Imaging of Brain Tumors with Histological Correlations
Antonios Drevelegas (Editor)
Imaging of Brain Tumors with Histological Correlations Second Edition
Editor Antonios Drevelegas Nikis 25, N. 751 552 36 Panorama Thessaloniki Greece
[email protected]
ISBN 978-3-540-87648-9 e-ISBN 978-3-540-87650-2 DOI 10.1007/978-3-540-87650-2 Springer Heidelberg Dordrecht London New York Library of Congress Control Number: 2010930084 © Springer-Verlag Berlin Heidelberg 2011 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Product liability: The publishers cannot guarantee the accuracy of any information about dosage and application contained in this book. In every individual case the user must check such information by consulting the relevant literature. Cover design: eStudio Calamar, Figueres/Berlin Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
This book is dedicated to: My parents My wife Mary and to my children Helen and Konstantinos All my colleagues who have stood by me all these years A. Drevelegas
v
Preface
The purpose of this book is to bring a new understanding to bear on the diagnosis of brain tumors by linking radiographic image characteristics to the underlying pathology. Brain tumors are relatively uncommon compared with other neoplasms (e.g., lung, breast, gastrointestinal). They require special study, since they are pathologically complicated, difficult to diagnose, and account for high morbidity. Athough many excellent neuroradiological books have been written, few of them focus especially on the diagnosis of brain tumors. In this book, brain tumors are discussed in detail. Special emphasis is placed on CT and MRI findings in relation to the pathology of each tumor. As pathology is the “mother” of radiology, this approach may be the best way to understand in depth the imaging manifestations of brain tumors. The illustrative examples herein, were chosen on the basis of their clarity or complexity, their teachability, and their significance for diagnosis and treatment. In the second edition of this book, all chapters have been revised and updated with new clinical information and new imaging material providing the scientists, interested in the field of Neuro-imaging and Neuro-onocology, with knowledge that will enhance their service to the patients. The latest developments in the field of MRI, the tendency to move to higher MRI fields (3T), as well as the introduction in the clinical practice of advanced imaging techniques such as, diffusion, diffusion tensor imaging, perfusion, spectroscopy, and functional imaging represent the new tools in the hands of the neuroscientists to help them in the diagnosis, treatment, and follow-up of brain tumors. The book includes 14 chapters. Chap. 1 deals with the epidemiology and classification of brain tumors. Chap. 2 discusses different imaging modalities and their contribution to the diagnosis of brain tumors. Special emphasis is laid on the latest developments and on potential future applications of MRI. Chap. 3 is regarding the genetic and molecular basis of gliomas. In Chap. 4, we have included a new section about application of f(MRI) and DTI in presurgical planning for tumor resertion. Chapters 5–12 constitute an in-depth study of imaging characteristics of different brain tumors on CT and MRI. The images’ contribution to diagnosis and their correspondence to certain pathologic appearances are particularly stressed. Finally, a state-of-the-art chapter on nuclear medicine is included to cover the impact of SPECT and PET imaging on brain tumor diagnosis. We hope that this book will serve as a teaching tool and practical reference for the diagnosis of brain tumors and will enhance the reader’s diagnostic performance.
vii
viii
Preface
I would also like to thank all the distinguished authors for their valuable contribution. Many thanks to the people of Springer, without their contribution this book could not have been completed. Thessaloniki, 2010
A. Drevelegas
Contents
1 Epidemiology, Histologic Classification, and Clinical Course of Brain Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . George Karkavelas and Nickolaos Tascos
1
2 Imaging Modalities in Brain Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . Antonios Drevelegas and Nickolas Papanikolaou
13
3 Molecular Abnormalities in Gliomas . . . . . . . . . . . . . . . . . . . . . . . . . . . Anna C. Goussia, Konstantinos Polyzoidis, Maria Bai, and Athanasios P. Kyritsis
35
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sofie Van Cauter, Silvia Kovacs, Caroline Sage, Ron Peeters, Judith Verhoeven, Sabine Deprez, and Stefan Sunaert 5 Low-Grade Gliomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . G.A. Christoforidis, Antonios Drevelegas, Eric C. Bourekas, and George Karkavelas
49
73
6 High-Grade Gliomas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Antonios Drevelegas and George Karkavelas 7 Pineal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 Antonios Drevelegas, Argyris K. Strigaris, and Christiana H. Samara 8 Embryonal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215 Guido Wilms, Antonios Drevelegas, Philippe Demaerel, and Raf Sciot 9 Tumours of the Cranial Nerves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 Hervé Tanghe, Paul M. Parizel, and Antonios Drevelegas 10 Meningeal Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255 Antonios Drevelegas, George Karkavelas, Danai Chourmouzi, Glykeria Boulogianni, and Anastasios Petridis
ix
x
11 Central Nervous Lymphomas and Hemopoietic Neoplasms . . . . . . . . . 303 Julia Frühwald-Pallamar, Negar Fakhrai, Majda M. Thurnher, and Antonios Drevelegas 12 Masses of the Sellar and Junxtasellar Region . . . . . . . . . . . . . . . . . . . . 325 Eric C. Bourekas, H. Wayne Slone, and Abhik Ray-Chaudhury 13 Brain Metastasis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373 Nicholas J. Patronas 14 Scintigraphy for Brain Tumors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401 George N. Sfakianakis, Efrosyni Sfakianaki, and Hilton Gomes Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 427
Contents
1
Epidemiology, Histologic Classification, and Clinical Course of Brain Tumors George Karkavelas and Nickolaos Tascos
1.1 Introduction
Contents 1.1 Introduction..............................................................
1
1.2 Epidemiology............................................................. 1.2.1 Pediatric Brain Tumors............................................... 1.2.2 Adult Brain Tumors....................................................
1 1 2
1.3 Clinical Course..........................................................
2
1.4 Histologic Classification and Grading....................
3
1.5 Incidence and Clinical Course by Age and Location.............................................................. 5 1.5.1 Adult Brain Tumors.................................................... 8 1.5.2 Pediatric Brain Tumors............................................... 10 References............................................................................ 11
Brain tumors are divided into primary (70%) and secondary (30%). About 24,000 primary brain tumors are diagnosed each year in the United States and account for 20% of malignancies before 15 years. The estimated incidence is 8.2 per 100,000 people. Although these tumors are not common (<1.5% of all cancers) in comparison with tumors of other organs (e.g., lung, breast, or colon), they deserve a unique position in tumor oncology because of their histopathologic complexity and biologic behavior. Noteworthy is the fact that despite their low incidence they have a high mortality among adult cancer patients and even more in children [1]. Topographical and histological classifications, as well as correlations of topography to pathology, serve for the better studying, understanding, and handling of brain tumors. They are found in any location and in patients of any age. Differences in epidemiology, location, and pathology between children and adults make it necessary to take into consideration the age parameter, and to consider adult and pediatric brain tumors separately. As a result for any therapeutic decision or prognostic determination, all these factors, that is, age, location, and pathology have to be combined.
1.2 Epidemiology 1.2.1 Pediatric Brain Tumors G. Karkavelas (*) Department of Pathology, Aristotle University of Thessaloniki, School of Medicine, Thessoloniki, Greece e-mail:
[email protected] N. Tascos B’ Department of Neurology, Aristotle University of Thessaloniki, School of Medicine, Thessoloniki, Greece
Fifteen to 20% of all intracranial tumors occur in children under 15 years of age, with the peak occurrence between 4 and 8 years of age. The incidence of brain tumors in children has been estimated to be 2.5 per 100,000 people per year, the vast majority of them
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_1, © Springer-Verlag Berlin Heidelberg 2011
1
2
being primary. Only 1–2% of all brain tumors occur in children under 2 years of age [2–4]. In neonates, brain tumors are uncommon, represent congenital tumors, and have a different histologic and topographic distribution compared with those that occur in young children and adolescents. They are more likely to be developed in the supratentorial region than in the posterior fossa. The most common primary brain tumors in the neonatal period are teratomas, embryonal tumors, and congenital glioblastoma multiforme [5]. In children between the ages of 2 and 10 years, primary brain tumors are generally more benign than those found in neonates. Seventy percent of these tumors are developed beneath the tentorium [3, 6]. About 30% of posterior fossa tumors in children are cerebellar astrocytomas (most often pilocytic), 35–40% are medulloblastomas (arising for more than 90% in the vermis of the cerebellum), 25% are brain stem gliomas, and 10–15% are ependymomas of the fourth ventricle. In children younger than 3 years of age, 30% of all intracranial tumors are ependymomas [7, 8].
1.2.2 Adult Brain Tumors The incidence of primary brain tumors appears to have increased, particularly in the elderly people of developed countries, during the last 25 years. The incidence of primary malignant brain tumors increased by 40% in the general population and 100% in the elderly (more than 65 years of age) in the United States and Canada [9–13]. It is debatable whether this is due to better diagnostic procedures, or reflects an actual increased incidence. The most dramatically increased incidence concerns primary central nervous system lymphoma. These tumors appear to have increased by approximately 300% in the immunocompetent and immunocompromised population during the last two decades [14–16]. Gliomas in man and meningiomas in woman are the most common primary brain tumors constituting respectively 60 and 20% of all intracranial tumors in adults. In adults, tumors of glial origin are the most frequent primary neoplasms [17, 18]. In young adults, low-grade gliomas occur more often. In middle age patients, anaplastic astrocytomas are more common [19, 20]. In the elderly, glioblastoma multiforme is the most common primary tumor. Increasing age generally correlates with increasing malignancy [11, 21]. In the adult patients, the most common supratentorial intra-axial neoplasms are
G. Karkavelas and N. Tascos
glioblastoma multiforme and metastatic tumors, whereas metastases and hemangioblastomas are the most frequent intra-axial neoplasms in the posterior fossa [10]. Among extra-axial lesions, the most common neoplasm to be found in the supratentorial region is meningioma. Schwanomma predominates infratentorially [17, 22]. Approximately 80–85% of all intracranial tumors in adults occur in the supratentorial region. The primary tumors are 50–75% of all tumors, while metastases account for the remainder. The infratentorial tumors in adults are mainly extra-axial lesions [10, 17]. Approximately 50% of the intracranial tumors are gliomas. Most of them (70%) are astrocytic tumors, and more than 50% of these are anaplastic or glioblastoma multiforme [21]. Approximately 15–20% of primary tumors in adults are meningiomas, most of them (75%) found supratentorially [23]. Pituitary adenomas are the most common tumors in the sellar-parasellar region. Tumors in this area are about 8% of all intracranial tumors [24]. Oligodendrogliomas represent about 5% of the intracranial tumors, and most of them are supratentorial [25]. The primary posterior fossa tumors in adults are either intra- or extra-axial. The three most common tumors are schwannoma, meningioma, and epidermoid tumor. All of these tumors are extra-axial. The most often intra-axial primary brain tumor in adults is hemangioblastoma and brain stem glioma. The metastatic intra-axial tumor in the posterior fossa is the most common tumor in this region. In this location, metastatic tumors account for 15–20% of all intracranial metastases [22, 23, 26–28].
1.3 Clinical Course Brain tumors produce symptoms and signs due to direct tissue destruction, local brain infiltration, or secondary effects of increased intracranial pressure. The symptoms of patients are dependent on the anatomic location of the tumor. Negative symptoms include loss of function, and positive symptoms seizures or headache. Headache is the first symptom in 35% of patients and with growing of the tumor it is present in up to 70% of patients. Usually, it is associated with vomiting or nausea, papilledema due to increased intracranial pressure or focal cerebral signs. Facial pain may be present with tumors at the base of the skull or nasopharynx.
1 Epidemiology, Histologic Classification, and Clinical Course of Brain Tumors
Seizures are the first symptom in 30% of brain tumors and account for 5% of all patients with epilepsy. Seizure frequency is seen in up to 95% of patients with oligodendrogliomas and occurs in only 18% of patients with metastatic brain tumors [29]. Focal clinical signs: The clinical manifestations of brain tumors depend on the location of tumor in the brain (Fig. 1.1). • Frontal lobe tumors cause personality changes, gait abnormalities, hemiparesis, and aphasia. • Temporal lobe tumors more often cause seizures. Temporal hemianopsia and aphasia may be present. • Parietal lobe tumors cause sensory loss, neglect, anosoagnosia, hemiparesis, and disturbances of visuospatial abilities. • Occipital lobe tumors cause visual field changes. • Brain stem tumors commonly present with cranial nerve and pyramidal track signs, as well as symptoms and signs associated with increased intracranial pressure.
3
• Thalamic tumors present with sensory loss, personality and mental changes, as well as symptoms and signs of increased intracranial pressure. • Pineal tumors produce symptoms and signs of hydrocephalus, Parinaud syndrome, and precocious puberty. • Pituitary tumors may present with hormonal alterations, symptoms and signs of cranial nerves, bitemporal hemianopsia, and obstructive hydrocephalus. • Cerebellar tumors present with gait ataxia, and symptoms and sings of obstructive hydrocephalus. • Intraventricular tumors present with symptoms and sings of obstructive hydrocephalus.
1.4 Histologic Classification and Grading Although modern techniques have greatly improved the diagnostic ability for CNS tumors, histopathology remains crucial for the accurate diagnosis. The histologic
Fig. 1.1 Coronal (upper left) and sagittal (right) autopsy specimens of the brain show the clinical signs of the brain tumors in relation to their location
4
classification and grading of these tumors is controversial, but on the other hand it is also critical for the assessment of their prognosis and treatment. Virchow was the first who used the term “gliomas” for the most common forms of primary tumors of the brain. Although under “gliomas” all the brain tumors were initially categorized, this term is now applied only to the tumors of neuroglial cell origin. The first attempt to classify gliomas was that of Bailey and Cushing [30] in 1926 in their book “A Classification of the Tumors of the Glioma Group.” They classified these tumors in 14 groups in accordance to their embryogenesis, following the Spanish School. Astrocytic tumors were categorized as astrocytomas, astroblastomas, and spongioblastomas multiforme. More primitive embryonal tumors were classified as meduloepitheliomas, neuroepitheliomas, and blastomas (e.g., meduloblastomas, neuroblastomas). Terms that are in use today, such as oligodendrogliomas and ependymomas were also found in their classification. In an effort to simplify the previous classification and having in mind the four-tiered Broders’ grading system of general pathology, Kernohan and Sayre in 1952 graded gliomas by the degree of their dedifferentiation in four grades, from 1 to 4. Their grading system, as was proposed in the first Armed Forces Institute of Pathology (AFIP) fascicle on brain tumors, was simple and easily adopted by neuropathologists and clinicians [31]. The four grades of this system were coordinated with the well-known transformation of gliomas to higher grades of malignancy and correlated with patient prognosis. Prerequisite for this grading and classification was the acceptance that gliomas arise from adult cells still capable of proliferation after dedifferentiation or anaplasia [32]. The glioma types were now reduced to five groups: astrocytomas (grades 1–4), ependymomas (grades 1–4), oligodendrogliomas (grades 1–4), neuroastrocytomas (grades 1–4), and medulloblastomas. Their grading system was applied mostly to astrocytic tumors. Despite the reservations for this classification system and the adoption of more recent systems, this is still in use by general pathologists and clinicians. To simplify the grading of gliomas, Ringertz [33] proposed in 1950 a system of three grades. Under this grading system, astrocytic tumors were categorized as astrocytomas, intermediate type astrocytomas, and glioblastomas multiforme. The system of Ringertz, as
G. Karkavelas and N. Tascos
well as other subsequently proposed three-tiered grading systems, were easily adopted and widely used with modifications (e.g., anaplastic instead of intermediate astrocytomas) [34–36]. The main reason for their popularity is the close relation of these grades to the survival of patients and the reproducibility of these systems [37] The value of these three-tiered grading systems has been supported by large cooperative oncology groups [35, 38] Almost two decades later, in 1972, Rubinstein, in the second fascicle of the AFIP [39], and later in “Pathology of Tumors of the Nervous System” [40], classified CNS tumors taking in consideration the existence of embryonic tumors. He proposed that these tumors were the result of neoplastic transformation in neurocytogenesis. Rubinstein adopted the term “embryonal tumors” of Bailey and Cushing, with two additional tumors: ependymoblastoma and polar spongioblastoma. He adopted also the fourtiered system of Kernohan and Sayre for gliomas (with reservations for the application to all gliomas) but he separated pilocytic from diffuse astrocytomas, emphasizing their better biologic behavior. The term primitive neuroectodermal tumor (“PNET”) was proposed in 1973 by Hart and Earle [41] for a number of undifferentiated supratentorial tumors of infancy. Despite the subsequent immunohistological character ization of many of these tumors, a number was still retained under this term. Furthermore, Rorke [42] as well as Becker and Hinton [43] in 1983 proposed the term PNET to be used for all embryonal neuroectodermal tumors of the CNS. In 1979, in the fascicle “Histological Classification of Tumors of the Central Nervous System,” the blue series of World Health Organization (WHO), Zulch [44] classified the CNS tumors in a way that combined the views of an international committee of neuropathologists. By this classification, grade I was applied to pilocytic astrocytomas, while diffuse astrocytic tumors (astrocytomas, anaplastic astrocytomas, and glioblastomas multiforme) were graded as II, III, and IV, respectively. The subjectivity in the appreciation of morphologic features is the main problem of histologic grading. In 1988, Daumas-Duport and collaborators looked at the grading of astrocytomas in a different angle. Their grading system (known as the St Anne-Mayo scheme) is a four-tired one again, but they use four histological
5
1 Epidemiology, Histologic Classification, and Clinical Course of Brain Tumors
features to grade each astrocytoma. These features are cellular pleomorphism, mitotic activity, microvascular proliferation, and necrosis. Absence of any of these features characterize grade I astrocytomas. The recognition of one feature characterizes grade 2, two features grade 3, and three or four features grade 4 [45]. In the revised edition of WHO (1993), which still is the most widely accepted classification system, some new entities have been added (e.g., central neurocytoma, pleomorphic xanthoastrocytoma, dysembryoblastic neuroepithelial tumor, and others). By this system, in grading astrocytomas, the criteria proposed by the St Anne-Mayo scheme were taken into consideration [36]. Burger and Scheithauer adopted elements of previous classification and grading systems in their own one, as presented in the tenth fascicle of the “Tumors of the Central Nervous System,” published by the Armed Forces of Pathology. They used verbal designation instead of numerical grading for malignancy; they rejected the “PNET” concept and expanded the classification of mesenchymal tumors [46]. In the sixth edition of “Russel and Rubinstein’s Pathology of Tumors of the Nervous System,” the authors adopted, with some modifications, the WHO revised classification of 1993 [37]. The WHO classification of tumors of the nervous system, as it is presented in the new edition of Kleihues and Cavenee “Pathology and Genetics: Tumours of the Nervous System,” is the result of an editorial and consensus conference, held in Lyon, in July 1999 [47]. In the fourth edition of the “WHO Classification of Tumors of the Central Nervous System” edited in 2007 by Louis, Ohgaki, Wiestler na Cavenee, was presented the latest classification of tumors of the nervous system (Table 1.1) [47a].
1.5 Incidence and Clinical Course by Age and Location Since incidence and clinical course of brain tumors are not only related to pathology, but also to age of patients and location of tumors, a correlation of these parameters is a useful tool in the differential diagnosis, narrowing the list possible tumors and significantly affecting the differential diagnosis (Fig. 1.2).
Table 1.1 WHO classification of tumors of the nervous system 1. TUMORS OF NEUROEPITHELIAL TISSUE 1.1 Astrocytic tumors Pilocytic astrocytoma Pilomyxoid astrocytoma Subependymal giant cell astrocytoma Pleomorphic xanthoastrocytoma Diffuse astrocytoma Fibrillary astrocytoma Gemistocytic astrocytoma Protoplasmic astrocytoma Anaplastic astrocytoma Glioblastoma Giant cell glioblastoma Gliosarcoma Gliomatosis celbri 1.2 Oligodendroglial tumors Oligodendroglioma Anaplastic oligodendroglioma 1.3 Ependymal tumors Subependymoma Myxopapillary ependymoma Ependymoma Cellular Papillary Clear cell Tanycytic Anaplastic ependymoma 1.4 Choroid plexus tumors Choroid plexus papilloma Atypical choroid plexus papilloma Choroid plexus carcinoma 1.5 Other neuroepithelial tumors Astroblastoma Chordoid glioma of the third ventricle Angiocentric glioma 1.6 Neuronal and mixed neuronal-glial tumors Dysplastic gangliocytoma of cerebellum (Lhermitte-Duclos) Desmoplastic infantile astrocytoma/ganglioglioma Dysembryoplastic neuroepithelial tumor Gangliocytoma Ganglioglioma Anaplastic ganglioglioma Central neurocytoma Extraventricular neurocytoma Cerebellar liponeurocytoma Papillary glioneuronal tumor Rosette-forming glioneuronal tumor of the fourth ventricle Paraganglioma 1.7 Tumors of the pineal region Pineocytoma Pineal parenchymal tumor of intermediate differentiation Pineoblastoma Papillary tumor of the pineal region (continued)
6
G. Karkavelas and N. Tascos
Table 1.1 (continued) 1.8 Embryonal tumors Ependymoblastoma Medulloblastoma Desmoplastic/nodular medulloblastoma Medulloblastoma with extensive nodularity Anaplastic medulloblastoma Large cell medulloblastoma CNS primitive neuroectodermal tumor CNS Neuroblastoma CNS Ganglioneuroblastoma Meduloepithelioma Ependymoblastoma Atypical teratoid/rhabdoid tumor 2. TUMORS OF CRANIAL AND PARASPINAL NERVES 2.1 Schwannoma (Neurilemoma, Neurinoma,) Cellular Plexiform Melanotic 2.2 Neurofibroma Plexiform 2.3 Perineurioma Intraneural perineurioma Soft tissue perineurioma 2.4 Malignant peripheral nerve sheath tumor (MPNST) Epithelioid MPNST MPNST with mesenchymal differentiation Melanotic MPNST MPNST with glandular differentiation 3. TUMORS OF THE MENINGES 3.1 Tumors of the meningothelial cells Meningioma Meningothelial Fibrous (fibroblastic) Transitional (mixed) Psammomatous Angiomatous Microcystic Secretory Lymphoplasmacyte rich Chordoid Clear cell Atypical Papillary Rhabdoid Anaplastic (malignant) 3.2 Mesenchymal tumors Lipoma Angiolipoma Hibernoma Liposarcoma Solitary fibrous tumor
Fibrosarcoma Malignant fibrous histiocytoma Leiomyoma Leiomyosarcoma Rhabdomyoma Rhabdomyosarcoma Chondroma Chondrosarcoma Osteoma Osteosarcoma Osteochondroma Hemangioma Epithelioid hemangioendothelioma Hemangiopericytoma Anaplastic hemangiopericytoma Angiosarcoma Kaposi sarcoma Ewing sarcoma-PNET 3.3 Primary melanocytic lesions Diffuse melanocytosis Melanocytoma Malignant melanoma Meningeal melanomatosis 3.4 Other neoplasms related to the meninges Hemangioblastoma 4. LYMPHOMAS & HEMOPOIETIC NEOPLASMS Malignant lymphomas Plasmacytoma Granulocytic sarcoma 5. GERM CELL TUMORS Germinoma Embryonal carcinoma Yolk sac tumor Choriocarcinoma Teratoma Mature Immature Teratoma with malignant transformation Mixed germ cell tumors 6. TUMORS OF THE SELLAR REGION Craniopharyngioma Adamantinomatous Papillary Granular cell tumor Pituicytoma Spindle cell oncocytoma of the adenohypophysis 7. METASTATIC TUMORS
1 Epidemiology, Histologic Classification, and Clinical Course of Brain Tumors
7
Fig. 1.2 Sagittal autopsy specimen of the brain shows the distribution most of the primary intracranial tumors in childhood and adulthood
8
1.5.1 Adult Brain Tumors 1.5.1.1 Supratentorial Intra-axial The most common intra-axial tumors are these of glial origin and metastases. More than half of all glial tumors are high-grade astrocytic tumors (anaplastic astrocytoma and glioblastoma multiforme).
G. Karkavelas and N. Tascos
Oligodendroglioma The frequency of oligodendrogliomas among gliomas ranges from 4 to 15%. The peak incidence is between 35 and 45 years of age. Oligodendrogliomas are commonly located in cortical or subcortical regions. These tumors arise from a subcortical location of the frontal, temporal, or parietal lobe and infiltrate the cortex. The most frequent initial symptoms that patients with oligodendrogliomas have are seizures in 50–75%, and headache in 9–48%. The duration of symptoms may be up to 20 years before tumor diagnosis [25].
High-Grade Astrocytic Tumors Anaplastic astrocytoma is among the most common primary malignant brain tumors. Glioblastoma multiform is the most common of all primary brain tumors. Although they can be found at any age, their peak incidence is in the fifth and sixth decades of life. They produce symptoms and signs due to direct tumor effects or secondary effects of brain edema, hydrocephalus or increased intracranial pressure. These tumors rarely grow into the ventricular system, so when they produce increased intracranial pressure, this is usually due to mass effect. The most frequent initial symptoms are headache and seizures. The most common presenting neurologic signs are hemiparesis in 61–83%, papilledema in 32–66%, confusion in 18–40%, and aphasia in 25–32% of patients. Approximately 6% of patients are present with acute onset of symptoms due to intracranial hemorrhage. Only 1% of patients with glioblastoma multiforme were without any neurological sign on presentation [11, 21].
Primary Central Nervous System Lymphoma
Low-Grade Astrocytic Tumors
Pituitary Region Tumors
Low-grade astrocytic tumors (pilocytic and diffuse astrocytomas) account for 10–20% of brain tumors in adults and 8–25% in children. Approximately two-thirds of patients are younger than 40 years of age. A total of 75–80% of pilocytic astrocytomas, occur in the first two decades of life and are located in midline structures. Patients with low-grade astrocytomas have seizures as presenting symptoms in 60%, and headache in 38–46%. They may have years of seizures with nondiagnostic studies. The neurologic deficit depends on tumor location. Midline tumors are less likely to present with seizures, and more likely to present with symptoms and signs of increased intracranial pressure [48, 49].
Pituitary adenomas represent 10% of all intracranial tumors. There is a high female preponderance. Twothirds of them are found in women. The tumors of pituitary region may present with symptoms and signs due either to endocrine dysfunction or to a mass effect on the pituitary and its surrounding neural and vascular structures. Mass effects include headache, hypopituitarism, diplopia, and any pattern of visual loss. In about 5% of pituitary tumors the first symptoms are those of “pituitary apoplexy,” due to hemorrhage or infarction of the adenoma. Symptoms include sudden onset of severe headache, diplopia, vomiting, altered
Primary central nervous system lymphomas account for 1–4% of primary brain tumors. The incidence of these rare tumors has been increasing, in both the immunocompetent and in the immunocompromised populations. Three percent of AIDS patients will develop primary central nervous system lymphomas, either before or during the course of their illness. In immunocompetent patients, the peak incidence is in the sixth and seventh decades of life, and in AIDS patients, in the fourth decade of life. The majority of these tumors are cerebral in location and in 40% of patients they are bilateral. They can also be seen in occular, spinal, and leptomeningeal site. The most frequent presenting symptoms are mental changes in 36% of patients, headaches in 22% at onset and 50% during their illness, cerebellar signs in 31%, seizures in 20%, motor disabillity in 17%, and visual defects in 12% of patients [50, 51].
1 Epidemiology, Histologic Classification, and Clinical Course of Brain Tumors
consciousness, and rapidly progressive visual loss. This is usually an indication for emergency surgery. The syndromes associated with hypersecretion of pituitary hormones by functional pituitary tumors include Cushing’s disease, acromegaly, hyperprolactenemia, and Nelson’s syndrome [24]. Metastatic Tumors The number of new cases of brain metastases in the United States is approximately 170,000 per year. Magnetic resonance imaging (MRI) reveals that twothirds to three-fourths of brain metastases are multiple. The lung is the most common primary source of the cerebral metastases, with the majority of them being supratentorial in location. When the primary tumor is located in the gastrointestinal tract or pelvic area, the posterior fossa is the site of the metastases in 50% of patients. Posterior fossa is involved in only 10% of the metastases from other tumors. The most common presenting symptoms of brain metastases are headache in 26–57% of patients, focal weakness in 26–75%, mental impairment in 22–77%, and seizures in 6–21%. The most common cause of altered mental status is metabolic encephalopathy in 61% of patients, with only 15% having metastases. Pain is the first symptom in 95% of patients with vertebral metastases. Other neurological signs are limb weakness in 76%, autonomic disturbances in 57%, and sensory dysfunction in 51% of the patients [52, 53].
Extra-axial The most common extra-axial tumors in the supratentorial region are meningioma and metastases.
9
The most common presenting symptoms are headache in 36%, mental changes in 21%, and paresis in 22% of the patients. Meningiomas are rare in children and located in the posterior fossa or intraventricularly, producing obstructive hydrocephalus [23, 54].
1.5.1.2 Infratentorial Intra-axial The most common intra-axial tumors of the posterior fossa are hemangioblastoma and metastases. Hemangioblastoma Hemangioblastomas most commonly affects adults and occur in isolation or in association with Von Hippel– Lindau disease. They represent uncommon tumors, acco unting for 2.5% of all primary CNS tumors. The peak incidence is between 40 and 60 years of age and they are very rare in children. Cerebellum is by far the most frequent site of location, producing symptoms and signs of limp and truncal ataxia and increased intracranial pressure. They can also be found in the medulla, the spinal cord, and very rarely, in the supratentorial meninges [28].
Extra-axial Schwannomas are the most common tumor of the posterior fossa in the extra-axial region, and meningiomas are the second most common. Arachnoid and epidermoid cysts are other common tumors in this region. Schwannoma
Meningioma Meningiomas are the most common non-glial primary brain tumors. They represent approximately 20% of all symptomatic intracranial tumors. The peak occurrence of meningiomas is between 40 and 60 years of age. Only 1–2% of these tumors occur in children less than 15 years of age. The majority of meningiomas are incidental findings on imaging studies or found at autopsy. Only 25% of these tumors are symptomatic on presentation. In 85–90% the meningiomas are supratentorially located.
A total of 6–8% of all intracranial tumors are schwannomas. They account for more than 75% of all tumors in the cerebellopontine angle. Five percent are bilateral and are, almost always, associated with neurofibromatosis II (NF2). Gradually, progressive unilateral hearing loss occurs in 97% of patients. Both, tinnitus and unsteadiness of gait occur in 70%, and headache in 38% of the patients. With large tumors there may be abnormal eye movements, cerebellar or pyramidal signs, papilledema or lower cranial nerve disorders [22, 55].
10
1.5.2 Pediatric Brain Tumors 1.5.2.1 Supratentorial The most frequent supratentorial tumors in the cerebral hemispheres are astrocytic tumors (50%), embryonal tumors (15%), and ependymomas (15%). In the suprasellar region, 75% of tumors are craniopharyngiomas, the most common non-neuroepithelial tumor in children. Ten percent of all pediatric tumors are located in the pineal region, and 50–70% of these arise from germ cells [7].
Astrocytic Tumors Most of the supratentorial astrocytic tumors are pilocytic or low-grade astrocytomas. The peak occurrence is between 2 and 4 years and 7 and 8 years. They are mainly located in the cerebral hemispheres but they can occur in the opticochiasmatic-hypothalamic area. Seizures and focal neurological deficits are the most common symptoms. High-grade astrocytomas are more aggressive tumors and show areas of hemorrhage, necrosis, peritumoral edema, and mass effect.
G. Karkavelas and N. Tascos
Direct midbrain compression can cause Parinaud syndrome with loss of upward gaze. Endocrine dysfunction is due to tumor spread to the hypothalamic region; it produces diabetes insipidus, somnolence, weight gain, temperature fluctuations, and precocious puberty [57].
1.5.2.2 Infratentorial The most common tumors in this region in the pediatric population are pilocytic astrocytoma, medulloblastoma, ependymoma, and brain stem glioma.
Pilocytic Astrocytoma Pilocytic astrocytoma is the most common tumor in the cerebellum and is characterized by cystic formation. Pilocytic astrocytomas represent approximately 80% of all cerebellar astrocytic tumors and occur equally within the vermis and cerebellar hemispheres. They usually present with signs of increased intracranial pressure.
Medulloblastoma Craniopharyngioma Craniopharyngiomas are congenital cystic tumors accounting for more than 50% of all childhood suprasellar tumors and approximately 10% of all pediatric brain tumors. Their peak incidence is between 5 and 15 years of age. They are mainly located in the suprasellar region, depressing the optic chiasm and extending up into the third ventricle. The presenting symptoms may be those of increased intracranial pressure, or symptoms of a combined hypopituitary, hypothalamic, and chiasmal derangement [56].
Pineal Region Tumors Pineal region tumors represent 0.5–1.6% of all adult brain tumors and 10% of all pediatric brain tumors. These tumors can produce symptoms and signs by increased intracranial pressure, by direct brain stem and cerebellar compression or by endocrine dysfunction.
Medulloblastoma is the second most common primary central nervous system tumor of childhood, accounting for 15–30% of all childhood brain tumors and 30–40% of all childhood posterior fossa tumors. Medulloblastomas in two-thirds of children arise from the cerebellar vermis. In adults, the tumor more frequently originates in the cerebellar hemispheres. After surgical resection the 5-year survival is approximately 80–90%. The tumor typically presents in almost 90% of children with headache, vomiting, diplopia, papilledema, and truncal unsteadiness. In adults, the tumor presents with headache, nausea and vomiting in 40% of patients and truncal or limb ataxia in 25%. The cranial nerves in the cerebellopontine angle may be also involved [58].
Ependymoma Ependymomas account for 2–6% of all intracranial tumors. In children, ependymomas are the third most
1 Epidemiology, Histologic Classification, and Clinical Course of Brain Tumors
common intracranial tumor, accounting for 6–12% of intracranial tumors. In children younger than 3 years of age, 30% of all intracranial tumors are ependymomas. Approximately 60% of intracranial ependymomas are located in the posterior fossa, and 40% are found above the tentorium. The vast majority of infratentorial ependymomas occur in the fourth ventricle, while the majority of supratentorial ependymomas are intraparenchymal. The symptoms and signs of ependymomas are likely to be identical of medulloblastomas, but ependymomas usually have a longer duration of symptoms for 6–12 months, versus 4 months of medulloblastomas [59].
Brain Stem Glioma Brain stem gliomas account for 10–20% of pediatric central nervous system tumors, and 75% of these occur before the age of 20 years. The most common neurologic signs at diagnosis, in 75% of patients, are seventh nerve palsy, horizontal nystagmus, as well as cerebellar and pyramidal signs, unilateral or bilateral [60].
References 1. Stiles CD (1998) Cancer of the central nervous system. Review of an AACR special conference in cancer research with joint section on tumors of the AANS/CNS. Biochem Biophys Acta 1377:R1–R10 2. Gurney JG, Davis S, Severson RK et al (1996) Trends in cancer incidence among children in the U.S. Cancer 78:532–541 3. Kaatsch R, Haaf G, Michaelis JF (1995) Childhood malignancies in Germany methods and results of a nationwide registry. Eur J Cancer 31A:993–999 4. Miltenburg D, Louw DF, Sutherland GR (1996) Epidemiology of childhood brain tumors. Can J Neurol Sci 23:118–122 5. McKinney PA, Parsow RC, Lane SA et al (1998) Epidemiology of childhood brain tumors in Yorkshire, UK, 1974–95: geographical distribution and changing patterns of occurrence. Br J Cancer 78:974–979 6. Mosso ML, Colombo R, Giordano L et al (1992) Childhood cancer registry of the province of Torino, Italy. Cancer 69: 1300–1306 7. Pollack IF (1994) Brain tumors in children. N Engl J Med 331:1500–1507 8. Pollack IF, Claasen D, Al-Shboul Q et al (1995) Low grade gliomas of the cerebral hemispheres in children: an analysis of 71 cases. J Neurosurg 82:536–547 9. CBTRUS (1998) 1997 Annual Report Chicago, Central Brain Tumor Registry of the United States, Chicago
11
10. Counsell CE, Grant R (1998) Incidence studies of primary and secondary intracranial tumors: a systematic review of their methodology and results. J Neurooncol 37:241–250 11. Greig NH, Ries LG, Yancik R et al (1990) Increasing annual incidence of primary malignant brain tumors in the elderly. J Natl Cancer Inst 89:1621–1624 12. Lowry J, Snyder J, Lowry P (1998) Brain tumors in the elderly: recent trends in a Minnesota cohort study. Arch Neurol 55: 922–928 13. Mao Y, Desmeules M, Semenciw RM et al (1991) Increasing brain cancer rates in Canada. Can Med Assoc J 145: 1583–1591 14. Cote TR, Manns A, Hardy CR et al (1996) Epidemiology of brain lymphoma among people with or without acquired immunodeficiency syndrome. J Natl Cancer Inst 88: 675–679 15. DeAngelis LM (1998) Primary central nervous system lymphoma. In: Gilman S, Goldstein G, Waxman S (eds) Neurobase. Arbor, San Diego 16. Eby N, Gruffennan S, Flannelly CM et al (1998) Increasing incidence of primary brain lymphoma in the US. Cancer 62:2461–2465 17. Davis R, Malinski N, Haenszel W et al (1995) Primary brain tumor incidence rates in four U.S. regions, 1985–89: a pilot study. Neuroepidemiology 15:103–112 18. Davis F, McCarthy B, Jukish P (1999) The descriptive epidemiology of brain tumors. Neuroimaging Clin N Am 9: 581–594 19. Preston-Martin S (1996) Epidemiology of primary CNS neoplasms. Neurol Clin 14:273–290 20. Radhakrishnan K, Mokri B, Parisi JE et al (1995) The trends in incidence of primary brain tumors in the population of Rochester, Minnesota. Ann Neurol 37:67–73 21. Wemer MH, Phuphanich S, Lyman GH (1995) The increasing incidence of malignant gliomas and primary central nervous system lymphomas in the elderly. Cancer 76:1634–1642 22. Macfarlane R, King TT (1995) Acoustic neurinoma (vestibular schwannoma). In: Kaye AH, Laws ER Jr (eds) Brain tumors. An encyclopedic approach. Churchill Livingstone, New York, pp 577–622 23. DeMonte F, Al-Mefty 0 (1991) Meningiomas. In: Kaye AH, Laws ER Jr (eds) Brain tumors. An encyclopedic approach. Churchill Livingstone, New York, pp 675–704 24. Faglia G (1993) Epidemiology and pathogenesis of pituitary adenomas. Acta Endocrinol 129(suppl 1):1–4 25. Shaw EG, Scheithauer BW, O’Fallon JR et al (1992) Oligodendrogliomas: the Mayo Clinic experience. J Neurosurg 76:428–434 26. Packer RJ, Nicholson HS, Vezina LG et al (1992) Brainstem gliomas. Neurosurg Clin N Am 3:863–879 27. Sawaya R, Bindal RK (1995) Metastatic brain tumors. In: Kaye AH, Laws ER Jr (eds) Brain tumors. An encyclopedic approach. Churchill Livingstone, New York, pp 923–946 28. Thapar K, Laws ER (1997) Vascular tumors: hemangioblastomas, hemangiopericytomas and cavernous hemangiomas. In: Sheaves R, Jenkins PT, Wass JAH (eds) Clinical endocrine oncology. Blackwell Scientific, Oxford, pp 264–274 29. Villemure JG, de Tribolet N (1996) Epilepsy in patients with central nervous system tumors. Curr Opin Neurol 9:424–428 30. Bailey P, Cushing H (1926) A classification of tumors of the glioma group. Lippincot, Philadelphia
12 31. Kernohan JW, Sayere GP (1952) Tumors of the central nervous system. Atlas of tumor pathology, Section X, Fascicle 35. Armed Forces Institute of Pathology, Washington, DC 32. Graham D (1980) Primary malignant tumours of the cerebral hemispheres of adults. In: Thomas DG, Graham DI (eds) Brain tumours. Scientific basis, clinical investigation and current therapy. Butterwoths, London 33. Ringetz N (1950) Grading of gliomas. Acta Path Microbiol Scand 27:51–64 34. Fulling KH, Nelson JS (1984) Cerebral astrocytic neoplasms in the adult: contribution of histologic examination to the assessment of prognosis. Semin Diagn Pathol 1: 152–163 35. Burger PC, Vogel FS, Green SB et al (1985) Glioblastoma multiforme and anaplastic astrocytoma. Pathologic criteria and prognostic implications. Cancer 56:1106–1112 36. Kleihues P, Burger PC, Scheithauer BW (1993) Histological typing of tumours of the central nervous system. In: WHO international histological classification of tumours, 2nd edn. Springer Verlang, New York 37. Russel DS, Rubinstein LJ (1998) Pathology of tumors of the nervous system. In: Bigner DB, McLendon RE, Bruner JM (eds), 6th edn. Arnold, London 38. Burger PC (1986) Malignant astrocytic neoplasms: classification, pathologic anatomy and response to treatment. Semin Oncol 13:16–20 39. Rubinstein LJ (1972) Tumors of the central nervous system. Atlas of tumor pathology, Fascicle 6, Second Series, Armed Forces Institute of Pathology, Washington, DC 40. Russel DS, Rubinstein LJ (eds) (1989) Pathology of tumours of the nervous system, 5th edn. Williams & Wilkins, Baltimore 41. Hart MN, Earle KM (1973) Primitive neuroectodermal tumors of the brain in children. Cancer 32:890–897 42. Rorke LB (1983) Presidential address. The cerebellar medulloblastoma and its relationship to primitive neuroectodermal tumors. J Neuropathol Exp Neurol 42:1–15 43. Becker LE, Hinton D (1983) Primitive neuroectodermal tumors of the central nervous system. Hum Pathol 14: 538–555 44. Zulch KJ (1979) Histological typing of tumours of the central nervous system. International histological classification of tumours Geneva. World Health Organization, Geneva 45. Daumas-Duport C, Scheithauer BW, O’Fallon J et al (1988) Grading of astrocytomas. A simple and reproducible method. Cancer 62:2152–2165
G. Karkavelas and N. Tascos 46. Burger PC, Scheithauer BW (1994) Tumors of the central nervous system. Atlas of tumor pathology. Armed Forces Institute of Pathology, Washington, DC 47. Kleihues P, Cavenne WK (2000) Pathology and genetics. Tumours of the nervous system. World Health Organization classification of tumours. IARC, Lyon 47a. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK (2007) WHO classification of tumours of the central nervous system. IARC, Lyon 48. Janny P, Cure H, Mohr M et al (1994) Low-grade supratentorial astrocytomas. Management and prognostic factors. Cancer 73:1937–1945 49. Whitton AC, Bloom HJ (1990) Low grade glioma of the cerebral hemispheres in adults: a retrospective analysis of 88 cases. Int J Radiat Oncol Biol Phys 18:783–786 50. Gail MH, Pluda JM, Rabkin CS et al (1991) Projections of the incidence of non-Hodgkin’s lymphoma related to acquired immunodeficiency syndrome. J Natl Cancer Inst 83:695–701 51. Grant JW, Isaacson PG (1992) Primary central nervous system lymphoma. Brain Pathol 2:97–109 52. De Angelis LM (1994) Management of brain metastases. Cancer Invest 12(2):156–165 53. Young B, Patchell RA (1996) Brain metastases. In: Youmans J (ed) Neurological surgery, 4th edn. W.B. Saunders, Philadelphia, pp 2748–2760 54. Rohringer M, Sutherland GR, Louw DF et al (1989) Incidence and clinicopathological features of meningioma. J Neurosurg 71:665–672 55. National Institutes of Health (1991) Acoustic neuroma. Consens Statement 9:1–24 56. Recht LW (1995) Craniopharyngiomas. In: Gilman S, Goldstein G, Waxman S (eds) Neurobase. Arbor Publishing Corporation, San Diego 57. Bruce JN, Connolly EZ, Stein BM (1995) Pineal cell and germ cell tumors. In: Kaye AH, Laws ER (eds) Brain tumors. Churchill Livingstone, Edinburgh, pp 725–755 58. Packer RJ (1998) Medulloblastoma. In: Gilman S, Goldstein G, Waxman S (eds) Neurobase. Arbor, San Diego 59. Nazar GB, Hoffman HJ, Becker LE et al (1990) Infratentorial ependymomas in childhood: prognostic factors and treatment. J Neurosurg 72:408–417 60. Maria BL, Render K, Eskin TA et al (1993) Brainstem glioma: I. Pathology, clinical features and therapy. J Child Neurol 8:112–128
2
Imaging Modalities in Brain Tumors Antonios Drevelegas and Nickolas Papanikolaou
Contents 2.1 Introduction.................................................................. 13 2.2 Computed Tomography............................................... 13 2.3 Magnetic Resonance Imaging..................................... 15 References............................................................................ 32
A. Drevelegas (*) Department of Radiology, Aristotle University of Thessaloniki, Thessaloniki, Greece e-mail:
[email protected] N. Papanikolaou Department of Radiology, University Hospital of Heraklion, Medical School of Crete, Greece
2.1 Introduction Imaging plays an important role in the evaluation of patients with brain tumors. CT and MRI represent the two most important and commonly used imaging modalities. They have a significant impact on patient care. The technical improvement of CT and MRI, the utility of contrast material in the imaging of brain tumors as well as the introduction of new imaging techniques, improved significantly the detection and the evaluation of brain neoplasms. Once a brain tumor is clinically suspected, radiologic evaluation is required to determine the location, the extent of the tumor and its relationship to the surrounding structures. This information is very important and critical in deciding between the different forms of therapy such as surgery, radiation, and chemotherapy. In this chapter we will give an overview of the role of CT and MRI in the diagnosis of brain tumors. New imaging techniques that evaluate tissue blood flow (perfusion imaging), water motion (diffusion imaging), brain metabolites (Proton magnetic resonance spectroscopy) and blood oxygen level dependent (BOLD) imaging have also been included.
2.2 Computed Tomography Computed tomography (CT) was introduced in the clinical practice in 1972 and rapidly became a very important factor in the radiological diagnosis. With the advent of CT in neuroradwiology direct images of the brain could be produced, and a new era in cerebral studies started. CT of the brain, which became the procedure of choice for evaluation and diagnosis of brain tumors, has
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_2, © Springer-Verlag Berlin Heidelberg 2011
13
14
replaced invasive procedures such as pneumoencephalography or cerebral angiography. Progressive improvement of the image quality, reduction of costs, and reduction of scan times has resulted in significant expansion of CT applications. The utility of contrast material in the imaging of the brain improved the efficacy of CT in the diagnosis of brain tumors. Enhancement is the increased difference in an imaging characteristic between a lesion and surrounding normal tissue after administration of contrast agent. This is due to the disruption of the blood–brain barrier (BBB) of the tumor vessels, which permits the passage of the contrast material into the extracellular spaces of the tumor (Fig. 2.1). On CT, the increased concentration of the contrast material within the tumor interstitium results in higher attenuation values within the tumor than in the surrounding brain. The majority of the brain tumors enhance after the administration of contrast material. The enhancement characteristics of different types of brain tumors will be discussed in the following chapters. Progress in CT development continued rapidly and new technology has revolutionized the field.
a
Fig. 2.1 Enhancement of tumor after IV contrast administration. (a) CT before and (b) after contrast administration. The neoplasm is clearly demonstrated on post-contrast CT. The con-
A. Drevelegas and N. Papanikolaou
Spiral and multislice CT allow faster acquisition times with substantially improved 3D spatial resolution. CT angiography provides images of excellent quality in a noninvasive way and is of great importance in the assessment of the relationship between the tumor and the vessels. Perfusion imaging techniques enable accurate measurement of CBV and CBF values in a variety of clinical and experimental settings [1]. CT-guided stereotactic biopsy is a reliable method for histological diagnosis of brain tumors and showed to be valuable in planning the appropriate treatment for each patient. Although MR is the main diagnostic tool for diseases of the central nervous system, CT is still a valuable modality in the imaging of brain tumors. CT is superior in detecting calcification, hemorrhage, and in evaluating bone changes related to a tumor (Fig. 2.2). Patients with pacemakers or metallic devices as well as critically ill, pediatric or unstable patients represent some of the specific areas where CT is the diagnostic modality of choice [2].
b
trast material identifies areas of BBB disruption facilitating the detection of the neoplastic tissue
15
2 Imaging Modalities in Brain Tumors
a
b
Fig. 2.2 CT versus MRI in calcified meningioma. (a) Axial CT. (b) Axial T1-weighted MRI. Densely calcified tumor is clearly demonstrated on CT. Most calcification is isointense to brain on T1WI
2.3 Magnetic Resonance Imaging Magnetic resonance imaging (MRI) is the modality of choice for evaluating patients who have symptoms and signs suggesting a brain tumor. Its multiplanar capability superior contrast resolution and flexible protocols allow it to play an important role in assessing tumor location and extent, in directing biopsies, in planning the proper therapy, and in evaluating the therapeutic results. The standard protocol most commonly used between institutions includes: spin-echo T1-weighted image (T1WI), proton density-weighted image (PDWI), T2-weighted image (T2WI), and T1WI after the administration of paramagnetic agent. Most brain tumors have prolonged T1 and T2 relaxation times and appear hypointense relative to normal brain on T1WI and hyperintense on T2WI. On PDWI the tumors show intermediate hyperintensity. However, the presence of fat, hemorrhage, necrosis, and calcification are responsible for the heterogeneous appearance of some tumors. As in CT, the utility of contrast material in MRI facilitates the detection of many brain tumors and can help distinguish some tumors from the adjacent normal brain parenchyma. The MRI contrast
agents most commonly used for central nervous system (CNS) tumor imaging are gadolinium (Gd) chelates. Although in normal brain Gd cannot pass from intravascular compartment to the interstitial space, in brain tumors, where the normal BBB may be disrupted, Gd is accumulated into the extracellular space of the tumor. As result in post-contrast T1WI the tumor becomes brighter than the surrounding normal brain tissue due to the shortening of T1 relaxation time. However, histologic examination of samples obtained of patients with brain tumors showed that there are regions with tumor cells outside the Gd-enhancing area. The accepted standard dose for Gd is 0.1 mmol/kg and has proved to be valuable in the evaluation of CNS tumors (deletion). In an attempt to improve the delineation of the extent of primary brain tumors as accurately as possible to guide potential surgical or radiation therapy, several studies have been performed and have shown that the administration of higher doses of contrast agent improved significantly the enhancement of most intracranial tumors [3–7] (Fig. 2.3). This has important therapeutic implications because the zone of glioma cells delineated by enhancement after high-dose Gd likely is a better estimate of microscopic tumor extent [8].
16
a
A. Drevelegas and N. Papanikolaou
b
Fig. 2.3 Comparison of standard and high dose contrast material. (a) T1WI after administration of 0.1 mmol/kg Gd shows a ring-like enhanced tumor in the right parietal lobe. (b) After
high dose of Gd. the ring-like enhancement of the tumor tissue is thicker and sharper
In brain metastases better lesion delineation and increase in the number of visible metastases is achieved by using double or triple doses of Gd [4, 9]. However, increasing the dose of contrast medium unfortunately increases imaging costs. In order to minimize the cost of the contrast agent, new sequences have been introduced as the standard dose magnetization transfer (MT) T1-weighted imaging which is equally effective as the triple dose T1-weighted imaging in terms of lesions conspicuity and detectability (Fig. 2.4). This magnetization transfer T1-weighted sequence is generated by suppressing the signal of the (nonenhancing) background tissue either by applying an off-resonance radio frequency (RF) prepulse or binomial on-resonance pulseto preferentially saturate bounded protons, which then transfer magnetization to mobile free protons. As a result, the signal of the white matter will be reduced since an increased amount of bounded protons in the myelin is present while the signal of pathologic tissue, containing more free protons, will remain unchanged; therefore, it will be presented with higher conspicuity [10]. In tumors MT improves the accuracy of tumor classification and allows differentiation between low-grade astrocytomas, hemangioblastomas and craniopharyngiomas [11]. In addition, MT T1WIs may be used in postoperative patients to
define enhancing residual tumor not seen on standard T1WIs. A drawback of MT images is the lower sensitivity in depicting cerebral edema. As information regarding edema is much more readily available on T2WIs, we do not consider this a major disadvantage of the MT T1WIs [10]. Although the traditional spin-echo MR sequences in conjunction with the post-contrast MRI are clearly effective in detecting and delineating brain neoplasms, an additional number of MR techniques have been applied in an attempt to improve the diagnostic efficacy for tumor imaging both before and after treatment. These MR techniques may ultimately supplant conventional MR spin-echo imaging and are designed to produce a high level of contrast (instead of a certain contrast) and improved image quality and data acquisition. They are fast spin echo (FSE), inversion recovery (IR), short tau inversion recovery (STIR), fluid attenuated inversion recovery (FLAIR), gradient echo pulse sequences, and echo-planar imaging (EPI). FSE is a spin-echo pulse sequence but with scan times shorter than the conventional spin echo. Since the scan time is greatly reduced FSE sequence allows greater patient throughout, which may be critical in clinical practice. FSE imaging is equal to SE imaging in the detection of white matter lesions larger than 5 mm
17
2 Imaging Modalities in Brain Tumors
a
b
Fig. 2.4 Effect of MT image on the detection of brain metastasis. (a) Post-contrast T1WI shows a large enhancing lesion. (b) MT image shows an additional small lesion (arrow). Note also that MT image clearly delineates the contour of the lesion (arrowhead)
and is slightly less sensitive in the detection of smaller than 5 mm lesions [12]. Thus FSE sequence offers a faster alternative to conventional spin-echo in routine MRI of the brain. There are some differences between FSE and conventional SE images in terms of contrast, and these can be summarized into: (a) brighter fat appearance due to J coupling effects, (b) increased MT effects that result in darker appearance of normal white matter, and (c) less sensitivity to hemorrhagic lesions due to the presence of multiple refocusing pulses. IR is a pulse sequence that begins with a 180° inverting pulse followed by a 90° excitation pulse, and by a 180° refocusing pulse. IR can be used to produce heavily T1WIs to demonstrate anatomy. In IR images the white matter has a short T1 and appears white, the gray matter has a longer T1 and appears gray and the cerebrospinal fluid has a very long T1 and appears dark. This sequence provides an excellent gray–white matter contrast, which is important in localization and assessing mass effects (Fig. 2.5). STIR is an IR sequence with a short inversion time ranging from 130 to 200 ms depending on the field strength and is used to achieve suppression of the fat signal in a T1WI. Spin preparation not only eliminates
the signal from fat, it also adds inverted T1 contrast to the image. Tissue with a long T1 appears brighter than tissue with short T1. STIR should not be used in conjunction with contrast because the signal from the enhancing tissue may be nulled. FLAIR imaging is another variation of the IR sequence with an inversion time ranging from 2,000 to 2,500 ms and may be used to suppress the high CSF signal in T2- and proton density-weighted images so that the pathology adjacent to the CSF is seen more clearly. The suppression of the CSF signal is achieved by applying an inversion pulse with a long recovery time between this pulse and the start of the measurement. With this sequence, CSF artifacts are reduced and heavily T2WIs are obtained with a long echo time. FLAIR images enable better delineation of the lesions adjacent to the ventricles. Additionally, subtle lesions near the cortex stand out against a background of attenuated CSF [13]. FLAIR images provide better definition between edema and tumor. Cerebral edema associated with brain tumors is also better delineated on FLAIR image. Therefore, they may be used as an adjunct to T2-weighted or proton density-weighted spin-echo
18
a
A. Drevelegas and N. Papanikolaou
b
Fig. 2.5 The contrast between gray and white matter is significantly improved on the coronal inversion recovery image (a) compared to the conventional T1-weighted spin-echo image (b)
images [14]. Contrast-enhanced FLAIR MR imaging has been successfully used by taking advantage of the T1 effect to achieve a particularly high contrast between tumor and background tissue [15]. They allow an exact delineation of enhancing and nonenhancing tumor parts in one sequence (Fig. 2.6). Although FLAIR technique is simple to implement, its disadvantages include long imaging times and a limited number of sections. In gradient echo pulse sequence the 180° refocusing pulse is omitted and a flip angle other than 90° is used. After the RF pulse is withdrawn, the free induction decay (FID) signal is immediately produced due to inhomogeneities in the magnetic field and T2* dephasing occurs. The magnetic moments within the transverse component of magnetization dephase, and are then rephased by a gradient .The gradient rephases the magnetic moments so that a signal can be received by the coil, which contains T1 and T2 information and is called gradient echo [16]. In gradient echo pulse sequence, the repetition time (TR) is reduced due to the absence of 180° rephasing pulse. The TR can also be reduced because flip angles other than 90° can be used. As a consequence, the imaging time is reduced and the motion artifacts are decreased. Therefore, gradient echo pulse sequences (instead of they) can be valuable for examining
critically ill, anxious or uncooperative patients whose conventional or fast spin-echo images show considerable motion artifacts [17]. Gradient echo images are very sensitive to flow, produce angiographic types of images, and may be used to clarify focal or linear regions of signal void within a mass whether they represent dense calcification or flow within tumor vessels. Calcified neoplasms in gradient-echo images appear as focal regions of signal void, while intratumoral vessels appear as round or linear areas of high signal intensity. Gradient echo pulse sequences are also very sensitive in the detection of hemorrhage. They are also particularly suited to 3D imaging, which is used when high resolution and thin contiguous slices are required. 2D and 3D GRE sequences are essential for time-of-flight MR angiography (MRA). The most important disadvantage is that there is no compensation for magnetic field inhomogeneities, and therefore, they are very sensitive to magnetic susceptibility artifacts. The steady state is a GE pulse sequence where the TR is shorter than the T1 and T2 times in tissues. In the steady-state sequence coexist both the longitudinal and the transverse magnetizations. Fast imaging with steady precession (FISP) and constructive interference of steady state (CISS) are steady-state gradient-echo techniques that produce heavy T2-weighting images. The
19
2 Imaging Modalities in Brain Tumors
a
b
c
d
Fig. 2.6 Left parietal glioblastoma. (a) Post-contrast T1WI shows an irregular ring-like enhancement. (b) T2WI shows a high signal mass surrounded by peritumoral edema (c) FLAIR image shows the mass and the peritumoral edema which is more
prominent than on T2WI (d). Post-contrast FLAIR image clearly demonstrates the ring-like enhanced tumor (arrows) as well as the surrounding edema
20
Fig. 2.7 Axial 3D CISS image shows clearly the facial (arrowhead) and the vestibular nerve (arrow)
CISS sequence is used for the imaging of basal cisterns and/or the discrimination of the facial-vestibulocochlear nerve complex [4] (Fig. 2.7). EPI is the fastest MR imaging technique and is achieved by means of rapid gradient switching, which maps all phase and frequency points in K-space during a single echo period. It allows one to collect all the data required to reconstruct an image from a single RF excitation. Individual images may be acquired on the order of 50–100 ms and so an entire brain survey can be completed in as little as 1 s. To keep the total time for data collection brief, gradients with high slew rate are used. In EPI can be used any combination of RF pulses used in conventional spin-echo technique. Alternatively a T2*-dependent gradient echo imaging can be applied (GRE EPI). An echo-planar image can be obtained either with a single-shot technique, where all data are collected after one excitation or with a multi-shot technique in which K-space is broken up into several sections and each section is scanned during subsequent TRs. With single-shot EPI a study of the entire brain can be performed in as little as 2 s [18, 19]. However, the sensitivity of single-shot EPI is lower compared with proton-density and T2-weighted conventional spinecho imaging for the detection of small brain lesions. Multi-shot EPI proved to be superior to single shot echo planar sequences in terms of lesion conspicuity and delineation [20]. In a study single-shot EPI depicted up to 70% of multiple sclerosis lesions larger than 1cm and only 23% of smaller lesions (<5 mm) [21]. When multi-shot
A. Drevelegas and N. Papanikolaou
echo-planar technique is used, the sensitivity for lesion detection increases to 98% for lesions larger than 1cm and 77% for lesions smaller than 5 mm. Despite the increased spatial resolution provided with multi-shot technique, EPI is still suffering from decreased spatial resolution, poor fat suppression, and increased ghosting and susceptibility artifacts (signal loss and geometric distortion). Therefore, conventional spin-echo or fast spin-echo imaging remains the preferred imaging technique. However, the greater sensitivity of echoplanar images to magnetic susceptibility variations makes them more sensitive to small amounts of hemorrhage in tissue. In addition, EPI reduces imaging time and motion artifacts allowing the MR examination of uncooperative claustrophobic and pediatric patients. An alternative approach to the use of EPI sequences is to combine EPI and FSE techniques to produce a combined gradient-echo (GRASE) image. The GRASE sequence produces reasonable quality T2WIs. At present, GRASE does not provide the image quality and contrast spectrum of conventional or fast spin-echo sequences; nevertheless, it might be useful for uncooperative patients whose conventional or spin-echo sequences show considerable motion artifacts [17]. The combination of half-Fourier acquisition and single-shot turbo spin-echo (HASTE) offers also a rapid imaging technique. This sequence is T2-weighted and is excellent for rapid screening of the brain. The introduction of EPI and other subsecond imaging techniques has allowed the fast progression of functional MR imaging (fMRI). Diffusion, perfusion, and BOLD functional MR imaging allows a better understanding of pathophysiology of various pathologic states. Cerebral ischemia, brain tumors, multiple sclerosis, metabolic diseases, and neurocognitive disorders represent a spectrum of diseases where fMRI provides useful diagnostic information and may allow better monitoring of the effects of therapy. Diffusion-weighted imaging is a unique tissue contrast technique based on the diffusion of water molecules, which move along random pathways (Brownian motion). Usually a spin-echo EPI sequence is utilized for DWI, where a pair of identical gradient pulses is added before and after the application of the refocusing pulse. The physical principle behind DWI is analogous to that of phase contrast MRA, although PCA is referred to macroscopic motion while DWI is referred to microscopic motion. More specifically, the first gradient pulse induces phase shifts to water molecules
2 Imaging Modalities in Brain Tumors
while the second gradient cancels the phase shifts by fully rephasing the spins. Moving molecules will acquire a phase shift due to their motion during the time interval in between the application of diffusion gradients. In other words, stationary spins will be fully refocused that means no phase shifts, while moving spins will be partially refocused that means a specific amount of phase shift that causes signal loss. The signal amplitude for the MR signal is exponential and is given by the equation Signal = Soexp (−bD), where So is the attenuation factor, D is the diffusion coefficient of tissue, which characterizes the rates of diffusional motion, and b is the diffusion coefficient factor. The diffusion coefficient is dependent on a number of factors including time, orientation of the imaging plane, tissue being imaged, and the energy state of the imaged tissue [22]. In biologic systems, factors such as perfusion, water transport, or bulk motion might contribute to the signal loss, so that the term apparent diffusion coefficient (ADC) is used instead of diffusion coefficient. The ADC can be calculated on a pixel-by-pixel basis, allowing the generation of a parametric map that reflects the diffusion influence eliminating the T2 effects, which prevents misinterpretation from the so-called T2 shine through effect [21, 22]. Differences in ADC are related to changes in cellularity, cell membrane permeability, intracellular and extracellular diffusion, and tissue structure. Diffusion-weighted MR imaging is a powerful tool in characterization of brain neoplasms. Tumor cellularity and tumor grade have been correlated with ADC values. Brain neoplasms with higher cellularity or higher grades show a significant reduction in the rate of the ADC and a marked increase in the signal of diffusion-weighted images (Fig.7.5). DWI can also be used in assessing high cellularity of other neoplasms. Lymphoma is a hypercellular tumor that has been found to present with high signal intensity on DWI and low ADC values. Medulloblastoma is a primitive neuro ectodermal tumor that also shows restricted diffusion pattern due to the densely packed cells and high nuclearto-cytoplasmic ratio. Diffusion-weighted MR images can be used to discriminate the tumor tissue from edema, cyst, or necrosis. The cystic or necrotic portion of the tumor in relation to the normal brain parenchyma appears hypointense on diffusion-weighted images and show much higher ADC values, whereas the areas of enhancing tissue on conventional MRI show high signal intensity on DWI [23–25] (Fig. 2.8).
21
Atypical and malignant meningiomas also tend to be markedly hyperintense on diffusion-weighted MR images and exhibit lower ADC values, while benign meningiomas have a variable appearance on diffusionweighted images and higher ADC values compared with normal brain, with the exception of densely calcified or psammomatous meningiomas, which have low ADC values [26]. Diffusion-weighted imaging may also be used to differentiate brain abscess from necrotic or cystic tumor. The abscesses show high signal intensity on DWI and low ADC values due to the presence of the pus, a viscous material that consists of inflammatory cells, debris, and fibrinogen, which leads to reduced water mobility within the cavity [27]. Epidermoid and arachnoid cysts can also be discriminated on the basis of diffusion-weighted images. On conventional spin-echo images, both show long T1 and T2. On diffusion-weighted images, epidermoid cysts show high signal intensity due to restricted motion of protons by the presence of membranes of densely layered epithelium, while arachnoid cysts are hypointense due to their free water motion [28] (Fig. 2.9). Since diffusion is a 3D process, to acquire more detailed information about anisotropic diffusion properties of the underlying tissue, simple standard DWI is not sufficient. Diffusion tensor imaging (DTI) is a more direct imaging technique to study microarchitecture of brain tissue. Through the application of diffusion sensitization in at least 6 non-colinear directions, it is possible to extract the diffusion tensor and quantify physical parameters like fractional anisotropy or mean diffusivity. Fractional anisotropy reflects the directionality of tissue water diffusion, and therefore, the degree of alignment or integrity of tissue structure within a given voxel (Fig. 2.10a). Mean diffusivity is a measure of overall diffusion in a voxel; its magnitude depends on the total size of extracellular space and the existence of diffusion barriers such as cell membranes or myelin sheaths [29–31] (Fig. 2.10b). MR DTI allows identification and characterization of white matter tracts according to the direction and degree of their anisotropic water diffusion. Quantifying the degree of anisotropy in terms of metrics such as the fractional anisotropy offers insight into white matter development and degradation. There are fractional anisotropy changes in the white matter of brain neoplasms that might indicate cellular infiltration beyond the area of the tumor enhancement. Within the tumor
22
a
A. Drevelegas and N. Papanikolaou
b
c
Fig. 2.8 Cystic metastasis. (a) T1WI shows a hypointense lesion in the right parietal lobe. The thick capsule of the metastatic lesion is isointense to the gray matter (arrowheads).
(b) On the diffusion-weighted image the lesion is hypointense. (c) Gross specimen of the lesion shows a central necrotic area surrounded by a thick capsule
center white matter fibers are displaced by cellular infiltration and fractional anisotropy is reduced, whereas in the periphery and in a narrow rim of white matter rim surrounding the tumor, this parameter could preserve or even increased by fiber compression due to space occupying effect of the tumor [32, 33].
Technical advances in diffusion-weighted imaging enable the assessment of the water diffusivity in 3D space [34, 35]. In vivo, white matter tracts show comparatively higher diffusion anisotropy because water diffusion is more facilitated along the direction of the fibers as compared to diffusion behavior perpendicular
23
2 Imaging Modalities in Brain Tumors
a
b
Fig. 2.9 Epidermoid cyst. (a) T1WI shows low signal intensity. (b) Diffusion-weighted image shows high signal intensity indicative of restricted diffusion
Fig. 2.10 (a) Fractional anisotropy map of a normal brain presenting with high signal intensity WM areas of dense myelin like splenium and genu of corpus callosum, while CSF and gray matter exhibit low signal intensity. This can be explained in the basis of local diffusion patterns, where in WM diffusion anisotropy is
dominating therefore FA value increases, while in gray matter diffusion becomes isotropic resulting in FA reduction. (b) Mean diffusivity of ADC map showing CSF bright due to increased water mobility in fluids, while more solid tissues like gray and white matter presents with significantly lower ADC values
24
to the fibers [34, 35]. In this context, it is possible to generate 3D representations of the major white matter tracts like corpus callosum (Fig. 2.11). The first step concerning data processing is to estimate each diffusion matrix component values utilizing multiple linear regression methods. The diagonalization of the diffusion tensor provides eigen vectors and eigen values, which correspond respectively to the main diffusion directions and associated diffusivities. Consequently, various indices can be calculated by using combinations of the eigen values. The most commonly used in
Fig. 2.11 3D representation of the corpus callosum based on diffusion tensor data acquired in 64 directions. The tractography algorithm that was used is the second order Runge–Kutta [55]
Fig. 2.12 Diffusion properties of CSF can be modeled by almost a sphere due to low anisotropy. On the contrary posterior corona radiata is modeled by elipsoids typical for anisotropic diffusion. Note that the size of either the sphere or the elipsoid
A. Drevelegas and N. Papanikolaou
clinical practice are the mean diffusivity which characterize the mean squared displacement of the water molecules and this parameter is rotationally invariant that means that is independent of the orientation of the reference frame and fractional anisotropy index that describes how much molecular displacements vary in space (ellipsoid eccentricity). In other words, it reflects the degree of alignment of cellular structures within fiber tracts and their structural integrity (Fig. 2.12). Streamline or deterministic fiber tracking algorithms are based on the computation of the eigen system (eigen vectors and their corresponding eigen values) for each voxel. A fiber bundle has a strong anisotropy due to the alignment of the fibers and the local tensors of that bundle have a first eigen value much greater than the others. The diffusion of water protons is most important along the direction first eigen vector (Fig 2.13). Consequently, the detection of strongly anisotropic voxels (based on a threshold value of fractional anisotropy) and representation of the eigen vectors on these voxels can be used to map the white matter fibers since the direction of the first eigen vector is the same with the long axis of the fibers [36]. A major problem of the latter approach is the crossing fibers in the dimensions of a single voxel. In that case, the algorithm fails to delineate two crossing fibers
represents the value of ADC while the shape represents the type of diffusion, ellipsoid in case of anisotropic diffusion and sphere in case of isotropic diffusion
2 Imaging Modalities in Brain Tumors
25
Fig. 2.13 Water mobility is facilitated along the direction of the tracts; therefore the principle diffusion eigenvalue l1 is considerably higher than the l2 and l3 that describe water mobility perpendicular to the fibers
and local errors in the generation of tracts might take place. An alternative strategy to deterministic streamline algorithms is probabilistic tractography. With the latter techniques, it is possible to track fibers in areas of high uncertainty; either these represent areas with low FA values (Fig. 2.14) or areas where crossing of fibers is taking place. Perfusion-weighted imaging provides information about the perfusion status of microcirculation. This technique requires the dynamic intravenous administration of a MR contrast agent. As the paramagnetic contrast agent passes through the intravascular compartment local field inhomogeneities are created that result in magnetic susceptibility effects with a decrease in signal on T2*-images that can be measured. This signal drop depends on both the vascular concentration of contrast agent and the concentration of small vessels per voxel tissue [37, 38]. Changes in signal intensity may be used to calculate an image of the relative cerebral blood volume (rCBV). Echo planar MR imaging systems, which use strong rapidly switching magnetic field gradients, permit the fast simultaneous acquisition of multiple T2-weighted slices during the administration of contrast material (Fig. 2.15). In brain tumor, rCBV maps are particularly sensitive for depicting the microvasculature of a tumor and therefore its aggressiveness and proliferative potential. Previous
studies correlating histopathologic grading of gliomas with rCBV showed a positive correlation of rCBV with tumor grading. Especially low-grade gliomas had homogeneous low rCBV, while high-grade tumors exhibited varying degrees of high rCBV [39, 40] (Fig. 4.2). rCBV maps may also be used to delineate tumor margins as well as to differentiate tumor recurrence from enhancing non-neoplastic tissue such as radiation necrosis which would be useful for surgical planning and targeting of biopsies (Fig. 2.16) and radiation therapy [18] (Fig. 2.17). An enhancing lesion with a normalized rCBV ratio higher than 2.6 or lower than 0.6 may suggest tumor recurrence or non-neoplastic contrast enhancing-tissue, respectively [38, 41]. Perfusion MR may also be used in AIDS patients, to differentiate toxoplasmosis from lymphoma. In toxoplasmosis, the surrounding edema shows vasoconstriction with reduced rCBV, while in lymphomas there are areas with increased cerebral blood volume correlating, with hypervascularity of active neoplastic tissue [42, 43] Conversely, a recent study of perfusion MR imaging in eight patients with lymphoma showed that cerebral lymphomas had a tendency to have low rCBV values [44]. Thus, rCBV mapping may be of limited value in grading lymphoma patients. Finally, perfusion MR imaging may be used in evaluating the pathological changes of chemotherapy in patients with brain tumors.
26
A. Drevelegas and N. Papanikolaou
a
b
Fig. 2.14 (a) Patient with low-grade glioma. Three dimensional representation of cingulate bundles based on deterministic tractography algorithm. The left cingulate is barely seen, most probably, due to lower FA value than the threshold (0.2) utilized by the
algorithm to avoid erroneous fiber drawing [55]. (b) Probabilistic tractography demonstrate both cingulate bundles. The left cingulate is shown intact but displaced from the tumor [56]
Cortical activity may be studied by fMRI techniques that are mostly based on the detection of the focal blood flow and oxygenation changes following neuronal activity. The BOLD effect is the most commonly used to study cortical function in the brain. Neural activation leads to an increase in local blood flow and
thus to an increase of oxygenated hemoglobin in the capillaries of activated brain tissue. As a result the oxy/ deoxygenated blood ratio is increased. The drop in the concentration of the paramagnetic deoxyhemoglobin leads to a focal signal increase in the affected tissue using T2*-weighted sequences [19, 45]. This effect is
2 Imaging Modalities in Brain Tumors
Fig. 2.15 (a) Serial echo planar images of the brain during contrast administration. Single slice images with 1 s interval between the images. From left to right the T2 signal intensity is gradually dropped due to the T2* susceptibility effect of gadolinium. (b) MR signal versus time curves show signal drop with
27
passage of bolus of contrast material. Upper and lower curves correspond to regions of interest drawn in white and gray matter, respectively. Note greater decrease in signal intensity in gray matter compared with the white matter
28
A. Drevelegas and N. Papanikolaou
known as BOLD effect and was first used to show functionally activated brain regions as result of sensory or motor stimulation [46]. In order to depict the
BOLD effect, echo planar sequences must be used due to a very short time window of the BOLD effect and the reduced sensitivity to motion that EPI sequences may provide with. BOLD imaging can be useful for many applications such as: localization of neural activities in the brain, display areas of the brain activate by sensory or motor activation (Fig. 2.18), and as a noninvasive tool for the presurgical mapping of cortical function in patients with intracranial tumors [47–49] (Fig. 2.19). Therefore, functional MRI can contribute to more efficient surgical removal of both benign and malignant brain tumors with an increase in patient survival and a decrease in surgical morbidity [50]. Proton magnetic resonance spectroscopy (1HMRS) is a noninvasive method to study various chemical compounds found on human brain tissue. It has been demonstrated that this technique provides with biochemical information, which can be useful in differentiating normal from abnormal brain tissue and in a certain extent provide information that might be important for differential diagnosis. Two different approaches have been implemented, namely, single voxel spectroscopy (SVS) and chemical shift imaging
Fig. 2.16 Patient with glioblastoma multiform. In the upper row, a series of postgadolinium T1-,weighted consecutive axial slices showing peripheral enhancement at the genu of corpus
callosum. Biopsy guidance can be more accurately performed when rCBV maps are utilized since an area of considerably higher neovascularity is only depicted on rCBV maps (arrow)
b
400 350 300
MR Signal
250 A: 0.4 cm2 A: 0.4 cm2
200
150
100
15
30 Time (sec)
45
60
Fig. 2.15 (continued)
2 Imaging Modalities in Brain Tumors
29
Fig. 2.17 Patient with high-grade glioma after surgery and radiation therapy. Two enhancing areas are shown on postgadolinium T1-weighted axial images (upper row, arrows). On rCBV maps, the lesion located posterior to the surgical cavity presents with high rCBV value (red color) that corresponds to tumor recurrence, while the lesion anteriorly to the surgical cavity is manifested with low rCBV that is compatible with radiation necrosis
(CSI). According to the first, a 3D area or volume of tissues is excited and the signals detected from this volume are transformed to a spectrum. In the second technique, multiple voxels are utilized either in a plane (2D CSI) or in a volume (3D CSI); therefore,it is possible to study larger areas with a single experiment. Metabolic maps can be calculated based on the information derived from each voxel. SVS produces a single spectrum from a single voxel that is typically 8 cm3 in volume, whereas CSI measures spectra from multiple voxels that are typically 1–1.5 cm3 in volume. CSI data may be presented in a variety of displays including individual spectra, spectral maps, or colored metabolite images overlaid on anatomical images (Fig. 2.20).
In MR spectroscopy, different echo time (TE) values can be utilized to control the “T2 contrast” of spectral peaks in the same way tissue T2 contrast is controlled in conventional imaging sequences. Metabolites with short T2 relaxation times decay faster, and the corresponding spectral peaks are not seen on long TE spectra. This type of metabolites can only be detected on short TE acquisitions. The major healthy brain metabolite peaks that are seen on long TE spectra include N-acetyl aspartate (NAA) at 2.02 and 2.6 ppm which is a neuronal marker, choline (Cho) at 3.20 ppm which is a membrane marker, and creatine (Cr) at 3.02 and 3.9 ppm which is an energy marker and generally is stable. Short TE spectra contain additional peaks, which include glutamine and glutamate (Glx) between 2.05–2.5 ppm and
30
A. Drevelegas and N. Papanikolaou
Fig. 2.18 Motor cortex activation during a finger tapping experiment in a patient with a high-grade tumor. Areas of activation are superimposed on axial T2-weighted images
3.65–3.8 ppm, scyllo-inositol (sI) at 3.36 ppm, glucose at 3.43 and 3.8 ppm, and myo-inositol (mI) at 3.56 and 4.06 ppm which is a glial marker. The most important advantages of SVS over CSI are: (a) shorter acquisition times, (b) better localized shimming, (c) simpler in terms of post-processing and precise volume definition. The most important disadvantage is that SVS can provide spectra only from one voxel and it can be time consuming when multiple, remote areas should be evaluated. In many disease processes, biochemical changes are preceding morphologic alterations in tissues, therefore MR spectroscopy is a powerful technique to identify early changes comparing to conventional MRI morphologic techniques. As a general rule, brain gliomas show increase of Cho and decrease of NAA
peaks compared to normal brain tissue (Fig. 2.21). According to tumor grading the relative Cho/Cr and Cho/NAA ratios show significant increase from low to high-grade gliomas. The most important clinical applications of MR Spectroscopy, either as a stand-alone technique or in combination with diffusion and perfusion weighted imaging techniques, can be summarized into the differentiation between (a) low and high-grade gliomas, (b) radiation induced necrosis and tumor recurrence, (c) primary and secondary malignant tumors and (d) abscesses and tumors. Another important clinical application of MR spectroscopy is the assessment of the therapeutic outcome by performing a baseline evaluation and follow-up experiments to identify therapeutic-induced changes and guide the therapeutic scheme [51–54].
2 Imaging Modalities in Brain Tumors Fig. 2.19 Language centers activation with fMRI in a patient with a high-grade tumor. Activation areas are evident both in Wernicke (yellow arrow) and Broca (blue arrow) areas
Fig. 2.20 (a) Choline (Cho) over creatine and (b) lipids metabolite maps generated with chemical shift imaging experiment
31
32
a
A. Drevelegas and N. Papanikolaou
b
Fig. 2.21 Anaplastic astrocytoma shows high signal intensity on axial FLAIR image (a) and elevated choline and reduced NAA on the corresponding spectroscopic image (b)
References 1. Nabavi DG, Cenic A, Craen RA et al (1999) CT assessment of cerebral perfusion: experimental validation and initial clinical experience. Radiology 213:141–149 2. Whelan HT, Clanton JA, Wilson RE et al (1988) Comparison of CT and MRI brain tumor imaging using a canine glioma model. Pediatr Neurol 4(5):279–283 3. Runge VM, Kirsch JE, Burke VJ et al (1992) High dose gadoteridol in MR imaging of intracranial neoplasm. J Magn Reson Imaging 2:9–18 4. Yoursy I, Camelio S, Schmid UD et al (2000) Visualization of cranial nerves I-XII: value of 3D CISS and T2 –weighted FSE sequences. Eur Radiol 10(7):1061–1067 5. Yuh WT, Fisher DJ, Engelken JD et al (1991) MR evaluation of CNS tumors: dose comparison study with gadopentate dimeglumine and gatoteridol. Radiology 180:485–491 6. Yuh WT, Fisher DJ, Runge et al (1994) Phase III multicenter trial of high-dose gadoteridol in MR evaluation of brain metastases. AJNR Am J Neuroradiol 15:1037–1051 7. Yuh WT, Nguyen HD, Tali ET et al (1994) Delineation of gliomas with various doses of MR contrast material. AJNR Am J Neuroradiol 15:983–989 8. Abdulach ND, Mathews VP (1999) Contrast issues in brain tumor imaging. Neuroim Clin North Am 9(4):733–749 9. Van Dijk P, Sijens PE, Schmitz PIM et al (1997) Gd-enhanced MR imaging of brain metastases: contrast as a function of dose and lesion size. Magn Reson Imaging 15:535–541 10. Knauth M, Forsting M, Hartmann M (1996) MR enhancement of brain lesions: increased contrast dose compared
with magnetization transfer. AJNR Am J Neuroradiol 17:1853–1859 11. Kurki T, Niemi P, Valtonen S (1995) Tissue characterization of intracranial tumors: the value of magnetization transfer and conventional MRI. Neuroradiology 37:515–521 12. Olson EM, Healy JF, Wong WHM et al (1994) MR detection of white matter disease of the brain in patients with HIV infection: fast spin-echo vs conventional spin-echo pulse sequences. AJNR Am J Neuroradiol 162:1199–1204 13. Essig M, Schlemmer HP, Tronnier V et al (2001) Fluidattenuated inversion recovery MR imaging of gliomatosis cerebri. Eur Radiol 11:303–308 14. Tsuchiya K, Mizutani Y, Hachiya J (1996) Preliminary evaluation of fluid-attenuated inversion-recovery MR in the diagnosis of intracranial tumors. AJNR Am J Neuroradiol 17:1081–1086 15. Essig M, Knopp MV, Schoenberg SO et al (1999) Cerebral gliomas and metastases: assesment with contrast-enhanced fast fluid-attenuated inversion-recovery-imaging. Radiology 210:551–557 16. Westbrook C, Kaut C (1993) Image weighting and contrast. In: Westbrook C (ed) MRI in practise. Blackwell Scientific Publications, pp 17–46 17. Fellner F, Fellner C, Held P et al (1997) Comparison of spinecho MR pulse sequences for imaging of the brain. AJNR Am J Neuroradiol 18:1617–1625 18. Wong JC, Provenzale JM, Petrella JR (2000) Perfusion MR imaging of brain neoplasms. AJR Am J Roentgenol 174:1147–1157 19. Edelman RR, Wiclopolski P, Schmitt F (1994) Echo-planar MR. Radiology 192:600–612
2 Imaging Modalities in Brain Tumors 20. Patel MR, Siewert B, Klufas R et al (1999) Echo planar MR imaging for Ultrafast detection of brain lesions. AJR Am J Roentgenol 173:479–485 21. Sievert B, Patel MR, Mueller MF et al (1995) Brain lesions in patients with multiple sclerosis: detection with echo-planar imaging. Radiology 196:765–777 22. Baird AE, Warach S (1998) Magnetic resonance imaging of acute stroke. J Cereb Blood Flow Metab 18:583–609 23. Nelson SJ, Nat D (1999) Imaging of brain tumors. Neuroimaging Clin N Am 9(4):801–819 24. Okamoto K, Ito J, Ishikawa K et al (2000) Diffusion-weighted echo-planar imaging in the differential diagnosis of brain tumors and tumor-like conditions. Eur Radiol 10(8):1342–1350 25. Sugahara T, Korogi Y, Kochi M et al (1999) Usefulness of diffusion-weighed MRI with echo planar technique in the evaluation of cellularity in gliomas. J Magn Reson Imaging 9(1):53–60 26. Filippi CG, Edgar MA, Ulu AM et al (2001) Appearance of meningiomas on diffusion-weighted images: correlating diffusion constants with histopathologic findings. AJNR Am J Neuroradiol 22:65–72 27. Kim YJ, Chang KH, Song IC et al (1998) Brain abscess and necrotic or cystic brain tumor discrimination with signal intensity on diffusion-weighted MR imaging. AJR Am J Roentgenol 171:1487–1490 28. Tsuruda JS, Chew WM, Moseley ME et al (1990) Diffusionweighted MR imaging of the brain: value of differentiating between extraaxial cysts and epidermoid tumors Am J Neuroradiol 11:925–931 29. Le Bihan D, Breton E, Lallemand D, Grenier P, Cabanis E, Laval-Jeantet M (1986) MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders. Radiology 161:401–407 30. Chenevert TL, Brunberg JA, Pipe JG (1990) Anisotropic diffusion in human white matter: demonstration with MR techniques in vivo. Radiology 177:401–405 31. Basser PJ, Pierpaoli C (1996) Microstructural and physiological features of tissues elucidated by quantitative- diffusion-tensor MRI. J Magn Reson B 111(3):209–219 32. Maier SE, Mamata H (2008) Diffusion Imaging of Brain Tumors. In: Newton EB, Jolesz FA (eds) Handbook of neurooncology neuroimaging. Academic Press, Elsevier, pp 239–247 33. Celso Hygino Cruz L Jr, Domingues RC, Sorensen AG (2008) Diffusion Magnetic Resonance Imaging in Brain Tumors. In: Newton EB, Jolesz FA (eds) Handbook of neurooncology neuroimaging. Academic Press, Elsevier, pp 215–238 34. Basser PJ, Pajevic S, Pierpaoli C, Duda J, Aldroubi A (2000) In vivo fiber tractography using DT-MRI data. Magn Reson Med 44(4):625–632 35. Catani M, Howard RJ, Pajevic S, Jones DK (2002) Virtual in vivo interactive dissection of white matter fasciculi in the human brain. Neuroimage 17(1):77–94 36. Mori S, Crain BJ, Chacko VP, van Zijl PCM (1999) Three dimensional tracking of axonal projections in the brain by magnetic resonance imaging. Ann Neurol 45:265–269 37. Baird AE, Benfield A, Schlaug G et al (1997) Enlargement of human cerebral ischemic lesion volumes measured by diffusion-weighted magnetic resonance imaging. Ann Neurol 41:581–589 38. Demaerel PH (ed) (2000) Recent advances in diagnostic neuroradiology. Springer Verlag, Berlin, pp 119–135 39. Aromen JH, Gazit IE, Louis DN et al (1994) Cerebral blood volume maps of gliomas: comparison with tumor grade and histologic findings. Radiology 191:41–51
33 40. Roberts HC, Roberts TPL, Brasch RC et al (2000) Quantitavive measurement of microvascular permeability in human brain tumors achieved using dynamic contrastenhanced MR imaging: correlation with histologic grade. AJNR Am J Neuroradiol 21:891–899 41. Sugahara T, Korogi Y, Tomiguchi S et al (2000) Posttherapeutic intraaxial brain tumor: the value of perfusion sensitive contrast-enhanced MR imaging for differentiating tumor recurrence from nonneoplastic contrast-enhancing tissue. AJNR Am J Neuroradiol 21:901–909 42. Miszkiel KA, Waldan AD (2000) Imaging in AIDS. In: Demaerel PH (ed) Recent advances in neuroradiology. Springer-Verlag, pp 249–273 43. Ernst TM, Chang L, Witt MD et al (1998) Cerebral toxoplasmosis and lymphoma in AIDS: perfusion MR imaging experience in 13 patients. Radiology 208:663–669 44. Sugahara T, Korogi Y, Shigematsu Y et al (1999) Perfusion sensitive MRI of cerebral lymphomas:a preliminary report. J Comput Assist Tomogr 23(2):232–237 45. Sunaert S, Dymarkowski S, Van Oostende S et al (1998) Functional magnetic resonance imaging (fMRI) visualizes the brain at work. Acta Neurol Belg 98:8–16 46. Kwong KK, Belliveau JW, Chesler DA et al (1992) Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation. Proc Natl Acad Sci 29:5675–5679 47. Mueller WM, Yetkin FZ, Hammeke TA et al (1996) Functional MRI mapping of the motor cortex in patients with cerebral tumors. Neurosurgery 39:515–520 48. Schreiber A, Hubbe U, Ziyeh S et al (2000) The influence of gliomas and non-glial space-occupying lesions on bloodoxygen-level-dependent contrast enhancement. AJNR Am J Neuroradiol 21:1055–1063 49. Shuber M, Maldjian JA, Liu WC et al (1998) Functional image-guided surgery of intracranial tumors located in or near the sensorimotor cortex. J Neurosurg 89:412–448 50. Wilms G, Sunaert S, Flamen P (2000) Recent developments in brain tumor diagnosis. In: Demaerel PH (ed) Recent advances in diagnostic neuroradiology. Springer Verlag, Berlin, pp 119–135 51. Weybright P, Sundgren PC, Maly P, Hassan DG, Nan B, Rohrer S, Junck L (2005) Differentiation between brain tumor recurrence and radiation injury using MR spectroscopy. AJR Am J Roentgenol 185(6):1471–1476 52. Lichy MP, Bachert P, Henze M, Lichy CM, Debus J, Schlemmer HP (2004) Monitoring individual response to brain-tumour chemotherapy: proton MR spectroscopy in a patient with recurrent glioma after stereotactic radiotherapy. Neuroradiology 46(2):126–129 53. Howe FA, Barton SJ, Cudlip SA, Stubbs M, Saunders DE, Murphy M, Wilkins P, Opstad KS, Doyle VL, McLean MA, Bell BA, Griffiths JR (2003) Metabolic profiles of human brain tumors using quantitative in vivo 1H magnetic resonance spectroscopy. Magn Reson Med 49:223–232 54. Moller-Hartmann W, Herminghaus S, Krings T, Marquardt G, Lanfermann H, Pilatus U, Zanella FE (2002) Clinical application of proton magnetic resonance spectroscopy in the diagnosis of intracranial mass lesions. Neuroradiology 44:371–378 55. Wang R, Wedeen VJ (2007) ISMRM abstract. Proc Intl Soc Mag Reson Med 15:3720 56. Woolrich MW, Jbabdi S, Patenaude B, Chappell M, Makni S, Behrens T, Beckmann C, Jenkinson M, Smith SM (2009) Bayesian analysis of neuroimaging data in FSL. Neuroimage 45:S173–S186
3
Molecular Abnormalities in Gliomas Anna C. Goussia, Konstantinos Polyzoidis, Maria Bai, and Athanasios P. Kyritsis
Contents
3.1 Introduction
3.1 Introduction.............................................................. 35
Gliomas constitute the most common type of primary brain tumour. They are derived from glial cells of astrocytic, oligodendroglial and ependymal origin. According to histologic type, the most frequently reported neoplasms are the astrocytic tumours, which account for approximately 40% of the cases reported. Oligodendrogliomas and ependymomas are less common tumours accounting for 3.2% and 3–9% of all primary central nervous system (CNS) neoplasms, respectively. Gliomas may manifest at any age, but preferentially affect adults. They are slightly more common in men than women and more common in white than black people. Gliomas can affect any part of the CNS, but they usually occur more supratentoriarly in adults and infratentorialy in children. Histological classification has been the ground on which clinicians base their therapeutic strategies for gliomas [1]. In recent years, there is a need for a more comprehensive understanding of glioma genesis, progression and, particularly, invasion, in order to develop more effective therapeutic strategies. Technical advances and a genomic approach applied to both clinical and basic research have yielded more information about the genetic and molecular basis of gliomas, which is certain to result in optimized drug development and in the discovery of more effective treatment strategies for the more aggressive neoplasms [2–6].
