV 1 percent
126
3 3.1
S.R. Digumarthy and S.L. Aquino
Imaging in Lung Cancer Staging Primary Tumor
According to Swensen, et al., contrast enhanced CT is 85 percent accurate in detecting the presence of malignancy in SPNs. According to their group, they found that enhancement of a SPN of more than 20 HU was sensitive in predicting malignancy, while enhancement of less than 15 HU was more characteristic of a benign lesion [23]. The assessment of a tumor’s involvement of the visceral pleural invasion is difficult, especially in the absence of a pleural effusion or pneumothorax. Although contrast-enhanced CT has excellent anatomic resolution, its usefulness in detecting malignancy in a pleural effusion is limited with an accuracy of only 50 percent. MRI has been reported to have a better accuracy of 91 percent in detecting parietal pleural invasion, where the involved pleura will display an increased signal intensity similar to the tumor on T1-weighted images [24]. T2 and Gadolinium enhanced T1-weighted images provide no additional information [24]. Chest wall invasion may be assessed on CT or MRI (Fig. 5.4). On CT, rib destruction, contact of ³ 5 cm with chest wall, obtuse angle (>90 degrees) of mass with chest wall, increased attenuation of subpleural fat plane and presence of tumor
Fig. 5.4 Chest wall invasion. Coronal CT of large cell carcinoma of right lung extending into the lower right chest wall
5 Imaging of Thoracic Malignancies
127
in deeper tissues of chest wall indicate invasion [25-29]. Thin section CT with 1 mm slice thickness reconstruction improves the accuracy in detecting chest wall invasion [30]. MRI with STIR images has been shown to be useful in clarifying the absence of chest wall invasion, when compared to Gadolinium-DTPA enhanced T1-weighted images [31]. Shiotani, et al. compared Breathing Dynamic Echo Planar (BDEPI) MRI with thin-section CT and conventional MRI and concluded that BDEPI and conventional MRI have a 90 percent specificity, in contrast to CT at 50 percent, in excluding chest wall invasion [31].
3.2
Lymph Nodes
Lymph nodes in the short-axis greater than 1 cm are considered to be abnormal on thoracic CT studies. However, this morphologic criteria leads to false positive and false negative results. Frequently, lymph nodes in the thorax are enlarged due to reactive hyperplasia. In contra-distinction, lymph nodes with early metastatic invasion may still be within 1 cm in size on CT imaging. 18-FDG-PET has been reported to be more accurate in detecting malignancy in lymph nodes with a reported sensitivity and specificity of 85 percent and 90 percent respectively, compared to 61 percent and 79 percent of CT [32]. False positive results on FDG-PET may occur from acute inflammation, infection or granulomatous disease. False negative results on lymph node staging by FDG-PET may occur due to inaccurate localization of lymph node anatomy, especially involving the subcarinal and aortopulmonary window nodes, in the absence of accompanying CT scan information [33]. This pitfall is largely overcome with combined PET/CT imaging [34, 35]. Because of the risk for falsely upstaging any patient with lung cancer based on radiologic imaging, most studies recommend that lymph nodes that are positive on FDG-PET scan, or >1 cm in short axis on CT, should be biopsied [36].
3.3
Metastatic Disease
Metastases from lung cancer commonly spread to the brain, bone, liver and adrenal glands [37] (Fig. 5.5). Adenocarcinoma, large cell carcinoma and tumors with advanced local disease are more likely to have metastatic disease [38, 39]. Brain metastasis in patients who are asymptomatic has been reported in up to 15 percent of patients at initial diagnosis. Gadolinium-enhanced MRI is superior to CT in the detection of brain metastases [40]. Liver metastases occur in 2.3 percent to 16 percent of patients with lung cancer at the time of diagnosis [37, 41]. They may be seen as areas of decreased attenuation, of varying sizes, with irregular and peripheral enhancement. MRI with liverspecific contrast agents, improves the detection of hepatic metastases [42]. In a study by Danet, et al. on the MRI appearance of untreated liver metastases, all the hepatic metastatic lesions from primary lung cancer were hypovascular, with faint
128
S.R. Digumarthy and S.L. Aquino
Fig. 5.5 Rib metastasis. Axial CT of expansile osteolytic metastasis of posterior left rib
or negligible peripheral ring enhancement during the arterial phase, and negligible enhancement during the portal phase [43]. Accurate assessment of small metastases on FDG-PET is limited due to the heterogeneous background activity in the liver and, therefore, MRI is more useful in evaluating such lesions [44]. Adrenal glands should always be included on CT of the chest when evaluating a pulmonary nodule or mass. Lesions greater than 10 HU in density, and with less contrast washout on delayed images (less than 60 percent at 15 minutes), should be considered suspicious for malignancy [45, 46]. These should be biopsied if there is no evidence for metastatic disease elsewhere. Though the presence of isolated adrenal metastasis confers a higher stage, surgical resection of the adrenal lesion may still be beneficial and improve long-term survival [47] (Fig. 5.6). FDG-PET imaging is more sensitive than CT in the detection of extrathoracic metastases. Up to 12 percent of patients with lung cancer were found to have metastases on PET imaging, compared to conventional staging methods [48, 49]. Therefore whole-body FDG-PET scan for staging NSLC still may play a significant role in initial staging of patients with lung cancer, despite the absence of metastatic disease on other imaging modalities [50] (Table 5.2).
5 Imaging of Thoracic Malignancies
129
Fig. 5.6 Adrenal metastasis. Axial CT of bilateral bulky adrenal metastases from adenocarcinoma of lung
Table 5.5 The role of imaging in staging NSLC Imaging in Staging NSLC CECT: >20 HU of enhancement characteristic of malignancy. Better visualization of small endobronchial lesions Bronchial obstruction resulting in atelectasis/pneumonitis Mediastinal lymph node assessment Identification of pleural metastases, with pleural effusion Chest wall invasion (better seen on thin multiplanar reconstructions) MRI: Visceral pleural invasion Chest wall invasion (better seen on STIR images) Exclude tumor fixation to pleura/chest wall (Breathing Dynamic Echo Planar MRI) PET: No metastatic disease on other studies
130
4
S.R. Digumarthy and S.L. Aquino
Small Cell Lung Carcinoma
Small Cell Lung Carcinoma (SCLC) accounts for 20 percent to 25 percent of lung cancer, with an incidence of 8.6 cases/100,000 in the United States. This tumor type is strongly associated with smoking and is highly aggressive. Eighty-five percent of small cell lung cancers arise within a central location near the mediastinum. On CT these tumors show a central distribution, e.g. as a hilar or perihilar mass, with bulky hilar and mediastinal adenopathy (Fig. 5.7). Compression of the trachea, bronchi, SVC, innominate veins or pulmonary arteries may also occur. SCLC is staged as (a) limited to thorax or (b) extensive, beyond thorax. TNM system for staging is used only if the lesion is surgically resectable, however, the majority of patients have advanced disease at presentation.
b
a
c
Fig. 5.7 Small cell lung cancer. (a) Axial CT of bulky anterior mediastinal adenopathy. (b) Axial CT of primary lung cancer in the anterior segment of left upper lobe. (c) PET scan with increased 18 FDG uptake in the nodes with central photopenic necrosis
5 Imaging of Thoracic Malignancies
131
FDG-PET-CT imaging is highly accurate in the staging of SCLC and may be used in place of other imaging modalities such as CT and MRI [51]. A study by Pandit, et al. found that PET might provide prognostic information as well. They found that those tumors with a higher standard uptake value (SUV) had a poorer prognosis with lower survival rates [52]. SCLC tends to demonstrate a rapid response to therapy, and PET scan has been shown to also provide essential prognostic information and response to treatment at follow-up imaging [52]. However, in many instances, routine follow-up imaging is not advised since the majority (71 percent) of patients usually develop symptoms indicating recurrence [53].
5
Carcinoid
A carcinoid tumor is a neuroendocrine neoplasm and represents 1 percent to 2 percent of all lung tumors. It is classified into two forms, typical and atypical. Typical carcinoid tumors are low-grade compared to the atypical form, which is more aggressive. Ninety percent of thoracic carcinoid tumors are endobronchial in origin. The remaining 10 percent originate in the lung parenchyma [54]. Symptoms upon presentation may include cough, hemoptysis, difficulty in breathing or pneumonia due to bronchial obstruction. 6.7 percent to 10 percent of patients with this tumor will present with a carcinoid syndrome [54]. Carcinoid syndrome occurs in patients with metastases from carcinoid tumors, usually to the liver, and is characterized by facial flushing, diarrhea and asthma attacks caused by release of vasoactive substances like serotonin, histamine, vasoactive peptides, etc. into systemic circulation. Radiologically the carcinoid tumors may be seen as hilar/perihilar or endobronchial masses. The involved bronchus may be lobar, segmental or subsegmental, and may have co-existing post-obstructive atelectasis or consolidation (Fig. 5.8). Pulmonary oligemia, airtrapping and mucoid impaction may also be reported [55]. On CT 43 percent of the centrally located carcinoid tumors contain calcium [56, 57]. These calcifications tend to be eccentric in location; however, diffuse, punctuate calcifications resembling broncholithiasis have been described [58, 59]. Sixty percent of carcinoid tumors strongly enhance on CECT; however, not all carcinoid tumors enhance and atypical carcinoids are more likely to have less uniform enhancement [58, 59, 60]. Asymptomatic patients usually have a solitary homogeneous pulmonary nodule or mass of varying size. Radiologically typical versus atypical carcinoid tumors are indistinguishable. Regional lymphadenopathy and metastases may occur with either tumor. Octreoscan is scintigraphy with octroetide, a radionuclide analogue of somatostatin, which is taken up by somatostatin receptor positive carcinoid tumors. Octreoscan may be used for follow-up or detection of somatostatin receptor-positive bronchial carcinoids. However, CT is superior to Octreotide scan in the visualization of primary tumor and liver metastases [61]. PET imaging is less useful in the characterization of carcinoid tumors. Erasmus, et al. report that the SUV of these tumors may vary from a low 1.6 to a high SUV of 6.6 [62].
132
S.R. Digumarthy and S.L. Aquino
Fig. 5.8 Carcinoid tumor. Axial CT of endobronchial carcinoid in the posterior segmental bronchus of left lower lobe
6
Lymphoma
Lymphoma is broadly classified as Hodgkin’s disease (HD) and Non-Hodgkin’s Lymphoma (NHL). The patient population involved with Hodgkin’s disease has a bimodal peak at 30 and 70 years. Within the thorax the anterior mediastinal lymph nodes, especially the prevascular and paratracheal nodes, are most frequently involved and spread to adjacent nodal groups. This is typically contiguous [63]. Unlike NHL the lung is less frequently involved as the primary site [64]. Secondary involvement of the lungs has been reported in 10 percent to 15 percent of patients and is almost always the result of spread from adjacent nodal disease in the mediastinum or hilum [64]. NHL has a greater incidence in patients who are immunosuppressed. Lymph node involvement is more characteristically non-contiguous with posterior mediastinal lymph nodes being the most common region of involvement (Fig. 5.9). Extranodal
5 Imaging of Thoracic Malignancies
133
b a
c
Fig. 5.9 Non-Hodgkins lymphoma. (a) Axial CT of bulky mediastinal adenopathy compressing the left upper lobe. (b) Sagittal CT reconstruction showing anterior mediastinal lymphadenopathy. (c) PET scan with intense 18 FDG uptake
disease, such as the lungs, may occur as the primary site and the presence of primary disease in the lungs is commonly without the presence of lymph node involvement [65, 66] (Fig. 5.10). On CT, both HD and NHL may appear in the lungs as nodules or masses of varying sizes, consolidation, ground glass opacities or an interstitial reticular pattern. Cavitation of pulmonary nodules can be seen in both HD and NHL. Chest wall involvement with pleural plaques is more common with NHL. Pleural effusions may occur due to direct tumor involvement of the pleura, or from lymphatic obstruction by tumor with resultant accumulation of fluid in the pleural space.
134
S.R. Digumarthy and S.L. Aquino
Fig. 5.10 Non-Hodgkin’s lymphoma. Axial CT of primary MALT lymphoma of the right middle lobe presenting as consolidation
7 7.1
Metastases Endobronchial Metastases
Endobronchial metastases from extrapulmonary primary tumors is rare with only 204 cases described in literature [67]. In the literature reviewed by Sorensen, et al. the cancers that most frequently metastasize to the airways originate from the breast, kidney, colon and rectum. Five percent of cases occurred in the trachea and 8 percent at multiple sites, with equal distribution between both lungs [67]. In 27 percent of patients with endobronchial metastases this was the only metastatic focus from a previously treated extrapulmonary primary tumor; 69 percent had additional metastases and 4 percent had extrapulmonary primary simultaneously [67]. On imaging the endobronchial tumor itself, or associated findings like atelectasis, mediastinal lymphadenopathy and pleural effusion, are present. Mucus filling of distal occluded airways can result in “finger-in-glove” or “tree-in-bud” appearance.
7.2
Pulmonary Metastases
Imaging has a crucial role in the diagnosis of metastatic disease, which is necessary for staging, planning management and determining the overall prognosis.
5 Imaging of Thoracic Malignancies
135
Most pulmonary metastases are hematogenous in origin and present as multiple nodules [68-72] (Fig. 5.11). Pulmonary metastases may vary in appearance, depending on the source and type of tumor [Table 5.6]. Not all pulmonary nodules identified in a patient with a known primary tumor are metastases. Studies have shown that only 5 percent to 20 percent of the pulmonary nodules in patients with cancer are metastatic [76-77]. The chances of nodules being metastatic increases with the number of nodules present, as well as the advanced stage of the primary tumor on diagnosis [78-80]. Follow-up CT is useful in detecting early metastases, and studies in colon cancer and sarcoma have shown that such surveillance may improve patient survival [78, 81, 82].
Fig. 5.11 Lung metastases. Axial CT of multiple pulmonary metastases from osteosarcoma
Table 5.6 Types of Pulmonary Metastases Miliary Thyroid, Melanoma, Renal cell Large Sarcoma, colon, renal Calcified Osteosarcoma, chondrosarcoma, Mucinous adenocarcinoma from Pancreas, ovary, small intestine Cavitary Transitional cell, sarcoma, lymphoma, post chemotherapy Ground glass (a) Due to hemorrhage- choriocarcinoma, melanoma, renal cell, angiosarcoma, Kaposi’s sarcoma [73-75] (b) Due to air space disease Tumor emboli, in Sarcoma, renal, hepatocellular, melanoma pulmonary arteries
136
S.R. Digumarthy and S.L. Aquino
Most pulmonary metastases are discrete nodules in the lungs. However, in certain tumor types metastases may appear as cavitary, calcified, ground glass or branching (Table 5-6). Metastases with a branching configuration are seen with tumor emboli. Tumor emboli are due to hematogeneous spread of tumors localized to the pulmonary arteries. These endovascular metastases manifest as branching, lobulated enlargement of small-medium vessels on CT [83]. Secondary findings on CT may include infarction distal to the vascular metastasis, which can be ground glass or consolidation [83, 84]. Rarely, interstitial disease can occur from lymphangitic involvement [85]. On ventilation-perfusion scanning, there may be several subsegmental perfusion defects [86, 87]. On FDG-PET scan these nodules, when large enough, will show FDG uptake. FDG-PET imaging has been found to have varying sensitivity and specificity for nodule detection, depending on the size, type and location of the primary extrathoracic tumor. For PET the sensitivity ranges from 64 percent to 100 percent, and the specificity from 98 percent to 100 percent. CT has a sensitivity and specificity of 87 percent and 91 percent [88, 89]. Because the threshold of detection by PET is 8 mm or greater, a combined PET/CT has the potential to better detect pulmonary involvement.
7.3
Lymphangitic Carcinomatosis
Lymphangitic carcinomatosis is commonly seen with adenocarcinoma of lung, breast, gastrointestinal tract, lymphoma and melanoma [90-94]. Involvement of pulmonary lymphatics and perilymphatic tissue by direct spread of hilar lymphadenopathy or emboli is the usual mechanism [85, 95]. On plain radiograph perihilar bronchovascular thickening and subpleural Kerley B lines in an asymmetric distribution are common patterns. HRCT is more sensitive and specific in diagnosing this condition. Smooth or nodular thickening of interlobular septae and bronchovascular interstitium results from either edema from lymphatic obstruction or direct neoplastic interstitial involvement (Fig. 5.12). Reticular lines and polygonal structures are also typical features [96, 97]. These patterns are usually present in a background of normal architecture, which helps to distinguish this from interstitial fibrosis.
7.4
Nodal Metastases
Lymph node metastases in the thorax are commonly seen with carcinoma of breast, kidney, head and neck and melanoma. CT is superior to chest radiography in detecting nodal involvement [98]. However, since the CT criteria for abnormal nodes is based on size of short-axis diameter of greater than 1 cm, false positive and negative interpretations frequently occur. PET has been found to be more accurate
5 Imaging of Thoracic Malignancies
137
Fig. 5.12 Lymphangitic carcinomatosis in non small cell lung cancer. Axial CT of thickening of the axial interstitium and interlobular septa in left upper lobe
than CT [35, 99]. However, PET has a low sensitivity for detection of disease in sentinel nodes in breast cancer and melanoma, as well as in lymph nodes less than 5 mm in size [35, 100, 101].
8 8.1
Tumors of Pleura Malignant Pleural Mesothelioma (MPM)
Asbestos exposure is a major risk factor for malignant mesothelioma. Patients with this pleural tumor may present with a dry cough, difficulty in breathing, chest pain, weight loss and a pleural effusion. On CT rind-like pleural thickening of more than a centimeter, interlobar fissure involvement and pleural effusion are characteristic features of MPM [102, 103]. Irregular margins along the thickened pleura and nodules are other common imaging findings. MRI with contrast-enhanced T1 fat-saturated sequences is superior to CT in identifying involvement of the interlobar fissures, peritoneum and skeleton [104, 105]. MRI also provides better assessment of diaphragmatic and chest wall invasion (Fig. 5.13). Currently studies on FDG-PET are limited, but this imaging modality appears to be sensitive in detecting malignancy in the pleural space, including the presence of MPM. Limitations in specificity have been reported due to the increased uptake of
138
S.R. Digumarthy and S.L. Aquino
b a
Fig. 5.13 Malignant mesothelioma. (a) Axial CT shows right pleural mass indenting the liver, calicied pleural plaques on the left. (b) Coronal T2 W MRI shows diaphragmatic invasion on the right and circumferential tumor
FDG in infection and inflammation. However, for MPM, PET has been shown to be useful in detecting distant metastases in supraclavicular lymph nodes or within the abdomen [106, 107].
8.2
Pleural Metastases
Metastatic involvement of the pleura manifests as pleural effusion, pleural thickening and/or nodules. The most common sources for malignant pleural effusion are lung carcinoma at 36 percent to 43 percent, breast cancer at 9 percent to 25 percent, lymphoma at 7 percent to 10 percent and unknown primary at 7 percent to 10 percent [108-110]. Ultrasonography is more accurate in the detection, characterization and quantification of pleural effusion. Complex, echogenic fluid with septations is characteristic of exudative effusion [111, 112]. However, only the presence of a pleural mass was specific for malignancy [112]. Pleural nodules and nodular thickening of pleura are other manifestations of malignancy that can be seen by ultrasound or CT. When present on CT these features are very accurate in defining the presence of pleural involvement [113]. However, the absence of these features does not exclude a malignant effusion. FDG-PET, on the other hand, is more sensitive than CT in identifying malignant pleural disease with a sensitivity and specificity of 88.8 percent and 94.1 percent [114] (Fig. 5.14). MRI has a similar sensitivity to CT (96 percent) in the detection of pleural metastatic disease [115]. However, MRI, especially with contrast-enhanced sequences, is more sensitive in detecting involvement of chest wall, diaphragm and mediastinum [116].
5 Imaging of Thoracic Malignancies
a
139
b
Fig. 5.14 Pleural metastases. (a) Axial CT shows left pleural effusion with pleural nodule. (b) PET shows increased 18 FDG uptake in the pleural effusion and the nodule
Key Points ●
●
●
●
●
●
●
Bronchogenic carcinoma can be a solitary pulmonary nodule or a mass or an area of consolidation with variable enlargement of mediastinal or hilar nodes Ground glass nodules have a higher rate of malignancy, more specifically bronchioloalveolar cell carcinoma, than solid nodules Revisions to the TNM System of Staging divide Stage I into IA and IB reflecting their different prognosis; stage IIIB and IV are unresectable Role of contrast-enhanced CT: Better evaluation/detection of small endobronchial lesions, atelectasis/pneumonitis, mediastinal lymph nodes, pleural metastases Role of MRI: To assess for visceral pleural or chest wall invasion; superior sulcus tumors and their extension; Breathing Dynamic Echo Planar Imaging to exclude tumor fixation to pleura/chest wall 18-FDG-PET: More accurate, sensitive and specific than CT for lymph node assessment. Dual PET/CT is even more accurate. Also more sensitive for detection of metastases than other modalities alone Not all nodules (only 5 percent to 20 percent) with a known primary tumor are metastatic
Conclusion In conclusion, current imaging modalities provide important information on the staging and subsequent clinical management of patients with head and neck malignancies. Presently, CT tends to be the study of choice for malignancies below the level of the soft palate, whereas MRI is advantageous in imaging the nasopharynx and soft palate.
140
S.R. Digumarthy and S.L. Aquino
References 1. Ginsberg MS. Epidemiology of lung cancer. Semin Roentgenol 2005; 40(2): 83-84 2. Pawel DJ, Puskin JS: The U.S. Environmental Protection Agency’s assessmet of risks from indoor radon. Health Phys 2000; 87: 68-74 3. Doll R, Peto R. The causes of cancer: quantitative estimates of avoidable risks of cancer in the United States today. J Natl Cancer Inst 1981: 66:1191-1308 4. Fry WA, Menck HR, Winchester DP. The National Cancer Data Base report on lung cancer. Cancer 1996; 77:1947-55. 5. Zwirewich CV, Vedal S, Miller RR,et al. Solitary pulmonary nodule: high resolution CT and radiologic-pathologic correlation. Radiology 1991; 179: 469-481 6. Zerhouni EA, Stilik FP, Siegelman SS, et al. CT of the pulmonary nodule: a co-operative study. Radiology 1986; 160: 319-327 7. Mirtcheva RM, Vazquez M, Yankelevitz DF, et al. Bronchioloalveolar carcinoma and adenocarcinoma with bronchioloalveolar features presenting as ground-glass opacities on CT. J Clin Imaging 2002; 26:95-100. 8. Gaeta M, Caruso R, Barone M, et al. Ground-glass attenuation in nodular bronchioloalveolar carcinoma: CT patterns and prognostic value. J Comput Assist Tomogr 1998; 22:215-9. 9. Henschke CI, Yankelevitz DF, Mirtcheva R, et al. CT screening for lung cancer: frequency and significance of part-solid and nonsolid nodules. Am J Roentgenol 2002; 178:1053-7. 10. Kim EA, Johkoh T, Lee KS, et al. Quantification of ground-glass opacity on high-resolution CT of small peripheral adenocarcinoma of the lung: pathologic and prognostic implications. Am J Roentgenol 2001; 177:1417-22. 11. Asamura H, Suzuki K, Watanabe S, et al. A clinicopathological study of resected subcentimeter lung cancers: a favorable prognosis for ground glass opacity lesions. Ann Thorac Sug 2003; 76:1016-22. 12. Aquino SL, Chiles C, Halford P. Distinction of consolidative bronchioloalveolar carcinoma from pneumonia: do CT criteria work? AJR Am J Roentgenol 1998; 171:359-63. 13. Jung JI, Kim H, Park SH, et al. CT differentiation of pneumonic-type bronchioloalveolar cell carcinoma and infectious pneumonia. Br J Radiol 2001; 74:490-4. 14. Shah RM, Friedman AC. CT angiogram sign: incidence and significance in lobar consolidations evaluated by contrast-enhanced CT. Am J Roentgenol 1998; 170:719-21. 15. Murayama S, Onitsuka H, Murakami J, et al. “CT angiogram sign” in obstructive pneumonitis and pneumonia. J Comput Assist Tomogr 1993; 17:609-12. 16. Naruke T, Goya T, Tsuchiya R, Suemasu K. Prognosis and survival in resected lung carcinoma based on the new international staging system. J Thorac Cardiovasc Surg.1988 Sep;96(3):4407. Erratum in: J Thorac Cardiovasc Surg 1989 Mar;97(3):350 17. Regnard JF, Santelmo N, Romdhani N, et al. Bronchioloalveolar lung carcinoma: results of surgical treatment and prognostic factors. Chest 1998; 114:45-50. 18. Takasugi JE, Rapoport S, Shaw C. Superior sulcus tumors: the role of imaging. J Thorac Imaging 1989; 4:41-8. 19. Heelan RT, Demas BE, Caravelli JF, et al. Superior sulcus tumors: CT and MR imaging. Radiology 1989;170:637-41. 20. Laissy JP, Soyer P, Sekkal SR, et al. Assessment of vascular involvement with magnetic resonance angiography (MRA) in Pancoast syndrome. Magn Reson Imaging 1995; 13:523-30. 21. Mountain CF. Revisions in the international system for staging lung cancer. Chest 1997; 111:1710-7. 22. ChestXray.com. Staging Lung Cancer. June 2000 23. Swensen SJ, Brown LR, Colby TV, et al. Lung nodule enhancement at CT: prospective findings. Radiology 1996; 201: 447-455. 24. Padovani B, Mouroux J, Seksik L, et al. Chest wall invasion by bronchogenic carcinoma: evaluation with MR imaging. Radiology 1993; 187:33-8. 25. Glazer HS, Duncan-Meyer J, Aronberg DJ, et al. Pleural and chest wall invasion in bronchogenic carcinoma: CT evaluation. Radiology 1985; 157:191-4.
5 Imaging of Thoracic Malignancies
141
26. Ratto GB, Piacenza G, Frola C, et al. Chest wall involvement by lung cancer: computed tomographic detection and results of operation. Ann Thorac Surg 1991; 51:182-8. 27. Pennes DR, Glazer GM, Wimbish KJ, et al. Chest wall invasion by lung cancer: limitations of CT evaluation. AJR Am J Roentgenol 1985; 144:507-11. 28. Pearlberg JL, Sandler MA, Beute GH, et al. Limitations of CT in evaluation of neoplasms involving chest wall. J Comput Assist Tomogr 1987;11:290-3. 29. Scott IR, Müller NL, Miller RR, et al. Resectable Stage III lung cancer: CT, surgical, and pathologic correlation. Radiology 1988; 166(1 Pt 1):75-9. 30. Uhrmeister P, Allmann KH, Wertzel H, et al. Chest wall infiltration by lung cancer: value of thin-sectional CT with different reconstruction algorithms. Eur Radiol 1999; 9:1304-9. 31. Shiotani S, Sugimura K, Sugihara M, et al. Diagnosis of chest wall invasion by lung cancer: useful criteria for exclusion of the possibility of chest wall invasion with MR imaging. Radiat Med 2000; 18:283-90. 32. Gould MK, Kuschner WG, Rydzak CE, et al. Test performance of positron emission tomography and computed tomography for mediastinal staging in patients with non-small-cell lung cancer. Ann Intern Med 2003; 139:879-92. 33. Cerfolio RJ, Ojha B, Bryant AS, et al. The role of FDG-PET scan in staging patients with non-small cell carcinoma. Ann Thorac Surg 2003; 76:861-6. 34. Antoch G, Stattaus J, Nemat AT, et al. Non-small cell lung cancer: dual-modality PET/CT in preoperative staging. Radiology 2003; 229:526-33. 35. Lardinois D, Weder W, Hany TF, et al. Staging of non-small cell lung cancer with integrated positron-emission tomography and computed tomography. N Engl J Med 2003; 348: 2500-2507. 36. Pfister DG, Johnson DH, Azzoli CG, et al. American Society of Clinical Oncology treatment of unresectable non-small cell lung cancer guideline: update 2003. J Clin Oncol 2004; 22:330-53. 37. Quint LE, Tummala S, Brisson LJ, et al. Distribution of distant metastases from newly diagnosed non-small cell lung cancer. Ann Thorac Surg 1996;62:246-50. 38. Cox JE, Chiles C. Extrathoracic staging of non-small cell lung cancer. In: Taveras JM, Ferrucci JT, editors. Radiology diagnosis-imaging-intervention. Philadelphia: LippincottRaven Publishers, 1998:1-17 39. Salbeck R, Grau HC, Artmann H. Cerebral tumor staging in patients with bronchial carcinoma by computed tomography. Cancer 1990; 66:2007-11 40. Sze G, Milano E, Johnson C, Heier L. Detection of brain metastases: comparison of contrastenhanced MR with unenhanced MR and enhanced CT. AJNR Am J Neuroradiol 1990; 11:785-91 41. Baker ME, Pelley R.Hepatic metastases: basic principles and implications for radiologists. Radiology 1995 Nov;197(2):329-37 42. Del Frate C, Bazzocchi M, Mortele KJ etal. Detection of liver metastases: comparison of gadobenate dimeglumine-enhanced and ferumoxides-enhanced MRI examinations. Radiology 2002;225:766-772 43. Danet IM, Semelka RC, Leonardou P et al. Spectrum of MRI appearances of untreated metastases of the liver. Am J Roentgenol. 2003 Sep;181(3):809-17. 44. Gilman MD, Aquino SL. State-of-the-Art FDG-PET imaging of lung cancer. Semin Roentgenol 2005 Apr;40(2):143-53. Review. 45. Dunnick NR, Korobkin M. Imaging of adrenal incidentalomas: current status. Am J Roentgenol 2002;179:559-68. 46. Caoili EM, Korobkin M, Francis IR, et al. Adrenal masses: characterization with combined unenhanced and delayed enhanced CT. Radiology 2002;222:629-33 47. K.-Y. Lam, C.-Y. Lo. Metastatic tumors of the adrenal glands: a 30-year experience in a teaching hospital. Clin Endocrinol 2002;56(1) 95–101 48. Pieterman RM, van Putten JWG, Meuzelaar JJ, et al. Preoperative staging of non-small cell lung cancer with positron-emission tomography. N Engl J Med 2000; 343:254-61. 49. Cerfolio RJ, Ojha B, Bryant AS, et al. The role of FDG-PET scan in staging patients with non-small cell carcinoma. Ann Thorac Surg 2003; 76:861-6.
142
S.R. Digumarthy and S.L. Aquino
50. Pfister DG, Johnson DH, Azzoli CG, et al. American Society of Clinical Oncology treatment of unresectable non-small cell lung cancer guideline: update 2003. J Clin Oncol 2004;22:330-53. 51. Schumacher T, Brink I, Mix M, et al. FDG-PET imaging for the staging and follow-up of small cell lung cancer. Eur J Nucl Med 2001; 28:483-8. 52. Pandit N, Gonen M, Krug L, et al. Prognostic value of [18F] FDG-PET imaging in small cell lung cancer. Eur J Nucl Med 2003; 30:78-84. 53. Perez EA, Loprinzi CL, Sloan JA, et al. Utility of screening procedures for detecting recurrence of disease after complete response in patients with small cell lung carcinoma. Cancer 1997; 80:676-80. 54. Iglesias M, Belda J, Baldo X et al. Bronchial Carcinoid Tumor: a retrospective Analysis of 62 Surgically Treated Cases. Arch Bronconeumol. 2004 May;40 (5):218-21. 55. Jeung MY, Gasser B, Gangi A, et al. Bronchial carcinoid tumors of the thorax: spectrum of radiologic findings. RadioGraphics 2002; 22:351-65. 56. Magid D, Siegelman SS, Eggleston JC, et al. Pulmonary carcinoid tumors: CT assessment. J Comput Assist Tomogr 1989; 13:244-7. 57. Zwiebel BR, Austin JH, Brimes MM. Bronchial carcinoid tumors: assessment with CT of location and intratumoral calcification in 31 patients. Radiology 1991;179:483-6. 58. Paillas W, Moro-Sibilot D, Lantuejoul S, Brichon PY, Coulomb M, Ferretti G. Bronchial carcinoid tumors: role of imaging for diagnosis and local staging. J Radiol 2004; 85: 1711-9. 59. Naidich DP. CT/MR correlation in the evaluation of tracheobronchial neoplasia. Radiol Clin North Am 1990; 28:555-571 60. Aronchick JM, Wexler JA, Christen B, et al. Computed tomography of bronchial carcinoid. J Comput Assist Tomogr 1986; 10:71-4. 61. Granberg D, Sundin A, Janson ET, Oberg K, Skogseid B, Westlin JE. Octreoscan in patients with bronchial carcinoid tumors. Clin Endocrinol (Oxf) 2003 Dec; 59(6):793-9. 62. Erasmus, JJ, McAdams HP, Patz EF Jr, et al. Evaluation of primary pulmonary carcinoid tumors using FDG PET. Am J Roentgenol 1998; 170:1369-73. 63. A.J. van der Molan, M.Prokop. Neck. Chapter in Spiral and multislice computed tomography of the body. M. Prokop & M. Galanski. Thieme 2003. 261-62 64. C.Schaefer-Prokop,M. Prokop.Lungs and tracheobronchial system. Chapter in Spiral and multislice computed tomography of the body. M. Prokop & M. Galanski. Thieme 2003. 324-327 65. Berkman N, Breuer R, Kramer M, Polliack A. Pulmonary involvement in lymphoma. Leuk Lymphoma 1996;20:229-237. 66. Lewis E, Caskey C, Fishman E. Lymphoma of the lung: CT findings in 31 patients. Am J Roentgenol 1991; 156:711-714. 67. Sorensen J.B. Endobronchial metastases from extrapulmonary solid tumors. Acta Oncol 2004; 43(1):73-79 68. Al-Mehdi AB, Tozawa K, Fisher AB, et al. Intravascular origin of metastasis from the proliferation of endothelium-attached tumor cells: a new model for metastasis. Nat Med 2000; 6:100-2. 69. Wong CW, Song C, Grimes MM, et al. Intravascular location of breast cancer cells after spontaneous metastasis to the lung. Am J Pathol 2002; 161:749-53. 70. Milne EN, Zerhouni EA. Blood supply of pulmonary metastases. J Thorac Imaging 1987; 2:15-23. 71. Stackpole CW. Intrapulmonary spread of established B16 melanoma lung metastases and lung colonies. Invasion Metastasis 1990; 10:267-80. 72. Alterman AL, Fornabaio DM, Stackpole CW. Metastatic dissemination of B16 melanoma: pattern and sequence of metastasis. J Natl Cancer Inst 1985; 75:691-702. 73. Primack SL, Hartman TE, Lee KS, et al. Pulmonary nodules and the CT halo sign. Radiology 1994; 190:513-5. 74. Patel A, Ryu J. Angiosarcoma in the lung. Chest 1993; 103:1531-1535. 75. Benditt JO, Farber HW, Wright J, et al. Pulmonary hemorrhage with diffuse alveolar infiltrates in men with high-volume choriocarcinoma. Ann Intern Med 1988; 109:674-5.
5 Imaging of Thoracic Malignancies
143
76. Chalmers N, Best JJ. The significance of pulmonary nodules detected by CT but not by chest radiography in tumor staging. Clin Radiol 1991; 44:410-412. 77. Kronawitter U, Kemeny NE, Heelan R, et al. Evaluation of chest computed tomography in the staging of patients with potentially resectable liver metastases from colorectal carcinoma. Cancer 1999; 86:229-235. 78. Picci P, Vanel D, Briccoli A, et al. Computed tomography of pulmonary metastases from osteosarcoma: the less poor technique. A study of 51 patients with histological correlation. Ann Oncol 2001; 12:1601-4. 79. Lim DJ, Carter MF. Computerized tomography in the preoperative staging for pulmonary metastases in patients with renal cell carcinoma. J Urol 1993;150:1112-4. 80. Heaston DK, Putman CE, Rodan BA, et al. Solitary pulmonary metastases in high-risk melanoma patients: a prospective comparison of conventional and computed tomography. AJR Am J Roentgenol 1983; 141:169-74. 81. Yamada H, Katoh H, Kondo S, et al. Surgical treatment of pulmonary recurrence after hepatectomy for colorectal liver metastases. Hepatogastroenterology 2002; 49:976-979. 82. Koong HN, Pastorino U, Ginsberg RJ. Is there a role for pneumonectomy in pulmonary metastases? Ann Thorac Surg 1999; 68:2039-2043. 83. Shepard J, Moore E, Templeton P, et al. Pulmonary intravascular tumor emboli: dilated and beaded peripheral pulmonary arteries at CT. Radiology 1993; 187:797-801. 84. Kang C, Choi J, Kim H, et al. Lung metastases manifesting as pulmonary infarction by mucin and tumor embolization: radiographic, high-resolution CT, and pathologic findings. J Comput Assist Tomogr 1999; 23:644-646. 85. Odeh M, Oliven A, Misselevitch I, et al. Acute cor pulmonale due to tumor cell microemboli. Respiration 1997; 64:384-7. 86. Chan CK, Hutcheon MA, Hyland RH, et al. Pulmonary tumor embolism: a critical review of clinical, imaging, and hemodynamic features. J Thorac Imaging 1987; 2:4-14. 87. Schriner RW, Ryu JH, Edwards WD. Microscopic pulmonary tumor embolism causing subacute cor pulmonale: a difficult antemortem diagnosis. Mayo Clin Proc 1991; 66:143-8. 88. Majhail NS, Urbain JL, Albani JM, et al. F-18 fluorodeoxyglucose positron emission tomography in the evaluation of distant metastases from renal cell carcinoma. J Clin Oncol 2003; 21:3995-4000. 89. Imdahl A, Reinhardt MJ, Nitzsche EU, et al. Impact of 18F-FDG- positron emission tomography for decision making in colorectal cancer recurrences. Langenbecks Arch Surg 2000; 385:129-34. 90. Hirakata K, Nakata H, Haratake J. Appearance of pulmonary metastases on high-resolution CT scans: comparison with histopathologic findings from autopsy specimens. AJR Am J Roentgenol 1993; 161:37-43. 91. Munk PL, Muller NL, Miller RR, Ostrow DN. Pulmonary lymphangitic carcinomatosis: CT and pathologic findings. Radiology1988;166:705-9. 92. Stein MG, Mayo J, Muller N, Aberle DR, Webb WR, Gamsu G. Pulmonary lymphangitic spread of carcinoma: appearance on CT scans. Radiology 1987; 162:371-5. 93. Berkman N, Breuer R, Kramer M, Polliack A. Pulmonary involvement in lymphoma. Leuk Lymphoma 1996; 20:229-237. 94. Tanaka N, Matsumoto T, Miura G, et al. CT findings of leukemic pulmonary infiltration with pathologic correlation. Eur Radiol 2002; 12:166-74. 95. Heyneman LE, Johkoh T, Ward S, et al. Pulmonary leukemic infiltrates: high-resolution CT findings in 10 patients. AJR Am J Roentgenol 2000; 174:517-21. 96. Zerhouni EA, Naidich DP, Stitik FP, et al. Computed tomography of the pulmonary parenchyma. Part 2: Interstitial disease. J Thorac Imaging 1985; 1:54-64. 97. Ren H, Hruban RH, Kuhlman JE, et al. Computed tomography of inflation-fixed lungs: the beaded septum sign of pulmonary metastases. J Comput Assist Tomogr 1989; 13:411-6. 98. Williams MP, Husband JE, Heron CW. Intrathoracic manifestations of metastatic testicular seminoma: a comparison of chest radiographic and CT findings. AJR Am J Roentgenol 1987; 149:473-5.
144
S.R. Digumarthy and S.L. Aquino
99. Schirrmeister H, Kuhn T, Guhlmann A, et al. Fluorine-18 2-deoxy-2-fluoro-D-glucose PET in the preoperative staging of breast cancer: comparison with the standard staging procedures. Eur J Nucl Med 2001; 28:351-8. 100. Barranger E, Grahek D, Antoine M, et al. Evaluation of fluorodeoxyglucose positron emission tomography in the detection of axillary lymph node metastases in patients with earlystage breast cancer. Ann Surg Oncol 2003; 10:622-7. 101. van der Hoeven JJ, Hoekstra OS, Comans EF, et al. Determinants of diagnostic performance of [F-18]fluorodeoxyglucose positron emission tomography for axillary staging in breast cancer. Ann Surg 2002; 236:619-24. 102. Kawashima A, Libshitz HI. Malignant pleural mesothelioma: CT manifestations in 50 cases. AJR Am J Roentgenol 1990; 155:965-9. 103. Metintas M, Ucgun I, Elbek O, et al. Computed tomography features in malignant pleural mesothelioma and other commonly seen pleural diseases. Eur J Radiol 2002; 41:1-9. 104. Knuuttila A, Halme M, Kivisaari L, et al. The clinical importance of magnetic resonance imaging versus computed tomography in malignant pleural mesothelioma. Lung Cancer 1998; 22:215-25. 105. Heelan RT, Rusch VW, Begg CB, et al. Staging of malignant pleural mesothelioma: comparison of CT and MR imaging. AJR Am J Roentgenol 1999; 172:1039-47. 106. Gerbaudo VH, Sugarbaker DJ, Britz-Cunningham S, et al. Assessment of malignant pleural mesothelioma with 18F-FDG dual-head gamma-camera coincidence imaging: comparison with histopathology. J Nucl Med 2002; 43:1144-1149. 107. Flores RA, Akhurst TB, Gonen MC, et al. Positron emission tomography defines metastatic disease but not locoregional disease in patients with malignant pleural mesothelioma. J Thorac Cardiovasc Surg 2003; 126:11-16. 108. Monte SA, Ehya H, Lang WR. Positive effusion cytology as the initial presentation of malignancy. Acta Cytol 1987; 31:448-52. 109. Sahn SA. Malignant pleural effusions. Clin Chest Med 1985; 6:113-25. 101. 110. O’Donovan PB, Eng P. Pleural changes in malignant pleural effusions: appearance on computed tomography. Cleve Clin J Med 1994; 61:127-31; quiz 162. 111. Yang PC, Luh KT, Chang DB, Wu HD, Yu CJ, Kuo SH. Value of sonography in determining the nature of pleural effusion: analysis of 320 cases. AJR Am J Roentgenol 1992; 159:29-33. 112. Gorg C, Restrepo I, Schwerk WB. Sonography of malignant pleural effusion. Eur Radiol 1997; 7:1195-8. 113. Arenas-Jimenez J, Alonso-Charterina S, Sanchez-Paya J, et al. Evaluation of CT findings for diagnosis of pleural effusions. Eur Radiol 2000; 10:681-90. 114. Gupta NC, Rogers JS, Graeber GM, et al. Clinical role of F-18 fluorodeoxyglucose positron emission tomography imaging in patients with lung cancer and suspected malignant pleural effusion. Chest 2002; 122:1918-1924. 115. Hierholzer J, Luo L, Bittner RC, et al. MRI and CT in the differential diagnosis of pleural disease. Chest 2000; 118:604-9 116. Carlsen SE, Bergin CJ, Hoppe RT. MRI to detect chest wall and pleural involvement in patients with lymphoma: effect on radiation therapy planning. AJR Am J Roentgenol 1993; 160:1191-5.
6
Imaging of Mediastinal Tumors Scott Moore1 MD, Hetal Dave-Verma2 MD, and Ajay Singh3, MD
Key Points ●
●
●
●
●
●
The most common cause of anterior mediastinal masses is thymoma that is often associated with myasthenia gravis and other paraneoplastic syndromes. Radiographically, thymic tumors and lymphoma present as smooth or lobulated masses in the upper half of the chest. With the presence of calcification, the lesion is most likely of thymic origin since untreated lymphoma rarely calcifies. Mediastinal and hilar lymph nodes are more often enlarged in lymphoma. GCTs, with the exception of NSGCT, tend to be more heterogeneous with multiple densities. NSGCTs are similar to GCTs with the absence of fat and calcification. Neurogenic tumors include those arising from the peripheral nerves, such as schwannomas, neurofibromas and their malignant counterparts: those arising from the sympathetic ganglia, such as ganglioneuroma, ganglioneuroblastoma and neuroblastoma, and those arising from the parasympathetic ganglia such as pheochromocytoma and chemodectoma. Neurogenic tumors are well-defined rounded masses in the costovertebral junction with cortical disruption of the adjacent ribs and vertebrae. They have low density on CT and enhance after contrast administration, most strikingly seen with the paragangliomas. MRI will demonstrate spinal cord involvement. Definitive diagnosis of a mediastinal mass involves histologic sampling that can be performed via US or CT-guided percutaneous biopsy, as well as other imageguided or surgical techniques. PET-CT is an evolving imaging technique that is used for re-staging of lymphoma and seminoma and likely NSGCT. Therapy monitoring is a promising new application of PET-CT that is currently under research.
1,2,3
The University Of Massachusetts Memorial Hospital, University Campus, Department Of Radiology, 55 Lake Avenue North, Worcester, Massachusetts 01655 3
Massachusetts General Hospital, Boston, MA-02114
Corresponding Author: Ajay Singh, 10 Museum Way,# 524, Cambridge, MA-02141
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
145
146
1
S. Moore et al.
Introduction
The mediastinum comprises the region extending from the thoracic inlet to the diaphragm in the central thorax, interposed between the two pleural cavities [1]. Traditionally, the mediastinum is separated into three compartments (anterior, middle and posterior) as a classification scheme, since various tumor types are more common in certain locations. These compartments are not actual anatomic locations divided by fascial planes, but are hypothetical regions radiographically. No universal approach to subdividing the mediastinum is utilized, but the scheme developed by Fraser, et al. [2], which is based on the lateral radiograph, is applied in this review. The anterior mediastinum is anterior to the pericardium and brachiocephalic vessels, and includes the thymus gland, lymph nodes and fat. The posterior mediastinum is posterior to the heart and trachea and extends from the posterior pericardial reflection to the posterior border of the vertebral bodies. It includes the autonomic ganglia, descending thoracic aorta, azygous vein, esophagus, lymph nodes, thoracic duct and fat. The remaining middle mediastinum is the space between the aforementioned spaces and includes the heart, intrapericardial great vessels, pericardium, trachea, and lymph nodes. The risk of malignancy is primarily related to which compartment the tumor occupies, as masses in the anterior compartment are statistically favored to be malignant [3]. Age and symptomatology are additional variables to be considered. Most mediastinal tumors in infants and children are neurogenic tumors; adults between the ages of 20 and 40 years are more likely to have a germ cell tumor or lymphoma. These populations of patients are more at risk of having a malignant mediastinal mass. Almost 75 percent of patients displaying symptoms have malignant medastinal tumors [3]. The most common symptoms include chest pain, cough, dyspnea, fever, chills and rarely, superior vena cava syndrome.
2
Imaging of Mediastinal Tumors
The initial radiologic investigation of a mediastinal tumor is the traditional posteroanterior (PA) and lateral chest radiograph. Plain film radiography confirms the diagnosis of a mediastinal mass. It also helps in defining the compartment the tumor occupies, estimating the size of the tumor and evaluating the composition of the tumor. For example, the presence of teeth or bone within a mediastinal mass is suggestive of a mediastinal teratoma. However, plain film radiography is rarely diagnostic. Computed tomography (CT) is the gold standard for further evaluation of mediastinal tumors. CT can determine tumor composition (cystic, fat, soft tissue) and characterize tumor enhancement [4] (Fig. 6.1). Furthermore, CT defines the tumor in relation to adjacent tissues and structures while providing further confirmation of specific compartment involvement. Additionally, it is valuable in determining a suitable approach for diagnostic biopsy (discussed later in this chapter). Iodinated contrast should be administered unless there is a contraindication, such as previous contrast anaphylaxis or renal failure.
6 Imaging of Mediastinal Tumors
147
Fig. 6.1 Mediastinal thymoma. (a) Thymoma in an 18-year-old male presenting with cough and found to have a mediastinal mass on CXR. Subsequent contrast-enhanced CT demonstrates an anterior mediastinal mass with both cystic and solid components. (b) Contrast-enhanced CT in a 75-year-old male reveals a heterogeneous anterior mediastinal mass (arrow) adjacent to the right heart border. This was a biopsy proven WHO Type A thymoma. (c) Contrast-enhanced CT in a 76-year-old female demonstrates an anterior mediastinal mass (arrow) consistent with a thymoma
Magnetic resonance imaging (MRI) usually provides complementary information with respect to CT evaluation of mediastinal tumors. However, MRI is recommended if CT is contraindicated, and for investigation of posterior mediastinal tumors [5]. Also, vascular and cardiac invasion is best identified with MRI. Although the role of ultrasound in the evaluation of mediastinal tumors is limited, it helps differentiate solid from cystic masses, and can be utilized for imageguided biopsy to determine the histology of a mediastinal tumor. Finally, nuclear medicine examination of mediastinal tumors includes the administration of radiotracers such as I-131, I-123, and historically, gallium-67. Radiolabeled I-131 and I-123 can evaluate tumors suspected to be of thyroid origin. Also, gallium-67 was long the radiotracer of choice for staging/re-staging of mediastinal lymphomas, but has been surpassed in recent years by the development of fusion PET/CT. Positron emission tomography (PET), in conjunction with CT
148
S. Moore et al.
utilizing FDG (fluorodeoxyglucose), has exploded onto the scene revolutionizing the evaluation of mediastinal tumors (especially lymphomas). FDG-PET is superior to Gallium-67 in lymphoma staging and re-staging, and provides a quicker scanning time and decreased radiation to the patient [6]. These aforementioned imaging modalities will be discussed in further detail as they pertain to the work-up of each individual tumor as we proceed into our discussion.
3 3.1
Thymic Tumors Thymoma
Thymomas are the most common cause of anterior mediastinal masses in adults, constituting 20 percent of all adult mediastinal masses and 50 percent of anterior mediastinal masses. They are rare in children, occur with about the same frequency in males and females, and there is no predilection for a particular race or geographic distribution. Although most are benign, 10 percent to 40 percent can be malignant [7, 8]. Thymomas are most commonly seen in the fifth and sixth decades of life [9, 10]. The etiology of thymomas has not been elucidated; however, they have been associated with various systemic syndromes. These include myasthenia gravis, red cell aplasia, hypogammaglobulinemia, polymyositis and, much less commonly, systemic lupus erythematosus, rheumatoid arthritis, thyroiditis, hyperthyroidism and other cytopenias [11]. Approximately 30 percent to 40 percent of patients with a thymoma have symptoms suggestive of myasthenia gravis, compared to 10 percent to 15 percent of patients with myasthenia gravis who have a thymoma [12, 13]. Thymomas with concurrent myasthenia gravis tend to be less aggressive tumors than those without this associated disorder [7, 8]. Thymomas are classified based on cell type predominance into the categories of lymphocytic, epithelial or spindle cell variants. These tumors are typically bland in appearance and can demonstrate mild to moderate cellular atypia [14]. No clear histologic distinction between benign and malignant thymomas exists, although there is an association between histologic subtype and degree of invasion. Most thymomas are solid tumors, but up to one-third may have components that are hemorrhagic, necrotic or cystic [15, 16]. The propensity of a thymoma to be malignant is determined by the invasiveness of the tumor. Malignant thymomas can directly extend through the capsule into adjacent structures such as lung, mediastinal soft tissue, pericardium, pleura, or via transdiaphragmatic extension into the abdomen [15, 16]. Over one-third of thymomas invade through their own capsules extending into surrounding structures with lymphogenous and hematogenous spread rarely seen [12, 17]. The extent of capsule invasion and the involvement of thoracic and extrathoracic structures determines the stage, which is correlated to the risk of recurrence and survival. Thus, extensive tissue sampling of the resected tumor is essential to define microscopic and macroscopic invasion through the fibrous capsule. Other prognostic
6 Imaging of Mediastinal Tumors
149
factors include the completeness of excision, tumor size, histologic typing, involvement of the great vessels and performance status [12, 18]. The World Health Organization classification (see Table 6.1) is based on cytologic differences, which may be helpful in determining treatment regimens and predicting survival, given the potential importance of histologic subtype on prognosis [18, 19]. Another staging system used for thymomas as proposed by Masaoka [12] is outlined in Table 6.2. The Masaoka staging system has proven to be a useful independent predictor of survival in patients with thymoma, since it is based primarily on extent of invasion [12, 18]. Typically, a thymoma is an incidental finding on a chest radiograph. One-third of patients manifest symptoms of chest pain, cough, dyspnea, dysphagia, fever, weight loss or anorexia related to tumor compression or invasion [2].
Table 6.1 World Health Organization Classification of Thymomas Class of Thymoma Cytologic Features Type A Type AB Type B1 Type B2 Type B3 From Wilkins, et al. [19]
Spindle cell, medullary Mixed Lymphocytic, predominantly cortical, organoid Cortical Atypical, squamous, epithelial, well-differentiated thymic carcinoma
Table 6.2 Masaoka Staging System of Thymoma 5-yr Survival Rate, % Treatment
Stage
Degree of invasion
1
Complete encapsulation 96–100 macroscopically and no capsular invasion microscopically Invasion into the sur86–95 rounding fatty tissue or mediastinal pleura macroscopically or invasion into the capsule microscopically Invasion into neighbor- 56–69 ing organs macroscopically
2
3
4a 4b
Pleural or pericardial 11–50 dissemination Lymphogenous or hematogenous metastasis
From Shamji, et al. [20].
Complete surgical excision
Complete surgical excision and postoperative radiotherapy to decrease the incidence of local recurrence
Complete surgical excision and postoperative radiotherapy to decrease the incidence of local recurrence Surgical debulking, radiotherapy, and chemotherapy Surgical debulking, radiotherapy, and chemotherapy
150
3.2
S. Moore et al.
Thymic Carcinoma
Thymic carcinomas are highly aggressive and invasive epithelial neoplasms of the thymus that are characterized by a high degree of cytologic atypia [3]. These tumors are rare, predominantly presenting in middle-aged men [21]. Most patients present with cough, shortness of breath, chest/shoulder discomfort, fatigue and weight loss. Superior vena cava syndrome and cardiac tamponade have also been described [22, 23]. Unlike thymomas, paraneoplastic syndromes are generally not associated with thymic carcinoma. However, well-differentiated thymic carcinoma has been reported in association with myasthenia gravis [24]. In contrast to thymomas, thymic carcinomas appear cytologically malignant with features of cellular atypia, cellular necrosis and mitoses. More than half of thymic carcinomas are undifferentiated. Other tumor subtypes include spindle cell, squamous cell, lymphoepithelioma-like, mucoepidermoid, basaloid, clear cell and adenoid cystic tumor [25]. Histologically, thymic carcinomas are large, firm, infiltrating masses with areas of cystic change and necrosis. The spindle cell variety is the most aggressive subtype, with reported mortality rates approaching 50 percent within five years [25]. There is also some evidence that the Epstein-Barr virus (EBV) may play a role in the development of a lymphoepithelioma-like carcinoma of the thymus gland. Southern blot analysis has shown the EBV viral genome in the cells of thymic lymphoepithelioma-like carcinoma, and EBV nuclease antigen has been found in tumor cells [26]. These tumors tend to have a poor prognosis because they attain very large size before discovery [25]. The other listed forms of thymic carcinoma are rare. Radiographically, thymic carcinomas are heterogeneous anterior mediastinal masses with necrosis and calcifications. They appear similar to thymomas, and tissue sampling is needed for definitive diagnosis.
3.3
Thymic Carcinoid
Thymic carcinoids, or primary thymic neuroendocrine carcinomas, are rare tumors which account for less than 5 percent of all anterior mediastinal neoplasms. They predominantly affect men aged 40 to 60 years and are much more aggressive than the neuroendocrine tumors originating in other locations. About half of patients are asymptomatic or have symptoms associated with local growth. The typical carcinoid syndrome is rarely associated with thymic carcinoid [27]. The most common endocrine abnormality that affects about 50 percent of patients is Cushing syndrome due to ectopic ACTH production or multiple endocrine neoplasia (MEN) syndrome [28]. According to one prospective study of patients with MEN1, thymic carcinoid developed in 8 percent of patients [29]. Thymic carcinoids are histologically similar to carcinoid tumors found at other sites. These thymic tumors may be designated as well-differentiated, moderately differentiated or poorly differentiated tumors, and are characterized by tumor cells that
6 Imaging of Mediastinal Tumors
151
form into into organoid clusters with tumor rosettes and ribbons. The vast majority of cells are positive for neuroendocrine markers such as chromagranin and synaptophysin, with no association found between prognosis and histologic features [30]. Metastasis is common with spread to regional lymph nodes, as well as distant metastasis developing in two-thirds of patients. Invasion of regional lymph nodes and distant metastases are reported in most cases and tend to occur late [27]. Whenever possible, complete surgical resection is the treatment of choice. For a locally invasive tumor, radiation and chemotherapy are used despite unsatisfactory results [31]. The prognosis of these tumors is generally poor.
3.4
Thymolipoma
Thymolipomas are very rare, benign thymic tumors containing mature fat and normal-looking or involuted thymic tissue. There is a broad age range from three to 60 years, but it is most common in young adults of both sexes [32]. Similar to thymomas, individual cases have been reported in association with a variety of conditions which include Graves’ disease, myasthenia gravis, aplastic anaemia and hypogammaglobulinaemia [33]. Thymolipomas can grow to a very large size and often mould themselves to the adjacent mediastinum and diaphragm. They may mimic cardiomegaly or lobar collapse [34]. CT and MRI demonstrate a fat density with areas of thymic tissue and fibrous septa [34]. The treatment of choice is surgical excision.
3.5
Imaging Features
Frontal and lateral chest radiographs can detect most thymic tumors as anterior mediastinal masses. On the frontal view, the lesion typically appears as a smooth mass in the upper half of the chest, overlying the superior aspect of the cardiac shadow near the junction of the heart and great vessels (Fig. 6.2). A few may be seen more inferiorly projecting over the right or left heart border or even as inferior as in the region of the cardiophrenic angles. On the lateral projection, there is often just a vague opacity [33]. Punctate or curvilinear calcifications can be seen in benign or malignant thymic tumors [35]. CT scans better delineate the features of thymic masses and can detect a smaller tumor missed on radiographs. A chest CT scan is the most sensitive imaging procedure for detection of thymoma in patients with myasthenia gravis [36]. A scan with intravenous contrast dye is preferred to show the relationship between the mass and surrounding vascular structures, to determine the degree of vascularity and to delineate the anatomy of the tumor to help guide the surgeon in removal. Thymomas, thymic carcinomas and thymic carcinoids usually demonstrate uniform enhancement [33]. Thymomas and thymic carcinomas will typically appear as well-defined, encapsulated, soft tissue masses that often contain hemorrhage,
152
S. Moore et al.
Fig. 6.2 Thymolipoma. Frontal CXR in a 20-year-old male shows fullness of the left AP window region (arrowheads) secondary to an anterior mediastinal thymolipoma
necrosis or cyst formation resulting in a heterogeneous density. Less likely, they can be predominantly cystic with a nodular component [21]. Thymic carcinoma can be associated with pleural and pericardial effusions. Thymic carcinoids present as large, lobulated, invasive masses of the anterior mediastinum with or without hemorrhage and necrosis. These tumors and characteristics are indistinguishable from the epithelial tumors, based on plain film and CT findings, but are not encapsulated [30]. In adults over the age of 40, thymic lesions as small as 1.5 to 2.0 cm in diameter can be identified because the rest of the thymus is atrophic. Before age 30, diagnosing a small thymic mass can be difficult because the normal gland is variable in size. Evaluation is more difficult in patients with myasthenia gravis in which the associated hyperplasia may cause a bulky gland. Visualizing asymmetrical focal swelling is important in these circumstances. Thymomas are very infrequent in children, so the difficult problem of finding a thymoma in a child rarely arises [33]. Invasion of the adjacent pleura and fat may be identified with a malignant thymoma. However, CT cannot distinguish a benign from a malignant thymoma if the tumor remains confined to the thymus [33]. MRI usually provides similar information to CT, but may be a useful adjunct to show mediastinal spread when CT results are questionable. Thymomas have a signal intensity similar to that of muscle and the adjacent normal thymic tissue on T1weighted images. On T2-weighted images the signal intensity is higher, making it difficult to distinguish a thymoma from adjacent mediastinal fat [21].
6 Imaging of Mediastinal Tumors
3.6
153
Mediastinal Lymphoma
Lymphoma is one of the most common mediastinal tumors (second most common anterior mediastinal tumor) and can manifest as a primary tumor or represent generalized disease. More often, lymphoma affects the mediastinum in combination with a broader scope of disease. Primary lymphoma affects the mediastinum in less than 10 percent of patients, usually occurring in the anterior mediastinum, with the remainder of cases occupying the middle mediastinum. Hodgkin’s disease (HD) and NonHodgkin’s Lymphoma (NHL) both affect the mediastinum, with HD affecting the mediastinum in approximately 50 percent to 70 percent of cases, and NHL in approximately 15 percent to 25 percent [37, 38]. Although any histologic variant of lymphoma may affect the mediastinum, the most common types include Hodgkin’s disease (nodular sclerosing subtype), large B-cell lymphoma and lymphoblastic lymphoma (the latter two entities are subtypes of Non-Hodgkin’s Lymphoma) [32].
3.7
Hodgkin’s Disease (HD)
Classic Hodgkin’s disease is based on the Rye classification and can be subdivided into four categories based on histopathologic investigation: lymphocyte-rich, lymphocyte-depleted, mixed cellularity and nodular sclerosis. HD is the most common lymphoma to affect the mediastinum with the nodular sclerosing variant representing the majority of cases. Uniquely, the nodular sclerosing variant involves the anterior mediastinum, most typically the thymus gland. The other subtypes of HD may affect the mediastinum, but they most frequently affect lymph nodes and do not manifest as discrete mediastinal masses. HD has a bimodal incidence with one peak in adolescence/young adulthood and the other after the age of 50 years [39]. Interestingly, patients presenting with isolated mediastinal involvement typically are younger with women in the third decade of life most often affected [40]. Clinically, patients with HD typically present with cervical or supraclavicular lymphadenopathy and constitutional (B symptoms) symptoms such as weight loss, night sweats, and fever. Also, patients with mediastinal involvement may present with cough, chest pain, dyspnea, dysphagia or even superior vena cava syndrome [40]. Histologically, the pathognomonic feature of HD is the presence of Reed-Sternberg cells which are bilobed nuclei with prominent eosinophilic nucleoli. The immunohistochemical profile of Reed-Sternberg cells classically displays biomarker positivity for Leu-M1 (CD15) and Ki-1 (CD30) [41]. The staging of HD is based on the modified Cotswold Ann Arbor staging system (see Tables 6.3 and 6.4) that determines prognostic data and treatment plan options.
3.8
Non-Hodgkin’s Lymphoma (NHL)
There are many subdivisions of NHL, but large B-cell lymphoma and lymphoblastic lymphoma are the two most common types to involve the mediastinum [42].
154
S. Moore et al.
Table 6.3 Ann Arbor Staging System Stage Extent of Invasion 1 2 3 4 From Duwe, et al. [21]
1 lymph node region involved 2 or more lymph node regions involved on one side of the diaphragm Spread across the diaphragm with lymph nodes on both sides involved Extranodal involvement
Table 6.4 Cotswold Modifications Stage Disease Characteristic A B
Asymptomatic ‘B’ symptoms including fever, night sweats and weight loss of >10 percent in six months X Large tumor size causing one-third widening of the mediastinum or >10cm diameter of a nodal mass E Extranodal involvement From Duwe, et al. [21]
Affected patients with NHL have a mean age of 55 years at diagnosis with a genetic disposition for Caucasian men [42]. Almost 85 percent of patients with NHL display constitutional (B symptoms) symptoms of fever, night sweats and weight loss. These patients also present with generalized lymphadenopathy and/or extensive extranodal and advanced disease at diagnosis [43]. Conversely, patients with NHL often present clinically as a rapidly enlarging mediastinal mass with invasion of the airway, chest wall and superior vena cava. However, like HD, patients with mediastinal involvement are younger. For large B-cell lymphoma, young adults (mostly women) are affected at a mean age of 26 years [44]. Lymphoblastic lymphoma typically presents in the first to second decade of life (most frequently affecting males) and bears resemblance to ALL (Acute Lymphoblastic Leukemia) as both share common clinical features. They have been postulated to be the solid and circulating phases of the same malignancy [45]. The staging of NHL can be based on the Ann Arbor staging system, but since these tumors are so aggressive they are most likely systemic upon presentation. Due to this fact and that NHLs are a diverse population of tumors, the histological classification offers more information on prognosis and outcome than anatomic extent [43]. The Cotswold modifications are often included in the staging system and include certain clinical factors that affect prognosis (see Table 6.4).
3.9
Imaging Features
As in the evaluation of other mediastinal tumors, PA and lateral radiography remains the initial imaging modality of choice. However, plain film radiography cannot distinguish between Hodgkin’s and Non-Hodgkin’s Lymphoma. On chest
6 Imaging of Mediastinal Tumors
155
radiography, mediastinal lymphoma presents as a lobulated mass with enlargement of the hilar and mediastinal lymph nodes (Fig. 6.3). This lobulated, mediastinal mass appearance can also be found in cases of thymoma and germ cell tumor as well. The presence of calcification may be beneficial as calcifications are quite rare in the absence of treated lymphoma. However, CT will be performed in cases of mediastinal lymphoma and the advantages have been well documented in the preceding sections. In cases of lymphoma, CT is also beneficial in determining radiation fields and is the gold standard for the staging of lymphomas. On CT, a conglomerate of lymph nodes or discrete enlarged nodes are often present in both HD and NHL with cystic degeneration [46] (Fig. 6.4). MRI offers similar anatomic structural information with diffusely low signal intensity masses on T1-weighted imaging demonstrating high signal intensity on T2-weighted imaging. An area of recent development concerns the post-treatment follow-up imaging of lymphoma patients. Many patients will have residual masses that are slow to
Fig. 6.3 Mediastinal lymphoma. (a, b) PA and lateral CXR in a 32-year-old female with the nodular sclerosing subtype of Hodgkin’s disease demonstrates an anterosuperior mediastinal mass (arrowheads). (c) Frontal CXR demonstrates a large anterior mediastinal mass (arrowhead) in a 48-year-old female with large cell lymphoma
156
S. Moore et al.
Fig. 6.4 Mediastinal lymphoma. (a) 63-year-old male presenting with cough. CXR showed a lobulated mass in the right suprahilar region. Subsequent contrast-enhanced CT shows a large right mediastinal mass that was biopsy-proven to be diffuse large B-cell lymphoma. (b) Contrastenhanced CT in a 22-year-old male with an anterior mediastinal mass (curved arrow) consistent with mediastinal lymphoma. (c) Contrast-enhanced CT shows a large, heterogeneous anterior mediastinal mass (arrowheads) that was found to be large cell NHL
resolve or that contain fibrous tissue. CT and MRI are not able to definitively distinguish between active disease and residual fibrosis since they offer more anatomic/structural information. CT demonstrates soft tissue density, often with calcifications within residual masses, but active disease cannot be excluded. Likewise, MRI of residual masses demonstrates low signal intensity on T1weighted and T2-weighted imaging. High signal intensity on T2-weighted imaging reflects active disease, but this appearance can be found in inflammation and necrosis as well [47]. PET-CT is superior for re-staging of mediastinal lymphomas due to the fact that this modality offers details on the metabolic activity of the viable residual mass [48]. Internal changes within the mass may be demonstrable in the absence of morphological changes in size, and this can help differentiate between a viable tumor and necrosis
6 Imaging of Mediastinal Tumors
157
or fibrosis. Further research will need to be performed, but initial evidence shows that PET-CT may have a potential role in pretreatment staging, as well as therapy monitoring which ultimately would affect initial treatment plans and predict response to therapies [48, 49].
4
Mediastinal Germ Cell Tumors (GCTs)
Germ cell tumors of the mediastinum originate from primitive germ cells that fail to migrate completely during embryogenesis, or from primordial cells of the thymus with germ cell potential [50]. However, the exact mechanism whereby germ cell tumors originate in the mediastinum remains unknown. They are most frequently located in the anterior mediastinum, comprising approximately 10 percent to 15 percent of all mediastinal tumors [3, 50]. Initially, mediastinal GCTs were thought to represent metastases from a primary gonadal origin, but general consensus at this time is that most of these tumors have an extragonadal derivation [51]. However, a primary gonadal malignancy must be excluded when a mediastinal GCT has been detected. Although histologically similar, mediastinal GCTs are separately distinct from their gonadal counterparts, with differing biological and clinical manifestations. For example, gonadal GCTs are frequently detected by the patient at a relatively small size, while mediastinal GCTs fail to be recognized until they have grown to a large size producing symptoms such as dyspnea, cough, chest pain or even superior vena cava syndrome. Furthermore, serum tumor markers can aid in differentiating the various mediastinal GCTs while also allowing detection of recurrent disease/ monitoring response to therapy. The anterior mediastinum is the most common site of extragonadal germ cell tumors with other common sites including the retroperitoneum, sacrococcygeal region, and the pineal gland [52]. These tumors can be benign or malignant depending on the particular subtype. Histologically, mediastinal GCTs are divided into three groups based on the cell type involved: teratoma, seminoma and embryonal tumors (NSGCTs or Nonseminomatous GCTs). The latter entity reflects a diverse population of distinct neoplasms including embryonal cell carcinoma, mixed GCTs, endodermal sinus tumor and choriocarcinoma [51, 53]. These tumors most often secrete serologic markers such as B-hCG (beta subunit of human chorionic gonadotropin), AFP (alpha-fetoprotein) and LDH (lactate dehydrogenase), which is helpful in the diagnostic evaluation [32].
4.1
Mediastinal Teratoma
Teratomas are the most common germ cell tumor within the mediastinum [21, 50]. Histologically, they are comprised of tissue that arises from the three embryonic primitive germ cell layers (ectoderm, mesoderm and endoderm). Ectodermal tissue
158
S. Moore et al.
is most prevalent in these tumors and includes structures such as hair, skin, sweat glands and teeth. Tissues derived from mesoderm and endoderm – such as fat, bone, cartilage, muscle and respiratory/intestinal epithelium – are less common [21]. Teratomas can be subclassified into mature teratomas, immature teratomas and teratocarcinomas, based on histopathologic features. Mature teratomas are welldifferentiated and benign [21, 50] while those found to include fetal tissue are termed immature teratomas and are malignant. Teratomas that contain a focus of carcinoma, malignant germ cell tumor or sarcoma are called teratocarcinomas. Of the various types of teratoma, the mature teratoma is the most common subtype accounting for 60 percent to 70 percent of germ cell tumors of the mediastinum. It is a benign tumor that is well-differentiated and is usually asymptomatic. Young adults are most frequently affected with no sex predilection identified. Although the course is usually benign, mature teratomas have the potential to undergo transformation into a malignant entity [54]. Of note, if a communication between the mature teratoma and the bronchus is present, then the patient may complain of a productive cough with hair (trichoptysis) or sebaceous material. This presentation, while rare, is virtually pathognomonic of a mature teratoma [55].
4.2
Mediastinal Seminoma
Mediastinal seminoma is the second most common mediastinal GCT, accounting for approximately 25 percent to 50 percent of malignant germ cell tumors [56]. Caucasian males in the second to fourth decade of life are most often affected and can present with symptoms of dyspnea, cough, chest pain, weakness or weight loss. For a germ cell tumor to be classified as a seminoma, no other tumor types can be present histologically. This distinction is important since pure seminomas are more radiosensitive, while tumors comprised of a mixture of seminoma and other tumor types are labeled and treated as NSGCTs since they are more aggressive and less radiosensitive [57]. Furthermore, the use of serum tumor markers can provide additional information. For example, approximately 10 percent of seminomas produce B-hCG, while the AFP level should never be elevated [32].
4.3 Mediastinal Nonseminomatous Malignant Germ Cell Tumors The mediastinal nonseminomatous germ cell tumors (NSGCTs) represent a collection of neoplasms (nonteratomatous/nonseminomatous), including embryonal cell caricinoma, mixed GCTs, endodermal sinus tumor and choriocaricinoma. These tumors are malignant, most often found in young males and are often metastatic at presentation [50, 53]. They are less common than their mediastinal GCT counterparts and their prognosis is much worse. The endodermal sinus tumor is the most common subtype of mediastinal NSGCT. At diagnosis, 85 percent of patients are
6 Imaging of Mediastinal Tumors
159
symptomatic, which includes complaints of chest pain, fever, hemoptysis, cough, weight loss or even superior vena cava syndrome [53]. Serum tumor markers can aid in differentiating the NSGCTs. For example, choriocarcinomas produce B-hCG, and these patients can clinically present with gynecomastia due to these elevated B-hCG levels. Also, an elevated AFP level is suggestive of either an endodermal sinus tumor or embryonal cell carcinoma, and effectively eliminates the possibility of a seminoma. Furthermore, the incidence of elevated AFP is higher in patients with NSGCTs than with gonadal metastases to the mediastinum. Of note, the NSGCTs are associated with numerous hematologic malignancies such as myeloid and lymphoid acute leukemia, malignant histiocytosis and myelodysplastic/myeloproliferative distorders [58, 59]. Also, an association between NSGCTs and Klinefelter’s syndrome (with typical clinical features and the abnormal karyotype 47 XXY) has also been reported in approximately 20 percent of patients [58].
4.4
Staging of Mediastinal GCTS
No specific staging system has been established for primary mediastinal GCTs, though a well-established staging system has been described for gonadal GCTs. The well-established systems to stage other mediastinal tumors, such as thymomas, lymphomas and neuroblastomas, can be applied to mediastinal GCTs.
4.5
Imaging Features
On PA and lateral radiography, mediastinal teratomas are rounded and well-defined with about 26 percent containing calcifications [60] (Fig. 6.5). As mentioned previously, the presence of teeth or bone within the mass on plain film is virtually diagnostic. CT and MRI are useful in further characterization, localizing lesions and evaluating resectibility. With respect to further characterization, additional densities within the mass suggestive of fat, cystic elements and sebum may be identified and are better defined. On CT, these tumors demonstrate a well-defined, thick-walled, cystic mass comprised of fat, soft tissue or water. The presence of fat within the mass is a helpful diagnostic feature further supportive of a diagnosis of benign teratoma with a fat-fluid level sometimes encountered [61] (Fig. 6.5). MRI provides similar information and can be performed in cases of suspected vascular involvement. The plain film findings in seminoma and NSGCT are similar to the mature teratoma, except for the fact that the former demonstrates a lobulated contour and there is notable absence of fat and calcifications. With regard to CT, seminomas are lobulated, asymmetrical and homogeneous tumors that demonstrate metastasis to lymph nodes and, less commonly, to the lungs, bone and liver [32]. NSGCTs are bulky, asymmetrical
160
S. Moore et al.
Fig. 6.5 Mediastinal teratoma. (a, b) Frontal CXR and contrast-enhanced CT demonstrate an anterior mediastinal mass adjacent to the right heart border which contains fat and curvilinear calcifications (arrowheads) consistent with a mediastinal teratoma. (c) Contrast-enhanced CT in a 2-year-old girl reveals an anterior mediastinal immature teratoma (curved arrow) located adjacent to the thymus (arrowhead)
masses that demonstrate areas of contrast enhancement within the mass, interwoven with areas of necrosis and hemorrhage [62]. MRI provides similar information for the malignant GCTs, aiding in suspected vascular invasion.
5
Neurogenic Tumors
Neurogenic tumors are usually located in the posterior mediastinum and constitute about 20 percent of all primary tumors and three-quarters of primary posterior mediastinal masses in adults. In children, they constitute a proportionally greater percentage (34 percent) of mediastinal masses [63]. They originate from the sympathetic ganglia and peripheral nerves. They can be found in both children and adults and range in their
6 Imaging of Mediastinal Tumors
161
behavior from benign to highly aggressive. Approximately 75 percent are benign in adults and around half of the patients are asymptomatic [27] with a greater percentage of malignant cases identified in children. When symptoms are present, they are usually caused by local extension such as pain, cough and dyspnea, as well as Pancoast’s or Horner’s syndromes resulting from involvement of the brachial and/or cervical sympathetic chain. Systemic symptoms related to production of neurohormonal agents may also occur. Schwannoma (neurilemoma), neurofibroma and malignant nerve sheath tumors (MNST) arise from the peripheral nerves [64]. Malignant nerve sheath tumors are uncommon and tend to be associated with neurofibromatosis. Ganglioneuroma, ganglioneuroblastoma and neuroblastomas arise from the sympathetic ganglia [65]. Sympathetic ganglia tumors are more common in children while nerve sheath tumors are most common in adults. Paragangliomas arise from the parasympathetic ganglia and include chemodectomas and pheochromocytomas. With the exception of pheochromocytomas, chemodectomas and neurofibromas of the vagus nerve, these tumors appear as well-defined round or oval masses in the costovertebral gutter. Ganglioneuromas demonstrate a broad base on the spine while the nerve sheath tumors tend to be more spherical. Displacement of adjacent vertebrae or ribs with possible thickening and scalloping of the cortex is often seen. The rib spaces and intervertebral foramina are widened by the tumor. These bony changes are diagnostic of a neurogenic lesion with the only differential diagnosis being a lateral thoracic meningocele. Bony destruction, rather than indentation or displacement, suggests neuroblastoma or a malignant nerve sheath tumor. Calcification is common in the sympathetic nerve tumors in both the benign and malignant forms. Fine or coarse punctuate calcification is more easily visualized on CT than on plain film. Nerve sheath tumors rarely calcify. CT or MRI help define the full extent of the lesion [66]. Prior to intravenous contrast administration, neurogenic tumors may be of decreased density on CT due to the lipid elements of the nerve sheaths. Contrast enhancement of the lesion is a striking feature on CT and is most noticeable with a pheochromocytoma. MRI is very useful for providing accurate information about the size and extent of the mass and its possible extent into the spinal canal. Also, MRI has the advantage of demonstrating the spinal cord and surrounding cerebrospinal fluid without the administration of intrathecal contrast, which is the imaging procedure of choice. The signal intensity on MRI is complex [67].
5.1
Schwannomas and Neurofibromas
Schwannomas and neurofibromas are the most common mediastinal neurogenic tumors and are frequently associated with neurofibromatosis (NF-1) [68]. Neurofibromas can also be associated with tuberous sclerosis, Sturge-Weber syndrome or Von Hippel-Lindau disease. [69, 70]. Both are benign, slow-growing
162
S. Moore et al.
neoplasms that usually arise from a spinal nerve root, but can also involve thoracic nerves. Schwannomas are encapsulated tumors that are often heterogeneous with areas of cystic degeneration, hemorrhage, myelin and calcifications [27]. Neurofibromas tend to be homogeneous, unencapsulated and well-marginated tumors that also originate from the Schwann cells, but contain myelinated and unmyelinated nerve fibers and fibroblasts. Plexiform neurofibromas infiltrate along an entire nerve trunk or plexus. Both schwannomas and neurofibromas are usually solitary except in association with NF-1 [71]. Both tumors have similar clinical features with each sex equally affected and most tumors presenting between 30 to 50 years of age. About 10 percent of schwannomas and neurofibromas grow through the adjacent intervertebral foramen and extend into the spinal canal leading to a “dumbbell” configuration resulting from the large intraspinal and paraspinal portions which are connected by a narrow strip of tissue traversing the intervertebral foramen. About 60 percent of patients with a dumbbell tumor have neurologic symptoms related to epidural cord compression [63]. Cytologically, schwannomas demonstrate an organized pattern of palisading spindle cells (Antoni A) or loose reticular areas with scattered spindle cells (Antoni B). Conversely, neurofibromas histologically demonstrate randomly arranged spindleshaped cells on a matrix of collagen and mucoid material [63]. Many of these tumors are found incidentally on chest radiographs in asymptomatic patients. Both schwannomas and neurofibromas present grossly as lobulated, spherical masses, the extent of which is well-defined with CT or MRI [72] (Fig. 6.6). An MRI should be performed in all patients with a suspected neurogenic tumor to
Fig. 6.6 Neurofibroma. 73-year-old female presenting with hearing loss and intracranial tumors. CXR showed masses in the upper thoracic paraspinal region. Subsequent contrast-enhanced CT scan demonstrates multiple bilateral neurofibromas
6 Imaging of Mediastinal Tumors
163
definitively exclude intraspinal tumor extension. Schwannomas demonstrate a nonhomogeneous, high-intensity appearance on T2-weighted images corresponding to alternating Antoni A and Antoni B areas. A central, high-intensity region in the tumor represents an area of cystic degeneration. On the other hand, neurofibromas demonstrate nodular areas of low signal intensity corresponding to collagenous fibrous tissue and peripheral high-intensity regions corresponding to cystic degeneration on T2-weighted magnetic resonance imaging [67]. The treatment of choice for these tumors is surgical resection.
5.2
Malignant Peripheral Nerve Sheath Tumors
Malignant schwannomas or malignant peripheral nerve sheath tumors (MPNST) are aggressive neoplasms arising from nerve trunks in the posterior mediastinum, or from peripheral nerves. They rarely result from malignant transformation of benign schwannomas, but can arise from either degeneration of neurofibromas in neurofibromatosis (NF-1) or in the region of prior radiation therapy. Clinically, in addition to local symptoms seen with their benign counterparts, patients can present with constitutional symptoms such as fatigue, weight loss, anorexia and fever. There are often multiple local recurrences with local spread to the heart, great vessels, vertebral bodies and intervertebral foramina, and distant hematogenous spread most often to the lung, liver, bone and skin [27]. On histologic examination, these malignant tumors demonstrate a variably cellular spindle cell and myxoid matrix with frequent mitotic figures. Elements such as fat, cartilage, bone and muscle are occasionally found [67]. Radiographically these tumors appear similar to their benign counterparts and are treated with surgical resection whenever possible. Resection is often difficult, secondary to invasion of adjacent vital structures. Due to unresponsiveness to adjuvant therapies the prognosis is extremely poor.
5.3 Neuroblastoma, Ganglioneuroblastoma and Ganglioneuroma Neuroblastoma, ganglioneuroblastoma and ganglioneuroma tumors arise from the sympathetic ganglion cells. These tumors are most often found in children and are rare in adults. [73, 74]. They differ from each other only in their degree of differentiation, with neuroblastoma being a high-grade, undifferentiated neoplasm and ganglioneuroma being a benign tumor with mature ganglion cells. A mediastinal presentation is seen in 15 percent of all neuroblastoma cases, predominantly in the very young [75]. When in the mediastinum, it usually occurs in the posterior compartment. Neuroblastoma is the most frequent cause of childhood mediastinal neurogenic masses. It is a highly aggressive tumor thought to arise from primitive neural crestderived cells called neuroblasts [76]. These neoplasms are highly invasive and frequently
164
S. Moore et al.
metastasize before diagnosis. Their course can be quite variable, sometimes spontaneously regressing, maturing or proliferating aggressively [77]. Unfortunately the majority presents at an advanced stage and do not regress spontaneously or mature. Patients with neuroblastoma or ganglioneuroblastoma can be asymptomatic, but about two-thirds present with dyspnea, malaise, cough and spinal cord compression [27]. Disseminated disease is found in more than half of the patients with mediastinal presentation [75]. Common sites of metastases are the bone marrow, regional lymph nodes, brain, liver and lung. Several paraneoplastic syndromes have been found to be associated with neuroblastoma and ganglioneuroblastoma. Opsoclonus-polymyoclonus syndrome may be related to an autoimmune complex and is characterized by acute cerebellar and truncal ataxia and rapid darting movements of the eyes. Profuse watery diarrhea and abdominal pain related to vasoactive intestinal polypeptide production and pheochromocytoma syndrome caused by catecholamine secretion have also been reported. A 24-hour urine collection should be obtained in children with posterior mediastinal masses to measure catecholamine levels [63]. Histologically, neuroblastomas are composed of small, round, immature cells organized in a rosette pattern. They can be undifferentiated, poorly differentiated or differentiated. Ganglioneuroblastomas exhibit further differentiation in comparison to neuroblastomas and are composed of mature and immature ganglion cells. Two different histologic patterns can be found: intermixed in which the neuroblastic component is seen as multiple microscopic foci, and nodular in which the neuroblastic component is seen as macroscopic nodules [77]. The intermixed subtype ganglioneuroblastoma has a significantly better prognosis than those with the nodular category [78]. The stage of the disease determines therapy (see Table 6-5): stage I, surgical excision; stage II, excision and radiation therapy; stages III and IV, multimodality therapy with surgical debulking, radiation therapy and chemotherapy. Children under the age of one year have an excellent prognosis, even when widespread disease is present. Increasing age and extent of involvement worsens prognosis. Mediastinal neuroblastomas tend to have a better prognosis than neuroblastomas occurring elsewhere, possibly from earlier detection secondary to local symptoms. Ganglioneuromas are mature and benign tumors composed of nerve fibers and mature sympathetic ganglion cells, and are usually located in the paravertebral region. They tend to develop in children and young adults that are older Table 6.5 Staging of Neuroblastoma [64, 65] and Ganglioneuroblastoma [63, 79] Stage Characteristics I IIA IIB III
IV IVS
Well-circumscribed, non-invasive ipsilateral tumor Local invasion without extension across the midline, no nodal involvement Local invasion without extension across the midline, ipsilateral lymph node involvement Tumor extension across the midline with bilateral lymph node involvement or no extension of tumor across the midline with involvement of contralateral lymph nodes Metastatic disease Stage I or II with metastatic disease to the liver, skin and/or bone marrow
6 Imaging of Mediastinal Tumors
165
than individuals presenting with neuroblastoma and ganglioneuroblastoma, and are usually asymptomatic. They can be differentiated radiographically from other neurogenic tumors since they usually do not contain calcifications. Histologically, ganglioneuromas are well encapsulated and exhibit areas of cystic degeneration. Two histologic subtypes exist: maturing and mature [77]. Surgical excision is curative.
5.4
Paraganglioma
There are two types of mediastinal paragangliomas: chemodectomas and pheochromocytomas, both of which may be benign or malignant. While pheochromocytomas are paraganglionic tumors of the autonomic nervous system that store and secrete catecholamines, chemodectomas are non-functioning tumors. Almost all intrathoracic chemodectomas are located close to the aortic arch and are classified as aortic body tumors [80]. The posterior mediastinum is the usual site of intrathoracic pheochromocytoma. They have also rarely been reported to occur in the middle mediastinum, with involvement of the left atrial wall or interatrial septum and aortic arch. All chemodectomas and approximately one-third of mediastinal pheochromocytomas are non-functioning and systemically asymptomatic. The remainder of pheochromocytomas present with findings of catecholamine overproduction such as hypertension, and symptoms of hypermetabolism such as weight loss, hyperhidrosis, palpitations and headaches [81]. Occasionally cardiomyopathy can result from catecholamine excess. Most mediastinal pheochromocytomas are clinically benign, but share histological similarities with their malignant counterparts such as nuclear atypia and pleomorphism. Metastatic disease is the only reliable way to define malignancy. The malignant forms can be locally aggressive and can metastasize to bone, lymph nodes, liver, lung and brain. The paragangliomas have similar appearances on plain chest radiography, CT and MRI. They form rounded, soft tissue masses which are very vascular and enhance brightly on CT [82]. On MRI pheochromocytomas usually show a signal intensity similar to muscle on T1-weighted images, and very high signal intensity on T2-weighted images [83]. Radioiodine MIBG (131I metaiodobenzylguanidine) and somatostatin receptor scintigraphy both show increased activity in pheochromocytomas [83].
6 6.1
Other Mediastinal Tumors Thyroid Masses
Masses of thyroid origin arising in the mediastinum usually represent the downward extension of a multinodular goiter and less often an adenoma or carcinoma.
166
S. Moore et al.
Infrequently, thyroid tumors (adenomas and carcinomas) are completely intrathoracic and develop from heterotopic thyroid tissue in the mediastinum [84]. Most commonly this heterotopic thyroid tissue is located in the anteriosuperior mediastinum, but is present to a lesser degree in the middle and posterior mediastinum, respectively. Establishing continuity between the mediastinal mass and the thyroid gland is a key imaging objective in distinguishing between the two previous entities. An intrathoracic extension of a substernal goiter will also have a vascular supply connected to the thyroid gland, while a completely intrathoracic thyroid tumor derives vascular supply from vessels in the thoracic region. Thyroid carcinoma arising from these heterotopic rests of tissue may represent any of the forms identified in the normal gland such as papillary, follicular, medullary, anaplastic and other histological morphologies. Mediastinal thyroid carcinomas usually affect older women and are invasive neoplasms causing symptoms such as dyspnea, chest pain, cough and wheezing secondary to compression of the trachea. With respect to thyroid masses, plain film radiography demonstrates a welldefined lobular mass that invariably displaces the trachea. Calcifications are commonly seen in benign disease, but may also be found in the malignant entities as well. CT can identify these calcifications much better than plain films, as well as identify multiple rounded low density regions most common in multinodular goiter. Welldefined rounded calcifications favor a benign process while amorphous “cloudlike” calcifications are seen in carcinomas. CT is also helpful in defining the size, shape, composition and location of the mass. A mass derived from the thyroid is usually well-defined and located in the paratracheal or retrotracheal region surrounded by the brachiocephalic vasculature, and most commonly connected to the cervical thyroid gland. Futhermore, on CT thyroid tissue demonstrates higher attenuation than adjacent muscle on pre-contrast and post-contrast imaging [85]. Determining a benign from a malignant mass on CT is impossible unless the tumor has spread beyond the boundaries of the thyroid gland. Although MRI, like CT, can distinguish between solid and cystic regions, it has the disadvantage of not identifying calcifications. Cystic regions corresponding to goiter demonstrate increased signal intensity on T2-weighted imaging. Also, adenomas and carcinomas are not distinguishable since both often demonstrate increased signal intensity on T2-weighted imaging. Lastly, radionuclide investigation with radioactive iodine (I-131 and I-123) can be utilized to detect functioning thyroid tissue within a mediastinal mass suspected to be of thyroid origin. However, CT is generally regarded as the test of choice as it can provide more anatomic information in cases of non-thyroid related tumors, and it can be used to detect the thyroid origin of a tumor.
6.2
Primary Mesenchymal Malignant Tumors
These are uncommon benign and malignant tumors of mesenchymal origin with malignant capability, especially in children [27, 86]. The malignant entities include
6 Imaging of Mediastinal Tumors
167
Fig. 6.7 Mediastinal metastases. 46-year-old male with a history of testicular carcinoma and CXR finding of a mediastinal mass. Contrast-enhanced CT demonstrates a large heterogeneous mass in the posterior mediastinum consistent with metastatic disease
liposarcoma, fibrosarcoma, malignant fibrous histiocytoma, hemangioendothelioma, angiosarcoma, leiomyosarcoma, rhabdomyosarcoma, chondrosarcoma and osteosarcoma.
6.3
Metastases
Metastatic disease should always be considered in the differential diagnosis of a mediastinal tumor. Metastases to the mediastinum are most commonly found in bronchogenic carcinoma, head and neck cancers, genitourinary tract, thyroid, melanoma and breast cancers. Imaging features of metastases can be difficult to distinguish from other etiologies of mediastinal tumors (Fig. 6.7).
7
Interventional Techniques for Histologic Sampling
Although history/physical examination, radiology (plain film radiography, CT, MRI) and serum tumor markers assist in narrowing the broad spectrum of entities in the mediastinum, definitive diagnosis is based on histopathological correlation.
168
S. Moore et al.
This is required since treatment options and prognosis vary, depending on specific tumor type. Percutaneous biopsy of the mediastinum is most often performed in conjunction with Ultrasound (US) and CT guidance. The other techniques to obtain biopsies include mediastinoscopy, transbronchial needle aspiration biopsy, US-guided transbronchial endoscopic biopsy, US-guided transesophageal endoscopic biopsy and US-guided supraclavicular lymph node biopsy [87, 90]. The above techniques are inherently limited by various factors including location of tumors, operator skill and relative invasiveness. In contrast, percutaneous mediastinal biopsies allow the physician access to virtually all mediastinal regions, including those that are inaccessible with the aforementioned procedures. Imaging-guided percutaneous needle biopsy of mediastinal tumors is a safe and effective technique for obtaining tissue samples. Biopsy needles can be classified as small caliber (20 to 25 gauge) or large caliber (14 to 19 gauge), and as aspiration or cutting needles [91]. Aspiration needles provide specimens suitable for cytologic evaluation, while cutting needles provide core specimens for histologic evaluation. The coaxial technique – which involves initial placement of a guide needle close to the target lesion, followed by advancement of the biopsy needle through this needle to obtain tissue samples – is the most commonly used technique for mediastinal biopsies [91] (Fig. 6.8). Both fine needle aspiration and core biopsy are well accepted techniques for obtaining diagnostic tissue from mediastinal lesions. Since
Fig. 6.8 CT-guided biopsy of mediastinal mass. CT-guided core biopsy of an anterior mediastinal mass. Axial non-contrast CT image demonstrates a 17-gauge coaxial biopsy needle (curved arrow) within a large anterior mediastinal mass (arrowhead) with a final pathologic diagnosis of hemorrhagic thymic cyst
6 Imaging of Mediastinal Tumors
169
adequate sampling is required to assess mediastinal tumors secondary to the complex histology of this diverse population of neoplasms, a percutaneous core needle biopsy is favored with regards to initial diagnosis of thymoma, mediastinal lymphoma, germ cell tumors and neurogenic tumors [92, 93]. A direct mediastinal approach to biopsy is the preferred method of obtaining tissue samples for histological analysis. This entails placement of the biopsy needle through an extrapleural space (medial to lung tissue) to prevent the risk of pneumothorax. A transpulmonary approach can also be performed, but the risk of pneumothorax is greater and this technique is employed if the direct mediastinal path is not feasible. The various methods of a direct mediastinal approach revolves around the site of advancement of the needle and includes the following approaches: parasternal, paravertebral, transsternal and suprasternal [91, 94, 95, 96]. Numerous factors affect the decision to biopsy a mediastinal tumor via US or CT, including location of the mass and institutional preference. US guidance aids the clinician with reference to constant real-time imaging of the needle, tumor and for evaluating adjacent vessels that lie in the proposed path of biopsy. The availability of coronal and oblique needle paths, as well as the ability to perform the biopsy in positions that would not be possible with CT, are other advantages of US guidance. The latter scenario is especially useful in debilitated patients at the bedside or in a slightly upright position (due to the preclusion of lying prone due to dyspnea or oxygen requirements) [97]. However, CT is the most common technique employed, as it provides more options for biopsy approach and operator preference.
8
Future Developments
PET-CT is an evolving imaging technique with exciting potential breakthroughs in tumor imaging. Through the depiction of the metabolic state of tumor cells PET, in conjunction with CT, might provide insights into the proliferative or malignant potential of disease. Although research is ongoing, this modality has recently been at the forefront of radiological development. The most researched application of this imaging modality, with respect to the mediastinum, lies in the re-staging of lymphomas. As mentioned previously, PET-CT is extensively utilized for re-staging of mediastinal lymphomas. Although research is limited, a promising new application of PET-CT concerns pretreatment staging and assessment of response during therapy (therapy monitoring). The use of PET-CT for therapy monitoring may prove to be the most beneficial application for patient management and outcome, so as to provide an early assessment of response with the goal of therapy adjustment. In seminomas PET-CT has proven to be a very useful tool in evaluating postchemotherapy masses. Some studies have also shown PET-CT can be used, in addition to current imaging techniques, to detect post-chemotherapy residual germ cell tumor of testicular and extragonadal origin. Research is still ongoing in the use of PET-CT in NSGCT, but it may help determine the need for surgery in patients with a posttreatment residual masses seen on CT and normal serum tumor markers [98].
170
S. Moore et al.
Fig. 6.9 Fused PET/CT image in a patient with mediastinal thymoma. The transverse PET/CT image shows an anterior mediastinal mass with increased metabolic activity (arrow)
Only a few studies have determined the usefulness of PET-CT in thymoma and thymic carcinomas (Fig. 6.9). They have shown promise in the use of PET-CT to differentiate the thymic epithelial subgroups, and to determine the extent of disease spread to the lymph nodes and pleura [99]. Finally, there is limited research on PET-CT and neuroblastoma. Findings on PET-CT appear to correlate with disease status (as determined by various other imaging modalities) as it may provide information on the proliferative or malignant potential of disease, based on metabolic activity. Furthermore, in chemotherapy patients who have measurable lesions on standard imaging modalities, PET-CT may demonstrate normal or minimally abnormal distribution of radiotracers. This finding could represent inactive or regressing, rather than active, disease which would give supportive evidence for the current treatment regimen [100].
Conclusion In summary, tumors of the mediastinum reflect an array of neoplastic processes, each with a certain predilection for various anatomic compartments and demographic populations. Patients may present with symptoms such as chest pain, cough, dyspnea, fever, chills and, rarely, superior vena cava syndrome. Plain film radiography is the starting point for evaluation, but CT is the gold standard technique for further characterization of mediastinal tumors. As we have seen, this information, along with history/physical examination and serum tumor markers, can assist in narrowing the differential diagnosis, but a definitive answer rests with histopathological correlation. Advancements in interventional techniques, such as percutaneous biopsy of the mediastinum via US or CT guidance, have led to the development of a safe and effective
6 Imaging of Mediastinal Tumors
171
alternative for gathering tissue samples in practically all mediastinal regions. Finally, continued research, especially with regard to PET-CT, promises further exciting breakthroughs in tumor imaging (with emphasis in oncologic practice) via pretreatment staging and assessment of patient response during therapy.
References 1. Wychulis AR, Payne WS, Clagett OT, et al. Surgical Treatment of mediastinal tumors. J Thorac Cardiovasc Surg 1972; 62:379-91. 2. Fraser RS, Pare JAP, Fraser RG, et al. The normal chest. In: Fraser RS, Pare JAP, Fraser RG, et al, eds. Synopsis of diseases of the chest. 2nd ed. Philadelphia: WB Saunders, 1994; 1-116. 3. Davis, RD Jr, Oldham, HN Jr, Sabiston, DC Jr. Primary cysts and neoplasms of the mediastinum: Recent changes in clinical presentation, methods of diagnosis, management and results. Ann Thorac Surg 1987; 44:229-237. 4. Gaerte, SC, Meyer, CA, Winer-Muram, HT, et al. Fat-containing lesions of the chest. RadioGraphics 2002; 22 Spec No:S61. 5. Grillo, HC, Ojemann, RG, Scannell, JG, et al. Combined approach to “dumbbell” intrathoracic and intraspinal neurogenic tumors. Ann Thorac Surg 1983; 36:402. 6. Friedberg JW, Fishman A, Neuberg D, et al. FDG-PET is superior to gallium scintigraphy in staging and more sensitive in the follow-up of patients with de novo Hodgkin’s Lymphoma: a blinded comparison. Leuk Lymphoma. 2004; 45:85–92. 7. Osserman KE, Genkins G. Studies in myasthenia gravis: review of a 20-year experience in over 1200 patients. Mt Sinai J Med 1971; 38:497–537. 8. Marx A, Muller-Hermelink HK, Strobel P. The role of thymomas in the development of myasthenia gravis. Ann N Y Acad Sci 2003; 998:223–236. 9. Thomas CR, Wright CD, Loehrer PJ. Thymoma: State of the art. J Clin Oncol 1999; 17(7):2280-2289. 10. Patterson GA. Thymomas. Semin Thorac Cardiovasc Surg 1992; 4(1):39-44. 11. Souadjian JV, Enriquez P, Silverstein MN, et al. The spectrum of diseases associated with thymoma. Coincidence or syndrome? Arch Intern Med 1974; 134(2):374-379. 12. Masaoka A, Monden Y, Nakahara K, et al. Follow-up study of thymomas with special reference to their clinical stages. Cancer 1981; 48:2485. 13. Monden Y, Uyama T, Taniki T, et al. The characteristics of thymoma with myasthenia gravis: A 28-year experience. J Surg Oncol 1988; 38(3):151-154. 14. Chalabreysse L, Roy P, Cordier JF, et al. Correlation of the WHO schema for the classification of thymic epithelial neoplasms with prognosis: A retrospective study of 90 tumors. Am J Surg Pathol 2002; 26(12):1605-1611. 15. Rosai J, Levine GD. Tumors of the thymus. In: Firminger HI, ed. Atlas of tumor pathology. Washington, DC: Armed Forces Institute of Pathology, 1976; 34–212. 16. Lewis JE, Wick MR, Scheithauer BW, et al. Thymoma: a clinicopathologic review. Cancer 1987; 60:2727–2743. 17. Verstandig AG, Epstein DM, Miller WT, et al. Thymoma-report of 71 cases and a review. Crit Rev Diagn Imaging 1992; 33:201–230. 18. Okumura M, Ohta M, Tateyama H, et al. The World Health Organization histologic classification system reflects the oncologic behavior of thymoma: a clinical study of 273 patients. Cancer 2002; 94:624–632. 19. Wilkins EW Jr, Edmunds L Jr, Castleman B. Cases of thymoma of the Massachusetts General Hospital. J Thorac Cardiovasc Surg 1966; 52:322–330. 20. Shamji F, Pearson FG, Todd TR, et al. Results of surgical treatment for thymoma. J Thorac Cardiovasc Surg 1984; 87:43–47.
172
S. Moore et al.
21. Duwe B, Sterman D, Musani A. Tumors of the Mediastinum. Chest 2005; 128(4):2893-2909. 22. Tamura Y, Kuroiwa T, Doi A, et al. Thymic carcinoma presenting as cranial metastasis with intradural and extracranial extension: case report. Neurosurgery 2004; 54:209-211. 23. Yaqub A, Munn NJ, Wolfer RS. Thymic carcinoma presenting as cardiac tamponade. South Med J 2004; 97:212–213. 24. Chung DA. Thymic carcinoma-analysis of nineteen clinicopathological studies. Thorac Cardiovasc Surg 2000; 48(2):114-119. 25. Walker AN, Mills SE, Fechner RE. Thymomas and thymic carcinomas. Semin Diagn Pathol 1990; 7(4):250-265. 26. Leyvraz S, Henle W, Chahinian AP, et al. Association of Epstein-Barr virus with thymic carcinoma. N Engl J Med 1985; 312(20):1296-1299. 27. Macchiarini P, Ostertag H. Uncommon Primary Mediastinal Tumours. The Lancet Oncology 2004; 5(2): 107-118. 28. Chaer R, Massad MG, Evans A, et al. Primary neuroendocrine tumours of the thymus. Ann Thorac Surg 2002; 74:1733-40. 29. Gibril F, Chen YJ, Schrump DS, et al. Prospective study of thymic carcinoids in patients with multiple endocrine neoplasia type 1. J Clin Endocrinol Metab 2003; 88:1066–1081. 30. Moran CA, Suster S. Neuroendocrine carcinomas (carcinoid tumor) of the thymus. A clinicopathologic analysis of 80 cases. Am J Clin Pathol 2000; 114(1):100-110. 31. Wick MR, Bernatz PE, Carney JA, et al. Primary mediastinal carcinoid tumors. Am J Surg Pathol 1982; 6:195–205. 32. Strollo DC, Rosado-de-Christenson ML, Jett JR, et al. Primary mediastinal tumors: Part 1. Tumors of the anterior mediastinum. Chest 1997; 112:511-522. 33. Armstrong P, Padley S. The mediastinum. Grainger RG, Allison DJ, Dixon AK 4th ed. Diagnostic Radiology: A Textbook of Medical Imaging 2001 New York (NY): Churchill Livingstone: pp 289-299; 353-376. 34. Chew FS, Weissleder R. Mediastinal thymolipoma. Am J Roentgenol 1991; 157:468. 35. Ellis K, Gregg HE. Thymomas: Roentgen considerations. Am J Roentgenol 1964; 91:105-119. 36. Kelleher P, Misbah SA. What is Good’s syndrome? Immunological abnormalities in patients with thymoma. J Clin Pathol 2003; 56:12–16. 37. Strickler JG, Kurtin PJ. Mediastinal lymphoma. Semin Diagn Pathol 1991; 8:2–13. 38. Lichtenstein AK, Levine A, Taylor CR, et al. Primary mediastinal lymphoma in adults. Am J Med 1980; 68:509–514. 39. Cartwright R, Brincker H, Carli PM, et al. The rise in incidence of lymphomas in Europe. Eur J Cancer 1999; 35:627–633. 40. Vaeth JM, Moskowitz SA, Green JP. Mediastinal Hodgkin’s disease. AJR Am J Roentgenol 1976; 126:123–126. 41. Kornstein MJ, DeBlois, et al. Pathology of the thymus and mediastinum. 1st ed. Philadelphia, PA: WB Saunders, 1995. 42. Sutcliff SB. Primary mediastinal malignant lymphoma. Semin Thorac Cardiovasc Surg 1992; 4:55–67. 43. Castellino, RA. The non-Hodgkin’s lymphomas: practical concepts for the diagnostic radiologist. Radiology 1991; 178:315-321. 44. Lazzarino M, Orlandi E, Paulli M, et al. Primary mediastinal B-cell lymphoma with sclerosis: an aggressive tumor with distinctive clinical and pathologic features. J Clin Oncol 1993; 11:2306-2311. 45. Harris NL, Jaffe ES, Stein H, et al. A revised European-American classification of lymphoid neoplasms: a proposal from the international lymphoma study group. Blood 1994; 84:1361-1392. 46. Blank N, Castellino RA. The mediastinum in Hodgkin’s and Non-Hodgkin’s Lymphomas. J Thorac Imaging 1987; 2:66-71. 47. Rahmouni A, Tempany C, Jones R, et al. Lymphoma: monitoring tumor size and signal intensity with MR imaging. Radiology 1993; 188:445-451.
6 Imaging of Mediastinal Tumors
173
48. Juweid, M. Utility of Positron Emission Tomography (PET) Scanning in managing patients with Hodgkin’s Lymphoma. Hematology 2006; 1:259-265. 49. Jhanwar Y, Straus D. The role of PET in Lymphoma. J. Nucl. Med. 2006; 47:1326-1334. 50. Nichols DR. Medisatinal germ cell tumors: clinical features and biologic correlates. Chest 1991; 99:472-479. 51. Luna MA, Valenzuela-Tamariz J. Germ-cell tumors of the mediastinum, postmortem findings. Am J Clin Pathol 1976; 65(4):450-454. 52. Hainsworth JD, Greco FA. Extragonadal germ cell tumors and unrecognized germ cell tumors. Semin Oncol 1992; 19(2):119-127. 53. Knapp, RH, Hart, RD, Payne, WS. Malignant germ cell tumors of the mediastinum. J Thorac Cardiovasc Surg 1985; 89:82. 54. Donadio AC, Motzer RJ, Bajorin DF, et al. Chemotherapy for teratoma with malignant transformation. J. Clin Oncol 2003; 21:4285-4291. 55. Adebonojo SA, Nicola ML. Teratoid tumors of the mediastinum. Am Surg 1976; 42:361. 56. Javadpour, N. Significance of elevated serum alphafetoprotein (AFP) in seminoma. Cancer 1980; 45:2166. 57. Schantz A, Sewell W, Castleman B. Mediastinal germinoma: A study of 21 cases with an excellent prognosis. Cancer 1972; 30:1189. 58. Dexeus FH, Logothetis CJ, Chong C, et al. Genetic abnormalities in men with germ cell tumors. J Urol 1988; 140(1):80-84. 59. Nichols CR, Roth BJ, Heerema N, et al. Hematologic neoplasia associated with primary mediastinal germ-cell tumors. N Engl J Med 1990; 322(20):1425-1429. 60. Lewis BD, Hurt RD, Payne WS, et al. Benign teratoma of the mediastinum. J Thorac Cardiovasc Surg 1983; 86:727-731. 61. Moeller KH, Rosado-de-Christenson ML, Templeton PA. Mediastinal mature teratoma: imaging features, AJR 1997; 169:985-990. 62. Lee KS, Im JG, Han CH, et al. Malignant primary germ cell tumors of the mediastinum: CT features. AJR 1989; 153:947-951. 63. Lau CL, Davis RD. The Mediastinum. Townsend CM, Beauchamp RD, Evers BM, Mattox KL 17th ed. Townsend: Sabiston Textbook of Surgery 2004 Philadelphia (PA): Elsevier: pp 1739-1759. 64. Evans AE, D’Angio GJ, Sather HN, et al. A comparison of four staging systems for localized and regional neuroblastoma: A report from the Children’s Cancer Study Group. J Clin Oncol 1990; 8:678-688. 65. Brodeur GM, Pritchard J, Berthold F, et al. Revisions of the international criteria for neuroblastoma diagnosis, staging, and response to treatment [comment]. J Clin Oncol 1993; 11:1466-1477. 66. Reed JC, Hallet KK, Feigin DS. Neural tumours of the thorax: subject review from the AFIP. Radiology 1978; 126:9-17. 67. Sakai F, Sone S, Kiyono K, et al. Intrathoracic neurogenic tumors: MR-pathologic correlation. Am J Roentgenol 1992; 159:279-283. 68. Brasfield RD, Das Gupta TK. Von Recklinghausen’s disease: A clinicopathological study. Ann Surg 1972; 175(1):86-104. 69. D’Agostino AN, Soule EH, Miller RH. Sarcomas of the peripheral nerves and somatic soft tissues associated with multiple neurofibromatosis. Cancer 1963; 16:1015. 70. Thomas JV, Schwartz PL, Gragoudas ES. Von Hippel’s disease in association with von Recklinghausen’s neurofibromatosis. Br J Ophthalmol 1978; 62(9):604-608. 71. Izumi AK, Rosato FE, Wood MG. Von Recklinghausen’s disease associated with multiple neurolemomas. Arch Dermatol 1971; 104(2):172-176. 72. Coleman BG, Arger PH, Dalinka MK, et al. CT of sarcomatous degeneration in neurofibromatosis. Am J Roentgenol 1983; 140:383-387. 73. Tateishi U, Hasegawa T, Makimoto A, et al. Adult neuroblastoma: Radiologic and clinicoathologic features. J Comput Assist Tomogr 2003; 27(3):321-326.
174
S. Moore et al.
74. Prece V, Bertagni A, Gallinaro L, et al. Neurogenic tumors of the mediastinum. Ann Ital Chir 2002; 73(2):125-127. 75. Grosfeld JL, Baehner RL. Neuroblastoma: an analysis of 160 cases. World J Surg 1980; 4(1):29-37. 76. Pizzo PA, Poplack DG. Principles and Practice of Pediatric Oncology, 4th ed.. Philadelphia: Lippincott Williams & Wilkins; 2002:1692. 77. Shimada H, Ambros IM, Dehner LP, et al. Terminology and morphologic criteria of neuroblastic tumors: Recommendations by the international neuroblastoma pathology committee. Cancer 1999; 86:349-363. 78. Shimada H, Umehara S, Monobe Y, et al. International neuroblastoma pathology classification for prognostic evaluation of patients with peripheral neuroblastic tumors: A report from the Children’s Cancer Group. Cancer 2001; 92:2451-2461. 79. Crapo JD, Glassroth J, Karlinsky J, et al. Baum’s textbook of pulmonary diseases. Philadelphia, PA: Lippincott Williams & Wilkins, 2004; 883–912. 80. Olson JL, Salyer WR. Mediastinal paragangliomas (aortic body tumour): a report of four cases and a review of the literature. Cancer 1978; 41:2405-2412. 81. Scott Jr HW, Reynolds V, Green N, et al. Clinical experience with malignant pheochromocytomas. Surg Gynecol Obstet 1982; 154(6):801-818. 82. Spizarny DL, Rebner M, Gross BH, et al. CT evaluation of enhancing mediastinal masses. J Comput Assist Tomogr 1987; 11:990-993. 83. Van Gils APG, Falke THM, van Erkel AR, et al. MR imaging and MIBG scintigraphy of pheochromocytomas and extraadrenal functioning paragangliomas. RadioGraphics 1991; 11:37-57. 84. Hall, TS, Caslowitz P, Popper C, et al. Substernal goiter versus intrathoracic aberrant thyroid: a critical difference. Ann Thorac Surg 1988; 46; 684-685. 85. Glazer GM, Axel L, Moss AA. CT diagnosis of mediastinal thyroid. AJR 1982; 138:495-498. 86. Swanson PE. Soft tissue neoplasms of the mediastinum. In: Wick MR, Santa Cruz DJ, editors. Seminars in diagnostic pathology: Mediastinal Pathology Philadelphia: WB Saunders Company; 1991. p. 14-34. 87. Larsen SS, Krasnik M, Vilmann P, et al. Endoscopic ultrasound guided biopsy of mediastinal lesions has a major impact on patient management. Thorax 2002; 57:98–103. 88. Herth F, Becker HD, Ernst A. Conventional vs endobronchial ultrasound-guided transbronchial needle aspiration: a randomized trial. Chest 2004; 125:322–325. 89. Hsu LH, Liu CC, Ko JS. Education and experience improve the performance of transbronchial needle aspiration: a learning curve at a cancer center. Chest 2004; 125:532–540. 90. Van Overhagen H, Brakel K, Heijenbrok MW, et al. Metastases in supraclavicular lymph nodes in lung cancer: assessment with palpation, US, and CT. Radiology 2004; 232:75–80. 91. Gupta S, Seaberg K, Wallace MJ, et al. Imaging-guided percutaneous biopsy of mediastinal lesions: different approaches and anatomic considerations. RadioGraphics 2005; 25:763–788. 92. Protopapas Z, Westcott JL. Transthoracic hilar and mediastinal biopsy. Radiol Clin North Am 2000; 38:281–291. 93. Yonemori, K, Tsuta U, Tateishi, H, et al. Diagnostic accuracy of CT-guided percutaneous cutting needle biopsy for thymic tumours. Clinical Radiology 2006 61:771-775. 94. Gupta S, Wallace MJ, Morello FA Jr, Ahrar K, Hicks ME. CT-guided percutaneous needle biopsy of intrathoracic lesions by using the transsternal approach: experience in 37 patients. Radiology 2002; 222:57–62. 95. Gupta S, Gulati M, Rajwanshi A, Gupta D, Suri S. Sonographically guided fine-needle aspiration biopsy of superior mediastinal lesions by the suprasternal route. Am J Roentgenol 1998; 171:1303–1306. 96. Astrom KG, Ahlstrom KH, Magnusson A. CT-guided transsternal core biopsy of anterior mediastinal masses. Radiology 1996; 199:564–567.
6 Imaging of Mediastinal Tumors
175
97. Rubens DJ, Strang JG, Fultz PJ, Gottlieb RH. Sonographic guidance of mediastinal biopsy: an effective alternative to CT guidance. Am J Roentgenol 1997; 169:1605–1610. 98. Johns Putra L, Lawrentschuk N, Ballok Z, Hannah A, Poon A, Tauro A, et.al.: 18F-fluorodeoxyglucose positron emission tomography in evaluation of germ cell tumor after chemotherapy. Urology 2004; 64(6): 1202-7. 99. Sung Y, Lee K, Kim BChoi J, Shim Y, and Yi C. 18F-FDG PET/CT of Thymic Epithelial Tumors: Usefulness for Distinguishing and Staging Tumor Subgroups. J. Nucl. Med. 2006; 47:1628-1634. 100. Kushner, Brian H. Neuroblastoma: A Disease Requiring a Multitude of Imaging Studies. J. Nucl. Med. 2004; 45:1172-1188.
7
Imaging Cardiac Tumors Mannudeep K. Kalra, MD and Suhny Abbara, MD
Key Points ● ● ●
●
●
● ●
● ●
1
Cardiac neoplasms are rare Metastases are the most common intracardiac tumors Cross-sectional imaging can help in localization and characterization of cardiac tumors Most literature on the role of imaging in cardiac tumors comes from retrospective case series and review articles Transthoracic echocardiography is generally sufficient for detecting cardiac masses and should be the first line of imaging MRI is critical for localization and characterization of cardiac tumors CT can be used in patients in whom MRI cannot be performed, such as those with pacemakers, metallic foreign body in the eyes, ferromagnetic or electronically operated stapedial implants and cerebral hemostatic clips CT is also the best imaging test for detecting calcification Imaging is also helpful in follow-up for recurrence or residual lesions
Introduction
Imaging of the heart has evolved rapidly in the last decade. With rapid advances in computed tomography (CT) and the introduction of multislice CT, it is possible to image the heart in less than a 10 seconds breath-hold [1]. Magnetic resonance imaging (MRI) of the heart has evolved to become the imaging modality of choice for evaluation of pericardial and cardiac tumors [2]. This chapter addresses the role of cross-sectional imaging, particularly CT and MRI, for evaluation of cardiac tumors.
Division of Cardiac Imaging, Massachusetts General Hospital, 25 New Chardon St, Boston, MA 02114. email:
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
177
178
2
M.K. Kalra and S. Abbara
Classification of Tumors of the Heart
For sake of simplicity, we have classified tumors of the heart into those arising from the pericardium and those from the heart itself. Metastatic tumors represent the most common tumors of the pericardium [3]. Most frequently, metastases occur from lung or breast cancers, lymphoma and leukemia. Autopsy studies suggest that almost one-quarter of all patients dying from cancer have pericardial metastasis. Primary neoplasms of the pericardium are exceedingly rare [4]. The most common primary pericardial tumor is the primary pericardial mesothelioma. Other primary neoplasms include benign and malignant teratoma, pheochromocytoma, angiosarcoma and fibrosarcoma [5]. Metastases are also the most common cardiac neoplasms, and may involve the heart via direct extension from local juxta-cardiac malignancies, lymphatic spread, venous extension or hematogenous spread [6]. Primary cardiac tumors are uncommon and include benign and malignant neoplasms originating from the myocardium, endocardium or cardiac valves [7-9]. Myxoma, lipoma, papillary fibroelastoma, fibroma, rhabdomyoma and hemangioma are some of the benign cardiac tumors. Malignant cardiac tumors include angiosarcoma, rhabdomyosarcoma, malignant fibrous histiocytoma, intracardiac lymphoma and chondrosarcoma. Thomas-de-Montpreville, et al. have proposed that, irrespective of patients’ age, cardiac tumors may be classified as congenital tumors with spontaneous non-progressive or regressive lesions possibly needing surgery for mass effect, acquired benign tumors needing surgery for risk of thromboembolism and, finally, the remaining primary and secondary neoplasms with globally poor prognosis, but with some indications for resection nevertheless [10].
3
Clinical Presentation
Cardiac tumors are rare. Several autopsy series in unselected patients have a reported incidence between 0.0017 percent and 0.19 percent [11]. The first antemortem recognition of a intracardiac tumor was reported in 1934 although the first surgical excision of an intracardiac tumor, a left atrial myxoma, did not happen until 1955. Dr. Crafoord worked with Viking Olov Bjork and Ake Senning to improve the heart-lung machine and used it to perform the successful resection of myxoma of the left atrium, with long-term survival [12]. With remarkable growth in the use of cross-sectional imaging, particularly in CT and MRI, cardiac masses may be picked up in asymptomatic patients. However, patients with primary cardiac tumors may present with one or more symptoms from the classic triad of symptoms related to right or left ventricular outflow or inflow obstruction (such as those of congestive heart failure, atypical chest pain or palpitation), symptoms and signs of systemic thrombo-embolization (from stroke), and nonspecific constitutional symptoms [5]. Clinical presentation with intracardiac obstruction and thrombo-embolization occurs in only about half of the cases [8]. In
7 Imaging Cardiac Tumors
179
general, signs and symptoms depend on the location of the tumor and its spread. Pericardial location or spread may be marked by pericardial pain, effusion, tamponade, constriction or predominantly atrial arrhythmias. Myocardial location or involvement may be associated with arrhythmias, conduction disturbances, heart blocks, congestive heart failure and EKG changes. Coronary involvement may present as angina pectoris or with myocardial infarction. Endocardial tumors may lead to valve obstruction, valve damage, thromboembolism and constitutional symptoms. Treatment of intra-cardiac metastases is generally dictated by the status and extent of the primary tumor. For benign tumors surgery is the treatment of choice, whereas chemotherapy is indicated in the presence of unresectable or widespread malignancies [5].
4
Differential Diagnosis from Non-Tumorous Masses
One important consideration with imaging of a suspected cardiac or pericardial tumor is the possibility of a normal cardiac structure or a non-neoplastic mass lesion mimicking neoplasm [13]. In the pericardium lesions such as hematoma, abscess, pleuropericardialcysts, and hydatid cysts may present as masses and confound the diagnosis. Often, crosssectional imaging modalities such as MRI (preferably) and CT scanning (as an alternative to MRI) can help in differentiating these lesions from solid pericardial masses or tumors involving the pericardium. In the heart intracardiac thrombus, septal aneurysm, crista terminalis, prominent trabeculae or papillary muscles, hydatid cyst, abscess, vegetations, benign lipomatous hypertrophy of the interatrial septum and aneurysms can be misinterpreted as cardiac neoplasms. Most often, echocardiography, transthoracic and/or transesophageal, can help in this differential diagnosis, particularly with reference to a left atrial thrombus. MRI and CT can also help in making a decision regarding characterization of such non-neoplastic cardiac masses. MRI is particularly useful in accurate characterization of ventricular thrombi, as thrombi usually do not show contrast enhancement after gadolinium administration on first pass or delayed MRI [14]. Post-contrast MRI in such cases can also show the underlying myocardial scar and wall motion abnormality, and help in differential diagnosis from cardiac tumor. However, at times, artifacts from slow flowing blood may also simulate cardiac masses on MRI.
5
Imaging of Cardiac Tumors
Cross-sectional imaging techniques such as echocardiography, CT and MRI can help in characterizing some tumors based on the patient’s age, medical history, location of the tumor, tumor extension, morphology and mobility, attenuation value or signal intensity, and contrast enhancement pattern [15-20].
180
M.K. Kalra and S. Abbara
Cross sectional imaging techniques such as echocardiography, CT and MRI provide information about the size, shape, location, vascularity and mobility of cardiac tumors along with their relation and local extension to adjoining cardiac and noncardiac thoracic structures. CT and MRI can also aid in the detection and evaluation of primary extracardiac malignancy with metastasis to the heart or pericardium. Additionally, MRI helps in tissue characterization of the cardiac tumors and its differentiation from thrombi and normal structures simulating a tumor.
5.1
Chest Radiography and Fluoroscopy
Chest radiography is limited as an evaluation tool for cardiac tumors, as chest radiographs may be completely normal in many cases. However, chest radiographs may show a lung mass silhouetting the cardiac border and provide a clue towards a possible malignant lung mass invading the pericardium. Although not as sensitive as CT, radiographs can depict lung, lymph nodal or bone metastases of breast, lymphoma and melanoma, and corroborate with a diagnosis of metastatic deposit to the heart. In addition, tumors presenting with left-sided obstructive physiology can be assessed for pulmonary venous hypertension (redistribution of pulmonary vascularity, interstitial or alveolar edema), and those presenting with right-sided obstructive physiology can be assessed for signs of systemic venous hypertension (superior vena cava and azygos vein enlargement) on chest radiography. Barium swallow with fluoroscopy provides valuable information about suspected esophageal cancer invading the pericardium. It is relatively contraindicated, however, in the presence of suspected tracheo-esophageal fistula or broncho-esophageal fistula secondary to tumoral extension.
5.2
Catheter Angiography
Catheter angiography has a limited role in imaging of patients with suspected cardiac tumors and has been replaced by echocardiography, MRI and CT in most instances. Major limitations of the technique include inability to directly visualize or characterize the tumor, risk of tumor embolization and the possibility of missing small or intramural tumors. Since tumors are not directly visualized on catheter angiography, it is generally not possible to differentiate non-neoplastic cardiac lesions, such as thrombi and cysts, from neoplasms. On the other hand, catheter angiography can provide information on the vascular supply to the tumor and the status of coronary circulation. Some authors suggest that coronary angiography may be done when patients are older than 40 years of age, have risk factors for coronary artery disease or suspected involvement of the coronary arteries [11]. The most important role of cardiac catheterization in cardiac tumors is that it may be used for performing biopsy.
7 Imaging Cardiac Tumors
5.3
181
Echocardiography
Transthoracic echocardiography is an important evaluating modality for cardiac tumors. It is a widely available and relatively inexpensive way of evaluating cardiac tumors in an accurate, real-time manner. Generally, in most cases, information about tumor size, location, extent, mobility and its relation to adjacent structures can be obtained with echocardiography [21]. Additionally, cardiac valves and functions can also be assessed. Transesophageal echocardiography can help in evaluating cardiac tumors, especially those in the left atrium. Limitations of transthoracic echocardiography include the lack of tumor characterization and suboptimal quality in large patients, and in those with chronic obstructive pulmonary disease. Evaluation of pericardial tumors and other intrathoracic tumors involving the heart or the pericardium is also limited with echocardiography.
5.4
Magnetic Resonance Imaging
MRI has established a unique position in imaging of the heart and pericardium. It is often performed to confirm a mass seen on echocardiography and provide further information about the mass. MRI helps in accurate evaluation of cardiac and pericardial tumors for their size, morphology, location, extent and relation to important structures such as valves, septum and ventricular outflow and inflow tracts. Extension into the myocardium and pericardium and involvement of lung and mediastinum can be better assessed with MRI [22, 23]. Of all the cross-sectional imaging modalities for the heart, MRI is the most valuable in terms of characterizing cardiac masses as solid or cystic, mobile or immobile, thrombus or fat-containing masses. MRI can specifically characterize benign tumors of the heart such as myxoma, lipoma, fibroma and hemangioma [24]. In addition, CINE MRI pulse sequences provide information on the mobility of the mass during the cardiac cycle. MRI also aids in the evaluation of cardiac function, in the presence of obstructive cardiac tumors. The ability to generate direct images within any cardiac plane enables interpreting physicians to evaluate the relationship between tumor and adjacent contiguous and non-contiguous cardiac and non-cardiac structures. The absence of ionizing radiation with MRI makes it more suitable for follow-up of patients with cardiac tumors, particularly for young patients with benign tumors. Contrary to echocardiography, MRI provides a much wider field of view, and is not limited by acoustic windows or large body habitus. Massachusetts General Hospital’s MRI protocol for imaging of patients with known or suspected cardiac tumors is summarized in Table 7.1. The presence of cardiac pacemakers, metallic foreign body in the eyes, ferromagnetic or electronically operated stapedial implants and cerebral aneurysm clips are contraindications for MRI. In such circumstances, CT may provide some of the required information [25].
182
M.K. Kalra and S. Abbara
Table 7.1 MRI protocol for imaging of patients with known or suspected cardiac tumor used at Massachusetts General Hospital. To focus the study on the location of the mass, all relevant prior studies (CT, echocardiography, cardiac catheterization) are checked prior to the MRI. (Key: SSFP, single shot free precession; ETL, echo train length; SE, spin echo; PD, proton density; MDE, myocardial delayed enhancement; TE, echo time; FSE, fast spin echo; TI, inversion time) Pulse-Sequence
Orientation
Comments
3-plane localizer Sequential 2D SSFP localizer Cine SSFP
Axial, Coronal, Saggital Axial
Entire chest in one breath hold
4-chamber
Cine SSFP
Through mass
Double IR T2 FSE
In best plane visualizing mass
PD FSE with fat sat
Use orthogonal view (2 or 4 chamber view) Axial (may change orientation depending on mass location) Same view as prior sequence
T1 SE
T1 SE post double dose of gadolinium (0.2 mmol/kg) 3D MDE (2D MDE Short axis optional)
6 to 8 slices to cover left ventricle (varies according to tumor location) To cover mass if visible on prior sequences 4 to 5 slices; TE ~ 100-120; ETL ~ 20
Look for enhancement of the mass Determine appropriate TI time at base, repeat at apex and through mass
Gadolinium-based MRI contrast agents are relatively contraindicated in patients with advanced renal insufficiency, due to risk of nephrogenic systemic fibrosis.
5.5
CT
In the past five years, multi-detector-row CT scanning has emerged as one of the most promising modalities for non-invasive imaging of the heart and coronary vasculature [26-29]. Cardiac CT examinations are done with electrocardiographic (EKG) gating within a single-breath-hold duration [30]. Iodinated contrast agents are used to opacify the cardiac chambers and coronary arteries. EKG-gated cardiac CT studies allow evaluation of cardiac chambers and coronary vessels. CT is the investigation of choice for evaluation of juxta-cardiac malignancies invading or metastasizing to the heart or the pericardium. In patients who cannot undergo MRI due to any reason as discussed above, CT may be used to assess suspected cardiac tumors. CT is also the best imaging technique for evaluation of cardiac calcifications and may be used as a problem-solving technique (e.g., cardiac masses versus tumoral mitral annular calcification).
7 Imaging Cardiac Tumors
183
Although CT does provide superior spatial resolution and faster scanning, the information on tumor characterization is limited, compared to MRI. The functional cardiac information obtained from MRI is also superior to CT in terms of cardiac output and flow mapping. It is important to remember that cardiac CT studies are done with iodinated contrast agents, which are nephrotoxic in the presence of prior or existent renal dysfunction. Patients with compromised renal function secondary to widespread malignant disease, co-existing diabetes mellitus and chemotherapy may be more vulnerable to the contrast media inducednephropathy. In general, estimated glomerular filtration rate (eGFR) is a better index of the risk of contrast-induced nephropathy than serum creatinine. Proper hydration prior to administration of the contrast injection may help in decreasing the risk of nephropathy in these patients.
6
Level of Evidence for Imaging of Cardiac Tumors
Neoplastic cardiac masses are rare entities and that may explain the lack of prospective randomized controlled trials for imaging of cardiac neoplasms. Most published studies are indeed retrospective reviews of imaging findings, case series or case reports of rare cardiac tumors [2-24, 31, 32].
7
Malignant Cardiac Tumors
Cardiac masses are more likely to be malignant if they are present on the right side of the heart, and have extracardiac extension, inhomogeneous signal intensity and associated pericardial effusion [24, 31]. In addition, evidence of a wide mural attachment, destruction of the cardiac chamber wall, invasion of the pericardium, pulmonary arteries, vein, or vena cava, involvement of two cardiac chambers and presence of multiple lesions also favor malignancy. Imaging of malignant cardiac tumors can help in detection and localization of the lesions, in assessment of their relation to adjoining critical structures, local invasion and distant metastases. The ultimate diagnosis about the exact type of malignancy frequently requires invasive procedures such as biopsy or surgery.
7.1
Metastases
The reported incidence of cardiac metastases in different published series varies widely [11, 33]. Cardiac metastases are about 100- to 1,000-fold more common than primary cardiac tumors [34]. Cardiac metastases most frequently originate from the lungs, hemopoietic system, breast and gastrointestinal and genitourinary tracts [35]. Almost 50 percent of patients who die from melanoma have cardiac
184
M.K. Kalra and S. Abbara
metastases at autopsy [11]. At autopsy, almost 40 percent of patients dying from hematogenous tumors (leukaemia > lymphoma) have cardiac involvement [34]. About 90 percent of patients with cardiac metastases are clinically silent. In a patient with known malignancy development of tachycardia, arrhythmias, cardiomegaly or congestive cardiac failure, cardiac metastases must be suspected. Twothirds of cardiac metastases occur in the pericardium, one-third in the myocardium and about 5 percent in the endocardium. Most patients with pericardial metastasis clinically present with a pericardial effusion or pericarditis. Pathologically, pericardial metastases may appear as massive infiltration of the pericardium, fibrinohemorrhagic pericarditis, pericardial infiltrate, direct invasion of pericardium or one or more pericardial nodules and masses. MRI provides the most comprehensive information about pericardial tumors, including metastases. In patients with pericardial invasion from tumors originating in contiguous structures such as lungs, mediastinum and esophagus, obliteration of the fat plane between the tumor and pericardium, pericardial enhancement and effusion are suggestive of invasion on MRI and CT. In cases of transvenous extension of tumor thrombi, for example with renal or hepatocellular cancer extending into the inferior vena cava, or lung cancer extending into the superior vena cava and pulmonary veins, both MRI and CT can provide critical information regarding the presence and extent of tumor thrombi in the veins as well as into the cardiac chambers (Fig. 7.1). In addition, both modalities allow differentiation between thrombus and tumor, as only the latter shows contrast enhancement on either imaging modality. On the other hand, endocardial metastases can present as multiple small endocardial lesions, a large intracardiac cavitary mass, massive neoplastic thrombosis or those invading the coronary veins or arteries. The right atrium and ventricles tend to be more commonly involved with endocardial metastases [33]. These lesions have no specific CT or MRI features to differentiate them from primary malignant tumors of the heart. On MRI they have non-specific characteristics, such as low T1 signal intensity and high T2 signal intensity, with variable contrast enhancement [22]. A notable exception is melanoma metastases, which have high signal intensity on both T1- and T2-weighted MR images. Also, in patients with carcinoid syndrome, tricuspid valve disease can be evaluated with echocardiography and MRI [36, 37]. Echocardiography demonstrates specific tricuspid valve abnormalities such as thickening, shortening and decreased mobility of regurgitant leaflets, and thickening and doming in the presence of a stenotic valve [36]. MRI has a distinct advantage over CT, as it can image valve motion and valvular dysfunction [37]. This is particularly true for right-sided valves which are difficult to assess with CT. Limited valvular functional evaluation by CT is possible for the aortic and mitral valves [38, 39]. There are no specific appearances of metastases or direct extension to the heart; malignant tumors, in general, have low signal intensity on T1-weighted images and high signal intensity on T2-weighted images with varying degrees of enhancement after contrast material administration [22]. The only exception, again, is metastatic melanoma, which may be bright on both T1- and T2-weighted images due to the presence of large amounts of paramagnetic melanin.
7 Imaging Cardiac Tumors
185
Fig. 7.1 80-year-old man with widespread metastases from malignant melanoma. Contrastenhanced CT images show a large, enhancing, broad–based, well-defined mass (*) representing a metastatic deposit in the right atrium (1a, mediastinal window), and a lung metastasis (arrow) in the right lower lobe (1b, lung window)
7.2
Angiosarcoma
Angiosarcomas are the most common primary malignant tumors of the heart, and second most common primary cardiac tumor after myxoma (the third most common cardiac tumor is lipoma), constituting slightly more than one-third of all such tumors. These tumors present with non-specific clinical signs and symptoms, typically in the fifth and sixth decade of life, with men being two times more likely to have the tumor, compared with women, and the right atrium being the preferential location. Almost 50 percent to 90 percent of patients with angiosarcoma develop metastases to lungs, brain, bone and colon, which are quite frequently present at the time of presentation [5, 11]. Pathologically, these tumors are poorly defined, often hemorrhagic, aggressive lesions that invade contiguous structures such as vena cava
186
M.K. Kalra and S. Abbara
and tricuspid valve. Microscopic examination of angiosarcoma shows atypical mesenchymal cells lining the anastomotic vascular spaces. Two morphologic types of angiosarcomas have been described: a well-defined exophytic mass-like lesion of the right atrium and a diffusely infiltrative type of lesion extending along the pericardium [16]. Both CT and MRI reflect the pathologic features of angiosarcomas. On MRI, the tumor has a heterogeneous T1 signal intensity due to interspersed regions with solid tumor tissue, necrosis and hemorrhage (methemoglobin). Heterogeneous contrast uptake in the tumor is common with marked peripheral surface enhancement and little to no central enhancement corresponding to an area of necrosis. In addition, both MRI and CT can show the invasion of adjoining structures, including the pericardium and great vessels. Pericardial involvement may be depicted as discrete enhancing mass or masses, or effusion. Since lungs are the most frequent sites of metastases from cardiac angiosarcoma, a chest CT may be included in the management. There is little literature on use of fluorine-18 fluordeoxyglucose (FDG) positron emission tomography (PET) and PET-CT for evaluation of right atrial angiosarcoma, its local recurrence and detection of metastatic lesions [40, 41].
7.3
Rhabdomyosarcoma
There are two types of rhabdomyosarcoma: the embryonal type – the most common primary malignant tumor of the heart in children – and the adult type, which is more pleomorphic and less common. Contrary to other primary cardiac malignancies, these tumors have equal prevalence on either side of the heart, and are more likely to involve the cardiac valves. The epicenter of these tumors lies in the myocardium, although they frequently involve the pericardium. They frequently become quite bulky in size, measuring up to 10 cm. On MRI the tumor is isointense to the myocardium on T1-weighted sequences with relatively homogeneous contrast enhancement, although an area of necrosis within the tumor may give rise to a more heterogeneous enhancement pattern.
7.4
Other Sarcomas: Undifferentiated Sarcoma, Osteosarcoma, Fibrosarcoma, Liposarcoma and Leiomyosarcoma
In the older literature undifferentiated sarcoma was reported to be the most common sarcoma of the heart [16]. The diagnosis of undifferentiated pleomorphic sarcoma is made after excluding other sarcomas, with appropriate use of tissue sampling and other ancillary diagnostic techniques (Fig. 7.2). In the heart the most common location of origin of the undifferentiated sarcoma, osteosarcoma or leiomyosarcoma is the left atrium, and they present most frequently with pulmonary congestion. The most common site for liposarcoma is an atrial chamber without any side preference. Liposarcoma and fibrosarcoma can also originate directly from the pericardium and invade the underlying myocardium.
7 Imaging Cardiac Tumors
187
Fig. 7.2 47-year-old man with pathologically proven spindle cell sarcoma of the left atrium. T2(2a) and post-contrast T1- (2b) weighted MR images show an ill-defined infiltrating mass in the superior portion of the left atrium involving the ostium of left superior pulmonary vein (arrow). A three-dimensional projection PET image (2c) shows an area of increased glucose uptake in the area corresponding to left atrium (arrow). (Note: the left ventricle shows variable uptake under normal circumstances). (Ao, aorta; LA, left atrium; RVOT, right ventricle outflow tract)
These five sarcomas do not have any specific gross pathology features and may show occasional areas of hemorrhage. Whereas calcification on CT may help in suggesting the diagnosis of osteosarcoma, identification of macroscopic fat in an invasive or aggressive tumor on CT or MRI may help confirm a diagnosis of liposarcoma. The role of imaging, however, is not to diagnose the specific cell type of these tumors, but to help in determining the anatomical limits of the tumors (local spread to cardiac, pericardial and mediastinal structures, and distant metastases),
188
M.K. Kalra and S. Abbara
their functional implications on the heart (valvular motion, venous return, cardiac output) and follow-up after surgery or chemotherapy.
7.5
Primary Cardiac Lymphoma
Up to 20 percent of patients with widely disseminated malignant lymphoma have cardiac or pericardial involvement on autopsy [42]. Primary cardiac lymphomas are rare in immunocompetent patients, and represent 1.3 percent of all cardiac tumors [42]. However, its incidence is increasing due to acquired immunodeficiency syndrome (AIDS) and patients who have received transplantations [11]. There is some controversy regarding the definition of primary cardiac lymphoma. Some require complete absence of lymphoma outside the pericardium on autopsy, while others assert that, for diagnosis, the bulk of tumor must be within the pericardium or cardiac symptoms must be present at the time of initial presentation [43]. Imaging, particularly CT and MRI, plays a vital role in staging of the tumor, as well as defining the extent of spread of the tumor. These tumors are more common on the right side of the heart, with right ventricle and right atrium being the most common sites of origin. They frequently invade the pericardium. Both MRI and CT have a distinct advantage over echocardiography due to a wider field of view and the ability to image tumor extension into the pericardium, or the great vessels. On CT and MRI, primary cardiac lymphomas have non-specific features and cannot be differentiated from other malignant tumors of the heart. Like most malignant neoplasms of the heart, lymphomas are also isointense to myocardium on T1-weighted images, and heterogeneous on T2weighted images with heterogeneous contrast enhancement.
7.6
Pericardial Mesothelioma
Primary pericardial mesothelioma is a rare malignancy which is often lethal. Most patients are males between 30 to 50 years of age [32]. A primary pericardial mesothelioma originates from mesothelial cells of the pericardium. MRI is the imaging modality of choice for demonstrating the nature and, more importantly, the extent of the tumor, and the infiltration into the cardiac wall and great vessels [44]. Most often, this tumor forms multiple coalescing masses in the pericardium, which are isointense to the myocardium on T1weighted images and heterogeneous on T2-weighted images, with marked contrast enhancement [22]. MRI also depicts the presence of constriction of the pericardium, secondary to the malignancy [45]. These findings can help guide the surgical resection, which is more likely to be successful in the presence of a localized tumor.
8
Benign Cardiac Tumors
Most primary cardiac tumors are benign. In contrast to the malignant tumors of the heart, benign tumors tend to have better prognosis. Surgery is the mainstay of treatment for benign tumors, whereas chemotherapy is the preferred treatment for most
7 Imaging Cardiac Tumors
189
malignant cardiac tumors due to the usual presence of distant metastases at presentation and the extent of local invasion. The benign tumors tend to be more frequent on the left side, particularly from the interatrial septum or roof of left atrium, although they can occur in any cardiac chamber. Likewise, a mobile or pedunculated tumor is also more likely to represent a benign tumor such as a myxoma (Fig. 7.3) or papillary fibroelastoma (Fig. 7.4). On the other hand, a broad based lesion may be a benign or malignant tumor. Clinically the signs and symptoms of benign tumors depend on their location. Some benign tumors are related to syndromes including tuberous sclerosis (rhabdomyoma), Gorlin syndrome (fibroma) and Carney’s complex (cardiac myxomas, endocrine hyperfunction and areas of skin pigmentation). Cross-sectional imaging is important for tumor detection, as well as treatment planning. Table 7.2 summarizes the types of benign cardiac tumors along with site
Fig. 7.3 (continued)
190
M.K. Kalra and S. Abbara
Fig. 7.3 46-year-old woman presented with worsening of shortness of breath with histopathologically proven left atrial myxoma. Cine-SSFP (steady state free precession) MR images (3a, ventricular systole; 3b, diastole) and contrast–enhanced, cardiac-gated CT (3c, ventricular systole; 3d, diastole) in four-chamber plane demonstrates a fairly large left atrial mass (arrow) arising from the interatrial septum and prolapsing into the mitral valve annulus in diastole (LA, left atrium; RV, right ventricle; RA, right atrium; LV, left ventricle)
preponderance, and most frequent pathologic manifestations [4, 5, 7, 8, 11]. Imaging features of these tumors are described in Table 7.3 [4, 16, 19, 22, 32, 46, 47]. Cardiac teratomas are rare primary tumors that occur most frequently in infants and children. Generally, cardiac teratoma originates from the pericardium and
7 Imaging Cardiac Tumors
191
Fig. 7.4 60-year-old woman presented with transient neurologic deficit. A small pedunculated mass arising from the aortic side (contrary to vegetations, which occur on ventricular surface) of the non-coronary cusp of the aortic valve (4a, sytole; 4b, diastole; 4c, LVOT in systole) was detected on EKG-gated cardiac CT angiography. This mass was resected and turned out to be a papillary fibroelastoma (Ao, aorta; LVOT, left ventricle outflow tract)
intracardiac location is very rare [32]. As these tumors are rarely malignant, surgery is generally curative despite the frequently bulky size of the tumor. Imaging helps in detection, localization and surgical planning. Thus, echocardiography and MRI are the mainstay imaging modalities for evaluation of benign cardiac tumors. They provide information on size, location, extent and important relations of the tumors. In patients with multiple rhabdomyomas, echocardiography can help in follow-up of the tumors for progression or regression.
30-60 years
Middle-aged, elderly
Any
Myxoma (5.7 cm)
Papillary fibroelastoma (<1 cm)
Lipoma
30-40 years
Children
Lymphangioma
Common
Left atrium
Broad based infiltrative or cir- Rare cumscribed lesion Pericardial space Multilocular cystic tumor Rare
Circumscribed firm tumors
Rare
Rare
Capsulated Multiple mural masses
Left ventricle
Rare
Common
Rare
Rare
Common
Endothelium lined thin wall spaces
Enlarged vacuolated cells Spindle cells (fiboblasts) Chromaffin cells
Adipose tissue and muscles
Hyperplastic endothelium
Myxoma cells
Hemorrhage and Cell type Calcification necrosis
Rare
Typical gross features
Interatrial septum Soft, gelatinous or firm Common 75 percent left smooth, bosselated surface atrium 15-20 percent right atrium Aortic and mitral Resembles a fuzzy ball Rare valvular endoattached to short pedicle cardium Any Broad based tumor Rare
Most common location
Children, young adults Left ventricle
Paraganglioma
Fibroma (5 cm)
Rhabdomyoma (4 cm) Children
Age at Presentation
Type
Table 7.2 Pathologic features, age of presentation and most common locations of benign cardiac tumors. Figures in the parenthesis suggest mean size of the lesion [4, 5, 7, 8, 11]
192 M.K. Kalra and S. Abbara
7 Imaging Cardiac Tumors
193
Table 7.3 Imaging features of benign cardiac tumors (# T2 bright relative to the myocardium) [4, 16, 19, 22, 32, 46, 47] Type
Echocardiography
Myxoma
Broad based myxoma- non-specific
CT and MRI
Narrow base supports diagnosis of myxoma Narrow stalk, mobility and distensi- Heterogeneous on T1-weighted MR, bility - are diagnostic Bright on T2#, and shows heteroegeneous contrast enhancement Small, mobile, valvular masses with May not be apparent. Papillary fibroelasstalk (<1 cm) Small pedunculated tumor attached to toma the valves Stippled edge with “shimmer” at interface with blood Lipoma Pericardial – Uniformly echogenic Permit unequivocal diagnosis of fat and or hypoechoic or with hypoelipoma when capsule is identified choic areas; Intracardiac – hyper(lipomatous hypertrophy in atrial echoic septum is dumb-bell shape from sparing of fossa ovale) Rhabdomyoma Homogeneous masses with nonHyperintense on T1 and T2 images, contractile myocardium strong enhancement Cine MR or CT: Non-contractile area Fibroma Large, noncontractile, solid mass in CT: Low attenuation lesion with dysa ventricular wall trophic calcification MRI: Hypointense on T2, isointense on T1 images Calcification may appear as hypointense T1 and T2 core Either: Little or no contrast enhancement Paraganglioma Echogenic mass, immobile CT: Low attenuation mass, strong enhancement MRI: Very bright on T2 images, marked enhancement Lymphangioma Heterogeneous, internal septae, hyp- CT: Low attenuation heterogeneous oechoic intrapericardial mass mass, ± internal septae MRI: T2 bright lesion with septations and heterogeneous enhancement
The degree of cardiac functional impairment can also be judged on echocardiography and MRI.
Summary Cross-sectional imaging techniques such as echocardiography and MRI provide important information about cardiac masses. Echocardiography can provide information about cardiac tumors with regard to size, location and response to treatment. MRI is helpful in determining tumor extent, mobility, location, local invasion, relationship with critical cardiac and non-cardiac structures and tumor characterization. For pericardial tumors and metastatic lesions, MRI appears to be the imaging modality of choice.
194
M.K. Kalra and S. Abbara
References 1. Kalra MK, Maher MM, D’Souza R, Saini S. Multidetector computed tomography technology: current status and emerging developments. J Comput Assist Tomogr. 2004;28 Suppl 1:S2-6. 2. Schvartzman PR, White RD. Imaging of cardiac and paracardiac masses. J thorac Imaging. 2000;15:265-73. 3. Basso C, Valente M, Poletti A, Casarotto D, Thiene G. Surgical pathology of primary cardiac and pericardial tumors. Eur J Cardiothorac Surg. 1997;12:730-7 4. Grebenc ML, Rosado de Christenson ML, Burke AP, Green CE, Galvin JR. Primary cardiac and pericardial neoplasms: radiologic-pathologic correlation.Radiographics. 2000;20:1073-103 5. Majano-Lainez RA. Cardiac tumors: a current clinical and pathological perspective. Crit Rev Oncog. 1997;8:293-303. 6. Hoffmeier A, Schmid C, Deiters S, Drees G, Rothenburger M, Tjan TD, Schmidt C, Loher A, Maintz D, Spieker T, Mesters RM, Scheld HH. Neoplastic heart disease – the Muenster experience with 108 patients. Thorac Cardiovasc Surg. 2005;53:1-8. 7. Endo A, Ohtahara A, Kinugawa T, Nawada T, Fujimoto Y, Mashiba H, Shigemasa C. Clinical incidence of primary cardiac tumors. J Cardiol. 1996;28:227-34. 8. Odim J, Reehal V, Laks H, Mehta U, Fishbein MC. Surgical pathology of cardiac tumors. Two decades at an urban institution. Cardiovasc Pathol. 2003 Sep;12:267-70. 9. Piazza N, Chughtai T, Toledano K, Sampalis J, Liao C, Morin JF. Primary cardiac tumours: eighteen years of surgical experience on 21 patients. Can J Cardiol. 2004;20:1443-8. 10. Thomas-de-Montpreville V, Nottin R, Dulmet E, Serraf A. Heart tumors in children and adults: clinicopathological study of 59 patients from a surgical center. Cardiovasc Pathol. 2007;16:22-8. 11. Butany J, Nair V, Naseemuddin A, Nair GM, Catton C, Yau T. Cardiac tumors: diagnosis and management. Lancet Oncol. 2005;6:219-28. 12. Limet R. The beginnings of cardiac surgery. Rev Med Liege. 2006;61:812-9. 13. Broderick LS, Brooks GN, Kuhlman JE. Anatomic pitfalls of the heart and pericardium. Radiographics. 2005;25:441-53. 14. Kaminaga T, Takeshita T, Kimura I. Role of magnetic resonance imaging for evaluation of tumors in the cardiac region. Eur Radiol. 2003;13 Suppl 4:L1-10 15. Gulati G, Sharma S, Kothari SS, Juneja R, Saxena A, Talwar KK. Comparison of echo and MRI in the imaging evaluation of intracardiac masses. Cardiovasc Intervent Radiol. 2004;27(5):459-69. 16. Araoz PA, Mulvagh SL, Tazelaar HD, Julsrud PR, Breen JF. CT and MRI of benign primary cardiac neoplasms with echocardiographic correlation. Radiographics. 2000;20:1303-19. 17. Krombach GA, Spuentrup E, Buecker A, Mahnken AH, Katoh M, Temur Y, Higgins CB, Gunther RW. Heart tumors: magnetic resonance imaging and multislice spiral CT. Rofo. 2005;177:1205-18. 18. Chiles C, Woodard PK, Gutierrez FR, Link KM. Metastatic involvement of the heart and pericardium: CT and MRI. Radiographics. 2001;21:439-49. 19. Restrepo CS, Largoza A, Lemos DF, Diethelm L, Koshy P, Castillo P, Gomez R, Moncada R, Pandit M.CT and MRI findings of malignant cardiac tumors. Curr Probl Diagn Radiol. 2005;34:1-11. 20. Restrepo CS, Largoza A, Lemos DF, Diethelm L, Koshy P, Castillo P, Gomez R, Moncada R, Pandit M.CT and MRI findings of benign cardiac tumors. Curr Probl Diagn Radiol. 2005;34:12-21. 21. Ragland MM, Tak T.The role of echocardiography in diagnosing space-occupying lesions of the heart. Clin Med Res. 2006;4:22-32. 22. Sparrow PJ, Kurian JB, Jones TR, Sivananthan MU. MRI of cardiac tumors. Radiographics. 2005;25:1255-76. 23. Baldassarre S, Cappeliez O, Leone A, Divano L.Cardiac magnetic resonance. Myth and reality. Rev Med Brux. 2004;25:80-6.
7 Imaging Cardiac Tumors
195
24. Luna A, Ribes R, Caro P, Vida J, Erasmus JJ. Evaluation of cardiac tumors with magnetic resonance imaging. Eur Radiol. 2005;15:1446-55. 25. Nazarian S, Roguin A, Zviman MM, Lardo AC, Dickfeld TL, Calkins H, Weiss RG, Berger RD, Bluemke DA, Halperin HR. Clinical utility and safety of a protocol for noncardiac and cardiac magnetic resonance imaging of patients with permanent pacemakers and implantablecardioverter defibrillators at 1.5 tesla. Circulation. 2006;114:1277-84. 26. Mollet NR, Cademartiri F, Van Mieghem C, Meijboom B, Pugliese F, Runza G, Baks T, Dikkeboer J, McFadden EP, Freericks MP, Kerker JP, Zoet SK, Boersma E, Krestin GP, de Feyter PJ. Adjunctive value of CT coronary angiography in the diagnostic work-up of patients with typical angina pectoris. Eur Heart J. 2007 Mar 9; [Epub ahead of print] 27. Schussler JM, Smith ER. Sixty-four-slice computed tomographic coronary angiography: will the “triple rule out” change chest pain evaluation in the ED? Am J Emerg Med. 2007;25:367-75. 28. Rispler S, Keidar Z, Ghersin E, Roguin A, Soil A, Dragu R, Litmanovich D, Frenkel A, Aronson D, Engel A, Beyar R, Israel O. Integrated single-photon emission computed tomography and computed tomography coronary angiography for the assessment of hemodynamically significant coronary artery lesions. J Am Coll Cardiol. 2007;49:1059-67. 29. Meijboom WB, Mollet NR, van Mieghem CA, Weustink AC, Pugliese F, van Pelt N, Cademartiri F, Vourvouri E, de Jaegere P, Krestin GP, de Feyter PJ. 64-slice Computed Tomography Coronary Angiography in Patients with Non-ST Elevation Acute Coronary Syndrome. Heart. 2007 Mar 7; [Epub ahead of print] 30. Dewey M, Hamm B. CT Coronary Angiography: Examination Technique, Clinical Results, and Outlook on Future Developments. Rofo. 2007;179:246-260. 31. Kar AK, Roy S, Chatterjee A, Banerjee A, Panja M, Mitra S. Cardiac tumors: an observational study. Indian Heart J. 1996;48:257-60. 32. Tatli S, Lipton MJ.CT for intracardiac thrombi and tumors. Int J Cardiovasc Imaging. 2005;21:115-31. 33. Bussani R, De-Giorgio F, Abbate A, Silvestri F. Cardiac metastases. J Clin Pathol. 2007;60:27-34. 34. Sarjeant JM, Butany J, Cusimano RJ. Cancer of the heart: epidemiology and management of primary tumors and metastases. Am J Cardiovasc Drugs. 2003;3:407-21 35. Butany J, Leong SW, Carmichael K, Komeda M. A 30-year analysis of cardiac neoplasms at autopsy. Can J Cardiol. 2005;21:675-80. 36. Forman MB, Byrd BF 3rd, Oates JA, Robertson RM. Two-dimensional echocardiography in the diagnosis of carcinoid heart disease. Am Heart J. 1984;107:492-6. 37. Mollet NR, Dymarkowski S, Bogaert J. MRI and CT revealing carcinoid heart disease. Eur Radiol. 2003;13 Suppl 6:L14-8. 38. Boxt LM. CT of valvular heart disease. Int J Cardiovasc Imaging. 2005;21:105-13. 39. Vogel-Claussen J, Pannu H, Spevak PJ, Fishman EK, Bluemke DA. Cardiac valve assessment with MRI and 64-section multi-detector row CT. Radiographics. 2006 Nov-Dec;26(6): 1769-84. 40. Freudenberg LS, Rosenbaum SJ, Schulte-Herbruggen J, Eising EG, Lauenstein T, Wolff A, Bockisch A. Diagnosis of a cardiac angiosarcoma by fluorine-18 fluordeoxyglucose positron emission tomography. Eur Radiol. 2002;12 Suppl 3:S158-61. 41. Juergens KU, Hoffmeier A, Riemann B, Maintz D. Early detection of local tumour recurrence and pulmonary metastasis in cardiac angiosarcoma with PET-CT and MRI. Eur Heart J. 2007;28:663. 42. Ryu SJ, Choi BW, Choe KO. CT and MR findings of primary cardiac lymphoma: report upon 2 cases and review. Yonsei Med J. 2001;42:451-6. 43. Ikeda H, Nakamura S, Nishimaki H, Masuda K, Takeo T, Kasai K, Ohashi T, Sakamoto N, Wakida Y, Itoh G. Primary lymphoma of the heart: case report and literature review. Pathol Int. 2004;54:187-95. 44. Eren NT, Akar AR. Primary pericardial mesothelioma. Curr Treat Options Oncol. 2002;3:369-73.
196
M.K. Kalra and S. Abbara
45. Oreopoulos G, Mickleborough L, Daniel L, De Sa M, Merchant N, Butany J. Primary pericardial mesothelioma presenting as constrictive pericarditis. Can J Cardiol. 1999;15:1367-72. 46. Wintersperger BJ, Becker CR, Gulbins H, Knez A, Bruening R, Heuck A, Reiser MF. Tumors of the cardiac valves: imaging findings in magnetic resonance imaging, electron beam computed tomography, and echocardiography. Eur Radiol. 2000;10:443-9. 47. Grebenc ML, Rosado-de-Christenson ML, Green CE, Burke AP, Galvin JR. Cardiac myxoma: imaging features in 83 patients. Radiographics. 2002;22:673-89.
8
Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies Unni Udayasankar, MD, Abbas Chamsuddin, MD, Pardeep Mittal, MD, and William C. Small, MD, PhD
The unique histological characteristics, combined with a high vascularity and biochemical load, increase the vulnerability of hepatic tissue for the development of primary and metastatic tumors. The number of cases of hepatocellular carcinoma (HCC) has increased over the past two decades [1], and approximately 10,000 people are estimated to die annually from primary hepatic and biliary cancers in the United States [2]. Hepatic metastases contribute to mortality in a large number of patients. Recent advances in imaging techniques, including molecular and cellular biology, assist in early diagnosis, treatment and follow-up of hepatobiliary tumors [3-7]. Imaging plays an important role, since the diagnosis and management of many liver tumors has witnessed a sharp change in the last two decades. Imaging modalities such as ultrasound and computerized tomography (CT) are frequently used for staging cancer patients for hepatic metastatic tumor spread, and for evaluating patients with preexisting liver disease (chronic hepatitis or cirrhosis) to look for development of potential primary malignancies. Contrast-enhanced CT, magnetic resonance imaging (MRI) or molecular imaging techniques, including positron emission tomography (PET), are used to estimate the tumor burden, vascular or biliary invasion and residual normal liver volume. Imaging also facilitates diagnosis and treatment by guided procedures including percutaneous biopsy, radiofrequency ablation (RFA) and transarterial chemoembolisation (TACE).
1
Hepatic Metastases
Autopsy studies reveal the presence of liver metastases in approximately 25 percent to 50 percent of all patients who die of malignant disease [8]. Tumors arising from colon, lung, pancreas, melanoma and sarcoma have a relatively high propensity for blood-borne metastasis to the liver [9].
Emory University School of Medicine, Atlanta, Georgia
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
199
200
U. Udayasankar et al.
The size, morphology, contrast enhancing characteristics, growth rate and response of adjacent liver tissue vary widely in liver metastases, but some characteristic patterns are present [10]. Treatment options vary depending on the site and histology of the primary cancer and overall tumor staging. Preoperative work-up of any patient with metastatic disease should include quantification of lesion number, identification of hepatic segment(s) involved, proximity to hepatic arterial, portal venous and bile duct branches, and extra-capsular extension and invasion of adjacent organs. Pathological and vascular characteristics of most metastases match those of the primary tumor. Most metastases are hypovascular, show lack of glycogen, and increased water content. Consequently, large proportions of liver metastases are hypodense on non-enhanced CT studies. A hypodense lesion with amorphous calcification is highly suggestive of metastatic spread from mucinous adenocarcinoma arising from the stomach or colon [10, 11] (Fig. 8.1). Cystic metastatic hepatic lesions may be present, irrespective of the morphology of primary neoplasm. The primary tumors may be cystic (ovarian carcinoma, mucinous cystadenocarcinoma of the pancreas) or solid (gastrointestinal stromal tumor, leiomyosarcoma, malignant melanoma, carcinoid, neuroendocrine tumor, pheochromocytoma and neuoblastoma) (Fig. 8.2). Cystic secondaries may also be seen when the primary tumor undergoes radio- or chemotherapy, especially with newer anti-cancer drugs including imatinib (Gleevac) [12-14]. Most liver imaging protocols benefit from the use of intravenous contrast agents. Optimum use of these agents requires an appreciation of the fact that the liver receives both arterial and portal (venous) blood supplies, allowing imaging to be
Fig. 8.1 Cystic metastases. Trasverse CT image obtained in a 54-year-old-man with carcinoid tumor of distal ileum shows multiple intrahepatic cystic lesions. Note fluid-fluid levels in some of the lesions (arrow)
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
201
Fig. 8.2 Calcified solitary liver metastasis. Transverse CT image of a 66-year-old-man with biopsy proven mucinous adenocarcinoma of the colon shows ill-defined high density calcified lesion in the right lobe of liver (arrow)
performed during phases of either predominantly arterial or portal liver enhancement. Liver metastases, particularly the extremely common example of spread of colorectal primaries, are usually demonstrated as hypovascular, low density masses during the portal venous phase of imaging (Fig. 8.3). Hypervascular hepatic metastases are seen in primary breast, renal cell, thyroid carcinomas, pancreatic neuroendocrine tumors, pheochromocytoma, malignant melanoma, carcinoid tumor and sarcomas [15-17], and require a different scanning protocol for optimal lesion detection, traditionally liver imaging during the arterial phase of contrast enhancement. The enhancement pattern of hypervascular hepatic metastases may overlap with those of benign liver lesions, including hemangioma, focal nodular hyperplasia (FNH) and hepatic adenomas. Dynamic imaging show both hemangiomas and metastases tend to fill in towards the center of the lesion with contrast material over time, but incomplete filling in is more characteristic of metastases. Hemangiomas characteristically present with a broken globular or nodular ring of enhancement. Globular enhancement has a sensitivity of 62 percent to 88 percent, and a specificity of 84 percent to 100 percent for the diagnosis of hemangioma [20, 21] (Fig. 8.4). In the portal venous phase hemangiomas tend to remain enhanced, whereas contrast material usually washes out of metastases [22]. Delayed images in some metastatic lesions may show a peripheral band of low density surrounding increased contrast attenuation in the center. This peripheral washout of contrast is due to compact
202
U. Udayasankar et al.
Fig. 8.3 Liver metastases from colon cancer. Transverse CT images obtained in 58-year-oldwoman with a known history of surgically resected adenocarcinoma of the colon. (a) Portal venous phase image shows multiple intrahepatic low-attenuation non-enhancing lesions (arrows) surrounded by normally enhancing liver parenchyma. Note the lesions are either subtle (arrow) or inconspicuous (arrowhead) in arterial phase image (b)
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
203
Fig. 8.4 Cavernous hemangioma of the liver. Contrast-enhanced CT image clearly shows a hypodense subcapsular lesion (arrowheads) with a peripheral enhancing nodule (arrow). This imaging appearance is characteristic of a hemangioma
packing of abundant tumor cells in the peripheral areas of certain metastatic lesions [23]. Both FNH and adenoma predominantly occur in younger individuals and have attenuation or signal intensity properties typically similar to that of surrounding liver tissue at CT or MRI, respectively, on all phases of hepatic imaging [24]. MRI is an excellent modality with which to both identify and characterize hepatic lesions. Liver metastases are often hypointense on T1-weighted images and hyperintense on T2-weighted images. Since most metastatic lesions contain abundant extracellular water, these lesions are often clearly visualized as well-marginated hyperintense lesions on T2-weighted sequences. Metastases with intratumoral hemorrhage, coagulative necrosis or mucin production may exhibit mixed signal intensity on T1-weighted images [14]. Hypovascular metastases may show slowly progressive irregular enhancement on post-gadolinium T1-weighted images, but do not show the characteristic interrupted nodular progression of contrast accumulation and slow washout characteristic of hemangiomas [25]. Additionally, hemangiomas are known to be predominantly hyperintense on T2-weighted scans (Fig. 8.5). Hypervascular metastases have variable T2 signal intensity, and on T1-weighted images are usually hypointense. A notable exception is metastases from malignant melanoma which shows elevated T1-weighted signal intensity due to melanin deposits. Arterial and portal phase contrast-enhanced MRI shows lesion enhancement characteristics similar to that seen with CT, including rapid enhancement and corresponding rapid washout. A relatively new technique of diffusion-weighted MRI has shown to be useful in characterizing focal liver lesions [26, 27]. However, significant overlaps have been shown in apparent diffusion coefficient (ADC) values of liver metastases, HCC, benign hepatocellular lesions and cavernous hemangiomas [27-29].
204
U. Udayasankar et al.
Fig. 8.5 Cavernous hemangioma of the liver. Transverse single shot fat suppressed T2-weighted MR image of upper abdomen shows a well-marginated subcapsular hyperintense lesion (“Lightbulb sign” – white arrow). Note that the lesion intensity is same as that of CSF in the spinal canal
18-F Fluro-deoxy-glucose (FDG) PET has proven to be highly sensitive in detecting hepatic metastases from different primaries, with highest sensitivity for hepatic metastases from colorectal, gastric and esophageal cancers [30]. Increased glucose metabolism in cancer cells leads to accumulation of FDG in tumor tissues, which are shown as areas of intense uptake on PET studies (Fig. 8.6). FDG is absorbed by tumor cells by facilitated glucose transporters (Glut). Glucose transporter 1 (Glut 1) was shown to be over-expressed in significant proportion of liver metastases, which facilitates easier detection on FDG-PET studies [31]. Many studies have demonstrated greater sensitivity of FDG-PET in diagnosing hepatic metastases when compared to MDCT, especially if CT findings are indeterminate [32-34]. Several studies have also shown the utility of FDG-PET in identifying additional metastatic lesions in cases where initial CT showed single hepatic metastases and, thus, changed the management strategy in these patients [35-37]. Patients considered for partial hepatectomy for colon cancer and evaluated with FDG-PET to assess metastatic tumor load showed a longer survival rate than those evaluated with classical cross-sectional imaging techniques [38]. FDG-PET has also proven to have higher accuracy than helical CT in detecting recurrent metastatic deposits. False negative PET for hepatic metastases may be due to a number of factors, including the inherent lower image resolution of PET, compared to other crosssectional imaging techniques (CT and MRI), underestimation of radiopharmaceutical uptake, inadequate attenuation correction, mislocalization of uptake foci into
Fig. 8.6 Hepatic metastases from endometrial cancer in a 54-year-old woman. (a) Transverse arterial phase CT image shows a large multi-cystic lesion in the right lobe of liver with irregular peripheral enhancement. (b) Whole-body PET image again shows the large mutilobulated lesion in the liver with a rim of increased FDG uptake (thin arrows). Focal increased uptake is also noted in a right mediastinal lymph node (thick arrow)
206
U. Udayasankar et al.
adjacent hepatic flexure or lung, physiological movement of liver during emission scan and recent completion of chemotherapy [39]. Examples of false positive findings include intrahepatic abscess, ruptured gallbladder empyema or inflammatory lesions such as regenerative nodules in cirrhotic liver [32, 33, 35].
2
Hepatocellular Carcinoma (HCC) and Lesions Associated with Chronic Liver Disease
HCC represents the most common primary hepatic malignancy worldwide. Cirrhosis is the most important risk factor for development of HCC. Cirrhosis associated with hepatitis B and C has a higher incidence of HCC. Early-stage HCC is typically clinically silent, and HCC is often advanced at first manifestation. Without treatment, the five-year survival rate is less than 5 percent [40]. The selected treatment depends on the presence of comorbidity, tumor size, location and morphology; and the presence of metastatic disease. A significant proportion of patients are beyond the scope of surgical resection at the time of diagnosis and most other therapeutic strategies are essentially palliative. Small unresectable HCCs that are fewer in number are managed with radio frequency ablation. More widespread disease is treated with percutaneous therapies such as chemoembolization and selective internal radiation therapy [41]. Detection of HCC in a cirrhotic liver is often considered a difficult task because of changes in both liver parenchymal (increased fibrosis, development of nodules and fatty replacement) and vascular characteristics (portal hypertension and development of arterial–portal venous shunts) [42, 43]. In a cirrhotic liver HCC usually develops in a multi-step fashion, from dysplastic nodules to dysplastic nodules with malignant foci, and to early well differentiated HCC [44] (Fig.8-7). Increased anaplasia in dysplastic nodules leads to hepatocarcinogenesis with resultant neoangiogenesis and capillarization. Longitudinal studies have documented the development of HCC within a dysplastic nodule in as early as four months from the date of detection [45]. With increasing anaplasia hepatic nodules tend to receive less blood supply from the portal vein and more from hepatic arteries [46]. In both cirrhotic and non-cirrhotic liver, HCC may also arise de novo from normal hepatocytes. Contrast-enhanced helical CT and MRI have been identified as accurate, noninvasive imaging techniques in the detection of HCC in both cirrhotic and noncirrhotic livers. Neovascularity is often used as an important marker for early identification of HCC. However, overlapping blood supply patterns and imaging findings make differentiation of various hepatic nodules a challenging imaging task. Non-enhanced CT studies are helpful in certain specific indications for demonstrating fatty replacement, hemorrhage within the tumor and presence of iodized oil after chemoembolization. Regenerating nodules are characteristically hyperdense on non-enhanced CT. In most centers MDCT evaluation of the liver for HCC is performed using a biphasic protocol. For contrast-enhanced CT most authors recommend a rapid iodinated contrast material injection rate (4 to 5 mL/sec) to maximize hypervascular tumor enhancement and, consequently, tumor conspicuity [47, 48].
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
207
Normal Hepatocytes
Dysplastic nodule with malignant foci Dysplastic nodule
Early differentiated HCC
Fig. 8.7 Multi-step process of development of hepatocellular carcinoma in cirrhotic liver
The later phase of hepatic arterial phase (HAP) of enhancement usually occurs 25 to 30 seconds after the initiation of contrast injection and is the period when a brightly enhancing lesion is best appreciated against a background of still relatively minimally enhanced liver parenchyma – the result of predominantly portal venous and as yet unenhanced supply of blood to the liver (75 percent of total blood flow). This is followed by the portal venous phase (PVP) which occurs approximately 60 seconds after the initiation of IV contrast when the majority of contrast is now in a venous phase of circulation and peak background hepatic parenchymal enhancement occurs. Though this second portal venous phase is less sensitive for tumor detection, it allows assessment of washout characteristics of HCC that facilitates lesion characterization/specificity and best demonstrates other important imaging information, such as vascular invasion not uncommonly seen in larger HCCs. Conventionally, MRI of the liver consists of T1-weighted and T2-weighted protocols, followed by a dynamic gadolinium-enhanced series. Commonly used MRI pulse sequences include T2-weighted fat-saturated fast spin echo (FSE), single shot FSE without fat-suppression, T-1 weighted opposed-phase and in-phase and 3-D GRE sequence with fat suppression before and after administration of intravenous gadolinium. At least three phases are acquired (arterial, portal and delayed) during dynamic post-gadolinium imaging [49]. Current MRI protocols for HCC evaluation images of liver are obtained during a breath-hold in most sequences and include T1-weighted fast spoiled gradient echo (SGE) and breath-hold half-Fourier transform single shot spin echo techniques [50-54]. MRI provides the best combination of sensitivity and characterization of the various levels of lesions of neoplastic degeneration found in the cirrhotic liver.
208
U. Udayasankar et al.
2.1
Imaging of Various Cirrhotic Nodules
2.1.1
Regenerative Nodule
Regenerative nodules (RN) usually receive blood supply from portal venous distribution [46] and are usually isodense to liver parenchyma on arterial and portal venous phase imaging of MDCT studies. Regenerative nodules are isointense on both T1- and T2-weighted MR images and have no enhancement on hepatic arterial phase images. RNs with thick septa are visualized as round nodules surrounded by enhancing fibrous septa on delayed-phase MRI. Siderotic RNs appear on unenhanced CT as predominantly high-attenuation nodules throughout the liver. On MRI studies siderotic nodules are characteristically hypointense on both T1- and T2-weighted gradient-echo images. 2.1.2
Dysplastic Nodule
Dysplastic nodules (DN) are usually larger than regenerative nodules and, again, usually derive their blood supply from portal venous branches. However, a few dysplastic nodules are supplied by branches of the hepatic artery [46]. Imaging features of dysplastic nodules vary depending on the degree of anaplasia. At dynamic contrast-enhanced CT, DNs are difficult to visualize because they are similar to the surrounding parenchyma in the arterial and portal phase images. On MRI signal intensity characteristics of DNs considerably overlap with those of early and small HCCs [55-58]. Commonly, DNs are demonstrated as homogeneously hyperintense on T1-weighted images, and hypointense on T2-weighted images. A characteristic and helpful feature which may help in differentiating HCC from DNs is that DNs are never hyperintense on T2-weighted images [59]. Assessing the level of enhancement is often difficult in dysplastic nodules since the nodules are mostly hyperintense on T1-weighted images. However, subtraction techniques may be useful in assessing the true level of enhancement, which may mirror enhancement of surrounding normal liver, but if supplied by predominantly hepatic arterial supply may be correspondingly bright on arterial phase imaging and isointense during portal venous phase. 2.1.3
Dysplastic Nodule with Malignant Foci
These nodules are also known as adenomatous hyperplasia with microscopic HCCs or macroregenerative nodules with microscopic HCCs. Dysplastic nodules with foci of HCC are commonly seen as isoattenuating nodules on dynamic contrastenhanced CT studies. However, MRI appearance of these lesions is classical and consists of a high signal intensity focus within a low intensity nodule on T2-weighted images. The central nodule of high signal intensity may also show enhancement during the HAP and represents the focus of HCC [60, 61].
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
2.1.4
209
Hepatocellular Carcinoma
Imaging features of HCC vary widely, based on the size of the lesions. HCC may present as a solitary mass lesion, multiple nodules or diffuse infiltrative disease. Large HCCs often show definitive diagnostic features of a large infiltrative tumor with vascular invasion and arterioportal shunting. Unfortunately most of these tumors are unresectable at the time of diagnosis. However, potentially curable, smaller HCCs can be detected at an early stage using imaging techniques. On unenhanced MDCT most HCCs have equal or lower attenuation than the adjacent liver parenchyma. HCCs are usually hypervascular and, therefore, show increased contrast enhancement during the arterial phase MDCT studies. Arterial phase imaging, thus, demonstrates the majority of HCCs [62-65]. There is rapid washout of contrast with the portal venous phase showing a remnant of peripheral capsular enhancement (Fig. 8.8). Portal venous phase imaging is usually less successful in identifying small HCCs since this phase provides lower sensitivity due to a combination of rapid lesion washout and significant enhancement of surrounding liver parenchyma. However, a few HCCs may be detected only on portal venous phase images or may be depicted more conspicuously during this phase than the arterial phase [66, 67]. Portal venous imaging is also useful for assessing portal venous complications, such as tumor thrombus.
Fig. 8.8 Hepatocelluar carcinoma. Transverse CT images in a 58-year old man. (a) Arterial phase image shows a relatively well-defined homogeneously enhancing mass in the right lobe of liver (arrowheads). A tiny branch of right hepatic artery is seen supplying the tumor (arrow). (b) Portal venous phase image shows early washout of contrast from the tumor. (c) An enhancing rim around the tumor (pseudo-capsule sign) (arrow) is noted in equilibrium phase image
210
U. Udayasankar et al.
Small HCCs are often hypovascular when compared to their larger counterparts [68]. A delayed phase imaging has been shown to be effective in accurately demonstrating these tumors, as these are often minimally enhancing on arterial phase studies, but have rapid contrast washout [48, 67]. As such, delayed phase imaging has been proposed as a useful component of MDCT protocols for detecting small HCCs in patients with cirrhosis [48]. Since some of the small well-differentiated HCCs receive blood supply from both hepatic and portal venous branches, another imaging strategy – a triple phase protocol, comprising an arterial phase, a portal venous inflow phase (essentially an early portal phase) and traditional portal venous phase – has also been attempted [43, 69] with maximum sensitivity seen in the middle inflow phase. As described above, this portal venous inflow phase has also incidentally shown to be better in demonstrating hypervascular metastases [70]. However, a recent study showed no added benefit of routine triple phase imaging in HCC detection in cirrhotic patients [71], so the value of such an extension of multiphase studies is still in question. The combination of hypointensity on T1-weighted images, hyperintensity on T2-weighted images and diffuse heterogeneous enhancement was the most common appearance of HCC on MR images in a multi-institutional patient population in North America [72]. Similar to MDCT, demonstration of arterial blood supply of most HCCs facilitate the diagnosis on MRI. Small HCCs measuring less than or equal to 1.5 cm were frequently isointense on both T1-weighted and T2-weighted images, and may be detected on immediate gadolinium-enhanced images only as diffuse homogeneously enhancing lesions. Dynamic imaging shows rapid washout of contrast extending to the portal venous phase [72-74], and delayed images demonstrate peripheral rim enhancement, also known as pseudo-capsule sign [25]. Larger HCCs show a heterogeneous pattern of enhancement in the arterial phase and often show enhancing capsule in delayed phases [72]. Large HCCs may have a number of other characteristic features including capsular bulging, extracapsular extension, formation of satellite nodules and vascular invasion. Diffuse or multifocal HCC may have imaging features of diffuse parenchymal liver disease and, at times, make it difficult to differentiate chronic liver disease from diffuse infiltrative malignancy. A potential source of confusion, particularly when dealing with the smaller HCC, is the cavernous hemangioma which does not show rapid washout and tends to remain isointense to the hepatic vasculature throughout multiple phases. Transient hepatic intensity difference (THID), a focal peripheral wedge-shaped area of increased signal intensity on hepatic arterial phase MRI, may mimic the heterogeneous enhancement pattern of large HCCs [75]. THIDs are almost always subcapsular in location, and are usually due to arterio-portal shunts or aberrant venous drainage [76]. The signal intensity characteristics of small HCCs considerably overlap those of dysplastic nodules. A pattern of increased signal intensity on T2-weighted images and diminished signal on T1-weighted images increases the likelihood of HCC, as compared to dysplastic nodules. High-grade dysplastic nodules may show similar signal intensity changes, but do not show enhancing capsules on delayed images. The tumor capsule is hypointense on both T1- and T2-weighted images in most cases. Increased signal intensity on T1-weighted images is shown to correlate with a more favorable histology [57]. Possible explanations for increased signal intensity
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
211
Fig. 9 HCC with portal vein thrombosis. (a) Transverse arterial phase contrast-enhanced CT image shows a large ill-marginated hypervascular lesion in the left lobe of liver (arrows). (b) Main portal vein and its left branch do not show any intraluminal contrast. Note the increase in caliber of left portal vein (arrow)
in HCC are fatty change, intratumoral copper and increased zinc content in surrounding liver parenchyma [77]. Imaging features which help to distinguish HCCs from dysplastic nodules [55] include: 1. 2. 3. 4. 5.
Moderate hyperintensity on T2-weighed images Lesions larger than 3 cm Enhancement during hepatic arterial phase Presence of a capsule Rapid interval growth
Portal vein thrombosis (PVT) is commonly observed in patients with long-term liver disease and cirrhosis. CT is a useful tool in identifying PVT, however it lacks accuracy in differentiating benign from malignant PVT [78]. Intrathrombus neovascularity, venous expansion (main portal vein greater than or equal to 23 mm) and direct invasion of the portal vein are suggested as independent imaging features diagnostic of malignant PVT (Fig. 8.9). PVT, contiguous with parenchymal HCC and generalized enhancement of a PVT, are two other CT signs strongly suggestive of malignant PVT. However, these findings are not considered absolute evidence of malignant PVT. Contrast enhancement of PVT on MRI studies helps distinguish malignant PVT from bland (non-enhancing) thrombosis in the cirrhotic liver. Subtraction MRI techniques help to increase the conspicuity of contrast enhancement, which also assists in determining the extent of tumor thrombi as distinguished from bland thrombus [79]. 2.1.5
Fibrolamellar HCC
Fibrolamellar HCC is an uncommon tumor that occurs primarily in young patients with no history of cirrhosis or chronic liver disease. The prognosis is better than classical HCC, with a higher rate of cure with resection and a prolonged survival rate. Serum tumor markers are usually absent.
212
U. Udayasankar et al.
Fig. 8.10 Focal nodular hyperplasia. (a) Transverse T2-weighted MR image of upper abdomen in a 38-year-old woman shows an ill-defined left lobe liver mass with hyperintense central scar (white arrow) (b) Immediate post gadolinium T1-weighted study shows uniform enhancement of the tumor with relative sparing of central scar (arrow). (c) Delayed enhancement of central scar is noted (arrow). Note washout of contrast from the rest of the lesion
Imaging studies demonstrate fibrolamellar HCC as typically large sharply defined lobulated masses (greater than 10 cm), often centered in the left lobe [24]. At CT, the tumor is often heterogeneous with calcification seen in 68 percent, usually within the central scar. This tumor is often heterogeneous with large areas of necrosis, hemorrhage and fibrous tissue. Two close differentials include FNH and hepatic adenoma, which often show homogenous enhancement on post-contrast studies. Hepatic adenomas often show evidence of spontaneous intratumoral hemorrhage and pockets of fat. Presence of calicification is rare in FNH. The small central scar in FNH is characteristically hyperintense on T2-weighted images, whereas the scar of fibrolamellar HCC is larger and often hypointense on T1- weighted and T2- weighted studies. Delayed phase enhancement of the scar is a frequent finding in FNH (Fig. 8.10).
2.1.6
Ultrasound and Liver Cancer Screening in the Cirrhotic Patient
The role of ultrasound (US) is important in the periodic surveillance of cirrhotic patients, and is a recommended imaging technique for screening patients at risk for HCC [80, 81]. Helical CT screening for HCC in patients with cirrhosis has shown a
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
213
substantial false positive detection rate. A few of the focal lesions with increased contrast enhancement were later proven not to be HCCs [82]. However, characterization of a nodule with US is also often limited because the gray scale appearance of HCC is variable and nonspecific. On gray scale US, HCC is usually observed as a discrete nodule with heterogeneous echo pattern, and may have a hypoechoic fibrous capsule [83]. As mentioned earlier, most of the imaging modalities rely on neovascularity and consequent contrast enhancement to detect primary malignant liver lesions. Even combined gray scale ultrasound and Doppler have a very low sensitivity and specificity for focal hepatic lesions (around 50 percent for individual lesions, but better for global staging), so that other imaging or biopsy usually is required after ultrasound scanning [84, 85]. The recent addition of microbubble contrast agents has expanded the role of US in liver lesion detection [86]. The most basic form of contrast-enhanced liver imaging employs color Doppler, and uses the signal enhancement produced by microbubbles to improve depiction of the vasculature of masses [87]. Recent contrast-enhanced US (CEUS) studies show a higher percentage of hypervascularity in HCCs when compared to CECT, and more intratumoral dysmorphic arteries were depicted on with CEUS [88]. Further studies with a large population of patients are needed to clarify the definite role of CEUS, but current results suggest great potential in the detection and characterization of hepatic tumors.
3 3.1
Biliary Lesions Cholangicarcinoma
In the United States cholangiocarcinoma is the second most common primary malignancy arising from the hepatobiliary system. However, the overall incidence is much less than HCC [2]. Histologically the tumors are adenocarcinomas arising from bile duct epithelium. Cholangiocarcinomas are broadly classified as intrahepatic and extrahepatic. Intrahepatic cholangiocarcinomas can be further classified as peripheral cholangiocarcinomas, which originate from an interlobular biliary duct, and hilar cholangiocarcinoma which originates from a main hepatic duct or the bifurcation of a common hepatic duct (known as Klatskin tumor). Hilar cholangiocarcinomas account for 40 percent to 60 percent of all cases [89]. The prognosis of cholangiocarcinoma is generally poor, but variable depending on location; peripheral cholangiocarcinomas show a four-fold better five-year survival rate when compared to hilar cholangiocarcinomas [90]. Intrahepatic cholangiocarcinomas are usually larger masses on imaging studies, whereas most extrahepatic cholangiocarcinomas tend to be small at the diagnosis, because the bile ducts are occluded during the early stage and patients present with jaundice at an earlier stage [91, 92]. Bile ducts proximal to the tumor are dilated, and the severity of this dilatation depends on the degree and duration of the obstruction. US is the initial imaging modality of choice for identifying hilar and distal extrahepatic cholangiocarcinomas. US clearly demonstrates the level of biliary obstruction, but the mass itself may be masked by adjacent structures or duodenal gas [93]. Sonographic
214
U. Udayasankar et al.
features are often unable to differentiate intrahepatic cholangiocarcinoma from other solid liver lesions. CT studies often demonstrate intrahepatic cholangiocarcinomas as irregular mass lesions that do not enhance either in arterial or portal venous phase images. Many tumors show uniform enhancement on delayed images due to their slow uptake and slow washout of intravenous contrast. Additionally, capsular retraction and segmental biliary duct dilatation and thickening are often noted [94]. Hilar and distal cholangiocarcinomas classically present as a hypodense thickening of the duct wall on portal venous phase CT, with delayed enhancement. Non-union of the right and left hepatic ducts with or without a visibly thickened wall is highly suggestive of a Klatskin tumor. Superficial spreading cholangiocarcinomas are hard to detect on routine imaging. When used in combination, features suggestive of unsuspected intrahepatic cholangiocarcinoma include segmental atrophy without obvious mass, diffuse fibrosis (seen as progressive late enhancement on contrastenhanced studies), irregular central narrowing of the dilated peripheral duct and the absence of cirrhosis [58]. Three distinct morphological variants of extrahepatic cholangiocarcinoma, as demonstrated on imaging studies, are mass-forming, periductal infiltrating and intraductal growing varieties [92]. Since both US and CT have low accuracy in determining the cause of distal common bile duct obstruction, invasive procedures such as endoscopic retrograde cholangiopancreatography (ERCP) and percutaneous transhepatic cholangiography (PTC) are widely used in jaundiced patients to confirm the site of obstruction and determine the nature of the disease to guide management steps. Widespread application of MRI, combined with the ability of magnetic resonance cholangiopancreatography (MRCP) to accurately map the biliary tree, has led to a drastic reduction in the number of invasive diagnostic techniques (Fig. 8.11). A complete preoperative lesion assessment and determination of its resectability is possible with new combined MRI protocols (MRI, MRCP and MR angiography) [95, 96]. In addition to identifying the primary lesion and the extent of obstruction, contrast-enhanced MRI allows accurate evaluation of periductal soft tissue infiltration, appreciation of possible vascular infiltration and study of metastatic involvement of the lymph nodes, liver and peritoneum [97]. Increased FDG uptake on PET studies is shown in intrahepatic cholangiocarcinomas, with the peripheral type showing more intense uptake when compared to the hilar variety. FDG-PET is also useful in detecting metastatic lymph nodes, although the yield with this technique is controversial [98, 99]. Currently, FDGPET is considered valuable in discovering unsuspected distant metastases in patients with peripheral cholangiocarcinoma because of the likelihood of distant metastases at the time of diagnosis, and the high FDG uptake in the peripheral variety and its metastatic deposits [99].
3.2
Imaging of Suspected Distant Malignant Biliary Obstruction
Many hepatobiliary and other malignant diseases lead to obstruction of the biliary tree. Imaging techniques aim to detect the presence, level and cause of biliary obstruction. US, CT and MRI have proven useful in the diagnosis and staging of pancreatic and
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
215
Fig. 8.11 Intrahepatic cholangiocarcinoma. (a) Transverse T2-weighted image obtained in a 74year old man with jaundice shows dilatation of the intrahepatic biliary ducts of the left lobe with abrupt cut off near the hilum (arrow). (b) Transverse delayed post-gadolinium image shows mildly enhancing irregular mass lesion at the hilum (arrow). (c) Hilar cholangiocarcinoma, better depicted on this coronal image was diagnosed on histopathological examination
hepatobiliary diseases, and obviate ERCP in many cases. US is often used as the initial non-invasive imaging modality in patients with suspected malignant biliary obstruction. Ultrasound has a higher sensitivity (71 percent) in delineating the level of obstruction than in defining the etiology (57 percent) [100]. Endoscopic ultrasound shows a higher accuracy in identifying the level of common bile obstruction with an added benefit of obtaining an immediate biopsy from the obstructing disease process 101. Intraductal ultrasound (IDUS) has been shown to have utility in detecting extrahepatic biliary stone disease as a benign mimic of an obstructive malignancy. When the cause of obstruction is malignant, IDUS has also proven to be useful in precisely identifying the degree of longitudinal spread of biliary cancer [102]. CT cholangiography is a novel technique of imaging the biliary tract after administration of intravenous biliary contrast media. Maximum intensity or volume-rendered projections enable visualization of the gallbladder and biliary tract lumen [102, 103]. Although this technique has proven useful in patients with biliary stone disease, its utility in malignant biliary obstruction is often limited due to reduced cholangiographic contrast media excretion in these patients [104].
216
U. Udayasankar et al.
MRCP is highly accurate in identifying diseases of the gallbladder, the biliary tree and the upper bile duct. Lower bile duct and periampullary lesions are better demonstrated on ERCP, with MRCP showing altered signal intensity changes near the duodenal wall as a result of interference from intraluminal gas [105, 106]. In a study comparing MRCP with ERCP for the diagnosis of malignant perihilar biliary obstructions, including hilar cholangiocarcinoma, MRCP and ERCP were very effective in detecting the presence of biliary obstructions, but MRCP was superior in its investigation of anatomic extent and the cause of the jaundice, when compared with ERCP [107]. MRCP was advantageous because it displayed the biliary tree cephalad to the lesion site and better characterized the intraductal filling defect. In addition, the level of obstruction in papillary-type cholangiocarcinoma or tumor thrombi dislodged to the common bile duct were more precisely estimated by MRCP than by ERCP [107].
4
Gallbladder Carcinoma
Gallbladder (GB) carcinoma is the most frequent tumor arising from the biliary tree worldwide. The most important risk factor for the development of GB carcinoma is the presence of GB stones, and the risk increases with increasing episodes of inflammatory reaction within the GB wall. An increased incidence of GB carcinoma is also noted in patients with porcelain GB, chronic salmonella infections and anomalous biliary/pancreatic duct system [108, 109]. GB carcinoma commonly manifests as an advanced inoperable tumor. Occasionally GB carcinoma presents as well-defined masses contained within the wall or, more frequently, as an histopathological finding after cholecystectomy for benign disease [110]. US is the initial imaging technique and it usually demonstrates an irregular heterogeneous mass in the gallbladder bed, which commonly infiltrates the adjacent hepatic parenchyma. In early stages ultrasound findings of irregular and asymmetric thickening of the GB wall may warrant further investigation [111]. Demonstrating gallstones trapped within the mass is a useful sign in differentiating GB carcinoma from a hepatic tumor [112, 113]. GB carcinoma may present occasionally as a mass or polyp extending to the GB lumen. US findings highly suggestive of malignancy include absence of displacement with patient movements, nodular or smooth shape and absence of posterior acoustic shadowing [114]. US is also useful in follow-up of GB polyps to assess the growth rate or changes in morphology. CT is widely used in GB carcinomas for identifying the primary neoplasm, assessing the margins and tumor staging. Contrast-enhanced CT commonly shows a heterogeneous mass lesion with irregular arterial phase enhancement [115] (Fig. 8.12). Earlier stages of the tumor may be shown as irregular wall thickening and polypoidal mass. The greatest benefits of CT lie in detecting potential local tumor extension to surrounding liver, bile duct and other adjacent structures, distant hepatic and peritoneal metastases and extent of lymphatic dissemination [112]. Survival of patients with GB carcinoma depends on the depth of primary tumor
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
217
Fig. 8.12 Gallbladder carcinoma. (a) Unenhanced transverse CT image shows evidence of thickening of GB wall (arrow). (b) However, irregularly enhancing thickened GB wall and nodular mass lesion (arrow) are better shown on the contrast-enhanced CT study
invasion and presence of metastatic lymph nodes [116, 117]. However, accurate preoperative staging is often difficult, especially in early stage cancers which are usually flat and, therefore, difficult to demonstrate on imaging studies [118]. Endoscopic ultrasound has a slightly higher sensitivity and specificity in detecting early stage GB carcinomas [119, 120]. MRI with MRCP is a useful modality in revealing bile duct involvement or liver invasion in advanced cases. However, it adds little information in distinguishing early stages of GB carcinoma [121]. FDGPET has been used to distinguish benign from malignant GB lesions, with a high sensitivity and specificity [122]. Since many patients with GB carcinoma are diagnosed incidentally after cholecystectomy, FDG-PET is commonly used for initial staging or re-staging when recurrence is suspected. Initial reports indicate a higher accuracy in the detection of local recurrent disease, compared to the ability to detect metastatic disease [123].
5 5.1
Image-Guided Interventions in Hepatobiliary Cancer Tumor Destruction
Surgery remains the mainstay of treatment in HCC. However, less than 20 percent of HCC can be treated surgically because of multifocal diseases, proximity of the tumor to key vascular or biliary structures precluding a margin-negative resection and inadequate functional hepatic reserve with cirrhosis. Percutaneous techniques alone, or in combination with surgery, are, therefore, considered the best treatment options for unresectable HCCs [124-126]. Tumor destruction is achieved by chemical (alcohol, acetic acid) or thermal (radiofrequency, microwave, laser and cryo-) ablation of the tumor substance. Percutaneous techniques have also proven beneficial in metastatic disease.
218
U. Udayasankar et al.
Percutaneous Ethanol Injection (PEI) is a widely recognized technique with few adverse effects (Fig. 8.13). PEI is a relatively inexpensive, simple technique usually performed under ultrasound guidance. PEI can attain responses of 90 percent to 100 percent in HCC smaller than 2 cm, to 70 percent in those of 3 cm and 50 percent in HCC of 5 cm in diameter [124, 127]. Disadvantages of this technique include the need for multiple sessions because of incomplete spread of ethanol within the entire tumor due to intratumoral fibrous septae, and incomplete responses in larger tumors. Using high frequency alternating current, radiofrequency ablation (RFA) raises the intratumoral temperature, leading to protein denaturation and coagulation necrosis [128]. This technique can be performed percutaneously, laparoscopically or during laparotomy using single or multiple probes. RF thermal ablation was shown to be superior to PEI in achieving complete tumor response and requiring fewer treatment sessions [129]. RFA is usually preferred over PEI in large unresectable tumors (larger than 3 cm). RFA is associated with lower mortality (0.05 percent) and fewer complications (8 percent), compared to surgical techniques [130].
Fig. 8.13 Percutaneous alcohol ablation of HCC. (a) Transverse arterial phase contrast-enhanced CT image obtained in an 80-year-old man shows peripherally enhancing intrahepatic mass lesion (arrows) (biopsy-proven HCC). (b) Percutaneous alcohol ablation was performed through an anterior approach. Note the position of needle in the tumor (arrow). (c) Follow-up contrastenhanced CT performed after a month post the procedure shows hypodense region with foci of air (arrows). There is interval reduction in tumor size with no significant abnormal enhancement to suggest residual or recurrent tumor
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
219
Needle track seeding is a recognized complication of RFA and is often seen in tumors with poorly differentiated morphology or subcapsular location. Rarely, RFA fails to achieve sufficiently high intratumoral temperature due to the cooling effect of flowing blood in a large adjacent blood vessel. Other percutaneous techniques including acetic acid injection, microwave coagulation therapy, laser-induced interstitial thermotherapy, cryoablation and high intensity-focused ablation are associated with increased risk of complications, often lack added benefits over RFA or PEI or are still experimental. CT, CEU and MRI are currently utilized to follow-up percutaneously treated liver lesions [131]. Inadequately treated lesions are identified as residual foci of enhancement on contrast-enhanced imaging techniques, whereas the absence of tumor perfusion is highly indicative of coagulated tissue. An immediate postprocedural increase in size and absence of enhancement on contrast-enhanced follow-up scans were significantly associated with successful treatment of tumors. A nodular or wedge-shaped enhancement, however, indicated unsuccessful treatment or recurrent tumor [132, 133]. A rim of reactive peripheral enhancement may be noted in some cases, and may affect the accuracy of immediate post-treatment scans to detect viable residual neoplastic tissue. Transarterial chemo-embolization and radiation therapy are increasingly used in treatment of unresectable primary and secondary liver malignancies. Several studies have demonstrated favorable outcomes in patients treated with arterially delivered cytotoxic agents and 90Yttrium incorporated into glass microspheres [134-136]. These techniques result in selective delivery of high doses of drug or radiation to the tumor through the hepatic arterial system with relative sparing of normal liver, which is predominantly supplied by the portal venous system. In many centers Transcatheter arterial lipiodol chemoembolization (TACE) is now the treatment of choice for non-curative HCC. Recent reports show that the overall survival rate of patients who underwent TACE as an initial treatment were 47 percent at three years, 26 percent at five years and 16 percent at seven years [137]. Transarterial 90Yttrium treatment is a more recent technique which is well tolerated, does not induce significant systemic toxicity and can be administered on an outpatient basis. The efficacy of transarterial techniques is frequently evaluated by dynamic contrast-enhanced CT. Conventionally, a decrease in tumor size and areas of necrosis are considered CT signs of tumor response. The iodized oil (lipiodol) used in TACE may be retained within the tumor and affect delineation of a residual lesion. Most authors consider deposition of iodized oil within the lesion as a sign of necrosis [138]. Dynamic MRI has been advocated as an alternative in these patients. TACE is administered with caution in patients with PVT, since occlusion of the hepatic artery may lead to hepatic failure.
5.2
Biliary Interventions
Percutaneous biliary opacification is typically perfomed when less invasive diagnostic tools fail to reach a conclusive diagnosis. It allows direct access to the biliary
220
U. Udayasankar et al.
tree and can be converted into a therapeutic intervention. Current percutaneous biliary interventions include (a) percutaneous transhepatic cholangiography (PTC) and biliary drainage to manage benign and malignant obstruction, (b) percutaneous cholecystostomy, percutaneous treatment of biliary stone disease with or without choledochoscopy, (c) cholangioplasty for biliary strictures, (d) biopsy of the biliary duct, (e) management of complications from laparoscopic cholecystectomy and liver transplantation and (f) treatment of malignant and benign strictures with plastic or metallic stent placement. PTC is more invasive and painful than ERCP, mainly because the PTC procedure involves puncturing the liver capsule. It also poses the risks of hemoperitoneum and bile peritonitis. PTC is now usually reserved for patients who had had unsuccessful ERCP when the biliary system cannot be cannulated, or when the obstructing lesion prevents contrast material from opacifying the cephalic portions of the biliary system (Fig. 8.14A). PTC, however, is the preferred procedure in patients with a history of anatomy-altering surgeries. These include the Billroth II procedure, which involves antrectomy with gastrojejunostomy. This procedure makes using an endoscope to cannulate the ampulla difficult because the endoscope must be passed through the gastrojejunostomy anastomosis and then retrograde towards the duodenum. The failure rate for ERCP in this situation is high, as is the complication rate. Other indications for PTC include the management of postoperative or posttraumatic bile leakage. In many cases PTC is followed by the placement of percutaneous biliary drainage (PBD); this is needed in patients with biliary obstruction of any cause. Relieving obstruction can be a life-saving procedure in these patients, especially in those patients where obstruction is due to unresectable malignant tumors (Fig. 8.14 B,C), as well as in patients with various types of
Fig. 8.14 Percutaneous puncture in the right biliary duct in a patient with pancreatic cancer and contrast injection reveals. (a) complete obstruction of the common bile duct (arrow). (b) Subsequent placement of an external drainage catheter (arrow). Two weeks later, an internal/external biliary drainage catheter (arrows) is placed to allow internal drainage of bile (c)
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
221
Fig. 8.14 (continued)
Fig. 8.15 Ultrasound-guided cholecystotomy. (a) Note the tip of the puncture needle in the lumen of the gallbladder (white arrow). This patient has a Klatskin tumor with severe biliary obstruction and sepsis. (b) Placement of a pigtail catheter in the lumen of the gall bladder for adequate drainage (arrow). Note the left and right biliary drainage catheters (arrowheads)
benign strictures (including postoperative strictures), primary sclerosing cholangitis and liver transplants. Other indications include diversion for bile leaks while the patient is awaiting surgery, and transhepatic brachytherapy for cholangiocarcinoma. Percutaneous cholecystostomy (Fig. 8.15) is used as a temporary measure in critically ill patients with acute cholecystitis who cannot undergo cholecystectomy. After the symptoms resolve and the patient’s condition is stabilized, definitive
222
U. Udayasankar et al.
treatment is still gallbladder removal. In acalculous cholecystitis, percutaneous drainage may be the only treatment required. The procedure is performed under ultrasonographic and fluoroscopic guidance.
Key Points ●
●
●
●
Imaging techniques greatly assist in the early diagnosis, treatment and follow-up of patients with hepatobiliary tumors. US and CT are the main modalities used for staging cancer patients for hepatic metastatic tumor spread, and for assessing patients with known chronic liver disease. Contrast-enhanced CT, MRI, PET and PET/CT are used to estimate the extent of malignancy, vascular or biliary invasion and residual normal liver volume. Imaging also permits diagnosis and treatment by guided procedures such as percutaneous biopsy, RFA and TACE.
References 1. El-Serag HB, Mason AC. Rising incidence of hepatocellular carcinoma in the United States. N Engl J Med 1999;340(10):745-50. 2. Jemal A, Murray T, Ward E, et al. Cancer statistics, 2005. CA Cancer J Clin 2005;55(1):10-30. 3. Bruix J, Hessheimer AJ, Forner A, Boix L, Vilana R, Llovet JM. New aspects of diagnosis and therapy of hepatocellular carcinoma. Oncogene 2006;25(27):3848-56. 4. Korner M, Hayes GM, Rehmann R, et al. Secretin receptors in the human liver: expression in biliary tract and cholangiocarcinoma, but not in hepatocytes or hepatocellular carcinoma. Journal of hepatology 2006;45(6):825-35. 5. Libbrecht L. Hepatic progenitor cells in human liver tumor development. World J Gastroenterol 2006;12(39):6261-5. 6. Llovet JM, Chen Y, Wurmbach E, et al. A molecular signature to discriminate dysplastic nodules from early hepatocellular carcinoma in HCV cirrhosis. Gastroenterology 2006;131(6):1758-67. 7. Ramos E, Llado L, Serrano T, et al. Utility of cell-cycle modulators to predict vascular invasion and recurrence after surgical treatment of hepatocellular carcinoma. Transplantation 2006;82(6):753-8. 8. Haaga JR, Lanzieri CF, Gilkeson RC. CT and MRI of the whole body. 4th ed. St. Louis, Mo.: Mosby; 2003. 9. Bluemke DA, Soyer P, Fishman EK. Helical (spiral) CT of the liver. Radiologic clinics of North America 1995;33(5):863-86. 10. Scatarige JC, Fishman EK, Saksouk FA, Siegelman SS. Computed tomography of calcified liver masses. Journal of computer assisted tomography 1983;7(1):83-9. 11. Hale HL, Husband JE, Gossios K, Norman AR, Cunningham D. CT of calcified liver metastases in colorectal carcinoma. Clin Radiol 1998;53(10):735-41. 12. Chen MY, Bechtold RE, Savage PD. Cystic changes in hepatic metastases from gastrointestinal stromal tumors (GISTs) treated with Gleevec (imatinib mesylate). Ajr 2002;179(4):1059-62.
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
223
13. Soyer P, Poccard M, Boudiaf M, et al. Detection of hypovascular hepatic metastases at triplephase helical CT: sensitivity of phases and comparison with surgical and histopathologic findings. Radiology 2004;231(2):413-20. 14. Kanematsu M, Kondo H, Goshima S, et al. Imaging liver metastases: review and update. Eur J Radiol 2006;58(2):217-28. 15. Blake SP, Weisinger K, Atkins MB, Raptopoulos V. Liver metastases from melanoma: detection with multiphasic contrast-enhanced CT. Radiology 1999;213(1):92-6. 16. Bressler EL, Alpern MB, Glazer GM, Francis IR, Ensminger WD. Hypervascular hepatic metastases: CT evaluation. Radiology 1987;162(1 Pt 1):49-51. 17. Sica GT, Ji H, Ros PR. CT and MR imaging of hepatic metastases. Ajr 2000;174(3):691-8. 18. Oliver JH, 3rd, Baron RL, Federle MP, Jones BC, Sheng R. Hypervascular liver metastases: do unenhanced and hepatic arterial phase CT images affect tumor detection? Radiology 1997;205(3):709-15. 19. Semelka RC, Hussain SM, Marcos HB, Woosley JT. Perilesional enhancement of hepatic metastases: correlation between MR imaging and histopathologic findings-initial observations. Radiology 2000;215(1):89-94. 20. Leslie DF, Johnson CD, MacCarty RL, Ward EM, Ilstrup DM, Harmsen WS. Single-pass CT of hepatic tumors: value of globular enhancement in distinguishing hemangiomas from hypervascular metastases. Ajr 1995;165(6):1403-6. 21. Leslie DF, Johnson CD, Johnson CM, Ilstrup DM, Harmsen WS. Distinction between cavernous hemangiomas of the liver and hepatic metastases on CT: value of contrast enhancement patterns. Ajr 1995;164(3):625-9. 22. Hanafusa K, Ohashi I, Himeno Y, Suzuki S, Shibuya H. Hepatic hemangioma: findings with two-phase CT. Radiology 1995;196(2):465-9. 23. Muramatsu Y, Takayasu K, Moriyama N, et al. Peripheral low-density area of hepatic tumors: CT-pathologic correlation. Radiology 1986;160(1):49-52. 24. Ichikawa T, Federle MP, Grazioli L, Madariaga J, Nalesnik M, Marsh W. Fibrolamellar hepatocellular carcinoma: imaging and pathologic findings in 31 recent cases. Radiology 1999;213(2):352-61. 25. Martin DR, Danrad R, Hussain SM. MR imaging of the liver. Radiologic clinics of North America 2005;43(5):861-86, viii. 26. Ichikawa T, Haradome H, Hachiya J, Nitatori T, Araki T. Diffusion-weighted MR imaging with a single-shot echoplanar sequence: detection and characterization of focal hepatic lesions. Ajr 1998;170(2):397-402. 27. Kim T, Murakami T, Takahashi S, Hori M, Tsuda K, Nakamura H. Diffusion-weighted singleshot echoplanar MR imaging for liver disease. Ajr 1999;173(2):393-8. 28. Namimoto T, Yamashita Y, Sumi S, Tang Y, Takahashi M. Focal liver masses: characterization with diffusion-weighted echo-planar MR imaging. Radiology 1997;204(3):739-44. 29. Taouli B, Vilgrain V, Dumont E, Daire JL, Fan B, Menu Y. Evaluation of liver diffusion isotropy and characterization of focal hepatic lesions with two single-shot echo-planar MR imaging sequences: prospective study in 66 patients. Radiology 2003;226(1):71-8. 30. Kinkel K, Lu Y, Both M, Warren RS, Thoeni RF. Detection of hepatic metastases from cancers of the gastrointestinal tract by using noninvasive imaging methods (US, CT, MR imaging, PET): a meta-analysis. Radiology 2002;224(3):748-56. 31. Zimmerman RL, Burke M, Young NA, Solomides CC, Bibbo M. Diagnostic utility of Glut-1 and CA 15-3 in discriminating adenocarcinoma from hepatocellular carcinoma in liver tumors biopsied by fine-needle aspiration. Cancer 2002;96(1):53-7. 32. Bohm B, Voth M, Geoghegan J, et al. Impact of positron emission tomography on strategy in liver resection for primary and secondary liver tumors. Journal of cancer research and clinical oncology 2004;130(5):266-72. 33. Delbeke D, Martin WH, Sandler MP, Chapman WC, Wright JK, Jr., Pinson CW. Evaluation of benign vs malignant hepatic lesions with positron emission tomography. Arch Surg 1998;133(5):510-5; discussion 5-6.
224
U. Udayasankar et al.
34. Vitola JV, Delbeke D, Sandler MP, et al. Positron emission tomography to stage suspected metastatic colorectal carcinoma to the liver. American journal of surgery 1996;171(1):21-6. 35. Arulampalam TH, Francis DL, Visvikis D, Taylor I, Ell PJ. FDG-PET for the pre-operative evaluation of colorectal liver metastases. Eur J Surg Oncol 2004;30(3):286-91. 36. Schiepers C, Penninckx F, De Vadder N, et al. Contribution of PET in the diagnosis of recurrent colorectal cancer: comparison with conventional imaging. Eur J Surg Oncol 1995;21(5):517-22. 37. Ogunbiyi OA, Flanagan FL, Dehdashti F, et al. Detection of recurrent and metastatic colorectal cancer: comparison of positron emission tomography and computed tomography. Ann Surg Oncol 1997;4(8):613-20. 38. Fernandez FG, Drebin JA, Linehan DC, Dehdashti F, Siegel BA, Strasberg SM. Five-year survival after resection of hepatic metastases from colorectal cancer in patients screened by positron emission tomography with F-18 fluorodeoxyglucose (FDG-PET). Annals of surgery 2004;240(3):438-47; discussion 47-50. 39. Khandani AH, Wahl RL. Applications of PET in liver imaging. Radiologic clinics of North America 2005;43(5):849-60, vii. 40. Llovet JM, Bustamante J, Castells A, et al. Natural history of untreated nonsurgical hepatocellular carcinoma: rationale for the design and evaluation of therapeutic trials. Hepatology 1999;29(1):62-7. 41. Clark HP, Carson WF, Kavanagh PV, Ho CP, Shen P, Zagoria RJ. Staging and current treatment of hepatocellular carcinoma. Radiographics 2005;25 Suppl 1:S3-23. 42. Miller WJ, Baron RL, Dodd GD, 3rd, Federle MP. Malignancies in patients with cirrhosis: CT sensitivity and specificity in 200 consecutive transplant patients. Radiology 1994;193(3):645-50. 43. Oliver JH, 3rd, Baron RL, Federle MP, Rockette HE, Jr. Detecting hepatocellular carcinoma: value of unenhanced or arterial phase CT imaging or both used in conjunction with conventional portal venous phase contrast-enhanced CT imaging. Ajr 1996;167(1):71-7. 44. Jeong YY, Yim NY, Kang HK. Hepatocellular carcinoma in the cirrhotic liver with helical CT and MRI: imaging spectrum and pitfalls of cirrhosis-related nodules. Ajr 2005;185(4): 1024-32. 45. Takayama T, Makuuchi M, Hirohashi S, et al. Malignant transformation of adenomatous hyperplasia to hepatocellular carcinoma. Lancet 1990;336(8724):1150-3. 46. Hayashi M, Matsui O, Ueda K, et al. Correlation between the blood supply and grade of malignancy of hepatocellular nodules associated with liver cirrhosis: evaluation by CT during intraarterial injection of contrast medium. Ajr 1999;172(4):969-76. 47. Kamel IR, Liapi E, Fishman EK. Multidetector CT of hepatocellular carcinoma. Best practice & research 2005;19(1):63-89. 48. Monzawa S, Ichikawa T, Nakajima H, Kitanaka Y, Omata K, Araki T. Dynamic CT for detecting small hepatocellular carcinoma: usefulness of delayed phase imaging. Ajr 2007;188(1):147-53. 49. Ito K. Hepatocellular carcinoma: conventional MRI findings including gadolinium-enhanced dynamic imaging. Eur J Radiol 2006;58(2):186-99. 50. Naganawa S, Jenner G, Cooper TG, Potchen EJ, Ishigaki T. Rapid MR imaging of the liver: comparison of twelve techniques for single breath-hold whole volume acquisition. Radiation medicine 1994;12(6):255-61. 51. Semelka RC, Kelekis NL, Thomasson D, Brown MA, Laub GA. HASTE MR imaging: description of technique and preliminary results in the abdomen. J Magn Reson Imaging 1996;6(4):698-9. 52. Coates GG, Borrello JA, McFarland EG, Mirowitz SA, Brown JJ. Hepatic T2-weighted MRI: a prospective comparison of sequences, including breath-hold, half-Fourier turbo spin echo (HASTE). J Magn Reson Imaging 1998;8(3):642-9. 53. Bradley WG, Jr. Optimizing lesion contrast without using contrast agents. J Magn Reson Imaging 1999;10(3):442-9. 54. Helmberger TK, Schroder J, Holzknecht N, et al. T2-weighted breathold imaging of the liver: a quantitative and qualitative comparison of fast spin echo and half Fourier single shot fast spin echo imaging. Magma (New York, NY 1999;9(1-2):42-51.
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
225
55. Krinsky GA, Lee VS. MR imaging of cirrhotic nodules. Abdom Imaging 2000;25(5):471-82. 56. Mitchell DG. Focal manifestations of diffuse liver disease at MR imaging. Radiology 1992;185(1):1-11. 57. Ebara M, Fukuda H, Kojima Y, et al. Small hepatocellular carcinoma: relationship of signal intensity to histopathologic findings and metal content of the tumor and surrounding hepatic parenchyma. Radiology 1999;210(1):81-8. 58. Martin DR, Brown MA, Semelka RC. Primer on MR imaging of the abdomen and pelvis. Hoboken, N.J.: John Wiley & Sons, Inc.; 2005. 59. Matsui O, Kadoya M, Kameyama T, et al. Adenomatous hyperplastic nodules in the cirrhotic liver: differentiation from hepatocellular carcinoma with MR imaging. Radiology 1989;173(1):123-6. 60. Sadek AG, Mitchell DG, Siegelman ES, Outwater EK, Matteucci T, Hann HW. Early hepatocellular carcinoma that develops within macroregenerative nodules: growth rate depicted at serial MR imaging. Radiology 1995;195(3):753-6. 61. Mitchell DG, Rubin R, Siegelman ES, Burk DL, Jr., Rifkin MD. Hepatocellular carcinoma within siderotic regenerative nodules: appearance as a nodule within a nodule on MR images. Radiology 1991;178(1):101-3. 62. Hollett MD, Jeffrey RB, Jr., Nino-Murcia M, Jorgensen MJ, Harris DP. Dual-phase helical CT of the liver: value of arterial phase scans in the detection of small (< or = 1.5 cm) malignant hepatic neoplasms. Ajr 1995;164(4):879-84. 63. Baron RL, Oliver JH, 3rd, Dodd GD, 3rd, Nalesnik M, Holbert BL, Carr B. Hepatocellular carcinoma: evaluation with biphasic, contrast-enhanced, helical CT. Radiology 1996;199(2):505-11. 64. Oliver JH, 3rd, Baron RL. Helical biphasic contrast-enhanced CT of the liver: technique, indications, interpretation, and pitfalls. Radiology 1996;201(1):1-14. 65. Oto A, Tamm EP, Szklaruk J. Multidetector row CT of the liver. Radiologic clinics of North America 2005;43(5):827-48, vii. 66. Kim T, Murakami T, Takahashi S, et al. Optimal phases of dynamic CT for detecting hepatocellular carcinoma: evaluation of unenhanced and triple-phase images. Abdom Imaging 1999;24(5):473-80. 67. Lim JH, Choi D, Kim SH, et al. Detection of hepatocellular carcinoma: value of adding delayed phase imaging to dual-phase helical CT. Ajr 2002;179(1):67-73. 68. Takayasu K, Furukawa H, Wakao F, et al. CT diagnosis of early hepatocellular carcinoma: sensitivity, findings, and CT-pathologic correlation. Ajr 1995;164(4):885-90. 69. Hu H, He HD, Foley WD, Fox SH. Four multidetector-row helical CT: image quality and volume coverage speed. Radiology 2000;215(1):55-62. 70. Murakami T, Kim T, Takamura M, et al. Hypervascular hepatocellular carcinoma: detection with double arterial phase multi-detector row helical CT. Radiology 2001;218(3):763-7. 71. Ichikawa T, Kitamura T, Nakajima H, et al. Hypervascular hepatocellular carcinoma: can double arterial phase imaging with multidetector CT improve tumor depiction in the cirrhotic liver? Ajr 2002;179(3):751-8. 72. Kelekis NL, Semelka RC, Worawattanakul S, et al. Hepatocellular carcinoma in North America: a multiinstitutional study of appearance on T1-weighted, T2-weighted, and serial gadolinium-enhanced gradient-echo images. Ajr 1998;170(4):1005-13. 73. Peterson MS, Baron RL, Murakami T. Hepatic malignancies: usefulness of acquisition of multiple arterial and portal venous phase images at dynamic gadolinium-enhanced MR imaging. Radiology 1996;201(2):337-45. 74. Fujita T, Ito K, Honjo K, Okazaki H, Matsumoto T, Matsunaga N. Detection of hepatocellular carcinoma: comparison of T2-weighted breath-hold fast spin-echo sequences and highresolution dynamic MR imaging with a phased-array body coil. J Magn Reson Imaging 1999;9(2):274-9. 75. Itai Y, Matsui O. Blood flow and liver imaging. Radiology 1997;202(2):306-14. 76. Matsui O, Kadoya M, Yoshikawa J, et al. Aberrant gastric venous drainage in cirrhotic livers: imaging findings in focal areas of liver parenchyma. Radiology 1995;197(2):345-9.
226
U. Udayasankar et al.
77. Kelekis NL, Semelka RC, Woosley JT. Malignant lesions of the liver with high signal intensity on T1-weighted MR images. J Magn Reson Imaging 1996;6(2):291-4. 78. Tublin ME, Dodd GD, 3rd, Baron RL. Benign and malignant portal vein thrombosis: differentiation by CT characteristics. Ajr 1997;168(3):719-23. 79. Yu JS, Rofsky NM. Dynamic subtraction MR imaging of the liver: advantages and pitfalls. Ajr 2003;180(5):1351-7. 80. Kim TK, Jang HJ, Wilson SR. Imaging diagnosis of hepatocellular carcinoma with differentiation from other pathology. Clinics in liver disease 2005;9(2):253-79. 81. Nguyen MH, Keeffe EB. Screening for hepatocellular carcinoma. J Clin Gastroenterol 2002;35(5 Suppl 2):S86-91. 82. Brancatelli G, Baron RL, Peterson MS, Marsh W. Helical CT screening for hepatocellular carcinoma in patients with cirrhosis: frequency and causes of false-positive interpretation. Ajr 2003;180(4):1007-14. 83. Choi BI, Kim CW, Han MC, et al. Sonographic characteristics of small hepatocellular carcinoma. Gastrointestinal radiology 1989;14(3):255-61. 84. Wernecke K, Rummeny E, Bongartz G, et al. Detection of hepatic masses in patients with carcinoma: comparative sensitivities of sonography, CT, and MR imaging. Ajr 1991;157(4): 731-9. 85. Cosgrove D, Blomley M. Liver tumors: evaluation with contrast-enhanced ultrasound. Abdom Imaging 2004;29(4):446-54. 86. Choi BI, Kim AY, Lee JY, et al. Hepatocellular carcinoma: contrast enhancement with Levovist. J Ultrasound Med 2002;21(1):77-84. 87. Jang HJ, Kim TK, Wilson SR. Imaging of malignant liver masses: characterization and detection. Ultrasound quarterly 2006;22(1):19-29. 88. Giorgio A, Ferraioli G, Tarantino L, et al. Contrast-enhanced sonographic appearance of hepatocellular carcinoma in patients with cirrhosis: comparison with contrast-enhanced helical CT appearance. Ajr 2004;183(5):1319-26. 89. Slattery JM, Sahani DV. What is the current state-of-the-art imaging for detection and staging of cholangiocarcinoma? The oncologist 2006;11(8):913-22. 90. Madariaga JR, Iwatsuki S, Todo S, Lee RG, Irish W, Starzl TE. Liver resection for hilar and peripheral cholangiocarcinomas: a study of 62 cases. Annals of surgery 1998;227(1): 70-9. 91. Lim JH. Cholangiocarcinoma: morphologic classification according to growth pattern and imaging findings. Ajr 2003;181(3):819-27. 92. Lim JH, Lee WJ, Takehara Y, Lim HK. Imaging of extrahepatic cholangiocarcinoma. Abdom Imaging 2004;29(5):565-71. 93. Robledo R, Muro A, Prieto ML. Extrahepatic bile duct carcinoma: US characteristics and accuracy in demonstration of tumors. Radiology 1996;198(3):869-73. 94. Kim TK, Choi BI, Han JK, Jang HJ, Cho SG, Han MC. Peripheral cholangiocarcinoma of the liver: two-phase spiral CT findings. Radiology 1997;204(2):539-43. 95. Magnuson TH, Bender JS, Duncan MD, Ahrendt SA, Harmon JW, Regan F. Utility of magnetic resonance cholangiography in the evaluation of biliary obstruction. Journal of the American College of Surgeons 1999;189(1):63-71; discussion -2. 96. Zidi SH, Prat F, Le Guen O, Rondeau Y, Pelletier G. Performance characteristics of magnetic resonance cholangiography in the staging of malignant hilar strictures. Gut 2000;46(1): 103-6. 97. Guarise A, Venturini S, Faccioli N, Pinali L, Morana G. Role of magnetic resonance in characterising extrahepatic cholangiocarcinomas. La Radiologia medica 2006;111(4): 526-38. 98. Kluge R, Schmidt F, Caca K, et al. Positron emission tomography with [(18)F]fluoro-2-deoxyD-glucose for diagnosis and staging of bile duct cancer. Hepatology 2001;33(5): 1029-35. 99. Kim YJ, Yun M, Lee WJ, Kim KS, Lee JD. Usefulness of 18F-FDG-PET in intrahepatic cholangiocarcinoma. European journal of nuclear medicine and molecular imaging 2003;30(11):1467-72.
8 Diagnostic Imaging and Image-Guided Interventions of Hepatobiliary Malignancies
227
100. Blackbourne LH, Earnhardt RC, Sistrom CL, Abbitt P, Jones RS. The sensitivity and role of ultrasound in the evaluation of biliary obstruction. The American surgeon 1994;60(9):683-90. 101. Songur Y, Temucin G, Sahin B. Endoscopic ultrasonography in the evaluation of dilated common bile duct. J Clin Gastroenterol 2001;33(4):302-5. 102. Tamada K, Yasuda Y, Nagai H, et al. Limitations of three-dimensional intraductal ultrasonography in the assessment of longitudinal spread of extrahepatic bile duct carcinoma. Journal of gastroenterology 2000;35(12):919-23. 103. Yeh BM, Breiman RS, Taouli B, Qayyum A, Roberts JP, Coakley FV. Biliary tract depiction in living potential liver donors: comparison of conventional MR, mangafodipir trisodiumenhanced excretory MR, and multi-detector row CT cholangiography–initial experience. Radiology 2004;230(3):645-51. 104. Stroszczynski C, Hunerbein M. Malignant biliary obstruction: value of imaging findings. Abdom Imaging 2005;30(3):314-23. 105. David V, Reinhold C, Hochman M, et al. Pitfalls in the interpretation of MR cholangiopancreatography. Ajr 1998;170(4):1055-9. 106. Kondo H, Kanematsu M, Shiratori Y, Moriwaki H, Hoshi H. Potential pitfall of MR cholangiopancreatography: right hepatic arterial impression of the common hepatic duct. Journal of computer assisted tomography 1999;23(1):60-2. 107. Yeh TS, Jan YY, Tseng JH, et al. Malignant perihilar biliary obstruction: magnetic resonance cholangiopancreatographic findings. Am J Gastroenterol 2000;95(2):432-40. 108. Rodriguez-Fernandez A, Gomez-Rio M, Medina-Benitez A, et al. Application of modern imaging methods in diagnosis of gallbladder cancer. Journal of surgical oncology 2006;93(8):650-64. 109. Sherlock S, Dooley J. Diseases of the liver and biliary system. 11th ed. Oxford: Blackwell Science; 2002. 110. Matsusaka S, Yamasaki H, Kitayama Y, Okada T, Maeda S. Occult gallbladder carcinoma diagnosed by a laparoscopic cholecystectomy. Surgery today 2003;33(10):740-2. 111. Soiva M, Aro K, Pamilo M, Paivansalo M, Suramo I, Taavitsainen M. Ultrasonography in carcinoma of the gallbladder. Acta Radiol 1987;28(6):711-4. 112. Levy AD, Murakata LA, Rohrmann CA, Jr. Gallbladder carcinoma: radiologic-pathologic correlation. Radiographics 2001;21(2):295-314; questionnaire, 549-55. 113. Gandolfi L, Torresan F, Solmi L, Puccetti A. The role of ultrasound in biliary and pancreatic diseases. Eur J Ultrasound 2003;16(3):141-59. 114. Sugiyama M, Atomi Y, Yamato T. Endoscopic ultrasonography for differential diagnosis of polypoid gall bladder lesions: analysis in surgical and follow up series. Gut 2000;46(2):250-4. 115. Kumaran V, Gulati S, Paul B, Pande K, Sahni P, Chattopadhyay K. The role of dual-phase helical CT in assessing resectability of carcinoma of the gallbladder. European radiology 2002;12(8):1993-9. 116. Bartlett DL, Fong Y, Fortner JG, Brennan MF, Blumgart LH. Long-term results after resection for gallbladder cancer. Implications for staging and management. Annals of surgery 1996;224(5):639-46. 117. Henson DE, Albores-Saavedra J, Corle D. Carcinoma of the gallbladder. Histologic types, stage of disease, grade, and survival rates. Cancer 1992;70(6):1493-7. 118. Tsuchiya Y. Early carcinoma of the gallbladder: macroscopic features and US findings. Radiology 1991;179(1):171-5. 119. Kokudo N, Makuuchi M, Natori T, et al. Strategies for surgical treatment of gallbladder carcinoma based on information available before resection. Arch Surg 2003;138(7):741-50; discussion 50. 120. Roseau G. [The application of digestive endoscopic ultrasonography in the gallbladder pathology]. Presse Med 2004;33(14 Pt 1):954-60. 121. Schwartz LH, Black J, Fong Y, et al. Gallbladder carcinoma: findings at MR imaging with MR cholangiopancreatography. Journal of computer assisted tomography 2002;26(3):405-10.
228
U. Udayasankar et al.
122. Koh T, Taniguchi H, Yamaguchi A, Kunishima S, Yamagishi H. Differential diagnosis of gallbladder cancer using positron emission tomography with fluorine-18-labeled fluorodeoxyglucose (FDG-PET). Journal of surgical oncology 2003;84(2):74-81. 123. Anderson CD, Rice MH, Pinson CW, Chapman WC, Chari RS, Delbeke D. Fluorodeoxyglucose PET imaging in the evaluation of gallbladder carcinoma and cholangiocarcinoma. J Gastrointest Surg 2004;8(1):90-7. 124. Livraghi T, Giorgio A, Marin G, et al. Hepatocellular carcinoma and cirrhosis in 746 patients: long-term results of percutaneous ethanol injection. Radiology 1995;197(1):101-8. 125. Llovet JM, Burroughs A, Bruix J. Hepatocellular carcinoma. Lancet 2003;362(9399): 1907-17. 126. Rossi S, Di Stasi M, Buscarini E, et al. Percutaneous RF interstitial thermal ablation in the treatment of hepatic cancer. Ajr 1996;167(3):759-68. 127. Lencioni R, Pinto F, Armillotta N, et al. Long-term results of percutaneous ethanol injection therapy for hepatocellular carcinoma in cirrhosis: a European experience. European radiology 1997;7(4):514-9. 128. Buscarini L, Buscarini E, Di Stasi M, Vallisa D, Quaretti P, Rocca A. Percutaneous radiofrequency ablation of small hepatocellular carcinoma: long-term results. European radiology 2001;11(6):914-21. 129. Lencioni RA, Allgaier HP, Cioni D, et al. Small hepatocellular carcinoma in cirrhosis: randomized comparison of radio-frequency thermal ablation versus percutaneous ethanol injection. Radiology 2003;228(1):235-40. 130. Mulier S, Mulier P, Ni Y, et al. Complications of radiofrequency coagulation of liver tumours. The British journal of surgery 2002;89(10):1206-22. 131. Kim SK, Lim HK, Kim YH, et al. Hepatocellular carcinoma treated with radio-frequency ablation: spectrum of imaging findings. Radiographics 2003;23(1):107-21. 132. Filippone A, Iezzi R, Di Fabio F, Cianci R, Grassedonio E, Storto ML. Multidetector-row computed tomography of focal liver lesions treated by radiofrequency ablation: spectrum of findings at long-term follow-up. Journal of computer assisted tomography 2007;31(1):42-52. 133. Chopra S, Dodd GD, 3rd, Chintapalli KN, Leyendecker JR, Karahan OI, Rhim H. Tumor recurrence after radiofrequency thermal ablation of hepatic tumors: spectrum of findings on dual-phase contrast-enhanced CT. Ajr 2001;177(2):381-7. 134. Keppke AL, Salem R, Reddy D, et al. Imaging of hepatocellular carcinoma after treatment with yttrium-90 microspheres. Ajr 2007;188(3):768-75. 135. Geschwind JF, Salem R, Carr BI, et al. Yttrium-90 microspheres for the treatment of hepatocellular carcinoma. Gastroenterology 2004;127(5 Suppl 1):S194-205. 136. Nakamura H, Hashimoto T, Oi H, Sawada S. Transcatheter oily chemoembolization of hepatocellular carcinoma. Radiology 1989;170(3 Pt 1):783-6. 137. Takayasu K, Arii S, Ikai I, et al. Prospective cohort study of transarterial chemoembolization for unresectable hepatocellular carcinoma in 8510 patients. Gastroenterology 2006;131(2): 461-9. 138. Takayasu K, Arii S, Matsuo N, et al. Comparison of CT findings with resected specimens after chemoembolization with iodized oil for hepatocellular carcinoma. Ajr 2000;175(3):699-704.
9
Recent Advances in Imaging of Pancreatic Neoplasms Chad B. Rabinowitz, MD, Hima B. Prabhakar, MD, and Dushyant V. Sahani, MD
Key Points ●
●
●
●
●
●
●
●
The use of specific imaging modalities in the workup of pancreatic neoplasms is dependent on local expertise, and, thus, familiarity of the ordering physician with multiple imaging techniques is paramount. While clinical symptoms can be suggestive of pancreatic cancer, pancreatic lesions are often detected incidentally. Some of these can be definitively characterized by imaging. Common diagnostic problems in pancreatic imaging include differentiating post-operative changes versus recurrent disease, and adenocarcinoma versus chronic pancreatitis. Multi-detector CT is the mainstay of abdominal imaging, and many surgeons will operate based on CT findings of neoplasm alone. MRI and endoscopic ultrasound are utilized as problem-solving tools. Endoscopic ultrasound has the highest sensitivity and specificity for locoregional extension of tumors, however must be performed in conjunction with other cross-sectional imaging (e.g., MDCT) to exclude peritoneal disease. FDG-PET has average sensitivity and specificity for pancreatic adenocarcinoma, and ongoing research suggests that FLT-PET may be more tumor-specific than FDG-PET. This modality shows promise in assessment of therapeutic tumor response and disease recurrence. Although spatial resolution of some current imaging techniques can reach submillimeter level, novel imaging techniques will require the exploitation of cellular differences between normal and abnormal tissue in order to improve resolution. Morphologic changes occur later than cellular changes in cancer treatment, and newer perfusion imaging techniques (CT perfusion, DCE-MRI) attempt to
Department of Abdominal Imaging and Interventional Radiology Massachusetts General Hospital - White 270, 55 Fruit Street, Boston MA 02114 Corresponding author: Dushyant V. Sahani, MD
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
229
230
●
C.B. Rabinowitz et al.
quantify changes in tissue perfusion as tumoral angiogenesis is targeted by newer drugs. The multiple techniques available in pancreatic imaging should be viewed as complementary to answer the clinical question.
Oncologic imaging of the pancreas is a challenging entity due to a large number of primary pancreatic neoplasms, as well as benign entities of the pancreas that simulate neoplasms, such as inflammatory and cystic disease. While clinical and laboratory data are able to distinguish many of the disease processes affecting the pancreas, imaging is inevitably tied to diagnosis and treatment, given the significant overlap of patient symptoms in benign and malignant pancreatic disorders. In general primary pancreatic neoplasms can be divided into three categories: solid, cystic and neuroendocrine tumors. All three of these can overlap in common imaging findings, and familiarity with available imaging modalities can help differentiate these tumors from benign disease. This chapter will discuss current and emerging techniques in pancreatic imaging, as well as their integration with oncologic care.
1
Epidemiology
Pancreatic ductal adenocarcinoma is the fourth most common cause of death among malignancies, with 33,370 deaths projected in 2007. In this year 37,170 new cases are expected, and the small difference between these two values reflects the aggressiveness of this tumor and its poor prognosis [1]. Early detection of pancreatic malignancy is paramount, as five-year survival falls off sharply, from a low of 20 percent to less than 5 percent, as local disease progresses to regional or metastatic disease [1]. Surgical resection currently offers the only chance of cure. Hereditary pancreatitis confers the highest cumulative risk at 30 percent to 40 percent, and chronic pancreatitis from multiple causes is also significantly contributory [2-4]. Specific pancreatic lesions, which include intraductal papillary mucinous neoplasm (IPMN) and mucinous cystic neoplasms (MCNs), are also associated with the subsequent development of adenocarcinoma at the site of the lesion [5], or at sites remote from the primary detected lesion [6]. We are beginning to understand that precursor lesions to pancreatic adenocarcinoma exist. A new nomenclature system has been devised [7], incorporating the term pancreatic intraepithelial neoplasia (panIN), which are non-invasive cellular changes present in pancreatic ducts. PanINs contain mutations associated with invasive adenocarcinomas, and further study into the carcinogenesis and the simultaneous existence of adenocarcinoma and other precursor lesions is necessary [8] (Fig. 9.1). While adenocarcinoma is the most lethal solid pancreatic neoplasm, a significant proportion of pancreatic lesions are of the cystic variety. Pancreatic cystic lesions are present in 15 percent to 24 percent of the population, based on recent studies
9 Recent Advances in Imaging of Pancreatic Neoplasms
231
Fig. 9.1 Multifocal pancreatic neoplasia. (a) Axial contrast-enhanced MDCT image of the pancreas shows a mucinous cystic neoplasm with a focal mural nodule (arrowheads). This was treated by surgical resection. (b) Follow-up axial MDCT image shows absence of the pancreatic body and tail consistent with surgery. A new hypodense lesion in the pancreatic head (arrowheads) was biopsy-proven adenocarcinoma, remote from the original disease
[9, 10]. Cystic neoplasms include entities such as serous and mucinous tumors. Serous cystadenomas are a common lesion, consisting of up to 25 percent of all cystic pancreatic tumors. Mucinous tumors represent 2 percent to 5 percent of all exocrine neoplasms, and consist of mucinous cystadenoma, mucinous cystadenocarcinoma and intraductal papillary mucinous neoplasm (IPMN). IPMNs vary in incidence from 1 percent to 8 percent. The serous cystadenoma is a common benign neoplasm, while mucinous cystic neoplasms range from benign to malignant. These can, however, be borderline or low-grade malignancies which are associated with the development of adenocarcinoma, depending on the amount of cellular atypia present [11, 12]. Differentiating cystic pancreatic neoplasms from benign cystic-appearing lesions is a primary clinical and imaging concern, and can be challenging if the natural history of these lesions is not understood. One area in which solid and cystic lesions can have imaging overlap is in the setting of neuroendocrine tumors. Neuroendocrine tumors are generally rare, and 85 percent of patients present with a clinical syndrome, depending on the type of tumor [13]. The majority of these tumors are small solid lesions. Cystic neuroendocrine tumors are even less common, but still comprise up to 4 percent of all pancreatic tumors [11]. The clinical presentation of these tumors overlaps with other clinical disorders, which is why imaging can be useful in confirming the presence of these lesions, especially when they are small.
1.1
Imaging Principles and Diagnostic Dilemmas
The goal of pancreatic imaging is the early detection and characterization of clinically relevant pancreatic lesions. Unfortunately, incidental cystic pancreatic lesions detected by multi-detector computed tomography (MDCT) are increasingly common
232
C.B. Rabinowitz et al.
and can range from benign incidental lesions to malignant. The resection of all cystic lesions is impractical, as a significant proportion of these lesions are benign. Given the high prevalence of pancreatic cysts, current imaging recommendations are being developed to guide management decisions. As the technology of existing imaging modalities improves, and new modalities are developed, it is important to recognize that all modalities have certain proven clinical uses, as well as limitations. Physicians must recognize and have an appreciation for the expertise that is available in their medical community to best integrate imaging findings with subsequent patient management. Lesion detection and characterization are the primary goal of radiologists. Imagers attempt to divide the aforementioned pancreatic abnormalities into “malignant or benign,” and “cystic or solid.” Occasionally, it can be difficult to answer these questions, and we need to utilize multiple imaging modalities to troubleshoot.
2 2.1
Imaging Features Solid and Neuroendocrine Tumors
Pancreatic adenocarcinoma is a dense fibrotic tumor with decreased vascularity, compared to the remainder of the gland and, thus, tumor to glandular contrast is an essential goal of imaging by whatever modality utilized for detection [14]. It is known that five-year survival in the setting of adenocarcinoma is significantly higher when lesions are detected < 1 cm [15]. Thus, this tumor is utilized as the reference lesion, which we need to exclude in cases of characterization and detection. When focal solid pancreatic lesions are detected, diagnosis can be aided by clinical and laboratory history, especially in the case of functioning neuroendocrine tumors or with invasive tumors causing biliary or pancreatic ductal obstruction. Additionally, certain imaging features on CT and MRI can be useful in narrowing a differential diagnosis. Examples of helpful imaging features include a “hypervascular” lesion, which narrows the differential diagnosis to include neuroendocrine tumors; hypervascular metastases such as renal cell carcinoma, thyroid cancer and melanoma, as well as occasionally pseudopapillary tumors. “Hypovascular” masses include adenocarcinoma and lymphoma, although the latter is a rare lesion. Lesions which are large include pseudopapillary tumors and non-functioning islet cell tumors and, less commonly, adenocarcinoma as this lesion tends to present earlier, secondary to patient symptoms [16]. It is important to recognize that large solid tumors can necrose and appear cystic. Calcification is also useful, as this can occur in large lesions when they undergo necrosis, such as in non-functioning islet cell tumors and pseudopapillary lesions. MRI is able to detect hemorrhage which is commonly seen in pseudopapillary lesions, although patient demographics (young females) will also help identify this lesion [16, 17]. Included in focal pancreatic abnormalities is the increasingly recognized non-neoplastic entity of autoimmune pancreatitis. This consists of hypergammaglobulinemia,
9 Recent Advances in Imaging of Pancreatic Neoplasms
233
mild/no clinical symptoms and occasional association with other autoimmune disorders [18]. Imaging features can include focal or diffuse pancreatic enlargement. In the presence of focal glandular enlargement, the features of vascular invasion seen in adenocarcinoma are characteristically absent. Common bile duct obstruction and pancreatic duct narrowing are common and, thus, imaging overlap with adenocarcinoma is important to recognize. While not a neoplasm, this condition is becoming increasingly diagnosed by imaging [19]. Additionally, as autoimmune pancreatitis is steroid-responsive, it is of critical importance for this differential diagnosis be included in the evaluation of focal pancreatic abnormalities.
2.2
Cystic Lesions
Cystic and ductal mucinous tumors also have radiologic and clinical features which can help weight the differential diagnosis. Microcystic and macrocystic patterns have been described corresponding to serous cystadenoma and mucinous cystadenomas. If a cystic lesion can be shown to connect to the pancreatic duct with MDCT, endoscopic retrograde cholangiopancreatography (ERCP) or magnetic resonance cholangiopancreatography (MRCP), it is likely a variant of IPMN. With a history of pancreatitis, pancreatic pseudocyst can be suggested, especially if there are imaging findings of pancreatitis such as glandular calcifications seen in chronic pancreatitis. When complexity is noted within a lesion, such as mural nodularity or rim-like calcification, pre-malignant or malignant lesions should be suspected, as in cases of mucinous cystic neoplasms (mucinous cystadenocarcinoma or malignant IPMN) [12, 20] (Fig. 9.2). In the case of IPMN, other features predictive of malignancy include main duct dilatation, diffuse ductal/multifocal involvement and large lesion size for branch type lesions [21]. A recent study looking at cysts < 3 cm on CT and magnetic resonance imaging (MRI) has shown that unilocular cysts are usually benign (97 percent PPV), while septations are associated with low-grade malignancy in 20 percent of cases [22]. The presence of any visible solid component is associated with invasive carcinoma [23, 24]. In IPMN, malignancy is reported in 7 percent to 46 percent of cases, varying from carcinoma in situ to frank adenocarcinoma [21].
2.3
Surgical and Clinical Principles
When imaging and clinical features suggest malignancy, it is important to recognize findings that indicate tumoral unresectability from a surgical standpoint. While visualized contact of neoplasm with adjacent vasculature on cross-sectional imaging is predictive of invasion based on the amount of contact [32, 33], a more recent paper has suggested that arteries and veins may need different criteria for invasion by cross-sectional imaging, as arteries are likely more resistant to invasion based upon their inherently stronger and thicker muscular wall [34, 35].
234
C.B. Rabinowitz et al.
Fig. 9.2 Mucinous neoplasms. (a) Side-branch type IPMN. Note the pancreatic duct (white arrowhead) and connections of the mucinous low density lesions to the duct (black arrowheads). If this connection can be demonstrated, the diagnosis is solidified. (b) Main duct type IPMN. Note the dilated main pancreatic duct (white arrowheads), tapering at the ampulla (black arrow). This dilated and tapered duct pattern, with isolated pancreatic ductal enlargement, is consistent with this diagnosis. (c) Mucinous cystadenoma. A focal cystic pancreatic mass is noted in the pancreatic body (black arrowheads). Thin septae are seen internally, one with focal calcification (white arrowhead). Peripheral or septal calcification is fairly specific for mucinous lesions. (d) Malignant IPMN. A complex cystic pancreatic head mass shows thick septations, and mural nodularity (arrowheads). These features are associated with malignancy
Ductal involvement in pancreatic adenocarcinoma may not be easily detected by routine cross-sectional imaging unless macroscopic ductal changes are present, such as focal ductal dilatation. When ductal changes are evident, difficulty in differentiating changes of chronic pancreatitis from adenocarcinoma occurs when tumor markers are not elevated, and when other glandular findings of chronic pancreatitis are not present on cross-sectional imaging, such as diffuse calcifications. A change in ductal caliber can either be due to focal stricture or an underlying mass that is too small to detect. Even with endoscopic ultrasound (EUS)-guided fine needle aspiration (FNA), diagnostic success is not perfect, often requiring multiple passes to exclude malignancy in the setting of a benign lesion [40, 41]. Direct comparison of EUSguided pancreatic FNA with CT-guided pancreatic FNA suggest that EUS-guided FNA may have the highest sensitivity [42, 43], however other studies have shown
9 Recent Advances in Imaging of Pancreatic Neoplasms
235
that diagnostic rates and sensitivity are similar for both EUS and CT [44]. Given that EUS has been shown to be the most sensitive imaging examination for small tumors, FNA at the time of the study may be more prudent. Additionally, the complication of peritoneal seeding has been raised using percutaneous methods of sampling [45]. The issue of pancreatitis (either chronic or focal acute) versus tumor is one of the most important issues in pancreatic imaging today. Up to 20 percent of patients with chronic pancreatitis can develop a focal mass which simulates tumors [28] (Fig. 9.3). While the presence of a non-obstructed pancreatic duct coursing through a focal pancreatic lesion has good accuracy for diagnosing a pseudotumor of chronic pancreatitis [46], this sign as described by MRCP is often equivocal in clinical practice, necessitating further study. Other significant problems affecting radiologic imaging are in the detection of micrometastatic disease to liver and peritoneum, as well as in the underestimation of vascular invasion [35]. One approach to this problem is to have patients who are candidates for resection undergo laparoscopy prior to or at the time of surgery. This is a controversial topic with its proponents stating that its routine use prior to pancreatic tumor resection will increase sensitivity for peritoneal disease in 15 percent to 51 percent of patients. Currently, even with advanced imaging techniques, unresectable disease is found at surgery in 20 percent to 57 percent when disease was deemed resectable by imaging. Some feel that routine laparoscopy is not cost-effective, and that the few studies relating to its use have many limitations [47]. Other papers suggest specific criteria for when laparoscopy should be used preoperatively, such as for tumors in the pancreatic body and tail [48, 49] and tumors greater than 3 cm [50].