3.2 Astrocytic Tumours.................................................. 3.2.1 Diffusely Infiltrating Astrocytomas............................ 3.2.2 Pilocytic Astrocytoma................................................ 3.2.3 Subependymal Giant Cell Astrocytoma..................... 3.2.4 Pleomorphic Xanthoastrocytoma...............................
35 36 41 42 42
3.3 Oligodendroglial Tumours....................................... 42 3.4 Ependymal Tumours................................................ 44 3.5 Oligoastrocytic Tumours.......................................... 45 References............................................................................ 45
A.C. Goussia and M. Bai Departments of Pathology, Medical School, University of Ioannina, Ioannina, Greece K. Polyzoidis Department of Neurosurgery, Medical School, Aristotle University of Thessaloniki, Thessaloniki, Greece and Neurosurgical Institute, Medical School, University of Ioannina, Ioannina, Greece A.P. Kyritsis (*) Neurosurgical Institute, Medical School, University of Ioannina, Ioannina, Greece and Neurology, Medical School, University of Ioannina, Ioannina, Greece e-mail:
[email protected]
3.2 Astrocytic Tumours Astrocytic tumours comprise a wide range of neoplasms that differ in their location within the CNS, age and gender distribution, growth potential, extent of
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_3, © Springer-Verlag Berlin Heidelberg 2011
35
36
invasiveness, morphological features, tendency for progression and clinical course. There is increasing evidence that these differences reflect the type and sequence of genetic alterations acquired during the process of transformation. Based on clinicopathologic features, astrocytic tumours can be divided into two major categories: (a) the diffusely infiltrating astrocytomas and (b) a heterogeneous group comprising pilocytic astrocytoma (PA), subependymal giant cell astrocytoma (SEGA) and pleomorphic xanthoastrocytoma (PXA). These latter three entities are mutually distinctive but share several characteristics: occurrence at an early age, little tendency for anaplastic progression and a relatively favourable prognosis compared to the diffusely infiltrating astrocytomas.
A.C. Goussia et al.
Fig. 3.1 Diffuse astrocytoma. Neoplastic astrocytes with minimal nuclear atypia in a loosely tumour matrix
3.2.1 Diffusely Infiltrating Astrocytomas Diffuse astrocytic tumours are infiltrative tumours composed of fibrillary neoplastic astrocytes. They represent a morphologic and biologic continuum ranging from relatively indolent examples (diffuse astrocytoma) to ones more aggressive (anaplastic astrocytoma) or frankly high grade (glioblastoma). These tumours arise at any site in the CNS, especially in the cerebral hemispheres; usually manifest in adults, and have a wide range of histopathological features and biological behaviour. Therefore, the grading of anaplasia and the malignant growth potential are salient aspects of their histologic classification and designation as diffuse astrocytoma (WHO grade II), anaplastic astrocytoma (WHO grade III) and glioblastoma multiforme (WHO grade IV) [1]. In this classification, diffuse astrocytoma represents an infiltrative tumour with mild or moderate increase in cellularity and mild nuclear pleomorphism (Fig. 3.1). Anaplastic astrocytomas, in addition to increased cellularity and nuclear pleomorphism, usually exhibit mitotic activity (Fig. 3.2). The presence of microvascular proliferation and necrosis advances the grade from anaplastic astrocytoma to the most malignant form, the glioblastoma multiforme (Fig. 3.3). The glioblastomas could arise ‘de novo’ (primary glioblastomas), or could arise after progression of an anaplastic astrocytoma (secondary glioblastoma). A correlation is noted between histologic grade and a number of clinical variables including patient age, duration of symptoms, neurologic performance status and length
Fig. 3.2 Anaplastic astrocytoma. High cellularity, nuclear atypia and mitotic activity
of post-operative survival. As a rule, the relatively more indolent, diffuse astrocytoma occurs in younger patients and is associated with more chronic symptoms and longer survival. On the other hand, the anaplastic tumours and particularly glioblastomas, which occur in older patients are rapidly progressive and have poor prognosis. A cardinal feature of the group of diffuse astrocytic tumours is their tendency to undergo spontaneous change to a more malignant variety. Thus, diffuse astrocytomas tend to become anaplastic, and anaplastic tumours tend to become glioblastomas. This histologic behaviour is a consequence of multiple genetic changes that accumulate during stepwise progression [7–10]. During the past years, a great progress has
37
3 Molecular Abnormalities in Gliomas
been made in our understanding of the critical events that accompany astroglial transformation and malignant progression (Table 3.1).
3.2.1.1 Alterations of Genes Involved in Cell Cycle Control Uncontrolled cellular proliferation is the hallmark of malignant neoplasms, and a number of recent reports have shown alterations in cell cycle gene expression in human brain tumours. It is known that the progression of
the cell cycle is controlled by positive and negative regulators. Cyclins and cyclin-dependent kinases (CDKs) comprise the major positive growth regulators of this process [11]. The CDKs phosphorylate key substrates, such as the protein of the retinoblastoma gene (Rb), facilitates the passage of the cell through the G1 phase of the cell-cycle [6]. The activity of CDKs can be modulated by a family of inhibitory cell cycle regulators (cyclin-dependent kinase inhibitors or CDKIs). Loss or inactivation of such inhibitors can result in uncontrolled growth; therefore CDKIs are candidate tumour suppressor genes. The CDKIs are classified into two major categories: those of the INK4 family (p16INK4A, p15INK4B, p18INK4C, p19INK4D) and the CIP/KIP family which includes p21CIP1, p27KIP1 and p57KIP2.
The p16/p15/CDK4/CDK6/Rb Pathway The p16INK4A Gene
Fig. 3.3 Glioblastoma multiforme. High degree of anaplasia, microvascular proliferation and necrosis
The p16INK4A gene (orp16CDKN2A) maps to chromosome 9p21, a locus commonly deleted in a variety of human malignancies. Inactivation of the p16INK4A gene results in uninhibited phosphorylation of the pRb, and subsequently to uncontrolled cell growth. Alterations in expression of p16INK4A protein have been described to be caused either by a decrease in mRNA or protein stability or by a decreased transcription of the p16INK4A gene due to methylation of CpG islands or rarely by mutation [12–14]. The correlation between frequency of p16INK4A alterations and malignancy suggests a crucial role for p16INK4A in the malignant progression of astrocytomas.
Table 3.1 Genetic abnormalities in astrocytic tumours Tumour type
Chromosome deviation
Gene alteration
Growth factors amplification/overexpression
Astrocytoma
+7q, −22q
−p53
PDGFR-a
Anaplastic astrocytoma
1, +7q, −9p, −10, −13q, +19, +20, −22q
−p53 −p16 −Rb
Primary glioblastoma
1, −4q, −6q,+7q, 8q, −9p, −10, −13q, −17p, 19, +20, −22q
+MDM2 −PTEN −p16 −Rb
EGFR
Secondary glioblastoma
−4q, −9p, −10, −13q, −17p
−p53 −p16 −Rb
PDGFR-a
+ chromosomic gains or gene amplification/overexpression; − chromosomic losses or gene inactivation; PDGFR-a platelet derived growth factor receptor alpha; EGFR epidermal growth derived factor receptor
38
A.C. Goussia et al.
Data collected by polymerase chain reaction (PCR), sequencing, or interface fluorescence in situ hybridization (FISH) at several laboratories showed homozygous p16INK4A gene deletions in 50–70% of high grade astrocytomas and rarely in low-grade tumours [15–17]. Deletion of the gene was not frequently observed in secondary glioblastomas [10]. Moreover, overexpression of the p16INK4A protein has been found to reduce glioma cells’ ability to invade normal cell [2]. Findings regarding the predictive value of p16INK4A deletions have been contradictory. In some studies, deletion of the gene was associated with unfavourable survival of glioblastoma patients while other studies failed to show any predictive value [16, 18].
nuclear protein that regulates cell-cycle progression through the G1 checkpoint. The CDK4/cyclin D1 complex phosphorylates pRb, thereby inducing release of the E2F transcription factor that activates genes involved in the G1 to S transition. Rb alterations are late genetic events during astrocytoma progression and are mostly associated with high-grade tumours, including anaplastic astrocytomas and glioblastomas [16, 21]. Alterations of the gene are more frequently in primary than secondary glioblastomas [21]. Clinical follow-up data showed a short post-operative survival for glioblastoma patients with a dysfunctional Rb pathway [22].
The p15INK4B Gene
The p53 Gene
The p15INK4B gene (or p15CDKN2B) is also located at 9q21 and is structurally homologous to the p16INK4A gene. Deletion of the p16INK4A gene occurs frequently with deletions of the p15INK4B gene in a subset of glioblastomas [19], suggesting that both genes may be targets of 9p21 deletions. It has been reported that alternative splicing may be a mechanism of p15INK4B inactivation complementing loss of p16INK4A expression during the development of these tumours [19].
The p53 gene, located on chromosome 17p13.1, is a tumour suppressor gene that plays a role in genomic stability, cell-cycle control, DNA repair after damage and apoptosis. Mutations, mainly missense, of the p53 gene are frequent genetic aberrations in astrocytic tumours and represent an early genetic event [10, 23, 24]. p53 mutations are observed in approximately one-third of all three grades of adult astrocytomas, suggesting that inactivation of p53 is important in the formation of the grade II tumours. Studies indicate that p53 gene plays a role in the progression of a low-grade disease towards secondary glioblastomas. p53 mutations or p53 protein overexpression are common in secondary glioblastomas compared with primary glioblastomas. There is evidence that the type and distribution of p53 mutations differ between glioblastoma subtypes. In secondary glioblastomas, 57% of mutations are located in the codons 248 and 273, while in primary glioblastomas mutations are more evenly distributed. G:C→A:T mutations at CpG sites are more frequent in secondary than primary glioblastomas, suggesting that the acquisition of p53 mutations at these glioblastoma subtypes may occur through different mechanisms [10]. Germline p53 mutations consist of a predisposing factor among some familiar cases of gliomas [25]. A mutation rate of 43–67% has been found in patients with multifocal glioma, additional primary tumour, or family history of cancer [25]. This rate was dependent on the presence of two or three factors; however, no germline mutations were seen in patients with unifocal gliomas or in patients without another primary tumour of family history of cancer.
The CDK4 and CDK6 Genes The CDK4 and CDK6 genes are located at 12q13–14 and 7q21–22 chromosomes, respectively, and encode proteins that are inhibited by p16INK4A and p15INK4B. The CDK4 gene is amplified in approximately 15% of high grade gliomas, especially in those without p16 INK4A and p15INK4B alterations [15]. A subset of gliomas exhibits CDK6 amplification without CDK4 amplification. Alterations of the genes are common in both primary and secondary glioblastomas. Recently, it has been suggested that CDK4 provides a growth advantage for astrocytes in vivo in concert with genetic alterations in the p53 pathway [20]. The Rb Gene The retinoblastoma (Rb) gene is a tumour suppressor gene that is inactivated in a number of cancers. A variety of different genetic lesions, including deletions or splicing mutations resulting in loss of an exon, can lead to this inactivation. The Rb protein (pRb) is a 110-kDa
The p53/MDM2/p21/p14ARF Pathway
39
3 Molecular Abnormalities in Gliomas
Recently, it has been found that the majority of lowgrade astrocytomas (92%) with MGMT methylation contained a p53 mutation, whereas only 39% of cases without MGMT methylation exhibited a p53 mutation [26]. Moreover, G:C→A:T transition mutations at CpG sites were significantly more common in low-grade astrocytomas with MGMT methylation than in those without [26]. MGMT (O6-methylguanine-DNA methyltranferase) is a repair protein that specifically removes promutagenic alkyl groups from the O6position of guanine in DNA. Therefore, MGMT protects cells against carcinogenesis induced by alkylating agents. Loss of MGMT expression may be caused by methylation of promoter CpG islands and has been observed in human malignancies, including gliomas. MGMT promoter methylation was detected in 75% of secondary glioblastomas and only in 36% of primary glioblastomas [26]. The above observations suggest that loss of MGMT expression due to promoter methylation occurs commonly at an early stage in the pathway leading to secondary glioblastoma, and appears to be related with increased frequency of p53 mutations in G:C→A:T transitions. The MDM2 Gene The MDM2 gene (called also HDM2) is located in the 12q14.3-q15 chromosomal region. Amplification and overexpression of MDM2 is an important mechanism of p53 inactivation in the absence of p53 mutations. The MDM2 protein binds to both wild-type and mutant p53, thereby inhibiting p53-mediated transactivation. In the largest series of tumours tested, MDM2 was amplified in about 10% of primary glioblastomas [27]. Overexpression of MDM2 was observed immunohistochemically in more than 50% of primary glioblastomas, but the fraction of immunoreactive cells varied considerably. To the contrary, less than 10% of secondary glioblastomas showed MDM2 overexpression. Thus, overexpression of MDM2, regardless of gene amplification, is a genetic hallmark of primary glioblastomas. The p21CIP1 Gene The p21CIP1 gene is located on 6p chromosome and in most cases is under the transcriptional control of the p53 gene. There is also a p53-independent activation of p21CIP1, and this pathway appears to be inducible by a variety of growth factors and differentiating agents. Overexpression of p21CIP1 protein causes growth arrest
suggesting its role in cell cycle progression. It has been reported increased p21 protein expression in most gliomas, regardless of their grade [28]. The simultaneous p53 and p21CIP1 overexpression has been associated with a shorter overall survival of astrocytoma patients [28]. The p14ARF Gene The p14ARF gene is involved in tumour suppression in the p53 pathway, but since p14ARF promoter is responsive to the E2F-1 transcription, it has been suggested that p14ARF is a plexus between both p53 and Rb pathways. During astrocytoma progression, hypermethylation of p14ARF has been detected in a considerable fraction of low-grade astrocytomas, independently to p53 mutations, suggesting that the gene may be present in a subset of tumours with intact p53 [29]. Homozygous deletion of p14ARF has been shown to be associated with shorter survival of these patients, especially when it was combined with p53 alterations [29]. It seems that there is no difference in the overall frequency of p14ARF alterations between primary and secondary glioblastomas; however, p14ARF promoter methylation is more common in secondary than primary glioblastomas.
Alterations of the p27KIP1 Gene The p27KIP1 gene regulates progression through the cell cycle by binding to and inactivating specific cyclinCDK complexes. Post-transcriptional control through the ubiquitin-proteasome pathway is thought to be the main process involved in the decreased expression of p27KIP1 protein observed in several human tumours. In gliomas, conflicting data regarding the p27KIP1 expression status and its prognostic role have been reported. Low expression of p27KIP1 protein was found to be related with poor prognosis of patients with high-grade astrocytomas [30]; however, this observation has not been confirmed by some studies.
3.2.1.2 Signal Transduction Growth Factors and Receptors Growth factors bind and activate receptors with a protein kinase activity, usually receptor tyrosine kinases, or receptors that transmit signals by interacting with
40
TP-binding proteins. The first type of receptors G dimerize upon binding growth factors, such as epidermal growth factor (EGF), platelet-derived growth factor (PDGF) or fibroblast growth factor (FGF). Signalling through EGF, PDGF and FGF seems to play an important role in the regulation of gliogenesis during normal brain development. Moreover, alterations of these growth factors and receptors have been reported in a number of gliomas. The epidermal growth factor receptor (EGFR) is a transmembrane receptor encoded by the EGFR cellular oncogene located on human chromosome 7. Amplification of EGFR occurs in approximately 40–60% of astrocytic tumours and is mainly associated with glioblastomas [31–33]. EGFR amplification constitutes a hallmark of primary glioblastomas, since more than 60% of these tumours show EGFR overexpression [31]. In contrast, secondary glioblastomas only rarely exhibit EGFR overexpression (<10%). In about half of the glioblastoma cases with receptor amplification, the event is coupled with gene rearrangement. The most common rearrangement results, is a variant form called EGFRvIII or delta EGFR (mutant form), that is expressed in 24–67% of glioblastoma cases [33]. The delta EGFR variant promotes activation of various intracellular signalling pathways, including the phosphatidylinositol 3-kinase (PI3 kinase)/Akt pathway [34]. In vivo experiments demonstrated that EGFRvIII can cause an increase in cell proliferation and cell motility. Moreover, the combination of EGFR activation and loss of both p16INK4A and p19INK4D was found to dedifferentiate astrocytes and to induce high-grade gliomas. Clinical studies are inconclusive concerning the prognostic role of EGFR amplification and /or overexpression [32, 33]. The prognostic effect of EGFRvIII has not been extensively studied, but in some reports its presence was found to be an unfavourable prognosticator of survival [32]. Expression of the platelet-derived growth factor-a (PDGF-a) receptor has been observed in all grades of astrocytomas, suggesting that this overexpression is important from the initial stages of astrocytoma formation [35]. Loss of chromosome 17p in the region of p53 gene is closely related with PDGF-a receptor, suggesting that p53 mutations may have an oncogenic effect mainly in the presence of PDGF-a receptor overexpression. The FGF is abnormally expressed in gliomas [36]. In vitro, the FGF plays an important role in promoting the malignant transformation of cultured astrocytes and in vivo administration of FGF significantly
A.C. Goussia et al.
increases the number of experimentally induced brain tumours. A strong association between malignancy in human astrocytic tumours and increased expression of FGF has been reported.
Transforming Growth Factor-b (TGF-b) Signalling The TGF-b contributes to tumour progression by regulating key mechanisms including proliferation, angiogenesis and metastasis. The traditional Smad pathway and the recently discovered MARK pathway are the most important pathways for TGF-b-related intracellular signal production, mediating differential pathobiological effects [37]. The TGF-b/Smad pathway was found to promote proliferation of gliomas through the induction of PDGF-B [38]. High TGF-b/Smad activity has been reported to be present in aggressive, highly proliferative gliomas and confers poor patients’ prognosis [38]. The administration of TGF-b antisense constructs to glioma cells had a growth inhibitory effect, indicating that TGF-b might contribute to glioma progression. Recent data suggest that the combined antagonization of the TGF-b and MARK pathways may be a promising approach for glioma therapy [37].
Phosphatidylinositol 3-Kinase Pathway/PTEN The phosphatidylinositol 3-kinase (PI3K) pathway is relevant to proliferation and cell survival of malignant gliomas. Activation of this pathway protects glioma and other malignant cells against apoptosis by phosphorylating and functionally inactivating several pro-apoptotic proteins as well as by inhibiting TRAIL-induced apoptosis [39]. In gliomas, the PI3K pathway is activated through several mechanisms, such as amplification/ overexpression of growth factors and mutation or LOH of PTEN. Loss of all or part of chromosome 10q where PTEN and other suppressor genes, such as MXI1 (Max interacting protein 1) and DMBT1 (deleted in malignant brain tumours 1), occurs in approximately 70–90% of glioblastomas [40–44]. Although overexpression of wild-type of PTEN in glioma cells can cause growth suppression, loss of PTEN does not seem to be sufficient to induce neoplastic transformation of glial cells. Therefore, loss of PTEN seems to be a late event and plays a role in the conversion of low-grade astrocytomas to glioblastomas.
41
3 Molecular Abnormalities in Gliomas
PTEN mutations have been detected commonly in primary glioblastomas and rarely in secondary glioblastomas [41]. PTEN homozygous deletion may occur in glioblastomas but it is rare. Promoter methylation may be an alternative mechanism of loss of PTEN expression, but the significance of PTEN methylation in the evolution of glioblastomas remains to be determined. In several studies, LOH at 10q has been associated with poorer prognosis of glioblastoma patients; however, PTEN mutations were not clearly associated with patients’ prognosis [22]. A recent study showed that loss of wild-type PTEN, in association with abnormalities in the Rb pathway, was related with a shorter overall survival in glioblastoma patients than were mutations of the Rb pathway alone [22].
3.2.1.3 Factors implicated in Angiogenesis and Invasion The most important factor implicated in glioma angiogenesis is vascular endothelial growth factor (VEGF), a secreted ligand with highly specific mitogenic and chemotactic activity on endothelial cells [9]. Other commonly involved factors include: FGF, PDGF and TGF-b [9]. As VEGF is hypoxia induced, a major trigger of angiogenesis in gliomas appears to be cellular hypoxia. Due to its dual function, VEGF may be responsible for both angiogenesis (microvascular proliferation) and vascular permeability (peritumoural oedema) in malignant gliomas. In situ hybridization studies have demonstrated expression of VEGF mRNA at high levels in the hypervascularized glioblastomas, especially by those cells surrounding the perinecrotic zones [45]. Intermediate levels of VEGF mRNA production have been observed in low-grade tumours and low levels in normal brain [45]. Genomics studies coupled with microarray experiments showed that insulin-like growth factor binding protein 2 (IGFBP2) is a novel gene overexpressed in high-grade gliomas. There are six\members in the IGFBP family and they have very different functions, especially those that are IGF independent. Among these, IGFBP2 is overexpressed in glioblastomas. Moreover, it has been shown that IGFBP2 and IGFBP5 are closely clustered, suggesting that both proteins may contribute to glioma invasion [5]. In vitro studies have shown that IGFBP2 promotes glioma cell migration and invasion by forming a complex with integrin a5 protein and activating expression of the matrix
metalloproteinase 2 gene [46]. Moreover, it has been reported that IGFBP2 plays a key role in activation of the Akt pathway and collaborates with KRas or PDGFR in the development and progression of gliomas [47].
3.2.2 Pilocytic Astrocytoma PA is the most common glioma in children and comprises about 20% of all childhood brain tumours. It is a low-grade tumour that corresponds to WHO grade I [1]. The classic locations of the neoplasm are cerebellum, hypothalamic/third ventricular region, optic nerves/chiasm, dorsal brain stem and spinal cord. By far the most common site is the cerebellum (85%). This site also carries the most favourable prognosis, due in large part to the potential for complete surgical resection. Histologically, PA exhibits a distinctive biphasic appearance, with varying proportions of compacted bipolar cells with Rosental fibres and loose/microcystic areas (Fig. 3.4). Microvascular proliferation is often a prominent component of the tumour and constitutes the presumed morphologic basis for the contrast enhancement seen in neuroimaging studies. Notably, mitotic activity or necrosis is inconspicuous and very rarely the tumour may undergo anaplastic progression. Reduced expression of neurofibromin (protein product of the neurofibromatosis type 1/NF1 gene) is seen in NF1-associated but not in sporadic PAs [48]. Therefore, these histologically identical tumours appear to arise through distinct molecular mechanisms.
Fig. 3.4 Pilocytic astrocytoma. Long, bipolar tumour cells, loose microcystic areas, Rosenthal fibres and granular bodies
42
Some NF1-associated PAs also appear to have germline mutations in the NF1 gene; however, the relative frequency of these mutations is unclear [48]. Studies aimed at identifying genetic and molecular changes in sporadic PAs have reported a variety of abnormalities, including LOH on chromosome 17, gains on chromosomes 7 and 8 and case reports of PTEN and KRAS gene mutations [49]. However, none of these changes are consistently found in PAs, suggesting that other molecular changes are involved in tumour pathogenesis. Gene expression data suggest the presence of a unique gene expression signature in PAs arising in patients with NF1 [50]. The newly revised WHO classification of Tumours of the Central Nervous System has introduced pilomyxoid astrocytoma (PMA) as a variant of PA [1]. PMA is typically a tumour of early childhood and has been designated WHO grade II. The hypothalamic/ chiasmal region is the most characteristic location. The histologic appearance is dominated by monomorphous bipolar cells, which lie within a rich myxoid matrix and often show an angiocentric arrangement. Molecular studies have shown no evidence of p53 gene alterations and one recent case study reported a PMA in a NF1 patient [51]. An insertion on chromosome 17 that involved disruption of the BCR gene has been detected [52].
3.2.3 Subependymal Giant Cell Astrocytoma SEGA is a benign, slowly growing intraventricular tumour, encountered nearly exclusively in the setting of tuberous sclerosis. SEGA corresponds to WHO grade I and histologically is composed of spindled and epithelioid cells arranged in sweeping fascicles or perivascular pseudorosettes. The cells often have a mixed glioneuronal appearance. Cellular pleomorphism and occasional mitoses do not indicate malignant behaviour. No consistent molecular alterations are known.
3.2.4 Pleomorphic Xanthoastrocytoma The PXA is an astrocytic neoplasm of late childhood and adolescence that accounts for less than 1% of all
A.C. Goussia et al.
astrocytic tumours. It occurs, most frequently, superficially in the cerebral hemispheres, particularly in the temporal lobe and has a relatively favourable prognosis. Most PXAs are classified as grade II in the WHO classification [1], although by conventional criteria they exhibit some ‘malignant’ histological features, such as cellular and nuclear pleomorphism. If mitotic activity becomes prominent they are regarded as having progressed to anaplastic tumours or glioblastomas. There are limited data concerning molecular abnormalities of PXAs. P53 mutations and amplification of the EGFR gene have been observed occasionally. Loss of chromosome 9 seems to be the most common imbalance in these tumours [53].
3.3 Oligodendroglial Tumours Oligodendroglial tumours are rare primary brain tumours accounting for 3.2% of all primary CNS neoplasms. They have a peak incidence in the fourth and fifth decades of life; they are comparatively rare in children, constituting 1–2% of all paediatric CNS tumours. Most cases are located in the cerebral hemispheres and typically originate in the white matter with secondary invasion of the cortex. They occur infrequently in the cerebellum, brain stem and spinal cord. The WHO classification recognizes two grades of oligodendroglial tumours: WHO grade II and WHO grade III for oligodendrogliomas and anaplastic oligodendrogliomas, respectively [1]. Histologically, oligodendrogliomas are moderately cellular neoplasms and are composed of tumour cells with round nuclei, bland chromatin and a clear perinuclear halo (‘fried-egg’ appearance). Additional features include microcalcifications, mucin-rich microcystic spaces and a rich branching capillary network reminiscent of ‘chicken wire’ (Fig. 3.5). Significant mitotic activity, microvascular proliferation or necrosis indicates progression to anaplastic oligodendrogliomas (Fig 3.6). Although median survival intervals vary among patient groups, generally, tumours that possess many or all of the histological signs of anaplasia are considered high-grade tumours and are likely to behave aggressively. Significant attention has recently been focused on the molecular genetic approaches for oligodendroglial tumours [54–70]. Losses of chromosomes 1p and 19q
43
3 Molecular Abnormalities in Gliomas
Fig. 3.5 Oligodendroglioma showing the typical chicken wirelike branching capillary pattern and tumour cells with clear cytoplasm (‘fried-egg’ appearance)
Fig. 3.6 Anaplastic oligodendroglioma is characterized by nuclear atypia and marked microvascular proliferation
are the most common abnormalities in oligodendroglial tumours [55–57, 61, 62, 64–69] (Table 3.2). Allelic loss at 1p is seen in 70–85% of oligodendrogliomas and loss at 19 q in 50% to more than 80% of the cases. Combined loss on chromosomes 1p and 19q was found in up to 80% of oligodendrogliomas. Importantly, this combined genetic alteration has emerged as an independent predictive marker of better response to radio- and chemotherapy as well as longer survival in patients with anaplastic oligodendroglial tumours [55, 56, 61, 63]. Based on these findings, ancillary testing for 1p and 19q status has become routine in some medical centres. The most commonly used techniques include FISH and LOH. The frequent LOH on 1p and 19 q indicates that these chromosomal arms carry tumour suppressor genes whose inactivation is of importance for oligodendroglioma development. Initial studies suggested several candidate genes, such the cyclin-dependent kinase inhibitor 2C (CDKN2C) gene at 1p32, that was found to be mutated or deleted in a small subset of anaplastic oligodendrogliomas [56]. A recent study pointed to the calmodulin-binding transcription activator 1 gene (CAMTA1) gene at 1p36, which showed reduced expression but no mutations in oligodendrogliomas with 1p deletion [66]. On 19q13.3, the product of the p190RhoGAP gene, which is a regulator of Rho kinases, has been reported to inhibit PDGFinduced murine oligodendrogliomas [64]. A candidate gene on 19q13.3, the myelin-related epithelial membrane protein gene 3 (EMP3), has been shown to demonstrate promoter hypermethylation and transcriptional downregulation in a subset of oligodendrogliomas [67]. By using cDNA microarrays, a distinct set of
Table 3.2 Genetic abnormalities in oligodendroglial and ependymal tumours Tumour type
Chromosome deviation
Gene alteration
Oligodendroglioma
−1p, −19q
Anaplastic oligodendroglioma
−1p, −19q, −9p, −10
−p16
Ependymoma
−22q, −6q, +7, −9p, −11, −13, −17p
−NF2
Anaplastic ependymoma
−22q, +7, −10, −9p, −17p
Oligoastrocytoma
−1p, −19q
Anaplastic oligoastrocytoma
−9p, −10q
Growth factors amplification/overexpression EGFR
−p16
EGFR
EGFR
+ chromosomic gains; − chromosomic losses or gene inactivation; EGFR epidermal growth derived factor receptor; NF2 neurofibromatosis type 2
44
novel candidate genes located on 1p or 19q were identified [68]. Among them, a total of 8 genes, located within 1p36.13-p36.31 (MGC4399, SRM, ICMT) or 19q13.2-q13.33 (RPL18, FTL, ZIN, FLJ10781, DBP), demonstrated lower expression in gliomas with LOH 1p/19q when compared to gliomas with retention of heterozygosity on both chromosomal arms. Moreover, a set of 35 genes discriminating between WHO grade II and WHO grade III oligodendrogliomas has been reported [68]. The most interesting gene was EP300, a histone acetyltransferase and tumour suppressor gene which is involved in the regulation of several tumourigenic pathways, such as TGF-b, retinoblastoma protein (pRb) and p53 pathways. Molecular studies in malignant glioma cell lines, xenograft tumour mouse model systems and human tumour specimens showed LOH for the stathmin, a microtubule-associated protein, which seems to be associated with improved outcome of patients with anaplastic oligodendrogliomas and LOH 1p [69]. Similarly to astrocytomas, p16INK4A deletions are common progression-associated alterations in oligodendroglial tumours [60]. The p16INK4A gene is homozygously deleted in about 25% of anaplastic oligodendrogliomas, including tumours with or without allelic loss on 1p and/ or 19q. It has been reported, that patients with anaplastic tumours and deletion of the p16INK4A gene had a significantly worse prognosis than patients without the deletion, with median survival times of less than 2 years. In addition, an inverse relationship between p16INK4A gene deletion and losses of chromosomes 1p and 19q has been observed, suggesting that there are two independent genetic subtypes of anaplastic oligodendroglioma with differential clinical behaviours. PTEN alterations were found in a subset of anaplastic oligodendrogliomas and their presence has been associated with unfavourable prognosis [59]. MGMT promoter methylation is considered as a late event in progressive oligodendrogliomas and was associated with either loss of 1p or combined loss of 1p and 19q chromosomes [60, 70]. Both low-grade and anaplastic oligodendrogliomas show expression of EGFR mRNA and protein in the absence of EGFR gene amplification and a small subset of anaplastic tumours exhibit amplification of the CDK4 gene [62]. Other growth factors, such as FGF, TGF-b, PDGF, VEGF, and insulin-like growth factor I (IGF-I) have been involved in the regulation of proliferation of oligodendroglial cells, but their significance in oligodendroglial tumourigenesis is unclear.
A.C. Goussia et al.
3.4 Ependymal Tumours This group of neoplasms originate from the ependymal lining of the ventricular system and from the remnants of the central canal of the spinal cord. Ependymal tumours occur predominantly in children and adolescents. In children, they are the third most frequent brain neoplasms following low-grade astrocytomas and medulloblastomas. They can occur at any site in the ventricular system, but they are commonly found in the posterior fossa, lateral ventricles and spinal cord. The following entities of ependymal tumours can be distinguished: ependymoma, anaplastic ependymoma, myxopapillary ependymoma and subependymoma. Histologically, ependymomas are designated as WHO grade II or III depending on the presence of anaplastic features (increased cellularity, nuclear pleomorphism, mitotic activity, vascular proliferation and/or necrosis) [1] (Fig.3.7). However, myxopapillary ependymomas and subependymomas, which are slow-growing tumours with a favourable prognosis, are considered as WHO grade I neoplasms. There is no general agreement regarding the prognostic value of the histological grade in patients with ependymomas and anaplastic ependymomas, and in more studies, histological parameters alone appear to be of limited significance for the prognosis of patients. Up to now, the only potential factors associated with prognosis have been the tumour location (spinal ependymomas are associated with a more favourable prognosis than intracranial ependymomas of the respective WHO grade), and the ability to achieve complete
Fig. 3.7 Ependymoma WHO grade II. Monomorphic nuclear morphology and perivascular pseudorosettes
45
3 Molecular Abnormalities in Gliomas
tumour resection, although the chances of recurrence continue to be high in a large proportion of patients. Therefore, it would be of interest to determine if any of the genetic changes observed in these tumours can predict their biological behaviour. In contrast to astrocytic and oligodendroglial tumours, in which molecular alterations associated with tumourigenesis are relatively well established, less is known about molecular changes in ependymal tumours. Chromosome arm 22q has been the most frequently described region of genomic loss in ependymomas [71–74] (Table 3.2). Some investigators have presented evidence of mutations of the neurofibromatosis 2 (NF2) tumour suppressor gene at 22q12, whereas others have been unable to identify such mutations. Moreover, a high incidence of LOH on 22q and NF2 gene mutations have been reported more often in ependymomas WHO grade II with a spinal than intracerebral localization [73, 74]. These findings suggest that the NF2 gene is the target of loss of chromosome 22 in a subset of ependymomas. In addition, the more favourable clinical course of spinal ependymomas may relate to a distinct pattern of genetic alterations different, from that of intracerebral ependymomas. Additional tumour suppressor genes are also suspected on 22q. Other abnormality in ependymomas involves monosomy or deletion of chromosomes 6. A study analyzing samples of paediatric ependymomas using comparative genomic hybridization found a frequent deletion of chromosome arm 6q, suggesting the presence of a tumour suppressor gene that may contribute to the development of ependymomas [75]. Gain of chromosomes 7 and 1q as well losses of chromosomes 6q, 9 and 13 are more frequently observed in intracranial tumours [76, 77]. The involvement of chromosome 9 in ependymal neoplasma led to the study of the 9p21 located tumour suppressor genes [78, 79]. Deletions of p16INK4A were found in 25% of the cases [78] and promoter methylation of p16INK4A, p15 INK4B and p14ARF genes was detected in 20–30% of the tumours [79].
3.5 Oligoastrocytic Tumours Oligoastrocytic tumours are composed of a conspicuous mixture of two distinct neoplastic cell types, morphologically resembling the tumour cells of oligodendroglioma
and astrocytoma. They account for less than 1% of brain tumours; most of these occur in the cerebral hemispheres. Histologically, oligoastrocytomas may be divided into ‘biphasic’ or ‘compact’ with separate areas resembling oligodendroglioma and astrocytoma and ‘intermingled’ or ‘diffuse’ in which the two elements are intermingled. According to WHO classification, oligoastrocytomas are grade II tumours and in the presence of focal or diffuse histological features of malignancy, the term anaplastic oligoastrocytoma is used, that corresponds to WHO grade III [1]. Molecular studies have shown that most cases of oligoastrocytic tumours are either genetically similar to pure astrocytomas (e.g. p53 mutations, monosomy 10, EGFR gene amplification) or pure astrocytomas (1p/19q deletions); however, some cases have none of these alterations [66]. Combined LOH on chromosomal arms 1p and 19q has been found in 20–50% of oligoastrocytomas [66]. Whether this alteration has the same prognostic implication as it does with pure oligodendrogliomas remains to be elucidated. About 30% of oligoastrocytomas have mutations of the p53 gene and /or LOH on 17p. Interestingly, those tumours with p53 mutations do not have LOH on 1p and 19q, and vice versa [80]. With respect to progressionasssociated molecular alterations, anaplastic oligoastrocytomas have been found to share many abnormalities that also implicated in the progression of oligodendrogliomas and astrocytomas, including losses of chromosomes 1p and 19q, deletion of p16INK4A gene and occasional amplification of the EGFR gene.
References 1. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK (eds) (2007) World Health Organization classification of tumours of the central nervous system. IARC, Lyon 2. Adachi Y, Chandrasekar N, Kin Y, Lakka SS, Mohanam S, Yanamandra N, Mohan PM, Fuller GN, Bingliang F, Fueyo J, Dinh DH, Olivero WC, Tamiya T, Ohmoto T, Kyritsis AP, Rao JS (2002) Suppression of glioma invasion and growth by adenovirus-mediated delivery of a bicistronic construct containing antisense uPAR and sense p16 sequences. Oncogene 21:87–95 3. Levin VA, Hess KR, Choucair A, Flynn PJ, Jaeckle KA, Kyritsis AP, Yung WK, Prados MD, Bruner J, Ictech S, Gleason MJ, Kim HW (2003) Phase III randomized study of postradiotherapy chemotherapy with combination alphadifluoromethylornithine-PCV versus PCV for anaplastic gliomas. Clin Cancer Res 9:981–990
46 4. Ribalta T, Wang H, Fuller GN (2004) Tissue microarrays: applications in neuro-oncology research. In: Zhang W, Fuller GN (eds) Genomic and molecular neuro-oncology. Jones and Bartlett, Boston 5. Jiang R, Mircean C, Shmulevich I, Cogdell D, Jia Y, Tabus I, Aldape K, Sawaya R, Bruner J, Fuller GN, Zhang W (2006) Pathway alterations during glioma progression revealed by reverse phase protein lysate arrays. Proteomics 6:2964–2971 6. Brandes AA, Tosoni A, Cavallo G, Reni M, Franceschi E, Bonaldi L, Bertorelle R, Gardiman M, Ghimenton C, Iuzzolino P, Pession A, Blatt V, Ermani M (2006) Correlations between O6-methylguanine DNA methyltransferase promoter methylation status, 1p and 19q deletions, and response to temozolomide in anaplastic and recurrent oligodendroglioma: a prospective GICNO study. J Clin Oncol 24:4746–4753 7. Goussia AC, Agnantis NJ, Rao JC, Kyritsis AP (2000) Cytogenetic and molecular abnormalities in astrocytic gliomas. Oncol Rep 7:401–412 8. Konopka G, Bonni A (2003) Signalling pathways regulating gliomagenesis. Curr Mol Med 3:73–84 9. Kargiotis O, Rao JS, Kyritsis AP (2006) Mechanisms of angiogenesis in gliomas. J Neuro-Oncol 78:281–293 10. Ohgaki H, Kleihues P (2007) Genetic pathways to primary and secondary glioblastomas. Am J Pathol 170:1445–1453 11. Murray A (2004) Recycling the cell cycle. Cell 116:221–234 12. Fueyo J, Comez-Manzano C, Bruner JM, Saito Y, Zhang B, Zhang W, Levin VA, Yung WK, Kyritsis AP (1996) Hypermethylation of the CpG island of p16/CDKN2 correlates with gene inactivation in gliomas. Oncogene 13:1615–1619 13. Kyritsis AP, Zhang B, Zhang W, Xiao M, Takeshima H, Bondy ML, Cunningham JE, Levin VA, Bruner J (1996) Mutations of the p16 gene in gliomas. Oncogene 12:63–67 14. Nakamura M, Watanabe T, Klangby U, Asker C, Wiman K, Yonekawa Y, Kleihues P, Ohgaki H (2001) p14ARF deletion and methylation in genetic pathways to glioblastomas. Brain Pathol 11:159–168 15. Nishikawa R, Furnari FB, Lin H, Arap W, Berger MS, Cavenee WK, Su Huang HJ (1995) Loss of p16INK4 expression is frequent in high grade gliomas. Cancer Res 55:1941–1945 16. Puduvalli VK, Kyritsis AP, Hess KR, Bondy ML, Fuller GN, Kouraklis GP, Levin VA, Bruner JM (2000) Patterns of expression of Rb and p16 in astrocytic gliomas, and correlation with survival. Int J Oncol 17:963–969 17. Labuhn M, Jones G, Speel EJ, Maier D, Zweifel C, Gratzl O, Van Meir EG, Hegi ME, Merlo A (2001) Quantitative realtime PCR does not show selective targeting of p14(ARF) but concomitant inactivation of both p16(INK4A) and p14(ARF) in 105 human primary gliomas. Oncogene 20:1103–1109 18. Kamiryo T, Tada K, Shiraishi S, Shinojima N, Nakamura H, Kochi M, Kuratsu J, Saya H, Ushio Y (2002) Analysis of homozygous deletion of the p16 gene and correlation with survival in patients with glioblastoma multiforme. J Neurosurg 96:815–822 19. Simon M, Koster G, Ludwig M, Mahlberg R, Rho S, Watzka M, Schramm J (2001) Alternative splicing of the p15 cdk inhibitor in glioblastoma multiform. Acta Neuropathol 102:167–174 20. Z-y H, Baldwin R, Hedrick NM, Gutmann DH (2002) Astrocyte-specific expression of CDK4 is not sufficient for tumor formation, but cooperates with p53 heterozygosity to
A.C. Goussia et al. provide a growth advantage for astrocytes in vivo. Oncogene 21:1325–1334 21. Nakamura M, Yonekawa Y, Kleihues P, Ohgaki H (2001) Promoter hypermethylation of the RB1 gene in glioblastomas. Lab Invest 81:77–82 22. Backlund LM, Nilsson BR, Goike HM, Schmidt EE, Liu L, Ichimura K, Collins VP (2003) Short postoperative survival for glioblastoma patients with a dysfunctional Rb1 pathway in combination with no wild-type PTEN. Clin Cancer Res 9:4151–4158 23. Vital A, Loiseau H, Kantor G, Daucourt V, Chene G, Cohadon F, Rougier A, Rivel J, Vital C (1998) p53 protein expression in grade II astrocytomas: immunohistochemical study of 100 cases with long-term follow-up. Pathol Res Pract 194:831–836 24. Maher EA, Furnari FB, Bachoo RM, Rowitch DH, Louis DN, Cavenne WK, DePinho RA (2001) Malignant glioma: genetics and biology of a grave matter. Gene Dev 15:1311–1333 25. Kyritsis AP, Bondy ML, Xiao M, Bernan EL, Cunningham JE, Lee PS, Levin VA, Saya H (1994) Germline p53 gene mutations in subsets of glioma patients. J Natl Cancer Inst 86:344–349 26. Nakamura M, Watanabe T, Yonekawa Y, Kleihues P, Ohgaki H (2001) Promoter methylation of the DNA repair gene MGMT in astrocytomas is frequently associated with G:C→A:T mutations of the TP53 tumor suppressor gene. Carcinogenesis 22:1715–1719 27. Reifenberger G, Ichimura K, Reifenberger J, Elkahloun AG, Meltzer PS, Collins VP (1996) Refined mapping of 12q13– q15 amplicons in malignant gliomas suggests CDK4/SAS and MDM2 as independent amplification targets. Cancer Res 56:5141–5145 28. Korkolopoulou P, Kouzelis K, Christodoulou P, Papanikolaou A, Thomas-Tsagli E (1998) Expression of retinoblastoma gene product and p21 (WAF1/Cip1) protein in gliomas: correlations with proliferation markers, p53 expression and survival. Acta Neuropathol 95:617–624 29. Watanabe T, Katayama Y, Yoshino A, Komine C, Yokoyama T (2003) Deregulation of the TP53/p14ARF tumor suppressor pathway in low-grade diffuse astrocytomas and its influence on clinical course. Clin Cancer Res 9:4884–4890 30. Kirta RM, Haapasalo HK, Kalimo H, Salminen EK (2003) Low expression of p27 indicates a poor prognosis in patients with high-grade astrocytomas. Cancer 97:644–648 31. Watanabe K, Tachibana O, Sata K, Yonekawa Y, Kleihues P, Ohgaki H (1996) Overexpression of the EGF receptor and p53 mutations are mutually exclusive in the evolution of primary and secondary glioblastomas. Brain Pathol 6:217–223 32. Shinojima N, Tada K, Shiraishi S, Kamiryo T, Kochi M, Nakamura H, Makino K, Saya H, Hirano H, Kuratsu J, Oka K, Ishimaru Y, Ushio Y (2003) Prognostic value of epidermal growth factor receptor in patients with glioblastoma multiforme. Cancer Res 63:6962–6970 33. Heimberger AB, Hlatky R, Suki D, Yang D, Weinberg J, Gilbert M, Sawaya R, Aldape K (2005) Prognostic effect of epidermal growth factor receptor and EGFRvIII in glioblastoma multiforme patients. Clin Cancer Res 11:1462–1466 34. Klingler-Hoffmann M, Bukczynka P, Tiganis T (2003) Inhibition of phosphatidylinositol 3-kinase signaling negates the growth advantage imparted by a mutant epidermal
3 Molecular Abnormalities in Gliomas growth factor receptor on human glioblastoma cells. Int J Cancer 105:331–339 35. Hermanson M, Funa K, Koopmann J, Maintz D, Waha A, Westermark B, Heldin C-H, Wiestler OD, Louis DN, von Deimling A, Nister M (1996) Association of high plateletderived growth factor (PDGF) a receptor expression with loss of heterozygosity (LOH) on chromosome 17p in human malignant gliomas. Cancer Res 56:164–171 36. Gately S, Soff GA, Brem S (1995) The potential role of basic fibroblast growth factor in the transformation of cultured primary human fetal astrocytes and the proliferation of human glioma (U-87) cells. Neurosurgery 37:723–730 37. Nickl-Jockschat T, Arslan F, Doerfelt A, Bogdahn U, Bosserhoff A, Hau P (2007) An imbalance between Smad and MARK pathways is responsible for TGF-beta tumor promoting effects in high-grade gliomas. Int J Oncol 30:499–507 38. Bruna A, Darken RS, Rojo F, Ocana A, Penuelas S, Arias A, paris R, Tortosa A, Mora J, Baselga J, Seoane J (2007) High TGFbeta-Smad activity confers poor prognosis in glioma patients and promotes cell proliferation depending on the methylation of the PDGF-B gene. Cancer Cell 11:147–160 39. Puduvalli VK, Sampath D, Bruner JM, Nagia J, Xu R, Kyritsis AP (2005) TRAIL-induced apoptosis in gliomas is enhanced by Akt-inhibition and is independent of JNK activation. Apoptosis 10:233–243 40. Wechsler DS, Shelly CA, Petroff CA, Dang CV (1997) MXI1, a putative suppressor gene, suppresses growth of human glioblastoma cells. Cancer Res 57:4905–4912 41. Fujisawa H, Reis RM, Nakamura M, Colella S, Yonekawa Y, Kleihues P, Ohgaki H (2000) Loss of heterozygosity on chromosome 10 is more extensive in primary (de novo) than in secondary glioblastomas. Lab Invest 80:65–72 42. Terada K, Tamiya T, Daido S, Mambara H, Tanaka H, Ono Y, Matsumoto K, Ito S, Ouchida M, Ohmoto T, Shimizu K (2002) Prognostic value of loss of heterozygosity around three candidate tumor suppressor genes on chromosome 10 q in astrocytomas. J Neurooncol 58:107–114 43. Pang JC, Dong Z, Zhang R, Liu Y, Zhou L, Chan B, Poon W, Ng H (2003) Mutation analysis of DMBT1 in glioblastoma, medulloblastoma and oligodendroglial tumors. Int J Cancer 105:76–81 44. Hu X, Pandolfi PP, Li Y, Koutcher JA, Rosenblum M, Holland EC (2005) mTOR promotes survival and astrocytic characteristics induced by Pten/AKT signaling in glioblastoma. Neoplasia 7:356–368 45. Plate KH, Breier G, Weich HA, Risau W (1992) Vascular endothelial growth factor is a potential tumor angiogenesis factor in human gliomas. Nature 359:845–848 46. Wang GK, Hu L, Fuller GN, Zhang W (2006) An interaction between insulin-like growth factor-binding protein 2 (IGFBP2) and integrin alpha5 is essential for IGFBP2induced cell mobility. J Biol Chem 281:14085–14091 47. Dunlap S, Celestino J, Wang H, Jiang R, Holland E, Fuller GN, Zhang W (2007) Insulin-like growth factor binding protein 2 promotes glioma development and progression. PNAS 104:11736–11741 48. Gutmann DH, James CD, Poyhonen M, Louis DN, Ferner R, Guha A, Hariharan S, Viskochil D, Perry A (2003) Molecular analysis of astrocytomas presenting after age 10 in individuals with NF1. Neurology 61:1397–1400
47 49. Sanoudou D, Tingby O, Ferguson-Smith MA, Collins VP, Coleman N (2000) Analysis of pilocytic astrocytoma by comparative genomic hybridization. Br J Cancer 82:1218–1222 50. Sharma MK, Mansur DB, Reifenberger G, Perry A, Leonard JR, Aldape KD, Albin MG, Emnett RJ, Loeser S, Watson MA, Nagarajan R, Gutmann DH (2007) Distinct genetic signatures among pilocytic astrocytomas relate to their brain region origin. Cancer Res 67:890–900 51. Khanani MF, Hawkins C, Shroff M, Dirks P, Capra M, Burger PC, Bouffet E (2006) Pilomyxoid astrocytoma in a patient with neurofibromatosis. Pediatr Blood Cancer 46:377–380 52. Melendez B, Fiano C, Ruano Y, Hernandez-Moneo JL, Mollejo M, Martinez P (2006) BCR gene disruption in a pilomyxoid astrocytoma. Neuropathology 26:442–446 53. Weber RG, Hoischen A, Ehrler M, Zipper P, Kaulich K, Blaschke B, Becker AJ, Weber-Mangal S, Jauch A, Radlwimmer B, Schramm J, Wiestler OD, Lichter P, Reifenberger G (2007) Frequent loss of chromosome 9, homozygous CDKN2A/p14(ARF)/CDKN2B deletion and low TSC1 mRNA expression in pleomorphic xanthoastrocytomas. Oncogene 26:1088–1097 54. Weber RG, Sabel M, Reifernberger J, Sommer C, Oberstrab J, Reifernberger G, Kiessling M, Cremer T (1996) Characterization of genomic alterations associated with glioma progression by comparative genomic hybridization. Oncogene 13:983–994 55. Cairncross JG, Ueki K, Zlatescu MC, Lisle DK, Finkelstein DM, Hammond RR, Silver JS, Stark PC, MacDonald DR, Iuo Y, Ramsay DA, Louis DN (1998) Specific genetic predictors of chemotherapeutic response and survival in patients with anaplastic oligodendrogliomas. J Natl Cancer Inst 90:1473–1479 56. Pohl U, Cairncross JG, Louis DN (1999) Homozygous deletions of the CDKN2C/p18INK4C gene on the short arm of chromosome 1 in anaplastic oligodendrogliomas. Brain Pathol 9:639–643 57. Smith JS, Perry A, Borell TJ, Lee HK, O’Fallon J, Hosek SM, Kimmel D, Yates A, Burger PC, Scheithauer BW, Jenkins RB (2000) Alterations of chromosome 1p and 19q as predictors of survival in oligodendrogliomas, astrocytomas, and mixed oligoastrocytomas. J Clin Oncol 18:636–645 58. Goussia AC, Kyritsis AP, Mitlianga P, Bruner JM (2001) Genetic abnormalities in oligodendroglial and ependymal tumors. J Neurology 248:1030–1035 59. Sasaki H, Zlatescu MC, Betensky RA, Ino Y, Cairncross JG, Louis DN (2001) PTEN is a target of chromosome 10q loss in anaplastic oligodendrogliomas and PTEN alterations are associated with poor prognosis. Am J Pathol 159:359–367 60. Dong S-M, Pang JC-S, Poon W-S, Hu J, To K-F, Chang AR, Ng H-K (2001) Concurrent hypermethylation of multiple genes is associated with grade of oligodendroglial tumors. J Neuropathol Exp Neurol 60:808–816 61. Ino Y, Betensky RA, Zlatescu MC, Sasaki H, MacDonald DR, Stemmer-Rachamimov AO, Ramsay DA, Cairncross JG, Louis DN (2001) Molecular subtypes of anaplastic oligodendroglioma: implications for patient management at diagnosis. Clin Cancer Res 7:839–845 62. Reinfenberger G, Louis DN (2003) Oligodendroglioma: toward molecular definitions in diagnostic neuro-oncology. J Neuropathol Exp Neurol 62:111–126
48 63. Puduvalli VK, Hashmi M, McAllister LD, Levin VA, Hess KR, Prados M, Jaeckle KA, Alfred YWK, Buys SS, Bruner JM, Townsend JJ, Davis R, Sawaya R, Kyritsis AP (2003) Anaplastic oligodendrogliomas: prognostic factors for tumor recurrence and survival. Oncology 65:259–266 64. Wolf RM, Draghi N, Liag X, Dai C, Uhrbom L, Eklof C, Westermark B, Holland EC, Resh MD (2003) p190RhoGAP can act to inhibit PDGF-induced gliomas in mice: a putative tumor suppressor encoded on human chromosome 19q13.3. Genes Dev 17:476–487 65. Gelpi E, Ambros IM, Birner P, Luegmayr A, Drlicek M, Fischer I, Kleinert R, Maier H, Huemer M, Gatterbauer B, Anton J, Rossler K, Budka H, Ambros PF, Hainfellner JA (2003) Fluorescent in situ hybridization on isolated tumor cell nuclei: a sensitive method for 1p and 19q deletion analysis in paraffin-embedded oligodendroglial tumor specimens. Mod Pathol 16:708–715 66. Barbashina V, Salazar P, Holland EC, Rosenblum MK, Ladanyi M (2005) Allelic losses at 1p36 and 19q13 in gliomas: correlation with histologic classification, definition of a 150-kb minimal deleted region on 1p36, and evaluation of CAMTA1 as a candidate tumor suppressor gene. Clin Cancer Res 11:1119–1128 67. Alaminos M, Davalos V, Ropero S, Setien F, Paz MF, Herranz M, Fraga MF, Mora J, Cheung NK, Gerald WL, Esteller M (2005) EMP3, a myelin related gene located in the critical 19q13.3 region, is epigenetically silenced and exhibits features of a candidate tumor suppressor in glioma and neuroblastoma. Cancer Res 65:2565–2571 68. Tews B, Felsberg J, Hartmann C, Kunitz A, Hahn M, Toedt G, Neben K, Hummerich L, von Deimling A, Reifenberger G, Lichter P (2006) Identification of novel oligodendrogliomaassociated candidate tumor suppressor genes in 1p36 and 19q13 using microarray-based expression profiling. Int J Cancer 119:792–800 69. Ngo TB, Peng T, Liang X-J, Akeju O, Pastorino S, Zhang W, Kotliarov Y, Zenklusen JC, Fine HA, Maric D, Wen PY, De Girolami U, Black PMcL WuWW, Shen R-F, Jeffries NO, Kang D-W, Park JK (2007) The 1p-encoded protein stathmin and resistance of malignant gliomas to nitrosoureas. J Natl Cancer Inst 99:639–652 70. Lavon I, Zrihan D, Zelikovitch B, Fellig Y, Fuchs D, Soffer D, Siegai T (2007) Longitudinal assessment of genetic and epigenetic markers in oligodendroglioma. Hum Cancer Biol 13:1429–1437
A.C. Goussia et al. 71. Wernique C, Thiel G, Lozanova T, Vogel S, Kintzel D, Janisch W, Lehmann K, Witkowski R (1995) Involvement of chromosome 22 in ependymomas. Cancer Genet Cytogenet 79:173–176 72. Bijlsma EK, Voesten AM, Bijleveld EH, Troost D, Westerveld A, Merel P, Thomas G, Hulsebos TJ (1995) Molecular analysis of genetic changes in ependymomas. Genes Chromosomes Cancer 13:272–277 73. Ebert C, von Haken M, Meyer-Puttlitz B, Wiestler OD, Reifebenberger G, Pietsch T, von Deimling A (1999) Molecular genetic analysis of ependymal tumors. NF2 mutations and chromosome 22q loss occur preferentially in intramedullary spinal ependymomas. Am J Pathol 155:627–632 74. Lamszus K, Lachenmayer L, Heinemann U, Kluwe L, Finckh U, Hoppner W, Stavrou D, Fillbrandt R, Westphal R (2001) Molecular genetic alterations on chromosomes 11 and 22 in ependymomas. Int J Cancer 91:803–808 75. Reardon DA, Entrekin RE, Sublett J, Ragsdale S, Li H, Boyett J, Kepner JL, Look AT (1999) Chromosome arm 6q loss is the most common recurrent autosomal alteration detected in primary pediatric ependymoma. Genes Chromosomes Cancer 24: 230–237 76. Hirose Y, Aldape K, Bollen A, James CD, Brat D, Lamborn K, Berger M, Feuerstein BG (2001) Chromosomal abnormalities subdivide ependymal tumors into clinically relevant groups. Am J Pathol 158:1137–1143 77. Dyes S, Prebble E, Davison V, Davies P, Ramani P, Ellison D, Grundy R (2002) Imbalances in pediatric intracranial ependymomas define clinically relevant groups. Am J Pathol 161:2133–2141 78. Rajaran V, Leuthardt EC, Singh PK, Ojemann JG, Brat DJ, Prayson RA, Perry A (2004) 9p21 and 13q14 dosages in ependymomas. A clinicopathologic study of 101 cases. Mod Pathol 17:9–14 79. Rousseau E, Ruchoux MM, Scaravilli F, Chapon F, Vinchon M, De Smet C, Godfraind C, Vikkula M (2003) CDKN4A, CDKN4B and p14ARF are frequently and differentially methylated in ependymal tumours. Neuropathol Appl Neurobiol 29: 574–583 80. Maintz D, Fiedler K, Koopmann J, Rollbrocker B, Nechev S, Lenartz D, Stangl AP, Louis DN, Schramm J, Wiestler OD, von Deimling A (1997) Molecular genetic evidence for subtypes of oligoastrocytomas. J Neuropathol Exp Neurol 56:1098–1104
4
The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors Sofie Van Cauter, Silvia Kovacs, Caroline Sage, Ron Peeters, Judith Verhoeven, Sabine Deprez, and Stefan Sunaert
Abbreviations
Contents 4.1 Introduction.............................................................. 49 4.2 Basics of fMRI........................................................... 4.2.1 Biophysical Basis of fMRI......................................... 4.2.2 fMRI Acquisition........................................................ 4.2.3 fMRI Paradigms......................................................... 4.2.4 General Limitations of fMRI......................................
50 50 52 52 53
4.3 Basics of DTI............................................................. 4.3.1 Biophysical Basis of DTI........................................... 4.3.2 DTI Acquisition.......................................................... 4.3.3 DTI Data Visualization............................................... 4.3.4 General Limitations of DTI........................................
56 56 57 58 59
4.4 Presurgical fMRI and DTI....................................... 4.4.1 Presurgical fMRI and DTI in Practice........................ 4.4.2 Challenges and Limitations of Presurgical fMRI....... 4.4.3 Challenges and Limitations of Presurgical DTI.........
60 60 64 66
4.5
Conclusion................................................................. 67
Reference............................................................................. 67
BOLD Blood oxygenation level dependent CBF Cerebral blood flow CBV Cerebral blood volume CMRGlu Cerebral metabolic rate of glucose CMRO2 Cerebral metabolic rate of oxygen D Diffusion constant D Diffusion tensor DTI Diffusion tensor imaging FA Fractional anisotropy fMRI Functional magnetic resonance imaging GE-SS-EPI Gradient-echo single-shot echoplanar imaging IAT Intracarotid amobarbital test ICS Intraoperative cortical stimulation ISS Intraoperative subcortical stimulation MD Mean diffusivity MRI Magnetic resonance imaging PMC Premotor cortex PPC Posterior parietal cortex ROI Region-of-interest SM1 Primary sensorimotor cortex SMA Supplementary motor area T Tesla T2 Transverse relaxation time
4.1 Introduction S.Van Cauter, S. Kovacs, C. Sage, R. Peeters, J. Verhoeven, S. Deprez, and S. Sunaert (*) Department of Radiology, University Hospitals of the Catholic University of Leuven, Herestraat 49, 3000, Leuven, Belgium e-mail:
[email protected]
The choice of treatment of brain tumors largely depends on the precise localization of the lesion and its relationship to eloquent cortex or white matter tracts. In this view, neuroimaging plays a crucial role in establishing the diagnosis, planning the therapy, as well as evaluating
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_4, © Springer-Verlag Berlin Heidelberg 2011
49
50
therapeutic effects and detecting early recurrence of brain tumors. Many neuroimaging techniques have been developed over the last decades, all of which are characterized by their specific resolution, extent of invasiveness, and type of information that can be obtained. Whereas in the past, neuroimaging yielded mainly morphological information, it currently allows the multimodal assessment of brain lesions, incorporating biochemistry (e.g., indicators of cellular integrity) and physiologic parameters (e.g., hemodynamic variables). Advanced magnetic resonance imaging (MRI) can provide information on tumor cellularity, metabolism, and angiogenesis, all of which are important predictors for tumor grading, therapy, and prognosis. Functional MRI (fMRI) allows to localize important brain functions noninvasively, such as sensory or motor functions, language and memory in clinical (presurgical) setting. This way, fMRI may predict deficits in cognitive, sensory, or motor functions that might arise from surgical intervention. Diffusion tensor imaging (DTI) is a MRI technique that can be used to characterize the orientational properties of the diffusion process of water molecules in biological tissues. With this new imaging method, the orientation and integrity of white matter fibers can be demonstrated in vivo as the reconstruction of the trajectories of white matter tracts is possible using fiber tractography. In the case of brain tumors, DTI, and especially fiber tractography, allows to assess the anatomical relationship between the lesion and the white matter tracts and helps to discriminate between displacement and infiltration of highly relevant white matter tracts, thus contributing to presurgical risk assessment of surgical intervention. By superimposition of the fMRI and DTI results on high-resolution anatomical MR images within the submillimeter range, these techniques can also contribute to surgical planning by assuring precise stereotactic proceedings or by providing information for neuronavigation. Additionally, the results of fMRI and DTI can be used to delineate specific areas of interest for intraoperative (sub) cortical stimulation. Apart from the many possible applications of fMRI and DTI in presurgical settings, other advantages of fMRI and DTI include: (1) the possibility to acquire functional and anatomic maps concurrently and preoperatively, (2) its sensitivity to both superficial and deep regions, (3) its noninvasiveness, (4) its cost-effectiveness and (5) its ease-of-use as it can be performed with routine clinical scanners.
S. Van Cauter et al.
In this chapter, we first give an overview of the basis and practical implementation of both techniques. Further more, we provide an extensive section on the clinical applicability of fMRI and DTI for presurgical risk assessment and utilization of these techniques for intraoperative (sub)cortical stimulation and for neuronavigation. Finally, we discuss the challenges and limitations of fMRI and DTI that are specific for the use of these techniques in clinical settings.
4.2 Basics of fMRI fMRI is an imaging technique that can be used to visualize brain activity induced by motor, sensory, or cognitive tasks in a noninvasive way. The technique is based on the indirect measurement of neuronal activity by mapping the concomitant vascular changes or blood oxygenation-level dependent (BOLD) contrast that was first described by Ogawa and Lee [1, 2]. In the following sections, we will discuss the biophysical basis, the practical implementation, and the general limitations of fMRI.
4.2.1 Biophysical Basis of fMRI Activation of a neuronal cell population (i.e., more than baseline neuronal firing) leads to an increase in cerebral metabolic rate of oxygen (CMRO2), an increase in cerebral metabolic rate of glucose (CMRGlu), and an increased production of CO2. These are powerful physiological stimuli for cerebral vasodilatation, resulting in extended cerebral blood flow (CBF), cerebral blood volume (CBV) and blood oxygenation. This connection – or “coupling” – between neural activity and energy metabolism/CBF is the foundation of functional neuroimaging and has already been clearly established in animal studies and, to a lesser degree, in human studies [3–6]. The function of this “neurovascular coupling” seems obvious, as increased neuronal metabolism requires increased oxygen and glucose supplies (Fig. 4.1). The reality, however, is more complicated: during increased neuronal activity, the rising amount of oxygen does not match the extended amount of oxygen consumption (Fig. 4.1a, b). During the first second(s) of brain activation, the oxygen concentration of the blood drops slightly (the “early response”), which reflects an increase in CMRO2 (Fig. 4.1c). However, this is followed by a
51
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors b
a
c
[oxyhemoglobin]
12 10 8 6 4 2 0 -2 0
5
10
15
20
time(s)
d
Blood Oxygenation Level Dependent contrast
Activation neuronal cell population
CMRO2: small increase CBF: huge increase
Increased blood oxygenation
Increased T2∗
Increased MR signal
Fig. 4.1 The biophysical basis of fMRI. (a) At rest, neurons extract oxygen (black circles) from hemoglobin molecules (white circles) in the blood. The blood flow maintains a constant level of blood oxygenation and oxygen/glucose supply. (b) When activated, the neurons extract a small additional amount of oxygen from the blood, while the CBF is hugely increased. This mismatch between oxygen extraction and CBF forms the basis of the blood oxygen level-dependent (BOLD)
contrast. (c) Time course describing the change in the concentration of oxyhemoglobin in the blood upon neuronal activation; immediately after neuronal activation, there is a slight dip in the concentration of oxyhemoglobin (early response, indicated by the dotted rectangle), which is followed by a huge increase (late response, indicated by the dashed rectangle). (d) Schematic representation of the events leading to a BOLD response
huge increase in CBF that overcompensates the metabolic demand for oxygen, which results in a local increase of oxygen concentration in the blood (the “late response”) (Fig. 4.1c). Although the early response corresponds both temporally and spatially better to the neuronal activity, this is difficult to measure directly. The physiological motive of the CBF overshoot is still not fully understood. Some authors believe it is a mechanism to discharge potentially deleterious by-products of cerebral metabolism and metabolic produced heat [6]. In fMRI, increased neuronal activity is measured indirectly using the rise in blood oxygenation originating from the secondary hemodynamic changes in the “late response” of the neurovascular coupling. When measuring BOLD responses in an fMRI experiment, three parameters can contribute to the recorded fMRI signal: the blood flow velocity, the blood flow volume, and the blood oxygenation level. By carefully choosing MR imaging parameters, the contribution of blood flow velocity and blood flow volume can be minimized, while
maximizing the blood oxygenation dependency of the fMRI signal, hence the name “blood oxygenation leveldependent” (BOLD) contrast. The BOLD imaging contrast has its basis in the observation of Thulborn et al. [7], who demonstrated that the transverse relaxation time (T2) of water protons provides information about the oxygenation state of the blood. In blood, oxygen is transported by hemoglobin. The magnetic state of the iron contained in the heme group of hemoglobin is dependent on the oxygenation state. In oxyhemoglobin, it is diamagnetic, whereas in deoxyhemoglobin it is paramagnetic due to the more free exposure of the iron molecule. Deoxyhemoglobin creates magnetic field gradients around the red blood cells and in the tissue surrounding the vessels. These field gradients shorten T2* of the tissue and reduce the MRI signal at rest from what it would be if there were no deoxyhemoglobin present. When the concentration of oxyhemoglobin increases, the distorting field gradients are reduced. Therefore, T2* becomes longer and the signal measured with a T2*-weighted pulse sequence
52
increases by a few percent. The T2 of blood is thus dependent on the degree of oxygenation of the blood, which provides an endogenous contrast agent detectable with MRI. In summary (Fig. 4.1d), the BOLD signal changes measured with fMRI originate from the imbalance between the increase in CBF and the increase of CMRO2 resulting from a change in neural activity. Following neuronal activity, CBF increases more than CMRO2, which results in a relative rise in the oxyhemoglobin/deoxyhemoglobin ratio in the venous and capillary blood. The MRI signal can be made sensitive for this change, as deoxyhemoglobin is paramagnetic and therefore reduces the MRI signal at rest. The MR signal will thus increase slightly during neural activity.
4.2.2 fMRI Acquisition Susceptibility-weighted sequences – T2*-weighted images – are very suitable to measure the local susceptibility changes induced by the fMRI BOLD contrast upon neuronal activation. Although BOLD activations have been demonstrated with many different T2*-weighted imaging schemes, most fMRI experiments are performed with gradient echo single-shot echo planar imaging (GE-SS-EPI) sequences. EPI allows to acquire multiple adjacent slices in a very limited amount of time: a single slice MRI image can be obtained in as little as 70–100 ms, whereas covering the entire brain takes about 2–3 s [8].
4.2.3 fMRI Paradigms In fMRI, “differential activity” is measured. Anatomical regions involved in motor, sensory, or cognitive processing can be visualized by comparing fMRI images obtained during the execution of a specific task to those obtained during a control condition. Therefore, specific task designs or “paradigms” have been developed in order to disambiguate different patterns of neuronal activity that reflect specific tasks or cognitive states. There are two major types of paradigms: blocked and event-related paradigms (Fig. 4.2). Blocked fMRI paradigms or “box-car designs” are based on an “on-off principle.” Periods of task performance are alternated with a control task (rest or an
S. Van Cauter et al.
alternative task), each lasting approximately 20–40 s (Fig. 4.2a). Blocked designs are the most common paradigms used in clinical settings because of its distinct advantages. The robust design is easy to explain and perform. The data processing is relatively simple. However, as the conditions are alternated in a regular pattern, habituation, fatigue, and anticipation possibly occur. This type of paradigm is widely used for mapping of motor, sensory (like tactile, auditory, visual perception), and language functions. Functional brain areas involved in motor execution can be mapped by alternating movement with rest. Due to the somatotopical organization of the central sulcus (Fig. 4.3), different tasks can be performed depending on the relation between the site of the lesion and the central sulcus. Patients can, for example, be asked to perform fingertapping, toe flexion/extension, or lip pouting [9–12]. In paretic patients, specific movements can be difficult or even become impossible. In such cases, paradigms can be simplified (e.g., opening and closing of the hand(s) instead of finger tapping). In paralyzed patients, indirect mapping of the primary motor cortex can be performed by sensory stimulation of the limb [13]. This induces activation in the adjacent primary sensory cortex, which is indicative for localization of the primary motor cortex. Language functions can be mapped using different tasks, depending on whether principally expressive language areas or comprehensive language areas are to be visualized. A typical activation pattern of the language network is shown in Fig. 4.4. Language expression tasks include verbal fluency, picture naming and verb-to-noun generation, [14] and activate preferably the pars opercularis and triangularis of the inferior frontal gyrus (Broca’s area) and additional expressive language areas in the middle and superior frontal gyrus [15, 16]. Language comprehension tasks, for example, semantic judgments tasks, reading or listening to spoken language, preferably activate the gyrus temporalis superior and medius (Wernicke’s area) [17, 18]. Additionally, activation in the dominant parieto-occipito-temporal junction (angular gyrus) is often visualized in language tasks. We refer to this focus as Geschwind’s area as Norman Geschwind described the dominant angular gyrus as a visual association area in language processing [19, 20]. It should be noted, however, that most of these language tasks, elicit activation of the different areas of the language network since both language comprehension and expression are required during language tasks.
53
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors
a REST
Fingertap
REST
Fingertap
ON
Fingertap
ON
OFF
0s
REST
REST
Fingertap
ON
OFF
ON
OFF
OFF
30s
60s
90s
120s
150s
180s
210s
240s
10s
20s
30s
40s
50s
60s
70s
80s
b
0s
Fig. 4.2 Schematic representation of a blocked and an eventrelated paradigm. (a) In a blocked fMRI design, periods of task performance (e.g., fingertapping) are alternated with a control task (e.g., rest), each lasting a fixed amount of time (e.g., 30 s). (b) In an event-related fMRI design, periods of task performance
(e.g., looking at fearful faces) are alternated with control tasks (e.g., looking at neutral or angry faces). All stimuli are shown pseudorandomly during a limited amount of time (e.g., 1 s). The pictures were taken from the series of Ekman and Friesen [121]
Event-related fMRI refers to a technique for detecting the brain’s response to brief stimuli or “events.” Stimuli are pseudorandomly presented for a short period of time (one to a few seconds) (Fig. 4.2b), which prevents habituation, fatigue, and/or anticipation to occur. However, event-related designs require more advanced and elaborate postprocessing and are statistically less powerful. Consequently, event-related designs have only limited implementations in clinical settings. These paradigms have been applied extensively in psychological and psychiatric settings to study emotion or behavior [21–23]. For example, to identify brain regions activated during an emotional state of anxiety, pictures of fearful faces can be shown pseudorandomly for a few seconds, alternated with, for example, pictures of faces with neutral expression (Fig. 4.2b). This example also clearly illustrates the importance of the control condition: if the control
condition would consist of looking at a black screen, looking at fearful faces would not only activate the limbic system but also different components of the visual system that are involved in face processing. By using looking at neutral faces as a control condition, the visual components are not visualized when comparing the task to the control condition.
4.2.4 General Limitations of fMRI FMRI has the distinct advantage of its noninvasiveness in investigation of brain function. However, fMRI has some general limitations, regardless of whether fMRI is performed in clinical or research settings. Limitations of fMRI that are specific for performing fMRI in clinical settings will be discussed later (see Sect. 4.2). The
54
Fig. 4.3 Somatotopy of the primary sensorimotor cortex (SM1) and cerebellum, visualized by fMRI on axial T1-weighted 3D-TFE slices. A healthy volunteer performed three series of motor tasks: lip pouting, bilateral fingertapping, and bilateral feet movement, shown in orange, magenta, and blue, respectively.
S. Van Cauter et al.
The somatotopical organization of SM1 and the cerebellum are clearly demonstrated, with the foot representation located most medially, the lip representation most laterally, and the hand representation in between
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors
55
Fig. 4.4 Language network visualized by fMRI on coronal (upper), axial (middle), and sagittal (lower) T1-weighted 3D-TFE slices. A healthy volunteer performed a verb-to-noun generation task, during which nouns were presented visually and the subject had to think of matching verbs. Significant activa-
tions can be seen in (1) left inferior frontal gyrus (Broca’s area), (2) left middle temporal gyrus (Wernicke’s area), (3) left angular gyrus (Geschwind’s area), and (4) SMA. Also note the additional activation of (5) right inferior frontal gyrus (homologue of Broca’s area). R right; L left
general limitations relate to the origin of the BOLD contrast itself, the effect of the static magnetic field of MRI scanners used for fMRI, the sensitivity to susceptibility of MRI acquisition sequences, and the effect of head motion. The BOLD contrast finds its basis in changes in blood oxygenation resulting from hemodynamic changes secondary to increased neuronal firing, which is thus an indirect way to visualize brain function. This feature gives rise to a certain spatial en temporal uncertainty, when compared to invasive electrophysiological methods to measure brain function. The spatial specificity that can be achieved with fMRI has been shown to be techniquedependent [24–26]. Implementation of imaging parameters confining the fMRI signal toward the site of neuronal activity is thus a prerequisite for clinical fMRI examinations [27]. Current MRI technology confines the spatial uncertainty to approximately 5 mm [28, 29].
FMRI investigations have mainly been performed at 1.5 Tesla (T) systems. Currently, fMRI examinations are increasingly obtained on MRI scanners of higher static magnetic field (3T) which exploits the advantages of a better signal-to-noise ratio and more accurate spatial localization. These can mainly be attributed to the fact that T2*, and subsequently the BOLD effect correlate linearly with the static magnetic field for a given vascular deoxyhemoglobin concentration for water protons in veins and large vessels, and quadratic for water protons in capillaries [30–32]. Consequently, the BOLD signal is directed more toward the site of neuronal activation. However, fMRI data acquired at higher static magnetic fields suffer from increased susceptibility artifacts, which reduce spatial accuracy [33]. As previously mentioned, fast susceptibility-weighted sequences are most suitable to measure BOLD-changes upon increased neuronal activity. These MRI sequences,
56
however, are prone to distortion and susceptibility artifacts. Compared to high-resolution T1-weighted anatomical images, image distortions of 10–20 mm can occur due to susceptibility-related field inhomogeneities [34, 35]. Signal drops near air–tissue boundaries like the temporal pole or the orbitofrontal cortex are especially problematic. In patients, metal implants, fixation material from previous surgery, and lesions containing hemosiderin (e.g., arteriovenous malformations, necrosis or bleeding) further distort the MRI signal and might even result in false positive or absent activations. Therefore, fMRI results should be interpreted with great caution in both research and clinical settings, as (the absence of) an “activation” may not always be induced by proper neuronal activity. In this view, the original fMRI images should be checked in order to exclude signals related to distortion or susceptibility artifacts. Some solutions have been proposed to partially overcome this problem. In parallel MRI, spatial information related to several coil elements (2–32) is used to reconstruct images. Parallel reconstruction techniques result in improved image quality by increasing signal-to-noise ratio, spatial resolution, and temporal resolution in dynamic MRI scans while reducing artifacts [36–39]. Finally, head motion is an important issue in fMRI, as this may prevent the correct visualization of the BOLD response. Head motion not only results in displacement of anatomical areas, it may even lead to complete signal loss. In the former case, voxels that correspond to a specific anatomical area will no longer correspond to that area after occurrence of head motion, which renders the data analysis problematic. Furthermore, since the fMRI signal change upon neuronal activation is only 2–5% [27], it will become very difficult to detect such changes in data that is corrupted by head motion. Head motion is most problematic when it occurs either correlated with the paradigm, which results in “false” activations, or in a direction through the image acquisition plane, which results in signal loss. A practical tip to distinguish between “real” and “false” activations is to consider the location of the activation: most of the motion-induced “false activations” are localized at the borders of different structures, which show a large signal intensity gradient between them [40]. Also, motioninduced signal changes typically have a ring-like appearance at these boundaries. Some solutions have been proposed to deal with head motion. First, fixation of the head in the coil of the MR scanner is a simple, though helpful, remedy [41]. Second, the scan plane can be adjusted to match the
S. Van Cauter et al.
expected direction of head motion, thus avoiding the occurrence of through-plane head motion. For example, if a subject is expected to nod during the task, a sagittal acquisition plane is more favorable than a transversal one. Finally, some processing algorithms can be applied to correct for head motion. During preprocessing, head motion can be corrected by performing realignment of the data [42], whereas during postprocessing, the head motion can be filtered out in the statistical analysis [43]. However, real activation foci may be filtered out as well, especially when head motion is correlated with the onset of task performance in the paradigm. A more detailed description of data processing of fMRI and the challenges thereof is beyond the scope of this chapter and is discussed elsewhere (for a review see Turner et al. [44]).
4.3 Basics of DTI DTI is an MRI technique that can be used to characterize the orientational properties of the diffusion process of water molecules in biological tissues [45–47]. In this imaging technique, the orientation and integrity of white matter fibers can be demonstrated in vivo [48, 49]. In the following sections, we will discuss the biophysical basis, the acquisition, the processing, and the general limitations of DTI.
4.3.1 Biophysical Basis of DTI Diffusion is the random motion of molecules due to their thermal energy and is described by a diffusion constant D. For pure water at 37°C, D is approximately 3.4 × 10−3 mm2/s. In a homogeneous fluid of infinite extent, diffusion is similar in all spatial directions (“isotropic diffusion”). The extent to which water molecules diffuse in tissues is affected not just by the physicochemical properties of the cytosol itself, but also by the presence of cellular structures that provide barriers to free diffusion (Fig. 4.5). Water molecules will thus be impeded in their movement by natural barriers such as cell membranes, other macromolecules, or subcellular structures. In tissues that have a highly organized structure, diffusion may be more restricted in one direction than in another. Diffusion that has strong directional dependence of this type is said to be “anisotropic” (Fig. 4.5a).
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors
In cerebral white matter, water molecules diffuse more freely along the direction of axonal fascicles than across them. The restriction of free water diffusion is mainly due to the axonal membrane, axonal microtubule, and the axonal myelin sheet [50]. White matter tracts therefore show a high degree of anisotropy.
4.3.2 DTI Acquisition To measure diffusion with MRI, the data acquisition has to be made sensitive to Brownian motion of water molecules in biological tissues. In this view, the most common approach is to use a spin-echo pulse sequence in which strong rectangular gradient pulses, equal in direction and duration but opposite in magnitude, are implemented before and after the 180° refocusing pulse to obtain diffusion-weighted images [51]. In the absence of incoherent motion, these diffusion sensitizing gradients would not affect the MRI signal, as the first gradient dephases the MRI signal, while the second rephases it. However, if the water molecules move between one gradient pulse and the other, the rephasing of the MRI signal is incomplete and the measured signal is decreased [52]. By deliberately applying large magnetic field gradients in particular directions, molecular diffusion can be made the dominant image contrast mechanism, enabling variations in diffusion to be visualized, including their directional dependence. In highly ordered structures, diffusion cannot be described by a single quantity such as a diffusion constant. In anisotropic structures, the diffusion process can be characterized by a diffusion tensor D [47]. Theoretically, to calculate this diffusion tensor, at least six independent measurements with diffusion gradients applied along six noncollinear directions are required. In doing so, a diffusion tensor can be calculated for each voxel. The diffusion tensor can be summarized by three eigenvalues (l1, l2, l3), which represent the magnitude of the diffusion in the direction of the corresponding eigenvectors. In anisotropic tissues organized in parallel bundles, the largest eigenvalue (l1) represents the diffusion coefficient along the direction parallel to the fibers, while the two remaining eigenvectors (l2 and l3) represent the transverse diffusion coefficients. The diffusion tensor is often visualized as an ellipsoid. In the ellipsoid, the three eigenvectors can be used to construct a “local frame of reference” within a voxel (Fig. 4.5b). In anisotropic fibrous tissues, the principal directions of
57
D must coincide with the orthotropic directions of that tissue. The eigenvector of the largest eigenvalue (l1) will thus define the “local fiber tract axis” of that voxel, while the two remaining eigenvectors perpendicular to it should define the two remaining orthotropic axes (Fig. 4.5b). Using these eigenvalues, different quantitative scalar measures can be calculated. These measures are invariant to the rotations of the frame of reference, which
Fig. 4.5 Basics of DTI. (a) Representation of isotropic (upper part) and anisotropic (lower part) diffusion by a sphere and an ellipsoid, respectively. (b) The diffusion ellipsoid represents the surface of constant mean squared displacement of water molecules that would arise in a hypothetical experiment in which they were released at the center of the voxel and were allowed to diffuse for a certain time. It displays the mean diffusion distances in each of the three principal directions during the diffusion time. The three eigenvectors and eigenvalues (l1, l2, l3, with l1 > l2 > l3) that describe the ellipsoid can be used to construct a “local frame of reference” within a voxel
58
S. Van Cauter et al.
enables comparison between different results. These include: (1) the fractional anisotropy (FA), which is a scalar index of the amount of anisotropy (directionality) that scales from 0 ( isotropic diffusion) to 1 (diffusion occurs in one direction only) and (2) the mean diffusivity (MD), which is a measure of the magnitude of diffusion (unit: mm²/s) [53]. These measures allow to differentiate between different brain tissues, as different tissue types have different diffusion characteristics. White matter is characterized by a high degree of anisotropy and a low MD. The combination of low anisotropy measures and low MD is suggestive of gray matter, while the cerebrospinal fluid will show low FA values and high MD values. Furthermore, these measures can be used to assess the integrity of white matter. For example, Pierpaoli et al. [54] established that the observation of reduced FA
and increased MD is suggestive of Wallerian degeneration and it has even been demonstrated that FA and MD can be used to evaluate the time course of Wallerian degeneration occurring after ischemic stroke from the early subacute to the chronic stage [55].
4.3.3 DTI Data Visualization The quantities derived from the diffusion tensor can easily be displayed individually as grayscale images, for example, FA maps or MD maps (Fig. 4.6a). However, interpretation of such individual images is rather difficult and it would be much easier if the information of the three components of the vector could be displayed in one single image in which this information is encoded in
superior
anterior left right posterior
inferior
Fig. 4.6 Different visualization schemes of diffusion tensor data. On the left, an FA and an MD map are shown as grayscale images. In the middle, a color-encoded FA map is shown, with the color encoding according to the color conventions of Pajevic and Basser [56]: red = right–left; blue = anterior–superior; green = inferior–superior, as illustrated by the arrows (lower middle).
The degree of anisotropy is represented by the intensity of the voxels. On the right, a part of the color-encoded FA map is enlarged (indicated by the white square) to show the diffusion ellipsoids (lower right) that were also color encoded according to the same color-encoding scheme for this visualization
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors
some way. This can be achieved by color-encoded FA maps, in which colors are assigned according to the eigenvector associated with the largest eigenvalue. By convention, red is assigned to the x-direction (left–right), green to the y-direction (anterior–posterior) and blue to the z-direction (superior–inferior) [56]. Intensities on these maps are scaled in proportion to the magnitude of FA (Fig. 4.6b). By combining FA with the directionality, it is possible to obtain estimates of fiber orientation. This has lead to the development of fiber tractography in which three-dimensional trajectories of white matter tracts can be reconstructed. Different algorithms have been developed for fiber tractography. In Fig. 4.7, we illustrate how the fasciculus arcuatus can be reconstructed using fiber tractography. First, a seed region-of-interest (ROI) is defined from which the tracking is initiated (Fig. 4.7a). By defining an additional AND ROI, the tract reconstruction can be constrained to
Fig. 4.7 Reconstruction of the fasciculus arcuatus using fiber tractography. (a) First, a seed ROI is defined on a sagittal slice (orange), which includes the inferior frontal and precentral gyrus. (b) Then, a second ROI is defined on a sagittal slice (green), which includes the inferior parietal lobulus and the temporal lobe. (c) By performing an AND operation, the reconstructed tract will only consist of streamlines that pass through both ROIs. (d) Doing so, a correct reconstruction of the fasciculus arcuatus (blue) can be obtained
59
only reconstruct the tracts that pass in both ROI (Fig. 4.7b). Doing so, a correct reconstruction of the fasciculus arcuatus can be obtained (Fig. 4.7c). In the preoperative planning of the resection of tumors, fiber tractography is used to evaluate the location of white matter tracts nearby the tumor. This way, damaging of the tracts by the operation can be avoided and vital motor and or cognitive functions can be spared.