Fig. 9.3 Focal Acute on Chronic Pancreatitis – (a, b) Axial plane and curved multiplanar reformatted image of the pancreas shows a pancreatic head “mass,” (black arrowheads) a dilated pancreatic duct (white arrows) in the pancreatic body and tail and involutional changes of the pancreatic tail. Pancreatic tail atrophy and ductal obstruction occurs in the setting of proximal obstruction from neoplasm, and can also be seen in chronic pancreatitis. The etiology in this case was determined only by endoscopic ultrasound and biopsy
236
2.4
C.B. Rabinowitz et al.
Imaging Modalities
Currently, there are multiple imaging modalities for evaluating the pancreas, including multidetector CT, MRI, positron emission tomography, transabdominal ultrasound, endoscopic ultrasound and scintigraphy. Newer techniques being explored include PET/CT and PET/MR fusion, CT perfusion, optical coherence tomography (OCT) and molecular imaging. Expertise in these modalities will vary locally, although there is significant overlap in the information that the modalities can provide.
3
MDCT
CT is the initial imaging test most commonly performed when abnormalities of the pancreas are clinically suspected. Its high sensitivity and specificity for pancreatic disease and non-invasive nature make it a good screening test for malignancy, and it can often assist the radiologist in diagnosing benign pancreatic disease. As the workhorse of abdominal imaging, helical CT and now MDCT have had numerous studies evaluating their use in the setting of pancreatic cancer staging. Generally, CT studies address the detection and characterization of pancreatic tumors, as well as the predictive value of CT for resectability. The most significant problems affecting radiologic imaging, and CT in general, are again in the detection of micrometastatic disease to liver and peritoneum, as well as in the underestimation of vascular invasion [35]. The primary CT sign of pancreatic neoplasm is a focal mass, however focal enlargement of the pancreatic gland is not uncommon in the absence of a discrete visualized tumor. Approximately 10 percent of tumors are not seen by CT because they are isoattenuating to surrounding parenchyma. In these cases secondary signs need to be carefully examined, including ductal dilatation, ductal interruption, pancreatic tail atrophy and abnormal pancreatic contour [53]. When there is isolated pancreatic head enlargement seen on CT, MRI will reveal a focal mass in a significant percentage of these patients [54]. CT has high negative predictive value (NPV) for cancer resectability. In a study of 84 patients with adenocarcinoma, nonresectability was established in 96 percent of cases when helical CT was performed in conjunction with pancreatic CTA [55]. This has been mirrored in several studies, including a study by Lu revealing a 93 percent NPV of CT for pancreatic mass nonresectability [32]. Later MDCT studies have shown high rates of nonresectability as well [56, 57]. When cancer is truly nonresectable radiologists are effective in identifying this. Many of these patients who subsequently undergo laparotomy are, in fact, non-operable due to micrometastases detected at the time of surgery In primary pancreatic cancer, detection by CT correlates with tumor size [58]. Overall, helical CT has a sensitivity for detection of 76 percent to 92 percent [28]. Prior studies show poor tumor sensitivities for detection of lesions smaller than
9 Recent Advances in Imaging of Pancreatic Neoplasms
237
2 cm, ranging from 58 percent to 67 percent [58, 59]. A more recent MDCT study improves on these sensitivities slightly, with a sensitivity of 72 percent to 77 percent for tumors < 2 cm [14]. More recent studies including MDCT show higher lesion sensitivity, with improvement in CT technology [60, 61]. CT is effective in the detection of small pancreatic cystic neoplasms, however characterization of these lesions is difficult utilizing this modality. As stated above, there are certain CT signs which are more suggestive of malignancy. Regarding specific lesion characterization, peripheral calcification is a feature typically seen in mucinous tumors. Overall, a blinded retrospective study of 50 patients showed a diagnostic accuracy of CT in separating serous from mucinous neoplasms ranging from only 23 percent to 41 percent [62]. A recent study had more success in distinguishing lesions such as macrocystic cystadenoma from mucinous tumors, including IPMN [63]. In the detection of neuroendocrine tumors the majority of patients have clinical symptoms, however lesion detection still remains difficult. Currently, intraoperative ultrasound has the highest sensitivity for lesion detection at 83 percent and should be considered the gold standard [64] (Fig. 9.4). CT has been shown to be able to detect lesions as small as 4 mm [65]. This study had a sensitivity of 82
Fig. 9.4 Insulinoma on CT, MRI and intra-operative US. (a) Axial contrast-enhanced MDCT image shows a hypervascular mass in the tail of the pancreas (arrow). (b) This mass is confirmed by Gadolinium-enhanced MRI (arrow). (c) Intra-operative ultrasound has the highest sensitivity for focal pancreatic lesions of these three examinations, and was able to detect this lesion before it was resected
238
C.B. Rabinowitz et al.
percent for lesions, but two lesions less than 5 mm were not detected. A more recent MDCT study showed a sensitivity of 84 percent [66], which exceeded both EUS (79 percent) and somatostatin scintigraphy (58 percent). In the setting of insulinomas a recent study showed a sensitivity of thin-section MDCT of 94 percent which was equivalent to EUS. Sensitivity of MDCT was significantly decreased when thin sections were not utilized. When CT and EUS were combined, all lesions were detected [67]. Older literature supports the superiority of EUS compared to CT in the detection of small lesions, however as MDCT improves, detection rates will undoubtedly increase. Non-functioning neuroendocrine lesions are typically larger at presentation, and detection approaches 100 percent [68].
4
MRI
MRI is typically used as a problem-solving modality, for example, when a pancreatic mass is suspected, but not identified on MDCT. It should also be considered an excellent imaging study in patients with an iodinated contrast allergy (Fig. 9.5). Generally, imaging principles of MDCT translate over to MRI, including principles of contrast enhancement (hypervascular versus hypovascular) and spatial resolution. MRI has increased tissue contrast resolution over CT, which is its primary imaging advantage. Studies reveal that the sensitivity of MRI for cancer detection utilizing contrast is similar, if not better, than that of helical CT [28], however the majority of these direct comparative studies were not performed utilizing MDCT. A recent study comparing MRI, including MRCP to MDCT in the assessment of locoregional
Fig. 9.5 Adenocarcinoma on CT and MRI. (a) axial non-contrast CT image shows focal prominence of the pancreatic neck (arrowheads). Note involution of the pancreatic tail which occurs with ductal obstruction. This patient was unable to get a contrast-enhanced CT scan. (b) Contrastenhanced MRI (using Mn-DPDP contrast) shows a focal mass responsible for the pancreatic enlargement (arrowheads). MRI can be utilized as a problem-solving modality
9 Recent Advances in Imaging of Pancreatic Neoplasms
239
extension, reports that MDCT including MPR images has significantly increased sensitivity for disease over MRI (96 percent versus 83 percent), whereas MRI was only minimally more specific (98 percent versus 97 percent) for disease [72]. This study correctly notes that, due to rapid advances in CT technology, many of the prior comparative studies need reevaluation. Additionally, an interesting study comparing MDCT to MRI in the detection of subcentimeter hepatic lesions notes that of 178 MDCT detected subcentimeter hepatic lesions, MRI was able to improve on CT specificity of lesions (97.5 percent versus 77.3 percent). Sensitivity of MRI and MDCT for subcentimeter lesions was similar (83.3 percent versus 81.2 percent) [73]. While this study was not without limitations, the specificity and accuracy of MRI may be able to help characterize lesions in cases of suspected liver metastases, especially with the high prevalence of subcentimeter hepatic metastatic disease in patients with known malignancy [74]. Contrast-enhanced MRI has been shown in a separate study to increase detection of hepatic metastases versus CT [75]. In the setting of neuroendocrine tumors, sensitivity of MRI has been reported to be 85 percent. In the setting of hepatic metastases, MRI was shown to outperform CT and somatostatin SPECT with sensitivities of 95 percent, 79 percent and 49 percent, respectively [76]. This is likely due to the higher specificity of MRI, as described above, and the lower spatial resolution of SPECT. The addition of a diffusion sequence to pancreatic MRI may be sensitive and specific for adenocarcinoma, as demonstrated in a recent study. The authors were able to obtain a sensitivity and specificity of 96.2 percent and 98.6 percent for adenocarcinoma [71]. The smallest tumor in this study was 16 mm and, thus, additional study is required for the detection of smaller tumors using this technique, as well as for its use in the detection of metastatic disease. An effective use could be in the differentiation of adenocarcinoma from chronic pancreatitis, and this sequence only requires a nominal increase in overall study time. MRCP performed after secretin administration is found to improve the detection of IPMNs. In addition to facilitating the depiction of the morphological characteristics of the lesions, they also help in detection of the communication of the branch duct IPMNs with the main pancreatic duct (Fig. 9.6). Secretin MRCP is fast emerging as the most suitable imaging modality in the diagnosis and follow up of IPMNs of the collateral branches.
5
FDG-PET and PET-CT
F18-fluorodeoxyglucose positron emission tomography (FDG-PET) is a rapidly evolving technique that can detect pancreatic tumors as small as 7 mm in size and additional distant metastases in 40 percent of patients [77]. While FDG-PET can show increased uptake within primary pancreatic tumors, as well as detect metastatic organ and nodal disease, its poor spatial localization does not allow for assessment of vascular invasion. FDG-PET has a sensitivity and
240
C.B. Rabinowitz et al.
Fig. 9.6 Secretin-enhanced MRCP. (a) Pre-Secretin MRCP maximum intensity projection (MIP) image shows the pancreatic duct (arrow) and a focal bright lesion (arrowhead) adjacent to the duct. (b) Secretin administration allows for ductal distention allowing for improved visualization. The lesion was shown to connect with the duct (not shown), consistent with side-branch IPMN
specificity for adenocarcinoma detection of 71 percent to 100 percent, and 53 percent to 100 percent, respectively, based on a 2004 metanalysis [78]. Typically, with FDG-PET scans, foci of uptake must be correlated topographically with separately acquired MRI or CT studies (Fig. 9.7). Studies have shown that when CT and FDGPET are integrated into one machine, diagnostic accuracy in the detection and localization of multiple tumor types is clearly better than with either modality alone [79]. When an integrated PET/CT system is not available, retrospective image fusion techniques can help improve the accuracy and sensitivity of FDG-PET and CT for lesion detection [80]. Retrospective fusion can also be performed utilizing FDG-PET and MR images [81]. An integrated PET/MRI system is on the horizon. The clinical role of FDG-PET in pancreatic malignancy is to detect unsuspected metastatic disease and to increase the specificity of visualized lesions, especially in the liver. In the detection of liver metastases, one study showed 70 percent sensitivity and 95 percent specificity [82], although another study by the same authors showed that the detection rate fell when lesions were < 1 cm, decreasing from 97 percent to 43 percent [82, 83]. In the detection of nodal metastases, FDG-PET demonstrated a 49 percent sensitivity and 63 percent specificity [82]. Sahani, et al. [84] confirmed that contrast-enhanced liver MRI with a hepatocyte specific contrast agent (Mangafodipir, Amersham Health, Oslo, Norway) was able to detect more metastatic lesions than whole body FDG-PET, especially when under 1 cm. It is still unclear how dual PET-CT would compare to MRI in metastatic lesion detection and characterization. Many institutions do not utilize intravenous contrast in their dual PET-CT studies, and use only noncontrast CT scans for attenuation correction. This lowers sensitivity of the CT portion of the exam for liver metastases, which are most often only evident on contrast-enhanced images. Yang, et al. did not show a significant difference in the detection of hepatic metastases between FDG-PET and MRI, but conceded that MRI is more specific [85]. This study included tumor sources other than pancreas, which may have favorably altered the FDG-PET lesion detection sensitivity. Microscopic peritoneal
9 Recent Advances in Imaging of Pancreatic Neoplasms
241
Fig. 9.7 Pancreatic adenocarcinoma on CT and FDG-PET. (a) Axial contrast-enhanced CT image demonstrates a focal hypodense mass in the pancreatic head. (b) FDG-PET shows focal tracer uptake in the pancreatic head corresponding to the mass. Separately acquired CT and PET images must be “topographically correlated” unless a dual PET/CT machine is utilized. Fusion software can also be used. FDG-PET is more often used as a problem-solving tool, as opposed to initial adenocarcinoma workup at our institution
metastases are beneath the resolution of FDG-PET. One study detected only 25 percent of peritoneal metastases [82]. Other investigations on the clinical utility of FDG-PET focus on therapeutic response and assessment of prognosis in pancreatic adenocarcinoma. The goal of FDG-PET is to separate treatment responders from non-responders, as a cellular response to treatment detectable by FDG-PET can precede morphologic changes by CT. In a small subset of patients undergoing arterial chemoinfusion and external radiotherapy, Yoshioka, et al. demonstrated a lag time in visualized tumor response by CT, with changes detected earlier by FDG-PET [86]. A pilot study by Maisey, et al. showed that a decrease in FDG uptake from baseline to zero after one month of therapy correlated with increased survival [87]. In another study of 93 patients with ductal adenocarcinoma, 15 underwent chemoradiation and were assessed by CT and FDG-PET, both pre- and post-therapy. FDG-PET was superior to CT in assessing tumor response in five of 15 patients, whereas these patients showed no response by CT [88]. Rose, et al. studied the role of FDG-PET in patients undergoing neoadjuvant chemoradiation; six patients showed a change in disease extent that was not detectable by CT [89]. Interestingly, these studies also assessed whether staging FDG-PET studies would change pre-operative management and found it did in 21.5 percent and 43 percent of patients, respectively. Additional uses FDG-PET in the evaluation of pancreatic neoplasms have been assessed. While cellular uptake of FDG and the acquisition of counts vary slightly, “dual phase” studies have been investigated which allow more time for cells to accumulate tracers before additional uptake values are established. Lychick, et al. showed that combining staging information with ratios of FDG uptake at one and two hours after injection was predictive of patient survival [90]. Another group showed that in delayed scans at two hours, uptake was significantly increased over one hour uptake
242
C.B. Rabinowitz et al.
in malignancy [91]. Conversely, cancers can show washout of tracers and have decreased uptake of FDG at two hours [77], thus diminishing the specificity of this sign.
5.1
FLT-PET
While FDG-PET has variable specificity for different diseases, attempts to improve on specificity are currently being studied, especially in the role of distinguishing between adenocarcinoma and chronic pancreatitis. 18F-flouro-3’deoxy-3’-L-fluorothymidine (18F-FLT) is a proliferation tracer which is phosphorylated by thymidine kinase 1, and is incorporated into cells which utilize a salvage pathway for DNA synthesis. 18F-FLT-PET, compared with 18F-FDG-PET, has shown to be less sensitive for disease than FDG-PET in multiple studies utilizing various cell lines and different neoplasms [92]. For example, FLT-PET shows significant liver and bone uptake, limiting its utilization for detection of metastases to these organs. FLT-PET, however, is more tumor-specific than FDG-PET [93], and its role in evaluating pancreatic cancer is currently being studied (Fig. 9.8). Since 18F - FLT is not
Fig. 9.8 Pancreatic adenocarcinoma on FLT-PET. (a) axial contrast-enhanced MDCT image through the pancreas shows a focal hypodense obstructing mass in the pancreatic head which was stented (arrowheads). (b) FLT-PET image (left to right: non-contrast CT, fused non-contrast CT / FLT-PET, FLT-PET image only) shows avid tracer uptake in the pancreatic head (arrows). This technique may be more tumor-specific than FDG-PET, although notice that significant liver and bone tracer uptake limits sensitivity for tumors in these organs
9 Recent Advances in Imaging of Pancreatic Neoplasms
243
directly incorporated into DNA, new analogues are being created that would be incorporated, perhaps more accurately reflecting cellular turnover [92]. In the setting of cystic pancreatic lesions, a recent study has shown that the sensitivity of FDG-PET was 57 percent and specificity 85 percent for malignancy [94]. In the FDG-PET-detected malignant lesions, cross-sectional imaging was able to detect malignant features and, thus, PET only confirmed these findings and did not aid in the detection of occult malignancy. Therefore, the authors have suggested that PET does not play a role in determining malignancy in pancreatic cystic lesions.
6
Transabdominal Ultrasound
The utility of transabdominal ultrasound in the diagnosis of pancreatic cancer in the United States is limited. Limiting technical factors often relate to large patient habitus and overlying bowel gas. Additionally, in the United States, ultrasound technologists are the primary imagers performing the examinations [95]. Contrast-enhanced ultrasound, with intravenously injected microbubbles, is utilized in Europe as an alternative to more expensive imaging modalities such as EUS and MRI. Although histology is the reference, there are ultrasound characteristics suggestive of various tumor types [96]. While contrast-enhanced ultrasound deserves attention, it is currently not approved for clinical use in this country. Interesting research is underway that will target microbubbles to specific pancreatic tumor vasculature. Antibodies against tumor vasculature can be synthesized, conjugated to microbubbles, and then imaged by ultrasound [97].
7
Endoscopic Ultrasound (EUS)
Endoscopic ultrasound clearly has a role in current diagnosis and staging of pancreatic abnormalities. Aside from tumor detection, EUS-guided FNA may be indicated when tissue sampling is required prior to adjuvant or palliative therapy, or when the diagnosis of carcinoma versus inflammatory disease is unclear. EUS is superior to MDCT in the detection of pancreatic lesions less than 2 to 3 mm, with a sensitivity of > 90 percent [98] for lesions this size. A recent metanalysis by DeWitt showed that studies comparing EUS and MDCT have intrinsic limitations, and that newer studies incorporating advances in imaging are required [99]. Nonetheless, EUS and MDCT appear to be similar for assessing local extension and tumor respectability, and in detecting nodal disease. An additional advantage of EUS is that fine needle aspiration can be performed at the time of study (Fig. 9.9). This technique has a negative predictive value that approaches 100 percent [100], although pathology results are dependent on pathologist experience [101], and a recent metanalysis showed an overall 88 percent sensitivity and 96 percent specificity of EUS for cancer in solid pancreatic masses [102]. Additional work is being performed assessing the use of EUS in the local administration of chemotherapeutic agents [103]. Palliative celiac neurolysis can be administered by EUS [104], although this technique is also
244
C.B. Rabinowitz et al.
Fig. 9.9 Acinar cell neoplasm on CT, MRI, and EUS – (a) Axial image from a contrast-enhanced MDCT of the pancreas shows a dilated pancreatic duct in the tail, without a clear focal obstructing mass (arrows). (b, c) T2 and T1 post-Gadolinium axial MR images mirror the CT scan findings of a dilated duct. Again, no focal mass is seen. (d) Endoscopic ultrasound revealed a 1.9 cm obstructing mass. This was biopsy-proven acinar cell neoplasm
straightforward utilizing CT-guided techniques [105]. Other EUS-guided treatment options, including tumor ablation, are being explored [102]. A reasonable approach is to avoid the use of EUS in the presence of known metastatic disease, unless tissue diagnosis is required for treatment. If a suspicious pancreatic mass is present on MDCT and the clinical context is correct, the patient can proceed to surgery if the tumor is deemed resectable. Given the high sensitivity for small lesions, EUS can be performed if the CT shows a focal abnormality without a clear mass, or if ductal signs suggest an infiltrative process and there is a suspicion for chronic pancreatitis [102]. In summary, as EUS is unable to detect distant disease, conjunction with other cross-sectional modalities is necessary for its use. EUS is excellent in the detection of neuroendocrine lesions with a sensitivity and specificity of 93 percent and 95 percent [106], however FNA is needed to make the diagnosis at the time of exam. This is especially useful in the setting of insulinomas, which are not easily seen at scintigraphy [13], and if they are also not detected by CT or MRI.
9 Recent Advances in Imaging of Pancreatic Neoplasms
8
245
Endoscopic Retrograde Cholangiopancreatography (ERCP)
This modality is currently utilized for obtaining high resolution images of the pancreatic duct, and is considered the gold standard for evaluating the pancreatic duct [107]. ERCP can also be utilized for several purposes, including endoscopic sphincterotomy, removal of stones, as well as the insertion of stents and dilation of biliary or pancreatic strictures. ERCP has high sensitivity and specificity for cancer, and is useful in detecting tumors if there is main pancreatic ductal involvement. When the main pancreatic duct is dilated in the setting of side-branch IPMN (mixed type) or in isolation (main duct type), this technique can confirm the findings of IPMN and exclude other causes of ductal dilatation, such as stricture from chronic pancreatitis. Focal uncinate or tail lesions can be missed if they do not involve the main pancreatic duct. In the setting of obstruction, ERCP is usually considered a palliative procedure. Comparisons of ERCP and EUS-guided FNA for the evaluation of biliary strictures has shown that EUS is superior to ERCP [102]. It should be noted that newer techniques such as intraductal ultrasonography, pancreatoscopy and optical coherence tomography can be performed using ERCP hardware [108]. New technology may allow the ERCP and EUS scope to be combined into one instrument (also known as EURCP) to provide both diagnostic and therapeutic techniques in the same setting, although these studies are currently often performed sequentially.
9
Intraductal Ultrasound (IDUS)
The use of a 2 mm probe fed into the pancreatic duct during ERCP over a guidewire can help to detect focal ductal lesions < 1 mm in height. This technique differs from EUS and transabdominal US in that a high frequency probe is utilized to maximize tissue resolution [108]. A limitation of this technique is when a tight ductal stricture does not allow cannulation of the duct. In one study comparing IDUS to standard ERCP, EUS and MDCT, the authors showed that IDUS is more sensitive and specific in identifying the cause of strictures than either of the other techniques (100 percent sensitivity, 93 percent specificity for cancer). The other modalities – ERCP, EUS and CT – were 86 percent, 90 percent and 67 percent sensitive, and 67 percent, 58 percent and 67 percent specific for cancer [109]. This author, in another larger study, showed similar values in the detection of pancreatic cancer [110].
10
Optical Coherence Tomography (OCT)
This is a newer technique utilizing existing ERCP hardware for evaluation of ductal epithelium. The probe can be added to an ERCP accessory port and, thus, adds only a short amount of time to the procedure after cannulation of the pancreatic duct. This examination is similar in principle to ultrasound, except that light is utilized for imaging
246
C.B. Rabinowitz et al.
instead of sound waves [111]. Visualization of the pancreatic duct epithelium directly utilizing infrared light is able to detect changes in epithelial architecture relating to interruption of the normal ductal epithelium. This optical technique allows pancreatic duct mucosal and submucosal structures to be evaluated to a depth of 1 to 2 mm. While earlier studies were able to differentiate tumors from normal epithelium [112], a more recent study was not able to distinguish tumors from other inflammatory and non-neoplastic ductal conditions [113]. Further research by the authors has been able to differentiate changes in ductal epithelium in chronic pancreatitis from neoplastic disease [114] and, should this prove to be true with further investigation, this exciting prospect may significantly change pancreatic imaging in a small subset of patients. Accuracy for malignancy detection by OCT was 100 percent versus only 67 percent for positivity of brushings in an in vivo study of 15 patients.
11
Scintigraphy
Nuclear medicine detection of hormone secreting lesions such as gastrinoma, insulinoma and other tumors of pancreatic endocrine origin is possible utilizing gamma or other photon-emitting substances coupled to a substance that binds to lesion receptors. Octreotide and Pentetreotide are somatostatin analogues that bind to somatostatin receptors, and are found on many different neuroendocrine neoplasms. These can be bound to Indium-111, which decays and emits photons that can be detected by a gamma detector. [Octreotide (Sandostatin, Novartis Pharmaceuticals), Pentetreotide (Octreoscan, Malinkrodt)]. Previously we stated that neuroendocrine lesions less than 2 cm are not consistently detected by cross-sectional imaging. This is because a significant percentage of these lesions are not located in the pancreas proper. For example, occult Gastrinomas are often located in an area called the “Gastrinoma triangle,” defined by the junction of the cystic and common bile ducts superiorly, the junction of the second and third portions of the duodenum inferiorly and the junction of the neck and body of the pancreas medially [117]. Sensitivity varies on the type of tumor being imaged, however one study assessing gastrinoma showed a 58 percent lesion sensitivity with scintigraphy, which was better than all other modalities (CT, MRI, US). In fact, the other modalities only increased the overall detection rate to 68 percent [118].
11.1
Emerging Technologies: Molecular/Angiogenesis Imaging
While lesion detection by imaging continues to approach the histologic and surgical standard, we note that the spatial resolution of the above modalities is only on the order of millimeters in the best possible scenario. Detecting changes of panIN and precursor lesions will require novel approaches that exploit cellular abnormalities, such as alterations in cellular receptor sites.
9 Recent Advances in Imaging of Pancreatic Neoplasms
247
For example, utilizing a mouse model, one set of researchers were able to utilize a magneto/fluorescent nanoparticle conjugate that was targeted to bombesin receptors of pancreatic ductal cells [119]. Pancreatic ductal adenocarcinomas, unlike normal pancreatic ductal cells, do not express this binding receptor [120]. The experimental model was able to demonstrate a decrease in MRI T2 signal intensity in normal pancreatic tissue, thus increasing signal intensity of the pancreatic tumor and the affected pancreatic specimen. Translation into clinical imaging may be promising. Numerous other molecular agents are in development to increase the sensitivity and specificity of MRI, as well as other imaging. Research has shown that tumor growth is dependent on the production of a vascular network beyond 2 mm3, and numerous growth factors that regulate this, such as vascular endothelial growth factor (VEGF), tumor necrosis factor (TNF), fibroblast growth factor (FGF) as well as interleukins (IL) have been demonstrated [121]. Anti-angiogenesis agents are in development, and the ability to monitor tissue perfusional changes before macroscopic changes are evident will be critical in patient management.
11.2
CT Perfusion
Utilizing a workstation and imaging software, one can dynamically assess tumoral bloodflow and monitor response to chemotherapy and radiation changes using bloodflow and tissue perfusion parameters. Since adenocarcinoma of the pancreas is hypovascular compared to normal pancreatic parenchyma, the tumor would involve an area of hypoperfusion, compared to normal pancreatic tissue. CT perfusion measurements of pancreatic tissue have been shown to be technically feasible [122]. A recent study from China has also shown the utility of pancreatic CT perfusion in the characterization and detection of insulinomas, a typically hypervascular tumor [123]. These lesions show increased blood flow and blood volume, compared with normal pancreatic tissue. Perfusion CT of the pancreas is an exciting and emerging technique which demands further study.
11.3
Dynamic Contrast-Enhanced MRI (DCE-MRI)
Similar to perfusion CT, tissue perfusion and permeability can be assessed before, during and after contrast administration. Tissue parameters as measured by T1 and T2 signal characteristics can be quantified by the use of imaging software. The main use for this technique has been in the development of anti-angiogenic drugs, however DCE-MRI has shown some difficulty with reproducibility across trials, and issues of reliable measurements in heterogeneous tumors have been noted. Further validation is also needed regarding protocol design [121, 124]. It should be noted that a study by Johnson, et al reported that this technique was unable to distinguish between chronic pancreatitis and cancer [125].
248
11.4
C.B. Rabinowitz et al.
Conclusion
When pancreatic neoplasm is suspected, there are a large number of imaging modalities available which should be viewed as complementary. MDCT imaging should be considered as the initial imaging study, regardless of suspected tumor type, with MRI utilized for further lesion characterization and to increase specificity in the setting of cystic pancreatic lesions and small liver lesions. When clinically apparent lesions are not detected, EUS should be utilized for troubleshooting and tissue diagnosis, due to its high sensitivity for local disease, with ERCP utilized for palliative stent placement as necessary. When hormonal syndromes suggest a neuroendocrine tumor, and if CT does not detect the lesion, scintigraphy should be performed before the patient has localization by EUS or intraoperative ultrasound, as it is non-invasive. Additionally, it should be noted that operative techniques such as laparoscopy and ultrasound have a high sensitivity for disease and, thus, can play a role in conjunction with imaging. Current challenges include avoiding excessive imaging in patients with incidentally detected cystic lesions, increasing early detection in aggressive disease with more tumor specific imaging and improving staging accuracy without increasing the amount of imaging needed to do so. There are numerous unresolved issues in the world of pancreatic tumoral imaging, however new research in tumor pathogenesis will hopefully add to patient care, and eventually improve survival in those patients with advanced disease.
References 1. Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 2007. CA Cancer J Clin. 2007 Jan-Feb;57(1):43-66. 2. Lowenfels AB, Maisonneuve P. Epidemiology and prevention of pancreatic cancer. Jpn J Clin Oncol. 2004 May;34(5):238-44. 3. Lowenfels AB, Maisonneuve P, Cavallini G, et al. Pancreatitis and the risk of pancreatic cancer. International Pancreatitis Study Group. N Engl J Med 1993;328:1433-7. 4. Lowenfels AB, Maisonneuve P, DiMagno EP, et al. Hereditary pancreatitis and the risk of pancreatic cancer. International Hereditary Pancreatitis Study Group. J Natl Cancer Inst.1997;89:442-6. 5. Spinelli KS, Fromwiller TE, Daniel RA, et al. Cystic pancreatic neoplasms: observe or operate. Ann Surg. 2004 May;239(5):651-7 6. Tada M, Kawabe T, Arizumi M, et al. Pancreatic cancer in patients with pancreatic cystic lesions: a prospective study in 197 patients. Clin Gastroenterol Hepatol. 2006 Oct;4(10):1265-70. 7. Hruban RH, Adsay NV, Albores-Saavedra J, et al. Pancreatic intraepithelial neoplasia: a new nomenclature and classification system for pancreatic duct lesions. Am J Surg Pathol. 2001 May;25(5):579-86. 8. Maitra A, Adsay NV, Argani P, et al. Multicomponent analysis of the pancreatic adenocarcinoma progression model using a pancreatic intraepithelial neoplasia tissue microarray. Mod Pathol. 2003 Sep;16(9):902-12.
9 Recent Advances in Imaging of Pancreatic Neoplasms
249
9. Kimura W, Nagai H, Kuroda A, Muto T, Esaki Y. Analysis of small cystic lesions of the pancreas. Int J Pancreatol. 1995 Dec;18(3):197-206. 10. Zhang XM, Mitchell DG, Dohke M, Holland GA, Parker L. Pancreatic cysts: depiction on single-shot fast spin-echo MR images. Radiology. 2002 May;223(2):547-53. 11. Brugge WR. Cystic pancreatic lesions: can we diagnose them accurately? What to look for. FNA marker molecular analysis resection, surveillance, or endoscopic treatment? Endoscopy. 2006 Jun;38 Suppl 1:S40-7. 12. Sahani DV, Kadavigere R, Saokar A, Fernandez-del Castillo C, Brugge WR, Hahn PF. Cystic pancreatic lesions: a simple imaging-based classification system for guiding management. Radiographics. 2005 Nov-Dec;25(6):1471-84. 13. Nichols MT, Russ PD, Chen YK. Pancreatic imaging: current and emerging technologies. Pancreas. 2006 Oct;33(3):211-20. 14. Bronstein YL, Loyer EM, Kaur H, et al. Detection of small pancreatic tumors with multiphasic helical CT. AJR Am J Roentgenol. 2004 Mar;182(3):619-23. 15. Ariyama J, Suyama M, Satoh K, Sai J. Imaging of small pancreatic ductal adenocarcinoma. Pancreas. 1998 Apr;16(3):396-401. 16. Manoharan P, Sheridan M B. Neoplasms of the Pancreas. Imaging 2004. (16): 323-337. 17. Mergo PJ, Helmberger TK, Buetow PC, Helmberger RC, Ros PR. Pancreatic neoplams: MR imaging and pathologic correlation. Radiographics. 1997 Mar-Apr;17(2):281-301. 18. Finkelberg DL, Sahani D, Deshpande V, Brugge WR. Autoimmune pancreatitis. N Engl J Med. 2006 Dec 21;355(25):2670-6. 19. Sahani DV, Kalva SP, Farrell J, et al. Autoimmune pancreatitis: imaging features. Radiology. 2004 Nov;233(2):345-52. 20. Brugge WR, Lauwers GY, Sahani D, Fernandez-del Castillo C, Warshaw AL. Cystic neoplasms of the pancreas. N Engl J Med. 2004 Sep 16;351(12):1218-26. 21. Kawamoto S, Lawler LP, Horton KM, Eng J, Hruban RH, Fishman EK. MDCT of intraductal papillary mucinous neoplasm of the pancreas: evaluation of features predictive of invasive carcinoma. AJR Am J Roentgenol. 2006 Mar;186(3):687-95. 22. Sahani DV, Saokar A, Hahn PF, Brugge WR, Fernandez-Del Castillo C. Pancreatic cysts 3 cm or smaller: how aggressive should treatment be? Radiology. 2006 Mar;238(3):912-9. 23. Megibow AJ, Lombardo FP, Guarise A, et al. Cystic pancreatic masses: cross-sectional imaging observations and serial follow-up. Abdom Imaging. 2001 Nov-Dec;26(6): 640-7. 24. Allen PJ, Jaques DP, D’Angelica M, Bowne WB, Conlon KC, Brennan MF. Cystic lesions of the pancreas: selection criteria for operative and nonoperative management in 209 patients. J Gastrointest Surg. 2003 Dec;7(8):970-7. 25. Duraker N, Hot S, Polat Y, Hobek A, Gencler N, Urhan N. CEA, CA 19-9, and CA 125 in the differential diagnosis of benign and malignant pancreatic diseases with or without jaundice. J Surg Oncol. 2007 Feb 1;95(2):142-7. 26. Goonetilleke KS, Siriwardena AK. Systematic review of carbohydrate antigen (CA 19-9) as a biochemical marker in the diagnosis of pancreatic cancer. Eur J Surg Oncol. 2006 Nov 8; [Epub ahead of print]. 27. Albert MB, Steinberg WM, Henry JP. Elevated serum levels of tumor marker CA19-9 in acute cholangitis. Dig Dis Sci. 1988 Oct;33(10):1223-5. 28. Schima W, Ba-Ssalamah A, Kolblinger C, Kulinna-Cosentini C, Puespoek A, Gotzinger P. Pancreatic adenocarcinoma. Eur Radiol. 2007 Mar;17(3):638-49. 29. Leach SD, Lee JE, Charnsangavej C, et al. Survival following pancreaticoduodenectomy with resection of the superior mesenteric-portal vein confluence for adenocarcinoma of the pancreatic head. Br J Surg. 1998 May;85(5):611-7. 30. Cameron JL, Riall TS, Coleman J, Belcher KA. One thousand consecutive pancreaticoduodenectomies. Ann Surg. 2006 Jul;244(1):10-5. 31. Geer RJ, Brennan MF. Prognostic indicators for survival after resection of pancreatic adenocarcinoma. Am J Surg. 1993 Jan;165(1):68-72.
250
C.B. Rabinowitz et al.
32. Lu DS, Reber HA, Krasny RM, Kadell BM, Sayre J. Local staging of pancreatic cancer: criteria for unresectability of major vessels as revealed by pancreatic-phase, thin-section helical CT. AJR Am J Roentgenol. 1997 Jun;168(6):1439-43. 33. O’Malley ME, Boland GW, Wood BJ, Fernandez-del Castillo C, Warshaw AL, Mueller PR. Adenocarcinoma of the head of the pancreas: determination of surgical unresectability with thin-section pancreatic-phase helical CT. AJR Am J Roentgenol. 1999 Dec;173(6):1513-8. 34. Li H, Zeng MS, Zhou KR, Jin DY, Lou WH. Pancreatic adenocarcinoma: the different CT criteria for peripancreatic major arterial and venous invasion. J Comput Assist Tomogr. 2005 Mar-Apr;29(2):170-5. 35. Li H, Zeng MS, Zhou KR, Jin DY, Lou WH. Pancreatic adenocarcinoma: signs of vascular invasion determined by multi-detector row CT. Br J Radiol. 2006 Nov;79(947):880-7. 36. Kalser MH, Barkin J, MacIntyre JM. Pancreatic cancer. Assessment of prognosis by clinical presentation. Cancer. 1985 Jul 15;56(2):397-402. 37. Vogt DP. Pancreatic cancer: a current overview. Curr Surg. 2000 May 1;57(3):214-220. 38. Reber HA. Small pancreatic tumors: Is size an indication of curability? J hepatobiliary Pancreat Surg 1995; 2(4): 384-386. 39. Spencer MP, Sarr MG, Nagorney DM. Radical pancreatectomy for pancreatic cancer in the elderly. Is it safe and justified? Ann Surg. 1990 Aug;212(2):140-3. 40. Farrell JJ. Diagnosing pancreatic malignancy in the setting of chronic pancreatitis: is there room for improvement? Gastrointest Endosc. 2005 Nov;62(5):737-41. 41. Varadarajulu S, Tamhane A, Eloubeidi MA. Yield of EUS-guided FNA of pancreatic masses in the presence or the absence of chronic pancreatitis. Gastrointest Endosc. 2005 Nov;62(5):728-36. 42. Volmar KE, Vollmer RT, Jowell PS, Nelson RC, Xie HB. Pancreatic FNA in 1000 cases: a comparison of imaging modalities. Gastrointest Endosc. 2005 Jun;61(7):854-61. 43. Horwhat JD, Paulson EK, McGrath K, et al. A randomized comparison of EUS-guided FNA versus CT or US-guided FNA for the evaluation of pancreatic mass lesions. Gastrointest Endosc. 2006 Jun;63(7):966-75. 44. Erturk SM, Mortele KJ, Tuncali K, Saltzman JR, Lao R, Silverman SG. Fine-needle aspiration biopsy of solid pancreatic masses: comparison of CT and endoscopic sonography guidance. AJR Am J Roentgenol. 2006 Dec;187(6):1531-5. 45. Micames C, Jowell PS, White R, et al. Lower frequency of peritoneal carcinomatosis in patients with pancreatic cancer diagnosed by EUS-guided FNA versus percutaneous FNA. Gastrointest Endosc. 2003 Nov;58(5):690-5. 46. Ichikawa T, Sou H, Araki T, et al. Duct-penetrating sign at MRCP: usefulness for differentiating inflammatory pancreatic mass from pancreatic carcinomas. Radiology. 2001 Oct;221(1):107-16. 47. Stefanidis D, Grove KD, Schwesinger WH, Thomas CR Jr. The current role of staging laparoscopy for adenocarcinoma of the pancreas: a review. Ann Oncol. 2006 Feb;17(2):189-99. 48. Fernandez-del Castillo C, Warshaw AL. Laparoscopy for staging in pancreatic carcinoma. Surg Oncol. 1993;2 Suppl 1:25-9. 49. Liu RC, Traverso LW. Diagnostic laparoscopy improves staging of pancreatic cancer deemed locally unresectable by computed tomography. Surg Endosc. 2005. May; 19(5):638-42. 50. Ichikawa T, Erturk SM, Sou H, et al. MDCT of Pancreatic Adenocarcinoma: Optimal Imaging Phases and Multiplanar Reformatted Imaging. AJR Am J Roentgenol. Dec 2006 (187): 1513-1520 51. Morganti AG, Brizi MG, Macchia G, et al. The prognostic effect of clinical staging in pancreatic adenocarcinoma.Ann Surg Oncol. 2005 Feb;12(2):145-51. 52. Schueller G, Schima W, Schueller-Weidekamm C, et al. Multidetector CT of pancreas: effects of contrast material flow rate and individualized scan delay on enhancement of pancreas and tumor contrast. Radiology. 2006 Nov;241(2):441-8. 53. Prokesch RW, Chow LC, Beaulieu CF, Bammer R, Jeffrey RB Jr. Isoattenuating pancreatic adenocarcinoma at multi-detector row CT: secondary signs. Radiology. 2002 Sep;224(3): 764-8. 54. Semelka RC, Kelekis NL, Molina PL, Sharp TJ, Calvo B. Pancreatic masses with inconclusive findings on spiral CT: is there a role for MRI? J Magn Reson Imaging. 1996 Jul-Aug;6(4):585-8.
9 Recent Advances in Imaging of Pancreatic Neoplasms
251
55. Raptopoulos V, Steer ML, Sheiman RG, Vrachliotis TG, Gougoutas CA, Movson JS. The use of helical CT and CT angiography to predict vascular involvement from pancreatic cancer: correlation with findings at surgery. AJR Am J Roentgenol. 1997 Apr;168(4):971-7. 56. Ellsmere J, Mortele K, Sahani D, et al. Does multidetector-row CT eliminate the role of diagnostic laparoscopy in assessing the resectability of pancreatic head adenocarcinoma? Surg Endosc. 2005 Mar;19(3):369-73. 57. Vargas R, Nino-Murcia M, Trueblood W, Jeffrey RB Jr. MDCT in Pancreatic adenocarcinoma: prediction of vascular invasion and resectability using a multiphasic technique with curved planar reformations. AJR Am J Roentgenol 2004 Feb;182(2):419-25. 58. Legmann P, Vignaux O, Dousset B, et al. Pancreatic tumors: comparison of dual-phase helical CT and endoscopic sonography. AJR Am J Roentgenol. 1998 May;170(5):1315-22. 59. Ichikawa T, Haradome H, Hachiya J, et al. Pancreatic ductal adenocarcinoma: preoperative assessment with helical CT versus dynamic MR imaging. Radiology. 1997 Mar;202(3):655-62. 60. Grenacher L, Klauss M, Dukic L, et al. Diagnosis and staging of pancreatic carcinoma: MRI versus multislice-CT – a prospective study. Rofo. 2004 Nov;176(11):1624-33. 61. DeWitt J, Devereaux B, Chriswell M, McGreevy,et al. Comparison of endoscopic ultrasonography and multidetector computed tomography for detecting and staging pancreatic cancer. Ann Intern Med. 2004 Nov 16;141(10):753-63. 62. Curry CA, Eng J, Horton KM, et al. CT of primary cystic pancreatic neoplasms: can CT be used for patient triage and treatment? AJR Am J Roentgenol. 2000 Jul;175(1):99-103. 63. Kim SY, Lee JM, Kim SH, et al. Macrocystic neoplasms of the pancreas: CT differentiation of serous oligocystic adenoma from mucinous cystadenoma and intraductal papillary mucinous tumor. AJR Am J Roentgenol. 2006 Nov;187(5):1192-8. 64. Kang CM, Park SH, Kim KS, Choi JS, Lee WJ, Kim BR. Surgical experiences of functioning neuroendocrine neoplasm of the pancreas. Yonsei Med J. 2006 Dec 31; 47(6):833-9. 65. Van Hoe L, Gryspeerdt S, Marchal G, Baert AL, Mertens L. Helical CT for the preoperative localization of islet cell tumors of the pancreas: value of arterial and parenchymal phase images. AJR Am J Roentgenol. 1995 Dec;165(6):1437-9. 66. Rappeport ED, Hansen CP, Kjaer A, Knigge U. Multidetector computed tomography and neuroendocrine pancreaticoduodenal tumors. Acta Radiol. 2006 Apr;47(3):248-56. 67. Gouya H, Vignaux O, Augui J, Dousset B, Palazzo L, Louvel A, et al. CT, endoscopic sonography, and a combined protocol for preoperative evaluation of pancreatic insulinomas. AJR Am J Roentgenol. 2003 Oct;181(4):987-92. 68. King CM, Reznek RH, Dacie JE, Wass JA. Imaging islet cell tumours. Clin Radiol. 1994 May;49(5):295-303. 69. Sodickson A, Mortele KJ, Barish MA, et al. Three-dimensional fast-recovery fast spin-echo MRCP: comparison with two-dimensional single-shot fast spin-echo techniques. Radiology. 2006; 238(2):549-59. 70. Hellerhoff KJ, Helmberger H 3rd, Rosch T, et al. Dynamic MR pancreatography after secretin administration: image quality and diagnostic accuracy. AJR Am J Roentgenol. 2002; 179(1):121-9. 71. Ichikawa T, Erturk SM, Motosugi U, et al. High-b value diffusion-weighted MRI for detecting pancreatic adenocarcinoma: preliminary results. AJR Am J Roentgenol. 2007 Feb;188(2):409-14. 72. Erturk SM, Ichikawa T, Sou H, et al. Pancreatic Adenocarcinoma: MDCT Versus MRI in the Detection and Assessment of Locoregional Extension. JCAT 2006; 30(4):583-590 73. Holalkere NS, Sahani DV, Blake MA, Halpern EF, Hahn PF, Mueller PR. Characterization of small liver lesions: Added role of MR after MDCT. J Comput Assist Tomogr. 2006 Jul-Aug;30(4):591-6. 74. Schwartz LH, Gandras EJ, Colangelo SM, Ercolani MC, Panicek DM. Prevalence and importance of small hepatic lesions found at CT in patients with cancer. Radiology. 1999 Jan;210(1):71-4. 75. Schima W, Fugger R, Schober E, et al. Diagnosis and staging of pancreatic cancer: comparison of mangafodipir trisodium-enhanced MR imaging and contrast-enhanced helical hydroCT. AJR Am J Roentgenol. 2002 Sep;179(3):717-24
252
C.B. Rabinowitz et al.
76. Dromain C, de Baere T, Lumbroso J, et al. Detection of liver metastases from endocrine tumors: a prospective comparison of somatostatin receptor scintigraphy, computed tomography, and magnetic resonance imaging. J Clin Oncol. 2005 Jan 1;23(1):70-8. 77. Higashi T, Saga T, Nakamoto Y, et al. Diagnosis of pancreatic cancer using fluorine-18 fluorodeoxyglucose positron emission tomography (FDG PET) –usefulness and limitations in “clinical reality”. Ann Nucl Med. 2003 Jun;17(4):261-79. 78. Orlando LA, Kulasingam SL, Matchar DB. Meta-analysis: the detection of pancreatic malignancy with positron emission tomography. Aliment Pharmacol Ther. 2004 Nov 15;20(10):1063-70. 79. Rosenbaum SJ, Stergar H, Antoch G, Veit P, Bockisch A, Kuhl H. Staging and follow-up of gastrointestinal tumors with PET/CT. Abdom Imaging. 2006 Jan-Feb;31(1):25-35. 80. Lemke AJ, Niehues SM, Hosten N, et al. Retrospective digital image fusion of multidetector CT and 18F-FDG PET: clinical value in pancreatic lesions–a prospective study with 104 patients. J Nucl Med. 2004 Aug;45(8):1279-86. 81. Ruf J, Lopez Hanninen E, Bohmig M, et al. Impact of FDG-PET/MRI Image Fusion on the Detection of Pancreatic Cancer. Pancreatology. 2006 Nov 13;6(6):512-519. 82. Diederichs CG, Staib L, Vogel J, et al. Values and limitations of 18F-fluorodeoxyglucose-positron-emission tomography with preoperative evaluation of patients with pancreatic masses. Pancreas. 2000 Mar; 20(2):109-16. 83. Frohlich A, Diederichs CG, Staib L, Vogel J, Beger HG, Reske SN. Detection of liver metastases from pancreatic cancer using FDG PET. J Nucl Med. 1999 Feb;40(2):250-5. 84. Sahani DV, Kalva SP, Fischman AJ, et al. Detection of liver metastases from adenocarcinoma of the colon and pancreas: comparison of mangafodipir trisodium-enhanced liver MRI and whole-body FDG PET. AJR Am J Roentgenol. 2005 Jul;185(1):239-46. 85. Yang M, Martin DR, Karabulut N, Frick MP. Comparison of MR and PET imaging for the evaluation of liver metastases. J Magn Reson Imaging. 2003 Mar;17(3):343-9. 86. Yoshioka M, Sato T, Furuya T, et al. Role of positron emission tomography with 2-deoxy-2[18F]fluoro-D-glucose in evaluating the effects of arterial infusion chemotherapy and radiotherapy on pancreatic cancer. J Gastroenterol. 2004 Jan;39(1):50-5. 87. Maisey NR, Webb A, Flux GD, et al. FDG-PET in the prediction of survival of patients with cancer of the pancreas: a pilot study. J Cancer. 2000 Aug;83(3):287-93. 88. Bang S, Chung HW, Park SW, et al. The clinical usefulness of 18-fluorodeoxyglucose positron emission tomography in the differential diagnosis, staging, and response evaluation after concurrent chemoradiotherapy for pancreatic cancer. J Clin Gastroenterol. 2006 Nov-Dec;40(10):923-9. 89. Rose DM, Delbeke D, Beauchamp RD, et al. 18Fluorodeoxyglucose-positron emission tomo graphy in the management of patients with suspected pancreatic cancer. Ann Surg. 1999 May;229(5):729-37. 90. Lyshchik A, Higashi T, Nakamoto Y, et al. Dual-phase 18F-fluoro-2-deoxy-D-glucose positron emission tomography as a prognostic parameter in patients with pancreatic cancer. Eur J Nucl Med Mol Imaging. 2005 Apr;32(4):389-97. 91. Nishiyama Y, Yamamoto Y, Monden T, et al. Evaluation of delayed additional FDG PET imaging in patients with pancreatic tumour. Nucl Med Commun. 2005 Oct;26(10):895-901. 92. Been LB, Suurmeijer AJ, Cobben DC, Jager PL, Hoekstra HJ, Elsinga PH. [18F]FLT-PET in oncology: current status and opportunities. Eur J Nucl Med Mol Imaging. 2004 Dec;31(12):1659-72. 93. van Waarde A, Jager PL, Ishiwata K, Dierckx RA, Elsinga PH. Comparison of sigma-ligands and metabolic PET tracers for differentiating tumor from inflammation. J Nucl Med. 2006 Jan;47(1):150-4. 94. Mansour JC, Schwartz L, Pandit-Taskar N, et al. The utility of F-18 fluorodeoxyglucose whole body PET imaging for determining malignancy in cystic lesions of the pancreas. J Gastrointest Surg. 2006 Dec;10(10):1354-60. 95. Cronan JJ. Ultrasound: Is there a future in diagnostic imaging? JACR 2006; 3(9): 645-6 96. Rickes S, Monkemuller K, Malfertheiner P. Contrast-enhanced ultrasound in the diagnosis of pancreatic tumors. JOP. 2006 Nov 10;7(6):584-92.
9 Recent Advances in Imaging of Pancreatic Neoplasms
253
97. Korpanty G, Carbon JG, Grayburn PA, Fleming JB, Brekken RA. Monitoring response to anticancer therapy by targeting microbubbles to tumor vasculature. Clin Cancer Res. 2007 Jan 1;13(1):323-30. 98. Calculli L, Pezzilli R, Casadei R, Fiscaletti, et al. The imaging of pancreatic exocrine solid tumors: the role of computed tomography and positron emission tomography. JOP. 2007 Jan 9;8(1 Suppl):77-84. 99. Dewitt J, Devereaux BM, Lehman GA, Sherman S, Imperiale TF. Comparison of endoscopic ultrasound and computed tomography for the preoperative evaluation of pancreatic cancer: a systematic review. Clin Gastroenterol Hepatol. 2006 Jun;4(6):717-25 100. Klapman JB, Chang KJ, Lee JG, Nguyen P. Negative predictive value of endoscopic ultrasound in a large series of patients with a clinical suspicion of pancreatic cancer. Am J Gastroenterol. 2005 Dec;100(12):2658-61. 101. Alsibai KD, Denis B, Bottlaender J, Kleinclaus I, Straub P, Fabre M. Impact of cytopathologist expert on diagnosis and treatment of pancreatic lesions in current clinical practice. A series of 106 endoscopic ultrasound-guided fine needle aspirations. Cytopathology. 2006 Feb;17(1):18-26. 102. De Angelis C, Repici A, Carucci P, et al. Pancreatic cancer imaging: the new role of endoscopic ultrasound. JOP. 2007 Jan 9;8(1 Suppl):85-97. 103. Micames CG, Gress FG. Local EUS-guided injection of chemotherapeutic agents as adjuvant to systemic treatment: the first steps are made. Gastrointest Endosc. 2006 Dec 13; [Epub ahead of print]. 104. Tran QN, Urayama S, Meyers FJ. Endoscopic ultrasound-guided celiac plexus neurolysis for pancreatic cancer pain: a single-institution experience and review of the literature. J Support Oncol. 2006 Oct;4(9):460-2, 464; discussion 463-4. 105. Wang PJ, Shang MY, Qian Z, Shao CW, Wang JH, Zhao XH. CT-guided percutaneous neurolytic celiac plexus block technique. Abdom Imaging. 2006 Dec 7; [Epub ahead of print]. 106. Anderson MA, Carpenter S, Thompson NW, Nostrant TT, Elta GH, Scheiman JM. Endoscopic ultrasound is highly accurate and directs management in patients with neuroendocrine tumors of the pancreas. Am J Gastroenterol. 2000 Sep;95(9):2271-7. 107. Kwon RS, Scheiman JM. New advances in pancreatic imaging. Curr Opin Gastroenterol. 2006 Sep;22(5):512-9. 108. Fujita N, Noda Y, Kobayashi G, Kimura K, Ito K. Endoscopic approach to early diagnosis of pancreatic cancer. Pancreas. 2004 Apr;28(3):279-81. 109. Furukawa T, Tsukamoto Y, Naitoh Y, Hirooka Y, Hayakawa T. Differential diagnosis between benign and malignant localized stenosis of the main pancreatic duct by intraductal ultrasound of the pancreas. Am J Gastroenterol. 1994 Nov;89(11):2038-41. 110. Furukawa T, Oohashi K, Yamao K, et al. Intraductal ultrasonography of the pancreas: development and clinical potential. Endoscopy. 1997 Aug;29(6):561-9. 111. Testoni PA, Mariani A, Mangiavillano B, et al. Main pancreatic duct, common bile duct and sphincter of Oddi structure visualized by optical coherence tomography: An ex vivo study compared with histology. Dig Liver Dis. 2006 Jun;38(6):409-14. 112. Testoni PA, Mangiavillano B, Albarello L, et al. Optical coherence tomography to detect epithelial lesions of the main pancreatic duct: an Ex Vivo study. Am J Gastroenterol. 2005 Dec;100(12):2777-83. 113. Testoni PA, Mariani A, Mangiavillano B, Arcidiacono PG, Masci E. Preliminary data on the use of intraductal optical coherence tomography during ERCP for investigating main pancreatic duct strictures. Gut. 2006 Nov;55(11):1680-1. 114. Testoni PA, Mariani A, Mangiavillano B, Arcidiacono PG, Di Pietro S, Masci E. Intraductal Optical Coherence Tomography for Investigating Main Pancreatic Duct Strictures. Am J Gastroenterol 2007;102:269-274 115. Roach PJ, Schembri GP, Ho Shon IA, Bailey EA, Bailey DL. SPECT/CT imaging using a spiral CT scanner for anatomical localization: Impact on diagnostic accuracy and reporter confidence in clinical practice. Nucl Med Commun. 2006 Dec;27(12):977-87.
254
C.B. Rabinowitz et al.
116. Ingui CJ, Shah NP, Oates ME. Endocrine neoplasm scintigraphy: added value of fusing SPECT/CT images compared with traditional side-by-side analysis. Clin Nucl Med. 2006 Nov;31(11):665-72. 117. Stabile BE, Morrow DJ, Passaro E Jr. The gastrinoma triangle: operative implications. Am J Surg. 1984 Jan;147(1):25-31. 118. Gibril F, Reynolds JC, Doppman JL, et al. Somatostatin receptor scintigraphy: its sensitivity compared with that of other imaging methods in detecting primary and metastatic gastrinomas. A prospective study. Ann Intern Med. 1996 Jul 1;125(1):26-34. 119. Montet X, Weissleder R, Josephson L. Imaging pancreatic cancer with a peptide-nanoparticle conjugate targeted to normal pancreas. Bioconjug Chem. 2006 Jul-Aug;17(4):905-11. 120. Fleischmann A, Laderach U, Friess H, Buechler MW, Reubi JC. Bombesin receptors in distinct tissue compartments of human pancreatic diseases. Lab Invest 2000. Dec;80(12):1807-17. 121. Rehman S, Jayson GC. Molecular imaging of antiangiogenic agents. Oncologist. 2005 Feb;10(2):92-103. 122. Miles KA, Hayball MP, Dixon AK. Measurement of human pancreatic perfusion using dynamic computed tomography with perfusion imaging. Br J Radiol. 1995 May;68(809):471-5. 123. Xue HD, Jin ZY, Liu W, Wang Y, Zhao WM. Perfusion characteristics of normal pancreas and insulinoma on multi-slice spiral CT. Zhongguo Yi Xue Ke Xue Yuan Xue Bao. 2006 Feb;28(1):68-70. 124. Kiessling F, Morgenstern B, Zhang C. Contrast agents and applications to assess tumor angiogenesis in vivo by magnetic resonance imaging. Curr Med Chem. 2007;14(1):77-91. 125. Johnson PT, Outwater EK. Pancreatic carcinoma versus chronic pancreatitis: dynamic MR imaging. Radiology. 1999 Jul. 212(1):213-8.
10
Imaging of Colorectal Carcinoma Jorge A. Soto, MD
1
Introduction and Epidemiology
Colorectal cancer is the second most common cause of cancer-related death in the United States, where the annual incidence is estimated at 150,000 cases [1, 2]. American adults have a 5 percent chance of developing a colorectal carcinoma and approximately a 2 percent chance of dying from the disease [3]. Colorectal cancer is also an important cause of cancer-related mortality in many other Western countries, although distribution of the disease varies widely throughout the world. Mortality from colorectal cancer is similar for males and females. Importantly, as is the case for many malignancies, mortality is directly related to the stage at the time of diagnosis, with five-year survival decreasing from over 80 percent for early-stage disease, to less than 10 percent for patients with distant metastases [2]. Unfortunately, less than 40 percent of colorectal carcinomas are diagnosed before the disease has spread beyond the wall of the colon or rectum. Factors that increase the risk for developing colorectal cancer include genetic predisposition, such as familial adenomatous polyposis syndrome (FAP), hereditary non-polyposis colorectal cancer (HNPCC), family history of colorectal cancer in a first-degree relative (especially if younger than 60 years) and personal history of colon cancer. Individuals with a first-degree relative with colorectal cancer have a lifetime risk of 12 percent to 15 percent [4]. Risk increases with age (more than 90 percent of colorectal carcinomas occur in patients older than 50 years) and with underlying conditions such as chronic inflammatory bowel disease (especially ulcerative colitis), diabetes, smoking and alcohol consumption. It is also widely believed that diets low in fiber and high in fat content and animal protein are also associated with a higher risk of developing colorectal cancer.
Boston Medical Center, 88 E. Newton Street, #H2504, Boston, MA, 02118-2308 e-mail:
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
255
256
2
J.A. Soto
Pathophysiology
It is well accepted that the vast majority of carcinomas arising from the mucosa of the colon and rectum originate from a precursor, namely the adenomatous polyp [5-9]. The theory that adenomas progress to carcinomas (“adenoma - carcinoma sequence”) is supported by the fact that the relative frequency with which both adenomas and carcinomas are found in the rectum and the various segments of the colon is very similar, although the mean age of appearance of adenomas occurs several years before that of carcinomas. Approximately one-half of the cancers are in the rectum and sigmoid colon, whereas the remainder are scattered throughout the proximal segments of the colon [10]. Patients with large (> 1 cm) colonic adenomas develop carcinomas with a frequency that surpasses that of adults without adenomas or family history of adenomas or carcinomas (“average-risk” adults) [10-12]. Histologically, adenomas are classified as tubular, tubulovillous and villous. Most colonic adenomas begin as tubular adenomas. As they grow, however, mutations can lead adenomas to develop foci of displasia or villous changes and, when the villous component predominates, they are referred to as villous adenomas. The risk of carcinoma is directly related to the presence of villous changes. When transformation of adenomas to carcinomas does occur, this process takes place over a long period of time estimated between seven and 10 years, depending upon the size of the adenoma. Thus, the risk of harboring foci of high-grade dysplasia or carcinoma is directly proportional to the size of the adenoma. This risk is estimated at less than 1 percent for polyps less than 1 cm in size, 10 percent for polyps between 1 and 2 cm in size, and greater than 25 percent for polyps larger than 2 cm in size [3, 8]. Malignant polyps grow faster than benign polyps. Removing intermediate size and large polyps decreases the frequency of colorectal cancer. Thus, much of the effort spent in screening for colorectal cancer hinges upon the identification of advanced adenomas, the vast majority of which are 1 cm or greater in diameter.
3
Screening
Colorectal cancer is especially well suited for successfully decreasing the diseasespecific mortality with the implementation of broad screening strategies. The main reason is that screening methods are directed towards detection and removal of precancerous lesions (adenomas) or early stage carcinomas. This differs from strategies used for detecting other tumors, such as breast or prostate cancer, where the lesion sought is the cancer itself. It should be noted that the terms “polyp” and “adenoma” are not interchangeable, as a “polyp” refers to any focal protrusion arising from the wall into the colonic lumen, whereas an “adenoma” refers to a neoplastic epithelial lesion. Other non-neoplastic, histological types of polyps include inflammatory or hyperplastic. For individuals with an average risk of developing colorectal cancer, it is recommended that screening start at the age of 50 years. Various tests and methods have
10 Imaging of Colorectal Carcinoma
257
been extensively studied as a means for detecting colorectal neoplasia [13]. These include fecal occult blood test (FOBT), endoscopic techniques such as sigmoidoscopy and optical colonoscopy and imaging tests such as double contrast barium enema and, more recently, CT colonography. Data have proved that screening with any method is better than no screening at all, and that the incidence of and mortality from colorectal cancer can both be decreased with adequate screening. The American Cancer Society recommends one of the following as acceptable strategies for screening: yearly FOBT, flexible sigmoidoscopy every five years, DCBE every five years or optical colonoscopy every 10 years. Unfortunately, public compliance with these strategies for colorectal cancer screening strategies is suboptimal and continues to be a main focus of attention of multiple agencies, especially the American Cancer Society. From the imaging point of view, enthusiasm about screening with double contrast barium enema has diminished considerably in recent years [14]. Many factors are responsible for this, but the most important one is the growing doubt about the performance of the test for detecting intermediate size and large polyps [14, 15]. There is, however, growing evidence that the performance of CT colonography exceeds that of double contrast barium enema and, in fact, may rival that of optical colonoscopy [16-20]. In the near future it is expected that CT colonography will be added to the list of acceptable options by the ACS. The expectation is that this will result in an increase in the fraction of eligible adults that are screened, by attracting individuals who have refused other methods.
4
Clinical Presentation
Colorectal cancer is a slow-growing tumor. Presenting symptoms vary with the specific location, size and stage of the tumor. Bleeding is a common presenting sign, and this may occur overtly as bright red blood per rectum or insidiously as iron deficiency anemia. Other presenting symptoms include abdominal pain secondary to developing bowel obstruction, changes in bowel habits or less specific symptoms such as weight loss, fever and malaise.
5
Imaging Detection of Colorectal Neoplasia
Traditionally, imaging methods have played a critical role in the detection, staging and surveillance of patients with colorectal neoplasia. The two imaging techniques commonly used today for the detection of polyps and tumors are the double contrast barium enema (DCBE) and, more recently, CT colonography. Contrast-enhanced CT and MRI techniques or PET/CT are preferred for local staging, and for evaluating regional and distant spread of cancers. Finally, high resolution MRI methods or intracavitary ultrasonography are used when accurate determination of the depth of
258
J.A. Soto
wall invasion is important for therapeutic decisions. Although a thorough description of the technical details that are necessary to ensure good quality imaging examinations is beyond the scope of this book, it is important to emphasize that good technique is critical for accurate detection of colorectal neoplasms.
6
Colonic Polyp Detection
On double contrast barium enema the appearance of a polyp depends upon its morphologic characteristics, the location within the colonic wall relative to the X-ray beam and the variable contact with barium and/or air of the polyp surface. Sessile polyps have a broad base of attachment to the colonic mucosa and are seen on enema examinations as filling defects (Fig. 10.1), rings or contour deformities. Sessile polyps, by definition, are fixed to the colonic wall and can be separated from fecal residue that is freely movable and almost always lies against the dependent wall. Conversely,
Fig. 10.1 Sessile polyp in the sigmoid colon demonstrated on a double contrast barium enema. The polyp has a broad base of attachment to the colonic mucosa and is sharply outlined by barium
10 Imaging of Colorectal Carcinoma
259
pedunculated polyps are attached to the wall by a stalk (Fig. 10.2) that allows the free end of the polyp (head) to move easily within the lumen. Polyps with villous elements typically exhibit a more irregular surface with a frond-like appearance. On CT colonography, sessile polyps are seen as focal protrusions of colonic mucosa-based lesions into the lumen of the bowel. Characteristically, sessile polyps have a smooth, cap-like surface and are seen on both the supine and prone image sets in the same location and do not move with changes in patient position (Fig. 10.3), unless the colon itself moves or rotates. CT colonography images demonstrate pedunculated polyps as focal lesions arising from the wall of the colon as well, but the free portion of the lesion changes in location when the patient moves from a supine to a prone position (Fig. 10.4). Villous adenomas are typically larger than tubular adenomas and on CT colonography also tend to have a more irregular surface (Fig. 10.5). The internal composition of colonic polyps is homogeneous and of soft tissue attenuation. On the contrary, stool residues have an irregular surface and a more heterogeneous internal attenuation with fatty and gas components. Furthermore, typical residual stool changes in position between supine and prone images, and tend to be located on the most dependent aspect of the colon. The best method to avoid misdiagnosing stool residue as polyps is to ensure adequate and complete cleansing of the colon with a full cathartic preparation. More recently, methods for stool and
Fig. 10.2 A spot image of a single contrast barium enema shows a polypoid lesion in the descending colon (arrow). The polyp is attached to the wall of the colon by a stalk (open arrow)
260
J.A. Soto
Fig. 10.3 8 mm sessile polyp seen on CT colonography. The broad-based polypoid lesion does not modify its location between the supine (a, arrow) and prone (b, arrow) positions. Note gravitational change in position of high density, iodine-tagged fluid. The endoluminal 3-D volume rendered image (c) confirms the sessile morphology of the polyp
fluid tagging have been added to the preparation regime for CT colonography, thus reducing the likelihood of false positive interpretations from this source (Fig. 10.3). Extensive work by several groups aims at testing the feasibility of performing CT colonography without a cathartic preparation (“prep-less” technique) [21-23].
7
Colorectal Carcinoma Detection
The search for colonic carcinomas on double contrast barium enema or CT colonography entails an exercise that is similar to that of the search for polyps. Early cancers have the appearance of large polyps, more commonly sessile and
10 Imaging of Colorectal Carcinoma
261
Fig. 10.4 10 mm pedunculated polyp seen on CT colonography. The polypoid lesion appears to slightly modify its location between the supine (a, arrow) and prone (b, arrow) positions. The short stalk is also seen on b (open arrow)
Fig. 10.5 Supine image of a CT colongraphy examination demonstrates a polypoid lesion with an irregular, frond-like surface in the descending colon (arrow). This lesion was histologically proven to be a villous adenoma
262
J.A. Soto
Fig. 10.6 Double contrast barium enema demonstrates circumferential narrowing of the splenic flexure caused by a mass with irregular, ulcerated, surfaces and overhanging edges. This is the typical “apple core” appearance of annular carcinomas
with a flat or irregular surface. Unfortunately, the majority of colonic carcinomas are diagnosed when they are in an advanced stage. These tumors are often polypoid or mass lesions that displace the column of barium or cause large, irregular contour defects in the colonic wall. With double contrast, the irregular surface, presence of ulcerations and broad base of attachment are better demonstrated. On CT colonography carcinomas manifest as fixed, irregular areas of wall thickening with an ulcerated surface, and cause a variable degree of lumen narrowing. As they grow, carcinomas commonly involve the wall in a circumferential fashion, leading to annular tumors which produce the typical “apple core” lesions on barium enema examinations (Fig. 10.6). Importantly, approximately 5 percent of patients with colon cancer harbor additional (synchronous) foci of carcinoma and an even larger percentage have adenomatous polyps. On barium enema examinations, it may be difficult or impossible to differentiate between strictures caused by carcinomas and complicated diverticular disease. Thus, if there is any doubt about the nature of a wall abnormality, sigmoidoscopy or colonoscopy are recommended. The reported sensitivity of double contrast barium enema for detecting colorectal cancer vary in the literature from 60 percent to
10 Imaging of Colorectal Carcinoma
263
Fig. 10.7 Large bowel obstruction caused by colon cancer. The scout topogram (a) shows marked dilatation of the colon, with little or no gas in the rectum. Axial CT image of the pelvis demonstrates the annular obstructing sigmoid mass (arrow), as well as marked dilatation and retention of gas and fluid in the segments proximal to the mass
70 percent, to higher than 90 percent. However, as mentioned in preceding sections, the sensitivity of double contrast barium enema for detection of polyps is much lower [24-26]). Better performance results are typically obtained by radiologists who have a special interest in gastrointestinal imaging. Complications of colorectal cancer include obstruction, perforation, fistula formation and bleeding. Colon cancer is a common cause of large bowel obstruction, especially when localized in the sigmoid or descending segments (Fig. 10.7). Perforation more commonly manifests as a pericolonic abscess, or may be the origin of a fistula communicating the lumen of the colon with nearby organs such as the urinary bladder, duodenum, stomach, gallbladder or vagina. The imaging findings of these complications will vary, depending upon the specific imaging technique used.
8
Staging of Colorectal Carcinoma
As is the case with other hollow viscera, staging of colorectal cancer takes into account the depth of invasion of the colonic wall, spread into pericolonic tissues and nearby organs, regional spread to draining lymph nodes and involvement of distant organs via hematogenous or peritoneal invasion. Even for patients who
264
J.A. Soto
undergo resection and are, thus, staged surgically, pathology can only identify metastases within the resection specimens and has no capability for detecting remote disease. As a result of this, many patients undergo futile operations for disease that could never have been cured by surgery alone. Several classifications of tumor stage have been described, but the TNM classification (Table 10.1) is currently the most used clinical standard to guide therapy. Prognosis and choice of type of therapy are determined by the stage of the tumor at the time of diagnosis. Accurate preoperative staging of colorectal cancer determines the surgical approach, which differs between colon and rectal cancer. Additionally, patient eligibility for clinical trials often hinges on accurate staging. In colon cancer, generous resections are generally performed; this achieves wide tumor-free margins and includes resection of multiple regional lymph node chains, including the mesenteric root. In rectal cancer, wide tumor-free margins are more difficult to achieve. Rectal tumors with only superficial involvement of the rectal wall may be susceptible to transanal resection. Deeper or transmural involvement generally require a total mesorectal excision, in which all the mesorectal tissues enveloped by the intact visceral layer of the pelvic fascia are resected. More advanced rectal tumors, with direct invasion of perirectal tissues, may be susceptible to neoadjuvant chemotherapy or radiation therapy (or both) prior to resection [27, 28]. Preoperative radiation therapy has also been proposed prior to mesorectal excisions [29]. From the preceding discussion, it is apparent that imaging plays a critical role in TNM staging and, therefore, in determining the type of therapy offered to colorectal cancer patients. CT and MRI have been used extensively for the preoperative staging Table 10.1 TNM Classification of Colorectal Carcinoma Stage Finding Tumor T1 T2 T3 T4
Tumor invades submucosa Tumor invades muscularis propria Tumor invades muscularis propria into subserosa or nonperitonealized pericolic or perirectal tissue Tumor directly invades other organs or structures and/or perforates visceral peritoneum
Regional nodal metastasis NX N0 N1 N2 N3
Regional lymph nodes cannot be assessed No nodal metastasis Metastasis in one to three pericolic or perirectal nodes Metastasis in four or more pericolic or perirectal nodes Metastasis in any node along course of a named vascular trunk and/or metatasis to apical node(s)
Distant metastasis MX M0 M1
Presence of distant metastasis cannot be assessed No distant metastasis Distant metastasis
10 Imaging of Colorectal Carcinoma
265
of colorectal carcinoma, with variable results. Findings associated with transmural spread of tumor include an irregular, serrated or spiculated outer contour of the mass (Fig. 10.8), loss of fat planes between the large bowel and surrounding muscles, a mass directly invading a nearby organ, poor definition of fascial planes or strands of soft tissue extending to the perirectal or pericolonic fat tissues. The tumor can directly invade the seminal vesicles, prostate, bladder, uterus, small bowel, bones or other organs. However, fat planes between the mass and surrounding tissues or organs can be obliterated by inflammation or fibrous reaction to the tumor without actual invasion. CT and MRI have benefited from technological advances in hardware and software, such as multi-detector technology (CT) and high resolution surface coils and parallel imaging (MRI). Unfortunately, correlation with operative findings and histopathological findings is imperfect, as definite invasion demonstrated by imaging findings is usually obvious upon macroscopic dissection, whereas microscopic invasion eludes preoperative diagnosis. Early studies showed sensitivity performance of CT between 55 percent and 60 percent for determining local
Fig. 10.8 Coronal CT reformation of axial CT data demonstrates a soft tissue mass in the wall of the cecum (arrow). Note the irregular, serrated outer contours of the mass with strands of soft tissue, indicating transmural spread of tumor, which was confirmed at laparotomy
266
J.A. Soto
invasion, as compared to the TNM classification [30, 31]. Multi-detector technology, with higher spatial resolution, may allow a more accurate estimation of the depth of mural invasion[32]. In general, CT is more accurate in detecting T4 and T3 lesions than T2 and T1 lesions. High resolution multi-planar MRI with surface coils compares favorably with CT for accurate staging of local extension of disease.
9
Rectal Cancer Staging
Rectal cancer is associated with a poor prognosis because of the risk both for metastases and for local recurrence after surgery. Incomplete removal of the lateral spread of the tumor is the cause of the majority of these recurrences. Results of several histopathologic studies have revealed the importance of extramural tumor spread and the influence of this spreading on prognosis [33-36]. In one of the largest series published, T3 tumors with extramural spread of more than 5 mm were associated with a five-year cancer-specific patient survival rate of only 54 percent, but T3 tumors with 5 mm or less of extramural spread—regardless of whether lymph node involvement was present—were associated with a five-year cancer-specific survival rate of greater than 85 percent [33]. With the increasing availability of newer preoperative (neoadjuvant) therapy options, an accurate and reproducible staging technique is, therefore, essential to enabling colorectal specialist multidisciplinary teams to consider potentially complex treatment options. The challenge for preoperative imaging in rectal cancer is to accurately determine the depth of mural involvement by the tumor (T stage), and the distance from the tumor to the circumferential mesorectal resection plane. Endorectal US, MRI and CT (Fig. 10.9) have been used for this purpose [37].
Fig. 10.9 Axial CT scan of rectal cancer. The mass involves nearly the entire circumference of the rectal wall. Note the strands of soft tissue extending to the peri-rectal fat (arrows), suggesting transmural extension of tumor. At surgery, this was, in fact, proven to represent tumor extending beyond the rectal wall (T3 disease)
10 Imaging of Colorectal Carcinoma
267
Endorectal US is now an established modality for evaluation of the integrity of the rectal wall layers. With accuracies for T staging varying between 65 percent and 95 percent [38-40], endorectal US is very accurate for staging of superficial rectal tumors, but is not as useful for staging of advanced rectal cancer [19]. The overall staging accuracy for US in bulky tumors is less because the limited depth of acoustic penetration prevents accurate assessment of local tumor extent. Thus, although endorectal US is useful for staging of superficial rectal cancer, it is less suitable for evaluation of the mesorectal excision plane. Moreover, endoluminal US is not able to depict lymph nodes that are outside the range of the transducer, and cannot discriminate between lymph nodes inside or outside the mesorectal fascia, since the fascia is not identified at endoluminal US. This may explain the more recent widespread use of MRI, since these limitations do not apply to MRI with external coils. CT has the advantage of evaluating the whole pelvis. Although early studies [41, 42] with CT reported high accuracy for staging locally advanced rectal cancer, more recent work, including a larger percentage of less advanced tumors, showed less encouraging results [43, 44] with accuracies varying between 52 percent and 74 percent. The low contrast and spatial resolution of CT protocols does not allow a detailed evaluation of the different layers of the rectal wall and may contribute to the low performance of CT for staging of superficial tumors. It is possible that the new-generation multi–detector row CT scanners, with improved spatial resolution and reconstructions in multiple planes, may provide better performance than conventional CT scanners [45, 46]. MRI is the most widely used technique for the local staging of rectal cancer [47-50]. The two major advantages of thin-section MRI are the ability to differentiate malignant tissue from the muscularis propria, allowing differentiation between T2 and T3 lesions (Fig. 10.10) and clear delineation of the mesorectal fascia (Fig. 10.11), which forms the circumferential resection margin at total mesorectal excision. This is a definite advantage over US, as determining the relationship of tumors with the mesorectal fascia has become increasingly important, perhaps as important as T stage determination. A standard protocol for MRI of rectal cancer consists of high-resolution T2-weighted fast spin-echo sequences, with or without the addition of contrast-enhanced sequences. Although endorectal coils have been used [51-53], most institutions prefer surface phased-array coil [54-56]. Staging failures, however, have been known to occur with MRI in the differentiation of T2 tumors (e.g., those confined to the rectal wall) and borderline T3 tumors (e.g., those that infiltrate the mesorectum). There is also a tendency for overstaging that is mainly attributed to desmoplastic reaction, which can cause spiculations in the perirectal fat that may or may not contain viable tumor cells. In a recent large multi-center study that compared high resolution MRI with mesorectal excision specimens, the depth of tumor spread depicted on the thin-section MR images was within 5 mm of the histopathologic measurement in the majority of patients [57]. Early work with 3 Tesla MRI suggests that improvement in accuracy for rectal cancer staging is only marginal [58].
268
J.A. Soto
Fig. 10.10 These two cases illustrate the ability of MRI to differentiate between T2 and T3 rectal tumors. A coronal T2-weighted image (a) demonstrates neoplastic thickening of the right rectal wall (arrow), without transmural extension. On a different patient (b), an axial T2-weighted image shows a tumor infiltrating beyond the outline of the rectal wall (arrow). Both cases courtesy of Michael A. Blake, MD
10
Lymph Node Detection
One of the major roles of preoperative imaging in colorectal cancer is the identification of a tumor that has spread beyond the wall of the colon. At any phase in the evaluation of patients with colorectal cancer, demonstration of systemic metastasis has profound therapeutic and prognostic implications. In the absence of systemic metastases nodal status become important, and when unresectable nodal metastases have been excluded, T-stage becomes important. However, identification of nodal disease is still a diagnostic problem for the radiologist. To determine the nodal stage of colorectal carcinoma, a radiologist must be aware of the predictable patterns of lymph node drainage from the affected portion of the colon [40, 59, 60]. The distribution of regional lymph node metastases in carcinoma of the left side of the colon, rectum and anus can be well shown with CT or MRI. Recognizing the location of nodes in the mesocolic, left colic and inferior mesenteric artery nodal groups is helpful for developing a systematic approach for detecting nodal metastases [60]. Carcinomas of the cecum, right colon and proximal transverse colon can metastasize to local mesenteric nodes (Fig. 10.12), and then to peripancreatic lymph nodes, simulating primary pancreatic cancer [61]. Tumors arising from the upper portion of the rectum drain to the inferior mesenteric nodal chain, whereas those arising from the lower rectum drain laterally and into the internal iliac node groups (Fig. 10-11). Imaging is capable only of depicting enlarged lymph nodes, recognizing that enlargement can also be secondary to reactive or hyperplastic nodes from associated inflammation. Lymph nodes should be measured in short axis, and the upper limit of normal varies with the specific location but, in general, is accepted to be 10 mm for retroperitoneal, mesenteric, external iliac and inguinal nodes, 8 mm for internal iliac, obturator and lateral sacral nodes and 5 mm for perirectal nodes. The
10 Imaging of Colorectal Carcinoma
269
Fig. 10.11 Axial T2-weighted MRI image demonstrates a tumor mass involving the complete circumference of the rectum. The mesorectal fascia is preserved on the left side (arrow), but appears to be involved on the right side (open arrow). In addition, note multiple enlarged perirectal nodes (arrowheads). Other images (not shown) demonstrated clear evidence of invasion of the prostate gland (T4 stage). Case courtesy of Michael A. Blake, MD
addition of [18F] Fluorodeoxyglucose (FDG) positron emission tomography (PET) aids in increasing the specificity of CT by adding a functional element to the purely anatomical and morphological information provided by CT [62]. Sensitivity for detecting of tumors in normal-size lymph nodes can also be improved by MRI after administration of ultra-small particles of iron oxide [63]. Early experience with this agent indicates that high resolution T2-weighted images can detect foci of rectal cancer in mesorectal lymph nodes 3 to 4 mm in size [63].
11
Search for Liver Metastases: US, CT, MRI
Hematogenous spread of colorectal cancer tumor cells to the liver is a common problem in clinical practice and is likely the result of the dual blood supply of the liver through the hepatic artery and portal vein. The liver serves as the first end-capillary bed and can easily trap the tumor cells or emboli. Liver metastases ultimately develop in approximately 40 percent of patients who undergo curative resection of colorectal
270
J.A. Soto
Fig. 10.12 Axial contrast-enhanced CT scan demonstrates a bulky mass arising from the cecum, with enlarged lymph nodes in the regional mesentery (arrows)
cancer. The development of liver metastases is a poor prognostic sign. For other cancers this usually indicates that disease is no longer curable. However, aggressive resection of a limited number of colorectal cancer liver metastases may be associated with long term survival [64-66]. Therefore, detection and accurate determination of the precise number and size of liver metastases is particularly important. Survival rates of up to 20 percent to 40 percent have been reported after wide resections of liver metastases from colorectal cancer. As image- guided therapy of liver tumors increases in popularity, the need for accurate staging will also increase. Thus, in a patient with newly diagnosed colorectal cancer, a thorough evaluation of the liver to rule out metastases is mandatory prior to bowel resection with curative intent. As metastases grow they become progressively easier to detect with imaging modalities. Blood tests that are commonly used to follow patients with colorectal cancer, and to identify those patients that require additional evaluation, include measurements of serum carcinoembryonic antigen (CEA) and liver function tests. Unfortunately, the sensitivity of CEA measurements is low (50 percent to 60 percent) [67, 68], and its use in practice is limited. Many imaging modalities have been used for detecting liver metastases with variable success. Regardless of the technique used, the ability to detect a focal space- occupying lesion in the liver depends on the size of the tumor, the spatial and contrast resolution of the imaging method, the difference in contrast and perfusion between the tumor and background liver parenchyma, and the adequacy of the method used for displaying images after
10 Imaging of Colorectal Carcinoma
271
acquired [69]. In general, a test is useful if sensitivity remains high at an acceptable specificity level. In a meta-analysis that studied the detection rate of liver metastases with multiple modalities, Kinkel, et al. [70] suggested that the minimum acceptable specificity of imaging tests for this indication should be 85 percent. CT and MRI are the most widely used techniques for evaluating the liver in the initial staging and follow-up of colorectal cancer patients. For detecting liver metastases, carefully performed CT and MRI studies with state-of-the-art equipment and interpretation by experienced radiologists afford similar good results [44, 71, 72]. Other modalities such as ultrasonography and, more recently, PET imaging are also used in specific circumstances. The sensitivity and specificity of ultrasonography improve substantially with the addition of microbubble contrast agents, which essentially augment the doppler and harmonic ultrasound signal [73, 74]. Ultrasound contrast materials, however, are not widely used due to limited availability and a general perception that the examination becomes excessively time consuming and elaborate. Intraoperative ultrasonography has higher sensitivity than transabdominal ultrasound, CT and MRI [75, 76]. Therefore, during resection of liver metastases, intraoperative ultrasound provides valuable information that may alter the preoperative surgical plan. CT is usually preferred because it is more widely available and because it is well established for evaluating the extra-hepatic abdominal organs and other tissues. On CT the typical colorectal cancer metastasis is hypovascular and appears hypoattenuating relative to background liver parenchyma (Fig. 10.13). Thus, for adequate detection, administration of intravenous contrast material and scanning during the peak of liver enhancement are critical. Peak enhancement typically occurs during the portal venous dominant phase, which occurs approximately 60 to 80 seconds after the initiation of contrast injection. Parenchymal attenuation should increase by at least 50 Hounsfield units with intravenous contrast for an adequate CT examination. Therefore, good CT technique requires administration of appropriate volume and concentration of iodine in the contrast material used, as well as adequate technique for contrast delivery. Studies using intraoperative palpation and ultrasound as standard of reference have reported high per-lesion sensitivity of greater than 85 percent [77, 78]. With the recent introduction of multi-detector CT scanners, it is likely that sensitivity may increase to 90 percent to 95 percent on a per-lesion basis using intraoperative findings, with ultrasonography as the standard of reference. Early data suggests that this is the case [79]. Enthusiasm about the use of CT during arterial photography, an invasive technique that requires catheterization of the superior mesenteric or splenic artery for direct injection of contrast, has decreased since the arrival of the latest generation CT scanners. Detection of metastases with MRI requires the acquisition of multiple sequences and administration of intravenous contrast. Although the appearance of colorectal cancer metastases on MRI is variable, the T1 and T2 relaxation times of metastases are prolonged relative to normal liver parenchyma. This typically results in hipointensity on T1-weighted sequences and hyperintensity on T2-weighted images (Fig. 10-14). Metastases can also have a perilesional halo of high signal, indicating viable tumor, or demonstrate a “doughnut” or “target” appearance (Fig. 10.14). An advantage of MRI is the superior ability to characterize multiple lesions and
272
J.A. Soto
Fig. 10.13 Axial contrast-enhanced CT image demonstrates multiple low attenuation, poorly circumscribed masses in the liver. Some lesions demonstrate central areas of necrosis
differentiate solid, benign lesions such as cysts and hemangiomas from metastases. On heavily T2-weighted scans, fluid-containing lesions (cysts and hemangiomas) typically remain hyperintense, whereas metastases drop signal and demonstrate lower intensity. Similar to CT, detection of metastases with contrast-enhanced MRI is maximized during the portal venous phase (Fig. 10-14). The reported sensitivity of MRI using multiple combinations of sequences and gadolinium chelates as contrast material varies between 65 percent and 95 percent [70, 80-82], with a mean of approximately 80 percent. Administration of liver-specific contrast agents that are taken up selectively by the hepatocytes or, less often, Kupffer cells provide a modest increase in sensitivity [83-85]. Benefits of their use have not been broadly accepted, though their use in specific circumstances is likely to increase in the future.
12
PET and PET/CT
FDG-PET is a useful imaging tool in the management of patients with colorectal carcinoma. This technique is able to measure and visualize metabolic changes in tumor cells. Interestingly, avidity for FDG is not limited to carcinomatous cells, but is also seen in adenomatous polyps [86]. This feature results in the theoretical ability to distinguish viable tumors from scar tissue, and in the detection of tumor foci at an earlier stage than generally possible with CT or MRI. There are now
10 Imaging of Colorectal Carcinoma
273
Fig. 10.14 Typical appearance of colorectal cancer metastases on MRI. The lesions demonstrate low signal intensity on T1-weighted images (a), intermediate signal intensity with a perilesional halo of high signal on T2-weighted images (b) and peripheral, annular, enhancement after intravenous administration of gadolinium chelates (c)
accumulating data that PET/CT could be used as the first test to assess metastatic disease and lymphadenopathy (M- and N-stage, respectively) for evaluating cancers with an intermediate to high pre-test likelihood of metastatic disease [62, 87, 88]. In this setting there is great opportunity for subsequently selecting and tailoring the performance of CT or MRI to define the structural relations of abnormalities identified by PET, when this information would be of relevance to management planning. FDG-PET plays a pivotal role in staging patients before surgical resection of recurrence and metastases, in the localization of recurrence in patients with an unexplained rise in serum carcinoembryonic antigen (Fig. 10-15) and in assessment of residual masses after treatment. In the presurgical evaluation FDG-PET is also best used in conjunction with anatomic imaging to combine the benefits of both anatomical and functional information, which leads to improvements in preoperative staging and preoperative judgment on the feasibility of resection. Another advantage of FDG-PET is the ability to evaluate the whole body with a single examination (Fig. 10-15). Although the ability of FDG-PET to detect small
10 Imaging of Colorectal Carcinoma
275
(subcentimeter) liver metastases is inferior to high resolution MRI or state-of-theart CT, it increases the specificity of cross-sectional imaging methods for detecting extra-hepatic disease in the abdomen [89]. It also appears that FDG-PET (especially when combined with CT) has great potential in monitoring the success of ablative therapies, and in the prediction and evaluation of response to radiotherapy, systemic therapy and combinations. Integration of FDG-PET into the management algorithm of colorectal cancer patients alters and improves therapeutic decisions, and may also reduce morbidity due to unnecessary surgery.
13
Post-treatment Follow-up
Imaging re-staging of colorectal carcinoma after treatment with surgery, radiation and/or chemotherapy poses additional challenges. The sequelae of prior treatment can be difficult to differentiate from residual cancer, and the likelihood of successful salvage therapy is even less than at presentation. Falsely assigning post-therapy changes to recurrent disease may potentially lead to subjecting patients to additional morbid treatments when cure has already been achieved. Thus, in post-treatment follow-up, the presence and extent of disease is equally critical to treatment selection and patient outcome as it is in primary staging. Unfortunately, in most patients receiving chemotherapy for colorectal metastases, a complete response on CT scan does not mean cure [90]. As stated in the preceding section, there is increasing evidence that FDG-PET (combined with CT or MRI for anatomical correlation of findings) may be the best modality for a comprehensive imaging monitoring of progression or regression of disease.
Key Points ● ●
●
●
Colorectal cancer continues to be a common and deadly disease. Since many of the disease-specific cancer deaths are potentially preventable by timely removal of adenomatous polyps, continued efforts focus on educating the public to achieve population-wide screening of average risk adults. It is expected that CT colonography will play a major role in achieving this goal. However, once colorectal cancer develops, the most important role of imaging is accurately staging the disease.
Fig. 10.15 Utility of FDG-PET following resection of colorectal carcinoma. Coronal CT reformation (a) demonstrates slightly enlarged retroperitoneal lymph nodes (arrows). Whole-body FDG-PET image shows hypermetabolic foci matching the location of these lymph nodes. Additional hypermetabolic foci are seen in the mediastinum and hila. Tumor recurrence was confirmed in this patient with prior left hemicolectomy for colon cancer and rising CEA levels
276 ● ●
●
J.A. Soto
The TNM classification is currently the preferred method for staging. Precise delineation of depth of mural involvement, transmural extension, lymph node invasion and detection of liver metastases are specific tasks that the various imaging techniques and methods are expected to perform. Recent developments that have improved performance of imaging tests include MDCT, high-resolution MRI with endocavitary coils in some cases, high resolution endosonography, PET and PET/CT and organ-specific contrast agents for MRI.
References 1. Ransohoff DF, Sandler RS. Clinical practice: screening for colorectal cancer. N Engl J Med 2002;346:40-44. 2. Jemal A, Tiwari RC, Murray T, et al. Cancer statistics, 2004. CA Cancer J Clin 2004;54:8-29. 3. Winawer SJ, Fletcher RH, Miller L, et al. Colorectal cancer screening: clinical guidelines and rationale. Gastroenterology 1997;112:594-601. 4. Bresalier RS, Kim YS. Malignant neoplasms of the large intestine. In: Sleisenger MH, Fordtran JS, eds. Gastrointestinal disease. 5th ed. Philadelphia, Pa: Saunders, 1993:1449-1494. 5. Winawer SJ. Natural history of colorectal cancer. Am J Med 1999;106(1A):3S-6S. 6. Luk GD. Epidemiology, etiology and diagnosis of colorectal neoplasia. Curr Opin Gastroenterol 1993;9:19-27. 7. Lane N, Fenolglio CM. The adenoma-carcinoma sequence in the stomach and colon. I. Observations on the adenoma as precursor to ordienary large bowel carcinoma. Gastrointest Radiol 1976;1:111-119. 8. Muto T, Bussey HJR, Morson BC. The evolution of cancer of the colon and rectum. Cancer 1975; 36:2251-2270. 9. Morson BC. Genesis of colorectal cancer. Clin Gastroenterol 1976;5:505-525. 10. Rickert RR, Auerbach O, Garfinkel L, et al. Adenomatous lesions of the large bowel: an autopsy survey. Cancer 1979; 43:1847-1857. 11. Ransohoff DF, Lang CA. Screening for colorectal cancer. N Engl J Med 1991;325:37-41. 12. Winaver SJ, Zauber AG, Ho MN, et al. Prevention of colorectal carcinoma by colonoscopic polypectomy. N Engl J Med 1993;329:1977-1981. 13. Winawer SJ. Screening of colorectal cancer. Surg Oncol Clin N Am 2005;14:699-722. 14. Rollandi GA, Biscaldi E, DeCicco E. Double contrast barium enema: technique, indications, results and limitations of a conventional imaging methodology in the MDCT virtual endoscopy era. Eur J Radiol 2007;61:382-387. 15. Ferrucci JT. Double contrast barium enema: use in practice and implications for CT colonography. Am J Roentgenol 2006;187:170-173. 16. Pickhardt PJ, Choi JR, Hwang I, et al. Computed tomographic virtual colonoscopy to screen for colorectal neoplasia in asymptomatic adults. N Engl J Med 2003;349:2191-2200. 17. Macari M, Bini EJ, Xue X, et al. colorectal neoplasms: prospective comparison of thin-section low-dose multi–detector row CT colonography and conventional colonoscopy for detection. Radiology 2002;224:383-392. 18. Yee J, Akerkar GA, Hung RK, Steinauer-Gebauer AM, Wall SD, McQuaid KR. Colorectal neoplasia: performance characteristics of CT colonography for detection in 300 patients. Radiology 2001;219:685-692. 19. Dachman AH, Kuniyoshi JK, Boyle CM, et al. CT colonography with three-dimensional problem solving for detection of colonic polyps. Am J Roentgenol 1998;171:989-995
10 Imaging of Colorectal Carcinoma
277
20. Rockey DC, Paulson E, Niedzwiecki D, Analysis of air contrast barium enema, computed tomographic colonography, and colonoscopy: prospective comparison. Lancet 2005;365:305-311. 21. Iannaccone R, Laghi A, Catalano C, et al. Computed tomographic colonography without cathartic preparation for the detection of colorectal polyps. Gastroenterology 2004;127:1300-1311. 22. Lefere P, Gryspeerdt S, Baekelandt M, Van Holsbeeck B. Laxative-free CT colonography. AJR Am J Roentgenol 2004;183:945–948. 23. Callstrom MR, Johnson CD, Fletcher JG, et al. CT colonography without cathartic preparation: feasibility study. Radiology 2001;219:693–698. 24. Winawer SJ, Stewart ET, Zauber AG, et al. A comparison of colonoscopy and double contrast barium enema for surveillance after polypectomy. National Polyp Study Work Group. N Engl J Med 2000;342:1766-1772. 25. Fletcher R. End of the barium enema? N Engl J Med 2000;342:1772-1773. 26. Rex DK, Rahmani EY, Haseman GT, et al. Relative sensitivity of colonoscopy and barium enema for detection of colorectal cancer in clinical practice. Gastroenterology 1997;112:24-28. 27. Minsky BD. Role of adjuvant therapy in adenocarcinoma of the rectum. Semin Surg Oncol 1999;17:189-198. 28. Swedish Rectal Cancer Trial. Improved survival with preoperative radiotherapy in resectable rectal cancer. N Engl J Med 1997;336:980-987. 29. Kapiteijn E, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001;345:638-646. 30. Balthazar EJ, Megibow AJ, Hulnick D, Naidich DP. Carcinoma of the colon: detection and preoperative staging by CT. Am J Roentgenol 1988;150:301-306. 31. Freeny PC, Marks WM, Ryan JA, Bolen JW. Colorectal carcinoma evaluation with CT: preoperative recurrence. Radiology 1986;158:347-353. 32. Smith NJ, Bees N, Barbachano Y, Norman AR, Swift RI, Brown G. Preoperative computed tomography staging of nonmetastatic colon cancer predicts outcome: implications for clinical trials. Br J Surg 2007;96:1030-1036. 33. Harrison JC, Dean PJ, el-Zeky F, Vander Zwaag R. From Dukes through Jass: pathological prognostic indicators in rectal cancer. Hum Pathol 1994;25:498–505. 34. Willett CG, Badizadegan K, Ancukiewicz M, Shellito PC. Prognostic factors in stage T3N0 rectal cancer: do all patients require postoperative pelvic irradiation and chemotherapy? Dis Colon Rectum 1999;42:167–173. 35. Shepherd NA, Baxter KJ, Love SB. Influence of local peritoneal involvement on pelvic recurrence and prognosis in rectal cancer. J Clin Pathol 1995;48:849–855. 36. Merkel S, Mansmann U, Siassi M, Papadopoulos T, Hohenberger W, Hermanek P. The prognostic inhomogeneity in pT3 rectal carcinomas. Int J Colorectal Dis 2001;16:298–304. 37. Beets-Tan RGH, Beets GL. Rectal Cancer: Review with Emphasis on MRI. Radiology; 232:335-346. 38. Hulsmans FJ, Tio TL, Fockens P, Bosma A, Tytgat GN. Assessment of tumor infiltration depth in rectal cancer with transrectal sonography: caution is necessary. Radiology 1994;190:715-720. 39. Mackay SG, Pager CK, Joseph D, Stewart PJ, Solomon MJ. Assessment of the accuracy of transrectal ultrasonography in anorectal neoplasia. Br J Surg 2003;90:346–350. 40. Bipat S, Glas AS, Slors FJ, Zwinderman AH, Bossuyt PM, Stoker J. Rectal cancer: local staging and assessment of lymph node involvement with endoluminal US, CT, and MRI—a metaanalysis. Radiology 2004;232:773–783. 41. Thoeni RF, Moss AA, Schnyder P, Margulis AR. Detection and staging of primary rectal and rectosigmoid cancer by computed tomography. Radiology 1981;141:135-138. 42. van Waes PF, Koehler PR, Feldberg MA. Management of rectal carcinoma: impact of computed tomography. Am J Roentgenol 1983;140:1137-1142. 43. Shank B, Dershaw DD, Caravelli J, Barth J, Enker W. A prospective study of the accuracy of preoperative computed tomographic staging of patients with biopsy-proven rectal carcinoma. Dis Colon Rectum 1990;33:285-290.
278
J.A. Soto
44. Zerhouni EA, Rutter C, Hamilton SR, et al. CT and MRI in the staging of colorectal carcinoma: report of the Radiology Diagnostic Oncology Group II. Radiology 1996;200:443-451. 45. Sinha R, Verma R, Rajesh A, Richards CJ. Diagnostic value of multi-detector row CT in rectal cancer staging: comparison of multiplanar and axial images with histopathology. Clin Radiol 2006;61:924-931. 46. Taylor A, Slater A, Mapstone N, Taylor S, Halligan S. Staging rectal cancer: MRI compared to MDCT. Abdom Imaging. 2006 Sep 12; [Epub ahead of print]. 47. Laghi A, Ferri M, Catalano C, et al. Local staging of rectal cancer with MRI using a phased array body coil. Abdom Imaging 2002;27:425-431. 48. Beets-Tan RG, Beets GL, Borstlap AC, et al. Preoperative assessment of local tumor extent in advanced rectal cancer: CT or high resolution MRI? Abdom Imaging 2000;25:533-541. 49. Blomqvist L, Rubio C, Holm T, Machado M, Hindmarsh T. Rectal adenocarcinoma: assessment of tumour involvement of the lateral resection margin by MRI of resected specimen. Br J Radiol 1999;72:18-23. 50. Brown G, Radcliffe AG, Newcombe RG, Dallimore NS, Bourne MW, Williams GT. Preoperative assessment of prognostic factors in rectal cancer using high resolution magnetic resonance imaging. Br J Surg 2003;90:355-364. 51. Chan TW, Kressel HY, Milestone B, et al. Rectal carcinoma: staging at MRI with endorectal surface coil—work in progress. Radiology 1991;181:461-467. 52. Schnall MD, Furth EE, Rosato EF, Kressel HY. Rectal tumor stage: correlation of endorectal MRI and pathologic findings. Radiology 1994;190:709-714. 53. Zagoria RJ, Schlarb CA, Ott DJ, et al. Assessment of rectal tumor infiltration utilizing endorectal MRI and comparison with endoscopic rectal sonography. J Surg Oncol 1997; 64:312-317. 54. Brown G, Richards CJ, Newcombe RG, et al. Rectal carcinoma: thin-section MRI for staging in 28 patients. Radiology 1999;211:215–222. 55. Blomqvist L, Holm T, Rubio C, Hindmarsh T. Rectal tumours: MRI with endorectal and/or phased-array coils, and histopathological staging on giant sections—a comparative study. Acta Radiol 1997;38:437–444. 56. Brown G, Kirkham A, Williams GT, et al. High-resolution MRI of the anatomy important in total mesorectal excision of the rectum. Am J Roentgenol 2004;182:431–439. 57. MERCURY Study Group. Extramural depth of tumor invasion at thin-section MR in patients with rectal cancer: results of the MERCURY study. Radiology 2007;243:132-139. 58. Chun HK, Choi D, Kim MJ, Lee J, Yun SH, Kim SH, Lee SJ, Kim CK. Preoperative staging of rectal cancer: comparison of 3-T high-field MRI and endorectal sonography. Am J Roentgenol. 2006 Dec;187:1557-1562. 59. Charnsangavej C, Whitley NO. Metastases to the pancreas and peripancreatic lymph nodes from carcinoma of the right side of the colon: CT findings in 12 patients. AJR 1993;160:49-52. 60. Granfield CA, Charnsangavej C, Dubrow RA, et al. Regional lymph node metastases in carcinoma of the colon and rectum: CT demonstration. AJR 1992;159:757-761. 61. Kerner BA, Oliver GC, Eisenstat TE, Rubin RJ, Salvati EP. Is preoperative computerized tomography useful in assessing patients with colorectal carcinoma? Dis Colon Rectum 1993; 36:1050-1053. 62. Abdel-Nabi H, Doerr RJ, Lamonica DM, et al. Staging of colorectal carcinoma with fluorine18 fluorodeoxyglucose whole-body PET: correlation with histopathologic and CT findings. Radiology 1998;206:755-760. 63. Koh DM, Brown G, Temple L, et al. Rectal cancer: mesorectal lymph nodes at MRI with USPIO versus histopathologic findings-initial observations. Radiology 2004;231:91-99. 64. Scheele J, Stang R, Altendorf-Hofmann A, Paul M. Resection of colorectal liver metastases. World J Surg 1995;19:59-71 65. Fusai G, Davidson BR. Strategies to increase the resectability of liver metastases from colorectal cancer. Dig Surg 2003;20:481-496.
10 Imaging of Colorectal Carcinoma
279
66. Gazelle GS, Hunink MG, Kuntz KM, et al. Cost-effectiveness of hepatic metastasectomy in patients with metastatic colorectal carcinoma: a state-transition Monte Carlo decision analysis. Ann Surg 2003;237:544-555. 67. Ohlsson B, Tranberg KG, Lundstedt C, Ekberg H, Hederstrom E. Detection of hepatic metastases in colorectal cancer: a prospective study of laboratory and imaging methods. Eur J Surg 1993;159:275-281. 68. Moertel CG, Fleming TR, Macdonald JS, Haller DG, Laurie JA, Tangen C. An evaluation of the carcinoembryonic antigen (CEA) test for monitoring patients with resected colon cancer. JAMA 1993;270:943-947. 69. Pijl ME, Wasser MN, Joekes EC, van de Velde CJ, Bloem JL. Metastases of colorectal carcinoma: comparison of soft- and hard-copy helical CT interpretation. Radiology. 2003;227:747-751. 70. Kinkel K, Lu Y, Both M, Warren RS, Thoeni RF. Detection of hepatic metastases from cancers of the gastrointestinal tract by using noninvasive imaging methods (US, CT, MRI, PET): a meta-analysis. Radiology 2002;224:748-756. 71. Semelka RC, Shoenut JP, Ascher SM, Kroeker MA, Greenberg HM, Yaffe CS, Micflikier AB. Solitary hepatic metastasis: comparison of dynamic contrast-enhanced CT and MRI with fatsuppressed T2-weighted, breath-hold T1-weighted FLASH, and dynamic gadoliniumenhanced FLASH sequences. J Magn Reson Imaging 1994;4:319-323. 72. Semelka RC, Worawattanakul S, Kelekis NL, John G, Woosley JT, Graham M, Cance WG. Liver lesion detection, characterization, and effect on patient management: comparison of single-phase spiral CT and current MR techniques. J Magn Reson Imaging. 1997;7:1040-1047. 73. Hohmann J, Albrecht T, Oldenburg A, Skrok J, Wolf KJ. Liver metastases in cancer: detection with contrast-enhanced ultrasonography. Abdom Imaging. 2004;29:669-681. 74. Larsen LP, Rosenkilde M, Christensen H, Bang N, Bolvig L, Christiansen T, Laurberg S. The value of contrast enhanced ultrasonography in detection of liver metastases from colorectal cancer: A prospective double-blinded study. Eur J Radiol 2007; 62:302-307. 75. Conlon R, Jacobs M, Dasgupta D, Lodge JP. The value of intraoperative ultrasound during hepatic resection compared with improved preoperative magnetic resonance imaging. Eur J Ultrasound 2003;16:211-216. 76. Hartley JE, Kumar H, Drew PJ, Heer K, Avery GR, Duthie GS, Monson JR. Laparoscopic ultrasound for the detection of hepatic metastases during laparoscopic colorectal cancer surgery. Dis Colon Rectum 2000;43:320-324. 77. Soyer P, Poccard M, Boudiaf M, Abitbol M, Hamzi L, Panis Y, Valleur P, Rymer R. Detection of hypovascular hepatic metastases at triple-phase helical CT: sensitivity of phases and comparison with surgical and histopathologic findings. Radiology 2004;231:413-420. 78. Valls C, Andia E, Sanchez A, Guma A, Figueras J, Torras J, Serrano T. Hepatic metastases from colorectal cancer: preoperative detection and assessment of resectability with helical CT. Radiology 2001;218:55-60. 79. Mainenti PP, Cirillo LC, Camera L, et al. Accuracy of single phase contrast enhanced multidetector CT colonography in the preoperative staging of colo-rectal cancer. Eur J Radiol 2006;60:453-459. 80. Seneterre E, Taourel P, Bouvier Y, Pradel J, Van Beers B, Daures JP, Pringot J, Mathieu D, Bruel JM. Detection of hepatic metastases: ferumoxides-enhanced MRI versus unenhanced MRI and CT during arterial portography. Radiology 1996;200:785-792. 81. Kanematsu M, Hoshi H, Itoh K, Murakami T, Hori M, Kondo H, Yokoyama R, Nakamura H. Focal hepatic lesion detection: comparison of four fat-suppressed T2-weighted MRI pulse sequences. Radiology 1999;211:363-371. 82. Vogl TJ, Schwarz W, Blume S et al. Preoperative evaluation of malignant liver tumors: comparison of unenhanced and SPIO (Resovist)-enhanced MRI with biphasic CTAP and intraoperative US. Eur Radiol 2003 Feb;13:262-272. 83. Ward J, Naik KS, Guthrie JA, Wilson D, Robinson PJ. Hepatic lesion detection: comparison of MRI after the administration of superparamagnetic iron oxide with dual-phase CT by using
280
84.
85. 86.
87.
88.
89.
90.
J.A. Soto alternative-free response receiver operating characteristic analysis. Radiology 1999;210:459-466. Said B, McCart JA, Libutti SK, Choyke PL. Ferumoxide-enhanced MRI in patients with colorectal cancer and rising CEA: surgical correlation in early recurrence. Magn Reson Imaging 2000;18:305-309. Matsuo M, Kanematsu M, Itoh K, et al. Detection of malignant hepatic tumors: comparison of gadolinium-and ferumoxide-enhanced MRI. Am J Roentgenol 2001;177:637-643. Gollub MJ, Akhurst T, Markowitz AJ, et al. Combined CT colonography and 18F-FDG PET of colon polyps: potential technique for selective detection of cancer and precancerous lesions. AJR 2007;188:130-138. Kantorova I, Lipska L, Belohlavek O. Routine (18)F-FDG PET preoperative staging of colorectal cancer: comparison with conventional staging and its impact on treatment decision making. J Nucl Med 2003;44:1784-1788. Abdel-Nabi H, Doerr RJ, Lamonica DM, Cronin VR, Galantowicz PJ, Carbone GM, Spaulding MB. Staging of primary colorectal carcinomas with fluorine-18 fluorodeoxyglucose whole-body PET: correlation with histopathologic and CT findings. Radiology 1998;206:755-760. Sahani DV, Kalva SP, Fischman AJ, Kadaviqere R, Blake M, Hahn PF, Saini S. Detection of liver metastases from adenocarcinoma of the colon and pancreas: comparison of mangafodipir trisodium-enhanced liver MRI and whole-body FDG PET. Am J Roengenol 2005;185: 239-246. Benoit S, Brouquet A, Penna C, et al. Complete response of colorectal liver metastases after chemotherapy: does it mean cure? J Clin Oncol 2006;24:3939-3945.
11
Imaging of Primary Malignant Tumors of Peritoneal and Retroperitoneal Origin J. Louis Hinshaw, MD and Perry J. Pickhardt, MD
Key Points ●
●
●
Primary malignancies arising from the peritoneal, subperitoneal and retroperitoneal spaces occur much less frequently than metastatic involvement from primary organ-based tumors or lymphoproliferative diseases. Nonetheless, these rare primary lesions should be considered in the absence of a known or suspected organ-based malignancy. Cross-sectional imaging can be useful for detection, characterization, staging, guiding biopsy for tissue diagnosis and evaluating response to therapy.
Abstract Peritoneal carcinomatosis and metastatic involvement of the retroperitoneum are relatively common manifestations of many organ-based malignancies and lymphoproliferative disorders. Primary malignancies of peritoneal and retroperitoneal origin occur much less frequently, and can be difficult to distinguish from metastatic disease. In many cases, a precise diagnosis based on imaging findings alone is not possible. However, the imaging features of these primary tumors, in combination with the clinical and demographic data, can be utilized to narrow the scope of the differential diagnosis. This chapter will present the clinical and imaging features of primary peritoneal and retroperitoneal tumors arising from the various tissue components that comprise the ligaments, mesenteries and connective tissues of these anatomic spaces.
1
Introduction
Peritoneal carcinomatosis is a relatively common manifestation of many organ-based malignancies, particularly of the GI tract and ovaries. Likewise, metastatic retroperitoneal lymphadenopathy and direct extension from an organ-based primary From the Department of Radiology, University of Wisconsin Medical School, Madison, WI Corresponding author: Perry J. Pickhardt, MD, Department of Radiology, University of Wisconsin Hospital and Clinics, E3/311 Clinical Science Center, 600 Highland Avenue, Madison, WI 53792, e-mail:
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
281
282
J.L. Hinshaw and P.J. Pickhardt
tumor are also common findings at imaging evaluation. Primary tumors of peritoneal and retroperitoneal origin occur much less frequently, but are often first identified on cross-sectional radiologic imaging studies, such as computed tomography (CT), ultrasound (US) or magnetic resonance imaging (MRI). Neoplastic involvement of the peritoneum and retroperitoneum generally manifest with an abnormal increase in soft tissue, which can appear infiltrative or tumerous and be associated with variable amounts of cystic change, calcification, fatty composition, intravenous contrast enhancement and surrounding fluid. However, since many non-neoplastic and metastatic processes demonstrate similar imaging findings, the appearance of many primary malignancies of the peritoneum and retroperitoneum is nonspecific [1]. As a result, even in the absence of a known organ-based primary malignancy, metastatic disease is often the first consideration when confronted with an abnormal soft tissue process arising within the peritoneal or retroperitoneal space. However, primary malignancies should also be considered in this setting. This chapter will present the salient clinical and imaging features of the majority of the primary neoplasms (Table 11.1) arising from the various tissue components that comprise the ligaments, mesenteries and connective tissues of the peritoneal and retroperitoneal spaces. The differential diagnosis for peritoneal-based and retroperitoneal-based neoplasms can often be refined by combining the imaging features with the patient’s relevant clinical and demographic information. In addition to detection and characterization, crosssectional imaging is useful for directing biopsy for tissue diagnosis.
2
Anatomic Considerations
The visceral and parietal peritoneum enclose the large potential space referred to as the peritoneal cavity. Pathologic processes that gain access to the peritoneal cavity can disseminate throughout this space via the relatively unrestricted movement of fluid and cells. Pathologic processes can also be disseminated within the subperitoneal space, which lies deep to the surface lining of the visceral and parietal peritoneum, omentum Table 11.1 Primary Peritoneal and Retroperitoneal Malignancies Primary Peritoneal Malignancies Mesothelioma Papillary serous carcinoma Desmoplastic small round cell tumor Malignant fibrous histiocytoma Liposarcoma Other mesenchymal tumors Primary Retroperitoneal Malignancies Liposarcoma Leiomyosarcoma Malignant fibrous histiocytoma Other mesenchymal tumors Paraganglioma Extragonadal germ cell tumors
11 Imaging of Primary Malignant Tumors of Peritoneal and Retroperitoneal Origin
283
and the various peritoneal ligaments and mesenteries [2]. The subperitoneal space has both intraperitoneal and extraperitoneal components that bridge the peritoneum and retroperitoneum, which can result in bi-directional spread of disease processes. This concept helps to explain the involvement of both the peritoneal and retroperitoneal space that is sometimes encountered. The retroperitoneal space is not defined by specific anatomic structures delineating its borders, but rather as the space posterior to the peritoneal cavity. Retroperitoneal structures may be defined as primary (e.g., retroperitoneal from the beginning of embryogenesis) or secondary (e.g., an area initially suspended by a mesentery during early embryogenesis that subsequently migrated and fused to become retroperitonealized). The extraperitoneal pelvis essentially represents the inferior continuation of the retroperitoneal space. The retroperitoneum and extraperitoneal pelvis represent a crossroads for a number of organ systems containing portions of the gastrointestinal and genitourinary tracts, as well as major vascular structures. The retroperitoneum, however, also contains intrinsic connective tissues, fat and neural elements. This chapter will focus on primary malignancies arising directly from the supporting tissues of the peritoneal, subperitoneal and retroperitoneal spaces, rather than tumors that arise from the organs contained within these spaces.
3 3.1
Primary Peritoneal Malignancies Mesothelioma
Clinical Features. Mesothelial cells line the internal body cavities, including the pleura, peritoneum, pericardium and paratesticular space. Mesothelioma is a rare tumor which arises from these cells and most frequently involves the pleural space. However, approximately 30 percent arise solely from the peritoneum [3]. There are benign, borderline and malignant variants, but benign cystic mesothelioma is not related to malignant mesothelioma. Compared with the pleural form, malignant mesothelioma of the peritoneum is less often associated with asbestos exposure [3, 4]. However, cases with both pleural and peritoneal involvement are usually asbestosrelated. In general, malignant mesothelioma of the peritoneum is an aggressive tumor with a rapidly progressive clinical course and a universally poor prognosis. Imaging Features. The imaging features of peritoneal mesothelioma are variable [4]. The “dry” appearance consists of single or multiple peritoneal-based soft tissue masses that may be large or confluent (Fig. 11.1). The “wet” appearance consists of peritoneal thickening that may be nodular and/or diffuse and is associated with peritoneal fluid (ascites) (Fig. 11.1). Scalloping and mass effect upon adjacent abdominal organs can be seen. Calcification, either within the mass or associated with peritoneal plaques, is uncommon and one should consider other etiologies in the setting of extensive calcification in a peritoneal-based tumor.
284
J.L. Hinshaw and P.J. Pickhardt
Fig. 11.1 Malignant peritoneal mesothelioma. (a) Contrast-enhanced CT image shows a large, confluent peritoneal-based mass with heterogeneous attenuation, but no calcification. There is no associated ascites present, reflecting the “dry form” of peritoneal mesothelioma. (b) Contrastenhanced CT image from a different patient shows ascites with diffuse thickening of both the visceral and parietal peritoneum, as well as omental and mesenteric soft tissue infiltration. This appearance reflects the so-called “wet form”
4
Papillary Serous Carcinoma
Clinical Features. Primary papillary serous carcinoma of the peritoneum is a rare malignancy that predominately affects post-menopausal women [5]. This tumor is histologically identical to serous ovarian papillary carcinoma and is clearly distinguishable when the ovaries are either not involved or only superficially involved [6]. Treatment generally consists of an abdominal hysterectomy, bilateral salpingooophorectomy and debulking surgery, which are followed by combination chemotherapy. Despite these interventions, the prognosis is dismal. Imaging Features. Cross-sectional imaging often shows extensive, multifocal involvement of the peritoneum, with omental caking, ascites and, importantly, no primary ovarian mass (Fig. 11.2). There is often extensive calcification within the omental masses, which can be a useful CT finding for differentiating this tumor from peritoneal mesothelioma [6].
5
Desmoplastic Small Round Cell Tumor
Clinical Features. Desmoplastic small round cell tumor is a rare, highly aggressive malignancy that was first described relatively recently [7]. It generally behaves like a high-grade soft tissue sarcoma, but has a predilection for primary peritoneal involvement. The disease tends to progress rapidly and metastatic disease to the liver, lungs, lymph nodes and bones are often present at diagnosis. Unlike most
11 Imaging of Primary Malignant Tumors of Peritoneal and Retroperitoneal Origin
285
Fig. 11.2 Primary peritoneal papillary serous carcinoma. (a) Contrast-enhanced CT image from a middle-aged female shows a heterogeneous infiltrative soft tissue mass involving the omentum; the ovaries were normal in appearance (not shown). (b) Diagnosis was confirmed by ultrasoundguided core biopsy of the thickened omentum (arrowheads)
other primary peritoneal neoplasms discussed herein, this tumor most often affects adolescents and young adults, particularly males. Treatment is relatively ineffective, but attempted therapy often includes surgical debulking, chemotherapy and radiation therapy. Imaging Features. The most common imaging appearance is that of multiple, bulky rounded peritoneal-based masses [7] (Fig. 11.3). There can be associated ascites, and heterogeneous enhancement of the masses with areas of central necrosis is common. The omentum and paravesical regions are often involved. Although the lesions are usually discrete, an infiltrative appearance is sometimes seen. Calcifications and lymphadenopathy are not usually present.
6
Malignant Fibrous Histiocytoma
Clinical Features. Primary sarcomas of the peritoneal/subperitoneal space, such as malignant fibrous histiocytoma (MFH) and liposarcoma, occur less frequently than their retroperitoneal counterparts [4, 8]. These tumors are most frequently seen in adults and are typically quite large at diagnosis. MFH accounts for approximately 20 percent of all soft tissue sarcomas, most commonly arising in the extremities and retroperitoneum [9]. However, MFH is also reported by some sources to be the single most common primary peritoneal sarcoma [8]. It occurs more frequently in males and has a peak incidence in the fifth and sixth decades of life. The mass is often clinically silent until it is quite large. Constitutional symptoms such as fever,
286
J.L. Hinshaw and P.J. Pickhardt
Fig. 11.3 Desmoplastic small round cell tumor. (a and b) Contrast-enhanced CT images show multiple, rounded, enhancing peritoneal-based masses, which is the most common appearance of this disease. (c) Coronal image from FDG-PET study shows multiple hypermetabolic foci corresponding to the rounded peritoneal-based masses seen on CT
malaise and weight loss can occur, but are nonspecific. The only treatment, if possible, is complete surgical resection. Metastatic disease most often involves the lungs, bone and liver. Prognosis is related to tumor grade, size and the presence or absence of metastatic disease. Specifically, high-grade tumors and tumors larger than 10 cm in size have a poor outcome with 10-year survival rates of less than 50 percent [9]. Imaging Features. Radiographically, MFH typically manifests as a large heterogeneous soft tissue mass (Fig. 11.4), as do most sarcomas. Biopsy is required to make a specific diagnosis. The mass is frequently lobulated with peripheral nodular
11 Imaging of Primary Malignant Tumors of Peritoneal and Retroperitoneal Origin
287
Fig. 11.4 Peritoneal malignant fibrous histiocytoma (MFH). Contrast-enhanced CT image shows a large lobulated heterogeneous peritoneal-based mass occupying the left subphrenic space. Ascites is also present
enhancement, can have associated calcifications (in approximately 10 percent), and may demonstrate heterogeneity from central necrosis, hemorrhage or myxoid degeneration. Fatty components are not seen in MFH [10]. The tumor may directly invade the abdominal musculature, but vascular invasion is rare.
7
Liposarcoma
Clinical Features. Fat-containing tumors are very common in general and account for approximately half of all soft tissue tumors in most surgical series [11, 12]. However, the vast majority of these represent benign lipomas, and differentiating these tumors from liposarcoma is not a trivial matter. Although liposarcoma is one of the most common primary retroperitoneal malignancies, primary peritoneal liposarcoma is relatively rare [13]. The clinical presentation is usually delayed due to the lack of associated symptoms. Ultimately, the mass may become palpable, create symptoms related to mass effect on adjacent structures, or may be incidentally identified at the time of imaging. Treatment is surgical resection, with or without chemotherapy and radiotherapy. Prognosis is inversely related to cellular differentiation of the tumor and directly related to completeness of resection.