4.3.4 General Limitations of DTI Although DTI allows to assess the white matter integrity and even to reconstruct various cerebral white matter tracts, some general limitations of DTI should be taken into consideration. Limitations of DTI that are specific for performing DTI in clinical settings will be discussed later (see Sect. 4.4.3).
60
One of the most important limitations stems from the fact that DTI reflects the averaged water diffusion within a voxel and is therefore an indirect indicator of the axonal structures. Within a single voxel, different tissue types may be represented in varying degrees. In many cases, DTI results are biased by the dominant axonal component. Especially in regions where different white matter fibers cross, the combined vector may not represent either direction, which would prevent the fiber tractography algorithm from following the correct tract [57]. Another limitation of fiber tractography is that in current imaging and processing strategies, it is almost impossible to make the differentiation between “kissing” (><), “crossing” (X) and “branching” (Y) tracts. This may hamper adequate reconstruction of the entire white matter tract, as the tracking may be stopped at a fiber crossing, or the tracking may yield an erroneous reconstruction if the dominant direction of diffusion does not match the expected direction of the tract. In this view, more elaborate modeling of the diffusion tensor has been proposed to solve the fiber crossings [57], but such techniques are not readily available for clinical examinations. Therefore, the fiber tractography results should always be interpreted with great caution. It should also be emphasized that fiber tractography provides tract reconstructions, which consist of a number of streamlines. These streamlines are only visualizations and do not represent actual axonal bundles. Finally, DTI sequences are prone to distortion and susceptibility artifacts, although DTI images are less severely affected by these artifacts than fMRI images [58].
4.4 Presurgical fMRI and DTI FMRI and DTI are increasingly employed in presurgical planning and therapeutic optimization in patients with brain pathology. The goals of presurgical fMRI and DTI are threefold: (1) assessing the risk of neurological deficit that follows a surgical procedure and determination of the operative trajectories, (2) selecting patients for invasive intraoperative mapping, and (3) guiding of the surgical procedure itself using neuronavigation. However, presurgical DTI and fMRI have some specific limitations. These topics will be further addressed here.
S. Van Cauter et al.
4.4.1 Presurgical fMRI and DTI in Practice 4.4.1.1 Preoperative Risk Assessment The ultimate goal of neurosurgery of patients with brain pathology is to achieve maximal resection of the lesion, while avoiding surgically induced neurological deficits. In this view, it is important to know the spatial relation between the pathology and functional brain areas and/or structures. In undistorted brain anatomy, different anatomical landmarks can be used to identify eloquent cortical areas [59–63] (for an extensive review of anatomy of white matter tracts and an overview of white matter pathology: see [64]). However, in patients with brain tumors, the delineated aspect of gyri and sulci is often erased and the deep gray and white matter is often distorted, due to the mass-effect of the lesion and/or the concomitant edema. In such patients, the use of anatomical landmarks is very limited. In presurgical settings, fMRI and DTI are mostly used to assess the distance between the margin of the lesion and specific functional areas and/or white matter structures. This way, fMRI and DTI have the potential to predict possible deficits in motor, sensory, or cognitive functions due to tumor resection or lesion growth, thus allowing preoperative risk assessment. In this view, it is important to define “safe distances” for fMRI and DTI. In fMRI, there is a general conception that a minimal distance of 10–15 mm between the margin of the lesion and the activation focus is required to perform a safe resection [65]. In DTI, safety margins of about 5 mm have to be taken into account when approaching eloquent tracts [66]. It should be noted however, that there are still no generally accepted standards concerning the distance that is to be considered “safe” for either of these techniques. Furthermore, it should also be emphasized that tract reconstructions and fMRI activation foci result from statistical analysis of DTI and fMRI data. By adjusting the statistical thresholds of fMRI results, the extent of an activation focus may increase or diminish (Fig. 4.8a). On the other hand, by changing the algorithm settings and/or the ROI placement for the fiber tractography, the resulting tract reconstructions may vary in their spatial extent (Fig. 4.8b) [67, 68]. We illustrate the combined use of fMRI and DTI as complementary techniques in the preoperative risk assessment by the following case: a 28-year-old female
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors
61
Fig. 4.8 The effect of thresholding on fMRI and DTI results. (a) fMRI results of a bilateral fingertapping task in a patient with multiple metastases, with the largest lesion located in the right postrolandic region. By applying a strict threshold (lower), only limited activation of the primary sensorimotor cortex (SM1) is visualized, whereas a more extensive activation of SM1 is found
when applying a less stringent threshold (upper). Also note the activation of the SMA at a lower thresholds, which is absent at a stricter threshold. (b) Reconstruction of the right corticospinal tract using fiber tractography in the same patient. By varying the FA threshold for the fiber reconstruction, the tract reconstruction will also vary in spatial extent. R right; L left
right-handed patient with a large meningioma with mass-effect on the frontal lobe was referred for presurgical fMRI and DTI. The fMRI examination consisted of three separate experiments, in which she was asked to perform a bilateral fingertapping task, lip pouting, and toe flexion/extension separately. The results of these experiments are shown in Fig. 4.9. Note the activation foci in the pre- and postcentral gyrus representing the
primary sensorimotor cortex (SM1) in both hemispheres. The somatotopic organization of the pre- and postcentral gyrus is illustrated, with the activation in SM1 due to lip pouting being located most laterally and the activation in SM1 due to toe flexion/extension most medially (Fig. 4.9a). In the right lesioned hemisphere, activation foci corresponding to the hand and lip representation in SM1 are slightly displaced posteriorly. All
62
S. Van Cauter et al.
Fig. 4.9 Presurgical fMRI and DTI results in a patient with a meningioma with mass effect on the frontal lobe. (a) Three fMRI experiments were performed: lip pouting (upper), bilateral fingertapping (middle), and toe flexion/extension (lower). The somatotopical organization of the primary sensorimotor cortex (SM1) is clearly demonstrated with the foot representation located most medially, the lip representation most laterally, and the hand representation in between. In the right lesioned
hemisphere, activation foci corresponding to the hand and lip representation in SM1 are slightly displaced posteriorly. Note that the activation of the SMA is located in close vicinity of the lesion. (b) The corticospinal tract was reconstructed using fiber tractography. It is demonstrated that this white matter tract is compressed by the meningioma and runs closely to the border of the lesion
activation foci representing SM1 are located +/− 10 mm posterior from the lesion. Additionally, activation of the supplementary motor area (SMA), a secondary motor area for planning motor action and bimanual control, is
located adjacently and medially to the lesion. The reconstruction of the corticospinal tract with DTI shows that this white matter tract is compressed by the meningioma and runs adjacently to the border of the lesion
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors
(Fig. 4.9b). Surgical intervention is highly riskful, as both the SMA and the corticospinal tract may be damaged. Besides measuring the distance between the border of the lesion and an activation foci, fMRI has established an important role in assessing language lateralization in patients with intractable temporal or frontal lobe epilepsy. Surgical resection of epileptic brain tissue in patients with intractable seizures has been shown to improve outcome in comparison with medical therapy, provided that accurate localization of the epileptogenic focus as well as adjacent functional zones could be obtained [69]. The intracarotid amobarbital test (IAT) or the Wada test has been the standard procedure in presurgical evaluation of language and memory lateralization. In this test, sodium amobarbital is injected unilaterally into the internal carotid, which temporally anesthetizes the ipsilateral hemisphere. Language and/ or memory functions of the contralateral hemisphere can thus be assessed. This procedure is invasive and uncomfortable and so both hemispheres have to be examined separately. FMRI, due to its noninvasiveness and the possibility of testing both hemispheres simultaneously, is gradually replacing the Wada test [14, 70, 71]. FMRI findings have shown promising correlations with the IAT, as illustrated in more than 25 studies comparing both techniques [70]. For example, Binder et al. investigated 22 patients undergoing evaluation for epilepsy surgery with a semantic decision task and showed concordant results of the Wada test and fMRI in all cases [72].
4.4.1.2 Intraoperative Mapping When a lesion is considered to be at a close distance to either a functional area or a white matter tract, the neurosurgeon may use intraoperative mapping to determine the exact resection margin during the surgery itself. Direct intraoperative cortical and subcortical electro-stimulation (ICS and ISS) are considered reference techniques for localization of eloquent cortical and subcortical brain areas and allow submaximal resection of the lesion while minimizing the resulting functional deficit [73–75]. The patient is awake during this specific part of the surgery and is asked to perform certain tasks, while specific brain areas are reversibly and locally disrupted [76]. Several studies focused on the concordance between electro-stimulation and fMRI
63
to validate presurgical fMRI results [77–81]. On the contrary, only limited studies are available on the combination of direct subcortical electro-stimulation and DTI, as this is still an emerging application [67, 82]. ICS and ISS are mostly used to validate fMRI or DTI results intraoperatively when an eloquent cortical or subcortical area is located adjacent to the lesion, which represents a high risk of induction of postoperative functional deficits. Furthermore, ICS and ISS are indispensable for discrimination of essential versus expandable regions. Although a certain task may elicit activation in multiple (sub)cortical areas, these areas are not at all essential for performance of the task. For example, in the case discussed above (see Fig. 4.9), activation of both the SM1 and the SMA could be demonstrated during the different motor tasks. If the SM1 is damaged during surgery, a severe long-lasting functional deficit (e.g., paresis or paralysis) will be induced. The SM1 is thus considered an “essential” area. On the other hand, if the SMA is damaged during surgery, the induced functional deficits have been reported to be less severe (depending on the extent of the damage) and to recover more rapidly [83]. Therefore, the SMA can be considered as an “expandable” brain region. Although these techniques are considered reference standards for localization of functional brain areas and structures, such a procedure cannot be performed in any subject. The patient is required to be awake for a part of the surgery and not all subjects may be able to tolerate this. In such cases, the surgeon has to rely on the combined results of presurgical fMRI and DTI. A surplus value of fMRI and DTI is that these techniques allow to perform the ICS and ISS more precisely and well-directed, which limits both the duration of the surgery and the extent of the craniotomy.
4.4.1.3 Surgical Guidance Using Functional Neuronavigation The third goal of presurgical fMRI is to provide guidance for the surgical procedure. Frame-based or frameless stereotactic surgery or neuronavigation has been widely applied in neurosurgery already for more than 20 years [84]. Preoperatively obtained CT or MRI scans are integrated in a neuronavigation station and help the neurosurgeon to perform safe and efficient surgery. The integration of fMRI activation maps or
64
DTI-based tract reconstructions coregistered to high resolution MRI anatomic scans into a neuronavigation station is referred to as “functional neuronavigation.” Using functional neuronavigation, the surgeon can rely on both intraoperative structural data (anatomical location of the lesion, location of adjacent white matter tracts) and functional data (location of essential areas) [76, 85–88]. This way, the extent of the resection can be limited, which may also limit the risk of postoperative deficits. Wu et al. demonstrated the contribution of DTI-based functional neuronavigation to ideal presurgical planning and intraoperative image guidance for maximal safe resection of cerebral gliomas with pyramidal tract involvement in a randomized control trial of 238 patients. Postoperatively, motor deficits were fewer and less severe and high-quality survival in the patients with high grade glioma was increased by the use of DTI-based neuronavigation [89]. Neuronavigation, based on preoperatively acquired CT or MR images has one major disadvantage. During surgery, a certain amount of brain shift occurs after craniotomy when cerebrospinal fluid is lost and brain tissue is removed [90]. Therefore, the accuracy of preoperatively obtained data decreases significantly during surgery. This brain shift is also one of the main reasons to perform ICS and ISS, as the structures visualized in previously obtained fMRI activation maps or DTI-based tract reconstructions may have shifted after craniotomy. Roberts reported brain shifts of up to 10 mm, with the largest component being in the direction of gravity [91], whereas Hastreiter reported brain shifts of up to 24 mm after craniotomy [92]. The extent of brain shift can be analyzed quantitatively by intraoperative imaging methods such as ultrasound or intraoperative use of MRI. Intraoperative MRI images (obtained with MR scanner of static magnetic field 0.2 T up to 3 T, located in the operating theater) update neuronavigation images to compensate for brain shift [93, 94]. The use of MR scanners in the operating theater seems to be safe and reliable, although the surgical procedure has to be adapted to the specific requirements of an MRI environment. Updating preoperative functional neuronavigation data by coregistration to intraoperative 3D ultrasound, is an alternative method to correct brain shift and has been shown to be flexible and inexpensive [95–97]. However, despite recent advances, reliable prediction of brain shift remains impossible.
S. Van Cauter et al.
4.4.2 Challenges and Limitations of Presurgical fMRI In clinical settings, additional factors may complicate the execution and interpretation of an fMRI experiment. These factors relate to the biochemical environment of the lesion, the patient himself/herself and the lack of standardization. First, due to the presence of brain pathology or alteration of the cerebrovascular system, the physiological BOLD contrast may be altered or even absent. Several factors, including vasogenic edema, vascular compression by the mass effect of the tumor and tumoral hemorrhage, may contribute to the observation of a decreased BOLD response in near-lesion brain tissue. For example, the BOLD response in the vicinity of glial tumors may not accurately reflect the electrical neuronal activity, as this tumor type can induce abnormal vessel proliferation in adjacent brain tissue, thereby altering regional CBF, CBV, vasoactivity, and potentially, the BOLD contrast [98, 99]. Congenital vascular malformations show vascular steal effects, compression, and loss of cerebral autoregulation, all of which will have profound deleterious effects on the BOLD contrast [100, 101]. Similar effects may be present in the case of highly vascularized metastases, in which aberrant vasculature may prevent a correct BOLD response to be elicited. Additionally, metabolic changes in some brain tumors could induce environmental changes, which may again eliminate the physiological hemodynamic response. Biological tumoral characteristics may lead to alterations in pH, lactate, glucose, ATP, K+ levels, and NO release by macrophages and astrocytes, resulting in loss of autoregulation and abnormal response to various physiological and pharmacological substances [102–104]. In this view, it should be noted that the absence of the typical BOLD response does not automatically mean that there is no neuronal activity; it may well be that there is correct neuronal activity, but that the BOLD response does not occur or that the vascular response upon neuronal activity does no longer resemble the typical BOLD response. We illustrate this by the following case (Fig. 4.10): A patient presented with a tumor (glioma grade II) in the left postcentral gyrus, extending to the central sulcus and precentral gyrus in the hand knob. For the fMRI examination, the patient was asked to perform a bilateral fingertapping task versus rest. In
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors
65
the right nonpathological hemisphere, activation foci were shown in the pre- and postcentral gyrus (i.e., primary sensorimotor cortex, SM1), the premotor cortex (PMC), and the right posterior parietal cortex (PPC). In the left hemisphere, only activation in the PMC could be demonstrated (Fig. 4.10a). This might be interpreted as absence of neuronal activity within the left SM1 or the PPC and might be explained by the contralateral, nonlesioned right hemisphere taking over the function due to plasticity changes. However, when looking at the MRI signal changes over time, we clearly see a decrease of MRI signal during task performance and an MRI signal increase during the rest condition (Fig. 4.10b). This phenomenon has been coined the “inverse” BOLD response [105]. Different mechanisms have been proposed to explain the occurrence of inverse BOLD responses, but larger patient studies are required to fully explore this phenomenon [106, 107]. This negative or inverted BOLD response should be kept in mind when interpreting presurgical fMRI results, as an inappropri-
ate therapeutic approach may be chosen due to misinterpretation of these fMRI results. Second, the BOLD signal can also be significantly influenced by various pharmacological agents [108]. For example, antihistaminics might reduce the BOLD response [109], while caffeine is a known booster of the BOLD response [110] (Fig. 4.11). Furthermore, cocaine, nicotine, pain medication, and anesthetics have been reported to affect the BOLD response [111–113]. In this view, it is important to know with which medication the patient is being treated at the time of the fMRI examination. Furthermore, not only the biological characteristics and effects of tumors may have a profound effect on the outcome of an fMRI examination, but also the patient himself/herself may influence the final results. One of the main limitations of fMRI is that the patient is actively involved in the examination and that not all patients are able to perform the tasks correctly. In this view, it is paramount to make sure that the patient is well instructed before the start of the scan session and that the patient is
Fig. 4.10 (permission required) Presurgical fMRI results in a patient with a glioma (grade II) in the left postcentral gyrus (upper middle). The activation pattern induced by bilateral fingertapping included activation of the right primary sensorimotor cortex (SM1), the right posterior parietal cortex (PP), and bilateral premotor cortices (PM). No significant activation of the left
SM1 could be demonstrated (lower). When assessing the MRI signal changes over time, it can be demonstrated that instead of an increase of fMRI signal during the fingertapping periods (onperiod in upper left time course), there is a decrease of fMRI signal in the left SM1 (on-period in upper right time course): an “inverted BOLD” signal
66
S. Van Cauter et al.
3 visual cortex (before caffeine) visual cortex (after caffeine) motorcortex (before caffeine) motorcortex (after caffeine)
Percent signal change
2.5 2 1.5 1 0.5 0 -0.5
0
5
10 time (sec)
15
20
Fig. 4.11 (permission required) fMRI activation maps and mean percent signal change of the hemodynamic response function (HRF), following a visual flash and a consequent fingertap,
before and after drinking three cups of coffee, showing activation in the visual and motor cortex and an increase of the HRF after caffeine intake
kept motivated throughout the session. Also, before the scan session, it is important to assess the patient’s cognitive (e.g., attention in a confused patient) and physical (e.g., movement of the limbs in a paralyzed patient) abilities so that the task that is to be performed by the patient can be adapted to the patient’s abilities. It should be mentioned, however, that fMRI paradigms still lack standardization. Therefore, the results obtained with a specific paradigm should be interpreted in view of that paradigm. More research on optimization, standardization, and evaluation of paradigms that are applicable in clinical settings is needed in larger cohorts. Lastly, the patient should be instructed to perform minimal to no head motion during the fMRI examination, as this may complicate the data processing and thus the quality of the fMRI results (see Sect. 4.2.4). Considering these factors, in addition to the general limitations of fMRI, it is clear that it is of utmost importance to realize that the observation of an absence of fMRI activity in a particular brain region does not directly imply
that electrical activity within that area is nonexisting, and thus that it is safe to surgically remove this region.
4.4.3 Challenges and Limitations of Presurgical DTI In clinical settings, the results of DTI may also be influenced by the biochemical environment of the lesion and the patient himself/herself. Furthermore, there is also a lack of standardized DTI processing protocols. First, demyelinating and degenerative diseases can significantly reduce FA focally, which has, for example, been demonstrated in Wallerian degeneration [54], amyotrophic lateral sclerosis [114, 115], and multiple sclerosis [116]. In the case of brain tumors, reduction of FA can be caused by several mechanisms. For instance, a tumoral lesion can destroy white matter fibers, thereby reducing the absolute number of axons
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors
within the fiber. Vasogenic edema spreads intact fibers, reducing the fiber density. Both a reduction of the number of axons and of the fiber density will result in a reduction of FA. Reduction of FA may prevent the fiber tractography algorithm from reconstructing specific white matter tracts that lie in the vicinity of these lesions giving rise to these reductions [117]. In contrast, high FA values have been shown in abscess cavities [118], which are unexpected observations, since we do not expect any tissue organization within an abscess. Therefore, the interpretation of tract reconstructions should always be made with great caution. Second, the positioning of the ROI for fiber tractography is highly user-dependent, even in healthy volunteers. Standardized ROI definition protocols have long been lacking. Recently, Wakana et al. proposed robust ROI definition protocols for the reconstruction of eleven major white matter tracts that may increase reliability and reproducibility of fiber tractography results [119]. However, in the case of distorted anatomy, these protocols may lose their applicability. In this view, it might be interesting to explore the use of tracking seeds that are based on the location of fMRI activations, thereby using the functional information to obtain further structural information [120]. Finally, again, the patient himself/herself might influence the final DTI results. Although a DTI examination does not require active involvement of the patient, the patient should be instructed to perform minimal to no head motion during the DTI examination. As mentioned for fMRI, head motion will both affect the data quality and the quality of the DTI results. In summary, DTI results will be affected by the biochemical environment and the patient himself/herself. Most limitations stem from the fact that in fiber tractography, mathematical algorithms are used for the reconstruction of specific white matter tracts and that these algorithms require specific ROI and algorithm settings to be defined. Therefore, one should always keep in mind that a tract reconstruction may not visualize the entire tract.
4.5 Conclusion Combining fMRI and DTI allows visualization of gray and white matter in a noninvasive way. In this chapter, we reviewed the role of fMRI and DTI in presurgical
67
planning and surgical guidance in patients with brain lesions. Despite some limitations, fMRI and DTI are powerful imaging techniques for optimizing therapeutic guidance in patients with brain pathology. Future research and further standardization will contribute to gain acceptance for the combination of fMRI and DTI as clinically valid and applicable techniques.
References 1. Ogawa S, Lee TM (1990) Magnetic resonance imaging of blood vessels at high fields: in vivo and in vitro measurements and image simulation. Magn Reson Med 16:9–18 2. Ogawa S, Lee TM, Kay AR, Tank DW (1990) Brain magnetic resonance imaging with contrast dependent on blood oxygenation. Proc Natl Acad Sci U S A 87:9868–9872 3. Girouard H, Iadecola C (2006) Neurovascular coupling in the normal brain and in hypertension, stroke, and Alzheimer disease. J Appl Physiol 100:328–335 4. Kida I, Smith AJ, Blumenfeld H, Behar KL, Hyder F (2006) Lamotrigine suppresses neurophysiological responses to somatosensory stimulation in the rodent. Neuroimage 29: 216–224 5. Malonek D, Dirnagl U, Lindauer U, Yamada K, Kanno I, Grinvald A (1997) Vascular imprints of neuronal activity: relationships between the dynamics of cortical blood flow, oxygenation, and volume changes following sensory stimulation. Proc Natl Acad Sci U S A 94:14826–14831 6. Mangia S, Giove F, Tkac I, Logothetis NK, Henry PG, Olman CA, Maraviglia B, Di SF, Ugurbil K (2009) Metabolic and hemodynamic events after changes in neuronal activity: current hypotheses, theoretical predictions and in vivo NMR experimental findings. J Cereb Blood Flow Metab 29:441–463 7. Thulborn KR, Waterton JC, Matthews PM, Radda GK (1982) Oxygenation dependence of the transverse relaxation time of water protons in whole blood at high field. Biochim Biophys Acta 714:265–270 8. Mansfield P, Maudsley AA (1977) Medical imaging by NMR. Br J Radiol 50:188–194 9. Dymarkowski S, Sunaert S, Van OS, van HP, Wilms G, Demaerel P, Nuttin B, Plets C, Marchal G (1998) Functional MRI of the brain: localisation of eloquent cortex in focal brain lesion therapy. Eur Radiol 8:1573–1580 10. Hirsch J, Ruge MI, Kim KHS, Correa DD, Victor JD, Relkin NR, Labar DR, Krol G, Bilsky MH, Souweidane MM, DeAngelis LM, Gutin PH (2000) An integrated functional magnetic resonance imaging procedure for preoperative mapping of cortical areas associated with tactile, motor, language, and visual functions. Neurosurgery 47:711–721 11. Lehericy S, Duffau H, Cornu P, Capelle L, Pidoux B, Carpentier A, Auliac S, Clemenceau S, Sichez JP, Bitar A, Valery CA, Van Effenterre R, Faillot T, Srour A, Fohanno D, Philippon J, Le Bihan D, Marsault C (2000) 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 92:589–598
68 12. Yetkin FZ, Mueller WM, Morris GL, McAuliffe TL, Ulmer JL, Cox RW, Daniels DL, Haughton VM (1997) Functional MR activation correlated with intraoperative cortical mapping. Am J Neuroradiol 18:1311–1315 13. Lee CC, Ward HA, Sharbrough FW, Meyer FB, Marsh WR, Raffel C, So EL, Cascino GD, Shin CS, Xu YC, Riederer SJ, Jack CR (1999) Assessment of functional MR imaging in neurosurgical planning. Am J Neuroradiol 20:1511–1519 14. Stippich C, Rapps N, Dreyhaupt J, Durst A, Kress B, Nennig E, Tronnier VM, Sartor K (2007) Localizing and lateralizing language in patients with brain tumors: feasibility of routine preoperative functional MR imaging in 81 consecutive patients. Radiology 243:828–836 15. De CD, Garreffa G, Colonnese C, Giulietti G, Labruna L, Briselli E, Ken S, Macri MA, Maraviglia B (2007) Identifica tion of activated regions during a language task. Magn Reson Imaging 25:933–938 16. Keller SS, Crow T, Foundas A, Amunts K, Roberts N (2009) Broca’s area: nomenclature, anatomy, typology and asymmetry. Brain Lang 109:29–48 17. Ashtari M, Perrine K, Elbaz R, Syed U, Thaden E, McIlree C, Dolgoff-Kaspar R, Clarke T, Diamond A, Ettinger A (2005) Mapping the functional anatomy of sentence comprehension and application to presurgical evaluation of patients with brain tumor. AJNR Am J Neuroradiol 26: 1461–1468 18. Gaillard WD, Balsamo L, Xu B, McKinney C, Papero PH, Weinstein S, Conry J, Pearl PL, Sachs B, Sato S, Vezina LG, Frattali C, Theodore WH (2004) fMRI language task panel improves determination of language dominance. Neurology 63:1403–1408 19. Galaburda AM, Sanides F, Geschwind N (1978) Human brain. Cytoarchitectonic left-right asymmetries in the temporal speech region. Arch Neurol 35:812–817 20. Mani J, Diehl B, Piao Z, Schuele SS, Lapresto E, Liu P, Nair DR, Dinner DS, Luders HO (2008) Evidence for a basal temporal visual language center: cortical stimulation producing pure alexia. Neurology 71:1621–1627 21. Bowden EM, Jung-Beeman M (2007) Methods for investigating the neural components of insight. Methods 42:87–99 22. Josephs O, Henson RN (1999) Event-related functional magnetic resonance imaging: modelling, inference and optimization. Philos Trans R Soc Lond B Biol Sci 354: 1215–1228 23. Ragland JD, Yoon J, Minzenberg MJ, Carter CS (2007) Neuroimaging of cognitive disability in schizophrenia: search for a pathophysiological mechanism. Int Rev Psychiatry 19: 417–427 24. Duyn JH, Moonen CT, van Yperen GH, de Boer RW, Luyten PR (1994) Inflow versus deoxyhemoglobin effects in BOLD functional MRI using gradient echoes at 1.5 T. NMR Biomed 7:83–88 25. Gao JH, Miller I, Lai S, Xiong TJ, Fox PT (1996) Quantitative assessment of blood inflow effects in functional MRI signals. Magn Reson Med 36:314–319 26. Haacke EM, Hopkins A, Lai S, Buckley P, Friedman L, Meltzer H, Hedera P, Friedland R, Klein S, Thompson L (1994) 2D and 3D high resolution gradient echo functional imaging of the brain: venous contributions to signal in motor cortex studies. NMR Biomed 7:54–62 27. Krings T, Reinges MHT, Erberich S, Kemeny S, Rohde V, Spetzger U, Korinth M, Willmes K, Gilsbach JM, Thron A (2001) Functional MRI for presurgical planning: problems,
S. Van Cauter et al. artefacts, and solution strategies. J Neurol Neurosurg Psychiatry 70:749–760 28. Hill DL, Smith AD, Simmons A, Maurer CR Jr, Cox TC, Elwes R, Brammer M, Hawkes DJ, Polkey CE (2000) Sources of error in comparing functional magnetic resonance imaging and invasive electrophysiological recordings. J Neurosurg 93:214–223 29. Turner R (2002) How much cortex can a vein drain? Downstream dilution of activation-related cerebral blood oxygenation changes. Neuroimage 16:1062–1067 30. Kruger G, Kastrup A, Glover GH (2001) Neuroimaging at 1.5 T and 3.0 T: comparison of oxygenation-sensitive magnetic resonance imaging. Magn Reson Med 45:595–604 31. Okada T, Yamada H, Ito H, Yonekura Y, Sadato N (2005) Magnetic field strength increase yields significantly greater contrast-to-noise ratio increase: Measured using BOLD contrast in the primary visual area. Acad Radiol 12:142–147 32. Tieleman A, Vandemaele P, Seurinck R, Deblaere K, Achten E (2007) Comparison between functional magnetic resonance imaging at 1.5 and 3 Tesla: effect of increased field strength on 4 paradigms used during presurgical work-up. Invest Radiol 42:130–138 33. Meindl T, Born C, Britsch S, Reiser M, Schoenberg S (2008) Functional BOLD MRI: comparison of different field strengths in a motor task. Eur Radiol 18:1102–1113 34. Jezzard P, Balaban RS (1995) Correction for geometric distortion in echo-planar images from B-0 field variations. Magn Reson Med 34:65–73 35. Zeng HR, Constable RT (2002) Image distortion correction in EPI: comparison of field mapping with point spread function mapping. Magn Reson Med 48:137–146 36. Blaimer M, Breuer F, Mueller M, Heidemann RM, Griswold MA, Jakob PM (2004) SMASH, SENSE, PILS, GRAPPA: how to choose the optimal method. Top Magn Reson Imaging 15:223–236 37. Larkman DJ, Nunes RG (2007) Parallel magnetic resonance imaging. Phys Med Biol 52:R15–R55 38. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P (1999) SENSE: sensitivity encoding for fast MRI. Magn Reson Med 42:952–962 39. Weiger M, Pruessmann KP, Osterbauer R, Bornert P, Boesiger P, Jezzard P (2002) Sensitivity-encoded single-shot spiral imaging for reduced susceptibility artifacts in BOLD fMRI. Magn Reson Med 48:860–866 40. Jezzard P, Clare S (1999) Sources of distortion in functional MRI data. Hum Brain Mapp 8:80–85 41. Debus J, Essig M, Schad LR, Wenz F, Baudendistel K, Knopp MV, Engenhart R, Lorenz WJ (1996) Functional magnetic resonance imaging in a stereotactic setup. Magn Reson Imaging 14:1007–1012 42. Friston KJ, Holmes AP, Worsley KJ, Poline JP, Frith CD, Frackowiak RSJ (1995) Statistical Parametric Maps in functional imaging: a general linear approach. Hum Brain Mapp 2:189–210 43. Friston KJ, Williams S, Howard R, Frackowiak RSJ, Turner R (1996) Movement-related effects in fMRI time-series. Magn Reson Med 35:346–355 44. Turner R, Howseman A, Rees GE, Josephs O, Friston K (1998) Functional magnetic resonance imaging of the human brain: data acquisition and analysis. Exp Brain Res 123:5–12 45. Basser PJ (1995) Inferring microstructural features and the physiological state of tissues from diffusion-weighted images. NMR Biomed 8:333–344
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors 46. Basser PJ, Pierpaoli C (1996) Microstructural and physiological features of tissues elucidated by quantitative-diffusion-tensor MRI. J Magn Reson B 111:209–219 47. Pierpaoli C, Jezzard P, Basser PJ, Barnett A, DiChiro G (1996) Diffusion tensor MR imaging of the human brain. Radiology 201:637–648 48. Mori S, Crain BJ, Chacko VP, van Zijl PCM (1999) Threedimensional tracking of axonal projections in the brain by magnetic resonance imaging. Ann Neurol 45:265–269 49. Wakana S, Jiang HY, Nagae-Poetscher LM, van Zijl PCM, Mori S (2004) Fiber tract-based atlas of human white matter anatomy. Radiology 230:77–87 50. Beaulieu C (2002) The basis of anisotropic water diffusion in the nervous system – a technical review. NMR Biomed 15:435–455 51. Hajnal JV, Doran M, Hall AS, Collins AG, Oatridge A, Pennock JM, Young IR, Bydder GM (1991) MR imaging of anisotropically restricted diffusion of water in the nervous system: technical, anatomic, and pathologic considerations. J Comput Assist Tomogr 15:1–18 52. Basser PJ, Mattiello J, Lebihan D (1994) Estimation of the effective self-diffusion tensor from the NMR spin echo. J Magn Reson B 103:247–254 53. Pierpaoli C, Basser PJ (1996) Toward a quantitative assessment of diffusion anisotropy. Magn Reson Med 36:893–906 54. Pierpaoli C, Barnett A, Pajevic S, Chen R, Penix L, Virta A, Basser P (2001) Water diffusion changes in Wallerian degeneration and their dependence on white matter architecture. Neuroimage 13:1174–1185 55. Thomalla G, Glauche V, Koch MA, Beaulieu C, Weiller C, Rother J (2004) Diffusion tensor imaging detects early Wallerian degeneration of the pyramidal tract after ischemic stroke. Neuroimage 22:1767–1774 56. Pajevic S, Pierpaoli C (1999) Color schemes to represent the orientation of anisotropic tissues from diffusion tensor data: application to white matter fiber tract mapping in the human brain. Magn Reson Med 42:526–540 57. Mori S, van Zijl PC (2002) Fiber tracking: principles and strategies – a technical review. NMR Biomed 15:468–480 58. Truong TK, Chen B, Song AW (2008) Integrated SENSE DTI with correction of susceptibility- and eddy currentinduced geometric distortions. Neuroimage 40:53–58 59. Fesl G, Moriggl B, Schmid UD, Naidich TP, Herholz K, Yousry TA (2003) Inferior central sulcus: variations of anatomy and function on the example of the motor tongue area. Neuroimage 20:601–610 60. Naidich TP, Blum JT, Firestone MI (2001) The parasagittal line: An anatomic landmark for axial imaging. Am J Neuroradiol 22:885–895 61. Naidich TP, Hof PR, Yousry TA, Yousry I (2001) The motor cortex: anatomic substrates of function. Neuroimaging Clin N Am 11:171–193 62. Naidich TP, Kang E, Fatterpekar GM, Delman BN, Gultekin SH, Wolfe D, Ortiz O, Yousry I, Weismann M, Yousry TA (2004) The insula: Anatomic study and MR Imaging display at 1.5 T. Am J Neuroradiol 25:222–232 63. Naidich TP, Valavanis AG, Kubik S (1995) Anatomic relationships along the low-middle convexity: part I–normal specimens and magnetic resonance imaging. Neurosurgery 36:517–532 64. Schmahmann JD, Smith EE, Eichler FS, Filley CM (2008) Cerebral white matter: neuroanatomy, clinical neurology, and neurobehavioral correlates. Ann N Y Acad Sci 1142:266–309
69
65. Yetkin FZ, Ulmer JL, Mueller W, Cox RW, Klosek MM, Haughton VM (1998) Functional magnetic resonance imaging assessment of the risk of postoperative hemiparesis after excision of cerebral tumors. Int J Neuroradiol 4:253–257 66. Nimsky C, Ganslandt O, Merhof D, Sorensen AG, Fahlbusch R (2006) Intraoperative visualization of the pyramidal tract by diffusion-tensor-imaging-based fiber tracking. Neuroimage 30:1219–1229 67. Bello L, Gambini A, Castellano A, Carrabba G, Acerbi F, Fava E, Giussani C, Cadioli M, Blasi V, Casarotti A, Papagno C, Gupta AK, Gaini S, Scotti G, Falini A (2008) Motor and language DTI Fiber Tracking combined with intraoperative subcortical mapping for surgical removal of gliomas. Neuroimage 39:369–382 68. Stadlbauer A, Nimsky C, Buslei R, Salomonowitz E, Hammen T, Buchfelder M, Moser E, Ernst-Stecken A, Ganslandt O (2007) Diffusion tensor imaging and optimized fiber tracking in glioma patients: histopathologic evaluation of tumorinvaded white matter structures. Neuroimage 34:949–956 69. Wiebe S, Blume WT, Girvin JP, Eliasziw M (2001) A randomized, controlled trial of surgery for temporal-lobe epilepsy. N Engl J Med 345:311–318 70. Abou-Khalil B (2007) Methods for determination of language dominance: the Wada test and proposed noninvasive alternatives. Curr Neurol Neurosci Rep 7:483–490 71. Pelletier I, Sauerwein HC, Lepore F, Saint-Amour D, Lassonde M (2007) Non-invasive alternatives to the Wada test in the presurgical evaluation of language and memory functions in epilepsy patients. Epileptic Disord 9:111–126 72. Binder JR, Swanson SJ, Hammeke TA, Morris GL, Mueller WM, Fischer M, Benbadis S, Frost JA, Rao SM, Haughton VM (1996) Determination of language dominance using functional MRI: a comparison with the Wada test. Neurology 46:978–984 73. Bello L, Gallucci M, Fava M, Carrabba G, Giussani C, Acerbi F, Baratta P, Songa V, Conte V, Branca V, Stocchetti N, Papagno C, Gaini SM (2007) Intraoperative subcortical language tract mapping guides surgical removal of gliomas involving speech areas. Neurosurgery 60:67–80 74. Keles GE, Lundin DA, Lamborn KR, Chang EF, Ojemann G, Berger MS (2004) Intraoperative subcortical stimulation mapping for hemispherical perirolandic gliomas located within or adjacent to the descending motor pathways: evaluation of morbidity and assessment of functional outcome in 294 patients. J Neurosurg 100:369–375 75. Taylor MD, Bernstein M (1999) Awake craniotomy with brain mapping as the routine surgical approach to treating patients with supratentorial intraaxial tumors: a prospective trial of 200 cases. J Neurosurg 90:35–41 76. Signorelli F, Guyotat J, Schneider F, Isnard J, Bret P (2003) Technical refinements for validating functional MRI-based neuronavigation data by electrical stimulation during cortical language mapping. Minim Invasive Neurosurg 46:265–268 77. Bizzi A, Blasi V, Falini A, Ferroli P, Cadioli M, Danesi U, Aquino D, Marras C, Caldiroli D, Broggi G (2008) Presurgical functional MR imaging of language and motor functions: validation with intraoperative electrocortical mapping. Radiology 248:579–589 78. Fandino J, Kollias SS, Wieser HG, Valavanis A, Yonekawa Y (1999) Intraoperative validation of functional magnetic resonance imaging and cortical reorganization patterns in patients with brain tumors involving the primary motor cortex. J Neurosurg 91:238–250
70 79. Majos A, Tybor K, Stefanczyk L, Goraj B (2005) Cortical mapping by functional magnetic resonance imaging in patients with brain tumors. Eur Radiol 15:1148–1158 80. Roux FE, Boulanouar K, Lotterie JA, Mejdoubi M, Lesage JP, Berry I (2003) Language functional magnetic resonance imaging in preoperative assessment of language areas: correlation with direct cortical stimulation. Neurosurgery 52: 1335–1345 81. Xie J, Chen XZ, Jiang T, Li SW, Li ZX, Zhang Z, Dai JP, Wang ZC (2008) Preoperative blood oxygen level-dependent functional magnetic resonance imaging in patients with gliomas involving the motor cortical areas. Chin Med J (Engl ) 121:631–635 82. Kamada K, Todo T, Masutani Y, Aoki S, Ino K, Takano T, Kirino T, Kawahara N, Morita A (2005) Combined use of tractography-integrated functional neuronavigation and direct fiber stimulation. J Neurosurg 102:664–672 83. Freund HJ (1987) Differential effects of cortical lesions in humans. Ciba Found Symp 132:269–281 84. Apuzzo ML, Chandrasoma PT, Cohen D, Zee CS, Zelman V (1987) Computed imaging stereotaxy: experience and perspective related to 500 procedures applied to brain masses. Neurosurgery 20:930–937 85. Cosgrove GR, Buchbinder BR, Jiang H (1996) Functional magnetic resonance imaging for intracranial navigation. Neurosurg Clin N Am 7:313–322 86. Nimsky C, Ganslandt O, Fahlbusch R (2007) Implementation of fiber tract navigation. Neurosurgery 61:306–317 87. Schneider JP, Schulz T, Schmidt F, Dietrich J, Lieberenz S, Trantakis C, Seifert V, Kellermann S, Schober R, Schaff ranietz L, Laufer M, Kahn T (2001) Gross-total surgery of supratentorial low-grade gliomas under intraoperative MR guidance. Am J Neuroradiol 22:89–98 88. Schulder M, Maldjian JA, Liu WC, Holodny AI, Kalnin AT, Mun IK, Carmel PW (1998) Functional image-guided surgery of intracranial tumors located in or near the sensorimotor cortex. J Neurosurg 89:412–418 89. Wu JS, Zhou LF, Tang WJ, Mao Y, Hu J, Song YY, Hong XN, Du GH (2007) Clinical evaluation and follow-up outcome of diffusion tensor imaging-based functional neuronavigation: a prospective, controlled study in patients with gliomas involving pyramidal tracts. Neurosurgery 61:935–948 90. Hall WA, Truwit CL (2008) Intraoperative MR-guided neurosurgery. J Magn Reson Imaging 27:368–375 91. Roberts DW, Hartov A, Kennedy FE, Miga MI, Paulsen KD (1998) Intraoperative brain shift and deformation: a quantitative analysis of cortical displacement in 28 cases. Neurosurgery 43:749–758 92. Hastreiter P, Rezk-Salama C, Soza G, Bauer M, Greiner G, Fahlbusch R, Ganslandt O, Nimsky C (2004) Strategies for brain shift evaluation. Med Image Anal 8:447–464 93. Foroglou N, Zamani A, Black P (2009) Intra-operative MRI (iop-MR) for brain tumour surgery. Br J Neurosurg 23:14–22 94. Jolesz FA (2005) Future perspectives for intraoperative MRI. Neurosurg Clin N Am 16:201–213 95. Coenen VA, Krings T, Weidemann J, Hans FJ, Reinacher P, Gilsbach JM, Rohde V (2005) Sequential visualization of brain and fiber tract deformation during intracranial surgery
S. Van Cauter et al. with three-dimensional ultrasound: an approach to evaluate the effect of brain shift. Neurosurgery 56:133–141 96. Rasmussen IA Jr, Lindseth F, Rygh OM, Berntsen EM, Selbekk T, Xu J, Nagelhus Hernes TA, Harg E, Haberg A, Unsgaard G (2007) Functional neuronavigation combined with intra-operative 3D ultrasound: initial experiences during surgical resections close to eloquent brain areas and future directions in automatic brain shift compensation of preoperative data. Acta Neurochir (Wien ) 149:365–378 97. Unsgaard G, Rygh OM, Selbekk T, Muller TB, Kolstad F, Lindseth F, Hernes TA (2006) Intra-operative 3D ultrasound in neurosurgery. Acta Neurochir (Wien ) 148:235–253 98. Holodny AI, Schulder M, Liu WC, Wolko J, Maldjian JA, Kalnin AJ (2000) The effect of brain tumors on BOLD functional MR imaging activation in the adjacent motor cortex: implications for image-guided neurosurgery. Am J Neuroradiol 21:1415–1422 99. Schreiber A, Hubbe U, Ziyeh S, Hennig J (2000) The influence of gliomas and nonglial space-occupying lesions on blood-oxygen-level-dependent contrast enhancement. Am J Neuroradiol 21:1055–1063 100. Lehericy S, Biondi A, Sourour N, Vlaicu M, du Montcel ST, Cohen L, Vivas E, Capelle L, Faillot T, Casasco A, Le BD, Marsault C (2002) Arteriovenous brain malformations: is functional MR imaging reliable for studying language reorganization in patients? Initial observations. Radiology 223: 672–682 101. Young WL, Prohovnik I, Ornstein E, Ostapkovich N, Sisti MB, Solomon RA, Stein BM (1990) The effect of arteriovenous malformation resection on cerebrovascular reactivity to carbon dioxide. Neurosurgery 27:257–266 102. Hossmann KA, Linn F, Okada Y (1992) Bioluminescence and fluoroscopic imaging of tissue pH and metabolites in experimental brain tumors of cat. NMR Biomed 5: 259–264 103. Linn F, Seo K, Hossmann KA (1989) Experimental transplantation gliomas in the adult cat brain. 3. Regional biochemistry. Acta Neurochir (Wien) 99:85–93 104. Whittle IR, Collins F, Kelly PAT, Ritchie I, Ironside JW (1996) Nitric oxide synthase is expressed in experimental malignant glioma and influences tumour blood flow. Acta Neurochirurgica 138:870–875 105. Bryan RN, Kraut M (1998) Functional magnetic resonance imaging: you get what you (barely) see. Am J Neuroradiol 19:991–992 106. Harel N, Lee SP, Nagaoka T, Kim DS, Kim SG (2002) Origin of negative blood oxygenation level-dependent fMRI signals. J Cereb Blood Flow Metab 22:908–917 107. Sunaert S, Dymarkowski S, Van OS, van HP, Wilms G, Marchal G (1998) Functional magnetic resonance imaging (fMRI) visualises the brain at work. Acta Neurol Belg 98:8–16 108. Iannetti GD, Wise RG (2007) BOLD functional MRI in disease and pharmacological studies: room for improvement? Magn Reson Imaging 25:978–988 109. Noseworthy MD, Alfonsi J, Bells S (2003) Attenuation of brain BOLD response following lipid ingestion. Hum Brain Mapp 20:116–121 110. Laurienti PJ, Field AS, Burdette JH, Maldjian JA, Yen YF, Moody DM (2002) Dietary caffeine consumption modulates fMRI measures. Neuroimage 17:751–757
4 The Clinical Applicability of fMRI and DTI in Patients with Brain Tumors 111. Lowen SB, Nickerson LD, Levin JM (2009) Differential effects of acute cocaine and placebo administration on visual cortical activation in healthy subjects measured using BOLD fMRI. Pharmacol Biochem Behav 92:277–282 112. Pattinson KT, Rogers R, Mayhew SD, Tracey I, Wise RG (2007) Pharmacological FMRI: measuring opioid effects on the BOLD response to hypercapnia. J Cereb Blood Flow Metab 27:414–423 113. Smolka MN, Buhler M, Klein S, Zimmermann U, Mann K, Heinz A, Braus DF (2006) Severity of nicotine dependence modulates cue-induced brain activity in regions involved in motor preparation and imagery. Psychopharmacology (Berl) 184:577–588 114. Sage CA, Peeters RR, Gorner A, Robberecht W, Sunaert S (2007) Quantitative diffusion tensor imaging in amyotrophic lateral sclerosis. Neuroimage 34:486–499 115. Sage CA, Peeters RR, Gorner A, Robberecht W, Sunaert S (2009) Quantitative diffusion tensor imaging in amyotrophic lateral sclerosis: revisited. Hum Brain Mapp 30:3657–3675 116. Reich DS, Zackowski KM, Gordon-Lipkin EM, Smith SA, Chodkowski BA, Cutter GR, Calabresi PA (2008) Corticospinal
71
tract abnormalities are associated with weakness in multiple sclerosis. AJNR Am J Neuroradiol 29:333–339 117. Sinha S, Bastin ME, Whittle IR, Wardlaw JM (2002) Diffusion tensor MR imaging of high-grade cerebral gliomas. Am J Neuroradiol 23:520–527 118. Gupta RK, Hasan KM, Mishra AM, Jha D, Husain M, Prasad KN, Narayana PA (2005) High fractional anisotropy in brain abscesses versus other cystic intracranial lesions. Am J Neuroradiol 26:1107–1114 119. Wakana S, Caprihan A, Panzenboeck MM, Fallon JH, Perry M, Gollub RL, Hua K, Zhang J, Jiang H, Dubey P, Blitz A, van ZP, Mori S (2007) Reproducibility of quantitative tractography methods applied to cerebral white matter. Neuroimage 36:630–644 120. Staempfli P, Reischauer C, Jaermann T, Valavanis A, Kollias S, Boesiger P (2008) Combining fMRI and DTI: a framework for exploring the limits of fMRI-guided DTI fiber tracking and for verifying DTI-based fiber tractography results. Neuroimage 39:119–126 121. Ekman P, Friesen WV (1976) Pictures of facial affect. Consulting Psychologists Press, Palo Alto
5
Low-Grade Gliomas G.A. Christoforidis, A. Drevelegas, E.C. Bourekas, and G. Karkavelas
Contents 5.1 Introduction............................................................. 73 5.2 Physiologic Imaging................................................ 73 5.3 Astrocytomas........................................................... 78 5.3.1 Well-Differentiated, Diffuse, Infiltrative Astrocytoma.............................................................. 78 5.3.2 Noninfiltrative, Diffuse, or Circumscribed Astrocytomas............................................................ 80 5.4 Oligodendroglioma................................................. 98 5.4.1 Pathology.................................................................. 98 5.4.2 Imaging..................................................................... 104 5.5 Choroid Plexus Papilloma...................................... 111 5.5.1 Pathology.................................................................. 111 5.5.2 Imaging..................................................................... 112 5.6 Ependymoma.......................................................... 113 5.7 Dysembryoplastic Neuroepithelial Tumor............ 125 5.7.1 Imaging..................................................................... 125
G.A. Christoforidis Dept. of Radiology University of Chicago-Medical Center Chicago IL USA, e-mail:
[email protected] A. Drevelegas Dept. of Radiology AHEPA University Hospital, Aristotles University of Thessaloniki Greece e-mail:
[email protected] E.C. Bourekas Dept. of Radiology Ohio State University Medical Center Columbus Ohio G. Karkavelas Dept. of Pathology Aristotles University of Thessaloniki Thessaloniki-Greece
5.8 Subependymoma..................................................... 127 5.8.1 Pathology.................................................................. 128 5.8.2 Imaging..................................................................... 128 5.9 Ganglion Cell Tumors............................................ 129 5.9.1 Gangliogliomas......................................................... 129 5.9.2 Gangliocytoma.......................................................... 135 5.9.3 Dysplastic Cerebellar Gangliocytoma...................... 138 5.9.4 Desmoplastic Infantile Astrocytoma and Ganglioglioma.......................................................... 141 5.10 Neurocytoma........................................................... 144 5.10.1 Imaging..................................................................... 144 References............................................................................ 148
5.1 Introduction Low-grade primary brain tumors include those tumors arising from the brain parenchyma, such as astrocytomas, oligodendrogliomas, ependymomas, subependymomas, choroid plexus papillomas (CPPs), ganglion cell tumors, and neurocytomas. Distinction of histopathologic varieties of primary brain tumors on imaging is based on tumor location and imaging features such as presence of a cyst, contrast enhancement pattern, signal intensity, and presence of calcifications. In this chapter, new imaging modalities which attempt to distinguish high from low-grade tumors are briefly reviewed and the imaging and pathologic characteristics of the low-grade tumors listed above are described.
5.2 Physiologic Imaging Many imaging modalities have been employed in recent years in an attempt to noninvasively differentiate high and low-grade gliomas. Hydrogen proton
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI 10.1007/978-3-540-87650-2_5, © Springer-Verlag Berlin Heidelberg 2011
73
74
magnetic resonance spectroscopy (MRS), positron emission tomography (PET) imaging, cerebral blood volume (CBV) mapping and single-photon emission computed tomography (SPECT) are all modalities which have been employed towards this end. Contrast enhancement is often associated with more aggressive tumors; however, this is not always the case [1]. MRS can be used to quantify various metabolites within a sample of tumor tissue. These metabolites include the neuronal marker N-acetylaspartate (NAA);
G.A. Christoforidis et al.
choline, a cell membrane component; lactate, for glycolysis and necrosis; and creatine, a marker for energy metabolism. Hydrogen MRS of brain tumors has been studied for over a decade now [2]. In general, primary brain tumors demonstrate reduced levels of NAA and increased levels of choline relative to normal brain tissue (Fig. 5.1). Elevated choline levels are thought to represent areas of increased tumor cellularity and proliferative activity [3–7], whereas decreased NAA is thought to represent decreased neuronal density and
b a
d
c
Fig. 5.1 Magnetic resonance spectroscopy (MRS) obtained from normal brain (a, d), a low-grade tumor (b, e), and a highgrade tumor (c, f). N-acetylaspartate (NAA) peaks are decreased in areas of neoplasia relative to normal brain. Note the elevated lipid lactate (LL) peak in the high-grade brain tumor (c). The
choline (Ch) peak is elevated in the low-grade tumor relative to creatine (Cr). The voxel obtained for the high-grade tumor predominantly includes necrotic tissue; as a result NAA, choline, and creatine are all significantly reduced. (Courtesy of Wayne Slone, The Ohio State University Medical Center)
75
5 Low-Grade Gliomas
e
f
Fig. 5.1 (continued)
viability found in gliomas [4, 5]. Elevated levels of lactate have been identified in higher-grade tumors, whereas no elevation of the lactate peak has been associated with low-grade tumors [8, 9]. There is however confounding evidence to support that lactate may not be as accurate an indicator of malignancy [6, 10]. Intravenous administration of contrast reagents enhances the conspicuity of brain tumors on imaging. Furthermore, patterns of enhancement can be used to characterize brain tumors. The resulting contrast enhancement has been shown to depend on the degree of disruption of the blood–brain barrier as well as tumor microvascularity. As field strength increases both T1 and T1 shortening increase. As a result the degree of contrast enhancement increases as well. Tumor to brain contrast after gadolinium administration at 3 T imaging compared to 1.5 T has been shown to be significantly higher [11]. Contrast to noise (CNR) has been shown to be 2.8 times higher at 3 T than at 1.5 T at the same contrast dosing and 1.3 times when using half the contrast dose at 3.0 T and full dose at 1.5 T. Extent of
contrast enhancement increases with contrast dose [12]. Time course of MR imaging following contrast administration affects signal intensity. Peak signal intensity occurs at 25–35 min; however, no significant change occurs after 5 min. It is generally accepted that contrast-enhanced MR imaging of brain tumors should not begin until 2–5 min after contrast administration [13]. Comparison of gadolinium dimeglumine vs. gadopentetate dimeglumine contrast reagents in a multisite study has shown that gadopentetate was superior in both qualitative and quantitative assessments of brain to tumor contrast [14]. Contrast enhancement has been shown to occur in 20% of low-grade gliomas, whereas lack of contrast enhancement gives a two in three chance of low tumor grade [15, 16]. Dynamic MR imaging studies attempt to characterize tumor vascular dynamics by imaging differences in the signal intensity of a tumor as contrast reagents course through the vascular bed using rapidly acquired T1-weighted, T2-weighted, and T2*-weighted techniques. Contrast leakage across a vascular bed is
76
influenced by vascular permeability, contrast binding within the vascular bed, hematocrit, and contrast reagent concentration within the blood. Choice of postprocessing methods influences signal to noise ratio and overall precision and accuracy obtained with dynamic contrast imaging [17]. This information can then be used to study the CBV, blood flow, blood–brain barrier permeability, and water exchange kinetics of the neoplasm relative to normal brain tissue. Relative cerebral blood volume (rCBV) mapping uses MRI perfusion imaging techniques in order to construct a map of the relative perfusion of blood within the brain and is considered a surrogate marker for tumoral microvascularity. Animal studies have indicated that rCBV corresponds to vascular concentration of contrast agent [18, 19]. Areas of low rCBV within a tumor, compared to normal white matter or gray matter, have been suggested to correlate with lower tumor grade, whereas areas of higher perfusion have been correlated to higher-grade tumors [18, 19] (Fig. 5.2). As a result, surgical biopsies may be directed toward more aggressive areas of the tumor using this information. rCBV maps have also been shown to correlate with vascular endothelial growth factor expression in nonenhancing gliomas and tumor grade [20]. Finally, rCBV measurements have been shown to assist in identifying low-grade gliomas that are either high-grade gliomas misdiagnosed as a result of sampling error, or low-grade gliomas undergoing transformation to a higher grade [21]. Dynamic contrast enhancement methods (DCE-MRI) use T1 and T2* techniques to assess vascular permeability within a tumor bed and tumor grade [22]. Vessel permeability within the vascular bed of an astrocytoma has been suggested to be associated with the presence of vascular endothelial growth factor and angiogenesis [23]. Importantly, corticosteroid administration intended to reduce peritumoral edema has been shown to influence the extent of contrast enhancement also; however, it does not affect tumoral blood flow or rCBV [23]. Diffusion-weighted imaging (DWI) techniques provide additional insight when characterizing tumor behavior. Isotropic DWI characterizes diffusion in three directions. Calculated diffusion values are displayed on apparent diffusion coefficient (ADC) maps. Diffusion tensor imaging is able to quantify water diffusion in multiple directions. This multiplicity of direction allows for postprocessed formation of images displaying the relationship between intraaxial tumors and adjacent white matter fiber tracts and tumor white matter infiltration patterns. Preoperative identification
G.A. Christoforidis et al.
of the anatomic relationship between the neoplastic tissue and eloquent cortical and white matter regions may assist the surgeon when planning the extent to which a brain tumor can be removed and preserve functionality. It has been hypothesized that identification of these patterns may be able to differentiate tumor types [24]. The DWI appearance of a tumor has been suggested to be influenced by tumor cellularity and total nuclear area and thus has the potential to assist in the characterization and grading of brain tumors. Lowgrade tumors are moderately cellular with loose intercellular connections and widened extracellular spaces providing more room for the diffusion of water protons, resulting in higher ADC values. Although in theory, ADC values may assist the differentiation of high- from low-grade astrocytomas or help distinguish nongliomatous tumors from metastases or lymphoma, this has not always been shown to be a consistent reliability. Some authors suggest that ADC maps can help determine the extent of tumor invasion; however, this remains controversial. ADC maps can be used to identify early brain tumor response to therapy and differentiate radiation necrosis from tumor recurrence [25]. Fractional anisotropy (FA) is a measure of water diffusion favoring one direction over others. Because FA values differ depending on location, they are compared to similar regions of interest in the normal appearing brain in the contralateral hemisphere. Unlike ADC values, FA values have been shown to differ in peritumoral signal abnormality associated with infiltrating tumors such as gliomas when compared to noninfiltrating tumors such as meningiomas [26]. Although FA values have been found to be low within the tumor beds of both high and low-grade gliomas, interrogation of the central portion of the tumor compared with the peripheral portion of the tumor appear to differ slightly in low but not high-grade gliomas [27]. This suggests greater disorganization within the peripheral portion of tumor bed of high-grade gliomas than low-grade gliomas. It is critical to point out that measurements obtained via DWI methods could potentially be influenced by therapeutic effects such as radiation, chemotherapy, and steroid use. PET imaging has also been used to separate high from low-grade tumors by distinguishing hypometabolic from hypermetabolic tumors. Metabolic activity within brain tumors using PET has been studied extensively with a large number of markers. Applications include measurement of blood flow, blood volume, oxygen use,
77
5 Low-Grade Gliomas
a
c
Fig. 5.2 Axial postgadolinium T1 MRI (a), and relative cerebral blood volume (rCBV) MRI map (b) of a low-grade oligodendroglioma are compared with axial T2 MRI (c) and rCBV MRI map (d) of a high-grade oligodendroglioma. Relative blood volume is demonstrated here on the basis of a colorized scale in
b
d
which red indicates areas of higher blood volume and blue areas of low blood volume. These images demonstrate increased blood volume in the high-grade tumor relative to normal white matter and lower uptake in the low-grade tumor. (Courtesy of Dr. S.S. Kollias, University of Zürich, Switzerland)
78
glucose use, glucose transport, amino acid uptake, protein synthesis, blood–brain barrier integrity, cerebral pH, membrane metabolism, and nucleic acid synthesis [28]. Tumor grade has been studied by measuring glucose uptake using 18F-flurodeoxyglucose (FDG) and amino acid uptake by using 11C-methylmethionine (MET) or 3¢-Deoxy-3¢-18F-fluorothymidine (FLT). Uptake of FDG in LGG is typically lower than in the surrounding brain, which can render tumor areas photopenic or isometabolic in PET scans, whereas uptake of FLT, MET, or other radiolabeled amino acids are transported specifically into tumors without significant uptake by normal brain [29, 30]. For glioma grading, FDG and MET are both good predictors of grade and patient survival. In histologically confirmed LGG, MET uptake can differentiate long-term and short-term survivors and can therefore be helpful in selecting a therapeutic strategy [31, 32]. More recently, PET imaging with VEGFR has been shown to be a promising agent, which may be able to demonstrate angiogenesis in a manner which differs from relative CBV mapping which correlates with microvascularity [33]. SPECT imaging of brain tumors has been studied predominantly with thalium-201 (201-Tl) and Technetium-99m hexamethyl propyleneamine oxime (99mTc-HMPAO). Because thallium behaves biologically like potassium, imaging of 201-thalium can act as a marker for the function of the sodium–potassium pump. Thalium can thus act as a measure of metabolic activity. The ratio of 201-thallium uptake in the tumor in question relative to normal brain has been used to predict which tumors are low grade. Furthermore, thalium can be used to differentiate tumor necrosis from recurrent tumor [34, 35]. 99mTc-HMPAO can act as a measure of tissue perfusion. Much like perfusion MRI, tumor recurrence shows an increased uptake of 99mTcHMPAO and can help increase the accuracy of 201-Tl-SPECT differentiation of recurrent tumor from radiation necrosis.
G.A. Christoforidis et al.
and noninfiltrative types (25%). Noninfiltrative astrocytomas include pilocytic astrocytomas, pleomorphic xanthoastrocytomas, subependymal giant cell astrocytomas (SGCAs), and desmoplastic cerebral astrocytoma of infancy [38–41]. Infiltrative astrocytomas include anaplastic and well-differentiated types [36, 37]. The differentiation of low-grade from high-grade astrocytomas relies on certain histopathologic features. These include cell density, nuclear atypia, mitotic activity, necrosis, and vascular proliferation. According to the Mayo-St. Anne’s system of astrocytoma grading, the presence of any mitotic figures, necrosis, or vascular proliferation places a tumor in the high-grade category [42]. Low-grade astrocytomas are differentiated from reactive astrocytes or normal brain tissue on the basis of cellularity and size of abnormal astrocytes [36, 37, 42]. Furthermore, a diverse pattern of genomic lesions can lead to the transformation of low-grade astrocytomas into higher tumor grades [43].
5.3.1 Well-Differentiated, Diffuse, Infiltrative Astrocytoma Well-differentiated diffuse infiltrative astrocytomas represent WHO classification grade II astrocytomas [36, 37] and comprise approximately 25% of all gliomas. They most frequently arise from the supratentorial brain and the brain stem. Brain stem involvement is more frequently seen with children. When they occur within the brainstem they are frequently referred to as brainstem gliomas (BSG). Mean age at presentation is 34 years with a slight male predilection. The tumors infiltrate adjacent and distant brain irrespective of histologic grade and tend to progress to a more malignant phenotype. Seizures are a common clinical symptom. The median survival rate is between 7 and 10 years.
5.3.1.1 Pathology
5.3 Astrocytomas Astrocytomas comprise those primary brain tumors that arise from astrocytes. A variety of tumor types are included in this group and they affect different locations, age-groups, and gender distributions [36, 37]. Astrocytomas can be divided into infiltrative types (75%)
Macroscopic examination may typically identify an unencapsulated ill-defined tumor with a firm rubbery consistency, which expands the involved brain (Fig. 5.3) and sometimes contains calcifications. Occasionally, intratumoral cysts filled with clear fluid are recognizable, although cyst formation is more common in pilocytic astrocytomas.
79
5 Low-Grade Gliomas
Markers for proliferative activity such as MIB-1 index have been found to be useful in predicting which lowgrade diffuse astrocytomas will transform into those of higher grades [37, 44]. Most astrocytomas show immunoreactivity with antibody to glial fibrillary acid protein (GFAP).
5.3.1.2 Imaging
Fig. 5.3 Diffuse astrocytoma. Gross specimen shows an expansive neoplasm of the medulla
Fig. 5.4 Diffuse astrocytoma characterized by mild cellularity of well-differentiated astrocytes and profound microcystic formation. Hematoxylin–eosin, original magnification ×400
On microscopic examination, astrocytomas are characterized by their modest or more profound hypercellularity and indistinct tumor borders. They are also characterized by a high degree of cellular differentiation located in an environment of neuroglial fibrils and are often accompanied by degenerative microscopic cysts. Microcyst formation is a feature of low-grade gliomas and helps in differential diagnosis of nonneoplastic reactive lesions (Fig. 5.4). The most frequently encountered subtype is the fibrillary astrocytoma. Other subtypes include the protoplasmic and the gemistocytic astrocytoma. Despite their low grade, these tumors may have a poor prognosis because of a tendency to dedifferentiate into higher grades with age.
On imaging these tumors most often appear homogeneous and infiltrating. Focal, circumscribed astrocytomas can also occur. They can be difficult to detect on CT, may manifest with only a slight density difference, and usually do not enhance (Fig. 5.5). MRI is more sensitive than CT in detecting these tumors [2, 45–47]. It has been well-documented that they may extend well beyond the margins of the tumor as identified on MRI [45–50]. Indeed, a threshold cellular density has been calculated in determining the detectability of these tumors on MRI [51]. They display a low signal on T1WI and a higher signal on T2WI with little or no enhancement following gadolinium administration (Fig. 5.6); however, because contrast enhancement has been reported in up to 40% of cases, it is not thought to be a reliable marker for high histopathologic grade [44, 48, 52]. CT reveals tumoral calcifications in 20% of cases whereas none are identified on MRI [44, 53]. Well-differentiated diffuse astrocytomas may spread to gray matter and may have cystic foci on imaging, but typically lack peritumoral edema and have nearly no mass effect [47, 48]. Most lowgrade astrocytomas transform into higher-grade tumors. This is more probable in older than in younger patients. When the tumor infiltrates the cortex, it may be confused with infarction or cortical-based tumors such as oligodendroglioma or ganglion cell tumor. Imaging has been shown to have the potential to predict progression of WHO grade II astrocytomas to higher grades or poor outcome. Larger tumor volumes and involvement of more than one lobe as depicted on T2 MRI have been shown to be associated with worse outcomes and transformation to higher grades [54]. In a 35 patient study, relative CBV maps were shown to predict progression as an adjunct to histopathologic findings [55]. Low-grade astrocytomas show no or minimally elevated tumor rCBV values when compared with the contralateral uninvolved brain and have significantly lower rCBV values than high-grade
80
a
G.A. Christoforidis et al.
b
Fig. 5.5 (a, b) Brain stem astrocytoma. (a) Noncontrast CT shows a hypodense lesion expanding the pons and compressing the IV ventricle. (b) On postcontrast CT the lesion does not enhance
gliomas. Higher relative CBVs are thought to be associated with microvascularity and predictive of angiogenesis. Inconsistencies between histopathology and dynamic susceptibility contrast-enhanced MRI may be attributable to misdiagnosis or sampling error on biopsy. In a 25 patient study, fibrillary WHO grade II astrocytomas associated with higher regional CBVs were shown to be associated with earlier recurrence following fractionated stereotactic radiotherapy relative to those with lower regional CBVs [56]. Histologic studies demonstrate that astrocytomas can infiltrate well past the margins of enhancement [57]. Vessel permeability within the vascular bed of an astrocytoma has been suggested to relate to the presence of vascular endothelial growth factor and angiogenesis [58]. The extent and pattern of enhancement can vary depending on the type of contrast agent used, the used of corticosteroid, pulse sequence used, dosage of contrast reagent used, and field strength [15, 16, 20]. Sudden change in contrast enhancement may be an
indication of conversion to higher grade (Fig. 5.7) or radiation necrosis. MR spectroscopy of low-grade astrocytomas displays a wide range of spectra [59, 60]. In general, these tumors display elevated choline to NAA and choline to creatine ratios. Furthermore, lipid lactate peaks, myoinositol and guanidinoacetate peaks are reduced relative to normal brain. Choline to NAA and choline to creatine ratios decrease following response to treatment and increase at the time of relapse [59, 60].
5.3.2 Noninfiltrative, Diffuse, or Circumscribed Astrocytomas These tumors are relatively well circumscribed when compared to diffuse astrocytomas. In general, they have a good prognosis. Several subtypes have been defined.
81
5 Low-Grade Gliomas
a
b
c
d Fig. 5.6 Sagittal T1 MRI (a), axial proton density MRI (b), axial T2 MRI (c), and post-gadolinium axial T1 MRI (d) images of a patient with a well differentiated diffuse infiltrative astrocytoma centered in the left superior temporal lobe. Note the lack of
enhancement as well as the homogenous appearance of the tumor on T1, T2 and proton density imaging. All of these features indicate low grade
82
G.A. Christoforidis et al.
a
b
c
d
Fig. 5.7 Anaplastic transformation of a low-grade astrocytoma. Sixty-one-year-old female patient who presented with aphasia was discovered to have a WHO grade II astrocytoma involving left temporal and parietal lobes. The lesion did not enhance on gadolinium MRI (a) but demonstrates extensive signal abnormality on FLAIR MRI (b). The tumor is much less conspicuous on a contrast-enhanced CT of the head (c). She underwent con-
formal radiation and chemotherapy. One year later the imaging characteristics of the tumor changed. The tumor enhanced (d), but the extent of the signal abnormality on FLAIR MRI did not change significantly (e). There was some hypodensity noted on CT within the tumor bed (f). Three months later the tumor showed significant progression on CT (g) and the patient expired shortly thereafter
83
5 Low-Grade Gliomas
e
g
Fig. 5.7 (continued)
f
84
5.3.2.1 Pilocytic Astrocytoma These tumors comprise WHO classification type I tumors classically composed of compact and spongy tissue. Pilocytic astrocytomas are slow growing and have been found to spontaneously stabilize or regress. They represent 2–6% of all primary brain tumors. Pilocytic astrocytomas usually present in the first two decades of life and they are the most common tumor of the cerebellum in this age-group (in other series medulloblastoma is the most common). They often arise from the cerebellum, hypothalamus, optic nerve, optic chiasm, and brain stem, or less commonly in the cerebral hemispheres. They are included in the few tumors that characteristically arise from the corpus callosum.
5.3.2.1.1 Pathology The gross appearance of pilocytic astrocytoma varies with location. In optic chiasm or optic nerve, pilocytic astrocytomas produce a fusiform expansion or a rather globular mass, respectively. In the hypothalamus and third ventricle, they are rather distinct tumors which may protrude intaventricularly. In brain stem, they are bulky protruding dorsally or causing diffuse expansion of the pons or medulla. In cerebral hemispheres or cerebellum, pilocytic astrocytomas are discrete compact tumors or cystic with a mural nodule. The name “pilocytic” (directly translated as “hair cell”) is derived from the long hair-like projections
a
G.A. Christoforidis et al.
emanating from the neoplastic astrocytes. Micro scopically, a combination of compact and loose areas characterizes most pilocytic astrocytomas (Fig. 5.8a). The former are composed of elongated cells with elongated nuclei, and the latter of microcysts with a spongiform appearance and stellate astrocytes with rounded nuclei. Pleomorphism, infrequent mitoses, or vascular proliferation may be found, but are rather degenerative and do not have an ominous prognosis as in other astrocytomas. Hyalinization of the blood vessels is another feature of pilocytic astrocytomas. Microvascular proliferation is also a frequent component of pilocytic astrocytomas and accounts for the contrast enhancement accompanying these tumors on cross-sectional imaging. Other histopathologic features which may accompany this tumor include Rosenthal fibers, eosinophilic granular bodies, and ganglion cells. Rosenthal fibers represent sausage-like or corkscrew-shaped filaments found in the cell processes of these tumor cells but can also be seen with reactive astrogliosis (Fig. 5.8b). The presence of eosinophilic granular bodies in these tumors is considered an important marker for lowgrade neoplasms such as this one. Pilocytic astrocytomas may spread locally, but are occasionally found to spread via cerebrospinal fluid (CSF) [38, 44, 61].
5.3.2.1.2 Imaging On cross-sectional imaging they are identified as wellcircumscribed tumors. Classically, they present as a
b
Fig. 5.8 Pilocytic astrocytoma. (a) A combination of mildly cellular regions and loose areas with microcysts. Hematoxylin–eosin, original magnification ×200. (b) The characteristics Rosenthal fibers (arrows). Hematoxylin–eosin, original magnification ×400
85
5 Low-Grade Gliomas
cystic tumor with an enhancing mural nodule (Fig. 5.9). Cystic components are identified in approximately 68% of cases and may develop on follow-up exam (Fig. 5.10). In general, chiasmatic location is associated with a lesser incidence of cyst formation [38, 44, a
62]. On CT the solid component of the tumor typically appears hypodense (43%) or isodense (51%) or, less commonly, hyperdense (6%) (Fig. 5.11). Calcifications have been reported in approximately 11% of tumors examined by CT [38, 44, 62] (Fig. 5.14a). On MRI, b
c
d
Fig. 5.9 Sagittal T1 (a), axial proton density (b), post gadolinium axial T1 (c) and sagittal T1 (d) MR images of a pilocytic astrocytoma of the cerebellum. Note the classic cyst and mural
nodule appearance of this neoplasm in a characteristic location of the posterior fossa. The nodule is enhanced while the adjacent cyst wall remains unenhanced
86
G.A. Christoforidis et al.
a
b
c
d
Fig. 5.10 Pilocytic astrocytoma within the left cerebellar hemisphere of a boy who presented at age 9 with headaches and developed a cyst 1 year later. Gadolinium-enhanced T1 MRI (a) demonstrated an enhancing focus within the tumor as well as T2 signal abnormality (b) beyond the enhancing borders of the
tumor. The tumor is less conspicuous on CT (c, arrowheads). One year later, the tumor developed a cystic component with ring enhancement on gadolinium-enhanced T1 imaging (d), which follows cerebrospinal fluid intensity on T2 (e) but not proton density (f) imaging
they present as well-circumscribed tumors (96%) with benign morphologic features and rare evidence for vasogenic edema pattern (5%). They display contrast enhancement in 94% of cases, which is thought to be related to the vascular nature of these tumors [44, 62, 63] (Figs. 5.9c, d, 5.12b, 5.13c, d, 5.15d). Despite the presence of increased vascularity in these tumors, this is not known to be a sign for higher grade in this tumor
type. On T1-weighted imaging, they tend to be of lower signal intensity relative to gray matter (Fig. 5.12a). On T2-weighted imaging, they are of higher signal relative to gray matter (Fig. 5.9b). The T2 signal intensity of the solid component is similar to that of CSF in 50% of cases. This differs from meduloblastomas, for example, which also occur in children but are isointense to gray matter on T2-weighted
87
5 Low-Grade Gliomas
e
f
Fig. 5.10 (continued)
a
b
Fig. 5.11 (a, b) Pilocytic astrocytoma. (a) Noncontrast CT shows a cystic lesion with an isodense nodule. (b) Postcontrast CT shows enhancement of the nodule
imaging and can thus help distinguish meduloblastomas from pilocytic astrocytomas [64]. Pilocytic astrocytomas display low rCBV values and show a characteristic type of signal intensity time curve with an increase in signal intensity above the baseline
due to massive leakage of contrast medium into the interstitial space. There are two major histological features that could explain this perfusion profile. One is the extensive hyaline mural thickening classically described in this tumor type leading to a possible vascular wall
88
a
G.A. Christoforidis et al.
b
Fig. 5.12 (a, b) Pilocytic astrocytoma of the cerebellum. (a) T1-weighted image shows a cystic lesion with a mural nodule, which is of lower signal intensity than the adjacent medulla. (b) On T2-weighted image the tumor appears hyperintense
fragility. The second is the rarity of endothelial proliferation defined as apparent multilayering of endothelial cells. Ultrastructural studies have shown that PA endothelial cells have open tight junctions and fenestrae allowing contrast medium extravasation. The strong enhancement of the solid component of PA may be due to a focal disturbance of the blood–brain barrier [65]. Varying imaging presentations may include solid tumors as well as multicystic tumors [62, 63]. In most locations they have a round or oval shape; however, in the chiasmatic location, they have been found to have a multilobulated shape (Fig. 5.13). In corpus callosum, pilocytic astocytomas should be differentiated from glioblastoma multiforme (Fig. 5.14). MR spectroscopy of these tumors can reveal an elevated lactate peak in these tumors; however, this is not a sign of malignancy for pilocytic astrocytomas [10]. In general, they are distinguished from other cystic tumors on the basis of location and patient age [2, 38, 44, 62, 63]. Although involvement of the subarachnoid space is frequent with pilocytic astrocytoma, leptomeningeal dissemination is rare [37, 66]. Pilocytic astrocytomas with atypical features on histopathology may display aggressive features such as invasion of adjacent structures and rapid growth on imaging (Fig. 5.15).
5.3.2.2 Pilomyxoid Astrocytoma Pilomyxoid astrocytoma (PMA) is a low-grade astrocytoma previously considered to be classified as a pilocytic astrocytoma. They tend to occur in early childhood but have been identified in adults [67]. After resection, progression-free survival has been reported in 38.7% of cases [68, 69].
5.3.2.2.1 Pathology This group of astrocytomas has a monophasic pilomyxoid angiocentric histologic pattern and lacks the Rosenthal fibers that are characteristic of pilocytic astrocytomas and rarely have eosinophilic granular bodies.
5.3.2.2.2 Imaging These tumors tend to occur in the hypothalamic/chiasmatic region (Fig. 5.16) and are likely to be solid enhancing tumors with homogeneously high signal on T2-weighted images with signal extending into
89
5 Low-Grade Gliomas
b
a
d c
Fig. 5.13 Sagittal T1 (a), axial T2 (b), postgadolinium axial T1 (c), and coronal T1 (d) MR images of a pilocytic astrocytoma of the opticochiasmatic-hypothalamic area. This location is not unusual for a pilocytic astrocytoma. Unlike well-differentiated,
diffuse infiltrative astrocytomas, pilocytic astrocytomas demonstrate enhancement areas of tumoral necrosis. The well-circumscribed appearance is more typical of low-grade tumors
the adjacent white matter and CSF dissemination. The larger and homogeneously enhancing solid component, T2 signal abnormality extending into the gray matter and white matter, CSF dissemination, and attendant hydrocephalus help distinguish PMA
from pilocytic astrocytomas. Proton MRS of PMA in two cases reported in the literature have shown a decreased concentration of choline, creatine, and NAA, which differs from pilocytic astrocytomas that tend to display elevated choline and decreased
90
a
G.A. Christoforidis et al.
b
d
c
Fig. 5.14 Axial CT (a), sagittal T1 (b), axial T2 (c) and axial post gadolinium T1 (d) MR images of a pilocytic astrocytoma centered in the corpus callosum. Among neoplasms that occur in young adults within the corpus callosum, pilocytic astrocytomas
should be considered. Note the calcific focus which occasionally accompanies this tumor type (arrows). The well-circumscribed appearance is suggestive of low grade and may help distinguish it from the “butterfly” pattern seen with glioblastoma
91
5 Low-Grade Gliomas
b
a
c
Fig. 5.15 (a) Atypical pilocytic astrocytoma. A sagittal T1 (a), axial T2 (b), axial FLAIR (c), and axial postgadolinium T1 (d) images of a pilocytic astrocytoma which displayed aggressive features on histopathology including an increase in proliferative activity. The neoplasm spans between the corpus callosum (black arrowheads) and the thalamus (black arrows). There is apparent invasion of adjacent centrum semiovale (white arrows)
d
suggestive of a more aggressive neoplasm. (b) Fig. 5.16 Pilomyxoid astrocytoma; 15-year-old male with pilomyxoid astrocytoma. Noncontrast axial CT (a), axial FLAIR (b), and postgadolinium axial T1 (c, d) images through the brain of a patient with a hypothalamic cystic lesion. Note calcifications (a), infiltrative appearance as well as remote spread (d)
92
G.A. Christoforidis et al.
a
b
c
d
Fig. 5.16 Pilomyxoid astrocytoma; 15-year-old male with pilomyxoid astrocytoma. Noncontrast axial CT (a), axial FLAIR (b), and postgadolinium axial T1 (c, d) images through the brain
of a patient with a hypothalamic cystic lesion. Note calcifications (a), infiltrative appearance as well as remote spread (d)
creatine and NAA. Because it is only recently described imaging findings are not frequently reported in the literature [70].
classified as WHO classification grade II neoplasms, but have been known to undergo malignant transformation. Like pilocytic astrocytomas, PXAs occur more frequently in the first two decades of life. Unlike pilocytic astrocytomas, they occur more commonly in the cerebral hemispheres with a predilection in the temporal lobes followed by the parietal, occipital, and frontal lobes. Rarely, these tumors may involve the cerebellum and the spinal cord [74–78].
5.3.2.3 Pleomorphic Xanthoastrocytoma Pleomorphic xanthoastrocytomas (PXA) are rare tumors of children and young adults comprising less than 1% of all gliomas [71–73]. They are generally
93
5 Low-Grade Gliomas
a
b
Fig. 5.17 (a, b) Pleomorphic xanthoastrocytoma. (a) Low magnification. A cellular, pleomorphic tumor with a focus of perivascular lymphocytes (asterisk). Hematoxylin–eosin, original
magnification ×100. (b) Higher magnification. Large pleomorphic plump cells (arrowheads), and cells with lipid droplets (arrows). Hematoxylin–eosin, original magnification ×400
5.3.2.3.1 Pathology
5.3.2.3.2 Imaging
PXA are tumors with a varied histologic appearance. Gross evaluation indicates that they involve the leptomeninges as well as the underlying brain. Their peripheral location suggests that they arise from subpial astrocytes [79]. These tumors have been rarely known to transform into gangliogliomas [39, 42, 80, 81]. If only a small amount of tumor is provided to the pathologist for analysis, PXA may be confused with a glioblastoma. The MRI and CT appearance of the tumor may assist pathologic interpretation of this tumor [82]. Microscopically, pleomorphism is the hallmark of PXAs in which spindle and rounded cells (large or giant, mono- or multinuclear) are the main population [83, 84]. PXAs are more cellular and pleomorphic in their compact areas and have less density at the periphery, resembling diffuse infiltrating astrocytomas [85]. The superficial part of the tumor usually extends into the subarachnoid space [86]. Compact areas consist of pleomorphic, plump, and eosinophilic “glassy” cells, as well as cells with lipid droplets (Fig. 5.17). Although cellular lipidization is a feature of this tumor, this is not always overt or even present. Despite nuclear pleomorphism, necrosis is not found and mitoses, if present, are rare. Pericellular reticulin, most abundant in the compact areas of the tumor, is constantly found. Vascular sclerosis (but not microvascular proliferation) and foci of perivascular lymphocytes are also features of this tumor. Eosinophilic granular bodies, found in slow-growing tumors, may be encountered in the superficial areas of PXAs. Most of the neoplastic cells are immunoreactive to GFAP.
PXA presents either as a cyst with a mural nodule or, less commonly, as a completely solid tumor [50]. The mural nodule is usually attached to the leptomeninges [87]. On unenhanced CT, the mural nodule or the solid portion of the tumor appears hypo- or hyperdense. After the administration of contrast medium PXAs enhance markedly [88, 89] (Figs. 5.18, 5.19a). Angiography reveals that these tumors are hypervascular and receive supply from the meningeal arteries [75]. MRI reveals that relative to gray matter, the solid component of these tumors is of similar signal intensity on T1-weighted images and increased signal on T2-weighted sequences [77, 90]. Imaging features of low grade such as lack of peritumoral edema and calvarial scalloping frequently accompany these tumors. Postcontrast T1WI shows intense enhancement of the mural nodule or of the solid tumor. The wall of the cyst may or may not be enhanced. Dural leptomeningeal or gyriform enhancement may be present [55, 74, 77, 87, 88, 90] (Figs. 5.19b–d, 5.20, 5.21). On FDG-PET, PXAs are hypermetabolic for glucose with respect to white matter. It has been mainly used to diagnose recurrent PXA and could be predictive of aggressive clinical behavior. Considering the possibility of the malignant transformation of PXA, even if histologic findings indicate benign tumor, follow-up by FDG-PET is mandatory in order to detect potentially malignant recurrence. FDG-PET findings may indicate that the grading of PXA has been underestimated. Because of their peripheral location, they can be confused with meningioma on imaging [2, 77]. Other
94
a
G.A. Christoforidis et al.
b
Fig. 5.18 (a, b) Pleomorphic xanthoastrocytoma. (a) Unenhanced CT shows a hyperdense, left parietal lesion surrounded by hypodense edema. (b) Postcontrast CT shows marked enhancement of the solid tumor
differential considerations when identifying a tumor with characteristics of PXA include pilocytic astrocytoma, ganglioglioma, and oligodendroglioma.
5.3.2.4 Subependymal Giant Cell Astrocytoma Subependymal giant cell astrocytoma (SGCA) is a low-grade primary brain tumor assigned a WHO grade I classification. These tumors invariably occur in the setting of tuberous sclerosis and affect the region near the foramen of Monro eventually obstructing this structure and causing hydrocephalus. Less frequently, the tumors are found in the subependymal tissue at or near the atria or temporal horns. In addition to identifying this tumor in its typical location near the foramen of Monro, the identification of stigmata related to tuberous sclerosis in the same patient confirms the diagnosis of SGCA. SGCAs occur in 6–16% of patients with tuberous sclerosis. [15, 16, 20, 44, 55–97]. The peak age of occurrence is in patients aged 8–18 years. In patients with tuberous sclerosis screening is recommended every 1–2 years [98]. Clinically, the typical symptom of SGCA is increased intracranial pressure due to the obstruction of the foramen of Monro. Since SGCA grows into the
ventricular lumen and not into the brain parenchyma, the most common clinical presentation is increased intracranial pressure due to obstruction of the foramen of Monro. Bilateral obstruction typically causes asymmetric hydrocephalus. Spontaneous intratumoral hemorrhage can lead to death [91].
5.3.2.4.1 Pathology SGCAs are mainly composed of spindle to epithelioid large cells with abundant glassy eosinophilic cytoplasm. Intermingled, smaller elongated cells are present. Nuclear pleomorphism and multinucleation are easily found (Fig. 5.22). The tumor cells are usually arranged in perivascular pseudorosettes. Although mitoses are usually rare, increased mitotic activity may also be recognized. Despite their astrocytic appearance, these cells reveal a varying positivity to GFAP and limited reactivity with neuronal markers. This dual reactivity suggests a hybrid nature of the tumor cells. In general, more aggressive histopathologic features such as mitosis and cellular pleomorphism in these neoplasms have not been associated with shorter survival times as they would in other tumors [40, 42, 80, 81].
95
5 Low-Grade Gliomas
a
b
d c
Fig. 5.19 Axial CT (a), axial FLAIR MRI (b), coronal T2 (c), and coronal postgadolinium T1 (d) of a low-grade pleomorphic xanthoastrocytoma (PXA) located in the right parietal lobe cortex. Note the hyperdense appearance on CT (arrow, a). Features which
may help distinguish this as a low-grade PXA include a well- circumscribed appearance, cortical location, and contrast enhancement. It would be difficult to distinguish this tumor from other cortically based tumors such as metastasis or ganglioglioma
96
G.A. Christoforidis et al.
b
a
c
Fig. 5.20 Sagittal T1 (a), axial T2 (b), axial postgadolinium T1 (c), and axial relative cerebral blood volume (rCBV) MRI map (d) of a high-grade PXA centered in the inferior right frontal gyrus. The tumor is not readily distinguished from a low-grade PXA (Fig. 5.15). Note the cortical location as well as the presence of a cystic-appearing component (arrowheads) and an
d
enhancing component as well as isointense signal relative to gray matter on T1 (a), all features of PXA. The CBV map indicates higher blood volume in part of the neoplasm (arrow). Although this feature is suggestive of higher grade in astrocytomas, in a PXA, it may just be an indication of increased vascularity unrelated to grade
97
5 Low-Grade Gliomas
a
b
Fig. 5.21 (a, b) Pleomorphic astrocytoma of the left temporal lobe. (a) Coronal postcontrast T1-weighted image shows a cystic lesion with a markedly enhanced peripheral nodule
(b) Coronal T1 weighted image at a posterior level shows leptomeningeal enhancement
5.3.2.4.2 Imaging
features include the presence of calcifications (Fig. 5.23) and a hyperdense appearance relative to cortex (Fig. 5.24a). Tumoral calcifications are thought to relate to small areas of hemorrhage. Associated stigmata of tuberous sclerosis include presence of cortical tubers and calcified subependymal nodules (Fig. 5.24a) On MRI, SGCAs show mixed-signal intensities on both T1- and T2-weighted imaging. Most of them are isointense on T1-weighted images and hyperintense on T2-weighted images. Contrast enhancement is common with these tumors on both CT and MRI (Figs. 5.24b–d, 5.25). In general, cortical tubers are more readily apparent on MRI, whereas calcified subependymal nodules are more readily identified on CT [55, 94, 95]. The extent of brain involvement with cortical tubers has been shown to correlate with the severity of disease in these patients [15, 20, 58]. Patients with tuberous sclerosis likely benefit from annual surveillance for these tumors during childhood [95]. In this manner, early resection of these tumors when they arise results in improved overall outcome. Differential considerations for this tumor on imaging include other intraventricular tumors such as central neurocytoma, metastasis, oligodendroglioma, pilocytic astrocytoma, and meningioma. SGCA can be distinguished from
Neonatal ultrasound imaging has identified this tumor as one of the few brain tumors which may be identified at birth and should therefore be included in the differential diagnosis of neonatal tumors when appropriate. The mass tends to be isoechoic with hyperechoic foci representing calcification or hemorrhage. Intrinsic CT
Fig. 5.22 Subependymal giant-cell astrocytoma. Large plump and fusiform cells with prominent thick processes. Hematoxylin– eosin, original magnification ×400
98
G.A. Christoforidis et al.
5.4.1 Pathology
Fig. 5.23 Precontrast CT shows a partially calcified mass at the foramen of Monro (arrows)
these tumors on imaging by the identification of features of tuberous sclerosis [97, 98].
5.4 Oligodendroglioma Oligodendrogliomas are typically slow-growing tumors with a propensity to involve the cortex. Imaging features associated with this tumor include calcifications, cyst formation, calvarial scalloping, and heterogeneous signal intensities on T2. Contrast enhancement with foci over 5 mm was seen only with anaplastic grade tumors. Histopathologic grading of the tumor, although often accurate, is not always predictive of biologic behavior. Since this is in part related to tumor heterogeneity, imaging may be very useful in directing biopsy sampling to areas that are more suspicious for higher grade. In this section, we review the pathologic features, histogenesis, grading, and imaging of oligodendrogliomas and provide some insight into the diagnostic challenges these tumors provide.
Oligodendrogliomas were first defined by Bailey and Cushing to represent glial neoplasms arising from oligodendrocytes [99]. They are commonly associated with prolonged survival relative to other CNS neoplasms [100]. Gross observation reveals an unencapsulated tumor that is gelatinous, soft, with a gray to pink hue. Oligodendrogliomas often contain cysts, calcifications, and foci of hemorrhage [81, 101] (Fig. 5.26). Classic microscopic features include cells with a characteristic retraction artifact (fried egg appearance) creating a perinuclear halo in a background of a rich plexiform, delicate capillary network (chicken feet appearance) (Fig. 5.27a). These neoplastic cells have a tendency to invade the cortex, a feature that is reminiscent of the oligodendrocyte’s migration pattern. At the cortex they are inclined to congregate around neurons (perineuronal satellitosis) (Fig. 5.27b). A high degree of cellularity out of proportion to the degree of nuclear pleomorphism may be confusing at first look but is a common feature of this neoplasm. Microcyst formation and mucin production often seen in oligodendrogliomas are thought to represent a vestige of the myelin formation in oligodendrocytes [81, 100–102]. There is no immunohistochemical marker specific for oligodendrogliomas [100, 103, 104]. The diagnosis of this tumor thus relies more heavily on the morphologic features of these neoplastic cells. Oligodendrogliomas do, however, provide variable expressions of GFAP, matrix glycoproteins, myelin basic protein (MBP), tetanus toxin protein (A2B5), galactocerebroside (GC), leu-7, vimentin, and S-100 protein [100, 103, 104]. GFAP-positive oligodendroglial cells (GFOC) have been of considerable interest. This subpopulation of cells constitutes the glial fibrillary oligodendroglioma (GFOG) subtype of oligodendroglioma [103, 105]. It is postulated that these cells give rise to transitional cells observed within oligodendrogliomas, which subsequently form gemistocytic (astrocytic) elements (Fig. 5.27c) within the tumor and are thus implicated in the transformation of an oligodendroglioma into an oligoastrocytoma [103, 105, 106] (Fig. 5.28). Astrocytic neoplastic cells are commonly found in oligodendrogliomas. A pure oligodendroglioma is defined to contain at least 75% oligodendroglioma cells. An oligodendroglial tumor with 25% or more content of astrocytic neoplastic cells is designated
99
5 Low-Grade Gliomas
a
b
c
d
Fig. 5.24 Axial CT (a), axial T1 (b), axial T2 (c), and axial postgadolinium T1 (d) MRI of a subependymal giant-cell astrocytoma (SGCA) located in the right foramen of Monro in a patient with tuberous sclerosis. Note that subependymal nodules are more conspicuous on CT examination (a) than on MRI. The tumor is hyperdense on CT, has a heterogeneous appearance on
T1 (b) and T2 (c) MRI, and enhances following contrast on MRI (d). These features, although typical of SGCA, are also features found in most lateral ventricular tumors. The diagnosis is made on the basis of stigmata of tuberous sclerosis such as the subependymal nodules
100
G.A. Christoforidis et al.
a
b
c
d
Fig. 5.25 (a–d) Subependymal giant-cell astrocytoma in a patient with tuberous sclerosis. (a) Axial T1-weighted image shows an isointense mass at the foramen of Monro. (b) Axial T2-weighted image shows high signal intensity of the mass (arrowhead). The low signal intensities represent intratumoral
focal calcifications. Also note the calcified subependymal nodules (arrows). (c, d) Axial and coronal postcontrast T1-weighted images show marked enhancement of the mass (arrow). A subependymal nodule is also enhanced (arrowhead)
5 Low-Grade Gliomas
Fig. 5.26 Intraoperative photograph of oligodendroglioma. Note the paucity of vascularity in the region of the tumor (arrows), the gelatinous appearance of the tumor, and the presence of a cyst on the cortical surface (arrowheads). This is the same patient as in Fig. 5.27
as oligoastrocytoma [107]. Described variants of oligodendrogliomas include a rare polymorphous variety in which uninucleated or multinucleated giant cells form and may have a familial occurrence [108]. A highly vascular variety (4%) was formerly referred to as angioglioma, but is angiographically occult and does not differ in clinical or prognostic terms from other oligodendrogliomas [81, 109]. Multicentric oligodendrogliomas [110] and metastatic forms [111] have also been described. As the oligodendroglioma infiltrates brain tissue, it engulfs reactive astrocytic cells, which in the midsection of the tumor bed gradually lose their cytoplasmic processes and morphologically resemble neoplastic astrocytic cells. This provides another diagnostic challenge to the pathologist in identifying tumors as oligodendrogliomas [109]. Immunohistochemical studies using monoclonal antibodies to A2B5, GC, GFAP, and MBP have given new insights into the histogenesis of oligodendrogliomas. Oligodendroglioma cells are thought to originate either directly or indirectly from O2A progenitor cells. These cells carry the propensity to differentiate into mature oligodendrocytes or type-2 astrocytes [101, 102, 112]. These progenitor cells initially express A2B5 antigens. The normal oligodendrocytic cell gradually loses its A2B5 expression as it differentiates and its ability to replicate. With further differentiation, it begins to form GC antigens, and lastly, loses its ability to migrate and begins
101
to form myelin (Fig. 5.29). Type-2 astrocytes, unlike the type 1 variety, do not form scar. They do, however, express GFAP. The neoplastic oligodendroglial cell expresses antigens from the O2A lineage. Immature cells are thought to represent dedifferentiated forms, which often do not express A2B5 [104, 112, 113]. Higher-grade oligodendrogliomas do not express GC or GFAP, which are frequently found in more mature oligodendroglioma cells [114] (Fig. 5.29). According to the concept of the window of neoplastic vulnerability, a glial cell is still replicating or capable of replicating at the time of the first step towards neoplastic transformation [113]. This requires either a reservoir of stem cells or the ability of differentiated cells to reenter the proliferative pool. Experimental evidence has recently shown that a reservoir of O2A cells does exist in humans and can thus serve as target cells toward the first step of the multistep process of neoplastic transformation. Tumor cells may arise elsewhere than the site of first strike [104, 114]. The use of histologic features in grading oligodendrogliomas and predicting clinical outcome has been difficult. Until 15 years ago, prediction of biologic behavior based on histopathologic grading was poor [115, 116]. The most widely accepted grading system is the Smith (AFIP) system [105]. Histologic features used to determine grade include pleomorphism, necrosis (Fig. 5.27d), nuclear to cytoplasmic ratio, endothelial proliferation, and cell density. If all five features are absent, the tumor is grade A, if all five are present, the tumor is grade D; only pleomorphism was found to independently correlate with survival. Median survival periods for each of these grades were: grade A, 94 months; grade B, 51 months; grade C, 45 months; and grade D, 17 months. Most large series do not concur on the prognostic significance of histologic grading criteria. Mitosis, necrosis, and pleomorphism (features suggestive of anaplasia) are the most consistently described features and correlate with shorter survival [117, 118]. Other features such as low cell density and microcyst formation have inconsistently correlated with longer survival [119–121]. Furthermore, in oligodendrogliomas, pleomorphism is difficult to evaluate since most lesions lack this feature to any significant degree [117]. Because Smith grades B and C show similar survival, many centers use a three-tiered system in describing oligodendrogliomas: well-differentiated, intermediate grade, and anaplastic, depending on the anaplastic features present [122]. Often, a small anaplastic focus and larger
102
a
G.A. Christoforidis et al.
c
b
d
e
103
5 Low-Grade Gliomas
Fig. 5.27 (a–e) Different histologic features of oligodendroglioma. (a) Classical oligodendroglioma with perinuclear halos (arrows) and delicate capillary vascularity. (b) Perineural satellitosis is a sign of cortical invasion. This perineural congregation occurs in both normal (white arrows) and neoplastic (black arrows) oligodendroglial cells. (c) Mini-gemistocytic astrocyte demonstrated on glial fibrillary acid protein (GFAP) stain counterstained with hematoxylin–eosin. The cytoplasm of this neoplastic cell is “stuffed” with GFAP staining substance. This cell
is thought to be involved in the transformation of oligodendrogliomas into oligoastrocytomas. (d) Tumor necrosis. Cell destruction is seen within this figure (arrows) associated with congregation of tumor nuclei around the area of necrosis (arrowheads). (e) Mitotic index. The proportion of cells staining positive for Ki-67 monoclonal antibodies [specific for nuclear antigens in active phases of the cell cycle (G1, S, G2)] is used to measure the proportion of cells capable of proliferating (arrows)
02A progenitor cell
ganglioside producing cell
Fig. 5.28 Hypothetical transformation of neoplastic oligodendrocytes into oligodendrogliomas and mixed oligoastrocytomas (see text). GFOC Glial fibrillary acid proteinpositive oligodendroglial cell; GFOG glial fibrillary oligodendroglioma
ganglioside & GFAP producing type 2 astrocyte
ganglioside producing neoplastic cell
nonmitotic migrating gangliocerebroside producing cell
galactocerebroside or GFAP + well differentiated neoplastic cell
galactoside + less mature neoplastic cell
nonmitotic nonmigrating myelin producing oligodendrocyte
GFAP + astrocytic neoplastic cell
immature antigen poor neoplastic cell
Neoplastic oligodendrocyte
GFOC
Transitional cell
neoplastic gemistocyte
Classic oligodendroglioma
GFOG subtype
Transitional tumar subtype
Gemistocytoma subtype
Oligodendroglioma
Fig. 5.29 Lineage of oligodendroglioma cells (see text)
Astrocytoma
Mixed Oligoastrocytoma
104
well-differentiated areas coexist; however, tumor grade is based on the most anaplastic focus. Recently, immunohistochemical stains using Ki-67 antibodies against nuclear antigens have been used to measure mitotic index and show promise in improving tumor grading [123] (Fig. 5.27e). Age has been determined in many series to have an inverse relation to survival [81, 100, 115, 117, 118, 124, 125]. Of note, tumor volumetric size, location, and presence of calcifications have not been found to correlate with survival in most large series [100, 121]. The concept that imaging can be used as an equivalent of macroscopic examination of a tumor has been proposed previously [126, 127]. Imaging evaluation prior to histologic examination can direct the surgeon’s and the pathologist’s attention to suspicious areas. Substitution of imaging for the macroscopic examination during the histopathologic examination of brain tumors may be useful because the tumor tends to be resected in a piecemeal fashion. Careful imaging evaluation in a tumor suspected on imaging to represent an oligodendroglioma can thus help suggest its diagnosis prior to pathologic examination. Directed biopsy may allow for more appropriate analysis of the features. Differentiating a glioma as an oligodendroglioma not only has prognostic implications but also treatment implications. Life-prolonging treatment regimens specific for oligodendrogliomas are available today.
5.4.2 Imaging These tumors are almost always supratentorial and distributed equally in all lobes on the basis of the size of the lobe, although they have been identified in the posterior fossa and within the ventricular system [44, 80, 81, 128–131]. They tend to involve both gray and white matter [44, 78, 79, 100, 107, 117]. Subarachnoid spread and multifocality have been described in the literature [107, 110]. These tumors have been found in patients 3–80 years of age, with a mild peak at the fourth and fifth decades and a slight male predilection [44, 80]. CT typically demonstrates a peripherally located hypodense tumor (Fig. 5.30). Hypodensity is observed in 57–70% of cases [132, 133]; however, intraventricular olidogendrogliomas have a tendency toward hyperdensity [130]. Calcifications have been identified in 40–90% of cases on CT [125, 130, 132, 133], although
G.A. Christoforidis et al.
pathologic series report calcifications in 45% of cases [121]. Calcifications tend to be coarse but more punctate or linear calcification may also be present (Figs. 5.30, 5.31). Plain films demonstrate the calcifications in 28–40% of cases [120]. Although calcifications have been correlated with a better prognosis in small radiologic series [124, 133], this has been disputed in larger pathologic series [117, 121]. Cysts have been identified in 20% of oligodendrogliomas on CT [133] and 32% in pathologic series [121] (Fig. 5.30a). The tumor is well-circumscribed in 49–57% of CT examination and enhances on CT in 24–66% of cases. Calvarial erosion indicative of long-standing tumor has been identified in 17% of CT examinations [125, 133]. MRI better delineates the tumor extent than CT. The tumor is usually (73%) hypointense on T1 MRI and hyperintense or heterogeneous on proton density and T2 images (Fig. 5.34a–c). On MRI, the tumor is found to be well-circumscribed more often than on CT [133] and foci of enhancement are more readily visible (Fig. 5.32a). Marked enhancement tends to be associated with anaplastic grades and milder or nodular enhancement does not appear to have a particular predilection for any grade [134] (Fig. 5.32b, c). MRI may show calvarial scalloping or calcification (Fig. 5.33b, c) but is generally less sensitive than CT. Calcifications may be more conspicuous on gradient-echo acquisitions [30, 135]. In our personal experience of 28 MRIs, we found cystic components similar to CSF on all pulse sequences associated with an oligodendroglioma in 21% of cases [134] (Fig. 5.33). Pathologically proven microcysts, which have been correlated with a better prognosis on pathologic series [119–121], give high signal intensity on both proton density and T2 imaging equal to that of CSF with a low T1 signal intensity (Fig. 5.34). The tumor can be seen on MRI to spread along white matter tracts and occasionally into and through the corpus callosum, but this does not correlate with tumor grade [134] (Fig. 5.35). Although hemorrhage has been noted to occur within these tumors, it may be difficult to distinguish on imaging, as it is not commonly reported in the imaging literature nor in our personal experience [80, 81, 125, 132, 133]. Contrast enhancement may occur in low-grade oligodendrogliomas; however, in our experience, enhancing foci over 5 mm in size occur more often in anaplastic oligodendrogliomas. Recurrence is not unusual, and especially in lower-grade tumors, nonenhancing recurrence may be difficult to distinguish from postoperative or postradiation change.
105
5 Low-Grade Gliomas
a
Fig. 5.30 CT demonstrating a hypodense, solid (arrowheads), and cystic (double arrow) component with clump-like calcifications (arrow) in anaplastic oligodendroglioma (a) and a
Fig. 5.31 Periatrial low-grade oligodendroglioma with intraventricular extension. Coronal CT demonstrates a calcific mass (arrow) with a nonenhancing solid component (arrowheads) and a cystic component (not seen). There was little or no evidence for cortical invasion
b
hypodense right frontal mass (arrowheads) with punctate calcifications (arrows) in intermediate-grade oligodendroglioma (b)
Gadolinium-based rCBV maps have been shown to have a very high predictive value for excluding the presence of high-grade glial tumor in untreated patients. However, oligodendrogliomas, which often display highblood volume foci even when benign, can confound the accuracy of rCBV mapping in glial tumor grading. Given the fine capillary network that is typical of even low-grade oligodendrogliomas, it is not surprising that their perfusion pattern differs from that of low-grade astrocytomas, and therefore can confound the reliability of nCBV values in distinguishing high-grade from low-grade untreated glial cell tumors [136]. Alternatively, higher rCBV in oligodendrogliomas may, in part, be related to their cortical location. Because the normal cortical gray matter contains a greater number of blood vessels compared with that of white matter, tumors involving the gray matter may have higher vascular density [137]. As such, rCBV is less reliable for grading and differentiation of oligodendroglioma than it is for astrocytomas. PET Imaging may help to differentiate low grade from anaplastic oligodendroglioma. MET seems to be more sensitive than FDG to
106
a
G.A. Christoforidis et al.
b
c
Fig. 5.32 (a–c) Two different forms of contrast enhancement. (a) Patient has a well-differentiated oligodendroglioma that is demonstrated on T1-weighted image as a well-circumscribed low-intensity lesion with punctate enhancement. (b) Patient has an anaplastic oligodendroglioma with marked enhancement,
cystic components, and hemorrhage. (c) Guided biopsy from the enhancing region of patient in (b) (arrow) does not demonstrate any distinguishing features and looks similar to biopsies from nonenhancing sites
107
5 Low-Grade Gliomas
a
b
c
Fig. 5.33 Cyst in a patient with low-grade oligodendroglioma (arrows) as depicted on T1 with contrast (a), proton density (b), and T2 (c) MR images. Also note clump-like calcifications (curved arrows) and calvarial scalloping (arrowheads)
108
a
G.A. Christoforidis et al.
b
c
d
Fig. 5.34 Axial MR images demonstrate a cortical tumor with low signal intensity on T1 (a) and marked signal intensity on proton density and T2-weighted (b, c) images. Biopsy demon-
strates microcystic changes (d, arrows) which contributes to the signal intensity in this patient with a well-differentiated oligodendroglioma
109
5 Low-Grade Gliomas
a
b
Fig. 5.35 Infiltration through the corpus callosum. (a) This patient presented with this infiltrative mass which demonstrated marked enhancement (not shown) and calcifications (arrow) He
was found to have anaplastic oligodendroglioma. He underwent partial resection and radiation. (b) Nine years later his exam is stable with postoperative changes
detect proliferation of oligodendroglioma. The increased uptake of MET could be due to the high rate of apoptosis in oligodendroglioma, which is supported by the finding that metabolites of methionine are involved in the apoptotic cascade. MET uptake also correlates with the Smith and Daumas-Duport classification and could be helpful to direct biopsies [138]. Oligodendrogliomas typically slowly infiltrate the surrounding tissues with a predilection for the cortex (Fig. 5.34). In our experience, cortical invasion is almost always present including occasional spread along the pia. This finding is a well-described feature of oligodendrogliomas in the literature [100, 109]. Tumor spread through white matter tracts is well-known. While infiltration into the corpus callosum does occur, there does not appear to be a correlation with tumor grade (Fig. 5.35). Although the tumor extends to the surface of the ventricle in nearly half the cases, this does not appear to have a prognostic significance. Recurrence with spread to the
ventricle, however, may result in CSF dissemination (Fig. 5.36). Behavior of oligodendroglioma is sometimes not predictable. Although in our experience, most recurrences occurred with the anaplastic grade, this is not consistent. One patient with an anaplastic grade tumor (Fig. 5.35) did not show recurrence, while one patient with a lowgrade tumor recurred within 1 year after surgery (Fig. 5.36). This underscores the problem in attempting to prognosticate oligodendrogliomas on the basis of a grading system. Anecdotal forms of oligodendroglioma include multifocal tumors (Fig. 5.37) and predominantly periventricular tumors (Fig. 5.31) rather than the more typical oligodendroglioma which invades gray matter. The differential considerations of low-grade oligodendrogliomas and low-grade mixed oligoastrocytomas on imaging include astrocytoma, ganglioglioma, gangliocytoma, and dysembryoplastic neuroepithelial tumor (DNT). Distinguishing features of oligodendrogliomas include gray and white matter involvement,
110
G.A. Christoforidis et al.
a
b
c
d
Fig. 5.36 (a–d) Cerebrospinal fluid (CSF) dissemination. This patient had a well-differentiated oligodendroglioma (a, T1 with contrast) that was resected but developed a local recurrence 11 months later. Biopsy and excision revealed low-grade tumor.
Following a full course of radiation and PCV (procarbizine, CCNU, vincristine) chemotherapy, the tumor recurred 14 months later locally (arrows, b proton density MR) and via CSF dissemination (arrows, c, d T1 MR with contrast)
111
5 Low-Grade Gliomas
Fig. 5.37 Multifocal intermediate-grade oligodendroglioma. Note the two tumor foci in the T1 sagittal MR (arrows). Biopsy in the MR normal-appearing parietal lobe between the two foci revealed sparse neoplastic cells
the presence of calcifications, and signal heterogeneity on MRI.
5.5 Choroid Plexus Papilloma Choroid plexus tumors include those neoplasms arising from choroid plexus epithelium. This includes CPPs, choroid plexus carcinomas, and atypical choroid plexus tumors. CPPs are WHO classification grade I tumors and are considered benign. Choroid plexus carcinoma is considered a WHO classification III neoplasm. The term atypical CPP is reserved for those tumors that do not clearly fall into the papilloma or the carcinoma categories [61, 80, 81, 139]. CPPs tend to occur in the first couple of decades of life, although they may develop at any age. Up to onehalf of all CPPs are found in the lateral ventricle, usually the atrium of the lateral ventricle. Of lateral ventricular CPPs, 80% occur within the first two decades of life. A small number of these tumors have been detected during the prenatal period. Fourth ventricular CPPs make up 40–48% of CPPs and are more evenly distributed with respect to age than lateral ventricular CPPs. The rest of the CPPs develop in the third ventricle and occasionally in multiple ventricles or in the vetricular foramina and the cerebrellopontine angle cistern arising from the small choroid tufts that normally project outside the foramen of Luschka [140].
The clinical findings include increased intracranial pressure due to hydrocephalus. Prior to closure of the cranial sutures, the findings include increased head size, frontal bossing, widening of the cranial sutures, and engorgement of the scalp veins. After closure of the cranial sutures, headache, vomiting, ataxia, and strabismus are the most common symptoms. Focal motor or sensory findings are uncommon [141, 142]. CPPs of the cerebellopontine angle may cause cranial nerve palsies [143]. A variety of mechanisms have been proposed to explain the hydrocephalus encountered in patients with choroid plexus neoplasms. The most widely accepted explanation for the development of hydrocephalus is the increased production of CSF by the tumor [142, 144]. Additionally, hydrocephalus can be caused by obstruction of the CSF flow due to tumor mass in the third or fourth ventricle. Typically, CPPs cause asymmetric hydrocephalus, which is more marked proximal to the obstructing mass. Other suggested causes of the hydrocephalus include increased protein content of CSF around the tumor, decreased CSF absorption by arachnoid granulation related to frequent tumoral hemorrhages, elevated intraventricular pressure, and adhesions around the exit foramina of the fourth ventricle caused by highly proteinaceous or hemorrhagic CSF [145]. Total resection of a CPP usually results in cure.
5.5.1 Pathology Choroid plexus neoplasms are tumors ranging from well-differentiated forms to carcinomas with epithelial differentiation. Most of them are papillomas composed of delicate fibrovascular stalks covered by a layer of columnar epithelium without or with sparse mitoses. Occasionally, focal ependymal differentiation is recognized. Bone and/or cartilage may also be found [146]. Atypical papillomas are characterized by cytologic atypia, increased nuclear:cytoplasmic ratio, and a limited number of mitoses (Fig. 5.38). Carcinomas are cellular tumors with poorly formed papillae, nuclear pleomorphism, and brisk mitotic activity. Neoplastic cells are arranged in atypical glands or show a cribriform pattern. The tumor cells are immunoreactive to cytokeratins as well as to vimentin, S-100 protein, and occasionally GFAP [147, 148].
112
a
G.A. Christoforidis et al.
b
Fig. 5.38 (a, b) Choroid plexus papilloma. (a) Low magnification. Delicate fibrovascular stalks covered by epithelium. Hematoxylin–eosin, original magnification ×100. (b) Higher
magnification. The epithelial cells that cover the stalks are tall and columnar. Hematoxylin–eosin, original magnification ×400
5.5.2 Imaging
hydrocephalus (Figs. 5.41, 5.43). Prenatal and neonatal ultrasound of CPPs reveals a hyperechoic mass with large vascular channels and associated ventriculomegaly [136, 156]. CSF seeding is known to accompany these tumors [139, 157]. As a result, consideration of spinal MRI surveillance for CSF seeding may be useful. Choroid plexus carcinomas tend to be less homogeneous and to invade the adjacent brain parenchyma [158] (Fig. 5.43). Central hypointense regions of cystic degeneration can be seen within choroid plexus carcinomas, which help to distinguish them from CPPs [142]. MR spectroscopy may be used to distinguish choroid plexus carcinoma from CPP. Both tumors may display markedly elevated choline peaks; however, the distinguishing feature is the myoinositol peak elevation in the CPP [159]. Uptake of methionine on PET imaging is significantly higher in CPP than in other low-grade gliomas, whereas no clear difference could be determined for FDG PET imaging. This implies that amino acid metabolism differs between CPP and low-grade glioma. There may be several explanations for this difference. Because methionine uptake is thought to depend on cell proliferation, increased methionine uptake may suggest that every CPP has the potential for aggressive proliferation. Alternatively, because the choroid plexus is the site for CSF production and the protein within CSF amino acid metabolism in CPP may be increased. Finally, CPP lacks a blood–brain barrier. Even though methionine transport though the
On imaging, CPPs acquire the imaging characteristics of the normal choroid plexus. On CT, they are hyperdense or isodense relative to cortex and frequently contain prominent calcifications and display intense contrast enhancement [149–154] (Fig. 5.39a). Angiography (Fig. 5.39d) displays a vascular blush in these tumors with typical arterial supply from branches of the anterior and posterior choroidal arteries when located above the tentorium, and the anterior inferior cerebellar artery and the posterior inferior cerebellar artery when located in the posterior fossa [153, 154]. On MRI CPPs tend to have a heterogeneous appearance with multiple lobulations. Areas of hypointensity within the tumor may represent calcifications or flow voids. On T2-weighted imaging they display heterogeneous but hyperintense signal relative to cortex. Contrast administration shows intense enhancement (Figs. 5.39, 5.40, 5.41). A cleft is frequently identified between the tumor and the adjacent ependymal surface. Clear evidence for attachment to the normal choroid helps confirm the diagnosis of a choroid plexus tumor. Hemorrhagic byproducts are often identified within these tumors and are not a sign of higher grade. Occasionally, a small area of brain parenchymal involvement with or without a vasogenic pattern of edema may accompany these tumors. Rarely, cystic areas (Fig. 5.42) have been reported to be present within the tumor [155]. CPPs and carcinomas are usually associated with ventriculomegaly and clinical
113
5 Low-Grade Gliomas
a
b
c
d
Fig. 5.39 An 8-month-old child with a biopsy-proven choroid plexus papilloma. CT imaging (a) demonstrates a left sided intraventricular mass isodense to gray matter associated with ventriculomegaly. T2-weighted imaging (b) reveals tumoral signal heterogeneity and a sizeable draining vein (arrow).
Postgadolinium T1 imaging (c) demonstrates signal heterogeneity and intense contrast enhancement and a draining vein (arrow). Lateral view of a left internal carotid angiogram (d) displays an enlarged anterior choroidal artery(arrows) and a tumor blush (arrowheads)
blood–brain barrier is carrier facilitated, membrane permeability may play a large role [160]. Differential considerations when identifying these patients on imaging include ependymoma, oligodendroglioma, meningioma, and subependymoma. In general, the patient’s age helps distinguish these tumors from other intraventricular tumors.
5.6 Ependymoma Ependymomas are WHO classification II neoplasms consisting of ependymal cells. WHO classification III ependymal neoplasms are referred to as anaplastic ependymomas. Although they may be found in any agegroup, brain ependymomas have a tendency to develop
114
a
G.A. Christoforidis et al.
b
d
c
Fig. 5.40 Sagittal T1 (a), axial T2 (b), and postgadolinium axial T1 (c) at the level of the pons and axial T1 (d) at the level of the medulla MR images of a choroid plexus papilloma of the fourth ventricle. Notice the lobulated appearance on the sagittal
view (arrows) as well as the heterogeneous appearance of the tumor on T1- and T2-weighted MRI. Varying patterns of enhancement are identified within different lobules of the tumor
in the first couple of decades of life, in contrast to spinal ependymomas, which comprise the most common glioma in adults. The overwhelming majority of ependymomas arise in the ventricular system, most commonly within the posterior fossa followed by the lateral ventricles and then the third ventricle. Ependymomas are usually found along the ependymal lining of the ventricular system. Approximately 58% are infratentorial and 42% are supratentorial, occurring most commonly
in the fourth ventricle followed by the lateral ventricles and less commonly (8%) in the third ventricle [161–164]. In general, infratentorial ependymomas are found most commonly in children [165]. Supratentorial ependymomas can be intraventricular or intraparenchymal. Parenchymal ependymomas may arise from infratentorial ependymal cell rests [166, 167]. Clinically, the most common symptoms of infra tentorial ependymomas, i.e., nausea, vomiting, and
115
5 Low-Grade Gliomas
a
c
Fig. 5.41 (a–d) Choroid plexus papilloma. (a) Axial T1-weighted image shows a large, almost isointense intraventricular mass in the left atrium causing severe hydrocephalus. (b) Axial T2-weighted image shows heterogeneous, high-signal intensity mass. Curvilinear structures of signal loss represent
b
d
intratumoral blood vessels (arrows). (c, d) Axial and coronal postcontrast T1-weighted images show a large, markedly enhanced intraventricular mass. Note the vascular pedicle (black arrow) and the lobulated appearance of the tumor (white arrows)
116
headache, are related to increased intracranial pressure. Supratentorial ependymomas usually cause headache, while focal motor symptoms are seen in 15–20% of cases [168]. The treatment of choice is surgical resection followed by irradiation. The prognosis depends on the
G.A. Christoforidis et al.
presence of residual tumor on the postoperative MRI. The overall 5-year survival rate has improved to approximately 64% [163]. Pathology Ependymomas are characterized by variations of cellularity and architecture among different tumors as well as in different areas of the same tumor.
a
b
c
d
Fig. 5.42 A CPP associated with a cyst. Axial CT (a), sagittal T1 (b), axial T2 (c), postgadolinium T1 (d), and diffusionweighted imaging (e) through an intensely enhancing (d) choroid plexus papilloma associated with a cystic component (arrows) whose signal characteristics are similar to that of cere-
brospinal fluid. The enhancing solid component is hyperdense to gray matter on CT. (a) and isodense to gray matter on T1-weighted imaging (b). Note the extensive vasogenic edema associated with this benign neoplasm.
117
5 Low-Grade Gliomas
Fig. 5.42 (continued)
a
Fig. 5.43 Axial T2 MRI (a) and axial postgadolinium T1 MRI (b) of a choroid plexus carcinoma (CPC) involving the trigone of the left lateral ventricle. Venticulomegaly as seen here is a
In classic ependymomas, glial and “epithelial” features may be found in different proportions. When the former predominate, neoplastic cells with glial features are diffusely arranged. Among these cells and around blood vessels, clear zones consisting of cytoplasmic processes are easily recognizable at low magnifications. These “perivascular pseudorosettes,” a hallmark of ependymomas, are a useful feature in the differential diagnosis from other gliomas in paraffin sections as well as in frozen sections. The “epithelial features” are a safe criterion in the diagnosis of ependymomas, but occur less frequently. Neoplastic cells retain their ependymal-epithelial properties and are disposed in linear, ependymal canalicular or pseudoglandular forms known as “true ependymal rosettes” (Fig. 5.44). Although diagnostic, unfortunately, this feature is uncommonly found and usually in infratentorial ependymomas [168]. Histopathologic variants of ependymoma include the cellular ependymoma, the rare papillary ependymoma, the tanycytic ependymoma, and the clear cell ependymoma, which tends to occur
b
frequent finding among patients with choroid plexus tumors. Distinguished from a CPP, a CPC tends to invade adjacent brain more readily as seen in this case
118
a
G.A. Christoforidis et al.
b
c
Fig. 5.44 (a, b) Ependymomas. (a) Low magnification. An area of high cellularity and perivascular pseudorosettes. (asterisks). Hematoxylin–eosin, original magnification ×100. (b) Higher
magnification. A perivascular pseudorosette.Hematoxylin– eosin, original magnification ×400. (c) True ependymal rosettes (asterisks). Hematoxylin–eosin, original magnification ×400
at the foramen of Monro [80, 81, 161]. True (genuine) papillary ependymomas are rare tumors characterized by a pseudoepithelial architecture around vessels. Tanycytic ependymomas consist of neoplastic cells with long processes arranged in a fascicular pattern. An oligodendroglial appearance in some areas of typical ependymomas is not rare and does not necessarily define a tumor as mixed. When an ependymoma is exclusively composed of oligodendroglial-like cells, it is characterized as clear cell ependymoma [169]. When focal, cellularity and atypia alone, occurring in an otherwise typical ependymoma, are not sufficient criteria for the diagnosis of malignancy. Ependymomas are characterized as anaplastic (malignant) when, in addition to marked cellularity and atypia, mitoses are abundant, necrosis is present, and endothelial proliferation is overt. A loss of true ependymal rosettes or
perivascular pseudorosettes is found in these anaplastic ependymomas [81, 168]. Due to the heterogeneity of the tumor, the borderline between typical and anaplastic ependymomas is not always clear. On immunohistochemistry, ependymomas are positive for GFAP, especially in the perivascular pseudorosettes. Reactivity for epithelial membrane antigen (EMA) is found on the epithelial surfaces of some ependymomas. Imaging. On CT, infratentorial ependymomas are most often isodense but can be slightly hyperdense. CT studies of ependymoma indicate that they often contain calcific foci. Up to 45% of posterior fossa ependymomas contain calcifications. After the administration of contrast material, they show enhancement varying from inhomogeneous to homogeneous [170–172] (Fig. 5.45). On MRI the typical appearance of posterior fossa ependymoma is a heterogeneous mass filling the fourth
119
5 Low-Grade Gliomas
a
b
Fig. 5.45 (a, b) CT of infratentorial intraventricular ependymoma. (a) Unenhanced CT shows a slightly hyperdense fourth ventricular mass with small calcifications (arrows). (b) Postcontrast CT shows mild homogeneous enhancement of the mass
ventricle and causing obstructive hydrocephalus. The solid portion of ependymomas is hypointense to isointense in relation to white matter on T1-weighted images and hyperintense on T2-weighted images. The heterogeneous appearance is due to necrosis, calcification, hemorrhage, or tumor vascularity [173]. Following contrast administration they are enhanced inhomogeneously; however, rarely, ependymomas do not enhance [91, 174] (Fig. 5.46). The distinguishing imaging feature for these tumors is their morphologic plasticity. These tumors have a tendency to conform to the ventricles that they are associated with and may indeed herniate through the ventricular foramina. An ependymoma, for example, may squeeze through the foramen of Lushka and find itself involving both the fourth ventricle and the cerebellomedullary angle cistern [175] (Fig. 5.47). A small percentage of ependymomas have been found to display CSF seeding [161–165]. Supratentorial ependymomas are most commonly intraparenchymal and very often contain a cystic component. The incidence of a cystic component in supratentorial ependymomas is significantly greater than in infratentorial ependymomas (Fig. 5.48). On
unenhanced CT they appear hypodense or isodense. About 50% of them show dense, punctuate calcification (Fig. 5.49a, Fig. 5.50a). After the administration of contrast material, they demonstrate variable enhancement, which can be homogeneous, inhomogeneous, or ring enhancing Fig. 5.50b. On MRI, the supratentorial ependymomas are usually hypointense to isointense on T1-weighted images and hyperintense on T2-weighted images. The signal intensities of cystic ependymomas on T1- and T2-weighted images are similar to those of CSF. The solid portion of the tumor may show signal heterogeneity on both T1and T2-weighted images due to the presence of hemosiderin, necrosis, vessels, and calcification. Enhancement is usually homogeneous or inhomogeneous (Fig. 5.49b–d). Only one-third of supratentorial ependymomas are intraventricular (Fig. 5.50). On MR spectroscopy ependymomas display a marked reduction in NAA. Glutamate relative to creatine is elevated more in anaplastic ependymomas than low-grade ependymomas [59, 60]. The differential diagnosis of supratentorial ependy momas includes low-grade astrocytoma, glioblastoma multiforme, primitive neuroectodermal tumor (PNET),
120
a
G.A. Christoforidis et al.
b
c
Fig. 5.46 (a–c) MRI of infratentorial intraventricular ependymoma. (a) Axial T1-weighted image shows an isointense mass filling the fourth ventricle. (b) T2-weighted image shows a
heterogeneous appearance of the mass. (c) Postcontrast T1-weighted image shows intense heterogeneous enhancement
121
5 Low-Grade Gliomas
a
b
c
Fig. 5.47 Sagittal T1 (a), axial T2 (b), and axial postgadolinium T1 (c) MR images of an ependymoma of the posterior fossa. The neoplasm herniates through the foramina of Luschka (arrows). This plasticity is a characteristic feature of ependymoma.
Other MRI features of ependymoma include a heterogeneous appearance and enhancement. In this case, the tumor signal on T1 (a) is difficult to distinguish from the signal of normal brain
giant-cell astrocytoma, and CPP. Low-grade astrocytomas usually do not enhance and infrequently show calcification [172]. Glioblastomas and PNETs may mimic a supratentorial ependymoma [176]. On the basis of imaging characteristics alone, intraventricular ependymomas cannot be distinguished from CPPs, while giant-cell astrocytomas are found in patients with tuberous sclerosis adjacent to the foramen of Monro [166]. Infratentorial ependymomas should be differentiated from medulloblastomas, cerebellar astrocytomas,
and BSG. Unlike ependymomas, medulloblastomas are typically homogeneous tumors, show homogeneous enhancement, and calcify only in 15% of the cases. Cerebellar astrocytomas, on the other hand, are usually located off-midline, extend into the cerebellar hemisphere, and tend to occur in older children. BSG involves the pons, causes local expansion, and rarely calcify [91]. In ependymomas, seeding through the CSF pathways with metastases occurs less frequently than in PNET. Generally, ependymomas may be difficult to distinguish from other neoplasms on imaging.
122
a
G.A. Christoforidis et al.
b
c
Fig. 5.48 Supratentorial ependymoma. Axial T1-weighted (a) and coronal T2-weighted (b) images show the solid portion of the tumor (arrow), the cystic portion (double arrows) and the adjacent edema (arrowheads). Coronal postcontrast T1-weighted
image (c) shows intense enhancement of the solid module (arrow), while the cystic portion of the tumor shows ring-like enhancement (double arrows)
123
5 Low-Grade Gliomas
a
b
c
d
Fig. 5.49 (a–d) Supratentorial ependymoma. (a) Postcontrast CT shows a right parietal mass with hypodense (arrows) and hyperdense component (double arrow). Also note the dense calcification (arrowhead). (b) On axial T1-weighted image, the mass appears inhomogeneous with hypo- and hyperintense areas. (c) On axial T2-weighted image, the focal calcification
appears dark, the adjacent solid tumor slightly hyperintense, while the cystic tumor appears markedly hyperintense. (d) Postcontrast axial T1-weighted image shows intense enhancement of the solid portion of the tumor. Note the extension of the tumor along the enhanced deep medullary veins
124
G.A. Christoforidis et al.
a
c
Fig. 5.50 (a–e) Supratentorial intraventricular ependymoma. (a) Unenhanced CT shows an intraventricular isodense mass with central calcification. (b) Postcontrast CT shows marked enhancement of the tumor. (c) On axial T1-weighted image the
b
d
tumor is isointense to the brain parenchyma. (d) T2-weighted image shows a heterogeneous mass with periventricular edema. (e) Postcontrast T1-weighted image shows intense homogeneous enhancement of the tumor
125
5 Low-Grade Gliomas
(79%) in one series [178] and 6–40% in the frontal lobe. Cerebellar involvement [179], and multifocal involvement of thalami, pons, basal ganglia, and third ventricle have been described [183]. Rare mitotic activity, vascular proliferation, and necrosis may be seen in complex histological forms of DNT.
5.7.1 Imaging
Fig. 5.50 (continued)
The age distribution of these neoplasms assists in favoring one tumor over the other. The morphologic plasticity of the tumor described earlier may also help distinguish an ependymoma from other tumors.
5.7 Dysembryoplastic Neuroepithelial Tumor DNT is a recently described cortically based tumor (WHO grade I) that is defined by a multinodular architecture which may include specific glioneuronal elements, a nodular component of glial cells, and cortical dysplasia [177]. Some 7.5–14% of cases of intractable epilepsy have been attributed to DNT [177, 178]. Patients typically present in childhood, though the tumor has been found in patients up to age 61 [179] and there is a slight male predominance. Daumas-Dupont et al. [177], in their original description of DNT in 39 epilepsy patients, found the tumor in the temporal lobe in 62%, frontal lobe in 31%, and parietoccipital lobe in 9% of the cases. Subsequent series [178, 180–182] have found an incidence of 40–91% in the temporal lobe, with predominance in the mesial temporal lobe
CT typically demonstrates a well-circumscribed hypodense lesion giving a “pseudocyst” appearance [177, 178, 182]. CT has been found to be normal in approximately 10% of cases [177, 178, 180] and occasionally isodense to gray matter or of mixed density [177, 184, 185]. A cystic appearance on CT has been noted in 28% of cases [186]. Calcific hyperdensities occur in only 20–36% of DNTs [178, 180, 181], though calcospherites on histopathologic examination have been found in 81% of cases [178]. Focal contrast enhancement on CT was noted in 18% of patients [177, 178]. Calvarial scalloping or temporal fossa erosion, signs of a slow growing neoplasm, are reported in 9–60% of patients with corticalbased tumors or larger-sized tumors [177, 181]. MRI experience has demonstrated that this tumor is usually well circumscribed and cortically based with gyral or nodular configuration, hypointense on T1, hyperintense on T2 relative to gray matter [177, 178, 180–185] (Fig. 5.51a), and homogeneous in 57% of cases [178]. Signal on proton density is increased in 66% and decreased in 33% relative to gray matter. MRI can identify a cystic component in 31% of cases [178]. Contrast enhancement on MR imaging was typically focal and punctate and identified in 16–66% of cases [178–181], though diffuse enhancement has been reported in 33% of lesions in one series [178] (Fig. 5.51b). The cause of punctate contrast enhancement is not known, though speculation includes the presence of vascular arcades on microscopic examination and breakdown of the blood–brain barrier due to frequent seizures [181]. White matter extension is noted in 43% of cases and blurring of the gray–white interface has been attributed to edema, invasion, cortical dysgenesis, and dysmyelination, noted on pathologic exam [36, 178]. The nodularity of the tumor has been speculated to be related to the histopathologically demonstrated architecture of the tumor with foci of cortical dysplasia and hypercellularity [178, 179]. Microcystic change accounts for at least some of the focal hyperintensities [181].
126
G.A. Christoforidis et al.
a
b
c
d
Fig. 5.51 Axial T2 (a) and postgadolinium axial T1 (b) MR images of a dysembryoplastic neuroepithelial tumor (DNT) centered in the right medial temporal lobe are compared with axial T2 (c) and postgadolinium axial T1 MR image (d) of a focus of cortical dysplasia adjacent to the trigone of the left lateral ventricle. The DNT is relatively well-circumscribed and predomi-
nantly involves cortex. A focus of contrast enhancement (arrow, b) is occasionally seen within these tumors but is not known to change the prognosis. Unlike DNT, cortical dysplasia tends to follow the signal of gray matter on all pulse sequences (arrowheads, c, d)
127
5 Low-Grade Gliomas
On perfusion imaging DNTs display characteristically low rCBV values, which allow their differentiation from low-grade oligodendrogliomas. Furthermore, the lower rCBV values of DNTs than those of normal parenchyma may also be helpful in distinguishing them from other low-grade cortical astrocytomas, which have rCBV values similar to or minimally higher than those of normal parenchyma [187, 188]. The lower rCBV values of DNTs than those of normal parenchyma has been explained by increased water content of the tumor, resulting from a loose eosinophilic matrix and an increased amount of fluid extravasation or by the lack of angiogenesis, which correlates well with their normal appearance on conventional angiography [189]. Diffusion imaging has demonstrated significantly higher ADC values in DNTs than normal parenchyma. DNTs have the highest ADC values (>2 × 10−3 mm2 s−1) among all low-grade tumors. The degree of malignancy is inversely correlated with ADC value likely because of the increasing tumoral cellularity with tumor grade [187, 188, 190]. In DNTs, a wide space for the Brownian motion of unbound water protons is created by the loose mucinous matrix due to an increased amount of fluid extravasation to the interstitial space and the floating appearance of neurons and bundles of axons attached to oligodendroglia-like cells [191]. The increased tumoral water content of DNTs is probably the major cause of their high ADC values. The ADC values were shown to have the highest correlation with the diagnosis of DNT. On spectroscopy, DNTs have lower NAA values than the contralateral normal parenchyma with a relative increase in Cho:NAA ratio which is due to a reduction in NAA level rather than to an increase in Cho. The NAA decrease in DNTs is probably due to a reduction in neurons per unit volume secondary to the presence of specific glioneuronal elements and increased mucinous material in the tumor [192]. The absence of lactate and lipid peaks also supports their less-malignant nature. One study could demonstrate a significant increase in myoinositol (mI) levels in DNT. Myoinositol is an astrocytic rather than a neuronal marker. The mI:Cr ratio in DNTs was higher than that of normal parenchyma. Evaluation of mI:Cr ratio can provide information on glial tumor grading. Increased mI:Cr ratios in glial tumors are always accompanied by decreased Cho:Cr ratios, independent of whether tumors are low or high grade [190, 193]. A relative increase in mI:Cr ratio with normal Cho:Cr ratio is quite diagnostic for DNT.
Angiography has typically been unremarkable with no neovascularity. Occasionally a vascular mass effect can be demonstrated [177]. SPECT imaging with N-isopropyl[123I]-p-iodoamphetamine (IMP) and 99m Tc-HMPAO has demonstrated marked hypoperfusion and no thallium uptake, unlike other low-grade gliomas, which demonstrate moderate hypoperfusion and low thallium uptake. PET using 18F-FDG has been shown to demonstrate hypometabolism within these tumors. SPECT using Tc99m-HMPAO has demonstrated hypoperfusion during the interictal period and hyperperfusion during ictus [194]. Differential considerations on MRI should include low-grade gliomas such as astrocytoma, oligodendroglioma, ganglioglioma, and pleomorphic xanthoastrocytoma. Distinguishing features of DNT when compared to these tumors include a thick nodular or gyral configuration with little or no white matter extension, rarely seen in other glial tumors [181, 184, 185]. Well-demarcated lobulated tumor margins without mass effect is seen in 80% of these tumors [181]. Preoperative suspicion of DNT may prompt the surgeon to submit the entire specimen to the pathologist since the pathologic diagnosis, in part, depends on the multiple nodular components of this tumor [177]. Distinguishing DNT tumor on pathologic analysis often presents a challenge. Indeed, the coexistence of DNT with ganglioglioma has been reported, which may indicate that these two tumors have a similar histogenetic origin [195, 196]. Of importance, cortical dysplasia, a congenital nonneoplastic entity, may be distinguished from DNT, because it closely follows the signal intensity of gray matter on all pulse sequences [178, 181] (Fig. 5.51c,d). Thin-section volume acquisition with multiplanar reformatting in patients with intractable epilepsy may help identify the tumor’s relationship to the mesial temporal structures. This technique may also demonstrate a small DNT in areas harder to discern on routine MRI exam such as the upper convexities. Furthermore, this approach may help identify foci of cortical dysgenesis, areas of cystic degeneration, and calcifications [178].
5.8 Subependymoma Subependymomas are biologically benign, slow-growing intraventricular tumors consisting of astrocytes
128
and ependymal cells. They are usually seen in adults with a mean age of 50 years. The term subependymoma was used first by Mark Scheinker in 1945 to describe a tumor of the fourth ventricle [197]. They are most commonly found in the fourth ventricle (30–60%), where they arise from the floor, and in the lateral ventricle (40–75%), where they are attached to the septum pellucidum [36, 198], although they have also been identified in the third ventricle and even the spinal cord [199]. Subependymomas are commonly identified in autopsy series and rarely produce symptoms. Based on critical location and size hydrocephalus is the most common presentation [36, 198].
5.8.1 Pathology The tumor is characterized by small groups of ependymal cells in a rather dense, delicately fibrillar stroma with prominent microcystic changes (Fig. 5.52). Calcifications, hemorrhages, and/or microvascular proliferation may be found.
5.8.2 Imaging The imaging appearance of subependymomas has been noted to vary based on location [146]. CT of fourth ventricular subependymomas demonstrates variable
Fig. 5.52 Subependymoma. Uniform nuclei in clusters (arrowhead) and microcystic changes (arrow) in a fibrillary background. Hematoxylin–eosin, original magnification ×400
G.A. Christoforidis et al.
density compared to gray matter with calcifications in 50–100% of cases and enhancement in 62–92% [198, 200–205] (Fig. 5.53a). Lateral ventricle subependymomas, on the other hand, vary in density, but are more often hypodense, usually do not enhance, and calcifications are seen in less than 10% [198, 201] (Fig. 5.54a). Unlike subependymomas, other lateral ventricle tumors such as ependymoma, choroid plexus papilloma, and central neurocytoma are more likely to demonstrate calcification or contrast enhancement. MRI of fourth ventricular subependymomas demonstrates the origin from the floor of the fourth ventricle almost always with extension through the foramina of Lushka or Magendie. They are either hypointense or isointense to gray matter on T1, and isointense or hyperintense to gray matter on T2. Fourth ventricular subependymomas almost always demonstrate heterogeneous enhancement on MRI (Fig. 5.53b–d). It is thus difficult to distinguish subependymomas from more aggressive tumors of the fourth ventricle based on imaging unless tumor is demonstrated to invade the adjacent brain parenchyma, which is distinctly unusual for subependymoma [206]. Fourth ventricular subependymomas have a close relationship to the brain stem and nearby cranial nerves and should thus be carefully scrutinized prior to surgery as even with incomplete excision recurrence or CSF dissemination is unusual [198, 206]. Although MRI may demonstrate encasement or displacement of adjacent blood vessels, subependymomas are usually dissected away from blood vessels at surgery [199]. Lateral ventricular subependymomas are typically hypointense to gray matter on T1 and hyperintense on T2. As with CT, they seldom demonstrate enhancement and are thus readily distinguished from other lateral ventricular tumors, which typically do enhance [198, 199, 203, 204, 207]. Subependymomas do not demonstrate paraventricular extension, unlike other ventricular tumors [207] (Fig. 5.54). On spectroscopy, 1H-MR spectra show modest elevation of the choline:creatine ratio, severe reduction of the NAA peak, and presence of lactate/ lipid peaks. The choline:creatine ratio of recurrent subependymoma has been reported to be as high as that of high-grade glioma, and higher than that of nonrecurrent subependymoma (2.66 vs. 0.48) [155]. Angiography of subependymoma inconsistently demonstrates an angiographic blush [155]. PET using 18F-FDG has disclosed that these tumors are very hypometabolic, which has been attributed to their low cellular density and slow growth [157].
129
5 Low-Grade Gliomas
a
b
c
d
Fig. 5.53 Contrast-enhanced axial CT (a), and axial T2 (b), postgadolinium axial T1 (c) and postgadolinium coronal T1 (d) MR images of a cerebellar subependymoma (arrows) in a 77-year-old male. The tumor displays contrast enhancement as well as the calcific deposits identified on the CT image (arrow-
heads). These features make the fourth ventricular location different than the lateral ventricular location of this tumor. The relative lack of invasion of the brain parenchyma suggest lowgrade tumor. Much like the ependymoma, this tumor also displays plasticity as it courses through the foramen of Luschka
Subependymoma are characterized by the absence of 201Tl uptake on SPECT, which seems to be related to the low cellularity of the tumor cells and the low activity of sodium potassium adenosine triphosphatase on the cell membrane [157].
gangliocytomas, which are also referred to as ganglioneuroma, and the more malignant ganglioneuroblastoma [80, 81].
5.9.1 Gangliogliomas 5.9 Ganglion Cell Tumors Tumors containing neoplastic ganglion cells are termed ganglion cell tumors and include gangliogliomas,
Gangliogliomas are formed by an admixture of neoplastic ganglion cells and glial cells. Unlike DNTs, the glial component is predominantly astrocytic and this tumor involves both gray and white matter [208].
130
G.A. Christoforidis et al.
a
b
c
d
Fig. 5.54 Contrast enhance axial CT (a), and axial proton density (b), axial postgadolinium T1 (c) and postgadolinium coronal T1 (d) MR images of a right lateral ventricular subependymoma. Lack of contrast enhancement distinguishes
this neoplasm from others in this location. Note the heterogeneous appearance of the tumor, a feature found in most tumors in this location. The tumor does not appear to invade adjacent brain parenchyma, indicating that it may be less aggressive
131
5 Low-Grade Gliomas
Gangliogliomas are relatively low-grade neoplasms that behave in a benign fashion and have a favorable prognosis. They are classified as WHO grade I or II tumors. Malignant transformation is very uncommon and metastasis in the CNS is exceptional. They frequently express cysts, which follow the CT density and MR signal pattern of CSF on imaging, as well as fibrosis and calcifications, which are hyperdense on CT and hypointense on MRI [209]. Solid components are often poorly defined and involve adjacent subarachnoid space in 47% of cases [210–212]. Patients affected range in age from 3 months to 80 years [213] with the median age at diagnosis reported among different series to be 14–25 years. There appears to be a slight male predilection [206, 209, 210, 214]. Gangliogliomas produce chronic intractable seizures more frequently than other tumors. Other presenting symptoms may include cranial nerve deficits and headache [212–215]. They represent 0.3–0.6% of all brain tumors and 1.2–7.8% of pediatric brain tumors [80, 81, 211]. A recent retrospective look at pediatric spinal tumors and adult cerebral gliomas using immunohistochemical neuronal markers revealed that the tumor may be more common than previously thought in the spinal cord and brain stem [209]. These two locations have an increased risk for recurrence [213, 214, 216–218]. Intracranially, the most common location is the temporal lobe, variously reported to occur in 30–84% of cases. Other frequently reported sites include the floor of the third ventricle, cerebellum, and brain stem, though any part of the brain including the optic nerves may be affected, as well as within the lateral ventricles [81, 208, 210, 212, 214, 216, 217, 219– 221]. Multifocal involvement is an infrequent occurrence [222].
Fig. 5.55 Ganglioglioma. Ganglion cells (arrowheads), neoplastic astrocytes (arrows), and eosinophilic granular bodies (gray arrowhead). Hematoxylin–eosin, original magnification ×160
a feature of slow-growing tumors, are common in ganglion cell tumors (Fig. 5.55). Rosenthal fibers may also be found. Calcospherites and perivascular lymphocytic infiltrates are common findings [209, 214, 216]. The diversity of the glial and the ganglion components of gangliogliomas occasionally provides difficulties in the differential diagnosis of these tumors from gangliocytomas or astrocytomas. To make things more complicated, histologic variations may coexist within the same tumor. For this reason the term “ganglion cell tumors” seems more appropriate [86]. When the neoplastic neurons are not recognizable among the glial cells, their immunoreactivity with antibodies to neurofilament, synaptophysin, class III b-tubulin, or chromogranin is a useful tool. Although anaplastic gangliogliomas may occasionally be encountered, this is unusual. In these cases, the element that has undergone neoplastic change is the glial one [81, 223].
5.9.1.1 Pathology 5.9.1.2 Imaging Diversity is a feature of ganglion cell tumors, which are characterized as gangliocytomas or gangliogliomas according to their cell population. In gangliogliomas, neoplastic glial cells are easily found in varying percentage and type, in addition to the ganglion cells. Most of them are astrocytes, although oligodendroglial cells may be recognized as well. In rare cases, the glial cells are morphologically similar to pilocytic cells. Eosinophilic granular bodies,
CT of gangliogliomas in a series of at least 12 patients exhibit a hypodense tumor relative to gray matter in 38–77% of cases or isodense in the rest [210, 216, 217, 220]; furthermore, 38–47% of the gangliogliomas appear cystic on CT [210, 216–218, 224–226]. This cyst-like component has been found to be solid intraoperatively [217, 226]. A mildly hyperdense tumor is identified in 0–23% of cases [159, 164, 165].
132
Fig. 5.56 Ganglioglioma. Noncontrast CT shows a left temporal lobe cystic lesion (thin arrows) with prominent bizarre calcifications (thick arrows)
Calcifications are identified in 31–69% of cases examined by CT [208, 210, 216–218, 224] and 8–21% by skull X-rays [210, 217, 218]. However, the typical appearance of ganglioglioma is a cyst with a mural nodule that is often calcified (Figs. 5.56, 5.57a). Contrast-enhanced CT exams demonstrate enhancement in 18–70% of cases [210, 216–218, 224] (Fig. 5.58a, b). CT exam is negative in 0–33% of cases [208, 210, 216, 218]. MRI is superior to CT in demonstrating the full extent of the tumor, location, and cysts; however, calcification is much better demonstrated on CT. Of note, there are at least two reported cases of patients who underwent temporal lobectomy for partial complex seizures who had normal MRI exams [218]. In general, MRI of these tumors exhibits a well-defined cysticappearing component and a less well-defined solid component [210] (Fig. 5.57b–d). Lesions which are primarily cystic occurred in 31–57% of large series, whereas completely solid lesions occurred 43–56% of the time [210, 216, 218]. Cystic components are hypointense (38%) or isointense (62%) on T1-weighted images and hyperintense (75%) on T2-weighted images [210, 216, 220] (Fig. 5.58c–d). However, in some cases, they show high signal on T1 due to the presence of hemorrhage, cholesterol, or proteins (Fig. 5.59). These cystic components on MRI have not always been confirmed as cysts intraoperatively [218]. Cystic tumors appear to be more common in early childhood (83% in patients with a mean age of 5.5 years) than in young adults (63% in
G.A. Christoforidis et al.
patients with a mean age of 25.6 years) with a higher percentage of the overall tumor size attributed to a cyst in childhood relative to adulthood (67 vs. 30%) [225]. Solid components typically demonstrate low (20–33%) or intermediate signal (33–73%) on T1, high (68–89%) or intermediate (11–29%) signal on T2, and high (98%) signal on proton density [210, 216] (Figs. 5.57b, 5.58d). Gadolinium enhancement has been variously reported to occur in 44% of cases in larger series [216]. Nodular, solid, ring-like, and mixed type of enhancement may be seen (Figs. 5.58e, 5.60). Enhancement typically involves one or more solid components of the tumor in a homogeneous fashion [216]. Intraventricular ganglioglioma may occur very exceptionally [210, 220, 225] (Fig. 5.61). Diffuse leptomeningeal spread of ganglioglioma on gadolinium-enhanced MRI is a rare occurrence [227]. Conversion to a higher grade may occur (Fig. 5.62). Gangliogliomas are typically avascular on angiography; however, a case of a highly vascular ganglioglioma has been reported in the literature [228]. Gangliogliomas demonstrate higher CBV compared with other low-grade gliomas, but the degree of vascular permeability in gangliogliomas is similar to other low-grade gliomas. The finding of increased perfusion in gangliogliomas indicates increased microvascularity, and hence gangliogliomas with higher rCBV measurements may behave more aggressively that those with a low rCBV [229]. MR spectroscopy of ganglioma has been shown to demonstrate a high choline to creatine and choline to NAA ratio in the face of low proliferative activity on histopathology [230]. Scintigraphic descriptions of gangliogliomas vary. PET using 18F-FDG has revealed hypermetabolism in one case [173]. In other recent studies, a group of lowgrade gangliogliomas were found to be hypometabolic on FDG-PET [231], whereas another series found hypermetabolism in one case [173] and heterogeneous metabolic activity in FDG-PET [232]. In one study, 201 T1-SPECT demonstrated high tracer uptake in two low-grade gangliogliomas [230]. Discrepant findings indicative of higher-grade malignancy with high metabolic activity on scintigraphic imaging and high cell turnover based on elevated choline to NAA ratio and choline to creatine ratio suggest that metabolic characteristics of this tumor may be influenced by activity other than growth [230]. Differential considerations in gangliogliomas include DNT, PXA, low-grade astrocytoma, oligodendroglioma, and gangliocytoma. Suggestive imaging
133
5 Low-Grade Gliomas
a
b
c
Fig. 5.57 Axial CT (a), and sagittal T1 (b), axial T2 (c) and postgadolinium axial T1 (d) MR images of a ganglioglioma containing cysts (arrows), calcifications (arrowheads), and solid
d
components (white arrows). All of these are features frequently found in a ganglioglioma
134
G.A. Christoforidis et al.
a
b
c
d
Fig. 5.58 (a–e) Ganglioglioma in a 30-year-old patient with headache. (a) Axial CT shows a hypodense left frontal mass with an isodense solid nodule (arrow). (b) Postcontrast CT shows intense enhancement of the nodule. (c) On T1-weighted image the left frontal mass is hypointense (arrows). (d) On
T2-weighted image, the cystic lesion appears hyperintense while the mural nodule shows intermediate signal intensity (arrow). (e) Axial postcontrast T1-weighted image shows marked enhancement of the nodule (arrow)
5 Low-Grade Gliomas
Fig. 5.58 (continued)
features for ganglioglioma include temporal lobe or posterior fossa location, involvement of both gray and white matter, combination of well-defined cystic and ill-defined solid components, calcifications, and enhancing nodule(s).
5.9.2 Gangliocytoma Gangliocytomas are extremely rare, purely neuronal tumors that can occur throughout the central nervous system. Intracranially, those typically occurring in the cerebral hemispheres and brain stem are distinctive from sellar gangliocytomas and dysplastic gangliocytomas of the cerebellum (Lhermitte–Duclos disease) described below. The age range is 5–52 years with an average age of 11 years at presentation. There is a slight male predilection [80]. In gangliocytomas (ganglioneuromas), which probably represent a hamartomatous process, dysmorphic neurons are clustered in a fibrillar background of
135
Fig. 5.59 Ganglioglioma of the left temporal lobe. Axial T1-weighted image shows high signal of the cystic lesion due to the presence of hemorrhage. The calcified portion of the tumor appears dark (arrows)
spindle, nonneoplastic cells (Fig. 5.63). At least some of these neurons are bi- or multinucleated. The nuclei are usually large and vesicular, with a prominent nucleolus. These abnormal ganglion cells are usually traced only with immunohistochemial techniques. Calcospherites may be found, but necrosis is absent [85]. Only limited imaging analysis has been reported on cerebral gangliocytomas. They tend to be slightly hyperdense on CT with little or no contrast enhancement, and no mass effect. These tumors tend to be difficult to identify on T1 MRI, though, if detected, are of mixed signal intensity. Signal intensity is intermediate to high on proton density MRI and intermediate or low signal on T2-weighted MRI [145, 146], although high signal on T2-weighted imaging has also been reported with this tumor [147] (Figs. 5.64, 5.65). Because of their signal characteristics these lesions can thus be confused with hemorrhagic foci. The intermediate and low T2 signal has been speculated to be related to dense congregations of large nuclei with prominent nucleoli with long-chain fatty acids, which can increase the T2 relaxation rate [233–237]. The hyperdense appearance on CT and the
136
a
G.A. Christoforidis et al.
b
c
Fig. 5.60 (a–c) Gangliogliomas with different patterns of enhancement. (a) Ring-like enhancement (arrows). (b) Solid-type and (c) mixed-type enhancement with solid (arrow) and ring-like (arrowhead) components
hypointense T2 signal of gangliocytomas may serve to differentiate them from other CNS neoplasms, which tend to be hyperintense on T2 and hypodense or isodense on CT. Heterotopia can be distinguished based on its tendency to follow gray matter on all pulse sequences [44]. This tumor has also been observed to express cystic components and enhancing nodules, which can look similar in appearance to gangliogliomas, PXA, or lowgrade astrocytomas [238, 239].
When occurring in the sellar region, gangliocytomas should be distinguished from hypothalamic hamartomas, which are congenital. Gangliocytomas, unlike hamartomas, demonstrate neoplastic cells and growth and are associated with pituitary adenomas in 65% of cases. Acromegaly is the most common presenting symptom in these patients. Sellar gangliocytomas cannot be distinguished from adenomas on imaging. They are hyperdense on CT in 90% of cases, enhance with
137
5 Low-Grade Gliomas
a
b
c
d
Fig. 5.61 (a–d) Intraventricular ganglioglioma in a 35-year-old patient. Axial T1-weighted image (a), axial proton densityweighted image (b), postcontrast axial (c), and coronal (d) T1-weighted images show an intraventricular mass in the left atrium, which is slightly hypointense on T1-weighted, hyperin-
tense on proton density-weighted, and markedly enhanced on postcontrast T1 proton density-weighted images. The small intratumoral signal voids represent blood vessels. Also note the ependymal enhancement on postcontrast T1-weighted image (arrows)
138
a
Fig. 5.62 Axial CT without contrast (a) and axial CT with contrast (b) demonstrate a high-grade ganglioglioma located in the right frontal lobe. The irregular-shaped cystic-appearing mass (arrowheads) with irregular margins and associated vasogenic
G.A. Christoforidis et al.
b
edema (arrowheads) is suspicious for a more aggressive tumor. Differentiation between this tumor and other high-grade tumors is not possible on the basis of this exam
5.9.3 Dysplastic Cerebellar Gangliocytoma
Fig. 5.63 Gangliocytoma. Clustered abnormal ganglion cells (arrows) in a fibrillar background. Hematoxylin–eosin, original magnification ×400
contrast, and demonstrate calcifications in 8% of cases. On MRI they are round, typically intrasellar, though they may involve the hypothalamus and be hyperintense on T1 and hypointense on T2 [240–242].
Dysplastic cerebellar gangliocytoma (DCG, Lhermitte–Duclos disease) is a tumor characterized by thickened cerebellar folia due to hypertrophy of granular cell neurons, hypermyelination in the molecular layer, and Purkinje cell loss and white matter atrophy [72, 243]. It is considered by some to represent a hamartomatous lesion [244, 245]. Conditions coexisting with DCG include holoprosencephaly, neurofibromatosis, Cowden’s disease, and multiple hamartoma syndrome [246–250]. Patients typically present with headache and hydrocephalus and range in age from newborn to 74 years, with an average age of 34 years [243]. No large series of this entity are available in the imaging literature. CT demonstrates a hypodense cerebellar lesion [243, 244]. Alternating layers of isodensity and hypodensity relative to gray matter involving the cerebellar cortex have been delineated on high resolution CT [243]. Calcifications are usually but not always absent
139
5 Low-Grade Gliomas
a
b
c
d
Fig. 5.64 Axial proton density (a), axial postgadolinium T1 (b), sagittal T1 (c), and sagittal postgadolinium T1 (d) MR images of a gangliocytoma involving the left caudate nucleus and left lateral ventricle. Note the heterogeneous appearance of
this tumor, a feature that is not unusual in these tumors. Lowsignal foci within the tumor bed (arrows) represent calcifications, which are a less common feature for these neoplasms
and no enhancement is discernible on CT [243, 244, 250]. MRI demonstrates the lesion to better advantage, especially since beam-hardening artifacts can hinder posterior fossa imaging on CT. The lack of clearly distinctive color, consistency, and structure relative to normal cerebellum makes it difficult to identify the lesion’s margins intraoperatively. MRI can thus help define the resection margins [250]. MRI demonstrates a laminated
lesion of T2 hyperintensity and T1 hypointensity with mass effect involving a cerebellar hemisphere. There is often involvement of the vermis. Hydrocephalus is a frequent finding in these patients. In addition, Chiari I and syrinx have also been observed with this lesion [243, 244, 251]. The striated appearance represents an isointense molecular layer with sulcal effacement observed along the cerebellar folia, with signal abnormality
140
a
c
Fig. 5.65 Axial CT (a), axial T2 (b), coronal proton density (c), and axial postgadolinium T1 (d) MR images of a gangliocytoma involving the right frontal periventricular white matter does not follow the characteristics presented in the literature. It is of low
G.A. Christoforidis et al.
b
d
density on CT (a), and high signal on T2-weighted (b) and proton density-weighted (c) imaging. It has features of low-grade, does not enhance following gadolinium administration (d), and is well circumscribed
141
5 Low-Grade Gliomas
(increased on T2 and decreased on T1) in the granular layer, deep molecular layer, and underlying white matter [243, 244]. Vascular proliferation in the associated pia is thought to represent a potential source of calcification and enhancement, occasionally observed in these patients [243–245, 252]. The striated appearance on MRI is felt to be characteristic for DCG; however, it can be confused with acute or subacute cerebellar infarcts. Infarct can be differentiated from this lesion based on clinical history and imaging evolution of infarcts over time [243, 244]. The unique appearance of DCG helps distinguish this lesion from other tumors. Recurrence after initial resection has been observed on long-term (12 years) follow-up and is occasionally symptomatic. Thus long-term follow-up may help detect early recurrence [243, 244, 253–256].
5.9.4 Desmoplastic Infantile Astrocytoma and Ganglioglioma Desmoplastic infantile ganglioglioma (DIG) and desmoplastic infantile astrocytomas are recently described, voluminous WHO grade I tumors, consisting of extensive desmoplasia, and either predominantly neoplastic astrocyts or mixed neoplastic ganglion and glial cells. They are typically associated with large cysts. They are considered to be benign despite their high mitotic activity, rapid growth, and aggressive appearance both on imaging and microscopy. They usually involve more than one lobe [41]. Typical locations include the frontal and parietal lobes, although it has also been observed to involve temporal and occipital lobes. The tumor is superficially located. Growth into the subarachnoid space and adjacent meninges is uniformly present in these tumors and a good deal of the desmoplastic component occurs in association with the meninges [257–259]. Extension into the lateral ventricles is unusual but has been reported with this tumor [260]. Almost all cases have occurred in children under 18 months of age, although there are isolated reports of DIG in adults [257, 258, 261–263]. The tumor has remarkable similarities to other recently described desmoplastic tumors of childhood including gliofibroma, pleomorphic xanthoastrocytoma (which have also been identified infratentorially and in the spinal
cord), and desmoplastic astrocytoma. Although the lack of ganglionic cells in these other desmoplastic tumors provides a distinction for DIG, these tumors may also represent varying expressions of a single desmoplastic tumor type [264–268].
5.9.4.1 Pathology Histologically, the tumor is characterized by deposition of dense collagen in combination with neuroepithelial and fibroblastic elements. The only distinguishing feature of DIG from DIA is the presence of neuronal cell differentiation in the DIG [258]. The neuro-epithelial component shows a variable proportion of glial astrocytic and neuronal ganglionic cell populations. The neoplastic astrocytes are moderately pleomorphic ranging from elongated to polygonal cells with irregular nuclei and eosinophilic cytoplasm. These cells show intense immunoreactivity for GFAP. The neuronal differentiated cells range from atypical ganglioid-cells to small polygonal cell types with copious cytoplasm and irregular nuclei. These neuronal cells express neuron-specific enolase (NSE) and synaptophysin. The superficial part of the tumor involves the leptomeninges and consists of elongated astrocytes intermixed with collagen and reticulin fibers surrounding the tumor [269]. Although angiomatoid vessels may be present, microvascular proliferation is not evident [19]. These tumors are associated with a low proliferative index, vementin and desmin expression. Because pleomorphic xanthoastrocytoma also expresses desmoplastic features, and clinico pathologic features of DIA, some authors suggest that these tumors may have common origins with differential expressions [270, 271]. Pleomorphic xanthoastrocytomas have a higher recurrence rate than DIA [269].
5.9.4.2 Imaging The most striking imaging feature of DIG/DIA is its relatively large size. CT reveals a large tumor with formation of a large hypodense cyst and a hyperdense solid component, which enhances intensely. Calcifications, and calvarial scalloping adjacent to the tumor have been noted to occur in about half the cases [272].
142
G.A. Christoforidis et al.
On MRI, the cystic component, as expected, has low T1 and high T2 signal. The signal characteristics of the solid component have been variably reported as hypo-, hyper-, or isointense relative to gray matter. The solid component markedly enhances and typically is adjacent to the meninges which may also thicken and enhance (Figs. 5.66, 5.67). A ring-like pattern of enhancement has been described but it is uncertain if this is characteristic of DIG [261, 262, 267, 273]. The variable signal characteristics of the solid component as well as its enhancement pattern can be explained by the intermixed desmoplasia found in this tumor [267].
a
b
d
e
Fig. 5.66 A desmoplastic infantile astrocytoma (DIA) discovered in a 6-year-old boy. Postgadolinium axial T1 (a), axial diffusion-weighted imaging (b), Axial FLAIR (c), coronal T2 (d), and postgadolinium coronal T1 (e) images localize the DIA (arrowheads) on both sides of the tentorial dura. Note the mild
The cystic cmponent typically surrounds the solid part of the tumor. Edema is usually absent or moderate [267, 269, 270]. Angiography has demonstrated a large avascular mass with a small tumor stain [260, 262, 269]. MET and FDG PET have displayed increased uptake in one case [269]. SPECT using I-123 a-methyl tyrosine (an amino acid marker) has identified a higher uptake in these tumors, indicating hypermetabolism [274]. Differential considerations include PNET, PXA, ependymoma, and astrocytoma. DIG should be suggested in infants presenting with a large superficial cerebral mass with large cystic components and an
c
meningeal reaction along the tentorium cerebelli (arrows). This patient has an atypically older age presentation for this tumor type. The more central location and lack of a cystic component make this particular tumor less distinctive relative to other patients presenting with this diagnosis
143
5 Low-Grade Gliomas
a
b
c
Fig. 5.67 Sagittal T1 (a), axial proton density (b) and postgadolinium axial T1 (c) MR images of a desmoplastic infantile ganglioglioma (DIG) centered in the right middle cranial fossa involving the right hemisphere. Involvement of the meninges (white arrrows) is a typical feature for these neoplasms. This
particular example contains large cysts separated be septations (arrowheads) and enhancing solid component adjacent to the meninges with heterogeneous appearance on proton density (arrows). These imaging features are typical of DIG
144
Fig. 5.68 Neurocytoma. Round and uniform nuclei with a “salt and pepper” chromatin appearance. Hematoxylin–eosin, original magnification ×400
enhancing solid component adjacent to meninges [261, 267]. Its identification is important since it has a significantly better prognosis and different management considerations relative to other infantile brain tumors.
5.10 Neurocytoma Neurocytoma is a recently defined, usually benign neoplasm of neuronal origin that is seen in young adults. Neurocytomas are slow-growing neoplasms and are classified as low-grade malignant tumors (WHO grade II) [275, 276]. A few cases with morphological features of malignancy have been described which are referred to as anaplastic neurocytoma [277–279]. They are almost exclusively intraventricular and have thus been termed central neurocytomas [278– 283]. Recently, intracerebral neurocytomas have been described and designated cerebral neurocytomas [284, 285]. Neurocytomas comprise 0.1–0.5% of all brain tumors [280, 281]. The age range for these tumors is 7–53 years [281, 286, 287] and mean reported ages among various small series span from 25 to 30 years with no sexual predilection [280–282]. Initial clinical findings can develop slowly over several months with symptoms of increased intracranial pressure attributed to obstructive hydrocephalus. Occasionally, the tumor can present with seizures, a stroke-like onset, or with intracranial hemorrhage [280–282, 284, 285]. Case reports in the literature suggest that these tumors have
G.A. Christoforidis et al.
the potential to hemorrhage and may present as such on imaging [283, 288–290]. Pathology. Neurocytomas are characterized by dense cellularity with a “honeycomb” appearance. The neoplastic cells are small and uniform, producing a streaming architecture. Their nuclei are characteristically uniform and finely speckled (“salt and pepper” pattern) (Fig. 5.68). The cytoplasm tends to be clear or eosinophilic with indistinct borders. Mitoses, pleomorphism, and necrosis are not found in neurocytomas. Features of these tumors are their perivascular fibrillarity (reminiscent of pseudorosettes) and the immunoreactivity to synaptophysin and NSE. GFAP staining can be positive in some neurocytomas. However, because of the presence of dendritic processes, the GFAPpositive cells are thought to be entrapped by nonneoplastic astrocytes. Positive staining with GFAP might suggest a more malignant course. Perinuclear halos, arcuating vascularity, and granular calcifications are features that raise the issue of differential diagnosis from oligodendrogliomas [81, 85, 277, 290]. Alternating fibrillary and cellular areas, scant mitotic activity, and a tendency to form ill-defined rosettes are features against oligodendroglioma. However, the tumor can also present with varying degrees of mitotic activity, vascular endothelial proliferation, and necrosis. These features are suggestive of atypical central neurocytoma. Aggressive behavior of the tumor has been described, such as fast progression, recurrence, extraventricular extension, and craniospinal dissemination. However, these aggressive patterns do not necessarily correlate with anaplastic features in original histology.
5.10.1 Imaging CT of central neurocytomas demonstrates a well-circumscribed, intraventricular, isodense (25–71%) or hyperdense (29–75%) tumor which enhances homogeneously (100%) (Fig. 5.69a). Calcifications are present in 52–75% of cases and tiny cystic components are present in 67–100% [280–282, 290, 291]. MR, due to its multiplanar imaging capabilities, can better demonstrate the intraventricular location of the tumor and its site of attachment [292]. MRI shows a large intraventricular tumor with frequent extension into the third ventricle through the foramen of Monro.
145
5 Low-Grade Gliomas
a
b
c
Fig. 5.69 (a–c) Central neurocytoma in a 40-year-old patient. (a) CT shows an intraventricular hyperdense mass. (b, c) Axial and sagittal postcontrast MR images show mild enhancement of
the mass causing displacement of the septum pellucidum (arrowhead). Note the small linear low signals representing intratumoral vessels (arrows)
146
G.A. Christoforidis et al.
Surgically demonstrated points of attachment more commonly include the septum pellucidum and lateral wall of the lateral ventricle, although central neurocytomas have also been reported to attach to the roof and inferomedial wall of the lateral ventricle as well as the third ventricle [280–282, 290, 291]. Extension into or origin from the adjacent brain parenchyma has been reported and has been associated with anaplastic forms of central neurocytoma [274, 281, 293]. Central neurocytomas are typically isointense to gray matter on T1 MRI (58–75%), but they can be hyperintense, hypointense, or mixed. T2-weighted imaging demonstrates isointensity (40–75%), mixed isointensity, and hyperintensity or hyperintensity relative to gray matter. Proton density sequences demonstrate isointense, hyperintense, or mixed signal relative to gray matter [280–282, 290]. Serpiginous and punctate signal voids can frequently be identified within the tumor [282]. In addition, a dilated thalamostriate vein or internal cerebral vein can often be seen on CT or MRI [280].
a
Fig. 5.70 (a–b) Intraventricular neurocytoma. (a) Axial T2-weighted image shows an intraventricular mass, isointense to the gray matter. (b) Postcontrast coronal T1-weighted image
Enhancement with gadolinium diethylenetriamine penta-acetic acid (Gd-DTPA) has been described as mild, homogeneous, or not seen [283, 293] (Figs. 5.69b,c, 5.70, 5.71a–d). On angiography, a vascular stain is shown (71%), and when demonstrated, the feeding arteries are typically choroidal or lenticulostriate. Draining veins are less frequently identified [282] (Fig. 5.70e). MR spectroscopy of central neurocytoma has shown significantly elevated ratios of choline to creatine: phosphocreatine and choline to NAA relative to normal brain. In addition, a lactate peak has also been observed in these patients [294]. Recurrence of neurocytomas has been recorded in the literature as early as 8 months and as late as 6 years [290, 295, 296], which suggests that appropriate follow-up imaging may help early detection. Isolated cases of ventricular and spinal dissemination following resection have also been reported and identified on follow-up imaging despite benign histology [295].
b
shows intense enhancement of the mass with a small cystic component (arrow)
147
5 Low-Grade Gliomas
a b
c
d
e Fig. 5.71 Sagittal T1 (a), axial T2 (b), postgadolinium axial T1 (c), and coronal T1 (d) MR images and lateral venous phase angiogram (e) of a central neurocytoma. Note attachment to the septum pellucidum (arrowheads), heterogeneous appearance and presence of
calcifications (white arrows), and homogeneous contrast enhancement. Note the enlarged thalamostriate vein (e arrow), a feature that may also be seen with other large neoplasms in this region
148
The few reported cases of cerebral neurocytomas indicate that these tumors can occur in the frontal and temporal lobes and involve both gray and white matter. They are hypodense and sometimes cystic in appearance on CT without calcifications. MRI demonstrates a tumor hyperintense on T2 and hypointense on T1 relative to gray matter [284, 285]. Differential considerations for central neurocytoma include intraventricular tumors centered around the frontal horn, foramen of Monro, and body of the lateral ventricle. These include meningioma, CPP, ependymoma, subependymoma, astrocytoma, oligodendroglioma, metastasis, and lymphoma. Intraventricular meningiomas are mainly found in older adults. Calcification is common, but is usually seen in the trigone of the lateral ventricles without close spatial relation to the septum pellucidum. Calcifications within meningioma are more common and larger than in central neurocytoma. Meningiomas also show strong enhancement after administration of contrast agent [297]. CPPs usually have an irregular surface and tend to involve the atria of the lateral ventricles. They enhance homogeneously with contrast agent and typically occur in children less than 10 years of age [298]. Ependymoma are most commonly found in the fourth ventricle and the surrounding brain is often invaded by the tumor. These tumors are mainly seen in children. Subependymomas may look similar to central neurocytoma on nonenhanced studies, but their lack of enhancement and older age distribution tends to differentiate them [299]. Astrocytoma may show calcifications; however, they usually lack intratumoral cysts and are often associated with a peritumoral edema which is uncommon in central neurocytoma [299]. Unlike neurocytoma, oligodendroglioma are rarely seen in the lateral ventricles. They usually arise in the frontotemporal brain and can erode the inner table of the calvarium, which is one of the distinguishing features [300]. Neuroblastoma and GBM tend to have a more aggressive appearance though may, on occasion, present a diagnostic difficulty. Many of these tumors tend to have imaging characteristics similar to those described for central neurocytomas thus making them difficult to distinguish [277, 283, 301]. Neurocytoma should strongly be considered as a diagnostic likelihood in a young adult patient with a lateral ventricle tumor associated with calcifications and moderate contrast enhancement.
G.A. Christoforidis et al.