288
J.L. Hinshaw and P.J. Pickhardt
Fig. 11.5 Peritoneal liposarcoma. Contrast-enhanced CT image shows a heterogeneous fatty mass with enhancing soft tissue elements. Note that it is displacing the adjacent small bowel
Imaging Features. Fat-containing tumors are readily and confidently recognized on CT and MRI when demonstrable macroscopic fat is present, which significantly limits the differential diagnosis. If the mass is homogeneous, well-defined and consists almost entirely of fat with only minimal if any soft tissue component, the diagnosis of a benign lipoma is almost certain. Liposarcomas are typically less well-defined, have indistinct borders and contain variable but increased amounts of soft tissue [14] (Fig. 11.5). In fact, some poorly differentiated liposarcomas have no demonstrable fat on cross-sectional imaging and are, therefore, indistinguishable from other sarcomas.
8
Other Malignant Mesenchymal Tumors
Clinical Features. The remaining malignant mesenchymal tumors beyond MFH and liposarcoma essentially lack any distinguishing clinical or radiographic features. As a result, tissue biopsy or surgical resection is required for definitive diagnosis. Malignant nerve sheath tumors and gastrointestinal stromal tumors (GIST) in the setting of neurofibromatosis type 1 (NF-1), however, can be an exception since the patient will often have clinical stigmata of NF-1 (e.g., café au lait spots and cutaneous neurofibromas), or will already carry the diagnosis of NF-1.
11 Imaging of Primary Malignant Tumors of Peritoneal and Retroperitoneal Origin
289
Peritoneal involvement by leiomyosarcoma and malignant GIST are most frequently due to metastatic spread from a primary gastrointestinal site, but primary peritoneal tumors can and do occur [15]. In the past malignant GISTs were incorrectly classified as leiomyosarcomas (see section on retroperitoneal leiomyosarcomas). Fibrosarcoma of the mesentery and omentum in young patients can be difficult to differentiate from inflammatory pseudotumor, both at imaging and at pathologic evaluation [16]. Angiosarcoma can develop from the vascular elements of the subperitoneal space. Even synovial sarcomas can arise within the peritoneum, and these tumors can have associated dystrophic calcifications [17]. Imaging Features. Malignant nerve sheath tumors are often multifocal and have a branching or coalescent appearance. They are frequently of low attenuation on CT and have high signal on T2-weighted MR images. As a result they are sometimes mistaken for cystic lesions. Frequently, there are associated nerve root lesions or other findings of NF-1. GIST should also be considered for peritoneal or retroperitoneal tumors in the setting of NF-1 (Fig. 11.6). The remaining sarcomas are usually indistinguishable from each other on crosssectional imaging, usually presenting as large soft tissue masses. Synovial sarcomas may have associated dystrophic calcifications (Fig. 11.7), and angiosarcomas are typically hypervascular, but these features are not always present and significant overlap in imaging features exists.
Fig. 11.6 Peritoneal GIST in the setting of NF-1. Contrast-enhanced CT image shows large heterogeneously enhancing mesenteric soft tissue mass with areas of necrosis or cystic change. Note the mesenchymal dysplasia involving the lumbar spine with associated lateral meningocele, as well as numerous cutaneous neurofibromas. This combination of findings is essentially diagnostic for NF-1. The two most likely considerations for the complication tumor would be malignant nerve sheath tumor and GIST. This lesion proved to be a large mesenteric GIST
290
J.L. Hinshaw and P.J. Pickhardt
Fig. 11.7 Primary peritoneal synovial sarcoma. Contrast-enhanced CT image shows a mixed cystic and solid mass arising within the gastrocolic ligament. Dystrophic calcification is present within the anterior soft tissue component of the mass (arrowhead). Primary peritoneal origin is a rare extra-articular location for this tumor
9 9.1
Primary Retroperitoneal Neoplasms Liposarcoma
Clinical Features. As a group, sarcomas are the most common primary malignancies of the retroperitoneum. The three most common cell types include liposarcoma, leiomyosarcoma and malignant fibrous histiocytoma (MFH). Beyond liposarcoma where the presence of fat usually provides a specific clue [18], most of these malignant mesenchymal tumors arising within the retroperitoneum are difficult to differentiate on imaging or clinical grounds. As with peritoneal fat-containing tumors, differentiating benign lipomas from liposarcomas can sometimes be extremely difficult. Liposarcoma is one of the most common primary retroperitoneal malignancies [19]. The mass is often large at the time of diagnosis due to the lack of associated clinical manifestations. It is treated by surgical resection and the decision to administer additional chemotherapy and radiotherapy is made on a case-by-case basis. Prognosis is related to the grade of the tumor and completeness of resection.
11 Imaging of Primary Malignant Tumors of Peritoneal and Retroperitoneal Origin
291
Fig. 11.8 Retroperitoneal liposarcoma. (a and b) Contrast-enhanced transverse and coronal CT images show a large retroperitoneal mass that contains both fatty and soft tissue components. Note that the tumor displaces the right kidney anteromedially, which confirms its retroperitoneal origin
Imaging Features. A well-defined, homogeneous fatty mass is likely to represent a benign lipoma. However, characteristics that are associated with a higher risk of liposarcoma include a lesion size > 10 cm, thick septations, globular and/or nodular nonadipose regions and a relative proportion of fat < 75 percent [14] (Fig. 11.8). Note that thin septations are seen in both benign and malignant lesions and are not predictive. There is also significant overlap in both the imaging findings and histologic findings of lipomas and well-differentiated liposarcomas.
10
Leiomyosarcoma
Clinical Features. The majority of retroperitoneal leiomyosarcomas occur in women, usually in the fifth or sixth decade of life. The retroperitoneum represents the most common primary site of origin, followed by the uterus [20]. A significant fraction of these tumors arise from the inferior vena cava [21]. Clinical presentation largely relates to whether or not an intravascular component is present; most cases are large heterogeneous tumors demonstrating an extraluminal growth pattern. Tumors with an intravascular component may present with symptoms relating to
292
J.L. Hinshaw and P.J. Pickhardt
venous compromise or thrombosis. The lungs are the most frequent site of metastatic involvement [22]. Treatment is difficult because surgical resection is often limited by the size and extent of the mass, while adjuvant chemotherapy and radiation therapy are relatively ineffective. Until recently many GISTs were incorrectly classified as smooth muscle tumors (leiomyomas and leiomyosarcomas), but recent advances in immunohistochemistry and electron microscopy have shown that these tumors are indeed unique [23]. In comparison, true retroperitoneal GISTs are extremely rare and although primary peritoneal origin is more common, it is still quite rare compared with a primary gastrointestinal tract origin. Imaging Features. Three growth patterns can be seen at imaging: extravascular (most common), completely intravascular (least common) and combined extra- and intravascular [21] (Fig. 11.9). US and angiography (by CT, MRI or conventional means) may be useful in cases with an intravascular component (Fig. 11.9). Although involvement of the inferior vena cava is suggestive of leiomyosarcoma, other sarcomas can secondarily invade this structure, reducing the specificity of this finding somewhat.
Fig. 11.9 Retroperitoneal leiomyosarcoma. (a) Contrast-enhanced CT image shows an enhancing retroperitoneal mass, which appears to arise from the adjacent inferior vena cava (IVC, arrowhead). (b) Image from direct venography from a different patient shows obstruction of the IVC by an intravascular leiomyosarcoma, which gives rise to a large filling defect (arrowhead). There is associated collateralization into the azygos system. Although this appearance is suggestive of leiomyosarcoma, other malignancies can secondarily invade the IVC
11 Imaging of Primary Malignant Tumors of Peritoneal and Retroperitoneal Origin
11
293
Malignant Fibrous Histiocytoma
Clinical Features. MFH contains both fibroblastic and histiocytic cells in various proportions. Pleomorphic is the most common histiologic subtype, but there are also myxoid, giant cell, inflammatory and angiomatoid subtypes [9]. MFH generally presents in the fifth and sixth decades and the most common symptoms are fever, malaise and weight loss. Metastatic disease most frequently involves the lungs, but osseous and hepatic metastases are sometimes seen. Treatment is surgical resection and although the risk of local recurrence is directly related to completeness of resection, the overall prognosis is more closely related to tumor grade (low, intermediate or high), tumor size and the presence of metastases [9]. Imaging Features. MFH usually manifests as a large heterogeneous soft tissue mass on cross-sectional imaging. The tumor heterogeneity is related to a variable combination of central necrosis, hemorrhage and myxoid degeneration. Enhancement is variable, but often nodular and peripheral. Calcifications are seen in approximately 10 percent of patients, but no fatty component or vascular invasion should be present [10]. MFH is often locally aggressive and invades adjacent structures.
12
Other Malignant Mesenchymal Tumors
Clinical Features. Similar to peritoneal sarcomas, other retroperitoneal sarcomas generally lack distinguishing features. However, there are some characteristics that may be helpful in differentiating the various subtypes. As previously discussed, malignant nerve sheath tumors and GISTs are associated with NF-1, although primary retroperitoneal GISTs are rare. Fibrosarcoma is another rare retroperitoneal tumor, which can have variable biologic behavior. Inflammatory fibrosarcoma in children can be difficult to distinguish from benign myofibroblastic tumor (inflammatory pseudotumor); this malignancy can be locally aggressive and has the potential for metastasis [16]. Angiosarcoma, rhabdomyosarcoma and hemangiopericytoma are rare, aggressive neoplasms that have an extremely poor prognosis. Imaging Features. Most of these sarcomas manifest as large, heterogeneous, locally invasive masses and the imaging findings are generally nonspecific. The role of imaging is more to evaluate for the extent of disease, to guide biopsy for tissue diagnosis and to assess response to therapy. Malignant nerve sheath tumors in the setting of NF-1 often have a branching, plexiform morphology, with low attenuation on CT and high signal on T2-weighted MRI that can mimic a cystic appearance [24].
13
Malignant Paraganglioma
Clinical Features. Paragangliomas arise from neuroendocrine cells derived from the embryologic neural crest. These tumors can occur anywhere along the sympathetic chain, including both adrenal (e.g., pheochromocytoma) and extra-adrenal
294
J.L. Hinshaw and P.J. Pickhardt
(e.g., paraganglioma) origin [25, 26]. Paragangliomas can be hormonally active, secreting catecholamines which can result in labile hypertension, palpitations, sweating and headaches. Most paragangliomas are benign, but up to 10 percent metastasize and display malignant behavior. They are most likely to occur between the ages of 30 and 45. Imaging Features. Other than a characteristic location adjacent to the aorta (including the organ of Zuckerkandl) (Fig. 11.10), there are no imaging-specific features for extra-adrenal paragangliomas. These tumors are generally hypervascular, which can be a suggestive imaging feature. The tumor can be solitary or multifocal, and
Fig. 11.10 Malignant retroperitoneal paraganglioma. (a) Contrast-enhanced CT image shows a large hypervascular retroperitoneal mass, which is relatively homogeneous in appearance considering its large size. (b) Contrast-enhanced CT image from a second patient shows a large paraganglioma with a predominately cystic appearance, likely from central necrosis. Adjacent low-attenuation retroperitoneal lymphadenopathy is present. (c) Sagittal T2-weighted MRI from a pregnant female with hypertension and palpitations shows a large mass with central high signal from cystic change arising from the organ of Zuckerkandl. Note the gravid uterus inferior and adjacent to the mass
11 Imaging of Primary Malignant Tumors of Peritoneal and Retroperitoneal Origin
295
is often relatively homogeneous in appearance, although malignant lesions tend to be larger and demonstrate areas of central necrosis [27]. Associated calcifications are present in about 15 percent of cases. Local invasion or distant metastases are diagnostic of malignancy.
14
Extragonadal Germ Cell Tumors
Clinical Features. Extragonadal germ cell tumors (EGGCT) represent 5 percent to 10 percent of all germ cell tumors, and are characterized by a midline location extending from the pineal gland to the coccyx [28, 29]. Approximately 20 percent to 40 percent of EGGCTs are seminomas, with nonseminomatous germ cell tumors (e.g., embryonal carcinoma, yolk-sac tumor, choriocarcinoma, teratoma or combined) representing the remaining 60 percent to 80 percent. The majority of these tumors occur in the mediasteinum, but the second most common site of involvement is the retroperitoneum (30 percent to 40 percent). Because metastatic retroperitoneal involvement from a testicular primary germ cell tumor is much more common than a primary EGGCT, males should undergo testicular US to exclude this possibility [28]. The most common clinical symptoms from retroperitoneal EGGCT include a palpable abdominal mass, abdominal or back pain and weight loss. Treatment generally includes primary chemotherapy, followed by surgical resection of any significant residual mass. Although controversial, any residual mass measuring > 3 cm is usually resected and, if residual disease is identified in the pathologic specimen, then further chemotherapy is given [28]. Although generally treated like a metastatic gonadal germ cell tumor, the prognosis of primary retroperitoneal EGGCT is somewhat worse, but still rather favorable overall. Negative prognostic factors include nonseminomatous histology, elevated tumor markers at the time of diagnosis and the presence of metastatic disease. Imaging Features. The cross-sectional imaging appearance of these tumors varies according to the underlying tissue type. Teratomas often have a markedly heterogeneous appearance due to varying combinations of soft tissue, calcification, fat and fluid. In contrast, seminomas tend to be large, homogeneous lobulated soft tissue masses. Nonseminomatous germ cell tumors are often very irregular in morphology and heterogeneous in appearance, with variable amounts of necrosis and hemorrhage (Fig. 11.11).
Conclusion Primary malignancies of the peritoneal, subperitoneal and retroperitoneal spaces occur much less frequently than metastatic involvement from primary organ-based tumors or lymphoproliferative diseases. Cross-sectional imaging techniques, such as CT and PET/CT, can be useful for detection, characterization, staging, guiding biopsy for tissue diagnosis and evaluating response to therapy.
296
J.L. Hinshaw and P.J. Pickhardt
Fig. 11.11 Extragonadal primary retroperitoneal germ cell tumor. (a and b) T2-weighted and contrast-enhanced T1-weighted MR images show a complex T2 hyperintense retroperitoneal mass that causes left-sided obstructive hydronephrosis. At biopsy, this proved to be a malignant mixed mullerian tumor (MMMT) arising within the retroperitoneum
References 1. Pickhardt PJ, Bhalla S. Unusual non-neoplastic peritoneal and subperitoneal conditions: CT findings. Radiographics 2005;25:719-730 2. Pickhardt PJ. Peritoneum and retroperitoneum. In: Body CT: a practical approach. Slone RM, Fisher AJ, Pickhardt PJ, Gutierrez, FR, Balfe DM, eds. New York: McGraw-Hill, 2000:159-177 3. Busch JM, Kruskal JB, Wu B. Malignant peritoneal mesothelioma. Radiographics 2002;22:1511-1515 4. Pickhardt PJ, Bhalla S. Primary neoplasms of peritoneal and sub-peritoneal origin: CT Findings. Radiographics 2005;25:983-995 5. Altaras MM, Aviram R, Cohen I, Cordoba M, Weiss E, Beyth Y. Primary peritoneal papillary serous adenocarcinoma: clinical management and aspects. Gynecol Oncol 1991;40:230-236 6. Stafford-Johnson DB, Bree RL, Francis IR, Korobkin M. CT appearance of primary papillary serous carcinoma of the peritoneum. AJR 1998;171:687-689 7. Pickhardt PJ, Fisher AJ, Balfe DM, Dehner LP, Huettner PC. Desmoplastic small round cell tumor of the abdomen: radiologic-histopathologic correlation. Radiology 1999;210:633-638 8. Bodner K, Bodner-Adler B, Mayerhofer S, et al. Malignant fibrous histiocytoma (MFH) of the mesentery: a case report. Anticancer Res 2002;22:1169-1170 9. Yip D, Stacy GS. Malignant Fibrous Histiocytoma, Soft Tissue. Emedicine. http://www. emedicine.com/Radio/topic420.htm. Accessed 11/17/06 10. Ros PR, Viamonte M Jr, Rywlin AM. Malignant fibrous histiocytoma: mesenchymal tumor of ubiquitous origin. AJR 1984;142:753-759 11. Rydholm A, Berg NO. Size, site and clinical incidence of lipoma: factors in the differential diagnosis of lipoma and sarcoma. Acta Orthop Scand 1983;54:929-934 12. Myhre-Jensen O. A consecutive seven-year series of 1331 benign soft tissue tumors: clinicopathologic data-comparison with sarcomas. Acta Orthop Scand 1981;52:287-293 13. Kim T, Murakami T, Oi H, et al. CT and MR imaging of abdominal liposarcoma. AJR 1996;166:829-833 14. Kransdorf MJ, Bancroft LW, Peterson JJ, Murphey MD, Foster WC, Temple HT. Imaging of fatty tumors: distinction of lipoma and well-differentiated liposarcoma. Radiology 2002;224:99-112
11 Imaging of Primary Malignant Tumors of Peritoneal and Retroperitoneal Origin
297
15. Kim HC, Lee JM, Kim SH, et al. Primary gastrointestinal stromal tumors in the omentum and mesentery: CT findings and pathologic correlations. AJR 2002;182:1463-1467 16. Meis JM, Enzinger FM. Inflammatory fibrosarcoma of the mesentery and retroperitoneum: a tumor closely simulating inflammatory pseudotumor. Am J Surg Pathol 1991;15:1146-1156 17. Ko SF, Chou FF, Huang CH, et al. Primary synovial sarcoma of the gastrocolic ligament. Br J Radiol 1998;71:438-440 18. Granstrom P, Unger E. MR imaging of the retroperitoneum. Magn Reson Imaging Clin N Am 1995;3:121-142 19. Engelken JD, Ros PR. Retroperitoneal MR Imaging. Magn Reson Imaging Clin N Am 1997;5:165-178 20. Clary BM, DeMatteo RP, Lewis JJ, Leung D, Brennan MF. Gastrointestinal stromal tumors and leiomyosarcomas of the abdomen and retroperitoneum: a clinical comparison. Ann Surg 2001;8:290-299 21. Hartman DS, Hayes WS, Choyke PL, Tibbetts GP. Leiomyosarcoma of the retroperitoneum and inferior vena cava: radiologic-pathologic correlation. Radiographics 1992;12:1203-1220 22. Sondak V, Economou J, Eilber F. Soft tissue sarcomas of the extremity and retroperitoneum: advances in management. Adv Surg 1991;24P:333-359 23. Erlandson R, Klimstra D, Woodruff J. Sub-classification of gastrointestinal stromal tumors based on evaluation by electron microscopy and immunohistochemistry. Ultrastruct Pathol 1996;20:373-393 24. Matsuki K, Kakitsubata Y, Watanabe K, Tsukino H, Nakajima K. Mesenteric plexiform neurofibroma associated with Recklinghausen’s disease. Pediatr Radiol 1997;27:255-256 25. Melicow MM. One hundred cases of pheochromocytoma (107 tumors) at the ColumbiaPresbyterian Medical Center, 1926-1976: a clinicopathological analysis. Cancer 1977;40:1987-2004 26. Glenn F, Gray GF. Functional tumors of the organ of Zuckerkandl. Ann Surg 1976;183:578-586 27. Hayes WS, Davidson AJ, Grimley PM, Hartman DS. Extraadrenal retroperitoneal paraganglioma: clinical, pathologic, and CT findings. AJR 1990;155:1247-1250 28. Makhoul I, Curti B. Extragonadal germ cell tumors. Emedicine. http://www.emedicine.com/ Radio/topic759.htm. Accessed 11/14/06 29. Nichols CR, Fox EP. Extragonadal and pediatric germ cell tumors. Hematol Oncol Clin North Am 1991;5:1189-1209
12
Imaging of Urinary Tract Tumors Michael A. Blake, FFR (RCSI), FRCR, MRCPI and Mannudeep K. Kalra, MD, DNB
1
Key Points
1.1 ●
●
●
●
CT is generally considered to be the most robust and comprehensive imaging modality for evaluation of renal tumors. When a proper technique is used, CT provides high accuracy in the diagnosis and staging of a renal mass. CT also provides useful diagnostic information for treatment planning and follow-up. MRI and ultrasound have certain advantages, but generally function clinically as valuable problem-solving tools.
1.2 ●
● ●
●
●
Imaging of Renal Tumors
Imaging of Bladder Malignancies
Imaging now has an important role in the evaluation of patients with invasive bladder cancer. Imaging is useful in staging of bladder cancer despite having some limitations. CT, as well as traditional MRI, rely on morphological criteria and are both useful in the detection of metastases to the lymph nodes, liver and bone MRI is more accurate than CT in determining the depth of bladder wall invasion and for staging bladder cancer. MRI is also considered superior in follow-up of patients with bladder cancer post-therapy as it can distinguish biopsy changes more accurately than CT.
Department of Radiology, White 270, Massachusetts General Hospital, 55 Fruit St, Boston MA 02114 Corresponding author: Michael A. Blake e-mail:
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
299
300
2
M.A. Blake and M.K. Kalra
Introduction
Advances in imaging have given radiology an increasingly significant role in the diagnosis, staging and re-staging of patients with urinary tract tumors. In this chapter we emphasize the value of current imaging and briefly discuss the potential applications of novel imaging techniques in the management of patients with urinary tract tumors. We will focus primarily on renal cell tumors and bladder transitional cell cancers. Multi-detector-row CT (MDCT) offers greater speed, improved spatial resolution and wider coverage, and allows screening of multiple organs in the abdomen for metastatic disease and for complications. MRI, with its superior soft tissue resolution, offers advantages in imaging some urinary tumors. Ultrasound can also play a complementary role to CT or MRI in imaging of urinary tumors. We will limit discussion of nuclear medicine techniques, but will mention the emerging role of FDG-PET and Positron Emission Tomography/CT (PET/CT) for the detection of distant metastases of several GU malignancies.
2.1
Kidney Cancer
Kidney cancers account for about 3 percent of all cancer cases, as well as about 3 percent of all cancer deaths in the United States [1]. Newer imaging techniques are detecting renal tumors more frequently and at lower disease stages, when tumors can be resected for cure [2, 3]. Most renal tumors arise from the renal parenchyma (renal cell tumors), with smaller numbers arising from the mesenchyma or the urothelium of the renal collecting system. This chapter will focus on the renal cell tumors. Renal cell tumors of different subtypes are associated with distinctively different disease progression and metastatic potential [4, 5].
3
Detection and Diagnosis
Diagnostic imaging of renal masses has undergone dramatic change over the last 20 years. CT, MRI and ultrasonography have essentially replaced traditional diagnostic imaging tests, such as intravenous urography and angiography. CT is now considered the modality of choice for detection and diagnosis of renal cortical tumors, with MRI and ultrasound often used for problem-solving or in patients with contraindications to iodinated contrast. Again, the advent of multi-detector CT has led to faster acquisition times, improved spatial resolution and greater numbers of CT examinations being performed; this in turn has led to a significant increase in the early diagnosis of renal tumors [2, 3]. Furthermore, renal cell tumors are also being detected incidentally in patients undergoing CT scanning for non-renal clinical indications.
12 Imaging of Urinary Tract Tumors
4
301
Role of CT
A dedicated imaging protocol with thin slice thickness (e.g., 2.5 mm) through the kidneys is needed for optimal evaluation of a renal mass by CT. Pre-contrast images are helpful to assess for calcifications and provide a baseline attenuation value for evaluating subsequent post-contrast enhancement in cystic or solid renal masses [6]. Cortico-medullary images (scan delay 70 to 85 seconds after injection of contrast medium) are superior for the evaluation of lesion vascularity, renal vascular anatomy and tumor involvement of venous structures. Not all renal tumors are well delineated during this corticomedullary phase, however, and images obtained during a later phase of enhancement (e.g., the nephrographic or excretory phase) facilitate the detection of small renal masses, especially those involving the medullary region [6, 7, 8, 9]. Excretory phase images (at about three minutes post-injection) are useful for demonstrating tumor involvement of the collecting system. Sensitivity up to 100 percent and specificity of 95 percent has been reported in series for the detection of renal masses when such a technique is used [9]. CT can also provide multiplanar reformations and an interactive display of a three-dimensional model of the affected kidney and its vascular supply, which is particularly helpful for surgeons to view before planned resection of a locally invasive tumor, nephron-sparing partial nephrectomy or venous thrombectomy (Fig. 12.1). It is important to remember that CT evaluation of a renal mass requires administration of iodinated contrast agents, which can lead to contrast-induced nephropathy in patients with compromised renal functions. Contrast-induced nephropathy is associated with significant morbidity and mortality. Therefore, patients with suspected renal dysfunction must undergo renal function evaluation and receive appropriate prophylactic treatment for contrast-induced nephropathy, if indicated. In patients with allergic contraindications to iodinated contrast, gadolinium-enhanced MRI of the abdomen and pelvis may be considered.
Fig. 12.1 RCC with renal vein involvement. Coronal CT images show right renal hypoenhancing mass (arrow) (a) which is extending into the right renal vein (arrow) (b)
302
4.1
M.A. Blake and M.K. Kalra
Solid Renal Tumors
Certain imaging features help in characterization of renal masses. For example, clear cell carcinomas enhance post-intravenous contrast administration to a greater degree than other subtypes of malignant renal lesions [10, 11] (Fig. 12.2). Papillary RCCs, in particular, demonstrate relatively little enhancement (20 to 30 HU) postcontrast and are typically homogeneous [10] (Fig. 12.3), whereas cystic degeneration is more evident in the clear cell subtype [11] (Fig. 12.4). Benign oncocytomas may overlap, however, with clear cell RCC in terms of imaging features and degree of enhancement [12]. A significant pitfall in characterizing renal lesions by CT is the presence of pseudo enhancement in renal cysts on contrast-enhanced images. This pseudo enhancement is greater in smaller renal cysts and is thought to be due to volume averaging and beam hardening effects [13, 14].
Fig. 12.2 Clear cell RCC. US scan (a) showing an echogenic mass (arrow) in medial left kidney. Pre-contrast CT image (b) shows 3 cm left renal mass (arrow) measuring 45 HU which markedly enhances to 104 HU post-contrast (arrow) (c) consistent with an RCC
Fig. 12.3 Papillary RCC. Pre-contrast CT image (a) shows 2 cm right renal mass (arrows) measuring 43 HU which enhances only to 52 HU on post-contrast axial (b) and coronal (c) images, but was papillary RCC on removal
Fig. 12.4 RCC in patient with Von Hippel Lindau Syndrome. T2-weighted MRI (a) showing a slightly T2 hyperintense right renal mass (arrows) with multiple simple T2 bright cysts in the kidneys and pancreas consistent with VHL. The right renal mass shows irregular enhancement post-gadolinium with irregular cystic areas (b) and represents a clear cell RCC
304
4.2
M.A. Blake and M.K. Kalra
Cystic Renal Tumors
The widely used Bosniak classification system categorizes a cystic renal mass for the likelihood of malignancy based on the complexity of the lesion [15]. When any solid enhancing component is present, the cystic renal mass is a Bosniak type 4 lesion (CT classification of cystic lesions based on likelihood of cystic lesion being cancerous), and is highly suspicious for malignancy. Cystic tumors with thin walls and septations without solid components in an adult may represent benign cystic nephroma, multilocular cystic RCC or, rarely, cystic hamartoma of the renal pelvis [16]. These tumors are similar in their gross appearances, but when such a cystic mass represents multilocular cystic RCC, however, it usually has little malignant potential [16] and carries a much better prognosis than other forms of RCC [17].
4.3
Applications of MRI
MRI is useful in the detection and differentiation of cystic and solid renal lesions, with accuracy comparable to that of CT [18]. MRI may function as an excellent tool, both for initial diagnosis and post-treatment follow-up in these patients. It is reliable for evaluation of small renal masses [19] due to its superior soft tissue contrast (Fig. 12-4). In addition, because of its direct multiplanar acquisition capability, MRI has been considered superior to CT for assessing the origin of a renal mass [12], although reformations now possible from isotropically acquired CT datasets is neutralizing this former advantage. State-of-the-art MRI also has been shown to be accurate for the identification and characterization of renal neoplasms amenable to partial nephrectomy [20]. Regarding the use of IV gadolinium, the United States Food and Drug Administration (FDA) has issued a public health advisory regarding the possibility of Nephrogenic Systemic Fibrosis or Nephrogenic Fibrosing Dermopathy (NSF/ NFD) occurring in patients with moderate to end-stage kidney disease after they have had an MRI scan with a gadolinium-based contrast agent [21, 22]. Further research is being conducted regarding this phenomenon, and policy guidelines are being developed. Given the latent period associated with developing NSF after administration of gadolinium, it is not clear at this time as to what strategy must be adopted when imaging a patient with advanced malignant disease.
4.4
Solid Renal Tumors
Solid renal tumors are usually slightly hypointense on T1-weighted images [23, 24], although some renal tumors may show T1 hyperintensity due to hemorrhage or a lipid/proteinaceous component. Renal cortical tumors tend to be mildly hyperintense
12 Imaging of Urinary Tract Tumors
305
[24] on T2-weighted images and enhance following IV contrast administration as on CT (Fig. 12-4).
4.5
Cystic Renal Tumors
Superior soft tissue contrast resolution of MRI does help in characterization of renal masses. Simple cysts are uniformly hypointense on T1-weighted images and hyperintense on T2-weighted images with no enhancement in cysts after administration of contrast. Some complex cysts may show a higher pre-contrast T1 signal and lower T2 signal owing to hemorrhage, debris or proteinaceous material. Identification of contrast enhancement is essential in diagnosing a solid renal neoplasm. It has been reported that the optimal percentage of enhancement threshold for distinguishing cysts from solid tumors on MRI is 15 percent when measurement is performed two to four minutes after administration of contrast material [25]. CT and MRI perform similarly in classifying most cystic renal masses [26]. In some cases, however, MR images may depict additional septa, thickening of the wall or septa or enhancement, which may lead to an upstaged Bosniak cyst classification and, thus, affect patient management [26].
5
Ultrasonography
One of the most important roles of ultrasound is in the characterization of renal lesions as cystic or solid. For a diagnosis of a simple renal cyst, it must be completely anechoic, have a round or oval shape, have a thin imperceptible wall, have posterior acoustic enhancement and be avascular. Ultrasound features suspicious for a malignant cystic lesion include a thickened cystic wall, thickened or nodular septations, irregular or central calcifications and the presence of flow in the septations or cystic wall on Doppler imaging. Most RCCs on ultrasound are solid, but cystic areas and calcifications may be present. Small (3 cm or less) renal masses are more likely to be hyperechoic than larger tumors, and can sometimes be mistaken for an angiomyolipoma which are also echogenic [27] (Fig. 12-2). If fat is present in the lesion on CT or MRI, in most cases an angiomyolipoma can then be diagnosed. On occasion, however, RCC may engulf the perirenal or sinus fat, or a liposarcoma may contain fatty components. Ultrasound and CT and MRI, in general, complement each other in renal lesion characterization. If a renal lesion is “indeterminate” on ultrasound, a dedicated renal protocol CT or MRI is helpful to characterize the lesion further [28]. Conversely, ultrasound may sometimes prove useful for renal lesions that are considered indeterminate on CT [24]. However, renal ultrasound is not considered a useful screening modality because small lesions can be easily missed [29]. CT detects more and smaller renal masses than ultrasound, but the two modalities are
306
M.A. Blake and M.K. Kalra
considered to be relatively similar in the characterization of 1- to 3-cm lesions [30]. Ultrasound may also assist in the preoperative and intraoperative evaluation of renal cortical tumors to help determine whether a partial nephrectomy is appropriate.
6
PET Imaging
Presently, there is insufficient evidence to support a role of FDG-PET or PET/CT for evaluation of renal cell tumors. Varying sensitivity of FDG-PET for the detection of renal malignancy has been reported, ranging from 40 percent to 94 percent [31-33]. Several factors may explain the false negative results of FDG-PET when used to detect renal malignancy, including the physiologic renal excretion of FDG decreasing the contrast between tumor and surrounding kidney, and several other factors, such as histologic subtypes, Fuhrman grades or tumor vascularity. In addition benign oncocytomas have been reported as showing increased FDG activity on PET [34]. FDG-PET may have a potential role in the evaluation of distant metastases, and in the differentiation between recurrence and post-treatment changes [3133, 35]. In addition, PET with new radiopharmaceutical agents such as radioisotope-labeled monoclonal antibodies with specificity for RCC, may become clinically available in the near future.
7
Staging
In staging renal cell carcinoma the extent of disease must be accurately delineated to allow optimal surgical planning. Cancer-specific survival for patients treated with surgery correlates well with the tumor stage [36]. The most commonly used staging system is the TNM system [37] of the American Joint Committee on Cancer. The patient’s overall disease stage is determined by American Joint Committee on Cancer stage groupings. An overall staging accuracy of 91 percent by preoperative CT has been reported, with most staging errors due to the diagnosis of perinephric extension of tumors [38]. Currently, there is no reliable indicator for perinephric tumor spread on CT. Unfortunately, perinephric stranding may be present in the absence of tumor spread [38]. The presence of an intact pseudocapsule, composed of compressed normal renal parenchyma and fibrous tissue surrounding the renal mass best detected by T2-weighted imaging, suggests a lack of perinephric fat invasion and has been reported to be helpful in local staging of renal cortical tumors [39]. The overall accuracy of MRI in staging is comparable to that of CT [12, 24, 40]. Evaluation of the venous system in patients with renal malignancy is crucial for treatment planning. A thrombus involving the renal vein or inferior vena cava in a patient with malignant renal tumor may represent tumor thrombus directly extending from the primary location or a bland thrombus or both (Fig. 12.1). Tumor
12 Imaging of Urinary Tract Tumors
307
thrombus enhances after contrast administration, whereas bland thrombus does not. In some cases the tumor thrombus can invade the wall of the inferior vena cava [41]. If tumoral spread within the inferior vena cava is identified, determination of the superior extent of the thrombus is essential to plan the optimal surgical strategy for thrombectomy, and to minimize the risk of intraoperative tumoral embolism. MRI with MRV or MDCT can be used to image such extension into the vena cava. On CT and MRI assessment of lymph nodes primarily relies on anatomic size criteria, and cross-sectional imaging is limited for detecting normal-sized lymph nodes that harbor microscopic metastatic disease or differentiating metastatic adenopathy from reactive benign lymph node enlargement. CT’s sensitivity for detecting regional lymph node metastases has been reported to be 95 percent [42]. False positive findings of 58 percent have been reported, however, when a size criterion of 1 cm is used for determining nodal metastasis, due to reactive or other benign nodal changes [42]. The most common metastatic sites from malignant renal cortical tumors are the lung, bone (Fig. 12.5), brain, liver and mediastinum [43]. Small renal tumors are unlikely to present initially with metastases. CT of the chest, abdomen and pelvis, and a chest radiograph are now usually done in the initial workup. RCC is associated with hypervascular liver metastases, particularly when the primary tumor is a clear cell type. It has been reported that portal venous phase imaging detects 90 percent of liver metastases from RCC, with the addition of pre-contrast or hepatic arterial phase imaging increasing the sensitivity in lesion detection to almost 100 percent [44]. Bone metastases from RCC are most commonly lytic without osteoblastic activity, so bone scans may be negative in these cases. MRI may be considered in evaluation of symptomatic patients. Similarly, imaging of the brain with CT or MRI is performed when brain metastasis is suggested by patient symptoms or signs.
Fig. 12.5 RCC bone metastasis. Plain film (a) shows large lytic right iliac defect (arrow). CT scan (b) shows a corresponding large soft tissue mass (arrow) destroying the lateral aspect of the right iliac bone, consistent with a metastasis
308
8
M.A. Blake and M.K. Kalra
Re-staging
Up to 30 percent of patients relapse after surgical treatment of localized renal tumors, with pulmonary metastases being the most common distant recurrence site, occurring in up to 60 percent of patients. Other common sites of recurrence include bone, the nephrectomy site, brain, liver, pancreas (Fig. 12.6) and the contralateral kidney. While most bilateral renal tumors present synchronously, asynchronous lesions may occur many years later, requiring that the patient maintain long-term follow-up [45]. For imaging surveillance, CT is the modality of choice for detection of local recurrence and distant metastases. In patients with allergic contraindications to iodinated contrast, gadolinium-enhanced MRI of the abdomen and pelvis may be considered. A chest CT study may be obtained for surveillance of pulmonary metastases, based on pathologic stage and as clinically indicated. FDG-PET may have a potential role in the evaluation of distant metastases, especially with equivocal findings from conventional studies, and in the differentiation of recurrence from post-treatment changes [31, 32, 33, 35]. Because of the high specificity and positive predictive value of FDG-PET, a positive result is strongly suggestive of local recurrence or metastasis, although a negative result unfortunately cannot reliably rule out metastatic disease [35].
Fig. 12.6 RCC pancreatic metastasis. Contrast-enhanced CT showing avidly enhancing 1.5 cm mass (arrow) in the pancreatic head representing a metastasis
12 Imaging of Urinary Tract Tumors
9
309
Summary
CT is generally considered to be the most robust and comprehensive imaging modality for evaluation of renal tumors. When a proper technique is used, CT provides high accuracy in the diagnosis and staging of a renal mass. CT also provides useful diagnostic information for treatment planning and follow-up. MRI and ultrasound have certain advantages, but generally function clinically as valuable problem-solving tools.
10
Bladder Cancer
Bladder cancer is the fourth most common tumor of the urinary tract in the United States accounting for 6 percent to 8 percent of malignancies in men, and 2 percent to 3 percent of malignancies in women [46]. CT’s primary role in bladder cancer is tumor staging and screening for distant metastases. MDCT improves the evaluation of bladder tumors by overcoming the difficulties of previous generations of CT in detecting invasion of contiguous organs and nodal staging [47. MRI, however, is still considered superior to CT for primary staging of bladder carcinoma [48]. The direct multiplanar acquisition capability of MRI, with its superior soft tissue discrimination, offers improved evaluation of local staging of bladder tumors. Positron Emission Tomography/CT (PET/CT) is emerging as a useful tool for the detection of distant metastases.
10.1
Role of CT
CT is a widely accessible and non-invasive imaging modality for the assessment of patients with bladder cancer. MDCT allows rapid screening for metastatic disease and for complications such as hydronephrosis. The inherent contrast between the bladder and extraperitoneal fat on CT facilitates the detection of extravesical spread. Patients with known or suspected bladder cancer should undergo a noncontrast CT, followed by a contrast-enhanced CT scan. This differentiates a blood clot from a mass based on its enhancement characteristics. Bladder cancers also enhance more intensely than the adjacent normal bladder wall following IV contrast administration (Fig. 12.7) [47]. CT is useful in assessing incomplete transurethral resection of the bladder tumor, and is also an excellent method for the detection of local and distant tumor recurrence following cystectomy [49-51]. Potential pertinent CT findings include a mass at the cystectomy site, retroperitoneal lymphadenopathy and hepatic or bone metastases [52]. The overall reported accuracy of CT in detecting and staging bladder cancer varies from 64 percent to 97 percent, whereas that reported for perivesical
310
M.A. Blake and M.K. Kalra
Fig. 12.7 Bladder TCC pre-contrast. CT shows a large right bladder wall mass (arrows) which shows irregular enhancement post-contrast with extension through the right bladder wall
Fig. 12.8 Synchronous TCC. Coronal CT images showing left-sided hydronephrosis with enhancing left pelvic-ureteral (arrow) (a) and inferior bladder (arrow) (b) masses consistent with synchronous TCC
invasion and for lymph node metastases ranges from 83 percent to 93 percent, and 73 percent to 92 percent, respectively [50]. High quality multiplanar reconstructions with MDCT have improved the sensitivity of CT in the detection of bladder tumors, especially for tumors at the base and dome (Fig. 12.8). Traditionally lymph nodes were considered abnormal if the lymph node measures 10 mm or more in its short-axis dimension, although microscopic involvement of normal sized nodes can lead to false negatives and subsequent under staging using this threshold. These criteria are likely to be modified with advancing technology, and the upper limits of “normal” may be reduced, particularly in patients with known bladder cancer. Lymph nodes greater than 8 mm in diameter in the obturator and internal iliac groups are now generally considered metastatic. Lymph
12 Imaging of Urinary Tract Tumors
311
nodes tend to become rounded with metastatic involvement [53]. Intensely enhancing lymph nodes not only occur in patients with bladder cancer, but also with concurrent inflammatory processes in the pelvis [54]. Para aortic adenopathy renders the patient’s cancer unresectable. Extra nodal metastases involve the liver, bones and lungs, and such metastases, if greater than 1 cm, can be well characterized on CT. CT should be obtained especially during follow-up because recurrent disease manifesting as remote metastases can occur [52].
10.2
Pitfalls of CT
A major shortcoming of CT is its inability to detect microscopic invasion of perivesical fat and tumors in normal sized lymph nodes, which is a major cause of under staging. For bladder cancer, the bladder must be adequately distended for accurate interpretation. However, over-distension of the bladder may result in underestimation of the bladder wall thickness and effacement of fat planes between the bladder and adjacent structures. Over staging may result from misinterpretation of normal fat planes between the posterior bladder wall and the seminal vesicles. Early postoperative CT, especially following transurethral resection of the bladder (TURB), can result in inaccurate staging and pseudo lesions [55]. This can be avoided by performing the CT examination an adequate interval after surgery. Similarly, CT should ideally be performed before any intervention such as cystoscopically guided biopsy; otherwise perivesical fibrosis can result in over staging on CT.
11
Applications of MRI
Superior localization and characterization of bladder tumors is possible on MRI due to its excellent intrinsic tissue contrast. The ability to acquire direct multiplanar imaging has been cited in the past as a specific MRI advantage, but multiplanar reconstructions of 3-D datasets are now routinely available with MDCT. Higher field-strength magnets can provide superior image quality through fast scanning techniques, but can pose problems due to its stronger chemical shift artifacts and lower T1 contrast. Phased-array external surface body coils offer higher signal-tonoise ratio and a smaller field of view, producing high spatial resolution image quality [56]. Multiple phased-array detectors allow the application of parallel imaging techniques with resultant higher-resolution images, or shorter acquisition times. Endoluminal coils offer excellent high spatial resolution images of the prostate, seminal vesicles and the inferior posterior bladder wall for accurate delineation of tumor invasion, but provide a more limited field of view. Administration of intravenous
312
M.A. Blake and M.K. Kalra
Fig. 12.9 Pelvic TCC. CT showing enhancing mass (arrow) with surface calcifications in the left renal pelvis representing a pelvic TCC
contrast helps improve overall staging accuracy. Fat saturation or subtraction techniques improve tumor distinction from perivesical fat; otherwise both have high signal intensity on post-contrast T1-weighted images. Transitional cell carcinoma is known to be multicentric in nature (Fig. 12.9), but MRI urography permits evaluation of the entire urinary tract in patients with TCC. Heavily T2-weighted sequences can demonstrate the urine in the pyelum and ureter, especially when they are dilated due to a distal obstruction. A breath hold T1-weighted MRI angiographic sequence also can be performed in a delayed fashion, after contrast injection, to demonstrate gadolinium excretion in the ureters in a fashion similar to intravenous urography (IVU). MRI is considered superior to CT scanning for local staging of carcinoma of the urinary bladder. Dedicated multiplanar imaging and superior intrinsic tissue contrast allows better visualization of the bladder dome, trigone and adjacent structures such as the prostate and seminal vesicles (Fig. 12.10). The reported accuracy of MRI in overall staging of bladder cancer varies from 60 percent to 85 percent, whereas that of local staging varies from 73 percent to 96 percent [57]. The reported overall staging accuracy of gadolinium-enhanced MRI for staging extravesical extension in bladder cancer is 73 percent to 100 percent.
12 Imaging of Urinary Tract Tumors
313
Fig. 12.10 Linitis plastica of bladder wall due to extensive TCC. Coronal CT (a) shows thickened bladder wall (arrow) with bilateral ureteral stents. Sagittal T2-weighted MRI image (b) shows marked irregular thickening of the bladder wall (arrows) which shows diffusely abnormal enhancement on the coronal post-gadolinium image (c)
12
Recent Advances
Virtual endoscopy with CT or MRI is a recently developed technique and preliminary studies have demonstrated its feasibility for imaging the urinary tract [58]. CT cystoscopy can be performed by insufflating air into the urinary tract or following intravenous contrast administration, also known as IVU virtual cystoscopy [58, 59]. There is a concern for the cumulative radiation exposures from CT cystoscopy given the repeated, continual surveillance required of patients with a history of transitional cell carcinoma raises [60]. MRI cystoscopy could overcome this limitation of CT and has also shown encouraging results [61]. MRI studies evaluating ultra small, super paramagnetic iron oxide (USPIO) particles have shown that normal lymph nodes take up this contrast material and show
314
M.A. Blake and M.K. Kalra
a selective decrease in signal intensity on T2- or T2*-weighted MR images, whereas nodes infiltrated with metastases lack uptake and retain their high signal intensity on USPIO- enhanced MR images. This technique, although not yet FDAapproved, promises to greatly improve the accuracy of MRI in the characterization of lymph nodes [62]. The role of (18) F-fluorodeoxyglucose (FDG)-PET and PET/CT in the evaluation of bladder cancer is somewhat limited. FDG is not a useful tracer for the detection of primary tumors because of its renal excretion and accumulation in the bladder. It may be useful, however, in determining the early spread of disease in patients with aggressive primary tumors, and in monitoring response to treatment in advanced disease [63].
Summary In summary, imaging now has an important role in the evaluation of patients with invasive bladder cancer. It is useful in staging of bladder cancer, despite having some limitations. CT is widely accessible and has enjoyed rapid advances in multidetector technology which have far-reaching applications. MRI, due to its intrinsic tissue characterization, is reported to have a higher accuracy for staging bladder cancer. It is superior to CT in determining the depth of bladder wall invasion. CT, as well as traditional MRI, rely on morphological criteria and are both useful in the detection of metastases to the lymph nodes, liver and bone. MRI, however, can now also take advantage of the tremendous advance in lymph node evaluation brought about by the advent of USPIO nodal imaging. MRI is also considered superior in follow-up of patients with bladder cancer post-therapy as it can distinguish biopsy changes more accurately than CT. In conclusion, it is clear that imaging plays a growing and increasingly important role in the evaluation of patients with bladder cancer.
References 1. American Cancer Society. Cancer facts and figures 2006. Atlanta (GA): American Cancer Society; 2006 2. Bosniak MA. Observation of small incidentally detected renal masses. Semin Urol Oncol. 1995;13(4):267-272. 3. Kassouf W, Aprikian AG, Laplante M, et al. Natural history of renal masses followed expectantly. J Urol. 2004;171(1):111-113 4. Beck SD, Patel MI, Snyder ME, et al. Effect of papillary and chromophobe cell type on diseasefree survival after nephrectomy for renal cell carcinoma. Ann Surg Oncol. 2004;11(1):71-77. 5. McKiernan J, Yossepowitch O, Kattan MW, et al. Partial nephrectomy for renal cortical tumors: pathologic findings and impact on outcome. Urology. 2002;60(6):1003-1009. 6. Zhang J, Lefkowitz RA, Bach A. Imaging of Kidney Cancer. Radiologic Clinics of North America 2007 45 (1) 119-147
12 Imaging of Urinary Tract Tumors
315
7. Yuh BI, Cohan RH, Francis IR, et al. Comparison of nephrographic with excretory phase helical computed tomography for detecting and characterizing renal masses. Can Assoc Radiol J. 2000;51(3):170-176. 8. Yuh BI, Cohan RH. Different phases of renal enhancement: role in detecting and characterizing renal masses during helical CT. AJR Am J Roentgenol. 1999;173(3):747-755. 9. Kopka L, Fischer U, Zoeller G, et al. Dual-phase helical CT of the kidney: value of the corticomedullary and nephrographic phase for evaluation of renal lesions and preoperative staging of renal cell carcinoma. AJR Am J Roentgenol. 1997;169(6):1573-1578. 10. Herts BR, Coll DM, Novick AC, et al. Enhancement characteristics of papillary renal neoplasms revealed on triphasic helical CT of the kidneys. AJR Am J Roentgenol. 2002;178(2):367-372. 11. Sheir KZ, El-Azab M, Mosbah A, et al. Differentiation of renal cell carcinoma subtypes by multislice computerized tomography. J Urol. 2005;174(2):451-455 12. Prasad SR, Humphrey PA, Catena JR, et al. Common and uncommon histologic subtypes of renal cell carcinoma: imaging spectrum with pathologic correlation. Radiographics. 2006 Nov-Dec;26(6):1795-806 13. Birnbaum BA, Maki DD, Chakraborty DP, et al.. Renal cyst pseudo enhancement: evaluation with an anthropomorphic body CT phantom. Radiology. 2002;225(1):83-90. 14. Abdulla C, Kalra MK, Saini S, et al. Pseudo enhancement of simulated renal cysts in a phantom using different multidetector CT scanners. AJR Am J Roentgenol. 2002;179(6):1473-1476. 15. Bosniak MA. The current radiological approach to renal cysts. Radiology. 1986; 158(1):1-10. 16. Eble JN, Bonsib SM. Extensively cystic renal neoplasms: cystic nephroma, cystic partially differentiated nephroblastoma, multilocular cystic renal cell carcinoma, and cystic hamartoma of renal pelvis. Semin Diagn Pathol. 1998;15(1):2-20. 17. Brinker DA, Amin MB, de Peralta-Venturina M, et al. Extensively necrotic cystic renal cell carcinoma: a clinicopathologic study with comparison to other cystic and necrotic renal cancers. Am J Surg Pathol. 2000;24(7):988-995. 18. Semelka RC, Shoenut JP, Kroeker MA, et al. Renal lesions: controlled comparison between CT and 1.5-T MRI with non-enhanced and gadolinium-enhanced fat-suppressed spin-echo and breath hold FLASH techniques. Radiology. 1992;182(2):425-430. 19. Scialpi M, Di Maggio A, Midiri M, et al.. Small renal masses: assessment of lesion characterization and vascularity on dynamic contrast-enhanced MRI with fat suppression. AJR Am J Roentgenol. 2000;175(3):751-757. 20. Pretorius ES, Siegelman ES, Ramchandani P, et al. Renal neoplasms amenable to partial nephrectomy: MRI. Radiology. 1999;212(1):28-34. 21. Broome DR, Girguis MS, Baron PW, Cottrell AC, Kjellin I, Kirk GA. M. Gadodiamide-assocaited nephrogenic systemic fibrosis: why radiologists should be concerned. Am J Roentgenol. 2007;188(2):586-92. 22. Elizabeth A. Sadowski, Lindsey K. Bennett, Micah R. Chan, Andrew L. Wentland, Andrea L. Garrett, Robert W. Garrett, and Arjang Djamali Nephrogenic Systemic Fibrosis: Risk Factors and Incidence Estimation Radiology 2007 243: 148-157 23. Eilenberg SS, Lee JK, Brown J, et al. Renal masses: evaluation with gradient-echo Gd-DTPAenhanced dynamic MRI. Radiology. 1990;176(2):333-338. 24. Fein AB, Lee JK, Balfe DM, et al. Diagnosis and staging of renal cell carcinoma: a comparison of MRI and CT. AJR Am J Roentgenol. 1987;148(4):749-753. 25. Ho VB, Allen SF, Hood MN, et al.. Renal masses: quantitative assessment of enhancement with dynamic MRI. Radiology. 2002;224(3):695-700. 26. Israel GM, Hindman N, Bosniak MA. Evaluation of cystic renal masses: comparison of CT and MRI by using the Bosniak classification system. Radiology. 2004;231(2):365-371. 27. Forman HP, Middleton WD, Melson GL, et al. Hyperechoic renal cell carcinomas: increase in detection at US. Radiology. 1993;188(2):431-434. 28. Prasad SR, Saini S, Stewart S, et al. CT characterization of “indeterminate” renal masses: targeted or comprehensive scanning?. J Comput Assist Tomogr. 2002;26(5):725-727.
316
M.A. Blake and M.K. Kalra
29. Warshauer DM, McCarthy SM, Street L, et al. Detection of renal masses: sensitivities and specificities of excretory urography/linear tomography, US, and CT. Radiology. 1988;169(2):363-365. 30. Jamis-Dow CA, Choyke PL, Jennings SB, et al. Small (< or = 3-cm) renal masses: detection with CT versus US and pathologic correlation. Radiology. 1996;198(3):785-788. 31. Aide N, Cappele O, Bottet P, et al. Efficiency of [(18)F] FDG-PET in characterizing renal cancer and detecting distant metastases: a comparison with CT. Eur J Nucl Med Mol Imaging. 2003;30(9):1236-1245. 32. Montravers F, Grahek D, Kerrou K, et al. Evaluation of FDG uptake by renal malignancies (primary tumor or metastases) using a coincidence detection gamma camera. J Nucl Med. 2000;41(1):78-84. 33. Ramdave S, Thomas GW, Berlangieri SU, et al. Clinical role of F-18 fluorodeoxyglucose positron emission tomography for detection and management of renal cell carcinoma. J Urol. 2001;166(3):825-830. 34. Blake MA, McKernan M, Setty B, et al. Renal oncocytoma displaying intense activity on 18F-FDG PET. AJR Am J Roentgenol. 2006;186(1):269-270. 35. Schoder H, Larson SM. Positron emission tomography for prostate, bladder, and renal cancer. Semin Nucl Med. 2004;34(4):274-292. 36. Javidan J, Stricker HJ, Tamboli P, et al.. Prognostic significance of the 1997 TNM classification of renal cell carcinoma. J Urol. 1999;162(4):1277-1281. 37. Kidney. In: American Joint Committee on Cancer. AJCC Cancer Staging Manual. 6th edition. New York: Springer; 2002;p. 323-325. 38. Johnson CD, Dunnick NR, Cohan RH, et al. Renal adenocarcinoma: CT staging of 100 tumors. AJR Am J Roentgenol. 1987;148(1):59-63. 39. Roy C, El Ghali S, Buy X, et al.. Significance of the pseudocapsule on MRI of renal neoplasms and its potential application for local staging: a retrospective study. AJR Am J Roentgenol. 2005;184(1):113-120. 40. Semelka RC, Shoenut JP, Magro CM, et al. Renal cancer staging: comparison of contrastenhanced CT and gadolinium-enhanced fat-suppressed spin-echo and gradient-echo MRI. J Magn Reson Imaging. 1993;3(4):597-602. 41. Didier D, Racle A, Etievent JP, et al. Tumor thrombus of the inferior vena cava secondary to malignant abdominal neoplasms: US and CT evaluation. Radiology. 1987;162(1 Pt 1):83-89. 42. Studer UE, Scherz S, Scheidegger J, et al. Enlargement of regional lymph nodes in renal cell carcinoma is often not due to metastases. J Urol. 1990;144(2 Pt 1):243-245. 43. Hilton S. Imaging of renal cell carcinoma. Semin Oncol. 2000;27(2):150-159. 44. Raptopoulos VD, Blake SP, Weisinger K, et al. Multiphase contrast-enhanced helical CT of liver metastases from renal cell carcinoma. Eur Radiol. 2001;11(12):2504-2509. 45. Rabbani F, Herr HW, Almahmeed T, et al. Temporal change in risk of metachronous contralateral renal cell carcinoma: influence of tumor characteristics and demographic factors. J Clin Oncol. 2002;20(9):2370-2375. 46. Barentsz JO, Witjes JA, Ruijs JH. What is new in bladder cancer imaging? Urol Clin North Am 1997; 24: 583-602. 47. MacVicar AD. Bladder cancer staging. BJU Int 2000; 86 [suppl 1]: 111-122. 48. Barentsz JO, Jager GJ, Witjes JA, et al. Primary staging of urinary bladder carcinoma: the role of MRI and a comparison with CT. Eur Radiol 1996; 6:129-133. 49. Stein JP, Lieskovsky G, Cote R, et al. Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. J Clin Oncol 2001; 19:666-675. 50. Kim J K, Park SY, Ahn HJ, et al. Bladder Cancer: Analysis of Multi–Detector Row Helical CT Enhancement Pattern and Accuracy in Tumor Detection and Perivesical Staging. Radiology 2004; 231:725-731. 51. Kundra V, Silverman PM. Imaging in oncology from the University of Texas M. D. Anderson Cancer Center. Imaging in the diagnosis, staging, and follow-up of cancer of the urinary bladder. AJR Am J Roentgenol. 2003; 180(4): 1045-54.
12 Imaging of Urinary Tract Tumors
317
52. Ellis JH, McCullough NB, Francis IR, et al. Transitional cell carcinoma of the bladder: patterns of recurrence after cystectomy as determined by CT. AJR Am J Roentgenol. 1991; 157(5): 999-1002. 53. Vinnicombe SJ, Norman AR, Nicholson V, et al. Normal pelvic lymph nodes: evaluation with CT after bipedal lymphangiography. Radiology 1995; 194:349-55 54. Husband JE, Robinson L, Thomas G. Contrast enhancing lymph nodes in bladder cancer: a potential pitfall on CT. Clin Radiol. 1992; 45(6): 395-8. 55. Kim B, Semelka RC, Ascher SM, et al. Bladder tumor staging: comparison of contrastenhanced CT, T1- and T2-weighted MRI, dynamic gadolinium-enhanced imaging, and late gadolinium-enhanced imaging. Radiology 1994; 193: 239-245. 56. Maeda H, Kinukawa T, Hattori R, et al. Detection of muscle layer invasion with sub millimeter pixel MR images: staging of bladder carcinoma. Magn Reson Imaging 13:9, 1995. 57. Tekes A, Kamel I, Imam K, et al. Dynamic MRI of bladder cancer: evaluation of staging accuracy. AJR Am J Roentgenol. 2005; 184 (1): 121-7. 58. Lammle M, Beer A, Settles M, et al. Reliability of MRI-based virtual cystoscopy in the diagnosis of cancer of the urinary bladder. AJR Am J Roentgenol. 2002; 178(6): 1483-8. 59. Kawai N, Mimura T, Nagata D, et al. Intravenous urography-virtual cystoscopy is a better preliminary examination than air virtual cystoscopy. BJU Int. 2004; 94(6): 832-6. 60. Tsili ACh, Tsampoulas C, Chatziparaskevas N, et al. Computed tomographic virtual cystoscopy for the detection of urinary bladder neoplasms. Eur Urol. 2004; 46(5): 579-85. 61. Nambirajan T, Sohaib SA, Muller-Pollard C, et al. Virtual cystoscopy from computed tomography: a pilot study. BJU Int. 2004; 94(6): 828-31. 62. Kawashima A, Glockner JF, King BF Jr. CT urography and MR urography. Radiol Clin North Am. 2003; 41(5): 945-61. 63. Drieskens O, Oyen R, Van Poppel H, Vankan Y, Flamen P, Mortelmans L. FDG-PET for preoperative staging of bladder cancer. Eur J Nucl Med Mol Imaging. 2005; 32(12): 1412-7.
13
Current Status of Imaging for Adrenal Malignant Involvement Michael A. Blake, FFR (RCSI), FRCR, MRCPI and Mannudeep K. Kalra, MD, DNB
Key Points ●
●
1
CT, MRI, PET and PET/CT are useful in differentiating benign from malignant adrenal involvement. Image-guided adrenal biopsy should be considered if needed for treatment planning, and for the now relatively uncommon lesions that remain indeterminate by imaging.
Introduction
Adrenal masses are relatively common in the general population, with a mean prevalence determined from several large autopsy studies of 2.3 percent [1]. Given the propensity for and the clinical importance of adrenal metastatic involvement, accurate diagnosis of adrenal masses is of particular important in oncologic patients. Fortunately non-invasive radiology can usually determine whether a mass is benign or likely malignant (indeterminate lesion), based on recent research into the imaging characteristics of adrenal masses.
2
Role of CT
CT examinations, using specialized adrenal protocols, have been shown to characterize solid adrenal masses as benign or indeterminate (likely malignant) with a high degree of accuracy [2, 3, 4], thus providing the means to diagnose the vast majority of non-functional adrenal lesions in a single step, usually without the need for invasive procedures. Department of Radiology, White 270, Massachusetts General Hospital, 55 Fruit St, Boston MA 02114, Corresponding author: Michael A. Blake, e-mail:
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
319
320
M.A. Blake and M.K. Kalra
Adrenal CT protocols exploit two significant differences between benign and malignant adrenal masses: their fat content and their vascular properties. Adenomas and myeloplipomas have relatively high fat content, compared to malignancies, and generally appear to be of low density in CT images (Fig.13.1). This difference was first taken advantage of in 1991, when Lee, et al. [5] demonstrated that the attenuation of adenomas on non-contrast CT images differed significantly from malignancies and, indeed, was superior to size measurements in this regard. Several studies have shown that malignant adrenal lesions with attenuations less than 10 Hounsfield units (HU) are extremely rare [5-9]. Consequently, a threshold value of 10 HU is now generally accepted as a practical cut-off value to distinguish an adenoma from a possible malignancy (indeterminate lesion) [9, 10] (Fig. 13.2). There is considerable overlap in the enhanced attenuation values of malignant and benign adrenal masses following intravenous contrast administration. However, it has been shown that the contrast agent washes out from adenomas significantly more rapidly than that from metastatic masses (p < .001) [11]. Studies have since demonstrated that the washout of contrast from adenomas is rapid and reaches a plateau within 10 to 15 minutes, whereas much of the enhancement remains in nonadenomas after 45 minutes [12], including metastases, most pheochromocytomas [13] and adrenocortical carcinomas [3] (Fig. 13.3). Furthermore, adrenocortical carcinomas have a propensity to involve the adrenal veins and IVC.
Fig. 13.1 Adrenal myelolipoma. Contrast-enhanced CT scan showing fat containing mass in right adrenal (arrow) consistent with a myelolipoma
13 Current Status of Imaging for Adrenal Malignant Involvement
321
Fig. 13.2 Adrenal adenoma on CT. Large left adrenal mass measuring 1 HU pre-contrast (a), 43 HU on dynamic imaging (b) and 29 HU on 10-minute delayed imaging (c). The low pre-contrast attenuation of the lesion <10 HU is consistent with a lipid rich adenoma although the washout values are not very high
Malignancies have abnormal vasculature with high microvascular density, accompanied by slow flow and abnormally high vascular endothelial permeability [14]. Due to these vascular abnormalities, more contrast agent is likely to accumulate in and to be retained for a longer period in malignant tissue. These differences explain why the contrast agent washout rate from benign adenomas is significantly faster, compared to malignant masses. Dedicated adrenal CT protocols, which combine non-contrast, early and delayed enhancement, avail of both the physiological differences described above and have been shown to be both highly sensitive and specific [2, 4]. The attenuation (HU) of the adrenal mass is measured on all three scans and the washout rate is calculated from pre-contrast attenuation (P), contrast-enhanced attenuation (E) measured during the portal venous phase and delayed contrast enhancement (D) attenuation measured 10 to 15 minutes later (Fig. 13.4.). The absolute percentage washout
322
M.A. Blake and M.K. Kalra
Fig. 13.3 Adrenal carcinoma. Large irregularly enhancing adrenal mass (a) on contrast-enhanced CT with evidence of invasion of the IVC (arrow) (b) and demonstrated delayed retention of contrast consistent with an adrenal carcinoma
Fig. 13.4 Lipid-poor adrenal adenoma on washout analysis. Right adrenal mass (arrows) measuring 20 HU pre-contrast (a) 80 HU on dynamic imaging (b) and 40 HU on delayed 10-minute images. The lesion is indeterminate by non-contrast criteria > 10 HU, but as the RPW = 50 percent and APW = 66.6 percent it is consistent with a lipid-poor adenoma
13 Current Status of Imaging for Adrenal Malignant Involvement
323
(APW) and the relative percentage washout (RPW) are calculated from the following formulae: APW = 100 x ([E – D]/[E – P] RPW = 100 x [E – D]/E We have demonstrated that the combination of pre-contrast attenuation (threshold: < 0 HU for adenomas, > 43 HU for malignancies), 10-minute relative percentage washout (threshold: > 37.5 percent) and 10-minute absolute percentage washout (threshold: > 52.0 percent) gives a sensitivity of 100 percent and a specificity of 98 percent for differentiating between benign and malignant lesions [2]. As expected, these cut-off values are somewhat lower than those reported (40 percent and 60 percent, respectively) in an earlier study from the University of Michigan in which the final CT was at a 15 minute, rather than a 10 minute delay after the contrast injection [4]. Other researchers have reported the sensitivity and specificity for detection of adenomas by using the percentage washout of contrast after 10 or 15 minutes to be 83 percent to 98 percent and 93 percent to 100 percent, respectively [15-17]. Using these protocols, the vast majority of adrenal masses can thus be diagnosed as indeterminate (suspicious for malignancy) or benign and, if the latter, require no further diagnostic workup, reducing the number of patients who require invasive biopsy evaluation or surgery. However, if no conclusive categorization of an incidentally adrenal mass has been obtained, follow-up CT imaging has been advised at six, 12 and 24 months after the initial discovery of the adrenal lesion, although there are no data from long-term studies to provide supportive evidence [18]. Recent initial reports of CT histogram analysis of individual pixel attenuations have not yet reached a clinical consensus, but suggest it may also give clinically useful adrenal diagnostic information [19, 20]. The imaging characteristics of pheochromocytomas are variable, but they usually contain little fat and, therefore, usually have attenuations >10 HU. Pheochromocytomas typically enhance avidly with contrast, but can be heterogeneous or show no enhancement due to cystic changes. In addition, washout rates are inconsistent, and it is possible to misclassify pheochromocytomas as adenomas or metastases on imaging [13, 21] (Fig. 13.5). Pheochromocytomas can also demonstrate hemorrhage and any hemorrhagic adrenal lesion needs to be carefully studied with follow-up imaging to exclude an underlying tumor that has bled (Fig. 13.6) Extra-adrenal pheochromocytomas are called paragangliomas and can occur anywhere along the sympathetic chain (Fig. 13.7).
3
Applications of MRI
MRI is not considered quite as accurate as adrenal protocol CT, but may be useful if a CT examination is equivocal, especially if an unenhanced CT has been performed, and the use of CT contrast agent is contraindicated. Chemical shift MRI exploits the same physiological difference between adenomas and malignancies as
324
M.A. Blake and M.K. Kalra
Fig. 13.5 Low density pheochromocytoma. Right adrenal mass representing pathologically proven pheochromocytoma which measured 10 HU on non-contrast CT (arrow)
Fig. 13.6 Adrenal hemorrhagic cyst. Large left adrenal lesion (arrow) with hematocrit level and dense component in non-dependent component on non-contrast supine (a) and prone (b) CT scans, respectively. Contrast-enhanced CT scan (c) shows no underlying enhancing lesion (arrow) and further follow-up scans were also negative of underlying tumor
13 Current Status of Imaging for Adrenal Malignant Involvement
325
Fig. 13.7 Extra-adrenal pheochromocytoma/paraganglioma. CT scan showing 2.2 cm-enhancing mass (arrow) anterolateral to the aorta consistent with an extra-adrenal paraganglioma
non-contrast CT imaging, namely the high fat content of adenomas. In this method, T1-weighted images are acquired at echo times that are in-phase and out-of-phase to take advantage of the consequences of the different resonant frequency rates of protons in fats and protons. In adenomas, out-of-phase signal intensity is lower due to cancellation of the signals for fat and water protons than that on in-phase images where the signals combine. The reported sensitivity of chemical shift MRI ranges from 81-100 percent and the specificity from 94-100 percent [23,24]. Simple cysts can be distinguished from necrotic malignancy on MRI by their uniformly low signal on T1 weighted images, uniformly hyperintense on T2 weighted images and lack of enhancement on post contrast images (Figure 8).
3
PET and PET/CT
PET using the tracer 2-[18F]-fluoro-deoxyglucose (FDG) is a very good method for detecting metastatic cancer to the adrenals [25-29]. Studies report that sensitivities and specificities of FDG-PET for detecting adrenal malignancy are in the range of 93 percent to 100 percent and 78 percent to 100 percent, respectively. False negatives have been reported due to hemorrhage and necrosis [27], while in one case a renal cell carcinoma metastasis was not identified with FDG-PET [30]. Combined PET/CT scanning has been used to show that the combination of unenhanced CT
326
M.A. Blake and M.K. Kalra
Fig. 13.8 Left adrenal cyst. Coronal T1-weighted (a), axial T2-weighted (b) and gadoliniumenhanced (c) MR images demonstrating a T1 hypointense, T2 hyperintense non-enhancing 2.5 cm left adrenal lesion (arrows) consistent with a cyst
with PET is better than PET alone for diagnosing malignant adrenal masses [26, 28] (Fig. 13.9), while one study found that PET/CT when using an adrenal protocol CT scan was even better, with a sensitivity and specificity of 100 percent in that study population [26]. Pheochromocytomas are also metabolically active and can also be detected with FDGPET. In addition, some new agents – 18F-fluorodopamine and 11C-hydro-xyephedrine – show promise as more specific and sensitive agents for pheochromocytomas [21].
4
Adrenal Biopsy
Adrenal biopsy should be considered if needed for treatment planning and for the relatively few cases in which CT, MRI or PET imaging do not provide a definitive diagnosis [23, 31-34]. CT-guided percutaneous needle aspiration biopsy (PNAB) is
13 Current Status of Imaging for Adrenal Malignant Involvement
327
Fig. 13.9 Adrenal metastasis from lung cancer on PET/CT. Axial and coronal PET/CT images demonstrating intense FDG uptake in the primary left upper lobe lung carcinoma (black arrow) and in the left adrenal metastasis (white arrows)
Fig. 13.10 Collision tumors. Contrast-enhanced CT scan (a) shows a low density left adrenal lesion stable, compared with previous scans consistent with an adenoma in a patient with breast cancer. CT scan (b) six months later shows new enhancing mass (arrow) in left adrenal consistent with breast cancer metastasis displacing the low density adenoma (collision tumors)
a well-established technique and the method of choice. However, pheochromocytoma must be recognized and, prophylaxized against if indicated, to avoid a hypertensive crisis provoked by the PNAB. Histological samples can be useful for the evaluation of metastasis in patients with no other signs of metastases and a heterogenous adrenal mass with a high attenuation value (> 20 HU). However, sampling error can sometimes lead to false negative PNAB results. Collision tumors affecting the adrenal gland (Fig. 13.10) when two different tumors co-exist in the adrenal can occasionally
328
M.A. Blake and M.K. Kalra
occur and PET/CT has been shown to help identify and direct appropriate biopsy in such a circumstance [35].
Summary CT, MRI, PET and PET/CT have all been shown to be clinically useful in differentiating benign from malignant adrenal involvement. Image-guided adrenal biopsy should be considered if needed for treatment planning and for the now relatively uncommon lesions that remain indeterminate by imaging.
References 1. Barzon L, Sonino N, Fallo F, Palu G, Boscaro M. Prevalence and natural history of adrenal incidentalomas. Eur J Endocrinol 2003; 149:273-85. 2. Blake MA, Kalra MK, Sweeney AT, et al. Distinguishing benign from malignant adrenal masses: multi-detector row CT protocol with 10-minute delay. Radiology 2006:578-585. 3. Szolar DH, Korobkin M, Reittner P, et al. Adrenocortical carcinomas and adrenal pheochromocytomas: mass and enhancement loss evaluation at delayed contrast-enhanced CT. Radiology 2005; 234:479-85. 4. Caoili EM, Korobkin M, Francis IR, et al. Adrenal masses: characterization with combined unenhanced and delayed enhanced CT. Radiology 2002; 222:629-33. 5. Lee M, Hahn P, Papanicolaou N, et al. Benign and malignant adrenal masses: CT distinction with attenuation coefficients, size, and observer analysis. Radiology 1991; 179:415-8. 6. Korobkin M, Brodeur FJ, Yutzy GG, et al. Differentiation of adrenal adenomas from nonadenomas using CT attenuation values. AJR Am J Roentgenol 1996; 166:531-6. 7. Singer A, Obuchowski N, Einstein D, Paushter D. Metastasis or adenoma? Computed tomographic evaluation of the adrenal mass. Cleve Clin J Med 1994; 61:200-5. 8. van Erkel A, van Gils A, Lequin M, Kruitwagen C, Bloem J, Falke T. CT and MR distinction of adenomas and nonadenomas of the adrenal gland. J Comput Assist Tomogr 1994; 18:432-8. 9. Boland GW, Lee MJ, Gazelle GS, Halpern EF, McNicholas MM, Mueller PR. Characterization of adrenal masses using unenhanced CT: an analysis of the CT literature. AJR Am J Roentgenol 1998; 171:201-4. 10. Choyke P. ACR Appropriateness Criteria on Incidentally Discovered Adrenal Masses. J Am Coll Radiol 2006; 3:498-504. 11. Korobkin MF, Brodeur FJ, Francis IR, Quint LE, Dunnick NR, Goodsitt M. Delayed enhanced CT for differentiation of benign from malignant adrenal masses. Radiology 1996; 200:737-42. 12. Korobkin MF, Brodeur FJ, Francis IR, Quint LE, Dunnick NR, Londy F. CT time-attenuation washout curves of adrenal adenomas and nonadenomas. AJR Am J Roentgenol 1998; 170: 747-52. 13. Blake MA, Krishnamoorthy SK, Boland GW, et al. Low-density pheochromocytoma on CT: a mimicker of adrenal adenoma. AJR Am J Roentgenol 2003; 181:1663-8. 14. Jain RK, Munn LL, Fukumura D. Dissecting tumor pathophysiology using intravital microscopy. Nat Rev Cancer 2002; 2:266-76. 15. Boland GW, Hahn PF, Pena C, Mueller PR. Adrenal masses: characterization with delayed contrast-enhanced CT. Radiology 1997; 202:693-6. 16. Caoili EM, Korobkin M, Francis IR, Cohan RH, Dunnick NR. Delayed enhanced CT of lipidpoor adrenal adenomas. AJR Am J Roentgenol 2000; 175:1411-5.
13 Current Status of Imaging for Adrenal Malignant Involvement
329
17. Pena CS, Boland GW, Hahn PF, Lee MJ, Mueller PR. Characterization of indeterminate (lipid-poor) adrenal masses: use of washout characteristics at contrast-enhanced CT. Radiology 2000; 217:798-802. 18. Young WF. The incidentally discovered adrenal mass. N Eng J Med 2007; 356:601-10. 19. Bae KT, Fuangtharnthip P, Prasad SR, Joe BN, Heiken JP. Adrenal masses: CT characterization with histogram analysis method. Radiology 2003; 228:735-42. 20. Remer EM, Motta-Ramirez GA, Shepardson LB, Hamrahian AH, Herts BR. CT histogram analysis in pathologically proven adrenal masses. AJR Am J Roentgenol 2006; 187:191-6. 21. Blake MA, Kalra MK, Maher MM, et al. Pheochromocytoma: an imaging chameleon. Radiographics 2004; 24:S87-99. 22. Israel GM, Korobkin M, Wang C, Hecht EN, Krinsky GA. Comparison of unenhanced CT and chemical shift MRI in evaluating lipid-rich adrenal adenomas. AJR Am J Roentgenol 2004; 183:215-9. 23. Mayo-Smith WW, Boland GW, Noto RB, Lee MJ. State-of-the-art adrenal imaging. Radiographics 2001; 21:995-1012. 24. Haider MA, Ghai S, Jhaveri K, Lockwood G. Chemical shift MR imaging of hyperattenuating (>10 HU) adrenal masses: does it still have a role? Radiology 2004; 231:711-716. 25. Boland GW, Goldberg MA, Lee MJ, et al. Indeterminate adrenal mass in patients with cancer: evaluation at PET with 2-[F-18]-fluoro-2-deoxy-D-glucose. Radiology 1995; 194:131-4. 26. Blake MA, Slattery JM, Kalra MK, et al. Adrenal lesions: characterization with fused PET/CT image in patients with proved or suspected malignancy–initial experience. Radiology 2006; 238:970-7. 27. Kumar R, Xiu Y, Yu JQ, et al. 18F-FDG PET in evaluation of adrenal lesions in patients with lung cancer. J Nucl Med 2004; 45:2058-62. 28. Metser U, Miller E, Lerman H, Lievshitz G, Avital S, Even-Sapir E. 18F-FDG PET/CT in the Evaluation of Adrenal Masses. J Nucl Med 2006; 47:32-37. 29. Erasmus JJ, Patz EF, Jr., McAdams HP, et al. Evaluation of adrenal masses in patients with bronchogenic carcinoma using 18F-fluorodeoxyglucose positron emission tomography. AJR Am J Roentgenol 1997; 168:1357-60. 30. Minn H, Salonen A, Friberg J, et al. Imaging of adrenal incidentalomas with PET using (11)Cmetomidate and (18)F-FDG. J Nucl Med 2004; 45:972-9. 31. Mansmann G, Lau J, Balk E, Rothberg M, Miyachi Y, Bornstein SR. The clinically inapparent adrenal mass: update in diagnosis and. Endocr Rev 2004; 25:309-40. 32. Grumbach MM, Biller BM, Braunstein GD, et al. Management of the clinically inapparent adrenal mass (“incidentaloma”). Ann Intern Med 2003; 138:424-9. 33. Schteingart DE, Doherty GM, Gauger PG, et al. Management of patients with adrenal cancer: recommendations of an international consensus conference. Endocr Relat Cancer 2005; 12:667-80. 34. Stone J. Incidentalomas–clinical correlation and translational science required. N Engl J Med 2006; 354:2748-9. 35 Blake MA, Sweeney AT, Kalra MK, Maher M. Collision Adrenal Tumors on PET/CT. Am J Roentgenol. 183(3):864-5, 2004.
14
Recent Advances in Imaging of Male Reproductive Tract Malignancies Jurgen J. Fütterer1, MD PhD and J. Roan Spermon2, MD PhD
Key Points ●
● ● ●
●
●
1
Testicular ultrasound is the initial investigative tool with regard to scrotal masses. Testicular cancer has a five-year survival rate exceeding 95 percent. Ninety-five percent of all testicular tumors are germ cell tumors. The sensitivity of testicular ultrasound in detecting testicular tumors is almost 100 percent. Computed tomography is used for staging metastatic disease and for follow-up after therapy in patients with disseminated disease. Positron emission tomography and MRI adds little to the management of clinical stage I non-seminoma germ cell cancer.
Introduction
The male reproductive system includes those organs whose function is to accomplish reproduction. This consists of testes, which produce spermatoza and hormones, a series of ducts that store and transport the sperm, seminal vesicles, the prostate and the penis. Cancer of the male reproductive system includes testicular, prostatic and penile neoplasms. Testicular cancer is the most common cancer in men between 15- to 35-years–old, and about 36,000 men are diagnosed with testicular cancer each year. Prostate cancer is the most frequently diagnosed malignancy in males. Cancer of 1
Departments of Radiology
2
Departments of Urology
1,2
Radboud University Nijmegen Medical Centre, Geert Grooteplein zuid 10, NL 6500 HB, Nijmegen, The Netherlands Corresponding author: Jurgen J. Fütterer (
[email protected])
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
331
332
J.J. Fütterer and J.R. Spermon
the penis is rare in western males, but more common in South East Asia and India. It is most often diagnosed in men over the age of 60 years. This chapter will present an overview of imaging of male reproductive tract malignancies.
2
Prostate Cancer
Prostate cancer is the most frequently diagnosed malignancy in western males and the incidence is increasing [1]. It is predicted that in 2007 in the United States alone 218,890 men will be diagnosed with prostate cancer [1]. This is partly due to a growing population of elderly man, but a major factor is the expanding use of the prostate-specific antigen (PSA) test as a prostate cancer biomarker. Between 1989 and 2002, the age-standardized incidence rate of prostate cancer increased by 21.3 percent in the United States. However, at the same time epidemiological surveys demonstrated decreased prostate cancer mortality in several countries since 1993. This decrease in mortality is mostly attributed to earlier diagnosis with a reduction in the number of men with distant metastases. From autopsy studies it is known that prostate cancer can be found in 55 percent of men in their fifth decade and 64 percent in their seventh decade, respectively [2-4]. Prostate cancer is very common in elderly males, and it occurs with a lifetime risk of one in 10 [1]. However, only one in eight of these men will die from this disease [1]. All patient and tumor characteristics must be evaluated to determine the treatment that optimally suits the individual patient. Most often, PSA level, the results of digital rectal examination and histopathological biopsy findings are used for this purpose. However, imaging plays an important role to detect, localize and to stage prostate cancer. This directly influences the diagnostic work-up and may lead to important changes in treatment strategy.
2.1
Prostate Anatomy
On the basis of its embryological origins the prostate is anatomically divided into three zones that are eccentrically located around the urethra: the innermost transition zone, the central zone and the outermost peripheral zone [5, 6]. In older patients the former two cannot be distinguished radiologically due to compression of the central zone by benign prostatic hyperplasia (BPH) in the transitional zone; therefore they are collectively referred to as the central gland, as opposed to the outer gland, which is composed of the peripheral zone. The prostate is divided into the apex and the base. The latter is directed upward and is applied to the inferior surface of the bladder. The apex is directed downward and is in contact with the superior fascia of the urogenital diaphragm. There is still a debate about whether the prostate has a capsule or not. The prostate is surrounded by a thick layer of fibromuscular tissue corresponding to the capsule. The ‘true’ prostatic capsule, however, is a thin (0.5 to 2 mm) layer of connective tissue located externally to the peripheral zone. Around this layer there is a
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
333
pelvic fascia, often called the “false” prostatic capsule. Satter, et al. considered the prostate capsule as an extension of the prostate parenchyma itself [7, 8]. The periprostatic venous plexus surrounds the gland and drains into the internal iliac veins and the presacral veins. The neurovascular bundles course along the posterolateral aspect of the gland and is a preferential path for tumor spread due to small nerve branches penetrating the prostate capsule in this area. Knowledge of the zonal anatomy of the prostate is useful considering that many prostatic diseases have a zonal distribution. More than 70 percent of adenocarcinoma of the prostate arises in the peripheral zone, whereas about 20 percent emerge in the transitional zone and 10 percent in the central zone.
2.2
Detection and Localization of Prostate Cancer
In its early stage prostate cancer is commonly asymptomatic because most cancers are located in the peripheral zone. A few patients have symptoms of the lower urinary tract due to obstruction. Prostate cancer patients rarely present with symptoms of haematuria or haematospermia. Prostate cancer is suspected in patients with elevated PSA values. The urologic work-up in patients with elevated PSA consists of a digital rectal examination and transrectal ultrasound (TRUS). The positive predictive value of a digital rectal examination in the detection of prostate cancer depends on the patient’s age, race, and serum PSA value. In a screening population the positive predictive value varies from 4 percent to 11 percent (PSA 0 to 2.9 ng/mL), and from 33 percent to 83 percent (PSA > 3 ng/mL) [9, 10]. The reproducibility and the interobserver agreement of a digital rectal examination are limited [11, 12].
2.3
Transrectal Ultrasound (TRUS)
Grayscale TRUS appearance of prostate cancer is a hypoechoic lesion in the peripheral zone. Other conditions such as prostatitis and prostatic intraepithelial neoplasia may also present as hypoechoic lesions (Fig. 14.1) [13, 14]. It is important to note that over 40 percent of prostate cancer lesions are isoechoic while only 5 percent are hyperechoic [15]. The positive predictive value of the hypoechoic lesion in the average urologic population ranged from 18 percent to 53 percent [16]. The systematic TRUS-guided biopsy protocol (sample tissue at standard locations) has become the most common biopsy technique [17]. The number of cores taken per session varies across institutions. Prostate cancer detection rates have varied from 19 percent to 40 percent [18, 19] and repeat biopsy sessions are often necessary [20]. Color Doppler TRUS – Doppler imaging enables the detection of blood flow to or from the ultrasound probe. Increased blood flow due to neovascularity is one of the characteristics of prostate cancer. Doppler enhancement correlated with the microvessel density and Gleason score of a lesion in a study of 96 patients with lower urinary tract symptoms, and PSA levels over 4 ng/ml [21]. Prostate cancer detection rates up to 40 percent were detected using Doppler TRUS [22]. Doppler
334
J.J. Fütterer and J.R. Spermon
Fig. 14.1 Axial gray-scale transrectal ultrasound image of the prostate of a 55-year-old man (PSA level, 5.7 ng/mL, Gleason sum score, 7 and normal digital rectal examination). A hypoechoic lesion was observed in the right peripheral zone (arrows)
TRUS imaging resulted in a high inter-observer variability [23, 24] and wide variation in sensitivity and specificity of 27 percent to 92 percent and 46 percent to 84 percent, respectively. Contrast-enhanced TRUS – A new development is the application of gas-filled microbubble contrast agents (Fig. 14.2.). The microbubbles remain intravascular and, thus, act as blood pool agents. Disadvantages of using contrast agents are the longer duration and higher degree of invasiveness of the examination: however, the risk of hypersensitivity to the substance is rare. Contrast agent-specific imaging techniques have been developed to optimize microbubble signal reception while preserving the microbubbles. Until now, three studies directly compared systematic and contrast-enhancedtargeted TRUS biopsy [25-27]. These studies showed significantly higher positive biopsy core rates when directing biopsy, based on focal areas of contrast enhancement. Sensitivities and specificities of prostate cancer detection using contrast agents varied between 48 percent to 94 percent and 46 percent to 88 percent, respectively. Sonoelastography – A novel ultrasound technique that analyzes the compressional characteristics of prostate tissue is transrectal sonoelastography. In a recent study of 404 men undergoing biopsy based on real-time sonoelastography revealed a detection rate of 37.4 percent [28]. A drawback of the study was the heterogeneity of the population since more than half of the patients had already undergone one or more negative biopsy sessions. A study comparing real-time elastography with radical prostatectomy reported a localization sensitivity of 88 percent [29].
2.4
Computed Tomography (CT)
A study by Prando and Wallace revealed that contrast-enhanced CT scanning was able to detect only 58 percent of the 102 histologic prostate cancer sites documented
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
335
Fig. 14.2 Contrast-enhanced transrectal ultrasound image in contrast-harmonic mode. After a 2.4 ml bolus injection of microbubble contrast agent an area of enhancement in the right lateral peripheral zone (arrows) was visible and showed marked enhancement, compared with the rest of the peripheral zone. A symmetrical enhancement of the central gland (arrowheads) was observed
by TRUS-guided biopsies in 25 patients [30]. CT scanning has too little soft tissue contrast resolution to discern the subtle tissue changes due to prostate cancer. CT should not be used for prostate cancer detection and localization.
2.5
Magnetic Resonance Imaging (MRI)
Anatomical MRI – MRI of the prostate is performed using a combination of an endorectal and pelvic phased array coils. On T2-weighted MR images, in the peripheral zone normal prostate tissue appears as an intermediate to high signal intensity, while the central gland has lower signal intensity than the peripheral zone (Fig. 14.3.). Conversely, the prostate has a homogeneous, intermediate signal intensity on T1-weighted images. This means differentiation between the peripheral zone and central gland cannot be perceived. On MRI prostate cancer appears as an area of low signal intensity within the brighter, healthy peripheral zone using a T2-weighted sequence (Fig. 14.4.). In the central gland, prostate cancer is not as clearly discernable because the central gland generally has lower signal intensity than the peripheral zone, and it is more inhomogeneous due to BPH-induced architectural changes that may mimic prostate cancer. In addition to carcinoma, the differential diagnosis of an area of low signal intensity includes postbiopsy hemorrhage, prostatitis, BPH, effects of hormone or radiation treatment, scars, calcifications, smooth muscle hyperplasia and fibromuscular hyperplasia MRI plays no role as a screening imaging modality in patients with suspected prostate cancer. In patients with a prior negative TRUS-guided biopsy, T2-weighted
336
J.J. Fütterer and J.R. Spermon
Fig. 14.3 Normal prostate in a 28-year-old man. T2-weighted MRI image shows peripheral zone (PZ) with intermediate to high signal intensity. Small central gland (CG) has lower signal intensity than does the peripheral zone. The neurovascular bundle is located at the posterolateral aspect of the gland (curved arrow)
Fig. 14.4 55-year-old man (same patient as in Fig. 14.1) with stage T2a prostate cancer in the right peripheral zone. The T2-weighted MRI image shows that the tumor (arrows) has a lower signal intensity compared with the rest of the peripheral zone
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
337
MRI plays an important role. In this patient population an 83 percent sensitivity and a 50 percent positive predictive value for MRI have been established [31]. Proton MR spectroscopic imaging (MRSI) – provides quantitative metabolic data based on the citrate, choline and creatine levels, as well as their ratios. MRSI can be used for detection and localization of prostate cancer (Fig. 14.5) [32, 33]. The addition of MRSI to MRI increased the localization accuracy of MRI, particularly by raising specificity up to 91 percent [34]. However, a limitation of MRSI is its
Fig. 14.5 In (a), the position of the voxel of which spectrum (b) and (c) originate from is indicated. The axial T2-weighted image of this patient shows a low signal intensity in the right peripheral zone which is suspicious for prostate cancer. MRI spectra (b) from a voxel in healthy left peripheral zone (high level of citrate and normal low level of choline and creatine) and from a voxel (c) that contained prostate cancer (decreased level of citrate and increased level of choline and creatine)
338
J.J. Fütterer and J.R. Spermon
low spatial resolution. MRSI significantly increased the area under the receiver operating curve, from 0.68 with regular anatomical MRI to 0.80 [35]. Dynamic contrast-enhanced MRI (DCE-MRI) – DCE-MRI is a technique in which the contrast agent concentration is followed in time [36]. This technique is reported to be an effective tool in visualizing the pharmacokinetics of gadolinium uptake in the prostate [37-39]. Early contrast enhancement and high (relative) peak enhancement are the most accurate predictors of prostate cancer of the peripheral zone, while washout of the contrast agent and high permeability of the blood vessels are most sensitive for central gland prostate cancer [40, 41]. A recent study showed that the area under the receiver operating curve for localizing prostate cancer increased significantly, from 0.68 with anatomical T2-weighted MRI, to 0.91 by applying contrast agent (Fig. 14.6) [35].
Fig. 14.6 MR images of the prostate of 65-year-old man with prostate cancer (prostate-specific antigen level, 8.4 ng/mL, Gleason sum score, 6 and normal digital rectal examination). (a) Axial T2weighted MRI image through the prostate shows a low signal intensity lesion in the left peripheral (arrows). (b–d) Pharmacokinetic maps of calculated Ktrans (b) and kep (c) showing increased levels of Ktrans and kep in the left peripheral zone. (d) Pharmacokinetic map shows a negative wash-out area (red) in the left peripheral zone. Histopathology after radical prostatectomy revealed a T2a tumor in the left peripheral zone
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
2.6
339
Positron Emission Tomography (PET)
Positron emission tomography – the utility of PET scanning with 18-fluorine-labelled deoxyglucose (18FDG) in detecting prostate cancer is compromised by the relatively low uptake of 18FDG by prostate cancer cells [42], and significant overlap with marker uptake by benign prostatic hyperplasia, fibrosis and inflammation. Generally, 18 FDG-PET is not recommended for evaluation of the prostate due to sensitivities as low as 4 percent to 64 percent, with a specificity of 50 percent [43-45]. Another tracer – carbon-11 labelled choline (11C-choline) – accumulates in prostatic cells and has the advantage that, unlike 18FDG, it is not excreted via the urinary tract, and thereby does not interfere with the visualization of the prostate (Fig. 14.7). Furthermore, the prostate is the only organ in the pelvis to accumulate 11 C-choline. The 11C-choline uptake was higher in prostate cancer, compared with benign prostatic hyperplasia, but the difference was not statistically significant [46]. Drawbacks are the high costs of 11C-choline and its short, 20-minute half-life.
Fig. 14.7 11C-choline PET-CT image in a 58-year-old patient. 11C-choline uptake is visible in the right central gland which corresponded with a local prostate cancer
340
2.7
J.J. Fütterer and J.R. Spermon
Staging of Prostate Cancer
Clinical staging of prostate cancer currently entails the use of digital rectal examination, PSA as well TRUS. It is now common practice for clinicians treating prostate cancer patients to employ nomograms to determine therapeutic options [47-49]. The most frequently used nomogram, the Partin tables, estimates the chance of organ-confined disease, capsular penetration, seminal vesicle invasion and lymph node metastasis, based on the results of the traditional triad of digital rectal examination, biopsy Gleason score and PSA value [50]. The clinical stage is identified using these variables and is expressed in the TNM staging classification (Table 14.1) [51]. The current general opinion is that localized prostate cancer can be treated successfully by radical prostatectomy or radiation therapy. Nevertheless, advantages of aggressive treatment over watchful waiting in terms of quality-adjusted life expectancy are often small, leading to controversies about the adequate treatment.
Table 14.1 TNM Staging Classification of Prostate Cancer [51] Stage Primary Tumor TX T0 T1 T1a T1b T1c T2 T2a T2b T2c T3 T3a T3b T4
Primary tumor cannot be assessed No evidence of primary tumor Clinically the tumor is neither palpable or visible with imaging Tumor is an incidental histologic finding in 5 percent or less of tissue resected Tumor is an incidental histologic finding in > 5 percent of tissue resected Tumor identified with needle biopsy (e.g., because of an elevated PSA) Tumor confined within the prostate Tumor involves one-half of one lobe or less Tumor involves more than one-half of one lobe, but not both lobes Tumor involves both lobes Tumor extends through the prostate capsule Extra-capsular extension (unilateral or bilateral) Tumor invades seminal vesicle(s) Tumor is fixed or invades adjacent structures other than seminal vesicles: bladder neck, external sphincter, rectum, levator muscles and/or pelvic wall
Regional Lymph Nodes NX N0 N1
Regional lymph nodes were not assessed No regional lymph node metastasis Metastasis in regional lymph node(s)
Distant Metastasis MX M0 M1 M1a M1b M1c
Distant metastasis cannot be assessed (not evaluated with any modality) No distant metastasis Distant metastasis Non-regional lymph node(s) Bone(s) Other site(s) with or without bone disease
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
341
Clinical assessment by digital rectal examination and PSA level are not accurate in determining local stage, with underestimations in as many as 40 percent to 60 percent of cases [52, 53]. Accurate staging with additional imaging techniques is, therefore, an important issue for correct management of prostate cancer patients. TRUS – TRUS may enable correct assessment of locally advanced tumors, but it is not sensitive enough to detect initial extraprostatic extension across the capsule or into the seminal vesicles in clinically confined lesions [54, 55]. Any change of the prostatic capsule, like bulging or irregularity, adjacent to a hypoechoic lesion is suspicious of extracapsular extension. Accuracies of gray-scale TRUS in determining the local disease stage varied from 58 percent to 83 percent, with sensitivities and specificities ranging from 33 percent to 76 percent and 46 percent to 91 percent, respectively [54, 56-59]. Three-dimensional TRUS aids in assessing local disease extension [58]. Duplex Doppler TRUS and contrast-enhanced TRUS are new methods to study tumor vascularity. These blood flow-enhancing TRUS techniques have the potential to improve the local staging of prostate cancer [26]. Future research will indicate their exact role. CT – Few recent studies have been published on role of CT for staging prostate cancer [60–63]. A pre-radiation therapy staging study of 85 patients showed that CT staging had only a marginal effect on treatment decisions [60]. CT has no use in assessing clinically confined lesions [61]. Two other studies revealed low sensitivity of 26 percent to 29 percent, and specificity of 80 percent to 89 percent [62, 63]. MRI – A large number of studies have been performed over the last two decades to show the accuracy of MRI in local staging of the prostate. Two meta-analyses on local staging by MRI found combined maximum sensitivities and specificities of 71 percent to 74 percent, while sensitivity was 62 percent to 69 percent at a specificity of 80 percent [64, 65]. T2-weighted MRI in more than one plane, as well as utilizing an endorectal coil, resulted in significantly better staging performance. The use of endorectal-pelvic phased array coils is recommended. Significant improvement of anatomic details, extracapsular extension accuracy and specificity was found when an endorectal-pelvic phased-array coil is used [53]. MRI should be performed at least four weeks after prostatic biopsy. T1-weighted sequence should be acquired for evaluation of post-biopsy hemorrhage. The most reliable criteria for the detection of extracapsular extension of prostate carcinoma are asymmetry of the neurovascular bundle (Fig. 14.8), obliteration of the rectoprostatic angle and tumor bulge into the periprostatic fat (Fig. 14.9) [66]. Seminal vesicles on T2-weighted images appear as tubular structures with thin hypointense walls and filled with hyper-intense fluid. The diagnosis of seminal vesicle invasion is made when focal or diffuse thickening (hypo-intense) of the tubular walls, associated with focal hypo-intense luminal lesions, is present (Fig 14-10). The most cost-effective patient group to undergo local staging with endorectal MRI are those considered to have an intermediate risk of T3 disease, based on PSA level (between 4 to 20 ng/mL), and a Gleason score of five to seven [67]. Jager, et al. developed a decision analysis model that supported the position that MRI in the preoperative work-up of prostate cancer is cost-effective in patients with a moderate to high chance of extra-capsular disease, and should be performed with an emphasis on
Fig. 14.8 60-year-old man with stage T3a disease (T) in the left peripheral zone and central gland. T2-weighted MRI image shows invasion of the neurovascular bundle (curved arrow). Obliteration of the left rectoprostatic angle (arrow), but the right neurovascular bundle and rectoprostatic angle are intact
Fig. 14.9 51-year-old man with stage T3a disease in the right peripheral zone. T2-weighted MRI image shows that the tumor (T) has lower signal intensity than the normal peripheral zone and shows bulging (arrows) and broad surface contact with capsule
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
343
Fig. 14.10 Seminal vesicle invasion in 58-year-old patient. Axial T2-weighted MRI image through seminal vesicles shows a low signal intensity within the lumen of the seminal vesicles and thickening of the tubular walls
achieving high specificity [68]. Langlotz, et al. emphasized the need of high-specificity reading in prostate MRI to ensure that as few patients as possible are unnecessarily denied potential curative therapy because of false positive MRI results [69]. A substantial improvement in overall staging accuracy of endorectal MRI can be achieved by careful pathologic correlation and by considering the anatomic features of prostate cancer. A prospective study of 103 patients revealed a significant improvement in staging performance for the less experienced reader using multislice dynamic contrast-enhanced MRI [70]. Also, the addition of three-dimensional MRSI to MRI improved staging accuracies, particularly for less experienced readers [71]. Imaging at higher magnetic field strengths (e.g., 3 tesla) results in increased anatomical resolution. Two recent studies on local staging with 3T MRI reported a sensitivity and specificity of 80 percent to 88 percent and 94 percent to 100 percent, respectively [72, 73]. Positron emission tomography – The role of 18FDG-PET in local staging is very limited due to this technique’s low spatial resolution and the low uptake within the primary tumor [74].
344
2.8
J.J. Fütterer and J.R. Spermon
Lymph Node Staging
Pelvic lymph node metastases have a significant effect on the prognosis of patients with malignancies. One positive lymph node can turn prostate cancer from a local to a systemic disease unsusceptible to curative treatment [75, 76]. Surgical open pelvic lymph node dissection with histopathological examination is currently the most reliable method of assessing lymph node status. Abdominal ultrasound plays no role in this phase of staging [77]. Routine crosssectional imaging modalities, such as CT and MRI, have a limited sensitivity in identifying metastases [78-80]. CT and MRI interpretation of lymph nodes is essentially based on size and shape criteria. These techniques only use the size (8 to 10 mm) and shape (round – oval) criteria and, therefore, are limited [80]. In a study of 80 patients, Harisinghani, et al. found a 35 percent sensitivity and 90 percent specificity of detection of positive lymph nodes using anatomical MRI node-by-node [81]. MR lymphangiography (MRL) uses intravenously administered lymphotropic ultrasmall superparamagnetic iron-oxide (USPIO) particles (Ferumoxtran-10) with a long plasma circulation time and is a novel, non-invasive cellular imaging tool for the evaluation of nodal involvement. MRL is an accurate tool to differentiate benign from malignant lymph nodes [82, 83]. Post-ferumoxtran-10 MRI exam includes both a sequence which is insensitive for iron using T1- or proton-weighted turbo spin echo sequences, and a sequence which is sensitive for iron (Fig. 14.11). For the latter purpose, a good sequence is a high resolution T2-weighted gradient echo sequence. Ferumoxtran-10-enhanced MRI achieved a 97.3 percent accuracy with high sensitivity (90.5 percent) and specificity (97.8 percent) on a node-by-node basis [81]. Harisinghani, et al. achieved a sensitivity of 100 percent and a specificity of 96 percent for detection of 5 to 10 mm nodes with 1.5T MRI. However, when the metastatic lymph node was smaller than 5 mm, this sensitivity dropped to 41 percent. Ferumoxtran-10-enhanced MRI at a 3T field strength using a higher spatial resolution with improvement of image quality may allow detection of small metastatic nodes (<5 mm) in the future [84]. Although very promising in metastatic lung cancer, the role of 18FDG-PETscanning is limited in the urinary tract region, as 18F-fluorodeoxyglucose accumulates in the urinary bladder and kidneys. This makes an evaluation of metastases at these sites difficult. In prostate cancer this method is further limited by its low uptake in metastatic nodes. Although the sensitivity of 18FDG-PET is slightly better (67 percent), compared to those of CT and unenhanced MRI, this value is, however, not high enough to replace pelvic lymph node dissection (Fig. 14.12) [85].
2.9
Metastatic Bone Disease
The first diagnostic test to detect or exclude bone metastases is the technetium99m-diphosphonate bone scintigraphy (Fig 14.13). A meta-analysis of 23 prostate cancer studies deduced detection rates of 2.3 percent, 5.3 percent and 16.2 percent for patients with PSA levels below 10 ng/mL, between 10 and 19.9 ng/mL
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
345
Fig. 14.11 Normal size lymph node (8 mm) in the right obturator fossa in a 60-year-old male with biopsy proven prostate cancer (PSA 15.3 ng/mL; Gleason score 7). On post ferumoxtran-10 T2weighted gradient echo MRI image (which is iron sensitive) this normal sized node remains white (circle). On histopathology this node was completely metastatic
and between 20 to 49.9 ng/mL, respectively [86]. In a large study it was found that, in patients with levels below 20 ng/mL, the false negative rate was less than 2 percent [87]. Bone scintigraphy lacks specificity and, thus, primary skeletal diseases may cause false positive findings. X-ray can be used to exclude false positive findings on bone scintigraphy due to conditions such as trauma, degenerative joint disease or other chronic diseases. Conventional X-rays are too insensitive for the detection of metastatic bone lesions. Most metastatic bone lesions are sclerotic [88]. A 50 percent change in bone mineral density is needed for metastatic bone lesions to be visible on X-ray images [89]. CT has no place in determining metastatic bone disease. The high spatial resolution and excellent soft tissue contrast make MRI an ideal tool for the detection of osseous lesions. Whole-body MRI appears to be a very sensitive tool to determine bone marrow metastases and, in less than 15 minutes, is feasible for tumor staging [90]. Advantages of MRI are the absence of radiation exposure, as well as the ability to also detect non-skeletal metastases.
346
J.J. Fütterer and J.R. Spermon
Fig. 14.12 11C-choline PET-CT image of the same patient as in Fig. 14.11. The metastatic node in right obturator fossa demonstrated uptake of 11C-choline. This is indicative of lymph node metastasis
Purely sclerotic lesions take up 18F-FDG less avidly, compared to purely lytic or mixed metastases. Bone metastases in prostate cancer are commonly sclerotic lesions. 18F-FDG-PET is considered to be inferior to bone scintigraphy [91]. An important role of PET imaging may lie in its early ability to detect treatment response in patients with metastatic disease who are receiving chemotherapy.
2.10
Conclusions
TRUS remains the primary imaging tool for the detection of prostate cancer and for guiding prostate biopsy. Functional MRI achieves high localization rates. MRI of the prostate can be used as a problem-solving tool in patients with rising PSA and repetitive negative biopsies. MRI at 1.5T, using an endorectal coil combined with a pelvic phased-array coil, is currently the optimal choice for determining the local disease stage in prostate cancer patients. Dynamic contrast-enhanced MRI and MR spectroscopic imaging may be used to increase the staging accuracy for less experienced readers. The
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
347
Fig. 14.13 Technetium-99 m-diphosphonate bone scintigraphy in a patient with bone metastases (PSA 82 ng/mL). Anterior and posterior whole body delayed planar imaging was performed. Planar bone scan imaging demonstrates a substantial focus of increased uptake at the level of thoracic spine level 2/3/4, as well as at the left aspect of the approximate T7 and T12. There is also moderate increased focal uptake at L3, the right sacroiliac joint, sacrum, as well as the level of the fifth (left) and seventh (right) rib
role of PET in local staging is limited. MR lymphangiography using ultrasmall superparamagnetic iron-oxide particles is the most sensitive and specific method for detecting lymph node metastases. Bone scintigraphy is the most sensitive method for detection of bone metastases. However, FDG-PET and whole body MRI are promising modalities for detection and assessment of response to therapy.
Key Points ●
●
The urologic work-up in patients with elevated PSA consists of a digital rectal examination and transrectal ultrasound. Transrectal ultrasound is the primary imaging tool for the detection of prostate cancer.
348 ●
●
●
●
●
J.J. Fütterer and J.R. Spermon
More than 70 percent of adenocarcinoma of the prostate arises in the peripheral zone, whereas about 20 percent emerge in the transitional zone and 10 percent in the central zone. Computed tomography should not be used for prostate cancer detection, localization and staging. Dynamic contrast-enhanced MRI and MR spectroscopic imaging can be used for localization of prostate cancer, and to increase the staging accuracy for less experienced readers. The optimal sensitivity and specificity for the detection of lymph node metastases is achieved by using MRI using ultra-small superparamagnetic iron-oxide particles. Bone scintigraphy remains the single most sensitive method of detecting bone metastases.
2.11
Testicular Cancer
Testicular germ cell cancer accounts for only 1 percent of all cancer in males [92]. The peak prevalence occurs between 25 and 35 years of age. The incidence of testicular cancer has doubled in the last 40 years. Bilateral tumors are found in 0.7 percent of men with germ cell tumors at diagnosis, and 1.5 percent of patients develop metachronous lesions within five years [93]. Once the leading cause of cancer death in men between 15 and 35 years of age, it has now proved to be a model of success with a five-year survival rate exceeding 95 percent [94]. This success in treatment is related to improved staging and treatment methods. Imaging plays a central role in assessment of tumor bulk, sites of metastases, monitoring response to therapy, surgical planning and accurate assessment of disease at relapse [95].
2.12
Clinical Symptoms of Testicular Cancer
The most common sign of testicular germ cell cancer is a painless unilateral mass in the scrotum, which is inseparable from the testis (up to 95 percent of cases)[96]. In 20 percent of the cases, the first symptom is scrotal pain, followed by back and flank pain in 11 percent. Gynecomastia appears in 7 percent of the cases [97]. Although differential diagnosis must be established with any other intrascrotal mass or disease, any scrotal complaint at a young age needs to be thoroughly investigated to rule out testicular germ cell cancer.
2.13
Pathology of Testicular Germ Cell Cancer
Ninety-five percent of all testicular tumors are germ cell tumors (Fig. 14.14). The remaining are lymphomas (4 percent) and Leydig or Sertoli cell tumors. Testicular
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
349
Fig. 14.14 (a) Seminomatous germ cell cancer. (b) Histopathology of normal testis (left) with adjacent seminomatous germ cell cancer (right)
germ cell tumors are derived from spermatogenic cells and may be classified as unipotential or totipotential. Unipotential tumors are seminomas, which comprise 35 percent to 50 percent of all germ cell tumors. Nonseminomatous germ cell tumors are considered to be totipotential. Serum tumor markers are especially helpful to differentiate germ cell tumors from each other and from other malignancies. Serum concentrations of alpha-fetoprotein (AFP) and/or beta-human chorionic gonadotropin (beta-hCG) are elevated in 80 percent to 85 percent of non-seminomas. In contrast, serum beta-hCG is elevated in fewer than 25 percent of testicular seminomas, and AFP is not elevated in pure seminomas. However, these tumor markers cannot accurately assess disease bulk or locate sites of tumor spread [95].
2.14
Diagnosis of Testicular Cancer
The first step in diagnosing testicular cancer is usually through self-examination. Testicular ultrasound is used to confirm the presence of a testicular mass (Fig. 14.15), to distinguish from other scrotal abnormalities and to explore the contralateral testis [98-102]. The sensitivity of testicular ultrasound in detecting testicular tumors is almost 100 percent [103]. Furthermore, ultrasound is almost 100 percent sensitive in differentiating intratesticular from extratesticular lesions [102, 104, 105] and is able to detect microlithiasis (Fig. 14.16). Extratesticular lesions are commonly benign in adults, whereas in children these lesions are often malignant [106]. Microlithiasis should be cautiously followed up, since it can be associated with testicular germ cell cancer [107]. Serum tumor markers contribute to the diagnosis. At the time of the diagnosis a chest radiography is used to evaluate the mediastinum for lymphadenopathy, and the lungs for haematogenous metastases. After a testicular tumor has been clinically diagnosed, the inguinal ablation of the testis is indicated (Fig. 14.17).
350
J.J. Fütterer and J.R. Spermon
Fig. 14.15 Testicular ultrasound of a testis with increased vascular flow at the cancerous part
Fig. 14.16 The appearances of testicular microlithiasis (‘snow storm’ appearance on ultrasound)
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
351
Fig. 14.17 Right-sided inguinal orchiectomy. The incision is high inguinal and the whole spermatic cord should be removed up to the internal ring with separation of ductus deferens and gonadal vessels
2.15
Staging of Testicular Cancer
As soon as the diagnosis of germ cell cancer has been pathologically confirmed, further staging examinations are warranted to examine the extent of disease. Staging is of utmost importance as it is the cornerstone for further treatment after orchiectomy. The European Germ Cell Cancer Consensus Group (EGCCCG) recommends that TNM staging be used [51] (Table 14.2). The most commonly used staging system in Europe for dissemination is the Royal Marsden Hospital Classification system (Table 14.3) [108]. Today, computerized tomography (CT) of the abdomen and chest is the standard technique in initial staging. The most common sites for metastases are via the lymphatic system to the retroperitoneal nodes, and via the hematogenous route to the lungs and, less commonly, to the liver, brain and bone. In general, advanced stage disease will be treated primarily with chemotherapy. Nonseminoma germ cell tumors appear as multiple small peripheral nodules, whereas seminoma metastases tend to be larger masses [95]. Other sites of hematogenous metastases, though rarely seen and usually only in the setting of advanced disease, include the adrenals, kidneys, spleen, pleura, pericardium and peritoneum [95]. Lymphatic spread occurs via lymphatic channels (from spermatic cord and testicular vessels to retroperitoneal lymph nodes). Usually, right-sided testicular neoplasms spread to the right side of the retroperitoneum. Lymph node metastases can be
352
J.J. Fütterer and J.R. Spermon
Table 14.2 TNM Staging Classification of Testicular Tumors [51] Stage Primary Tumor The extent of the primary tumor is classified after radical orchidectomy (pT) pTX Primary tumor cannot be assessed (if no radical orchidectomy has been performed, TX is used) pT0 No evidence of primary tumor (e.g., histological scar in testis) pTis Intratubular germ cell neoplasia pT1 Tumor limited to testis and epididymis without vascular/lymphatic invasion; tumor may invade into the tunica albuginea, but not the tunica vaginalis pT2 Tumor limited to testis and epididymis with vascular/lymphatic invasion, or tumor extending through tunica albuginea with involvement of tunica vaginalis pT3 Tumor invades spermatic cord with or without vascular/lymphatic invasion pT4 Tumor invades scrotum with or without vascular/lymphatic invasion Regional Lymph Nodes Clinical Involvement NX N0 N1 N2
N3
Regional nodes cannot be assessed No regional lymph node metastasis Metastasis with a lymph node mass ≤2 cm in greatest dimension or multiple lymph nodes none >2 cm in greatest dimension Metastasis with a lymph node mass >2 cm, but <5 cm in greatest dimension, or multiple lymph nodes, any one mass >2 cm, but ≤5 cm in greatest dimension Metastasis with a lymph node mass >5 cm in greatest dimension
Pathological Involvement pN0 pN1 pN2
pN3 Distant Metastases MX M0 M1 M1a M1b
No regional lymph node metastases Metastasis with a lymph node mass ≤2 cm in greatest dimension and five or fewer positive nodes, none >2 cm in greatest dimension Metastasis with a lymph node mass >2 cm, but ≤5 cm in greatest dimensions; or more than five nodes positive, none >5 cm; or evidence of extranodal extension of tumor Metastasis with a lymph node mass >5 cm in greatest dimension Distant metastasis cannot be assessed No distant metastasis Distant metastasis Non-regional lymph node or pulmonary metastasis Distant metastasis other than to non-regional lymph nodes and lungs
identified around the inferior vena cava, and between the level of right renal hilum and the aortic bifurcation. Lymph node metastases of left-sided testicular cancer may be found adjacent to the abdominal aorta and just below the left renal vein. Contralateral involvement is uncommon, but may occur with a larger disease burden [109]. Pelvic lymphadenopathy is uncommon in the absence of bulky disease [110].
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
353
Table 14.3 The Royal Marsden Hospital Classification System for Germ Cell Tumors Stage I Tumor limited to testis
Stage II IIA IIB IIC IID
Infradiaphragmatic lymph node involvement Metastases <2 cm in diameter Metastases 2 to 5 cm in diameter Metastases 5 to 10 cm in diameter Metastases >10 cm in diameter
Stage III A-D
Supradiaphragmatic lymph node involvement See stage II
Stage IV
Extralymphatic involvement of lung (L), liver (H), brain and bone
2.16
Ultrasound
Testicular ultrasound (linear 6 to 12 MHz probe) is performed in at least two planes. The homogenous, low-to-medium echogenicity of the testicle noted in boys increases after puberty [111]. Testicular tumors are usually well defined and hypoechoic relative to the normal testicle. Some testicular tumors may show a heterogeneous echotexture, calcification or cystic change. Color and power Doppler ultrasound may be helpful in delineating areas of malignant involvement, but this is not specific and may not be demonstrated in small tumors [112]. If a malignantappearing mass is encountered, sonography of the retroperitoneum may identify associated lymphadenopathy [113].
2.17
Computed Tomography
Computed tomography (CT) is used for staging metastatic disease and for follow-up after therapy in patients with disseminated disease. The abdominal CT examination offers a sensitivity of 30 percent to 35 percent in the evaluation of retroperitoneal lymph nodes in the landing zone by using a threshold of 1 cm. Contrast-enhanced
354
J.J. Fütterer and J.R. Spermon
CT of the thorax, abdomen and pelvis is recommended according to the guidelines of the EGCCCG. CT of the brain is only performed in patients with suspected disease and patients with high risk factors for metastases. CT is limited in distinguishing residual tumors from hematoma, fibrosis and/or necrosis [113].
2.18
Magnetic Resonance Imaging
Magnetic resonance imaging (MRI) can be used as a problem-solving tool in inconclusive ultrasound cases. MRI is performed in supine positioning and surface coils (phased-array) are positioned above the testicles. T1- and T2-weighted sequences in at least two planes are acquired. Dynamic contrast-enhanced subtraction MRI can be used to differentiate testicular diseases from scrotal disorders [114]. The normal testicle has an intermediate homogenous signal intensity on T1weighted images, and homogeneous high signal intensity (less than fluid signal intensity) on T2-weighted images. Signal intensity of the epididymis is low signal intensity on both T1- and T2-weighted images. The tunica albuginea and testicular septa appear as low signal intensity structures [115]. Testicular neoplasms present with low signal intensity on T2-weighted images and intermediate to low signal intensity on T1-weighted images. MRI cannot predict the histological type [116].
2.19
Positron Emission Tomography (PET)
Examining the role of 18-fluorine-labelled deoxyglucose 18F-FDG-PET in testicular germ cell cancer (Fig. 14.18) shows a sensitivity of 82 percent, a specificity of 94 percent, and a negative predictive value of 94 percent [117-120]. However, lymph node metastases smaller than 1 cm can be missed with 18F-FDG-PET. Seminomatous germ cell tumors have a significantly higher uptake of FDG, compared to nonseminomatous lesions. The role of 18F-FDG-PET in primary staging is minimal if metastatic disease has already been diagnosed [95]. 18F-FDG-PET is of incremental value in assessing residual disease or recurrence [121].
2.20
Imaging in Clinical Stage I Testicular Cancer
Patients with clinical stage I tumors have disease that is confined to the testis. However, approximately 30 percent of the clinical stage I non-seminomas are understaged by radiological imaging, and are found to have metastatic disease at retroperitoneal surgery [122]. The abdominal CT scan offers a sensitivity of 30 percent to 35 percent in the evaluation of retroperitoneal lymph nodes in the landing zone by using a threshold
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
355
Fig. 14.18 Patient with a nonseminomatous testicular cancer. Through clinical imaging a lesion of 9 mm was found on CT. Increased uptake of 18FDG suggests the presence of a retroperitoneal metastasis
of 1 cm. Lowering this threshold results in an increased sensitivity, but a decreased specificity (with a criterion of 4 mm, the sensitivity increases to 93 percent, but the specificity decreases to 58 percent) [123]. New generation CT scans do not seem to improve the sensitivity [124]. Although pulmonary involvement rarely occurs in the absence of retroperitoneal disease, a chest X-ray is mandatory and the preferred imaging modality. Routine CT of the chest, although highly sensitive, produces a significant number of false positive scans (detecting 2 mm sized lesions, but 70 percent of those are benign) [125]. Alternative imaging methods like PET and MRI add little to the management of clinical stage I non-seminoma germ cell cancer. The accuracy of MRI is in line with CT examination [126-127]. Currently, the additional value of intravenous ferumoxtran10 administration before MRI has been evaluated. Ferumoxtran-10 is an ultrasmall nanoparticle given intravenously, which moves into the reticulo-endothelial system. Benign nodes only take up ferumoxtran-10, leaving the cancerous lymph nodes without enhancement. Ferumoxtran-10-enhanced MRI yields a higher sensitivity and specificity when compared with unenhanced MRI (sensitivity: 88.2 percent vs 70.5 percent, specificity: 92 percent vs 68 percent). Although the results are encouraging, the precise role of this tool in clinical stage I testicular germ cell cancer remains to be determined [128]. The old-fashioned method of imaging lymph nodes through lymphangiography has gained new interest via new contrast agents. Lymphangiography allows visualization of the three main lymphatic channels (paracaval, interaortacaval and paraaortal, Fig. 14.19). The major goal of the new contrast agents is to investigate the feasibility and accuracy of radio-guided mapping of sentinel lymph nodes (SLNs) in clinical
356
J.J. Fütterer and J.R. Spermon
Fig. 14.19 Landing zone for retroperitoneal metastases of testicular germ cell cancer. In patients with right-sided tumors (a), the limits of dissection for the modified nerve-sparing template include the right ureter, the renal veins, the right lateral wall of the aorta, the inferior mesenteric artery and the iliac bifurcation. For left-sided tumors (b), the limits of dissection are the left ureter, left renal vein, left mid-wall of vena cava, the inferior mesenteric artery and iliac bifurcation
Stage I testicular tumors. For a left-sided testicular tumor the primary landing zone (e.g., SLN) includes the nodes in the para-aortic region below the renal vessels and the ipsilateral lateral distribution of the para-aortic, pre-aortic and left common iliac nodes. For right-sided tumors the primary landing zone is in the interaortacaval region below the renal vessels and the ipsilateral lymph nodes in the paracaval, preaortic and right common iliac region. Satoh, et al. injected (99 m) Technetium-labeled phytate around the testicular tumor in 22 patients. In 21 of them the SLN was detected by laparoscopic retroperitoneal lymph node dissection. Only in two patients were micrometastases found in the SLN. Both patients were free of disease after adjuvant chemotherapy [129]. As in two other patients lymph node relapses were detected, the real value of radio-guided mapping of SLNs with laparoscopy can be questioned. In clinical stage I seminomas approximately 15 percent of patients have subclinical metastatic disease [130]. In accordance with nonseminomas, FDG-PET and MRI provide no additional value above CT scan. [118, 128, 131]
2.21
Imaging in Advanced Stage Testicular Germ Cell Cancer
The most common sites for metastases are via the lymphatic system to the retroperitoneal nodes, and via the hematogenous route to the lungs and, less commonly,
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
357
Fig. 14.20 Coronal 18FDG-PET scan shows metastases of nonseminomatous testicular germ cell tumor in the retroperitoneum, and in lungs with increased uptake of 18FDG (arrow). The lesion in the retroperitoneum shows no uptake in the center. The patient showed partial radiological response during treatment: both decrease in 18FDG uptake and volume reduction on CT scan retroperitoneally, and disappearance of lung metastases. Surgery of residual retroperitoneal mass showed necrotic and teratomatous tissue in the center, and inflammatory tissue at the rim of the retroperitoneal mass
to the liver, brain and bone. In general, advanced stage disease will be treated primarily with chemotherapy. Today, CT is the standard in initial staging. Though FDG-PET has the potential to improve clinical staging, more studies are warranted to establish its definitive value [131, 132]. Following completion of chemotherapy, residual tumorous lesions are found in up to 15 percent of patients with seminomas [133], compared to 20 percent of patients with non-seminomas [134]. Furthermore, 40 percent of the nonseminomatous residual masses contain mature teratoma (pre-malignant disease). The key to success is complete surgical removal of these masses. A major challenge is finding the optimal method for differentiating patients with post-chemotherapy (pre-) malignant residual masses from those with fibrotic lesions. Again, CT and the change in size of the mass has been the standard for assessing residual masses. PET is of incremental value in assessing residual seminomatous disease. A study of 56 scans by De Santis, et al. reveals that PET had a sensitivity, specificity, positive predictive value and negative predictive value of 100 percent, 80 percent, 100 percent and 96 percent, respectively, versus 74 percent, 70 percent, 37 percent and 90 percent for CT [121]. In contrast, in nonseminomas there is no real additional value as PET cannot differentiate between fibrosis and mature teratoma (Fig. 14.20) [119].