References 1. Chamberlain MC, Murovic JA, Levin VA (1988) Absence of contrast enhancement on CT brain scans of patients with supratentorial malignant gliomas. Neurology 38:1371–1374 2. Sanders WP, Christoforidis GA (1999) Imaging of low-grade primary brain tumors. In: Rock JP, Rosenblum ML, Shaw EG, Caincross JG (eds) The practical management of lowgrade primary brain tumors. Lippincott Williams & Wilkins, Philadelphia, pp 5–32 3. Gupta RK, Sinha U, Cloughesy TF et al (1999) Inverse correlation between choline magnetic resonance spectroscopy signal intensity and the apparent diffusion coefficient in human glioma. Magn Reson Med 41:2–7 4. Pomper MG, Port JD (2000) New techniques in MR imaging of brain tumors. Magn Reson Imaging Clin N Am 8:691–713 5. Castillo M, Kwock L (1998) Proton MR spectroscopy of common brain tumors. Neuroimaging Clin N Am 8:733–752 6. Ott D, Hening J, Ernst T (1993) Human brain tumors: assessment with in vivo proton MR spectroscopy. Radiology 186:745–752 7. Shimizu H, Kumabe T, Shirane R, Yoshimoto T (2000) Correlation between level measured by proton MR spectroscopy and Li-67 labeling index in gliomas. AJNR Am J Neuroradiol 21:659–665 8. Fulham MJ, Bizzi A, Dietz MJ et al (1992) Mapping of brain tumor metabolites with proton spectroscopic imaging: clinical relevance. Radiology 185:675–686 9. Gotsis ED, Fountas K, Kapsalaki E, Toulas P, Peristeris G, Papadakis N (1996) In vivo proton MR spectroscopy: the diagnostic possibilities of lipid resonances in brain tumors. Anticancer Res 16:1565–1568 10. Hwang JH, Egnaczyk GF, Ballard E, Dunn RS, Holland SK, Ball WS (1998) Proton MR spectroscopic characteristics of pediatric pilocytic astrocytomas. AJNR Am J Neuroradiol 19:535–540 11. Trattnig S, Ba-Ssalamah A, Noebauer-Huhmann IM, Barth M, Wolfsberger S, Pinker K, Knosp E (2003) MR contrast agent at high-field MRI (3 Tesla). Top Magn Reson Imaging 14(5):365–375 12. Yuh WT, Fisher DJ, Engelken JD (1991) MR evaluation of CNS tumors: dose comparison study with gadopentate dimeculmine and gadoteridol. Radiology 180:485–491 13. Abdullah ND, Mathews VP (1999) Contrast issues in brain tumor imaging. Neuroimaging Clin N Am 9:733–749 14. Knopp MV, Runge VM, Essig M, Hartman M, Jansen O, Kirchin MA, Moeller A, Seeberg AH, Lodemann KP (2004) Primary and secondary brain tumors at MR imaging: bicentric intraindividual crossover comparison of gadobenate dimeglumine and gadopentetate dimeglumine. Radiology 230(1):55–64 15. Barker FG, Chang SM, Huhn SL, Davis RL, Gutin PH, McDermott MW, Wilson CB, Prados MD (1997) Age and the risk of anaplasia in magnetic resonance-nonenhancing supratentorial cerebral tumors. Cancer 80:936–941 16. Scott JN, Bracher PM, Sevick RJ, Rewcastle NB, Forsyth PA (2002) How often are nonenhancing supratentorial gliomas malignant? A population study. Neurology 59:947–949
5 Low-Grade Gliomas 17. Kotys MS, Akbudak E, Markham J, Conturo TE (2007) Precision, signal-to-noise ratio, and dose optimization of magnitude and phase arterial input functions in dynamic susceptibility contrast MRI. J Magn Reson Imaging 25:598–611 18. Aronen HJ, Gazit IE, Louis DN et al (1994) Cerebral blood volume maps of gliomas: comparison with tumor grade and histologic findings. Radiology 191:41–51 19. Lev MH, Rosen R (1999) Clinical applications of intracranial perfusion MRI imaging. Neuroimaging Clin N Am 9:309–331 20. Maia ACM, Malheiros SMF, da Rocha AJ, da Silva CJ, Gabbai AA, Ferraz APF, Stavale JN (2005) MR cerebral blood volume maps correlated with vascular endothelial growth factor expression and tumor grade in nonenhancing gliomas. AJNR Am J Neuroradiol 26:777–783 21. Law M, Oh S, Johnson G, Babb JS, Zagzag D, Golfinos J, Kelly P (2006) Perfusion magnetic resonance imaging predicts patient outcome as an adjunct to histopathology: a second reference standard in the surgical and nonsurgical treatment of low-grade gliomas. Neurosurgery 58:1099–1107 22. Provenzale JM, Wang GR, Brenner T, Petrella JR, Sorensen AG (2002) Comparison of permeability in high-grade and low grade brain tumors using dynamic susceptibility contrast MR imaging. AJR Am J Roentgenol 178:711–716 23. Bastin ME, Carpenter TK, Armitage PA, Sinha S, Wardlaw JM, Whittle IR (2006) Effects of dexamethasone on cerebral perfusion and water diffusion in patients with high-grade glioma. AJNR Am J Neuroradiol 27:402–408 24. Talos IF, Zou KH, Kikinis R, Jolesz FA (2007) Volumetric assessment of tumor infiltration of adjacent white matter based on anatomic MRI and diffusion tensor tractography. Acad Radiol 14:431–436 25. Moritani T, Ekholm S, Westessib PL (2004) Diffusionweighted MR imaging of the brain. Springer, Heidelberg, pp 161–179 26. Provenzale JM, McGraw P, Mhatre P, Alexander CG, Delong D (2004) Peritumoral brain regions in gliomas and meningiomas: investigation with isotropic diffusion weighted imaging and diffusion-tensor MR imaging. Radiology 232:451–460 27. Goebell E, Paustenbach S, Vaeterlein P, Ding XQ, Heese O, Fiehler J, Kucinski T, Hagel C, Westphal M, Zeumer H (2006) Low-grade and anaplastic gliomas: differences in architecture evaluated with diffusion-tensor MR imaging. Radiology 239:217–222 28. Budinger TF, Brennan KM (2000) Metabolic imaging. In: Berenstein M, Berger MS (eds) Neuro-oncology: the essentials. Thieme Medical Publishers, New York, pp 79–93 29. Chung JK, Kim YK, Kim S et al (2002) 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 29:176–182 30. Jacobs AH, Thomas A, Kracht LW, Li H, Dittmar C, Garlip G, Galldiks N, Klein JC, Sobesky J, Hilker R, Vollmar S, Herholz K, Wienhard K, Heiss WD (2005) 18F-fluoro-L-thymidine and 11C-methylmethionine as markers of increased transport and proliferation in brain tumors. J Nucl Med 46:1948–1958
149 31. De Witte O, Goldberg I, Wikler D et al (2001) Positron emission tomography with injection of methionine as a prognostic factor in glioma. J Neurosurg 95:746–750 32. Ribom D, Eriksson A, Hartman M et al (2001) Positron emission tomography 11C-methionine and survival in patients with low grade gliomas. Cancer 92:1541–1549 33. Cai W, Chen K, Mohamedali KA, Cao Q, Gambhir SS, Rosenblum MG, Chen X (2006) PET of vascular endothelial growth factor receptor expression. J Nucl Med 47(12):2048–2056 34. Schwartz RB, Varvalho PA, Alexander ED (1991) Radiation necrosis vs high grade recurrent glioma: differentiation by using dual-isotope SPECT with 201Tl and 99mTc-HMPAO. AJNR Am J Neuroradiol 12:1187–1192 35. Hustinix R, Alavi A (1999) SPECT and PET imaging of brain tumors. Neuroimaging Clin N Am 9:751–766 36. Cavanee WK, Furnari FB, Nagane M, Huang HJS, Newcomb EW, Bigner DD, Weller M, Berens ME, Plate KH, Israel MA, Noble MD (2000) Diffusely infiltrating astrocytomas. In: Kleihues P, Cavanee WK (eds) Pathology and genetics of tumours of the nervous system. IARC, Lyon, pp 9–21 37. Kleihues P, Davis RL, Ohgaki H, Burger PC, Westphal MM, Cavanee WK (2000) Diffuse astrocytoma. In: Kleihues P, Cavanee WK (eds) Pathology and genetics of tumours of the nervous system. IARC, Lyon, pp 22–26 38. Burger PC, Sheithauer BW, Paulus W, Szymas J, Giannini C, Kleihues P (2000) Pilocytic astrocytoma. In: Kleihues P, Cavanee WK (eds) Pathology and genetics of tumours of the nervous system. IARC, Lyon, pp 45–51 39. Kepes JJ, Louis DN, Giannini C, Paulus W (2000) Pleomorphic xanthoastrocytoma. In: Kleihues P, Cavanee WK (eds) Pathology and genetics of tumours of the nervous system. IARC, Lyon, pp 52–54 40. Wiestler OD, Lopes BS, Green AJ, Vinters HV (2000) Tuberous sclerosis complex and subependymal giant cell astrocytoma. In: Kleihues P, Cavanee WK (eds) Pathology and genetics of tumours of the nervous system. IARC, Lyon, pp 227–230 41. Taruto AL, VandenBerg SR, Rorke LB (2000) Desmoplastic infantile astrocytoma andganglioglioma. In: Kleihues P, Cavanee WK (eds) Pathology and genetics of tumours of the nervous system. IARC, Lyon, pp 99–102 42. Daumas-Duport C, Scheihauer B, O’Fallon J, Kelly P (1988) Grading of astrocytomas. A simple and reproducible method. Cancer 62:2152–2165 43. Pollack IF, Hamilton RL, Finkelstein SD, Lieberman F (2002) Molecular abnormalities and correlations with tumor response and outcome in glioma patients. Neuroimaging Clin N Am 12:627–639 44. Atlas SW, Lavi E (1996) Intraaxial brain tumors. In: Atals SW (ed) Magnetic resonance imaging of the brain and spine, 2nd edn. Lippincott-Raven, Philadelphia, pp 423–488 45. McGinnis BD, Brady TJ, New PF, Buonanno FS, Pykett IL, DeLaPaz RL, Kistler JP, Taveras JM (1983) Nuclear magnetic resonance (NMR) imaging of tumors in the posterior fossa. J Comput Assist Tomogr 7:575–584 46. Garin von Eisiedel H, Loffler W (1982) Nuclear magnetic resonance imaging of brain tumors unrevealed by CT. Eur J Radiol 2:226–234
150 47. Holland BA, Brandt-Zawadzki M, Norman D, Newton TH (1985) Magnetic resonance imaging of primary intracranial tumors: a review. Int J Radiat Oncol Biol Phys 11:315–321 48. Earnest F, Kelly PJ, Scheithauer BW, Kall BA, Cascino TL, Ehman RL, Forbes GS, Axley PL (1988) Cerebral astrocytomas: histopathologic correlation of MR and CT contrast enhancement with stereotactic biopsy. Radiology 166:823–827 49. Kelly P, Caumas-Duport C, Kispert D, Kall B, Sheihauer B, Illig J (1987) Imaging-based stereotactic serial biopsies in untreated intracranial glial neoplasms. J Neurosurg 66:865–874 50. Greene G, Hitchon P, Schelper R, Yuh W, Dyste G (1989) Diagnostic yield in CT guided stereotactic biopsy of gliomas. J Neurosurg 71:494–497 51. Swanson KR, Alvord EC, Murray JD (2000) A quantatative model for differential motility of gliomas in grey and white matter. Cell Prolif 33:317–329 52. Marks JE, Gado M (1977) Serial computed tomography of primary brain tumors following surgery, irradiation, and chemotherapy. Radiology 125:119–125 53. Holland BA, Kucharcyzk A, Brandt-Zawadzki M, Norman D, Haas DK, Harper PS (1985) MR imaging of calcified intracranial lesions. Radiology 157:353–356 54. Mariani L, Siegenthaler P, Guzman R, Friedrich D, Fathi AR, Ozdoba C, Weis J, Ballinari P, Seiler RW (2004) The impact of tumour volume and surgery on the outcome of adults with supratentorial WHO grade II astrocytomas and oligoastrocytomas. Acta Neurochir (Wien) 146:441–448 55. Altman NR, Purser RK, Post MJD (1988) Tuberous sclerosis: characteristics at CT and MR imaging. Radiology 167:527–532 56. Fuss M, Wenz F, Essig M, Muenter M, Debus J, Herman TS, Wannenmacher M (2001) Tumor angiogenesis of low-grade astrocytomas measured by dynamic susceptibility contrastenhanced MRI (DSC-MRI) is predictive of local tumor control after radiation therapy. Int J Radiat Oncol Biol Phys 51(2):478–482 57. Earnest F, Kelly PJ, Sheithauer BW (1988) Cerebral asrtocytomas: histopathologic correlation of MR and CT contrast enhancement with stereotactic biopsy. Radiology 166: 823–827 58. Dvorak HF, Brown LF, Detmar M, Dvorak AM (1995) Vascular permeability factor/vascular endothelial growth factor, microvascular hyperpermeability and angiogenesis. Am J Pathol 146:029–1039 59. Panigrahy A, Krieger MD, Gonzalez-Gomez I, Liu X, McComb JG, Finlay JL, Nelson MD Jr, Gilles FH, Blüml S (2006) Quantitative short echo time 1H-MR spectroscopy of untreated pediatric brain tumors: preoperative diagnosis and characterization. AJNR Am J Neuroradiol 27:560–572 60. Poussaint TY, Rodriguez D (2006) Advanced neuroimaging of pediatric brain tumors: MR diffusion, MR perfusion, and MR spectroscopy. Neuroimaging Clin N Am 16:169–192 61. Guitierrez JA (1999) Classification and pathobiology of lowgrade glial and glioneuronal neoplasms. In: Rock JP, Rosenblum ML, Shaw EG, Caincross JG (eds) The practical management of low-grade primary brain tumors. Lippincott Williams & Wilkins, Philadelphia, pp 33–67
G.A. Christoforidis et al. 62. Lee YY, Van Tassel P, Bruner JM, Moser RP, Share JC (1989) Juvenile pilocytic astrocytomas: CT and MR characteristics. AJR Am J Roentgenol 10:363–370 63. Coakley KJ, Huston J, Sheithauer BW, Forbes G, Kelly PJ (1995) Pilocytic astrocytomas: well-demarcated magnetic resonance appearance despite frequent infiltration histologically. Mayo Clin Proc 70:747–751 64. Arai K, Sato N, Aoki J, Yagi A, Taketomi-Takahashi A, Morita H, Koyama Y, Oba H, Ishiuchi S, Saito N, Endo K (2006) MR signal of the solid portion of pilocytic astrocytoma on T2-weighted images: is it useful for differentiation from medulloblastoma? Neuroradiology 48(4): 233–237 65. Grand SD, Tropres IM, Chabardes SJ et al (2007) Perfusionsensitive imaging of pilocytic astrocytomas: initial results. Neuroradiology 49:545–550 66. Abel TJ, Chowdhary A, Thapa M, Rutledge JC, Geyer JR, Ojemann J, Avellino AM (2006) Spinal cord pilocytic astrocytoma with leptomeningeal dissemination to the brain. Case report and review of the literature. J Neurosurg 105(6 Suppl):508–514 67. Komotar RJ, Mocco J, Zacharia BE, Wilson DA, Kim PY, Canoll PD, Goodman RR (2006) Astrocytoma with pilomyxoid features presenting in an adult. Neuropathology 26(1):89–93 68. Tihan T, Fisher PG, Kepner JL, Godfraind C, McComb RD, Goldthwaite PT, Burger PC (1999) Pediatric astrocytomas with monomorphous pilomyxoid features and a less favorable outcome. J Neuropathol Exp Neurol 58(10): 1061–1068 69. Arslanoglu A, Cirak B, Horska A, Okoh J, Tihan T, Aronson L, Avellino AM, Burger PC, Yousem DM (2003) MR imaging characteristics of pilomyxoid astrocytomas. AJNR Am J Neuroradiol 24(9):1906–1908 70. Cirak B, Horska A, Barker PB, Burger PC, Carson BS, Avellino AM (2005) Proton magnetic resonance spectroscopic imaging in pediatric pilomyxoid astrocytoma. Childs Nerv Syst 21(5):404–409 71. Allegranza A, Ferrari S, Brurone M et al (1991) Cerebromeningeal pleomorphic xanthoastrocytoma: report of four cases: clinical radiologic and pathologic features (including a case with malignant evolution). Neurosurg Rev 14:43–49 72. Hosokawa Y, Tsuchihashi V, Okabe H et al (1991) Pleomorphic xanthoastrocytoma: ultrastructual, immumohistochemical and DNA cytoflurometric study of a case. Cancer 68:853–859 73. Buciero A, De Caro M, De Stefano V et al (1997) Pleomorphic xanthoastrocytoma: clinical, imaging and pathological features of four cases. Clin Neurol Neurosurg 99:40–45 74. Tien RD, Cardena CA, Rajagopalan S et al (1992) Pleomorphic xanthoastrocytoma of the brain: MR findings in six patients. AJR Am J Roentgenol 159:1287–1290 75. Rosenberg S, Rotta JM, Yassuda A et al (2000) Pleomorphic xanthoastrocytoma of the cerebellum. Clin Neuropathol 19(5):238–242 76. Brown JH, Chew FS (1993) Pleomorphic xanthoastrocytoma. AJR Am J Roentgenol 160:1972 77. Yoshino MT, Lucio R (1992) Pleomorphic xanthoastrocytoma. AJR Am J Roentgenol 13:1330–1332
5 Low-Grade Gliomas 78. Herpers MJHM, Freling G, Beuls EAM (1994) Pleomorphic xanthoastrocytoma in the spinal cord. J Neurosurg 80: 564–569 79. Weldon-Linne CM, Victor TA, Groothuis DR, Vick NA (1983) Pleomorphic xanthoastrocytoma: ultrastructural and immunohistochemical study of a case with a rapidly fatal outcome following surgery. Cancer 52:2055–2063 80. Zulch KJ (1986) Brain tumors: their biology and pathology, vol. 15, 3rd edn. Springer, Berlin, pp 210–341 81. Russel DS, Rubinstein LJ (1989) Pathology of tumors of the nervous system, vol. 3, 5th edn. Lippincott Williams & Wilkins, Baltimore, pp 83–350 82. Rippe DJ, Boyko OB, Radi M, Worth R, Fuller GN (1992) MRI of temporal lobe pleomorphic xanthoastrocytoma. J Comput Assist Tomogr 16:856–859 83. Keppes JJ (1993) Pleomorphic xanthoastrocytoma: the birth of a diagnosis and a concept. Brain Pathol 3:269–274 84. Keppes JJ, Louis DN, Paulus W (1997) Pleomorphic xanthoastrocytoma In Kleihues P and Cavenee W (ed), Pathology and genetics of tumours of the nervous system Lyon, France: International Agency for Research on Cancer. pp 34-36 85. Burger PC, Scheithauer BW, Vogel FS (1991) Surgical pathology of the nervous system and its coverings, 3rd edn. Churchill Livingstone, New York. pp 193–324 86. Burger PC, Scheithauer BW (1994) Tumors of the central nervous system. In: Rosai J (ed) Atlas of tumor pathology. Armed Forces Institute of Pathology, Washington, DC, pp 25–77 87. Toun JC, Paulus W, Warthmuth-Metz M et al (1997) Pleomorphic xanthoastrocytoma: report of six cases with special consideration of diagnostic and therapeutic pitfalls. Surg Neurol 47:162–169 88. Osborn AG (1994) Brain tumors and tumor like masses: classification and differential diagnosis in diagnostic neuroradiology. Mosby, St Louis 89. Blom RJ (1988) Pleomorphic xanthoastrocytoma: CT appearance. J Comp Assist Tomogr 12:31–352 90. Petropoulou K, Whiteman ML, Altman NR, Bruce J, Morrison G (1995) CT and MRI of pleomorphic xanthoastrocytoma: unusual biologic behavior. J Comput Assist Tomogr 19:860–865 91. Luh GY, Bird CR (1999) Imaging of brain tumors in the pediatric population. Neuroimaging Clin N Am 9(4):691–716 92. Lipper MH, Gherhard DA, Phillips CD et al (1993) Pleomorphic xanthoastrocytoma, a distinctive astroglial tumor: neuroradiologic and pathologic features. AJNR Am J Neuroradiol 14(6):1397–1404 93. Turgut M, Akalan N, Ozgen T, Ruacan S, Erbengi A (1996) Subependymal giant cell astrocytoma associated with tuberous sclerosis: diagnostic and surgical characteristics of five cases with unusual features. Clin Neurol Neurosurg 98:217–221 94. Hahn JS, Bejar R, Gladson CL (1991) Neonatal subependymal giant cell astrocytoma with tuberous sclerosis: MRI, CT, and ultrasound correlation. Neurology 41: 124–128 95. Nabbout R, Santos M, Rolland Y, Delalande O, Dulac O, Chiron C (1999) Early dignosis of subependymal giant cell astrocytoma in children with tuberous sclerosis. J Neurol Neurosurg Psychiatry 66:370–375
151 96. Jeong MG, Chung TS, Coe CJ, Jeon TJ, Kin DI, Joo AY (1997) Application of magnetization transfer imaging for intracranial lesions of tuberous sclerosis. J Comput Assist Tomogr 21:8–14 97. Jelenik J, Smirniotopoulos JG, Parisi JE, Kanzer M (1990) Lateral ventricular neoplasms of the brain: differential diagnosis based on clinical, CT, and MR findings. Am J Neuroradiol 11:567–574 98. Smirniotopoulos JG et al (1999) Image interpretation session. Radiographics 19(1):226–229 99. Bailey P, Cushing H (1926) A classification of the tumors of the glioma group on a histogenetic basis, with correlative study of prognosis, 1st edn. Lippincott Williams & Wilkins, Philadelphia 100. Bruner JA (1987) Oligodendroglioma: diagnosis and prognosis. Semin Diagn Pathol 4(3):251–261 101. Bruner JM (1994) Neuropathology of malignant gliomas. Semin Oncol 21(2):126–138 102. Ludwin SK (1997) The pathobiology of the oligodendrocyte. J Neuropathol Exp Neurol 56(2):11–23 103. Kros JM, Jong AW, Kwars TH (1992) Ultrastructural characterization of oligodendroglioma. J Neuropathol Exp Neurol 51(2):186–193 104. De la Monte S (1989) Uniform lineage of oligodendrogliomas. Am J Pathol 135(3):529–540 105. Herpers MJH, Budka H (1984) GFAP in oligodendroglial tumors: gliofibrillary oligodendroglioma and transitional oligoastrocytoma as subtypes of oligodendroglioma. Acta Neuropathol 64:265–272 106. Kros JM, Schouten PJ, Ja J et al (1996) Proliferation of gemistocytic cells and glial fibrillary acidic protein (GFAP)positive oligodendroglial cells in gliomas: a MIB-1/GFAP double labeling study. Acta Neuropathol 91:99–103 107. Smith MT, Ludwig CL, Godfrey AD et al (1983) Grading of oligodendroglioma. Cancer 52:2107–2114 108. Kros JM, Lie S, Stefanko S (1994) Familial occurrence of polymorphous oligodendroglioma. Neurosurgery 34(4): 732–736 109. Kros JM, Van Eden CG, Stefanko SZ et al (1990) Prognostic implications of GFAP cell types in oligodendrogliomas. Cancer 66:1204–1212 110. Osawara H, Kiya K et al (1990) Multiple oligodendroglioma: case report. Neurol Med Chir 30:127–131 111. Macdonald DR, Brien RA, Gilbert JJ et al (1989) Metastatic oligodendroglioma. Neurology 39:1593–1596 112. Bishop M, de la Monte SM (1989) Dual lineage of astrocytomas. Am J Pathol 135:517–523 113. Rubinsteion LJ (1989) Glioma cytogeny and differentiation viewed through the window of neoplastic vulnerability. In: Salomon M (ed) Neurobiology of brain tumors. Lippincott Williams & Wilkins, Baltimore, pp 35–47 114. Min KW, Scheithauer BW (1994) Oligodendroglioma: ultrastructural spectrum. Ultrastruct Pathol 18:47–60 115. Bullard DE, Rawlings CE, Phillips B et al (1987) Oligodendroglioma analysis of the value of radiation therapy. Cancer 60:2179–2188 116. Kros JM (1995) Oligodendrogliomas: clinicopathologic correlations. J Neurooncol 24:29–31 117. Burger PC, Rawling CE, Cox ER et al (1987) Clinicopathologic correlation in the oligodendroglioma. Cancer 59:1345–1352
152 118. Kros JM, Pieterdman H et al (1994) Oligodendroglioma: the Rotterdam’Dijkzigt experience. Neurosurgery 34:959–965 119. Mork SJ, Halvorsen TB, Lindegaard KF et al (1986) Oligodendroglioma. Histologic evaluation and prognosis. J Neuropathol Exp Neurol 45:65–78 120. Shaw EG, Scheithauer BW, O’Falllon JR et al (1992) Oligodendrogliomas: the Mayo clinic experience. J Neuro surg 76:428–434 121. Mork SJ, Lindergard KF, Halvoresen TB et al (1985) Oligodendroglioma incidence and biologic behavior in a defined population. J Neurosurg 63:881–889 122. Kros JM, Troost D et al (1988) Oligodendroglioma comparison of two grading systems. Cancer 61:2251–2259 123. Kros JM, Hop WC et al (1996) Prognostic value of proliferation related antigen Ki-067 in oligodendrogliomas. Cancer 78:1107–1113 124. Shimizu KT, Tran LM, Mark RJ et al (1993) Management of oligodendrogliomas. Radiology 186:569–572 125. Tice H, Barnes PD, Goumnerova L et al (1993) Pediatric and adolescent oligodendrogliomas. AJNR Am J Neuroradiol 14:1293–1300 126. Daumas-Dupont C, Varlet P, Tucker ML, Beuvon F, Cervera P, Chodkiewicz JP (1997) Oligodendrogliomas, part I. Patterns of growth, histological diagnosis, clinical and imaging correlations: a study of 153 cases. J Neurooncol 34:37–59 127. Daumas-Dupont C, Tucker ML, Kolles H, Cervera P, Beuvon F, Varlet P, Koziak M, Chodkiewicz JP (1997) Oligodendrogliomas, part II. A new grading system based on morphological and imaging criteria. J Neurooncol 34:61–78 128. Kros RG, JM BPC, Louis DN, Collins VP (2000) Oligodendroglioma. In: Kleihues P, Cavanee WK (eds) Pathology and genetics of tumours of the nervous system. IARC, Lyon, pp 56–61 129. Kros RG, JM BPC, Louis DN, Collins VP (2000) Anaplastic oligodendroglioma. In: Kleihues P, Cavanee WK (eds) Pathology and genetics of tumours of the nervous system. IARC, Lyon, pp 62–64 130. Donliskas CA, Simeone FA (1987) CT characteristics of intraventricular oligodendrogliomas. AJNR Am J Neuro radiol 8:1077–1082 131. Packer RJ, Sutton LN, Rorke LB et al (1985) Oligodendroglioma of the posterior fossa in childhood. Cancer 56:195–199 132. Vonafakos D, Marcu H, Hacker H (1979) Oligodendrogliomas: CT patterns with emphasis on features indicating malignancy. J Comput Assist Tomogr 3:783–789 133. Lee YY, Tassel PV (1989) Intracranial oligodendrogliomas imaging findings in 35 untreated cases. AJR Am J Roentgenol 10:119–127 134. Christoforidis GA, Mehta, BA, Agrawal R, Georganos SA, Patel, SC (1997) Imaging of Oligodendrogliomas. ASNR Proceedings 361 135. Atlas SW, Grossman RI, Hackney DB (1988) Calcified intracranial lesions: detection with gradient-echo-acquisition rapid MR imaging. AJR Am J Roentgenol 9:253–259 136. Lev MH, Ozsunar Y, Henson JW et al (2004) Glial tumor grading and outcome prediction using dynamic spin-echo MR susceptibility mapping compared with conventional contrast-enhanced MR: confounding effect of elevated
G.A. Christoforidis et al. rCBV of oligodendrogliomas [corrected]. AJNR Am J Neuroradiol 25(2):214–221. Erratum in: AJNR Am J Neuroradiol 25(3):B1 137. Cha S, Tihan T, Crawford F et al (2005) Differentiation of low-grade oligodendrogliomas from low-grade astrocytomas by using quantitative blood-volume measurements derived from dynamic susceptibility contrast-enhanced MR imaging. AJNR Am J Neuroradiol 26(2):266–273 138. Leaver HA, Whittle IR, Wharton SB, Ironside JW (1998) Apoptosis in human primary brain tumors. Br J Neurosurg 12:539–546 139. Aguzzi A, Brandner S, Paulus W (2000) Choroid plexus tumors. In: Kleihues P, Cavanee WK (eds) Pathology and genetics of tumours of the nervous system. IARC, Lyon, pp 84–86 140. Chan RC, Thompson GB, Durity FA (1983) Primary choroid plexus papilloma of the cerebellopontine angle. Neurosurgery 12:334–336 141. Ellenbogen RG, Winston KR, Kupsky WJ (1989) Tumors of the choroid plexus in children. Neurosurgery 25: 327–335 142. Johnson DL (1989) Management of choroid plexus tumors in children. Pediatr Neurosci 15:195–206 143. Ken JG, Sobel DF, Copeland B et al (1991) Choroid plexus papillomas of the foramen of Luschka: MR appearance. AJNR Am J Neuroradiol 12:1201–1202 144. Milhorat TH, Hammock MK, Davis DA et al (1976) Choroid plexus papilloma. I. Proof of cerebrospinal fluid overproduction. Childs Brain 2:273–289 145. Fitz CR, Rao KLVG (1983) Primary tumors in children. In: McGraw RP, Boynton SD, Cowell MW (eds) Cranial computed tomography. Mc Graw-Hill, New York, pp 329–332 146. Doran SE, Blaivas M, Dauser RC (1995) Bone formation within a choroids plexus papilloma. Pediatr Neurosurg 23:216–218 147. Miettinen M, Clark R, Virtanen I (1986) Intermediate filament proteins in choroid plexus and ependyma and their tumors. Am J Pathol 123:231–240 148. Ang LC, Taylor AR, Bergin D et al (1990) An immunohistochemical study of papillary tumors in the central nervous system. Cancer 65:2712–2719 149. Guermazi A, De Kerviler E, Sagdanski AM, Frija J (2000) Diagnostic imaging of choroid plexus disease. Clin Radiol 55:503–516 150. Tasdemiroglu E, Awh MH, Walsh JW (1996) MRI of cerebellopontine angle choroid plexus papilloma. Neuro radiology 38:38–40 151. Sahar A, Feinsod M, Beller AJ (1980) Choroid plexus papilloma: hydrocephalus and cerebrospinal fluid dynamics. Surg Neurol 13:476–478 152. Sarkar C, Sharma MC, Gaikwad S, Sharma C, Singh VP (1999) Choroid plexus papilloma: a clinicopathological study of 23 cases. Surg Neurol 52:30–36 153. Wagle V, Melanson D, Ethier R, Bertrand G, Villemure JG (1987) Choroid plexus papilloma: magnetic resonance, computed tomography, and angiographic observations. Surg Neurol 27:466–468 154. Furuya K, Sasaki T, Saito N, Atsuchi M, Kirino T (1995) Primary large choroid plexus papillomas in the cerebellopontine angle: radiological manifestations and surgical management. Acta Neurochir 136:144–149
5 Low-Grade Gliomas 155. Tacconi L, Delfini R, Cantore G (1996) Choroid plexus papillomas: consideration of a surgical series of 33 cases. Acta Neurochir 138:802–810 156. D’Addario V, Pinto V, Meo F, Resta M (1998) The specificity of ultrasound in the detection of fetal intracranial tumors. J Perinat Med 26:480–485 157. Schellhas KP, Siebert R, Heithoff KB, Franciosi RA (1988) Congenital choroid plexus papilloma of the third ventricle: diagnosis with real time sonography and MR imaging. AJNR Am J Neuroradiol 9:797–798 158. Berger C, Thiesse P, Lellouch-Tubiana A et al (1998) Choroid plexus carcinomas in childhood: clinical features and prognostic factors. Neurosurgery 42:470–475 159. Krieger MD, Panigrahy A, McComb JG, Nelson MD, Liu X, Gonzalez-Gomez I, Gilles F, Bluml S (2005) Differentiation of choroid plexus tumors by advanced magnetic resonance spectroscopy. Neurosurg Focus 18(6A):E4 160. Sunada I, Tsuyuguchi N, Hara M, Ochi H (2002) 18F-FDG and 11C-methionine PET in choroid plexus papilloma – report of three cases. Radiat Med 20:97–100 161. Wiestlerr OD, Schiffer D, Coons SW, Prayson RA, Rosenblum MK (2000) Ependymoma. In: Kleihues P, Cavanee WK (eds) Pathology and genetics of tumours of the nervous system. IARC, Lyon, pp 72–76 162. Spoto GP, Press GA, Hesselink JR, Solomon M (1990) Intracranial ependymoma and subependymoma: MR manifestations. AJR Am J Neuroradiol 11:83–91 163. Tortori-Donati P, Fondelli MP, Cama A, Garre ML, Rossi A, Andreussi L (1995) Ependymomas of the posterior cranial fossa: CT and MRI findings. Neurodiology 37:238–243 164. Oppenheim JS, Strauss RC, Mormino J, Sachdev RAS (1994) Ependymomas of the third ventricle. Neurosurgery 34:350–353 165. Kudo H, Oi S, Tamaki N, Nishida Y, Matsumoto S (1990) Ependymoma diagnosed in the first year of life in Japan in collaboration with the International Society for Pediatric Neurosurgery. Childs Nerv Syst 6:375–378 166. Armington WG, Osborn AG, Gubberley AD et al (1985) Supratentorial ependymoma: CT appearance. Radiology 157:367–372 167. Naidich TP, Zimmerman RA (1984) Primary brain tumors in children. Semin Roentgenol 19:100–114 168. Rawlings CE, Giangaspero BPC et al (1988) Ependymomas: a clinicopathological study. Surg Neurol 29:271–281 169. Kawano N, Yada K, Yagishita S (1989) Clear cell ependymoma: a histological variant with diagnostic implications. Virchows Arch 415:467–472 170. Swartz JD, Zimmerman RA, Billaniuk LT (1982) Computed tomography of intracranial ependymomas. Radiology 143: 97–101 171. Lizak PF, Woodruff WW (1992) Posterior fossa neoplasms: multiplanar imaging. Semin Ultrasound CT MR 13:182–206 172. Schott LH, Naidich TP, Gan J (1983) Common pediatric brain tumors: typical computed tomographic appearance. J Comput Tomogr 7:3–15 173. Spoto GP, Press GA, Hesselnik JR et al (1990) Intracranial ependymoma and subependymoma: MR manifestations. AJNR Am J Neuroradiol 11:83–91 174. Furie DM, Provenzale JM (1995) Supratentorial ependymomas and subependymomas: CT and MR appearance. J Comput Assist Tomogr 19:518–526
153 175. Courville CB, Broussalian SL (1961) Plastic ependymomas of the lateral recess. Report of eight verified cases. J Neurosurg 18:792 176. Lee SH, Rao KCVG (1983) Primary tumors in adults. In: Lee SH, Rao KCVG (eds) Cranial computed tomography. McGraw-Hill, New York, pp 241–293 177. Daumas-Dupont C, Scheihauer BW, Chodkiewicz JP et al (1988) Dysembryoplastic neuroepithelial tumor: a surgically curable tumor of young patients with intractable partial seizures: report of thirty-nine cases. Neurosurgery 23:545–556 178. Raymond AA, Halpin SFS, Alsanjari N et al (1994) Dysembryoplastic neuroepithelial tumour: features in 16 patients. Brain 117:461–475 179. Reiche W, Kolles H, Eymann R, Feiden W (1996) Dysembryoplastic neuroepithelial tumor (DNT): pattern of neuroradiologic findings. Radiologe 36:884–889 180. Koeller KK, Dillon WP (1992) Dysembryoplastic neuroepithelial tumors: MR appearance. AJNR Am J Neuroradiol 13:1319–1325 181. Kuroiwa T, Bergey GK, Rothman MI et al (1995) Radiologic appearance of the dysembryoplastic neuroepithelial tumor. Radiology 197:233–238 182. Lemsle M, Borsotti JP, Justrabo E et al (1996) Dysembryoplastic neuroepithelial tumors: a benign tumor cause of partial epilepsy in young adults. Rev Neurol 152:451–457 183. Leung SY, Gwi E, Ng HK et al (1994) Dysembryoplastic neuroepithelial tumor: a tumor with small neuronal cells resembling oligodendroglioma. Am J Surg Pathol 18:604–614 184. Abe M, Tabuchi K, Tsuji T et al (1995) Dysembryoplastic neuroepithelial tumor: report of three cases. Surg Neurol 43:240–245 185. Kuroiwa T, Kishikawa T, Kato A et al (1994) Dysembryoplastic neuroepithelial tumors: MR findings. J Comput Assist Tomogr 18:352–356 186. Taratuto AL, Pomata H, Sevlever G et al (1995) Dysembryoplastic neuroepithelial tumor: morphological, immunocytochemical, and deoxyribonucleic acid analyses in a pediatric series. Neurosurgery 36:474–481 187. Di Constanzo A, Scarabino T, Trojsi F et al (2006) Multiparametric 3T MR approach to the assessment of cerebral gliomas: tumor extent and malignancy. Neuro radiology 48:622–631 188. Rollin N, Guyotata J, Streichenberger N et al (2006) Clinical relevance of diffusion and perfusion magnetic resonance imaging in assessing intra-axial brain tumors. Neuroradiology 48:150–159 189. Ostertun B, Wolf HK, Campos MG et al (1996) Dysembryoplastic neuroepithelial tumors: MR and CT evaluation. AJNR Am J Neuroradiol 17:419–430 190. Bulakbasi N, Kacaoglu M, Ors F et al (2003) Combination of single voxel proton MR spectroscopy and apparent diffusion coefficients calculation in the evaluation of common brain tumors. AJNR Am J Neuroradiol 24:225–233 191. Daumas-Duport C, Pietsch T, Lantos PL (2000) Dysembryoblastic neuroepithelial tumor. In: Cavenee WK, Kleihues P (eds) WHO classification of tumours: pathology and genetics of tumours of the nervous system. IARC, Lyon, pp 103–106
154 192. Wang L, Li K, Chen L et al (2005) Perfusion MR imaging and proton MR spectroscopy in a case of dysembryoplastic neuroepithelial tumor. Chin Med J 118: 1134–1136 193. Castillo M, Smith JK, Kwock L (2000) Correlation of myoinositol levels and grading of cerebral astrocytoma. AJNR Am J Neuroradiol 21:1645–1649 194. Lee DY, Chung CK, Hwang YS, Choe G, Chi JG, Kim HJ, Cho BK (2000) Dysembryoplastic neuroectodermal tumor: radiological findings (including PET, SPECT, and MRS) and surgical strategy. J Neurooncol 47:167–174 195. Prayson RA (1999) Composite ganglioglioma and dysembryoplastic neuroectodermal tumor. Arch Pathol Lab Med 123:247–250 196. Hirose T, Sheithauer BW (1998) Mixed dysembryoplastic neuroectodermal tumor and ganglioglioma. Acta Neuropathol 95:649–654 197. Scheinker IM (1945) Subependymoma: an unrecognized tumor of subependymal derivation. J Neurosurg 2:232–240 198. Chiechi MV, Smirniotopoulos JG, Jones RV (1995) Intracranial subependymomas: CT and MR imaging features in 24 cases. AJR Am J Roentgenol 165:1245–1250 199. Hoeffel C, Boukobza M, Polivka M et al (1995) MR manifestations of subependymomas. AJNR Am J Neuroradiol 16:2121–2129 200. Jooma R, Torrens MJ, BradshawJ BB (1985) Subepe ndymomas of the fourth ventricle: surgical treatment in 12 cases. J Neurosurg 62:508–512 201. Labato RD, Sarabia M, Castro S et al (1986) Symptomatic subependymoma: report of four new cases studied with computed tomography and review of the literature. Neurosurgery 19:594–598 202. Lombardi D, Scheithauer BW, Meyer FB et al (1991) Symptomatic subependymoma: a clinicopathological and flow cytometric study. J Neurosurg 75:585–588 203. Kim OG, Han MH, Lee SH et al (1993) MRI of intracranial subependymoma: report of a case. Neuroradiology 35: 185–186 204. Maiuri F, Gangemi M, Iaconetta G, Signorelli F, Del Basso De Caro M (1997) Symptomatic subependymomas of the lateral ventricles. Report of eight cases. Clin Neurol Neurosurg 99:17–22 205. Silverstein JE, Lenchik L, Stanciu MG, Shimkin PM (1995) MRI of intracranial subependymomas. J Comput Assist Tomogr 19:264–267 206. Furie DM, Provenzale JM (1995) Supratentorial ependymomas and subependymomas: CT and MR appearance. J Comput Assist Tomogr 19:518–526 207. Nishio S, Morioka T, Mihara F, Fukui M (2000) Subependymoma of the lateral ventricles. Neurosurg Rev 23:98–103 208. Smith NM, Carli MM, Hanieh A et al (1992) Gangliogliomas in childhood. Childs Nerv Syst 8:238–242 209. Miller DC, Lang FF, Epstein FJ (1993) Central nervous system gangliogliomas part 1: pathology. J Neurosurg 79:859–866 210. Castillo M, Davis PC, Takei Y, Hoffman JCJ (1990) Intracranial ganglioglioma: MR, CT, and clinical findings in 18 patients. AJNR Am J Neuroradiol 11:109–114 211. Demierre B, Stichnoth FA, Hori A, Spoerri O (1986) Intracerebral ganglioglioma. J Neurosurg 65:177–182
G.A. Christoforidis et al. 212. Chintagumpala MM, Armstrong D, Miki S et al (1996) Mixed neuronal-glial tumors (gangliogliomas) in children. Pediatr Neurosurg 24:306–313 213. Lang FF, Epstein FJ, Ransohoff J et al (1993) Central nervous system gangliogliomas part 2: clinical outcome. J Neurosurg 79:867–873 214. Hirose T, Scheithauer BW, Lopes MBS et al (1997) Ganglioglioma an intrastructural and immunohistochemical study. Cancer 79:989–1003 215. Morris HH, Estes ML, Gilmore R et al (1993) Chronic intractable epilepsy as the only symptom of primary brain tumor. Epilepsia 34:1038–1043 216. Zentner J, Wolf HK, Ostertum B et al (1994) Gangliogliomas: clinical, radiological, and histopathological findings in 51 patients. J Neurol Neurosurg Psychiatry 57:1497–1502 217. Dorne HL, O’Gorman AM, Melanson D (1986) Computed tomography of intracranial gangliogliomas. AJNR Am J Neuroradiol 7:281–285 218. Tampieri D, Moumdjian R, Melanson D, Ethier R (1991) Inracerebral gangliogliomas in patients with partial complex seizures: CT and MR imaging findings. AJNR Am J Neuroradiol 12:749–755 219. Lu WY, Goldman M, Young B, Davis DG (1993) Optic nerve ganglioglioma. J Neurosurg 78:979–982 220. Matsumoto K, Tamiya T, Ono Y, Furuta T, Asari S, Ohmoto T (1999) Cerebral gangliogliomas: clinical characteristics, CT and MRI. Acta Neurochir (Wien) 141:135–141 221. Majos C, Aguilera C, Ferrer I, Lopez L, Pons LC (1998) Intraventricular ganglioglioma: case report. Neuroradiology 40:377–379 222. Al-Sarraj ST (1997) Multifocal neurocytoma/ganglioglioma. Am J Surg Pathol 21:258–259 223. Wolf HK, Muller MB, Spanle M et al (1994) Ganglioglioma: a detailed histopathological and immunohistochemical analysis of 61 cases. Acta Neuropathol 88:166–173 224. Luo CB, Teng MM, Chen SS, Lirng JF, Guo WY, Lan GY, Chang T (1997) Intracranial ganglioglioma: CT and MRI findings. Kaohsiung J Med Sci 13:467–474 225. Provenzale JM, Ali U, Barboriak DP, Kallmes DF, Delong DM, McLendon RE (2000) Comparison of patient age with MR imaging features of gangliogliomas. AJR Am J Roentgenol 174:859–862 226. Johannsson JH, Rekate HL, Roessmann U (1981) Gangliogliomas: pathological and clinical correlation. J Neurosurg 54:58–63 227. Wacker MR, Cogen PH, Etzell JE et al (1992) Diffuse leptomeningeal involvement by a gangliogliomas in a child. J Neurosurg 77:302–306 228. Batltuch GH, Farmer JP, Meagher-Villemure K et al (1993) Ganglioglioma presenting as a vascular lesion in a 10-year old boy. J Neurosurg 79:920–923 229. Law M, Meltzer DE, Wetzel SG et al (2004) Conventional MR imaging with simultaneous measurements of cerebral blood volume and vascular permeability in ganglioglioma. Magn Reson Imaging 22:599–606 230. Kumabe T, Shimizu H, Sonoda Y, Shirane R (1999) Thallium-201 single-photon emission computed tomographic and proton magnetic resonance spectroscopic characteristics of intracranial ganglioglioma: three technical case reports. Neurosurgery 45:183–187
5 Low-Grade Gliomas 231. Im SH, Chung CK, Cho BK et al (2002) Intracranial ganglioglioma: preoperative characteristics and oncologic outcome after surgery. J Neurooncol 59:173–183 232. Provenzale JM, Arata MA, Turkington TG, McLendon RE, Coleman RE (1999) Gangliogliomas: characterization by registered positron emission tomography – MR images. AJR Am J Roentgenol 172:1103–1107 233. Altman NR (1982) MR and CT characteristics of gangliocytoma a rare cause of epilepsy in children. AJNR Am J Neuroradiol 9:917–921 234. Furie DM, Felsberg GJ, Tien RD, Friedman HS et al (1993) MRI of gangliocytoma of cerebellum and spinal cord. J Comput Assist Tomogr 17:488–491 235. Sherazi ZA (1998) Gangliocytoma: magnetic resonance imaging characteristics. Singapore Med J 39:373–375 236. Schorner W, Meencke HJ, Felix R (1987) Temporal lobe epilepsy: comparison of CT and MR imaging. AJNR Am J Neuroradiol 8:773–781 237. Ormson MJ, Kispert DB, Sharbrough FW et al (1985) Cryptic structural lesions in refractory partial epilepsy: MR imaging and CT studies. Radiology 160:215–219 238. Kawamoto K, Yamanouchi Y, Suwa J et al (1985) Ultrastructural study of a cerebral gangliocytoma. Surg Neurol 24:541–549 239. Izukawa D, Lach B, Benoit B (1988) Gangliocytoma of the cerebellum: ultrastructure and immunohistochemistry. Neurosurgery 22:576–581 240. Towfighi J, Salam MM, McLendon RE et al (1996) Ganglion cell-containing tumors of the pituitary gland. Arch Pathol Lab Med 120:369–377 241. Puchner MJA, Ludecke DK, Saeger W et al (1995) Gangliocytomas of the sellar region – a review. Exp Clin Endocrinol 103:129–149 242. McCowen KC, Glickman JN, Black PM, Zervas NT, Lidov HG, Garber JR (1997) Gangliocytoma masquerading as a prolactinoma. Case report. J Neurosurg 91:490–495 243. Kulkhantrakorn K, Awwad EE, Levy B et al (1997) MRI in Lhermitte–Duclos disease. Neurology 48:725–731 244. Meltzer CC, Smirniotopoulos JG, Jones RV (1995) The striated cerebellum: an MR imaging sign in Lhermitte– Duclos disease (dysplastic gangliocytoma). Radiology 194:699–703 245. da Silva AAD, Banerjee T, Coimbra RLM (1996) Lhermitte–Duclos disease (cerebellar gangliocytoma). South Med J 89:1208–1212 246. Richieri-Costa A, Frederigue U, Guion-Almelda ML (1993) Holoprosencephaly, hamartomatous growth of the cerebrum, dysplastic gangliocytoma of the cerebellum, unique brain anomalies, and renal agenesis in a Brazilian infant born to a diabetic mother: a clinical and pathologic study. Birth Defects Orig Artic Ser 29:389–394 247. Lindboe CF, Helseth E, Myhr G (1995) Lhermitte–Duclos disease and giant meningiomas as manifestations of Cowden’s disease. Clin Neuropathol 14:327–330 248. Thomas DW, Lewis MA (1995) Lhermitte–Duclos disease associated with Cowden’s disease. Int J Oral Maxillofac Surg 24:369–371 249. Rimbau J, Isamat F (1994) Dysplastic gangliocytoma of the cerebellum (Lhermitte–Duclos disease) and its relation to the multiple hamartoma syndrome (Cowden disease). J Neurooncol 18:191–197
155 250. Smith RR, Grossman RI, Goldberg HI et al (1987) MR imaging of Lhermitte–Duclos disease: a case report. AJNR Am J Neuroradiol 10:187–189 251. Marcus CD, Galeon M, Peruzzi P et al (1996) Lhermitte– Duclos disease associated with syringomyelia. Neuro radiology 38:529–531 252. Awwad EE, Levy E, Martin DS, Merenda GO (1995) Atypical MR appearance of Lhermitte–Duclos disease with contrast enhancement. AJNR Am J Neuroradiol 16: 1719–1720 253. Williams DW III, Elster AD, Ginsberg LE, Stanton C (1992) Recurrent Lhrmitte–Duclos disease: report of two cases and association with Cowden’s disease. AJNR Am J Neuroradiol 13:287–290 254. Rainov NG, Holzhausen HJ, Winfried B (1995) Dysplastic gangliocytoma of the cerebellum (Lhermitte–Duclos disease). Clin Neurol Neurosurg 97:175–180 255. Marano SR, Johnson PC, Speltzer RF (1988) Recurrent Lhermitte–Duclos disease in a child: case report. J Neurosurg 69:599–603 256. Hashimoto M, Fujimoto K, Shinoda S, Masuzawa T (1993) Magnetic resonance imaging of ganglion cell tumours. Neuroradiology 35:181–184 257. VandenBerg SR, May EE, Rubinstein LJ et al (1987) Desmoplastic supratentorial neuroepithelial tumors of infancy with divergent differentiation potential (“desmoplastic infantile gangliogliomas”): report on 11 cases of a distinctive embryonal tumor with favorable prognosis. J Neurosurg 66:58–71 258. Kato M, Yano H, Okumura A, Shinoda J, Sakai N, Shimokawa K (2004) A non-infantile case of desmoplastic infantile astrocytoma. Childs Nerv Syst 20:499–501 259. Ng THK, Fung CF, Ma LT (1990) The pathological spectrum of desmoplastic infantile gangliogliomas. Histopathology 16:235–241 260. Toshimitsu A, Hiroshi A, Terufumi I et al (1993) Desmoplastic infantile ganglioglioma. Neurol Med Chir 33:463–466 261. Tenreico-Picon OR, Kamath SV, Knorr JR et al (1995) Desmoplastic infantile ganglioglioma: CT and MRI features. Pediatr Radiol 25:540–543 262. Sperner J, Gottschalk J, Neumann K et al (1994) Clinical, radiological and histological findings in desmoplastic infantile ganglioglioma. Childs Nerv Syst 10:458–463 263. Kuchelmeister K, Bergmann M, von Wild K et al (1993) Desmoplastic ganglioglioma: report of two non-infantile cases. Acta Neuropathol 85:199–204 264. Prayson RA (1996) Gliofibroma: a distinct entity or a subtype of desmoplastic astrocytoma? Hum Pathol 27:610–613 265. Rushing EJ, Rorke LB, Sutton L (1993) Problems in the nosology of desmoplastic tumors of childhood. Pediatr Neurosurg 19:57–62 266. Duffner PK, Burger PC, Cohen ME et al (1994) Desmoplastic infantile gangliogliomas: an approach to therapy. Neurosurgery 34:583–589 267. Martin DS, Levy B, Awwad EE, Pittman T (1991) Desmoplastic infantile gangliogliomas: CT and MR features. AJNR Am J Neuroradiol 12:1195–1197 268. Kordek R, Biernat W, Alwasiak J, Liberski PP (1994) Pleomorphic xanthoastrocytoma and desmoplastic infantile ganglioglioma – have these neoplasms a common origin? Folia Neuropathol 32:237–239
156 269. Shao L, Tihan T, Burger PC (2002) Desmoplastic infantile ganglioglioma: a clinical and pathological review of eight cases. J Neuropathol Exp Neurol 61:466 270. Lellouch-Tubuiana MC, Salazar C, Cinalli G, Renier D, Sainte-Rose C, Pierre-Kahn A, Zerah M (2000) The management of desmoplastic neuroepithelial tumors in childhood. Childs Nerv Syst 16:8–14 271. Paulus W, Schlote W, Perentes E, Jacobi G, Warmuth-Metz M, Roggendorf W (1992) Desmoplastic supratentorial neuroepithelial tumours of infancy. Histopathology 21:43–49 272. Trehan G, Bruge H, Vinchon M, Khalil C, Ruchoux MM, Dhellemmes P, Ares GS (2004) MR imaging in the diagnosis of desmoplastic infantile tumor: retrospective study of six cases. AJNR Am J Neuroradiol 25:1028–1033 273. Chadarevian JP, Pattisapu JV, Faerber EN (1990) Desmoplastic cerebral astrocytoma of infancy: light microscopy, immunohistochemistry, and ultrastructure. Cancer 66:173–179 274. Cerda-Nicolas M, Kepes JJ (1993) Gliofibromas including malignant forms, and gliosarcomas: a comparative study and review of the literature. Acta Neuropathol 85:349–361 275. Barbosa MD, Balsitis M, Jaspan T et al (1990) Intraventricular neurocytoma: a clinical and pathological study of three cases and review of the literature. Neurosurgery 26:1045–1054 276. Coons SW, Asby LS (1999) Pathology of intracranial neoplasms. Neuroimaging Clin N Am 9(4):615–619 277. Yasargil MG, von Ammos K, von Deimling A et al (1992) Central neurocytoma: histopathological variants and therapeutic approaches. J Neurosurg 76:32–37 278. Tomura IV, Hirano N, Watanabe O et al (1997) Central neurocytoma with clinically malignant behavior. AJNR Am J Neuroradiol 18(6):1175–1178 279. Sgouros S, Carey M, Aluwihare N et al (1998) Central neurocytoma: a correlative clinicopathologic and radiologic analysis. Surg Neurol 49(2):197–204 280. Woesler B, Kuwerrt T, Kurlemann G, Morgenroth C, Probst-Cousin S, Lerch H, Gullotta F, Wassman H, Schober O (1998) High amino acid uptake in a low-grade desmoplastic infantile ganglioglioma in a 14-year old patient. Neurosurg Rev 21:31–35 281. Wichman N, Schubiger O, Von Deimling A et al (1991) Neuroradiology of central neurocytoma. Neuroradiology 33:143–148 282. Chang KH, Han MH, Kim DG, Chi JG et al (1993) MR appearance of central neurocytoma. Acta Radiol 34:520–526 283. Kim DG, Chi JG, Park SH et al (1992) Intraventricular neurocytoma: clinicopathological analysis of seven cases. J Neurosurg 76:759–765 284. Schweitzer JB, Davies KG (1997) Differentiating central neurocytoma. J Neurosurg 86:543–546
G.A. Christoforidis et al. 285. Rabinowicz AL, Abrey LE, Hinton DR, Couldwell WT (1995) Cerebral neurocytoma: an unusual cause of refractory epilepsy: case report and review of the literature. Epilepsia 36:1237–1240 286. Nishio S, Rabinowicz AL, Abrey LE, Hinton DR, Couldwell WT (1992) Cerebral neurocytoma: an Takeshita I, Kaneko Y, Fukui M. Cerebral neurocytoma: a new subset of benign neuronal tumors of the cerebrum. Cancer 70:529–537 287. Porter-Grenn LM, Silbergleit R, Stern HJ et al (1991) Intraventricular primary neuronal neoplasms: CT, MR, and angiographic findings. J Comput Assist Tomogr 15:365–368 288. Parker DR (1991) Neuroradiology case of the day: central neurocytoma. AJR Am J Roentgenol 156:1311 289. Okamura A, Goto S, Sato K, Ushio Y (1995) Central neurocytoma with hemorrhagic onset. Surg Neurol 43:252–255 290. Smoker WRK, Townsend JJ, Reichman MV (1991) Neurocytoma accompanied by intraventricular hemorrhage: case report and literature review. AJNR Am J Neuroradiol 12:755–770 291. Kim DG, Kim JS, Chi JG et al (1996) Central neurocytoma: proliferative potential and biological behavior. J Neurosurg 84:742–747 292. Drevelengas A, Polyzoides K, Kalaitzoglou I (1994) Intraventricular neurocytoma: case report and review. Eur J Radiol 19:14–18 293. Goergen SK, Gonzales MF, McLean CA (1992) Intraventricular neurocytoma: radiologic features and review of the literature. Radiology 182:787–792 294. Sgouros S, Walsh AR, Barber P (1994) Central neurocytoma of thalamic origin. Br J Neurosurg 8:373–376 295. Kim DG, Choe WJ, Chang KH, Song IC, Han MH, Jung HW, Cho BK (2000) In vivo proton magnetic resonance spectroscopy of central neurocytomas. Neurosurgery 46:329–333 296. Eng DY, Demonte F, Ginsberg L et al (1997) Craniospinal dissemination of central neurocytoma: report of two cases. J Neurosurg 86:547–552 297. Kendall B, Reider-Grosswater I, Valentine A (1983) Diagnosis of masses presenting within the ventricles on computed tomography. Neuroradiology 25:11–22 298. Tien RD (1991) Intraventricular mass lesions of the brain; CT and MRI findings. AJR Am J Roentgenol 157: 1283–1290 299. Spoto GP, Press GA, Hesselink JR, Solomon M (1990) Intracranial ependymoma and subependymoma: MR manifestations. AJR Am J Roentgenol 154:837–845 300. Muzumdar DP, Goel A, Pakhmode CK (2003) Oligodendroglioma causing calvarial erosion. Neurol India 51:129–130 301. Duong H, Sarazin L, Bourgouin P, Vezina JL (1995) Magnetic resonance imaging of lateral ventricular tumors. Can Assoc Radiol J 46:434–442
6
High-Grade Gliomas Antonios Drevelegas and George Karkavelas
Contents 6.1 Introduction.................................................................. 157 6.2 Anaplastic Astrocytoma.............................................. 158 6.3 Glioblastoma Multiforme............................................ 159 6.4 Gliosarcoma.................................................................. 188 6.5 Gliomatosis Cerebri..................................................... 189 References............................................................................ 196
A. Drevelegas (*) Department of Radiology, AHEPA university Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece e-mail:
[email protected] G. Karkavelas Department of Pathology, Aristotle University of Thessaloniki Medical School, Thessaloniki, Greece
6.1 Introduction Gliomas are the most common primary brain tumors in adults and account for 40–50% of all intracranial tumors. They may manifest at any age, but preferentially affect adults. Their peak incidence is in the fifth and sixth decade of life. They are slightly more common in men than women (1.5:1 ratio), and significantly more common in white than black people. Gliomas can affect any part of the CNS, but they usually occur more supratentorially in adults and infratentorially in children [1]. The clinical symptoms of the tumors depend on the anatomic location of the neoplasm in the brain. Headache, seizures, hemiparesis, personality changes, visual loss, gait disturbances, and signs of increased intracranial pressure are among the most common clinical manifestations. More than half of all glial tumors are astrocytic tumors. The pathologic classification and grading of astrocytomas is controversial, but on the other hand, it is also critical for assessment of their prognosis and treatment. A simple grading system for gliomas relies upon recognition of four parameters: nuclear atypia, mitoses, endothelial proliferation, and necrosis. The presence of two or more of the above-described features in a glioma would place the tumor in the high-grade category [2]. HGGs are tumors with both expansive and infiltrative growth [3]. They show some degree of anaplasia, without any cleavage plane, and in microscopic examination, tumor cells extend beyond the tumor margins. Anaplastic astrocytoma (AA) (WHO grade III) and glioblastoma multiforme (GBM) (WHO grade IV), the most common primary malignant brain tumors, are classified as high-grade tumors. Gliosarcoma, a rare (WHO grade IV) tumor composed of neoplastic glial cells and sarcomatous component is also reviewed with
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_6, © Springer-Verlag Berlin Heidelberg 2011
157
158
the high-grade tumors. Finally, the gliomatosis cerebri (GC) is considered as high-grade (WHO grade III) astrocytoma because of the poor prognosis of patients with this tumor, while histologically, the tumor consists of low-grade astrocytoma cells.
6.2 Anaplastic Astrocytoma AAs are infiltrating lesions with biology and average age of diagnosis intermediate between simple astrocytomas and glioblastomas multiforme (GBM). Almost all AAs originate as a benign tumor and have a tendency for malignant progression to glioblastoma multiforme [4]. Although some arise as new primaries, 75% result from differentiation of low-grade gliomas (LGGs). The course of progression of LGGs to AA varies considerably with time intervals ranging from less than 1 year to more than 10 years, the mean interval being 4–5 years. AAs correspond histologically to WHO grade III tumors. They represent one-third of fibrillary astrocytic tumors and about one-quarter of all gliomas [5]. AAs generally appear in a slightly higher age group than low-grade astrocytomas. Their peak incidence is in the fourth and fifth decades of life. The most common symptoms are seizures and focal neurologic deficits. These tumors have a poor prognosis with an average of 2-year survival rate. Pathology: AAs show enlargement and distortion of the invaded anatomical structures with blurring of the gross anatomical boundaries. Cystic areas and hemorrhage may be present (Fig. 6.1). AAs are more cellular
Fig. 6.1 Coronal gross section of the brain shows an anaplastic astrocytoma with hemorrhage
A. Drevelegas and G. Karkavelas
and pleomorphic (Fig. 6.2a), at least focally, than welldifferentiated astrocytomas (WHO grade II). On the other hand, they lack the necrosis and/or significant vascular proliferation that characterize glioblastomas. The range of cellularity and pleomorphism, as well as the number of mitoses, vary from low to high The central portions of these tumors are usually more cellular and anaplastic in comparison to the differentiated peripheral areas. Furthermore, multiple areas of increased cellularity and atypia may be found within the same tumor [6–8]. GFAP positivity is usually found in most, but not all, neoplastic cells (Fig. 6.2b). The term gemistocytic astrocytoma (from the Greek word gemistos that means stuffed or full) is used for fibrillary astrocytomas with profoundly eosinophilic or “glassy” cytoplasm and a considerable tendency for malignant transformation. Although not always malignant, they a
b
Fig. 6.2 Anaplastic astrocytoma. (a) The tumor is characterized by high cellularity and marked cellular pleomorphism (Hematoxylin– Eosin, original magnification ×100). (b) Neoplastic astrocytes positive to glial fibrillary acid protein (arrow) (GFAP, original magnification ×400)
6 High-Grade Gliomas
usually demonstrate sufficient features to be diagnosed as anaplastic [9]. Imaging: On CT AA presents as an ill-defined inhomogeneous lesion. Calcification is rarely encountered and only in cases of preexisting LGGs with malignant transformation. Peritumoral edema may be present as a hypodense area. After the administration of contrast material they show moderate or significant heterogeneous enhancement [10, 11] (Fig 6.3a). On MRI, AAs are heterogeneous on both T1- and T2-weighted images. However, the heterogeneous composition of the tumor is better reflected on T2-weighted images compared to T1-weighted images. On T2-weighted images, they often present with a hyperintense central area surrounded by an isointense rim with peripheral high signal intensity reflecting to the peritumoral edema. Following contrast administration they show heterogeneous or patchy enhancement [12] (Fig 6.3b–d). Tumor cells can be found either in the most lateral aspects of peritumoral edema or in areas depicted as normal on T2-weighted images outside the margins of the peritumoral edema. AAs may disseminate along the ependyma, leptomeninges, and CSF [12–14]. In terms of imaging characteristics, AAs may be difficult to differentiate from GBMs. However, AA margins are less defined and exhibit a moderate amount of mass effect, vasogenic edema, and heterogeneity. They show a minimal amount of hemorrhage, as opposed to findings in GBM. Necrosis, the imaging hallmark of GBM, is absent. AAs may also mimic the appearance of low-grade astrocytomas and can present as a well-demarcated, homogeneous nonenhancing mass [15, 16] (Fig. 6.4). Nonenhancing supratentorial neoplasm does not equate with low-grade malignancy. In one study, 40% of nonenhancing lesions proved to be AAs [17] (Fig. 6.5). MRI is the modality of choice for tumor surveillance and potential malignant transformation over time (see Fig. 6.3).
6.3 Glioblastoma Multiforme GBM is the most common primary intracranial CNS tumor accounting for more than half of all glial tumors and 15–20% of all intracranial tumors [18, 19]. About 50–60% of all astrocytic tumors are classified as GBMs. Although it represents only 1–2% of all malignancies, GBM is diagnosed in 15,000–20,000 patients
159
per year, most of who die from their disease [20]. Glioblastoma and its variants correspond to WHO grade IV tumors and is the most aggressive and least differentiated type of gliomas [21]. GBM may occur at any age with a peak incidence between 45 and 70 years. As with gliomas in general, these lesions show a male predominance of approximately 3:2. GBMs are most often found in cerebral hemispheres, particularly in frontal, parietal, and temporal lobes, although they can be situated in any lobe. They can also involve basal ganglia, and rarely, the posterior fossa. Intraventricular glioblastomas are very exceptional. Glioblastomas can arise “de novo” (primary glioblastomas), or after progression of an AA (secondary glioblastoma). Primary GBMs account for the vast majority of cases (60%) in adults older than 50 years. After a short clinical history, usually less than 3 months, they manifest de novo (that is, without clinical or histopathological evidence of a preexisting, less malignant precursor lesion). Secondary GBMs (40%) typically develop in younger patients (<45 years) through malignant progression from a low-grade astrocytoma (WHO grade II) or AA (WHO grade III). The most common clinical symptoms in GBMs include seizures, headaches, personality changes, focal neurologic deficits, and decreased intracranial pressure. Glioblastoma represent an aggressive tumor with an ominous prognosis. Despite the progress in surgery, radiation therapy, or chemotherapy, the mean survival time range between 6 and 12 months. Only exceptionally do patients survive beyond 2 years [22]. Pathology:Usually, GBMs are poorly delineated, heterogeneous tumors with necrosis, hemorrhage, and increased vascularity. Central necrosis is the hallmark of GBMs and may occupy as much as 80% of total tumor mass (Fig. 6.6). Intra- and intertumoral heterogeneity and pleomorphism, cellularity, and increased mitotic index are all features of GBMs. However, the hallmark, and prerequisite, for the diagnosis of these tumors with ominous prognosis is the presence of significant endothelial proliferation and/or necrosis [23, 24]. A range of malignant neoplastic cells can be recognized, from monotonously small to giant or “monstrous” ones (Fig. 6.7a). Occasionally, multinucleated neoplastic cells may also be found. Gemistocytes, granular or lipidized neoplastic cells may be present in GBMs, often with focal distribution [25]. Although in most GBMs, astrocytic features are apparent, in
160
a
c
Fig. 6.3 Anaplastic astrocytoma in a 68-year-old patient. (a) Postcontrast CT shows an inhomogeneous left parietal mass with moderate enhancement (arrow), and white matter edema. (b) Axial T1WI shows a heterogeneous left parietal mass. (c) Axial PDWI shows a high signal mass with extensive edema
A. Drevelegas and G. Karkavelas
b
d
and considerable mass effect. The curvilinear low signal structures represent intratumoral vessels (arrow). (d) Axial postcontrast T1WI shows intense, heterogeneous enhancement with subependymal involvement (arrow)
161
6 High-Grade Gliomas
a
b
c
Fig. 6.4 Anaplastic astrocytoma. (a) Axial T1WI shows a low signal left parietal mass (arrow). (b) On T2WI the mass shows homogeneous high signal intensity. (c) On postcontrast axial T1WI the mass remains unenhanced
162
a
A. Drevelegas and G. Karkavelas
b
c
Fig. 6.5 Gemistocytic astrocytoma. (a) T1WI shows a hypointense lesion involving the amygdala of the right temporal lobe. (b) T2WI shows a high signal intensity lesion that is more extensive than on T1WI. (c) After contrast administration the mass is not enhanced
163
6 High-Grade Gliomas
a
b
c
Fig. 6.6 Gross brain sections of different GBMs. (a) Typical large necrotic tumor. (b) Heterogeneous mass involving the left occipital lobe. (c) Large tumor with necrotic and hemorrhagic
areas. Note the extension of the tumor through the corpus callosum into the contralateral hemisphere
others, the origin of neoplastic cells is hardly recognizable. Additionally, areas with features of welldifferentiated, AA and GBM may coexist in the same tumor. Areas of increased collagen deposition may be found among neoplastic cells. Aggregates of perivascular lymphocytes is another feature found in GBMs. The number of mitoses varies from tumor to tumor, as well as from area to area in the same section [9, 26].
Vascular (microvascular, endothelial) proliferation is characterized by a combination of hypertrophy and hyperplasia of endothelial cells within small vessels, mainly the capillaries (Fig. 6.7b). Pericytes and vascular smooth muscle cells contribute as well. These proliferations, found throughout the neoplastic area, are usually more aggregated around necroses. A proliferation of capillaries resembling renal glomerular tuft (glomeruloids) is correlated with an ominous prognosis [27].
164
a
A. Drevelegas and G. Karkavelas
b
c
Fig. 6.7 Glioblastoma multiforme. (a) High power. Pleomorphism of neoplastic cells and giant neoplastic figure (arrowhead) (Hematoxylin–Eosin, original magnification ×400). (b) Low magnification. A cellular astrocytic tumor with overt endothelial proliferation (asterisks) (Hematoxylin–Eosin, original magnifi-
cation ×100). (c) Area of necrosis (asterisk) surrounded by neoplastic astrocytes (pseudopalisading necrosis) (arrow). Overt endothelial (microvascular), glomeruloid proliferation (arrowheads) (Hematoxylin–Eosin, original magnification ×100)
Necrosis alone is sufficient to distinguish glioblastomas from lower-grade astrocytic tumors. Although a characteristic rim of neoplastic cells at the margin of necrosis (known as “palisading” or “pseudopalisading”) is easily recognized in some GBMs, this is not a rule (Fig. 6.7c). The size of necroses ranges from tiny, hardly recognizable areas with round or serpiginous shape to large areas resembling infarcts. These large necroses, recognized also by neuroimaging, are associated with a sinister prognosis [28, 29]. The morphologic diversity of GBMs is also extended to the level of antigen expression, as revealed by immunohistochemical methods. Although GFAP-positive neoplastic cells are usually easily recognizable, nonreactive and weakly stained cells or areas may be found within the same tumor as well. Vimentin positivity is
nonspecific. S-100 protein as well as cytokeratin immunoreactivity may also be found [30, 31]. Imaging: On unenhanced CT GBM appears as a central low-density mass located usually in the centrum semiovale. Calcification is rare in GBMs. They are heterogeneous due to the reflecting sites of necrosis, hemorrhage, and increased cellularity. Necrosis is the imaging hallmark of GBMs [32, 33]. Besides necrosis and hemorrhage, another characteristic radiologic feature is the presence of edema, which surrounds the tumor, extends along the adjacent white matter tracts, and usually produces significant mass effect. After the administration of contrast material, they usually show marked heterogeneous rim enhancement with thick, shaggy, irregular, and nodular wall (Figs. 6.8–6.10).
165
6 High-Grade Gliomas
a
b
Fig. 6.8 (a) Unenhanced CT shows a low density, right temporoparietal mass with ill-defined borders. (b) Postenhanced CT shows an irregular ring-like enhancement of the mass with central necrosis
Fig. 6.9 GBM in a 46-year-old patient. Postcontrast CT shows a right frontal mass with thick, irregular enhancement causing significant mass effect. Note the extensive peritumoral edema along the white matter tracks and the central necrosis
Fig. 6.10 Postcontrast CT shows a low-density mass in the left centrum semiovale with necrosis, peritumor edema, and irregular ring-like enhancement
166
MR imaging is frequently ordered as the initial imaging study in patients suspected of having a brain neoplasm. MR findings reflect some of the pathologic features of GBMs. On T1-weighted images, they appear with low- or mixed-signal intensity, while on T2-weighted images, high signal is indicative of surrounding vasogenic edema or necrosis [34] Prominent inhomogeneous enhancement is seen (Fig. 6.11). The vasogenic edema is produced by abnormal neoplastic vessels, which lack the normal blood–brain barrier resulting in transudation of fluids and proteins into the extracellular space. Although MR imaging is the most sensitive method for depicting abnormal amounts of tissue water, discrimination of tumor tissue from edema in terms of signal characteristics has proved unreliable. The white matter edema produced by GBMs is very extensive and actually represent a tumor plus edema [13, 35, 36] (Figs. 6.11b, 6.12a). Among the various paths of dissemination, direct extension along white matter tracts is the most common route. Spread across the corpus callosum and anterior and posterior commissure is also typical. Symmetric extension through the corpus callosum gives rise to butterfly appearance. Subependymal spread of GBM can occur and is correlated with a poor prognosis (Fig. 6.13). Spread along the neuraxis via the CSF is also well documented and, incidentally, has been found to have a 6–20% incidence at autopsy series. The incidence of symptomatic metastases is certainly lower than the incidence seen at postmortem and is due to the short survival of the affected patients [37, 38]. The majority of GBMs are solitary lesions. Multifocal or multicentric tumors occur rarely in 0.5– 1% of cases [39]. Multicentric tumors are those with neither macroscopic nor microscopic connection. On the other hand, those with either gross or microscopic continuity are defined as multifocal [40–42] (Fig. 6.14). The most frequent dissemination route in the latter group is the meningeal-subarachnoid space, followed by the subepedymal, intraventricular route and direct brain penetration [43, 44]. Hemorrhage was reported in approximately 19% of patients with GBM and in approximately 12% of patients with low-grade lesions [33]. The presence of hemorrhage of different ages and necrosis is responsible for the heterogeneous pattern on T2-weighted images. T2*-weighted gradient-refocused images are the most effective for hemorrhage detection due to increased sensitivity of gradient-echo images for the detection of field heterogeneity and magnetic
A. Drevelegas and G. Karkavelas
susceptibility effects [39, 40]. On T2*-weighted gradient-echo MR imaging, intracerebral hemorrhage causes marked signal intensity loss because of magnetic susceptibility effect (Fig. 6.12b). Studies of human neoplasms have demonstrated that increased malignancy is associated with increased neo-vascularity. Thus HGGs are highly vascular and prominent flow-voids are often present [41–44]. Occasionally, GBMs show vascular features similar to those of arteriovenous malformation. High-power MRI systems (8 T) have shown promise in directly depicting areas of increased vascularity within foci of the tumor bed (Fig. 6.15). The therapeutic approaches to GBMs differ considerably according to the tumor extension and tumor grade. Both influence patient prognosis and therapeutic options. Although the extension of the lesion is best delineated by MRI, the imaging findings do not correlate well with the histopathologic features, because microscopic residual tumor extends beyond recognizable demarcation on CT or on MRI images. The inability to determine tumor margins reliably with either contrast-enhanced T1-weighted images or with T2-weighted MR images represents a significant problem in the management of patients with malignant glioma [13]. On the other hand, determination of glioma grade is based on histopathologic features, but sampling error in a limited biopsy may result in undergrading of some tumors. For this reason, imaging characteristics play an important role in predicting the grade of glial neoplasms. They include mass effect, heterogeneity, edema, contrast enhancement, hemorrhage, and cyst formation or necrosis [45]. However, grading of tumors with conventional MR imaging is not always accurate, with sensitivity in identifying HGGs ranging from 55.1–83.3% [46–48]. The development of techniques capable of accurately depicting tumor margins and grades in vivo is important for the determination of the most appropriate treatment for gliomas. The implementation of echo-planar imaging allowed the development of advanced imaging techniques such as diffusion-weighted imaging (DWI), dynamic contrast-enhanced perfusion (perfusion-weighted imaging (PWI)), and MR spectroscopy (MRS), providing physiologic information that complements the anatomic information available with conventional MR imaging. Today, advanced MR imaging is a key modality not only for lesion diagnosis, but also to evaluate the extension,
167
6 High-Grade Gliomas
a
b
c
Fig. 6.11 (a) Axial T1WI shows a heterogeneous right frontal lobe mass causing significant mass effect. (b) On axial T2WI, the mass shows central high signal intensity due to the necrosis.
Also note the surrounding vasogenic edema extending along the adjacent white matter. (c) Axial postcontrast T1WI shows central necrosis with thick and irregular ring-like enhancement
168
a
A. Drevelegas and G. Karkavelas
b
Fig. 6.12 (a) Axial T2-weighted image shows typical finger-like edema along the white matter tracts in a patient with GBM. (b) Axial T2* image shows low signal intensity in a patient with hemorrhagic GBM due to magnetic susceptibility effect
a
b
Fig. 6.13 Patterns of GBMs dissemination. (a) Axial post-contrast T1-weighted image shows extension of a GBM through the splenium of the corpus callosum (butterfly glioma). (b) Axial T1-weighted images shows subependymal spread of a GBM (arrow)
169
6 High-Grade Gliomas
a
b
Fig. 6.14 (a) Right parasagittal postcontrast T1-weighted image shows a parietal lesion with strong enhancement and peripheral edema. In the right temporal lobe a second separate enhanced lesion is seen. Autopsy proved a multicentric GBM. (b) Axial
T1-weighted images show a GBM with satellite-enhanced nodules. Note the spread across the ependyma. Biopsy proved a multifocal GBM
type, and grade of the tumor. Additionally, it provides some new insights into neuroradiology practice, such as, showing tumor areas before stereotactic biopsy, distinguishing radiation necrosis from tumor infiltration, and assessing tumor response to therapy [49, 50]. DW imaging uses a pair of strong magnetic gradient pulses to dephase and subsequently rephase protons. Protons with higher diffusion rates show a loss of phase coherence and a low MR signal while protons with slow or restricted diffusion will largely rephase and appear as a high MR signal. The information provided reflects the viability and structure of tissue on a cellular level. Quantitative information on restricted diffusion of water molecules can be obtained by calculating the apparent diffusion coefficient (ADC). DW imaging has proved to be clinically useful in the evaluation of cerebral ischemia, infection, and tumors [51, 52]. The role of DWI in patients with brain tumors has been investigated. ADCs could provide useful information in the diagnosis of patients with brain tumors, such as tumor malignancy and peritumoral infiltration. Tumor
cellularity and tumor grade have been correlated with ADC values. Brain neoplasms with higher cellularity or higher grades show a significant reduction in the value of the ADC and a marked increase in the signal of diffusionweighted images. The low ADC values in tumor probably reflect a decreased volume of extracellular space due to increased tumor cellularity and increased intracellular viscosity, with subsequent restriction of water motion. Lower ADCs indicate mostly HGGs, whereas higher ADCs LGGs [53–56] (Fig. 6.16a–c). Thus, low-grade astrocytomas show high ADC, AA intermediate ADC, and GBMs the lowest ADC. Arvinda et al. [57] found that normal ADC values were 75 ± 4.89 × 10−3 mm2 s−1 and was 134.84 ± 35.74 × 10−3 mm2 s−1 for LGGs and 112.92 ± 40.389 × 10−3 mm2 s−1 for HGGs. An ADC value of 98.50 mm2s−1 was defined as a threshold below which tumors were classified as HGGs and showed a sensitivity, specificity, PPV, and NPV of 90, 87.1, 81.81, and 93.10%, respectively. Other investigators failed to find a significant difference between the ADC values of HGGs and low-grade tumors [58, 59].
170
a
A. Drevelegas and G. Karkavelas
b
c
d
Fig. 6.15 High-grade astrocytoma. Comparison of conventional 1.5 T MRI T1 postgadolinium imaging with magnetization transfer (a) to 8 T gradient-echo high resolution MRI (b) and
perfusion MRI (c, d) demonstrates a good correlation between areas of increased perfusion and increased vascularity on 8 T high-resolution imaging (arrow). (Courtesy Gr. Christoforidis)
Recently, an effort was made to define glioma margins using diffusion tensor imaging (DTI). MR DTI is a noninvasive in vivo method allowing the mapping of white matter tracts in the human brain. DTI is based on
the concepts of iso- and anisotropic diffusion. When water molecules diffuse equally in all three directions, this is termed as isotropic diffusion. This is typical in the ventricles, but is also true in the gray matter. In the
171
6 High-Grade Gliomas
white matter, free water molecules move anisotropically, that is, water diffusion is not equal in all three directions. This is because, in the white matter tracts,
a
c
Fig. 6.16 High-grade glioma (GBM). Postcontrast T1WI shows a space-occupying and inhomogeneously enhanced mass involving the corpus callosum (a) with increased diffusion (b) and low corresponding ADC (c). The fractional anisotropy color coded
the myelin sheath surrounding the white matter causes the water molecules to move more along the long axis of a fiber bundle and less perpendicularly.
b
d
map (d) shows reduction of the FA values (arrow), while the tractography (e) demonstrates displacement and destruction of the white matter tracts. Perfusion color coded map (f) demonstrates increased rCBV
172
A. Drevelegas and G. Karkavelas
e
f
Fig. 6.16 (continued)
Information from DTI is presented in two formats, which are fractional anisotropy (FA) maps and 3-D reconstruction of the white matter tracts which is called tractography. FA can be quantified numerically or can be coded in color maps. When a white matter tract is destroyed by a tumor, there is loss of anisotropy and therefore a reduction in the FA values (Fig. 6.16d). There are FA changes in the white matter of
glioblastomas that might indicate cellular infiltration beyond the area of the tumor enhancement. The finding of reduced FA in the white matter around a tumor in the direction increases the probability of peritumorous white matter infiltration by tumor elements. White matter tracts can be displaced, invaded, or destructed by a brain tumor (Fig. 6.16e). DTI can be used in the current clinical practice to determine the
6 High-Grade Gliomas
relation of a tumor to the adjacent white matter tracts [60]. This in turn helps guide the surgical approach and extent of resection [61]. DTI demonstration of the corticospinal tracts is a useful adjunct to intraoperative fiber stimulation [62–64]. Preoperative tractography showing tumor involvement of the corticospinal tract has been correlated to motor deficits, even when the motor cortex is uninvolved [65]. Conversely, normalization at postoperative tractography was predictive of improvement in function, suggesting a role for intraoperative tractography [66, 67]. Several studies suggest that contrast enhancement alone is not sufficient to predict tumor grade, since 20% of LGGs demonstrate contrast enhancement while one fifth to one third of high-grade tumors do not [46]. Although contrast enhancement indicates disruption of the blood–brain barrier, it is not particularly effective at revealing the underlying regional vascularity. Precise determination of the regional vascularity is important, given that the degree of vascular proliferation is an important parameter for the histopathologic grading of gliomas. The introduction in the clinical practice of echo planar MR imaging allowed the estimation of parameters that reflect tissue vascularization. PWI provides useful information about the microcirculation in brain tissue. This technique requires the dynamic (bolus) intravenous administration of MR contrast agent. As the paramagnetic contrast agent passes through the intravascular compartment, it causes local field inhomogeneities that result in magnetic susceptibility effects with a decrease in signal on multiple repeated T2*-images that can be measured. This signal drop depends on both the vascular concentration of contrast agent and the concentration of small vessels per voxel of tissue [68, 69]. The postprocessed row data of the perfusion technique allow us to obtain information on the four perfusion parameters, which are cerebral blood volume (CBV), cerebral blood flow (CBF), mean transit time (MTT), and time to peak (TTP). Arterial spin labeling is a new perfusion technique that does not require exogenous contrast; instead, it exploits the spins of endogenous water protons that perfuse the imaging plane. In brain tumors, CBV maps are particularly sensitive for depicting the microvasculature of a tumor and therefore its aggressiveness and proliferative potential. Several studies have found statistically significant correlations between tumor rCBV and glioma grade. HGGs demonstrate higher rCBV values (Figs. 6.16 f, 6.17a–c) than low-grade tumors, and the low-grade tumors with
173
contrast enhancement present with low rCBV [70–73]. Several studies have shown that LGGs have maximal rCBV values between 1.11 and 2.14, while maximal rCBV values of HGGs range between 3.54 and 7.32 [74–78]. In one study, a threshold value of 2.91 for rCBV provided sensitivity, specificity, PPV, and NPV of 94.7, 93.75, 90.0 and 96.8, respectively in determining HGGs [57]. Lev and Rosen used an rCBV threshold value of 1.5 in discriminating 32 consecutive patients with glioma [79]. The sensitivity and specificity with the use of an rCBV of 1.5 as a threshold value were 100 and 69%, respectively. In clinical practice, 95–100% sensitivity has been reported for differentiating high-grade from LGGs using thresholds of 1.75 and 1.5 for rCBV, respectively. Although the rCBV values are a very reliable parameter for glioma grading, it has been found that the perfusion pattern of low-grade oligodendrogliomas, in contrast to that of low-grade astrocytomas, may show foci of high rCBV. As a result, glioma grading based on rCBV measurements may not be accurate if low-grade oligodendrogliomas are included in the sample of gliomas [80, 81]. Relative rCBV maps can also be used to reduce the sampling error in the histopathologic diagnosis of gliomas, improving the selection of targets for stereotactic biopsy. High CBV areas represent the best stereotactic biopsy site for accurate grading of astrocytomas [72, 79]. MRS is a noninvasive technique that allows in vivo measurements of certain tissue metabolites. By suppressing water signal, the relative concentrations of nonwater, proton-containing metabolites from discrete tissue regions can be quantified. There is certainly compelling evidence that MRS provides important supplementary information to that of conventional MR imaging. In normal brain, the principal metabolite signals that can be measured by MRS are choline (Cho), creatine (Cr), N-acetyl aspartate (NAA), and lactate (Lac) [63, 83]. Each metabolite resonates at a specific frequency (ppm), and each one reflects specific cellular and biochemical processes. Choline is a cell membrane marker. The increase in Cho in simple terms is attributed to cell membrane turnover and proliferation. An elevation in Cho may be due to either cell membrane synthesis, destruction, or both. Cho is most prominent in regions with high neoplastic cellular density and is progressively lower in moderate and low-grade tumors. Some highly malignant tumors and some GBMs may show low Cho because of extensive necrosis [84–86]. Cr is a marker
174
A. Drevelegas and G. Karkavelas
of “energy metabolism” and is reduced in tumors due to the increased metabolic activity of tumors consuming energy. NAA is a marker of neuronal density and viability, and it is generally felt to be neuron specific. The decrease in NAA represents the replacement of normal functioning neurons and axons by any disease
that adversely affects neuronal integrity [83, 84]. Lactate is a product of anaerobic glycolysis and is present only in minute amounts in normal brain. It appears when tumors outgrow their blood supply and start utilizing anaerobic glycolysis. Lactate levels increase significantly in necrotic and cystic tumors.
a
b
Fig. 6.17 Glioblastoma multiforme. (a) Axial postcontrast T1-weighted image shows intense inhomogeneous enhancement of the tumor. (b) On the perfusion MR image, the enhanced tumors shows low signal intensity due to the high rCBV. (c) Timesignal intensity curve of the tumor (red) drops markedly compared
to that of the opposite normal brain (yellow). (d) Long TE (135 ms) spectra obtained from the enhancing lesion shows marked elevation of the Cho, decrease of NAA, and increase of lipids. (e, f) Cho:Cr and Cho:NAA maps demonstrate significantly elevated ratios (Cho:Cr 8:06 and Cho:NAA 15:3)
6 High-Grade Gliomas Fig. 6.17 (continued)
175
c
d
176
A. Drevelegas and G. Karkavelas
e
f
Fig. 6.17 (continued)
Lipids are products of brain destruction and are found in necrotic portions of tumors. Myo-inositol is glial cell marker and osmolyte hormone receptor mechanism. The basic metabolite changes common to brain tumors include elevation in choline (Cho), lactate (Lac), lipids (L), decrease in N-acetyl aspartate (NAA), and decrease in creatine (Cr) (Fig. 6.17a–d). As a general rule, as malignancy increases, NAA and creatine decrease, and choline, lactate, and lipids increase. Myo-inositol can also be used to differentiate LGGs and HGGs. LGGs express higher levels of myo-inositol compared with HGGs. This may be due to the lack of activation of phosphatidylinositol metabolism resulting in accumulation of myo-inositol in LGGs [87–89]. There is extensive literature demonstrating the metabolite ratios of Cho:Cr, NAA:Cr, NAA:Cho, and Cho:NAA, and the presence of lipids and lactate to be useful in grading tumors and predicting tumor malignancy. A significant difference is noted in several investigations on Cho:Cr, Cho:NAA, and NAA:Cr ratios for differentiating between LGGs and HGGs. Law et al. [74] demonstrated a threshold value of 1.56 for Cho:Cr to provide sensitivity, specificity, and positive and negative predictive values of 75.8, 47.5, 81.2, and 39.6%, respectively, for the determination of
HGGs vs. LGGs. In the same study, a threshold value of 1.6 for Cho:NAA provided 74.2, 62.5, 85.6, and 44.6% for the sensitivity, specificity, and positive and negative predictive values, respectively, for determination of HGGs (Fig. 6.17e,f). The differential diagnosis of GBM includes abscess, metastasis, lymphoma, and tumefactive multiple sclerosis (MS). Brain abscess remains a diagnostic challenge, because the presenting clinical manifestations and neuroradiologic appearances are often nonspecific. Only 40–50% of patients are febrile on examination or have symptoms such as headache, nausea, or altered mental status. Imaging findings helpful in differential diagnosis include a thin wall with ring-like enhancement that is often thinner along the medial margin, daughter rings, and hypointense rim on T2-weighted images [90, 91] (Figs. 6.18, 6.19 a,b, 6.20 a,b). Diffusion-weighted images, perfusion, and spectroscopy can be used to differentiate abscesses from high-grade tumors. The typical DWI feature for brain abscess is restricted diffusion with markedly reduced ADC (Figs. 6.19c,d, 6.20c). Most studies have reported decreased ADC values in abscesses as opposed to the necrotic component of tumors. Noguchi et al. found that the necrotic components of tumors had ADC
177
6 High-Grade Gliomas
a
b
c
Fig. 6.18 Differential diagnosis. GBM from brain abscess. (a) Coronal T1-weighted image shows a ring-like thin-walled enhancement. (b) Axial T2-weighted image of the same patient shows a peripheral high-signal lesion with significant white matter edema. Biopsy proved a GBM. (c, d) Axial and coronal
d
T1-weighted images in a patient with brain abscess show ringlike enhanced satellite lesions (arrow). Also note the thin medial wall of the abscess (arrowheads). (e) On axial T2-weighted image of the above patient the collageneous capsule is hypointense (arrowheads)
178
Fig. 6.18 (continued)
values in the range of 2.2–3.2 × 10−3 mm2 s−1, in contrast to abscesses that had ADC values of less than 0.7 × 10−3 mm2 s−1 [92]. In another study, the calculated ADC values of the enhancing component of GBM were in the range of 0.78 × 10−3–1.79 × 10−3 (mean, 1.14 × 10−3) [93]. However, there are reported cases of tumors demonstrating marked hyperintensity on DW-MRI with extremely low ADC values in the necrotic centre, similar to what has been reported for abscesses. Restricted diffusion within these lesions could be due to a variety of causes, including intratumoral hemorrhage, cytotoxic edema in the early phase of cell death, thick sterile liquefaction, or pyogenic superinfection [94–97]. On perfusion images abscesses show low rCBV values. High-grade primary neoplasms and metastases can be differentiated from abscesses with perfusion MR imaging, in which the wall of necrotic or cystic neoplasms tends to have higher rCBV compared with the capsule of an abscess. MRS can also noninvasively contribute to the establishment of the differential diagnosis between tumors and abscesses. The spectrum of the abscess cavity shows elevation of lactate, acetate, succinate, alanine, and some amino acids, and this spectrum appears significantly different from that of necrotic or cystic brain tumor [98] (Fig. 6.20d).
A. Drevelegas and G. Karkavelas
Intracranial metastases and primary malignant gliomas are two common brain tumors encountered in adults. The management of these two tumors is different and can potentially affect the clinical outcome. When the clinical findings are nonspecific, conventional MR imaging characteristics may be similar and unable to differentiate the two entities. DWI techniques can be helpful in distinguishing solitary intraaxial metastatic lesions from HGGs. Studies showed that both contrast enhancing portions and peritumoral edema of metastasis have higher ADC than HGGs. However, the distinction between metastases and HGGs is often difficult to make based on ADC values, as some HGGs also exhibit high ADC values [55, 95, 99–101]. Perfusion MR imaging and spectroscopy may also be used to differentiate high-grade primary gliomas and solitary metastases. Law et al. found significant difference between the mean rCBV value within the peritumoral region in HGGs which was 1.31 ± 0.97, suggesting increased peritumoral perfusion due to tumor infiltration and the mean rCBV surrounding metastatic tumors which was 0.39 ± 0.19, consistent with compression of capillaries by vasogenic edema [102] (Fig. 6.21a–c). No difference has been demonstrated between the intratumoral rCBV values of HGGs and metastases. In the same study, they have found elevated Cho:Cr (2.28 ± 1.24) in the peritumoral regions of HGGs in keeping with tumoral infiltration. No increase in the Cho:Cr (0.76 ± 0.23) was found in the peritumoral region of metastases, which again suggests vasogenic edema (see Fig. 6.21d). In the same study, there was also no significant difference in the peritumoral NAA:Cr between the two groups because there is no neuronal replacement or destruction in the peritumoral region of either pathologic condition. The differential diagnosis of primary CNS lymphoma (PCNSL) from malignant astrocytoma can be based on the periventricular and deep gray matter location of lymphoma and on hypo- or isointensity of PCNSL on T2-weighted images while HGG shows high signal intensity. On postcontrast T1-weighted images, PCNSL shows intense homogeneous enhancement (Fig. 6.22a,b). However, less commonly, PCNSL may be hyperintense on T2-weighted images and may show ring-like enhancement especially in immunocompromised patients. In these cases, it can be difficult to distinguish PCNSL from HGG on the basis of conventional MR imaging features.
179
6 High-Grade Gliomas
a
b
d
c
Fig. 6.19 Diffusion study in a patient with cerebellar abscess. (a) Axial T2-weighted image shows a left cerebellar hyperintense lesion surrounded by a well-defined hypointense rim and perilesional edema. (b) Postgadolinium T1-weighted image
demonstrates ring-like enhancement of the capsule. (c) On diffusion weighted image (b = 1,000 mm2 s−1) the lesion shows high signal intensity. (d) The corresponding ADC map shows low signal intensity due to restricted diffusion
Diffusion-weighted imaging techniques showed that the mean ADC value in lymphomas is lower than that of HGGs reflecting the increased cellularity of lymphomas
[103, 104] (Fig.6.22c–e). Neovascularization is absent in PCNSLs in contrast to HGGs; therefore, perfusionweighted images can also be used to distinguish the two
180
A. Drevelegas and G. Karkavelas
a
c
Fig. 6.20 Brain abscess. (a) Postcontrast T1-weighted image shows a ring-like enhanced lesion compressing the adjacent ventricle. (b) On T2-weighted image the lesion appears hyperintense surrounded by a hypointense rim. Note the extensive
b
d
perilesional edema. (c) On diffusion-weighted image the lesion shows increased signal intensity. (d) Long TR spectrum (TE 135) shows acetate (Ac), alanine (Ala), lactate (Lac), lipids (Lip), and amino acids (AA)
181
6 High-Grade Gliomas
entities. The lower rCBV values of PCNSLs may be helpful in differentiating from other brain tumors such as HGGs [105] (see Fig. 6.22f). The MR spectrum of PCNSL is characterized by increased Cho, lactate, and lipids and is associated
a
with decreases in NAA, Cr, and myo-inositol levels (Fig. 6.22g). The latter metabolite pattern can also be found in malignant astrocytomas. However, if the analysis of the solid portion of the tumor shows a significant increase in lipids, it is probably lymphoma [106].
b
c
Fig. 6.21 Brain metastasis. Axial postcontrast T1-weighted (a) and T2-weighted (b) images show an enhancing lesion surrounded by significant edema. Time-intensity curves image (c) demonstrates marked signal drop at the site of the lesion (green) as opposed to the normal contralateral brain tissue (yellow). Note
that the peritumoral edema shows minimal drop (red curve) suggesting decreased perfusion due to capillaries compression by vasogenic edema. Long TR (135 ms) spectrum obtained from the peritumoral edema shows normal Cho, Cr, and NAA peaks
182
A. Drevelegas and G. Karkavelas
Fig. 6.21 (continued)
The conventional MRI features of tumefactive MS may mimic that of HGGs. Both lesions may show variable contrast enhancement, perilesional edema, mass effect, and central necrosis. The pathologic difference between tumefactive MS and HGGs is the absence of angiogenesis in MS, which demonstrates low rCBV values (Fig.6.23) as opposed to gliomas that are characterized by neovascularization and angiogenesis resulting in a significant elevation of rCBV [107]. In spectroscopy there is some overlap between tumefactive MS and HGGs. There is Cho elevation in MS from astrogliosis, demyelination, and inflammation; lactate is found from anaerobic glycolysis; and decrease of NAA from neuronal and axonal damage. However, NAA:Cr ratio is higher in both the enhancing and central regions in MS. Hence, the combination of reduced perfusion and moderate NAA reduction can help differentiate tumefactive MS from HGGs [108, 109] (Fig. 6.24). Posttherapeutic MR examinations are used routinely to differentiate therapy-induced necrosis from residual or recurrent tumor. On conventional MRI, abnormal contrast enhancement on postcontrast T1-weighted images and the pattern of hyperintensity on T2-weighted images are used to discriminate these
entities. Features that are inspected for assessment of tumor progression are regions of abnormal contrast enhancement on postcontrast T1-weighted images and the volume of hyperintensity on T2-weighted images. Other conventional MR imaging findings suggesting predominant gliomas progression are corpus callosum involvement, conjunction with multiple enhancing lesions without crossing of the midline, and subependymal spread [110]. Although such morphologic changes are indicative of the existence of the disease, it is often difficult to determine whether an enhanced lesion represents tumor recurrence or not, especially when the enhancement is initially observed. Perfusion MR imaging may be helpful for differentiating recurrent tumors from posttherapeutic necrosis. Elevated rCBV values most likely represent recurrent tumor while decreased rCBV values are likely to represent radionecrosis due to vascular injury (Figs. 6.17, 6.25a,b). Studies found that if the rCBV ratio of enhancing lesion is more than 2.6 or less than 0.6, tumor recurrence or radiation necrosis, respectively, should be strongly suspected [111]. MRS has a significant clinical impact on posttreatment evaluation of intracranial neoplasms and it is useful to identify recurrent tumor earlier than conventional
183
6 High-Grade Gliomas
a
c
Fig. 6.22 Primary cerebral non-Hodgkin B-cell lymphoma. (a) Axial T2-weighted image (a) shows a left occipital mass, isointense to the gray matter, surrounded by high signal edema. On postcontrast T1-weighted image (b) the mass shows intense homogeneous enhancement. On diffusion-weighted image (c) the tumor shows high signal intensity, while on ADC map (d) exhibits low signal intensity compared to the opposite area
b
d
of the normal brain due to restricted diffusion, and on the fractional anisotropy color coded map (e) shows reduction of the FA values. Perfusion color map and time-intensity curve (f) demonstrate low rCBV similar to the normal brain. Long TE spectra (g) of the mass shows elevation of choline, decrease of NAA and Cr, and increase of lipids
184
e
f
Fig. 6.22 (continued)
A. Drevelegas and G. Karkavelas
185
6 High-Grade Gliomas
Fig. 6.22 (continued)
a
b
Fig. 6.23 Paraventricular tumor-like MS lesions in a 30-year-old patient. (a) Axial T2WI shows multiple well-defined hyperintense lesions. (b) On perfusion color coded map image the lesions demonstrate decreased rCBV
MRI and differentiate residual or recurrent tumor from posttreatment abnormality. Evidence of radiation necrosis is typically observed within 3–6 months after therapy and is characterized
by decreases in Cho, Cr, NAA, and increased lipid and lactate levels (Fig. 6.25c). On the other hand, elevated Cho levels and greater Cho:NAA ratio suggest tumor recurrence (Fig. 6.26).
186
a
A. Drevelegas and G. Karkavelas
b
c
Fig. 6.24 (a) Axial FLAIR image shows a hyperintense lesion adjacent to the left atrium involving the splenium. (b) On axial postcontrast T1WI the lesion shows irregular, partial ring-like
enhancement. (c) Single-voxel MR proton spectrum shows mild depression of the NAA peak and elevation of the choline, findings consistent with MS
6 High-Grade Gliomas
187
a
b
Fig. 6.25 Radiation necrosis 6-months after radiotherapy and adjuvant chemotherapy in a patient with GBM. (a) Axial postcontrast T1WI shows ring-like enhancement of the tumor. (b) Perfusion color map and time-intensity curve show low
rCBV of the enhanced rim of the tumor (purple) as opposed to the normal contralateral brain (green). (c) Long TE spectroscopy obtained from the enhanced rim shows absence of Cho and NAA and increase of the lipids indicative of radiation necrosis
188
A. Drevelegas and G. Karkavelas
Fig. 6.25 (continued)
Fig. 6.26 Tumor recurrence in a patient with GBM after radiotherapy and adjuvant chemotherapy. MR spectroscopy of the enhanced area demonstrates increase of Cho, decrease of NAA,
and increase of lactate and lipids. Note the significant increase of Cho:NAA ratio (4:46)
Increases in lipid and lactate levels are less specific than elevated Cho levels because they may result from both tissue necrosis and radiation-induced necrosis. Follow-up of these lesions is crucial for accurate diagnosis [112].
6.4 Gliosarcoma In a small percentage of GBM, not more than 2%, a sarcomatous component is apparent, making, macroscopically, a relatively discreet appearance. These tumors are
189
6 High-Grade Gliomas
characterized as gliosarcomas [3, 26, 113, 114]. The most common types of differentiation of sarcomatous components in this tumor are fibrosarcomatous and malignant fibrous histiocytoma-like types, but various other lines of mesenchymal differentiation have been described in gliosarcoma [115–117]. They are peripherally located and involve the temporal, parietal, and occipital lobes [27, 118, 119]. Posterior fossa gliosarcomas have also been reported [114, 118]. Radiation-induced gliosarcoma may appear at the site of a treated intracranial neoplasm [118, 120]. Most patients with gliosarcoma are in their fifth to seventh decade. Gliosarcomas correspond histologically to WHO grade IV. Extracranial metastasis of the sarcomatous component is common occurring in 15–30% of all gliosarcomas [113, 121]. Pathology: This tumor, a variant of glioblastoma, is characterized by a combination of anaplastic glial and mesenchymal cells. Although oligodendroglial cells are occasionally recognized, in most cases, the glial cells are astrocytic, with a tissue pattern of a glioblastoma. As expected, heterogeneity is a feature of the glial component. Additional variation is recognized in the mesenchymal component. Malignant fibrous histiocytoma is the most frequent sarcomatous component [122]. Fibrosarcoma, smooth and striated muscle sarcoma as well as bone or cartilaginous sarcoma may also be found (Fig. 6.27). The pathologist should be aware of the possibility that the mesenchymal part of the tumor may be reactive and not neoplastic, and make a careful examination in order to
find the cytological features that characterize sarcomas [113, 123–126]. Imaging: On plain CT, most tumors appear as slightly hyperdense lesions because of their high vascularity and cellularity. After the contrast administration, gliosarcomas show marked enhancement and may mimic a meningioma, when the tumor is located near the skull or falx. Gliosarcomas, however, are less homogenously hyperdense than meningiomas, do not have a large base in contact with the skull, and are virtually always associated with peritumoral edema. In other cases, the CT appearance is that of an intracerebral mass with irregular and peripheral enhancement and large necrotic areas, similar to malignant astrocytomas or glioblastomas [127, 128]. On MR, they have inhomogeneous or cystic appearance with surrounded vasogenic edema. They are intraaxial but abutting a dural surface with intense heterogeneous tumor enhancement. On T2-weighted images, they have intermediate signal intensity with peripheral high-signal intensity due to the surrounding edema (Fig. 6.28) Hemorrhage and necrosis are common [129, 130]. Gliosarcoma should be included in the differential diagnosis of any tumor that appears to be intraaxial, but abuts a dural surface and shows imaging characteristics similar to gray matter on T2-weighted images. Gliosarcomas are considered to have an aggressive course, with prognosis rather similar to that of GBMs [113, 131].
6.5 Gliomatosis Cerebri
Fig. 6.27 Gliosarcoma. An area with features of osteogenic sarcoma (black asterisk) adjacent to the glial component of the tumor (white asterisk) (Hematoxylin–Eosin, original magnification ×100)
GC describes a diffuse overgrowth of the neuraxis by neoplastic glial cells with relative preservation of the underlying cytoarchitecture where no grossly discernible mass is recognizable [132, 133]. In the past, many terms have been used to describe this entity such as diffuse glioma, diffuse central Schwannosis, and diffuse cerebral gliomatosis, but in 1979, the World Health Organization’s classification of tumors of the CNS featured the term “GC,” which was first proposed by Nevins in 1938 [134–136]. The term gliomatosis is appropriate when at least two, but usually three, lobes of the brain are affected. GC involves contiguous areas and differs from multicentric glioma in which tumor masses occur at different sites [137]. Although the cerebellum, brain stem, and spinal cord may be involved, the cerebral hemispheres are primarily affected [39].
190
a
A. Drevelegas and G. Karkavelas
b
c
Fig. 6.28 Gliosarcoma. (a) Axial T1WI shows a low-signal lesion with prominent intratumoral vessels (arrow). (b) On axial T2WI the mass as well as the adjacent edema (arrow) shows high-signal intensity. (c) On axial postcontrast T1WI the tumor
shows intense homogeneous enhancement. Note the close relation of the mass with the dura that is enhanced in a way mimicking meningioma (arrow)
6 High-Grade Gliomas
Primary leptomeningeal gliomatosis, a form of GC, can also occur and may simulate meningeal carcinomatosis [138, 139]. GC is a separate histopathologic entity and is included in the WHO classification in the group of neuroepithelial neoplasm of unknown origin [138, 140]. In the majority of the cases, the overall biologic behavior classifies the GC as WHO grade III neoplasm. GC can affect all age groups, but the peak incidence is in the second to fourth decade. Men and women are affected with equal frequency [141]. The clinical symptoms of GC are nonspecific [142]; global, intellectual, and personality disturbances tend to precede local neurologic signs [143]. Pathology: Histologically, the tumor is usually moderately cellular, composed of rounded to elongated, usually bipolar cells with hyperchromatic nuclei in varying stages of differentiation. Occasional anaplastic areas may be found. Despite the astrocytic appearance of most cells, any glioma may have the features of GC. Microvascular proliferation and necrosis are absent in the classical form of GC [144]. Neoplastic cells may be accumulated around vessels and neurons or in the subpial and subependymal regions (Fig. 6.29). Although white matter is most commonly involved, gray matter can be affected as well. The underlying anatomical architecture is essentially preserved [133, 145–147]. The immunoreactivity to GFAP and S-100 protein in gliomatosis is not stable, with a fair number of neoplastic cells remaining unstained.
Fig. 6.29 Gliomatosis cerebri. Concentration of small glial cells in a diffuse pattern. Perivascular concentration of neoplastic cells (Hematoxylin–Eosin, original magnification ×100)
191
Imaging: GC appears on CT as a diffuse, poorly defined, iso- to hypodense area with more or less diffuse mass effect, which is usually not enhanced [132, 148]. Contrast enhancement is encountered only in some cases in the late stage of the disease. The pattern of enhancement varies from local to linear [149] (Fig. 6.30a). MRI is the modality of choice for evaluation of patients with GC and should be used as a primary imaging study in the evaluation of GC. On T1-weighted images GC appears as a poorly defined hypo- or isointense lesion (Fig. 6.30b). Proton-density and T2-weighted images show mild to moderate high signal intensity (Figs. 6.30c, 6.31a, 6.32a, b, 6.33a). Some parts of the lesion may show higher signal intensities. After the administration of contrast medium usually no or minimal enhancement is seen (Fig. 6.31b). However, in approximately 50% of cases, solitary or patchy enhancement may be seen as an indication of malignant progression [18, 150, 151] (Fig. 6.30d, 6.32c, 6.33b). The enhancement represents higher tumor grade and dense tumor infiltration. However, the key to the diagnosis is the diffuse, extensive, contiguous involvement with preservation of the overall cerebral structure (Fig. 6.32a,b). On perfusion images, the mean rCBV measurement is not elevated, compared with values in uninvolved white matter, reflecting the lack of vascular hyperplasia found at histolopathologic evaluation [152] (Figs. 6.31c, 6.33c). Areas with malignant degeneration show increased rCBV (Fig. 6.33c). The spectroscopic findings include mild elevation of Cho:Cr and Cho:NAA ratios as well as varying degrees of decreased NAA:Cr ratios. Areas with anaplastic transformation show an highly elevated Cho:Cr and Cho:NAA ratios (Fig. 6.33d). The presence of lactate or lipids is also indicative of malignant transformation. MRS may also be used for follow-up examinations [153, 154]. The differential diagnosis includes MS, encephalitis, ischemic disease, adenoleukodystrophy, metachromatic leukodystrophy, or subacute sclerosing panencephalitis. Diffuse infiltrating astrocytoma and glioblastoma multiforme must be considered in the differential diagnosis as well. In MS, the large plaques are more circumscribed than the lesions in GC. Differentiation from encephalitis may be difficult or even impossible. The acute onset of clinical symptoms in ischemic disease plays a determining role. Infiltrating
192
A. Drevelegas and G. Karkavelas
a
b
c
d
Fig. 6.30 Gliomatosis cerebri in a 50-year-old patient. (a) Post contrast CT shows a hypodense, slightly enhanced lesion involving the left temporal lobe. (b) On axial T1WI the lesion appears isointense with left temporal horn effacement. (c) Axial PD image shows a diffuse infiltrating hyperintense lesion involving
the entire left temporal lobe. Also note a high signal lesion involving the left cerebral peduncle (arrow). (d) Postcontrast coronal T1WI image shows partial enhancement of the lesion (crossed arrow)
193
6 High-Grade Gliomas
a
b
c
Fig. 6.31 Gliomatosis cerebri. (a) Axial T2-weighted image shows an extensive hyperintense lesion involving both temporal lobes. (b) On postcontrast T1 weighted image no enhancement is seen. (c) On perfusion color map the lesion shows decreased rCBV
194
a
A. Drevelegas and G. Karkavelas
b
c
Fig. 6.32 Gliomatosis cerebri in a 47-year-old patient. (a) PD axial image shows a diffuse hyperintense infiltration of the left temporal lobe. Also note the involvement of the right temporal lobe (arrow). (b) T2WI at a higher level shows diffuse infiltration of the left parietal lobe, right frontal lobe, left thalamus
(arrowhead), and splenium of the corpus callosum (arrow). Note the lack of sharp demarcation and the preservation of the cerebral structures. (c) Axial postcontrast T1WI shows only mild enhancement of the splenium of the corpus callosum (arrow)
astrocytoma or GBM show a focal mass and cause more neurological symptoms than the GC [155, 156]. Nevertheless, all criteria are relatively nonspecific and do not provide definite proof in the diagnosis of GC. Therefore, stereotactic biopsy in correlation with
radiological findings may provide the final diagnosis [141, 148, 149, 155, 156]. The prognosis is poor and the survival rate ranges from months to 1 or 2 years from the time of onset of symptoms [141, 148, 157, 158].
195
6 High-Grade Gliomas
a
b
c
d
Fig. 6.33 Gliomatosis cerebri. (a) Axial T2-weighted image shows diffuse periventricular hyperintensity with inhomogeneous involvement of the corpus callosum. Patient also had involvement of midbrain and right temporal lobe. (b) Postcontrast T1-weighted image shows focal enhancement of the CC indicative of malignant degeneration. (c) Perfusion and time-intensity curve image shows significant drop in the area of malignant
degeneration (green). Lower signal drop is seen in the area of diffuse infiltration (red) compared to the normal white matter (yellow). (d) The spectroscopic color map of the diffuse hyperintense area shows mild elevation of Cho:Cr, Cho:NAA, and decrease of NAA:Cr ratios. Note the significant elevation of Cho:Cr (3:11) and Cho:NAA (7:98) ratios of the enhanced area compatible with malignant degeneration
196
References 1. Davis FG, McCarthy B, Jukich P (1999) The descriptive epidemiology of brain tumors. Neuroimaging Clin N Am 9:581–594 2. Daumas-Duport C, Scheihauer B, O’Fallon J, Kelly P (1988) Grading of astrocytomas. A simple and reproducible method. Cancer 62:2152–2165 3. Coons SW, Ashby LS (1999) Pathology of intracranial neoplasms. Neuroimaging Clin N Am 9:615–649 4. Smirniotopoulos JG (1999) The new WHO classification of brain tumors. Neuroimaging Clin N Am 9(4):595–613 5. Parisi JE, Scheithauer BW (1993) Glial tumors. In: Nelson JS, Parisi JE, Scheithauer BW (eds) Principles and practice of neuropathology. Mosby, St Louis, pp 123–183 6. Okazaki H, Igaku Shoin (1989) Fundamentals of neuropathology. Morphologic basis of neurologic disorders, 2nd edn. Igaku-Shoin Medical, New York 7. Bruner JM (1994) Neuropathology of malignant gliomas. Semin Oncol 21:126–138 8. McLendon RE, Enterline DS, Tien RD et al (1998) Tumors of central neuroepithelial origin. In: Bigner DB, McLendon RE, Bruner JM (eds) Russel DS, Rubinstein LJ Pathology of tumors of the nervous system, 6th edn. Arnold, London, pp 307–571 9. Burger PC, Scheithauer BW (1994) Tumors of the central nervous system. Atlas of tumor pathology. Armed Forces Institute of Pathology, Washington, DC 10. Philippon JH, Clemenceau SH, Fauchon FH et al (1993) Supratentorial low-grade astrocytomas in adults. Neurosurgery 32:554–559 11. Castillo M, Seatlift JH, Buldin TW et al (1992) Radiologicpathologic correlation: intracranial astrocytoma. AJNR Am J Neuroradiol 13(6):1609–1616 12. Watanabe M, Tanaka R, Takeda N (1992) Magnetic resonance imaging and histopathology of cerebral gliomas. Neuroradiology 35:463–469 13. Earnest FIV, Kelly P, Scheithauer BW et al (1988) Cerebral astrocytomas: histopathologic correlation of MR and CT contrast enhancement with stereotactic. Radiology 166:823–827 14. Graff PA, Al A, Pang D (1992) Dissemination of supratentorial malignant gliomas via the cerebrospinal fluid in children. Neurosurgery 30:64–71 15. Barker FG II, Chang SM, Huhn SL et al (1997) Age and the risk of anaplasia in magnetic resonance-nonenhancing supratentorial cerebral tumors. Cancer 80:936–941 16. Kondziolka D, Lunsford LD, Martinetz AJ et al (1993) Unreliability of contemporary neurodiagnostic imaging in evaluation of suspected adult supratentorial (low-grade) astrocytoma. J Neurosurg 79:533–536 17. Ginsberg LE, Fuller GN, Hashmi M et al (1998) The significance of lack of MR contrast enhancement of supratentorial brain tumors in adults: histopathological evaluation of a series. Surg Neurol 49(4):436–440 18. Ricci PE (1999) Imaging of adult brain tumors. Neuroimaging Clin N Am 9(4):651–669 19. Russel D, Rubinstein L (1989) Tumors of contral neuroepithelial origin. In: Rubinstein LJ (ed) Pathology of tumors of the nervous system, vol 83. Williams & Wilkins, Baltimore, pp 83–350 20. Boring CC, Squires TS, Tong T (1993) Cancer statistics. CA Cancer J Clin 43:7–26
A. Drevelegas and G. Karkavelas 21. Mao Y, Desmeules M, Semenciw RM et al (1991) Increasing brain cancer rates in Canada. Can Med Assoc J 145: 1583–1589 22. Becker LE (1995) Central neuronal tumors in childhood: relationship to dysplasia. J Neurooncol 24:13 23. Paulus W, Pfeifer J (1989) Intratumoral histologic heterogeneity of gliomas. Cancer 64:442–447 24. Vandenberg ST (1992) Current diagnostic concepts of astrocytic tumors. J Neuropathol Exp Neurol 51:644–657 25. Kleihues P, Cavenee WK (1997) Tumors of the nervous system. International Agency for Research on Cancer, Lyon, pp 1–225 26. Burger PC, Scheithauer BW, Vogel FS (1991) Surgical pathology of the nervous system and its coverings, 3rd edn. Churchill Livingstone, New York 27. Wesseling P, Schlingemann RO, Rietveld FJ et al (1995) Early and extensive contribution of pericytes/vascular smooth muscle cells to microvascular proliferation in glioblastoma multiforme: an immuno-light and immunoelectron microscopy study. J Neuropathol Exp Neurol 54:304–310 28. Burger PC, Green SB (1987) Patients age, histologic features, and length of survival in patients with glioblastoma multiforme. Cancer 59:1617–1625 29. Barker FG, Davis RL, Chang SM et al (1996) Necrosis as a prognostic factor in glioblastoma multiforme. Cancer 77: 1161–1186 30. Scheithauer BW, Giordana MT, Germano I et al (1986) Anaplasia and heterogeneity of GFAP expression in gliomas. Tumori 72:163–170 31. Hirato J, Nagazato Y, Ogawa A (1994) Expression of nonglial interemediate filament proteins in gliomas. Clin Neuropathol 13:1–11 32. Burger PC, Heinz ER, Shibata T et al (1988) Topographic anatomy and CT correlations in the untreated glioblastomas multiforme. J Neurosurg 68:698–704 33. Lilja BK, Spännare B et al (1981) Reliability of CT in assesing histopathological features of malignant supratentorial gliomas. J Comput Assist Tomogr 5:625 34. Iwama T, Yamada H, Sakai N et al (1991) Correlation between magnetic resonance imaging and histopathology of intracranial gliomas. Neurol Res 13(1):48–54 35. Atlas SW (1990) Adult supratentorial tumors. Semin Roentgenol 25:130–154 36. Tovi M, Lilja A, Erickson A (1994) MR imaging in cerebral gliomas: tissue component analysis in correlation with histopahtology of whole-brain specimens. Acta Radiol 35: 495–505 37. Vestosick FT, Selker RG (1990) Brain stem and spinal metastases of supratentorial glioblastomas multiforme: a clinical series. Neurosurgery 27:516–522 38. Rees JH, Smirniotopoulos JG, Jones RV et al (1996) Glioblastoma mutliforme: radiologic-pathologic correlation. Radiographics 16:1413–1438 39. Barnard RO, Geddes JF (1987) The incidence of multifocal gliomas: a histologic study of large hemisphere sections. Cancer 60:1519–1531 40. Van Tassel P, Lee YY, Bruner JM (1988) Synchronous and metachronous malignant gliomas: CT findings. AJNR Am J Neuroradiol 9(4):725–732 41. Rao K, Levine H, Itani A et al (1980) CT findings in multicentric glioblastoma: diagnostic pathologic correlation. J Comput Tomogr 4(3):187–192
6 High-Grade Gliomas 42. Kyritsis AP, Levin VA, Yung WK et al (1993) Imaging patterns of multifocal gliomas. Eur J Radiol 16(3):163–170 43. Prather JL, Long JM, van Heertum R et al (1975) Multicentric and isolated multifocal glioblastomas multiforme simulating metastatic disease. Br J Radiol 48(565):10–15 44. Lafitte F, Morel-Precetti S, Martin-Duverneuil N et al (2001) Multiple glioblastomas: CT and MR features. Eur Radiol 11(1):131–136 45. Dean BL, Drayer BP, Bird CR et al (1994) Gliomas: classification with MR imaging. Radiology 174:411–415 46. Knopp EA, Cha S, Johnson G, Mazumdar A, Golfinos JG, Zagzag D, Miller DC, Kelly PJ, Kricheff II (1999) Glial neoplasm: dynamic contrast-enhanced T2*-weighted MR imaging. Radiology 211:791–798 47. Moller-Hartmann W, Herminghaus S, Krings T, Marquardt G, Lanfermann H, Pilatus U, Zanella FE (2002) Clinical application of proton magnetic resonance spectroscopy in the diagnosis of intracranial mass lesions. Neuroradiology 44: 371–381 48. Asari S, Katayama S, Itoh T, Tsuchida S, Ohmoto T (1994) Assessment of the pathological grade of astrocytic gliomas using an MRI score. Neuroradiology 36:308–310 49. Chang YW, Yoon HK, Shin HJ, Roh HG, Cho JM (2003) MR imaging of glioblastoma in children: usefulness of diffusion/perfusion-weighted MRI and MR spectroscopy. Pediatr Radiol 33:836–842 50. Moffat BA, Chenevert TL, Lawrence TS, Meyer CR, Johnson TD, Dong Q, Tsien C, Mukherji S, Quint DJ, Gebarski SS, Robertson PL, Junck LR, Rehemtulla A, Ross BD (2005) Functional diffusion map: a noninvasive MRI biomarker for early stratification of clinical brain tumor response. Proc Natl Acad Sci 102:5524–5529 51. Mullins ME, Schaefer PW, Sorensen AG, Halpern EF, Ay H, He J, Koroshetz WJ, Gonzalez RG (2002) CT and conventional and diffusion-weighted MR imaging in acute stroke: study in 691 patients at presentation to the emergency department. Radiology 224:353–360 52. Schaefer PW, Ozsunar Y, He J, Hamberg LM, Hunter GJ, Sorensen AG, Koroshetz WJ, Gonzalez RG (2003) Assessing tissue viability with MR diffusion and perfusion imaging. AJNR Am J Neuroradiol 24:436–443 53. Lam WW, Poon WS, Metreweli C (2002) Diffusion MR imaging in glioma: does it have any role in the preoperation determination of grading of glioma? Clin Radiol 57:219–225 54. Kono K, Inoue Y, Nakayama K, Shakudo M, Morino M, Ohata K, Wakasa K, Yamada R (2001) The role of diffusionweighted imaging in patients with brain tumors. AJNR Am J Neuroradiol 22:1081–1088 55. Krabbe K, Gideon P, Wagn P, Hansen U, Thomsen C, Madsen F (1997) MR diffusion imaging of human intracranial tumors. Neuroradiology 39:483–489 56. Tropine A, Vucurevic G, Delani P, Boor S, Hopf N, Bohl J, Stoeter P (2004) Contribution of diffusion tensor imaging to delineation of gliomas and glioblastomas. J Magn Reson Imaging 20:905–912 57. Arvinda HR, Kesavadas C, Sarma PS, Thomas B, Radhakrishnan W, Gupta AK, Kapilamoorthy TR, Nair S (2009) Glioma grading: sensitivity, specificity, positive and negative predictive values of diffusion and perfusion imaging. J Neurooncol 94(1):87–96 58. Tien RD, Felsberg GL, Brown M, MacFall J (1994) MR imaging of high-grade cerebral gliomas: value of diffusion-weighted
197 echoplanar pulse sequences. AJR Am J Roentgenol 162: 671–677 59. Fan G, Zang P, Jing Z, Guo Q (2005) Usefulness of diffusion/perfusion-weighted MRI in rat gliomas: correlation with histopathology. Acad Radiol 12:640–651 60. Goebell E, Fiehler J, Ding XQ, Paustenbach S, Nietz S, Heese O et al (2006) Disarrangement of fiber tracts and decline of neuronal density correlate in glioma patients: a combined diffusion tensor imaging and 1-H-MR Spectroscopy study. AJNR Am J Neuroradiol 27:1426–1431 61. Nucifora PG, Verma R, Lee SK, Melhem ER (2007) Diffusion tensor MR imaging and tractography. Radiology 245:367–384 62. Kamada K, Todo T, Masutani Y, Aoki S, Ino K, Takano T et al (2005) Combined use of tractography-integrated functional neuronavigation and direct fiber stimulation. J Neurosurg 102:664–672 63. Kinoshita M, Yamada K, Hashimoto N, Kato A, Izumoto S, Baba T et al (2005) Fiber-tracking does not accurately estimate size of fiber bundle in pathological condition: initial neurosurgical experience using neuronavigation and subcortical white matter stimulation. Neuroimage 25:424–429 64. Yu CS, Li KC, Xuan Y, Ji XM, Qin W (2005) Diffusion tensor tractography in patients with cerebral tumours: a helpful technique for neurosurgical planning and postoperative assessment. Eur J Radiol 56:197–204 65. Laundre BJ, Jellison BJ, Badie B, Alexander AL, Field AS (2005) Diffusion tensor imaging of the corticospinal tract before and after mass resection as correlated with clinical motor findings: preliminary data. AJNR Am J Neuroradiol 26:791–796 66. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T, Sorensen AG et al (2005) Intraoperative diffusion-tensor MR imaging: shifting of white matter tracts during neurosurgical procedures-initial experience. Radiology 234:218–225 67. Nimsky C, Ganslandt O, Hastreiter P, Wang R, Benner T, Sorensen AG et al (2005) Preoperative and intraoperative diffusion tensor imaging-based fiber tracking in glioma surgery. Neurosurgery 56:130–137 68. Baird AE, Benfield A, Schlaug G, Siewert B, Lovblad KO, Edelman RR, Warach S (1997) Enlargement of human cerebral ischemic lesions volumes measured by diffusionweighted magnetic resonance imaging. Ann Neurol 41: 581–589 69. Wilms G, Sunaert S, Flamen P (2001) Recent developments in brain tumour diagnosis. In: Demaerel P, Baert AL, Demaerel PH (eds) Recent advances in diagnostic neuroradiology. Springer, Berlin, pp 119–135 70. Cha S, Knopp EA, Johnson G, Wetzel SG, Litt AW, Zagzag D (2002) Intracranial mass lesions: dynamic contrast enhanced susceptibility-weighted echo-planar perfusion MR imaging. Radiology 223:11–29 71. Maia AC Jr, Malheiros SM, da Rocha AJ, da Silva CJ, Gabbai AA, Ferraz FA, Stavale JN (2005) MR cerebral blood volume maps correlated with vascular endothelial growth factor expression and tumor grade in nonenhancing gliomas. AJNR Am J Neuroradiol 26:777–783 72. Yang D, Korogi Y, Sugahara T, Kitajima M, Shigematsu Y, Liang L, Ushio Y, Takahashi M (2002) Cerebral gliomas: prospective comparison of multivoxel 2D chemical- shift imaging proton MR spectroscopy, echoplanar perfusion and diffusion-weighted MRI. Neuroradiology 44:656–666
198 73. Rollin N, Guyotat J, Streichenberger N, Honnorat J, Tran Minh VA, Cotton F (2006) Clinical relevance of diffusion and perfusion magnetic resonance imaging in assessing intra-axial brain tumors. Neuroradiology 48:150–159 74. Law M, Yang S, Wang H, Babb JS, Johnson G, Cha S, Knopp EA, Zagzag D (2003) Glioma grading: sensitivity, specificity and predictive value of perfusion MRI and proton spectroscopic imaging compared with conventional MR imaging. AJNR Am J Neuroradiol 24:1989–1998 75. Aronen HJ, Gazit IE, Louis DN et al (1994) Cerebral blood volume maps of gliomas: comparison with tumor grade and histologic findings. Radiology 191:41–51 76. Sugahara T, Korogi Y, Kochi M et al (1998) Correlation of MR imaging-determined cerebral blood volume maps with histologic and angiographic determination of vascularity of gliomas. AJR Am J Roentgenol 171:1479–1486 77. Knopp EA, Cha S, Johnson G et al (1999) Glial neoplasms: dynamic contrast-enhanced T2*-weighted MR imaging. Radiology 211:791–798 78. Shin JH, Lee HK, Kwunet BD et al (2002) Using relative cerebral blood flow and volume to evaluate the histopathologic grade of cerebral gliomas: preliminary results. AJR Am J Roentgenol 179:783–789 79. Lev MH, Rosen BR (1999) Clinical applications of intracranial perfusion MR imaging. Neuroimaging Clin N Am 9:309–331 80. Lev MH, Ozsunar Y, Henson JW et al (2004) Glial tumor grading and outcome prediction using dynamic spin-echo MR susceptibility mapping compared with conventional contrastenhanced MR: confounding effect of elevated rCBV oligodendrogliomas. AJNR Am J Neuroradiol 25:214–221 81. Cha S, Tihan T, Crawford F, Fischbein NJ, Chang S, Bollen A, Nelson SJ, Prados M, Berger MS, Dillon WP (2005) Differentiation of low-grade oligodendrogliomas from lowgrade astrocytomas by using quantitative blood-volume measurementsderivedfromdynamicsusceptibilitycontrast-enhanced MR imaging. AJNR Am J Neuroradiol 26:266–273 82. Law M (2005) Perfusion and MRS for brain tumor diagnosis. In: Edelman RR (ed) Clinical magnetic resonance imaging. Saunders, Philadelphia, pp 1215–1247 83. Pomper MG, Port JD (2000) New techniques in MR imaging of brain tumors. Magn Reson Imaging Clin N Am 8:691–713 84. Castillo M, Kwock L (1998) Proton MR spectroscopy of common brain tumors. Neuroimaging Clin N Am 8:733–752 85. Brandao LA (2004) Inroduction and Technique. In: Bandao LA (ed) MR Spectroscopy of the Brain. Lippincott Williams & Wilkins, Philadelphia, pp 1–15 86. Gupta RK, Cloughesy TF, Sinha U, Garakian J, Lazareff J, Rubino G, Rubino L, Becker DP, Vinters HV, Alger JR (2000) Relationships between choline magnetic resonance spectroscopy, apparent diffusion coefficient and quantitative histopathology in human glioma. J Neurooncol 50:215–226 87. Tedeschi G, Lundbom N, Raman R, Bonavita S, Duyn JH, Alger JR, Di Chiro G (1997) Increased choline signal coinciding with malignant degeneration of cerebral gliomas: a serial proton magnetic resonance spectroscopy imaging study. J Neurosurg 87:516–524 88. Urenjak J, Williams SR, Gadian DG, Noble M (1992) Specific expression of N-acetylaspartate in neurons, oligodendrocyte-type-2 astrocyte progenitors, and immature oligodendrocytes in vitro. J Neurochem 59:55–61
A. Drevelegas and G. Karkavelas 89. Negendank WG, Sauter R, Brown TR, Evelhoch JL, Falini A, Gotsis ED, Heerschap A, Kamada K, Lee BC, Mengeot MM, Moser E, Padavic-Shaller KA, Sanders JA, Spraggins TA, Stillman AE, Terwey B, Vogl TJ, Wicklow K, Zimmerman RA (1996) Proton magnetic resonance spectroscopy in patients with glial tumors: a multicenter study. J Neurosurg 84:449–458 90. Haimes AB, Zimmerman RD, Morgello S, Weingarten K, Becker RD, Jennis R, Deck MD (1989) MR imaging of brain abscesses. AJR Am J Roentgenol 152:1073–1085 91. Kim YJ, Chang KH, Song IC, Kim HD, Seong SO, Kim YH, Han MH (1998) Brain abscess and necrotic or cystic brain tumor: discrimination with signal intensity on diffusion-weighted MR imaging. AJR Am J Roentgenol 171:1487–1490 92. Noguchi K, Watanabe N, Nagayoshi T, Kanazawa T, Toyoshima S, Shimizu M, Seto H (1999) Role of diffusionweighted echo-planar MRI in distinguishing between brain brain abscess and tumor: a preliminary report. Neuroradiology 41:171–174 93. Stadnik TW, Chaskis C, Michotte A, Shabana WM, van Rompaey K, Luypaert R, Budinsky L, Jellus V, Osteaux M (2001) Diffusion-weighted MR imaging of intracerebral masses: comparison with conventional MR imaging and histologic findings. AJNR Am J Neuroradiol 22:969–976 94. Batra A, Tripathi RP (2004) Atypical diffusion-weighted magnetic resonance findings in glioblastoma multiforme. Australas Radiol 48:388–391 95. Tung GA, Evangelista P, Rogg JM, Duncan JA 3rd (2001) Diffusion-weighted MR imaging of rim-enhancing brain masses: is markedly decreased water diffusion specific for brain abscess? AJR Am J Roentgenol 177:709–712 96. Hakyemez B, Erdogan C, Yildirim N, Parlak M (2005) Glioblastoma multiforme with atypical diffusion-weighted MR findings. Br J Radiol 78:989–992 97. Holtas S, Geijer B, Stromblad LG, Maly-Sundgren P, Burtscher IM (2000) A ring-enhancing metastasis with central high signal on diffusion-weighted imaging and low apparent diffusion coefficients. Neuroradiology 42: 824–827 98. Lai PH, Li KT, Hsu SS, Hsiao CC, Yip CW, Ding S, Yeh LR, Pan HB (2005) Pyogenic brain abscess: findings from in vivo 1.5-T and 11.7-T in vitro proton MR spectroscopy. AJNR Am J Neuroradiol 26:279–288 99. Stadnik TW, Demaerel P, Luypaert RR, Chaskis C, Van Rompaey KL, Michotte A, Osteaux MJ (2003) Imaging tutorial: differential diagnosis of bright lesions on diffusion-weighted MR images. Radiographics 23:e7 100. Chiang IC, Kuo YT, Lu CY, Yeung KW, Lin WC, Sheu FO, Liu GC (2004) Distinction between high-grade gliomas and solitary metastases using peritumoral 3-T magnetic resonance spectroscopy, diffusion, and perfusion imagings. Neuroradiology 46:619–627 101. Lu S, Ahn D, Johnson G, Law M, Zagzag D, Grossman RI (2004) Diffusion-tensor MR imaging of intracranial neoplasia and associated peritumoral edema: introduction of the tumor infiltration index. Radiology 232:221–228 102. Law M, Cha S, Knopp EA, Johnson G, Arnett J, Litt AW (2002) High-grade gliomas and solitary metastases: differentiation by using perfusion and proton spectroscopic MR imaging. Radiology 222:715–721
6 High-Grade Gliomas 103. Guo AC, Cummings TJ, Dash RC, Provenzale JM (2002) Lymphomas and high-grade astrocytomas: comparison of water diffusibility and histologic characteristics. Radiology 224:177–183 104. Hiwatashi A (2003) Brain neoplasm. In: Moritani T, Ekholm S, Westesson PL (eds) Diffusion-weighted MR imaging of the brain. Springer, Heidelberg, pp 161–179 105. Hartmann M, Heiland S, Harting I, Tronnier VM, Sommer C, Ludwig R, Sartor K (2003) Distinguishing of primary cerebral lymphoma from high-grade glioma with perfusion-weighted magnetic resonance imaging. Neurosci Lett 338:119–122 106. Harting I, Hartmann M, Jost G, Sommer C, Ahmadi R, Heiland S, Sartor K (2003) Differentiating primary central nervous system lymphoma from glioma in humans using localised proton magnetic resonance spectroscopy. Neurosci Lett 342:163–166 107. Cha S, Pierce S, Knopp EA, Johnson G, Yang C, Ton A, Litt AW, Zagzag D (2001) Dynamic contrast-enhanced T2*-weighted MR imaging of tumefactive demyelinating lesions. AJNR Am J Neuroradiol 22:1109–1116 108. Arnold DL, Matthews PM, Francis GS, O’Connor J, Antel JP (1992) Proton magnetic resonance spectroscopic imaging for metabolic characterization of demyelinating plaques. Ann Neurol 31:235–241 109. Davie CA, Hawkins CP, Barker GJ, Brennan A, Tofts PS, Miller DH, McDonald WI (1994) Serial proton magnetic resonance spectroscopy in acute multiple sclerosis lesions. Brain 117:49–58 110. Mullins ME, Bares GD, Schaefer PW, Hochberg FH, Gonzalez RG, Lev MH (2005) Radiation necrosis versus glioma recurrence: conventional MR imaging clues to diagnosis. AJNR Am J Neuroradiol 26:1967–1972 111. Sugahara T, Korogi Y, Tomiguchi S, Shigematsu Y, Ikushima I, Kira T, Liang L, Ushio Y, Takahashi M (2000) Posttherapeutic intraaxial brain tumor: the value of perfusion sensitive contrast-enhanced MR imaging for differentiating tumor recurrence from nonneoplastic contrast-enhancing tissue. AJNR Am J Neuroradiol 21: 901–919 112. Law M, Hamburger M, Johnson G, Inglese M, Londono A, Golfinos J, Zagzag D, Knopp EA (2004) Differentiating surgical from non-surgical lesions using perfusion MR imaging and proton MR spectroscopic imaging. Technol Cancer Res Treat 3:557–565 113. Morantz RA, Feigin I, Ransohoff J (1976) Clinical and pathological study of 24 cases of gliosarcoma. J Neurosurg 45:398–408 114. Nitta H, Hayase H, Moriyama Y et al (1993) Gliosarcoma of the posterior cranial fossa: MRI findings. Neuroradiology 35(4):279–280 115. Odhaki H, Biernat W, Reis R, Hegi M, Kleihues P (2000) Gliosarcoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumors of the nervous system. IARC, Lyon, pp 42–44 116. Sreenan JJ, Prayson RA (1997) Gliosarcoma. A study of 13 tumors, including p53 and CD34 immunohistochemistry. Arch Pathol Lab Med 121:129–133 117. Perry JR, Ang LC, Bilbao JM, Muller PJ (1995) Clinicopathologic features of primary and postradiation cerebral gliosarcoma. Cancer 75:2910–2918
199 118. Beute BJ, Fobben GS, Hubschmann O et al (1991) Cerebellar gliosarcoma: report of a probable radiationinduced neoplasm. AJNR Am J Neuroradiol 12:554–556 119. Kim DS, Kang SK, Chi JG (1999) Gliosarcoma: a case with unusual epithelial feature. J Korean Med Sci 14(3):345–350 120. Lach M, Wallace CJ, Kreek J, Curry B (1996) Radiationassociated gliosarcoma. Can Assoc Radiol J 47(3):209–212 121. Cerame MA, Buthikonda M, Kohli CN (1985) Extraneural metastases in gliosarcoma. A case report and review of the literature. Neurosurgery 17:413–418 122. Meis JM, Martz KL, Nelson JS (1991) Mixed glioblastoma multiforme and sarcoma. A clinicopathologic study of 26 radiation therapy oncology group cases. Cancer 67: 2342–2349 123. Mathews T, Moossy J (1974) Gliomas containing bone and cartilage. J Neuropathol Exp Neurol 33:456–471 124. Barnard RO, Bradford R, Scott T et al (1986) Gliomyosarcoma: report of a case of rhabdomyosarcoma arising in a malignant glioma. Acta Neuropathol 69:23–27 125. Tada T, Katsuyama T, Aoki T et al (1986) Mixed glioblastoma and sarcoma with osteoid – chondral tissue. Clin Neuropathol 6:160–163 126. Haddad SF, Moore SA, Shelper RL et al (1992) Smooth muscle can comprise the sarcomatous component of gliosarcomas. J Neuropathol Exp Neurol 51:493–498 127. Kohshi K, Munaka M, Vamada H et al (1992) Gliosarcoma associated with von Recklinghausen’s disease: a case report. No Shinkei Geka 20(1)):1195–1198 128. Maiuri F, Stella L, Benvenuti D et al (1990) Cerebral gliosarcomas: correlation of computed tomographic findings, surgical aspects, pathological features and prognosis. Neurosurgery 26:261–267 129. Sakurai T, Abe J, Hayashi T et al (1993) A case of gliosarcoma associated with large cyst. No Shinkei Geka 21(7): 637–640 130. Dwyer KW, Naul LG, Hise JH (1996) Gliosarcoma: MR features. J Comput Assist Tomogr 20(5):719–723 131. Galanis E, Buckner JC, Dinapoli RP et al (1998) Clinical outcome of gliosarcoma compared with glioblastomas multiforme: North Central Cancer. Treatment group results. J Neurosurg 89:425–430 132. Ross IB, Robitaille Y, Villemure JG et al (1991) Diagnosis and management of gliomatosis cerebri: recent trends. Surg Neurol 36:31–40 133. Artigas J, Cervos Navaro J et al (1985) Gliomatosis cerebri: clinical and histological findings. Clin Neuropath 4: 135–148 134. Nevins (1938) Gliomatosis cerebri. Brain 61:170–191 135. Okaraki H (1989) Fundamentals of neuropathology, 2nd edn. Agaku-Shoin, Tokyo, pp 204–275 136. Scheinker I, Evans J (1943) Diffuse cerebral gliomatosis. J Neuropathol 2:178–189 137. Del Carpio O, Korah I, Salazar A et al (1996) Gliomatosis cerebri. Radiology 198:831–835 138. Kleihues P, Burger PC, Scheitauer BW (1993) The new WHO classification of brain tumors. Brain Pathol 3:255–268 139. Rippe DJ, Boyko OB, Fuller GN et al (1990) Gadopentetatedimiglumine-enhanced MR imaging of gliomatosis cerebral: appearance mimicking leptomeningeal tumor dissemination. AJNR Am J Neuroradiol 11:800–801
200 140. Leproux F, Melanson D, Mercier C et al (1993) Leptomeningeal gliomatosis. MR findings. J Comp Assist Tomogr 17:317–320 141. Felsberg GJ, Silver SA, Brown MT et al (1994) Radiologicpathologic correlation: gliomatosis cerebri. AJNR Am J Neuroradiol 15:1745–1753 142. Couch JR, Weiss SA (1976) Gliomatosis cerebri: report of four cases and review of the literature. Neurology 24: 504–511 143. Dickson DW, Horoupian DS, Thal LJ et al (1988) Gliomatosis cerebri presenting with hydrocephalus and dementia. AJNR Am J Neuroradiol 9:200–202 144. Yip M, Fisch C, Lamarche JB (2003) Gliomatosis cerebri affecting the entire neuraxis. Radiographics 23:247–253 145. Kandler RH, Smith CML, Broome JC et al (1991) Gliomatosis cerebri. A clinical radiological and pathological report of four cases. Br J Neurosurg 5:187–193 146. Balko MG, Blisard KS, Samaha FJ (1992) Oligodendroglial gliomatosis cerebri. Hum Pathol 23:706–707 147. Burger PC, Scheithauer BW (1994) Tumors of the central nervous system. Armed Forces Institute of Pathology, Washington, DC 148. Geremia GK, Wollman R, Foust R (1988) Computed tomography at gliomatosis cerebri. J Comput Assist Tomogr 12:698–701 149. Shin YM, Chang KH, Han MH et al (1993) Gliomatosis cerebri: comparison of MR and CT features. AJR Am J Roentgenol 161:859–862
A. Drevelegas and G. Karkavelas 150. Spagnoli MV, Grossman RI, Packer RJ et al (1987) Magnetic resonance imaging determination of gliomatosis cerebri. Neuroradiology 29:15–18 151. Freund M, Ha´ hnel S, Sommer C et al (2001) CT and MRI findings in gliomatosis cerebri: a neuroradiologic and neuropathologic review of diffuse infiltrating brain neoplasm. Eur Radiol 11(2):309–316 152. Yang S, Wetzel S, Cha S (2002) Dynamic contrast-enhanced T2*-weighted MR imaging of gliomatosis cerebri. AJNR Am J Neuroradiol 23:350–355 153. Guzman-de-Villoria JA, Sanchez-Gonzalez J, Muroz L, Reig S, Benito C, Garcna-Barreno P, Desco M (2007) 1H MR spectroscopy in the assessment of gliomatosis cerebri. AJR Am J Roentgenol 188:710–714 154. Bendszus M, Warmuth-Metz M, Klein R, Burger R, Schichor C, Tonn JC, Solymosi L (2000) MR spectroscopy in gliomatosis cerebri. AJNR Am J Neuroradiol 21:375–380 155. Fallentin E, Skriver E, Herning M, Broholm H (1997) Gliomatosis cerebri: an appropriate diagnosis? Acta Radiol 38:381390 156. Enterline DS, Davey NC, Tien RD (1995) Gliomatosis cerebri (case of the day). AJR Am J Roentgenol 165:212–215 157. Yanaka K, Kameraki T, Kobayashi E et al (1992) MR imaging of diffuse glioma. AJNR Am J Neuroradiol 13:349–351 158. Mena IX, Olivates DA, del Brutto OH et al (2000) Gliomatosis cerebri: clinico-pathological and neuroimaging characteristics and the results of treatment with radiotherapy. Rev Neurol 31(2):101–106
7
Pineal Tumors Antonios Drevelegas, Argyris K. Strigaris, and Christiana H. Samara
Contents 7.1 Introduction............................................................. 201 7.2 Normal Anatomy, Embryology, and Clinical Symptoms.......................................... 201 7.3 Germ Cell Tumors.................................................. 203 7.3.1 Germinomas.............................................................. 203 7.3.2 Teratomas.................................................................. 205 7.3.3 Endodermal Sinus or Yolk-Sac Tumors.................... 205 7.3.4 Choriocarcinoma...................................................... 205 7.3.5 Embryonal Cell Carcinoma...................................... 205 7.4 Tumors of Pineal Cell Origin................................. 207 7.4.1 Pineocytomas............................................................ 208 7.4.2 Parenchymal Pineal Tumors of Intermediate Differentiation (PPTID)............................................ 208 7.4.3 Pineoblastomas......................................................... 209 7.5 Other Tumors and Nonneoplastic Masses............ 210
7.1 Introduction The term “pineal area” refers either to the pineal gland, which is situated in the deep part of the brain, or to the pineal region, which includes the gland itself and its surrounding structures. Pineal tumors represent 3–8% of intracranial tumors in children and 0.4–1% of brain tumors in adults [1]. In Asia the pineal tumors are most often found among the pediatric population where the incidence is about 10–12.5% [2]. Although CT and MRI characteristics of pineal tumors are nonspecific, the localization of a tumor in the pineal region allows one to choose the most appropriate therapeutic method. In this chapter we will deal with the germ cell and the pineal cell tumors, which represent the vast majority of the pineal gland tumors.
References............................................................................ 213
7.2 Normal Anatomy, Embryology, and Clinical Symptoms
A.K. Strigaris Institute of Psychiatry, King’s College London, UK
The pineal gland or epiphysis is a small cone-shaped formation attached to the roof of the third ventricle. It consists of a network of rich vascular connective tissue trabeculae in the meshes of which are found glial cells and pineal cells. Pineal cells are of variable size with a pale nucleus, granular argentophilic cytoplasm, and relatively few branching processes. They may represent modified nerve cells since glial stains do not stain them. True nerve cells do not appear to be present [3]. The surrounding structures of the pineal gland are the following [4]:
C.H. Samara Department of Radiology, Athens General Hospital, Greece
• Tectum: superior-inferior colliculi. • Brainstem – thalamus – splenium of corpus callosum – posterior part of third ventricle, and aqueduct.
A. Drevelegas (*) Department of Radiology, AHEPA university Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece e-mail:
[email protected]
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_7, © Springer-Verlag Berlin Heidelberg 2011
201
202
• Subarachnoid cisterns as quatrigeminal plate, ambient, and velum interpositum. • Dura – tentorial apex and vessels: internal cerebral vein, vein of Galen, basal veins, posterior choroidal artery, and posterior cerebral artery (Fig. 7.1). a
A. Drevelegas et al.
The pathological findings could be masses originating from the pineal gland itself or from the surrounding structures. Tumors located in the pineal gland may originate either from germ cell tumors or from pineal parenchyma cells or from other cells. Germ cell tumors b
c
Fig. 7.1 Normal anatomy of the pineal region. (a) Midsagittal section of a gross specimen. (b) Sagittal T1-weighted image and (c) axial T1-weighted MR image. Pineal gland one, tectum two,
splenium of the corpus callosum three, thalamus four, internal cerebral vein five, basal vein of Rosenthal six, aqueduct (arrowheads), quadrigeminal plate cistern (asterisk)
203
7 Pineal Tumors Table 7.1 Pineal gland lesions Germ cell tumors
Germinoma, teratoma, endodermal sinus tumor (yolk sac tumor), choriocarcinoma, embryonal cell carcinoma, and mixed germ cell tumor
Pineal cell tumors
Pineocytoma, pineal tumor of intermediate differentiation (PTID), and pineoblastoma
Other cell origin tumors and nonneoplastic masses
Astrocytoma, glioma (midbrain, thalamus, corpus callosum), meningioma, metastasis, vascular malformations (cavernous angioma, AVM), miscellaneous (lipoma, epidermoid, dermoid), pineal cyst
represent the most common tumors in the pineal region, where 65% of them are pure germinomas. The pineal parenchyma cell origin tumors include the pineocytoma, pineal tumor of intermediate differentiation (PTID), and pineoblastoma. Tumors from other cells could be astrocytoma, metastasis, meningioma, etc. The pathology is summarized in the Table 7.1 The symptoms produced by the pineal gland tumors are: nausea, vomiting, headache, and Parinaud’s syndrome (supranuclear impairment of upward gaze, defective convergence, and slow pupillary reaction to light). Less commonly pineal region masses are associated with precocious puberty and hypogonadism [5].
7.3 Germ Cell Tumors Tumors of germ cell origin are the most common of the pineal gland (50–70%). Several different tumors constitute the family of germ-cell neoplasms including a variety of cell types. They appear most often between the ages of 10 and 20 years, the distribution being skewed toward 20. In the embryonic life, germ cells separate from the somatic cells and migrate to the region of the developing gonads. These cells are also widely distributed within the developing embryo and they regress in all locations except that of the gonads, the thymus and the pineal, and supracellar region. Germ cell tumors include germinoma, teratoma (including immature- and malignant-type teratocarcinoma), embryonal carcinoma, endodermal sinus – yolk sac tumor, choriocarcinoma, and mixed germ tumors. The malignancy of each of them depends on the embryonic stage of development: primordial germ cell
(germinoma), the embryonic differentiated derivative (teratoma) of the pluripotential stem cell of the embryo proper (embryonal carcinoma), and extra embryonic differentiated derivatives which form the yolk-sac tumor and trophoblast (choriocarcinoma) [1, 6].
7.3.1 Germinomas Germinomas are the most common tumors of pineal region (greater than 50%), but can also be found in other locations such as the supracellar region (infundibular stalk), the anterior third ventricle, the basal ganglia, and thalamus [7]. They occur most often during the second decade and almost exclusively in men. Pineal germinomas have a male/female ratio of 3:1 while suprasellar germinomas have a 1:1 ratio [8]. Germinomas grow slowly and spread either to the adjacent tissue or via subependymal or the subarachoid space [9]. Placental alkaline phospahatase in serum and cerebro spinal fluid (CSF) has been suggested as specific tumor marker for germinomas [10]. Central nervous system (CNS) germinomas are histologically almost identical to testicular seminomas and ovarian dysgerminomas. These tumors have a “two cells” pattern and they consist from lobules of primitive germ-cells embedded in a matrix of lymphocytes or lymphocyte-like cells. On computerized tomography (CT) scans a slightly hyperdense homogeneous mass appears in the majority of germinomas. The tumor often displaces or engulfs a calcified pineal gland. After the administration of contrast medium they show intense and homogeneous enhancement (Fig. 6.2a, b). On MRI the mass is slightly hypo- to isointense to the brain on T1 and iso- to hyperintense on T2-weighted images. Cystic areas may also be present. Contrast should be given to all patients because germinomas and their metastases (subarachnoid, subependymal) show a noticeable contrast enhancement [1, 4, 10] (Figs. 7.2c, d and 7.3). Germinomas tend to extend into the posterior aspect of the third ventricle (Fig. 6.2e). Invasion of the adjacent structures (thalami or midbrain) can be seen. Lepto meningeal dissemination of the tumor is frequently seen. It has been estimated that one third of patients show spinal dissemination at the time of diagnosis [11]. Since germinomas are radiosensitive, they should rapidly respond to radiation, while gliomas and teratomas often show persistent tumor mass after radiation
204
a
c
Fig. 7.2 Atypical germinoma: axial CT without (a) and with intravenous administration of contrast at a slightly higher level (b) show a hyperdense pineal mass displacing the calcified pineal gland. Note the obstructive hydrocephalus and the intense contrast enhancement of the solid component of the mass. On sagittal T1- (c) and axial T2-weighted (d) images the solid com-
A. Drevelegas et al.
b
d
ponent of the tumor is isointense relative to gray matter, while the cystic component is hypo- on T1WI and hyperintense on T2-weighted image. On postcontrast axial T1-weighted image (e) the solid component of the tumor shows intense enhancement. Note the extension of the tumor into the posterior aspect of the third ventricle
205
7 Pineal Tumors
component of the tumor appears on T1-weighted images as areas of high signal intensity. Both CT and MRI show heterogeneous or ring-like contrast enhancement [1, 4, 14, 16] (Fig. 7.4). Pineal teratomas may rupture either during surgery or spontaneously causing chemical meningitis [17]. Malignant teratomas tend to invade surrounding structures such as the tectum, the brainstem, and the splenium of corpus callosum.
7.3.3 Endodermal Sinus or Yolk-Sac Tumors
Fig. 7.2 (continued)
[12, 13]. The 5-year survival rate following radiation therapy in these patients ranges from 75% to 85% in the literature [14].
7.3.2 Teratomas Teratomas are, after germinomas, the second most common tumors of the pineal gland and account for 15% of all pineal region tumors. Teratomas usually occur in an earlier age group (first decade) and they exhibit the same male predominance as germinomas. Teratomas can be classified into two categories: benign (mature) teratomas, which contain fully differentiated tissue and malignant (immature) teratomas, which contain some primitive tissue [13]. Teratomas have a wide variation in the degree of histologic maturity and they demonstrate a variable biological behavior and clinical course. They usually are inhomogeneous masses because of the presence of calcifications, hemorrhage, fat, and cystic or necrotic areas [15]. Calcifications can be linear or nodular and are better detected with CT imaging [14]. On T1- and T2-weighted MR images teratoma shows a heterogeneous mass with areas of low and high signal intensities. The fatty or lipid
The tumors usually appear during the second and third decade and they are strongly male-predominant at a ratio of 5:1 [18]. The appearance of alpha-fetoprotein (AFP) is a strong indication for the presence of yolk-sac tumor [19]. The tumor has nonspecific CT imaging characteristics. On MR images, the tumor has iso-intense signal on T1-weighted images and an inhomogeneous hyperintense signal on T2-weighted images [1]. Abnormal calcifications and cystic components may be present [6].
7.3.4 Choriocarcinoma Choriocarcinomas usually appear between the ages of 10 and 15 years and they almost exclusively appear in males [18]. This type of germ cell tumor typically produces b-human chorionic gonadotropin in the serum and CSF [20]. Imaging findings show a heterogeneous mass with irregular margins invading the normal adjacent tissue. In precontrast CT the tumor may have a hyperdense appearance because of the presence of a hemorrhagic component. After administration of contrast medium, the tumor is intensively enhanced. On MRI, the tumor is heterogeneous with a large hemorrhagic component on T1- and T2-weighted images. On angiogram neovascularity and small areas of aneurysmal dilatation can appear [1, 16].
7.3.5 Embryonal Cell Carcinoma The appearance of both AFP and chorionic gonadotropin indicates the presence of the third type of tumor, the embryonal cell carcinoma [21]. It consists of
206
A. Drevelegas et al.
a
b
c
d
Fig. 7.3 Germinoma. (a) Axial T1- weighted image shows an oval mass in pineal region isointense relative to gray matter. (b) On T2-weighted image the mass is also isointense. (c) and
(d) axial and sagittal T1- weighted images after intravenous administration of contrast show intense homogeneous enhancement of the tumor
207
7 Pineal Tumors
a
b
c
d
Fig. 7.4 Teratoma. Axial T1-weighted image (a) show an inhomogeneous pineal mass with high and isointense signal intensities. Fatty portion of the mass has high intensity equal to the subcutaneous fat. Sagittal post-contrast T1-weighted MR image (b) shows inhomogeneous enhancement of mass. Low magnification of mature teratoma (c) shows several tissue elements such
as pseudostratified columnar epithelium (respiratory type), sweat glands, mucinus epithelium, fat tissue, and cartilage (H&E 20×). Higher magnification of mature teratoma (d) shows mucinus epithelium (left), columnar epithelium (right) and cartilage (middle) (H&E 100×)
undifferentiated embryonic epithelial cells [14]. Histo logically hemorrhage, mitoses and necrosis are usually present [22]. On CT the tumor appears as an isointense mass with parenchymal calcifications and intense contrast enhancement. On T1- weighted MR images emb ryonal cell carcinomas are usually isointense and on T2-weighted images, hyperintense.
7.4 Tumors of Pineal Cell Origin Pineal parenchymal tumors (PPTs) represent 15–30% of all pineal region tumors [14]. They are classified as pineocytoma (WHO grade II), pineal parenchymal tumor of intermediate differentiation (PPTID), and pineoblastoma (grade IV). They have equal distribution in males
208
and females. PPTs arise from neuroepithelial cells of the gland itself. Actually pineoblastomas arise from small cells, while pineocytomas from large cells. Although they derive from different cell types they often coexist within the same tumor. The cells in pineoblastomas and pineocytomas have a greater tendency to calcify than do germinal tumor cells [23–25].
A. Drevelegas et al.
pineoblastomas and their cells are richer in cytoplasm. On CT they appear as slightly hyperdense masses with a noticeable amount of calcification [7, 16, 28]. On MRI the tumors appear homogeneous with high signal intensity on T2-weighted images and low or isointense signal on T1-weighted images reflecting the high amount of cytoplasm (Fig. 7.5). The tumors exhibit significant contrast enhancement, and for large tumors heterogeneity is observed [1, 10, 16].
7.4.1 Pineocytomas Pineocytomas account for 0.4–1% of intracranial brain tumors and represent approximately 45% of all PPTs. They appear between the ages of 18 and 50 and show no sex predilection. Pineocytomas develop in pineal parenchymal cells; they are slow-growing benign tumors (WHO grade II), which do not usually seed across the CSF pathways [26]. On histology, the tumors have lobular architecture separated by fibrovascular septa and are composed of small mature cells resembling pineocytes, which often form pineocytomatous rosettes [27]. They are less cellular tumors than
a
7.4.2 Parenchymal Pineal Tumors of Intermediate Differentiation (PPTID) PPTIDs are rare tumors accounting for 10–20% of all PPTs. They affect all ages with a peak incidence in young adults and a slight female predominance. PPTs of intermediate differentiation are potentially aggressive tumors, which may correspond to WHO grade II or III classification, but definite grading criteria have yet to be established [26].
b
Fig. 7.5 Pineocytoma (a) axial PD-weighted MR image. The pineal mass shows homogeneous high signal intensity reflecting the increased water content of tumor. (b) Post-contrast axial T1-weighted MR image shows intense homogeneous enhancement of the mass
209
7 Pineal Tumors
Histologically, they arise from pineocytes or their precursors and are characterized by moderately high cellularity, mild-to-moderate nuclear atypia, and moderate mitotic activity [29–31]. On CT they appear iso- or hyperdense and demonstrate homogeneous or heterogeneous enhancement. Tumoral calcification may be seen (Fig. 7.6a). On MRI the tumor appears hypointense on T1- and hyperintense on T2-weighted images. After the administration of contrast medium intense homo- or heterogeneous enhancement is seen [30] (Fig. 7.6b–e).
7.4.3 Pineoblastomas Pineoblastomas (WHO grade IV) may appear at any age but most often during the first and second decade
a
Fig. 7.6 PPTID. Axial post-contrast CT (a) shows a homogeneously enhanced pineal tumor with peripheral calcification (arrows). On axial, T1- (b) and T2-weighted images (c), the solid component of the tumor appears isointense to the gray-
of life with a predilection for young children. They are a subtype of primitive neuroectodermal tumors (PNET) with biological behavior similar to that of medulloblastomas or neuroblastomas [4]. There is a rare association of pineoblastoma with bilateral retinoblastoma [32]. Often at the time of initial diagnosis metastases in subarachnoid or in subependymal spaces are present. They grow rapidly to a size of over 4 cm. They are unencapsulated tumors, often invade the adjacent brain parenchyma (corpus callosum, tegmentum, vermis, thalamus), and show CSF spread. The tumors are usually irregularly shaped and their morphology is heterogeneous due to the presence of hemorrhage and necrosis [24, 27, 29]. Calcification is rarely seen. On MRI the tumors show a heterogeneous pattern which is hypo- to isointense to gray matter on T1-weighted images and iso- to hyperintense on
b
matter, while the peripheral calcification hyperintense. On axial (d) and sagittal (e) post-contrast T1-weighted images the tumor shows homogeneous enhancement
210
A. Drevelegas et al.
c
d
e
Fig. 7.6 (continued)
T2-weighted images due to the high cellularity of the tumor. Pineoblastomas exhibit intense contrast enhancement (Fig. 7.7). The presence of small cystic necrotic areas is also occasionally observed [1, 14]. The 5-year survival rate of pineoblastomas is about 50% [24].
7.5 Other Tumors and Nonneoplastic Masses Other tumors include astrocytic tumors, meningiomas, ependymomas, dermoid/epidermoid cysts, pineal cysts (Fig. 7.8), and pineal metastases. Nonneoplastic masses
211
7 Pineal Tumors
a
b
c
d
e
Fig. 7.7 Pineoblastoma. Sagittal T1-weighted image (a) shows an isointense pineal mass which is iso- and hyperintense on axial PD-weighted image (b). Sagittal T1-weighted image (c) after contrast administration shows marked homogeneous enhancement. Note the invasion of the tumor to the adjacent structures. Coronal T1-weighted image (d) 2 years after the operation shows
local relapse and subependymal spread. The mass is more heterogeneous than before with necrotic areas and local invasion of surrounding structures (tectum, thalamus, brainstem). Microscopic view of cellular pinealoblastoma resembling medulloblastoma (e). The tumor consists of solid cords separated by delicate connective tissue stroma (H&E 40×)
212
a
A. Drevelegas et al.
b
c
Fig. 7.8 Pineal cyst. Oval lesion in the region of the pineal gland with low signal intensity on T1-weighted images (a) and high signal intensity on T2-weighted images (b). No enhance-
ment is seen on sagittal post-contrast T1-weighted images (c). The adjacent vessels are stretched around the lesion
7 Pineal Tumors
are cavernous angiomas, vein of Galen aneurysms, and congenital lesions such as lipomas.
References 1. Tien RD, Barkovich AJ, Edwards MSB (1990) MR imaging of pineal tumors. AJNR 11:557–565 2. Koide O, Watanabe Y, Sato K (1980) Pathological survey of intracranial germinoma and pinealoma in Japan. Cancer 45: 2119–2130 3. Truex RC, Carpenter MB (1969) Human neuroanatomy. Williams & Wilkins, Baltimore 4. Osborne AG (1994) Diagnostic neuroradiology. MosbyYear Book, St. Louis 5. Kirkwood JR (1990) Essential of neuroimaging. Churchill Livingstone, New York 6. Jennings MT, Gelman R, Hochberg F (1985) Intracranial germcell tumors: natural history and pathogenesis. J Neurosurg 63:155–167 7. Zimmerman RA, Bilanluk LT, Wood JH, Bruce DA, Schut L (1980) Computed tomography of pineal, parapineal, and histologically related tumors. Radiology 137:669–677 8. Hoffman HJ, Otsubo H (1991) Intracranial gem-cell tumors. J Neurosurg 74:545–551 9. Soejima T, Takeshita I (1987) Computed tomography of germinomas in basal ganglia and thalamus. Neuroradiology 29:366–370 10. Luh GY, Bird CG (1999) Imaging of brain tumors in the pediatric population. Radiol Clin N Am 9(4):691–716 11. Maity A, Shu HK, Janss A et al (2004) Craniospinal radiation in the treatment of biopsy proven intracranial germinomas: twenty-five years’ experience in a single center. Int J Radiat Oncol Biol Phys 58:1165–1170 12. Friedman NB (1947) Germinoma of the pineal. Its identity with germinoma (“seminoma”) of the testis. Cancer Res 7:363–368 13. Sung DMD, Harisiadis L, Chang Ch MD (1978) Midline pineal tumors and suprasellar germinomas: highly curable by irradiation. Radiology 128:745–751 14. Smirniotopoulos JG, Rushing EJ, Mena H (1992) Pineal region masses: differential diagnosis. Radiographics 12:577–596 15. Prahlow JA, Challa VR (1996) Neoplasms of the pineal region. South Med J 89:1081–1087
213 16. Zee CS, Segal H, Apuzzo M et al (1991) MR imaging of pineal region neoplasms. J Comput Assist Tomogr 15:56–63 17. Plassche McCormack TJ, Jr WM, Lin SR (1978) Ruptured teratoid tumor in the pineal region. J Comput Assist Tomogr 2:499–501 18. Chang T, Teng MMH, Guo WY, Sheng WC (1989) CT of pineal tumors and intracranial germ-cell tumors. AJNR 10:1039–1044 19. Arita NMD, Bitoh Sh MD (1980) Primary pineal endodermal sinus tumor with elevated serum and CSF alphafetoprotein levels. J Neurosurg 53:244–248 20. Hoffman HJ, Otsubo H, Hendrick EB et al (1991) Intracranial germ cell tumors in children. J Neurosurg 74:545–551 21. Raaijmakers C, Wilms G, Demaerel P, Bart AL (1992) Pineal teratocarcinoma with drop metastases: MR features. Neuroradiol 34:227–229 22. Herrick MK (1984) Pathology of pineal tumors. In: Neuwalt EA (ed) Diagnosis and treatment of pineal region tumors. Williams & Wilkins, Baltimore, pp 31–60 23. Chang CGS, Kageyama N (1981) Pineal tumors: clinical diagnosis, with special emphasis on the significance of pineal calcification. Neurosurgery 8:656–668 24. Schild RM, Scheithauer BW, Schomberg PJ et al (1993) Pineal parenchymal tumors. Cancer 72:870–880 25. Kumar P, Tatke M, Sharma A, Singh D (2006) Histological analysis of lesions of the pineal region: a retrospective study of 12 years. Pathol Res Pract 202:85–92 26. Nakagawa H, Iwasaki S, Kichikawa K et al (1990) MR imaging of pineocytoma: report of two cases. AJNR 11:195–198 27. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK (2007) WHO classification of tumours of the central nervous system. IARC, Lyon, pp 122–127 28. Ganti SR, Hilal SK (1986) CT of pineal region tumors. AJR 146:451–458 29. Silvera VM (2008) Pineal region tumors. In: Newton HB, Jolesz FA (eds) Handbook of neuro-oncology neuroimaging. Academic, USA, pp 449–459 30. Pustaszeri M, Pica A, Janzer R (2006) Pineal parenchymal tumors of intermediate differentiation in adults: case report and literature review. Neuropathology 26:153–157 31. Sasaki A, Horiguchi K, Nakazato Y (2006) Pineal parenchymal tumor of intermediate differentiation with cytologic pleomorphism. Neuropathology 26:212–217 32. Provenzale JM, Weber AL, Kinworth JK et al (1995) Radiologic-pathologic correlation: Bilateral retinoblastoma with coexistent pineoblastoma (trilateral retinoblastoma). AJNR 16:157–165
8
Embryonal Tumors Guido Wilms, Antonios Drevelegas, Philippe Demaerel, and Raf Sciot
Contents 8.1 Introduction.................................................................. 215 8.2 Medulloblastoma.......................................................... 215 8.3 Cerebral Neuroblastoma............................................. 221 8.4 Atypical Teratoid/Rhabdoid Tumors......................... 221 References............................................................................ 227
G. Wilms () Afdeling Radiologie, UZ Gasthuisberg, Herestraat 49, Leuven, Belgium e-mail:
[email protected] A. Drevelegas Department of Radiology, AHEPA university Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece e-mail:
[email protected] P. Demaerel Department of Radiology, University Hospital Gasthuisberg, Herestraat 49, Leuven, Belgium R. Sciot Department of Pathology, University Hospital, Catholic University of Leuven, Belgium
8.1 Introduction The concept of embryonal or primitive neuroectodermal tumors (PNETs) is based on the assumption that several tumors of the CNS show a common origin of multipotential neuroepithelial cells, which may undergo malignant transformation. This leads to tumors at various levels of the CNS, with identical histology, morphology, and biological behavior [1]. The most common neuroectodermal tumor is the medulloblastoma, making up 85% of this group of tumors [2]. Supratentorial PNETs are highly malignant tumors that have a much worse prognosis than medulloblastomas and include tumors as pineoblastoma, ependymoblastoma, neuroblastoma, ganglioneuroblastoma, and medulloepithelioma. Cerebellar PNETs can be associated with other tumors, as well as several syndromes such as basal cell-nevus syndrome, Gorlin syndrome, Turcot syndrome, and ataxia-teleangiectasia [3].
8.2 Medulloblastoma Medulloblastoma is a malignant primitive ectodermal tumor involving the cerebellum. It is the most common posterior fossa tumor in children accounting for one third of childhood pediatric central nervous system (CNS) tumors and about 7% of all brain tumors [4]. In adults medulloblastoma accounts for only about 1% of all brain tumors [5]. About three quarters of medulloblastomas occurs in children 5–15 years old, with a mean age at diagnosis of 7,3 years. Boys are more commonly affected (61.5%). They can also be encountered in young adults at 35 years or rarely older.
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_8, © Springer-Verlag Berlin Heidelberg 2011
215
216
Medulloblastomas are usually midline tumors and the most common site (80%) is the region of the vermis and inferior medullary velum. A less frequent location is the lateral cerebellar hemisphere, encountered in older children. A desmoplastic variant of medulloblastoma exists and is characterized by the presence of a sizable component of collageneous connective tissue within a neoplasm [6]. The desmoplastic variant is usually located in the cerebral hemispheres and is typically seen in adolescents or young adults. Approximately 80% of medulloblastomas present with hydrocephalus and the most common symptoms include headache, nausea, vomiting, and ataxia. These tumors frequently show dissemination along the cerebro spinal fluid CSF pathways and occasionally metastasize to bone and other sites outside the CNS [7]. Disseminated medulloblastoma is predominately located in the leptomeninges, while the spinal cord and the vertebral marrow are infrequently involved. Dissemination along the CSF pathways is associated with poor prognosis. The median time from initial detection to recurrence is 10, 5 months and 76% relapses occurred during the first 2 years [8]. After the surgical resection of the tumor both chemotherapy and radiation therapy are important in the treatment of medulloblastoma. The 5-year survival rate is 50–65%. The degree of surgical resection did not have a major effect on longterm survival; long-term survival was possible even in patients who had received only a biopsy [9]. Pathology: On gross pathology, the aspect of medulloblastomas is highly variable. The masses can be dense and well circumscribed, but other lesions are ill defined and softer; hemorrhage can be present [4] (Fig. 8.1a). At histopathological examination, medulloblastomas are highly cellular tumors composed of small neoplastic cells. Although these cells are often devoid of any architecture, inter and intratumoral heterogeneity is not uncommon. Therefore, tumor subtyping is imperative. The World Health Organization classifies medulloblastoma a grade IV lesion. Four subtypes are recognized: classic, desmoplastic, extensive nodular with advanced neuronal differentiation, and large cell medulloblastoma. In classic medulloblastomas the densely in diffuse sheets packed cells have hyperchromatic, rounded, or carrot-shaped nuclei and scanty, ill-defined cytoplasm. In less than 40%, neuroblastic (Homer-Wright) rosettes, a typical feature of these tumors, are recognized. (Fig. 8.1b). Nuclear pleomorphism, giant, and multinucleated cells
G. Wilms et al.
a
b
Fig. 8.1 Pathology of medulloblastoma. (a) Macroscopic autoptic image: notice the solid nodular mass originating from the vermis with anterior displacement and obstruction of the fourth ventricle. (b) Microscopic image: (hematoxylin-eosin, original magnification ×100) small anaplastic cells in a diffuse pattern. Inset: Homer-Wright rosettes (Hematoxylin-Eosin, original magnification ×400)
may be found. Ganglion cells are rarely recognized. In contrast to the rarity of vascular proliferation, calcifications and hemorrhages, an increased mitotic index is found in the majority of the cases. Apoptosis or geographic zones of necrosis may also be seen. Reticulin fibers are usually absent [10, 11]. In desmoplastic medulloblastomas (or medulloblastomas with a nodular pattern), more often found in adolescents or young adults, a nodular architecture is easily seen. This nodularity is characterized by centrally found, reticulin free, pale areas (“pale islands”) of primitive neuroepithelial cells, surrounded by mantles of dark cells intermingled in a network of reticulin fibers [12, 13]. Probably this variant arises from cells in the superficial part of the molecular layer. Therefore, an invasion of the leptomeningeal space is possible [14]. The third type, extensive nodular medulloblastomas with advanced neuronal differentiation is also known as “cerebellar neuroblastoma.” This subgroup of medullo
217
8 Embryonal Tumors
blastomas is found predominantly in children under 3 years of age, shows a “grape-like” nodularity and a lobular pattern of uniform cells with neuroblastic features streaming in a fine fibrillary background [15]. The large cell medulloblastoma is the least common and most malignant type, and presents with large round nuclei with prominent nucleoli and abundant cytoplasm. Transitional and mixed neuropathological forms are described. Synaptophysin, neurofilament proteins, and class III b-tubulin positivity characterizes tumor cells in medulloblastomas. Vimentin immunoreactivity may also be found [16, 17]. Several chromosomal abnormalities have been des cribed in medulloblastoma. The most common is a loss of material from chromosomal arm17p that seems to harbor a suppressor gene. Imaging: Medulloblastoma is as a well-circumscribed mass, centered near the midline filling the fourth ventricle. Midline extension through the foramen of Magendie into the cisterna magna may occur, while lateral extension is uncommon [18].
a
On CT medulloblastoma appears as a hyperdense or isodense mass [19]. The increased density is due to the hypercellularity of the tumor and the high nuclear-to-cytoplasm ratio of the tumor cells [18]. After the administration of contrast medium medulloblastomas show moderate to strong homogeneous enhancement [20] (Fig. 8.2). Bourgouin et al. [20] reported that in adult patients medulloblastomas showed a slight to moderate enhancement after injection of contrast medium, while in children the tumors enhanced markedly and homogeneously. Cystic and necrotic degeneration is most commonly encountered in adult medulloblastomas. Calcification is seen in 20% of cases [4, 21]. On MRI medulloblastoma is hypo- to isointense on T1WI and iso- to hyperintense on T2WI [22]. Cysts, hemorrhage, necrosis, and clump-like calcification are responsible for the signal heterogeneity on T2WI. Post contrast MRI shows intermediate to strong enhancement. (Fig. 8.3). In adults, cerebellar medulloblastoma may appear as well-demarcated hemispheric mass with mild to moderate enhancement. Medulloblastoma in adults
b
Fig. 8.2 Medulloblastoma. (a) Axial CT shows a round hyperdense midline posterior fossa mass. (b) Postcontrast CT shows moderate homogeneous enhancement of the mass. Note also the crescent sign of the fourth ventricle (arrow)
218
a
G. Wilms et al.
b
c
Fig. 8.3 MRI in medulloblastoma. (a) T1WI shows a low signal midline mass filling the fourth ventricle. (b) The mass shows high signal on PD and T2W images. (c) Contrast-enhanced MRI shows intense inhomogeneous enhancement of the mass
219
8 Embryonal Tumors
abuts the tentorium and may mimic a meningioma [5] (Fig. 8.4). Some authors have reported that the combination of a mass with a high density on CT and low signal on T1WI is highly suggestive of medulloblastoma [23]. Diffusion-weighted MRI in medulloblastoma shows a
a
marked increase in signal intensity reflecting the dense nature of the tumor, which restricts extracellular diffusion of water protons, and the high nuclear-to-cytoplasmic ratio of these neoplastic cells which limits intracellular motion [24] (Fig. 8.5). Leptomeningeal metastases either
b
c
Fig. 8.4 Medulloblastoma in a 39-year-old patient. (a) Axial T1WI shows a hypointense posterior fossa mass involving the left cerebellar hemisphere. (b) Coronal T2WI shows a slightly hyperintense cerebellar mass with high signal cystic compo-
nents. (c) Postcontrast coronal MRI shows intense enhancement of the solid component of the mass that may be mistaken for meningioma due to its close relation to the tentorium
220
to the brain or to the spine are better depicted on postcontrast T1WI [25] (Fig. 8.6). MR imaging had greater diagnostic accuracy than did CSF cytologic analysis in the early detection of disseminated tumor [7].
G. Wilms et al.
The differential diagnosis of medulloblastomas includes ependymoma, astrocytoma choroid plexus papilloma, metastasis, and rhabdoid tumors of the cerebellum.
a
b
c
d
Fig. 8.5 Medulloblastoma. (a) Axial T1WI shows a hypointense right cerebellar mass compressing the fourth ventricle. (b) Enhanced MRI shows mild heterogeneous enhancement of the mass. (c) On T2WI the mass shows iso- and hyperintense
areas. (d) Coronal diffusion-weighted image shows a markedly hyperintense mass. (e) On apparent diffusion coefficient image the mass shows low signal intensity. The perifocal high signal intensity is due to the surrounding edema
221
8 Embryonal Tumors
e
Fig. 8.5 (continued)
8.3 Cerebral Neuroblastoma Cerebral neuroblastoma is considered as a subset of PNETs although this term is controversial. Supraten torial PNETs are highly malignant tumors, and besides neuroblastomas they include also pineoblastomas and ependymoblastomas. They are accounting for less than 1% of all primary CNS tumors [26]. Cerebral neuroblastomas most commonly occur in the first decade of life. Adult’s neuroblastomas have also been reported [27]. The typical cerebral neuroblastoma is an inhomogeneous intraparenchymal supratentorial mass with little associated edema. Periventricular or intraventricular neuroblastomas may also be seen both in children and in adults [8]. Seizures and increased intracranial pressure are the most common clinical signs. Cerebral neuroblastomas are highly malignant tumors and the median survival is less than 24 months. Pathology: Neuroblastomas consist of small undifferentiated cells with hyperchromatic nuclei, increased mitotic activity, and scanty cytoplasm. These cells tend to cluster around a fibrinoid matrix, forming the Homer-Wright rosettes [21]. Imaging: CT shows a large heterogeneous intraparenchymal mass. Calcification, spontaneous hemorrhage, and cyst formation are common findings. After the administration of contrast material the mass shows inhomogeneous enhancement (Fig. 8.7). On MRI neuroblastomas show heterogeneous intensities on both T1- and T2-weighted images. Contrast enhanced MRI shows inhomogeneous enhancement and is also essential for revealing subarachnoid tumor seeding (Fig. 8.8). Differential diagnosis includes oligodendroglioma, astrocytoma, meningioma, and ependymoma.
8.4 Atypical Teratoid/Rhabdoid Tumors
Fig. 8.6 Leptomeningeal spread of medulloblastoma. Post contrast T1WI shows diffuse intradural enhancement around the conus medullaris and cauda equina arrows. In addition, small nodular enhancement of nerve roots is also seen (arrow heads)
Atypical teratoid/rhabdoid tumors (AT/RTs) are rare intracranial malignant neoplasms occurring in young children [28–33]. Originally described in the kidney as a highly malignant subgroup of renal tumors in children, the first description of the location of this tumor in the CNS dates back to 1987. Since a rhabdoid component is present the term “atypical AT/RT” was used to underline the unique combination of neuroepithelial, peripheral epithelial, and mesenchymal elements. AT/ RTs account for 1, 3% of pediatric CNS tumors and
222
a
G. Wilms et al.
b
Fig. 8.7 (a, b) Cerebral neuroblastoma. Postcontrast CT images at different levels show a large, heterogeneous mass with marked enhancement. Note the sharp demarcation and the linear calcification of the mass (arrows)
represent 6, 7% of the CNS tumors in children under the age of 2. Although highly variable in most reports, the number of infra- and supratentorial AT/RTs is grossly equal. Rarely the location is multiple at initial diagnosis. The tumor is mostly intra-axial with only one case of extra-axial location reported. The tumors are mostly large in size, especially in the supratentorial space. Dissemination is present in about half of the patients at the time of initial diagnosis (Parmar) and typically occurs along the subarachnoid space of the brain and spine. It is more frequent in children of the youngest age group [33]. The mean survival rates are in the range of 6–15 months, with a 5-year survival of 28%. Pathology: At histological examination, the lesion often contains areas indistinguishable from classic PNET-MB; nevertheless, the lesion shows clearly distinct histopathological and immuno-histochemical patterns, allowing a differential diagnosis. Besides the rhabdoid cells and PNET cells, the tumor contains varying percentages of malignant mesenchymal spindele-shaped cells and cells with epithelial differentiation (Fig. 8.9). Immunohistochemichal features show antibodies against epithelial membrane antigen, vimentin, actin, and cytokeratin in a high percentage of cases. Moreover, a variety of genetic abnormalities involving chromosome 22 have been described, typical for AT/RTs [34].