2.22
Follow-up of Testicular Germ Cell Cancer
Because most recurrences after curative therapy will occur in the first two years, follow-up should be most frequent and intensive during this time. Follow-up protocols
358
J.J. Fütterer and J.R. Spermon
vary by institution and by type, stage and treatment of the primary disease. After treatment all patients receive follow-up care through regular outpatient visits, during which physical examination, serum tumor markers, chest X-ray and CT scans are performed. Currently, efforts are made to optimize the follow-up schedule. [135,136]
2.23
Conclusions
Ultrasound is the initial investigative tool with regard to scrotal masses. Patients should undergo a CT examination of the chest, abdomen and pelvis when the histological diagnosis of testicular germ cell cancer has been confirmed. Clinical staging is hampered by the inability to detect micrometastatic disease because the sensitivity of conventional imaging studies is inversely proportional to tumor volume. To date, the metabolic tracer imaging studies have no additional value because the micrometastases do not show enough metabolic activity for detection. In re-assessing the extent of metastastic disease after chemotherapy, CT scan remains the first choice of imaging. PET can contribute to the management of residual seminoma lesions.
References 1. Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 2007. CA Cancer J Clin 2007; 57(1); 43-66. 2. Carter HB, Piantadosi S, Isaacs JT. Clinical evidence for and implications of the multi-step development of prostate cancer. J Urol 1990; 143:742-746. 3. Parkin DM, Bray FI, Devesa SS. Cancer burden in the year 2000. The global picture. Eur J Cancer 2001; 37 Suppl 8:S4-66. 4. Konety BR, Bird VY, Deorah S, Dahmoush L. Comparison of the incidence of latent prostate cancer detected at autopsy before and after the prostate specific antigen era. J Urol 2005; 174:1785-1788. 5. McNeal JE. Normal anatomy of the prostate and changes in benign prostatic hypertrophy and carcinoma. Semin Ultrasound CT MR 1988; 9:329-334. 6. Coakley FV, Hricak H. Radiologic anatomy of the prostate gland: a clinical approach. Radiol Clin North Am 2000; 38:15-30. 7. Sattar AA, Noël J-C, Vanderhaeghen J-J, Schulman CC, Wespes E. Prostate capsule: computerized morphometric analysis of its components. Urology 1995; 46:178-181 8. Greenhalgh R, Kirby RS. Anatomy and physiology of the prostate and benign prostatic hyperplasia. Atlas Urol Clin 2002; 10:1-9 9. Richie JP, Catalona WJ, Ahmann FR, et al. Effect of patient age on early detection of prostate cancer with serum prostate-specific antigen and digital rectal examination. Urology 1993; 42:365-374. 10. Schröder FH, van der Maas P, Beemsterboer P, et al. Evaluation of the digital rectal examination as a screening test for prostate cancer. Rotterdam section of the European Randomized Study of Screening for Prostate Cancer. J Natl Cancer Inst. 1998; 90:1817-1823.
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
359
11. Philips TH, Thompson IM. Digital rectal examination and carcinoma of the prostate. Urol Clin North Am. 1991; 18:459-465. 12. Freedland SJ, Mangold LA, Epstein JI, Partin AW. Biopsy indicator –a predictor of pathologic stage among men with preoperative serum PSA levels of 4.0 ng/mL or less and T1c disease. 2004; 63(5):887-891. 13. Lee F, Gray JM, McLeary RD, et al. Prostatic evaluation by transrectal sonography: criteria for diagnosis of early carcinoma. Radiology 1986; 158:91-95. 14. Meirelles LR, Billis A, Cotta AC, Nakamura RT, Caserta NM, Prando A. Prostatic atrophy: evidence for a possible role of local ischemia in its pathogenesis. Int Urol Nephrol 2002; 34:345-350. 15. Ellis WJ, Brawer MK. The significance of isoechoic prostatic carcinoma. J Urol 1994; 152:2304-2307. 16. Heijmink SW, van Moerkerk H, Kiemeney LA, Witjes JA, Frauscher F, Barentsz JO. A comparison of the diagnostic performance of systematic versus ultrasound-guided biopsies of prostate cancer. Eur Radiol 2006; 16:927-938. 17. Hodge KK, McNeal JE, Terris MK, Stamey TA. Random systematic versus directed ultrasound guided transrectal core biopsies of the prostate. J Urol 1989; 142:71-74. 18. Gore JL, Shariat SF, Miles BJ, et al. Optimal combinations of systematic sextant and laterally directed biopsies for the detection of prostate cancer. J Urol 2001; 165:1554-1559. 19. Scherr DS, Eastham J, Ohori M, Scardino PT. Prostate biopsy techniques and indications: when, where, and how? Semin Urol Oncol 2002; 20:18-31. 20. Djavan B, Remzi M, Schulman CC, Marberger M, Zlotta AR. Repeat prostate biopsy: who, how and when?. a review. Eur Urol 2002; 42:93-103. 21. Wilson NM, Masoud AM, Barsoum HB, Refaat MM, Moustafa MI, Kamal TA. Correlation of power Doppler with microvessel density in assessing prostate needle biopsy. Clin Radiol 2004; 59:946-950. 22. Okihara K, Kojima M, Nakanouchi T, Okada K, Miki T. Transrectal power Doppler imaging in the detection of prostate cancer. BJU Int 2000; 85:1053-1057. 23. Rifkin MD, Sudakoff GS, Alexander AA. Prostate: techniques, results, and potential applications of color Doppler US scanning. Radiology 1993; 186:509-513. 24. Lavoipierre AM, Snow RM, Frydenberg M, et al. Prostatic cancer: role of color Doppler imaging in transrectal sonography. AJR Am J Roentgenol 1998; 171:205-210. 25. Frauscher F, Klauser A, Volgger H, et al. Comparison of contrast enhanced color Doppler targeted biopsy with conventional systematic biopsy: impact on prostate cancer detection. J Urol 2002; 167:1648-1652. 26. Pelzer A, Bektic J, Berger AP, et al. Prostate cancer detection in men with prostate specific antigen 4 to 10 ng/ml using a combined approach of contrast enhanced color Doppler targeted and systematic biopsy. J Urol 2005; 173:1926-1929. 27. Halpern EJ, Ramey JR, Strup SE, Frauscher F, McCue P, Gomella LG. Detection of prostate carcinoma with contrast-enhanced sonography using intermittent harmonic imaging. Cancer 2005; 104:2373-2383. 28. Konig K, Scheipers U, Pesavento A, Lorenz A, Ermert H, Senge T. Initial experiences with real-time elastography guided biopsies of the prostate. J Urol 2005; 174:115-117. 29. Frauscher F, Klauser A, Koppelstaetter F, Mallouhi A, Horninger W, Zur ND. Real-time elastography for prostate cancer detection: Preliminary experience. Eur Radiol 2004; 14:150. 30. Prando A, Wallace S. Helical CT of prostate cancer: early clinical experience. AJR Am J Roentgenol. 2000; 175:343-346. 31. Beyersdorff D, Taupitz M, Winkelmann B, et al. Patients with a history of elevated prostatespecific antigen levels and negative transrectal US-guided quadrant or sextant biopsy results: value of MRI. Radiology 2002; 224:701-706. 32. Kurhanewicz J, Vigneron DB, Hricak H, Narayan P, Carroll P, Nelson SJ. Three-dimensional H-1 MR spectroscopic imaging of the in situ human prostate with high (0.24-0.7-cm3) spatial resolution. Radiology 1996; 198:795-805. 33. Heerschap A, Jager GJ, Graaf M van der, et al. In vivo proton MR spectroscopy reveals altered metabolite content in malignant prostate tissue. Anticancer Res 1997; 17:1455-1460.
360
J.J. Fütterer and J.R. Spermon
34. Scheidler J, Hricak H, Vigneron DB, et al. Prostate cancer: localization with three-dimensional proton MR spectroscopic imaging–clinicopathologic study. Radiology 1999; 213:473-480. 35. Futterer JJ, Heijmink SW, Scheenen TW, et al. Prostate cancer localization with dynamic contrast-enhanced MRI and proton MR spectroscopic imaging. Radiology. 2006; 241:449-458. 36. Barentsz JO, Engelbrecht M, Jager GJ, et al. Fast dynamic gadolinium-enhanced MRI of urinary bladder and prostate cancer. J Magn Reson Imaging 1999; 10:295-304. 37. Donahue KM, Weisskoff RM, Parmelee DJ, et al. Dynamic Gd-DTPA enhanced MRI measurement of tissue cell volume fraction. Magn Reson Med 1995; 34:423-432. 38. Huisman HJ, Engelbrecht MR, Barentsz JO. Accurate estimation of pharmacokinetic contrastenhanced dynamic MRI parameters of the prostate. J Magn Reson Imaging 2001; 13:607-614. 39. Kiessling F, Lichy M, Grobholz R, et al. Detection of prostate carcinomas with T1-weighted dynamic contrast-enhanced MRI. Value of two-compartment model. Radiologe. 2003; 43: 474-480. 40. Engelbrecht MR, Huisman HJ, Laheij RJ, et al. Discrimination of prostate cancer from normal peripheral zone and central gland tissue by using dynamic contrast-enhanced MRI. Radiology 2003; 229:248-254. 41. Van Dorsten FA, Van Der Graaf M., Engelbrecht MR, et al. Combined quantitative dynamic contrast-enhanced MRI and (1)H MR spectroscopic imaging of human prostate cancer. J Magn Reson Imaging 2004; 20:279-287. 42. Shreve PD, Grossman HB, Gross MD, Wahl RL. Metastatic prostate cancer: initial findings of PET with 2-deoxy-2-[F-18]fluoro-D-glucose. Radiology 1996; 199:751-756. 43. Effert PJ, Bares R, Handt S, Wolff JM, Bull U, Jakse G. Metabolic imaging of untreated prostate cancer by positron emission tomography with 18fluorine-labeled deoxyglucose. J Urol 1996; 155:994-998. 44. Haseman MK, Reed NL, Rosenthal SA. Monoclonal antibody imaging of occult prostate cancer in patients with elevated prostate-specific antigen. Positron emission tomography and biopsy correlation. Clin Nucl Med 1996; 21:704-713. 45. Liu IJ, Zafar MB, Lai YH, Segall GM, Terris MK. Fluorodeoxyglucose positron emission tomography studies in diagnosis and staging of clinically organ-confined prostate cancer. Urology 2001; 57:108-111. 46. Sutinen E, Nurmi M, Roivainen A, et al. Kinetics of [(11)C]choline uptake in prostate cancer: a PET study [correction for study]. Eur J Nucl Med Mol Imaging 2004; 31:317-324. 47. Ross PL, Scardino PT, Kattan MW. A catalog of prostate cancer nomograms. J Urol 2001; 165:1562-1568. 48. Ross PL, Gerigk C, Gonen M, et al. Comparisons of nomograms and urologists’ predictions in prostate cancer. Semin Urol Oncol 2002; 20:82-88. 49. Reckwitz T, Potter SR, Partin AW. Prediction of locoregional extension and metastatic disease in prostate cancer: a review. World J Urol 2000; 18:165-172. 50. Khan MA, Partin AW. Partin tables: past and present. BJU Int 2003; 92:7-11. 51. Sobin LH, Wittekind CH. UICC: TNM Classification of Malignant Tumors, 6th edn WileyLiss, New York, 2002. 52. Heiken JP, Forman HP, Brown JJ. Neoplasms of the bladder, prostate and testis. Radiol Clin North Am 1994; 32:81-98. 53. Futterer JJ, Engelbrecht MR, Jager GJ, et al. Prostate cancer: comparison of local staging accuracy of pelvic phased-array coil alone versus integrated endorectal-pelvic phased-array coils. Local staging accuracy of prostate cancer using endorectal coil MRI. Eur Radiol. 2007; 17:1055-1065. 54. May F, Treumann T, Dettmar P, et al. Limited value of endorectal magnetic resonance imaging and transrectal ultrasonography in the staging of clinically localized prostate cancer. BJU int 2001; 87:66-69. 55. Sauvain JL, Palasack P, Bourscheid D, et al. Value of power Doppler and 3D vascular sonography as a method for diagnosis and staging of prostate cancer. Eur Urol 2003; 44:21-31.
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
361
56. Presti JC, Jr., Hricak H, Narayan PA, Shinohara K, White S, Carroll PR. Local staging of prostatic carcinoma: comparison of transrectal sonography and endorectal MRI. AJR Am J Roentgenol 1996; 166:103-108. 57. Ekici S, Ozen H, Agildere M, et al. A comparison of transrectal ultrasonography and endorectal magnetic resonance imaging in the local staging of prostatic carcinoma. BJU Int 1999; 83:796-800. 58. Garg S, Fortling B, Chadwick D, Robinson MC, Hamdy FC. Staging of prostate cancer using 3-dimensional transrectal ultrasound images: a pilot study. J Urol 1999; 162:1318-1321. 59. Okihara K, Kamoi K, Lane RB, Evans RB, Troncoso P, Babaian RJ. Role of systematic ultrasound-guided staging biopsies in predicting extraprostatic extension and seminal vesicle invasion in men with prostate cancer. J Clin Ultrasound 2002; 30:123-131. 60. Burcombe RJ, Ostler PJ, Ayoub AW, Hoskin PJ. The role of staging CT scans in the treatment of prostate cancer: a retrospective audit. Clin Oncol (R Coll Radiol) 2000; 12:32-35. 61. Huncharek M, Muscat J. Serum prostate-specific antigen as a predictor of staging abdominal/ pelvic computed tomography in newly diagnosed prostate cancer. Abdom Imaging. 1996; 21:364-367. 62. Tarcan T, Turkeri L, Biren T, Kullu S, Gurmen N, Akdas A. The effectiveness of imaging modalities in clinical staging of localized prostatic carcinoma. Int Urol Nephrol 1996; 28:773-779. 63. Barbieri A, Monica B, Sebastio N, Incarbone GP, Di Stefano C. Value and limitations of transrectal ultrasonography and computer tomography in preoperative staging of prostate carcinoma]. Acta Biomed Ateneo Parmense 1997; 68:23-26. 64. Sonnad SS, Langlotz CP, Schwartz JS. Accuracy of MRI for staging prostate cancer: a metaanalysis to examine the effect of technologic change. Acad Radiol 2001; 8:149-157. 65. Engelbrecht MR, Jager GJ, Laheij RJ, Verbeek AL, van Lier HJ, Barentsz JO. Local staging of prostate cancer using magnetic resonance imaging: a meta-analysis. Eur Radiol 2002; 12:2294-2302. 66. Wang L, Mullerad M, Chen HN, et al. Prostate cancer: incremental value of endorectal MRI findings for prediction of extracapsular extension. Radiology 2004; 232:133-139. 67. Engelbrecht MR, Jager GJ, Severens JL. Patient selection for magnetic resonance imaging of prostate cancer. Eur Urol 2001; 40:300-307. 68. Jager GJ, Severens JL, Thornbury JR, de la Rosette JJ, Ruijs SH, Barentsz JO. Prostate cancer staging: should MRI be used?–A decision analytic approach. Radiology 2000; 215:445-451. 69. Langlotz CP, Schnall MD, Malkowicz SB, Schwartz JS. Cost-effectiveness of endorectal magnetic resonance imaging for the staging of prostate cancer. Acad Radiol 1996; 3 Suppl 1:S24-S27. 70. Fütterer JJ, Engelbrecht MR, Huisman HJ, et al. Staging Prostate Cancer with Dynamic Contrast-enhanced Endorectal MRI prior to Radical Prostatectomy: Experienced versus Less Experienced Readers. Radiology 2005; 237:541-549. 71. Yu KK, Scheidler J, Hricak H, et al. Prostate cancer: prediction of extracapsular extension with endorectal MRI and three-dimensional proton MR spectroscopic imaging. Radiology 1999; 213:481-488. 72. Fütterer JJ, Heijmink SW, Scheenen TW, et al. Prostate cancer: local staging at 3-T endorectal MRI–early experience. Radiology 2006; 238:184-191. 73. Heijmink SWTPJ, Fütterer JJ, Hambrock T, et al. Body Array versus Endorectal coil MRI of Prostate Cancer at 3 Tesla: Comparison of Image Quality, Localization, and Staging Performance with Whole-Mount Section Histopathology as Standard of Reference. Radiology 2007; in press. 74. Lawrentschuk N, Davis ID, Bolton DM, Scott AM. Positron emission tomography and molecular imaging of the prostate: an update. BJU Int. 2006; 97:923-931. 75. Messing EM, Manola J, Sarosdy M, Wilding G, Crawford ED, Trump D. Immediate hormonal therapy compared with observation after radical prostatectomy and pelvic lymphadenectomy in men with node-positive prostate cancer. N Engl J Med 1999; 341(24):1781-1788. 76. Walsh PC. Surgery and the reduction of mortality from prostate cancer. N Engl J Med 2002; 347(11):839-840.
362
J.J. Fütterer and J.R. Spermon
77. Magnusson A, Fritjofsson A, Norlen BJ, Wicklund H. The value of computed tomography and ultrasound in assessment of pelvic lymph node metastases in patients with clinically locally confined carcinoma of the prostate. Scand J Urol Nephrol 1988; 22:7-10. 78. Flanigan RC, McKay TC, Olson M, Shankey TV, Pyle J, Waters WB. Limited efficacy of preoperative computed tomographic scanning for the evaluation of lymph node metastasis in patients before radical prostatectomy. Urology 1996; 48:428-432. 79. Borley N, Fabrin K, Sriprasad S, et al. Laparoscopic pelvic lymph node dissection allows significantly more accurate staging in “high-risk” prostate cancer compared to MRI or CT. Scand J Urol Nephrol 2003; 37:382-386. 80. Jager GJ, Barentsz JO, Oosterhof GO, Witjes JA, Ruijs SJ. Pelvic adenopathy in prostatic and urinary bladder carcinoma: MRI with a three-dimensional TI-weighted magnetizationprepared-rapid gradient-echo sequence. AJR Am J Roentgenol 1996; 167:1503-1507. 81. Harisinghani MG, Barentsz J, Hahn PF, et al. Noninvasive detection of clinically occult lymph-node metastases in prostate cancer. N Engl J Med 2003; 348:2491-2499. 82. Weissleder R, Elizondo G, Wittenberg J. et al. Ultrasmall paramagnetic iron oxide: an intravenous contrast agent for assessing lymph nodes with MRI. Radiology 1990; 175:494-498. 83. Vassallo P, Matei C, Heston WDW, et al. AMI-227-enhanced MR Lymphography: usefulness for differentiating reactive from tumor-bearing lymph nodes. Radiology 1994; 193:501-506. 84. Heesakkers RA, Fütterer JJ, Hovels AM, et al. Prostate cancer evaluated with ferumoxtran-10enhanced T2*-weighted MRI at 1.5 and 3.0 T: early experience. Radiology 2006; 239:481-487. 85. Heicappell R, Muller-Mattheis V, Reinhardt M, et al. Staging of pelvic lymph nodes in neoplasms of the bladder and prostate by positron emission tomography with 2-[(18)F]-2deoxy-D-glucose. Eur Urol 1999; 36:582-587. 86. Lentle BC, McGowan DG, Dierich H. Technetium-99M polyphosphate bone scanning in carcinoma of the prostate. Br J Urol 1974; 46:543-548. 87. Elkin M, Mueller HP. Metastases from cancer of the prostate; autopsy and roentgenological findings. Cancer 1954; 7:1246-1248. 88. Abuzallouf S, Dayes I, Lukka H. Baseline staging of newly diagnosed prostate cancer: a summary of the literature. J Urol 2004; 171:2122-2127. 89. Oesterling JE, Martin SK, Bergstralh EJ, Lowe FC. The use of prostate-specific antigen in staging patients with newly diagnosed prostate cancer. JAMA 1993; 269:57-60. 90. Lauenstein TC, Goehde SC, Herborn CU, et al. Whole-body MRI: evaluation of patients for metastases. Radiology 2004; 233:139-148. 91. Shreve PD, Grossman HB, Gross MD, Wahl RL. Metastatic prostate cancer: initial findings of PET with 2-deoxy-2-[F-18]fluoro-D-glucose. Radiology 1996; 199:751-756. 92. Garner MJ, Turner MC, Ghadirian P, Krewski D. Epidemiology of testicular cancer: an overview. Int J Cancer. 2005;116(3):331-339. 93. Comiter CV, Benson CJ, Capelouto CC, et al: Nonpalpable intratesticular masses detected sonographically. J Urol 1995; 154:1367-1369. 94. National Cancer Institute. SEER Cancer Statistics Review, 1975-2001. Bethesda, MD, 2004. Available at: http://seer.cancer.gov/csr/1975_2001/. Accessed June 15, 2006. 95. Dalal PU, Sohaib SA, Huddart R. Imaging of testicular germ cell tumors. Cancer Imaging 2006; 6:124-134. 96. Coakley FV, Hricak H, Presti JC Jr. Imaging and management of atypical testicular masses. Urol Clin North Am 1998; 25:375–388. 97. Hernes EH, Harstad K, Fossa SD. Changing incidence and delay of testicular cancer in southern Norway (1981-1992). Eur Urol 1996; 30:349-357. 98. Hricak H. Imaging of the scrotum. Textbook and atlas, New York, Raven press, 1995;49-93. 99. Guthrie JA, Fowler RC. Ultrasound diagnosis of testicular tumors presenting as epididymal disease. Clin Radiol 1992; 46: 397–400. 100. Grantham JG, Charboneau JW, James EM, et al. Testicular neoplasms: 29 tumors studied by high-resolution US. Radiology 1985; 157: 775–780. 101. Senay BA, Stein BS. Testicular neoplasm diagnosed by ultrasound. J Surg Oncol 1986; 32: 110–112
14 Recent Advances in Imaging of Male Reproductive Tract Malignancies
363
102. Rifkin MD, Kurtz AB, Pasto ME, Goldberg BB. Diagnostic capabilities of high-resolution scrotal ultrasonography: prospective evaluation. J Ultrasound Med 1985; 4:13–19. 103. Oyen R. Imaging of testicular neoplasms. In: Carcinoma of the kidney and testis, and rare malignancies. Innovations in management. Springer Verlag, 1999;203-210. 104. Schwerk WB, Schwerk WN, Rodeck G: Testicular tumors: Prospective analysis of real-time US patterns and abdominal staging. Radiology 1987; 164:369-374. 105. Benson CB, Doubilet PM, Richie JP: Sonography of the male genital tract. Am J Roentgenol 1989; 153:705-713. 106. Ciftci AO, Bingol-Kologlu M, Senocak ME, et al. Testicular tumors in children. J Pediatr Surg 2001; 36:1796-1801. 107. Backus ML, Mack LA, Middleton WD, King BF, Winter TC, True LD. Testicular microlithiasis: imaging appearances and pathologic correlation. Radiology 1994; 192:781-785. 108. Peckham MJ, Barrett A, McEwlain TJ, Hendry WF, Raghaven D. Non-seminoma germ cell tumors (malignant teratomas) of the testis. BJU 1981; 53:162-172. 109. Donohue JP, Zachary JM, Maynard BR. Distribution of nodal metastases in nonseminomatous testis cancer. J Urol 1982; 128:315-320. 110. Mason MD, Featherstone T, Olliff J, Horwich A. Inguinal and iliac lymph node involvement in germ cell tumors of the testis: implications for radiological investigation and for therapy. Clin Oncol 1991; 3:147-150. 111. Siegel MJ: The acute scrotum. Radiol Clin North Am 1997; 35:959-976. 112. Horstman WG, Melson GL, Middleton WD, Andriole GL. Testicular tumors: findings with color Doppler US. Radiology 1992; 185:733–737. 113. Kundra V. Testicular cancer. Semin Roentgenol 2004; 39:437-450. 114. Watanabe Y, Dohke M, Ohkubo K, et al. Scrotal disorders: evaluation of testicular enhancement patterns at dynamic contrast-enhanced subtraction MRI. Radiology 2000; 217:219-227. 115. Schultz-Lampel D, Bogaert G, Thuroff JW, Schlegel E, Cramer B. MRI for evaluation of scrotal pathology. Urol Res 1991; 19:289-292. 116. Cramer BM, Schlegel EA, Thueroff JW. MRI in the differential diagnosis of scrotal and testicular disease. Radiographics. 1991; 11:9-21. 117. Albers P, Bender H, Yilmaz H, Schoeneich G, Biersack HJ, Mueller SC. Positron emission tomography in the clinical staging of patients with Stage I and II testicular germ cell tumors. Urology. 1999; 53:808-811. 118. Muller-Mattheis V, Reinhardt M, Gerharz CD, et al. Positron emission tomography with [18 F]-2-fluoro-2-deoxy-D-glucose (18FDG-PET) in diagnosis of retroperitoneal lymph node metastases of testicular tumors. Urologe A. 1998; 37:609-620. 119. Spermon JR, De Geus-Oei LF, Kiemeney LA, Witjes JA, Oyen WJ. The role of (18)fluoro2-deoxyglucose positron emission tomography in initial staging and re-staging after chemotherapy for testicular germ cell tumors. BJU Int. 2002; 89:549-556. 120. Gambhir SS, Czernin J, Schwimmer J, et al. A tabulated summary of the FDG PET literature. J Nucl Med 2001; 42:1S-93S 121. De Santis M, Becherer A, Bokemeyer C, et al. 2-18fluoro-deoxy-D-glucose positron emission tomography is a reliable predictor for viable tumor in postchemotherapy seminoma: an update of the prospective multicentric SEMPET trial. J Clin Oncol. 2004;22(6):1034-1039. 122. Donohue JP, Thornhill JA, Foster RS, Rowland RG, Bihrle R. Primary retroperitoneal lymph node dissection in clinical stage A non-seminomatous germ cell testis cancer. Review of the Indiana University experience 1965-1989. Br J Urol. 1993; 71:326-335. 123. Hilton S, Herr HW, Teitcher JB, Begg CB, Castellino RA. CT detection of retroperitoneal lymph node metastases in patients with clinical stage I testicular nonseminomatous germ cell cancer: assessment of size and distribution criteria. AJR Am J Roentgenol. 1997; 169: 521-525. 124. Fernandez EB, Moul JW, Foley JP, Colon E, McLeod DG. Retroperitoneal imaging with third and fourth generation computed axial tomography in clinical stage I nonseminomatous germ cell tumors. Urology 1994; 44:548-552.
364
J.J. Fütterer and J.R. Spermon
125. See W.A., Hoxie L. Chest staging in testis cancer patients: imaging modality selection based upon risk assessment as determined by abdominal computerized tomography scan results. J Urol 1993; 150:874-878. 126. Ellis JH, Bies JR, Kopecky KK, Klatte EC, Rowland RG, Donohue JP. Comparison of NMR and CT imaging in the evaluation of metastatic retroperitoneal lymphadenopathy from testicular carcinoma. J Comput Assist Tomogr. 1984; 8:709-719. 127. Hogeboom WR, Hoekstra HJ, Mooyaart, EL et al. The role of magnetic resonance imaging and computed tomography in the treatment evaluation of retroperitoneal lymph-node metastases of non-seminomatous testicular tumors. Eur J Radiol. 1991; 13:31-36. 128. Harisinghani MG, Saksena M, Ross RW, Tabatabaei S, A pilot study of lymphotrophic nanoparticle-enhanced magnetic resonance imaging technique in early stage testicular cancer: a new method for non-invasive lymph node evaluation. Urology. 2005; 66:1066-1071. 129. Satoh M, Ito A, Kaiho Y, et al. Intraoperative, radio-guided sentinel lymph node mapping in laparoscopic lymph node dissection for Stage I testicular carcinoma. Cancer. 2005; 103:2067-2072. 130. P Warde, MK Gospodarowicz, T Panzarella, CN Catton Stage I testicular seminoma: results of adjuvant irradiation and surveillance. J Clin Oncol. 1995; 13:2255-2262. 131. Cremerius U, Wildberger JE, Borchers H, et al. Does positron emission tomography using 18-fluoro-2-deoxyglucose improve clinical staging of testicular cancer?–Results of a study in 50 patients. Urology 1999; 54:900-904. 132. Hain SF, O’Doherty MJ, Timothy AR, et al. Fluorodeoxyglucose PET in the initial staging of germ cell tumors. Eur J Nucl Med 2000; 27:590-594. 133. Flechon A, Bompas E, Biron P, Droz JP. Management of post-chemotherapy residual masses in advanced seminoma. J Urol 2002; 168:1975-1979. 134. Steyerberg EW, Gerl A, Fossa SD, et al. Validity of predictions of residual retroperitoneal mass histology in nonseminomatous testicular cancer. J Clin Oncol 1998; 16:269-274. 135. Rustin GJ, Mead GM, Stenning SP, et al. National Cancer Research Institute Testis Cancer Clinical Studies Group. Randomized trial of two or five computed tomography scans in the surveillance of patients with stage I nonseminomatous germ cell tumors of the testis: Medical Research Council Trial TE08, ISRCTN56475197–the National Cancer Research Institute Testis Cancer Clinical Studies Group. J Clin Oncol 2007; 25:1310-1315. 136. Albers P, Albrecht W, Algaba F, Bokemeyer C, Cohn-Cedermark G, Horwich A, Klepp O, Laguna MP, Pizzocaro G. Guidelines on testicular cancer. Eur Urol 2005; 48:885-894.
15
Imaging of Malignant Skeletal Tumors Jay Pahade, MD, Aarti Sekhar, MD, and Sanjay K. Shetty, MD
1
Introduction
Malignant tumors of the skeleton represent a diverse group of primary and secondary neoplasms, each with unique imaging and clinical features. The radiologist encountering a lesion of the skeleton must apply a methodical approach to the analysis of imaging features to distinguish benign from malignant entities. This methodical approach can provide invaluable insight into the nature of the lesion, and will ultimately guide the final diagnosis; indeed, concordance between the imaging appearance and a preliminary histologic diagnosis is absolutely necessary to ensure that each lesion is appropriately diagnosed and managed. For the clinician, there is an ever-expanding array of potential imaging modalities that can characterize a lesion and evaluate its extent. Imaging will guide treatment, monitor response to therapy and facilitate discussions of prognosis. The purpose of this chapter is to familiarize the practicing clinician and radiologist with the most common malignant lesions of the skeleton. The chapter describes the major primary lesions of bone (osteosarcoma, chondrosarcoma, myeloma, Ewing’s Sarcoma and primary lymphoma of bone), as well as metastasis. Our goal is to familiarize the reader with the key imaging characteristics of each lesion, as well as the clinical features that may guide the differential diagnosis. The discussion incorporates all imaging modalities, including radiographs, magnetic resonance imaging (MRI), computed tomography (CT), positron emission tomography (PET) and bone scintigraphy, with a particular focus on the appropriate use of each modality in the diagnosis and staging of a newly detected lesion. Recent evidence, particularly focused on the newer modalities (MRI and PET), is presented to provide an evidence-based foundation for the imaging work-up.
Department of Radiology, Musculoskeletal Section, Beth Israel Deaconess Medical Center, Harvard Medical School, 330 Brookline Avenue, E/CC-4, Boston, MA 02215, 617-667-1658, Fax 617-667-8212 Direct correspondence to: SK Shetty,
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
367
368
2
J. Pahade et al.
General Approach: Radiographs
Radiographs are commonly the first imaging modality on which a skeletal lesion is detected and characterized. Despite the development of new imaging modalities, radiographs play a central role in the characterization of skeletal lesions. The majority of incidentally detected findings represent benign lesions, and it is the role of the radiologist to correctly differentiate these benign lesions to prevent the additional cost and morbidity associated with additional evaluation. When evaluating a lesion several key features should be assessed to guide the differential diagnosis. The revised Lodwick classification system is a simple methodology for evaluating the margin and growth pattern of a bone lesion that provides insight into tumor growth rate and host response. This classification system divides lesions into geographic with sclerotic margins (grade IA), geographic with nonsclerotic margins, at most partial cortical destruction, and/or greater than 1 cm of cortical expansion (grade IB), geographic with full cortical penetration and at most 1 cm of a moth-eaten margin (grade IC), mixed geographic and motheaten (grade II), and permeative or moth-eaten (grade III). Key to this classification system is evaluation of the margin: a geographic lesion is one in which a pencil could easily be used to trace the margin. In contrast, a moth-eaten lesion will have irregular margins, often with numerous areas of lucency (“holes”), and a permeative lesion is extremely subtle and ill-defined. This grading system is important because of its predictive value; the rate of malignancy as defined in previous studies is: grade IA (6 percent), grade IB (48 percent), grade IC (36 percent), grade II (97 percent), and grade III (100 percent) [1, 2]. Lesions with a more aggressive appearance in this system should be considered malignant until proven otherwise, although it is important to keep in mind that benign entities (including infection and Langerhans Cell Histiocytosis) can have an aggressive appearance based on imaging alone. Other characteristics can also help narrow the differential diagnosis. Location can be described in several ways: the affected bone (axial or appendicular skeleton, long or flat bone), location along the length of the bone (epiphyseal, metaphyseal or diaphyseal) and location within the bone (central, eccentric, cortical or juxtacortical). Internal matrix is either absent (no internal matrix), osseous (dense mineralization or cloud-like), chondroid (rings-and-arcs, popcorn) or fibrous (ground glass). Associated findings include the presence or absence of cortical destruction and endosteal scalloping, periosteal reaction and a contiguous soft tissue mass. In discussing each of the major categories of malignant skeletal tumors, these imaging features will be referenced to help develop a framework for approaching these lesions from an imaging and clinical perspective.
15 Imaging of Malignant Skeletal Tumors
3 3.1
369
Osteosarcoma Introduction
Osteosarcoma is the second most common primary bone tumor after multiple myeloma, accounting for 15 percent of all primary bone tumors. It is the most common bone tumor of young adults, coinciding with a period of increased bone development. While the etiology of osteosarcoma is unknown, the increased risk seen in patients with hereditary retinoblastoma and Li-Fraumeni syndrome suggests important pathogenetic roles of the tumor suppressor genes p53 and RB [3]. There is a predilection for males (gender distribution is 2:1) and the clinical presentation is usually nonspecific, with symptoms such as pain and swelling. Histologically, osteosarcoma is an aggressive osteoid-producing lesion in the metaphyses of fast-growing long bones. The most common locations are the distal femur, proximal tibia, and proximal humerus. Even if elements such as fibrous or cartilage matrix dominate the tumor tissue, any production of an osteoid matrix confers the diagnosis of osteosarcoma. Pathologic fractures are seen in 15 to 20 percent of lesions, and approximately 20 to 25 percent of patients have metastases at the time of presentation [4]. The vast majority of these metastases are found in the lungs (90 percent) or bones [3]. Radiographic appearance is usually suggestive of the diagnosis of osteosarcoma. Key imaging findings include the presence of osteoid matrix and an overall aggressive imaging appearance, including aggressive periosteal reaction. CT and MRI can add critical information for staging and preoperative planning [4]. Numerous subtypes of osteosarcoma have been identified and characterized (see Table 15.1). This section focuses on the main subtypes of osteosarcoma, including intramedullary, surface and secondary osteosarcomas. Distinguishing between these subtypes requires characterizing the lesion in terms of patient demographics, location, and distinct imaging features. From the perspective of the imager, the correct identification of a particular subtype is particularly important when the distinction has therapeutic or prognostic impact.
3.2
Conventional or High Grade Osteosarcoma
Conventional osteosarcoma, the most common subtype, is a large aggressive lesion affecting the metaphysis of long bones, particularly around the knee (Figs. 15.1, 15.2). The femur is most commonly involved (40 percent to 45 percent), followed by the tibia (16 percent to 20 percent) and the humerus (10 percent to 15 percent). The lesion usually starts in the metaphysis (90 percent to 95 percent), but often extends into the epiphysis, particularly when the physes are open. There is a slight male predilection, with a gender ratio of 1.5-2 to 1 [4].
370
J. Pahade et al.
Table 15.1 Types of Osteosarcoma I. Intramedullary (arise in medullary canal and occupy entire width of bone, usually higher grade) High grade or conventional 75% Telangiectatic 5-11% Low grade 5% Small cell 1-4% Osteosarcomatosis or multifocal 3-4% Gnathic 6-9% II. Surface or Juxtacortical (4-10% of all osteosarcomas, usually lower grade, arise in 3rd and 4th decade) Intracortical Rare Parosteal 5% (65% of surface osteosarcoma) Periosteal (25% of surface osteosarcoma) High-grade (10% of surface osteosarcoma) III. Secondary (5-7% of all osteosarcomas) Paget (67-90% of secondary osteosarcoma) Post-radiation (6-22% of secondary osteosarcoma) Chronic infection, osteonecrosis, fibrous dysplasia IV. Extraskeletal (4% of all osteosarcomas)
Fig. 15.1 Chondroblastic osteosarcoma in a 29-year-old male who presented with three months of persistent pain after trauma. (a) Lateral radiograph of the distal femur reveals a large, partially ossified lesion of the distal femur (arrows) with associated aggressive periosteal reaction. (b) Coronal T1-weighted and (c) sagittal T2-weighted MR images show a T1 isointense and T2 hyperintense mass of the distal femoral diaphysis and metaphysis with extraosseous extension (arrows). Note that the MRI clearly shows the extent of marrow involvement and clearly delineates associated soft tissue structures. (d) Frontal planar bone scintigraphy shows increased uptake in the distal femur, corresponding to the distal femoral mass (large arrow) and irregular uptake along the length of the tibia (small arrow) that reflects recent trauma. (e) Follow-up lateral radiograph after two cycles of neoadjuvent chemotherapy (methotrexate, Adriamycin, and cisplatin) demonstrates dense ossification of the mass. Images courtesy of Dr. Mary G. Hochman
15 Imaging of Malignant Skeletal Tumors
371
Fig. 15.1 (continued)
Pathologically, osteosarcomas are usually large tumors (5 to 10 cm) with frequent soft tissue extension. They are comprised of mesenchymal cells that produce an osteoid matrix. Three histologic patterns are described depending on the predominant cell type: osteoblastic (50 percent to 80 percent), fibroblastic/fibrohistocytic (7 percent to 25 percent), and chondroblastic (5 percent to 25 percent). Osteosarcoma can also be graded from I to IV, depending on the degree of anaplasia and mitotic rate. Conventional osteosarcoma often achieves a grade of III to IV.
372
J. Pahade et al.
Fig. 15.2 High-grade conventional osteosarcoma in a 21-year-old male presenting with knee pain. (a) Frontal radiograph of the right knee demonstrates an ill-defined lucent lesion of the proximal tibial epiphysis / metaphysis (white arrows) with medial cortical destruction. (b) Coronal and (c) axial proton density weighted and (d) coronal STIR MR images of the proximal tibia better show the extent of marrow involvement and the extraosseous extension through a violated cortex (arrows). Images courtesy of Dr. Mary G. Hochman
Radiographic characteristics include a mixed lytic-sclerotic pattern with a variable amount of fluffy, cloud-like opacities within the lesion representing osteoid matrix. Cortical breakthrough without expansion of the bone and an aggressive periosteal reaction with a laminated, hair-on-end or sunburst appearance are common. Soft tissue masses are seen in 80 to 90 percent of lesions [4, 5]. Poor prognostic factors include tumor size > 10 cm, advanced stage at presentation, and pathologic fracture [4].
15 Imaging of Malignant Skeletal Tumors
373
CT imaging is useful in defining the extent of tumor, evaluating anatomically complex areas (pelvis, mandible, maxilla), and characterizing small lesions that measure < 5 cm. Additionally, CT offers superior detection of subtle areas of mineralized matrix in mostly lytic lesions. On CT imaging, the tumor will be primarily low attenuation or soft tissue density, with areas of higher attenuation osteoid production. When there is extensive edema or necrosis, CT scanning may be superior to MRI for determining soft tissue involvement. Osteosarcoma shows marked uptake of radiotracer on bone scintigraphy; however, the pattern of uptake is nonspecific. The main role for scintigraphy is in evaluating for distant metastases, both osseous and extraosseous, including ossified pulmonary metastases. MRI is important for preoperative evaluation and staging, particularly for assessing the extent of marrow, soft tissue, epiphyseal, neurovascular and joint involvement [4, 5] (Fig. 15.3). It also allows better identification of viable tumors to improve biopsy accuracy. A tumor is seen as low to intermediate signal intensity on T1-weighted images and high signal intensity on STIR or T2-weighted images. Areas of low signal intensity on T1 and T2 images represent mineralized matrix, and areas of high signal intensity on both T1- and T2-weighted images can represent hemorrhage. Necrosis can be seen as low signal intensity on T1 and high signal intensity on T2. MRI should be performed before neoadjuvant chemotherapy, since edema can be misinterpreted as tumor, particularly on STIR images. MRI can also readily identify skip lesions, which are rare (occurring in < 5 percent of conventional osteosarcomas), but are important to identify because they necessitate a more extensive resection and are associated with an extremely poor prognosis [6]. There is currently insufficient data regarding the use of PET/CT in the evaluation and follow-up of osteosarcoma. However, preliminary work has been promising, suggesting that FDG-PET may be useful for the characterization of biologic features of osteosarcoma that relate to tumor grading and treatment follow-up [7]. Treatment of high-grade intramedullary osteosarcoma includes neoadjuvant chemotherapy followed by limb-salvage procedures and postoperative multi-drug chemotherapy, leading to a five-year survival rate of 60 percent to 80 percent.
3.3
Telangiectatic Osteosarcoma
Telangiectatic osteosarcoma is an uncommon variant, comprising 5 to 11 percent of osteosarcomas. It is characterized by cystic cavities with cavernous vessels and blood-filled spaces. Osteoid matrix can sometimes be seen in the periphery of the lesion or in the septations of these cavities. Like high-grade intramedullary osteosarcoma, the telangiectatic variant most commonly affects the metaphyseal regions of the long bones around the knee (48 percent in the distal femur and 14 percent in the proximal tibia), with the proximal humerus also being common (16 percent) [8]. The classic radiographic appearance is a large lytic and expansile lesion (mean lesion size is 6.8 × 11.2 cm) that is occasionally
374
J. Pahade et al.
15 Imaging of Malignant Skeletal Tumors
375
markedly aneurysmal and can mimic an aneurysmal bone cyst. Multiple small fluid/fluid levels, best seen on MRI, are a characteristic finding of telangiectatic osteosarcomas that also mimic aneurysmal bone cysts. The key distinguishing feature is that the telangiectatic osteosarcoma will have a rim of viable tumor cells along the periphery of its cystic spaces, seen best on contrast-enhanced CT and MRI as thick nodular peripheral enhancement [9]. Aggressive features such as cortical destruction (in a geographic pattern), periosteal reaction, wide zone of transition, and pathologic fracture are common [4, 9]. Contrast-enhanced CT shows a marrow-replacing lesion with heterogeneous attenuation. MRI shows very heterogenous signal intensity that is intermediate to high signal intensity on T1-weighted images, and high on T2-weighted images, with evidence of hemorrhage on all MRI pulse sequences. Both CT and MRI contrastenhanced imaging show thick peripheral and septal enhancement, corresponding to areas of sarcomatous tissue and osteoid matrix (the latter of which is optimally seen on CT). Biopsy should be directed towards these peripheral nodules. Soft tissue masses are also frequently seen on CT and MRI. Bone scintigraphy shows marked radionuclide uptake with central photopenia (termed the “donut pattern”), corresponding to a hemorrhagic center. Angiography, which is not routinely performed, can show a hypervascular peripheral stain with or without early venous drainage [9]. Telangiectatic osteosarcoma previously had a dismal prognosis, until the advent of chemotherapy. Now, prognosis is comparable to conventional osteosarcoma [10]. With chemotherapy and wide surgical resection, the five-year survival rate is 68 percent [4].
3.4
Low Grade Osteosarcoma
Low-grade osteosarcoma comprises 5 percent of intramedullary osteosarcomas and affects patients most commonly in the third decade of life. Distribution is most commonly in the metaphyseal region around the knee. Pathologically and radiographically, low-grade osteosarcoma simulates a benign process, including NOF, fibrous dysplasia, and chondromyxoid fibroma. The lesion can show well-defined sclerotic margins, with only subtle evidence of a more aggressive process, such as
Fig. 15.3 Sacral chondroblastic osteosarcoma in a 25-year-old who presented with left leg numbness. (a) AP radiograph of the pelvis demonstrates a very subtle ill-defined sclerosis involving the left sacrum (arrows). (b) Technicium-99 m bone scintigraphy demonstrates a focus of intense tracer uptake within the left sacrum, without distant osseous metastases. (c) Coronal T1, (d) coronal STIR, and (e) axial STIR MR images shows a large infiltrative mass that is T2 hyperintense and T1 isointense to muscle involving the left side of the sacrum at the S1 and S2 levels. The mass extends across midline, into the ventral soft tissues, and into the S1, S2, and S3 neural foramina (arrowheads)
376
J. Pahade et al.
small areas of cortical destruction and an associated soft tissue mass. This variant has locally aggressive behavior and will often recur unless a wide excision is performed. With complete resection, prognosis is excellent.
3.5
Small Cell Osteosarcoma
Small cell osteosarcoma, which accounts for 1 percent to 4 percent of osteosarcomas, is composed of small round blue cells. Histologically, these tumors are similar to Ewing’s sarcoma except that small cell osteosarcoma produces an osteoid matrix and lacks the cellular uniformity seen in Ewing’s. However, some investigators consider this tumor a Ewing’s variant, especially those tumors with positive CD99 membrane staining and chromosome 11-22 translocation, which are classic findings for Ewing’s sarcoma [10]. Again, the metaphysis of the distal femur is the most common location, followed by the proximal humerus and the pelvis. Imaging findings are nonspecific for small cell osteosarcoma (usually diagnosed by biopsy), but findings are similar to conventional osteosarcoma and suggest a highly aggressive lesion. These lesions can be permeative, lytic, and have aggressive features such as cortical breakthrough, aggressive periosteal reaction and soft tissue extension, the latter of which is best seen on cross-sectional imaging. Intramedullary sclerosis is also common. Prognosis is extremely poor [4, 11].
3.6
Gnathic Osteosarcoma
Gnathic osteosarcoma is an osteoid-producing tumor of the mandible and maxilla that is predominantly chondroblastic and affects a slightly older population (average age 34 to 36 years). These tumors are difficult to image and treat, due to the complex anatomic location. CT is often needed to detect the osteoid matrix. Opacification of the maxillary sinus is commonly seen in maxillary lesions. Treatment consists of surgical resection, radiation, and chemotherapy. Local recurrence is common, and the tumor carries a five-year survival rate of only 40 percent.
3.7
Intracortical Osteosarcoma
Intracortical osteosarcoma, first described by Jaffe in 1960, is the rarest subtype of osteosarcoma, with only a handful of cases reported in the literature [12]. The lesion is typically a cortically based lytic lesion that is < 4 cm in diameter and has a rim of perilesional sclerosis. The femur and tibia are, again, the most common location [12].
15 Imaging of Malignant Skeletal Tumors
3.8
377
Parosteal Osteosarcoma
Parosteal osteosarcoma, the most frequent type of surface lesion (comprising 65 percent of surface osteosarcomas) originates from the outer layer of the periosteum. Patients are usually in the third to fourth decade of life and present with a palpable mass in their distal posterior thigh. The lesion is typically a large lobulated cauliflower-like juxtacortical mass arising from the metaphyseal region of long bones. A histologically lowgrade tumor with possible high-grade regions within it, parosteal osteosarcoma occasionally demonstrates “back-growth” invasion into the medullary canal. Radiographically, parosteal osteosarcoma presents as a large centrally dense lesion, attached to the underlying bone by a stalk in earlier stages and with a broader base later in the progression of the disease. The classic location is the posterior distal femur (50 percent to 65 percent of cases). Cortical thickening without an aggressive periosteal reaction is common. Because of its appearance as an ossified mass outside of the bone, parosteal osteosarcoma must be differentiated from myositis ossificans, a lesion that is denser peripherally and not attached to the cortex. Additionally, a cartilage cap is seen in 25 percent to 30 percent of lesions, which may lead to mis-diagnosis as osteochondroma [10]. CT can be useful in demonstrating a radiolucent zone of periosteum and fibrous tissue that becomes trapped between the encircling tumor and cortex. MRI can define the extent of tumor extension to ensure complete surgical resection. Parosteal osteosarcoma has an excellent prognosis and is usually treated with local resection [4, 10]. There is a rare dedifferentiated variant, with a much poorer prognosis.
3.9
Periosteal Osteosarcoma
Periosteal osteosarcoma is a rarer surface lesion arising from the deep layer of periosteum of the femur, tibia, or humerus. Radiographically, these lesions present as a diaphyseal lesion with a thickened, scalloped cortex (without intramedullary invasion), involving over 50 percent of the osseous circumference and displaying a perpendicular periosteal reaction. This periosteal reaction can be seen as rays of low signal intensity on all MRI sequences. These lesions are considered intermediate grade with a fair prognosis. Treatment usually involves wide surgical resection.
3.10
High-grade Surface Osteosarcoma
High-grade surface osteosarcoma are rare tumors that are histologically identical to high-grade intramedullary osteosarcoma and radiographically similar to periosteal osteosarcoma. Often the entire circumference of the bone is involved. There is controversy over whether these lesions can invade the medullary canal
378
J. Pahade et al.
[4, 10]. Aggressive (hair-on-end) periosteal reaction has also been observed. The diaphysis of the femur, humerus, or fibula is involved. Prognosis and treatment are similar to conventional intramedullary osteosarcoma.
3.11
Multifocal Osteosarcoma
Multifocal osteosarcoma is a rare subtype which affects children in the first decade of life and is rapidly fatal. This entity is thought to be a metastatic process involving a dominant lesion with aggressive features, and multiple secondary foci that are smaller and more benign-appearing (sclerotic with well-defined margins). Pulmonary metastases are often seen with this subtype.
3.12
Secondary Osteosarcoma
Osteosarcoma can also result from malignant transformation of a benign process and is seen most frequently in the setting of Paget’s disease (67 percent to 90 percent) (Fig. 15.4) and previous radiation (6 percent to 22 percent) of greater than 1,000 cGy [4]. Other entities that have been reported to uncommonly predispose to osteosarcomatous degeneration include fibrous dysplasia and multiple chondromas. Patients with Paget’s disease tend to be older (typical age range: 55 to 80 years) and malignant transformation to osteosarcoma is suggested by new and progressive bone pain. There is large variation in the frequency of malignant transformation, depending on the extent of disease: patients with limited disease have only a 0.2 percent frequency of transformation, while the figure can be as high as 7.5 percent in those with extensive skeletal disease. Radiographically, aggressive bony destruction can be seen in the setting of the bony sclerosis and expansion that is characteristic of Paget’s disease. The most common sites of disease include the femur, pelvis, humerus, and craniofacial bones. Soft tissue masses are common. These lesions are high-grade and aggressive, with an extremely poor prognosis and five- to ten-year survival rate of less than 5 percent. The dismal prognosis is also a reflection of the poor health of the older population affected by the tumor, surgically inaccessible tumor sites and increased vascularity of the bone, which predisposes to hematologic metastases, particularly to the lungs [4, 13].
3.13
Osteosarcoma: Staging
Prognosis for osteosarcoma is determined by histologic grade and the site of the lesion. The presence of metastases confers a much poorer prognosis, and may require more intensive chemotherapy regimens that have their own risk, including
15 Imaging of Malignant Skeletal Tumors
379
Fig. 15.4 Secondary osteosarcoma in a 63-year-old female, occurring in the setting of Paget’s disease. (a) Composite frontal radiographs of the right femur demonstrate underlying Paget’s disease with cortical thickening and trabecular coarsening involving the proximal femur. A large secondary osteosarcoma is seen within the diaphysis, with “hair-on-end” periosteal reaction and an extensive ossified soft tissue component. (b) Coronal HASTE (T2) and (c) coronal HASTE (T2) with fat saturation reveal an expansile mass arising from the diaphysis of the right femur, extending from the lesser trochanter to the distal metaphysis. Extensive abnormal periosteal reaction and internal osseous matrix (low signal) are seen. The muscle groups of the thigh are displaced peripherally in all directions
380
J. Pahade et al.
high-dose alkylating agents which can increase risk for leukemia [3]. However, in general, neoadjuvant chemotherapy has dramatically improved survival rates for osteosarcoma, thus increasing the demands for imaging to provide detailed information on tumor staging and grading. Currently, staging for osteosarcoma includes conventional radiography of the lesion and biopsy for definitive diagnosis; MRI of the tumor and surrounding bone to assess the extent of disease and evaluate for skip lesions; high-resolution CT (HRCT) of the chest to evaluate for pulmonary metastases; and bone scintigraphy to evaluate for distant bone metastases [7, 14]. There is currently insufficient data on the role of PET/CT in staging of osteosarcoma [7]. Treatment involves neoadjuvant chemotherapy followed by wide local excision with limb-salvage procedures and postoperative chemotherapy, leading to a cure rate ranging from 58 percent to 76 percent [4]. Complete resection with negative margins at the initial surgery is crucial, as positive margins correlate with an increased likelihood of local recurrence, and a subsequently poor prognosis (fiveyear survival rate of 19.2 percent) [15]. Chemotherapy has become increasingly important in curbing hematologic spread and preventing lung metastases. In addition, response to chemotherapy is one of the most important prognostic factors: greater than 90 percent necrosis after chemotherapy is associated with a significantly higher survival rate. Radiation therapy should be used conservatively in the young population affected by osteosarcoma as these tumors are relatively insensitive to radiation, and there is an increased risk of secondary osteosarcoma and radiation-induced soft tissue sarcomas [3].
3.14
Osteosarcoma: Follow-up
Patients should be followed closely for local and systemic recurrence. Local recurrence confers a much poorer prognosis. For pulmonary metastases that are detected early, surgical removal of these lesions can confer a 20 percent to 50 percent chance of cure [7]. There is no established follow-up regimen. Some authors advocate plain radiographs of the affected extremity and HRCT of the chest every three to six months for the first two years following surgery, every six months for the second through fifth year, and then annual surveillance exams. Annual bone scintigraphy is recommended for the first two years [16]. Others advise follow-up with bone scintigraphy alone, since osteosarcomas usually incorporate bisphosphonates, and HRCT of the chest at six-month intervals [7]. Prostheses should also be evaluated for loosening, infection, and mechanical failure. In evaluating for local response to chemotherapy and local recurrence, modalities that evaluate morphologic changes (such as radiographs and CT) have been shown to be of limited value. Thallium-201 scintigraphy, which reflects tumor metabolic activity, has proven itself to be a powerful tool in monitoring tumor response to induction chemotherapy and for detecting local recurrence [17, 18]. MRI is also useful in evaluating treatment response due to its superior soft tissue
15 Imaging of Malignant Skeletal Tumors
381
contrast and the sensitivity for detection of enhancement following gadolinium administration, which may help distinguish viable from necrotic tumor. There is also considerable interest in the role of functional imaging, such as PET/CT, for following response to therapy. PET/CT may help target biopsy in large heterogeneous tumors by delineating highly metabolically active areas [7]. Bredella, et al. showed PET/CT to be helpful in distinguishing viable tumors from post-therapeutic changes when MRI was equivocal [19]. However, there is conflicting evidence from several small studies regarding correlation between SUV measurements and histologic response after chemotherapy [7, 20]. One major pitfall of PET is poor sensitivity for detection of pulmonary metastases, particularly those that are less than 9 mm, the detection of which can significantly improve survival. Larger prospective trials are needed to define the role of PET/CT in the management of osteosarcoma.
3.15
Osteosarcoma: Summary
Osteosarcoma is the most common primary bone tumor of young adults and is defined as any primary bone tumor with production of an osteoid matrix. It commonly occurs in the bones surrounding the knee joint and in the proximal humerus. A variety of subtypes have been described, each with characteristic radiologic features and widely varying prognoses. The most common subtype, conventional or intramedullary osteosarcoma, is a high-grade tumor characterized by cortical destruction, strong periosteal reaction, and associated soft tissue mass. Diagnostic work-up of osteosarcoma includes tumor characterization by radiographs and MRI, image-guided biopsy, evaluation of pulmonary metastases with HRCT of the chest, and bone scintigraphy to evaluate for bony metastases. Treatment involves neoadjuvant chemotherapy followed by wide local excision with limb-salvage therapies and postoperative chemotherapy. Chemotherapy has considerably improved survival rates in osteosarcoma over the past decade. Follow-up evaluation, including imaging of the affected area and HRCT of the chest, is crucial for early detection of local recurrence and metastasis. PET/CT remains an unproven but promising tool in assessing post-therapy response and residual disease.
4 4.1
Chondrosarcoma Introduction
Chondrosarcoma is a malignant tumor of bone characterized by cells that produce a cartilaginous tumor matrix. It is generally classified as primary or secondary on the basis of whether the cells arise de novo (primary) or are superimposed on a
382
J. Pahade et al.
preexistent benign condition such as enchondroma or osteochondroma (secondary). Regional classification systems divide chondrosarcomas based on osseous location into central or peripheral. Central defines tumors that are intramedullary in origin, including those with peripheral extension. Peripheral tumors are further subdivided into those that are secondary, arising from a preexisting osteochondroma, and juxtacortical, which arise from the bone surface [21]. Chondrosarcoma is the third most common type of primary malignant bone tumor. It is estimated to represent 20 percent to 27 percent of all primary malignant bone neoplasms [21]. Syndromic associations with increased risk of chondrosarcoma include the enchondromatosis syndromes (Ollier disease and Maffucci syndrome) and Hereditary Multiple Exostoses (HME). Primary chondrosarcoma is a broad term encompassing many pathologic subtypes with distinct clinical and radiological characteristics. We will discuss conventional intramedullary, clear cell, juxtacortical, myxoid, mesenchymal and dedifferentiated subtypes in this chapter. Radiographs or CT can often suggest the cartiligenous nature of these lesions, demonstrating a predominant chondroid matrix mineralization (termed “arcs-and-rings”). The key role of radiographic work-up is to characterize individual lesions (including assessment of benign and malignant features). The combined multimodality use of radiographs, bone scintigraphy, PET, CT, and MRI scanning can help with staging and guidance of surgical resection and overall treatment.
4.2
Conventional Intramedullary Chondrosarcoma
Conventional intramedullary chondrosarcoma (central chondrosarcoma) is the most common type of primary chondrosarcoma [21]. Patients most frequently present in the fifth and sixth decades and there is a 1.5-2 fold male predilection. Insidious, progressive pain that is worse at night is often the chief complaint; patients complain of pain in at least 95 percent of cases at presentation (Figs. 15.5, 15.6). Palpable mass is sometimes present. Pathologic fractures may often be the initial findings, occurring in 3 to 17 percent of patients [21]. New onset pain should also suggest the possibility of malignant degeneration of benign lesions such as enchondromas and ostechondromas, although pain may also be caused by local impingement or pathologic fracture and does not in itself reflect malignancy. Malignant lesions can also be clinically silent and detected incidentally [22]. Skeletal location can help differentiate conventional intramedullary chondrosarcoma from other subtypes. The proximal aspect of long tubular bones are most commonly affected, particularly the femur (20 percent to 35 percent of cases), upper extremity (10 percent to 20 percent, usually proximal humerus), and tibia (5 percent). The axial skeletal is also frequency affected, with the pelvic bones accounting for approximately 25 percent of lesions and ribs accounting for 8 percent. Any bone can be affected by conventional chondrosarcoma, including the spine, sesamoids and short tubular bones of the hand and feet [21]. In the long
15 Imaging of Malignant Skeletal Tumors
383
Fig. 15.5 Low-grade conventional chondrosarcoma. (a) Radiograph displays mostly lytic lesion involving the proximal fibula. Note the focal cortical destruction, expansion of the proximal fibula, endosteal scalloping extending along the fibular shaft and the faint “arcs-and-rings” cartilaginous matrix (arrow) in the adjacent soft tissues reflecting the extraosseous extension of the mass. (b) Coronal STIR MRI image reveals the hyperintense cartilagenous internal structure and the extension of the mass along the fibular shaft. (c) Pathologic specimen shows the full mass, including the soft tissue component and the distal extension of the mass along the fibular shaft (arrow)
384
J. Pahade et al.
Fig. 15.6 Intermediate grade chondrosarcoma in a 41-year-old male presenting with vague hip pain. (a) Frontal radiograph shows a subtle lucent lesion in the proximal femur with extension distally into the proximal diaphysis (white arrowheads) and endosteal scalloping (black arrowheads). (b) Frog-leg lateral radiograph reveals a focal area of cartilagenous matrix in the adjacent soft tissues, suggesting extraosseous extension. (c) Coronal and (d) axial STIR MR images show the full extent of involvement within the femoral shaft (arrows) and the nodules of extraosseous extension (arrowheads)
15 Imaging of Malignant Skeletal Tumors
385
bones, the metaphysis and diaphysis are involved in 49 and 36 percent of cases, respectively. Only 16 percent of cases are centered in the epiphysis [23]. Chondrosarcomas typically display a mixed lytic and sclerotic pattern, with the sclerotic pattern representing chondroid matrix mineralization. “An arcs-and-rings” pattern of mineralization is the classic description, although on occasion this mineralization may coalesce into a denser pattern. The chondroid matrix pattern of calcification often allows confident diagnosis of a cartilaginous lesion. The lytic component of the lesion often suggests the grade of tumor, a characteristic which correlates well with outcome [24]. The initial histologic grading system was initially suggested by Evans and colleagues [25]. Grade I (low-grade) lesions possess a predominantly chondroid stroma with sparse myxoid areas and chondrocytes with small dense nuclei. Grade II have less chondroid matrix and are more cellular, frequently with myxoid stroma. Grade III (high-grade) lesions exhibit greater cellularity with little to no chondroid matrix and small intercellular myxoid material. High-grade lesions display less chondroid matrix calcification with the radiolucent component displaying a more aggressive pattern of geographic multilobulated bony lysis [21]. Low-grade chondrosarcoma continues to be the most difficult lesion to differentiate from benign cartilaginous lesions [22, 26] in terms of both imaging and histology. Even percutaneous biopsy is subject to sampling error that may fail to identify focal regions of sarcoma. In practice histologic analysis of the entire lesion (following resection) may be necessary to definitively characterize the lesion as benign or malignant. Investigators have attempted to establish imaging parameters to distinguish the two entities, employing a variety of modalities. Endosteal scalloping (erosion of the inner cortex) leading to cortical penetration and soft tissue extension are frequently associated with conventional chondrosarcomas. The depth of scalloping may serve as one of the best distinguishing features between a chondrosarcoma and benign enchondroma: endosteal scalloping involving greater then two-thirds of the normal thickness of cortex is strong evidence of chondrosarcoma as it was identified in 75 percent of cases of chondrosarcoma, versus 9 percent of enchondromas [23]. Another distinguishing feature is the longitudinal extent of endosteal scalloping: chondrosarcoma typically has scalloping involving the entire length of the lesion versus partial involvement in enchondroma [23]. Cortical destruction is an important characteristic: up to 88 percent of conventional chondrosarcomas display cortical destruction, compared to only 8 percent of enchondromas. Soft tissue involvement essentially rules out the diagnosis of enchondroma [23]. Recent work has also suggested that analysis of chromosomal abnormalities may be able to reliably diagnose grade I chondrosarcoma [27]. Traditional work-up for a conventional intramedullary chondrosarcoma usually involves multimodality imaging to better define the osseous and soft tissue characteristics. CT possesses an inherent benefit relative to radiographs when defining osseous involvement, matrix pattern, depth and extension of endosteal scalloping, and particularly in demonstrating cortical destruction. Contrast-enhanced scanning generally shows the lesion’s peripheral rim and septal enhancement [21].
386
J. Pahade et al.
MRI provides the best assessment of marrow involvement. T1-weighted images display low to intermediate signal intensity in areas with marrow replacement relative to the high signal medullary fat. Focal areas of high T1 signal within the lesion, representing unaffected areas of marrow, is a more common feature of benign lesions, identified in 35 percent of conventional chondrosarcomas versus 65 percent of enchondromas [21]. Unmineralized portions of conventional chondrosarcomas display high signal intensity on T2-weighted images reflecting the high water content of hyaline cartilage. Matrix mineralization is better visualized via CT or radiographs and displays low signal on all MRI pulse sequences creating a heterogeneous pattern on T2-weighed sequences. Similar T2 patterns can also be identified in fibrous tissue with high collagen content or other types of generalized calcification [21]. MRI is superior to other modalities in its ability to identify soft tissue extension due to its superior soft tissue contrast. Approximately 76 percent of conventional chondrosarcomas display soft tissue extension, essentially excluding enchondroma [23]. Most agree that a larger soft tissue mass tends to be associated with a higher grade lesion. Often the soft tissue component displays similar imaging intrinsic characteristics as the intraosseous component. Peritumoral edema, identified on water sensitive MRI sequences, may also suggest the diagnosis of conventional chondrosarcoma versus enchondroma [28]. MRI enhancement patterns have been evaluated by several studies with mixed results. Nevertheless, enhancement patterns can suggest the diagnosis of chondrosarcoma versus an osteochondroma. Peripheral enhancement is typically seen in osteochondromas, though it can occasionally be seen in low-grade chondrosarcomas. High-grade chondrosarcomas, on the other hand, have a more diffuse enhancement pattern which can be homogeneous or heterogeneous [29]. Septal enhancement can be seen in both benign osteochondromas and low-grade chondrosarcoma [30, 31]. Subtraction MRI has revealed early and progressive enhancement in chondrosarcomas. When used in combination, these enhancement patterns may help differentiate malignant from benign lesions with increased certainty [31]. Enhancement patterns may also help pinpoint the most appropriate target for percutaneous biopsy. Bone scintigraphy will typically display increased tracer uptake in chondrosarcoma. However, the findings are not specific for the diagnosis, and up to 21 percent of enchondromas also show increased uptake. A heterogeneous uptake pattern may help better distinguish the two entities; 63 percent of intramedullary chondrosarcoma display this pattern versus 30 percent of enchondromas [23]. The role of 18FDG-PET is still being evaluated in clinical practice; however preliminary results appear favorable in using PET – or combination PET/CT – in initial diagnosis, detection of metastasis, and follow-up evaluation in chondrosarcoma [22]. While literature focused solely on PET for conventional intramedullary chondrosarcoma is scarce, several reviews have included multiple subtypes of chondrosarcoma. PET cannot reliably differentiate benign enchondroma from Grade I chondrosarcoma on the basis of SUV (standard uptake values) alone. However, PET may have a role in the classification of lesions as high-grade Grade II to III chondrosarcomas, as SUV values are significantly higher than in benign
15 Imaging of Malignant Skeletal Tumors
387
cartilaginous lesions. It has been suggested that a SUV value above 2.3 be the benchmark to characterize high-grade lesions, although there are case reports of benign lesions (such as giant cell tumors) displaying SUV values above 3. While some series have suggested that PET can reliably distinguish low-grade and highgrade chondrosarcoma [24], others have found overlap in maximal SUV values that preclude definitive separation of Grade I and Grade III tumors [32]. In one series, Grade I central medullary chondrosarcomas had a maximum SUV of 4.1 with a mean value of 2.8, which can overlap with higher grade lesions [32]. Tumor size did not reliably affect SUV values, as larger tumors do not necessarily have higher standardized uptake values than the smaller tumors [24, 32]. PET may play a more prominent role in detecting and evaluating metastases, as SUV values have been found to be extremely high [24]. The role of PET in predicting outcomes has also been investigated; pre-therapeutic tumor maximal SUV obtained by quantitative FDG-PET imaging may be a useful parameter for predicting patient outcome [32].
4.3
Clear Cell Chondrosarcoma
Clear cell chondrosarcoma (CCCS) is rare, constituting approximately 1 to 2 percent of all chondrosarcomas. Patients are most commonly affected in the third to fifth decade of life with a two-fold predilection for men [21]. The lesion tends to be slow-growing and less aggressive, leading to improved prognosis when compared to high-grade conventional chondrosarcoma. Distant metastasis and dedifferentiation is rare, but has been reported [33]. Pathologic fracture may be the presenting symptom in 25 percent of cases [21]. The long tubular bones are affected in 85 to 90 percent of cases; in particular, the proximal femur and proximal humerus are involved in 55 to 60 percent, and 15 to 20 percent of cases, respectively. One key radiographic finding for clear cell chondrosarcoma is its predilection for the epiphysis of long bones. The importance of distinguishing CCCS from other benign entities often centered in the epiphysis, such as chondroblastoma and giant cell tumor (GCT), is important. Conservative excision and curettage frequently results in CCCS recurrence, and en bloc resection may be required [33]. Radiographs reveal a lucent lesion with a variable zone of transition. Typical chondroid matrix mineralization is identified in only 30 percent of cases. A welldefined sclerotic margin may also be identified [33]. Bony remodeling and expansion may be identified in approximately 30 percent of cases, with soft tissue extension being rare but present more frequently in lesions involving the axial skeleton [21]. Periosteal new bone formation is rare [33]. CT examination, like in conventional intramedullary chondrosarcoma, will help identify matrix mineralization, osseous destruction, and soft tissue extension, particularly in anatomically challenging locations such as the flat bones or vertebrae. [33]. MRI shows homogenous or heterogeneous low to intermediate signal on T1 sequences with heterogeneous high signal on T2-weighted sequences [21, 33].
388
J. Pahade et al.
Low signal on T2-weighted sequences has also been reported [21]. Areas of heterogeneity on MRI appear to correlate with mineralization, hemorrhage, and cystic changes. Post-contrast T1 images display a diffuse or heterogeneous pattern of enhancement, which is nonspecific and is also observed in benign chondroblastoma [33, 34]. As noted above, important differential considerations for an epiphyseal lesion such as CCCS include chondroblastoma and giant cell tumor. On MRI, chondroblastoma tends to have low to intermediate signal on T1 sequences and T2 sequences, but high T2 signal has been observed in cystic areas of chondroblastoma and giant cell tumor [33, 34]. Chondroblastoma affects a younger population (third and fourth decade), has a similar male predilection, and tends to be confined solely to the epiphysis with distinct sclerotic margins [21, 33, 34]. It is commonly found about the knee or proximal humerus; rarely, it is found in the hands and feet [35]. Chondroblastoma tends to have extensive peritumoral bone marrow edema (not common in CCCS) and surrounding soft tissue edema (occasionally seen in CCCS) [33, 34]. Definite differentiation of CCCS from chondroblastoma is difficult, but a diagnosis of CCCS may be implied when considering the patient’s age, metaphyseal extension, and lack of bone marrow edema [33, 34]. Giant cell tumor, in contrast, is a lucent lesion that begins in the metaphysis and usually involves the epiphysis at the time of presentation (usually after skeletal maturity). These well-circumscribed lesions are characterized by the lack of a host response (such as a sclerotic margin or periosteal reaction) and the lack of internal matrix. Cortical breakthrough, an associated soft tissue mass, and internal low signal on MRI due to hemosiderin can also be seen [34, 36].
4.4
Juxtacortical Chondrosarcoma
This is a rare lesion, accounting for about 4 percent of all chondrosarcomas [21] (Fig. 15.7). Due to its origin on the bone surface, it has also been termed periosteal or parosteal chondrosarcoma. Patients tend to be in their fourth or fifth decade and there is a slight male predilection. Patients typically report a palpable, painless slowly growing mass [21, 37]. Lesions tend to arise on the surface of long bones, with the most frequent site being the posterior distal femoral metaphysis or diaphysis. Juxtacortical chondrosarcoma can recur after excision, and dedifferentiation has been reported [37]. Radiographs display a characteristic round or oval soft tissue mass on the bone surface with a chondroid matrix. The underlying bone cortex is often thickened with a Codman triangle pattern of periosteal reaction [38]. As with other chondrosarcomas, CT can help better define the matrix mineralization, with the non-mineralized portion showing attenuation values less than that of muscle [21]. MRI displays a lesion with low heterogeneous signal on T1-weighted
15 Imaging of Malignant Skeletal Tumors
389
Fig. 15.7 Low-grade juxtacortical chondrosarcoma in 34-year-old female. (a) Frontal radiograph displays a dense “arcs–and-rings” mineralization pattern associated with a partially calcified soft tissue mass. (b) Coronal STIR, (c) axial proton density, (d) axial STIR, and (e) axial T1 postgadolinium fat saturation images show the T2 hyperintense cartilage within the lesion (best seen on the STIR image) and nodular peripheral and septal enhancement (best seen on the T1 postcontrast image). Note the excellent soft tissue contrast which permits evaluation of neurovascular structures for resection planning
390
J. Pahade et al.
images and heterogeneous high signal on T2-weighted images. The bone marrow is typically spared. Contrast-enhanced scans reveal peripheral and septal enhancement. These imaging characteristics are nonspecific and the differential when approaching such a lesion includes a juxtacortical chondroma, parosteal osteosarcoma, and periosteal osteosarcoma. Chondromas occur three to four times more frequently, but can possess similar imaging characteristics including matrix calcification, intramedullary extension, bone edema, and irregular soft tissue margins [39]. Lesion size may be the only helpful characteristic in differentiating juxtacortical chondroma from juxtacortical chondrosarcoma; chondromas tend to be smaller, averaging 2 cm, while juxtacortical chondrosarcoma averages approximately 5 cm [21, 39]. Some have advocated that all lesions greater then 3 cm should, therefore, undergo wide surgical excision [39]. Differentiating juxtacortical chondrosarcoma from periosteal and parosteal osteosarcoma is also difficult. Periosteal osteosarcoma has a similar histologic appearance, but is usually present in younger patients (10 to 25 years) and is associated with periosteal reaction that occurs perpendicular to the cortex, which is uncommon in juxtacortical chondrosarcoma [40]. Parosteal osteosarcoma appears similar radiographically, except it usually displays a stalk of attachment to the cortex and does not contain chondroid tissue histologically.
4.5
Skeletal Myxoid Chondrosarcoma
Morphologically distinct myxoid chondrosarcoma of the bone (skeletal myxoid chondrosarcoma) is not a well-established entity, but myxoid components are common in conventional intramedullary chondrosarcoma. It is well known that extraskeletal myxoid chondrosarcoma frequently possesses recurrent translocation t(9;22)(q22–31;q11–12), although this genetic abnormality has not been consistently detected in myxoid chondrosarcoma of the bone, suggesting that the soft tissue and intraosseus entities are distinct [41]. Due to the limited number of cases, few conclusions can be drawn about the tumor, but it appears to have a male predilection and is often found in the femur [21]. Although initially believed to be less aggressive, the tumor has been shown to have a high recurrence rate and often develops distant metastases [21, 41]. Radiographically, myxoid chondrosarcoma of the bone does not possess any unique imaging features. It appears aggressive with a lytic permeative pattern, endosteal scalloping, cortical destruction, bony expansion and an associated soft tissue mass [21, 41]. Matrix mineralization may be noted, often more easily detected with CT. CT or MRI can better define soft tissue extent and cortical destruction. MRI displays high signal on T2-weighted sequences with most tumors displaying a component of hemorrhage, especially involving the soft tissue component. Mild contrast enhancement is typical [21].
15 Imaging of Malignant Skeletal Tumors
4.6
391
Mesenchymal Chondrosarcoma of Bone
Mesenchymal chondrosarcoma accounts for around 2 to 13 percent of chondrosarcoma of the bone [21]. It often presents like other malignant bone tumors, with pain and soft tissue swelling. Unlike other chondrosarcomas, it affects women and men equally and affects a younger population (third through fifth decades). In contrast to conventional chondrosarcoma, it most commonly involves the axial skeleton with the craniofacial region being most common (15 percent to 30 percent of cases). Other common sites include the femur, ribs, spine, pelvis, and humerus [21]. Prognosis is unpredictable, but overall survival is poor with a five-year survival rate of 42 to 55 percent, and 10-year survival rate under 30 percent. Distant metastases have been reported after resection [42-44]. Proliferative activity of the cells is being investigated for use as a prognostic factor [43]. Radical surgery is the primary treatment with adjuvant chemotherapy/radiation used pre-operatively or for recurrence and metastases [42, 44]. Radiographs typically display a nonspecific permeative pattern of bone destruction and ill-defined periosteal reaction. Extensive extraosseous components are common and, while not always prominent, an “arc-and-rings” chondroid calcification pattern is noted in up to 67 percent of cases. Most are centered within the medullary cavity, but 6 percent may be surface lesions [21, 42, 45]. CT is helpful in further characterizing the findings noted on plain radiographs. Aggressive bone destruction and an associated soft tissue mass are common. The calcification tends to be stippled, but may appear subtle or heavy on CT [42]. Tumors often have foci of low attenuation, and this is believed to represent necrosis [21]. MRI tends to display low to intermediate intensity on T1-weighted images and intermediate signal on T2-weighted images. Contrast-enhanced images are helpful with mesenchymal chondrosarcoma, as enhancement is often diffuse without the typical pattern of septal and peripheral enhancement seen in other forms of chondrosarcoma. Some may also display high flow serpentine vessels on MRI, a feature not seen in other chondrosarcomas [21, 42]. The features of intermediate T2 signal (lower than other chondrosarcomas) and more intense enhancement on MRI can suggest the diagnosis, although tissue diagnosis remains necessary. On histology, mesenchymal chondrosarcomas display a characteristic bimorphic appearance, virtually pathognomonic, with islands of differentiated cartilaginous tissue surrounded by highly cellular zones with plexiform vascular networks [42]. Nuclear imaging with thallium has been reported as a means of identifying metabolically active lesions, and as a screening method for metastasis [42].
4.7
Dedifferentiated Chondrosarcoma
Dedifferentiated chondrosarcoma, also known as spindle cell chondrosarcoma, represents approximately 9 percent to 10 percent of all chondrosarcomas [21]. Multiple theories exist as to how these types of tumors arise, the most popular being
392
J. Pahade et al.
that a high-grade non-cartilaginous component arises in a lower grade, longstanding chondrosarcoma. Patients tend to be older than in other forms of chondrosarcoma, averaging 60 years old, and there is no gender predilection. Most tend to present with pain or pathological fracture. A soft tissue mass is seen in around 55 to 87 percent of tumors [21, 46, 47]. Lesions often arise from conventional chondrosarcomas or enchondromas [48]. Locations mirror those of conventional chondrosarcoma, with an intramedullary location and most commonly involving the femur [47]. Case reports of a peripheral location in bone similar to juxtacortical chondrosarcoma have also been published [37]. The chondroid component tends to be sharply demarcated from the non-cartilaginous component, which is often (in descending order of frequency) osteosarcoma, fibrosarcoma, or malignant fibrous histiocytoma (MFH) [47, 49-51]. Radiographic appearance varies, based on the size of the non-cartilaginous component of the tumor. Lesions usually display the features of conventional chondrosarcoma as discussed previously, while the dedifferentiated components are associated with more aggressive bone lysis, cortical destruction, and absent chondroid matrix [49]. CT and MRI help to display the two components (chondroid and non-chondroid) involved in these tumors, referred to as tumor bimorphism. Overall, one review showed evidence of bimorphism in approximately 35 percent of radiographs, 48 percent of CT scans, and 33 percent of MR images [49]. The soft tissue mass and chondroid component may be missed by plain radiograph [49]. The extraosseous soft tissue component is more likely to harbor the high-grade neoplasm than the intraosseous component [21]. With contrast, most lesions display heterogeneous enhancement, with around 50 percent displaying a more diffuse, prominent enhancement pattern characteristic of dedifferentiation [49]. The high-grade non-cartilaginous component often has soft tissue attenuation on CT with variable enhancement. On MRI, an important finding is lower T2 signal intensity of the dedifferentiated component relative to the adjacent chondroid tissue [50, 52]. Recognition of dedifferentiation can provide important information whereby to discuss prognosis; dedifferentiated chondrosarcoma is associated with poor prognosis, an aggressive pattern, and the ability to metastasize [51]. The fiveyear survival is approximately 10.5 to 24 percent with a median survival of 13 months. Most demonstrate poor response to systemic chemotherapy [47, 51]. The association between the type of tumor in the dedifferentiated component and prognosis has not been well-defined, although metastasis at diagnosis and a higher percentage of dedifferentiated component in the lesion results in a poorer outcome [47]. Imaging may play its largest role in guiding biopsy in these patients. It is imperative that the dedifferentiated component is identified and targeted during biopsy to render an accurate diagnosis for treatment planning [50, 52]. Wide or radical surgical margins are mandatory for treatment; however, it is unclear whether radical resection improves long-term results [48].
15 Imaging of Malignant Skeletal Tumors
5 5.1
393
Multiple Myeloma Introduction
Multiple myeloma is the most common primary skeletal malignancy, with approximately 14,000 new cases in the United States per year [53]. The median age at diagnosis is 65 years, with a higher incidence in men and African Americans [54]. The disorder is caused by a clonal proliferation of plasma cells. While the exact cause is not yet defined, many patients have been found to have an abnormal karyotype, and having a chromosome 13 deletion has been found to correlate with patient outcome [53, 54]. Clinically, the diagnosis is based on an elevated level of gamma globulin on serum protein electrophoresis, and the presence of at least 10 percent abnormal plasma cells in a bone marrow aspirate specimen [53, 54]. Assessment of tumor burden aids treatment planning and prognosis determination; this assessment can include both imaging and measurement of serum markers such as Serum β2-Microglobulin, C-reactive protein, and lactose dehydrogenase (LDH). The radiographic findings of multiple myeloma range from subtle to prominent. The classic descriptors of multiple myeloma on radiographs are diffuse osteoporosis and multiple “punched out” lucent lesions. The lucent lesions of multiple myeloma are caused by the increased osteoclastic cell response induced by the invasion of myelomatous cells into bone marrow [54]. Importantly, approximately 50 percent bone destruction must occur before a lucent lesion is visible on radiographs [54].
5.2
Myeloma Subtypes
Imaging plays an important role in the diagnosis and differentiation of many of the subtypes of monoclonal gammopathy, including monoclonal gammopathy of undetermined significance (MGUS), asymptomatic myeloma, and symptomatic multiple myeloma. Additional subtypes also include solitary plasmacytoma and extraskeletal plasmacytoma. The diagnosis of MGUS is based on mild elevation of gamma globulin on serum protein electrophoresis (SPEP). The patients should be asymptomatic and have no other clinical findings of multiple myeloma. These patients were once thought to have no osseous involvement [55], although some reports suggest that this may not be entirely accurate. A small series of 37 patients with MGUS revealed MRI abnormalities in seven patients, of whom four had diffuse patchy spinal MRI marrow signal abnormalities and three displayed focal abnormalities [56]. The patients with abnormal MRI findings progressed to require treatment for myeloma faster than those without MRI abnormalities, suggesting that the subset of MGUS patients with abnormal marrow signal on MRI may require closer follow-up. MGUS has a
394
J. Pahade et al.
Table 15.2 Durie/Salmon PLUS Staging System for Multiple Myeloma. Staging of multiple myeloma using imaging characteristics. The distinction between substages A and B is based on serum creatinine and the presence or absence of extramedullary disease (EMD): Substage A is serum creatinine < 2.0 mg/dL and no EMD. Substage B is serum creatinine > 2.0 mg/dL and/or EMD. Adapted from Durie, et al. [55]. Classification MRI and/or FDG-PET MGUS Stage IA (Smoldering or Indolent) Multiple Myeloma Stage IB Stage IIA/B Stage IIIA/B
All negative Single Plasmacytoma or Limited Disease on Imaging < 5 Focal Lesions, Mild Diffuse Disease 5-20 Focal Lesions, Moderate Diffuse Disease >20 Focal Lesions, Severe Diffuse Disease
risk of progression to multiple myeloma and other entities such amyloidosis, macroglobulinemia, leukemia, and lymphoma [53]. Asymptomatic myeloma, also known as indolent or smoldering myeloma, has slightly higher levels of gamma globulin on SPEP, as well as other laboratory abnormalities, but patients are still asymptomatic. Newly published consensus guidelines state that these patients most commonly have no imaging abnormalities, although a subset will have one lesion detected on imaging [53]. Symptomatic myeloma accounts for the majority of multiple myeloma cases and includes several subtypes such as classic, generalized, osteosclerotic and leukemic [53]. Imaging plays the largest role with symptomatic myeloma and has now been incorporated into the recently updated Durie/Salmon PLUS staging system [55, 57] (see Table 15.2). For purposes of the staging system, moderate diffuse spine involvement was defined as diffuse marrow abnormality on T1-weighted sequences with signal intensity of vertebral marrow brighter than the adjacent intervertebral disks, and severe diffuse spine disease was defined as diffuse marrow abnormality on T1-weighted sequences with signal intensity of vertebral marrow lower than or equal to that in the adjacent intervertebral disks [55].
5.3
Imaging in Multiple Myeloma
The current standard approach in patients with multiple myeloma relies on the skeletal survey, which is abnormal in 80 percent to 90 percent of newly diagnosed patients, and may be the only test needed for staging purposes, based on the number of lesions [53]. This usually consists of standard AP and lateral radiographs of the skull and spine, with additional anteroposterior radiographs of the ribs, pelvis and all long bones. Newer techniques involve the use of whole-body imaging C-arms [58]. Targeted radiographs in regions that are difficult to assess (including the ribs and scapula) may be ordered based on clinical suspicion or questionable abnormalities viewed on survey [53]. Skeletal survey displays the most abnormalities in the
15 Imaging of Malignant Skeletal Tumors
395
classic and leukemic subtype of symptomatic myeloma, while the generalized subtype displays only diffuse osteopenia with no focal lytic lesions. The osteosclerotic subtype displays sclerotic lesions and is often seen in the POEMS syndrome (polyneuropathy, organomegaly, endocrine disorders, monoclonal gammopathy and skin changes). Some myeloma patients may display both lytic and sclerotic lesions without any of the additional findings associated with POEMS syndrome [53]. It is important to remember that, while the majority of lesions associated with multiple myeloma are lytic, sclerotic subtypes exist and should be considered when interpreting radiographs in patients with suspected or known multiple myeloma. MRI has been established as an appropriate imaging modality in the diagnosis and follow-up of multiple myeloma, although no standard imaging protocols (including standards of anatomic coverage) have been established (Figs. 15.8, 15.9). A recent review on myeloma noted that no definitive conclusion can be drawn based on the available evidence and, instead, recommended inclusion of the entire skeleton or to broaden coverage as much as could be tolerated by the patient or the constraints of resources at one’s institution [53]. Several large reviews on myeloma have suggested the use of MRI to evaluate the skull, entire spine and pelvis [54], or performing whole-body MRI examination [58, 59] (Fig. 15.10). If staging with whole-body MRI indicates stage III disease, no other imaging may be needed. In contrast, findings suggestive of stage I or II disease may indicate a need for additional imaging to prevent understaging in up to 10 percent of patients [58, 60]. A recent review showed whole-body MRI and skeletal survey to be discordant in 24 percent of cases, with 19 percent of cases having false negative skeletal surveys and stage III disease by spinal MRI [61]. The use of gadolinium is also controversial, with some employing the intravenous contrast on initial scans and others recommending its use only on follow-up scans, suggesting that enhancing areas are likely to harbor persistent disease [53, 54]. Before an MRI-based diagnosis is made it is important to understand the spectrum of normal marrow appearance, based on patient age. On T1-weighted images, fatty marrow is hyperintense and cellular marrow is hypointense relative to the intensity of skeletal muscles. With age, there is a replacement of cellular red marrow to a more T1 hyperintense fatty marrow. Caution must be used when identifying diffuse T1 hypointense areas with hyperintense nonuniform or band-like end plate changes as abnormal, because these findings can be normally seen in over 85 percent of patients aged 40 to 50 years [54]. Accurately differentiating myeloma from other marrow infiltration processes or metastasis is important. Iatrogenic processes are particularly important to consider in myeloma patients; for example, marrow changes have been noted in patients with primary musculoskeletal malignancies receiving GCSF. One small study found that red marrow conversion from treatment resulted in low signal on T1 and mildly increased signal on T2 images, when compared to normal yellow marrow, findings that can also be found in myeloma and tumor metastasis to the marrow. The pelvis and proximal long bones were noted to be affected most frequently [62]. MRI may be used to assess tumor burden, with low burden usually being associated with no MRI abnormalities and high burden associated with diffuse hypointensity on
396
J. Pahade et al.
15 Imaging of Malignant Skeletal Tumors
397
T1 sequences, high signal on fluid sensitive sequences, and enhancement with gadolinium. Marrow signal may be homogeneous or heterogeneous. Numerous studies have shown that the pattern of diffuse marrow involvement, as detected by MRI, correlates with increased marrow cellularity, increased plasmacytosis, anemia and overall poorer survival when compared to patients with a normal MRI pattern [54, 63]. Lecouvet, et al. reported 37 percent 60-month survival rates in stage III patients with diffuse MRI abnormalities, compared to 70 percent in patients with a normal MRI pattern. Prior studies in early stage myeloma patients have also found a diffuse MRI pattern to be associated with early progression [63]. Spinal fractures are commonly associated with multiple myeloma, and are seen in approximately 55 percent to 70 percent of patients. In one review of spinal fractures imaged with MRI, 80 percent of single fractures were associated with a focal lesion, although only 46 percent of fracture sites in patients with multiple fractures were associated with a focal lesion. These results suggest that fractures can occur at sites that appear normal on MR images. Additionally, patients with diffuse disease on MR images have a shorter fracture-free interval than those without diffuse disease [54]. Compression fractures without associated edema on MRI have been reported [64]. In addition to morphologic and signal characteristic criteria, in-phase and opposed-phase MRI may help accurately distinguish spinal fractures due to benign versus malignant causes [65]. Multiple studies have demonstrated the efficacy of vertebroplasty in treatment of myeloma-related fractures [64]. Recent work has evaluated the use of CT in diagnosis of myeloma, and in predicting the risk of impending spinal fractures. It is generally accepted that CT is superior to conventional radiographs in identifying lytic lesions in certain areas of the body, including the ribs, skull, and axial skeleton, and recent reports have found that whole-body CT detects more focal lesions and can lead to changes in staging [53, 54]. However, replacement of the conventional skeletal series with wholebody CT has not been universally accepted due to increased radiation exposure. Low dose protocols have been developed and found to be successful in diagnosis and identifying lesions at high risk for fracture [66]. CT can also be used to guide biopsy for lesions defined by MRI, as larger gauge needles are difficult to use in the MRI environment. Targeted biopsies can increase yield and improve detection of cytologic and cytogenetic abnormalities, resulting in the alteration of a patient’s treatment plan [54]. Bone scintigraphy with 99 mTc-MDP generally does not play a significant role in staging or diagnosis of myeloma, as the modality relies on osteoblastic and not osteoclastic activity; as a result, myelomatous lesions do not demonstrate increased
Fig. 15.8 Comparison of radiographs, CT and MRI in the imaging of the lumbar spine in an 88year-old male with multiple myeloma. (a) Lateral radiograph, (b) sagittal CT reformat, (c) and sagittal STIR MRI demonstrate varying sensitivity for detection of individual lesions. Note the different appearance of the dominant L2 lytic lesion (arrowheads). The MR images reveal numerous lesions that are subtle or occult on the other modalities
398
J. Pahade et al.
15 Imaging of Malignant Skeletal Tumors
399
Fig. 15.10 Normal whole-body STIR MRI scans from newly diagnosed 60-year-old man with multiple myeloma. Images courtesy of Dr. Michael Mulligan. Reprinted from [58]
Fig. 15.9 (a) Sagittal T1 and (b) sagittal STIR MR images of the thoracic spine, (c) axial CT image of the chest including the sternum, and (d) bone scintigraphy in a patient with multiple myeloma. Sagittal T1 MRI image shows multiple lesions replacing the high intensity marrow, consistent with imaging Stage II myeloma. STIR MR image shows multiple hyperintense lesions that contrast with the low signal marrow fat, as well as a compression fracture (white arrowhead). CT demonstrates a sternal lesion (arrow) that is evident as a region of a photopenia with surrounding uptake on the bone scan (white arrowhead)
400
J. Pahade et al.
uptake on scintigraphy. Multiple studies have confirmed that bone scans tend to underestimate the extent of disease [54]. 67Galium citrate, 99Tc MIBI and, 201Tl chloride have all been investigated as potential agents for use in myeloma with promising results, although none have been integrated into standard care for myeloma patients. Focal MIBI uptake appears to be better associated with active myeloma than diffuse uptake [67].
5.4
The Role of FDG-PET
FDG-PET appears to be an acceptable complementary imaging modality to stage and monitor treatment (Fig. 15.11). Mixed results have been obtained when comparing PET/CT to other imaging modalities, including skeletal survey, CT, and MRI. Breyer, et al. used SUV values of greater then 2.5 and found that PET/CT identified approximately 104 (14 soft tissue and 90 osseous lesions), of which 57 (55 percent) were new or previously undetected. In this series, conventional skeletal series missed 56/57 of the lesions, while CT failed to detect nine of these sites. PET had a low sensitivity in this study, as 133 sites felt to represent myeloma were identified by the other modalities (radiograph, CT or MRI) and not detected on PET. Many of these were missed due to their small size, as PET has limited spatial resolution and suffers from volume averaging effects [68]. However, Nanni et al. showed PET/CT to be more sensitive then whole-body X-ray by identifying lesions below the contrast resolution of X-ray and those with less then 50 percent bone resorption [69]. Comparing PET to MRI has also been mixed: while PET can detect lesions outside of MRI’s field of view (particularly when whole-body MRI is not available), there is diminished sensitivity in identifying diffuse spinal disease evident on MRI [68, 69]. PET is also subject to false positives caused by infection, inflammation, post-surgical or radiation changes, and hemangiomas [68, 69]. Detection of previously occult lesions on PET scan can upstage patients [68, 70]. When interpreting PET scans, an SUV value threshold of 2.5 may be inappropriate, as smaller myeloma lesions have been evaluated in retrospect to be FDG avid with values below the threshold of 2.5. This has prompted some to suggest that any FDG uptake (regardless of SUV value) in lesions smaller then 5 mm should be viewed with suspicion as an additional focus of disease [58]. At least one report suggests that following treatment, a decrease in uptake values on FDG-PET has predicted clinical outcome [70]. Future work with PET or PET/CT will likely focus on using imaging to help determine which lesions are “active,” and whether this finding can be incorporated into staging systems. The combination of PET/CT may also enable simultaneous evaluation of lesions with two distinct radiologic methods. Some institutions have recommended use of FDG-PET routinely in patients with non-secretory myeloma, solitary plasmacytoma of the bone, or extramedullary plasmacytoma and consider it as a potential modality for future use in routine radiologic follow-up in myeloma patients to examine disease response [58].
15 Imaging of Malignant Skeletal Tumors
401
Fig. 15.11 PET/CT in multiple myeloma. (a) Coronal CT and fused PET-CT images in 71-yearold female with myeloma and new neck mass. FDG avid focus (SUV=3.5) is noted within enlarged left neck lymph nodes (arrows), representing a biopsy proven extramedullary myeloma recurrence. Images courtesy of Dr. Michael Mulligan. Reprinted from (58). (b) CT and fused PET/CT axial images in 42-year-old male with biopsy proven lytic solitary plasmacytoma (arrows) with standardized uptake value (SUV) of 10
402
5.5
J. Pahade et al.
Solitary Plasmacytoma and Extramedullary Plasmacytoma
Solitary plasmacytoma accounts for 2 percent to 5 percent of most cases of myeloma [53, 54]. The most common site of involvement is the spine. Flat and long bone lesions can have almost any radiographic appearance, from benign-appearing to aggressive [53]. Patients present with bone pain, and treatment consists of local radiation, although patients are prone to developing symptomatic multiple myeloma within a short time of diagnosis. Radiologic work-up of these lesions starts with the classic skeletal survey. MRI has been suggested as an adjunctive imaging procedure, as it displays abnormalities not picked up on original survey in one-third of patients and helps better define soft tissue extent for radiation therapy [71]. Lesions greater than 5 cm have been associated with poorer prognosis [53]. Extramedullary plasmacytoma accounts for approximately 3 percent of myeloma cases and has a 3 to 1 male predominance. Most cases are found within the head/neck region, notably involving the paranasal sinuses and oropharynx [53]. Imaging work-up for these lesions is similar to solitary plasmacytoma, with the adjunctive role of MRI and FDG-PET not firmly established [53]. Treatment is radiation therapy or surgical excision.
5.6
Multiple Myeloma: Follow-up
Acceptable protocols for radiologic follow-up in myeloma patients are not well established. Some authors believe that routine radiological (in addition to serologic) follow-up is not indicated in myeloma, while others employ MRI to help determine treatment response. Defining an acceptable standard imaging follow-up regimen is still necessary in asymptomatic patients; changes in a patient’s symptoms is obviously an indication for re-assessment [58].
6 6.1
Metastasis Introduction
Metastases represent the most common type of malignant bone lesion, accounting for approximately 70 percent of bone tumors, and should always be considered in the differential diagnosis of a skeletal lesion, particularly in older patients. Bone metastasis is 25 times more common than primary bone tumors [72]. Following the lungs and liver, the skeletal system is the third most common location of distant metastases. The most common sources of osseous metastases include breast, prostate, lung, colon, stomach, rectum, uterus, bladder, renal and thyroid primary malignancies; of these, breast and prostate cancer represent the most common primary
15 Imaging of Malignant Skeletal Tumors
403
sites [73]. Clinically, patients can present with severe bone pain, bone tenderness, soft tissue mass, pathologic fracture and spinal cord compression, all of which reduce quality of life and worsen prognosis. Life-threatening hypercalcemia is another possible effect of bone metastasis [73]. Radiographs, scintigraphy, CT and MRI are all currently utilized for detection and monitoring of metastases, with PET and whole-body MRI providing promising diagnostic potential. Despite this variety of modalities, metastases present a diagnostic challenge for radiologists due to the wide variability in radiologic appearance and the difficulty in measuring response to treatment.
6.2
Mechanisms and Radiographic Appearance
Metastatic disease of the skeleton can arise from direct extension, hematogenous or lymphatic dissemination, or intraspinal spread of tumor. Bone represents an excellent site for metastatic tumor cells, due to high blood flow (particularly in red marrow), adhesive molecules on tumor cells that can induce production of angiogenic and resorptive factors when bound to bone marrow stromal cells or bone matrix, and immobilized growth factors that are present to support ongoing bone remodeling and resorption [73]. Direct extension, such as a lung cancer invading the ribs, typically involves a soft tissue mass and osseous destruction. Lymphatic spread is particularly relevant for pelvic cancers such as prostate, bladder or gynecologic cancers, where spread to local lymph nodes can then directly invade adjacent structures. Hematogenous spread can occur via arterial or venous routes. Intraspinal spread can occur when an intracranial neoplasm gains access to cerebrospinal fluid, allowing drop metastases to form within the spinal canal with secondary invasion of the vertebrae. The typical classification of bone metastasis distinguishes between osteolytic (lucent) and osteoblastic (sclerotic) lesions. These broad categories represent the two extremes of abnormal bone metabolism and remodeling that occurs in the presence of an osseous metastasis [73]. Osseous structures respond to neoplastic infiltration with varying degrees of bone resorption and formation. Resorption is thought to occur by osteoclast activating factors released by tumor cells, among other mechanisms. The mechanism of osteoblastic metastases is less understood. There is evidence, however, of a “vicious cycle” in breast and prostate cancer metastases in which tumor cells promote proliferation of osteoclasts and viceversa [73]. The distribution of the radiographic appearance of metastatic lesions includes osteolytic (75 percent of metastases), osteosclerotic (15 percent) and mixed lytic/ sclerotic (10 percent) (see Table 15.3). Lytic metastases can arise from the breast, thyroid, kidney, lung, breast, gastrointestinal tract, adrenal gland, uterus, Ewing’s sarcoma, squamous cell carcinoma and pheochromocytoma. Bubbly expansile lytic metastases are seen in renal and thyroid metastases. Sclerotic lesions are typically seen with prostate cancer, as well as carcinoid tumors, medulloblastoma and
404
J. Pahade et al.
Table 15.3 Common Appearance of Skeletal Metastases. Note that the table presents the most common appearance of each primary tumor type, although any individual lesion can have a different appearance, particularly following treatment Primary Tumor Type Imaging Appearance Breast Lung Prostate Kidney Thyroid Melanoma Bladder Esophagus Stomach Colon Pancreatic Uterus/Cervix Ovarian Neuroblastoma Retinoblastoma
Lucent or Mixed Lucent or Mixed (Occasionally Sclerotic with Small Cell Carcinoma and Adenocarcinoma) Sclerotic Lucent, Expansile Lucent, Expansile Lucent, often with soft tissue mass Lucent Lucent Lucent (Sclerotic in Mucinous subtypes) Lucent or Mixed (Sclerotic in Mucinous subtypes) Lucent Lucent Lucent, Mixed, or Sclerotic Lucent, Permeative Lucent, Permeative
osteosarcoma. Mixed lytic/sclerotic lesions can be seen in almost any metastasis, including from breast, lung, cervical, ovarian, colon and testicular tumors. Metastases have a predilection for the red marrow-rich axial skeleton, particularly the lumbar and thoracic vertebral bodies, pelvic bones, sternum and ribs. The vertebrae house 75 percent of the body’s bone marrow and are highly vascular, making them especially susceptible to metastatic spread. Vertebral metastases most commonly arise from carcinomas of the lung, breast, and prostate, as well as lymphoma and myeloma. Sclerosis of a vertebral body, such as an “ivory vertebrae” (increased density of the entire body) or partial sclerosis, can be seen most commonly in prostate metastases, as well as lymphoma, myeloma, and Paget’s disease. Spinal metastases may also present as a malignant vertebral body compression fracture; differential possibilities would include both malignant (myeloma) and benign (osteoporosis) etiologies. Characteristics more commonly seen in metastases than in these other entities include compression of an upper level thoracic vertebrae, associated soft tissue mass, and destruction of the pedicle, usually by direct extension from the vertebral body. Definitive diagnosis can be made by biopsy. In the appendicular skeleton metastatic disease is most commonly proximal in location; it is rare to have metastatic disease distal to the elbow or knee. Notable exceptions are bronchogenic carcinoma and breast cancer, which can present with widespread skeletal metastases and together account for 50 percent of metastases distal to the elbows and knees. Relatively common locations for metastases are the proximal metaphyses of the humerus or femur, usually presenting as a medullary
15 Imaging of Malignant Skeletal Tumors
405
lesion with later involvement of the cortex. These lesions usually do not have a significant soft tissue mass or periosteal reaction, in contradistinction to primary bone tumors. In long tubular bones, pathologic fracture can occur once more than 50 percent of the cortical thickness is destroyed.
6.3
Imaging Modalities
Each imaging modality confers a specific subset of information that adds to the diagnostic query. Plain radiography and CT can demonstrate bone structure, CT and MRI visualize the tumor and bone marrow, bone scanning and SPECT can reveal osteoblastic metabolism, and PET can visualize tumor metabolism. MRI and PET can potentially detect early bone marrow changes, before structural changes are visible. Radiographs: On radiographs, the appearance of metastases is highly variable, taking on any pattern of bony destruction (geographic, moth-eaten, permeative), with poorly or well-defined margins. Lesions can be lytic, sclerotic, or mixed lytic/sclerotic. Approximately 30 percent to 50 percent of normal bone mineral must be lost before a bone metastasis becomes visible on a plain radiograph, so metastases may remain occult for up to three to six months [72]. As most metastases spread hematogenously, tumor emboli tend to lodge in the marrow, with only later involvement of the cortex. Differentiating metastases from primary bone tumors, infection, or metabolic lesions can be very challenging due to variable radiologic characteristics. In general, multiple lesions are more suggestive of metastases, though solitary metastases can occur (particularly with renal and thyroid metastases), and it can be difficult to differentiate solitary metastases from primary bone tumors. When multiple, metastatic lesions are usually variable in size, in contrast to multiple myeloma lesions which tend towards uniformity. Periosteal reaction and soft tissue mass are usually limited, in contrast to primary bone tumors. Additionally, metastatic lesions tend to be smaller, averaging 2 to 4 cm in diameter. Cortical metastases, which usually affect the femur, can be seen with bronchogenic carcinoma, melanoma, and cancers of the bladder and kidney. Radiographically, a lytic lesion that is responding to chemotherapy or radiation therapy can develop peripheral sclerosis that moves centrally, with eventual resolution of the sclerotic area. Healing mixed lytic/sclerotic lesions become progressively more sclerotic. Meanwhile, lesion expansion or new zones of lysis usually indicate progression of disease. New areas of sclerosis can signify disease progression or healing of a previously unrecognized lesion. Bone scintigraphy: Bone scintigraphy is the most commonly used modality in screening for metastases, with an approximate sensitivity of 62 percent to 100 percent, and specificity of 78 percent to 100 percent [74]. Bone scintigraphy is a useful screening tool as it can provide rapid whole-body images at a reasonable cost. Most commonly, metastases are seen as areas of increased tracer uptake (“hot spots”),
406
J. Pahade et al.
which reflects hyperemia, reactive repair, and new bone formation at the periphery of the lesion. Both lytic and sclerotic lesions tend to have increased tracer uptake. When the bony destruction is extensive, an area of decreased tracer uptake (“cold spot”) can be seen. In challenging areas, such as the thoracolumbar spine and pelvis, tomographic information available using SPECT (single photon emission computed tomography) can be used to further define the anatomy. Diffuse uptake by the axial skeleton, referred to as a “superscan,” can be seen with diffuse prostate or breast metastases, and can be easily mistaken as normal because of the uniform increase in uptake. This pattern can be identified by the lack of tracer uptake in the genitourinary system and distal appendicular skeleton. Isolated foci of increased uptake (particularly involving the ribs or joints) in a patient with a known primary malignancy are usually benign lesions due to degenerative changes or old fractures. Limitations of skeletal scintigraphy include the broad list of differential possibilities for a focus of increased radiopharmaceutical uptake. A wide variety of neoplastic, infectious, inflammatory, and traumatic etiologies can lead to a false positive scan. Metastatic lesions that do not have aggressive new bone formation or are rapidly expanding, may not demonstrate increased uptake, resulting in a false negative. Correlation with radiographs or CT, as necessary, can increase specificity of bone scintigraphy up to 95 percent [72]. A specialized type of scintigraphy – bone marrow scintigraphy – uses technetium-99 labeled monoclonal antibodies such as the NCA095 antibody to detect early bone marrow infiltration. In particular, bone marrow scintigraphy has been shown to be superior to conventional bone scintigraphy for certain tumor subtypes, such as small cell lung cancer and breast cancer with osteolytic metastases [72]. During therapy, tracer uptake on sequential scans cannot reliably be correlated with tumor behavior. A lesion may demonstrate increased uptake due to either disease progression or tumor healing. In the first three months post-therapy, 75 percent of metastases have increased tracer uptake, dubbed the “flare phenomenon,” which subsequently subsides by six months post-therapy. Similarly, decreased tracer uptake may be due to healing or, less commonly, to advanced osseous destruction. Rather, the presence of new lesions is a more accurate indicator of disease progression. Again, correlation with the clinical picture and radiographic pattern are essential. Computed Tomography (CT): CT is more sensitive than radiographs for detecting subtle osseous destruction, as well as an associated soft tissue mass. The sensitivity of CT ranges from 71 percent to 100 percent, and it is particularly useful in cases of a positive bone scan and equivocal radiographs. While cortical destruction and large lytic lesions are easily visualized on CT, subtle marrow lesions may be difficult to identify prospectively without physiologic information that can be obtained on bone scintigraphy or PET scan. CT findings suggestive of metastases include lytic lesions with associated cortical and trabecular destruction and high attenuation areas of soft tissue density within the predominantly low attenuation fatty marrow. Like radiographs, sclerosis usually suggests response to tumor, while progressive lysis can indicate disease progression. It is particularly important to examine the
15 Imaging of Malignant Skeletal Tumors
407
entire imaged skeleton using display parameters optimized for visualizing bone (“bone windows”) and multiple plane reformats. Magnetic Resonance (MRI): MRI has a diagnostic sensitivity of 82 percent to 100 percent and specificity of 73 to 100 percent for bony metastases, and it is particularly sensitive in detecting bone marrow lesions. A number of recent studies have shown that MRI detects marrow changes even earlier than bone scintigraphy [72]. Metastases in the bone marrow lead to a longer T1 relaxation time (low T1 signal) due to edema and displacement of the marrow fat. Variable changes in the T2 relaxation time are seen, but lytic tumors tend towards high T2 signal, while sclerotic tumors have a low T2 signal. Solid, cell-rich tumors have intermediate T2 signal. Adding T2-weighted sequences with fat suppression results in suppression of the normal bone marrow and excellent contrast between the lesion and normal tissue. Phase shifted gradient echo imaging is very sensitive for detection of metastases. Gadolinium enhancement does not necessarily help with detection of metastases, although it can help delineate tumor extent and differentiate tumor necrosis from viable tumor. It can be challenging to differentiate metastases from reconversion of hematopoietic bone marrow, which also results in low T1 and high T2 signal. Reconversion represents change from yellow marrow (predominantly fat) to red marrow (more cellular), and usually follows a predictable pattern from proximal to distal. A basic rule-of-thumb is that red marrow still has higher signal than muscle on T1-weighted images, in contrast to marrow-replacing neoplastic processes. Schweitzer et al. described a characteristic rim of T2 hyperintensity (edema) around a low signal intensity metastatic lesion, dubbed the “halo sign” [75]. Benign lesions have more of a “bull’s eye” appearance with central T1 hyperintensity surrounded by a T1 hypointense lesion [75]. Gradient echo sequences are also useful in detecting normal hematopoietic bone marrow [72]. Another common diagnostic dilemma is differentiating pathologic vertebral body fractures from osteoporotic fractures. In general, the following findings are more characteristic of pathologic fractures: convex posterior contour of the posterior cortex of the vertebral body, abnormal signal in the pedicle or vertebral arch, associated soft tissue mass, normal signal in the adjacent disc, restricted diffusion on diffusion-weighted imaging and diffuse low T1 signal with corresponding increase in T2 signal intensity [72, 76]. Due to its high sensitivity for metastatic detection and ability to detect bone marrow disease earlier than bone scintigraphy, MRI has been proposed as a screening modality. Whole-body MRI (WBMRI) is a promising new staging method that can evaluate for both bony and parenchymal lesions while decreasing the time and expense involved in having multiple staging exams using several different modalities. In the past, WBMRI was technologically infeasible due to limited field of view, coil limitations, and unreasonably long exam times. Recent advances, including a whole-body coil system, new table concepts, and ultra-fast data acquisition have enabled researchers to attain high resolution images of the whole body within 14 minutes. Lauenstein et al. compared staging work-up in 51 patients using WBMRI
408
J. Pahade et al.
versus a multimodality approach employing bone scintigraphy, CT and MRI. WBMRI was able to detect all cerebral, pulmonary and hepatic metastases greater than 6 mm, resulting in sensitivity and specificity of 100 percent. Mean scan time was only 14.5 minutes. Tiny lung nodules that were missed by MRI did not change overall management [77]. There is limited data comparing WBMRI and PET/CT for detection of osseous metastases, although reported sensitivities range between 80 percent to 100 percent for both modalities, and there is evidence of overall concordance of 93 percent between the two modalities [72]. In a recent prospective study, Schmidt et al. compared WBMRI (using 32 channel parallel imaging) to PET/CT for the detection of bone metastases in 30 patients with a known primary tumor and suspected metastases. WBMRI had a sensitivity of 94 percent and specificity of 76 percent, while PET/CT had a sensitivity of 78 percent and specificity of 80 percent; at least part of this difference in performance is attributable to increased sensitivity for small lesions, with a minimum size threshold of 2 mm for WBMRI and 5 mm for PET/CT [78]. These results suggest that WBMRI may have a slightly superior diagnostic accuracy to PET/CT for metastatic screening, without exposing the patient to ionizing radiation [78]. Continued research in this modality should further define its performance and increase implementation at a broader range of centers. Positron Emission Tomography (PET): The use of PET in the evaluation of malignancy has continued to increase, though its sensitivity in evaluating for bone metastases is not well established. FDG-PET has a sensitivity ranging from 62 percent to 100 percent, with a specificity from 96 percent to 100 percent [74]. When compared to bone scintigraphy, PET has a similar sensitivity through poorer specificity in detecting breast and lung metastases. PET is less sensitive in detecting prostate metastases, due either to the decreased metabolic activity of prostate metastases or the increased sensitivity of bone scintigraphy for the osteoblastic activity characteristic of prostate metastases [72, 79]. One advantage of PET is increased spatial resolution relative to bone scintigraphy. PET also has increased sensitivity relative to radiograph or CT, which do not permit reliable visualization of subtle marrow lesions, suggesting that PET might be an important tool in detecting metastases before morphologic changes are apparent by CT or CR [80]. Disadvantages of PET include relatively lower specificity, high cost, and lack of availability.
6.4
Skeletal Metastases of Unknown Origin
In any particular patient, particularly an older patient, an aggressive lytic lesion of the skeleton is most likely to represent a metastatic lesion, given the relative rarity of primary bone malignancies. In patients with an unknown primary tumor, a protocol that includes screening with chest radiograph followed by CT of the torso (including chest, abdomen and pelvis) can identify 85 percent to 90 percent of primary tumors [81]. Employing mammography in the scenario in women with a tumor of unknown origin is controversial, particularly in the context of a normal
15 Imaging of Malignant Skeletal Tumors
409
breast exam, since the yield of this additional test is uncertain. Importantly, biopsy of skeletal metastases often fails to identify the primary tumor, with undifferentiated or poorly differentiated carcinoma as the histologic diagnosis in up to 65 percent of cases [81].
6.5
Image-Guided Treatment of Painful Bone Metastases
Treatment with external beam radiation is the current standard of care for patients with localized bone pain. Along with analgesics, chemotherapy, hormonal therapy and bisphosphonates, up to 70 percent of patients experience significant pain relief. Patients with spinal cord compression and a high life-expectancy may benefit from surgery, including debulking and vertebroplasty. However, the 30 percent of patients who are refractory to standard therapy provide a therapeutic challenge, and new image-guided treatment options, including radiofrequency ablation and cryoablation, show promise in helping this subgroup of patients (see below).
6.6
Summary of Bone Metastases
Bone metastases are most commonly seen with cancers of the breast, prostate, lung, kidney, and thyroid gland, and characteristically present as multiple lesions in the axial skeleton of a patient over the age of 40. Imaging of bone metastases presents a diagnostic challenge due to their highly variable pathophysiology and radiologic appearance. Solitary lesions must be differentiated from primary bone tumors, while multiple lesions must be differentiated from multiple myeloma and lymphoma. Treatment response can be very challenging to ascertain by all of the modalities due to the variable appearance of metastases and the variety of ways in which they respond. There are no current standardized criteria for following bone metastases, and these lesions are considered “non-target” lesions under the widely-used RECIST system.
7
Primary Bone Lymphoma
Primary bone lymphoma (PBL) is rare, accounting for only 5 percent of primary bone tumors (Figs. 15.12, 15.13). By definition these tumors are a diagnosis of exclusion and can only be considered when there is a focus of biopsy-proven lymphoma in a single bone, with no evidence of distal lymphatic or soft tissue involvement for six months. Regional lymph node involvement does not exclude a diagnosis of PBL. Non-Hodgkin’s Lymphomas represent the vast majority, usually
410
J. Pahade et al.
15 Imaging of Malignant Skeletal Tumors
411
Fig. 15.13 Lymphoma in a 36-year-old male. (a) Radiograph shows a permeative lesion involving the acromion and extending into the scapular body (arrowheads). (b) Axial T2 and (c) axial T1 post-gadolinium fat saturation MR images of the scapular body reveal a large soft tissue mass centered at the scapular body (arrowheads) that displaces the rotator cuff musculature
Fig. 15.12 Primary lymphoma of bone in a 59-year-old female. (a) Frontal radiograph of the knee shows an ill-defined lytic lesion of the proximal tibia (arrowheads). (b) Axial CT scan performed at the time of percutaneous biopsy shows the lytic lesion with ill-defined margins (black arrows) and cortical destruction (white arrowheads). (c) Coronal proton density, (d) coronal STIR and (e) coronal post-gadolinium T1-weighted MR images of the proximal tibia show the lesion extending through the tibial cortex and to the articular surface of the proximal tibia. (f) Bone scintigraphy shows increased uptake in the proximal tibia, corresponding to the site of the lesion
412
J. Pahade et al.
of the large B-cell or mixed small and large B-cell lineage. Hodgkin’s disease accounts for only 6 percent of cases [82]. There is a broad distribution of patients affected, ranging between 20 to 80 years old, with a peak prevalence in the sixth and seventh decades. Clinically, patients present with symptoms similar to many other primary bone tumors: either asymptomatically or with insidious bone pain or swelling. Rarely, systemic symptoms such as weight loss and fever can be present. Chemotherapy, with or without radiation, leads to a five-year survival of 83 percent to 90 percent. Younger patients with disease confined to a single location have a particularly excellent prognosis [83, 84]. Therefore, early radiographic detection and appropriate diagnosis is key. It is important to consider the possibility of lymphoma at the time of biopsy to ensure that appropriate cytologic analysis and immunophenotyping are performed during biopsy. There is wide variability to the radiographic appearance of PBL, with certain features being more characteristic. A retrospective analysis from the AFIP of 237 pathologically proven cases of primary bone lymphoma revealed that 70 percent of lesions were lytic with 74 percent of these lesions showing a permeative or motheaten appearance; 71 percent of lesions occurred in the long bones; periosteal reaction was seen in 58 percent, commonly in a layered pattern; sequestra were found in 16 percent; and soft tissue masses were seen in 48 percent [82]. The metadiaphysis of the femur is, by far, the most common location for PBL, accounting for 25 percent of cases [83]. The metadiaphysis of the proximal tibia is also frequently involved. Cortical destruction, pathologic fractures, and large soft tissue masses indicate a more aggressive pattern and poorer prognosis [83]. Sclerosis can appear following therapy, leading to a mixed lytic and sclerotic appearance. As primary lymphoma of the bone is a marrow-replacing process, findings can be subtle or even occult on radiography. If the patient has continued symptoms with a negative radiograph, a more sensitive study such as bone scintigraphy or MRI can be beneficial. On bone scintigraphy, primary bone lymphoma almost uniformly shows increased tracer uptake, reflecting increased osteoblastic activity. However, these findings are nonspecific and should be correlated with the radiographic findings. MRI is the most sensitive modality for identifying and defining the extent of marrow involvement and soft tissue mass seen in PBL. Marrow replacement presents as low signal intensity on T1 with varied, though usually high, signal intensity on T2. Peritumoral edema also produces high T2 signal, while fibrosis within the lesion will decrease the T2 signal intensity. STIR images are particularly useful in showing the soft tissue mass. These lesions will also enhance following contrast administration [83]. The differential for PBL includes other permeative lesions such as osteosarcoma, metastatic disease, round cell tumors such as Ewing’s Sarcoma and multiple myeloma, and secondary osseous lymphoma. Secondary osseous lymphoma can only be excluded after whole-body surveillance for distant lesions, either by PET or whole-body CT. Finally, it is important to distinguish PBL from osteomyelitis. In the AFIP series, 16 percent of PBL cases had a sequestra, and 5 percent of cases had involve-
15 Imaging of Malignant Skeletal Tumors
413
ment across the joint space to involve contiguous osseous structures, both of which are features that are more characteristic of infection [82]. Infection can be more definitively ruled out by a negative 99 mTc-WBC scintigraphy study. Clinical history is invaluable in narrowing the differential. Chemotherapy is the mainstay of treatment, with or without adjuvant local radiation therapy, leading to survival rates of 83 percent to 90 percent. Follow-up after therapy is usually performed with MRI. Similar to other bone tumors, it is challenging to differentiate residual tumor from granulation tissue and necrosis. One series demonstrated a dramatic decrease in tumor volume in the first three months after initiation of therapy with 71 percent to 96 percent reduction by five months [84], with no change in signal characteristics of the lesion and no development of necrosis. Additionally, a pattern resembling bone infarct was observed following treatment, with a linear hypointense rim on T1 and adjacent high signal on T2-weighted images. Paralleling the tumor volume reduction, the soft tissue component almost universally disappeared by three to four months of therapy. The authors concluded that follow-up MRI should be performed at two to three months and six to 12 months after initiation of therapy [84]. Primary bone lymphoma is a rare entity, but should be considered for any adult with a solitary metadiaphyseal lesion in the distal femur, or proximal tibia that has a permeative lytic appearance with associated soft tissue mass and minimal cortical destruction. Work-up of these lesions include plain radiographs and MRI. Correct diagnosis is essential for early and appropriate treatment, which can lead to an excellent survival rate.
8
Ewing’s Sarcoma
Ewing’s sarcoma, a pediatric small round cell blue tumor, is the second most common primary bone malignancy in children and adolescents, after osteosarcoma. It is slightly more common in boys, with the second decade of life being the most common age of diagnosis. Whites are most commonly affected, while the tumor is rare in the African American population. Pain is the most common presenting symptom and is often attributed to bone growth or traumatic injury. Pain without trauma, continuing at night, and lasting over one month should prompt further work-up [85]. Tumor growth eventually leads to a palpable mass. Most Ewing’s sarcoma occur in bone, with the pelvic bones (26 percent), femur (20 percent), tibia/fibula (18 percent) and ribs (10 percent) most commonly involved. Unlike osteosarcoma, Ewing’s sarcoma tends to originate in the diaphysis, rather then metaphysis. However, in clinical practice, localization in the metadiaphysis makes distinct classification difficult. Primary cranial, spinal (usually the sacrum or involving the posterior elements) and periosteal location without extension into the marrow is rare, but reported [86-88]. Primary metastasis to the lung, bone or bone marrow, or a combination of these sites is present in 25 percent of cases. Lymph node metastasis is rare [85].
414
J. Pahade et al.
The initial imaging modality for patients suspected of Ewing’s is plain radiographs. Classically one will see an aggressive-appearing, permeative, diaphyseal lesion with a raised periosteum (Codman Triangle) or onion-skin periostitis, and calcifications within a surrounding soft tissue mass [85]. The differential of such a lesion in a child includes infection and eosinophilic granuloma, with a benignappearing (thick, wavy) periosteal reaction suggesting eosinophilic granuloma (rare in Ewing’s) and a bony sequestrum suggesting infection. A purely permeative lesion may not always be seen, as sclerotic reaction may also occur in response to the tumor, creating a more patchy lytic appearance. Staging of Ewing’s usually includes a chest CT, as lung or pleural involvement is the most frequent site of metastatic disease, bone marrow aspiration and a bone scan to detect occult skeletal lesions. Contrast-enhanced MRI may be used to assess bone marrow involvement or soft tissue extent, with non-contrast T1-weighted images shown to correlate best with tumor size on pathologic examination [89, 90]. MRI also appears to be the best modality to assess rare skip lesions in the bone, and may also be performed prior to biopsy to help guide biopsy [89]. It is suggested that MRI be performed prior to any procedure, as post-procedural changes can be confused with tumor involvement. The use of FDG-PET is also becoming more frequent. PET and whole-body MRI have both been shown to detect more skeletal lesions than bone scan alone [90]. PET is slightly more sensitive than whole-body MRI in detecting skeletal lesions [90, 91]. The combination of whole-body MRI and bone scan may lead to results comparable with PET scanning alone, with less cost and fewer false positives [91]. Identifying distant lesions is important because they impact treatment protocols and are associated with poor prognosis (most notable for bone marrow metastasis) [85]. Confirmed local disease is amenable to surgery, while diffuse metastatic osseous disease will often alter the radiation therapy treatment plan. Imaging is also playing a larger role in evaluating response to therapy. A small study indicated that FDG-PET correlates with histological response to neoadjuvant chemotherapy, and SUV values less then 2.5 in lesions after therapy have been shown to be predictive of progression-free survival, regardless of initial stage [92]. Timing of PET exam acquisition post-therapy has not been well-defined. Whether these results will alter management is also yet to be concluded, although early identification of non-responders may allow change in chemotherapy, and alteration of radiation therapy and surgical planning [90, 92]. MRI has also been evaluated to help monitor response to chemotherapy, especially after induction chemotherapy where post-therapy necrosis, currently determined following surgery, serves as a prognostic factor [93]. Work is currently being done to evaluate the efficacy of dynamic MRI, which utilizes the enhancement characteristics of tumors to help determine which tumors may respond to chemotherapy, and to correlate changes after chemotherapy with tumor necrosis [89, 93]. High-grade necrosis of tumor following induction therapy with wide, negative surgical margins have been shown to correlate with low local recurrence [93]. T2 signal variations in suspect lesions were initially thought to be correlated with good histologic response to therapy;
15 Imaging of Malignant Skeletal Tumors
415
however, residual areas of viable Ewing’s sarcoma have been found in areas with both low and high T2 signal [89]. Cure from Ewing’s can only be achieved with chemotherapy and local control, with combination of multi-modality therapy resulting in cure rates of 50 percent or more in local disease. More widespread disease, often involving the bone or bone marrow, has a less than 20 percent chance of cure [85]. Other prognostic factors being evaluated include bone marrow micrometastasis and circulating tumor cells detected by reverse transcription PCR [89].
9 9.1
Image-Guided Procedures Percutaneous Biopsy
Percutaneous, image-guided biopsy of bone and soft tissue lesions is a widely used, minimally invasive, safe, and cost-effective approach to obtaining diagnostic tissue samples. The procedure can often be done with local anesthesia or moderate sedation, and has a low complication rate and rapid recovery time [94-96]. Most importantly, it can eliminate the need for open biopsy, which has been shown to have a 2 to 20 percent complication rate, allowing quicker treatment initiation [97, 98]. Complications can occur with percutaneous biopsy, and may be higher in nonspecialized oncology centers [94]. Conjoint planning and knowledge of compartmental anatomy cannot be over-emphasized, as adverse outcomes including unnecessary limb amputation have been documented following inappropriate percutaneous biopsy routes [97, 99-101]. Tumor cell seeding occurs rarely when compared to open biopsy, reinforcing the need to cooperatively plan the biopsy approach so there is no compromise of subsequent surgical procedure and needle track resection [94, 98, 102]. Indications include identification of primary or secondary tumors, metastasis or infection [94, 97]. There are few contraindications; important exceptions to this include the lack of an appropriate imaging approach (“target window”), overlying infection, and altered coagulation profile. Equally important is careful synthesis of available clinical and imaging information to avoid biopsy of classic benign (“do not touch”) lesions [94]. Accuracy rates of both fine needle aspiration (FNA) and core biopsy in osteolytic, osteosclerotic and soft tissue lesions has been extensively reviewed, with accuracy/clinically useful rates using CT guidance of 66 percent to 97 percent, where accuracy of tissue sample was compared to that of final pathologic diagnosis or clinical follow-up, and “clinically useful” was defined as allowing initiation of correct treatment based on biopsy result [95, 103-105]. Diagnostic yield, providing a specimen that can be accurately interpreted by the examining pathologist, varies, but averages around 80 percent to 90 percent [96, 98]. Deep-seated or difficult to access lesions also appear amenable to CT-guided biopsy [96]. It is unclear if the site of the lesion affects accuracy [103, 106]. Immediate analysis of
416
J. Pahade et al.
FNA specimens by a cytopathology technologist allows immediate assessment of samples for tissue adequacy, thereby allowing increased diagnostic yield by prompting repeat sampling when necessary [97]. Some institutions have also demonstrated successful and accurate use of ultrasound in biopsy of select lesions. The key component in use of ultrasound in primary bone lesions is the corresponding presence of an extraosseous mass [107]. Completing both FNA and core biopsies helps to increase accuracy/clinically useful samples and reduces non-diagnostic samples. Limitations appear to exist most notably in a subset of patients with infectious etiology or tissue containing myxoid components [95, 103, 106]. It is not clear whether biopsy of bone lesions is more accurate than that of soft tissue lesions, with some recent reviews indicating better accuracy with bone lesions, and others displaying more accuracy with soft tissue lesions [95, 98, 103, 106]. Clinical suspicion still plays a large role in whether a patient is sent for an open biopsy; patients with accurate percutaneous biopsy results may still be sent for open biopsy because of a questionable diagnosis or low confidence in the needle biopsy result [95, 103, 108]. Examination of all imaging studies obtained in the patient is recommended, as MRI or ultrasound characteristics can help target areas that are most suspicious for malignancy and guide avoidance of necrotic or cystic areas that are likely to be low yield [100]. Decreased accuracy of needle biopsy, compared to open biopsy, is largely a result of the small sample size afforded by needle biopsy. This is of particular concern in heterogeneous lesions and well-differentiated lipomatous, chondroid, and cyst-like tumors that are difficult to distinguish from benign entities [94, 98]. In attempting to predict which lesions are best suited for a particular biopsy approach, some investigators have found that open biopsy is as unsuccessful in a similar group of tumors as needle biopsy, especially lesions that demonstrate prominent blood or fluid levels [97]. MRI-safe equipment and “interventional magnets” are not widespread, but where available may help increase diagnostic yield by allowing MRI- guided biopsy of suspicious regions [94, 107].
9.2
Image-Guided Therapy
Percutaneous therapy is not limited to diagnosis, and treatment involving this approach has also increased. The target of percutaneous therapies is most commonly metastasis to bone, which has been shown to cause significant pain, decrease quality of life, lead to pathologic fractures, and induce depression and anxiety [109]. In the past, treatment for these lesions has been based on analgesics/opiates, local therapy (external beam radiation therapy or surgery), and systemic therapy (chemotherapy, hormonal, bisphosphonates) which usually proved effective [109]. However, 20 to 30 percent of patients treated with radiation do not experience pain relief. A recent multicenter trial revealed effectiveness in radiofrequency ablation in treatment of painful lytic or mixed lytic/blastic metastatic bone lesions from a multitude of cancers including renal, colorectal, lung, and thyroid carcinomas [109].
15 Imaging of Malignant Skeletal Tumors
417
Radiofrequency ablation (RFA) involves transmitting a high-frequency alternating current through a needle to cause frictional heating and necrosis of tumor. Disadvantages of RFA include increased pain during and immediately following treatment and a period of weeks before substantial pain reduction is achieved [109, 110]. Cryoablation involves delivery of argon gas through an insulated probe, with rapid expansion of the gas resulting in cooling that reaches −100 ° C within a few seconds. Subsequent thawing is achieved by instilling helium gas, followied by another cycle of freezing. In a small single-center trial, Callstrom et al. showed results similar to RFA, with a significant reduction in pain levels in 14 patients [111]. Larger prospective multicenter trials need to be performed before either RFA or cryoablation become standard of care for therapeutically challenging patients. Percutaneous vertebroplasty is widely used and effective in treatment of painful pathologic vertebral body fractures [64]. With increased use and performance, it is expected that accuracy and yield of percutaneous-guided biopsy with fluoroscopy, CT, MRI or ultrasound will continue to improve. Development of external CT guidance localization devices may also make the procedure more accurate and easier to perform, while at the same time reducing procedural radiation doses [112].
Key Points ●
●
●
In summary, the available imaging modalities have benefits and limitations and therefore physicians must carefully select the correct test or set of tests to efficiently and accurately evaluate a patient that presents with a new skeletal lesion. Close interactions with the radiologist will ensure that patients are appropriately triaged. Given rapid changes in imaging technology and ongoing, active research in oncologic imaging, the future of skeletal tumor imaging promises many new innovations that will further expand the role of imaging in the diagnosis, staging, and treatment of patients in the future.
References 1. Lodwick G S, Wilson A J, Farrell C, Virtama P, and Dittrich F. Determining growth rates of focal lesions of bone from radiographs. Radiology, 134: 577-583, 1980. 2. Lodwick G S, Wilson A J, Farrell C, Virtama P, Smeltzer F M, and Dittrich F. Estimating rate of growth in bone lesions: observer performance and error. Radiology, 134: 585-590, 1980. 3. Arndt C A and Crist W M. Common musculoskeletal tumors of childhood and adolescence. N Engl J Med, 341: 342-352, 1999. 4. Murphey M D, Robbin M R, McRae G A, Flemming D J, Temple H T, and Kransdorf M J. The many faces of osteosarcoma. Radiographics, 17: 1205-1231, 1997. 5. Miller S L and Hoffer F A. Malignant and benign bone tumors. Radiol Clin North Am, 39: 673-699, 2001.
418
J. Pahade et al.
6. Sajadi K R, Heck R K, Neel M D, et al. The incidence and prognosis of osteosarcoma skip metastases. Clin Orthop Relat Res: 92-96, 2004. 7. Brenner W, Bohuslavizki K H, and Eary J F. PET imaging of osteosarcoma. J Nucl Med, 44: 930-942, 2003. 8. Huvos A G, Rosen G, Bretsky S S, and Butler A. Telangiectatic osteogenic sarcoma: a clinicopathologic study of 124 patients. Cancer, 49: 1679-1689, 1982. 9. Murphey M D, wan Jaovisidha S, Temple H T, Gannon F H, Jelinek J S, and Malawer M M. Telangiectatic osteosarcoma: radiologic-pathologic comparison. Radiology, 229: 545-553, 2003. 10. Klein M J and Siegal G P. Osteosarcoma: anatomic and histologic variants. Am J Clin Pathol, 125: 555-581, 2006. 11. Nakajima H, Sim F H, Bond J R, and Unni K K. Small cell osteosarcoma of bone. Review of 72 cases. Cancer, 79: 2095-2106, 1997. 12. Jaffe H L. Intracortical osteogenic sarcoma. Bull Hosp Joint Dis, 21: 189-197, 1960. 13. Smith J, Botet J F, and Yeh S D. Bone sarcomas in Paget disease: a study of 85 patients. Radiology, 152: 583-590, 1984. 14. McCarville M B, Christie R, Daw N C, Spunt S L, and Kaste S C. PET/CT in the evaluation of childhood sarcomas. AJR Am J Roentgenol, 184: 1293-1304, 2005. 15. Rodriguez-Galindo C, Shah N, McCarville M B, et al. Outcome after local recurrence of osteosarcoma: the St. Jude Children’s Research Hospital experience (1970-2000). Cancer, 100: 1928-1935, 2004. 16. Wittig J C, Bickels J, Priebat D, et al. Osteosarcoma: a multidisciplinary approach to diagnosis and treatment. Am Fam Physician, 65: 1123-1132, 2002. 17. Imbriaco M, Yeh S D, Yeung H, et al. Thallium-201 scintigraphy for the evaluation of tumor response to preoperative chemotherapy in patients with osteosarcoma. Cancer, 80: 1507-1512, 1997. 18. Menendez L R, Fideler B M, and Mirra J. Thallium-201 scanning for the evaluation of osteosarcoma and soft tissue sarcoma. A study of the evaluation and predictability of the histological response to chemotherapy. J Bone Joint Surg Am, 75: 526-531, 1993. 19. Bredella M A, Caputo G R, and Steinbach L S. Value of FDG positron emission tomography in conjunction with MR imaging for evaluating therapy response in patients with musculoskeletal sarcomas. AJR Am J Roentgenol, 179: 1145-1150, 2002. 20. Hawkins D S, Rajendran J G, Conrad E U, 3rd, Bruckner J D, and Eary J F. Evaluation of chemotherapy response in pediatric bone sarcomas by [F-18]-fluorodeoxy-D-glucose positron emission tomography. Cancer, 94: 3277-3284, 2002. 21. Murphey M D, Walker E A, Wilson A J, Kransdorf M J, Temple H T, and Gannon F H. From the archives of the AFIP: imaging of primary chondrosarcoma: radiologic-pathologic correlation. Radiographics, 23: 1245-1278, 2003. 22. Feldman F, Van Heertum R, Saxena C, and Parisien M. 18FDG-PET applications for cartilage neoplasms. Skeletal Radiol, 34: 367-374, 2005. 23. Murphey M D, Flemming D J, Boyea S R, Bojescul J A, Sweet D E, and Temple H T. Enchondroma versus chondrosarcoma in the appendicular skeleton: differentiating features. Radiographics, 18: 1213-1237; quiz 1244-1215, 1998. 24. Lee F Y, Yu J, Chang S S, Fawwaz R, and Parisien M V. Diagnostic value and limitations of fluorine-18 fluorodeoxyglucose positron emission tomography for cartilaginous tumors of bone. J Bone Joint Surg Am, 86-A: 2677-2685, 2004. 25. Evans H L, Ayala A G, and Romsdahl M M. Prognostic factors in chondrosarcoma of bone: a clinicopathologic analysis with emphasis on histologic grading. Cancer, 40: 818-831, 1977. 26. Arsos G, Venizelos I, Karatzas N, Koukoulidis A, and Karakatsanis C. Low-grade chondrosarcomas: a difficult target for radionuclide imaging. Case report and review of the literature. Eur J Radiol, 43: 66-72, 2002. 27. Tallini G, Dorfman H, Brys P, et al. Correlation between clinicopathological features and karyotype in 100 cartilaginous and chordoid tumors. A report from the Chromosomes and Morphology (CHAMP) Collaborative Study Group. J Pathol, 196: 194-203, 2002.
15 Imaging of Malignant Skeletal Tumors
419
28. Janzen L, Logan P M, O’Connell J X, Connell D G, and Munk P L. Intramedullary chondroid tumors of bone: correlation of abnormal peritumoral marrow and soft tissue MRI signal with tumor type. Skeletal Radiol, 26: 100-106, 1997. 29. Geirnaerdt M J, Bloem J L, Eulderink F, Hogendoorn P C, and Taminiau A H. Cartilaginous tumors: correlation of gadolinium-enhanced MR imaging and histopathologic findings. Radiology, 186: 813-817, 1993. 30. Aoki J, Sone S, Fujioka F, et al. MRI of enchondroma and chondrosarcoma: rings and arcs of Gd-DTPA enhancement. J Comput Assist Tomogr, 15: 1011-1016, 1991. 31. Geirnaerdt M J, Hogendoorn P C, Bloem J L, Taminiau A H, and van der Woude H J. Cartilaginous tumors: fast contrast-enhanced MR imaging. Radiology, 214: 539-546, 2000. 32. Brenner W, Conrad E U, and Eary J F. FDG PET imaging for grading and prediction of outcome in chondrosarcoma patients. Eur J Nucl Med Mol Imaging, 31: 189-195, 2004. 33. Collins M S, Koyama T, Swee R G, and Inwards C Y. Clear cell chondrosarcoma: radiographic, computed tomographic, and magnetic resonance findings in 34 patients with pathologic correlation. Skeletal Radiol, 32: 687-694, 2003. 34. Kaim A H, Hugli R, Bonel H M, and Jundt G. Chondroblastoma and clear cell chondrosarcoma: radiological and MRI characteristics with histopathological correlation. Skeletal Radiol, 31: 88-95, 2002. 35. Davila J A, Amrami K K, Sundaram M, Adkins M C, and Unni K K. Chondroblastoma of the hands and feet. Skeletal Radiol, 33: 582-587, 2004. 36. Aoki J, Tanikawa H, Ishii K, et al. MRI findings indicative of hemosiderin in giant-cell tumor of bone: frequency, cause, and diagnostic significance. AJR Am J Roentgenol, 166: 145-148, 1996. 37. Kumta S M, Griffith J F, Chow L T, and Leung P C. Primary juxtacortical chondrosarcoma dedifferentiating after 20 years. Skeletal Radiol, 27: 569-573, 1998. 38. Schajowicz F. Juxtacortical chondrosarcoma. J Bone Joint Surg Br, 59-B: 473-480, 1977. 39. Robinson P, White L M, Sundaram M, et al. Periosteal chondroid tumors: radiologic evaluation with pathologic correlation. AJR Am J Roentgenol, 177: 1183-1188, 2001. 40. Seeger L L, Yao L, and Eckardt J J. Surface lesions of bone. Radiology, 206: 17-33, 1998. 41. Antonescu C R, Argani P, Erlandson R A, Healey J H, Ladanyi M, and Huvos A G. Skeletal and extraskeletal myxoid chondrosarcoma: a comparative clinicopathologic, ultrastructural, and molecular study. Cancer, 83: 1504-1521, 1998. 42. Amukotuwa S A, Choong P F, Smith P J, Powell G J, Thomas D, and Schlicht S M. Femoral mesenchymal chondrosarcoma with secondary aneurysmal bone cysts mimicking a small-cell osteosarcoma. Skeletal Radiol, 35: 311-318, 2006. 43. Nussbeck W, Neureiter D, Soder S, Inwards C, and Aigner T. Mesenchymal chondrosarcoma: an immunohistochemical study of 10 cases examining prognostic significance of proliferative activity and cellular differentiation. Pathology, 36: 230-233, 2004. 44. Chidambaram A and Sanville P. Mesenchymal chondrosarcoma of the maxilla. J Laryngol Otol, 114: 536-539, 2000. 45. Nguyen B D, Daffner R H, Dash N, Rothfus W E, Nathan G, and Toca A R, Jr. Case report 790. Mesenchymal chondrosarcoma of the sacrum. Skeletal Radiol, 22: 362-366, 1993. 46. Frassica F J, Unni K K, Beabout J W, and Sim F H. Dedifferentiated chondrosarcoma. A report of the clinicopathological features and treatment of seventy-eight cases. J Bone Joint Surg Am, 68: 1197-1205, 1986. 47. Staals E L, Bacchini P, and Bertoni F. Dedifferentiated central chondrosarcoma. Cancer, 106: 2682-2691, 2006. 48. Bruns J, Fiedler W, Werner M, and Delling G. Dedifferentiated chondrosarcoma–a fatal disease. J Cancer Res Clin Oncol, 131: 333-339, 2005. 49. Littrell L A, Wenger D E, Wold L E, et al. Radiographic, CT, and MR imaging features of dedifferentiated chondrosarcomas: a retrospective review of 174 de novo cases. Radiographics, 24: 1397-1409, 2004. 50. MacSweeney F, Darby A, and Saifuddin A. Dedifferentiated chondrosarcoma of the appendicular skeleton: MRI-pathological correlation. Skeletal Radiol, 32: 671-678, 2003.
420
J. Pahade et al.
51. Okada K, Hasegawa T, Tateishi U, Endo M, and Itoi E. Dedifferentiated chondrosarcoma with telangiectatic osteosarcoma-like features. J Clin Pathol, 59: 1200-1202, 2006. 52. Saifuddin A, Mann B S, Mahroof S, Pringle J A, Briggs T W, and Cannon S R. Dedifferentiated chondrosarcoma: use of MRI to guide needle biopsy. Clin Radiol, 59: 268-272, 2004. 53. Mulligan M E. Imaging techniques used in the diagnosis, staging, and follow-up of patients with myeloma. Acta Radiol, 46: 716-724, 2005. 54. Angtuaco E J, Fassas A B, Walker R, Sethi R, and Barlogie B. Multiple myeloma: clinical review and diagnostic imaging. Radiology, 231: 11-23, 2004. 55. Durie B G, Kyle R A, Belch A, et al. Myeloma management guidelines: a consensus report from the Scientific Advisors of the International Myeloma Foundation. Hematol J, 4: 379398, 2003. 56. Vande Berg B C, Michaux L, Lecouvet F E, et al. Nonmyelomatous monoclonal gammopathy: correlation of bone marrow MR images with laboratory findings and spontaneous clinical outcome. Radiology, 202: 247-251, 1997. 57. Baur A, Stabler A, Nagel D, et al. Magnetic resonance imaging as a supplement for the clinical staging system of Durie and Salmon? Cancer, 95: 1334-1345, 2002. 58. Mulligan M E and Badros A Z. PET/CT and MR imaging in myeloma. Skeletal Radiol, 36: 5-16, 2007. 59. Johnston C, Brennan S, Ford S, and Eustace S. Whole body MR imaging: applications in oncology. Eur J Surg Oncol, 32: 239-246, 2006. 60. Lecouvet F E, Dechambre S, Malghem J, Ferrant A, Vande Berg B C, and Maldague B. Bone marrow transplantation in patients with multiple myeloma: prognostic significance of MR imaging. AJR Am J Roentgenol, 176: 91-96, 2001. 61. Ghanem N, Lohrmann C, Engelhardt M, et al. Whole-body MRI in the detection of bone marrow infiltration in patients with plasma cell neoplasms in comparison to the radiological skeletal survey. Eur Radiol, 16: 1005-1014, 2006. 62. Hartman R P, Sundaram M, Okuno S H, and Sim F H. Effect of granulocyte-stimulating factors on marrow of adult patients with musculoskeletal malignancies: incidence and MRI findings. AJR Am J Roentgenol, 183: 645-653, 2004. 63. Lecouvet F E, Vande Berg B C, Michaux L, et al. Stage III multiple myeloma: clinical and prognostic value of spinal bone marrow MR imaging. Radiology, 209: 653-660, 1998. 64. Layton K F, Thielen K R, Cloft H J, and Kallmes D F. Acute vertebral compression fractures in patients with multiple myeloma: evaluation of vertebral body edema patterns on MR imaging and the implications for vertebroplasty. AJNR Am J Neuroradiol, 27: 1732-1734, 2006. 65. Erly W K, Oh E S, and Outwater E K. The utility of in-phase/opposed-phase imaging in differentiating malignancy from acute benign compression fractures of the spine. AJNR Am J Neuroradiol, 27: 1183-1188, 2006. 66. Horger M, Claussen C D, Bross-Bach U, et al. Whole-body low-dose multidetector row-CT in the diagnosis of multiple myeloma: an alternative to conventional radiography. Eur J Radiol, 54: 289-297, 2005. 67. Nandurkar D, Kalff V, Turlakow A, Spencer A, Bailey M J, and Kelly M J. Focal MIBI uptake is a better indicator of active myeloma than diffuse uptake. Eur J Haematol, 76: 141-146, 2006. 68. Breyer R J, 3rd, Mulligan M E, Smith S E, Line B R, and Badros A Z. Comparison of imaging with FDG PET/CT with other imaging modalities in myeloma. Skeletal Radiol, 35: 632-640, 2006. 69. Nanni C, Zamagni E, Farsad M, et al. Role of 18F-FDG PET/CT in the assessment of bone involvement in newly diagnosed multiple myeloma: preliminary results. Eur J Nucl Med Mol Imaging, 33: 525-531, 2006. 70. Bredella M A, Steinbach L, Caputo G, Segall G, and Hawkins R. Value of FDG PET in the assessment of patients with multiple myeloma. AJR Am J Roentgenol, 184: 1199-1204, 2005. 71. Moulopoulos L A, Gika D, Anagnostopoulos A, et al. Prognostic significance of magnetic resonance imaging of bone marrow in previously untreated patients with multiple myeloma. Ann Oncol, 16: 1824-1828, 2005.
15 Imaging of Malignant Skeletal Tumors
421
72. Ghanem N, Uhl M, Brink I, et al. Diagnostic value of MRI in comparison to scintigraphy, PET, MS-CT and PET/CT for the detection of metastases of bone. Eur J Radiol, 55: 41-55, 2005. 73. Roodman G D. Mechanisms of bone metastasis. N Engl J Med, 350: 1655-1664, 2004. 74. Hamaoka T, Madewell J E, Podoloff D A, Hortobagyi G N, and Ueno N T. Bone imaging in metastatic breast cancer. J Clin Oncol, 22: 2942-2953, 2004. 75. Schweitzer M E, Levine C, Mitchell D G, Gannon F H, and Gomella L G. Bull’s-eyes and halos: useful MRI discriminators of osseous metastases. Radiology, 188: 249-252, 1993. 76. Spuentrup E, Buecker A, Adam G, van Vaals J J, and Guenther R W. Diffusion-weighted MR imaging for differentiation of benign fracture edema and tumor infiltration of the vertebral body. AJR Am J Roentgenol, 176: 351-358, 2001. 77. Lauenstein T C, Goehde S C, Herborn C U, et al. Whole-body MR imaging: evaluation of patients for metastases. Radiology, 233: 139-148, 2004. 78. Schmidt G P, Haug A R, Schoenberg S O, and Reiser M F. Whole-body MRI and PET-CT in the management of cancer patients. Eur Radiol, 16: 1216-1225, 2006. 79. Fogelman I, Cook G, Israel O, and Van der Wall H. Positron emission tomography and bone metastases. Semin Nucl Med, 35: 135-142, 2005. 80. Nakamoto Y, Cohade C, Tatsumi M, Hammoud D, and Wahl R L. CT appearance of bone metastases detected with FDG PET as part of the same PET/CT examination. Radiology, 237: 627-634, 2005. 81. Rougraff B T, Kneisl J S, and Simon M A. Skeletal metastases of unknown origin. A prospective study of a diagnostic strategy. J Bone Joint Surg Am, 75: 1276-1281, 1993. 82. Mulligan M E, McRae G A, and Murphey M D. Imaging features of primary lymphoma of bone. AJR Am J Roentgenol, 173: 1691-1697, 1999. 83. Krishnan A, Shirkhoda A, Tehranzadeh J, Armin A R, Irwin R, and Les K. Primary bone lymphoma: radiographic-MR imaging correlation. Radiographics, 23: 1371-1383; discussion 1384-1377, 2003. 84. Mengiardi B, Honegger H, Hodler J, Exner U G, Csherhati M D, and Bruhlmann W. Primary lymphoma of bone: MRI and CT characteristics during and after successful treatment. AJR Am J Roentgenol, 184: 185-192, 2005. 85. Bernstein M, Kovar H, Paulussen M, et al. Ewing’s sarcoma family of tumors: current management. Oncologist, 11: 503-519, 2006. 86. Hatori M, Okada K, Nishida J, and Kokubun S. Periosteal Ewing’s sarcoma: radiological imaging and histological features. Arch Orthop Trauma Surg, 121: 594-597, 2001. 87. Ilaslan H, Sundaram M, Unni K K, and Dekutoski M B. Primary Ewing’s sarcoma of the vertebral column. Skeletal Radiol, 33: 506-513, 2004. 88. Li W Y, Brock P, and Saunders D E. Imaging characteristics of primary cranial Ewing sarcoma. Pediatr Radiol, 35: 612-618, 2005. 89. Brisse H, Ollivier L, Edeline V, et al. Imaging of malignant tumours of the long bones in children: monitoring response to neoadjuvant chemotherapy and preoperative assessment. Pediatr Radiol, 34: 595-605, 2004. 90. Furth C, Amthauer H, Denecke T, Ruf J, Henze G, and Gutberlet M. Impact of whole-body MRI and FDG-PET on staging and assessment of therapy response in a patient with Ewing sarcoma. Pediatr Blood Cancer, 47: 607-611, 2006. 91. Daldrup-Link H E, Franzius C, Link T M, et al. Whole-body MR imaging for detection of bone metastases in children and young adults: comparison with skeletal scintigraphy and FDG PET. AJR Am J Roentgenol, 177: 229-236, 2001. 92. Hawkins D S, Schuetze S M, Butrynski J E, et al. [18F]Fluorodeoxyglucose positron emission tomography predicts outcome for Ewing sarcoma family of tumors. J Clin Oncol, 23: 88288834, 2005. 93. Dyke J P, Panicek D M, Healey J H, et al. Osteogenic and Ewing sarcomas: estimation of necrotic fraction during induction chemotherapy with dynamic contrast-enhanced MR imaging. Radiology, 228: 271-278, 2003. 94. Choi J J, Davis K W, and Blankenbaker D G. Percutaneous musculoskeletal biopsy. Semin Roentgenol, 39: 114-128, 2004.
422
J. Pahade et al.
95. Ogilvie C M, Torbert J T, Finstein J L, Fox E J, and Lackman R D. Clinical utility of percutaneous biopsies of musculoskeletal tumors. Clin Orthop Relat Res, 450: 95-100, 2006. 96. Puri A, Shingade V U, Agarwal M G, et al. CT-guided percutaneous core needle biopsy in deep seated musculoskeletal lesions: a prospective study of 128 cases. Skeletal Radiol, 35: 138-143, 2006. 97. Jelinek J S, Murphey M D, Welker J A, et al. Diagnosis of primary bone tumors with imageguided percutaneous biopsy: experience with 110 tumors. Radiology, 223: 731-737, 2002. 98. Mitsuyoshi G, Naito N, Kawai A, et al. Accurate diagnosis of musculoskeletal lesions by core needle biopsy. J Surg Oncol, 94: 21-27, 2006. 99. Anderson M W, Temple H T, Dussault R G, and Kaplan P A. Compartmental anatomy: relevance to staging and biopsy of musculoskeletal tumors. AJR Am J Roentgenol, 173: 1663-1671, 1999. 100. Liu P T, Valadez S D, Chivers F S, Roberts C C, and Beauchamp C P. Anatomically based guidelines for core needle biopsy of bone tumors: implications for limb-sparing surgery. Radiographics, 27: 189-205; discussion 206, 2007. 101. Mankin H J, Mankin C J, and Simon M A. The hazards of the biopsy, revisited. Members of the Musculoskeletal Tumor Society. J Bone Joint Surg Am, 78: 656-663, 1996. 102. Davies N M, Livesley P J, and Cannon S R. Recurrence of an osteosarcoma in a needle biopsy track. J Bone Joint Surg Br, 75: 977-978, 1993. 103. Hau A, Kim I, Kattapuram S, et al. Accuracy of CT-guided biopsies in 359 patients with musculoskeletal lesions. Skeletal Radiol, 31: 349-353, 2002. 104. Leffler S G and Chew F S. CT-guided percutaneous biopsy of sclerotic bone lesions: diagnostic yield and accuracy. AJR Am J Roentgenol, 172: 1389-1392, 1999. 105. Stoker D J, Cobb J P, and Pringle J A. Needle biopsy of musculoskeletal lesions. A review of 208 procedures. J Bone Joint Surg Br, 73: 498-500, 1991. 106. Tsukushi S, Katagiri H, Nakashima H, Shido Y, and Arai E. Application and utility of computed tomography-guided needle biopsy with musculoskeletal lesions. J Orthop Sci, 9: 122125, 2004. 107. Saifuddin A, Mitchell R, Burnett S J, Sandison A, and Pringle J A. Ultrasound-guided needle biopsy of primary bone tumours. J Bone Joint Surg Br, 82: 50-54, 2000. 108. Yao L, Nelson S D, Seeger L L, Eckardt J J, and Eilber F R. Primary musculoskeletal neoplasms: effectiveness of core-needle biopsy. Radiology, 212: 682-686, 1999. 109. Goetz M P, Callstrom M R, Charboneau J W, et al. Percutaneous image-guided radiofrequency ablation of painful metastases involving bone: a multicenter study. J Clin Oncol, 22: 300-306, 2004. 110. Callstrom M R, Charboneau J W, Goetz M P, et al. Painful metastases involving bone: feasibility of percutaneous CT- and US-guided radio-frequency ablation. Radiology, 224: 87-97, 2002. 111. Callstrom M R, Atwell T D, Charboneau J W, et al. Painful metastases involving bone: percutaneous image-guided cryoablation–prospective trial interim analysis. Radiology, 241: 572-580, 2006. 112. Roberts C C, Morrison W B, Deely D M, Zoga A C, Koulouris G, and Winalski C S. Use of a novel percutaneous biopsy localization device: initial musculoskeletal experience. Skeletal Radiol, 36: 53-57, 2007.
16
Radiology of Soft Tissue Tumors Including Melanoma M.J. Shelly1, P.J. MacMahon1, and S. Eustace1, 2
Introduction Soft tissue tumors are defined as mesenchymal proliferations that occur in the extraskeletal, non-epithelial tissues of the body, excluding the viscera, coverings of the brain and lymphoreticular system [1]. The true frequency of soft tissue tumors is difficult to estimate because most benign lesions are not removed. A conservative estimate is that benign tumors outnumber their malignant counterparts by a ratio of at least 100:1. In the United States only 7,200 sarcomas are diagnosed annually (0.8 percent of invasive malignancies), yet they are responsible for 2 percent of all cancer deaths, reflecting their lethal nature [1]. Classification is based on the tissue from which the lesions arise (Table 16.1). The cause of most soft tissue tumors is unknown. There are documented associations between radiation therapy [2], and rare instances in which chemical burns, heat burns or trauma were associated with subsequent development of sarcoma [1]. Soft tissue tumors may arise in any location, with approximately 50 percent in the extremities (two-thirds of these in the lower extremities), 14 percent in the retroperitoneum, 15 percent in the viscera, 10 percent in the trunk and 11 percent in other sites [3]. Regarding sarcomas, males are affected more frequently than females (ratio 1.4:1), and the incidence generally increases with age. Fifteen percent arise in children and constitute the fourth most common malignancy in this age group [1]. Specific sarcomas tend to appear in certain age groups (e.g., rhabdomyosarcoma in children, synovial sarcoma in young adults and liposarcoma and malignant fibrous histiocytoma in mid- to late adult life) [1]. Some features of soft tissue tumors influence the prognosis. Accurate histologic classification significantly contributes to establishing the prognosis of a sarcoma. 1 Department of Radiology, Mater Misericordiae University Hospital, Eccles Street, Dublin 7, Republic of Ireland. 2 Department of Radiology, Cappagh National Orthopaedic Hospital, Finglas, Dublin 11, Republic of Ireland
Corresponding author: Martin J. Shelly Department of Radiology, Mater Misericordiae University Hospital, Eccles Street, Dublin 7, Republic of Ireland, e-mail:
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
423
424
M.J. Shelly et al.
Table 16.1 Classification of Soft Tissue Tumors
Soft Tissue Tumors n
n
n
n
n
n
n
n
Tumors of adipose tissue Lipoma Liposarcoma Tumors of fibrous tissue Fibromatosis (superficial and deep) Fibrosarcoma Fibrohistiocytic tumors Benign fibrous histiocytoma Malignant fibrous histiocytoma Tumors of skeletal muscle Rhabdomyoma Rhabdomyosarcoma Tumors of smooth muscle Leiomyoma Leiomyosarcoma Vascular tumors Hemangioma Angiosarcoma Peripheral nerve tumors Neurofibroma Schwannoma Malignant peripheral nerve sheath tumors Synovial sarcoma
Important diagnostic features are cell morphology and architectural arrangement. Whatever the type, the grade of a soft tissue sarcoma is of great importance. Mitotic activity and the extent of tumor necrosis (a reflection of growth rate) are thought to be particularly significant. The size, depth and stage of the tumor also provide important diagnostic information [4]. In general tumors arising in superficial locations (e.g., skin and subcutaneous tissues) have a better prognosis than deeper lesions. Overall, the 10-year survival rate for sarcomas is approximately 40 percent [1].
1
Advances in Soft Tissue Tumor Imaging
The role of radiological imaging in the diagnosis, staging and eventual follow-up of patients who have soft tissue tumors has greatly expanded in recent years. In particular, magnetic resonance imaging (MRI) provides excellent soft tissue contrast for anatomic imaging, and positron emission tomography (PET) demonstrates metabolic activity and functional imaging. This chapter presents an overview of the role of imaging, recent advances in technology and state-of-the-art techniques for evaluating soft tissue neoplasia.
16 Radiology of Soft Tissue Tumors Including Melanoma
1.1
425
Conventional Imaging Modalities
Radiographs have a limited use for soft tissue lesions [5], with a few noteworthy exceptions. Soft tissue tumors juxtaposed to bone may cause focal cortical erosion (e.g. fibrosarcoma) or reactive periosteal changes (e.g. hemangioma) [6]. Normal distinct fatty planes are typically obscured as neoplastic or inflammatory lesions displace them, while lipomas may be identified by their radiolucent appearance compared to surrounding tissues [6]. Calcification in the soft tissue may suggest liposarcoma [6], while phleboliths may be visible in vascular lesions [7]. Ackerman’s [8] zone phenomenon should be considered when evaluating soft tissue masses that mineralize. This implies more central, “fluffy” mineralization in soft tissue neoplasia, compared with more uniform mineralization from the periphery of the lesion in nonneoplastic processes such as myositis ossificans. Mineralization patterns aside, plain films are generally nonspecific for the evaluation of soft tissue tumors. Angiography was formally used to assess the vascularity of neoplasia, but its findings are nonspecific and cannot reliably distinguish benign from malignant lesions [9, 10]. Today, diagnostic angiography has been replaced by MRI and magnetic resonance angiography (MRA) [11], but some vascular tumors may need presurgical embolization before a definitive surgical procedure [7, 11]. Bone scintigraphy has been available for decades and the use of technetium-labeled nucleotides and more modern gamma cameras has led to a significant decrease in patient dose and an increase in diagnostic information, making it a very useful screening tool [6]. Bone scans are a sensitive modality for assessing abnormalities in bone formation and perfusion, and have been a reliable tool for detecting multifocal osseous lesions. They remain the mainstay for evaluating bone metastases due to the fact that, while plain films require approximately 50 percent loss of mineralization to allow detection of destructive lesions of bone, scintigraphy is an excellent screening modality to detect lesions not otherwise seen on routine radiographs. Unlike PET, however, which demonstrates metabolic activity, bone scan detects areas of bony repair and, therefore, could be negative in purely lytic lesions such as myeloma and, in some cases, of Ewing’s sarcoma [7]. In soft tissue lesions bone scintigraphy is relatively nonspecific for diagnostic purposes. Soft tissue lesions show variable radiopharmaceutical uptake, but in general, most malignant lesions exhibit increased uptake, while benign lesions tend to exhibit little or no intrinsic uptake [12].
1.2
Ultrasound
Ultrasound, with its lack of ionizing radiation and dynamic real-time imaging capabilities, is a readily available, noninvasive and relatively inexpensive method of detecting and determining the size and constituency of a soft tissue mass. For musculoskeletal oncology, the use of ultrasound is somewhat limited to differentiating cystic from solid masses, and evaluating postoperative fluid collections [7, 13]. Doppler ultrasound could be helpful in assessing the vascularity of soft tissue
426
M.J. Shelly et al.
masses, which is important for diagnosis and pre-surgical planning [14]. Ultrasound can be used for image guidance in biopsy procedures [11], but this is usually accomplished using computed tomography (CT), especially in complex anatomic sites [7]. One potential drawback of ultrasound is its operator dependence – only in experienced hands is it a useful adjunct imaging modality in the setting of soft tissue oncology.
1.3
Computed Tomography
Since the introduction of clinical CT scanning in the mid-1970s, numerous advances in technology have led to higher demand for CT-based diagnostic examinations. CT provides excellent discrimination of fat from other tissues and, when combined with intravenous contrast, provides a more straightforward evaluation of the vascularity of structures than MRI [15]. Lesions that contain calcification are easily detected on CT, while the low density of fat makes lipomas readily identifiable with discrete areas of soft tissue density within fatty lesions suggestive of liposarcoma [6]. The introduction of spiral or helical CT, and then multidetectorrow CT, has resulted in significant increases in scanning speed, diminishing the problems related to patient motion during examination. Picture archiving and communication systems (PACS) allow the interpreting radiologist to review the CT images on a monitor using different window-level adjustments to visualize soft tissue or bony detail, and to enable instant three-dimensional reconstructions in different planes [6, 7]. This capability significantly enhances diagnostic detail, compared with hard copy images. Although CT remains an important imaging modality, the superior contrast resolution of MRI has replaced the need for CT scans in many cases of soft tissue lesions [11]. There are exceptions, however, which are better evaluated using CT, such as evaluating cortical integrity, detecting subtle matrix mineral and establishing the presence or absence of a thin rim of cortex around expansive lesions or masses containing calcific densities – all of which have a bearing on the differential diagnosis [7]. Abdominal and chest CT remain particularly important tools in the staging process. CT is critical for evaluating pulmonary nodules. Chest CT should be obtained in all cases of known malignant soft tissue neoplasia to evaluate for the presence of pulmonary metastases and lymphadenopathy. Despite recent advances in MRI, CT remains the modality of choice to evaluate the abdomen or pelvis for masses, lymphadenopathy or other signs of metastatic disease [7].
1.4
Magnetic Resonance Imaging
The clinical use of MRI over the past two decades has had a profound effect on the initial imaging, staging and subsequent post-treatment follow-up of soft tissue tumors. MRI is currently the imaging modality of choice for evaluating soft tissue lesions
16 Radiology of Soft Tissue Tumors Including Melanoma
427
[5, 11]. The lack of ionizing radiation, the ability to image in multiple planes without loss of image resolution and the tissue characteristics provided by different pulse sequences are important advantages of MRI [7, 15]. Due to the versatility of MRI, choices must be made among the many studies and possible parameters to keep the length of the examination tolerable for the patient [15]. The contrast resolution of radiographs and CT is approximately 1 percent and 7 percent, respectively, whereas MRI has a soft tissue contrast resolution that exceeds 50 percent [16]. Additionally, this high resolution modality can be manipulated considerably by varying signal parameters to obtain a wide variety of images that better characterize tissue types based on their signal characteristics. The most commonly used pulse sequences are the spin echo sequences (T1- and T2-weighted images are obtained in this manner) [6]. This precise, high-resolution anatomic evaluation has had a significant impact on the ability to appropriately stage soft tissue tumors and adequately plan for limb salvage surgery [17]. Subcutaneous fat and lipomas demonstrate bright signal intensity on T1-weighted images. When evaluating fatty lesions, the more complex the lesion (internal stranding, heterogeneous signal, nodules, or areas of enhancement), the higher the likelihood that the lesion represents a liposarcoma rather than a simple lipoma. Intravenous gadolinium is distributed in areas of increased blood flow and vascular permeability. Although tumors may have internal areas of hemorrhage, gadolinium enhancement could be used to differentiate hematomas from hemorrhagic sarcomas, and edema from areas of hyperemia as well as suggesting overall tumor vascularity [11]. Hematomas typically demonstrate peripheral mild enhancement and have ill-defined areas of edema and fluid-like signal [6]. Tumors tend to show more complex heterogenous enhancement depending on the degree of necrosis and, in general, they have well-defined margins [7]. Assessment of tumor necrosis is objectively achieved by utilizing dynamic-enhanced MRI which allows the calculation of the tumor’s dynamic vector magnitude and kep [18]. Cystic or fluid-containing lesions are bright on T2-weighted images, but demonstrate low to intermediate signal on T1-weighted images. Cystic lesions that contain cellular debris or cystic lesions that contain two types of tissue, such as fluid and blood or fluid and fat, may create so-called “fluid-fluid” layering [6]. Despite its modest limitations (e.g., sensitivity to motion, inferiority to CT for bony detail), MRI is the investigation of choice for all indeterminate soft tissue masses as it can delineate the margins of a tumor and its relations to adjacent neurovascular and osseous structures [11]. Thus, with its excellent soft tissue contrast and superb anatomic detail, MRI is indispensable in the staging of these lesions.
1.5
Positron Emission Tomography
PET is a relatively new functional imaging technique that enables the evaluation of tissue metabolism in vivo with positron-emitting radionuclides. [18F] 2-deoxy-2fluoro-D-glucose (FDG), the most commonly used radiotracer for PET imaging, has demonstrated increased accumulation in several different types of neoplasia.
428
M.J. Shelly et al.
FDG-PET has been shown to be useful in detecting local recurrence [19] and metastatic disease in patients who have sarcoma [20] and is used to evaluate response to neoadjuvant chemotherapy [21-23]. The role of FDG-PET in the management of soft tissue sarcoma is still evolving. FDG-PET has been shown to be complementary to anatomic imaging, with high sensitivity for the detection of various malignant soft tissue lesions, for prediction of tumor grade and for directing the biopsy of large heterogenous masses [24-26]. FDG-PET has been shown to be effective in distinguishing high-grade soft tissue sarcoma from low-grade or benign lesions, and it has a complementary role, along with CT and MRI, in the evaluation of distant metastasis [27, 28]. (Fig. 16.1). Integrated PET/CT scanners became commercially available at the beginning of this decade and, while the published data are limited, recent studies have concluded that PET/CT has a modest, but nonetheless clinically relevant, impact on diagnostic performance compared with visually correlated PET and CT [29]. A semiquantitive index of glucose metabolism, the standard uptake value (SUV), may be used for lesion characterization as a marker of glucose metabolism. The SUV is calculated by placing a region of interest cursor over the lesion and dividing the value (in microcuries per cubic centimeter) by the injected dose (in microcuries), divided by the patient’s body weight (in grams) [30]. Initial publications reported a good correlation between glucose consumption measured by FDG-PET and the aggressive-
Fig. 16.1 Whole-body PET image of a patient with metastatic melanoma. Arrow indicates area of increased uptake within a splenic lesion. Note the intense uptake in the liver and multiple hilar lymph nodes
16 Radiology of Soft Tissue Tumors Including Melanoma
429
ness of musculoskeletal tumors [31], but recent reports have demonstrated a significant overlap of SUVs of benign and malignant tumors [32, 33]. In such cases correlation with anatomic imaging to identify obvious malignant characteristics of soft tissue masses is essential. FDG-PET has several significant limitations including high cost, long exam time and poor delineation of anatomy [7]. Furthermore, because FGD is an analogue of glucose, it may accumulate in normal and inflammatory tissues [34], and some malignant tumors have an intrinsically low uptake of FDG [29]. Nonetheless, FDG-PET is a promising imaging technique in musculoskeletal oncology because of its unique ability to semiquantitively measure metabolic activity of suspicious soft tissue masses, and as a method of assessing tumor response to chemo-radiotherapy [35, 36].
1.6
Image-Guided Interventions
CT guidance is commonly used for percutaneous biopsy [37, 38], however, conventional CT does not provide real-time guidance capability. CT fluoroscopy, with its real-time imaging, reduces the time required to target soft tissue masses [39]. With skin markers over the area of concern, exact localization of the area of interest, with respect to depth and proximity to vital structures, can be easily achieved. This is especially useful when core biopsy is required, as precise needle localization is essential for a successful outcome of the procedure. Currently, MRI interventions are in the initial stages of development and, in the near future, could play an increasingly important role in image-guided procedures because of the inherent benefits of MRI, namely lack of ionizing radiation, superior soft tissue contrast resolution compared to CT and the ability to demonstrate subtle bone marrow changes [7]. With this background we now turn to the pathology and radiographic appearances of individual tumors and tumor-like lesions.
2 2.1
Fatty Tumors Lipomas
They are the most common soft tissue tumors in adults. Typically asymptomatic, they are subclassified according to particular morphologic features as conventional lipoma, fibrolipoma, angiolipoma, spindle cell lipoma, myelolipoma and pleomorphic lipoma. The conventional lipoma, the most common subtype, is a well-encapsulated mass of mature adipocytes that varies considerably in size. They are most commonly found in the subcutaneous tissue, or between muscle and other connective tissue structures of the proximal extremities and trunk [15]. They arise most frequently during mid-adulthood and appear to be affected by body habitus and weight gain because they are more common in the obese and people with corticosteroid excess,
430
M.J. Shelly et al.
either endogenous or exogenous. Individual lesions may be stable or grow slowly over time. They can be differentiated from the normal body fat by the architecture of their stroma, by the mass effect they exert on adjacent structures, by their metabolic behavior and by cytogenetic features [40]. Histologically, they consist of mature fat cells and scant connective tissue [15] with no evidence of pleomorphism or abnormal growth [1]. Ultrasound is often the first modality used to evaluate soft tissue masses. The ultrasound characteristics of lipomas have been described [41, 42] and include a finely heterogenous echogenicity that is greater than that of muscle and a non-cystic, hypovascular nature. The majority of lipomas are well-defined sonographically, but a significant number have ill-defined margins that blend into the surrounding tissues [43]. Unfortunately, ultrasound alone is not specific enough to obviate the need for further imaging evaluation. On CT, lipomas are circumscribed, smoothly marginated masses with thin (< 2 mm) fibrous septa and almost uniform attenuation [15]. Lipomas are commonly encountered in routine MRI practice. MRI shows homogenous fat signal intensity throughout the lesion with high signal on T1-weighted (Fig. 16.2) and fast spin echo T2 images. A pseudocapsule is frequently present. Complete suppression of signal is expected with fat suppression sequences [7]. Fat suppression is important to evaluate the underlying stroma, which is otherwise obscured by the high signal intensity of the fat [44]. Atypical features are often seen in lipomas; these include septa thicker than 2 mm, septa with nodular components, hemorrhage,
Fig. 16.2 Coronal T1 MRI of the right upper limb demonstrating a well circumscribed high signal bi-lobed mass deep to the triceps muscle (black arrow)
16 Radiology of Soft Tissue Tumors Including Melanoma
431
calcification and areas of abnormal enhancement with intravenous contrast. These features should raise the possibility of well-differentiated liposarcoma or lipoma variants, and histologic examination of the atypical tissue is warranted [15, 45].
2.2
Liposarcoma
Liposarcoma are the second most common soft tissue sarcoma after malignant fibrous histiocytoma [15]. These masses vary from circumscribed lesions consisting predominantly of adipose tissue, to infiltrating masses without any macroscopically visible adipose element (Fig. 16.3). They are divided into four subtypes based on
Fig. 16.3 Whole-body MRI (STIR sequence) demonstrating a large soft tissue mass (arrow) within the anterior compartment of left thigh. Note the heterogeneity of the lesion, which is characteristic of a liposarcoma
432
M.J. Shelly et al.
histologic features: well-differentiated, myxoid, round cell and pleomorphic [1]. The well-differentiated and myxoid subtypes are low-intermediate grade lesions, which generally have a good prognosis and a five-year survival rate of between 75 percent and 100 percent. The round cell and pleomorphic subtypes are high-grade tumors with five-year survival rates of approximately 20 percent [46]. The appearance of liposarcoma on both CT and MRI is variable and differs with histologic subtype. Well-differentiated liposarcomas predominantly consist of fatty material and have such features as thick septa, foci of nonlipomatous material and contrast enhancement. The septa and other nonlipomatous elements are of soft tissue attenuation on CT. On MRI they are hypointense - isointense to muscle on T1-weighted images (Fig. 16.4a), and hyperintense to muscle on T2-weighted images (Fig. 16.b). The other liposarcoma subtypes have a much higher proportion of nonlipomatous elements, with a large study [47] reporting that none of the myxoid, round cell or pleomorphic tumors had more than 25 percent of their volume composed of adipose tissue. When only nonlipomatous material is present, liposarcomas cannot be distinguished from other soft tissue tumors by CT or MRI.
Fig. 16.4 Coronal T1 (a) and coronal STIR (b) MR images demonstrating a well-circumscribed mass in the posteromedial soft tissue of left lower limb (black arrow). Note that lesion appears isotense on T1-weighted image, and hyperintense on the T2-weighted STIR sequence (white arrow)
16 Radiology of Soft Tissue Tumors Including Melanoma
3 3.1
433
Fibrous Tumors and Tumor-like Lesions Myofibromatosis
Myofibromatosis is a benign proliferation of cells with characteristics of the myofibroblast, and is probably hamartomatous in nature. This process can be multifocal and produces small fibrous masses that are usually located in the dermis or subcutaneous tissues [48]. When lesions are multiple, the viscera are involved in about one-third of cases [49]. It is the most common tumor in infants and the prognosis is generally excellent with little morbidity [15]. These tumors are usually subcutaneous and removed without any imaging. In multifocal disease imaging provides important information on the extent of disease, with lesions usually appearing as discrete nodular masses without macroscopic invasion of adjacent tissues. On CT, the lesions may be solid or have necrotic or cystic portions, and calcification is common. The solid portion has attenuation similar to or slightly greater than skeletal muscle. On MRI the lesions are usually heterogenous on both T1- and T2-weighted sequences [50]. Lesions usually enhance after intravenous contrast administration.
3.2
Aggressive Fibromatoses (Deep Fibromatoses/ Desmoid Tumor)
Aggressive fibromatosis is a histologically benign, but locally aggressive fibroblastic lesion that arises from musculoaponeurotic structures [15]. Histologically they are composed of highly cellular, well-differentiated fibrous tissue whose biologic behavior is intermediate between that of a benign fibrous lesion and fibrosarcoma; although locally aggressive, they never metastasize [48]. They may occur at any age, but are most frequent in the teens to 30s. These lesions occur as unicentric, gray-white, firm, poorly demarcated masses varying in diameter from 1 to 15 cm. They are rubbery and tough, and infiltrate surrounding structures [1]. The imaging appearance of deep fibromatosis is both nonspecific and variable as the mass can be grossly infiltrative or have well-defined margins [51]. On CT examination, it has variable attenuation. On MRI the lesions may be hypointense or isointense relative to adjacent muscle on T1-weighted images, and be hypointense and distinctive on T2-weighted images and show an infiltrative pattern, although more cellular lesions may have areas of increased T2-weighted signal [7]. Hemorrhage and mineralization within lesions may yield a similar appearance [52]. Most lesions show contrast enhancement following intravenous administration of gadolinium, with enhancement corresponding to the cellular portions of the lesion [49].
434
3.3
M.J. Shelly et al.
Fibrosarcoma
Fibrosarcoma is a malignant neoplasm of fibroblasts composed of anaplastic spindle cells with varying degrees of differentiation. Fibrosarcomas are rare (accounting for approximately 5 percent of soft tissue sarcomas [49]), but may occur anywhere in the body, most commonly in the retroperitoneum, the thigh, the knee and the distal extremities [1]. They differ histologically from aggressive fibromatosis by their cellular atypia and relative lack of collagen [48]. Fibrosarcoma metastasizes in greater than 60 percent of cases; consequently, the five-year survival rate is 39 to 54 percent [48]. The imaging appearance of fibrosarcoma is nonspecific. On CT, calcification may be present and the lesion appears as a homogenous soft tissue attenuation mass that may erode adjacent bone. On MRI fibrosarcoma has been reported to be of low- to intermediate signal intensity on all imaging sequences, and to show moderate enhancement after intravenous contrast administration (Fig. 16.5) [53, 54]. Areas of necrosis and hemorrhage may be present.
4 4.1
Fibrohistiocytic Tumors Benign Fibrous Histiocytoma
Also known as a dermatofibroma, this is a relatively common benign lesion that usually occurs in the dermis and subcutaneous tissue. It is painless and slow-growing, and most often presents in mid-adult life as a firm, small (up to 1 cm) mobile
16 Radiology of Soft Tissue Tumors Including Melanoma
435
nodule. Most benign fibrous histiocytomas consist of cells resembling normal fibroblasts and histiocytes [15]. Other variants may contain numerous blood vessels and hemosiderin deposition, and are often called sclerosing hemangiomas. All are variations on a common theme, and in all the margins are infiltrative, but the tumor does not invade the overlying epidermis [1]. The subcutaneous lesions do not usually require radiographic evaluation and deeper histiocytomas have a nonspecific imaging appearance, not reliably differentiated from their malignant counterparts [49].
4.2
Malignant Fibrous Histiocytoma
This is a heterogenous group of aggressive soft tissue tumors with a background of inflamed collagenous stroma, often with foamy macrophages [1]. They are the most common soft tissue malignancy of adulthood [55] and the most common postradiation sarcoma [15]. Ultrastructural studies have shown many of these tumors to be variants of liposarcoma, leiomyosarcoma and rhabdomyosarcoma. It usually arises in the musculature of the proximal extremities and the retroperitoneum. These tumors are large (5 to 20 cm), unencapsulated, aggressive masses that have a metastatic rate of 30 percent to 50 percent [1], and a high propensity for local recurrence. The imaging appearance of malignant fibrous histiocytoma is nonspecific. They are commonly circumscribed masses, and overt invasion of bone is commonly seen. On CT, they are often of homogenous soft tissue attenuation, but may exhibit small foci of calcification, hemorrhage or necrosis [15]. On T1-weighted MR images they are usually of similar signal intensity to muscle. On T2-weighted images they are heterogenous and hyperintense to muscle [56, 57]. These tumors generally enhance following administration of intravenous contrast.
5 5.1
Tumors of Skeletal Muscle Rhabdomyoma
This is a benign neoplasm of skeletal muscle cells. It is most often found in the heart, where it is associated with tuberous sclerosis. Those found in the heart are considered to be hamartomatous, but extracardiac rhabdomyomas are generally considered neoplastic [15]. These lesions are usually small (<1 cm), solitary and encapsulated. They can occur at any age, but are most common in adults, and they show a male predominance [1]. The most common site for extracardiac rhabdomyomas Fig. 16.5 Coronal T1-weighted MR image (a) of a large parascapular fibrosarcoma demonstrating intermediate signal intensity (white arrow). MR angiographic sequence (b) of the same lesion demonstrating extensive vascular supply from the subclavian artery and its branches
436
M.J. Shelly et al.
is the head and neck (accounting for 70 percent of such tumors) [15]. They are much less common than their malignant striated muscle counterparts. They may be found within muscles or as separate soft tissue masses surrounded by fat or fascia [48]. Growth is slow and non-invasive. The appearance on both CT and MRI is nonspecific. They are circumscribed, smoothly marginated masses [15]. On CT their attenuation is similar to that of muscle. With MRI the few reported cases have been generally isointense or slightly hyperintense to muscle on T1-weighted images, and heterogeneously hyperintense to muscle on T2-weighted images [58]. They usually enhance following administration of intravenous contrast [59, 60].
5.2
Rhabdomyosarcoma
Rhabdomyosarcoma, the most common soft tissue sarcoma of childhood and adolescence, usually appears before the age 20 [1]. They may arise in any anatomic location, but most occur in the head and neck or genitourinary tract, where there is little if any skeletal muscle as a normal constituent [61]. Only in the extremities do they appear in relation to skeletal muscle. Rhabdomyosarcoma is histologically subclassified into embryonal, botryoid, alveolar and pleomorphic variants [48]. Embryonal rhabdomyosarcoma is the most common type, accounting for 66 percent of rhabdomyosarcomas, and occurs in children under the age of 10 years, typically arising in the nasal cavity, orbit, middle ear, prostate and paratesticular region. Alveolar rhabdomyosarcoma, accounting for about 20 percent of cases, is most common in adolescence and usually arises in the deep musculature of the extremities. Pleomorphic rhabdomyosarcoma is uncommon (approximately 5 percent of cases) and arises in the deep soft tissues of adults [1]. They are an aggressive tumor type with an overall survival of approximately 65 percent in children, while adults fare less well. [62] The imaging appearance on CT and MRI is of an aggressive soft tissue mass. No imaging features are specific for this neoplasm, but CT often shows local bone invasion, often with a component of bone remodeling [49]. Sclerosis and periosteal reaction may be present. MRI features include isointensity to muscle on T1weighted images, and hyperintensity to muscle on T2-weighted images. Marked contrast enhancement is commonly seen [63].
6 6.1
Tumors of Smooth Muscle Leiomyomas
Leiomyomas are benign tumors of smooth muscle occurring most commonly in the uterus where they represent the most common neoplasm in women, namely uterine fibroids. Leiomyomas may also arise in the skin and subcutaneous tissues of the skin,
16 Radiology of Soft Tissue Tumors Including Melanoma
437
nipples, scrotum and labia (genital leiomyomas), and less frequently in the deep soft tissues. They tend to occur in adolescence and early adult life and are usually not larger than 1 to 2 cm in greatest dimension [1]. Differentiation from their malignant counterparts, leiomyosarcomas, is difficult on both imaging studies and histology [64]. The imaging characteristics of leiomyomas differ, depending on their anatomic location. Leiomyomas generally occur as intramural eccentric lesions and, thus, produce characteristic signs of a well-defined, smooth filling defect on barium examination [64]. On CT leiomyomas appear as round, sharply defined masses of homogenous soft tissue attenuation and show uniform enhancement. CT is useful for establishing the intramural location of the tumor and distinguishing it from a lipoma or extrinsic mass [64]. On MRI leiomyomas typically enhance after intravenous contrast administration and are isointense to muscle on T1-weighted images, and of heterogenous signal intensity on T2-weighted images.
6.2
Leiomyosarcoma
Leiomyosarcomas account for 10 to 20 percent of soft tissue sarcomas. They occur in adults in the fifth to seventh decades of life and, in general, afflict women more frequently than men; a notable exception to this rule being leiomyosarcoma located in the skin and subcutaneous tissues [48]. The majority of lesions develop in the skin and deep soft tissues of the extremities, and the retroperitoneum (most commonly from the inferior vena cava [65]). They generally present late as painless, firm masses, with retroperitoneal tumors that are often large (up to 35 cm in size) and bulky, leading to abdominal symptoms [1, 48]. On CT leiomyosarcoma is seen as a large, well defined, heterogenous, hypodense mass, with or without central necrotic or cystic degeneration, that exhibits either homogenous or heterogenous enhancement after administration of intravenous contrast [66]. On MRI the lesion appears hypointense on T1-weighted images and shows intermediate signal intensity on T2-weighted images, with heterogenous enhancement apparent after administration of intravenous contrast [67] (Fig. 16.6). The tumor margin is often indistinct from the adjacent structures. On ultrasound leiomyosarcoma presents as a hypoechoic mass that can be heterogenous in echogenicity and may show centrally cystic areas corresponding to necrosis, with internal hemorrhage having a variable appearance as either hypo-, iso- or hyperechoic, depending on its stage of maturity [68].
7
Vascular Tumors
Tumors of the blood and lymphatic vessels constitute a spectrum from the benign hemangiomas, to relatively rare, highly malignant angiosarcomas. Benign tumors produce readily recognized vascular channels filled with blood. Malignant tumors are more solidly cellular and usually do not form well-organized vessels [69].
438
M.J. Shelly et al.
Fig. 16.6 Contrast-enhanced coronal T1 image (a) of a leiomyosarcoma (arrow) infiltrating the medial belly of the triceps muscle. Axial T1-weighted image (b) demonstrating the same lesion
7.1
Hemangioma
Hemangiomas may be difficult to distinguish with certainty from vascular malformations or hamartomas. Hemangiomas are most commonly localized; however, some involve large segments of the body, such as an entire extremity, for which the term angiomatosis is applied. The majority are superficial lesions, often of the head or neck, but they may occur internally, with nearly one-third in the liver. Malignant transformation occurs rarely, if at all. Hemangiomas are extremely common, particularly in infancy and childhood, constituting 7 percent of all benign tumors [69]. There are several histologic and clinical variants. Capillary Hemangiomas (the largest single group) usually occur in the skin, subcutaneous tissues and mucous membranes of the oral cavities and lips, but they may also occur in the internal viscera, such as the liver, spleen and kidneys. The strawberry type of capillary hemangioma is extremely common (1 in 200 births), and may be multiple. Capillary hemangiomas vary in size from a few millimeters up to several centimeters in diameter. Lesions are level with the surface of the skin or slightly elevated and they are usually encapsulated [69]. Cavernous Hemangioma are less common than the capillary variety and are distinguished by the formation of large, dilated vascular channels. They are usually larger (1 to 2 cm in diameter), less circumscribed,and more likely to involve deep structures than capillary hemangiomas. Intravascular thrombosis, with associated dystrophic calcification, is common [69]. Superficial hemangiomas are rarely imaged as excision without expensive imaging is the most appropriate management of these lesions. For deeper hemangiomas, MRI demonstrates intermediate signal intensity on T1-weighted (Fig. 16.7) spin echo images, and extremely bright signal on T2-weighted images (Fig. 16.8). Ultrasound
16 Radiology of Soft Tissue Tumors Including Melanoma
439
Fig. 16.7 Axial T1-weighted MR image (a) of a poorly marginated haemangioma within the extensor compartment of the right forearm (white arrow). Coronal T2-weighted MR image (b) of the same lesion
usually shows a hypoechoic soft tissue mass, and acoustic shadowing consistent with a calcified phlebolith. Increased color flow and low resistance is commonly found on Doppler ultrasound of lesions. CT imaging yields a nonspecific mass with phleboliths commonly visible [70].
7.2
Angiosarcoma
Angiosarcomas are malignant endothelial neoplasia that occur in both sexes, In older adults they can occur anywhere in the body, but most often in the skin, soft tissue, breast and liver. Approximately 90 percent of reported cases occur in the upper extremity after mastectomy for breast carcinoma, with a number of tumors occurring in the lower extremities in association with chronic lymphedema [71]. Angiosarcoma may begin as deceptively small, sharply demarcated, asymptomatic, often multiple red nodules, but eventually most angiosarcomas become large, fleshy masses. The tumor margins blend with surrounding structures and areas of necrosis and hemorrhage in the center of the lesion are commonly seen [69].
440
M.J. Shelly et al.
Fig. 16.8 Coronal T1-weighted MR image demonstrating a plexiform neurofibroma (white arrows) on the plantar surface of the right foot. Note that it extends into the tarsal tunnel and involves branches of the medial plantar nerve (black arrow)
The highly vascular nature of angiosarcomas leads to characteristic radiologic findings. On CT they appear as dense, irregularly shaped masses in the subcutaneous fat that significantly enhance after administration of iodinated contrast medium [72]. On MRI lesions show low signal intensity on T1-weighted images, and low to intermediate and heterogenous signal intensity on T2-weighted images, compared with fat. Angiosarcomas demonstrate heterogenous enhancement following
16 Radiology of Soft Tissue Tumors Including Melanoma
441
administration of intravenous gadolinium, with areas of stronger enhancement reflecting the component of the tumor with relatively large blood vessels and narrow stroma [72]. Dynamic contrast-enhanced MRI findings have been reported in angiosarcoma of the liver. This lesion was found to enhance in a peripheral nodular fashion on immediate post-gadolinium MR images, and showed centripetal progression of enhancement on delayed images [73]. This was thought to reflect the size and number of feeding blood vessels, which might also be an index of tumor angiogenesis [73].
8 8.1
Peripheral Nerve Sheath Tumors Schwannoma / Neurilemoma
These benign tumors arise from the neural crest-derived Schwann cell, and are associated with Neurofibromatosis Type 2 [74]. They are well-circumscribed, encapsulated masses that are attached to the nerve and tend to present in the fourth and fifth decades of life [48]. They are firm masses, often with cystic areas, that do not become malignant [74]. Ultrasound imaging of schwannomas shows a well-defined hypoechoic mass with variable acoustic enhancement [75, 76]; the presence of cavitation is virtually diagnostic of schwannoma [77]. On CT, schwannoma appears as a well-defined soft tissue mass which is less dense than surrounding muscle, and enhances heterogeneously before and after intravenous contrast [78].
8.2
Neurofibroma
Two histologically distinct lesions have been termed neurofibromas. The most common form occurs in the skin (cutaneous neurofibroma) or in peripheral nerve (solitary neurofibroma). These lesions arise sporadically or in association with Neurofibromatosis Type 1, and they present in the dermis and subcutaneous fat as well, delineated but un-encapsulated masses. They are not invasive, but adnexal structures can become enwrapped by the edges of the lesion. (Fig. 16.8) The skin lesions are evident as nodules that may grow to be large and become pedunculated. The risk of malignant transformation in these tumors is extremely small [48, 74]. The second type is the plexiform neurofibroma which only occurs in patients with Neurofibromatosis Type 1, and has a significant potential for malignant transformation. They can arise anywhere along the extent of a nerve (large nerve trunks are the most common site), and are frequently multiple. At the site of each lesion the host nerve is irregularly expanded, as each of its fascicles is infiltrated by neoplasm. The proximal and distal extremes of the tumor may have poorly defined margins [74].
442
M.J. Shelly et al.
Ultrasound imaging in neurofibroma demonstrates a well-defined, hypoechoic mass with variable acoustic enhancement [76, 79]. CT shows a well-defined soft tissue mass which is less dense than muscle with homogenous enhancement after intravenous contrast [80, 81]. MRI of peripheral nerve sheath tumors is nonspecific with no features capable of confidently differentiating benign from malignant tumors [82]. Most neurofibromas are mildly heterogenous with intermediate to moderately bright signal on T1-weighted images, and high signal on T2-weighted images [83]. A central low intensity “target” appearance is common in neurofibromas [84, 85]. Most have smooth, well-defined margins. An irregular infiltrative pattern is commonly seen in plexiform neurofibroma [83, 84]. CT shows a diffuse low attenuation mass which may envelope adjacent structures, and calcification may be present [86]. MRI shows single or multiple masses along peripheral nerves that may enhance and are isointense to muscle on T1-weighted and hyperintense on T2-weighted images [84, 86, 87]. FDG-PET has been shown to be useful in determining malignant change in plexiform neurofibromas in patients with neurofibromatosis Type 1 [88].
8.3
Malignant Peripheral Nerve Sheath Tumor (Malignant Schwannoma)
These are highly malignant sarcomas that are locally invasive, frequently leading to local recurrences and eventual metastatic spread [89]. They account for 10 percent of all soft tissue sarcomas with 10 percent of lesions associated with previous irradiation, and 50 percent with Neurofibromatosis Type 1 [48]. They are poorly defined heterogenous tumor masses consisting of spindle cells focally arranged in sweeping fascicles alternating with areas of necrosis [74, 89]. There is frequent infiltration of the tumor along the axis of the parent nerve, as well as invasion of the adjacent soft tissues [48, 74]. Ultrasound shows a hypoechoic mass which may be ill-defined, contain echogenic foci due to hemorrhage and necrosis, or areas of cystic degeneration [90]. Doppler studies show hypervascularity, but the malignant schwannoma may be ultrastructurally indistinguishable from benign nerve tumors [75]. Malignant schwannoma avidly takes up gallium citrate, but benign nerve sheath tumors rarely do, thus, this can be an important differentiating feature [91]. CT shows an enhancing, irregularly marginated soft tissue mass which may displace or invade adjacent soft tissue or bone. It is iso- or hypodense to muscle, usually homogenous and may contain calcification or cavities [80]. On MRI malignant schwannoma is seen as a heterogenous, irregular, infiltrative mass that is hyperintense to muscle on T2weighted images, and similar to muscle on T1-weighted images. The “target sign” is not a feature [84, 91]. In recent times, FDG-PET has been used to successfully distinguish benign from malignant nerve sheath tumors, with FDG uptake appreciably higher in malignant variants [92].
16 Radiology of Soft Tissue Tumors Including Melanoma
9
443
Synovial Sarcoma
Synovial sarcoma accounts for up to 10 percent of all primary malignant soft tissue neoplasms and ranks as the fourth most common type of sarcoma. Despite its name, only 10 percent of lesions occur in an intra-articular location with the remainder occurring near joints [1]. The tumor does not arise from the synovial membrane, but rather from primitive mesenchymal cells [93]. Synovial sarcomas typically affect the extremities of adolescents and young adults, particularly the knee in the popliteal fossa. These lesions are an intermediate- to high-grade neoplasm with extensive metastatic potential [94]. Patients usually present with a deep-seated mass that has been noted for several years. Uncommonly, these tumors may occur in the parapharyngeal region or in the abdominal wall. The imaging findings in synovial sarcoma, although not pathogenomic, frequently suggest the diagnosis. Radiographs appear normal in approximately 50 percent of synovial sarcoma cases, particularly with small lesions [95]. Lesions typically appear as nonspecific, round to oval juxta-articular soft tissue masses with calcification present on plain film radiology in up to 30 percent of cases. These calcifications are often eccentric or peripheral within the soft tissue mass. Involvement of the underlying bone is not uncommon, but the bone erosion often has an indolent non-aggressive appearance on radiographs, which can lead to misinterpreting the lesion as a benign process [94]. The lesion typically appears hypervascular on angiography. Isotope bone scan reveals prominent increased uptake of technetium-99 m on blood flow and blood pool images, a finding that reflects the increased vascularity of these lesions [96]. On CT, synovial sarcoma appears as a heterogenous, multinodular deep-seated soft tissue mass with attenuation similar to or slightly lower than that of muscle [94]. Areas of lower attenuation represent necrosis or hemorrhage, but smaller lesions may be more homogenous. CT, after administration of intravenous contrast, shows heterogenous enhancement in 90 percent to 100 percent of cases [97]. CT is also useful for detecting calcification and bone involvement, particularly in complex anatomical areas such as the pelvis or shoulder, or when lesions are small and subtle. On MRI T1-weighted images typically demonstrate a prominently heterogenous, multilobulated soft tissue mass with signal intensity similar to or slightly higher than that of muscle (Fig. 16.9). Prominent heterogeneity with predominant high signal intensity is also a feature of T2-weighted images [98]. Jones [99] described this signal heterogeneity as the triple sign, representing intermixed areas of low, intermediate and high signal intensity on long repetition time images. The triple sign is presumably the result of the mixture of solid cellular elements (intermediate signal intensity), hemorrhage or necrosis (high signal intensity), and calcified or fibrotic collagenous regions (low signal intensity). The triple sign has been described as occurring in up to 60 percent of cases of synovial sarcoma [94]. These lesions tend to be sharply marginated with a heterogenous, multiloculated appearance with varying degrees of internal septations, with or without fluid-fluid levels [93].
444
M.J. Shelly et al.
Fig. 16.9 Coronal oblique T1-weighted MR image of a synovial sarcoma (white arrow) within the soft tissue of the right upper limb adjacent to the elbow joint (star). Note its heterogeneity and similar signal intensity to muscle
10
Malignant Melanoma
Malignant melanoma is a relatively common neoplasm that was, until recently, considered uniformly deadly. Although the disease accounts for only 4 percent of all cancer cases, it accounts for approximately 79 percent of skin cancer-related deaths, and the incidence of cutaneous melanoma is on the rise with an estimated 60,000 new cases diagnosed in 2005 in the United States [100]. Although the great preponderance of melanomas arise in the skin, other sites of origin include the oral and anogenital mucosal surfaces, esophagus, meninges and the eye [101]. In this chapter we will focus on cutaneous melanomas. Today, as a result of increased public awareness of the earliest signs of skin melanomas, most are cured surgically [102], but unfortunately, as before the incidence of these lesions is increasing. Central to the understanding of the complicated histology and growth pattern of malignant melanoma is the concept of radial and vertical growth. Simply stated, radial growth
16 Radiology of Soft Tissue Tumors Including Melanoma
445
indicates the tendency of a melanoma to grow horizontally within the epidermal and superficial dermal layers, often for a prolonged time. During this stage of growth, melanoma cells do not have the capacity to metastasize. With time, the pattern of growth assumes a vertical component, and the melanoma grows downwards into the deeper dermal layers [101]. The probability of metastasis in such lesions may be predicted by simply measuring in millimeters the depth of invasion of this vertical growth phase nodule below the epidermis [103]. Predicting clinical outcome has been further refined by taking into account factors such as number of mitoses and degree of infiltrative lymphocytic response within the tumor nodule [104].
Fig. 16.10 Coronal PET/CT fused image of a patient with extensive metastatic melanoma. Arrows highlight areas of increased uptake within the liver and hilar lymph nodes consistent with metastatic disease
446
M.J. Shelly et al.
Malignant melanoma has a very strong propensity for loco-regional and distant metastases due to the biology of the tumor and its complex metastatic pathways that include local extension, regional spread to lymph nodes and distant spread to visceral organs [105]. Most patients develop regional lymph node metastases as their first manifestation of the disease’s spread [106]. Since melanoma metastases can affect every organ in the body, the appearance of these lesions varies according to the surrounding tissue. On MRI of the brain, for example, there are two broad divisions of melanoma’s appearance; the melanotic and amelanotic patterns. The melanotic pattern consists of high signal intensity on T1-weighted images and low signal intensity on T2-weighted images. In the amelanotic pattern, lesions appear hypointense or isointense to the cerebral cortex on T1-weighted images and hyperintense or isointense to the cortex on T2-weighted images [107]. Nuclear medicinebased techniques, such as lymph node mapping with sentinel lymphadenectomy, FDG-PET and, more recently, anatomic-functional imaging with hybrid devices such as PET/CT and single photon emission CT (SPECT)/CT, play an essential role in the overall management of patients with melanoma (Fig. 16.10). A large body of evidence shows the clinical usefulness of lymph node mapping and lymphadenectomy in the detection of metastases in the early stages of disease (stages 1 to 2) [106, 108, 109], whereas FDG-PET is preferably indicated in advanced disease (Stages 3 to 4) [110, 111]. Chest CT is the modality of choice for pulmonary surveillance as it is more informative than standard chest X-ray and more sensitive than FDG-PET, although FDG-PET is more specific [105]. The recent introduction of combined PET/CT scanners will allow for simultaneous assessment of both metabolic and anatomic characteristics of the primary tumor and its potential local, regional and distant spread [110]. Similarly, more accurate localization of sentinel lymph nodes before surgery is achievable with new hybrid SPECT/CT systems [112]. At the frontier of translational research, FDG-sensitive intraoperative probes have been designed and used with encouraging results to optimize the detection of melanoma tumor deposits in vivo and, thus, open new avenues for image-guided therapy, as well as radio-guided surgery [113].
Key Points ●
●
●
Imaging plays a vital role in the diagnosis, preoperative evaluation, treatment planning, treatment evaluation and long-term monitoring of soft tissue tumors. It is complementary to clinical examination and pathology in the initial work-up of a suspicious soft tissue mass. Each imaging modality has its strengths and weaknesses, and knowledge of the uses and shortcomings of these modalities can help in accurate, timely and costeffective patient evaluation. The role of plain film radiology, angiography and radioisotope bone scan in the evaluation of soft tissue tumors is limited as these modalities have been largely superseded by MRI and PET.
16 Radiology of Soft Tissue Tumors Including Melanoma ●
●
●
●
●
●
447
Ultrasound, with its lack of ionizing radiation and dynamic real-time imaging capabilities, is a readily available, non-invasive and relatively inexpensive method of detecting and determining the size and constituency of a soft tissue mass. The use of ultrasound is somewhat limited to differentiating cystic from solid masses, and evaluating tumor vascularity. MRI is the preferred modality for evaluating a soft tissue mass due to its excellent soft tissue contrast, multiplanar image acquisition and lack of ionizing radiation. It must be emphasized that MRI cannot reliably distinguish between benign and malignant lesions, and when radiologic evaluation is nonspecific, one is illadvised to suggest a lesion is benign or malignant solely on the basis of its MRI appearance. CT is useful in specific instances for the identification of subtle soft tissue mineralization and involvement of bone. It is the screening tool of choice for the detection of distant metastases, particularly in the lung. Newer technologies, such as PET/CT and SPECT/CT, allow for the simultaneous precise anatomic localization of primary and secondary lesions, and assessment of their metabolic activity. This information facilitates accurate and early staging of soft tissue tumors. Recent advances in molecular imaging have expanded the armamentarium of imaging modalities available to the clinician. The use of tumor-specific probes that are readily visible on PET/CT is an exciting development that will take the staging of soft tissue tumors to a new level of accuracy and specificity that has been heretofore unavailable.
References 1. Rosenberg A. Bones, Joints, and Soft Tissue Tumors. In: Cotran RS, Kumar V, Collins T, editors. Robbins Pathologic Basis of Disease. 6th ed. Philadelphia: WB Saunders; 1999. 1259-1267. 2. Brady MS, Gaynor JJ, Brennan MF. Radiation-associated sarcoma of bone and soft tissue. Arch Surg 1992; 127(12):1379-1385. 3. Brennan MF. The surgeon as a leader in cancer care: lessons learned from the study of soft tissue sarcoma. J Am Coll Surg 1996; 182(6):520-529. 4. Guillou L, Coindre JM, Bonichon F, et al. Comparative study of the National Cancer Institute and French Federation of Cancer Centres Sarcoma Group grading systems in a population of 420 adult patients with soft tissue sarcoma. J Clin Oncol 1997; 15(1):350-362. 5. Heslin MJ, Smith JK. Imaging of soft tissue sarcomas. Surg Oncol Clin N Am 1999; 8(1):91-107. 6. Sanders TG, Parsons TW, III. Radiographic imaging of musculoskeletal neoplasia. Cancer Control 2001; 8(3):221-231. 7. Ilaslan H, Sundaram M. Advances in musculoskeletal tumor imaging. Orthop Clin North Am 2006; 37(3):375-91, vii. 8. CKERMAN LV. Extra-osseous localized non-neoplastic bone and cartilage formation (so-called myositis ossificans): clinical and pathological confusion with malignant neoplasms. J Bone Joint Surg Am 1958; 40-A(2):279-298.
448
M.J. Shelly et al.
9. Levine E, Lee KR, Neff JR, Maklad NF, Robinson RG, Preston DF. Comparison of computed tomography and other imaging modalities in the evaluation of musculoskeletal tumors. Radiology 1979; 131(2):431-437. 10. Martel W, Abell MR. Radiologic evaluation of soft tissue tumors. A retrospective study. Cancer 1973; 32(2):352-366. 11. Kransdorf MJ, Jelinek JS, Moser RP, Jr. Imaging of soft tissue tumors. Radiol Clin North Am 1993; 31(2):359-372. 12. Chew FS, Hudson TM, Enneking WF. Radionuclide imaging of soft tissue neoplasms. Semin Nucl Med 1981; 11(4):266-276. 13. Fornage BD. Soft tissue masses. Clin Diagn Ultrasound 1995; 30:21-42. 14. Taylor GA, Perlman EJ, Scherer LR, Gearhart JP, Leventhal BG, Wiley J. Vascularity of tumors in children: evaluation with color Doppler imaging. AJR Am J Roentgenol 1991; 157(6):1267-1271. 15. Eskey CJ, Robson CD, Weber AL. Imaging of benign and malignant soft tissue tumors of the neck. Radiol Clin North Am 2000; 38(5):1091-104, xi. 16. Richardson ML, Gillespy T. Magnetic Resonance Imaging. In: Kricun ME, editor. Imaging of bone tumors. Philadelphia: WB Saunders; 1993. 358-446. 17. Cheng EY, Thompson RC, Jr. New developments in the staging and imaging of soft tissue sarcomas. Instr Course Lect 2000; 49:443-451. 18. Reddick WE, Wang S, Xiong X, Glass JO, Wu S, Kaste SC et al. Dynamic magnetic resonance imaging of regional contrast access as an additional prognostic factor in pediatric osteosarcoma. Cancer 2001; 91(12):2230-2237. 19. Kole AC, Nieweg OE, van Ginkel RJ, Pruim J, Hoekstra HJ, Paans AM et al. Detection of local recurrence of soft tissue sarcoma with positron emission tomography using [18F]fluorodeoxyglucose. Ann Surg Oncol 1997; 4(1):57-63. 20. Hawkins DS, Schuetze SM, Butrynski JE, Rajendran JG, Vernon CB, Conrad EU, III et al. [18F]Fluorodeoxyglucose positron emission tomography predicts outcome for Ewing sarcoma family of tumors. J Clin Oncol 2005; 23(34):8828-8834. 21. Schuetze SM, Rubin BP, Vernon C, Hawkins DS, Bruckner JD, Conrad EU, III et al. Use of positron emission tomography in localized extremity soft tissue sarcoma treated with neoadjuvant chemotherapy. Cancer 2005; 103(2):339-348. 22. Jones DN, McCowage GB, Sostman HD, Brizel DM, Layfield L, Charles HC et al. Monitoring of neoadjuvant therapy response of soft tissue and musculoskeletal sarcoma using fluorine-18FDG PET. J Nucl Med 1996; 37(9):1438-1444. 23. Hawkins DS, Rajendran JG, Conrad EU, III, Bruckner JD, Eary JF. Evaluation of chemotherapy response in pediatric bone sarcomas by [F-18]-fluorodeoxy-D-glucose positron emission tomography. Cancer 2002; 94(12):3277-3284. 24. Folpe AL, Lyles RH, Sprouse JT, Conrad EU, III, Eary JF. (F-18) fluorodeoxyglucose positron emission tomography as a predictor of pathologic grade and other prognostic variables in bone and soft tissue sarcoma. Clin Cancer Res 2000; 6(4):1279-1287. 25. McCarville MB, Christie R, Daw NC, Spunt SL, Kaste SC. PET/CT in the evaluation of childhood sarcomas. AJR Am J Roentgenol 2005; 184(4):1293-1304. 26. Wegner EA, Barrington SF, Kingston JE, Robinson RO, Ferner RE, Taj M et al. The impact of PET scanning on management of paediatric oncology patients. Eur J Nucl Med Mol Imaging 2005; 32(1):23-30. 27. Lucas JD, O’Doherty MJ, Cronin BF, Marsden PK, Lodge MA, McKee PH et al. Prospective evaluation of soft tissue masses and sarcomas using fluorodeoxyglucose positron emission tomography. Br J Surg 1999; 86(4):550-556. 28. Lucas JD, O’Doherty MJ, Wong JC, Bingham JB, McKee PH, Fletcher CD et al. Evaluation of fluorodeoxyglucose positron emission tomography in the management of soft tissue sarcomas. J Bone Joint Surg Br 1998; 80(3):441-447. 29. Siegel BA, Dehdashti F. Oncologic PET/CT: current status and controversies. Eur Radiol 2005; 15 Suppl 4:D127-D132.
16 Radiology of Soft Tissue Tumors Including Melanoma
449
30. Conrad EU, III, Morgan HD, Vernon C, Schuetze SM, Eary JF. Fluorodeoxyglucose positron emission tomography scanning: basic principles and imaging of adult soft tissue sarcomas. J Bone Joint Surg Am 2004; 86-A Suppl 2:98-104. 31. Aoki J, Endo K, Watanabe H, Shinozaki T, Yanagawa T, Ahmed AR et al. FDG-PET for evaluating musculoskeletal tumors: a review. J Orthop Sci 2003; 8(3):435-441. 32. Kole AC, Nieweg OE, Hoekstra HJ, van H, Jr., Koops HS, Vaalburg W. Fluorine-18-fluorodeoxyglucose assessment of glucose metabolism in bone tumors. J Nucl Med 1998; 39(5):810-815. 33. Aoki J, Watanabe H, Shinozaki T, Takagishi K, Tokunaga M, Koyama Y et al. FDG-PET for preoperative differential diagnosis between benign and malignant soft tissue masses. Skeletal Radiol 2003; 32(3):133-138. 34. Ben Arush MW, Bar SR, Postovsky S, Militianu D, Haimi M, Zaidman I et al. Assessing the use of FDG-PET in the detection of regional and metastatic nodes in alveolar rhabdomyosarcoma of extremities. J Pediatr Hematol Oncol 2006; 28(7):440-445. 35. Schuetze SM. Utility of positron emission tomography in sarcomas. Curr Opin Oncol 2006; 18(4):369-373. 36. Schuetze SM. Imaging and response in soft tissue sarcomas. Hematol Oncol Clin North Am 2005; 19(3):471-87, vi. 37. Yao L, Nelson SD, Seeger LL, Eckardt JJ, Eilber FR. Primary musculoskeletal neoplasms: effectiveness of core-needle biopsy. Radiology 1999; 212(3):682-686. 38. Dupuy DE, Rosenberg AE, Punyaratabandhu T, Tan MH, Mankin HJ. Accuracy of CT-guided needle biopsy of musculoskeletal neoplasms. AJR Am J Roentgenol 1998; 171(3):759-762. 39. Silverman SG, Tuncali K, Adams DF, Nawfel RD, Zou KH, Judy PF. CT fluoroscopy-guided abdominal interventions: techniques, results, and radiation exposure. Radiology 1999; 212(3):673-681. 40. Fletcher JA. Soft Tissue Tumors. St. Louis: Mosby; 1995. 41. Ahuja AT, King AD, Kew J, King W, Metreweli C. Head and neck lipomas: sonographic appearance. AJNR Am J Neuroradiol 1998; 19(3):505-508. 42. Gritzmann N, Schratter M, Traxler M, Helmer M. Sonography and computed tomography in deep cervical lipomas and lipomatosis of the neck. J Ultrasound Med 1988; 7(8):451-456. 43. Fornage BD, Tassin GB. Sonographic appearances of superficial soft tissue lipomas. J Clin Ultrasound 1991; 19(4):215-220. 44. Hosono M, Kobayashi H, Fujimoto R, Kotoura Y, Tsuboyama T, Matsusue Y et al. Septumlike structures in lipoma and liposarcoma: MR imaging and pathologic correlation. Skeletal Radiol 1997; 26(3):150-154. 45. Gaskin CM, Helms CA. Lipomas, lipoma variants, and well-differentiated liposarcomas (atypical lipomas): results of MRI evaluations of 126 consecutive fatty masses. AJR Am J Roentgenol 2004; 182(3):733-739. 46. Stewart MG, Schwartz MR, Alford BR. Atypical and malignant lipomatous lesions of the head and neck. Arch Otolaryngol Head Neck Surg 1994; 120(10):1151-1155. 47. Einarsdottir H, Soderlund V, Larsson O, Mandahl N, Bauer HC. 110 subfascial lipomatous tumors. MR and CT findings versus histopathological diagnosis and cytogenetic analysis. Acta Radiol 1999; 40(6):603-609. 48. Enzinger FM, Weiss SW. Soft Tissue Tumors. 3rd ed. St Louis: Mosby; 1995. 49. Kransdorf MJ, Murphey MD. Imaging of Soft Tissue Tumors. Philadelphia: WB Saunders; 1997. 50. Eich GF, Hoeffel JC, Tschappeler H, Gassner I, Willi UV. Fibrous tumors in children: imaging features of a heterogeneous group of disorders. Pediatr Radiol 1998; 28(7):500-509. 51. Quinn SF, Erickson SJ, Dee PM, Walling A, Hackbarth DA, Knudson GJ et al. MR imaging in fibromatosis: results in 26 patients with pathologic correlation. AJR Am J Roentgenol 1991; 156(3):539-542. 52. Hartman TE, Berquist TH, Fetsch JF. MR imaging of extraabdominal desmoids: differentiation from other neoplasms. AJR Am J Roentgenol 1992; 158(3):581-585. 53. Patel SC, Silbergleit R, Talati SJ. Sarcomas of the head and neck. Top Magn Reson Imaging 1999; 10(6):362-375.
450
M.J. Shelly et al.
54. O’Connell TE, Castillo M, Mukherji SK. Fibrosarcoma arising in the maxillary sinus: CT and MR features. J Comput Assist Tomogr 1996; 20(5):736-738. 55. White LM, Buckwalter KA. Technical considerations: CT and MR imaging in the postoperative orthopedic patient. Semin Musculoskelet Radiol 2002; 6(1):5-17. 56. Mahajan H, Kim EE, Wallace S, Abello R, Benjamin R, Evans HL. Magnetic resonance imaging of malignant fibrous histiocytoma. Magn Reson Imaging 1989; 7(3):283-288. 57. Munk PL, Sallomi DF, Janzen DL, Lee MJ, Connell DG, O’Connell JX et al. Malignant fibrous histiocytoma of soft tissue imaging with emphasis on MRI. J Comput Assist Tomogr 1998; 22(5):819-826. 58. Helmberger RC, Stringer SP, Mancuso AA. Rhabdomyoma of the pharyngeal musculature extending into the prestyloid parapharyngeal space. AJNR Am J Neuroradiol 1996; 17(6):1115-1118. 59. Freije JE, Gluckman JL, Biddinger PW, Wiot G. Muscle tumors in the parapharyngeal space. Head Neck 1992; 14(1):49-54. 60. Vermeersch H, van VP, Lemmerling M, Moerman M, De PC. Bilateral recurrent adult rhabdomyomas of the pharyngeal wall. Eur Arch Otorhinolaryngol 2000; 257(1):24-26. 61. Malogolowkin MH, Ortega JA. Rhabdomyosarcoma of childhood. Pediatr Ann 1988; 17(4):251, 254-251, 268. 62. Pappo AS, Shapiro DN, Crist WM. Rhabdomyosarcoma. Biology and treatment. Pediatr Clin North Am 1997; 44(4):953-972. 63. Yousem DM, Lexa FJ, Bilaniuk LT, Zimmerman RI. Rhabdomyosarcomas in the head and neck: MR imaging evaluation. Radiology 1990; 177(3):683-686. 64. Betts MT, Huo EJ, Miller FH. Gastrointestinal and genitourinary smooth-muscle tumors. AJR Am J Roentgenol 2003; 181(5):1349-1354. 65. Hines OJ, Nelson S, Quinones-Baldrich WJ, Eilber FR. Leiomyosarcoma of the inferior vena cava: prognosis and comparison with leiomyosarcoma of other anatomic sites. Cancer 1999; 85(5):1077-1083. 66. McLeod AJ, Zornoza J, Shirkhoda A. Leiomyosarcoma: computed tomographic findings. Radiology 1984; 152(1):133-136. 67. Ohnishi T, Yoshioka H, Ishida O. MR imaging of gastrointestinal leiomyosarcoma. Radiat Med 1991; 9(3):114-117. 68. Chen JJ, Changchien CS, Chiou SS, Tai DI, Lee CM, Kuo CH. Various sonographic patterns of smooth muscle tumors of the gastrointestinal tract: a comparison with computed tomography. J Ultrasound Med 1992; 11(10):527-531. 69. Schoen FJ, Cotran RS. Blood Vessels. In: Cotran RS, Kumar V, Collins T, editors. Robbins Pathologic Basis of Disease. 6th ed. Philadelphia: WB Saunders; 1999. 531-538. 70. Greenspan A, McGahan JP, Vogelsang P, Szabo RM. Imaging strategies in the evaluation of soft tissue hemangiomas of the extremities: correlation of the findings of plain radiography, angiography, CT, MRI, and ultrasonography in 12 histologically proven cases. Skeletal Radiol 1992; 21(1):11-18. 71. Woodward AH, Ivins JC, Soule EH. Lymphangiosarcoma arising in chronic lymphedematous extremities. Cancer 1972; 30(2):562-572. 72. Nakazono T, Kudo S, Matsuo Y, Matsubayashi R, Ehara S, Narisawa H et al. Angiosarcoma associated with chronic lymphedema (Stewart-Treves syndrome) of the leg: MR imaging. Skeletal Radiol 2000; 29(7):413-416. 73. Worawattanakul S, Semelka RC, Kelekis NL, Woosley JT. Angiosarcoma of the liver: MR imaging pre- and post-chemotherapy. Magn Reson Imaging 1997; 15(5):613-617. 74. De Girolami U, Anthony DC, Frosch MP. The Central Nervous System. In: Cotran RS, Kumar V, Collins T, editors. Robbins Pathologic Basis of Disease. 6th ed. Philadelphia: WB Saunders; 1999. 1352-1354. 75. Hughes DG, Wilson DJ. Ultrasound appearances of peripheral nerve tumors. Br J Radiol 1986; 59(706):1041-1043. 76. Fornage BD. Peripheral nerves of the extremities: imaging with US. Radiology 1988; 167(1):179-182.
16 Radiology of Soft Tissue Tumors Including Melanoma
451
77. Kuo CH, Changchien CS. Sonographic features of retroperitoneal neurilemoma. J Clin Ultrasound 1993; 21(5):309-312. 78. Cohen LM, Schwartz AM, Rockoff SD. Benign schwannomas: pathologic basis for CT inhomogeneities. AJR Am J Roentgenol 1986; 147(1):141-143. 79. Hughes DG, Wilson DJ. Ultrasound appearances of peripheral nerve tumors. Br J Radiol 1986; 59(706):1041-1043. 80. Chui MC, Bird BL, Rogers J. Extracranial and extraspinal nerve sheath tumors: computed tomographic evaluation. Neuroradiology 1988; 30(1):47-53. 81. Kumar AJ, Kuhajda FP, Martinez CR, Fishman EK, Jezic DV, Siegelman SS. Computed tomography of extracranial nerve sheath tumors with pathological correlation. J Comput Assist Tomogr 1983; 7(5):857-865. 82. Armstrong SJ, Watt I. Imaging of soft tissue tumors. Curr Opin Radiol 1992; 4(6):39-44. 83. Beggs I. Pictorial review: imaging of peripheral nerve tumors. Clin Radiol 1997; 52(1):8-17. 84. Stull MA, Moser RP, Jr., Kransdorf MJ, Bogumill GP, Nelson MC. Magnetic resonance appearance of peripheral nerve sheath tumors. Skeletal Radiol 1991; 20(1):9-14. 85. Burk DL, Jr., Brunberg JA, Kanal E, Latchaw RE, Wolf GL. Spinal and paraspinal neurofibromatosis: surface coil MR imaging at 1.5 T1. Radiology 1987; 162(3):797-801. 86. Gossios KJ, Guy RL. Case report: imaging of widespread plexiform neurofibromatosis. Clin Radiol 1993; 47(3):211-213. 87. Suh JS, Abenoza P, Galloway HR, Everson LI, Griffiths HJ. Peripheral (extracranial) nerve tumors: correlation of MR imaging and histologic findings. Radiology 1992; 183(2):341-346. 88. Ferner RE, Lucas JD, O’Doherty MJ, Hughes RA, Smith MA, Cronin BF et al. Evaluation of (18)fluorodeoxyglucose positron emission tomography ( (18)FDG PET) in the detection of malignant peripheral nerve sheath tumors arising from within plexiform neurofibromas in neurofibromatosis 1. J Neurol Neurosurg Psychiatry 2000; 68(3):353-357. 89. Terry DG, Sauser DD, Gordon MD. Intraosseous malignant peripheral nerve sheath tumor in a patient with neurofibromatosis. Skeletal Radiol 1998; 27(6):346-349. 90. Fornage BD. Peripheral nerves of the extremities: imaging with US. Radiology 1988; 167(1):179-182. 91. Levine E, Huntrakoon M, Wetzel LH. Malignant nerve-sheath neoplasms in neurofibromatosis: distinction from benign tumors by using imaging techniques. AJR Am J Roentgenol 1987; 149(5):1059-1064. 92. Cardona S, Schwarzbach M, Hinz U, mitrakopoulou-Strauss A, Attigah N, Mechtersheimer section et al. Evaluation of F18-deoxyglucose positron emission tomography (FDG-PET) to assess the nature of neurogenic tumors. Eur J Surg Oncol 2003; 29(6):536-541. 93. McCarville MB, Spunt SL, Skapek SX, Pappo AS. Synovial sarcoma in pediatric patients. AJR Am J Roentgenol 2002; 179(3):797-801. 94. Murphey MD, Gibson MS, Jennings BT, Crespo-Rodriguez AM, Fanburg-Smith J, Gajewski DA. From the archives of the AFIP: Imaging of synovial sarcoma with radiologic-pathologic correlation. Radiographics 2006; 26(5):1543-1565. 95. Horowitz AL, Resnick D, Watson RC. The roentgen features of synovial sarcomas. Clin Radiol 1973; 24(4):481-484. 96. Rice S, Stewart CA. Synovial sarcoma seen on bone scan. Clin Nucl Med 1990; 15(6):445-446. 97. Tateishi U, Hasegawa T, Beppu Y, Satake M, Moriyama N. Synovial sarcoma of the soft tissues: prognostic significance of imaging features. J Comput Assist Tomogr 2004; 28(1):140-148. 98. Morton MJ, Berquist TH, McLeod RA, Unni KK, Sim FH. MR imaging of synovial sarcoma. AJR Am J Roentgenol 1991; 156(2):337-340. 99. Jones BC, Sundaram M, Kransdorf MJ. Synovial sarcoma: MR imaging findings in 34 patients. AJR Am J Roentgenol 1993; 161(4):827-830.
452
M.J. Shelly et al.
100. Jemal A, Murray T, Ward E, Samuels A, Tiwari RC, Ghafoor A et al. Cancer statistics, 2005. CA Cancer J Clin 2005; 55(1):10-30. 101. Murphy GF, Mihm MC. The skin. In: Cotran RS, Kumar V, Collins T, editors. Robbins Pathologic Basis of Disease. 6th ed. Philadelphia: WB Saunders; 1999. 1177-1186. 102. Mihm MC, Jr., Clark WH, Jr., From L. The clinical diagnosis, classification and histogenetic concepts of the early stages of cutaneous malignant melanomas. N Engl J Med 1971; 284(19):1078-1082. 103. Breslow A. Thickness, cross-sectional areas and depth of invasion in the prognosis of cutaneous melanoma. Ann Surg 1970; 172(5):902-908. 104. Clark WH, Jr., Elder DE, Guerry D, Braitman LE, Trock BJ, Schultz D et al. Model predicting survival in stage I melanoma based on tumor progression. J Natl Cancer Inst 1989; 81(24):1893-1904. 105. Essner R, Belhocine T, Scott AM, Even-Sapir E. Novel imaging techniques in melanoma. Surg Oncol Clin N Am 2006; 15(2):253-283. 106. Morton DL, Thompson JF, Essner R, Elashoff R, Stern SL, Nieweg OE et al. Validation of the accuracy of intraoperative lymphatic mapping and sentinel lymphadenectomy for earlystage melanoma: a multicenter trial. Multicenter Selective Lymphadenectomy Trial Group. Ann Surg 1999; 230(4):453-463. 107. Escott EJ. A variety of appearances of malignant melanoma in the head: a review. Radiographics 2001; 21(3):625-639. 108. Cochran AJ, Balda BR, Starz H, Bachter D, Krag DN, Cruse CW et al. The Augsburg Consensus. Techniques of lymphatic mapping, sentinel lymphadenectomy, and completion lymphadenectomy in cutaneous malignancies. Cancer 2000; 89(2):236-241. 109. Gershenwald JE, Thompson W, Mansfield PF, Lee JE, Colome MI, Tseng CH et al. Multiinstitutional melanoma lymphatic mapping experience: the prognostic value of sentinel lymph node status in 612 stage I or II melanoma patients. J Clin Oncol 1999; 17(3):976-983. 110. Schoder H, Larson SM, Yeung HW. PET/CT in oncology: integration into clinical management of lymphoma, melanoma, and gastrointestinal malignancies. J Nucl Med 2004; 45 Suppl 1:72S-81S. 111. Schwimmer J, Essner R, Patel A, Jahan SA, Shepherd JE, Park K et al. A review of the literature for whole-body FDG PET in the management of patients with melanoma. Q J Nucl Med 2000; 44(2):153-167. 112. Schillaci O. Hybrid SPECT/CT: a new era for SPECT imaging? Eur J Nucl Med Mol Imaging 2005; 32(5):521-524. 113. Essner R, Hsueh EC, Haigh PI, Glass EC, Huynh Y, Daghighian F. Application of an [(18)F] fluorodeoxyglucose-sensitive probe for the intraoperative detection of malignancy. J Surg Res 2001; 96(1):120-126.
17
Reticuloendothelium Malignancy: Current Role of Imaging Sunit Sebastian1 MD and Brian C. Lucey2 MD
Key Points ●
●
● ● ●
Imaging plays a vital role in the diagnosis and staging of lymphoma, thereby influencing patient management and outcome. Older modalities like lymphangiography have been superseded by newer crosssectional imaging modalities. CT is probably the most widely used modality for this purpose. MRI is as useful as CT in the initial work-up of the patient. PET/CT has the ability to detect residual disease post-treatment, and is superior to both CT and MRI in this regard.
Introduction The reticuloendothelial system (RES), also known as the mononuclear phagocytic system (MPS), is comprised of lymphoid organs including the bone marrow, liver, spleen, lymph nodes, thymus, microglia of the brain, tonsils as well as MALT (mucosa associated lymphoid tissue), BALT (bronchus-associated lymphoid tissue) and GALT (gut-associated lymphoid tissue). The T-lymphocytes make up 75 percent and the B-lymphocytes constitute 25 percent of the total lymphocytes. The spleen contains both B- and T-lymphocytes. Lymphoid tissue associated with mucosa, bronchus and gut are termed MALT, BALT and GALT, respectively. Tonsils respond to antigens by producing B-lymphocytes. Since the basic tenets of imaging the reticuloendothelial system essentially remain the same, this chapter will be confined to imaging of the lymph nodes and spleen. Other organ systems (such as liver, thymus etc) will be dealt with elsewhere in this book. 1 Department of Radiology, Emory University School of Medicine, 1364 Clifton Road, Atlanta, GA 30324 2 Corresponding author: Brian C. Lucey Department of Radiology, VA Medical Center, Boston, MA, e-mail:
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
455
456
S. Sebastian and B.C. Lucey
The merits and disadvantages of all imaging modalities will be discussed. Moreover, the increasingly important role of newer imaging techniques such as PET-CT will also be emphasized, especially in the evaluation of post-treatment residual disease.
Lymph Nodes and Spleen: Imaging Armamentarium Sonography Sonography is a straightforward and convenient technique to investigate lymphadenopathy [1]. Superficial lymphadenopathy is a common manifestation of lymphoma. The head and neck regions are the most common sites of involvement in lymphoma [2]. In clinical practice it is important to differentiate between lymphomatous and metastatic nodes. Lymphomatous nodes have a consistent pattern of involvement that provides a clue to their diagnosis. Submandibular, submental and deep cervical nodes are most commonly involved in lymphoma. Metastatic lymph nodes usually involve the submandibular and upper cervical regions. Posterior triangle involvement is infrequent in the metastatic disease process [3]. Two distinct features can be used to differentiate between lymphomatous and metastatic nodes [4]. The presence of distal enhancement in lymphoma is a significant and consistent differentiating feature from metastatic nodes. The second distinguishing feature is the presence of intranodal necrosis (both coagulation and cystic necrosis), which is more frequent in metastatic head and neck nodes, than in lymphomatous nodes [5]. Thus, sonography may prove to be a useful initial investigation [6]. However, the main disadvantages of sonography are the poor spatial resolution, its limited use in the thorax and deep retroperitoneum, and high operator dependency. These limitations may be overcome by Doppler sonography which offers functional imaging of the lymph node. Since feeding vessels determine tumor growth, color/power Doppler sonography (US) may be used to differentiate lymphoma from metastatic carcinoma [7-10]. Giovagnorio et al. [11] have described that vessels could be identified in all lymph nodes in patients with lymphoma. The majority of the lymph nodes demonstrated hilar vascularity because lymphoma arises within the lymph nodes and progresses in a centrifugal fashion. Other studies have also reported a similar central perfusion pattern of lymphomatous nodes [12, 13]. In contrast, metastatic nodes show the presence of peripheral subcapsular vessels which access lymph nodes through afferent lymphatic vessels and invade marginal sinuses. Recent reports suggest that intravenously administered microbubbles help the diagnosis of lymphadenopathy with accurate demonstration of vascular flow within a lymph node [14]. In a relatively small series of patients contrast-enhanced color power Doppler sonography with Levovist® was used to study the vascular patterns of lymph nodes with different types of lymphoma. The patterns of vessel distribution were classified as “central” when hilar vessels were seen with minimal or no rim vessels; a “peripheral” pattern was assigned when vessels were observed only
17 Reticuloendothelium Malignancy: Current Role of Imaging
457
in the periphery; “capsular and central” when vessels were apparent in the parenchyma of the node, as well as in the periphery. B-cell lymphoma patients were noted to have a central vascular pattern in the lymph nodes. A peripheral vascular pattern was observed in the lymph nodes of T-cell lymphoma patients [15]. Sonography is a quick and noninvasive method for detecting splenic involvement in lymphoma [16]. Lymphomatous nodules and hematogenous metastases may have similar characteristics, appearing isoechoic or hypoechoic to normal spleen on unenhanced images [17, 18]. The different sonographic patterns observed in Hodgkin’s Lymphoma are: diffuse involvement, focal small nodular lesions, focal large nodular lesions and bulky disease. High-grade lymphoma is usually manifested as large nodular or small nodular lesions. The diffuse pattern is seen predominantly in low-grade lymphoma [19]. Different morphological patterns are also seen in Non-Hodgkin’s Lymphoma (NHL) with sonography. Diffuse infiltration is more common in NHL [20]. On sonography, lymphoma may appear anechoic, thus mimicking a cyst. The shape, echogenicity of the lesion and mode of posterior echo are not specific enough characteristics to differentiate between splenic lymphomas and splenic cysts. However, the boundaries of the lesions are indistinct in splenic lymphomas and distinct in splenic cysts. Blood flow signals and vascular penetration are also seen exclusively in splenic lymphoma [21]. Lymphomatous lesions in the spleen appear as clear hypoechoic defects after contrast medium injection. Regularly deposited vessels may be first seen encircling and then entering the nodule, especially during the early phase of opacification. The tumor tissue usually shows a lesser degree of enhancement, compared to the surrounding normal splenic parenchyma [22].
Computed Tomography The potentially curative chemotherapeutic agents used for the treatment of lymphoma require accurate staging for maximal efficacy. Computed tomography (CT) has been the primary imaging technique used for staging and follow-up of patients with lymphoma for many years. In many institutions CT examination is the standard to map disease sites and estimate tumor burden. Common CT criteria to assess Hodgkin’s and Non-Hodgkin’s Lymphoma in the lymph nodes and spleen are organomegaly, abnormal contrast enhancement, or presence of an abnormal mass. Enlarged nodes are the most common finding on CT of lymphoma (Fig. 17.1). Although absolute numbers are difficult to define, cervical, thoracic and pelvic lymph nodes are generally considered to be enlarged if they are greater than 10 mm in size. Abdominal lymph nodes are considered to be abnormal if they are greater than 5 mm, and inguinal lymph nodes are enlarged if they are greater than 15 mm in size. The sensitivity of CT in evaluating nodal and extranodal disease ranges between 60 percent and 90 percent [23]. The anterior mediastinal, pretracheal and hilar nodal chains are the most common nodal chains involving
458
S. Sebastian and B.C. Lucey
Fig. 17.1 Axial CT images of a 56-year-old patient with a known diagnosis of non-Hodgkin’s lymphoma. (a) The lymphadenopathy has conglomerated to form a soft tissue mass compressing part of the small bowel with infiltration of the bowel wall. (b) This image shows a large soft tissue mass in the pelvis that represents enlarged lymph nodes. This mass displaces the bowel and the right external iliac artery
lymphoma in the chest [24]. In the thorax, Hodgkin’s disease tends to spread to contiguous nodal groups. The subcarinal, peridiaphragmatic, periesophageal and internal mammary nodes are involved in decreasing order of frequency. Isolated hilar lymph node enlargement is a relatively unusual finding in the setting of lymphoma. The presence of large mediastinal adenopathy has a higher risk of relapse, and a dual treatment strategy with chemotherapy and radiation therapy is warranted regardless of tumor grade [25]. In most cases the enlarged lymphomatous nodes are homogeneous, although necrotic nodes are not uncommon. Hopper, et al. [26] have observed that the presence of necrosis has no significant effect on the patient’s clinical response to treatment or ultimate survival. Necrotic nodes, when present, are most commonly associated with nodular sclerosing Hodgkin’s disease. However, nodal necrosis can be identified subsequent to chemotherapy or radiation therapy. Enhancing nodes may be observed, though infrequently in NHL. Lymph node calcification before treatment is fairly rare, but can be associated with aggressive Hodgkin’s or Non-Hodgkin’s Lymphoma [27]. In lymphoma patients irregular or eggshell calcifications are seen in lymph nodes post-treatment [28]. Staging of lymphoma is as follows: Single-station nodal disease is defined as stage 1; multiple nodes restricted to one body area are defined as stage 2, and disease on both sides of the diaphragm is defined as stage 3. Visceral involvement is regarded as stage 4 (Fig. 17.2). Accurate description of the extent of disease is important for radiation therapy planning [29]. The extrapleural space can be involved by nodal disease, especially in patients with NHL [29]. Obstruction of the pleural lymphatics by the tumor is often associated with pleural effusion. Extranodal involvement is more commonly observed in NHL [30]. In Hodgkin’s disease, extranodal invasion of adjacent tissue is seen in up to 15 percent of cases [3]. Another important role of CT in patients with lymphoma is to assess response to therapy, evaluate recurrence and monitor patients before and after treatment [32].
17 Reticuloendothelium Malignancy: Current Role of Imaging
459
Fig. 17.2 Axial CT images of a 32-year-old patient with a known diagnosis of non-Hodgkin’s lymphoma. The image shows nodular and infiltrative involvement of the lymphoma throughout the mesentery
Disease recurrence is common in the pericardial and internal mammary lymph nodes, since these are usually not included in the radiation field. CT of the chest, abdomen and pelvis has also been shown to be valuable in the follow-up of lymphoma. Neumann, et al. have reported that CT enabled detection of unsuspected active disease in 43 percent of patients with NHL thought to be in remission [33]. Primary splenic lymphoma is rare. The spleen represents a “nodal organ” in Hodgkin’s disease and an extranodal organ in NHL [31]. However, most of the primary splenic lymphomas tend to be NHLs (marginal zone cell lymphoma). Splenic involvement is usually secondary to generalized lymphoma. The most common finding is splenomegaly, but it may be absent in up to 30 percent of lymphoma patients. However, if the spleen is markedly enlarged in the presence of involvement of other sites, splenic lymphoma is more likely [34]. Staging laparotomy has shown that the spleen is infiltrated in about 30 percent to 40 percent of patients at presentation [35]. Splenic lesions in lymphoma could present as diffuse splenic enlargement, or a solitary mass and multifocal lesions (Figs. 17.3 and 17.4), with diffuse infiltration being the most common manifestation. If nodules are noted, they demonstrate low attenuation with reduced contrast material enhancement, compared with normal splenic tissue, and are infrequently larger than 1 cm in diameter [36]. However, caution should be exercised in interpreting CT scans obtained during the early phase of a bolus injection of contrast material, due to the
460
S. Sebastian and B.C. Lucey
Fig. 17.3 CT image of a 42-year-old male patient with non-Hodgkin’s lymphoma. There are multiple low attenuation lesions within the spleen
Fig. 17.4 Axial CT image in a 30-year-old patient showing a solitary hypodense mass in the spleen. This proved to be a Hodgkin’s lymphoma
17 Reticuloendothelium Malignancy: Current Role of Imaging
461
heterogeneous enhancement of the spleen that mimics tumor infiltration. Diffuse involvement of the spleen in Hodgkin’s disease could be nonspecific, and splenomegaly may be present in the absence of lymphoma, or the spleen could be normal in size in spite of tumor infiltration [37-39]. The near isotropic images produced by multi–detector row spiral computed tomography (CT) also enables determination of splenic volume, accurate depiction of polar arteries, the presence, number and size of all accessory spleens, and of focal parenchymal lesions [40]. Despite advances in CT technology, CT has its limitations. It may not depict tumor activity within a post-therapy residual mass. Inability to detect disease in normal-sized lymph nodes could be a cause for false negatives. Moreover, splenic and bone marrow infiltration are not well depicted on CT [41]. CT assessment of lymphoma is frequently based only on size criteria. For example, lymph nodes less than 1 cm in diameter are not considered abnormal by most current criteria. In addition, CT may not be able to differentiate between residual tumor masses and fibrosis on post-treatment follow-up imaging [42].
PET/PET-CT As discussed earlier, CT has inherent limitations in the detection and follow-up of lymphoma. PET with 18F-FDG can provide functional information based on the increased metabolic demands of tumor cells requiring adenosine triphosphate generated by glycolysis [43]. In addition to detection of tumor foci in the lymph nodes and spleen, PET imaging has the ability to differentiate between aggressive and low-grade lymphomas. Aggressive lymphomas tend to have a higher 18F-FDG uptake with an SUV of more than 10 [44, 45]. The median sensitivity and specificity reported for PET is 90.3 percent and 91.1 percent, respectively. The maximum joint sensitivity and specificity was 87.8 percent [46-49]. 18FDG PET is very useful in the identification of patients with and without splenic disease and is superior to CT for this purpose. Splenic lymphoma, either diffuse or focal, tends to have an 18 FDG uptake greater than hepatic uptake [18]. Reported accuracies of 18FDG PET and CT for evaluating the spleen are 100 percent and 57 percent, respectively [50]. In newly diagnosed Hodgkin’s disease, FDG can identify splenic involvement precisely and is significantly more sensitive and accurate than Ga-67 for this purpose [51]. However, PET imaging has numerous limitations such as absence of precise anatomic landmarks for accurate localization of lesions, inherent lack of specificity since 18F-FDG can be taken up by lymphomatous nodes and sites of active inflammation and physiologically by some organs and low-grade lymphomas may not demonstrate uptake or have a low 18FDG uptake [52, 53]. The advent of PET-CT has remarkably improved the accuracy in the diagnostic work-up of patients with lymphoma. PET-CT provides dual modality imaging, which combines the functional information provided by PET and the excellent anatomic resolution offered by CT. Several studies [54, 55] have noted the role of PET/ CT in the staging of lymphoma. PET/CT can aid in the differentiation between
462
S. Sebastian and B.C. Lucey
tumoral and physiologic or inflammatory pathologic uptake, thereby overcoming false positives resulting from using PET alone in the diagnosis of lymphoma. PET/ CT plays an important role in the staging of lymphoma since it can identify pathologic lymphadenopathy accurately [54, 56-58]. In a prospective study PET/CT proved to be superior compared with CT and PET alone in nodal evaluation and detection of extranodal disease [59]. There were some previous issues indicating that the use of oral and intravenous contrast for the CT part of PET-CT caused problems in CT-based attenuation correction [60, 61]. However, recent studies have demonstrated that intravenous contrast material, when used at normal concentrations, does not interfere with CT-based attenuation correction [62, 63]. However, the optimal protocol of PET/CT is not yet determined, and controversy remains regarding the acquisition of a full diagnostic CT with oral and intravenous contrast [55]. A major drawback of PET/CT is the concern regarding the radiation dose incurred by the patient. Whole-body mean effective dose from FDG is approximately 10.73 mSv [64]. A whole-body diagnostic CT scan significantly increases the effective dose to 19.262 mSv for a whole-body PET/CT scan. Hence, protocol optimization is vital for the evaluation of lymphoma with PET/CT, and must be tailored according to the patient in question.
Magnetic Resonance Imaging The introduction of fast MR imaging techniques has reduced imaging time substantially without compromising the quality of MR images. MR imaging is considered to be as diagnostic as CT for staging Hodgkin’s disease. The excellent soft tissue contrast and the lack of exposure to ionizing radiation are formidable advantages offered by MR imaging. Lymphoma is usually hypointense or nearly isointense on T1-weighted images, and hyperintense on T2-weighted images. Injection of contrast medium may improve detection of splenic lymphoma [65]. Lymphomatous tissue has an increased water content that returns a high signal on the STIR sequence, enabling detection [66, 67]. Detection of splenic lymphoma at MR imaging is difficult because normal splenic parenchyma and lymphomatous tissue may have similar signal intensity [37]. Lymphomatous nodules are hypo- or isointense on T1-weighted MR images, and hyperintense on T2-weighted images. These nodules do not enhance as much as the normal spleen after administration of gadopentetate dimeglumine. MRI has a good capability in distinguishing nodal and extranodal involvement, both in Hodgkin’s Lymphoma and in NHLs [68]. Detection of nodal involvement in normal sized lymph nodes and residual tumor activity after therapy are some issues that need consideration. However, newly developed lymphotropic contrast agents for MR imaging might be helpful to answer these questions in the future [69]. Detection of splenic involvement can alter management. Lymphomatous deposits have T1 and T2 signal intensities similar to those of normal splenic parenchyma. Gadoliniumenhanced sequences are more sensitive for the evaluation of splenic lymphoma.
17 Reticuloendothelium Malignancy: Current Role of Imaging
463
Diffuse involvement may be seen as large irregularly enhancing regions. Multifocal disease is also common and can be seen as multiple focal lesions that are hypointense relative to the enhancing splenic parenchyma [70, 71].