This differentiation between PNET-MB and AT/RT is necessarily given the extremely poor prognosis requiring aggressive therapy combining neurosurgical resection, radio-and high-dose chemotherapy, different from the treatment of PNET-MB. Imaging: On imaging studies too there are also overlapping features with PNET-MBs [29–33]. On CT the tumor is typically spontaneously hyperdense compared to gray matter. Calcifications and small peripheral cysts are possible. Enhancement is strong in the majority of cases. On MR the signal intensity is very heterogeneous. This is not only due to the presence of necrosis, hemorrhage, and calcifications, but also reflects the heterogeneous cellular populations in the tumor. On short TR/ short TE images (“T1-weighted images”) the tumor is mostly iso-intense to gray matter with possible areas of hyperintensity due to calcification or hemorrhage. Flowvoids due to hypervascularization are possible. Necrotic areas will appear more hypointense. On long TR/long TE images (“T2-weighted images”) the lesion again is very heterogeneous, with a low to intermediate signal of the solid portions. High signal points to edema, hemorrhage, necrosis or cysts. Low T2 signal can be due to calcification, hemorrhage, or flowvoids in hypervascular lesions. On fluid-attenuation inversion-recovery
223
8 Embryonal Tumors
a
b
c
d
Fig. 8.8 Cerebral neuroblastoma in an adult. (a) Axial T2WI shows a markedly inhomogeneous tumor in the left pre-Rolandic region of the brain. The lesion is ovoid, and composed of cystic components alternating with more hypointense solid components. Minimal midline shift is seen. (b) Axial T1WI shows the
multilocular composition.of the tumor. (c) Axial and (d) Coronal enhanced MRI shows inhomogeneous enhancement, with a more intensely enhancing nodule laterally, diffuse moderate enhancement of the noncystic components and peripheral septalike enhancement among the cystic parts of the tumor
224
(FLAIR) images the lesion is iso- to slightly hyperintense to gray matter. On contrast-enhanced MR all lesions show at least some enhancement within the majority of cases a strong enhancement. This enhancement is inhomogeneous in the majority of cases (Figs. 8.10 and 8.11). a
G. Wilms et al.
On diffusion-weighted sequences the tumor shows restricted diffusion with high signal on DWI and low signal on ADC, similar to those reported for PNET-MB and related to the hypercellularity of the tumor [35]. b
Fig. 8.9 (a, b) Histopathology of AT/RT
a
Fig. 8.10 Atypical teratoid rhabdoid tumor of the posterior fossa. (a) Axial T2WI shows a large heterogeneous mass laterally in the left cerebellar hemisphere. Important mass effect on the fourth ventricle. (b) Axial FLAIR shows the lesion to be hypointense
b
with some surrounding edema. (c) Axial T1WI reveals the tumor as hypointense. (d) Axial and (e) coronal-enhanced MRI shows intense but inhomogeneous enhancement. Notice the herniation of the tumor through the foramen magnum
225
8 Embryonal Tumors
c
d
e
Fig. 8.10 (continued)
As usual in the pediatric age group, a standard MR examination should include images of the spinal region in the look for spinal dissemination.
The differential diagnosis of AT/RTs includes medulloblastoma, ependymoma, astrocytoma and in older patients choroid plexus papilloma and metastasis.
226
G. Wilms et al.
a
b
c
d
Fig. 8.11 Atypical teratoid rhabdoid tumor of the supratentorial region. (a) Axial T2WI shows a large inhomogeneous lesion at the periphery of the left frontal lobe with predominant high signal but manifest intratumoral hypointensities owing to hemor-
rhage and flow-voids. (b) Axial T1WI shows an isointense tumor with multiple hyperintensities due to hemorrhage (c) Axial and (d) sagittal-enhanced MRI shows intense but inhomogeneous enhancement of the lesion
8 Embryonal Tumors
References 1. Nishio S, Morioka T, Fukui M (1998) Primitive neuroectodermal tumors. Crit Rev Neurosurg 8:261–268 2. Becker LE, Halliday WC (1987) Central nervous system tumors of the childhood. Perspect Pediatr Pathol 10:86 3. Atlas SW (1991) Intraaxial brain tumors. In Atlas SW Mag netic resonance imaging of the brain and spine, 1st edn. Raven, New York, pp 223–326 4. Koeller K, Rushing (2003) From the archives of the AFIP: medulloblastoma: a comprehensive review with radiologicpathologic correlation. Radiographics 23:1613–1637 5. Koci TM, Chiang F, Mehringer CM et al (1993) Adults cerebellar medulloblastoma. Imaging features with emphasis on MR findings. AJNR 14:929–939 6. Levy RA, Blaivas M, Muraszkok et al (1997) Desmoplastic medulloblastoma: MR findings. AJNR 18:1364–1366 7. Meyers SP, Widenhain SL, Ja-Kwei C et al (2000) Postoperative evaluation for disseminated medulloblastoma involving the spine: contrast-enhanced MR findings, CSF cytologic analysis, timing of disease occurrence, and patient outcomes. AJNR 21:1757–1765 8. Bouffet E, Doz F, Demaille MC et al (1998) Improving survival in recurrent medulloblastoma: earlier detection, better treatment or still an impasse? Br J Cancer 77:1321–1326 9. Hubbard JL, Scheithauer BW, Kispert DB et al (1989) Adult cerebellar medulloblastomas: the pathological, radiolographic and clinical disease spectrum. J Neurosurg 70(4):536–544 10. Burger PC, Grahmann FC, Bliestle A et al (1987) Differentiation in the meduloblastoma. A histological and immunohistological study. Acta Neuropathol 73:115–123 11. Russel DS, Rubinstein LJ (1989) Pathology of tumours of the nervous system, 5th edn. Williams & Wilkins, Baltimore 12. Katsetos CD, Herman MM, Frankfurter A et al (1989) Cerebelar desmoplastic medulloblastomas. A further immunohistochemical characterization of the reticulin- free pale islands. Arch Pathol Lab Med 113:1019–1029 13. Burger PC, Scheithauer BW, Vogel PS (1991) Surgical pathology of the nervous system and its coverings, 3rd edn. Churchill-Livingstone, New York 14. McLendon RE, Enterline DS, Tien RD et al. (1998) In: Russel DS, Rubinstein LJ (eds) Pathology of tumors of the nervous system, 6th edn. Bigner DB, McLendon RE, Bruner JM. Arnold: London; pp 307–571 15. Schofield DE, Yunis EJ, Geyer JR et al (1992) DNA content and other prognostic features in childhood medulloblastoma. Proposal of a scoring system. Cancer 69:1307–1314 16. Burger PC, Scheithauer BW (1994) Tumors of the central nervous system. Armed Forces Institute of Pathology, Washington 17. Katsetos CD, Krishna L, Frankfurter A et al (1995) A cytomorphological scheme of differentiating neuronal phenotypes in cerebellar medulloblastomas based on immunolocalization of class III b-tubulin isotype and proliferating cell nuclear antigen (PCNA/cyclin). Clin Neuropathol 14:72–80 18. Smirniotopoulos JG (1999) The new classification of brain tumors. Neuroim Clin North Am 9(4):595–613
227 19. Blaser SI, Harwood-Nash DC (1996) Neuroradiology of pediatric posterior fossa medulloblastoma. J Neurooncol 29(1):23–34 20. Bourgouin PM, Tampieri D, Grahovac SZ et al (1992) CT and MR findings in adults with cerebellar medulloblastoma. AJR 159:609–612 21. Robles HA, Smirniotopoulos JG, Figueroa RE (1992) Understanding the radiology of intracranial primitive neuroectodermal tumors from a pathological perspective: a review. Semin Ultarsound CT MR 13:170–181 22. Luh GY, Bird CR (1999) Imaging of brain tumors in the pediatric population. Neuroim Clin North Am 9(4):691–716 23. Tortori-Donati P, Fondelli MP, Rossi A et al (1996) Medulloblastoma in children: CT and MRI findings. Neuroradiology 38(4):352–359 24. Kotsenas AL, Roth TC, Mannes WK et al (1999) Abnormal diffusion-weighted MRI in medulloblastoma: does it reflectsmall cell histology. Pediatr Radiol 29:524–526 25. Kochi M, Mihara Y, Takada A et al (1991) MRI of subarachnoid dissemination of medulloblastoma. Neuroradiology 33(3):264–268 26. David R, Lamki N, Fan S et al (1989) The many faces of neuroblastoma. Radiographics 9:859–882 27. Davis PC, Wichman RD, Takei Y et al (1990) Primary cerebral neuroblastoma: CT and MR findings in 12 cases. AJR 154:831–836 28. Burger PC, Yu I-T, Friedman HS et al (1998) Atypical teratoid/rhabdoid tumor of the central nervous system: a highly malignant tumor of infancy and childhood frequently mistaken for medulloblastoma – a pediatric oncology study. Am J Surg Pathol 22:1083–1092 29. Arslanoglu A, Aygun N, Tekhtani D et al (2004) Imaging findings of CNS atypical teratoid/rhabdoid tumors. AJNR Am J Neuroradiol 25:476–480 30. Cheng YC, Lirng JF, Chang FC et al (2005) Neuroradiological findings in atypical teratoid/rhabdoid tumor of the central nervous system. Acta Radio 46:89–96 31. Meyers SP, Khademian ZP, Biegel JA, Chuang SH, Korones DN, Zimmerman RA (2006) Primary intracranial atypical teratoid/rhabdoid tumors of infancy and childhood: MRI features and patient outcomes. AJNR Am J Neuroradiol 27:962–971 32. Parmar H, Hawkins C, Bouffet E, Rutka J, Shroff M (2006) Imaging findings in primary intracranial atypical teratoid/ rhabdoid tumors. Pediatr Radiol 36(2):126–132 33. Warmuth-Metz M, Bison B, Dannemann-Stern E, Kortmann R, Rutkowski S, Pietsch T (2008) (2008) CT and MR imaging in atypical teratoid/rhabdoid tumors of the central nervous system. Neuroradiology 50(5):447–452 34. Bruch LA, Hill A, Cai DX et al (2001) A role for fluorescence in situ hybridization detection for chromosome 22q dosage in distinguishing atypical teratoid/rhabdoid tumors from medulloblastoma/central primitive neuroectodermal tumors. Hum Pathol 32:156–162 35. Rumboldt Z, Camacho DL, Lake D, Welsh CT, Castillo M (2006) Apparent diffusion coefficients for differentiation of cerebellar tumors in children. AJNR Am J Neuroradiol 27:1362–1369
9
Tumours of the Cranial Nerves Hervé Tanghe, Paul M. Parizel, and Antonios Drevelegas
Contents 9.1 Tumours Related to the Cranial Nerves............... 229 9.2 Tumours Related to the Olfactory Nerve.............. 230 9.2.1 Esthesioneuroblastoma (Olfactory Neuroblastoma)........................................................ 230 9.3 Tumours Related to the Optic Nerve.................... 232 9.3.1 Introduction.............................................................. 232 9.3.2 Optic Nerve Glioma.................................................. 232 9.3.3 Optic Nerve Meningioma......................................... 235 9.3.4 Optic Nerve Arachnoid Cyst.................................... 239 9.3.5 Melanocytic Neoplasms: Melanocytoma.................. 240 9.4
Tumours Related to the Other Cranial Nerves........................................................ 241 9.4.1 Schwannoma............................................................. 241 9.4.2 Neurofibromatosis.................................................... 251 References............................................................................ 254
H. Tanghe Section of Neuroradiology, Department of Radiology, University Hospitals Dijkzigt/Sophia/Daniel, Erasmus University Medical Centre, Dr. Molenwaterplein 40, 3015 GD, Rotterdam, The Netherlands P.M. Parizel Department of Radiology, Universitair Ziekenhuis Antwerpen, University of Antwerp, Edegem, Belgium A. Drevelegas (*) Department of Radiology, AHEPA university Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece e-mail:
[email protected]
9.1 Tumours Related to the Cranial Nerves The cranial nerves are traditionally grouped together with a total number of 12. This is not completely correct, because the olfactory system (cranial nerve I) and the optic nerve (cranial nerve II) should be considered as embryologic evaginations of fibre tracts from the telencephalon and diencephalon, and therefore are not cranial nerves in the true sense of the word. This embryological difference is anatomically reflected in a different covering of the nerve and pathologically in different disease processes and tumours, which do not affect the other cranial nerves. The cranial nerves III through XII have a special transition zone between their central covering of neuroglia and their peripheral covering of Schwann cells (see later in this chapter). An increasing number of intra-axial schwannomas (intracerebral or intramedullary), not related to a major cranial nerve, are reported, including a review of 18 cases reported by our own hospital [1]. They presumably arise either from small myelinated peripheral nerve fibres that accompany blood vessels into the parenchyma, or from Schwann cells near the dorsal root entry zone [2] (Table 9.1). In this chapter, we shall discuss: 1. Tumours related to the olfactory nerve • Esthesioneuroblastoma or olfactory neuroblastoma 2. Tumours related to the optic nerve: • Glioma • Meningioma • Arachnoid cyst • Optic nerve sheath dilatation
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_9, © Springer-Verlag Berlin Heidelberg 2011
229
230
H. Tanghe et al.
Table 9.1 The relationship of the cranial nerves with their anatomical covering and the possible tumours Cranial nerve Covering or origin Related tumours Olfactory nerve (cranial nerve I)
None
Esthesioneuroblastoma
Optic nerve (cranial nerve II)
Meningeal sheath
Meningioma Glioma Arachnoid cyst Optic nerve sheath dilatation Many others: Table 9.2
Cranial nerves III through XII
Central part: neuroglia Peripheral part: Schwann cell
Primary cranial nerve root entry zone glioma Schwannoma Neurofibroma Malignant schwannoma
Not related to cranial or peripheral nerves
Peripheral nerve fibres accompanying blood vessels into the parenchyma or Schwann cells near the dorsal root entry zone
Primary intra-axial cerebral or intramedullary schwannoma
3. Tumours related to the other cranial nerves: • Schwannoma • The primary intracerebral schwannoma
9.2 Tumours Related to the Olfactory Nerve 9.2.1 Esthesioneuroblastoma (Olfactory Neuroblastoma) 9.2.1.1 Aetiology and Definition The esthesioneuroblastoma is a tumour composed primarily of neuroblasts, which arises in the vault of the nose overlying the cribriform plate. It arises from the neuroepithelium in the olfactory rim of the nasal cavity [2].
9.2.1.2 Incidence Esthesioneuroblastoma can occur at any age, but has a bimodal peak incidence. The first peak occurs in the age group between 11 and 20 years, and represents 17% of all esthesioneuroblastomas; the second, larger peak occurs in the age group of 51–60 years, and represents 22% of all cases [3, 4]. Esthesioneuroblastoma is a rare tumour that accounts for about 3% of all intranasal tumours. According to a review of Broich et al in 1997, only 945 cases were reported in the literature since the original description in 1924 [5].
9.2.1.3 Location The starting point of the tumour is always unilateral, and most tumours stay unilateral. It is only in advanced cases that a bilateral extension pattern is found [3, 4]. The most common locations are the superior meatus, the roof of the nasal fossa, or high in the ethmoid sinus. More rare primary locations are: the nasopharynx, the maxillary sinus, the parasellar region. From the superior meatus or the ethmoid sinus, the tumour can extend to the lower parts of the nasal fossa, the orbit, the maxillary sinus, the anterior cranial fossa. The intracranial extension is usually extradural, but invasion of the brain parenchyma is possible.
9.2.1.4 Prognosis Esthesioneuroblastoma is a slow-growing malignant tumour that is locally invasive. Metastases are rare (less than 20%) and the most frequent sites are the cervical lymph nodes, followed by the lungs and the bones [6]. For the staging of the tumour, two systems are used: 1. The Kadish clinical classification is based on the extension of the tumour at the time of diagnosis. In stage A, the tumour is limited to the nasal fossa. In stage B, there is extension into one or more sinuses. Extension beyond the nasal fossa and the sinuses, i.e. to the orbit or intracranially is called stage C [7]. 2. The Hayms grading system is a histological classification in which grade I tumours have a good prognosis and grade IV tumours are fatal [8].
231
9 Tumours of the Cranial Nerves
Several therapeutic regimes are used: (1) surgery alone, (2) radiotherapy alone, (3) combined surgery and radiotherapy, and (4) adjuvant chemotherapy. The best results are obtained by the combination of surgery and radiotherapy with a 5-year survival rate ranging from 72.5 to 66.3% [5, 9, 10].
3. Unusual imaging features are: (a) Hyperostosis in the lamina cribrosa and the nasal septum [11]. (b) Intratumoral calcifications: small, puntiform calcifications are regularly found, but gross calcifications are rare [12]. (c) Intratumoral cysts or necrotic parts. (d) Brain parenchyma invasion except in the late stage.
9.2.1.5 Imaging Features 1. On computed tomography(CT), the tumour has usually an homogeneous density, equal or hyperdense to muscle. After intravenous contrast, there is an intense enhancement, usually homogeneous. With extension of the tumour there is a bone destruction of the walls of the nasal fossa, lamina cribrosa, orbit. This is better depicted by CT than by magnetic resonance imaging (MRI). CT is also superior for the detection of intratumoral calcifications (Fig. 9.1a). 2. On MR imaging, the tumour is isointense relative to muscle on signal intensity (SI) in the T1W images, and hyperintense on T2W images. After intravenous injection of Gd-chelates, marked enhancement is observed. The intracranial extra axial extension is better seen with MRI (Fig. 9.1b).
a
Fig. 9.1 (a) Coronal CT; (b) coronal SE T1W with Gd. A 70- yearold male patient with an advanced stage Esthesioneuroblastoma, located bilateral in the nasal fossa with destruction of the medial
9.2.1.6 Pathology The histological presentation of esthesioneuroblastoma is highly variable. To some extent this can be explained by the fact that the olfactory epithelium consists of three different cell types: (1) basal cells; (2) supporting cells; and (3) the sensory neurons [6]. Rosettes of two varieties may be seen. Homer Wright rosettes are present in a minority of the lesions and the Flexner type is rare. The tumour cells are undifferentiated, small and with round nuclei. The nuclei are uniform. The pathological diagnosis of esthesioneuroblastoma can be difficult. Several tumours that can occur in this region are composed of small undifferentiated cells
b
wall of both orbits, and of the lamina cribrosa. The tumour extends into the extraconal orbital space and intracranially both extra- and intra-axial
232
and can look alike under the microscope. The principal pathological differential diagnosis is with the sinonasal undifferentiated carcinoma and with the neuroendocrine carcinoma. For the correct diagnosis, immunohistochemical examination (reactivity for S-100 protein and neuron-specific enolase) and electron microscopy are needed [2, 6, 8]. 9.2.1.7 Differential Diagnosis 1. Common tumours in the region of the lamina cribrosa (a) In children: (i) Rhabdomyosarcoma (ii) Lymphoma (iii) Metastasis of neuroblastoma (iv) Juvenile angiofibroma (v) Esthesioneuroblastoma (rare) (b) In adults: (i) Esthesioneuroblastoma (ii) Olfactory groove meningioma with eventually extracranial extension (iii) Metastasis (iv) Haemangiopericytoma (v) Sinonasal carcinoma (vi) Rare: plasmocytoma; melanoma; fibro sarcoma 2. Naso-ethmoidal soft tissue lesions with possible hyperostosis (osteoblastic reaction): (i) Meningioma with extracranial extension or primary extracranial (ii) Bone-forming sarcoma: osteosarcoma; chondrosarcoma (iii) Fibrous dysplasia (iv) Chronic sinusitis (v) Esthesioneuroblastoma
9.3 Tumours Related to the Optic Nerve 9.3.1 Introduction It should be remembered that the optic nerve is merely a long extension of brain tissue, and similar to the brain, it is surrounded by a subarachnoid space and its meningeal coverings. Therefore, the same lesions that involve the brain (such as glioma, demyelination), as well as
H. Tanghe et al. Table 9.2 Overview of optic nerve lesions with enlargement of the optic nerve/nerve sheath Common Rare tumours Miscellaneous tumours lesions Glioma
Medullloepithelioma
Arachnoid cyst
Malignant glioma
Ganglioglioma
Optic nerve sheath dilatation
Meningioma
Haemangioblastoma
Sarcoidosis
Metastasis
Cysticercosis
Lymphoproliferative disease Choristoma Melanocytoma
disease process extending into the subarachnoid space (like meningitis, sarcoidosis, metastasis, cysticercosis) or arising in the meninges (like meningioma) can occur in the optic nerve [13–20] (Table 9.2).
9.3.2 Optic Nerve Glioma 9.3.2.1 Incidence Optic nerve gliomas represent approximately 4% of all orbital tumours, 2% of all intracranial tumours and 4% of all gliomas. They are 3 or 4 times more frequent than optic nerve sheath meningioma. The peak incidence is from 2 to 8 years with 90% manifesting in the first two decades [20, 21]. Optic pathway glioma is related to neurofibromatosis type I (NF Type I). Between 10 and 40% of patients with NF type I develop an optic pathway glioma; conversely, 10–70% of the patients with an optic pathway glioma have NF type I. The growth rate of the tumour is variable in patients with or without NF type I. Bilateral involvement of the optic nerves by a glioma is more common in patients with NF Type I [21].
9.3.2.2 Location Optic pathway gliomas can occur anywhere along the optic pathway, from just behind the globe to the occipital cortex, but 50–85% involve the optic chiasm and or
233
9 Tumours of the Cranial Nerves
a
b
c
d
Fig. 9.2 Glioma of the right optic nerve in a 15-month-old boy. MR images through the orbits in planes. (a) Axial SE T1WI; (b) axial and sagittal SE T2 WI; (c) sagittal SE T1WI through the right orbit; (d) Gd-enhanced sagittal SE T1WI (same slice position as c). There is fusiform thickening of the right optic nerve.
The nerve is tortuous, with inferior buckling, best seen on the sagittal images. The subarachnoid spaces of the optic nerve sheath are markedly widened; this is presumably due to because of trapping or obstruction of the outflow of the fluid. After Gd-injection, the tumour enhances homogeneously and intensely
the hypothalamus. The optic chiasm gliomas/hypothalamic gliomas (OCHG) are often considered a singledisease entity. The optic nerve glioma can be unilateral or bilateral, and may involve only the orbital part of the nerve (Fig. 9.2) , or the orbital, intracanalicular and intracranial part and even extend to the optic chiasm and beyond [16, 20, 21] (Fig. 9.3). A normal imaging study in a patient with NF type I does not exclude the future development of an optic pathway glioma.
depends on the location. The OCHGs are more aggressive in children below 5 years. The prognosis of optic pathway glioma is better in patients with NF type I [21, 22]. On pathologic examination, the majority of these tumours are pilocytic astrocytomas (WHO grade I) [22]. Cystic degeneration is common. Necrosis and intratumoral haemorrhage are rare. Calcifications do not occur in non-radiated optic nerve gliomas (differential diagnosis with optic nerve sheath meningioma). By their growth, they cause a fusiform enlargement of the optic nerve, within an intact dura. 2. Malignant optic nerve glioma (MONG) MONG is a rare optic nerve tumour. It is a distinct entity from the benign pilocytic astrocytoma optic nerve tumour in childhood. It is more prevalent in males and the peak incidence is in the fifth decade.
9.3.2.3 Prognosis and Pathology 1. Optic nerve glioma Most optic pathway gliomas are benign, slowgrowing tumours. The tumour-related mortality
234
H. Tanghe et al.
a
c
Fig. 9.3 Bilateral visual pathway glioma, in a 10-year-old girl. (a) Axial SE T1W; (b) axial T1W; (c) axial T1W with Gd; (d) axial TSE T2W. The tumor is located in the intracanalicular
b
d
and intracranial parts of the optic nerve, in the chiasm and the tractus opticus. The tumour consists of non enhancing and enhancing solid parts and of cystic parts
235
9 Tumours of the Cranial Nerves
Histological examination reveals anaplastic astrocytoma or glioblastoma multiforme. Mortality is almost 100% with a mean survival of 9 months [20]. 9.3.2.4 Imaging Features On CT, an intraorbital optic nerve glioma causes a fusiform enlargement of the optic nerve. The tumour growth occurs in all directions, and this causes thickening and lengthening of the optic nerve. Within the confined space of the orbit, kinking and tortuosity of the optic nerve are often observed (Fig. 9.2). Moreover, the intraorbital mass effect causes exophthalmia. This growth pattern is different from that observed in optic nerve sheath meningioma, which stretches the nerve, causing an exophthalmia without buckling. The tumour is isodense with the brain parenchyma and the enhancement is variable. MR is the preferred imaging method. The tumour is hypointense on T1W images (Fig. 9.4a), and hyperintense on T2W images (Figs. 9.4b, c). A heterogeneous SI appearance on the T2W is possible, because the hyperintense peripheral portion can represent arachnoid hyperplasia. This component does not enhance after gadolinium. The extension to the intracanalicular (Fig. 9.4c) and the intracranial part of the optic nerve (Fig. 9.4d), and to the optic chiasm or even further in the visual pathways is better depicted by MR. Like on CT, the enhancement of the tumour is variable. Optic nerve gliomas may be associated with dilatation of the cerebrospinal fluid within the subarachnoid space of the optic nerve sheath, because of trapping or obstruction of the outflow of the fluid (Fig. 9.2).
9.3.3 Optic Nerve Meningioma
portions of the optic nerve. Their histology is that of syncytial or transitional meningioma. We will focus our discussion on this entity. The secondary meningioma originates in the intracranial dura surrounding the orbit, usually at the sphenoid ridge or tuberculum sellae. The tumour can enter the orbit through the optic canal, the superior orbital fissure or may invade and destroy the orbital bony wall, eventually breaching the periorbita. Approximately two-thirds of meningiomas involving the orbit have their origin outside it [23]. The ectopic meningioma originates from ectopic meningeal tissue that is pinched off within the orbit during intrauterine development. Such tumours may also occur outside the muscle cone or even adjacent to the optic nerve, but in an extradural location [15, 16]. PONSM represents less than 1% of all meningiomas and 3–5% of the orbital tumours. They occur predominantly between 30 and 50 years of age, but may occur at any age. PONSM in childhood is more aggressive than in adults. Most PONSM are unilateral, but bilateral cases do occur. The most frequent symptoms consist of progressive visual loss, exophthalmia, disc oedema or pallor [20].
9.3.3.2 Prognosis Surgical treatment of the PONSM is difficult. If the optic nerve is completely surrounded, then removal of the tumour almost invariably results in blindness. With modern imaging techniques, it is possible to define the proximal extent of the tumour and whether or not there is intracanalicular or intracranial extension. If the meningioma remains confined to the orbit, it is reasonable to adopt a conservative approach. Once vision is lost, a more aggressive treatment can be undertaken. The role of radiation therapy is controversial [23].
9.3.3.1 Classification, Incidence and Location The meningiomas involving the orbit and the optic nerve can be classified as [16]: 1. Primary optic nerve sheath meningioma (PONSM) 2. Secondary meningioma 3. Ectopic orbital meningioma The primary optic nerve sheath meningioma (PONSM) arises from the cap cells of the arachnoid, surrounding the intraorbital or, less commonly, the intracanalicular
9.3.3.3 Imaging Features On CT, the PONSM is hyperdense and frequently contains calcifications (globular, linear, plaque-like) (Fig. 9.5). The meningioma grows along the nerve, with stretching of the nerve. PONSM may penetrate the dura with expansion into the adjacent orbital fat. Usually, the growth is symmetrically circumferential, thereby creating a tubular configuration of the enlarged optic nerve/nerve sheath
236
H. Tanghe et al.
a
b
c
d
Fig. 9.4 A biopsy proven optic nerve pilocytic astrocytoma. (a) axial SE T1W; (b) axial TSE T2W, (c) coronal TSE T2W; (d) coronal SE T1W with Gd. Girl, 2 years. The tumor is located in all three parts of the right nerve. Homogeneous SI and
enhancement. The optic canal is enlarged (Fig. 9.2c). The normal non-thickened optic nerve sheath is visible as a dark line in the TSE T2W (Fig. 9.2b). Note the buckling of the nerve (Fig. 9.1a)
237
9 Tumours of the Cranial Nerves
a
c
b
e
Fig. 9.5 Meningioma of the right optic nerve sheath in a 58-year-old woman. CT and MR images through the orbits. (a) Non-contrast axial CT scan; (b) non-contrast coronal CT scan; (c) axial SE T1WI with fat saturation; (d) Gd-enhanced axial SE T1WI with fat saturation. The axial and coronal CT images
show a partially calcified spindle-shaped mass surrounding the right optic nerve. The tumour is of intermediate signal intensity on SE T1 WI and is hypointense on SE T2 WI. After Gd-injection, the tumour enhances intensely and homogeneously. Note the normal appearance of the right optic nerve “inside” the tumour
complex (Fig. 9.5). Sometimes, the tumour grows eccentrically, causing an asymmetrical enlargement of the complex. The optic nerve is embedded within the tumour and remains visible. This causes the so called tram-track sign (Fig. 9.5). Contrast enhancement is strong. The PONSM may extend from the orbital part to the canalicular part of
the nerve and may grow further intracranially around the anterior clinoid process. Extension into the optic canal can cause enlargement and occasionally hyperostosis of the walls of the canal [20]. On MR, the PONSM has a low SI on T1W and T2W images. The calcifications are not as well visible
238
as on CT. The extension into the optic canal and intracranially is better seen with contrast-enhanced MR than with CT. Fat-suppression techniques are important to differentiate the enhancing tumour from the intraorbital fat (Fig. 9.5). The superior contrast resolution, together with the multiplanar capabilities of MR, makes this technique superior to CT for imaging of optic nerve tumours. Small intracanalicular optic nerve meningiomas are rare and best seen on MR (Fig. 9.6). a
H. Tanghe et al.
9.3.3.4 Differential Diagnosis The most important differential diagnosis in optic nerve tumours is the distinction between an optic nerve glioma and a PONSM (see Table 9.3). Optic nerve sarcoidosis can be difficult to differentiate from meningioma, but is rare and other abnormalities, like in the chest, point to the correct diagnosis. Choristoma, ganglioglioma, lymphoma and haemangioblastoma of the optic nerve have no specific radiological features. b
c
Fig. 9.6 Intracanalicular optic nerve sheath meningioma in a 13-year-old girl with a normal intraorbital part of the nerve. (a) Parasagittal SE T1W after Gd.; (b) axial SE T1W after Gd;
(c) axial SE T1W after Gd. Strong enhancement (arrow) with dural tail sign (arrowhead) along the anterior clinoid process and tuberculum sellae. (Proved by operation and pathology)
239
9 Tumours of the Cranial Nerves Table 9.3 Differential diagnosis between optic nerve glioma and optic nerve meningioma Glioma Meningioma Kinking
Present
Absent
Enhancement
Variable
Strong
Calcifications
Absent
Possible
Tram-track sign
Absent
Possible
Enlargement of the perioptic fluid space
Possible
Absent
Intracranial extension
Follows the visual pathway
Follows the dura
a
9.3.4 Optic Nerve Arachnoid Cyst Cystic lesions of the optic nerve sheath within the orbit may occur in association with optic nerve pilocytic astrocytoma and haemangioblastoma. But a cyst of the nerve sheath may occur without any further demonstrable cause. CT and MR show a widening of the meninges by a cerebrospinal fluid-filled space (Fig. 9.7). They resemble the common intracranial arachnoid cyst. This entity should not be confused with a simple optic nerve sheath dilatation as can be caused by
b
c
Fig. 9.7 Arachnoid cyst in the optic nerve sheath in a 20-yearold women. (a) Axial SE T1W after Gd; (b) parasagittal SE T1W; (c) axial SE T2W. Intraorbital optic nerve sheath dilatation by a fluid filled space, with a SI equal to liquor on all pulse
sequences. No enhancement. The optic nerve is visible at the proximal site of the cyst, displaced medial and inferior, indicating a local asymmetric dilatation of the nerve sheath (differential diagnosis with a simple optic nerve sheath dilatation)
240
intracranial hypertension, optic nerve glioma (cfr. supra), hydrocephalus, optic nerve hydrops or idiopathic [24].
9.3.5 Melanocytic Neoplasms: Melanocytoma Melanocytes are normally present in the intracranial leptomeninges, usually in the posterior cerebral fossa, and can give rise to benign (meningeal melanocytoma)
a
H. Tanghe et al.
or malignant (malignant melanoma) tumours. The highest density of melanocytic neoplasms is found underneath the brainstem and along the upper cervical spinal cord. Orbital manifestation is rarely encountered. Histologically, most melanocytomas are spindle cell neoplasms, although epithelioid cytology may dominate in some cases. Tumour cells grow in sheets, nests, or fascicles (Fig. 9.8c). On CT, primary melanocytic lesions appear as wellcircumscribed, isoattenuating to hyperattenuating, extra-axial tumours with homogeneous enhancement. Even if there is no hyperostosis, the neoplasms can
b
c
Fig. 9.8 Axial T1-weighted MR image (a) and post contrast T1-weighted image (b) show a left intraconal mass with intermediate signal intensity witch enhances intensively and homo-
geneously. Photomicrograph of the mass. (c) The neoplasm contains sheets of cells, many of which contain dark brown melanin pigment (heimatoxylin-eosin ×400)
241
9 Tumours of the Cranial Nerves
mimic meningioma. MRI demonstrates variable SI on T1- and T2-weighted images, in proportion to the amount of melanin within the tumour . Therefore, melanocytic neoplasms are isointense or hyperintense to adjacent normal brain tissue on T1-weighted images and isointense or hypointense on T2-weighted images. The neoplasms enhance after intravenous administration of contrast material (Fig. 9.8b).
9.4 Tumours Related to the Other Cranial Nerves 9.4.1 Schwannoma 9.4.1.1 Definition and Incidence The schwannoma is a benign tumour, composed entirely of Schwann cells. The neurofibroma is a well-differentiated nerve sheath tumour composed predominantly of Schwann cells and, to a lesser extent, fibroblasts and perineural cells. Neurofibromas of cranial nerves are extremely rare, and they shall not be discussed in this chapter [2]. Schwannomas are the second most common extraaxial intracranial tumours, preceded only by meningiomas. They constitute 5–10% of all intracranial neoplasms. The peak incidence is between the third and sixth decades. 9.4.1.2 Pathology The peripheral and cranial nerves (except for the olfactory nerve and the optic nerve) are ensheathed by Schwann cells, starting from the transition zone down to their terminations [6]. The place of the transition zone, also called the glioneural junction, varies in its distance from the neuraxis, and the longest distance is encountered in the vestibulocochlear nerve, where the junction is located in the internal auditory canal (IAC) in 95% and in the cerebellopontine cistern in 5% of cases. Thus, the cranial nerves have a central part, which is not covered by Schwann cells, but by neuroglia [6]. Therefore, rare cases of primary cranial nerve root entry zone gliomas can occur [25]. Macroscopically, schwannomas are typically well circumscribed and more often globular than fusiform
Fig. 9.9 Biphasic pattern of Schwannoma: cellular small asterisks and hypocellular big asterisks areas
in configuration. In small lesions, the parent nerve can be detected within the tumour, but in larger tumours the relationship between the nerve and the tumour becomes obscured. The schwannoma is surrounded by a thick, completely collagenous capsule. Microscopically, two patterns can be distinguished, according to the morphology of the tumour cells and their spatial arrangements: the Antoni A and B types as described in 1920 by Antoni [6]. In the Antoni A type, tumour texture is compact and composed of interwoven bundles of long bipolar spindle cells. The type B Antoni architecture is often intermingled with type A, and has a loose texture and polymorphism (Fig. 9.9). Mucinous and cystic changes occur and when confluent, large cysts develop. Degenerative changes are frequent. Tumour-related cysts can occur in the centre or at the periphery of the tumour. Non-tumoural peripheral arachnoid cysts are frequently found in large vestibular schwannomas. Small tumours tend to be of high cellularity and of Antoni type A architecture, while larger tumours acquire an Antoni type B format, with degenerative and cystic changes. 9.4.1.3 Location Like their intraspinal counterparts, the intracranial schwannomas show a predilection for the sensory nerves, and most often involve the vestibular division of the eight nerve. The fifth cranial nerve is the second most common site of origin. Schwannomas of the cranial nerves three, four and six are rare. Schwannomas of the
242
Fig. 9.10 Axial 0.7 mm overlapping sections, 3D TSE T2W. Small medial vestibular schwannoma, not related to the IAC, in a 63-year-old man
jugular foramen usually originate from the ninth nerve. The presenting symptoms may be similar to those of a vestibular schwannoma, due to the growth in the posterior fossa. Signs of injury of the vagal or accessory nerve are frequently absent [26]. The facial nerve schwannoma can be located in the IAC or in the facial canal. The vestibular schwannoma usually starts at the transition zone located in 95% in the IAC. Further growth gives an enlargement of the IAC, and an extension in the cerebellopontine angle cistern, centred around the IAC. The so called “medial schwannoma” originates from a transition zone located in the cerebellopontine angle cistern (5% of the cases). This tumour has no part in the IAC (Fig. 9.10). Labyrinthine schwannomas are known from the pathologic literature since years, and are reported in the radiological literature since the early 1990s [27]. Unlike the counterpart in the IAC, the intralabyrinthine schwannoma usually originates from the cochlear part.
9.4.1.4 Imaging Features The intracanalicular vestibular schwannoma is characterised on MR imaging by the absence of the normal
H. Tanghe et al.
CSF signal in the IAC, and by the distorted anatomy of the neural bundle complex. Instead, an intracanalicular mass is seen, not always associated with an enlargement of the IAC, and with variable extension into the cerebellopontine angle cistern. The tumour enhances after gadolinium. With conventional CT these lesions cannot be reliably diagnosed (Fig. 9.11). The intralabyrinthine vestibular or cochlear schwannoma is characterised on T2-W MR images by the absence of the normal fluid signal in the involved part of the labyrinth. The diagnosis relies on the use of high resolution, ultra-thin section heavily T2-W images. These can be obtained by using ultra-thin 3D FT FSE T2W images or by using a 3D gradient echo sequence such as constructive interference in steady sequence (CISS). Moreover, the tumour enhances after intravenous gadolinium injection. Therefore, it is important to perform thin section T1-W images both before and after gadolinium injection. The T1W images without gadolinium can be used for the differential diagnosis with intralabyrinthine haemorrhage [27, 28] (Fig. 9.12). The large vestibular schwannoma is located in the IAC with extension in the cerebellopontine angle, centred around the IAC. The resulting appearance is that of a scoop of ice cream on the top of an ice cream cone. The SI depends on the histological composition of the tumour. Antoni type A lesions tend to present a homogeneous SI, are hypointense on T2-W images and enhance homogeneously. It is estimated that about 70% of schwannomas are homogeneously enhanced (Fig. 9.13). Antoni type B lesions contain more extracellular fluid and therefore, present a higher SI in the T2W image; they often contain intratumoral cysts or necrotic foci and enhance heterogeneously (Fig. 9.14). About 25% of schwannomas show heterogeneous enhancement. The majority of schwannomas show a biphasic pattern on T2-weighted images, which is described as “target sign”. They show central low SI secondary to fibrocollagen (Antoni type A) surrounded by peripheral high SI due to the presence of myxomatous matrix (Antoni type B) [29] (Fig. 9.15). A nearly complete cystic schwannoma is rare [30]. Associated arachnoidal cysts or loculations are encountered around the tumour in 5–10% of cases [30, 31]; they are especially observed in large schwannomas (Fig. 9.16). Intratumoral calcifications are rare and, when found, should favour the diagnosis of meningioma. The trigeminal nerve schwannoma can be located in every segment of the nerve: the cisternal segment in
243
9 Tumours of the Cranial Nerves
a
Fig. 9.11 Intracanalicular vestibular schwannoma in a 64-yearold man. (a) Axial 1 mm overlapping sections, 3D TSE T2W; (b) coronal 1.5 mm 3D FT GE T1W after Gd. This is a very
Fig. 9.12 Intralabyrinthine cochlear schwannoma in a 64-yearold man. Coronal 1.5 mm overlapping sections, 3D FT GE T1W after Gd. Pathological enhancement of the left cochlea (arrow)
the prepontine and cerebellopontine cistern, Meckel’s cave, cavernous sinus, the superior or inferior orbital fissure. From Meckel’s cave extension below the skull base through oval foramen is possible. Frequently, the
b
small nodular tumour, located inferoposterior in the IAC, clearly related to the inferior vestibular part of the nerve (arrows)
tumour presents a dumbbell configuration, with part of the lesion located in Meckel’s cave (supratentorial) and part of the tumour extending into the medial cerebellopontine angle cistern (infratentorial) (Fig. 9.17). The imaging characteristics are the same as for the vestibular schwannoma, but intratumoral cysts are more frequent [32]. The jugular foramen schwannomas are usually large at presentation with an extension in the posterior cranial fossa and below the skull base in the carotid loge (Fig. 9.18). The most common clinical presentation is hearing loss or symptoms relating to a posterior fossa mass. Glossopharyngeal deficit points to a large extension below the skull base. The tumour causes an enlargement of the jugular foramen, with rounded, sharp and sclerotic rims, without bony invasion or osteolysis on CT. In the differential diagnosis, large vestibular schwannomas should be considered: they can grow into the jugular foramen, but are associated with erosion and enlargement of the IAC. Conversely, a jugular foramen schwannoma may have a significant extension in the cerebellopontine angle but the IAC remains normal. Jugular foramen schwannomas traditionally present a low SI on T1W and a high signal on
244
a
H. Tanghe et al.
b
c
Fig. 9.13 (a) Axial T1WI; (b) Axial constructive interference in steady state sequence (CISS); (c) Coronal SE T1W after Gd.; Antoni A vestibular schwannoma appears hypointense on T1,
and isointense on T2W images. After the administration of contrast medium shows intense and homogeneous enhancement
245
9 Tumours of the Cranial Nerves
a
b
c
Fig. 9.14 (a) Axial T1WI, (b) axial T2WI, (c) axial PCT1WI. Antoni B vestibular schwannoma appears hypointense on T1- hyperintense on T2-weighted images and shows heterogeneous enhancement with intratumoral cysts on PCT1WI
246
a
H. Tanghe et al.
b
Fig. 9.15 (a) Axial T1WI and (b) axial T2WI. Vestibular schwannoma shows low signal intensity on T1-weighted images. On T2-weighted image the central area appears hypointense surrounded by peripheral high signal intensity
Fig. 9.16 Axial 1.5 mm overlapping sections, 3D FT GE T1W after Gd. Large vestibular schwannoma with an intratumoral cyst and a large secondary arachnoidal cyst (non-tumoural) in a woman 66-year-old
T2W images. The enhancement is strong or moderate. Cystic components are less frequent than in the trigeminal nerve schwannoma. In the differential diagnosis,
the following lesions must be considered: meningioma; glomus jugulotympanicum, metastasis, lymphoma, giant cell tumour. Meningioma causes an enlargement of the jugular foramen with possible hyperostosis and associated calcifications. On CT, the precontrast attenuation is higher than with schwannoma. On MR, the SI in the T2W image is lower. The contrast enhancement is similar to a schwannoma (Fig. 9.19). The glomus jugulare causes enlargement of the jugular foramen, with erosion of the borders and extension into the bone and into the hypotympanum. The tumour is highly vascular and on MR, intratumoral small flow voids are seen. Contrast enhancement is strong and homogeneous. Schwannomas are usually not very vascular. In glomus jugulare, tumours a preoperative angiogram with embolization can be necessary [26, 33]. The facial nerve schwannoma can arise in the IAC, at the level of the ganglion geniculi, in the middle ear or in the facial canal. Because the facial nerve is located in the anterosuperior quadrant of the IAC, bony erosion in this vicinity favours a facial rather than a vestibular schwannoma, as would extension of the mass in the labyrinthine segment of the facial canal. A facial nerve schwannoma causes enlargement
247
9 Tumours of the Cranial Nerves
a
b
c
Fig. 9.17 Trigeminal nerve schwannoma in a 48-year-old women. (a) Axial SE T1W after Gd; (b) coronal SE T1W after Gd; (c) axial TSE T2W. The tumour has a solid, enhancing part and a cystic part. There is a dumbbell shape with a component in Meckel’s cave and in the medial cerebellopontine angle cistern.
This last part encircles the cisternal segment of the optic nerve. Notice the cystic extension through the oval foramen, following the maxillary division of the nerve. The solid part of the tumour is hyperintens in the T2W image, favouring the diagnosis of a schwannoma, against a meningioma
248
H. Tanghe et al.
a
b
c
d
Fig. 9.18 Jugular fossa schwannoma in a 35-year-old man. MR images through the skull base. (a) Sagittal SE T1WI; (b) Gd-enhanced sagittal SE T1WI; (c) axial SE T2WI; (d) Gd-enhanced axial SE T1WI. A fusiform soft tissue mass extends from the right cerebello-pontine angle through the pars nervosa of the jugular foramen into the right parapharyngeal
space. A CT scan showed smooth erosion of the jugular foramen and the jugular tubercle (not shown). The tumour is of intermediate signal intensity on pre-contrast T1 WI, and is hyperintense on the T2WI. After Gd-administration, the lesion enhances intensely, though somewhat inhomogeneously
249
9 Tumours of the Cranial Nerves
a
b
c
Fig. 9.19 Calcified jugular foramen meningioma (arrows) in a 35-year-old man. (a) Axial CT; (b) axial SE T1W after Gd.; (c) coronal SE t1W after Gd. The jugular foramen is smoothly
enlarged without cortical erosion nor bony destruction The CT and even the MR shows clearly the intratumoral calcifications. The tumour extends below the skull base in the carotid space
250
H. Tanghe et al.
a
b
c
d
Fig. 9.20 Facial nerve schwannoma in a 39-year-old women. (a, b) Axial CT scans through the left petrous bone (1 mm slice thickness); (c, d) axial CT scans through the left petrous bone (1 mm slice thickness). A soft tissue mass involves the genicu-
late ganglion and the proximal tympanic segment of the left facial nerve. The tumour causes bone erosion with scalloping and enlargement of the geniculate fossa, and extends into the tympanic cavity, where it abutts the auditory ossicles (arrows)
251
9 Tumours of the Cranial Nerves
a
b
Fig. 9.21 Facial nerve schwannoma in a 40-year-old women. (a) Gd-enhanced axial SE T1WI; (b) Gd-enhanced coronal SE T1WI. An enhancing tumour is observed along the course of the left facial nerve. The tumour consists of a small nodular component in the left cerebellopontine angle (white arrow), a
wedge-shaped enhancing component in the left internal auditory canal (arrowhead), and a crescent-shaped component at the geniculate ganglion black arrow). The vertical portion of the facial nerve was not involved
of the facial canal with possible erosion and extension in the middle ear as a soft tissue mass (Figs. 9.20 and 9.21). The primary cerebral schwannoma arises in the brain parenchyma and is very rare. All regions of the brain may be affected. A correct radiological diagnosis is, of course, impossible. Our own recent case occurred in the occipital lobe in a boy of 13 years. The tumour had a homogeneous SI and enhancement and a lot of perifocal oedema [1] (Fig. 9.22).
9.4.2 Neurofibromatosis
9.4.1.5 Differential Diagnosis The differential diagnosis of the various types of schwannomas is described in Tables 9.4–9.8.
Most vestibular schwannomas are unilateral and sporadic. In 5% of cases, a vestibular schwannoma can occur as a manifestation of neurofibromatosis (NF) type I or type II. Although several variants of NF have been reported, to date the National Institute of Health (NIH) Consensus Development Conference has defined only distinct types, namely NF Type 1 (von Reckling hausen’s disease, sometimes called peripheral NF) and NF Type II (bilateral acoustic schwannomas or central NF) [34]. In NF type I (Von Recklinghausen’s disease), a vestibular schwannoma may occur as a rare event (less than 2% of patients with NF Type I); in these cases, the vestibular schwannoma is unilateral.
252
H. Tanghe et al.
a
b
Fig. 9.22 (a) Axial SE T1W after Gd; (b) Axial TSE T2W. Boy, 13 years. Primary cerebral schwannoma in the occipital lobe (proved by pathology). There is a lot of perifocal oedema. A correct radiological diagnosis is in this case impossible
Table 9.4 Differential diagnosis of intracanalicular schwannoma
Table 9.5 Differential diagnosis of a cerebellopontine angle tumour
Vestibular schwannoma
Vestibular schwannoma
80%
Facial schwannoma
Meningioma
10%
Meningioma
Others
10%
Metastasis
(Epidermoid)
(Lipoma)
Haemangioma
(Arachnoidal cyst)
(AVM)
Leptomeningeal disease
(Metastasis)
(Other schwannoma)
253
9 Tumours of the Cranial Nerves Table 9.6 Radiological differential diagnosis between schwannoma and meningioma in the cerebellopontine angle Schwannoma Meningioma Intracanalicular component
Almost always (95%)
Rare
Centred around the IAC
Yes
No, excentric
Calcifications
No
Possible
Necrotic/cystic parts
Frequent
Rare
Dural tail sign
Possible
Possible
Secondary arachnoid cyst
Possible in large tumours
No
Influence on the bone
Enlargement of the IAC
Hyperostosis, enostosis, invasion
Supratentorial extension
No
Possible
Contact with facies posterior
Sharp angle
Broad based contact
a
b Table 9.7 Possible causes of contrast enhancement in the labyrinth [5] Labyrinthine schwannoma Viral labyrinthitis Lues Sarcoidosis Other tumours in the cochlea: lymphoma Cochlear infarct Cogan’s syndrome
Table 9.8 Differential diagnosis of a tumour in the jugular foramen [10] Glomus jugulare Meningioma Schwannoma Metastasis Giant cell tumour Pseudomass of large or high jugular bulb Lymphoma
Fig. 9.23 Neurofibromatosis type 2 in a 24-year-old-man. (a) Gd-enhanced coronal SE T1WI through internal auditory canals; (b) Gd-enhanced axial SE T1WI through the orbits. There are bilateral acoustic nerve schwannomas, seen as enhancing lesions within the auditory canals. Moreover, this patient developed multiple meningiomas: large convexity meningioma in the right temporo-parietal region, meningioma “en plaque” with dural thickening and marked hyperostosis involving the left parietal bone, and a left optic nerve meningioma. In addition, the patient developed an ependymoma of the 4th ventricle. This case illustrates the wide spectrum of tumours associated with neurofibromatosis type 2
254
Bilateral vestibular schwannomas are the hallmark of NF Type II. The pathology of the schwannoma occurring in NF type I or II is identical to those of the sporadic cases. NF type II is a disease of the central nervous system with development of tumours of the coverings: meningioma and schwannoma. Presence of multiple tumours is the rule [35] (Fig. 9.23).
References 1. Stefanko SZ, Vuzevski VD, Maas AIR, van Vroonhoven CCJ (1986) Intracerebral malignant schwannoma. Acta Neuropathologica 71:321–325 2. Burger PC, Scheithauer BW (1994) Atlas of tumor pathology: tumors of the central nervous system. AFIP, Washington 3. Li C, Youssem DM, Hayden RE, Doty RL (1993) Olfactory neuroblastoma: MR evaluation. AJNR 14:1167–1171 4. Truwit CL, Kelly WM (1993) The olfactory system. Neuroimaging Clin N Am 3:47–69 5. Broich G, Pagliari A, Ottaviani F (1997) Esthesioneuro blastoma: a general review of the cases published since the discovery of the tumour in 1924. Anticancer Res 17:2683–2706 6. Russel DS, Rubinstein LJ (1989) Pathology of tumours of the nervous system, 5th edn. Edward Arnold, London 7. Kadish S, Goodman M, Wang CC (1976) Olfactory neuroblastoma, a clinical analysis of 17 cases. Cancer 37:1571–1576 8. Miyamoto RC, Gleich LL, Bidinger PW, Gluckman JL (2000) Esthesioneuroblastoma and sinonasal undiiferentiated carcinoma: impact of histological grading and clinical staging on survival and prognosis. Laryngoscope 110:1262–1265 9. Chao KS, Kaplan C, Simpson JR, Haughey B, Spector GJ, Sessions DG, Arquette M (2001) Esthesioneuroblastoma: the impact of treatment modality. Head Neck 23:749–757 10. Resto VA, Eisele DW, Forastiere A, Zahurak M, Dj L, Westra WE (2000) Esthesioneuroblastoma: the John Hopkins experience. Head Neck 22:550–558 11. Regenbogen VS, Zinreich SJ, Kim KS, Kuhajda FP, Applebaum BJ, Price JC, Rosenbaum AE (1988) Hyperostotic esthesioneuroblastoma: CT and MR findings. J Comput Assist Tomogr 12:52–56 12. Vanhoenacker P, Hermans R, Sneyers W, Vanderporre H, Soncke R, Baert AL (1993) Atypical aesthesioneuroblastoma: CT and MRI findings. Neuroradiology 35:466–467 13. Daxer A, Sailer U, Ettl A, Bleckenwegner G, Felber S (1998) Choristoma of the optic nerve: neuroimaging characteristics and association with spinal lipoma. Ophthalmologica 212:180–183 14. Ing EB, Garrity JA, Cross SA, Ebersold MJ (1997) Sarcoid masquerading as optic nerve sheath meningioma. Mayo Clin Proc 72:38–43 15. Khadem JJ, Werter JJ (1997) Melanocytomas of the optic nerve and uvea. Int Ophthalmol Clin 37:149–158
H. Tanghe et al. 16. Miller (1988) Walsh and Hoyt’s clinical neuro-ophthalmology, vol 3, 4th edn. Williams & Wilkins, Baltimore 17. O’Keefe M, Fulcher T, Kelly P, Lee W, Dudgeon J (1997) Medulloepithelioma of the optic nerve head. Arch ophthalmol 115:1325–1327 18. Rubio A, Meyers SP, Powers JM, Nelson CN, de Papp EW (1994) Hemangioblastoma of the optic nerve. Hum Pathol 25:1249–1251 19. Sadun F, Hinton DR, Sadun AA (1996) Rapid growth of an optic nerve ganglioglioma in a patient with neurofibromatosis 1. Ophthalmology 103:794–799 20. Weber AL, Klufas R, Pless M (1996) Imaging evaluation of the optic nerve and visual pathway. Neuroimaging Clin N Am 6:143–175 21. Luh GY, Bird CR (1999) Imaging of brain tumors in the pediatric population. Neuroimaging Clin N Am 9:671–716 22. Cumming TJ, Provenzale JM, Hunter SB, Friedman AH, Klintwort GK, Bigner SH, McLendon RE (2000) Gliomas of the optic nerve: histological, immunohistochemical (MIB-1 and p53), and MRI analysis. Acta Neuropathol 99:563–570 23. Newman SA, Jane JA (1991) Meningiomas of the optic nerve, orbit and anterior visual pathways. In: Al-Mefty O (ed) Meningiomas. Raven Press, New York, pp 461–494 24. Shields JA (1989) Diagnosis and management of orbital tumors. W.B. Saunders, Philadelphia 25. Arnoutovic KJ, Husain MM, Linskey ME (2000) Cranial nerve root entry zone primary cerebellopontine angle glioma: a rare and poorly recognized subset of extraparenchymal tumours. J Neurooncol 49:205–212 26. Eldevik OP, Gabrielsen TO, Jacobsen EA (2000) Imaging findings in schwannomas of the jugular foramen. AJNR Am J Neuroradiol 21:1139–1144 27. Mark SA, Seltzer S, Harnsberger HR (1993) Sensorineural hearing loss: more than meets the eye. AJNR Am J Neuroradiol 14:37–45 28. Casselman JW, Kuhweide R, Ampe W, Meeus L, Steyaert L (1993) Pathology of the membranous labyrinth: comparison of T1- and T2-weighted and gadolinium-enhanced spin-echo and 3DFT-CISS imaging. AJNR 69 14:59–69 29. Hsu L (2008) Intracranial schwannomas. In: Newton HB, Jolenz FA (eds) Handbook of neuro-oncology neuroimaging. Academic, New York, pp 408–418 30. Tali ET, Yuh WTC, Nguyen HD (1993) Cystic acoustic schwannoma: MR characteristics. AJNR Am J Neuroradiol 14:1241–1247 31. Ricci PE (1999) Imaging of adult brain tumors. Neuroimaging Clin N Am 9:651–669 32. De Marco JK, Hesselink JR (1993) Trigeminal neuropathy. Neuroimaging Clin N Am 3:105–128 33. Weber AL, McKenna MJ (1994) Radiologic evaluation of the jugular foramen; anatomy, vascular variants, anomalies and tumors. Neuroimaging Clin N Am 4:579–598 34. Parizel PM, Simoens WA, Matos C, Verstraete KL (2001) Tumors of peripheral nerves. In: De Schepper AM, Parizel PM, De Beuckeleer L, Vanhoenacker F (eds) Imaging of soft tissue tumors, 2nd edn. Springer, Berlin, pp 301–330 35. Akeson P, Holtas S (1994) Radiological investigation of neurofibromatosis type 2. Neuroradiology 36:107–110
Meningeal Tumors
10
Antonios Drevelegas, George Karkavelas, Danai Chourmouzi, Glykeria Boulogianni, and Anastasios Petridis
Contents 10.1 Meningioma............................................................. 255 10.1.1 Imaging..................................................................... 260 10.2 Hemangiopericytoma............................................. 289 10.3 Hemangioblastoma................................................. 292 References............................................................................ 298
A. Drevelegas Department of Radiology, AHEPA university Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece e-mail:
[email protected] G. Karkavelas Department of Pathology, Aristotle University of Thessaloniki, School of Medicine, Thessoloniki, Greece e-mail:
[email protected] D. Chourmouzi and G. Boulogianni Department of Radiology, Interbalkan Medical Center, Thessoloniki, Greece A. Petridis Department of Radiology, Papanicolaou General hospital, Thessoloniki, Greece
10.1 Meningioma Meningiomas are vascular, nonglial tumors of the central nervous system arising from the cellular elements of the leptomeninges. They generally receive their blood supply from branches of the external carotid arteries although large tumors will also recruit branches from the internal carotid arteries [1, 2]. Although their origin from the mesodermal meningoendothelial or neuroectodermic arachnoid capsule cell implies a general extraneuraxial embryology, they have been recently classified into three basic categories. The largest one includes tumors growing within the neuraxis called primary neuraxial meningiomas (PNM). The second group includes tumors that grow outside the neuraxis referred to as primary extraneuraxial meningiomas (PEM) and the third group comprises tumors extending directly outside the neuraxis or metastasize and called secondary meningiomas [3]. Thus, the great majority of meningiomas are included in the PNM group and have an apparent dural attachment, while the relatively small proportion of all meningiomas included in the two other groups, which lack dural attachment, is further divided into four subgroups: (a) wholly intraventricular tumors of the choroids plexus, (b) so-called subcortical meningiomas, partly embedded in the brain, most of which arise from the lateral margin of the superior tela choroidea, (c) tumors of the deep Sylvian cleft anchored to the internal carotid artery and its branches, and (d) exceptional free subtentorial meningiomas of inconclusive origin [4, 5]. The relative frequency of meningiomas regarding all intracranial tumors varies between 15% [1, 3, 6] and 20% [1, 7, 8]. Their incidence rises with advancing age and they account for 10–20% of all primary intracranial tumors in adulthood [9, 10]. On the other hand,
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_10, © Springer-Verlag Berlin Heidelberg 2011
255
256
a review of multiple large series reveals that meningiomas constitute on an average 23% of intracranial neoplasms in adult populations ranging between 15 and 40% [11]. According to other studies, the relative frequency of meningiomas regarding all symptomatic intracranial neoplasms is approximately 15% and about 33% of all incidental (asymptomatic) intracranial neoplasms [12, 13]. Symptomatic meningiomas occur 2–3 times more often in female patients, especially during middle-age [12–14] while childhood meningiomas are uncommon but not rare lesions with the predominance of female sex being rather similar to that in adults [9]. Childhood meningiomas represent 1.0–4.2% of central nervous system tumors and 1.5–1.8% of all intracranial meningiomas [15–19]. They are considered as having diverse characteristics relating to the clinical and biological behavior and outcome [9]. Although women tend to develop meningiomas more frequently than men, when the statistics are broken down according to location, the parasagittal and convexity meningiomas are of equal incidence in men and in women, whereas lesions clustered about the skull base are much more common in women than in men. In addition, it has been reported that there is no sign of female predominance in African Americans, whereas there is a female predominance in whites and Asians. In Nigeria, meningiomas have been found to be more common in men [11]. On the other hand, the male to female ratio of PEMs has been found to be 3:5, while the male to female ratio of intraosseous meningiomas is 3:7 [20]. Tumor location is the single most important feature regarding therapy since it practically defines the terms of surgical intervention. The frequency of meningiomas at various intracranial sites varies from study to study. In general, the convexity and parasagittal meningiomas tend to occur for approximately 50% of all intracranial meningiomas, while sphenoid ridge meningiomas account for approximately 20% the anterior cranial fossa for 10% and those of parasellar regions for approximately 10% [1, 7, 8]. According to other studies, excluding spinal meningiomas which constitute approximately 12% of all meningiomas, intracranial and juxtacranial meningiomas arise in the following locations in descending order of frequency: convexity meningiomas (lateral hemisphere) 20–34%, parasagittal (medial hemisphere) 18–22% (including falcine meningiomas 5%), sphenoid and middle cranial fossa 17–25% (including middle ear
A. Drevelegas et al.
meningiomas, which account for approximately 37% of all middle cranial fossa meningiomas), frontobasal 10%, posterior fossa 9–15% with tentorium cerebelli meningiomas representing the 2–4%. Cerebellar convexity meningiomas, account for approximately 5%, cerebellopontine angle (CPA) M for 2–4%, clivus less than 1%, intraventricular 2–5%, orbital less than 1–2%, and ectopic less than 1% [1, 10, 20–23]. Considering the location frequency, it is conceivable that true meningiomas tend to occur where meningothelial cells and arachnoid cap cells are most numerous. The arachnoid granulations or villi have a large number of cap cells and therefore, are common sites of origin for meningiomas especially along the dural venous sinuses where the villi are mostly concentrated [1], or along the cranial sutures where arachnoid granulations or rest of arachnoid cells are often present [2]. Although the frequency of meningiomas without dural attachment and ectopic meningiomas is very low, it is important to recognize the rare and atypical locations in order to avoid misdiagnosis. In cases of meningiomas far removed from the neuraxis, including mediastinum, lung and adrenal glands, arachnoid cells’ ectopy, and meningoendothelial differentiation from pluripotential mesenchymal cells might be implicated in their development [1]. On the other hand, direct extension of a primary intracranial meningioma, extension from arachnoid cells accompanying nerve sheaths, metastases from intracranial meningiomas [6], or arachnoid cell capture as the result of trauma analogous to the formation of leptomeningeal cyst [3] have been proposed as possible mechanisms of development. Van Tassel has described a mechanism that results from head molding during birth [24]. The presenting signs and symptoms of meningiomas, when neurofibromatosis is excluded, are related to the tumor location and size; they are often nonspecific and vague, primarily related to brain compression and edema from the adjacent neoplasm [1, 11]. Signs of increased cranial pressure (nausea, headache, and vomiting) are found in 50% of the patients. Confusion, focal weakness, and seizures are the most common symptoms while paresis is the most frequently found physical sign [1, 9, 25]. Headache has been reported as the most common single presenting symptom found in 36%, while normal physical examination has been reported to be found in 26% of the patients [11]. Symptoms due to compression of focal adjacent structures are also nonspecific, as these signs could be
10 Meningeal Tumors
caused by any compressive lesion. For example, motor weakness and left hemiparesis have been described in a case of a small mass in the right frontal region [26], a 2-year history of headache in a frontal meningioma [27], facial numbness and double vision were reportedly the presenting signs of a temporal lobe meningioma [28]. Considering the symptoms of CPA meningiomas, it is important to recognize that meningiomas in this region do not have the propensity to involve the internal auditory canal, which is a fairly constant feature of schwannomas [1]. The development of ocular symptoms due to optic nerve atrophy in orbital meningiomas usually lead the patient to seek medical care, while a slowly progressive proptosis is a common clinical sign in this unusual location [20]. Visual disturbance (as bitemporal hemianopia, severe unilateral visual loss or unilateral temporal hemianopia), is the most common symptom in tuberculum sellae meningiomas. Cases with mild hyperprolactinemia or moderate hypothyroidism have also been reported [29]. Most patients with cavernous meningiomas, present with progressive neuropathies of the optic and ocular nerves, expressed as visual disturbance and opthalmoplegia as the oculomotor nerve may be entrapped by the tumor [30, 31]. In cases of en plaque meningiomas, the hyperostosis is frequently responsible for the symptoms that the patient experiences as a result of impingement of adjacent structures [32]. Considering intraosseous meningiomas, slight pain of the affected skull region may be the presenting symptom, while a history of trauma might be related to the development of meningiomas [33]. It is worth noting that proptosis produced by an intraosseous meningiomas of the orbit may be clinically misleading thyroid disorder [20]. Destruction of the skull base with secondary intracranial extension is seen in over one-third of patients with nasopharyngeal and paranasal sinuses meningiomas [1]. In childhood meningiomas, the insidious onset of symptoms is usually mentioned as the cause of nonspecific symptomatology. The most common clinical manifestation of meningiomas in childhood is the sign of increased cranial pressure (45%) followed by seizure (21%), and other rare symptoms and signs [9]. In other studies, however, focal neurological deficit is reported as the most frequent symptomatology (33%), followed by seizure (25%), and increased cranial pressure symptoms (25%)
257
[9, 34]. This difference is probably related to the patient’s late referral to physicians. Despite their vascularity, meningiomas rarely result in hemorrhages. Spontaneous intracranial hemorrhage occurs in 3.9% of all brain tumors, mostly metastatic or malignant gliomas, while the incidence of intracranial hemorrhage is 1.3% of all meningiomas [35]. In 35% of cases, the location of hemorrhage is outside the tumor [36]. The most common type of bleeding is subarachnoid and/or subdural hemorrhage, followed by intracerebral, intraventricular or intratumoral hemorrhage [37]. Although the hemorrhage may be due to the release of humoral substances that produce a consumptive coagulopathy effect, a physical abnormality of the vasculature within the meningiomas is a more likely cause. Rarely, meningiomas have been associated with secondary disease states, such as hypercoagulable states induced by glycoproteins secreted by the tumors that can finally alter the homeostatic levels of fibrin [11]. Meningiomas are usually benign neoplasms and except for neurofibromatosis, there is no proven genetic predisposition for their development. Although some locations and histological subtypes could be correlated with “poor” long-term outcome, the prognosis is good if the tumor can be totally removed [9]. Resection of the tumor with a wide margin is necessary to achieve complete excision of meningioma and to avoid recurrence [38]. Childhood meningiomas are associated with significantly shorter survival time due to limitations in surgical intervention, association with neurofibromatosis, incr eased incidence of sarcomatous changes, large size of tumors [25, 39, 40], and inability to use adjuvant radiotherapy and chemotherapy for the residual tumor tissues. Adjuvant radiotherapy appears to be beneficial after incomplete excision of meningiomas in adults [41]. The tumor growth of the incidental meningiomas seems to be associated with the age of the patient and the size of the tumor at its initial diagnosis. Increased tumor growth rate has been reported in younger patients [38]. Generally 90–95% of meningiomas are considered benign, 5–7% are atypical, and 1–2% are frankly malignant [7, 8]. Meningiomas are well-circumscribed globular or lobulated dural-based tumors, clearly demarcated from the brain (Fig. 10.1). Pathology: Meningiomas are characterized by incr eased inter- and intratumoral heterogeneity. Although by recent classifications a number of types are recognized (and new variants are described), only a minority of them
258
A. Drevelegas et al.
Fig. 10.3 Fibrous meningioma consisting of parallel spindle cells and bands of collagen (Hematoxylin-Eosin, original magnification ×400)
Fig. 10.1 Gross specimen of a globular meningioma
Fig. 10.4 Transitional meningioma characterized by the formation of whorls (Hematoxylin-Eosin, original magnification ×100) Fig. 10.2 Meningothelial meningioma, composed of nests of cells with rounded nuclei and indistinct cytoplasmic borders (Hematoxylin-Eosin, original magnification ×200)
is found in pure form. Meningothelial, fibrous, and transitional meningiomas are the most commonly found forms. In meningothelial (syncytial) meningiomas, the neoplastic cells are uniform, arranged in varying sized lobules. Their cellular borders are indistinct, forming a syncytium, and their nuclei are oval or rounded (Fig. 10.2). Fibrous (fibroblastic) meningiomas are less cellular tumors consisting of elongated, spindle-shaped cells in
an abundant collagen matrix. These cells are arranged in interlacing bundles (Fig. 10.3). Transitional (mixed) meningiomas are characterized by the presence of neoplastic cells arranged in whorls, in the center of which collagenized vessels or psammoma bodies are often recognized (Fig. 10.4). Although psammoma bodies are often found in meningiomas, when abundant, they characterize the tumor as psammomatous meningioma (Fig. 10.5). Less commonly found subtypes are angiomatous meningiomas (characterized by numerous blood vessels), microcystic meningiomas (exhibiting a mucinous background) [42] (Fig. 10.6), and secretory meningiomas. The secretory meningioma is a rare subtype of meningioma,
10 Meningeal Tumors
259
Fig. 10.5 Psammomatous meningioma with multiple psammoma bodies (Hematoxylin-Eosin, original magnification ×400)
Fig. 10.7 Atypical meningioma with necrosis (HematoxylinEosin, original magnification ×400)
Fig. 10.6 Microcystic meningioma with multiple small cysts (Hematoxylin-Eosin, original magnification ×400)
Fig. 10.8 Papillary meningioma with the characteristic papillary architecture (Hematoxylin-Eosin, original magnification ×400)
characterized by eosinophilic, Pas-positive hyaline inclusions (pseudo-psammoma-bodies), and the expression of several immunohistochemically detectable markers, including IgA, IgM, CEA, and alpha-1-antitrypsin [43]. An extensive chronic inflammatory infiltrate is found in the lymphoplasmacyte-rich meningioma [44] and mesenchymal differentiation in metaplastic men ingiomas. The above-described subtypes are grade I by WHO classification, exposing a low risk of recurrence and aggressive growth. On the contrary, the WHO grade II (chordoid, clear cell and atypical) meningiomas, and especially the WHO grade III (papillary, rhabdoid, and anaplastic) meningiomas are more aggressive with a higher risk for recurrence [45].
In the chordoid subtype, among typical areas of meningiomas, neoplastic cells with features reminiscent of chordomas are found [46], whereas in clear cell meningiomas, the tumor cells exhibit a clear cytoplasm. Atypical meningiomas are cellular tumors exhibiting a sheet-like growth of small, atypical cells. An increased mitotic rate as well as necroses are features of this tumor (Fig. 10.7). Papillary and rhabdoid meningiomas are rare and aggressive subtypes where papillary forms [47] (Fig. 10.8) or rhabdoid cells, respectively are recognized [48]. Anaplastic (malignant) meningiomas are characterized by the presence of malignant features, such as overt cell anaplasia, high mitotic rate, areas of necroses and/or brain invasion.
260
All meningiomas are positive to vimentin and most of them to epithelial membrane antigen (EMA). A variable grade of S-100 protein positivity is also expressed in some meningiomas.
10.1.1 Imaging In 72–85% of intracranial meningiomas, computed tomography(CT) demonstrates typical diagnostic features [1, 23, 49]. On noncontrast CT (NCCT), meningioma appears as a unilobular, homogeneous high density mass in relation to brain parenchyma with attenuation values varying from 40 to 50 Hounsfield units (HU) [10, 14, 49, 50]. The hyperdensity on NCCT has been attributed to compact and dense cellularity with a relatively small amount of intercellular water and in part, to the presence of calcified psammoma bodies within the tumor [10, 50, 51]. After the administration of contrast material, they show intense and homogeneous contrast enhancement in approximately 72–80% [1, 10, 14, 52]. The use of multiplanar reconstructions generated from multislice spiral CT data sets improve the assessment of meningiomas (Fig. 10.9). Lesions that are significantly calcified on the NCCT scan show little to no density changes after contrast administration. Calcification is seen on CT in 20–27% of meningiomas [10, 49]. It is usually microscopic or punctuated, but may be large, conglomerate, peripheral, or central (Figs. 10.9a and 10.10). Among the histological subtypes, transitional and fibroblastic meningiomas most frequently show visible calcium deposits on CT with incidence of 39%. The presence of calcification indicates slowly progressive benign nature [50]. Malignant meningiomas are rarely calcified. Another typical imaging characteristic of meningioma is hyperostosis of the adjacent calvarium (Fig. 10.11). Hyperostosis has been found in 18–50% of cases [10]. The hyperostosis occurs only in those masses that are immediately adjacent to the bone. Pathologically, the hyperostosis is due to the new cortical bone formation in parallel sheets along the inner table of the skull. The reactive bone formation is usually accompanied by micro- or macroscopic invasion of the bone by the meningioma tissue (Fig. 10.20c). Osteoma, fibrous displasia, Paget’s disease, and hyperostosis frontalis may imitate the hyperostosis typically seen in meningiomas. En plaque meningiomas have a
A. Drevelegas et al.
greater association with hyperostosis seen on CT [53]. This type of meningioma consists of a thin layer of neoplasm that closely follows the adjacent inner table with disproportion to the size of the underlying neoplasm thickening of the bone (Fig. 10.12). En plaque meningiomas are flat or nodular and do not produce significant mass effect in the adjacent brain. Some tumors may extend throughout the table of the skull and present primarily as a scalp mass (Fig. 10.13). Bone destruction by meningiomas is an uncommon feature, found in approximately 3% of cases [54] (Figs. 10.14 and 10.27a, b). Benign as well as malignant meningiomas may invade the skull, causing bone destruction. These destructive meningiomas are typically associated with adjacent areas of hyperostosis. If a purely destructive skull lesion is identified, this is more likely due to metastasis, sarcoma, or myeloma. Characteristic angiographic findings for meningiomas include dilatation of meningeal feeding arteries, radiating intratumoral dural arteries also known as spoke wheel pattern and fairly homogeneous tumoral blush of varying intensity, which persists into the late venous phase known as “mother-in-law blush” (comes early, stays late – tumor blush) (Fig. 10.15). Degree of vascularity can vary depending on histologic subtype. Angioblastic and transitional types tend to be more vascular than fibroblastic or syncytial varieties. Psammomatous subtypes are, sometimes, avascular [55–57]. The morphologic characteristics of meningiomas on MRI are similar to those seen on CT studies. Typically, they are peripheral unilobular masses with broad-based dural attachments and smooth, welldefined borders. On T1-W images, meningiomas are usually isointense or mildly hypointense to normal gray matter [22, 58–60]. This finding contrasts with those from other intracranial tumors, which are usually moderately hypointense on T1W images, due to the increased water content of the neoplastic tissues. The relatively high cellularity and low water content of most meningiomas may account for the generally isointense appearance on T1W images. Although the signal intensity on T2W images varies, most tumors are reported to be isointense to mildly hyperintense compared with gray matter. Nearly all meningiomas enhance rapidly and intensely following contrast administration [61] (Figs. 10.16 and 10.20). Enhanced MR scans are particularly useful in detecting small, inconspicuous meningiomas that are isointense with adjacent cortex on all pulse sequences. The wide
261
10 Meningeal Tumors
a
b
c
d
Fig. 10.9 Typical meningioma. Noncontrast (CTNCCT) shows a well-marginated, high density mass with punctuate calcification (arrow) (a) which enhances homogeneously after the
administration of contrast material (b). Coronal (c) and sagittal (d) image reconstructions better depicts the tumor margins and location
variation of signal intensity among meningiomas reflects the diversity of histopathology in meningiomas. In general, T1WI values offer little to no specificity whereas T2W images can give information about histological subtype, vascularity, and consistency. Men ingiomas hyperintense to the cortex on T2WI are usually soft, more vascular, and more frequently of
syncytial or angiomatous subtype (Fig. 10.17). Tumors hypointense or hypo-isointense on T2WI tend to have harder consistency and are more of fibroblastic or transitional subtype [62, 63] (Fig. 10.18). Advanced MR imaging techniques are usually of little value in making the diagnosis in patients with typical imaging findings of meningioma.
262
A. Drevelegas et al.
a
b
c
d
Fig. 10.10 Calcification patterns of meningiomas in NCCT. (a) Ring like peripheral calcification. (b) Dense confluent calcifications. (c) Multiple psammomatous calcifications (arrowheads)
with partial rim-like calcification (arrow). (d) Almost completely calcified meningioma. (e) Multiple punctuate calcifications
10 Meningeal Tumors
Fig. 10.10 (continued)
Fig. 10.11 Axial CT with bone window shows hyperostosis of the skull (arrows) due to a frontal meningioma
Studies focus on appearance of meningiomas on diffusion-weighted images and the possible correlation with histopathologic findings. They postulated that atypical and malignant meningiomas tend to be markedly hyperintense on diffusion-weighted MR images and exhibit marked decrease in ADC values when compared with normal brain parenchyma (Fig. 10.19), Although benign meningiomas have a variable appearance on diffusion-weighted images, they tend to have higher ADC
263
values compared with normal brain (Fig. 10.20), with the exception of densely calcified or psammomatous meningiomas, which may have a low ADC (Fig. 10.21). In contrast, other investigators showed that the ADC may not be predictive of the degree of malignancy in meningiomas or of their histologic subtype [64–66]. Even more recently, investigators hypotheses that classic and atypical meningiomas have different patterns of intratumoral water diffusion and that differences in diffusion anisotropy as detected by DTI allow differentiation between them. Classic meningiomas have significantly lower intratumoral FA, and higher ADC compared with atypical meningiomas. These DTI findings indicate that intratumoral microscopic water motion is more disorganized in classic than in atypical meningiomas. They conclude that DTI may be helpful in the differentiation between classic meningiomas and atypical meningiomas [67]. MR spectroscopy may provide additional information in differential diagnosis. Cho reflects membrane turnover, correlates with malignancy in astrocytic tumors, and forms high peaks in meningioma. The most common proton spectrum found in meningiomas is a high Cho peak with low or absent NAA and Cr and variable amounts of lactate. Most importantly, an unusually high ratio of Ala to Cr has been found in meningiomas because of the high Ala and low Cr content, and this is a relatively specific finding for meningioma. Alanine is seen as doublet centered at 1.47 ppm and inverts on the long-TE sequence (Fig. 10.20f, g). MR spectroscopy has been shown to have some ability to differentiate histologically atypical meningiomas on the basis of lactate peak at 1.3 ppm. Although it is observed in a few benign meningiomas (Fig. 10.20g), the lipid peak signal is correlated with the extent of necrosis and tumor malignancy, and lipid/creatinine may be a valuable marker in the differential diagnosis of anaplastic meningiomas. Other series showed that there is considerable overlap of spectra between typical and atypical meningiomas [68, 69]. Perfusion MR has been studied recently in evaluation of meningiomas. Meningiomas are extraaxial hypervascular tumors that have dual arterial supply from branches of the external carotid artery and internal carotid artery. In the absence of a blood-brain barrier, permeability of their feeding arteries is high and mean rCBV and rCBF are the highest of all intracranial tumors (Fig. 10.20h). Perfusion MR imaging can
264
A. Drevelegas et al.
a
b
c
Fig. 10.12 En plaque meningioma. (a) CT scan reveals a marked bony hyperostosis in the right frontal region. (b) On axial T2-weighted image the hyperostosis shows low signal
intensity. (c) On axial postcontrast T1-weighted image a thin linear enhancement of the dura is seen across the inner table which is in disproportion to the adjacent thickened bone
provide useful functional information on meningiomas and help in the preoperative diagnosis of some subtypes of meningiomas. Not surprisingly, CBV maps show highest levels for angiomatous type and the lowest for fibrous meningiomas. Moreover, the mean rCBV value of benign meningiomas is usually higher than that of malignant meningiomas. However, the difference between the two groups shows no statistical significance. Furthermore, maps of relative rCBV provide
hemodynamic information in meningiomas and monitor the treatment effect of embolization in meningiomas more precisely than T1-weighted contrast-enhanced imaging [70, 71]. A useful feature in confirming the extra-axial location of the suspected meningioma is the inward bowing of the gray–white junction of the adjacent brain parenchyma often called ‘white matter buckling’ [72]. White matter buckling is especially well visualized on
265
10 Meningeal Tumors
a
b
c
d
Fig. 10.13 En plaque meningioma with extracranial component. (a) Axial T2-weighted image demonstrate a left occipital lesion with extracranial mass elevating the subcutaneous tissue. (b) Axial post contrast T1-weighted image demonstrate enhance-
ment of both meningeal and extracranial component. (c) CT scan reveals hyperostosis of the left occipital bone. (d) MR venography reveals thrombosis of the ipsilateral transverse and sigmoid sinus
266
a
A. Drevelegas et al.
b
Fig. 10.14 Axial postcontrast (a) and coronal (b) CT images show lobulated bilateral falcine mass with intense homogeneous enhancement. Note the destruction of the calvarium and the extracranial extension of the mass
MRI studies because of the superior recognition of the gray and white matter of the superficial brain on MRI examinations (Figs. 10.22 and 10.27c). Calcification within meningioma is rarely identified on MR images. However, focal large dystrophic calcifications may present as hypointense foci within the tumor (Fig. 10.18a). Another useful MR characteristic is the presence of signal void pseudocapsule. This pseudocapsule consists of linear signal void representing the dura itself, interposed between the tumor and the brain parenchyma, as well as of punctuate foci of signal void owing to the displaced vessels (Fig. 10.23). The presence of feeding artery as a signal void entering the tumor may also be seen [22, 59]. There may also be a CSF cleft trapped between the cortex and the meningioma, which demonstrates low-signal intensity on T1W images and highsignal intensity on T2W images (Fig. 10.24). A linear enhancement along the dura mater, on either side of meningioma on contrast-enhanced MRI called “dural tail sign” is considered an important finding in the diagnosis of meningioma (Figs. 10.20c and 10.25). This sign is not specific to meningioma and also observed
in several conditions including glioma [73, 74], brain metastasis [75], acoustic neuroma [76, 77], lymphoma [78, 79], adenoid cystic carcinoma [80], sarcoidosis [78], and aneurysm [81] (Figs. 10.39–10.41). The precise pathologic explanation for the “dural tail” sign is questionable, whether there is correlation with dural invasion of intracranial meningiomas or that is only reactive thickening of the dura mater. Some reports have found reactive vascular changes [82, 83] whereas other describe neoplastic invasion into the subarachnoid or subdural space [28, 84]. Hutzelmann et al. [38] found that the presence of a thickened enhanced linear structure or tail adjacent to a meningioma correlated in 64.5% of cases with tumor invasion and in 35.5% of cases with loose connective tissue proliferation, hypervascularity, and vascular dilatation. However, 30.4% of meningiomas without the meningeal sign also infiltrate the dural matter. Therefore, it is still not possible to determine with MRI whether there is dural infiltration of a meningioma or not. In medical practice, most neurosurgeons resect this enhancing dura, if possible, in an attempt to prevent tumor recurrence.
267
10 Meningeal Tumors
a
b
c
Fig. 10.15 Angiogram of a large convexity meningioma growing through the calvarium. Sagittal postcontrast T1-weighted image (a) demonstrates a large parieto-occipital meningioma with intense homogeneous contrast enhancement. Lateral view
of external carotid angiogram; arterial phase (b) shows “spoke wheel” appearance of tumoral vessels. Late venous phase (c) shows persistence of tumor blush
Several imaging features such as peritumoral edema, cystic changes, lipomatous transformation, intracranial hemorrhage, focal or diffuse; irregular contour, poorly defined margins, and ring enhancement are considered unusual or atypical. It is important
to recognize the variable features of these neoplasms so that an atypically appearing meningioma is not confused with other intracranial masses. Despite meningiomas representing extra-axial growing tumors, 60% of all are associated with brain
268
a
A. Drevelegas et al.
b
c
Fig. 10.16 Typical MR findings in a parasagittal meningioma. Axial T1WI (a) and coronal T2WI (b) show a well delineated, isointense mass on both sequences. After administration of contrast medium (c) the mass shows intense and homogeneous enhancement
edema [23, 85, 86]. It is more common with large lesions but may be extensive with small ones. Peritu moral brain edema in meningiomas can aggravate clinical symptoms, has a positive correlation with presence of seizures, adversely affect surgical outcome, has a positive correlation with higher recurrence rates,
as well as it may be incorrectly suggestive of an intra-axial tumor (glioma). Peritumoral brain edema may be related to the invading potential of meningiomas. Reports have found that severe edema is associated with more aggressive syncytial and angioblastic cell types, and mild-to-moderate edema associated
269
10 Meningeal Tumors
a
b
Fig. 10.17 Angioblastic meningioma appears hypo-/isointense on T1- and hyperintense on T2-weighted image
a
Fig. 10.18 (a, b) Fibroblastic meningioma. (a) Axial T1-weigh ted image shows a right parasagittal isointense mass. The focal low signal intensity is consistent with calcification (arrowhead).
b
(b) On T2-weighted image the mass is hypo- and isointense. Note also the extensive peritumoral edema
270
A. Drevelegas et al.
a
b
c
d
Fig. 10.19 Atypical meningioma. Axial T2-weighted (a) and post contrast T1-weighted images (b) show a large meningioma with both intraventricular and intraparenchymal component with intense inhomogeneous enhancement. On DWI (c) the intra-
parenchymal component of the lesions is hyperintense compared with the normal brain parenchyma. On apparent diffusion coefficient (ADC) image (d) the meningioma shows low ADC (0.7 × 10−3 mm/s)
with fibroblastic and transitional cell types. MRI is the modality of choice in detecting the extent of peritumoral edema.
Currently, it is widely accepted that the cause of edema is vasogenic rather than cytotoxic. Vasogenic edema, as a result of disrupted blood-brain barrier, is an
271
10 Meningeal Tumors
a
b
c
d
Fig. 10.20 Frontal meningioma. Axial T2-weighted (a) and axial T1-weighted (b) images show a parasagittal convexity meningioma isointense on both sequences compared to normal brain compressing the falx. On postcontrast T1-weighted image (c) the mass shows intensive inhomogeneous enhancement. Note dural tail (arrow) and linear enhancement extended to the calvarium (arrowheads). On pathology report bone invasion was mentioned. Note hyperostosis with marked thickening of diploic
space. On DWI (d) the lesion is isointense. On MRV (e) superior sagittal sinus thrombosis is revealed. H MR spectrum TE 135 (f) from meningioma shows many metabolites, among which are the prominent choline peak at 3.2 ppm and the alanine peak, which appears as a double peak at at 1.4 ppm and is inverted on the long-TE 270. (g) Note the lipid peak as well. On perfusion color map (h) the mass is hyperperfused (rCBV:10.5)
272
A. Drevelegas et al.
e
g
f
h
Fig. 10.20 (continued)
accepted mechanism in intra-axial tumors, like gliomas. In meningiomas, a vasogenic edema is more difficult to conceptualize, growing primarily in extra-axial spaces. It is thought to be due mainly to the presence of small feeding vessels from the territory of the internal carotid artery that penetrate the pia matter and facilitate transport of extracellular fluid back into the brain [87]. Regarding the pathogenesis of meningioma-related edema, Philipon et al. [87] discussed a secretory- excretory phenomenon. According to this hypothesis, tumor cells may produce and secrete edema-inducing substances into the adjacent brain parenchyma. Other
findings suggest a hydrodynamic mechanism of ede magenesis [88, 89]. According to this theory, a pressure gradient must be supposed between the extracellular space of the tumor and the interstitium of the brain. A disruption of the arachnoid membrane has been mentioned rendering an osmotic dispersion of edemainducing macromolecules from the tumor into adjacent brain [90]. Damage of brain tissue by chronic tumorrelated pressure causing ischemia is also regarded as a pathogenetic mechanism [91]. The influence of vascular endothelial growth factor (VEGF) in angiogenesis and brain edema in
273
10 Meningeal Tumors
intracranial meningiomas is also discussed in many recent studies [92]. Several studies suggest the importance of VEGF for angiogenesis in meningiomas as in glioblastomas [93, 94]. There is also a pathogenetic correlation between VEGF expression, tumor vascularization, and edemagenesis [92]. If further studies
a
c
Fig. 10.21 Benign meningothelial meningioma. Axial T2-weigh ted image (a) shows a large parasagittal isointense well delineated lesion with perilesional edema. On axial post contrast T1-weighted image (b) meningioma shows intense contrast enhancement. DW
confirm these results, the treatment with anti-VEGF antibodies could probably influence the extent of the peritumoral edema. In general, abnormalities in frontal lobes and anterior temporal lobes produce more edema than similar abnormalities in other parts of the brain, whereas
b
d
MR image (c) showing the tumor to be hyperintense. ADC map (d) showing decreased signal intensity compared with normal white matter. On CT image (e) meningioma is densely calcified
274
A. Drevelegas et al.
a
Fig. 10.21 (continued)
Fig. 10.22 Large parietal meningioma. Sagittal T1WI shows an (extra-axial isointense mass compressing the adjacent cortical convolutions (arrows)
structures in posterior fossa are the most resistant to the spread of edema (Figs. 10.21a, 10.26, and 10.27c). Meningiomas with hypervascular areas such as the angiomatous types, and those forms that are more biologically aggressive tend to display greater amount of edema regardless of location.
b
Fig. 10.23 Frontal meningioma. (a) Axial T2WI shows a rimlike low signal intensity (arrows) representing the dural pseudocapsule. (b) T1WI at a lower level shows the signal voids (arrows) at the periphery of the mass due to the displaced vessels
The term cystic meningioma has been used to describe two different morphologies: intratumoral cavities and extratumoral or arachnoid cysts [1]. The
275
10 Meningeal Tumors
a
b
Fig. 10.24 Parasagittal meningioma. Axial T1- (a) and T2-weighted (b) images show a CSF cleft surrounding the tumor that exhibits low signal on T1 and high signal intensity on T2WI (arrows)
a
b
Fig. 10.25 Postcontrast T1WI of different meningiomas. Axial (a), coronal (b) and sagittal (c) images show different types of “dural-tail” enhancement (arrows)
276
A. Drevelegas et al.
c
Fig. 10.25 (continued)
currently used classification of cystic meningiomas, suggested by Wasenko et al. [95] is subdivided into five types based on cyst characteristics or location with respect to the tumor: (1) The cyst lies centrally; (2) it is eccentric; (3) there is peritumoral cyst with a wall composed of fibrous tissue and nests of tumor cells; (4) the peritumoral cyst is adjacent to the tumor within brain parenchyma and (5) the peritumoral cyst lies between the tumor and brain, with trapped, loculated CSF within it. The hypotheses on intratumoral cyst formation in meningiomas are secretion of fluid by tumor cells, cystic degenerative changes within the tumor, and ischemic necrosis and hemorrhage. Peritumoral cyst may be formed by reactive gliosis as a response to the tumor, formation of fluid by adjacent glial cells, and a process of edema with cyst formation and loculation of CSF. More speculative causes include demyelination, perfusion deficit, and hemorrhage [96]. True intratumoral cystic meningiomas with large fluid-filled cysts, are an uncommon variant. Benign meningiomas with heterogeneous enhancement that contain small nonenhancing areas of cystic change or necrosis occur much more frequently (up to 8–23% of cases) (Fig. 10.27) [10, 14].