Lymphangiography Lymphangiography was used in the past to investigate lymphadenopathy and the staging of lymphoma. However, these indications are now evaluated using CT, PET/CT or MRI. Technologic advances leading to state-of-the-art CT and PET/CT scanning, coupled with the development of more effective chemotherapeutic regimens and the potential adverse effects of lymphangiography, have further led to its fallout in current clinical practice as a staging tool in patients with lymphoma. A study to determine the current value of lymphography in previously untreated patients with Hodgkin’s and Non-Hodgkin’s Lymphoma concluded that lymphographic findings did not significantly contribute to staging in these patients [72]. Detection of subtle nodal changes is no longer needed since the advent of potent chemotherapy and radiation therapy.
Interventional Radiology in Lymphoma Image-guided biopsy of abdominal or thoracic lymphoma plays an important role in patient management and can be performed either by sonographic or CT guidance. Endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) is feasible even in very small foci, when CT- or US-guided biopsy is not successful [73]. Fine needles (16 to 18 G) are just as likely as larger needles (20 to 24 G) to enable both determinations of tumor grade and treatment [74]. Flow cytometry is valuable in assessing immunophenotypic criteria that are helpful in the diagnosis of T-cell neoplasms. Features attributable for malignancy include loss or markedly dim expression of CD45; complete loss of one or more pan-T antigens and CD4/CD8 dual-positive or dual-negative expression [75].
Post-Treatment Follow-up Imaging The treatment of lymphoma usually involves the administration of chemotherapy until a complete clinical remission is attained. Approximately 85 percent of patients will respond satisfactorily to their primary chemotherapy [76]. It is important to identify patients early in the course of chemotherapy who do not attain complete remission to modify treatment strategies either with high-dose chemotherapy or adjuvant radiotherapy. Cheson’s criteria [77] are most commonly employed to follow up residual disease with CT scanning. The maximum short-axis dimension is most commonly
464
S. Sebastian and B.C. Lucey
used to measure residual disease following treatment [78]. However, a clear consensus has not been achieved regarding response criteria. The glaring limitation of using Cheson’s criteria is that disease involvement is based solely on size criteria. Shape, longitudinal/transverse diameter ratios and enhancement patterns are not used in this assessment. The growing role of tumor volumetrics to monitor tumor response following treatment promises to address the limitations posed by using size as the only criteria to measure residual disease [79]. The inability of CT to differentiate active lymphoma from necrosis and fibrosis is a significant limitation in the follow-up for residual disease or relapse post-treatment. 67-Gallium scintigraphy is a metabolic imaging technique to detect active tumor tissue; however, low spatial resolution and difficulty identifying residual abdominal masses are serious limitations [80]. MRI demonstrates high accuracy in the assessment of residual disease post-treatment if performed at least six months after the end of therapy, reaching the highest sensitivity and specificity values at 12-month follow-up. MRI can help in distinguishing fibrous from active residual masses in treated Hodgkin’s disease. Low signal intensity and low contrast enhancement are generally signs of inactive residues; homogeneous high signal intensity and high contrast enhancement are suggestive of active residual disease; heterogeneous signal intensity and heterogeneous contrast enhancement are indicative of partial remission or necrotic/inflammatory processes [81]. 18 F-FDG PET can distinguish between post-treatment fibrosis and viable tumor. FDG-PET/CT can improve re-staging in lymphoma leading to improved patient therapy and survival [82]. Hypermetabolic brown fat can lead to false positives in the assessment of tumoral activity of residual masses [83].
Summary and Conclusion Imaging has an important role in the diagnosis and staging of patients with lymphoma, with a major influence on both patient management and outcome. CT is probably the most widely used modality in patients with lymphoma bur MRI is probably as useful as CT in the initial work-up of the patient. Finally, PET/CT can detect residual disease in patients post-treatment better than both CT and MRI and is playing an increasing role in the imaging of patients with lymphoma.
References 1. Vassallo P, Wernecke K, Roos N, Peters PE. Differentiation of benign from malignant superficial lymphadenopathy: the role of high-resolution US. Radiology 1992; 183:215-220. 2. DePena CA, Van Tassel P, Lee YY. Lymphoma of the head and neck. Radiol Clin North Am 1990; 28:723-743. 3. Som PM. Lymph nodes of the neck. Radiology 1987; 165:593-600.
17 Reticuloendothelium Malignancy: Current Role of Imaging
465
4. Bruneton JN NF. Ultrasonography of the Neck. In: Springer-Verlag, 1987; 81-92. 5. Sakai F, Sone S, Kiyono K, et al. Computed tomography of neck lymph nodes involved with malignant lymphoma: comparison with ultrasound. Radiat Med 1991; 9:203-208. 6. Baatenburg de Jong RJ, Rongen RJ, Lameris JS, Harthoorn M, Verwoerd CD, Knegt P. Metastatic neck disease. Palpation vs ultrasound examination. Arch Otolaryngol Head Neck Surg 1989; 115:689-690. 7. Ho SS, Ahuja AT, Kew J, Metreweli C. Differentiation of lymphadenopathy in different forms of carcinoma with Doppler sonography. Clin Radiol 2000; 55:627-631. 8. Na DG, Lim HK, Byun HS, Kim HD, Ko YH, Baek JH. Differential diagnosis of cervical lymphadenopathy: usefulness of color Doppler sonography. AJR Am J Roentgenol 1997; 168:1311-1316. 9. Schroder RJ, Maurer J, Hidajat N, et al. [Signal-enhanced color-coded duplex sonography of reactively and metastatically enlarged lymph nodes]. Rofo 1998; 168:57-63. 10. Tschammler A, Hahn D. Multivariate analysis of the adjustment of the colour duplex unit for the differential diagnosis of lymph node alterations. Eur Radiol 1999; 9:1445-1450. 11. Giovagnorio F, Galluzzo M, Andreoli C, De CM, David V. Color Doppler sonography in the evaluation of superficial lymphomatous lymph nodes. J Ultrasound Med 2002; 21:403-408. 12. Schulte-Altedorneburg G, Demharter J, Linne R, Droste DW, Bohndorf K, Bucklein W. Does ultrasound contrast agent improve the diagnostic value of colour and power Doppler sonography in superficial lymph node enlargement? Eur J Radiol 2003; 48:252-257. 13. Shirakawa T, Miyamoto Y, Yamagishi J, Fukuda K, Tada S. Color/power Doppler sonographic differential diagnosis of superficial lymphadenopathy: metastasis, malignant lymphoma, and benign process. J Ultrasound Med 2001; 20:525-532. 14. Barrett T, Choyke PL, Kobayashi H. Imaging of the lymphatic system: new horizons. Contrast Media Mol Imaging 2006; 1:230-245. 15. Nakase K, Yamamoto K, Hiasa A, Tawara I, Yamaguchi M, Shiku H. Contrast-enhanced ultrasound examination of lymph nodes in different types of lymphoma. Cancer Detect Prev 2006; 30:188-191. 16. Munker R, Stengel A, Stabler A, Hiller E, Brehm G. Diagnostic accuracy of ultrasound and computed tomography in the staging of Hodgkin’s disease. Verification by laparotomy in 100 cases. Cancer 1995; 76:1460-1466. 17. Goerg C, Schwerk WB, Goerg K. Sonography of focal lesions of the spleen. AJR Am J Roentgenol 1991; 156:949-953. 18. Robertson F, Leander P, Ekberg O. Radiology of the spleen. Eur Radiol 2001; 11:80-95. 19. Gorg C, Weide R, Schwerk WB. Malignant splenic lymphoma: sonographic patterns, diagnosis and follow-up. Clin Radiol 1997; 52:535-540. 20. Gorg C, Weide R, Schwerk WB. [Ultrasound involvement of the spleen in non-Hodgkin’s lymphomas]. Ultraschall Med 1995; 16:104-108. 21. Ishida H, Konno K, Ishida J, et al. Splenic lymphoma: differentiation from splenic cyst with ultrasonography. Abdom Imaging 2001; 26:529-532. 22. Catalano O, Sandomenico F, Matarazzo I, Siani A. Contrast-enhanced sonography of the spleen. AJR Am J Roentgenol 2005; 184:1150-1156. 23. Lu P. Staging and classification of lymphoma. Semin Nucl Med 2005; 35:160-164. 24. North LB, Libshitz HI, Lorigan JG. Thoracic lymphoma. Radiol Clin North Am 1990; 28:745-762. 25. Greene FL PD, Fleming ID. AJCC Cancer Staging Manual. In: Springer-Verlag, 2002; 91-241. 26. Hopper KD, Diehl LF, Cole BA, Lynch JC, Meilstrup JW, McCauslin MA. The significance of necrotic mediastinal lymph nodes on CT in patients with newly diagnosed Hodgkin disease. AJR Am J Roentgenol 1990; 155:267-270. 27. Apter S, Avigdor A, Gayer G, Portnoy O, Zissin R, Hertz M. Calcification in lymphoma occurring before therapy: CT features and clinical correlation. AJR Am J Roentgenol 2002; 178:935-938. 28. Sharma A, Fidias P, Hayman LA, Loomis SL, Taber KH, Aquino SL. Patterns of lymphadenopathy in thoracic malignancies. Radiographics 2004; 24:419-434.
466
S. Sebastian and B.C. Lucey
29. Aquino SL, Chen MY, Kuo WT, Chiles C. The CT appearance of pleural and extrapleural disease in lymphoma. Clin Radiol 1999; 54:647-650. 30. Byun JH, Ha HK, Kim AY, et al. CT findings in peripheral T-cell lymphoma involving the gastrointestinal tract. Radiology 2003; 227:59-67. 31. Guermazi A, Brice P, de Kerviler EE, et al. Extranodal Hodgkin disease: spectrum of disease. Radiographics 2001; 21:161-179. 32. Lewis ER, Caskey CI, Fishman EK. Lymphoma of the lung: CT findings in 31 patients. AJR Am J Roentgenol 1991; 156:711-714. 33. Neumann CH, Robert NJ, Canellos G, Rosenthal D. Computed tomography of the abdomen and pelvis in Non-Hodgkin’s Lymphoma. J Comput Assist Tomogr 1983; 7:846-850. 34. Rademaker J. Hodgkin’s and Non-Hodgkin’s Lymphomas. Radiol Clin North Am 2007; 45:69-83. 35. Sandrasegaran K, Robinson PJ, Selby P. Staging of lymphoma in adults. Clin Radiol 1994; 49:149-161. 36. Fishman EK, Kuhlman JE, Jones RJ. CT of lymphoma: spectrum of disease. Radiographics 1991; 11:647-669. 37. Shirkhoda A, Ros PR, Farah J, Staab EV. Lymphoma of the solid abdominal viscera. Radiol Clin North Am 1990; 28:785-799. 38. Ahmann DL, Kiely JM, Harrison EG, Jr., Payne WS. Malignant lymphoma of the spleen. A review of 49 cases in which the diagnosis was made at splenectomy. Cancer 1966; 19:461-469. 39. Giovagnoni A, Giorgi C, Goteri G. Tumors of the spleen. Cancer Imaging 2005; 5:73-77. 40. Napoli A, Catalano C, Silecchia G, et al. Laparoscopic splenectomy: multi-detector row CT for preoperative evaluation. Radiology 2004; 232:361-367. 41. Vinnicombe SJ, Reznek RH. Computerised tomography in the staging of Hodgkin’s disease and Non-Hodgkin’s Lymphoma. Eur J Nucl Med Mol Imaging 2003; 30 Suppl 1:S42-55. 42. Dorfman RE, Alpern MB, Gross BH, Sandler MA. Upper abdominal lymph nodes: criteria for normal size determined with CT. Radiology 1991; 180:319-322. 43. Warburg O. On the origin of cancer cells. Science 1956; 123:309-314. 44. Rodriguez M, Rehn S, Ahlstrom H, Sundstrom C, Glimelius B. Predicting malignancy grade with PET in non-Hodgkin’s lymphoma. J Nucl Med 1995; 36:1790-1796. 45. Schoder H, Noy A, Gonen M, et al. Intensity of 18fluorodeoxyglucose uptake in positron emission tomography distinguishes between indolent and aggressive non-Hodgkin’s lymphoma. J Clin Oncol 2005; 23:4643-4651. 46. Isasi CR, Lu P, Blaufox MD. A meta-analysis of 18F-2-deoxy-2-fluoro-D-glucose positron emission tomography in the staging and restaging of patients with lymphoma. Cancer 2005; 104:1066-1074. 47. Lang O, Bihl H, Hultenschmidt B, Sautter-Bihl ML. Clinical relevance of positron emission tomography (PET) in treatment control and relapse of Hodgkin’s disease. Strahlenther Onkol 2001; 177:138-144. 48. Stumpe KD, Urbinelli M, Steinert HC, Glanzmann C, Buck A, von Schulthess GK. Wholebody positron emission tomography using fluorodeoxyglucose for staging of lymphoma: effectiveness and comparison with computed tomography. Eur J Nucl Med 1998; 25:721-728. 49. Mikosch P, Gallowitsch HJ, Zinke-Cerwenka W, et al. Accuracy of whole-body 18F-FDP-PET for restaging malignant lymphoma. Acta Med Austriaca 2003; 30:41-47. 50. Rini JN, Leonidas JC, Tomas MB, Palestro CJ. 18F-FDG PET versus CT for evaluating the spleen during initial staging of lymphoma. J Nucl Med 2003; 44:1072-1074. 51. Rini JN, Manalili EY, Hoffman MA, et al. F-18 FDG versus Ga-67 for detecting splenic involvement in Hodgkin’s disease. Clin Nucl Med 2002; 27:572-577. 52. Shreve PD, Anzai Y, Wahl RL. Pitfalls in oncologic diagnosis with FDG PET imaging: physiologic and benign variants. Radiographics 1999; 19:61-77; quiz 150-151. 53. Barrington SF, O’Doherty MJ. Limitations of PET for imaging lymphoma. Eur J Nucl Med Mol Imaging 2003; 30 Suppl 1:S117-127. 54. Schaefer NG, Hany TF, Taverna C, et al. Non-Hodgkin’s Lymphoma and Hodgkin’s disease: coregistered FDG PET and CT at staging and restaging–do we need contrast-enhanced CT? Radiology 2004; 232:823-829.
17 Reticuloendothelium Malignancy: Current Role of Imaging
467
55. Raanani P, Shasha Y, Perry C, et al. Is CT scan still necessary for staging in Hodgkin and Non-Hodgkin’s Lymphoma patients in the PET/CT era? Ann Oncol 2006; 17:117-122. 56. Allen-Auerbach M, Quon A, Weber WA, et al. Comparison between 2-deoxy-2-[18F]fluoroD-glucose positron emission tomography and positron emission tomography/computed tomography hardware fusion for staging of patients with lymphoma. Mol Imaging Biol 2004; 6:411-416. 57. Tatsumi M, Cohade C, Nakamoto Y, Fishman EK, Wahl RL. Direct comparison of FDG PET and CT findings in patients with lymphoma: initial experience. Radiology 2005; 237:1038-1045. 58. Ell PJ. The contribution of PET/CT to improved patient management. Br J Radiol 2006; 79:32-36. 59. Hernandez-Maraver D, Hernandez-Navarro F, Gomez-Leon N, et al. Positron emission tomography/ computed tomography: diagnostic accuracy in lymphoma. Br J Haematol 2006; 135:293-302. 60. Antoch G, Freudenberg LS, Egelhof T, et al. Focal tracer uptake: a potential artifact in contrast-enhanced dual-modality PET/CT scans. J Nucl Med 2002; 43:1339-1342. 61. Antoch G, Jentzen W, Freudenberg LS, et al. Effect of oral contrast agents on computed tomography-based positron emission tomography attenuation correction in dual-modality positron emission tomography/computed tomography imaging. Invest Radiol 2003; 38:784-789. 62. Beyer T, Antoch G, Bockisch A, Stattaus J. Optimized intravenous contrast administration for diagnostic whole-body 18F-FDG PET/CT. J Nucl Med 2005; 46:429-435. 63. Yau YY, Chan WS, Tam YM, et al. Application of intravenous contrast in PET/CT: does it really introduce significant attenuation correction error? J Nucl Med 2005; 46:283-291. 64. Deloar HM, Fujiwara T, Shidahara M, et al. Estimation of absorbed dose for 2-[F-18]fluoro2-deoxy-D-glucose using whole-body positron emission tomography and magnetic resonance imaging. Eur J Nucl Med 1998; 25:565-574. 65. Nyman R, Rhen S, Ericsson A, et al. An attempt to characterize malignant lymphoma in spleen, liver and lymph nodes with magnetic resonance imaging. Acta Radiol 1987; 28:527-533. 66. Hoane BR, Shields AF, Porter BA, Borrow JW. Comparison of initial lymphoma staging using computed tomography (CT) and magnetic resonance (MR) imaging. Am J Hematol 1994; 47:100-105. 67. Hoane BR, Shields AF, Porter BA, Shulman HM. Detection of lymphomatous bone marrow involvement with magnetic resonance imaging. Blood 1991; 78:728-738. 68. Tesoro-Tess JD, Balzarini L, Ceglia E, Petrillo R, Santoro A, Musumeci R. Magnetic resonance imaging in the initial staging of Hodgkin’s disease and Non-Hodgkin’s Lymphoma. Eur J Radiol 1991; 12:81-90. 69. Gossmann A, Eich HT, Engert A, et al. CT and MR imaging in Hodgkin’s disease–present and future. Eur J Haematol Suppl 2005:83-89. 70. Rabushka LS, Kawashima A, Fishman EK. Imaging of the spleen: CT with supplemental MR examination. Radiographics 1994; 14:307-332. 71. Kishimoto K, Koyama T, Kigami Y, et al. Primary splenic malignant lymphoma associated with hepatitis C virus infection. Abdom Imaging 2001; 26:55-58. 72. North LB, Wallace S, Lindell MM, Jr., Jing BS, Fuller LM, Allen PK. Lymphography for staging lymphomas: is it still a useful procedure? AJR Am J Roentgenol 1993; 161:867-869. 73. Fritscher-Ravens A, Mylonaki M, Pantes A, Topalidis T, Thonke F, Swain P. Endoscopic ultrasound-guided biopsy for the diagnosis of focal lesions of the spleen. Am J Gastroenterol 2003; 98:1022-1027. 74. Silverman SG, Lee BY, Mueller PR, Cibas ES, Seltzer SE. Impact of positive findings at image-guided biopsy of lymphoma on patient care: evaluation of clinical history, needle size, and pathologic findings on biopsy performance. Radiology 1994; 190:759-764. 75. Gorczyca W, Weisberger J, Liu Z, et al. An approach to diagnosis of T-cell lymphoproliferative disorders by flow cytometry. Cytometry 2002; 50:177-190. 76. Hoskin PJ. PET in lymphoma: what are the oncologist’s needs? Eur J Nucl Med Mol Imaging 2003; 30 Suppl 1:S37-41.
468
S. Sebastian and B.C. Lucey
77. Cheson BD, Horning SJ, Coiffier B, et al. Report of an international workshop to standardize response criteria for non-Hodgkin’s lymphomas. NCI Sponsored International Working Group. J Clin Oncol 1999; 17:1244. 78. Grillo-Lopez AJ, Cheson BD, Horning SJ, et al. Response criteria for NHL: importance of ‘normal’ lymph node size and correlations with response rates. Ann Oncol 2000; 11:399-408. 79. Prasad SR, Jhaveri KS, Saini S, Hahn PF, Halpern EF, Sumner JE. CT tumor measurement for therapeutic response assessment: comparison of unidimensional, bidimensional, and volumetric techniques initial observations. Radiology 2002; 225:416-419. 80. Zinzani PL, Zompatori M, Bendandi M, et al. Monitoring bulky mediastinal disease with gallium-67, CT-scan and magnetic resonance imaging in Hodgkin’s disease and high-grade nonHodgkin’s lymphoma. Leuk Lymphoma 1996; 22:131-135. 81. Di Cesare E, Cerone G, Enrici RM, Tombolini V, Anselmo P, Masciocchi C. MRI characterization of residual mediastinal masses in Hodgkin’s disease: long-term follow-up. Magn Reson Imaging 2004; 22:31-38. 82. Freudenberg LS, Antoch G, Schutt P, et al. FDG-PET/CT in re-staging of patients with lymphoma. Eur J Nucl Med Mol Imaging 2004; 31:325-329. 83. Castellucci P, Nanni C, Farsad M, et al. Potential pitfalls of 18F-FDG PET in a large series of patients treated for malignant lymphoma: prevalence and scan interpretation. Nucl Med Commun 2005; 26:689-694.
18
Pediatric Malignancies: Synopsis of Current Imaging Techniques Sabah Servaes1, Monica Epelman2, Avrum Pollock3, and Karuna Shekdar4
Key Points ●
●
●
●
1
PET/CT has proven utility in lymphoma, but studies are ongoing regarding its utility with pediatric neoplasms. CT is generally used initially with PET/CT for follow-up. Whole-body MRI is a promising technique to evaluate pathology such as metastases without using ionizing radiation with limits regarding specificity and small lesions. New MRI techniques are being developed and older techniques are being applied in new ways to improve the evaluation of pediatric neoplasms. MRI can also provide biochemical assessment of lesions. Chest CT is required for the staging of sarcomas. Metastases can be evaluated with bone scan and whole-body MRI, and PET/CT is currently under evaluation.
Introduction
The imaging evaluation of malignancies is directed towards the assessment of size, location and characterization of the neoplasm. Imaging children necessitates additional attention to the dose of radiation, given the radiosensitivity and the expected longevity of children. This chapter will present some of the latest technologies used to image pediatric malignancies, as well as methods to evaluate the most common pediatric neoplasms. 1 Department of Radiology, The Children’s Hospital of Philadelphia, 34th Street and Civic Center Blvd., Philadelphia, PA 19104,
[email protected] 2 Department of Radiology, The Children’s Hospital of Philadelphia, 34th Street and Civic Center Blvd., Philadelphia, PA 19104,
[email protected] 3 Director of Residency and Fellowship Programs, Department of Radiology, Division of Neuroradiology, The Children’s Hospital of Philadelphia, 34th Street and Civic Center Blvd., Philadelphia, PA 19104-4399,
[email protected] 4 Clinical Assistant – Neuroradiology, Hosp. of Univ. of Pennsylvania,, 219 Dulles Bldg., 3400 Spruce St., Philadelphia, PA 19104-4283,
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
469
470
2 2.1
S. Servaes et al.
Overview of Latest Imaging Technologies PET/CT
Staging of tumors, as well as response to therapy, is typically assessed by the size of the tumor and nodal location utilizing radiography, ultrasound, MRI and especially CT. With PET functional evaluation increases the diagnostic capabilities. Initially, PET and CT were always performed separately with interpreter comparison or software which retrospectively fused the images. Since 1998 PET/CT machines have minimized the problems which may arise from separately acquired images, such as misregistration due to differences in patient positioning [1]. PET enables clinicians to evaluate the presence of metabolically active tissue and determine the amount of metabolic activity by measuring the ratio of tumor activity concentration to the rest of the body (standardized uptake value or SUV). These measurements help to distinguish benign from malignant tumors, to evaluate the stage of a neoplasm, to evaluate response to therapy, to determine the optimal site for biopsy and to distinguish scar from residual neoplastic tissue following therapy. Fluorine-18 Fluorodeoxyglucose (18FDG)-PET is the radiopharmaceutical most commonly used clinically. 18FDG is a glucose analogue which localizes to sites where there is active metabolism and increased glucose uptake, such as malignant neoplasms. Glucose and fluordeoxyglucose are transported into cells and subsequently phosphorylated. Metabolism of fluordeoxyglucose is very slow at this point and, therefore, accumulates within the cell at a rate proportional to the rate of glucose metabolism. The one area of exception is in the liver, where fluorodeoxyglucose is readily dephosphorylated. Blood flow is another important physiologic feature which can be evaluated by measuring the accumulation of a freely diffusible tracer (such as oxygen-15-water), or by detecting markers of angiogenesis with isotop- labeled monoclonal antibodies. Other radiopharmaceuticals are currently being investigated for clinical utility [2]. In addition to increased uptake in malignant neoplasms, increased uptake may occur in normal tissues and benign entities that must be recognized. The gray matter of brain parenchyma has high 18FDG uptake because of its high glucose metabolism and, therefore, there is limited clinical benefit for intra-axial malignant disease, although facial and neck abnormalities may be detected. Liver, spleen and bone marrow also show homogeneous low-grade uptake, but malignant processes can still be seen within these tissues. Myocardial uptake is variable and fasting prior to an exam is helpful because the heart then uses fatty acids for metabolism. However, distinguishing myocardial activity from lymph nodes may be challenging. Normal lymphoid tissue in Waldeyer’s ring and pediatric thymic tissue have moderate uptake. Following chemotherapy in children and young adults, thymic rebound also demonstrates increased 18FDG uptake. Within the gastrointestinal tract activity is typically low in the esophagus, but benign entities such as esophagitis can increase 18FDG uptake. Similarly, gastritis can cause increased 18FDG uptake in the stomach. Although there is minimal uptake in the small bowel, the colon and particularly the cecum demonstrate avid uptake of 18FDG. 18FDG
18 Synopsis of Current Imaging Techniques
471
is excreted renally and, therefore, malignancies involving the urinary tract have limited clinical benefit. Bladder catheterization and appropriate placement of the bag which collects the urine are important in the evaluation of the pelvis. Children may also demonstrate increased 18FDG uptake in brown fat. To decrease skeletal muscle uptake patient motion, including speech, must be minimized and the temperature of the room adjusted to prevent shivering. Pathologic causes of increased 18FDG uptake include benign entities such as inflammation or infection. Bone which is healing from trauma or arthritis causes increased 18FDG uptake. Lymph nodes may demonstrate increased 18FDG uptake in tuberculosis and sarcoidosis. Similarly, recent infection, surgery or extravasation of radiotracer may cause the lymph nodes draining the involved region to demonstrate increased 18FDG uptake. Because the duration of time required for acquiring the PET images is more than 20 minutes and the CT can be obtained within a single breath-hold, the region about the diaphragm and nodules within the lungs are difficult to assess secondary to misregistration. Other technical factors which impact interpretation are the use of oral and intravenous contrast agents. High density oral contrast causes artifact, similar to metallic implants. Analogously, the arterial phase of intravenous enhancement causes more difficulty than venous phase. There is ongoing debate regarding the best technique, ranging from the absence of any contrast to those advocating for both oral and intravenous contrast [3-8]. Among the pediatric malignancies most often imaged with PET/CT are: Neuroblastoma [9], lymphoma [10-12] and soft tissue sarcomas [13, 14]. One advantage of PET/CT over CT alone is that PET provides information regarding the functionality of lymph nodes, allowing for appreciation of normal sized nodes that are involved in disease and the realization that some large lymph nodes are merely reactive. (Fig. 18.1) One of the pitfalls in the evaluation with PET is increased 18FDG uptake in nonneoplastic disorders [15]. Although there are issues regarding some of PET/CT’s findings and difficulties with interpretation, the initial studies indicate that PET/CT will play an important role in pediatric malignancy evaluation [16, 17].
3
Whole-Body MRI
Turbo short tau inversion recovery (STIR) whole-body MR imaging is a promising technique to study the entire body in a reasonable amount of time, and without utilizing ionizing radiation, a feature which is especially important in this population, in whom serial longitudinal follow-up studies are needed. This procedure has been used in pediatric oncology mainly for the detection of metastases or total tumor burden assessment [18-27]. Whole-body MRI has been shown to have better diagnostic accuracy in lesions that do not induce osteoblastic response, and is particularly helpful in the detection of bone marrow lesions, as compared with conventional skeletal scintigraphy and plain radiographs [18, 19, 25, 28]. However, whole-body MRI has other potential roles. Whole-body MRI may be particularly helpful in the
472
S. Servaes et al.
Fig. 18.1 (a) Coronal CT reformat of the chest, abdomen and pelvis in this girl with Hodgkin’s lymphoma and a mediastinal mass which is decreasing in size. (b) PET confirms that there is no uptake in the residual mediastinal mass
evaluation of multifocal avascular necrosis, especially following treatment with high-dose steroids. The detection of post-transplantation lymphoproliferative disorders, mainly after bone marrow transplantation, is another indication for wholebody MRI [29]. If follow-up whole-body MRI shows no response or tumor progression, prompt modifications in therapy become feasible due to this ability to evaluate initial treatment response [19]. Whole-body MRI can be performed on almost every body MRI system and may potentially replace other imaging studies, especially given recent improvements to decrease acquisition time. Once localizing scans are acquired, the entire body is imaged from the vertex to the heels by coronal turbo STIR sequences obtained in multiple overlapping stations. The field of view is adjusted to the size of the patient. The slice thickness is chosen to allow for complete anterior-to-posterior coverage [22, 25]. By using the turbo STIR sequences each set of coronal images may be acquired in less than four minutes, which can be further reduced to less than a minute if breath-hold techniques are applied [25]. Many parameters have not been standardized including, among others, the type of coil used (body versus surface
18 Synopsis of Current Imaging Techniques
473
coils), imaging techniques to reduce motion artifacts (e.g., respiratory triggering) and imaging plane selection [19]. The turbo STIR sequence is sensitive to soft tissue and skeletal pathology, due to the resultant added proton density (PD), T1 and T2 contrast with inherent suppression of signal from fat. Most pathologic tissues are proton-rich and have prolonged T1 relaxation and T2 decay times, resulting in high signal on turbo STIR images. This signal hyperintensity within proton-rich tumor metastasis becomes particularly noticeable against a background of hypointense suppressed fatty marrow, affording ease of image interpretation [22, 25, 26]. The turbo STIR sequence is sensitive to soft tissue and skeletal pathology, due to its added proton density, T1 and T2 contrast with inherent suppression of signal from fat without needing intravenous contrast administration. Most pathologic tissues are proton-rich and have prolonged T1 relaxation and T2 decay times, resulting in high signal on turbo STIR images. This signal hyperintensity within proton-rich metastastes becomes particularly noticeable against a background of hypointense suppressed fatty marrow, or isointense to muscle normal hematopoietic marrow, allowing ease of image interpretation [22, 25]. However, there are some limitations to whole-body MRI. STIR techniques are highly sensitive for detecting of pathologic lesions, but not specific for malignancy. Inflammatory, infectious, traumatic and necrotic changes, as well as benign lesions such as cysts and hemangiomas, cannot be differentiated from neoplastic lesions and appear hyperintense on STIR imaging. Whole-body STIR MR imaging may also be limited in the evaluation of oncologic patients following treatment, in that therapy-induced marrow changes such as edema, necrosis, fibrosis or red marrow hyperplasia (related to anemia or treatment with granulocyte-colony stimulating factor), cannot be differentiated from viable tumors. Although parenchymal lesions in the liver, spleen, kidneys, lungs and brain are readily detectable, smaller parenchymal lesions may be missed due to the low spatial resolution of the whole-body technique [19, 22]. PET/CT represents the greatest challenge to whole-body MRI. On the other hand, the major disadvantage of PET/CT is the increased radiation exposure sustained by young patients which may result in high cumulative radiation doses in children in whom oncologic imaging is performed repetitively and over relatively short time intervals. These may result in a higher cancer mortality risk, since children are more radiosensitive than adults [19, 30]. Therefore, whole-body MRI offers an alternative as a non-ionizing imaging modality for staging and monitoring tumor response [20].
4
CT Angiography – MRI Angiography
Although vascular invasion in childhood neoplasm is uncommon, vessel invasion may occur, particularly in Wilms’ tumors [31], soft tissue sarcomas and primary bone tumors [32]. Nonetheless, decision-making before surgical treatment in patients with
474
S. Servaes et al.
neoplasms, requires accurate delineation of the presence and level of vascular involvement. Unlike in adults, the role of CTA and MRA in pediatric oncologic imaging is yet to be defined. Many of the questions raised in conventional CT or MRI exams can be solved with Doppler ultrasound (US). However, CTA or MRA are indicated in cases when US results are inconclusive, confirmation is required before surgery to avoid complications or if clinical suspicion exists despite normal US results. The lesion’s appearance and its relationship to the vasculature is of use to the surgeons, particularly prior to resection of tumors encasing the renal arteries. Vascular mapping, by means of CTA or MRA, is extremely helpful for careful and safe surgical planning, and has a potential role in the preoperative planning of hepatoblastoma. The not-uncommon replaced right hepatic artery off the superior mesenteric artery is easily identified in the arterial phase, if right hepatectomy is required. While in the portal venous phase, the assessment of portal venous occlusion, cavernous transformation or portal vein tumor thrombus becomes more conspicuous. In addition, vascular mapping of the aortic vasculature, including accessory renal arteries, is of use in cases of Wilms’ tumor, especially if nephron sparing surgery is required, or in the surgical planning of subadventitial dissection of neuroblastoma and extensive retroperitoneal lymphadenopathy (Fig. 18.2) [20, 70, 71].
Fig. 18.2 Coronal CT reformat of the abdomen and pelvis in this 10-month-old with neuroblastoma
18 Synopsis of Current Imaging Techniques
5
475
MR Urography
With recent improvements in MRI equipment, including the development of faster gradients, stronger magnetic fields and parallel imaging, MR urography is now being used not only for the delineation of genitourinary anomalies, but also for the precise anatomic delineation of either intrinsic or extrinsic neoplasms affecting the genitourinary tract. In the pediatric population, and following initial US evaluation, MR urography, in combination with standard MR imaging, has the potential to reduce the need for intravenous pyelography, abdomino-pelvic CT and retrograde pyelography as a noninvasive, non-ionizing imaging option (Fig. 18.3) [33, 34].
6
MRI Spectroscopy
Proton MR Spectroscopy (MRS) is a powerful technique used to evaluate CNS neoplasms, which provides metabolite information that helps further characterize a tumor. Specifically, MRS is useful in differentiating tumors from normal tissue, and from other pathologies producing similar abnormal brain signal. It is useful in determining
Fig. 18.3 Coronal image from an MRU in this 5-year-old boy with prostatic rhabdomyosarcoma, which caused bilateral pelvicaliectasis, bilateral ureterectasis and a urinoma on the right
476
S. Servaes et al.
the tumor grade and in guiding the appropriate site for biopsy [35]. Another important use of MRS is in distinguishing radiation necrosis from recurrent neoplasm. MRS interrogates the chemical structure of areas in the brain. Spectroscopic data is obtained from a small volume of brain (approximately 1 square cm) to prevent the averaging of other normal metabolites within that volume. The sampling area or the cube of tissue which is evaluated is called a voxel. Single-voxel and multi-voxel MR spectroscopy techniques are utilized. Normal brain tissue has a characteristic neurochemical signature, depending on its location and the age of the patient [36]. MR spectroscopy can detect the various neurochemical spectral peaks. Among the neurochemicals which are most important in the analysis of brain tumors are N-acetyl acetate (NAA), choline compounds (Cho), creatine (Cr), lactate (Lac), myo-inositol (Myo) and amino acids (AA) [36]. NAA is considered a marker of normal functioning neurons. NAA is depleted in disorders which destroy the neuronal cells. Choline is an important component of the cell membrane. Choline levels are markers for cell membrane turnover which is increased in brain tumors and also in other disorders such as infection, inflammation and demyelination. Creatine peak results from creatine and phosphocreatine compounds which are high-energy buffers for ATP synthesis. Creatine is seen to be depressed in high-grade tumors which have an overwhelming energy requirement due to the hyper-proliferative tumor cells. In necrotic portions of the tumor, creatine is nearly depleted due to cell death. Lactate is a marker of anaerobic metabolism and, therefore, is elevated in highgrade or necrotic neoplasms, ischemia and in areas of injured brain tissue. MR spectroscopy is useful for grading astrocytic neoplasms based on the different magnitudes and ratios of the metabolites in the tumor [36]. Typically the NAA is decreased and choline is elevated (Fig. 18.4). The more the metabolite peaks vary
Fig. 18.4 (a) Localizing post-contrast T1 coronal MR image of the brain in a patient with glioblastoma multiforme. (b) The control spectrum demonstrates the normal spectral peaks
18 Synopsis of Current Imaging Techniques
477
Fig. 18.4 (continued) (c) The voxel centered within this patient’s tumor demonstrates loss of the normal spectral pattern within the tumor mass
from the normal values, the more aggressive the tumor. Increased lactate is associated with more aggressive neoplasms. Some brain tumors have characteristic spectral patterns, e.g. meningiomas, which have an alanine peak not seen in other tumors. In radiation injury, usually in the early stages, there is increased choline from destruction of the cellular membrane, but a normal NAA peak. Without the presence of active tumor there is usually global decrease in the peaks of NAA, choline and creatine in areas of treatment-related injury. Elevation of choline and reduction of NAA are suggestive of tumor recurrence. It is important, however, to realize that some tumors (such as juvenile pilocytic astrocytoma, the most benign and relatively common pediatric brain tumor) can demonstrate elevated choline and lactate with reduced NAA, a pattern which is seen with higher grade tumors [37]. Nonneoplastic disorders comprised of inflammatory processes, demyelination plaques, tuberculomas and encephalitis have MR spectroscopy features which are nearly identical to neoplasms. Hence, it is important that MR spectroscopy data be evaluated not in isolation, but in conjunction with anatomic MR imaging. MR spectroscopy findings may remain nonspecific. Further, spectroscopy data can be technically inadequate depending upon the size and the location of the neoplasm. Spectral data can be contaminated by adjacent CSF or fat in the calvarium/skull base. Voxel sampling may be inadequate if the tumor size is small or is extremely peripheral in location. MR spectroscopy is a valuable tool which supplements conventional MR imaging. MR spectroscopy imaging improves imaging of pediatric brain tumors because it gives a biochemical assessment of the tumor, and may distinguish residual or recurrent tumor from radiation necrosis [38].
478
7
S. Servaes et al.
MRI Perfusion
With the availability of high strength MRI gradients and ultrafast imaging sequences and, specifically, echo planar imaging, MRI perfusion imaging can be utilized for CNS neoplasms with a very small scan time penalty. It is possible to evaluate the dynamic changes which reflect tissue perfusion, relative cerebral blood volume (rCBV), cerebral blood flow (CBF) and mean transit time (MTT). Several methods can be utilized to derive perfusion parameters with MR imaging. Commonly, MRI perfusion imaging is performed using intravenously administered contrast medium (e.g., gadolinium-DTPA). The dynamic signal changes are due to passage of the bolus of contrast through the cerebral vessels into the capillary bed. The degree of signal loss is proportional to the concentration of gadolinium in the tissues of interest. This information is then utilized to generate perfusion maps of the area’s hemodynamic parameters [39]. Studies have shown that data obtained from perfusion MRI correlate with the degree of tumor angiogenesis and, hence, may be good predictors of the aggressive nature of the tumor. Therefore, tumors which have larger rCBV correlate to highergrade aggressive neoplasms [40]. Traditionally, contrast enhancement has been used as a crude marker of tumor vascularity. However, contrast enhancement reflects the breakdown in the blood-brain barrier and is not always representative of tumor vascularity. On the other hand, increased rCBV represents areas of increased vascularity in an aggressive neoplasm. Studies have shown that areas of increased rCBV may not always correspond to the contrast-enhancing portion of the tumor. Perfusion-weighted MR imaging is a more sensitive, noninvasive modality to predict tumor angiogenesis and vascularity [40]. Besides grading neoplasms, perfusion MRI can be used to guide stereotactic biopsies [41]. This capability is especially important in heterogeneous tumors. Prior to perfusion MRI, most of the brain tumor stereotactic biopsies would be directed to areas of contrast enhancement. Using perfusion MRI the areas within the tumor that demonstrate high rCBV can be identified as being the most aggressive, thereby guiding the interventionalist/surgeon to biopsy in this region. Serial perfusion-weighted MRI studies during therapy can noninvasively assess the changing tumor biology and dynamics. This function can potentially be used to determine optimal chemotherapy dosing [42]. Another important application of perfusion MR imaging is distinguishing treatment-induced brain injury from residual or recurrent neoplasm [41]. This difference in the pathophysiology of injured brain versus tumor tissue is exploited by perfusion MR imaging. With brain injury that results from damage to the endothelium, the end result is hypo-perfusion of the affected tissue, whereas growing tumor cells have an increased amount of blood supply for their proliferation and are associated with an increased rCBV. Studies have shown that these results, based on perfusion MRI, correlate well with identification of treatment-related brain injury and residual or recurrent tumors [41, 43]. MRI perfusion imaging is limited, especially in the younger pediatric population, by the need for adequate sized intravenous access to achieve an adequate contrast bolus.
18 Synopsis of Current Imaging Techniques
479
In summary, perfusion MR imaging is a valuable diagnostic tool to enhance anatomic MR imaging. It is clinically useful in accurately grading tumors, guiding stereotactic biopsies and helping distinguish between residual/progressive neoplasms from treatment-related brain injury. The potential use of MRI perfusion for mapping chemotherapeutic agent distribution and dosages are currently being assessed.
8
Diffusion MR Imaging
Diffusion-weighted MR imaging has been utilized for many years in neuroradiology, and recently has been promoted for abdominal imaging [44-46]. The studies suggest that alteration in water motion between compartments or diffusion, which is impacted by tumor response to treatment, can be detected with this MRI sequence and, thereby, monitor neoplastic processes. This technique, which measures the motion of water protons, can be effectively performed even with patient motion associated with breathing freely [44]. This technique has been used to distinguish a neoplastic process from a lymphatic malformation in a child [45]. Although more studies are needed, these results suggest that diffusion-weighted MR imaging will provide an alternative method to monitor response to therapy in oncology patients without using ionizing radiation.
9
9.1
Applications of Modern Imaging Techniques in Pediatric Malignancies Leukemia and Lymphoma
Leukemia is the most common childhood malignancy and lymphoma is the third most common. Together, these malignancies constitute nearly half of all childhood cancer. Approximately one-third of all childhood malignancy is due to leukemia, primarily acute leukemia. Acute lymphoblastic leukemia (ALL) peaks at two to three years of age, and acute myeloid leukemia (AML) peaks during the first two years of life and again during adolescence. Radiation exposure is one of the important risk factors associated with the development of leukemia [47, 48]. Lymphoma constitutes 10 percent to 12 percent of all childhood cancer. Lymphoma is classified into Non-Hodgkin’s Lymphoma and Hodgkin’s Lymphoma [49]. Hodgkin’s Lymphoma is the most common lymphoma, and, because of the high frequency in adolescence, has a greater incidence than brain neoplasms, which represent the second most common childhood malignancy. There are many imaging features which overlap between leukemia and lymphoma and, therefore, both are often suggested together in differentials provided with radiological interpretation. Lymphoma is the most common etiology for a mediastinal mass, which may be easily seen on chest radiographs (Fig. 18.5), and
480
S. Servaes et al.
Fig. 18.5 (a) PA and (b) lateral chest X-ray demonstrating a mediastinal mass in this patient with Hodgkin’s lymphoma
leukemia may also result in a mediastinal mass. CT is often used in the evaluation of a mediastinal mass. However, PET/CT plays an increasingly important role in imaging these patients [10-12, 50-56]. Gallium scans may provide similar information, but not with the same anatomic localization as PET/CT, and some studies indicate that gallium scanning is inferior to PET/CT [57-60]. Whole-body MRI has also been shown to be of value [61], providing evaluation without the use of ionizing radiation, which is particularly important in the pediatric population [62]. Although MRI offers the advantage of providing different signal with differing tissue characteristics, determination of lymph node pathology remains based upon size criteria, as with CT. Osseous abnormalities are also found with leukemia and lymphoma. On radiographs, metaphyseal lucencies are referred to as leukemic lines. (Fig. 18.6) MRI is especially sensitive at detecting marrow infiltration of leukemia and lymphoma. Fluid-sensitive sequences such as T2-weighted or STIR demonstrate abnormally increased signal, whereas T1-weighted images demonstrate decreased signal secondary to the increased water content of tumors [61]. (Fig. 18.7) The pediatric survival rate for lymphoma is 75 percent in the United States [63]. This cure rate is dependent upon accurate staging, for which there is a promising future of continued improvement.
18 Synopsis of Current Imaging Techniques
481
Fig. 18.6 AP X-ray of the ankle in this child with ALL demonstrates metaphyseal lucency
10
Central Nervous System
The prevalence rate for all primary brain and central nervous system tumors was estimated to be 9.5 per 100,000 in the United States for the year 2000 [64]. As with any cancer, early detection is essential to achieving a positive outcome. Cross-sectional imaging, e.g., CT and MRI, is the mainstay in diagnosing central nervous system tumors. CT is most often the initial imaging study obtained in the evaluation of suspected brain tumors. CT scanning involves ionizing radiation and should be used judiciously in the pediatric population. CT is especially valuable when a rapid study is needed to assess for hemorrhage, ventricular size or mass effect. With its multiplanar capability and excellent high spatial resolution, MRI is a superior imaging modality in the diagnosis and characterization of brain tumors, as well as in the follow-up to assess response to treatment. (Fig. 18.8) The development of new pulse sequences and ultrafast sequences has allowed for high resolution images to be obtained in a short time, which is particularly critical in pediatric imaging because it may obviate the need for sedation. MRI is the modality of choice in imaging spinal cord tumors. CT scan is a useful tool for assessment of bony vertebral body tumors. However, evaluation of spinal canal extension and cord involvement is best assessed with MRI.
Fig. 18.7 AP X-ray of the bilateral femurs, demonstrating periosteal reaction on the right (a) secondary to a healing fracture and a normal left (b) in this patient with lymphoma. (c) The marrow infiltration is only appreciated with the MRI (coronal proton density)
18 Synopsis of Current Imaging Techniques
483
Fig. 18.8 (a) T1 post-contrast midline sagittal view of the brain of a child with cerebellar astrocyoma. Note the large cyst (asterisk) centered within in the cerebellum surrounded by a rind of rim enhancing tumor (arrowheads). (b) T1 post-contrast fat-saturated axial view of the brain of a child with choroid plexus papilloma. Note the intensely enhancing frond-like tumor centered with the region of the third ventricle, with resultant hydrocephalus. (c) T1 post-contrast fat-saturated sagittal view of the brain of a child with craniopharyngioma. Note the suprasellar cystic lesion with rim enhancement, extending cephalad into the region of the third ventricle. (d) T1 postcontrast midline sagittal view of the brain of a child with medulloblastoma of the posterior fossa
MR imaging is extensively used in the diagnosis and follow-up of pediatric patients with brain tumors. However, conventional MR imaging does not provide information about the tissue biochemistry or grading of malignancy. An additional shortcoming of MR imaging is incomplete estimation of the extent of the tumor
484
S. Servaes et al.
[65]. Frequently a stereotactic biopsy, an invasive procedure, must be performed to obtain this information. The advanced imaging techniques of MR spectroscopy and MRI perfusion with ADC mapping can provide valuable information about tumor extent, tumor biochemistry and hemodynamics. These are noninvasive imaging tools which supplement the important anatomic information obtained by conventional gadolinium-enhanced MR imaging.
11
Neuroblastoma
The proper diagnostic imaging work-up for neuroblastoma is in constant evolution. Neuroblastoma has a primary location in the adrenal glands or sympathetic chain. Distant metastatic lesions at the time of presentation are seen in approximately 70 percent of children with abdominal neuroblastoma, with the most commonly affected sites being the bone marrow, followed by the lymph nodes, the liver and skin [66]. Conventional radiographs and ultrasound are routinely obtained in children with a suspicious space-occupying lesion. Chest radiographs may reveal a posterior mediastinal mass. Calcification may be seen in at least 30 percent of cases [67]. Ultrasound may show the solid nature of the tumor, which is usually of mixed echogenicity. Contrast-enhanced CT provides further detail on tumor architecture, extent of local disease and distant metastasis. The typical appearance is that of a large suprarenal mass with calcifications. MRI is particularly helpful in determining extension into the spinal canal. Metaiodobenzylguanidine (MIBG) scan is highly specific and sensitive for evaluating bone and bone marrow disease as MIBG is taken up by most neuroblastomas, but not by normal bone [68]. Additional use of technetium 99 m methylene diphosphonate bone scan is advocated by many since it may virtually eliminate any false negative MIBG results [69]. Eventually, confirmation or exclusion of the diagnosis of neuroblastoma is made with examination of tissue obtained by bone marrow aspiration or biopsy of primary or secondary disease. Metastatic lesions in the skin and lymph nodes are typically amenable to open biopsy. In primary tumors fine needle aspiration may provide reliable diagnosis; however, small sample size generally precludes important immunohistochemical and cytogenetic analysis. FDG-PET and whole-body MRI appear to be promising techniques [20]. Whole-body MRI is currently under investigation (www.acrin.org).
12
Wilms’ Tumor
Ultrasonography (US) is frequently the first imaging modality obtained. Whenever Wilms’ tumor is suspected, the inferior vena cava (IVC) should be carefully scrutinized for tumor thrombosis, since unrecognized caval thrombosis may result in fatal
18 Synopsis of Current Imaging Techniques
485
Fig. 18.9 (a) Gray scale ultrasound of a 4 year-old girl with left Wilms’ tumor demonstrates tumor invasion (arrowhead) of the IVC. AO = Aorta. (b) Color Doppler demonstratestumor (arrowhead) within the IVC. (c) Coronal CT reformat depicts tumor within the IVC (arrowhead). (d) Axial post-contrast MRI demonstrates tumor in the renal vein and IVC (arrowhead) and in the left kidney (arrow)
pulmonary embolus while cross-clamping the IVC during nephrectomy (Fig. 18.9) [20, 70, 71]. CT scan is currently the standard modality for diagnosis. CT is comprehensive, readily available and fast and, therefore, can possibly be performed without sedation. With multidetector scanners reconstructions/reformations in coronal and sagittal planes can assist in surgical planning of nephron-sparing surgery. CT depicts features of the renal lesion, determines the anatomic extent of the tumor and assesses the presence of a normal contralateral kidney versus bilateral tumor involvement. The scan should be obtained during the portal phase of enhancement at 70 seconds following the administration of intravenous contrast. At this point the renal cortex and medulla are attenuating; however, Wilms’ tumors are usually hypo-attenuating to normal kidney parenchyma during this phase of
486
S. Servaes et al.
enhancement, allowing for effortless identification of even subcentimeter lesions (Fig. 18.9C) [72]. The role of chest CT for the work-up and later management of pulmonary metastases remains controversial [72-74]. Traditionally, chest involvement is considered positive, if pulmonary nodules are identified on chest radiographs. It is expected that the Children’s Oncology Group (COG) will list chest CT as the modality of choice for the evaluation of pulmonary metastasis. MRI is suggested for the follow-up of bilateral lesions and nephroblastomatosis, as MRI can identify nephrogenic rests as small as 4 mm [20, 75, 76]. This ability is of significant value because it may obviate the need to perform contralateral exploration. Hence, MRI has the potential to replace CT as the pre-surgical imaging modality of choice in the evaluation of Wilms’ tumor (Fig. 18.9D) [20, 72].
13
Sarcoma
Rhabdomyosarcoma, osteosarcoma and Ewing’s sarcoma are the most common sarcomas in children [9]. Rhabdomyosarcoma is the most common soft tissue sarcoma, with 350 new cases per year in the United States [77]. Osteosarcoma is the most common primary malignant bone tumor and Ewing’s sarcoma is the second [78]. Rhabdomyosarcomas are subcategorized into two groups: embryonal rhabdomyosarcoma typically found in children less than six years of age and which constitutes 80 percent, and the remaining 20 percent are alveolar rhabdomyosarcoma typically found in children over six years of age [9]. The incidence of osteosarcoma is 5.6 per million children in the United States [9]. Most osteosarcomas arise from the medullary cavity of the metaphyses of long bones, predominantly about the knee. Although Ewing’s sarcoma is the second most common primary malignant bone tumor, the incidence is 225 each year in people under 20 years of age in North America [78]. Ewing’s sarcoma occurs most frequently in flat bones or the diaphysis of long bones. Staging of rhabdomyosarcoma requires assessment of the primary tumor size, location and invasiveness, and is typically done with CT or MRI [77]. The evaluation for metastatic disease includes chest CT and another modality for marrow infiltration such as bone scan or whole-body MRI; PET/CT is not yet part of routine staging for rhabdomyosarcoma, but is currently being investigated [61, 77]. The initial and the critical diagnostic imaging study in the evaluation of an osseous abnormality is a plain radiograph obtained in two orthogonal planes. Radiographs are crucial in making the diagnosis of an osseous lesion. MRI is the other modality of choice (Fig. 18.10). One of the goals of MRI is to evaluate the regional nerves and vessels. Imaging the entire bone is of the utmost importance in osteosarcoma to determine if skip lesions are present. One limitation with MRI is the inability to distinguish between edema surrounding the tumor and the actual extent of tumor. MRI frequently reveals adjacent soft tissue mass as part of Ewing’s sarcoma. Chest CT is typically performed to evaluate for metastases in both
18 Synopsis of Current Imaging Techniques
487
Fig. 18.10 AP (A) and lateral (B) X-rays of the left femur demonstrate periosteal reaction and the aggressive appearance of this osteosarcoma. (C) MRI was performed to evaluate the extent of the lesion and for skip lesions (sagittal STIR)
osteosarcoma and Ewing’s sarcoma. Whole-body 99 mTechnetium bone scan, PET/ CT and whole-body MRI have been utilized to determine the presence of osseous metastases. A 2001 study comparing these techniques demonstrated PET/CT to be the most sensitive modality, followed by whole-body MRI [79].
Conclusion CT and MRI imaging provide crucial information in childhood cancers. Currently, research is ongoing to further define the role of PET/CT imaging in childhood malignancy although its role in imaging subjects with lymphoma has been established.
References 1. Berry JD, Cook GJ. Positron emission tomography in oncology. Br Med Bull 2006. 2. Juweid ME, Cheson BD. Positron-emission tomography and assessment of cancer therapy. N Engl J Med 2006;354:496-507. 3. Dizendorf EV, Treyer V, Von Schulthess GK, Hany TF. Application of oral contrast media in coregistered positron emission tomography-CT. AJR Am J Roentgenol 2002;179:477-81.
488
S. Servaes et al.
4. Rodriguez-Vigil B, Gomez-Leon N, Pinilla I, et al. PET/CT in lymphoma: prospective study of enhanced full-dose PET/CT versus unenhanced low-dose PET/CT. J Nucl Med 2006;47:1643-8. 5. Schaefer NG, Hany TF, Taverna C, et al. Non-Hodgkin’s Lymphoma and Hodgkin’s disease: co-registered FDG PET and CT at staging and restaging–do we need contrast-enhanced CT? Radiology 2004;232:823-9. 6. Antoch G, Freudenberg LS, Beyer T, Bockisch A, Debatin JF. To enhance or not to enhance? 18F-FDG and CT contrast agents in dual-modality 18F-FDG PET/CT. J Nucl Med 2004;45 Suppl 1:56S-65S. 7. Antoch G, Freudenberg LS, Stattaus J, et al. Whole-body positron emission tomography-CT: optimized CT using oral and IV contrast materials. AJR Am J Roentgenol 2002;179: 1555-60. 8. Antoch G, Kuehl H, Kanja J, et al. Dual-modality PET/CT scanning with negative oral contrast agent to avoid artifacts: introduction and evaluation. Radiology 2004;230:879-85. 9. Nanni C, Rubello D, Castellucci P, et al. 18F-FDG PET/CT fusion imaging in paediatric solid extracranial tumours. Biomed Pharmacother 2006;60:593-606. 10. Hernandez-Pampaloni M, Takalkar A, Yu JQ, Zhuang H, Alavi A. F-18 FDG-PET imaging and correlation with CT in staging and follow-up of pediatric lymphomas. Pediatr Radiol 2006;36:524-31. 11. Miller E, Metser U, Avrahami G, et al. Role of 18F-FDG PET/CT in staging and follow-up of lymphoma in pediatric and young adult patients. J Comput Assist Tomogr 2006;30:689-94. 12. Furth C, Denecke T, Steffen I, et al. Correlative imaging strategies implementing CT, MRI, and PET for staging of childhood Hodgkin disease. J Pediatr Hematol Oncol 2006;28: 501-12. 13. Arush MW, Israel O, Postovsky S, et al. Positron emission tomography/computed tomography with (18)fluoro-deoxyglucose in the detection of local recurrence and distant metastases of pediatric sarcoma. Pediatr Blood Cancer 2007;. 14. Ben Arush MW, Bar Shalom R, Postovsky S, et al. Assessing the use of FDG-PET in the detection of regional and metastatic nodes in alveolar rhabdomyosarcoma of extremities. J Pediatr Hematol Oncol 2006;28:440-5. 15. Kavanagh PV, Stevenson AW, Chen MY, Clark PB. Nonneoplastic diseases in the chest showing increased activity on FDG PET. AJR Am J Roentgenol 2004;183:1133-41. 16. Blodgett TM, Casagranda B, Townsend DW, Meltzer CC. Issues, controversies, and clinical utility of combined PET/CT imaging: what is the interpreting physician facing? AJR Am J Roentgenol 2005;184:S138-45. 17. Schoder H, Gonen M. Screening for Cancer with PET and PET/CT: Potential and Limitations. J Nucl Med 2007;48 Suppl 1:4S-18S. 18. Daldrup-Link HE, Franzius C, Link TM, et al. Whole-body MR imaging for detection of bone metastases in children and young adults: comparison with skeletal scintigraphy and FDG PET. AJR Am J Roentgenol 2001;177:229-36. 19. Goo HW, Yang DH, Ra YS, et al. Whole-body MRI of Langerhans cell histiocytosis: comparison with radiography and bone scintigraphy. Pediatr Radiol 2006;36:1019-31. 20. Hoffer FA. Magnetic resonance imaging of abdominal masses in the pediatric patient. Semin Ultrasound CT MR 2005;26:212-23. 21. Kellenberger CJ, Miller SF, Khan M, Gilday DL, Weitzman S, Babyn PS. Initial experience with FSE STIR whole-body MR imaging for staging lymphoma in children. Eur Radiol 2004;14:1829-41. 22. Kellenberger CJ, Epelman M, Miller SF, Babyn PS. Fast STIR whole-body MR imaging in children. Radiographics 2004;24:1317-30. 23. Laffan EE, O’Connor R, Ryan SP, V D. Whole-body magnetic resonance imaging: a useful additional sequence in paediatric imaging. Pediatr Radiol 2004;34:472-80. 24. Mazumdar A, Siegel MJ, V N, Luchtman-Jones L. Whole-body fast inversion recovery MR imaging of small cell neoplasms in pediatric patients: a pilot study. AJR Am J Roentgenol 2002;179:1261-6.
18 Synopsis of Current Imaging Techniques
489
25. Walker RE, Eustace SJ. Whole-body magnetic resonance imaging: techniques, clinical indications, and future applications. Semin Musculoskelet Radiol 2001;5:5-20. 26. Eustace SJ, Walker R, Blake M, Yucel EK. Whole-body MR imaging. Practical issues, clinical applications, and future directions. Magn Reson Imaging Clin N Am 1999;7:209-36. 27. Siegel MJ, Luker GG. Bone marrow imaging in children. Magn Reson Imaging Clin N Am 1996;4:771-96. 28. Mentzel HJ, Kentouche K, Sauner D, et al. Comparison of whole-body STIR-MRI and 99 mTc-methylene-diphosphonate scintigraphy in children with suspected multifocal bone lesions. Eur Radiol 2004;14:2297-302. 29. Levine DS, Navarro OM, Chaudry G, Doyle JJ, Blaser SI. Imaging the complications of bone marrow transplantation in children. Radiographics 2007;27:307-24. 30. Brenner D, Elliston C, Hall E, Berdon W. Estimated risks of radiation-induced fatal cancer from pediatric CT. AJR Am J Roentgenol 2001;176:289-96. 31. Lowe LH, Isuani BH, Heller RM, et al. Pediatric renal masses: Wilms’ tumor and beyond. Radiographics 2000;20:1585-603. 32. Panicek DM, Go SD, Healey JH, Leung DH, Brennan MF, Lewis JJ. Soft tissue sarcoma involving bone or neurovascular structures: MR imaging prognostic factors. Radiology 1997;205:871-5. 33. Riccabona M. (Paediatric) magnetic resonance urography: just fancy images or a new important diagnostic tool? Curr Opin Urol 2007;17:48-55. 34. Nolte-Ernsting CC, Staatz G, Tacke J, Gunther RW. MR urography today. Abdom Imaging 2003;28:191-209. 35. Dowling C, Bollen AW, Noworolski SM, et al. Preoperative proton MR spectroscopic imaging of brain tumors: correlation with histopathologic analysis of resection specimens. AJNR Am J Neuroradiol 2001;22:604-12. 36. Hunter JV, Wang ZJ. MR spectroscopy in pediatric neuroradiology. Magn Reson Imaging Clin N Am 2001;9:165,89, ix. 37. Lazareff JA, Gupta RK, Alger J. Variation of post-treatment H-MRSI choline intensity in pediatric gliomas. J Neurooncol 1999;41:291-8. 38. Tzika AA, Zurakowski D, Poussaint TY, et al. Proton magnetic spectroscopic imaging of the child’s brain: the response of tumors to treatment. Neuroradiology 2001;43:169-77. 39. Cha S, Lu S, Johnson G, Knopp EA. Dynamic susceptibility contrast MR imaging: correlation of signal intensity changes with cerebral blood volume measurements. J Magn Reson Imaging 2000;11:114-9. 40. Roberts HC, Roberts TP, Brasch RC, Dillon WP. Quantitative measurement of microvascular permeability in human brain tumors achieved using dynamic contrast-enhanced MR imaging: correlation with histologic grade. AJNR Am J Neuroradiol 2000;21:891-9. 41. Cha S, Knopp EA, Johnson G, Wetzel SG, Litt AW, Zagzag D. Intracranial mass lesions: dynamic contrast-enhanced susceptibility-weighted echo-planar perfusion MR imaging. Radiology 2002;223:11-29. 42. Cha S, Knopp EA, Johnson G, et al. Dynamic contrast-enhanced T2-weighted MR imaging of recurrent malignant gliomas treated with thalidomide and carboplatin. AJNR Am J Neuroradiol 2000;21:881-90. 43. Siegal T, Rubinstein R, Tzuk-Shina T, Gomori JM. Utility of relative cerebral blood volume mapping derived from perfusion magnetic resonance imaging in the routine follow up of brain tumors. J Neurosurg 1997;86:22-7. 44. Olsen OE, Sebire NJ. Apparent diffusion coefficient maps of pediatric mass lesions with freebreathing diffusion-weighted magnetic resonance: feasibility study. Acta Radiol 2006;47:198-204. 45. Humphries PD, Wynne CS, Sebire NJ, Olsen OE. Atypical abdominal paediatric lymphangiomatosis: diagnosis aided by diffusion-weighted MRI. Pediatr Radiol 2006;36:857-9. 46. Murtz P, Flacke S, Traber F, van den Brink JS, Gieseke J, Schild HH. Abdomen: diffusionweighted MR imaging with pulse-triggered single-shot sequences. Radiology 2002;224:258-64.
490
S. Servaes et al.
47. Schubauer-Berigan MK, Daniels RD, Fleming DA, et al. Risk of chronic myeloid and acute leukemia mortality after exposure to ionizing radiation among workers at four U.S. nuclear weapons facilities and a nuclear naval shipyard. Radiat Res 2007;167:222-32. 48. Shuryak I, Sachs RK, Hlatky L, Little MP, Hahnfeldt P, Brenner DJ. Radiation-induced leukemia at doses relevant to radiation therapy: modeling mechanisms and estimating risks. J Natl Cancer Inst 2006;98:1794-806. 49. Steliarova-Foucher E, Stiller C, Lacour B, Kaatsch P. International Classification of Childhood Cancer, third edition. Cancer 2005;103:1457-67. 50. Amthauer H, Furth C, Denecke T, et al. FDG-PET in 10 children with Non-Hodgkin’s Lymphoma: initial experience in staging and follow-up. Klin Padiatr 2005;217:327-33. 51. Collins CD. PET in lymphoma. Cancer Imaging 2006;6:S63-70. 52. Elstrom R, Guan L, Baker G, et al. Utility of FDG-PET scanning in lymphoma by WHO classification. Blood 2003;101:3875-6. 53. Friedberg JW, Fischman A, Neuberg D, et al. FDG-PET is superior to gallium scintigraphy in staging and more sensitive in the follow-up of patients with de novo Hodgkin lymphoma: a blinded comparison. Leuk Lymphoma 2004;45:85-92. 54. Guermazi A, Juweid ME. Commentary: PET poised to alter the current paradigm for response assessment of non-Hodgkin’s lymphoma. Br J Radiol 2006;79:365-7. 55. Hermann S, Wormanns D, Pixberg M, et al. Staging in childhood lymphoma: differences between FDG-PET and CT. Nuklearmedizin 2005;44:1-7. 56. Montravers F, McNamara D, Landman-Parker J, et al. (18) FDG in childhood lymphoma: clinical utility and impact on management. Eur J Nucl Med Mol Imaging 2002;29:1155-65. 57. Rini JN, Nunez R, Nichols K, et al. Coincidence-detection FDG-PET versus gallium in children and young adults with newly diagnosed Hodgkin’s disease. Pediatr Radiol 2005;35:169-78. 58. Willkomm P, Palmedo H, Grunwald F, Ruhlmann J, Biersack HJ. Functional imaging of Hodgkin’s disease with FDG-PET and gallium-67. Nuklearmedizin 1998;37:251-3. 59. Juweid ME. Utility of Positron Emission Tomography (PET) Scanning in Managing Patients with Hodgkin’s Lymphoma. Hematology Am Soc Hematol Educ Program 2006;:259-65. 60. Jhanwar YS, Straus DJ. The role of PET in lymphoma. J Nucl Med 2006;47:1326-34. 61. Kellenberger CJ, Epelman M, Miller SF, Babyn PS. Fast STIR whole-body MR imaging in children. Radiographics 2004;24:1317-30. 62. Frush DP, Donnelly LF, Rosen NS. Computed tomography and radiation risks: what pediatric health care providers should know. Pediatrics 2003;112:951-7. 63. Hsu SC, Metzger ML, Hudson MM, et al. Comparison of treatment outcomes of childhood Hodgkin’s lymphoma in two US centers and a center in Recife, Brazil. Pediatr Blood Cancer 2006. 64. 2005-2006 Statistical Report: Primary Brain Tumors in the United States Statistical Report, 1998-2002 (Years Data Collected). Illinois, 2005 (http://www.cbtrus.org/reports//2005-2006/ 2006report.pdf.). 65. Wang Z, Zimmerman RA, Sauter R. Proton MR spectroscopy of the brain: clinically useful information obtained in assessing CNS diseases in children. AJR Am J Roentgenol 1996;167:191-9. 66. Papaioannou G, McHugh K. Neuroblastoma in childhood: review and radiological findings. Cancer Imaging 2005;5:116-27. 67. Lonergan GJ, Schwab CM, Suarez ES, Carlson CL. Neuroblastoma, ganglioneuroblastoma, and ganglioneuroma: radiologic-pathologic correlation. Radiographics 2002;22:911-34. 68. Andrich MP, Shalaby-Rana E, Movassaghi N, Majd M. The role of 131 iodine-metaiodobenzylguanidine scanning in the correlative imaging of patients with neuroblastoma. Pediatrics 1996;97:246-50. 69. Gordon I, Peters AM, Gutman A, Morony S, Dicks-Mireaux C, Pritchard J. Skeletal assessment in neuroblastoma–the pitfalls of iodine-123-MIBG scans. J Nucl Med 1990;31:129-34. 70. Ritchey ML, Kelalis PP, Haase GM, Shochat SJ, Green DM, D’Angio G. Preoperative therapy for intracaval and atrial extension of Wilms tumor. Cancer 1993;71:4104-10.
18 Synopsis of Current Imaging Techniques
491
71. Ritchey ML, Shamberger RC, Haase G, Horwitz J, Bergemann T, Breslow NE. Surgical complications after primary nephrectomy for Wilms’ tumor: report from the National Wilms’ Tumor Study Group. J Am Coll Surg 2001;192:63,8; quiz 146. 72. Grundy P, Perlman E, Rosen NS, et al. Current issues in Wilms tumor management. Curr Probl Cancer 2005;29:221-60. 73. Wilimas JA, Kaste SC, Kauffman WM, et al. Use of chest computed tomography in the staging of pediatric Wilms’ tumor: interobserver variability and prognostic significance. J Clin Oncol 1997;15:2631-5. 74. Owens CM, Veys PA, Pritchard J, Levitt G, Imeson J, Dicks-Mireaux C. Role of chest computed tomography at diagnosis in the management of Wilms’ tumor: a study by the United Kingdom Children’s Cancer Study Group. J Clin Oncol 2002;20:2768-73. 75. Rohrschneider WK, Weirich A, Rieden K, Darge K, Troger J, Graf N. US, CT and MR imaging characteristics of nephroblastomatosis. Pediatr Radiol 1998;28:435-43. 76. Gylys-Morin V, Hoffer FA, Kozakewich H, Shamberger RC. Wilms tumor and nephroblastomatosis: imaging characteristics at gadolinium-enhanced MR imaging. Radiology 1993;188:517-21. 77. Breitfeld PP, Meyer WH. Rhabdomyosarcoma: new windows of opportunity. Oncologist 2005;10:518-27. 78. Bernstein M, Kovar H, Paulussen M, et al. Ewing’s sarcoma family of tumors: current management. Oncologist 2006;11:503-19. 79. Daldrup-Link HE, Franzius C, Link TM, et al. Whole-body MR imaging for detection of bone metastases in children and young adults: comparison with skeletal scintigraphy and FDG PET. AJR Am J Roentgenol 2001;177:229-36.
19
Interventional Radiology in Oncology O.J. O’Connor, MRCSI, J.M. Buckley, MRCS, and M.M. Maher, MD, FRCR
1
Introduction
Since its development as a sub-speciality within radiology, Interventional Radiology (IR) has played an increasingly important role in caring for the patient with cancer. This role begins with initial diagnosis of cancer and involvement now extends into minimally invasive treatment of malignancy alone or in combination with other treatment modalities. IR has established a very important role in the management of complications incurred during many oncological treatments. This chapter provides an updated overview of the scope of IR in the management of the oncology patient.
2 2.1
Interventional Radiology in the Diagnosis of Cancer Biopsy
In the modern era, the interventional radiologist utilises an expanding range of imaging modalities, either alone or in combination, to assess the appropriateness of percutaneous biopsy in individual cases, to obtain a histological diagnosis and or definitively stage malignant disease. The modern interventionalist, therefore,
Cork University Hospital, Mercy University Hospital and University College Cork, Cork, Ireland Cork University Hospital, and University College Cork, Cork, Ireland Cork University Hospital, Mercy University Hospital, and University College Cork, Cork, Ireland Correspondence to: Michael M. Maher, Department of Radiology, University College Cork, Cork University Hospital, Wilton, Cork, Ireland e-mail:
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
493
494
O.J. O’Connor et al.
requires skills in interpretation of modern cross-sectional imaging techniques to ensure that percutaneous biopsy is indicated, that the correct lesion is biopsied when there are multiple lesions, and to offer an opinion regarding future management when histological diagnosis based on percutaneous biopsy would not appear representative of imaging appearances. Percutaneous biopsy was first described by Leiden in 1883 when the procedure was utilised to sample the causative microorganisms of pneumonia [1]. Percutaneous biopsy has been applied in most organ systems with excellent results and few complications [2]. The key to successful and safe biopsy is the use of image guidance which facilitates safe passage of needle into an organ or mass to facilitate histological or cytological analysis [2]. In oncology patients with febrile neutropenia, percutaneous biopsy is also increasingly being performed for microbiologic analysis of lesions within organs such as lung or liver suspicious for opportunistic infections such as fungal infection. With regard to choice of image-modality to guide percutaneous biopsy, ultrasound (US) has the advantage of real-time imaging, allowing accurate monitoring of the needle trajectory as it traverses tissues en route to the target lesion [3]. When a lesion is visible by ultrasound, with appropriate ultrasound equipment and operator experience, this modality offers much better real-time imaging than CT [3]. In addition, the use of US avoids radiation exposure to patients and staff during the course of a biopsy. The use of CT has the benefit of precise needle localization and better definition of regional anatomy when compared with US [3] (Fig. 19.1). This is particularly important in the case of pelvic or retroperitoneal biopsies that can frequently be difficult to perform using US guidance. The use of CT has the disadvantage of increased procedure duration and associated radiation dose to staff and patient. The utilization of CT fluoroscopy allows near real-time imaging of needle trajectory. Contraindications to percutaneous needle biopsy include coagulation defects which can increase the risk of bleeding following the procedure or lack of a safe access to the lesion [2]. For biopsy of intrabdominal or pelvic lesions, traversing bowel should be avoided. There are relative contra-indications which are specific to individual organ biopsies such as severe emphysema, pulmonary hypertension or previous pneumonectomy in the case of percutaneous lung biopsy. In these situations, the benefits of the procedure need to be weighed against the risks and discussion and consensus at multidisciplinary meetings is extremely helpful. Once it has been decided that the risk:benefit ratio is acceptable, a number of physiological parameters need to be measured and corrected in order to adequately prepare a patient for the procedure. The pathological samples obtained using IR take one of two main forms; histological and cytological. In general, the yield from cytology is less than that of histology. Although larger specimens are preferred where possible, the size of the sample that can be obtained depends on the size and location of the mass. In general, at our institution, visceral biopsies with the exception of lung biopsies are
19 Interventional Radiology in Oncology
495
Fig. 19.1 64-year-old man with lung nodule: (a) CT scan of chest in prone position shows lower lobe lung lesion. (b) A 19-gauge needle has been positioned within the lesion under CT guidance. Fine needle aspiration or core biopsy is then performed using coaxial technique
performed using a 17 or 18 G co-axial needle system. For percutaneous lung biopsy we use a 19 or 20 G co-axial needle system. In selected cases, with appropriate lesion selection, percutaneous biopsy can not only establish histological diagnosis but in addition, can facilitate staging of disease. For example if a patient has a malignant-appearing lung lesion in the presence of a liver or adrenal lesion, biopsy of the liver or adrenal lesion can establish histological diagnosis and stage the patient at the same time (Fig. 19.2).
2.2
Complications of Biopsy
Many of the complications described following percutaneous needle biopsy are common to biopsy of any organ and include bleeding and infection. Other complications are specific to the organ being biopsied such as pneumothorax and
496
O.J. O’Connor et al.
Fig. 19.2 72-year-old-man with right lung mass and adrenal mass: Biopsy of adrenal mass facilitates histological diagnosis and definitive staging in one procedure. (a) CT scan shows right upper lobe speculated nodule. There is severe background emphysema, which increases the risk associated with percutaneous lung biopsy. (b) CT scan of the upper abdomen shows a left adrenal mass (arrow), suspicious for metastatic disease. (c) Percutaneous biopsy of left adrenal gland in left lateral decubitas position achieves histological diagnosis and confirms advanced staging. The risk of pneumothorax is avoided
hemoptysis following lung biopsy, bowel perforation following pelvic mass biopsy and hematuria, urinary retention and prostatitis following transrectal prostate biopsy. Appropriate informed consent should include a description of common complications specific to the biopsy being performed. The accepted incidences of complications following a range of IR procedures, including percutaneous biopsy has been reported by the Society of Interventional Radiology (SIR) [2] Incidences of complications, however, can vary between institutions, depending on severity of disease in patient population being treated, and the incidence of co-morbidity.
19 Interventional Radiology in Oncology
497
Institutional audit should be performed and complication rates should be compared with threshold incidence values, which have been calculated for each procedure.
3
Interventional Radiology in the Treatment of Cancer
IR services may be used in a multidisciplinary setting to assist in the management and treatment of cancer. Perhaps the most common means by which IR can facilitate patient treatment is by the provision of image guided central venous access. In addition, IR procedures are currently expanding the range of oncological therapies. These procedures are generally less invasive than surgery and are therefore considered minimally invasive. Chemotherapy agents can be selectively administered as part of chemoembolization procedures and focused treatments using thermal ablation and gene therapy may be administered in the appropriate setting using image guidance.
3.1
Central Venous Catheters
Central venous catheters (CVC) provide a means of administering medications or parenteral nutrition to patients. There are 4 main types of CVC device. Temporary peripheral and central non-tunnelled catheters, tunnelled central catheters and implanted devices. Over 200,000 central access devices are inserted per year in the UK [4]. Although these devices have been inserted by anaesthetists and surgeons in the past, these devices are now commonly inserted using IR [5]. This is because real-time imaging guidance of the needle or catheter either by radiological screening or ultrasound reduces the incidence of complications related to insertion [6]. For these reasons central venous access using ultrasound guidance is recommended by the National Institute for Clinical Excellence in the UK since 2002. The right internal jugular vein is the most commonly used central access portal [7]. Complications related to central venous access procedures are dependent on choice of access route and are impacted by patient selection [8]. Complications that occur at the time of insertion are typically related to injury to surrounding structures or malposition of catheter and occur less commonly when performed with imageguidance than when performed blindly or using external landmarks [8].
3.2
Thrombosis
In addition, IR has the ability to diagnose complications related to CVC insertion and IR can also map alternative access sites in difficult cases. The common long term complications associated with CVC placement are thrombosis and infection. The overall
498
O.J. O’Connor et al.
long-term incidence of central venous thrombosis is between 30% and 70% [9, 10]. Trauma to the endothelium from the catheter tip is believed to cause the accumulation of thrombus [11]. Although the incidence of symptomatic CVC thromboembolism is less than 5%, asymptomatic thrombosis frequently occurs. Larger catheters with more lumens and catheters inserted into left sided veins are associated with an increased risk of thrombosis as is catheter insertion in patients with inherited prothrombotic tendencies [10]. Thrombosis is a potentially serious complication. It is the second leading cause of death in patients with cancer and 1 in 7 cancer patients who die in hospital do so as a result of venous thromboembolism which occurs for a variety of reasons including presence of CVC [12]. Unfortunately thrombotic prophylaxis using low dose warfarin and heparin has not been shown to reduce the incidence of CVC thrombosis [13]. Satisfactory data pertaining to the treatment of catheter related thrombosis are lacking. No uniformly accepted method of anticoagulation or duration of such treatments exists.
3.3
Infection
Nosocomial infection introduced at the time of catheter insertion is an important source of patient morbidity. Efforts to reduce the incidence of infection have concentrated on asepsis at the time of insertion and careful skin preparation [14]. Up to 20% of patients with catheter-related blood borne infections die. One third of these deaths are directly attributable to catheter-related infection [14]. Catheterrelated infections should be diagnosed by paired quantitative blood cultures. In the absence of shock, local infection or septic thrombophlebitis, a tunneled CVC with an external access device may be left in situ and the infection treated with parenteral antibiotics [15]. The deposition of infected clots within port devices means that once infected removal of the device will be required [15].
3.4
Embolisation
Minimally invasive image guided cancer treatments as an alternative or adjunct to surgery are currently under development. These cancer treatments consist of image guided procedures which sometimes require more than one IR session to compete therapy. Chemoembolization, radionuclide ablation and thermal ablation are being performed at present. Following contrast enhanced CT or MRI, transcatheter embolization may be performed in order to devascularise neoplastic tissue by occluding tumour arterial supply [16]. Mechanical occlusion is achieved using Gelfoam (Pharmacia & Upjohn, Kalamazoo, MI), polyvinyl alcohol or blood clots. These materials are introduced into the tumour bed following fluouroscopic guided selective arterial catheterisation by IR. Angiographic experience particularly with embolization techniques and in-depth knowledge of normal and variant anatomy
19 Interventional Radiology in Oncology
499
relevant to the procedure are vital before embarking on these techniques. These techniques can be used as the primary modality of treatment of hepatic metastases or in conjunction with ablative treatments or conventional surgery. Arterial embolization can also be employed prior to surgical resection in an effort to reduce operative blood losses, particularly when tumours appear hypervascular on preoperative imaging. Palliative embolization of inoperable tumours may also reduce tumour burden and help treat symptoms. This is most commonly performed to assist in the treatment of hepatic disease. For palliative treatment of hepatic disease, embolization materials may be administered in combination with chemotherapy agents and ethiodized oil (Ethiodol: Savage, Melville, NY). The term “chemoembolization” is used to describe this procedure [17]. Hepatic tumours rely on the hepatic artery for most of their blood supply. This has been demonstrated using radiolabelled albumen which quantified tumour uptake of radioisotope infused through the hepatic artery compared to the portal vein. The uptake of radioisotope was ten-fold greater following hepatic arterial infusion than following portal vein infusion [18]. Embolization of the hepatic artery allows chemotherapy agents to dwell within a target tumour for a longer period of time than infusion of a chemotherapy agent alone. This technique has been used successfully in the treatment of hepatocellular carcinoma (HCC), hypervascular metastases (e.g. ocular melanoma) and hepatic endocrine metastases. Hepatic transplantation for patients with HCC, when there is a single tumour measuring less than 5 cm in size or 2 tumours each less than 3 cm in diameter, results in a 70% 5-year survival [19]. For inoperable disease, the 2-year survival of patients with HCC is improved from 10 % without treatment to 30-40% following transcatheter oil chemoembolization [20]. Significantly increased 2-year survival from 27% to 63% has been demonstrated when chemoembolization has been performed by administering doxorubicin with gelatin [21]. Hormonal symptoms caused by endocrine hepatic metastases can be effectively treated with chemoembolization when somatostatin agents or other medical manipulations become less effective [22]. It is currently unclear whether chemoembolization is superior to embolization in the treatment of hepatic lesions [16]. The effects of chemotherapy alone versus chemotherapy combined with chemoembolization are currently being investigated in the setting of hepatic metastases secondary to colorectal carcinoma [23]. Since embolization causes tissue necrosis, complications such as sepsis, abscess formation and ischemia may occur. Approximately 10% of patients experience complications as a result of chemoembolization [17]. Post-embolization syndrome consists of fever, pain and elevated white cell count. This is experienced by large numbers of patients following embolization [17]. Prior to treating hepatic lesions, bowel preparation, drainage of obstructed biliary systems and octreotide administration for carcinoid tumours help to reduce the incidence of complications. Chemoembolization requires that the portal vein be patent or at least there must be good collateral blood supply to the liver. Otherwise hepatic necrosis is likely to occur following treatment. Interestingly periprocedural antibiotics are not a uniformly accepted method of reducing gram negative sepsis [24]. Ischemia of non-target
500
O.J. O’Connor et al.
organs such as the spine is another potential major complication. The incidence of spinal ischemia is reduced by careful planning angiography. Re-imaging should be performed 4-6 weeks following treatment. The absence of arterial phase enhancement of a lesion which was hypervascular on pre-procedural imaging is suggestive of successful treatment. Nodular enhancement of the portal vein and the appearance of an enhancing lesion usually signify residual disease. Treatment may involve a number of sessions until the entire tumour bed is devascularised. The ability to repeat treatments is an advantage of chemoembolization over surgical options. Radioembolization of tumours is another example of treatment that may be administered through a carefully placed catheter using IR. This method had not gained widespread recognition but offers the potential for focused treatment particularly of primary and secondary hepatic malignancies. Microspheres composed of glass, albumen or resin may be combined with radionuclides such as yttrium-90, rhenium or holmium and introduced directly into a tumour mass [25]. The type of radioisotope that is best suited to treating a tumour will depend on the nature of the tumour. These radioisotopes are primarily beta radiation emitters. This form of radiation has a very low penetrating power and its necrosing effects are localised. The emission of some gamma radiation is desirable as gamma radiation can penetrate the body facilitating detection by a gamma camera. Gamma radiation allows the distribution of the radiolabelled particles to be assessed. The efficacy of radioembolization has yet to be determined by randomised control trials, however, accurate IR transcatheter delivery has shown to be safe and preliminary results confirm its efficacy [26].
3.5
Thermal Ablation
In addition to chemoembolization and radioembolization, tumour necrosis may also be achieved using a variety of thermal ablation techniques. These methods include radiofrequency (RF), laser, microwave, ultrasound and cryoablation. IR mediated thermal ablation induces tumour necrosis by the application of energy. Radiofrequency ablation (RFA) is performed by applying electromagnetic energy with a frequency of less than 30 MHz to a tumour. Most devices apply energy between 375 and 500 kHz [27]. RFA may be administered using monopolar or bipolar energy sources. The most commonly used devices consist of multitined electrodes which have an umbrella appearance, clustered electrodes which consist of multiple internally cooled electrodes that are held together as a group and finally perfusion electrodes which allow fluids such as hypertonic saline to be instilled into the tissues being treated [28]. Tissues are heated to temperatures in excess of 60 degrees Celsius ensuring cell death. This method has been shown to be safe with a mortality rate of 0.3% and the rate of major complications is 2.2% [29]. RFA has gained acceptance as a method of treating hepatic and lung disease and is being used in the treatment of adrenal, renal and skeletal lesions [30, 31, 32]. RFA has been reported to be
19 Interventional Radiology in Oncology
501
effective in treating tumours up to 7 cm in size and a 1 cm margin of treated normal tissue surrounding the tumour is desirable. Other methods of ablation are less commonly used than RF [33]. Laser ablation, focused ultrasound ablation and microwave ablation, which uses electromagnetic radiation with a frequency in excess of 30 MHz, are all currently under investigation. Many of these techniques are still in development and the roles and relative success rates of these modalities in local treatment of malignant lesions at different sites have not yet been clearly defined. The involutional changes that occur following necrosis should be monitored by serial imaging following ablation. Specific post-ablation CT and MR imaging protocols are being developed at many institutions in an effort to confirm completeness of ablation and to detect residual or recurrent disease [34].
3.6
Gene Therapy
Gene therapy for cancer may be performed by stimulating tumoral immune response (tumor vaccines, cytokines), reducing the expression of oncogenes, restoring tumour suppressor gene function (p53), enhancing chemotherapeutic sensitivity (chemoembolization) and by modifying angiogenesis (retroviruses ) [35]. Selective arterial embolization following the delivery of genetic agents by IR reduces unwanted side effects and increases dwell time, achieving a better genetic transfer rate [36]. This therapy may be performed directly using cytokine and p53 genes, however, DNA crosses cell membranes poorly minimizing transfection rates. In order to adequately express a therapeutic molecule within a cell, vector agents that carry genetic agents across the cell membrane are required [37]. Plasmids and phospholipid agents are often used for this purpose but because only a small proportion of a plasmid enters the nucleus, it is not incorporated into the genome. In addition, plasmid mediated delivery is often shortlived. Retroviral, adenoviral and Epstein Barr viruses which cannot replicate are not without their limitations but can achieve greater and longer lasting genetic expression [38]. Newer therapies are likely to focus on these agents as potential methods of delivery [39].
4
Interventional Radiology in the Treatment of the Complications of Oncology
Many complications can occur as a result of organ dysfunction induced by malignancy. Although these complications can be very debilitating, many are reversible. These patients are often in ill health, malnourished and represent poor operative candidates. Minimally invasive methods of treating these complications are provided by IR. For example, obstruction of the renal, biliary and gastrointestinal systems may be
502
O.J. O’Connor et al.
relieved. Tumours that outgrow their blood supply may induce ischemia and necrosis not only inducing pain, but also resulting in perforation and occasionally abscess formation. Localised tumor abscess can be treated by percutaneous catheter drainage, but patient and referring physicians need to be made aware that although such treatments frequently reduce sepsis and alleviate pain and discomfort, high volumes of fluid typically drain persistently from these infected tumors and frequently the catheter must remain in situ for the remainder of the patient’s life [40].
4.1
Biliary Obstruction
In the oncology patient, biliary obstruction can occur as a result of intrinsic obstruction or extrinsic compression of the bile ducts. The majority of patients presenting with malignant biliary obstruction have obstructive jaundice due to distal common bile duct obstruction secondary to pancreatic neoplasm [41]. Other causes of malignant obstruction include cholangiocarcinoma or metastatic disease. Obstructive jaundice secondary to metastatic disease is frequently due to metastatic nodal disease at the liver hilum or in peripancreatic nodes. Unlike distal biliary obstruction which can be treated endoscopically, proximal obstruction usually requires percutaneous intervention. Percutaneous transhepatic cholangiography (PTC) involves the injection of contrast into an intrahepatic bile duct in order to image the biliary tree. Percutaneous transhepatic biliary drainage (PBD) may be either external or internal/external if the level of obstruction can be bridged by a catheter. Percutaneous treatment of biliary lesions is frequently staged requiring several sessions to achieve therapeutic goals [42]. Currently, metal stents are almost exclusively used for malignant disease with a 6 month patency rate of 50% [43]. In the majority of patients, indices of liver function improve following treatment [40]. There is a higher incidence of complications associated with PBD performed in patients with cancer than in general population [44]. This possibly relates to the presence of co-existing immunosuppression. Patients with cancer can have an incidence of cholangitis following PBD of up to 50% [45]. One third of these infections occur in patients drained externally and two-thirds occur in patients drained externally and internally. In addition, there is a positive association between the duration of PBD and the development of cholangitis [46]. Unfortunately patients requiring PBD are often in the later stages of their disease processes and the median survival following PBD in one series was 57 days among patients with pancreatic carcinoma [43]. Less frequent complications of PBD include catheter dislodgement, occlusion, leakage and electrolyte imbalance. Prophylactic antibiotics, effective in combating Escherichia coli, Klebsiella, Enterococcus, Streptococcus, Enterobacter and Pseudomonas aeruginosa, should be administered prior to percutaneous biliary procedures. In addition, careful manipulation of an infected system thus avoiding over-distension with contrast also decrease the risk of bacteremia [47]. PBD can be successfully completed in 95% of patients with dilated ducts but this rate
19 Interventional Radiology in Oncology
503
falls to 70% for non-dilated ducts [40]. Major complications of PBD include sepsis, hemorrhage and localized infection/inflammatory process (abscess, peritonitis, cholecystitis and pancreatitis) [43].
4.2
Renal Obstruction
The most common renal intervention performed by IR is percutaneous nephrostomy (PCN) (Fig. 19.3). The indications for PCN are urinary tract obstruction caused by intrinsic or extrinsic ureteral obstruction usually secondary to calculus disease, malignancy or iatrogenic causes [48]. Emergency PCN may be required for urinary tract sepsis, pyonephrosis, deteriorating renal function or metabolic disturbances such as hyperkalemia and metabolic acidosis. This procedure reduces the incidence of gram negative septicaemia due to renal obstruction and may improve impaired renal function and reduce inpatient admission times. Ureteric obstruction is not infrequent among oncology patients. In one series of 218 patients requiring PCN, 76 had underlying malignancy [49]. The most common malignancies causing obstruction in this series were cervical and prostate carcinoma. In cases of malignant ureteric obstruction, when retrograde stenting is unsuccessful or is not feasible, the ureter can be accessed antegradely through the PCN tract, and once the ureteric stricture is crossed, a ureteric stent can be placed. This may be either a 1 or a 2-step procedure. Obstructed renal collecting systems
Fig. 19.3 54-year-old woman with cervical cancer develops ureteric obstruction following surgical treatment: (a) Contrast injection immediately following percutaneous nephrostomy confirms 10 french catheter in good position within left renal collecting system. (b) Contrast-enhanced CT scan confirms percutaneous nephrostomy catheter in good position within left renal collecting system with good decompression
504
O.J. O’Connor et al.
without pyonephrosis are usually suitable for primary antegrade stenting as a onestep procedure with success rates of over 80% [50]. During a 2-step procedure either a PCN or an internal-external nephroureteral tube is inserted initially and a ureteric stent is placed on a subsequent visit to IR department. Retrograde stenting of an obstructed renal collecting system has the advantage of avoiding renal puncture. In the oncology patient, antegrade stenting may be more appropriate because retrograde techniques have a failure rate of 27% among oncology patients compared with 6 % among patients with benign disease. Failure is most commonly due to distortion of the ureteric orifices precluding stent insertion [51]. Plastic stents are favoured over metal ones because they induce less urothelial hyperplasia and they can be easily replaced. This is usually necessary every 3 to 6 months [52]. PCN can be successfully completed in 98-99% of patients. The rate of successful completion of PCN in oncology is mainly determined by degree of dilatation of collecting system and by patient’s body habitus [49]. Common major complications of PCN include septic shock and hemorrhage [49].
4.3
Upper Gastrointestinal Obstruction
Tumours of the head and neck often preclude oral feeding. Percutaneous gastrostomy or gastojejunostomy tubes may be inserted under fluoroscopy guidance as palliative measures (Fig. 19.4). These are most commonly performed by IR having fewer complications than endoscopically and surgically inserted feeding tubes [53]. Mild pain during feeding soon after gastrostomy insertion as well as infection are the most common early complications described with incidences of 33% and 23%, respectively [54, 55]. The most common long-term complication of gastrostomy tube insertion is tube dislodgement. If the gastrostomy tube falls out having been in situ for greater than two weeks, a tract has usually formed and it is frequently possible to access the tract and reinsert the tube without the need for re-puncture of the stomach. There is no significant difference between the complication rates associated with gastrostomy and gastrojejunostomy insertion. There have been reports of decompression IR gastrostomy and gastrojejunostomy insertion in patients with malignant small-bowel obstruction. These cases frequently have peritoneal carcinomatosis from malignant disease, with ovarian carcinoma being the most common primary tumor. Ryan et al reported 45 consecutive patients with metastatic ovarian cancer who underwent a radiologic gastrostomy or gastrojejunostomy with gastropexy [56]. A technical success rate of 98% was achieved and it was concluded that radiologic gastrostomy and gastrojejunostomy can be performed safely in patients with ascites and peritoneal carcinomatosis, if the patients undergo paracentesis first and if the re-accumulation of ascites is prevented after tube placement. In addition to paracentesis, fixation of the stomach to the anterior abdominal wall using gastropexy sutures also plays an important role in preventing pericatheter leakage.
19 Interventional Radiology in Oncology
505
Fig. 19.4 73-year-old man with pancreatic cancer: Percutaneous gastrojejunostomy catheter placed for feeding. (a) Contrast injection following placement of percutaneous gastrojejunostomy tube confirms that the tip of the tube is in excellent position in the jejunum (black arrow). Note the gas-filled stomach (white arrow) and the locking pigtail catheter in stomach (red arrow) which serves to maintain catheter in position and prevent dislodgement
4.4
Pain
Pain is a significant cancer related morbidity which is more common in the later stages of disease [57]. Opiates remain the mainstay of pain treatment. In order to reduce unwanted side effects the three-step ladder and medicine rotation methods are used [58]. With recent developments of new procedures, IR is assuming an expanding role in managing cancer associated pain. Vertebroplasty represents a new and effective treatment for vertebral body compression fractures refractory to medical therapy. To date, vertebroplasty has mainly been used in the treatment of osteoporotic fractures by introducing cement into the fracture thereby stabilising it. Osteolytic tumours often fracture resulting in instability and pain. Vertebroplasty has been shown to reduce requirements for analgesia and is now being utilized in the treatment of vertebral fractures which result from malignant osseous infiltration [59]. The incidence of major complications is 5% among oncology patients compared with an incidence of 1% in the general population [60]. The most
506
O.J. O’Connor et al.
significant complications are leakage of cement into the spinal canal, pulmonary embolus and pulmonary oedema. Vertebroplasty should relieve pain and improve mobility in 50-60% of oncology patients treated. Better results are achieved by treating sub-acute rather than chronic fractures [61]. Upper abdominal visceral tumours, and specifically pancreatic, gastric, esophageal, colorectal, gallbladder and cholangiocarcinoma are frequently associated with abdominal pain which is poorly responsive to analgesic therapy [62]. Celiac ganglion neurolysis and nerve block can be performed for palliation of pain, when resistant to analgesics. The success rate of the procedure has been reported to lie between 70-97% [62, 63]. Alcohol or phenol destroy nerve roots and triamcinalone reversibly blocks nocireceptors. There is controversy regarding timing of celiac plexus block, with some authors recommending that these interventions should only be performed in the later stages of disease and others arguing that better results have been demonstrated by treating disease in its earlier stage [63]. Successful celiac axis block can be performed using various methods of image guidance. CT guided blockade can be performed using an anterior approach or a posterior approach and the choice between these approaches is usually dependent on operator experience and anatomic considerations in individual patients [62, 64]. The most common reported minor complications are transient diarrhoea and orthostatic hypotension experienced by 73% and 12% of patients, respectively [64]. Although IR may be used to treat oncological pain it is worth noting that IR itself can induce pain among patients. The most common procedures which can result in intra-procedural and post procedural pain include biliary and nephrostomy drainages and percutaneous gastrostomy procedures [65]. It is important that interventional radiologists ensure optimal control of pain during interventional procedures and also work within a multidisciplinary team to ensure good pain relief following these procedures.
4.5
Pleural Space Intervention
Dyspnoea in the oncology patient is often caused by malignant pleural effusions (MPE) due to pleural and lymphatic involvement. These effusions may be treated by therapeutic thoracocentesis which can usually be performed as an outpatient procedure or by placement of a chest tube which usually requires admission to hospital. To definitively treat pleural effusions which re-accumulate following thoracocentesis or following chest tube removal, pleurodesis may be performed. A recent systematic review concluded that available evidence supports the need for chemical sclerosants for successful pleurodesis, with talc being the sclerosant of choice, and thoracoscopic pleurodesis as the preferred technique for pleurodesis based on efficacy [66]. Pleurodesis through chest tubes placed by IR can be performed at the bedside avoiding the need for general anaesthetic. However, the above systematic review suggested that risk of recurrence
19 Interventional Radiology in Oncology
507
of pleural effusion was less with thoracosopic versus bedside instillation through different sized chest tubes of various sclerosants including tetracycline, bleomycin, talc or mustine [66].
4.6
Venous Thromboembolism
Inferior vena cava (IVC) filter placement is an accepted method of managing venous thromboembolism (VTE) in the oncology patient. These filters are often inserted following lower limb deep venous thrombosis (DVT) in patients for whom anticoagulation is contraindicated, in whom a complication of anticoagulation has occurred or in whom recurrent PE’s occur in spite of adequate anticoagulation [67]. In the absence of anticoagulation, recurrent DVT occurs more commonly in oncologic patients but the incidence has been reported to be reduced to 4% with an inferior vena caval filter in situ [68]. Randomized control trails in this field are currently lacking. Although many devices are available, empirical use in oncology patients is not supported in the literature at present. The technical success of IVC filter placement is greater than 97% in experienced hands [67].
4.7
Abscess Drainage
The percutaneous aspiration and drainage of primary and postoperative abscesses and collections in the absence of indications for immediate surgery has helped reduce patient morbidity and morality and reduce hospital stay [69]. The method of imaging used to perform these procedures depends on the location of the collection. CT, CT fluoroscopy and ultrasound may all be used. Image-guided needle aspiration of fluid collections in oncology is frequently performed to investigate for infection or malignancy. Infected collections or collections causing pain or obstruction of the gastrointestinal, urinary or biliary tracts should be treated by image-guided catheter placement [40]. The incidence of complications associated with percutaneous drainage depends on patient health, location and nature of collection. It is generally reported to be 10% [70].
Key Points ●
●
Minimally invasive techniques in the hands of interventional radiologists are increasingly utilised in the management of oncology patients from initial diagnosis to treatment of primary tumor and metastatic disease and to management of cancer-related morbidity and complications related to treatment. The use of IR techniques should be evidence-based to ensure optimal outcome following utilization of these techniques in the oncology patient.
508
O.J. O’Connor et al.
References 1. Leyden OO, Under infectious pneumonie. Dtsch Med Wochenschr 1883; 9: 52. 2. Cardella JF, Bakal CW, Bertino RE, Burke DR, Drooz A, Haskal AZ at al, for the Society of Interventional Radiology Standards of Practice Committee. Quality improvement guidelines for image-guided percutaneous biopsy in adults. J Vasc Interv Radiol 2003; 14: S227-S230. 3. Arellano RS, Maher MM, Gervais DA, Hahn PF, Mueller PR. The difficult biopsy: let’s make it easier. Curr Probl Diagn Radiol 2003; 32:218-226. 4. Elliott TS, Faroqui MS, Armstrong RF and Hanson GC, Guidelines for good practice in central venous catheterization. Hospital Infection Society and the Research Unit of the Royal College of Physicians. J Hosp Infect 1994; 28: 163–176. 5. Tan PL, Gibson M. Central venous catheters: the role of radiology. Clin Radiol 2006; 61(1): 13-22. 6. Hind D, Calvert C, McWilliams R, et al., Ultrasonic locating devices for central venous cannulation: meta-analysis, BMJ 2003; 327: 361-367. 7. Trerotola SO, Kuhn-Fulton J, Johnson MS et al. Tunneled infusion catheters: increased incidence of symptomatic venous thrombosis after subclavian versus internal jugular venous access, Radiology 2000; 217: 89–93. 8. Lewis CA, Allen TE, Burke DR, Cardella JF, Citron SJ, Cole PE et al, for the Society of Interventional Radiology Standards of Practice Committee. Quality improvement guidelines for central venous access. J Vasc Interv Radiol 2003; 14: S231-S235. 9. Hoch JR. Management of the complications of long-term venous access. Semin Vasc Surg 1997; 10: 135–143. 10. Timsit JF, Farkas JC and Boyer JM et al. Central vein catheter-related thrombosis in intensive care patients: incidence, risks factors, and relationship with catheter-related sepsis, Chest 1998; 114: 207–213. 11. Tan P, Gibson M. Central venous catheters: the role of Radiology. Clinical Radiology 2006; 61(1): 13-22. 12. Pruemer J. Prevalence, causes, and impact of cancer-associated thrombosis. Am J Health Syst Pharm 2005; 62 (22 Suppl 5): S4-6. 13. Linenberger ML. Catheter-related thrombosis: risks, diagnosis, and management. J Natl Compr Canc Netw 2006; 4(9): 889-901. 14. Stiges-Serra A. Strategies for prevention of catheter-related bloodstream infections. Support Care Cancer 1999; 7: 391–395. 15. Standards, options et recommandations pour la prévention, le diagnostic et le traitement des infections liées aux voies veineuses en cancérologie. In: FNCLCC, editor. SOR 8: Infection et Cancer, John Libbey, France; 129–179. 16. Dyon D, Mouzon A, Jourde AN, Regensberg C, Frileux C. Hepatic, arterial embolization in patients with malignant liver tumours. Ann Radiol 1974;17:593-603. 17. Brown DB, Cardella JF, Dacks D, Goldberg SN, Gervais DA, Rajan D. et al. for the Society of Interventional Radiology Standards of Practice Committee. Quality improvement guidelines for transhepatic arterial chemoembolization, embolization, and chemotherapeutic infusion for hepatic malignancy. J Vasc Interv Radiol 2006; 17: 225-232. 18. Sigurdson ER, Ridge JA, Kemeny N, Daly JM. Tumor and liver drug uptake following hepatic artery and portal vein infusion. J Clin Oncol 1987; 5(11): 1836-40. 19. J Bruix and JM Llovet, Prognostic prediction and treatment strategy in hepatocellular carcinoma, Hepatology 2002; 35: 519–524. 20. Bronowicki JP, Vetter D, Dumas F, Boudjema K, Bader R, Weiss AM, et al. Transcatheter oily chemoembolization of hepatocellular carcinoma. A 4-year study of 127 French patients. Cancer 1994; 197: 101-8. 21. Llovet JM, Real MI, Montaña X, Planas R, Coll S, Aponte J et al.. Arterial embolisation or chemoembolisation versus symptomatic treatment in patients with unresectable hepatocellular carcinoma: a randomised controlled trial. The Lancet 2002; 359(9319): 1734-1739
19 Interventional Radiology in Oncology
509
22. Gupta S, Yao JC, Ahrar K, et al. Hepatic artery embolisation and chemoembolisation for treatment of patients with metastatic carcinoid tumours: the M.D. Anderson experience. Cancer J 2003; 9: 261-267. 23. Soulen MC. A randomized phase 3 study of systemic chemotherapy with or without hepatic chemoembolisation for liver-dominant metastatic adenocarcinoma of the colon and rectum. ACRIN protocol: 6655(www.acrin.org). 24. Reed RA, Teitelbaum GP, Daniels JR, et al. Prevalence of infection following hepatic chemoembolisation with cross linked collagen with administration of prophylactic antibiotics. J Vasc Interv Radiol 1994; 5: 367-371. 25. Nijsen JF, van Het Schip AD, Hennick WE, Rook DW, Rijk PP, de Klerk JM. Advances in nuclear oncology: microspheres for internal radionuclidetherapy of liver tumours. Current Medicinal Chemistry 2002; 9: 73-82. 26. Salem R, Thurston KG. Radioembolization with yttrium-90 microspheres: a state-of-the-art brachytherapy treatment for primary and secondary liver malignancies: part 3: comprehensive literature review and future direction. J Vasc Interv Radiol 2006; 17(10): 1571-93. 27. Goldberg SN, Dupuy DE. Image-guided radiofrequency tumour ablation: challenges and opportunities-part 1. J Vasc Interv Radiol 2001; 12: 1021-1032. 28. Kettenbach J, Kostler W, Rucklinger E at al. Percutaneous saline-enhanced radiofrequency ablation of unresectable liver tumours: initial experience in 26 patients. AJR 2003; 180: 1537-1545. 29. Livraghi T, Solbiati L, Meloni MF, Gazelle GS, Halpern EF, Goldberg SN. Treatment of focal liver tumours with percutaneous radiofrequency ablation: complications encountered in a multidisciplinary study. Radiology 2003; 226: 441-51. 30. de Meijer VE, Verhoef C, Kuiper JW, Alwayn IP, Kazemier G, Ijzermans JN. Radiofrequency ablation in patients with primary and secondary hepatic malignancies. J Gastrointest Surg 2006; 10(7): 960-73. 31. Rose SC, Thistlethwaite PA, Sewell PE, Vance RB. Lung cancer and radiofrequency ablation. J Vasc Interv Radiol 2006; 17(6): 927-51. 32. Brown DB. Concepts, considerations, and concerns on the cutting edge of radiofrequency ablation. J Vasc Interv Radiol 2005; 16(5): 597-613. 33. Goldberg SN, Grassi CJ, Cardella JF, Charboneau JW, Dodd GD 3rd, Dupuy DE, Gervais D, Gillams AR, Kane RA, Lee FT Jr, Livraghi T, McGahan J, Phillips DA, Rhim H, Silverman SG; Society of Interventional Radiology Technology Assessment Committee; International Working Group on Image-Guided Tumor Ablation. Image-guided tumor ablation: standardization of terminology and reporting criteria. Radiology 2005; 235(3): 728-39. 34. Dupuy DE, Goldberg SN. Image-guided radiofrequency tumor ablation: challenges and opportunities-part2. J Vasc Interv Radiol 2001; 12: 1135-1148. 35. Shiba H, Okamoto T, Futagawa Y, Ohashi T, Eto Y. Efficient and cancer-selective gene transfer to hepatocellular carcinoma in a rat using adenovirus vector with iodized oil esters. Cancer Gene Ther 2001; 8: 713-718. 36. Gerolami R, Cardoso J, Bralet MP, Cuenod CA, Clement O, Tran PL, Brechot C. Enhanced in vivo adenovirus-mediated gene transfer to rat hepatocarcinomas by selective administration into the hepatic artery. Gene Ther 1998; 5: 896-904. 37. Albelda SM, Wiewrodt R, Sterman DH. Gene therapy for lung neoplasms. Clin Chest Med 2002; 23:265-277. 38. Cohen ZR, Duvdevani R, Nass D, Hadani M, Ram Z. Intraarterial delivery of genetic vectors for the treatment of malignant brain tumors. Isr Med Assoc J 2001; 3:117-120. 39. Biceroglu S, Memis A. Gene therapy: applications in Interventional Radiology. Diagn Interv Radiol 2005; 11(2): 113-8. 40. Maher MM, Gervais DA, Kalra MK, Lucey B, Sahani DV, Arellano R, Hahn PF, Mueller PR. The inaccessible or undrainable abscess: how to drain it. Radiographics 2004; 24(3): 717-35. 41. Venbrux AC, Osterman FA. Malignant obstruction of the hepatobiliary system. In: Stanley Baum and Michael J Pentecost, editors.Abram’s Angiography: Interventional Radiology. Boston, MA, USA. Little, Brown and Co. 1997: 472-482.
510
O.J. O’Connor et al.
42. Burke DR, Lewis CA, Cardella JF, Citron SJ, Drooz AT, Haskal ZV et al, for the Society of Interventional Radiology Standards of Practice Committee. Quality improvement guidelines for percutaneous transhepatic cholangiography and biliary drainage. J Vasc Interv Radiol 2003; 14: S246-S246. 43. Rossi P, Bezzi M, Rossi M et al. Metal stents in malignant biliary obstruction: results of a multicenter European study of 240 patients. J Vasc Interv Radiol 1994; 5: 279-285. 44. Becker CD, Giatti A, Malbach R, Bauer HU. Percutaneous obstructive jaundice with the Wallstent endoprosthesis: follow-up and re-intervention in patients with hilar and non-hilar obstruction. J Vasc Interv Radiol 1993; 4: 597-604. 45. Carrasco CH, Zornoza J, Bechtel WJ. Malignant biliary obstruction: complications of percutaneous biliary drainage. Radiology 1984; 152(2): 343-6. 46. Nomura T, Shirai Y, Hatakeyama K. Bacteribilia and cholangitis after percutaneous transhepatic biliary drainage for malignant biliary obstruction. Dig Dis Sci 1999; 44(3): 542-6. 47. McNicholas MM, Lee MJ, Dawson SL, et al. Complications of percutaneous biliary drainage and stricture dilatation. Semin Interv Radiol 1994; 11: 242-253. 48. Ramchandani P, Cardella JF, Grassi CJ, Roberts AC, Sacks DSchwartzberg MS et al, for the Society of Interventional Radiology standards of practice committee. J Vasc Interv Radiol 2003; 14: S277-S281. 49. Lang EK, Price ET. Redefinitions of indications for percutaneous nephrostomy. Radiology 1983; 147: 419. 50. Watson GM, Patel U. Primary antegrade ureteral stenting: prospective experience and cost effectiveness analysis in 50 ureters. Clin Radiol 2001; 56:568-574. 51. Yossepowitch O, Lifshitz DA, Dekel Y, et al. Predicting the success of retrograde stenting for managing ureteral obstruction. J Urol 2001; 166:1746-1749. 52. Tolley D. Ureteric stents, far from ideal. Lancet 2000; 356:872-873. 53. Wollman BS, Horacio BD, Walus-Wigle JR, Easter DW, Beale A. Radiologic, endoscopic, and surgical gastrostomy: an institutional evaluation and meta-analysis of the literature. Radiology 1995; 197: 699-704. 54. Goncalves F, Mozes M, Saraiva I, Ramos C. Gastrostomies in palliative care. Support Care Cancer 2006; 14(11): 1147-51. 55. Silas AM, Pearce LF, Lestina LS, Grove MR, Tosteson A, Manganiello WD, Bettmann MA, Gordon SR. Percutaneous radiologic gastrostomy versus percutaneous endoscopic gastrostomy: a comparison of indications, complications and outcomes in 370 patients. Eur J Radiol 2005; 56(1): 84-90. 56. Ryan JM, Hahn PF, Mueller PR. Performing radiologic gastrostomy or gastrojejunostomy in patients with malignant ascites. AJR Am J Roentgenol 1998; 171(4): 1003-6. 57. Howard PH, Bonica JJ, Bergner M. The prevalence of pain in four cancers. Cancer 1987; 60: 2563 –2569 58. Ahmedzai S. New approaches to pain control in patients with cancer. Eur J Cancer 1997; 33 Suppl 6: S8-14. 59. McGraw JK, Cardella J, Barr JD, Mathis JM, Sanchez O, Schwartzberg MS, Swan TL, Sacks D et al, for the Society of Interventional Radiology Standards of Practice Committee. Society of Interventional Radiology quality improvement guidelines for percutaneous vertebroplasty. J Vasc Interv Radiol 2003; 14(9 Pt 2): S311-S315. 60. McGraw JK, Lippert JA, Minkus KD, Rami PM, Davis TM, Budzik RF. Prospective evaluation of pain relief in 100 patients undergoing percutaneous vertebroplasty: results and follow-up. J Vasc Interv Radiol 2002; 13(9 Pt 1): 883-6. 61. Crandall D, Slaughter D, Hankins PJ, Moore C, Jerman J. Acute versus chronic vertebral compression fractures treated with kyphoplasty: early results. Spine J 2004; 4(4): 418-24. 62. Titton RL, Lucey BC, Gervais DA, Boland GW, Mueller PR. Celiac plexus block: a palliative tool underused by Radiologists. Am J Roentgenol. 2002; 179(3): 633-6. 63. Akinci D, Akhan O. Celiac ganglia block. Eur J Radiol 2005; 55(3): 355-61.
19 Interventional Radiology in Oncology
511
64. Akhan O, Ozmen MN, Basgun N, Akinci D, Oguz O, Koroglu M, Karcaaltincaba M. Longterm results of celiac Ganglia block: correlation of grade of tumoral invasion and pain relief. AJR Am J Roentgenol 2004; 182(4): 891-6. 65. England A, Tam CL, Thacker DE, Walker AL, Parkinson AS, Demello W, Bradley AJ, Tuck JS, Laasch HU, Butterfield JS, Ashleigh RJ, England RE, Martin DF. Patterns, incidence and predictive factors for pain after Interventional Radiology. Clin Radiol 2005; 60(11): 1188-94. 66. Shaw P, Agarwal R. Pleurodesis for malignant pleural effusions. Cochrane Database Syst Rev 2004; (1): CD002916. 67. Grassi CJ, Swan TL, Cardella JF, Meranze SG, Oglevie SB, Omary RA et al, for the Society of Interventional Radiologists Standards of Practice Committee. Quality improvement guidelines for percutaneous permanent inferior vena cava filter placement for the prevention of pulmonary embolism. J Vasc Interv Radiol 2003; 14: S271-S275. 68. Streiff MB. Vena caval filters: a review for intensive care specialists. J Intensive Care Med 2003; 18(2): 59-79. 69. Gervais DA, Brown SD, Connolly SA, Brec SL, Harisinghani MG, Mueller PR. Percutaneous imaging-guided abdominal and pelvic abscess drainage in children. Radiographics 2004; 24(3): 737-54. 70. Bakal CW, Sacks D, Burke DR, Cardella JF, Chopra PS, Dawson SL et al; Society of Interventional Radiology Standards of Practice Committee. Quality improvement guidelines for adult percutaneous abscess and fluid drainage. J Vasc Interv Radiol 2003; 14(9 Pt 2): S223-5.
20
Breast Tumor Imaging Deirdre Coll, MD
Key points ●
●
●
●
●
1
Mammography remains as the gold standard for screening. Recent studies suggest that digital mammography is most beneficial for women under age 50, women of any age with heterogeneously or an extremely dense breast glandular pattern and pre- or perimenopausal women of any age. Breast tomosynthesis and breast CT offer more detailed images of the breast with thin sections being taken at various intervals through the breast. Both techniques require use of radiation. Their role in breast imaging remains to be defined. Breast ultrasound has undergone many technological advances. Harmonic imaging and compound imaging are clinically available and have been shown to improve image quality. Breast MRI has been shown to have a high sensitivity for the detection and demonstration of the extent of breast cancer. Its reported specificities however range from 30% to 80%, necessitating a means of tissue sampling and pathologic confirmation of MRI findings. There are now guidelines available to aid the physician when it is appropriate to order a breast MRI. Computer aided detection is now widely used for the interpretation of screening mammography. The largest clinical study performed showed reduced accuracy of interpretation of screening mammograms and an increased rate of biopsy that was not clearly associated with an improved detection of invasive breast cancer. This study may provoke research to reassess how this technology should be evaluated.
Introduction
Breast cancer is the second leading cause of cancer death in women in the United States. This means that every year in the U.S., approximately 40,000 women die of this disease. Statistics from the United Kingdom show that approximately 13,000 Department of Radiology, University of Maryland Medical Center email:
[email protected]
M.A. Blake and M.K. Kalra (eds.), Imaging in Oncology. © Springer 2008
515
516
D. Coll
women die of breast cancer each year. According to the World Health Organization, more than 1.2 million people will be diagnosed with breast cancer this year worldwide. The chance of developing invasive breast cancer during a woman’s lifetime is approximately 1 in 8 (about 13%). The importance of developing new imaging tools for the detection and diagnosis of breast disease is therefore self-evident. There have been encouraging advances in the treatment of breast cancer including newer minimally invasive surgical techniques and more advanced oncology and radiation therapy options. Physicians and surgeons today require more sophisticated information before making treatment decisions for their patients. Imaging techniques have evolved in response to these demands. This chapter will attempt to summarize and explain the application of existing and new imaging technology in the diagnosis, treatment and management of breast cancer.
2
Mammography
Mammography is the recognized test of choice for breast screening. In recent decades, the incidence of breast cancer has increased but mortality rates have stabilized or decreased [1]. This would suggest an improvement in prognosis likely due to screening Mammography [2]. It is now generally accepted that periodic mammographic screening results in earlier detection of breast cancers and reduces patient mortality [3]. Mammography is also the appropriate initial test for women over 35 (who are not pregnant) who present with a breast lump. The most common recommendation for mammography screening in the United States is annual testing starting at age 40. However, screening mammography has limitations in women with dense breasts, which occurs in 25-40% of patients including young patients and those on hormone replacement therapy [4]. Screening mammography also has significant false positive rate [5] of up to 33% in patients who undergo biopsy. The false negative rate of mammography ranges between approximately 20-35%.
3 3.1
Digital Mammography Background
Small field digital mammography had been available for many years but was limited to imaging only a focal area of the breast. Small field digital mammography was used very successfully for the performance of procedures such as stereo tactic biopsy or breast localization where only a small portion of the breast was required to be visualized [6]. The most significant recent development in the field of mammography has been the introduction of full field digital mammography (FFDM). FFDM images the entire breast and can be performed with computed radiography systems using photostimulable phosphor plates or systems that employ digital
20 Breast Tumor Imaging
517
flat-panel detectors. Systems that employ photostimulable phosphor plates require subsequent laser scanning of the plate to generate an image. Systems that employ digital flat-panel detectors include scintillator-photodetector devices, direct conversion devices and optically coupled CCD devices [7]. These flat panel detector systems record the images on a digital detector and display them almost immediately on a monitor. The introduction of FFDM allows the entire breast to be imaged digitally and offers many practical advantages: ●
● ●
●
●
Improved workflow over screen film mammography (SFM) as there is no film to be physically developed so the test can be performed more quickly. Near instantaneous viewing of images. Full integration to picture archiving and communication system (PACS) for storage and retrieval. This means that the problem of storing and retrieving prior screen film mammograms for comparison will eventually be phased out. Separation of the process of image acquisition from image display and storage, permitting post processing. Higher contrast resolution, less noise (as the digital detector captures more incoming photons) and equal or better dynamic range (Fig. 20.1).
From the diagnostic viewpoint, the use of digital imaging has opened up the world of advanced technology to mammographers. Image manipulation can now easily be performed on dedicated mammography workstations. The radiologist can manipulate the digital mammogram electronically to magnify an area, change contrast, alter the brightness, etc. With regard to FFDM viewing, it is a MQSA (Mammography Quality Standards Act) requirement that FFDM images be interpreted on a monitor that has 5 k × 5 k resolution [8]. In comparison to FFDM, SFM is limited by its narrow dynamic range, low contrast resolution, image noise, and film processing artifacts. SFM does however have a significantly higher spatial resolution than digital mammography. SFM can resolve between 15-20 line pairs per millimeter. This is approximately equivalent to a pixel size of 25 to 35 µm. FFDM systems offer a resolution of 5 to 10 line pairs per millimeter being equivalent to a pixel size of 50 to 100 µm. Given that one of the most important tasks in mammography is the detection and characterization of microcalcifications, this difference in resolution is a potential concern. The smallest calcifications generally seen in SFM are 100 to 200 µm. However, at a pixel size of 50 µm calcifications become very faint and it is argued that contrast resolution may be more important than spatial resolution [9].
3.2
Guidelines for the Use of Digital Mammography—The DMIST Study
The publication of the Digital Mammographic Imaging Screening Trial (DMIST) [10] in 2005 offered guidelines for the optimal use of digital mammography. This study enrolled 49,500 women at 33 sites in the United States and Canada. The
518
D. Coll
Fig. 20.1 43-year-old female with normal screening mammogram. (a) Mediolateral oblique film screen mammogram. This patient has dense fibroglandular tissue, this can obscure underlying lesions. (b) Mediolateral oblique digital mammogram of the same breast. On the digital mammogram, the imaging parameters can be manipulated to enable better visualization of this area
women in the trial underwent both FFDM and SFM examinations on the same day, each with a minimum of two views of each breast. Two different radiologists then interpreted these images independently. Breast cancer status was determined through available breast biopsy information within 15 months of study entry or through follow-up mammography ten months or later after study entry. It is important to note that DMIST showed that for the entire population of women studied that FFDM and SFM had very similar screening accuracy. Digital mammography was however significantly better in screening women within any of the following three categories: under age 50, women of any age with heterogeneously or an extremely dense breast glandular pattern and pre- or perimenopausal women of any age (this was defined as women who had a last menstrual period within 12 months of their mammogram). The results of the DMIST trial suggest that, for women who fall into these three subgroups, digital mammography may be better at detecting breast cancer than traditional screen-film mammography. However, although not analyzed as part of the DMIST study, because the overall results were similar, if FFDM performance was significantly better in the three subgroups identified above, then it necessarily implies that SFM might be significantly better in subgroups of the remaining patient population such as women with fatty-replaced breasts.
20 Breast Tumor Imaging
519
At present, only approximately 10 percent of the mammography units in the U.S. are digital systems, whereas it is estimated that 25-40% of women undergoing screening mammography have dense breasts. The limiting factor here is the price of digital systems. Digital units cost about $400,000 to $500,000 whereas screen-film units cost less than $100,000. The added costs of routine maintenance and image storage compound the price differential. There is recognition of this issue and some insurers, such as Medicare, have recognized these cost differences and reimburse FFDM at a slightly higher rate than SFM. However, the cost difference remains one of the greatest challenges in making FFDM technology available to every woman. There is also continuing debate over the validity of using age 50 as a “cut-off point” for recommending FFDM, the validity of the use of a 15 month follow-up period in the DMIST study and whether SFM is better for women who are over the age of 50, have fatty breast tissue, and those who are not still menstruating. Further analysis of these results is pending and will hopefully shed some light on these unresolved questions.
3.3
Advanced Applications of Digital Mammography
3.3.1
Contrast Enhanced Digital Mammography (CEDM)
Contrast enhanced breast MRI has utilized the fact that the development of a breast malignancy is accompanied by angiogenesis. In the early 1980’s Chang et al [11] demonstrated that breast cancer shows enhancement after intravenous administration of iodinated contrast. Contrast enhanced digital mammography (CEDM) simply means that digital mammography is performed before and after the injection of an iodinated contrast agent. For this technique to be successful, it requires subtraction of un-enhanced images from the enhanced images to highlight enhancing tissue. This can be accomplished either by using temporal subtraction (subtracting a precontrast or mask image from a post contrast image) or by dual energy subtraction. CEDM remains experimental, it is not FDA approved and only small pilot studies have been performed. Dromain et al [12] looked at 20 patients with malignant lesions using CEDM. A total of six contrast-enhanced craniocaudal views were acquired from 30 seconds to 7 minutes after the injection of a bolus of 100 mL of an iodinated contrast agent. Post processing included subtraction and analysis of enhancement kinetic curves. The mammography findings were then compared with biopsy results. The technique was successful in 18 patients and failed in two patients. The failures resulted from lesions that were located far posteriorally and were excluded from the field of view on the contrast enhanced images. The sensitivity for the detection of breast carcinoma in this study was 80%. This is similar to other published studies [13]. The authors concluded that contrast-enhanced digital mammography is able to depict angiogenesis in breast carcinoma but that further work is needed to improve technical aspects and assess specificity. Because injection of
520
D. Coll
contrast is involved with this imaging modality, it is unlikely that it will ever be accepted as a screening technique. Its role as a diagnostic technique remains to be defined.
3.3.2
Breast Tomosynthesis
Breast tomosynthesis systems are similar to that of conventional digital mammography, except that multiple projection images from different angles are acquired as the X-ray tube is moved through a limited arc above the stationary breast and digital detectors. X-ray exposures are obtained while the source is stationary at fixed sequential locations through the arc. It is possible to reconstruct any plane in the breast that is parallel to the detector [14]. Therefore, when the detector is positioned as in a craniocaudal (CC) view, multiple CC views are obtained as slices through the breast. When the detector is positioned as in a mediolateral (MLO) view, multiple MLO views are obtained as slices through the breast. The image obtained at each exposure angle is of low radiation dose, but the total radiation dose through the entire arc for obtaining slices through the breast is reported to be comparable to the dose used for a two-view film-screen mammogram. Tomosynthesis is intended to improve lesion conspicuity by minimizing the superimposition of overlying breast tissue. Tomosynthesis is not currently FDA approved. It had been hoped that tomosynthesis could be performed using approximately the same dose of radiation to each breast as a two-view screening mammogram. This had been thought possible by using just a single projection tomosynthesis— either the CC view only or the MLO view only. However, a recent small preliminary study questioned the need for both tomosynthesis views in order to evaluate the breast adequately. This was because some lesions were seen only on one view and not the other [15]. This is a significant question as it has dose implications for the patient and also workflow issues for the technologist and radiologist. Preliminary clinical results utilizing breast tomosynthesis have been encouraging. Rafferty et al. showed a 16% increase in sensitivity and an 85% decrease in false positives with digital breast tomosynthesis compared to digital mammography [16]. A question that remains to be answered is how well tomosynthesis will be able to depict individual microcalcifications (due to potential image blurring) and clusters of microcalcifications that are present on sequential tomosynthesis slices (due to blurring of portions of a cluster that are not present in the slice). The combination of contrast-enhanced digital mammography with tomosynthesis potentially combines the advantages of both. In this single test we could potentially have access to better morphological information and vascular kinetics. This technique also offers the advantage of direct correlation with the mammogram and would be less expensive and more widely available than breast MRI. Initial experience with pilot studies suggest that contrast enhanced digital breast tomosynthesis (CE-DBT) provides morphologic and vascular information that is concordant with other breast imaging techniques (Fig. 20.2) [17].
20 Breast Tumor Imaging
521
Fig. 20.2 43-year-old woman with palpable mass on physical examination in the upper quadrant of the right breast. (a) Craniocaudal mammogram with no obvious abnormality. (b) Temporal contrast enhanced digital mammography (CEDM) subtraction image obtained at 90 sec post contrast shows a round enhancing mass in the deep part of the breast (arrow). Histologic analysis confirmed an invasive ductal carcinoma. Images courtesy of Clarisse Dromain, MD, Institut Gustave Roussy – Villejuif, France
The clinical role of tomosynthesis remains unclear. It is unlikely given the current state of mammography in the United States that it will be accepted as a screening technique in the near future. The radiation dose, training requirements and the workflow issues of reading an increased number of images means that there are many significant practical problems to be overcome. The logistics of obtaining a sufficient amount of data to prove that tomosynthesis is superior to screening mammography would require many years of follow-up and very large multi-institution trials. Given that screening mammography has undergone such rigorous testing and its application is still questioned, the future role of tomosynthesis likely lies in a diagnostic application or in a clinically select patient population who are at high risk or who may benefit from a particular advantage of the technique.
522
3.3.3
D. Coll
Breast Computed Tomography (Breast CT)
There is currently great interest in the development of a low dose, high resolution dedicated Breast CT system. Early work on a dedicated breast CT scanner was unsuccessful. The first CT scanner designed for specifically imaging the breast was built as a prototype by General Electric in the 1970s. Unfortunately there were many problems including large slice thickness and large voxel size. Slip-ring technology was not available at that time so this led to a long scanning time and this system was never commercially marketed [18]. There have been similar attempts to use conventional body CT scanners for breast imaging. These non-dedicated breast CT scanners also suffered from large slice thickness, large voxel size, and excessive radiation dose to the breast [19]. There has been renewed interest in breast CT with the focus on developing a dedicated cone-beam breast CT system utilizing flat panel detectors. Clinical prototypes have been developed and initial trials imaging patients are ongoing. Generally, the breast is imaged with the patient lying prone and the breast hanging dependently through a padded hole in the table. For example, the scanner developed by Boone et al [20] generally takes from 10- 17 seconds for one scan and produces about 300 images with 512 × 512 resolution (Fig. 20.3). Preliminary studies have indicated that high quality CT images can be obtained with a radiation dose equal to that of a standard two-view mammogram [21].
Figs. 20.3 Right screening mammogram and right breast CT. (a) Heterogenously dense mammogram. There is a questions of architectural distortion in the upper portion of the breast (arrow). (b) CT of the same patient clearly shows that there is a spiculated mass (arrow) in this area which is worrisome for carcinoma. Images courtesy of the University of Rochester
20 Breast Tumor Imaging
523
The advantages of breast CT are that it provides exquisite anatomical detail of the breast, separates the overlying structures and allows for true 3D multiplanar reconstruction—a potential clear advantage over SFM or FFDM. However, breast CT has a serious potential dose disadvantage. Furthermore, a more relevant comparison would be with Breast MRI, which provides 3D multiplanar images, has been part of clinical practice for many years and has undergone extensive research. It is also important to keep in mind that both breast MRI and breast CT will likely remain predominantly used as diagnostic tools. One of the main problems with breast CT is the lack of image contrast, which means that on unenhanced images, a cancer and normal fibroglandular breast tissue typically have the same tissue attenuation. Therefore it can be very difficult to distinguish a small breast cancer from the surrounding tissue without the use of a contrast material. There are also other challenges facing the current breast CT systems including the difficulty of imaging axillary and posterior breast tissue, microcalcification resolution, noise reduction, scatter and dose. The dose consideration becomes more paramount if contrast material administration is necessary since multiple acquisitions will then be obtained. Furthermore, the time required to read large CT data sets must also be considered if this technique is to be successfully integrated into clinical practice. There are currently phase II clinical trials ongoing and it still remains to be determined what role this technology will play in the future of breast imaging.
4 4.1
Breast Ultrasound Introduction
Breast ultrasound has been used successfully for many years to distinguish solid from cystic masses with a near 100% accuracy [22]. The role of ultrasound in breast tumor imaging has greatly expanded over the last ten years. While currently not generally used for primary cancer screening, breast ultrasound has become an invaluable diagnostic tool most frequently used to further characterize a mammographic or palpable abnormality. The radiologist then assigns a level of suspicion to each imaging finding. The American College of Radiology has introduced a Breast Ultrasound Lexicon. The BIRADS-US terminology is used to classify the lesion into the same categories as those used in mammography.
4.2
Technical Advances
There have been a number of exciting technical advances in ultrasound, which have resulted in an improvement in image quality. These newer technologies are now available on most modern ultrasound machines and it is important to understand their applications and role in breast imaging.
524
4.2.1
D. Coll
Tissue Harmonic Imaging (THI)
When an ultrasound wave is transmitted into the body, reflected sound waves are produced at integral multiples of the transmission frequency. That is, if the transmission frequency is f, then sound is reflected at frequencies f, 2 f, 3 f, … nf, etc. These multiples of the transmission frequency are referred to as harmonics—the first harmonic being equal to the transmission frequency itself. Conventional ultrasound produces an image that represents the amplitude of reflected sound at the transmission frequency, or first harmonic. Tissue harmonic imaging (THI) produces an image that represents the amplitude of higher order harmonics. In general, the magnitude of the reflected sound drops off sharply for higher values of n so that THI is typically performed using the second harmonic. THI images images generally have improved axial and lateral resolution, less reverberation and side lobe artifacts and increased contrast resolution compared to conventional reflection sonography. Clinically, THI has proven very useful when scanning cystic breast lesions. The reduction of the “speckle artifact” which makes many cysts appear complicated has greatly helped with lesion characterization [23]. Rosen and Soo [24] compared 73 breast lesions with conventional imaging and tissue harmonic imaging. They found that radiologists preferred tissue harmonic imaging over conventional imaging for lesion conspicuity, margin evaluation and overall image quality. Disadvantages of THI include the fact that echoes at the second harmonic are weaker and can result in more image noise. THI also has less penetration than images obtained with conventional sonography and it may therefore be necessary to use conventional ultrasound in order to fully visualize the deeper tissues in patients with large breasts [25].
4.2.2
Real-time Compound Sonography
In conventional sonography, the image is constructed line-by-line at a constant angle of insonation. In compound sonography, the image is obtained by combining data from multiple different angles [26]. This can result in reduced image artifacts and improved image contrast. Similar to THI images, specular echoes in simple cysts are reduced and normal breast ligaments, capsules and spiculations around masses are better seen. Limitations of compound sonography include blurring from motion, as the image takes longer to acquire. It is important to note that with compound imaging there is also less posterior acoustic enhancement and shadowing and these artifacts are useful for lesion characterization (Fig. 20.4).
4.2.3
Color Flow, Power Doppler and Ultrasound Contrast Agents
The ability to assess blood flow with conventional ultrasound has been available for many years. It is well-known that tumor induced angiogenesis provides increased blood flow. These vessels are abnormal and show decreased impedance and
20 Breast Tumor Imaging
525
Fig. 20.4 54-year-old female with a biopsy proven infiltrating ductal carcinoma. Ultrasound with Power Doppler image shows an irregularly marginated hypoechoic mass with posterior acoustic shadowing. Note the tortous peripheral vasculature (arrow)
Fig. 20.5 35-year-old female with a simple breast cyst. (a) Conventional sonogram of the breast cyst. Note the internal reverberation artifact (arrow). (b) Songram of the same cyst with compound imaging. The overall image is smoother, there is less reverberation artifact and better definition of the cyst wall
increased permeability [27]. Tumor induced angiogenesis also increases the opportunity for tumor cells to enter the blood or lymph circulation. Studies evaluating blood flow to malignant tumors have generally shown overall increased vascularity, increased peripheral flow and low impedance flow (Fig. 20.5). The literature regarding use of peak systolic velocities and resistive indices to characterize breast lesion are confusing. Studies have shown an overlap in appearance between malignant and benign lesions and this has limited the usefulness of color and power Doppler for lesion characterization [28].
526
D. Coll
More recent work on the use of contrast enhanced ultrasound has shown the potential to improve ultrasound lesion characterization. Ultrasound specific intravenous contrast agents typically consist of an encapsulating shell material (albumin, phospholipid or polymer) surrounding air or a flourocarbon gas. These particles are relatively stable in the bloodstream and highly reflective of ultrasound at typical frequencies used in medical imaging. In general, the degree of increased echogenicity in a breast lesion will depend on the relative perfusion of the lesion compared to the parenchyma (thus yielding information similar to a contrast enhanced breast MRI). These agents have not been approved by the FDA for use in the breast in the United States although they have been used in Europe for several years. The blood vessels in a tumor vasculature are known to be highly disorganized, tortuous and dilated, with uneven diameter and excessive branching. Research is ongoing into analysis of these 3D morphologic features that when used in conjunction with more established parameters, can hopefully improve the specificity of the diagnosis [29]. Recent work by Caumo et al [30] has suggested that 3D mapping of tumor vessels with ultrasound contrast agents (termed angiosonography) is more accurate than color Doppler US in the correct assessment of biological behavior of suspicious breast lesions.
4.2.4
Elastography
It is well-known from clinical palpation that breast cancer usually feels “harder” than the surrounding tissue. Studies have shown that malignant breast lesions are in general less prone to deformation by pressure than normal breast tissue. The term “elasticity” describes how much pressure must be applied to a tissue for deformation to occur [31]. Use of sonography to measure elasticity, sonoelastography, offers a different method to generate tissue contrast and detect breast cancers [32]. Sonoelastography is generally performed using ultrasound before and after gentle palpation. Current sonoelastography technology allows calculation of tissue elasticity in real time. The information is displayed similar to a color Doppler image. The elasticity information is superimposed in color on the B-mode image. Preliminary studies indicate that while fibroadenomas can be up to two times stiffer than normal parenchyma, breast cancers can be up to 15 times stiffer than normal tissue [33]. Initial clinical studies using sonoelastography have been promising. Thomas et al. [34] demonstrated that sonoelastography of breast lesions has a higher specificity but a lower sensitivity in comparison with B-mode ultrasound. In a study of 300 focal breast lesions, mammography showed a specificity of 85% with a sensitivity of 87%; B mode ultrasound showed a specificity of 83% and a sensitivity of 94%; and elastography showed a specificity of 87% and a sensitivity of 82%. Another interesting feature of sonoelastography is a change of lesion size on sonoelastography images compared with conventional B-mode ultrasound. Breast cysts and fibroadenomas are the same size on the two modes of imaging but carci-
20 Breast Tumor Imaging
527
nomas appear to be larger on sonoelastography [35]. It has been suggested that the cancer appears larger because the tumor infiltrates the surrounding tissues also increasing their stiffness.
4.2.5
Breast Ultrasound Conclusion
The introduction of new ultrasound techniques offers exciting potential to increase specificity of breast ultrasound. Breast ultrasound is however a very difficult field in which to make a significant clinical impact in the United States. This is because ultrasound has to offer an extremely high specificity (to thereby avoid biopsy) in order to be successful. The performance of an ultrasound guided breast biopsy is in general a very quick and accurate option that most women will still choose when faced with uncertainty about the nature of a focal breast lesion. It is this author’s opinion however that these newer techniques could play an important role in decreasing the number of unnecessary biopsies performed. Medicine in general is moving to an evidence-based approach. If these newer technologies could sufficiently increase the probablity that a mass is benign, then this information could be used to lend support to the recommendation of follow up rather than biopsy in women with BIRADS 3 lesions.
5 5.1
Breast MRI Introduction
There has been an almost exponential rise in the use of Breast MRI over the last few years. It offers excellent tissue characterization, lack of use of ionizing radiation, multiplanar imaging and an exquisite depiction of breast anatomy. Breast MRI has become widely available throughout the United States. The strength of MRI is the near 100% sensitivity for the detection of breast cancer following contrast administration. However, the reported specificity and positive predictive value of MRI has varied greatly in reported series [36] due to the fact that many different types of benign lesions show enhancement and enhancement patterns of benign and malignant lesions can overlap. There is still debate in the literature regarding the use of morphologic and/or enhancement characteristics to distinguish benign from malignant breast lesions. In order to help standardize reporting of breast MRI studies, in 2003 the American College of Radiology (ACR) published the first ACR BIRADMRI lexicon to “aid clinicians in understanding the results of the breast MRI tests for subsequent patient management, and to aid scientific research by enabling investigators to compare studies based on similar breast MRI terminology.” The lexicon includes terminology to describe both morphologic characteristics and enhancement patterns of focal breast lesions.
528
5.2
D. Coll
Technique
The American College of Radiology (ACR) has published recommendations for obtaining breast MR images in the 4th edition of the Breast Imaging Reporting and Data System, (BI-RADS) [37]. At a minimum, high spatial resolution T1- and T2weighted images followed by pre and post gadolinium images are suggested. An effective method of fat suppression or subtraction is necessary to identify enhancing lesions following contrast administration. A 1 Tesla or higher strength magnet and a dedicated breast coil are required. Post-contrast enhanced images should be obtained at least every 2 minutes following contrast administration. The basic premise of contrast enhanced MRI is that malignant lesions are more likely to show early intense enhancement with a rapid washout. The patient should be scanned during the second week of her menstrual cycle to minimize hormonal effects. The time for each portion of the scan should not exceed 2 minutes with 90 seconds being a better time point to aim for. If the scan time is longer than 2 minutes, all of the tissues may enhance similarly. MRI guided biopsy and vacuum assisted biopsy devices are now widely available. It is prudent to refer patients to an institution that can complete the follow up of the patient and perform the MRI biopsy. This avoids a lot of patient anxiety and unnecessary waiting times. The standard practise is to initially perform a second look ultrasound in an attempt to identify an MRI-detected abnormality and then perform a biopsy with ultrasound guidance. If the abnormality is not identified on ultrasound then the patient undergoes an MRI guided biopsy or localisation.
5.3
Imaging Features & Controversies of Breast Cancer on MRI
The classical appearance of a breast cancer on MRI mimics that seen on mammography. It generally presents as an enhancing mass with an irregular or spiculated lesion margin (Fig. 20.6). It is generally of low signal on T2. Various subtypes of breast cancer can show different appearances. There has been much work done looking at contrast uptake within breast lesions (Figs. 20.7 and 20.8). A region of interest is selectively placed into the area of the lesion where the enhancement is the strongest. Lesion signal intensity is then plotted versus time to yield the signal intensity time course. There are three types of enhancement curves typically described [38]: type I, persistent ; type II, plateau; or type III, washout. A type I (persistent enhancement) curve is assigned if the signal intensity increases steadily throughout the dynamic period. This enhancement pattern is usually associated with a benign finding (83% benign, 9% malignant). A type II (plateau) curve is assigned if peak signal intensity is reached in the early post contrast period and is followed by a plateau of signal intensity in the remaining series. This pattern has a sensitivity of 42.6% and specificity of 75% for the detection of malignancy. A type III (washout) curve is assigned when peak signal intensity is reached in the early phase and is immediately followed by a loss of signal intensity in the early post contrast
20 Breast Tumor Imaging
529
Fig. 20.6 54-year-old female with infiltrating ductal carcinoma. Gadolinium image shows a spiculated mass (arrow) with adjacent satellite lesions. Note the skin retraction caused by fibrous spicules extending from the lesion. (a) Sagittal 3-D FSPGR MRI pre gadolinium. (b) Sagittal 3-D FSPGR MRI post gadolinium
period. This pattern is generally associated with malignant lesions having a specificity of 90.4% but a sensitivity of only 20.5%. The ACR BI-RADS-MRI recommends describing the initial enhancement pattern within the first two minutes (or when the curve begins to change) as “slow,” “medium” or “rapid”; and the delayed phase (i.e., the enhancement pattern after 2 minutes or after the curve begins to change) as persistent (“continued increase in signal over time”), plateau (“signal intensity does not change over time after its initial rise”) or washout (“signal intensity decreases after its highest point from its initial rise”): Unfortunately, breast cancers are composed of a heterogeneous group of tumors and they do not always conform to the typical post-contrast enhancement pattern malignant appearance with a fast initial phase and a washout on the delayed phase. Breast MRI also allows the use of morphologic features of focal abnormalities to help distinguish benign from malignant lesions. Morphologic features are included in the ACR BI-RADS-MR lexicon and are well-described in the literature [39]. These morphologic features include: margin, shape, homogeneity, internal septations and distribution of enhancement among others. There has been much debate in the literature over whether morphologic or kinetic features are the most important for the correct diagnosis of a breast malignancy. There now seems to be general agreement that morphologic analysis is more important than the contrast enhancement kinetic information and that the specificity is improved if both features are used for diagnosis [40]. Kuhl et al [41] wrote a fascinating paper evaluating the trade-off between temporal and spatial resolution in dynamic contrast material–enhanced bilateral magnetic resonance (MR) imaging of the breast. This impressive group had long been proponents of the kinetic evaluation of breast lesion. They looked at 54 lesions, 28 benign and 26 malignant. They imaged the women twice, once with a standard dynamic protocol and then a second time with a high spatial resolution protocol. They found that increased spatial resolution significantly improved diagnostic confidence and accuracy even if this improvement occurs at the expense of tempo-
530
D. Coll
Fig. 20.7 66-year-old female with invasive ductal carcinoma with an adjacent satellite lesion. (a) 3-D FSPGR post-gadolinium shows an enhancing mass (arrow) with a second satellite lesion seen anterior to the mass. Both are biopsy proven invasive ductal carcinomas. (b) The same image as in a with a color map applied. The large amount of color seen posteriorly reflects flow in the heart. The cross hair lies over the tumor which demonstrates increased uptake of contrast as compared to the surrounding tissue. (c) Time intensity kinetic curve of the mass shows a rapid uptake of contrast with a plateau or type II curve
20 Breast Tumor Imaging
531
Fig. 20.8 70-year-old female 2 years status post lumpectomy and radiation therapy. (a) Sagittal T1 weighted image shows scarring in the upper outer quadrant of her right breast. Note the irregular nodular linear stranding extending from her scar towards the nipple (arrow). (b) 3-D FSPGR postgadolinium subtraction image shows segmental linear enhancement of the linear stranding suspicious for DCIS (arrow). Biopsy confirmed these imaging findings
Delayed
Fa
st
Intitial
SI
Plateau
m
iu
ed
Persistent
M
Washout
ow Sl
Intitial slope within 2 minutes or when curve starts to change. Delayed slope after 2 minutes or after curve starts to change.
ral resolution. This was because of the broad overlap in enhancement rates for benign and malignant lesions. In other words the extra information gained by performing multiple scans to assess the uptake of contrast was not helpful. However it is important to note that the authors still felt that analysis of the time course pattern was still very helpful. They confirmed the time course analysis to have a high specificity and high positive predictive value.
532
D. Coll
5.4
Indications for use of MRI in the Breast Cancer Patient
5.4.1
Extent of Disease Evaluation for the Breast Cancer Patient
MRI has been shown to be the most sensitive imaging technique for extent of disease evaluation. For the patient with breast cancer, it offers an almost 100% sensitivity for the detection of invasive breast carcinoma. It has been shown to be the most accurate test for assessing multifocality, multicentricity, tumor size and evaluating the contralateral breast [42]. Many studies have demonstrated that breast MRI defines multifocality and multicentricity in many patients with breast cancer who present with what appears to be a solitary lesion on mammogram [43]. This is particularly true in patients with dense breasts in whom the mammogram is very difficult to interpret. Liberman et al. [44] reviewed multiple series totaling 1280 breasts with histopathologic correlation after a pre-operative diagnosis of unifocal cancer. In this review, 619 patients (48%) were found to have additional unsuspected tumor foci. It is known that incidental contralateral cancers occur in 3 to 6% of patients. In a recent large multicenter trial [45], MRI was shown to be superior to both mammograms and ultrasound to detect contralateral cancer. The study examined 969 women with a recent diagnosis of unilateral breast cancer and no abnormalities on mammographic and clinical examination of the contra-lateral breast. All women underwent breast MRI. The diagnosis of MRI-detected cancer was confirmed by means of biopsy within 12 months after study entry. MRI detected clinically and mammographically occult breast cancer in the contralateral breast in 3.1% of the 969 women. The authors concluded that MRI can detect cancer in the contralateral breast that is missed by mammography and clinical examination at the time of the initial examination. 5.4.2
Dense or heterogeneously dense mammographic pattern [46]
In dense and heterogeneously dense breasts, mammographic sensitivity is decreased with as few as 30%–48% of cancers depicted in extremely dense breasts. MRI allows evaluation of the dense breast and can easily depict cancers located within dense parenchyma following contrast administration. 5.4.3
Invasive lobular carcinoma (ILC)
MRI has proved especially helpful in patients with invasive lobular carcinoma because these tumors are often mammographically occult. On MRI ILC may manifest as an enhancing solitary mass with irregular margins, multiple enhancing lesions, or only enhancing septa [47]. Hilleren et al [48] showed in their series that more than one in three mammographically depicted ILCs was seen as a vague asymmetry, poorly defined opacity, or area of possible architectural distortion
20 Breast Tumor Imaging
533
which was difficult to detect mammographically. For the patient with newly diagnosed ILC, MRI has shown better correlation with pathological extent of disease. Weinstein et al showed that in 39% of women with ILC, MR imaging depicted more extensive disease than was suspected with conventional imaging.
5.4.4
Extensive Intraductal Component (EIC)
EIC is generally defined as ductal carcinoma in situ (DCIS) occupying 25% or more of the area encompassed by the infiltrating tumor and DCIS present in grossly normal adjacent breast tissue. Extensive intraductal component (EIC) is thought to play a role in local recurrence after treatment of early-stage breast cancer with lumpectomy and irradiation (Fig. 20.8). Mammography has been shown to underestimate the extent of DCIS, particularly when it is non-calcified. In this patient group, MR imaging can depict an extensive intraductal component that was otherwise occult [49, 50].
5.4.5
Scar Verus Tumor Recurrence
Breast MRI has proved very helpful in distinguishing scarring from recurrent carcinoma [51]. Mammography and ultrasound can be difficult to interpret as the appearance of a scar and a tumor can be similar. The sensitivity of mammography for detecting local recurrence at the lumpectomy site is around 60% to 70%. MRI can help to resolve this clinical difficulty by differentiating non enhancing scar or fat necrosis from enhancing recurrent tumour. The performance of MRI in differentiating scar from recurrence has been reported with sensitivities of 93% to 100% and specificities of 88% to 100% [52]. It is important to note that false-positive results can occur due to normal benign enhancement of granulation tissue for as long as 9 to 18 months after surgery. Therefore, in general in the absence of other clinical indications it is advisable to wait at least 6 months before imaging after surgery.
5.4.6
Positive Axillary Nodes with Negative Mammogram and Ultrasound
An occult primary breast carcinoma with positive axillary lymph nodes represents less then 1% of all breast cancers. In a review of six studies (n=113), where the mammogram was negative, the sensitivity of MRI was 94%, specificity 94-100% with an estimated PPV of 90% or greater [53].
5.4.7
Implant Assessment
Breast MRI is the definitive test for assessing implant integrity. It can distinguish silicone from fat or water using chemical selective techniques and it can depict the fine detail of the internal implant structure including in foldings (i.e., radial folds).
534
D. Coll
MRI can clearly delineate the fibrous capsule and therefore allows distinction between intracapsular and extracapsular rupture.
5.4.8
Indeterminate Mammogram And Ultrasound
Breast MRI is the test of choice when faced with an indeterminate mammogram. The usual clinical presentation is an abnormality seen on one view only. Such patients should undergo sonographic evaluation and if the clinical question is still not resolved, it is appropriate to refer the patient for a breast MRI.
5.4.9
Screening Of High Risk Populations
Studies have demonstrated the ability of MRI to detect cancers in high-risk women before they reach the age when mammography would be performed [54]. These women are those who are BRCA positive or who have a very strong family history. This includes multiple first- and second-degree relatives who have had breast or ovarian carcinoma, a first-degree relative who has had breast cancer before age 50 years, male relatives who have had breast cancer, and Ashkenazi Jewish women who have a family history of breast or ovarian cancer. The American Cancer Society has recently published guidelines for performing screening breast MRI in high-risk patients [55]. According to these guidelines: “Screening MRI is recommended for women with an approximately 20–25% or greater lifetime risk of breast cancer, including women with a strong family history of breast or ovarian cancer and women who were treated for Hodgkin disease. There are several risk subgroups for which the available data are insufficient to recommend for or against screening, including women with a personal history of breast cancer, carcinoma in situ, atypical hyperplasia, and extremely dense breasts on mammography.”
5.5
Magnetic Resonance Spectroscopy (MRS)
Proton MR spectroscopy (MRS) is approved by the U.S. Food and Drug Administration for imaging in the prostate and brain. Proton MR spectroscopy is based on the detection of elevated levels of choline compounds, which are a marker of active tumor [56]. Studies have shown that a resonance from choline containing compounds is commonly present in breast carcinomas but not in benign or normal breast tissue. Early studies have shown sensitivities of 70%– 100% and specificities of 67%–100% for breast MR proton spectroscopy [57]. This technique requires a very homogeous local field (the proton spectra of the breast often have large lipid signals that can give rise to contaminant peaks) and no motion artifact.
20 Breast Tumor Imaging
535
The advantages of MRS are that it can be performed at the end of a breast MRI examination without subjecting the patient to yet another MRI and it does not require contrast administration. Typically it is performed after the completion of the contrast enhanced sequences. The additional information gained from MRS may help serve to increase the specificity of the diagnosis. The other promising application of this technique is assessing tumor response to chemotherapy. Breast MRS offers the chance to assess response at a cellular level long before traditional assessment would be able to detect any change. Jagannathan et al [58] were the first to demonstrate that the tCho resonance disappeared or became smaller in 89% of subjects undergoing successful chemotherapy. Further research is ongoing into multi-voxel techniques, sodium (Na-23) MR imaging, chemical shift MR imaging (which would enable us to examine the whole breast) and higher-field-strength magnets. It is hoped that when used in combination, these techniques might provide a comprehensive data set to increase the accuracy of diagnosis. However much further work needs to be done for validation with large multi-center trials before this technique can be incorporated into routine clinical practice.
6 6.1
PET/Nuclear Medicine Background
Positron emission tomography (PET) scanning is widely used for the diagnosis, staging, and management of a variety of malignancies, including breast cancer. The vast majority of PET scans are performed using (F-18)-2-deoxy-2-fluoro-D-glucose (FDG) as the standard PET tracer. FDG is a glucose analog that is taken up by cells in proportion to their rate of glucose metabolism. It has been known for many years that increased glycolysis is a distinctive feature of malignant tumors compared with normal tissues. This increased utilization of glucose by malignant cells in comparison to normal tissue is the basis of the ability of FDG-PET imaging to differentiate cancer from benign tissue [59]. FDG-PET is not an appropriate primary imaging modality for the detection of primary breast cancer. In the U.S., the Centers for Medicare and Medicaid Services has not approved coverage of FDG-PET for the initial diagnosis of breast cancer and the staging of axillary lymph nodes due to insufficient sensitivity (64–79%) for lesion detection [60]. This is secondary to low spatial resolution and radiation exposure. Current state of the art whole body PET scans can only resolve approximately tumors of > 1 cm or more [61] (Fig. 20.9a, b). FDG uptake can also vary according to tumor type [62]. Invasive ductal carcinoma generally shows a high tumor to background ratio (TBR) whereas invasive lobular carcinoma shows a much lower TBR. Avril et al. [63] reported that up to two-thirds of lobular breast cancers demonstrate a false negative result.
536
D. Coll
Fig. 20.9 (a) Fluorine-18 fluorodeoxyglucose positron emission tomography shows strong glucose uptake in a left breast IDC (arrow). (b) Same patient as in a imaged with a specific breast coil. Note the better resolution enabling precise localization of the tumor (arrow) and a better evaluation of both breasts
6.2
Clinical Use
FDG PET has been reported to be clinically useful for the staging and restaging of breast cancer and also in assessment of response to treatment. The advantages of combining CT with PET have been well documented. This allows greater accuracy by combining functional with anatomic information.
6.2.1
Staging & Restaging Detection of metastases
FDG-PET and PET-CT is now commonly used for the staging and restaging of breast cancer. Studies have shown that the additional metabolic information provided by FDG-PET increases the accuracy of detecting recurrent or metastatic lesions, particularly for skeletal and regional nodal metastases. A meta-analysis by
20 Breast Tumor Imaging
537
Isasi et al. [64] looked at eighteen studies and determined that FDG-PET has a high diagnostic accuracy for the detection of breast cancer recurrence and metastases, with a summary true positive rate (TPR) of 90% and a summary false positive rate (FPR) of 12%, after the exclusion of outliers. They concluded that despite the 10% false negative rate that FDG-PET plays an important role in the follow up of patients with breast cancer. The Centers for Medicare & Medicaid Services have approved FDG-PET for insurance payment for the staging or restaging of patients with recurrent or metastatic breast cancer particularly when the results of conventional staging studies are equivocal. This decision was based in part on several retrospective studies showing the improved sensitivity and accuracy of FDG-PET compared with conventional imaging in restaging these patients [65].
6.2.2
Assessment of response to treatment.
Traditionally, we have used imaging procedures such as mammography, ultrasonography and MRI to measure tumor size in order to monitor response to therapy. However sequential measurement of tumour size frequently does not allow the determination of early response or differentiation between viable tumor tissue and scar tissue. The cellular uptake of FDG is a function of cell viability; animal models have shown that, after therapy, the amount of tumor FDG uptake reflects the number of viable tumor cells present. Follow up studies have shown that treatment-induced reduction in tumour metabolic activity has correlated with the clinical response [66]. On the basis of the results obtained to date, this would suggest that PET imaging may provide a way to evaluate the therapeutic response earlier than has been available to us with other imaging tools. This would significantly improve patient management by identifying ineffective therapies, preventing side-effects and allowing earlier introduction of alternative and more effective therapies [67]. The role of PET post radiotherapy and chemotherapy still needs to be investigated preferably with larger multi center trials.
6.3
Technology Update
There has been recent work towards developing breast specific PET imaging systems. This would allow improved spatial resolution. In general these systems image the breast in a prone position with a small field of view and a breast specific detector. These breast specific imaging systems have been combined with other breast imaging modalities such as digital mammography and breast CT. This would offer the advantages of combining functional and anatomical information. For these devices to be clinically useful, it is essential that these devices would have the ability to biopsy the abnormalities that they detect. Adler et al. [68] described a system that consists of an dedicated positron emission mammography (PEM) device
538
D. Coll
mounted on a stereotactic X-ray mammography system, permitting sequential acquisition of mammographic and emission images during a single breast compression. The authors were then able to directly perform a stereotactic biopsy of the suspicious area. There has also been investigation of new PET tracers which are aimed at targeting cellular processes that are more specific than glucose metabolism. For breast cancer, these tracers include thymidine analogs such as [F-18]fluoro-L-thymidine (FLT) that target DNA replication as a measure of cell proliferation, annexin V derivatives that evaluate apoptosis, and estrogen receptor (ER) tracers such as 16 -[F18] fluoroestradiol- 17 (FES).
6.3.1
Computer Aided Detection and Diagnosis
Errors made by radiologists during image interpretation can be cognitive, perceptual, technical or administrative. A radiologist commits a cognitive error when they perceive an abnormality but misinterprets the nature or significance of the abnormality due to incomplete knowledge [69] or faulty reasoning or judgment [70]. A radiologist commits a perceptual error when they fail to see an abnormality at the time of interpretation and that abnormality is “evident,” in retrospect, at a later time [71]. Technical errors include incomplete imaging, incorrect x-ray exposure, poor patient positioning, etc. Administrative errors include imaging the wrong body part or wrong region of the body, losing the films or failing to store the digital images properly [72], imaging the wrong patient, mislabeling the right side as the left side (or vice versa), failure to obtain or utilize adequate patient history, failure to compare the current test with prior tests, failure to correct an error in a dictated report, etc. Image interpretation can be broken down into 3 essential tasks: detection, description, and differential diagnosis [73]. Computers can be utilized during any of the three interpretation tasks To distinguish the use of computers during the detection task from the diagnosis task, the acronyms CADe (computer-aided detection) and CADx (computer-aided diagnosis) have been employed in the literature [74]. The term “computer-assisted” is used interchangeably with “computer-aided.” Diagnosis is also sometimes referred to as classification [75]. The goal of CADe is to help radiologists detect potential abnormalities (also referred to as “regions of interest”). Once a potential abnormality is detected, the goal of CADx is to help radiologists determine whether the potential abnormality is malignant, benign, indeterminate or otherwise requires a clinical action. A clinical action would include additional concurrent imaging, follow-up imaging, biopsy or surgery. Even when performing the detection task, a CADe device may be designed such that some component of diagnosis is also involved. For example, a CADe device designed to detect all potential abnormalities (i.e., to detect BI-RADS 2-5 lesions)
20 Breast Tumor Imaging
539
would have no component of diagnosis in the detection task. A CADe device designed to detect BI-RADS 3-5 lesions (i.e., designed to not detect benign abnormalities) would employ some component of diagnosis in the detection task. A CADe device designed to only detect BI-RADS 5 abnormalities (i.e., findings that have a ≥ 95% chance of malignancy) would employ a strong component of diagnosis in the detection task and might very well be considered a CADe and CADx device. Finally, a CADe device designed to only detect abnormalities that are in fact malignant would essentially be a CADx device. By design, a CADe device automatically detects potential abnormalities and brings them to the attention of the radiologist for subsequent performance of the diagnosis task. When performing diagnosis, it is crucial that the radiologist understand the design and limitations of the CADe device. For example, if a radiologist knows that a CADe device is designed to detect BI-RADS 3 or higher potential abnormalities, a radiologist might not hesitate to dismiss a potential abnormality marked by the CADe device. On the other hand, if a CADe device is designed to only detect BI-RADS 5 potential abnormalities, then a radiologist might be more hesitant to dismiss a potential abnormality marked by the CADe device [76]. Likewise, independent of the type of potential abnormalities that the CADe device is designed to detect and independent of the radiologist’s knowledge of the design of the CADe device, the number of CADe marks may independently affect the diagnosis task. For example, a large number of marks (regardless of the type of potential abnormality marked) may distract the radiologist during the diagnosis task. Even for the same potential abnormality, the physical characteristics of the mark itself might affect the diagnosis task [77]. An important limitation of currently available CADe devices is their inability to correlate CAD marks on different views of the same breast. This is a significant limitation because this form of CAD marking does not correspond to the paradigm of mammographic interpretation and severely limits the conclusions that can be drawn from CADe stand-alone performance testing (see below). There is extensive literature evaluating the performance of CADe devices. In order to properly interpret these studies it is important to have knowledge of the design of the CADe device as well as the methodology and gold standard used to measure performance. We have already discussed design issues above. With regard to methodology, CADe devices can be evaluated using stand-alone performance, clinical performance, outcome performance or some combination thereof. Stand-alone performance measures how well the CADe device marks regions of known mammographic abnormalities (and how well the CADe device avoids marking regions without abnormalities) in the absence of radiologist interaction. Clinical performance measures how well the radiologist performs when interpreting mammograms with CADe marks compared to interpretation without CADe marks. Outcome performance is a form of clinical performance and would compare detection rates of cancers without and with the use of CADe.
540
D. Coll
With regard to gold standards, there are at least three possible standards of truth that can be used to assess a region marked by a CADe device (radiologist-based assessment, pathologic-based assessment or a mix of the two): ●
●
●
A region marked by a CADe device can be correlated with radiologist-determined BI-RADS assessment categories. This approach directly parallels how radiologists currently interpret mammograms to classify findings and determine clinical action. However, this approach may be problematic because of the variability of BI-RADS assessment between different radiologists and because BI-RADS assessment categories may be associated with a wide range of probability of malignancy. A region marked by a CADe device can be correlated with the findings at pathologic examination (following surgery or biopsy). While this would correlate the CADe mark with the “ultimate” form of truth, even if a CADe device was able to mark the most subtle cancers (subtle in the sense that radiologists have a difficult time detecting and/or recognizing such a finding as suspicious), unless radiologists would ultimately assess the finding as at least BI-RADS 3 or higher, there would be no clinical action taken. A region marked by a CADe device can be correlated with both a radiologistdetermined BI-RADS assessment category and the findings at pathology. For example, one could argue that the ideal CADe device would detect only regions of interest that radiologists would assess as BI-RADS categories 3 or higher (or perhaps 4 or higher) AND that turn out to be cancers at pathologic examination.
CADe devices can reduce perceptual errors by bringing obvious or subtle visible abnormalities to the attention of the radiologist. In addition, CADe devices offer the potential to detect abnormalities that are otherwise not generally visible to radiologists. Stand-alone peformance testing may not be adequate to make these assessments because this type of testing cannot determine if a radiologist would recognize that a CADe device has in fact found a cancer that would otherwise go undetected. Some published studies have evaluated the performance of CADe devices based upon their ability to detect “missed cancers [78].” The literature has typically defined “missed cancers” by taking the most recent prior mammogram of a patient diagnosed with cancer by mammography and having panels of radiologists review the most recent prior mammogram. The panel of radiologists then determines in retrospect if the cancer was in fact actually visible on the prior mammogram and if it should have prompted a clinical action. The methodology of the study by Burnenne et al. [79], (and most other similar studies) can be inherently flawed for a variety of reasons: ● ●
The CADe device may not have been designed to detect “missed cancers.” The panel of radiologists may have knowledge of the precise location of the “missed cancer” when reviewing the prior mammograms thus introducing severe bias in the assessment.
20 Breast Tumor Imaging ●
541
Even if the panel radiologists only had knowledge that a cancer was in fact present, that would still introduce significant bias. To properly determine if a cancer is in fact a “missed cancer” would require that the panel of radiologists interpret the mammograms with cancers in a blinded fashion and with other mammograms intermixed so that the radiologists did not know which mammogram contained the cancer of interest (or for that matter that any of the mammograms necessarily contained cancers). Furthermore, the panel should perform the interpretations in a simulated clinical environment (with time and other constraints that are typically encountered in clinical practice).
Studies evaluating CADe clinical performance have shown conflicting results [80]. A more recent study [81] compared clinical performance and outcome of mammography with and without a CADe device in a large multi-center population (over 200,000 patients). Specificity was defined as the percentage of screening mammograms that were negative among patients who did not receive a diagnosis of breast cancer within 1 year after screening. The authors defined mammograms with BI-RADS assessment scores of 0, 4, or 5 as positive and mammograms with BI-RADS assessment scores of 1 or 2 as negative. Mammograms with a BI-RADS assessment score of 3 were defined as positive if the radiologist also recommended immediate evaluation and were defined as negative otherwise. Sensitivity was defined as the percentage of screening mammograms that were positive among patients who received a diagnosis of breast cancer within 1 year after screening. The positive predictive value was defined as the probability of a breast-cancer diagnosis within 1 year after a positive screening mammogram. Overall accuracy was assessed with the use of ROC curves. These authors found that: “the use of computer-aided detection is associated with reduced accuracy of interpretation of screening mammograms. The increased rate of biopsy with the use of computer- aided detection is not clearly associated with improved detection of invasive breast cancer.” This author would argue that the effectiveness of CADe devices can only be appropriately determined by study designs where the radiologist is in the loop—as in the study described above by Fenton et al. Stand-alone performance of a CADe device is a poor surrogate for actual clinical performance given the complex interaction between the radiologist and the CADe device output. Further work is needed to validate and quantify the actual clinical value of all mammographic CADe devices.
7
Conclusion
The field of breast imaging is an evolving and exciting one. There have been dramatic improvements in breast imaging over the last decade. As a screening test, mammography remains as the gold standard. Digital mammography has been shown to be of benefit in select patient groups and offers access to telemammography and other more advanced techniques. Ultrasound and breast magnetic resonance imaging are playing
542
D. Coll
ever more important roles in breast imaging. Newer techniques as detailed above, tomosynthesis, elastography & spectroscopy will hopefully increase the specificity of diagnosis and decrease the number of benign biopsies, which are currently performed. Further study is necessary but the future is an exciting one [82].
References 1. See e.g., JP Garne, K Aspegren, G Balldin, J Ranstam. Increasing incidence of and declining mortality frombreast carcinoma. Trends in Malmo, Sweden, 1961-1992. Cancer 1997; 79:6974; KC Chu, RE Tarone, LG Kessler, et al. Recent trends in US breast cancer incidence, survival and mortality rates. Journal of the National Cancer Institute 1996; 88:1571-1579. 2. AL Svendsen, AH Olsen, M von Euler-Chelpin, et al. Breast cancer incidence after the introduction of mammography screening: what should be expected? Cancer 2006; 106:1883-1890. 3. Warwick, L Tabar, B Vitak, SW Duffy. Time-dependent effects on survival in breast carcinoma: Results from 20 years of follow-up from the Swedish two-county study. Cancer 2004; 100:1331-1336. 4. See e.g., DB Kopans. 1992. The positive predictive value of mammography. Am. J. Roentgenology. 158:521_526; V Jackson, R Hendrick, S Feig, D Kopans. 1993. Imaging of the radiographically dense breast. Radiology. 188:297_301. 5. RD Rosenberg, et al. Performance Benchmarks for Screening Mammography. Radiology 2006; 241:55-66. 6. See e.g., DD Dershaw, RC Fleischman, L Liberman, B Deutch, AF Abramson, and L Hann. Use of digital mammography in needle localization procedures. Am. J. Roentgenol., Sep 1993; 161: 559 – 562; A practical approach to minimally invasive breast biopsy. SH Parker, F Burbank - Radiology, 1996. 7. See e.g., James JJ. The current status of digital mammography. Clin Radiol 2004; 59:1-10; Guidance for the Submission of 510(k)’s for Solid State X-ray Imaging Devices, available at http://www.fda.gov/cdrh/ode/ssxiguid.pdf (last accessed 7/3/07). 8. JM Lewin, C J. D’Orsi, R. E Hendrick et al. Clinical Comparison of Full-Field Digital Mammography and Screen-Film Mammography for Detection of Breast Cancer. Am. J. Roentgenol. Sep 2002; 179: 671-677. 9. See e.g., Yip WM, Pang SY, Yim WS, Kwok CS. ROC analysis of lesion detectability on phantoms: comparison of digital spot film mammography with conventional spot film mammography. Br J Radiol 2001;74:621-628; Comparison of Calcification Specificity in Digital Mammography Using Soft-Copy Display Versus Screen-Film Mammography. Hak Hee Kim1, Etta D. Pisano2, Elodia B. Cole2, et al. AJR 2006; 187:47-50. 10. ED Pisano et al. Diagnostic Performance of Digital versus Film Mammography for BreastCancer Screening. 2005; NEJM 353: 1773-1783. 11. Chang CHJ, Nesbit DE, Fisher DR, et al. Computed tomographic mammography using a conventional body scanner. Am J Roentgenol 1982;138:553-558. 12. Dromain C. Balleyguier C. Muller S, et al. Evaluation of tumor angiogenesis of breast carcinoma using contrast-enhanced digital mammography. AJR. American Journal of Roentgenology. 187(5):W528-37, 2006. 13. Lewin JM, Isaacs PK, Vance V, et al. Dual-energy contrast-enhanced digital subtraction mammography: feasibility. Radiology 2003;229: 261-268. 14. LT Niklason, et al. Digital Tomosynthesis in Breast Imaging. Radiology 1997; 205:399-406. 15. Rafferty E, Niklason L, Jameson-Meehan L. Breast Tomosynthesis: One view or Two. RSNA Scientific Paper 2006. 16. Rafferty EA. Tomosynthesis: New weapon in breast cancer fight. Decis Imaging Econ 2004: 17.
20 Breast Tumor Imaging
543
17. SC Chen et al. Initial Clinical Experience With Contrast-Enhanced Digital Breast Tomosynthesis. Acad Radiol 2007; 14:229–238. 18. CHJ Chang, JL Sibala, JH Gallagher et al. 1977. Computed tomography of the breast: A preliminary report. Radiology 124:827_829. 19. Chang, CH, DE Nesbit, and DR Fisher, Computed tomographic mammography using a conventional body scanner. AJR Am J Roentgenol, 1982. 138: p. 553-558. 20. JM Boone, TR Nelson, KK Lindfors, JA Siebert. 2001. Dedicated breast CT: radiation dose and image quality evaluation. Radiology. 221:657_667. 21. Boone JM Kwan AL, Seibert JA et al. Technique factors and their relationship to radiation dose in pendant geometry breast Med Phys. 32, 3767-3776 (2005). 22. ACR Appropriateness Criteria for Breast Ultrasound. 23. Szopinski KT, Pajk AM, Wysocki M, et al: Tissue harmonic imaging: utility in breast sonography. J Ultrasound Med 22:479-487, 2003. 24. Rosen EL, Soo MS: Tissue harmonic imaging sonography of breast lesions. Improved margin analysis, conspicuity, and image quality compared to conventional ultrasound. Clin Imaging 25:379-384, 2001. 25. Powers JE, Burns PN, Souquet J. Imaging instrumentation for ultrasound contrast agents. In: Nanda NC, Schlief R, Goldberg BB, eds. Advances in echo imaging using contrast enhancement. Dordrecht, The Netherlands: Kluwer Academic Publishers, 1997:139–170. 26. Jespersen SK, Wilhjelm JE, Sillesen H. Multiangle compound imaging. Ultrason Imaging1998; 20:81–102; Huber S, Wagner M, Medl M, et al.: Real time spatial compound imaging in breast ultrasound. Ultrasound Med Biol 28:155-163, 2002. 27. Sehgal CM, Arger PH, Rowling SE: Quantitative vascularity of breast masses by Doppler imaging: regional variations and diagnostic implications. J Ultrasound Med 19:427-440, 2000; Lee SW, Choi HY, Baek SY, et al.: Role of color and power Doppler imaging in differentiating between malignant and benign solid breast masses. J Clin Ultrasound 30:459-464, 2002. 28. Birdwell RL, Ikeda DM, Jeffrey SS, et al.: Preliminary experience with power Doppler imaging of solid breast masses. AJR Am J Roentgenol 169:703-707, 1997; Wilkens TH, Burke BJ, Cancelada DA, et al.: Evaluation of palpable breast masses with color Doppler sonography and gray scale imaging. J Ultrasound Med 17:109-115, 1998. 29. Chang RF, Huang SF, Moon WK, et al. Solid breast masses: neural network analysis of vascular features at three-dimensional power Doppler US for benign or malignant classification. Radiology. 2007; 243(1):56-62. 30. Caumo F, Carbognin G, Casarin A, et al. Angiosonography in suspicious breast lesions with non-diagnostic FNAC: comparison with power Doppler US. Radiol Med (Torino). 2006; 111(1):61-72. 31. J. Ophir, B. Garra and F. Kallel, et al., Elastographic imaging, Ultrasound Med Biol. 2000; 26: 23–29. 32. S.F. Levinson, M. Shinagawat and T. Satot, Sonoelastic determination of human skeletal muscle elasticity, J Biomechanics 10 (1995), pp. 1145–1154. S Catheline, F Wu and M Fink. A solution to diffraction biases in sonoelasticity: the acoustic impulse technique, J Acoust Soc Am 105 (1999), pp. 2941–2950. 33. Itoh A. Ueno E. Tohno E. Kamma H. Takahashi H. Shiina T. Yamakawa M. Matsumura T. Breast disease: clinical application of US elastography for diagnosis. Radiology. 2006; 239(2):341-50. 34. A. Thomas, T. Fischer and H. Frey et al., An advanced method of ultrasound—real-time elastography: first results in 108 patients with breast lesions, Ultrasound Obst Gyn 28 (2006), pp. 335–340. 35. Garra BS. Cespedes EI. Ophir J. Spratt SR. Zuurbier RA. Magnant CM. Pennanen MF. Elastography of breast lesions: initial clinical results. Radiology. 1997; 202(1):79-86. 36. See e.g., MJ Stoutjesdijk et al. Variability in the Description of Morphologic and Contrast Enhancement Characteristics of Breast Lesions on Magnetic Resonance Imaging. Invest Radiol 2005;40: 355–362; MD Schnall et al. Diagnostic Architectural and Dynamic Features at Breast MR Imaging: Multicenter Study. Radiology 2006; 238: 42-53.
544
D. Coll
37. American College of Radiology (ACR). ACR BI-RADS®—Mammography, ACR Breast Imaging Reporting and Data System, Breast Imaging Atlas. Reston, VA, American College of Radiology, 2003. 38. Kuhl CK, Mielcareck P, Klaschik S, et al. Dynamic breast MR imaging: are signal intensity time course data useful for differential diagnosis of enhancing lesions? Radiology 1999; 211:101–110. 39. MD Schnall et al. Diagnostic Architectural and Dynamic Features at Breast MR Imaging: Multicenter Study. Radiology 2006; 238: 42-53. 40. Schnall MD, Ikeda DM. Lesion diagnosis working group report. J Magn Reson Imaging 1999; 10(6): 982–990. 41. Christiane K. Kuhl, MD, Hans H. Schild, MD, Nuschin Morakkabati, MD. Dynamic Bilateral Contrast-enhanced MR Imaging of the Breast: Trade-off between Spatial and Temporal Resolution. Radiology 2005; 236:789–800. 42. See e.g., Davis PL, McCarty KS: Sensitivity of enhanced MRI for the detection of breast cancer: new, multicentric, residual, and recurrent. Eur Radiol 7:S289-S298, 1997; HeywangKöbrunner SH, Viehweg P, Heinig A, et al.: Contrast-enhanced MRI of the breast: accuracy, value, controversies, solutions. Eur J Radiol 24:94-108, 1997; Davis PL, Staiger MJ, Harris KB, et al.: Breast cancer measurements with magnetic resonance imaging, ultrasonography, and mammography. Breast Cancer Res Treat 37:1-9, 1996; and Yang WT, Lam WWM, Cheung H, et al.: Sonographic, magnetic resonance imaging and mammographic assessments of preoperative size of breast cancer. J Ultrasound Med 16:791-797, 1997. 43. See e.g., Drew PJ, Chatterjee S, Turnbull LW, et al.: Dynamic contrast enhanced magnetic resonance imaging of the breast is superior to triple assessment for the pre-operative detection of multifocal breast cancer. Ann Surg Oncol 6:599-603, 1999; Esserman L, Hylton N, Yassa L, et al.: Utility of magnetic resonance imaging in the management of breast cancer: evidence for improved preoperative staging. J Clin Oncol 17:110-119, 1999. 44. Liberman L, Morris EA, Dershaw DD, Abramson AF, Tan LK. MR imaging of the ipsilateral breast in women with percutaneously proven breast cancer. AJR Am J Roentgenol 2003; 180:901–910. 45. Lehman CD, Gatsonis C, Kuhl CK et al; ACRIN Trial 6667 Investigators Group. MRI evaluation of the contralateral breast in women with recently diagnosed breast cancer. N Engl J Med. 2007 Mar 29;356(13):1295-303. 46. Mandelson MT, Oestreicher N, Porter PL, et al. Breast density as a predictor of mammographic detection: comparison of interval- and screen-detected cancers. J Natl Cancer Inst 2000; 92:1081–1087. 47. Qayyum A, Birdwell RL, Daniel BL, et al. MR imaging features of infiltrating lobular carcinoma of the breast: histopathologic correlation. AJR Am J Roentgenol 2002;178:1227–1232. 48. Hilleren DJ, Andersson IT, Lindholm K, Linnell FS. Invasive lobular carcinoma: mammographic findings in a 10-year experience.Radiology 1991; 178:149–154. 49. See Hurd TC, Sneige N, Allen PK et al. Impact of extensive intraductal component on recurrence and survival in patients with stage I or II breast cancer treated with breast conservation therapy. Ann Surg Oncol 1997; 4: 119–124.[Abstract]; Schnitt SJ, Connolly JL, Harris JR et al. Pathologic predictors of early local recurrence in Stage I and II breast cancer treated by primary radiation therapy. Cancer 1984; 53: 1049–1057. 50. Van Goethem M, Schelfout K, Kersschot E, et al. MR mammography is useful in the preoperative locoregional staging of breast carcinomas with extensive intraductal component. Eur J Radiol. 2007 Jan 11; Sundararajan S, Tohno E, Kamma H, Ueno E, Minami M.Role of ultrasonography and MRI in the detection of wide intraductal component of invasive breast cancer – a prospective study. Clin Radiol. 2007 Mar;62(3):252-61. 51. Orel SG, Schnall MD. MR imaging of the breast for the detection, diagnosis, and staging of the breast cancer. Radiology 2001;220:13–30. 52. Orel SG, Schnall MD. MR imaging of the breast for the detection, diagnosis, and staging of the breast cancer. Radiology 2001;220:13–30.
20 Breast Tumor Imaging
545
53. Technology Evaluation Center (TEC). Breast MRI for detection or diagnosis of primary or recurrent breast cancer. Assessment Program 2004;vol19(No.1). 54. Role of MRI in screening women at high risk for breast cancer. J Magn Reson Imaging. 2006 Nov;24(5):964-70. 55. D Saslow et al. American Cancer Society Guidelines for Breast Screening with MRI as an Adjunct to Mammography. CA Cancer J Clin 2007; 57:75–89. 56. Negendank W. Studies of human tumors by MRS: a review. NMR Biomed 1992;5:303–324. 57. Yeung DK, Cheung HS, Tse GM. Human breast lesions: characterization with contrastenhanced in vivo proton MR spectroscopy— initial results. Radiology 2001;220: 40–46; Kvistad KA, Bakken IJ, Gribbestad IS, et al. Characterization of neoplastic and normal human breast tissues with in vivo (1)H MR spectroscopy. J Magn Reson Imaging 1999;10:159–164. 58. Jagannathan NR, Kumar M, Seenu, et al. Evaluation of total choline from in-vivo volume localized proton MR spectroscopy and its response to neoadjuvant chemotherapy in locally advanced breast cancer.Br J Cancer. 2001 Apr 20;84(8):1016-22. 59. Gambhir SS: Molecular imaging of cancer with positron emission tomography. Nat Rev Cancer 2:684-693, 2002. 60. Centers for Medicare and Medicaid Services. National Coverage Analysis (NCA), Positron Emission Tomography (FDG) for Breast Cancer (#CAG-00094N). 61. Wahl RL. Current status of PET in breast cancer imaging, staging, and therapy.Semin Roentgenol 2001;36:250–60. 62. Buck A, Schirrmeister H, Kuhn T, et al. FDG uptake in breast cancer: correlation with biological and clinical prognostic parameters. Eur J Nucl Med Mol Imaging 2002; 29:1317– 1323. 63. Avril N, Rose CA, Schelling M, et al. Breast imaging with positron emission tomography and fluorine-18 fluorodeoxyglucose: use and limitations. J Clin Oncol 2000; 18: 3495–3502. 64. A meta-analysis of FDG-PET for the evaluation of breast cancer recurrence and metastases. Carmen R. Isasi, Renee M. Moadel, and M. Donald Blaufox. Breast Cancer Research and Treatment (2005) 90: 105–112. 65. Dose J, Bleckmann C, Bachmann S, et al. Comparison of fluorodeoxyglucose positron emission tomography and “conventional diagnostic procedures” for the detection of distant metastases in breast cancer patients. Nucl Med Commun 2002;23:857–864. 66. Lonneux M, Borbath I, Berliere M, et al. The place of whole-body FDG PET for the diagnosis of distant recurrence of breast cancer. Clinical Positron Imaging 2000;3:45–49. 67. Schelling M, Avril N, Nahrig et al. Positron emission tomography using [18F]fluorodeoxyglucose for monitoring primary chemotherapy in breast cancer. J Clin Oncol 2000; 18:1689–1695. 68. Biersack HJ, Palmedo H. Locally advanced breast cancer: Is PET useful for monitoring primary chemotherapy? J Nucl Med 2003; 44:1815–1817. 69. Adler LP, Weinberg IN, Bradbury MS et al. The Breast Journal, Volume 9, Number 3, 2003 163–166. 70. Leonard Berlin, Malpractice Issues in Radiology: Possessing Ordinary Knowledge, 166 AJR 1027 (1996). 71. Leonard Berlin, Malpractice Issues in Radiology: Errors in Judgment, 166 AJR 1259 (1996). 72. Leonard Berlin, Malpractice Issues in Radiology: Perceptual Errors, 167 AJR 587, 589 (1997). 73. Leonard Berlin, Malpractice Issues in Radiology: Picture Archiving and Communications Systems (PACS) and the Loss of Patient Examination Records, 176 AJR 1381 (2001). 74. BJ Erickson & B Bartholmai, Computer-Aided Detection and Diagnosis at the Start of the Third Millennium, 15(2) Journal of Digital Imaging 59 (2002). 75. RS Rana, et al., Independent Evaluation of Computer Classification of Malignant and Benign Calcifications in Full-Field Digital Mammograms. 14 Acad Radiol 363 (2007). 76. EA Krupinski. Computer-aided Detection in Clinical Environment: Benefits and Challenges for Radiologists. 231 Radiology 7 (2004).
546
D. Coll
77. Such psychological effects could be measured by conducting studies where radiologists interpret the same mammograms using the same CADe device at three separate reading sessions separated in time. The radiologists could be told that they are testing three separate CADe devices: one CADe device designed to detect BI-RADS 3 or higher potential abnormalities; a second CADe device designed to only detect BI-RADS 5 potential abnormalities; and a third CADe device designed to only detect actual malignancies. At each reading session, the radiologists would interpret 1/3 of the mammograms with “each CADe device.” By having three reading sessions separated in time to avoid recall bias, each radiologist would ultimately interpret each mammogram with the same CADe device but with a different instruction as to the design of the CADe device. 78. EA Krupinski, et al., A Perceptually Based Method for Enhancing Pulmonary Nodule Recognition. 28(4) Investigative Radiology 289 (1993). 79. See e.g., RL Birdwell et al., Mammographic Characteristics of 115 Missed Cancers Later Detected with Screening Mammography and the Potential Utility of Computer-aided Detection. 219 Radiology 192 (2001); LJW Burhenne, Potential Contribution of Computeraided Detection to the Sensitivity of Screening Mammography. 215 Radiology 554 (2000); RF Brem et al., Improvement in Sensitivity of Screening Mammography with Computer-Aided Detection: A Multiinstitutional Trial. 181 AJR 687 (2003). 80. LJW Burhenne, Potential Contribution of Computer-aided Detection to the Sensitivity of Screening Mammography. 215 Radiology 554 (2000). 81. See RM Nishikawa and M Kallergi. Point/Counterpoint: Computer-aided Detection, in its Present Form, is not an Effective Aid for Screening Mammography. 33 Med Phys 811 (2006). 82. JJ Fenton, et al., Influence of Computer-Aided Detection on Performance of Screening Mammography. 356 NEJM 1399 (2007).
Index
A Abscess drainage, 507 Absolute percentage washout (APW), 321, 323 ACE. See Acetate Acetate (ACE), 86 Acinar cell neoplasms, 244 Acquired immune deficiency syndrome (AIDS), 188 ACTH. See Adrenocorticotropic hormone Acute benign compression fractures, 48–50 Acute lymphoblastic leukemia (ALL), 154, 479–480, 481 Acute myeloid leukemia (AML), 479–480 ADC. See Apparent diffusion coefficient Adenocarcinoma, 123, 232 CT in, 238 MRI in, 238 pancreatic ductal, 230, 241 Adenoid cystic carcinoma, 101 Adenomas, 256, 325 adrenal, 321 lipid-poor adrenal, 322 tubular, 256, 259 tubulovillous, 256 villous, 256, 259 Adenopathy, 311 Adenoviruses, 501 Adrenal adenoma, 321 lipid-poor, 322 Adrenal biopsy, 326–328 Adrenal carcinoma, 322 Adrenal cysts, 326 Adrenal glands, 128 Adrenal hemorrhagic cyst, 324 Adrenal masses, 319, 496 CT in, 319–323 MRI for, 323–325 PET in, 325–326
Adrenal metastasis, 129 from lung cancer, 327 Adrenal myelolipoma, 320 Adrenocortical carcinomas, 320 Adrenocorticotropic hormone (ACTH), 150 AFP. See Alpha-fetoprotein AIDS. See Acquired immune deficiency syndrome AJCC. See American Joint Commission on Cancer Albumin, 526 ALL. See Acute lymphoblastic leukemia Alpha-fetoprotein (AFP), 157, 349 American College of Radiology, 523, 527, 528 American Joint Commission on Cancer (AJCC), 94, 306 Amino acids, 476 metabolism, 80 radiolabeled, 32–33 transport, 80 AML. See Acute myeloid leukemia Amyloidosis, 394 Aneurysmal bone cyst, 55–56 Angiogenesis imaging, 246–247 tumor induced, 525 Angiography, 292 catheter, 180 CT, 473–474 MRI, 473–474 Angiomatosis, 438 Angiosarcoma, 178, 185–186, 289, 439–441 CT of, 186, 440–441 MRI of, 186, 440–441 Animal models, 30 Ann Arbor staging system, 154 Annular carcinomas, 262 Anterior spread, of NPC, 96 Antoni A, 163
547
548 Antoni B, 163 Apparent diffusion coefficient (ADC), 16, 23–24, 52, 203 tumor types and, 18 Appendicular skeleton, 404 APW. See Absolute percentage washout; Relative percentage washout Argon gas, 417 Arterial spin labeling (ASL), 11, 12 Asbestos exposure, 137 ASL. See Arterial spin labeling Astrocytoma, 5, 8, 59–60 anaplastic, 8, 18 cerebellar, 483 creatine in, 21 grading of, 12–13 intramedullary cervical, 60 low-grade fibrillary, 8 pilocytic, 8 Axial skeleton, 404, 406
B BAC. See Bronchioloalveolar cell carcinoma BALT. See Bronchus-associated lymphoid tissue Barium, 437 Barium enema. See Double contrast barium enema; Single contrast barium enema Barium swallow, 180 BBB. See Blood-brain barrier BDEPI. See Breathing Dynamic Echo Planar Benign cardiac tumors, 188–193 echocardiography for, 191–193 imaging features of, 193 MRI for, 191–193 pathologic features of, 192 Benign fibrous histiocytoma, 434–435 Benign prostatic hyperplasia (BPH), 332 beta-HCG. See Beta-human chorionioc gonadotropin Beta-human chorionioc gonadotropin (beta-HCG), 349 Bilateral tonsillar masses, 104 Biliary interventions, 219–222 Biliary lesions, 213–216 Biliary obstruction, 214–216, 502–503 Biopsy, 493–497 adrenal, 326–328 complications, 495–497 of mediastinal tumors, 169 percutaneous, 494 BI-RADS. See Breast Imaging Reporting and Data System
Index Bladder malignancies, 309–314 CT in, 309–311 FDG in, 314 imaging of, 299 lymph nodes in, 310 MRI in, 311 PET in, 309, 314 Bleomycin, 507 Blood-brain barrier (BBB), 7, 28, 29 Blood flow, 470 Blood oxygenation level-dependant (BOLD), 26 influence of tumors on, 26–27 BOLD. See Blood oxygenation level-dependant Bone cysts, 55–56 Bone erosion, 98, 99 Bone marrow, 386 edema, 388 hematopoietic, 407 scintigraphy, 406 Bone metastasis, 45, 344–346, 402–409 in appendicular skeleton, 404 in axial skeleton, 404 common appearance, 404 CT in, 406–407 imaging modalities, 405–408 MRI in, 345–346, 407 osteoblastic, 403 osteolytic, 403 painful, 409 PET in, 408 radiographic appearance of, 403–405 summary of, 409 of unknown origin, 408–409 whole-body MRI in, 407–408 Bonferroni correction, 26 Bowel obstruction, 263 Box plots, 52 BPH. See Benign prostatic hyperplasia Brain metastasis, 127 Brain tumors general features of, 4–7 grading of, 12–15 Breast cancer, 515–516 high risk populations, 534 Breast computed tomography (Breast CT), 522–523 advantages of, 523 cone-beam, 522 Breast cysts, 525 Breast Imaging Reporting and Data System (BI-RADS), 528, 529, 538–539, 541
Index Breast magnetic resonance imaging (Breast MRI), 527–535 controversies, 528–531 dense patterns in, 532 disease evaluation in, 532 in high risk populations, 534 imaging features, 528–531 implant assessment, 533–534 indications for, 532–534 initial enhancement pattern in, 529 invasive lobular carcinoma in, 532–533 morphologic features in, 529 positive axillary nodes, 533 scar versus tumor recurrence, 533 technique, 528 Breast positron emission tomography (Breast PET), 535–541 clinical use of, 536 computer aided detection and diagnosis, 538–541 in metastases detection, 536–537 technology update, 537–538 treatment assessment, 537 Breast tomosynthesis, 520–521 exposure angles in, 520 Breast ultrasound, 523–527 technical advances in, 523 Breathing Dynamic Echo Planar (BDEPI), 127 Bronchioloalveolar cell carcinoma (BAC), 122 Bronchogenic carcinoma, 121–123 Bronchus-associated lymphoid tissue (BALT), 455 Buccal carcinoma, 100
C CADe. See Computer-aided detection CADx. See Computer-aided diagnosis Calcification, 6, 166, 443 Carbon, 339 Carcinoembryonic antigen (CEA), 270 Carcinoid tumors, 131 CT of, 131 Carcinoma. See also Colorectal carcinoma; Squamous cell carcinoma adenoid cystic, 101 adrenal, 322 annular, 262 bronchogenic, 121–123 buccal, 100 floor of mouth, 102 gallbladder, 216–217 gingiva, 100–101 hard palate, 100–101
549 hypopharyngeal, 106 infiltrating ductal, 525, 529 invasive ductal, 530 invasive lobular, 532–533 lip, 100, 102 lung, 9 oral cavity, 102 oropharyngeal, 105 papillary serous, 284, 285 sinonasal carcinoma, 111–112 soft palate, 103 thymic, 150 tongue base, 103–104 tonsil, 103 Carcinomatosis, 62 lymphangitic, 136–137 Cardiac teratomas, 190–191 Cardiac tumors benign, 188–193 chest radiography of, 180 classification of, 178 clinical presentation of, 178–179 CT for, 183–184 differential diagnosis in, 179 echocardiography, 181 fluoroscopy of, 180 imaging of, 179–183 level of evidence for imaging of, 183 malignant, 183–188 metastatic, 183–184 MRI of, 181–182 C-arms, 394 Carney’s complex, 189 Catheter angiography, 180 CBF. See Cerebral blood flow CBV. See Cerebral blood volume CCCS. See Clear cell chondrosarcoma CD99, 376 CEA. See Carcinoembryonic antigen Cecum, 270 CEDM. See Contrast enhanced digital mammography Celiac ganglion neurolysis, 506 Cellular proliferation, 84–86 Centers for Medicare and Medicaid Services, 537 Central nervous system, 481–484 Central venous catheters (CVC), 497, 498 Cerebral blood flow (CBF), 32, 478 Cerebral blood volume (CBV), 32 Cerebral oxygen metabolic rate (CMRO2), 32 Cerebrospinal fluid (CSF), 6, 477 Cervical cancer, 503 CEUS. See Contrast enhanced ultrasound
550 Chemical shift selective excitation (CHESS), 20 Chemodectomas, 161 Chemoembolization, 498, 499 Chemotherapy, 95 in Ewing’s sarcoma, 415 in skeletal malignancies, 380 Cheson’s criteria, 463–464 CHESS. See Chemical shift selective excitation Chest radiography, cardiac tumor, 180 Chest wall invasion, 126 Children’s Oncology Group (COG), 486 Cholangiocarcinoma, 213–214 hilar, 213 intrahepatic, 215 Cholangiography, 215 Cholecystostomy, 221 Choline, 21, 339, 346, 476 Chondroblastic osteosarcoma, 370 Chondroblastoma, 388 Chondrocytes, 385 Chondroid matrix, 385, 388 Chondrosarcoma, 53, 178, 381–387, 388–390 clear cell, 387–388 conventional intramedullary, 382–387 CT in, 382 dedifferentiated, 391–392 high-grade, 386 intermediate grade, 384 juxtacortical, 388–390 low-grade, 383 mesenchymal, 391 MRI in, 53, 382 PET in, 382 radiographs in, 382, 383 skeletal location in, 382 skeletal myxoid, 390 Chordoma, 52 Chronic liver disease, 206–213 Cine-steady state free precession (SSFP), 190 Cirrhosis, 206 imaging of, 208–213 US in, 212–213 Clear cell chondrosarcoma (CCCS), 387–388 MRI in, 387–388 11 C Methionine (MET), 80–83 co-registered, 82 FDG-PET v., 81 uptake, 81 CMRO2. See Cerebral oxygen metabolic rate CNR. See Contrast-to-noise ratio Codman triangle pattern, 388, 414 COG. See Children’s Oncology Group
Index Collision tumors, 327 Colon cancer, liver metastases from, 202 Colonic polyp detection, 258–260 CT in, 259 Colonography, 257, 259, 260, 261 Colorectal carcinoma clinical presentation of, 257 complications, 263 CT in, 257, 264–265, 272–275 detection, 260–263 FDG in, 275 imaging detection in, 257–258 large bowel obstruction in, 263 lymph nodes in, 268–269 metastases from, 269–272 MRI in, 257, 264–265 pathophysiology of, 256 PET in, 257, 272–275 post-treatment follow-up, 275 screening, 256–257 staging, 263–266 Color flow, 524–526 Compression fractures, 48–50 clinical features of, 51 osteoporotic, 51 Computed tomography (CT), 3, 44, 48, 56–57, 93, 199. See also High resolution CT acinar cell neoplasms on, 244 adenocarcinoma, 238 of adrenal masses, 319–323 advances in, 313–314 angiography in, 473–474 of angiosarcoma, 186, 440–441 axial, 137 in bladder cancer, 309–310 in bone metastasis, 406–407 breast, 522–523 of carcinoid tumors, 131 of cardiac metastases, 184 for cardiac tumors, 182–183 of cecum, 270 central nervous system, 481–484 in colonic polyp detection, 259 in colorectal carcinoma, 257, 264–265, 272–275 contrast-enhanced, 375 in dedifferentiated chondrosarcoma, 392 in Ewing’s sarcoma, 414 in gallbladder carcinoma, 216–217 of GCTs, 159 in HCC, 206 in Hodgkin’s disease, 460, 472 leiomyosarcoma, 437 of liposarcoma, 288, 432
Index in liver lesion follow-up, 219 in liver metastases, 269–272 in lung metastasis, 135 of lymph nodes, 268, 457–461 of lymphoma, 459–461 mammogram v., 522 in mediastinal lymphoma, 155 mediastinal mass, 168 of mediastinal tumors, 146–148, 151–152 of melanoma, 446 of MFH, 435 in multiple myeloma, 396, 397 of neuroblastomas, 474 of neurofibromas, 162–163 in NHL, 458, 459, 460 in osteosarcoma, 373 pancreatic tumors, 236–237 perfusion, 247 in peritoneal tumors, 282 pitfalls of, 311 in pleural metastases, 139 in prostate cancer, 334–335 in prostate cancer staging, 341 of RCC, 303 of rectal cancer, 266–267 in renal tumors, 300, 301–305, 307 in retroperitoneal tumors, 282 of rhabdomyosarcoma, 436 of schwannomas, 162–163 in sinonasal carcinoma, 112 in skeletal tumors, 367 in soft tissue tumors, 426, 429 of synovial sarcomas, 443 in testicular cancer, 351, 353–355 of thymolipomas, 151 of thyroid masses, 166 Computer-aided detection (CADe), 538–541 evaluating, 541 limitations of, 539 performance of, 539–540 regions marked by, 540 Computer-aided diagnosis (CADx), 538–541 Contrast enhanced digital mammography (CEDM), 519–520, 521 Contrast enhanced ultrasound (CEUS), 213 Contrast enhancement, 8 Contrast-to-noise ratio (CNR), 7 Conventional intramedullary chondrosarcoma, 382–387 low-grade, 383 Conventional therapy, 62 Cord compression, 50–52 Corticospinal tract (CST), 19 Cotswold modifications, 154
551 C-reactive protein, 393 Creatine (tCr), 20, 21, 476 Cricohyoidopexy, 107 Cricoidectomy, 110 Cryoablation, 417 CSF. See Cerebrospinal fluid CST. See Corticospinal tract CT. See Computed tomography 11 C Tyrosine, 80 CVC. See Central venous catheters Cyclotrons, 68 Cystadenocarcinoma, 200 Cystic lesions, 233 Cystic metastases, 200 Cystic renal tumors, 304 MRI in, 305 Cysts adrenal, 326 breast, 525
D DCBE. See Double contrast barium enema DCE. See Dynamic contrast-enhanced DCIS. See Ductal carcinoma in situ Dedifferentiated chondrosarcoma, 391–392 CT in, 392 MRI in, 392 Deep venous thrombosis (DVT), 507 3′-Deoxy-3′18F-fluorothymidine (FLT), 70, 85 in pancreatic tumors, 242–243 Desmoid tumor, 433 Desmoplastic small round cell tumor, 284–285, 286 clinical features of, 284 imaging features of, 285 Diffusion tensor imaging (DTI), 17, 18, 27, 57 Diffusion-weighted imaging (DWI), 16–18, 50, 57, 479 basic principles of, 16 Digital Mammographic Imaging Screening Trial (DMIST), 517–518 Digital mammography, 516–523 advanced applications of, 519–520 background, 516–517 contrast enhanced, 519–520 guidelines for, 517–519 Digital rectal examination, 340 3,4-dihydroxy-6-18F-fluoro-1-phenylalanine (FDOPA), 70, 80, 84 Distant malignant biliary obstruction, 214–216 DMIST. See Digital Mammographic Imaging Screening Trial DN. See Dysplastic nodules
552 Double contrast barium enema (DCBE), 257, 262 Doxorubicin, 499 DSC. See Dynamic susceptibility contrast DTI. See Diffusion tensor imaging Ductal carcinoma in situ (DCIS), 533 Durie/Salmon PLUS Staging System, 394 DVT. See Deep venous thrombosis DWI. See Diffusion-weighted imaging Dynamic contrast-enhanced (DCE), 11, 12, 14, 247 MRI, 338 Dynamic susceptibility contrast (DSC), 11, 14 Dysplastic nodules (DN), 208 MRI of, 208 with malignant foci, 208–209
E Echocardiography for benign cardiac tumors, 191–193 cardiac tumor, 181 EES. See Extravascular extracellular space EGCCCG. See European Germ Cell Cancer Consensus Group EGFR. See Endothelial growth factor receptors eGFR. See Estimated glomerular filtration rate EGGCT. See Extragonadal germ cell tumors EIC. See Extensive intraductal component EKG. See Electrocardiograph Elastography, 526–527 Electrocardiograph (EKG), 182 Embolization, 498–500 Enchondroma, 385, 386 Enchondromatosis, 53 Endobronchial metastases, 134 Endocardial lesions, 184 Endometrial cancer, 205 Endorectal ultrasound, 266, 267 Endoscopic retrograde cholangiopancreatography (ERCP), 214, 215, 216, 220, 233 in pancreatic tumors, 245 Endoscopic ultrasound (EUS), 234–235 acinar cell neoplasms on, 244 FNA, 463 in pancreatic tumors, 243–244 Endoscopy, 95 Endothelial growth factor receptors (EGFR), 7 Enema. See Double contrast barium enema; Single contrast barium enema Ependymoma, 57–59 intramedullary cervical, 58 myxopapillary, 59
Index Epidermoid cysts, 10 Epiphysis, 388 Epstein-Barr Virus, 150, 501 ERCP. See Endoscopic retrograde cholangiopancreatography Esthesioneuroblastoma, 112 Estimated glomerular filtration rate (eGFR), 183 Ethiodized oil, 499 Ethmoidal soft tissue mass, 113 European Germ Cell Cancer Consensus Group (EGCCCG), 351, 354 EUS. See Endoscopic ultrasound Ewing’s sarcoma, 413–415 chemotherapy in, 415 CT in, 414 FDG in, 414 MRI in, 414 pediatric, 486–487 Experimental brain tumor models, 30 Extensive intraductal component (EIC), 533 Extra-axial tumors, 8–9, 10–11 intra-axial v., 7 Extradural tumors, 44–56 benign, 53–56 malignant, 44–53 Extragonadal germ cell tumors (EGGCT), 295, 296, 354 clinical features, 295 imaging features, 295 Extramedullary plasmacytoma, 402 Extramedullary tumors, 61–62 Extravascular extracellular space (EES), 12
F 2-18F Tyrosine, 80 FA. See Fractional anisotropy False positive rates (FPR), 537 False vocal cords, 107, 108 Familial adenomatous polyposis syndrome (FAP), 255 FAP. See Familial adenomatous polyposis syndrome Fast spin echo (FSE), 207 Fatty tumors, 429–432 FAZA. See 1-(5-fluoro 5-deoxy-alpha-Darabinofuranosyl)-2-nitroimidazole FCH. See 18F labeled choline FDOPA. See 3,4-dihydroxy-6-18F-fluoro-1phenylalanine Fecal occult blood test (FOBT), 257 Femurs, 482 Ferumoxtran, 355
Index FET. See 0-(2-[18F] fluor ethyl)-L-tyrosine FFDM. See Full field digital mammography FGF. See Fibroblast growth factor Fibroblast growth factor (FGF), 247 Fibrohistiocytic tumors, 434–435 Fibrolamellar HCC, 211–212 Fibroma, 178 Fibromatoses, 433 Fibrosarcoma, 186–188, 434 Fibrous tumors, 433–434 Fine needle aspiration (FNA), 234, 415, 416 EUS, 463 Fistula, 263 Five-year survival rates, 125 18 F labeled choline (FCH), 86 FLAIR. See Fluid-attenuated inversion recovery Floor of mouth, 99 imaging checklist, 102 FLT. See 3′-Deoxy-3′18F-fluorothymidine Fluid-attenuated inversion recovery (FLAIR), 5, 6, 8, 57 0-(2-[18F] fluor ethyl)-L-tyrosine (FET), 70, 80, 83–84 1-(5-fluoro 5-deoxy-alpha-Darabinofuranosyl)-2-nitroimidazole (FAZA), 86 Fluoro-2-deoxy-D-glucose (FDG), 32, 70, 72–73, 85, 128, 136, 137–138, 148, 214, 269 in adrenal masses, 326 in bladder cancer, 314 in bone metastatic disease, 346 in brain tumor evaluation, 70–71 breast, 535–541 in CNS lymphoma, 79 in colorectal carcinoma, 275 co-registered, 74, 77 in Hodgkin’s disease, 461 in initial diagnosis, 71 in liver metastases, 204 in lymph node staging, 344 in lymphoma, 79, 461–462, 464 of melanoma, 446 in metastatic brain lesions, 78–79 MET-PET v., 81 in multiple myeloma, 400 in nerve sheath tumors, 442 in pancreatic tumors, 239–242 in pediatric malignancies, 470–471 in pheochromocytomas, 326 in post-therapy, 75 in prognosis assessment, 74–75 in renal tumors, 306, 308
553 in residual tumor detection, 76 serial co-registered, 75–76 of soft tissue tumors, 427–429 in testicular cancer, 32 uptake of, 241 of various tumors, 79–80 18 F Fluoromisonidazole (FMISO), 86 Fluoroscopy, cardiac tumor, 180 FMISO. See 18F Fluoromisonidazole fMRI. See Functional magnetic resonance imaging FNA. See Fine needle aspiration FNH. See Focal nodular hyperplasia FOBT. See Fecal occult blood test Focal nodular hyperplasia (FNH), 201, 203, 212 FPR. See False positive rates Fractional anisotropy (FA), 16, 52 maps, 17 FSE. See Fast spin echo Full field digital mammography (FFDM), 516 Functional magnetic resonance imaging (fMRI), 24–27 tractography and, 27
G Ga-67, 461 Gadolinium, 126, 127, 182, 304, 312, 338, 397, 427, 462, 478, 529 Gallbladder carcinoma CT in, 216–217 US in, 216 Gallium-67, 147, 148 Gallium citrate, 400 GALT. See Gut-associated lymphoid tissue Gamma detectors, 246 Gamma radiation, 500 Gangliogliomas, 60 Ganglioneuroblastomas, 163–165 staging of, 164 symptoms of, 164 therapy, 164 Ganglioneuromas, 161, 163–165 Gastrinoma triangle, 246 Gastrointestinal obstruction, 504 Gastrointestinal stromal tumors (GIST), 288, 289, 292 Gastrojejunostomy, 504, 505 Gastrostomy, 504 GBM. See Glioblastoma multiforme GCTs. See Germ cell tumors Gd-DTPA, 85 Gelfoam, 498
554 Gene therapy, 501 Germ cell tumors (GCTs). See also Nonseminomatous malignant germ cell tumors classification of, 353 extragonadal, 295, 296 imaging features of, 159–160 mediastinal, 157–159 seminomatous, 349 staging of, 159 testicular, 356, 357–358 Giant cell tumor, 54–55, 56, 387 MRI of, 54 Gingiva carcinoma, 100–101 GIST. See Gastrointestinal stromal tumors Glioblastoma multiforme (GBM), 3, 8, 18–19, 67, 72, 75, 80 Gliomas, 3, 15 differentiation of, 15, 23 grading, 13, 14, 16–17, 22–23 high-grade, 5, 15, 22, 31 low-grade, 6, 31 multiparametric analysis of, 24, 25 parametric maps from, 15, 17 proton magnetic resonance spectrum from, 22 thalamic, 73 Gliomatosis, 10 Glossectomy, 99 Glottis, 109–110 TNM staging system for, 110 Glucose, 204 Glut, 204 GLUT1, 32 GLUT3, 32 Glutamate, 20, 21 Glutamine, 20, 21 Glutathione, 21 Gnathic osteosarcoma, 376 Gorlin syndrome, 189 Gradient echo (GRE), 12 GRE. See Gradient echo Gut-associated lymphoid tissue (GALT), 455 Gynecomastia, 348
H HAP. See Hepatic arterial phase Hard palate carcinoma, 100–101 HCC. See Hepatocellular carcinoma hCG. See Human chorionic gonadotropin HD. See Hodgkin’s disease Head neoplasms imaging goals, 94
Index specific locations, 95–113 technique, 94 tumor resectability, 94–95 Hemangioblastomas, 10, 60–61 Hemangioma, 178, 201, 438–439 capillary, 438 cavernous, 204, 438 MRI of, 438–439 strawberry type, 438 typical, 54 vertebral, 53–54 Hematomas, 427 Hematuria, 496 Hemoptysis, 496 Hemorrhage, 6 Hemosiderin, 388 Hepatic arterial phase (HAP), 207 Hepatic endocrine metastases, 499 Hepatic metastases, 199–206 from endometrial cancer, 205 Hepatobiliary cancer, image-guided interventions in, 217–219 Hepatocellular carcinoma (HCC), 199, 206–213, 499 CT in, 206 fibrolamellar, 211–212 large, 210 MDCT of, 209 MRI of, 210 percutaneous alcohol ablation, 218 small, 210 tumor destruction in, 217 with portal vein thrombosis, 211 Hereditary Multiple Exostoses (HME), 382 Hereditary non-polyposis colorectal cancer (HNPCC), 255 High-grade surface osteosarcoma, 377–378 High resolution CT (HRCT), 122, 380 Histiocytoma, 178. See also Malignant fibrous histiocytoma benign fibrous, 434–435 Histiocytosis, Langerhans Cell, 368 Histologic sampling, interventional techniques for, 167–169 HME. See Hereditary Multiple Exostoses HNPCC. See Hereditary non-polyposis colorectal cancer Hodgkin’s disease (HD), 132–133, 412, 457 CT in, 460, 472 FDG in, 461 mediastinum in, 153 MRI in, 462–463 spread of, 458 staging of, 458
Index Holmium, 500 Horner’s syndrome, 123, 161 Hounsfield units (HU), 320 HRCT. See High resolution CT HU. See Hounsfield units Human chorionic gonadotropin (hCG), 157 Hydatid cysts, 179 Hypercalcemia, 403 Hypergammaglobulinemia, 232 Hypermetabolism, 72, 74 Hypopharynx, 104–106 carcinoma, 106 SCCA, 105 TNM staging system for, 106 Hypoxia, 86–87
I IDUS. See Intraductal ultrasound IL. See Interleukins ILC. See Invasive lobular carcinoma Imatinib, 200 Implant assessment, 533–534 Infection, 498 Inferior spread, of NPC, 96 Inferior vena cava (IVC), 507 Insulinoma, 237 Interleukins (IL), 247 Interventional Radiology (IR), 493, 497–501 biliary obstruction and, 502–503 central venous catheters, 497 in complications treatment, 501–507 decompression, 504 embolization, 498–500 gene therapy, 501 infection, 498 pain, 505–506 pleural space, 506–507 renal obstruction, 503–504 thermal ablation, 500–501 thrombosis, 497–498 upper gastrointestinal obstruction, 504 venous thromboembolism, 507 Intra-axial tumors, 7–8, 10 extra-axial v., 7 Intracortical osteosarcoma, 376 Intraductal papillary mucinous neoplasm (IPMN), 230, 233, 237 Intraductal ultrasound (IDUS), 215 in pancreatic tumors, 245 Intradural tumors, 56–62 Intrahepatic cholangiocarcinoma, 215 Intramedullary tumors, 56–57 Intravenous urography (IVU), 312
555 Invasive lobular carcinoma (ILC), 532–533 IPMN. See Intraductal papillary mucinous neoplasm IR. See Interventional Radiology IVC. See Inferior vena cava IVU. See Intravenous urography
J Juxtacortical chondrosarcoma, 388–390
K Kaplan-Meier survival analysis, 83 Karnofsky performance status (KPS), 81 Ki-1, 153 Ki-67 proliferation index, 81, 87 Klatskin tumor, 213, 214 KPS. See Karnofsky performance status
L L4 lesion, vertebroplasty for, 63 Lactate, 21 Lactate dehydrogenase (LDH), 157 Lactose dehydrogenase (LDH), 393 Langerhans Cell Histiocytosis, 368 Laparoscopy, 235 Large bowel obstruction, 263 Laryngectomy, 107, 110 supracricoid, 107 Laryngopharyngectomy, 106 Larynx, 106–107 Laser ablation, 501 Lateral spread, of NPC, 96 LDH. See Lactate dehydrogenase, Lactose dehydrogenase Leiomyomas, 436–437 MRI of, 437 Leiomyosarcoma, 186–188, 200, 291–292 CT of, 437 imaging features, 292 US in, 292 Leukemia, 47–48, 394 pediatric, 479–480 Leu-M1, 153 Levovist, 456 Leydig cell tumors, 348 Linitis plastica, 313 Lip carcinoma, 100 imaging checklist, 102 Lipids, 22 Lipid synthesis, 86
556 Lipoma, 178, 429–431 MRI in, 430 US in, 430 Liposarcoma, 186–188, 287–288, 431–432 clinical features of, 287, 290 CT of, 288, 432 imaging features, 288, 291 MRI of, 432 Liver function tests, 270 Liver metastasis, 127, 199–206, 235 calcified solitary, 201 from colon cancer, 202 CT of, 269–272 FDG in, 204 hypervascular, 201 MRI in, 203, 269–272 US of, 269–272 Lodwick classification system, 368 Lumpectomy, 531 Lung cancer, 496 adrenal metastasis from, 327 lymph node staging in, 127 metastatic disease, 127–128, 135 nodules, 495 primary tumor imaging in, 126–127 TNM staging of, 124 Lymphadenopathy, 273, 352, 426, 456 Lymphangiography, 463 Lymphangitic carcinomatosis, 136–137 Lymphedema, 439 Lymph nodes in bladder cancer, 310 colorectal carcinoma in, 268–269 CT of, 268 detection, 268–269 imaging, 456–464 in lung cancer staging, 127 metastasis and, 151, 352 MRI of, 268 normal size, 345 sentinel, 355–356 staging, 344 Lymphoid tissue, 455 Lymphomas, 10, 46–47, 61, 103, 132–133, 232, 394, 471. See also Hodgkin’s disease; Non-Hodgkin’s lymphoma; Primary bone lymphoma B-cell, 47, 154 CT of, 459–461 FDG-PET study of, 79, 461–462, 464 interventional radiology in, 463 intracardiac, 178 mediastinal, 153, 155 MRI in, 462–463
Index pediatric, 479–480 PET in, 461–462 post-treatment follow-up imaging, 463–464 primary cardiac, 188 radiographs of, 410–411 staging of, 458 Lymphomatous lesions, 457 Lytic lesions, 45, 397, 405
M Macroglobulinemia, 394 Maffucci syndrome, 382 Magnetic resonance angiography (MRA), 425 Magnetic resonance cholangiopancreatography (MRCP), 214, 216, 233 acinar cell neoplasms on, 244 secretin-enhanced, 240 Magnetic resonance imaging (MRI), 3, 93, 199 acute benign compression fractures and, 48–50 adenocarcinoma, 238 for adrenal masses, 323–325 advances in, 313–314 anatomical, 335 angiography, 473–474 of angiosarcoma, 186, 440–441 for benign cardiac tumors, 191–193 in bladder malignancies, 311–312 in bone metastatic disease, 345–346, 407 breast, 527–535 of cardiac metastases, 184 cardiac tumor, 181–182 in CCCS, 387 chemical shift, 49–50 of chondrosarcoma, 53 CINE, 181 CNS, 481–484 in colorectal carcinoma, 257, 264–265 contraindications for, 181 conventional, 4–11 co-registered, 77 CT in, 303 in cystic renal tumors, 305 in dedifferentiated chondrosarcoma, 392 development of, 4 in diagnosis, 4–28 of dysplastic nodules, 208 endorectal, 341–342, 343 in Ewing’s sarcoma, 414 in experimental brain tumor models, 30 of GCTs, 159–160
Index of giant cell tumors, 54 HCCs on, 210 of hemangioma, 438–439 of Hodgkin’s disease, 462–463 leiomyomas on, 437 in lipomas, 430 of liposarcoma, 432 in liver lesion follow-up, 219 in liver metastases, 203, 269 of lymph nodes, 268 of lymphoma, 462–463 in marrow, 386 of mediastinal tumors, 147, 151–152 of melanoma, 446 of mesenchymal chondrosarcoma, 391 in metastasis detection, 44 in multiple myeloma, 395, 396, 397 of neurofibromas, 162–163 of NHL, 462–463 in osteosarcoma, 373 in pancreatic tumors, 238–239 perfusion, 478–479 in peritoneal tumors, 282 in pleural metastases, 139 of primary bone lymphoma, 412 in prostate cancer, 338 in prostate cancer staging, 341 protocol, 182 PVT on, 211 RCC in, 303 of rectal cancer, 266–267 in renal tumors, 300, 304, 307 in retroperitoneal tumors, 282 of rhabdomyosarcoma, 186, 436 of schwannomas, 162–163 in sinonasal carcinoma, 112 in skeletal tumors, 367, 416 of soft tissue tumors, 426–427 of spinal cord, 57 in stereotactic biopsy, 28–30 of synovial sarcomas, 443, 444 T2-weighted, 51–52 in testicular cancer, 354, 355 of thymolipomas, 151 of thyroid masses, 166 whole-body, 345, 399, 407–408, 431, 471–473, 480 Magnetic resonance spectroscopy (MRS), 19–23, 475–477 basic principle of, 19–20 breast, 534–535 in high-grade gliomas, 22 in normal brain, 20 Magnetic resonance urography (MRU), 475
557 Magnetization transfer (MT), 57 Male reproductive system, 331–332 Malignant fibrous histiocytoma (MFH), 285–287, 290, 392, 435 clinical features of, 285–286, 293 CT of, 435 imaging features of, 286–287, 293 Malignant nerve sheath tumors (MNST), 161 Malignant peripheral nerve sheath tumor (MPNST), 163 Malignant pleural effusions (MPE), 506 Malignant pleural mesothelioma (MPM), 137–138 MALT. See Mucosa-associated lymphoid tissue Mammography, 516 contrast-enhanced digital, 519–520 CT v., 522 digital, 516–523 full field digital, 516–517, 518–519 indeterminate, 534 normal screening, 518 screen film, 517, 518–519 Mammography Quality Standards Act (MQSA), 517 Mandibulectomy, 99 Masaoka staging system, 149 MCNs. See Mucinous cystic neoplasms MDCT. See Multidetector CT Mean transit time (MTT), 478 Mediastinal lymphoma, 153, 155 CT in, 155 PET in, 156–157 Mediastinal seminoma, 158 Mediastinal teratoma, 157–158, 160 Mediastinal thymoma, 147 Mediastinal tumors, 146 biopsy of, 169 CT of, 146–148, 151–152, 168 features, 151–152, 154–157 germ cell, 157–159 histologic sampling, 167–169 imaging of, 146–148 metastases, 167 MRI of, 147, 151–152 PET of, 147–148 Medulloblastoma, 5 Melanoma, 200, 444–446 CT of, 446 FDG of, 446 metastases from, 185, 445 MRI of, 446 SPECT of, 446 MEN. See Multiple endocrine neoplasia
558 Meningiomas, 7, 9, 13, 61–62 Mesenchymal chondrosarcoma, 391 MRI, 391 Mesenchymal malignant tumors, 166–167, 288–289 clinical features of, 288, 293 imaging features of, 289, 293 Mesothelioma, 284 clinical features, 283 imaging features, 283 pericardial, 188 peritoneal, 283 MET. See 11C Methionine Metaiodobenzylguanidine (MIBG), 165, 484 Metastatic tumors, 7, 14, 21, 62 adrenal, 327 bone, 307, 344–346, 402–409 breast, 536–537 cardiac, 183–184 cerebral, 9 from colorectal cancer, 269–272 cystic, 200 differentiation of, 15, 23 endobronchial, 134 FDG in, 78–79 hepatic, 199–206 liver, 127, 199–206, 235 lung cancer, 127–128, 135 lymph nodes, 151, 352 in mediastinum, 167 from melanoma, 185, 445 multiparametric analysis of, 25 pulmonary, 134–136 MFH. See Malignant fibrous histiocytoma MGUS. See Monoclonal gammopathy of undetermined significance MIBG. See Metaiodobenzylguanidine Microlithiasis, 350 MM. See Multiple myeloma MNST. See Malignant nerve sheath tumors Molecular imaging, 246–247 Monoclonal gammopathy of undetermined significance (MGUS), 393 Mononuclear phagocytic system (MPS), 455 Moth-eaten margin, 368 Mouth, floor of, 99 imaging checklist, 102 MPE. See Malignant pleural effusions MPM. See Malignant pleural mesothelioma MPNST. See Malignant peripheral nerve sheath tumor MPS. See Mononuclear phagocytic system MQSA. See Mammography Quality Standards Act
Index MRA. See Magnetic resonance angiography MRCP. See Magnetic resonance cholangiopancreatography MRI. See Magnetic resonance imaging MRL. See MR lymphangiography MR lymphangiography (MRL), 344, 347 MRS. See Magnetic resonance spectroscopy MRSI. See MR spectroscopic imaging MR spectroscopic imaging (MRSI), 337 MRU. See Magnetic resonance urography MS. See Multiple sclerosis MT. See Magnetization transfer MTT. See Mean transit time Mucinous cystic neoplasms (MCNs), 230, 234 Mucosa-associated lymphoid tissue (MALT), 455 Multidetector CT (MDCT), 231, 300, 310, 311 of HCC, 209 for pancreatic tumors, 236–238 Multifocal bronchoalveolar carcinoma, 123 Multifocal osteosarcoma, 378 Multiparametric analysis, 23–24 of high-grade glioma, 24 of low-grade glioma, 25 of metastasis, 25 Multiple endocrine neoplasia (MEN), 150 Multiple myeloma (MM), 48, 393–402 asymptomatic, 394 CT in, 396, 397 follow-up, 402 imaging in, 394–400 metastasis, 49 MRI in, 395, 396, 397 PET in, 401 spinal metastasis, 50 subtypes, 393–394 Multiple sclerosis (MS), 57 Multi-voxel techniques, 535 Mustine, 507 Myelography, 56–57 Myelolipoma, adrenal, 320 Myofibromatosis, 433 Myo-Inositol, 21 Myxoid chondrosarcoma, 390 radiographs, 390 Myxoma, 178, 189, 190
N NAA. See N-acetyl aspartate N-acetyl aspartate (NAA), 20, 21, 476, 477
Index Nasopharyngeal carcinoma (NPC), 95 anterior spread, 96 inferior spread, 96 lateral spread, 96 posterior spread, 96 superior spread, 96 Nasopharynx, 95–97 National Cancer Data Base (NCDB), 67 National Institute for Clinical Excellence, 497 NCA095, 406 NCDB. See National Cancer Data Base Neck neoplasms imaging goals, 94 specific locations, 95–113 technique, 94 tumor resectability, 94–95 Necrosis, 77–78, 413 Negative predictive value (NPV), 236 Nephrogenic Fibrosing Dermopathy (NFD), 304 Nephrogenic Systemic Fibrosis (NSF), 304 Nephropathy, 301 Nerve sheath tumors (NFT), 61. See also Peripheral nerve sheath tumors Neurilemoma, 441 Neuroblastomas, 163–165, 471 composition of, 164 CT of, 474 pediatric, 484 staging of, 164 symptoms of, 164 therapy, 164 Neuroendocrine tumors, 232–233, 244 Neurofibromas, 161–163, 441–442 cutaneous, 441 plexiform, 440, 441 solitary, 441 Neurofibromatosis type 1 (NF-1), 288 Neurogenic tumors, 160–165 NF-1. See Neurofibromatosis type 1 NFD. See Nephrogenic Fibrosing Dermopathy NFT. See Nerve sheath tumors NHL. See Non-Hodgkin’s lymphoma Nodal metastases, 136–137 Non-Hodgkin’s lymphoma (NHL), 46, 47, 132–133, 457, 479 CT of, 458, 459, 460 mediastinum in, 153–154 MRI in, 462–463 staging of, 458 Nonseminomatous malignant germ cell tumors (NSGCTs), 157, 158, 169. See also Germ cell tumors
559 Non-Small cell Lung Carcinoma (NSLC), 124 imaging in, 129 NPC. See Nasopharyngeal carcinoma NPV. See Negative predictive value NSF. See Nephrogenic Systemic Fibrosis NSGCTs. See Nonseminomatous malignant germ cell tumors NSLC. See Non-Small cell Lung Carcinoma Nuclear Medicine, 68
O Obstruction, large bowel, 263 OCT. See Optical coherence tomography OER. See Oxygen extraction 15 O labeled tracers, 87 Oligodendrogliomas, 8, 13, 23, 74 Ollier disease, 382 Oncocytomas, 302 Opiates, 505 Optical coherence tomography (OCT), 245–246 Oral cavity, 97 carcinomas, 102 TNM staging system for, 102 Oral tongue, 98 imaging checklist, 102 Orchiectomy, 351 Oropharynx, 101–103 carcinoma, 105 TNM staging system for, 105 Osseous abnormalities, 480 Osteoblastic lesions, 403 Osteochondroma, 53 Osteochondroradionecrosis, 107 Osteolytic lesions, 403 Osteomyelitis, 412 Osteosarcoma, 186–188, 369–381 chondroplastic, 370 conventional, 369–373 CT in, 373 diagnostic work-up, 381 follow-up, 380–381 gnathic, 376 high grade, 369–373 high grade surface, 377–378 intracortical, 376 low grade, 375–376 MRI in, 373 multifocal, 378 parosteal, 377 pediatric, 486–487 periosteal, 377 radiographic characteristics, 372
560 Osteosarcoma (cont.) sacral chondroblastic, 374 secondary, 378, 379 size of, 371 small cell, 376 staging, 378–380 telangiectatic, 373 types of, 370 Out-of-phase images, 52 Ovarian carcinoma, 200 Oxygen extraction (OER), 32
P p53 gene, 501 PA. See Posteroanterior PACS. See Picture archiving and communication systems Paget’s disease, 378, 379 Pain, 505–506 Pancoast tumor, 123, 124, 161 Pancreatic ductal adenocarcinoma, 230, 241 Pancreatic tumors, 505 CT for, 236–237 cystic lesions, 233 ERCP in, 245 EUS in, 243 FDG in, 239–242 FLT in, 242–243 IDUS in, 245 imaging, 232–235 MDCT, 236–238 MRI in, 238–239 OCT in, 245–246 PET in, 239–242 scintigraphy in, 246 surgical principles of, 233–235 Pancreatitis, 235 Papillary fibroelastoma, 178, 189 Papillary serous carcinoma, 284, 285 clinical features of, 284 imaging features of, 284 Paragangliomas, 161, 165, 293–295 clinical features, 293–294 extra-adrenal, 325 imaging features of, 294–295 Parenchyma, 526 Parosteal osteosarcoma, 377 Partin tables, 340 PBD. See Percutaneous biliary damage PBL. See Primary bone lymphoma PCL. See Primary cerebral lymphoma PCN. See Percutaneous nephrostomy PD. See Proton density Peak systolic velocities, 525
Index Pediatric malignancies, 469 Ewing’s sarcoma, 486–487 FDG in, 470–471 leukemia as, 479–480 lymphoma as, 479–480 neuroblastomas, 484 osteosarcoma, 486–487 PET in, 470–471 rhabdomyosarcoma, 486–487 sarcoma, 486–487 STIR in, 472–473 Wilm’s tumor, 484–486 PEM. See Positron emission mammography Pentetreotide, 246 Percutaneous alcohol ablation, 218 Percutaneous biliary damage (PBD), 220, 502–503 Percutaneous biopsy, 415–416 Percutaneous cholecystostomy, 221 Percutaneous Ethanol Injection (PEI), 218 Percutaneous needle aspiration biopsy (PNAB), 326–327 contraindications of, 494 Percutaneous needle biopsy, 168 Percutaneous nephrostomy (PCN), 503, 504 Percutaneous puncture, 220 Percutaneous transhepatic cholangiography (PTC), 214, 220, 502 Perfusion-weighted imaging, 11–15, 28 basic principle of, 11–12 Pericardial lesions, 184 Pericardial mesothelioma, 188 Periosteal osteosarcoma, 377 Periosteum, 414 Peripheral nerve sheath tumors, 441–442 FDG in, 442 malignant, 442 Peritoneal tumors, 281–282 anatomic considerations in, 282–283 desmoplastic small round cell tumor, 284–285 mesenchymal malignant, 288–289 mesothelioma, 283 MFH, 285–287 primary, 282, 283 Peritoneum, 235 PET. See Positron emission tomography Pheochromocytomas, 161, 165, 320, 323, 326 extra-adrenal, 325 FDG in, 326 imaging of, 323 low density, 324 Phospholipids, 501, 526 Picture archiving and communication systems (PACS), 426, 517
Index Plasmacytomas, 402 Plasmids, 501 Pleural nodules, 138 Pleural space intervention, 506–507 Pleural tumors, 137–139 Pleurodesis, 506 Pleuropericardial cysts, 179 PNAB. See Percutaneous needle aspiration biopsy PNET. See Primary neuroectodermal tumors POEMS syndrome, 395 Point-resolved spectroscopy (PRESS), 20 Polypoid lesions, 261 Polyps, 256 detection, 258–260 pedunculated, 261 sessile, 258, 260 Polyvinyl alcohol, 498 Portal vein thrombosis (PVT), 211 HCC with, 211 MRI of, 211 Portal venous phase (PVP), 207 Positive predictive values (PPV), 46 Positron emission mammography (PEM), 537–538 Positron emission tomography (PET), 3, 4, 30–33, 93, 94, 199, 269. See also Fluoro-2-deoxy-D-glucose adrenal metastasis in, 327 in adrenal tumors, 325–326 advantages of, 68 basics of, 67–68 in bladder cancer, 309, 314 in bone metastasis, 408 breast, 535–541 in colorectal carcinoma, 257, 272–275 coronal, 445 differentiation using, 31 in Ewing’s sarcoma, 414 fused, 170 general considerations in, 68 in lymphoma, 461–462 in mediastinal lymphoma, 156–157 in mediastinal tumors, 169–170 of mediastinal tumors, 147–148 in multiple myeloma, 401 occult lesions on, 400 in pancreatic tumors, 239–242 in pediatric malignancies, 470–471 in prostate cancer, 339 in prostate cancer staging, 343 radiopharmaceuticals, 70 in renal tumors, 306 in skeletal tumors, 367, 400 of soft tissue tumors, 427–429
561 in testicular cancer, 354, 355, 357 in thymic carcinoma, 170 in thymoma, 170 tracers, 69, 80 whole-body, 428 Posterior spread, of NPC, 96 Posteroanterior (PA), 146 Post obstructive changes, 122 Power Doppler, 524–526 PPV. See Positive predictive values PRESS. See Point-resolved spectroscopy Primary bone lymphoma (PBL), 409–413 MRI in, 412 radiograph of, 410, 412 Primary cardiac lymphoma, 188 Primary cerebral lymphoma (PCL), 13 Primary mesenchymal malignant tumors, 166–167 Primary neuroectodermal tumors (PNET), 62 Prostate cancer, 332–348 anatomy, 332–333 capsule in, 332 CT in, 334–335 CT in staging of, 341 detection of, 333 extracapsular extension of, 341 localization of, 333 lymph node staging in, 344 metastasis, 45 metastatic bone disease, 344–346 MRI in, 338 MRI in staging of, 341–342 PET in, 339 PET in staging of, 343 staging of, 340–343 TNM staging classification of, 340 TRUS in staging of, 341 Prostate-specific antigen (PSA), 332 Proton density (PD), 473 PSA. See Prostate-specific antigen Pseudo-capsule sign, 210 PTC. See Percutaneous transhepatic cholangiography Pterygomandibular raphe, 101 Pulmonary metastases, 134–136 PVP. See Portal venous phase PVT. See Portal vein thrombosis
R Radiation necrosis, recurrent tumors and, 27–30 Radiation therapy, 95, 109, 531 Radiculopathy, 123 Radioembolization, 500
562 Radiofrequency ablation (RFA), 199, 218, 219, 417, 500 Radiography, 56–57, 146 in chondrosarcoma, 382, 383 of lymphoma, 410–411 of myxoid chondrosarcoma, 390 in osteosarcoma, 372 of primary bone lymphoma, 410, 412 in skeletal tumors, 368 Radiology, biopsy in, 493–497 Radionuclides, 68 Radiopharmaceuticals, 70 rCBV. See Relative cerebral blood volume RCC. See Renal cell carcinoma Real-time compound sonography, 524 Rectal tumors, 264 CT of, 266–267 endorectal ultrasound of, 266–267 MRI of, 266–267 staging, 266–267 Recurrent tumors, 27–30 necrosis v., 77–78 Regenerative nodules (RN), 208 Region of interest (ROI), 19 Relative cerebral blood volume (rCBV), 11, 12–13, 28, 478 Relative percentage washout (APW), 323 Renal cell carcinoma (RCC), 301 bone metastases, 307 clear cell, 302 CT in, 303 MRI in, 303 pancreatic metastasis, 308 papillary, 302, 303 re-staging, 308 staging of, 306–307 in Von Hippel Lindau Syndrome, 303 Renal obstruction, 503–504 Renal tumors, 300 CT in, 300, 301–305, 307 cystic, 304, 305 detection and diagnosis, 300 FDG in, 306, 308 imaging of, 299 MRI in, 300, 304, 307 PET in, 306 re-staging, 308 solid, 302, 304–305 staging of, 306–307 US in, 300, 305–306 RES. See Reticuloendothelial system Residual tumor detection, 76 Reticuloendothelial system (RES), 455–464
Index Retromolar trigone, 101 Retroperitoneal tumors, 281–282, 356 anatomic considerations in, 282–283 leiomyosarcoma, 291–292 liposarcoma, 290 mesenchymal malignant tumors, 293 MFH, 293 paragangliomas, 293–295 primary, 282, 290–295 Retroviruses, 501 Reverse transcriptase pCR, 415 RFA. See Radiofrequency ablation Rhabdomyoma, 178, 435–436 Rhabdomyosarcoma, 178, 436 CT of, 436 MRI of, 186, 436 pediatric, 486–487 types of, 186 Rhenium, 500 Rib metastasis, 128 RN. See Regenerative nodules ROI. See Region of interest Royal Marsden Hospital Classification system, 351, 353
S Sacral chondroblastic osteosarcoma, 374 Sacrum, 54 Sarcoma, 186–188 Ewing’s, 413–415 pediatric, 486–487 spindle cell, 187 synovial, 289, 290, 443 undifferentiated, 186–188 Scars, 533 SCCA. See Squamous cell carcinoma Schwannomas, 9, 21, 161–163, 441 malignant, 442 Scintigraphy bone, 345, 375, 397–398, 405–406, 425 bone marrow, 406 in pancreatic tumors, 246 SCLC. See Small Cell Lung Carcinoma Scoliosis, 60 Screen film mammography (SFM), 517 SE. See Spin echo Seminal vesicle invasion, 343 Seminoma, 357 mediastinal, 158 Seminomatous germ cell cancer, 349 Sentinel lymph nodes (SLNs), 355–356 Sertoli cell tumors, 348 Serum protein electrophoresis (SPEP), 393, 394
Index SFM. See Screen film mammography SGE. See Spoiled gradient echo Short tau inversion recovery (STIR), 45–46, 127, 373, 462, 471, 480 in pediatric malignancies, 472–473 of thoracic spine, 398 whole body, 399 Single contrast barium enema, 259 Sinonasal carcinoma, 111–112 CT in, 112 MRI in, 112 SIR. See Society of Interventional Radiology Skeletal muscle tumors, 435–436 Skeletal tumors, 367 chemotherapy in, 380 CT in, 367 FDG in, 400 image-guided therapy in, 416–417 image-guide procedures, 415–417 MRI in, 367, 416 PET in, 367, 400 radiographs in, 368 SLNs. See Sentinel lymph nodes Small Cell Lung Carcinoma (SCLC), 130–131 Smooth muscle tumors, 436–437 Society of Interventional Radiology (SIR), 496 Sodium, 535 Soft palate carcinoma, 103 Soft tissue tumors, 423–424 advance in imaging, 424–429 conventional imaging modalities, 425 CT in, 426, 429 fatty, 429–432 FDG, 427–429 fibrous, 433–434 image-guided interventions in, 429 malignant melanoma, 444–446 MRI in, 426–427 peripheral nerve sheath, 441 PET of, 427–429 smooth muscle, 436–437 tumor-like lesions, 433–434 types of, 424 US of, 425–426 vascular, 437–441 Solid tumors, 232–235 renal, 302, 304–305 Solitary plasmacytoma, 402 Solitary pulmonary nodules (SPN), 122, 126 Sonoelastography, 334, 526 Sonography, 456–457. See also Ultrasound real-time compound, 524
563 SPARS. See Spatially resolved spectroscopy Spatially resolved spectroscopy (SPARS), 20 SPECT, 239, 406 of melanoma, 446 SPEP. See Serum protein electrophoresis Spinal compression fractures, 48 Spinal cord, MRI of, 57 Spindle cell sarcoma, 187 Spin echo (SE), 12 Spleen, 456–464 SPN. See Solitary pulmonary nodules Spoiled gradient echo (SGE), 207 Squamous cell carcinoma (SCCA), 97 in hila, 123 hypopharyngeal, 105 in major airways, 123 of supraglottis, 107 SSFP. See Cine-steady state free precession Standard uptake values (SUV), 131, 386–387, 428 STEAM. See Stimulated echo method Stem cell therapy, 30 Stereotactic biopsy, 28–30 optimal sites for, 72–74 Stimulated echo method (STEAM), 20, 57 STIR. See Short tau inversion recovery Sturge-Weber syndrome, 161–162 Subglottis, 110 TNM staging system for, 110 Superior spread, of NPC, 96 Supraglottis, 107–108 SCCA of, 107 TNM staging system for, 109 Survival rates, five-year, 125 SUV. See Standard uptake values Synovial sarcomas, 289, 290, 443 CT of, 443 MRI of, 443, 444 triple sign in, 443
T TACE. See Transarterial chemoembolisation Talarack coordinate system, 71 Talc, 507 tCho. See Total choline Tc MIBI, 400 tCr. See Creatine Technetium-99-m-diphosphonate, 347, 356, 406 Telangiectatic osteosarcoma, 373–375 Teratoma cardiac, 190–191 mediastinal, 157–158, 160
564 Testicular cancer, 348–358 clinical symptoms of, 348 CT in, 351, 353–355 diagnosis of, 349–350 FDG in, 32 germ cell, 356, 357–358 imaging in advanced stage, 356–357 imaging in stage 1, 354–356 MRI in, 354, 355 nonseminomatous, 355 pathology of, 348–349 PET in, 354, 355, 357 self-examination for, 349 staging of, 351–352 TNM staging of, 352 in ultrasound, 353 US in, 349, 350 Testicular microlithiasis, 350 Testis, histopathology of, 349 Tetracycline, 507 T/GM. See Tumor-to-gray-matter Thermal ablation, 500–501 THI. See Tissue harmonic imaging THID. See Transient hepatic intensity difference Thoracic malignancies classification of, 122 staging, 124–125 Thoracic spine, 398 Thromboembolism, 178 venous, 507 Thrombosis, 497–498 Thymic carcinoma, 150 PET in, 170 Thymic tumors, 148–149 Thymolipoma, 151, 152 Thymoma, 147, 148, 152, 170 classification of, 149 malignancy of, 148 Masaoka staging system of, 149 PET in, 170 Thyroid masses, 165–166 CT of, 166 MRI of, 166 Tissue harmonic imaging (THI), 524 Tissue hypoxia, 86 Tl chloride, 400 TNF. See Tumor necrosis factor TNM staging system, 264, 266, 306 descriptors, 125 for glottic carcinomas, 110 for hypopharyngeal carcinoma, 106 of lung cancer, 124 for oral cavity, 102
Index for oropharynx, 105 of prostate cancer, 340 for subglottic carcinomas, 110 for supraglottic laryngeal carcinoma, 109 of testicular tumors, 352 for thoracic malignancies, 124–125 T/N ratio, 81, 84 Tomosynthesis, 520–521 Tongue base carcinoma, 103–104 Tonsil carcinoma, 103 Total choline (tCho), 21, 24 TPR. See True positive rate Transabdominal ultrasound, 243 Transarterial chemoembolization (TACE), 199, 219 Transient hepatic intensity difference (THID), 210 Transient neurologic deficit, 191 Transplantation, 188 Transrectal ultrasound (TRUS), 333–334, 346–347 color doppler, 333–334 contrast-enhanced, 334 grayscale, 333, 334 in prostate cancer staging, 341 sonoelastography, 334 Transurethral resection of bladder (TURB), 311 Triamcinolone, 506 True positive rate (TPR), 537 True vocal cord, 106, 109 TRUS. See Transrectal ultrasound T-stage classification, 94 Tuberous sclerosis, 189 Tumor cell migration, 30 Tumor destruction, 217–219 in HCC, 217 Tumor grade assessment, 72 Tumor induced angiogenesis, 525 Tumor-like lesions, 433–434 Tumor malignancy, 3 Tumor necrosis factor (TNF), 247 Tumor progression, 76–77 Tumors. See specific types Tumor-to-gray-matter (T/GM), 72 Tumor-to-white-matter (T/WM), 72 TURB. See Transurethral resection of bladder T/WM. See Tumor-to-white-matter
U Ultra small super paramagnetic iron oxide (USPIO), 313–314, 344 Ultrasound (US), 168, 169, 494. See also Contrast enhanced ultrasound;
Index
565
Intraductal ultrasound; Transrectal ultrasound breast, 523–527 in cirrhosis, 212–213 contrast agents, 524–526 in gallbladder carcinoma, 216 gray scale, 485 indeterminate, 534 in leiomyosarcoma, 292 in lipomas, 430 in liver metastases, 269–272 of lymph nodes, 456–457 in peripheral nerve sheath tumors, 442 in peritoneal tumors, 282 in renal tumors, 300, 305–306 in retroperitoneal tumors, 282 of soft tissue tumors, 425–426 in testicular cancer, 349, 350, 353 transabdominal, 243 Wilm’s tumor, 484–486 Undifferentiated sarcoma, 186–188 Upper gastrointestinal obstruction, 504 Urothelial hyperplasia, 504 US. See Ultrasound USPIO. See Ultra small super paramagnetic iron oxide
Vascular tumors, 437–441 Vasculature, 321 VEGF. See Vascular endothelial factor VEGF-R1, 87 Venous thromboembolism, 507 Vertebrae, 404 Vertebral wall, 63 Vertebroplasty, 62–63, 397, 505 for L4 lesion, 63 Von Hippel Lindau Syndrome, 60–61, 161–162 RCC in, 303 Voxels, 337
V Vagus nerve, 161 Vascular endothelial factor (VEGF), 247
Y Yttrium, 219 Yttrium-90, 500
W Waldeyer’s ring, 103 White matter tractography, 18–19 WHO. See World Health Organization Wilm’s tumor, 484–486 US of, 484–486 World Health Organization (WHO), 3, 67 thymoma classification, 149
X X-ray beam, 258