Fig. 10.27 Intradiploic meningioma with a large cystic component (a) plain X-rays of the skull shows a poorly defined destructive lytic lesion of the calvarium. (b) CT image shows widening of diploic space with destruction of both inner and outer table of the calvarium. (c) Axial T2-weighted MR image shows diploic lesion with intermediate signal intensity, a large intracranial cyst
Fig. 10.26 Axial T2WI shows a left frontal meningioma with extensive white matter edema
A large cystic meningioma may have an atypical clinical presentation, in that they are more common in male and pediatric patients; these unusual clinical features often lead to a misdiagnosis of cystic or necrotic glioma [1, 97]. The presence of neoplastic cyst should be suspected when ring enhancement of the wall is seen [1] (Fig. 10.28). In surgical practice, in cases with contrast enhancement of the cystic component on CT or MRI, both the tumor and the cyst wall have to be excised in order to confirm histologically the presence of nest of neoplastic cells within the wall of the cyst [96]. Low signal intensity mass on T1W images associated with poor contrast enhancement are the characteristic signs of the rare microcystic variant of meningioma. The pathogenesis of the microcystic formation includes mechanisms such as degeneration
with high signal intensity compressing the adjacent cortical convolutions and perilesional edema. On axial (d) and sagittal (e) post contrast T1-weighted images an inhomogeneous enhancement of the calvarial mass is seen on both intra- and extracranial component. The cystic walls remain unenhanced
277
10 Meningeal Tumors
a
b
c
d
e
278
A. Drevelegas et al.
Fig. 10.28 Cystic meningioma. Axial postcontrast T1WI shows a left temporal meningioma with a large cystic component. The rim-like enhancement of the cyst is compatible with neoplastic involvement of the cyst wall (arrows)
a
and protein fluid transudation. The presence of typical meningothelial cells within the tumor and faint tumor staining showed in angiography suggest that microcystic meningioma might be a degenerative form of meningotheliomatous meningioma caused by poor blood supply [98]. Lipoblastic meningioma represents a variant in which there is a metaplastic change of meningothelial cells into adipocytes, through the accumulation of fat (mostly, triglycerides) within their cytoplasm [99]. Lipomatous meningiomas are markedly hypodense on CT (negative HU) and may have minimal to slight enhancement within the fatty regions [100] (Fig. 10.29). Xanthomatous change in meningioma can, histologically, be differentiated from the lipoblastic variant. Radiologic distinction, however, may be difficult since both contain excess lipid. A lipoblastic meningioma may be suggested when the fatty regions are larger, more confluent, and do not have prominent enhancement [100]. Secretory meningioma may be characterized by imaging features unusual for other subtypes of meningioma, such as low attenuation on CT, high (fat-tissue equivalent) signal intensity on T1W MRI, marked surrounding edema, and irregular contrast enhancement [27].
b
Fig. 10.29 (a) Postcontrast CT shows a low-attenuation (compatible with fat) lesion with ring-like enhancement. On sagittal T1-weighted (b) and axial T2-weighted (c) images the lesion
shows a signal intensity that is identical to that of the subcutaneous fat. (d) Gross specimen shows mass with fatty metaplasia (with permission)
279
10 Meningeal Tumors
c
d
Fig. 10.29 (continued)
Spontaneous intracranial hemorrhage associated with meningioma is an uncommon condition with incidence 1.3% of all meningiomas [101]; it is not related to sex, age, blood dyscrasia, hypertension or tumor location [102]. Many reports suggest that hemorrhagic meningiomas are most often of meningothelial or transitional types and rarely of the angiomatous type [4]. Hell reported a relatively high risk of bleeding in angiomatous and malignant meningiomas [103]; however; the majority of meningiomas manifesting hemorrhage are benign variants. The most common type of bleeding is subarachnoid hemorrhage, followed by intracerebral and intratumoral hemorrhage [37]. Less than 10% of meningioma-associated bleeding occurs in subdural space [103]. The most common hypothesis concerning the mechanism of hemorrhage is the rupture of the abnormal vascular networks of the tumor. This hypothesis is based on the histological findings, such as weak, thinwalled vessels or direct peritumoral vascular erosion by the tumor. Infarction of the tumor might be an important event in the process of peritumoral hemorrhage. As the tumor infarction progresses, the intratumoral pressure increases and rupture of the peritumoral vessels eventually occurs [101]. Intratumoral hemorrhage, when acute, appears on CT as a focal area of high density [50]. On MRI, the
acute intratumoral hemorrhage, shows hypointensity to isointensity on T1WI, while on T2WI, it is markedly hypointense due to the deoxyhemoglobin products of blood breakdown (Fig. 10.30). Hyperintensity on both T1W and T2W images is seen in the subacute phase while hypointensity on T2WI is seen in the chronic phase of hematoma due to hemosiderin deposition. In case of intracerebral peritumoral hemorrhage, CT may show an isodense or low density mass with crescenting intracerebral hematoma [101]. MR images reveal a homogeneous enhancing mass with peritumoral intracerebral hematoma. Diffusion-weighted images may confirm suspicion of an underlying lesion within the subdural hematoma [2]. Although meningiomas are usually homogeneous masses with homogeneous enhancement they may have an atypical ring enhancement. This unusual feature can be seen in both histologically typical meningiomas and in some malignant or aggressive histological variants that may have cyst formation, hemorrhage, or necrosis. The peripheral enhancement represents the meningeal neoplasm and the center, a necrotic region [1]. The differential diagnosis of this unusual imaging appearance includes necrotic glioma, metastasis or even an abscess. Although the vast majority of meningiomas occur along the large dural sinus (Figs. 10.13d and 10.20)
280
a
A. Drevelegas et al.
b
Fig. 10.30 Meningioma with intratumoral hemorrhage. (a) On NCCT the hemorrhage appears hyperdense. (b) On axial T2WI the intratumoral hematoma shows low signal intensity indicating acute/subacute bleeding
and at the skull base (Fig. 10.31), approximately 15% of them may occur in less common sites. CPA meningiomas represent the second most common mass lesions of the CPA, although less than 5% of all meningiomas occur in CPA [104, 105] (Fig. 10.32). Approximately, 80% of masses in the CPA are acoustic schwannomas and half of the remainder 20% are meningiomas. Orbital meningiomas account for less than 2% of cranial meningiomas, but constitute 10% of all intraorbital neoplasms, arising from the optic nerve sheath between the globe and the optic canal [21] (Fig. 10.33). However, according to Daffner et al. [20], orbital meningiomas represent 58% of all PEMs. Intraventricular meningiomas are the most common trigonal masses in an adult [106] accounting for approximately 2–5% of intracranial meningiomas [10, 23] (Fig. 10.34). They arise from the tela choroidea or the stroma of the choroid plexus itself and the 80% are located in the lateral ventricle, 15% in the third ventricle, and about 5% within the fourth ventricle [12, 107]. They are generally classified as meningiomas without dural attachment.
Subcortical meningiomas are mainly deep Sylvian meningiomas that arise from leptomeningeal infolding in the Sylvian fissure and involve branches of the MCA as they grow [108–110]. The mean age of appearance is reported to be 29.3 years [109], which is earlier than for ordinary meningiomas. They are also classified as meningiomas without dural attachment and have been reported to be more frequent in Japanese populations [111] (Fig. 10.35). En plaque meningiomas are a variant of PNM [32], which may infiltrate both the dura and the bone cloaking the inner table of the skull (Figs. 10.12 and 10.13). Diaphragma sellae (Fig. 12.13, Chap. 12) and tuberculum sellae meningiomas show clinical and radiological features similar to those of nonsecreting pituitary macroadenomas with suprasellar extension. However, preoperative differentiation is essential, because the transphenoidal approach used for adenoma’s removal is inappropriate for meningiomas, which should be approached via craniotomy [29]. Intraosseous meningiomas usually arise near the cranial sutures, particularly the coronal and pterional [20], originating from arachnoid cells trapped during
281
10 Meningeal Tumors
a Fig. 10.31 Parasellar meningioma. Axial (a), and coronal (b) postcontrast T1W images clearly show a strong enhanced parasellar mass extending into the orbital apex; the ethmoid cells, the
a
Fig. 10.32 Cerebellopontine angle (CPA) meningioma. Axial T1- (a), and T2-weighed (b) images show a broad based mass that is isointense on both sequences compressing the brain stem. Note the CSF cleft surrounding the tumor (arrowheads).
b sella Turkica, and thought the tentorium, into the posterior fossa compressing the midbrain
b
Postcontrast axial T1WI (c) shows intense homogeneous enhancement of the tumor. The obtuse angle with the adjacent dura and the “dural tail” (arrow) indicate the extra-axial origin of the tumor
282
A. Drevelegas et al.
c
Other sites of ectopic meningiomas include the outer table of the skull, the overlying skin, paranasal sinuses, parotid gland and parapharyngeal space, and they are all best considered as originated from inclusion of arachnoid cells disseminated during the formation of the skull [1]. Developments of meningiomas in patients infected with human immunodeficiency virus (HIV) have recently been studied. Meningiomas may have grown in these HIV-infected hosts because of either loss of immune function or dysregulation of cytokines [116] (Fig. 10.37). The frequency of multiple meningiomas is between 1 and 5% in surgical series and from 8 to 16% in autopsy and neuroradiologic series. Type 2 neurofibromatosis (NF2), inherited in an autosomal dominant manner with high penetrance, predisposes at an early age to multiple schwanomas, meningiomas, and spinal ependymomas, with bilateral vestibular schwannomas as the classic diagnostic hallmark [117] (Figs. 10.38 and 9.23). Differential diagnosis: There are multiple neoplastic and non-neoplastic entities that clinically and radiographically mimic meningiomas, including solitary fibrous tumors, hemangiopericytoma, gliosarcoma, leiomyosarcoma, dural metastases, Hodgkin’s disease, plasmocytoma, Rosai Dorfman disease, neurosarcoidosis, melanocytic neoplasms, plasma cell granuloma, Tolosa-Hund syndrome, pituitary macroadenoma,
Fig. 10.32 (continued)
skull development or after a cranial trauma [112, 113]. They are also referred to as calvarial [114] or intradiploic [115], their incidence is even lower than that of PEMs [20] and they are considered to be ectopic meningiomas (Figs. 10.27 and 10.36).
a
b
Fig. 10.33 Optic nerve sheath meningioma. Axial (a) and sagittal (b) post contrast T1-weighted with fat saturation images show a spindle shape enhancement along the length of the right intraorbital optic nerve sheath
283
10 Meningeal Tumors
a
b
c
Fig. 10.34 Intraventricular meningioma. Axial T1- (a), and T2-weighed (b) images show an oval, well-marginated mass in the right atrium that is isointense on both sequences. The mass causes local ventricular obstruction and mild periventricular edema.
Postcontrast axial T1WI (c) shows intense homogeneous enhancement. Note the relationship of the mass with the displaced choroid plexus (arrow)
284
a
A. Drevelegas et al.
b
c
Fig. 10.35 Subcortical meningioma Axial PD (a) and postcontrast T1-weighted (b) images show an extraaxial homogeneous enhanced mass without dural attachment. (c) Operative view. The
tumor (arrows) is seen on the surface of the brain cortex compressing the adjacent gyri. Note the reflected dura (arrowheads)
acoustic neuroma, choroid plexus tumors, ependymoma and rarely, intraventricular astrocytoma. The differential diagnosis of dural-based lesions in the brain varies from incidental and benign to symptomatic and malignant lesions. The above-mentioned entities involving the dura are rare and almost always diagnosed after tissue is obtained because of their
clinical and radiographic similarity to meningiomas. These lesions can closely resemble meningiomas in terms of signal characteristics, enhancement pattern, and location (Fig. 10.39). Moreover, the lesions may resemble meningiomas, even intra-operatively. Imaging characteristics alone can be misleading; neuropathological support is essential for accurate diagnosis.
285
10 Meningeal Tumors
b
a
d
c
Fig. 10.36 Intradiploic meningioma. (a) Lateral radiograph shows a lytic lesion of the calvarium (arrow). (b) Coronal CT with bone window shows an intradiploic mass expanding the diploic space. The tumor causes thinning of the inner and outer
table of the skull. (c) Sagittal T1WI demonstrates an isointense mass growing through the calvarium. (d) Coronal T1WI shows inhomogeneous enhancement of the mass that is extended both intra-and extracranially
286
a
A. Drevelegas et al.
b
c
Fig. 10.37 Meningioma in a patient with AIDS. Axial T1-weighted (a) and T2W (b) images show an extraaxial right parietal mass that is isointense on both sequences. (c) Following
the administration of contrast medium the tumor shows intense homogeneous enhancement with dura-tail sign
287
10 Meningeal Tumors
Fig. 10.38 Multiple meningiomas. Axial postcontrast T1W images in different levels show multiple enhanced meningiomas
Fig. 10.39 Solitary fibrous tumor (SFT). Postcontrast CT shows a dural-based lobulated mass with dural tail enhancement (arrow)
A thorough clinical evaluation can reveal likely diagnostic possibilities. In the presence of a heterogeneous dural-based mass with prominent internal vessels and bone erosion, the case of hemangiopericytoma should be considered [118] (Fig. 10.44). Leiomyosarcomas should be included in the differential diagnosis of duralbased lesions in HIV-infected patients [119]. Careful vigilance in patients with a history of cancer, presenting with new symptoms or imaging evidence of dural-based lesions, should raise the possibility of dural metastasis. Intracranial dural metastases (IDM) are found at autopsy in 9% of patients with advanced systemic cancer. In adults, metastasis from breast, lung, renal and prostate cancer are the most common (Fig. 10.40). In the pediatric age group, neuroblastoma or sarcoma are the most common neoplasms giving rise to skull or dura metastasis [120]. The diagnosis of lymphoma should be considered for lesions affecting the dura in high-risk immunocompromised patients. The presence of an apparent dural tail can be seen in a lymphoma. The absence of hyperostosis helps differentiate lymphoma from meningioma. The diagnosis of extramedullary hemopoiesis should be considered in patients with b-Thalassemia [121–123]. Hypercalcemia, hyperproteinemia, and serum gamma globulin peak in a patient with an extra-axial dural mass
288
a
Fig. 10.40 Dural metastasis in a 36-year-old woman with breast carcinoma. (a) Axial postcontrast T1-weighted image shows extra-axial enhanced lesions involving the calvarium and under-
should raise the possibility of plasmocytoma [124, 125] (Fig. 10.41). Rosai-Dorfman disease should be considered when a young adult with painless neck lymphadenopathy, fever, and anemia presents with a dural mass [126] (Fig. 10.42). Inflammatory lesions of leptomeninges such as tuberculosis, fungal infection, and syphilis may result in an extra-axial mass. The already known history of systemic sarcoidosis in a patient with a dural mass and leptomeningeal enhancement is suggestive of neurosarcoidosis [127]. Awareness that these lesions involve the dura may facilitate radiological and intra-operative recognition and, in some cases, preclude unnecessary additional surgery. Cavernous sinus meningiomas may affect cranial nerves III, IV, VI and the first and second division of cranial nerve V. Meningioma in such location must be differentiated by Tolosa-Hund syndrome, which may present a with the same clinical and radiological characteristics [128]. The differential diagnosis between diaphragma sellae meningioma and pituitary macroadenoma is difficult based on imaging findings. Marked and homogeneous contrast enhancement, pituitary gland displaced against
A. Drevelegas et al.
b
lying dura. (b) MR examination 4 months later shows marked enlargement of the left frontal mass with extracranial extension
Fig. 10.41 Plasmocytoma in a 62-year-old patient. Axial postcontrast T1-weighted image shows intense homogeneous enhancement of the lesion with a dural tail (arrow)
289
10 Meningeal Tumors
similar to those of gray matter and with dense contrast enhancement in a middle-aged patient favors the diagnosis of meningioma. One morphologic feature in differentiating meningiomas from choroid plexus papillomas is the magnitude of hydrocephalus commonly associated with the papilloma, which is out of proportion with the size of the mass itself. Alternatively, most meningiomas cause enlargement of the involved temporal horn owing to obstruction of CSF outflow, while the remainder of the ventricular system remains normal in size [11].
10.2 Hemangiopericytoma
Fig. 10.42 Rosai-Dorfman disease. Postcontrast CT shows a convex, homogeneously enhanced extra-axial mass
the floor of the sella and thickening or sclerosis of the adjacent osseous structures favors the diagnosis of meningiomas. Clinical and biochemical evidence of hyperfunctioning pituitary gland may also distinguish adenoma from meningioma [29]. CPA meningioma must be differentiated mainly from acoustic neuroma. An extra-axial mass in CPA, broadedbased against the dura, with obtuse angles, eccentric to the porus, extending through the tentorium to the middle cranial fossa and with signal intensity equal to or less than that of gray matter on T2W images is most likely to be meningioma [129]. Another feature that must be taken into account in differential diagnosis between CPA meningioma and acoustic neuroma is the ADC values of tumors. ADC of schwannomas is usually significantly higher than that of meningiomas. Histologically, schwannomas comprise Antoni type A and type B neurinomas, and their higher ADC may reflect the lower cell density of Antoni type B neurinomas. Some meningioma subtypes, however, such as microcystic meningiomas, may have higher ADC values than those of schwannomas [66]. As most intraventricular meningiomas are found in the atria of lateral ventricles, the differential diagnosis includes choroid plexus tumors, ependymoma, and rarely intraventricular astrocytoma. These tumors are most commonly seen in the first 10 years of life. An atrium mass with T1 and T2 signal characteristics that are
Hemangiopericytomas are tumors arising from pericytes that are modified smooth muscle cells, which surround capillaries [130, 131]. They are originating in the meninges and thought to be a variant of angioblastic meningioma [132]. The 1993 WHO classification has eliminated the term angioblastic meningioma in favor of hemangiopericytoma [133]. They are lowgrade malignant tumors (WHO grade II) with a tendency to bleed on resection. Hemangiopericytomas represent 0.4–1.0% of all intracranial tumors and their peak occurrence is between 37 and 44 years of age [134–136]. The location of hemangiopericytomas is similar to that of meningiomas. Hemangiopericytomas have a different biologic behavior from meningiomas. They are aggressive tumors, show local recurrence and metastasize extracranially particularly to bone, lung, liver, kidney, and adrenals [137, 138]. Pathology: Macroscopically, hemangiopericytomas are well-defined, usually lobulated, dural-based tumors, similar to meningiomas. They are markedly hypervascular and more heterogeneous than meningiomas [132, 133]. Microscopically, they are highly cellular and vascular tumors. They are composed of angular pericytes that surround often ill-defined capillaries in a branching pattern (staghorn vascularity) [132, 139]. Imaging: On CT, hemangiopericytomas appear as iso- or hyperdense, well-defined nodular tumors. They are not associated with calcifications or hyperostosis. Following the administration of contrast medium they show strong heterogeneous enhancement [133, 140]. On MRI, hemangiopericytomas are heterogeneous, predominantly isointense on T1-weighted and slightly hyperintense on T2-weighted images with internal flow voids. Strong heterogeneous enhancement occurs
290
A. Drevelegas et al.
a
b
c
d
Fig. 10.43 Hemangiopericytoma. (a) Axial T1W1 shows a multilobulated heterogeneous mass with internal flow voids (arrowheads). (b) Axial T2WI shows an extra-axial mass with internal hyperintensities. (c) Coronal contrast enhanced T1WI
shows strong heterogeneous enhancement of the mass. (d) Left carotid arteriogram shows multiple irregular vessels with intense tumor blush
291
10 Meningeal Tumors
a
b
c
Fig. 10.44 Axial (a) and coronal T1-weighted (b) images show an extra-axial isointense mass with internal flow voids (arrowheads) and calvarial destruction. On axial T2WI (c) the mass remains isointense with internal areas of hyperintensity (arrows)
after the administration of contrast medium (Fig. 10.43). They are lobulated, dural-based extra axial tumors and show white matter “buckling” [141]. Approximately, one-third of hemangiopericytomas show a narrow base
of dural attachment, with the remaining two-thirds showing broad-based attachment with dural tail sign [78, 133]. They may also show local recurrence and bone erosion (Fig. 10.44).
292
10.3 Hemangioblastoma Hemangioblastoma is a benign (WHO grade I) vascular neoplasm that constitutes about 1% of all intracranial tumors and approximately, 7% of posterior fossa tumors in adults. It appears most commonly (80%) as a solitary lesion or in association with von Hippel-Lindau (VHL) disease and represents the most common intra-axial tumor of the posterior fossa in adults after metastasis [12]. Approximately, 25% of hemangioblastomas are associated with VHL disease [142]. VHL disease is a hereditary disorder with an autosomal dominant mode of transmission, which includes multiple hemangioblastomas, similar retinal tumors, pancreatic or renal cysts, renal carcinoma, and pheochromocytoma. The gene causing VHL has been linked to a defect on chromosome 3 [143, 144]. Hemangioblastomas occur most commonly in adults ranging in age from 30 to 65 years; however, they may affect all ages ranging from 1 to 75 years. VHL-associated tumors occur in younger patients, with a mean age of 29 [145]. These tumors are usually located in the cerebellum (83–86%). Other sites include spinal cord (3–13%), medulla (2–5%), and cerebrum (1.5%) [146, 147]. Supratentorial hemangioblastomas are rare. The most common clinical symptom is headache, which is usually occipital in location. Other symptoms include disequilibrium, nausea/vomiting and dizziness/ vertigo [146, 148]. Hemangioblastoma can also express erythropoietin (Epo), which might be of importance for tumor development and progression [149]. Polycythemia, due to erythropoietin secretion, is seen in approximately 20% of patients with hemangioblastomas. It usually resolves following resection of the tumor but may return with tumor recurrence [147, 150]. The prognosis for cerebellar hemangioblastomas is quite good, with a median survival rate 5–20 years following surgical resection of the tumor [150]. Pathology: Hemangioblastomas are benign, welldemarcated tumors. They are usually cystic with a mural nodule (60%), but may also be solid (40%) [132, 146, 151] (Fig. 10.45). They are characterized by the presence of a network of capillary-like channels, separated by trabeculae or islands of “stromal cells.” These cells, which represent the neoplastic component of the tumor, are large, with pale or vacuolated cytoplasm (“clear cells”) that contains in varying numbers lipid droplets
A. Drevelegas et al.
(Fig. 10.46). Their nuclei are usually small and uniform, centrally placed. Atypical or hyperchromatic nuclei are occasionally present. Multinucleated cells may be found as well. Mitoses are uncommon or absent. Although necrosis and calcifications are not common findings, cystic changes and hemorrhages are not infrequent [12]. “Reticular” and “cellular” variants may be recognized, occasionally in the same tumor, when the capillary/ blood vessels network or the stromal cells in compact groups, respectively, predominate [152].
Fig. 10.45 Gross specimen of a resected cystic hemangioblastoma with a small mural nodule (arrow)
Fig. 10.46 Stromal cells with lipid droplets (HematoxylinEosin, original magnification ×400)
293
10 Meningeal Tumors
The stromal cells are immunoreactive for vimentin and neuron-specific enolase; GFAP and S100-protein positivity have been found in some cells as well [153, 154]. Imaging: On unenhanced CT, the hemangioblastoma usually appears as a thin-walled, well-marginated cystic lesion with a mural nodule. The cystic lesion appears hypodense, while the mural nodule is isodense with brain parenchyma. The mural nodule frequently abuts the pial surface. Following the administration of contrast medium, the nodule shows strong, homogeneous
a
c
Fig. 10.47 Typical appearance of cerebellar hemangioblastoma. (a) Unenhanced CT shows a hypodense cystic lesion with an isodense mural nodule (arrow). (b) Enhanced CT shows marked enhancement of the mural nodule (arrow). (c) T1WI shows a cystic lesion, which is slightly hyperintense compared
enhancement [155–157] (Fig. 10.47a, b). Approximately, 30–40% of hemangioblastomas are solid tumors, which appear hyperdense on unenhanced CT and show strong, homogeneous enhancement [158, 159]. MRI is the modality of choice in screening of hemangioblastomas because of the absence of beam-hardening effects and its multiplanar capability. On MRI, the cystic component of hemangioblastoma is either iso- or slightly hyperintense relative to CSF on T1WI and hyperintense on T2WI. This is due to the high protein content of the cyst. The solid component of the tumor
b
d
to CSF of IV ventricle, and an isointense mural nodule (arrow). (d) On coronal T2WI the cyst shows prominent high signal intensity, while the mural nodule is slightly hyperintense (arrow). (e) Axial and coronal postcontrast T1WI shows strong enhancement of the mural nodule
294
A. Drevelegas et al.
e
Fig. 10.47 (continued)
is hypo- or isointense on T1WI, slightly hyperintense on T2WI, and shows marked enhancement after the administration of contrast medium [160–162] (Fig. 10.47c–e). Spontaneous hemorrhage within the tumor may be the cause for the heterogeneous appearance of hemangioblastoma on T1WI. Sometimes, the tumor may appear completely cystic. The cystic component of the tumor does not usually enhance, but if it is lined by neoplasm the wall will enhance [146, 157]. Flow-voids within and at the periphery of the tumor represent abnormal tumor vessels (Fig. 10.48). As seen on enhanced CT, the solid pattern of the tumor shows marked enhancement on T1WI (Fig. 10.49). In patients
with VHL, the appearance of hemangioblastoma is that of a cerebellar cyst with or without enhancing wall or mural nodule (Fig. 10.50). On diffusion-weighted images, the cystic portion of the tumor appears to be hypointense, reflecting increased diffusion properties of the cyst content (Fig. 10.51). The differential diagnosis includes pilocytic astrocytoma, cystic metastasis, and meningioma in case of solid hemangioblastoma attached to the dura [158, 163]. The pilocytic astrocytoma occurs in younger patients than do hemangioblastomas while the wall in cystic metastasis, in contrast to hemangioblastomas, enhances after intravenous contrast administration.
295
10 Meningeal Tumors
a
b
c
Fig. 10.48 Cystic pattern of hemangioblastoma. (a) Axial T1WI shows a cerebellar cyst, which is slightly hyperintense, compared to CSF. (b) Axial T2WI shows high signal intensity of
the cyst. Note the flow-voids at the periphery of the cyst representing afferent and efferent tumor vessels. (c) Coronal T1WI shows ring-like enhancement of the wall of the cyst
296
A. Drevelegas et al.
a
b
Fig. 10.49 Solid cerebellar hemangioblastoma. Axial postcontrast T1WI shows a strong homogeneously enhanced tumor. Note the subpial location of the tumor (arrow)
c
Fig. 10.50 Hemangioblastoma in a patient with von HippelLindau (VHL). (a) Coronal T1WI shows a cystic lesion with enhancing wall and mural nodule in right cerebellum hemisphere. Note a second lesion (arrow) involving the medulla. (b) Abdominal CT shows multiple cystic lesions in the pancreatic body. (c) A solid lesion (arrows) in the pancreatic head represents a proved neuroendocrine tumor. Note additional renal cysts (arrowheads)
297
10 Meningeal Tumors
a
b
c
d
Fig. 10.51 (a) Axial T2WI shows a hyperintense cerebellar lesion with surrounding edema. (b) Axial postcontrast T1WI shows a ring-like enhancement of the cystic lesion. (c) On the diffusion-weighted image the cystic lesion is hypointense.
(d) On the ADC image the central cystic lesion exhibits the same high signal intensity with CSF, whereas the peritumoral edema is less hyperintense
298
References 1. Buetow M, Buetow P, Smirniotopoulos J (1991) Typical, atypical and misleading features in meningioma. Radiographics 11:1087–1106 2. Arbelaez A, Castillo M, Armao D (1999) Meningioma presenting as an acute subdural hematoma. Emerg Radiol 6: 149–152 3. Shuangshoti S (1991) Primary meningiomas outside the central nervous system. In: Al-Meft O (ed) Meningiomas. Raven Press, New York, pp 107–128 4. Hayashi Y, Hamada Y, Oki H, Yamashita J (1997) Pituitary stalk meningioma: case report. Neuroradiology 39:351–353 5. Cushing H, Eisenhardt L (1969) Meningiomas; their classification, regional, life history and surgical results. Hafner, New York, pp 133–168 6. Michel RG, Woodard BH (1979) Extracranial meningioma. Ann Otol 88:407–412 7. Castillo M (1998) Neuroradiology companion, 2nd edn. Lippincott, Philadelphia 8. Osborn AG (1994) Diagnostic neuroradiology. Mosby, St Louis 9. Amirjamshidi A, Mehrazin M, Abbassioun K (2000) Meningiomas of the central nervous system occurring below the age of 17: report of 24 cases not associated with neurofibromatosis and review of literature. Childs Nerv Syst 16: 406–416 10. Rohringer M, Sutherland GR, Louw DF, Sima AAF (1989) Incidence and clinicopathological features of meningioma. J Neurosurg 71:665–672 11. Sanders WP, Chundi VV (2000) Extra-axial tumors including pituitary and parasellar. In: Orrison WW (ed) Neuroimaging. WB Saunders, Philadelphia, pp 612–717 12. Russel DS, Rubinstein LJ (1989) Pathology of tumors of the nervous system, 5th edn. William and Wilkins, Baltimore, pp 449–483 13. Wood MW, White R, Kernohan J (1957) One hundred meningiomas found incidentally at necropsy. J Neuropathol Exp Neurol 16:337–340 14. Russel EJ, George AE, Kricheff II, Budzilovich G (1980) Atypical computed tomographic features of intracranial meningioma: radiological – pathological correlation in a series of 131 consecutive cases. Radiology 135:673–682 15. Doty JR, Schut L, Bruce DA, Sutton LN (1987) Intracranial meningiomas of childhood. Prog Exp Tumor Res 30:24–254 16. Drake JM, Hendrick EB, Becker LE, Chuang SH, Hoffman HJ, Humphreys RP (1986) Intracranial meningiomas in children. Pediatr Neurosci 12:134–139 17. Ferrante L, Acqui M, Mastronardi L, Rocchi G, Fortuna A (1989) Cerebral meningiomas in children. Childs Nerv Syst 5:83–86 18. Germano IM, Edwards MSB, Davia RL, Schiffer D (1994) Intracranial meningiomas of the first two decades of life. J Neurosurg 80:447–453 19. Kolluri VRS, Reddy DR, Reddy PK, Naidu MRC, Rao SBP, Sumethi C (1987) Meningiomas in children Childs Nerv Syst 3:271–273 20. Daffner R, Yakulis R, Maroon J (1998) Intraosseous meningioma. Skeletal Radiol 27:108–111 21. Bradac GB, Ferszt R, Kendall BE (1990) Cranial meningiomas. Springer, Berlin, pp 1–128
A. Drevelegas et al. 22. Zimmerman RD, Fleming CA, Saint-Louis LA, Lee BCP, Manning JJ, Deck MDF (1985) Magnetic resonance imaging of meningiomas. AJNR Am J Neuroradiol 6(2):149–157 23. New PFJ, Aronow S, Hesselink JR (1980) National Cancer Institute study: evaluation of computed tomography in the diagnosis of intracranial neoplasms IV. Meningiomas. Radiology 136:665–675 24. Van Tassel P, Lee Y-Y, Ayala A, Carrasco CH, Klima T (1991) Case report 680: intraosseous meningioma of the sphenoid bone. Skeletal Radiol 20:383–386 25. Erdincler P, Lena G, Sarioglou AC, Kuday C (1997) Intracranial meningiomas in children: review of 29 cases. Surg Neurol 49:136–141 26. Ijiri R, Tanaka Y, Hara M, Sekido K (2000) Radiation associated xanthomatous meningioma in a child. Childs Nerv Syst 16:304–308 27. Liebig T, Hoffmann T, Hosten N, Sander B, van Landeghem F, Stoltenburg-Didinger G, Lanksch WR (1998) Lipomatous secretory meningioma: case report and review of the literature. Neuroradiology 40:656–658 28. Kuroiwa T, Ohta T (2000) MRI appearances mimicking the dural tail sign: a report of two cases. Neuroradiology 42: 199–202 29. Cappabianca P, Cirillo S, Alfieri A, D’ Amico A, Maiuri F, Mariniello G, Caranci F, De Divitiis E (1999) Pituitary macroadenoma and diaphragma sellae meningioma: differential diagnosis on MRI. Neuroradiology 41:22–26 30. DeMonte F, Smith HK, Al-Mefty O (1994) Outcome of aggressive removal of the cavernous sinus meningiomas. J Neurosurg 81:245–251 31. Sen C, Hague K (1997) Meningiomas involving the cavernous sinus: histological factors affecting the degree of resection. J Neurosurg 87:535–543 32. Kim KS, Rogers LF, Goldblatt D (1987) CT features of hyperostosing meningioma en plaque. AJNR Am J Neuroradiol 8:853–859 33. Okamoto S, Hisaoka M, Aoki T, Kadoya C, Kobanawa S, Hashimoto H (2000) Intraosseous microcystic meningioma. Skeletal Radiol 29:354–357 34. Choux M, Lena G, Genitory L (1991) Meningioma in children. In: Schmideck HH (ed) Meningiomas and their surgical management. Saunders, Philadelphia, pp 93–102 35. Modesti LM, Binet EF, Collins GH (1976) Meningiomas causing spontaneous intracranial hematomas. J Neurosurg 45:437–444 36. Kohli CM, Crouch RL (1984) Meningioma with intracerebral hematoma. Neurosurgery 14:237–240 37. Lazaro RP, Messer HD, Brinker RA (1981) Intracerebral hemorrhage associated with meningioma. Neurosurgery 8:96–101 38. Hutzelmann A, Palmie S, Buhl R, Freund M, Heller M (1998) Dural invasion of meningiomas adjacent to the tumor margin on Gd-DTPA-enhanced MR images: histopathologic correlation. Eur Radiol 8:746–748 39. Chan BC, Thomson GB (1984) Intracranial meningiomas in childhood. Surg Neurol 21:319–322 40. Ghim TT, Seo JJ, O’Brien M, Meacham L, Crocker I, Krawiecki N (1993) Childhood intracranial meningiomas after high dose irradiation. Cancer 71:4091–4095 41. Goldsmoth BJ, Wara WM, Wilson CB, Larson DA (1994) Postoperative irradiation for subtotally resected meningiomas. A retrospective analysis of 140 patients treated from 1967 to 1990. J Neurosurg 80:195–201
10 Meningeal Tumors 42. Ng HK, Tse CC, Lo ST (1989) Microcystic meningiomas-an unusual morphological variant of meningiomas. Histopathology 14:1–9 43. Alguacil-Garcia A, Pettigrew NM, Sima AAF (1986) Secretory meningioma. A distinct subtype of meningioma. Am J Surg Path 10:102–111 44. Horten BC, Urich H, Stefoski D (1979) Meningiomas with conspicuous plasma cell-lymphocytic components. A report of five cases. Cancer 13:1353–1364 45. Louis DN, Scheithauer BW, Budka H et al (2000) In: Kleihues P, Cavenne WK (eds) Pathology and genetics. Tumours of the nervous system. World Health Organization classification of tumours. IARC Press, Lyon 46. Kepes JJ, Chen WY, Connors MH et al (1988) “Chordoid” meningeal tumors in young individuals with peritumoral lymphoplasmacellular infiltrate causing systemic manifestations of the Castleman syndrome. A report of seven cases. Cancer 62:391–406 47. Pasquier B, Gasnier F, Pasquier D et al (1986) Papillary meningioma. Clinicopathologic study of seven cases and review of the literature. Cancer 58:299–305 48. Perry A, Scheithauer BW, Stafford SL et al (1998) “Rhabdoid” meningioma: an aggressive variant. Am J Surg Pathol 22: 1482–1490 49. Claveria LE, Sutton D, Tress BM (1977) The radiological diagnosis of meningiomas, the impact of EMI scanning. Br J Radiol 50:15–22 50. Vassilouthis J, Ambrose J (1979) Computerized tomography scanning appearance of intracranial meningiomas. J Neurosurg 50:320–327 51. Kepes J (1975) Observation on the formation of psammoma bodies and pseudopsammoma bodies in meningioma. J Neuropathol Exp Neurol 20:255 52. Smirniotopoulos J, Lee HS (1992) Primary tumors in adults. In: Lee HS, Rao KCVG, Zimmerman RA (eds) Cranial MRI, CT. McGraw-Hill, New York, pp 333–345 53. Kim KS, Rogers LF, Lee C (1983) The dural lucent line: characteristic sign of hyperostosing meningioma en plaque. AJR Am J Roentgenol 141(6):1217–1221 54. Sutton L, Claveria L (1977) Meningiomas diagnosed by scanning: a review of 100 intracranial cases. In: DuBoulay G, Moseley I (eds) The first European seminar on computerized axial tomography in clinical practice. Springer, Heidelberg, p 102 55. Luis DN, Budka H, Von Deimling A (1997) Meningiomas. In: Kleihues P, Cavenee WK (eds) Pathology and genetics of tumors of the nervous system. International Agency for Research on Cancer, Lyon, pp 134–141 56. Hasso AN, Bell SA, Tadmor R (1994) Intracranial vascular tumors. Neuroimaging Clin N Am 4:849–870 57. Christoforidis GA (2008) Meningeal tumors. In: Newton HB, Jolesz FA (eds) Handbook of neuro-oncology neuroimaging. Academic, New York 58. Chakeres DW, Curtin A, Ford G (1989) Magnetic resonance imaging of pituitary and parasellar abnormalities. Radiol Clin North Am 27(2):265–281 59. Spagnoli MV, Goldberg HI, Grossman RI, Bilaniuk LT, Gomori JM, Hackney DB, Zimmerman RA (1986) Intracranial meningiomas: high-field MR imaging. Radiology 61(2): 369–375 60. Komiyama M, Yagura H, Baba M, Yasui T, Hakuba A, Nishimura S, Inoue Y (1987) MR imaging: possibility of
299 tissue characterization of brain tumors using T1 and T2 values. AJNR Am J Neuroradiol 8(1):65–70 61. Fujii K, Fujita N, Hirabuki N, Hashimoto T, Miura T, Kozuka T (1992) Neuromas and meningiomas: evaluation of early enhancement with dynamic MR imaging. AJNR Am J Neuroradiol 13(4):1215–1220 62. Maiuri F, Iaconetta G, de Divitiis O, Cirillo S, Di Salle F, De Caro ML (1999) Intracranial meningiomas: correlations between MR imaging and histology. Eur J Radiol 31(1):69–75 63. Elster AD, Challa VR, Gilbert TH, Richardson DN, Contento JC (1989) Meningiomas: MR and histopathologic features. Radiology 170(3 Pt 1):857–862 64. Filippi CG, Edgar MA, Ulug AM, Prowda JC, Heier LA, Zimmerman RD (2001) Appearance of meningiomas on diffusion-weighted images: correlating diffusion constants with histopathologic findings. AJNR Am J Neuroradiol 22:65–72 65. Nagar VA, Ye JR, Ng WH, Chan YH, Hui F, Lee CK, Lim CCT (2008) Diffusion-weighted MR imaging: diagnosing atypical or malignant meningiomas and detecting tumor dedifferentiation. AJNR Am J Neuroradiol 29:1147–1152 66. Yamasaki F, Kurisu K, Satoh K, Arita K, Sugiyama K, Ohtaki M, Takaba J, Tominaga A, Hanaya R, Yoshioka H, Hama S, Ito Y, Kajiwara Y, Yahara K, Saito T, Thohar MA (2005) Apparent diffusion coefficient of human brain tumors at MR imaging. Radiology 235:985–991 67. Toh CH, Castillo M, Wong AM, Wei KC, Wong HF, Ng SH, Wan YL (2008) Differentiation between classic and atypical meningiomas with use of diffusion tensor imaging. AJNR Am J Neuroradiol 29:1630–1635 68. Buhl R, Nabavi A, Wolff S, Hugo HH, Alfke K, Jansen O, Mehdorn HM (2007) MR spectroscopy in patients with intracranial meningiomas. Neurol Res 29(1):43–46 69. Qi ZG, Li YX, Wang Y, Geng DY, Li KC, Shen TZ, Chen XR (2008) Lipid signal in evaluation of intracranial meningiomas. Chin Med J 121(23):2415–2419 70. Zhang H, Rödiger LA, Shen T, Miao J, Oudkerk M (2008) Preoperative subtyping of meningiomas by perfusion MR imaging. Neuroradiology 50(10):835–840 71. Zhang H, Rödiger LA, Shen T, Miao J, Oudkerk M (2008) Perfusion MR imaging for differentiation of benign and malignant meningiomas. Neuroradiology 50(6):525–530 72. George AE, Russel EJ, Kricheff II (1980) White mutter buckling: CT sign of extra-axial intracranial mass. AJNR Am J Neuroradiol 1:425–430 73. Gupta S, Gupta RK, Banerjee D, Gujral RB (1993) Problems with the “dural tail” sign. Neuroradiology 35(7):541–542 74. Wilms G, Lammens M, Marchal G, Demaerel P, Verplancke J, Van Calenbergh F, Goffin J, Plets C, Baert AL (1991) Prominent dural enhancement adjacent to nonmeningiomatous malignant lesions on contrast-enhanced MR images. AJNR Am J Neuroradiol 12(4):761–764 75. Senegor M (1991) Prominent meningeal “tail sign” in a patient with a metastatic tumor. Neurosurgery 29(2):294–296 76. Kutcher TJ, Brown DC, Maurer PK, Ghaed VN (1991) Dural tail adjacent to acoustic neuroma: MR features. J Comput Assist Tomogr 15(4):669–670 77. Lunardi P, Mastronardi L, Nardacci B, Acqui M, Fortuna A (1993) “Dural tail” adjacent to acoustic neuroma on MRI: a case report. Neuroradiology 35(4):270–271 78. Tien RD, Yang PJ, Chu PK (1991) “Dural tail sign”: a specific MR sign for meningioma? J Comput Assist Tomogr 15(1):64–66
300 79. Bourekas EC, Wildenhain P, Lewin JS, Tarr RW, Dastur KJ, Raji MR, Lanzieri CF (1995) The dural tail sign revisited. AJNR Am J Neuroradiol 16(7):1514–1516 80. Morioka T, Matsushima T, Ikezaki K, Nagata S, Ohta M, Hasuo K, Fukui M (1993) Intracranial adenoid cystic carcinoma mimicking meningioma: report of two cases. Neuroradiology 35(6):462–465 81. Good CD, Kingsley DP, Taylor WJ, Harkness WF (1997) “Dural tail” adjacent to a giant posterior cerebral artery aneurysm: case report and review of the literature. Neuroradiology 39(8):577–580 82. Goldsher D, Litt AW, Pinto RS, Bannon KR, Kricheff II (1990) Dural “tail” associated with meningiomas on Gd-DTPA-enhanced MR images: characteristics, differential diagnostic value, and possible implications for treatment. Radiology 176(2):447–450 83. Tokumaru A, O’uchi T, Eguchi T, Kawamoto S, Kokubo T, Suzuki M, Kameda T (1990) Prominent meningeal enhancement adjacent to meningioma on Gd-DTPA-enhanced MR images: histopathologic correlation. Radiology 175(2):431–433 84. Wilms G, Lammens M, Marchal G, Van Calenbergh F, Plets C, Van Fraeyenhoven L, Baert AL (1989) Thickening of dura surrounding meningiomas: MR features. J Comput Assist Tomogr 13(5):763–768 85. Bradac GB, Ferszt R, Bender A, Schorner W (1986) Peritumoral edema in meningiomas. A radiological and histological study. Neuroradiology 28(4):304–312 86. Sigel RM, Messina AV (1976) Computed tomography: the anatomic basis of the zone of diminished density surrounding meningiomas. AJR Am J Roentgenol 127:139–141 87. Philipon J, Foncin JF, Grob R, Scour A, Poisson M, Pertuiset BF (1984) Cerebral edema associated with meningiomas: possible role of a secretory-excretory phenomenon. Neurosurgery 14: 295–301 88. Go KG, Kamman RL, Wilmink JT, Mooyaart EL (1994) A study on peritumoral brain edema around meningiomas by MRI and contrast CT. Acta Neurochir Suppl (Wien) 60:365–368 89. Bitzer M, Nagele T, Geist-Barth B, Klose U, Gronewaller E, Morgalla M, Heiss E, Voigt K (2000) Role of hydrodynamic processes in the pathogenesis of peritumoral brain edema in meningiomas. J Neurosurg 93(4):594–604 90. Ide M, Jimbo M, Kubo O, Yamamoto M, Takeyama E, Imanaga H (1994) Peritumoral brain edema and cortical damage by meningioma. Acta Neurochir 160:369–372 91. Tatagiba M, Mirzai S, Samii M (1991) Peritumoral blood flow in intracranial meningiomas. Neurosurgery 28(3):400–404 92. Bitzer M, Opitz H, Popp J, Morgalla M, Gruber A, Heiss E, Voigt K (1998) Angiogenesis and brain oedema in intracranial meningiomas: influence of vascular endothelial growth factor. Acta Neurochir (Wien) 140(4):333–340 93. Provias J, Claffey K, Lau N, Feldkamp M, Guha A (1997) Meningiomas: role of vascular endothelial growth factor/ vascular permeability factor in angiogenesis and peritumoral edema. Neurosurgery 40(5):1016–1026 94. Samoto K, Ikezaki K, Ono M, Shono T, Kohno K, Kuwano M, Fukui M (1995) Expression of vascular endothelial growth factor and its possible relation with neovascularization in human brain tumors. Cancer Res 55(5): 1189–1193 95. Wasenko JJ, Hochhauser L, Stopa EG, Winfield JA (1994) Cystic meningiomas: MR characteristics and surgical correlations. AJNR Am J Neuroradiol 15(10):1959–1965
A. Drevelegas et al. 96. Lohle PN, Wurzer HA, Seelen PJ, Kingma LM, Go KG (1999) Cystic lesions accompanying extra-axial tumours. Neuroradiology 41(1):13–17 97. Worthington C, Caron JL, Melanson D, Leblanc R (1985) Meningioma cysts. Neurology 35(12):1720–1724 98. Shimoji K, Yasuma Y, Mori K, Eguchi M, Maeda M (1999) Unique radiological appearance of a microcystic meningioma. Acta Neurochir (Wien) 141(10):1119–1121 99. Kepes JJ (1982) Meningiomas: biology, pathology, and differential diagnosis. New York, Masson, pp 75–109 100. LeRoux P, Hope A, Lofton S, Harris AB (1989) Lipomatous meningioma – an uncommon tumor with distinct radiographic findings. Surg Neurol 32(5):360–365 101. Kim DG, Park CK, Paek SH, Choe GY, Gwak HS, Yoo H, Jung HW (2000) Meningioma manifesting intracerebral haemorrhage: a possible mechanism of haemorrhage. Acta Neurochir 142(2):165–168 102. Chaskis C, Raftopoulos C, Noterman J, Flament-Durand J, Brotchi J (1992) Meningioma associated with subdural haematoma: report of two cases and review of the literature. Clin Neurol Neurosurg 94(3):269–274 103. Hell TL, Conley FK (1980) Haemorrhage associated with meningioma: a case report and review of the literature. J Neurosurg Psychiatry 43:725–729 104. Kendall B, Symon L (1997) Investigation of patients presenting with cerebellopontine angle syndromes. Neuroradiology 13:65–84 105. Mikhael MA, Ciric IS, Wolff AP (1985) Differentiation of cerebellopontine angle neuromas and meningiomas with MR imaging. J Comput Assist Tomogr 9:852–856 106. Jelinek J, Smirniotopoulos JG, Parisi JE, Kanzer M (1990) Lateral ventricular neoplasms of the brain: differential diagnosis based on clinical CT and MR findings. AJNR Am J Neuroradiol 11:567–574 107. Mani RL, Hedgcock MW, Mass SI, Gilmor RL, Enzmann DR, Eisenberg RL (1978) Radiographic diagnosis of meningiomas of the lateral ventricle: review of 22 cases. J Neurosurg 49:249–255 108. Cushing H, Eisenhardt L (1938) Meningiomas without dural attachment. In: Charles CT (ed) Meningiomas. Charles C. Thomas, Springfield, pp 133–168 109. Matsumoto S, Yamamoto T, Ban S, Sato S, Shingu T, Yoshida S, Tokuno T, Nakazawa K, Saiwai S, Shirane H (1995) A case of deep sylvian meningioma presenting temporal lobe epilepsy. No To Shinkei 47:503–508 110. Wada T, Suzuki M, Beppu T, Arai H, Yoshida Y, Ogawa A, Sasou A (2000) A case of subcortical meningioma. Acta Neurochir 142:209–213 111. Cho BK, Wang KC, Chang KH, Chi JG (1990) Deep sylvian meningioma in a child. Childs Nerv Syst 6:228–230 112. Ito H, Takagi H, Kawano N, Yada K (1992) Primary intraosseous meningioma: a case report. J Neurooncol 13: 57–61 113. Azar-kia B, Sarwar M, Alan Mare J, Schechter MM (1974) Intraosseous meningioma. Neuroradiology 6:246–253 114. Kuali A, Ilcayto R, Rahmanli O (1991) Primary calvarial ectopic meningiomas. Neurochirurgia 34:173–177 115. Halpin SFS, Britton J, Wilkins P, Uttley D (1991) Intradiploic meningioma: a radiological study of two cases confirmed histologically. Neuroradiology 33:247–250 116. Khurshid A, Joseph JT, Rachlin J, Cooley TP, Kleefield J, Dezube BJ (1999) Meningioma in four patients with human
10 Meningeal Tumors immunodeficiency virus infection. Mayo Clin Proc 4(3): 253–257 117. Antinheimo J, Sankila R, Carpen O, Pukkala E, Sainio M, Jaaskelainen J (2000) Population-based analysis of sporadic and type 2 neurofibromatosis-associated meningiomas and schwanomas. Neurology 54:71 118. Akiyama M, Sakai H, Onoue H, Miyazaki Y, Abe T (2004) Imaging intracranial haemangiopericytomas: study of seven cases. Neuroradiology 46(3):194–197 119. Lerdlum S, Lalitanantpong S, Numkarunarunrote N, Chao wanapanja P, Suankratay C, Shuangshoti S (2004) MR imaging of CNS leiomyosarcoma in AIDS patients. J Med Assoc Thai 87(Suppl 2):S152–S160 120. Maroldi R, Ambrosi C, Farina D (2005) Metastatic disease of the brain: extraaxial metastases (skull, dura, leptomeningeal) and tumour spread. Eur Radiol 15(3):617–626 121. Thurner MM, Rieger A, Popov CK, Schindler E (2001) Malignant lymphoma of the cranial vault in a HIV-positive patient: imaging findings. Eur Radiol 11:1506–1509 122. Singh S, Cherian RS, George B, Nair S, Srivastava A (2000) Unusual extra-axial central nervous system involvement of non-Hodgkin’s lymphoma: magnetic resonance imaging. Australas Radiol 44(1):112–114 123. Rodriguez LE, Rodriguez CY, Cardozo DP, Pena JA, Molina OM, Cardozo JJ (2000) The classical clinical and neuroimaging features of meningiomas are mimicked by other intracranial, supratentorial expansive lesions. Rev Neurol 30(10):907–910 124. Vaicys C, Schulder M, Wolansky LJ, Fromowitz FB (1999) Falcotentorial plasmacytoma. Case report. J Neurosurg 91(1):132–135 125. Gallina P, Mascalchi M, Mouchaty H, Buccoliero A, Perrini P (2004) Misleading imaging features of intracranial dural plasmacytoma: report of two cases. Br J Neurosurg 18(6): 643–646 126. Kim M, Provias J, Bernstein M (1995) Rosai-Dorfman disease mimicking multiple meningioma: case report. Neuro surgery 36:1185–1187 127. Sandhu FA, Schellinger D, Martuza RL (2000) A vascular sarcoid mass mimicking a convexity meningioma. Neuroradiology 42:195–198 128. Leijzer CT, Prevo RL, Hageman G (1999) Meningioma presenting as Tolosa-Hunt syndrome. Clin Neurol Neurosurg 101(1):19–22 129. Lo WWM, Solti-Bohman LG (1996) Tumors of the temporal bone and the cerebellopontine angle. In: Som PM, Curtin HD (eds) Head and neck imaging. Mosby, St. Louis, p 1460 130. Stout AP, Murray MR (1942) Hemangiopericytoma: a vascular tumor featuring Zimmenman’s pericytes. Ann Surg 116:26–33 131. Parker DR, Rabinov JD (1991) Recurrent meningeal hemangiopericytoma. AJR Am J Roentgenol 156:10307–11313 132. Burger PC, Scheithauer BW, Vogel FS (1991) Surgical pathology of the nervous system and its coverings, 3rd edn. Churchill-Livingstone, New York 133. Chiechi MV, Smirniotopoulos JG, Mena H (1996) Intracranial hemangiopericytomas: MR and CT features. AJNR Am J Neuroradiol 17:1365–1371 134. Guthrie BL, Ebersold MJ, Scheithauer BW, Shaw EG (1989) Meningeal hemangiopericytoma: histological features, treatment, and long-term follow-up of 44 cases. Neurosurgery 25:514–522
301 135. Goellner JR, Laws ER, Soule EH, Okazaki H (1978) Hemangiopericytoma of the meninges Mayo Clinic experience. Am J Clin Pathol 1978:375–380 136. Osborne DR, Dubois P, Drayer B et al (1981) Primary intracranial meningeal and spinal hemangiopericytoma: radiologic manifestations. AJNR Am J Neuroradiol 2:69–74 137. Mena H, Ribas JL, Pezeshkpour GH et al (1991) Heman giopericytoma of the central nervous system: a review of 94 cases. Hum Pathol 22:84–91 138. Pitkethly DT, Hardman JM, Kempe LG, Earle KM (1970) Angioblastic meningiomas: clinicopathologic study of 81 cases. J Neurosurg 32:539–544 139. Jaaskelainen J, Louis DN, Paulus W et al (1997) Hemangiopericytoma. In: Kleihues P, Canenee WK (eds) Pathology and genetics: tumors of the central nervous system. International Agency for Research on Cancer, Lyon, pp 146–148 140. Servo A, Jaaskelainen J, Wahlsrom T, Haltia M (1985) Diagnosis of intracranial hemangiopericytomas with angiography and CT scanning. Neuroradiology 27:38–43 141. Cosentino CM, poulton TB, Esguerra JV, Sands SF (1993) Giant cranial hemangiopericytoma: MR and angiographic features. AJNR Am J Neuroradiol 14:253–256 142. Neumann HP, Wiestler DD (1994) Von Hippel-Lindau disease: a syndrome providing insights into growth control and tumorigenesis. Nephrol Dial Transplant 9:1832–1833 143. Seizinger BR, Rouleaud GA, Ozelius LJ et al (1988) Von Hippel-Lindau disease maps to the region of chromosome 3 associated with renal cell carcinoma. Nature 332:268–269 144. Hosoe S, Brauch H, Latiff F et al (1990) Localization of the von Hippel-Lindau disease to a small region of chromosome 3. Genomics 8:634–640 145. Maher ER, Yates JR, Ferguson S (1990) Statistical analysis of the two-stage mutation model in Hippel-Lindau disease, and in sporadic cerebellar hemangioblastoma and renal cell carcinoma. J Med Genet 27:311–314 146. Ho VB, Smirniotopoulos JG, Murphy PM et al (1992) Radiologic-pathologic correlation: hemangioblastoma. AJNR Am J Neuroradiol 25:514–522 147. Constans JP, Meder F, Maiuri F et al (1986) Posterior fossa hemangioblastomas. Surg Neurol 23:269–275 148. Neumann HPH, Eggert HR, Weigel K et al (1989) Hemangioblastomas of the nervous system: a 10 year study with special reference to Hippel-Lindau syndrome. J Neurosurg 70:24–30 149. Krieg M, Marti HH, Plate KH (1998) Coexpression of erythropoietin and vascular endothelial growth factor in nervous system tumors associated with von Hippel-Lindau tumor suppressor gene loss of function. Blood 92(9):3388–3393 150. Rubinstein LJ (1972) Tumors of the central nervous system 2nd series. Fassicle 6. Armed Forces Institute of Pathology, Washington, pp 235–241 151. Bohling T, Hatva E, Plate KH et al (1997) Von HippelLindau disease and capillary hemangioblastoma. In: Kleihues P, Cavenee WK (eds) Pathology and genetics: tumors of the central nervous system. International Agency for Research on Cancer, Lyon, pp 179–181 152. Bruner JM, Tien RD, McLendon RE (1998) Tumors of vascular origin. In: Russel DS, Rubinstein LJ (eds) Pathology of tumors of the nervous system. Arnold, London 153. Deck JHN, Rubinstein LJ (1981) Glial fibrillary acidic protein in stromal cells of some capillary hemangioblastomas:
302 significance and possible implications of an immunoperoxidase study. Acta Neuropathol 54:173–181 154. Frank TS, Trojanowski JQ, Roberts SA et al (1989) A detailed immunohistochemical analysis of cerebellar hemangioblastoma. An undifferentiated mesenchymal tumor. Mod Pathol 2:638–651 155. Elster AD, Arthur DW (1988) Intracranial hemangioblastomas: CT and MR findings. J Comput Assist Tomogr 12:736–739 156. Ganti SR, Silver AJ, Hilal SK et al (1982) Computed tomography of cerebellar hemangioblastomas. J Comput Assist Tomogr 6:912–919 157. Filling-Katz MR, Choyke PL, Patronas NJ et al (1989) Radiologic screening for von Hippel-Lindau disease: the role of Gd-DPTA enhanced MR imaging of the CNS. J Comput Assist Tomogr 13:743–755
A. Drevelegas et al. 158. Young S, Ridchardson AE (1987) Solid hemangioblastomas of the posterior fossa: radiologic features and results of surgery. J Neurol Neurosurg Psychiatry 50:155–158 159. Smirniotopoulos JG, Murphy FM, Brown DC (1989) MR imaging of hemangioblastoma. Radiology 173:85 160. Isawa T, Horibe K, Nakatani S et al (1999) Hemangioblastoma of the third ventricle. Neurosurg Rev 22(2–3):140–144 161. Sato Y, Waziri M, Smith W et al (1988) Hippel-Lindau disease; MR imaging. Radiology 166:241–246 162. Lee SR, Sanches J, Mark AJ et al (1989) Posterior fossa of hemangioblastomas: MR imaging. Radiology 117:463–468 163. Ricci PE (1999) Imaging of adult brain tumors. Neuroimaging Clin N Am 9(4):651–669
Central Nervous Lymphomas and Hemopoietic Neoplasms
11
Julia Frühwald-Pallamar, Negar Fakhrai, Majda M. Thurnher, and Antonios Drevelegas
Contents 11.1 Primary Central Nervous System Lymphoma (PCNSL).................................................................. 303 11.1.1 Epidemiology............................................................ 303 11.1.2 Pathology.................................................................. 304 11.1.3 Clinical Features....................................................... 305 11.1.4 Imaging Findings in Immunocompetent Patients.................................. 305 11.1.5 Imaging Findings in Immunocompromised Patients............................. 310 11.2 Leukemia................................................................. 319 11.2.1 Lymphomatous and Leukemic Meningitis (LM)....................................................... 320 References............................................................................ 322
J. Frühwald-Pallamar Department of Radiology, Medical University of Vienna, Vienna, Austria N. Fakhrai Clinical Division of Oncology Department of Medicine I Medical University of Vienna, Vienna, Austria M.M. Thurnher (*) Department of Radiology, Medical University of Vienna, Vienna, Austria e-mail:
[email protected] A. Drevelegas Department of Radiology Aristotle University of Thessaloniki Medical School, Thessaloniki, Greece
11.1 Primary Central Nervous System Lymphoma (PCNSL) Lymphomas are aggressive malignancies that require rapid diagnosis. The first description of what we now recognize as a lymphoma is generally attributed to Thomas Hodgkin, who described it in 1832 [1]. In 1991, an international group of pathologists interested in lymphoma (Inter national Lymphoma Study Group [ILSG]) was formed. At their third meeting, in Berlin in 1993, the group reached a consensus on a long list of lymphoid neoplasms, and a new classification of lymphoid neoplasms, called the Revised European-American Classification of Lymphoid Neoplasms (REAL classification), was established [2]. The World Health Organization (WHO) classification has been updated in the past 2 years, with contributions from more than 130 authors from 22 countries. Lymphomas are now recognized as a heterogeneous group of distinct diseases, most of which are unrelated, and not a single disease with a spectrum of histological grade and clinical behavior. According to the WHO 2008 classification, lymphomas can be divided into four groups: (a) mature B-cell neoplasms, (b) mature T-cell and NK-cell neoplasms, (c) Hodgkin’s lymphoma, and (d) posttransplantation lymphoproliferative disorders (PTLD) [2].
11.1.1 Epidemiology Primary central nervous system lymphoma (PCNSL) is an extranodal form of non-Hodgkin’s lymphoma. Previously known as a rare tumor, it has significantly
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_11, © Springer-Verlag Berlin Heidelberg 2011
303
304
increased in frequency in the past 2 decades. The incidence increased from 2.7 in the 1970s to 7.5 cases per ten million population in the early 1980s [3–7]. The increase is most often attributed to acquired immunodeficiency syndrome (AIDS) and other conditions that cause a compromised immune system. The absolute incidence rate of brain lymphoma in AIDS patients is up to 3,600-fold higher than in the general population [8]. PCNSL has a strong association with Epstein–Barr virus (EBV) [9–11]. In the immunocompetent population, the frequency increases progressively with age, with the highest rates in those over the age of 65 years. Until 1987, PCNSL rates were similar for males and females. After 1987, lower PCNSL rates were observed for females compared to males. As a result of the increased frequency, the prevalence of PCNSL is now equivalent to that of meningioma and low-grade astrocytoma [12, 13]. Secondary central nervous system (CNS) lymphoma occurs less frequently and it is distinguished from its primary counterpart by its propensity to involve the dura mater and the leptomeninges [14].
11.1.2 Pathology Intracranial lymphoma is almost exclusively nonHodgkin’s lymphoma [15]. Hodgkin’s disease rarely involves the brain and, when it does, it is usually late in its course and arises from the dura mater [14]. The vast majority (almost 98%) of PCNSLs are of B-cell origin, with only a small percentage (not more than 2%) being T-cell lymphomas [16, 17]. Plasmacytomas, intravascular (angiocentric) lymphomas and MALT lymphomas, as well as Hodgkin’s disease, are rare as primary entities in the CNS. PCNSL can occur anywhere in the brain, but the most common locations are the periventricular white matter, the corpus callosum, and the basal ganglia [18, 19]. Less commonly, the tumor involves the corticomedullary junction and cerebellum. T-cell lymphomas appear to arise in the cerebellum more frequently. PCNSL has a distinct tendency for perivascular extension. Leptomeningeal or ependymal involvement occurs in about 12% of cases. Dural involvement is rare in patients with PCNSL. Macroscopically, the tumors are brown, gray, or yellow masses with mostly poorly defined margins.
J. Frühwald-Pallamar et al.
Fig. 11.1 Primary cerebral lymphoma. Gross specimen shows a large deeply located lesion extending into the ventricle
Some tumors are well-lineated, (Fig 11.1). In the highgrade lymphomas, the lymphoid neoplastic cells are large, with rounded nuclei and prominent nucleoli. In the uncommonly found low-grade lymphomas, the neoplastic lymphocytes are small or plasmacytoid. These cells invade the brain parenchyma either as compact aggregates or as diffuse infiltrates of single cells (Fig. 11.2). The main pathologic feature of PCNSLs is the angiocentric infiltration, characterized by the presence of perivascular neoplastic cell cuffs, combined with an increase of perivascular reticulin fibers. Invasion of the vessel wall may also be found. Small, reactive lymphocytes, histiocytes, as well as microglial cells and reactive astrocytes, are usually
Fig. 11.2 Primary B-cell CNS Lymphoma Rounded neoplastic cells in a diffuse pattern (Haematoxylin-Eosin, original magnification X400 Inset: Immunohistochemical stain of the same tumor (L-26, original magnification X400)
305
11 Central Nervous Lymphomas and Hemopoietic Neoplasms
admixed with neoplastic lymphoid cells in varying percentages. Necrosis may be recognized in single tumor cells or, more diffusely, in a geographic distribution without a pseudopalisading pattern. Necrotic areas are usually found around a perivascular cuff of neoplastic cells. Necrosis is more commonly found in AIDS patients [9]. Endothelial proliferation is not found in PCNLs. PCNSLs present a diffuse pattern of infiltration and indistinct borders. Neoplastic cells may be found far from the grossly recognized tumor boundaries. With the exception of follicular lymphoma, all other lymphoma subtypes found outside the CNS may be present in the brain as well [11]. The majority of PCNSL express the B-cell markers CD19, 20, and CD79a. Ninety percent of the tumor cells show expression of Ki-67.
11.1.3 Clinical Features The presenting clinical symptoms of PCNSL are variable, depending on the location of the tumor and intracranial pressure. Focal symptoms are present in about 50% of patients and include headache, seizures, personality changes, motor dysfunction, and cerebellar signs. Cranial nerve dysfunction is present in 10–40% [20, 21]. Cerebrospinal fluid (CSF) evaluation shows elevation of protein and mononuclear cells [22]. Patients under 50 years of age have a better outcome compared to older patients [22]. Patients with AIDS have a less favorable course. The overall survival ranges from 3 weeks to 21 months. Age and performance status at time of diagnosis are the most important prognostic factors. Poor prognostic factors include: location in deep brain structures, increased patient age, increased serum lactate dehydrogenase, and increased CSF protein [23]. With an increased number of risk factors, the 2-year survival falls from 80% to 48% to 15% [24]. An attempt at tumor resection is contraindicated because of the diffusely infiltrative nature of the lymphomas, and because surgery may cause neurologic deficits [25]. The use of proper radiation therapy or chemotherapy may prolong the median survival time. However, high-dose methotrexate is the single most active and important agent in the treatment of this disease [26].
11.1.4 Imaging Findings in Immunocompetent Patients 11.1.4.1 Computed Tomography (CT) The radiologic diagnosis of PCNSL is usually based on computed tomography (CT) and magnetic resonance imaging (MRI) findings [27, 28]. On nonenhanced CT (NECT), primary cerebral lymphoma in immunocompetent individuals appears as a round or oval iso- /hyperdense lesion. After the administration of contrast medium, PCNSL shows moderate, uniform, and homogeneous enhancement (Figs. 11.3 and 11.4). Nonenhancing lesions are very rare, and mimic white matter disease, seen primarily in immuncompromised patients.
11.1.4.2 Magnetic Resonance Imaging On T1-weighted MR images, lymphoma is homogeneous and iso- or slightly hypointense relative to the cortex. On T2-weighted images, PCNSL is usually hypo- or isointense to the cortex, and surrounded by mild hyperintense edema (Fig. 11.3, 11.5, 11.7) [27– 31]. T2 shortening is a result of high lymphoma cellularity (high nucleus-to-cytoplasm ratio and low intra-tumoral water content) (Fig. 11.3). Perifocal edema is much less extensive than is seen with primary glial tumors or metastases. The signal intensity on fluid-attenuated inversion-recovery (FLAIR) sequences is iso- to hypointense compared to the cortex (Fig. 11.3). On postcontrast T1WI, lymphoma shows marked homogeneous enhancement (Figs. 11.3, 11.5–11.7). When calcifications are present, low signal areas on T2WI and T2*GRE can be noted. Due to their high cellularity, CNS lymphomas will show high signal on diffusion-weighted MR images (DWI), representing restricted diffusion (Fig. 11.3) [32–34]. In one retrospective study with 11 patients with brain lymphomas (19 lesions) and 17 patients with astrocytomas (19 lesions), the mean ADC ratio of lymphomas was 1.15 and that of high-grade astrocytomas 1.68 [34]. Later studies have also shown that restricted diffusion is a consistent imaging finding in CNS lymphoma in immunocompetent patients [35]. In one recent study, the ADC values of lymphomas were compared with those measured in glioblastomas [36].
306
J. Frühwald-Pallamar et al.
a
b
c
d
Fig. 11.3 A 63-year-old male patient with histologically proven B-cell lymphoma. (a) On nonenhanced CT (NECT) a homogeneous hyperdense mass is demonstrated in the corpus callosum. (b) On axial FLAIR MR image a homogeneous mass is shown in the splenium of the corpus callosum. Note perifocal edema and compression of the ventricles. (c, d) Relatively high signal
is seen on trace DWI (c) with low ADC (d) representing restricted diffusion, due to the high cellularity of the tumor. (e, f) Marked homogeneous enhancement is observed on axial and sagittal postcontrast T1WI. (g) MRS shows low NAA, high choline and a lactate peak. Histology revealed large B cell lymphoma
11 Central Nervous Lymphomas and Hemopoietic Neoplasms
e
g
Fig. 11.3 (continued)
f
307
308
a
J. Frühwald-Pallamar et al.
b
Fig. 11.4 Primary CNS lymphoma. (a) Unenhanced CT shows a high-attenuation mass in the left temporo-parietal region (b) On the corresponding post-contrast CT image the mass iw markedly enhanced
The FA and ADC of lymphomas were significantly lower than those in the GBM, and cutoff values to differentiate lymphomas from GBM were 0.192 for FA, 0.33 for FA ratio, 0.818 for ADC, and 1.06 for ADC ratio [36]. On MR spectroscopy (MRS,) decreased NAA and elevation of choline will be present. Lipid and lactate peaks were also reported (Figs. 11.3 and 11.5) [37]. On perfusion MR, lymphomas tend to have a low regional cerebral blood volume (rCBV) ratio compared to gliomas (Fig. 6.22f). In one study, rCBV of primary intracranial lymphomas was 1.72, while that of highgrade gliomas was 4.86 [38]. In another study, primary and secondary CNS lymphomas showed relatively low values for the maximum rCBV ratio [39]. Intravascular (angiocentric) lymphoma is an extremely rare type of systemic large B-cell lymphoma [40]. Lymphoma is limited to the lumens of medium to
small blood vessels and capillary beds. Imaging findings in intravascular lymphoma include multifocal T2WI hyperintensity in the deep white matter, cortex, or basal ganglia, with variable enhancement (linear, punctate, and patchy) (Fig. 11.8). Differential diagnoses of this rare lymphoma type are vasculitis and vascular dementia. Leptomeningeal primary lymphomas with no associated cerebral parenchymal lesions are uncommon, comprising approximately 7% of PCNSLs, and with only a few cases reported in the literature. Lepto meningeal lymphoma is usually the result of secondary spread of the disease. Contrast T1-weighted and FLAIR MRI sequences are the most sensitive for detecting leptomeningeal lymphoma, which will demonstrate hyperintensity of the subarachnoid spaces on FLAIR, with leptomeningeal enhancement on postcontrast scans (Fig. 11.9).
309
11 Central Nervous Lymphomas and Hemopoietic Neoplasms
a
b
c
d
Fig. 11.5 A 27-year-old male patient with B-cell lymphoma. (a, b) A large low signal intensity mass is shown in the left periventricular region on axial and coronal T2WI. Note hyperintense perifocal edema. Low signal on T2WI is a result of the
high cellularity of lymphoma. (c) Marked enhancement is demonstrated on axial postcontrast T1WI. (d, e) MRS shows high choline peak, low NAA peak, lipid peak, and high Cho/Cr ratio
310
e
J. Frühwald-Pallamar et al.
demyelinating plaques will allow differentiation from CNS lymphoma (Fig. 11.10). Meningioma may be difficult to differentiate from peripheral lymphoma, which is in contact with the meningeal surface (Fig. 11.11). However, the calcification commonly seen in meningiomas is not found in lymphomas.
11.1.5 Imaging Findings in Immunocompromised Patients
Fig. 11.5 (continued)
Imaging findings in leptomeningeal lymphoma are nonspecific, and the differential diagnosis includes other secondary malignancies. Lymphomatous involvement of the cranial vault is a rare entity, described not only in AIDS patients, but also in non-AIDS patients. Malignant lymphomas of the cranial vault are recognized on CT/MR examinations as large scalp-enhancing lesions with bone destruction, and with an epidural component [41].
11.1.4.3 Differential Diagnosis The differential diagnosis of PCNSL in immunocompetent patients includes glioma, metastasis, primitive neuroectodermal tumor, multiple sclerosis, and meningioma. Glioma shows high signal intensity on T2-weighted images, while PCNSL is usually hypo- or isointense. Metastasis is usually present at the gray/white matter zone and shows significant edema. Primitive ectodermal tumors show almost similar imaging characteristics, but they appear in childhood or adolescence, whereas lymphomas usually appear in adults. Multiple sclerosis plaques may mimic PCNSL, because they are most commonly located in the periventricular region, and may show marked contrast enhancement. In addition, both may show significant improvement after steroid administration. However, the T2 hyperintensity of
There are distinct differences in imaging findings in immunocompromised patients with PCNSL. On CT scans, lymphomas tend to present as necrotic masses with peripheral or a ring-like enhancement (Fig. 11.12). On T1WI and T2WI, they may be more heterogeneous due to hemorrhage and necrotic changes [27, 28, 42–44]. In contrast to the immunocompetent population, lymphomas present more commonly as multifocal lesions with peripheral enhancement in immuno compromised patients (Fig. 11.13) [27]. The differential diagnosis of ring-enhancing lesions in immunocompromised patients include toxoplasmosis and bacterial or fungal abscesses. Studies have clearly shown that the number of lesions, the signal intensities, location, and enhancement pattern on conventional MR sequences are not reliable factors for the differentiation between lymphoma and toxoplasmosis. Subependymal spread is more common in lymphoma. In one study, periventricular involvement was seen in 50% of AIDS patients with lymphoma and only 3% of patients with toxoplasmosis, and subependymal spread was found in 38% of patients with lymphoma and in no patients with toxoplasmosis [42] (Fig. 11.14). In any case, antitoxoplasmosis therapy is always indicated when lesions with the appearance of toxoplasmosis are found on CT or MRI. Response to the antitoxoplasmosis therapy within 1–2 weeks favors the diagnosis of toxoplasmosis (Fig 11.15). If there is no response, there is a strong possibility of PCNSL, and biopsy is indicated [44]. Studies have suggested that perfusion MR may also be used in AIDS patients, to differentiate toxoplasmosis from lymphoma. Reduced rCBV in toxoplasmosis lesions is probably due to a lack of vasculature within the abscess; increased rCBV in lymphomas is probably
311
11 Central Nervous Lymphomas and Hemopoietic Neoplasms
a
b
c
Fig. 11.6 A 39-year-old male patient with a history of lymphoma of the right testis. He presented with personality changes. Lymphoma was found in the left basal ganglia region. (a, b) On axial FLAIR and coronal T2WI high signal intensity abnormal-
d
ity is shown in the left basal ganglia region, frontal cortex, corpus callosum, and white matter of both frontal lobes. (c, d) Only mild enhancement is observed on postcontrast T1WI
312
a
J. Frühwald-Pallamar et al.
b
c
Fig. 11.7 MRI of the same patient of figure 1. (a) Axial T1WI shows a low signal intensity left temporo-parietal mass (b) On axial T2WI the mass is isointense to the gray matter. Note the
moderate surrounding edema. (c) Contrast enhanced T1WI shows strong homogeneous enhancement
313
11 Central Nervous Lymphomas and Hemopoietic Neoplasms
a
b
c
d
e
Fig. 11.8 Patient with biopsy proven angiocentric lymphoma. (a, b) Axial FLAIR images show signal abnormality in the left cerebellum, white matter and corpus callosum. Post-contrast
f
T1WI axial (c, d) and coronal (e, f) show different types of enhancement (patchy, punctate and linear)
314
a
Fig. 11.9 A 43-year-old female patient with a history of Burkitt lymphoma, and secondary meningeal lymphomatous disease. (a) On axial FLAIR subtle high signal intensity is observed in
a
J. Frühwald-Pallamar et al.
b
the subarachnoid spaces. (b) Meningeal enhancement is noted on coronal postcontrast T1WI with magnetization transfer contrast (MTC)
b
Fig. 11.10 Multiple sclerosis mimicking PCL. (a) Post-contrast T1WI shows a homogeneously enhanced left parietal plaque. (b) Axial T2WI shows high signal intensity of demyelinating plaque
11 Central Nervous Lymphomas and Hemopoietic Neoplasms
315
due to hypervascularity in the foci of active tumor growth. Decreased rCBV in the edema is probably due to vasoconstriction associated with increased interstitial pressure [45]. MRS was introduced as a potentially noninvasive adjunctive tool for distinguishing toxoplasmosis from lymphoma. In one series of 11 toxoplasma lesions and eight lymphomas of the brain in AIDS patients, MRS correctly diagnosed 11/11 toxoplasmosis cases, and 8/8 lymphoma cases [46]. MR spectra from 18 toxoplasma and nine lymphoma lesions on 1.5 T were reported in one study [47]. Visual analysis in that study failed to differentiate between toxoplasma and lymphoma. The maturity of the lesions, as well as the presence of necrosis, are important variables for MRS. An early, immature toxoplasma lesion will show a nonspecific spectrum, with no lipid, increased choline, and decreased NAA. In a later stage (necrotic mass), the spectrum will be similar to necrotic lymphoma. The presence of lipids will be seen in both entities as a result of brain destruction. Fig. 11.11 Non Hodgkin lymphoma mimicking meningioma. F-18 fluorodeoxyglucose (FDG)-positron emission Axial post-contrast T1WI shows a markedly enhanced dural tomography (PET) and 201-thallium SPECT may be lesion with leptomeningeal infiltration used to differentiate PCNSL from toxoplasmasmosis. Thallium is a potassium analog with uptake in active tissue, and normally, there is no uptake of 201TI in the brain. One of the early studies showed that positive 201 TI brain SPECT is suggestive of CNS lymphoma, (Fig. 14.4) and negative uptake suggests infection (toxoplasmosis) in AIDS patients [48]. All lymphomas in this series showed an uptake of 201TI in contrast to infections, which showed a negative uptake [48]. However, the results from another prospective study on 14 patients with AIDS and focal brain lesions suggest the inability of 201TI SPECT to differentiate lymphoma from toxoplasmosis [49]. The accuracy was only 57% in that study, with a positive predictive value of 43% and a negative predictive value of 71% [48]. The potential use of FDG-PET in differentiating lymphoma from toxoplasmosis in AIDS patients has also been described in only a limited number of studies [50, 51]. The standardized uptake values (SUVs) were reported to be significantly higher in lymphomas than in toxoplasma lesions (Fig. 14.5) [52]. A combined approach using imaging findings and CSF analysis (toxoplasma IgG, and EBV DNA polymerase chain reaction in CSF) improves diagnostic Fig. 11.12 Primary PCL in a patient with AIDS. Contrast- accuracy in the differentiation of lymphoma and toxoenhanced CT shows a ring-like enhanced mass with severe mass plasmosis in HIV-positive patients. effect and surrounding white matter edema
316
J. Frühwald-Pallamar et al.
a
b
c
d
Fig. 11.13 A 47 year-old human immunodeficiency virus (HIV)positive male patient with headache and nausea. (a, b) On axial FLAIR images multiple low signal intensity lesions with high signal intensity edema are demonstrated. The lesions are located in the posterior fossa and in both cerebral hemispheres. (c, d) High
signal is observed on DWI with low ADC values, consistent with high cellularity tumor. (e, f) Peripheral enhancement with central necrotic parts is demonstrated on axial and coronal postcontrast T1WI. (g) The hypermetabolism in the TC99m SPECT–CT favors the diagnosis of lymphoma over toxoplasmosis
317
11 Central Nervous Lymphomas and Hemopoietic Neoplasms
e
g
Fig. 11.13 (continued)
f
318
a
Fig. 11.14 PCL in a patient with AIDS. Axial T2W (a) and post-contrast T1W (b) images show a large periventricular mass involving the left basal ganglia and the adjacent white matter.
a
J. Frühwald-Pallamar et al.
b
Note the ring-like enhancement and the subependymal extension of the mass
b
Fig. 11.15 CNS toxoplasmosis in a patient with AIDS. Coronal T1WI before (a) and after antitoxoplasmosis therapy (b). The number of the enhanced cerebral nodules is significantly decreased
11 Central Nervous Lymphomas and Hemopoietic Neoplasms
a
319
b
Fig. 11.16 CT in a patient with granulocytic sarcoma. (a) Non-contrast CT shows a hyperdense lesion in the left cerebellar hemisphere. (b). Post-contrast CT shows intense enhancement of the lesion
11.2 Leukemia Leukemias are a heterogeneous group of hematologic malignancies due to neoplastic proliferation of hema topoietic cells. CNS leukemia can present as (a) meningeal disease (usually in ALL), (b) intravascular aggregates, and (c) focal masses (chloroma). Granulocytic sarcoma (chloroma) is a solid tumor found in association with systemic leukemia, usually the myelogeneous type [53, 54]. They are also called extramedullary myeloid cell tumors (EMT). The original term chloroma was used by King in 1853 with regard to the lesion’s typical green color, which is caused by high levels of myeloperoxidase in immature cells [55]. Rappaport introduced the term “granulocytic sarcoma,” since approximately 25% of these tumors are white, brown, or gray, depending on the variety of cell types involved [56]. The color differences are due to concentrations of myeloperoxidase and its various oxidative states. Granulocytic sarcomas consist of primitive precursors of the granulocytic series of white blood cells that include myeloblasts, promyelocytes, and myelocytes [57, 58]. Granulocytic sarcomas of the brain are uncommon and are postulated to arise from neoplastic cells that traverse the calvarial marrow to involve the dura and then pass through the perivenous adventitial tissue to
invade the brain parenchyma [57, 58]. Patients with leukemia who develop granulocytic sarcoma have a poor prognosis. The chloromas may be dural based and less common intraparenchymally. On NECT, they are iso- to hyperdense, with homogeneous enhancement following the administration of contrast material. Differential diagnosis includes hemorrhages, in cases of marked hyperdensity. There may be some affinity for the posterior fossa (Fig. 11.16). The MR characteristics are variable, but, in most cases, the lesions are inhomogeneous, and hypo- to isointense on both T1-weighted and T2-weighted images. The signal intensity on T2-weighted images is due to the high levels of myeloperoxidase. FLAIR is the most sensitive sequence for simultaneous leptomeningeal changes. After the administration of contrast medium, chloromas show homogeneous enhancement [59–64] (Fig. 11.17). MR angiography may identify medium vessel vasculitis. Because of the dural-based location, meningioma and metastatic neuroblastoma are included in the differential diagnosis [63]. The dural tail is more common in meningioma, and the metastatic neuroblastoma rarely occurs without an extracranial manifestation. Treatment options include chemotherapy and radiation, as well as bone marrow transplantation (BMT).
320
a
J. Frühwald-Pallamar et al.
b
c
Fig. 11.17 Chloroma in a patient with leukemia after bone marrow transplantation. (a) Axial T1WI shows a hypointense lesion in the right parietal lobe, which is hypo- to isointense on T2WI
surrounded by perifocal high signal edema (b). After contrast administration the central part of the lesion is enhanced (c)
Patients with leukemia may develop disseminated intravascular coagulation (intravascular aggregates) and hypofibrinogenemia, resulting in multiple small hemorrhages in the subcortical white matter. On CT or MR, subacute hematomas, or multiple small intraparenchymal hemorrhages may be seen as the initial manifestation of leukemia (Fig. 11.18) [64].
11.2.1 Lymphomatous and Leukemic Meningitis (LM) In contrast to patients with solid primary tumors, patients with leukemia or lymphoma often present with meningitis, without evidence of systemic disease or during periods of remission [65]. In a review of
321
11 Central Nervous Lymphomas and Hemopoietic Neoplasms
a
b
c
d
Fig. 11.18 A 13-year-old boy presented with headache, multiple cortical hemorrhages were detected on CT and MRI. High white blood count was found, and acute leukemia was diagnosed. (a) NECT shows multiple cortical hyperdense areas, rep-
resenting hemorrhages. (b) Axial FLAIR and coronal T2WI (c) demonstrate multiple low signal intensity lesions with hyperintense rim. (d) On axial SWI multiple low signal intensity bleedings were found
63 cases, Kaplan et al. found that meningitis from solid tumors occurred in patients with advanced systemic disease in 90% of the cases, compared to patients with leukemia and lymphoma whose meningitis occurred without systemic disease (18 and 13%, respectively) or
during remission (35 and 27%, respectively) [66]. Acute leukemia, and particularly, acute lymphocytic leukemia (ALL), has the highest propensity to invade the meninges and result in leukemic meningitis (LM). Clinically, patients with hematologic malignancies
322
present with a higher frequency of cranial nerve signs as initial manifestations of neoplastic meningitis. The most useful laboratory test in diagnosing LM is the lumbar puncture and an examination of the CSF. In nearly all patients with LM, the CSF is abnormal, regardless of the results of CSF cytology. Multiple other intracerebral manifestiations and complications of leukemia must be mentioned: posterior reversible encephalopathy syndrome (PRES), fungal abscesses, vasculitis, and posttherapeutic changes, such as venous thrombosis, cavernous angiomas, and lymphoma after BMT [67].
References 1. Hodgkin T (1832) On some morbid experiences of the absorbent glands and spleen. Med Chir Trans 17:68–97 2. Jaffe ES, Harris NL, Stein H, Isaacson PG (2008) Classi fication of lymphoid neoplasms: the microscope as a tool for disease discovery. Blood 112:4384–4399 3. DeAngelis LM (1991) Primary central nervous system lymphoma: a new clinical challenge. Neurology 41:619–621 4. Miller DC, Hochberg FH, Harris NL, Gruber ML, Louis DN, Cohen H (1994) Pathology with clinical correlations of primary central nervous system non-Hodgkin’s lymphoma. The Massachusetts General Hospital experience 1958–1989. Cancer 74:1383–1397 5. Olson JE, Janney CA, Rao RD et al (2002) The continuing increase in the incidence of primary central nervous system non-Hodgkin lymphoma: a surveillance, epidemiology, and end results analysis. Cancer 95:1504–1510 6. Eby NL, Grufferman S, Flannelly CM, Schold SC Jr, Vogel FS, Burger PC (1988) Increasing incidence of primary brain lymphoma in the US. Cancer 62:2461–2465 7. Hao D, DiFrancesco LM, Brasher PM et al (1999) Is primary CNS lymphoma really becoming more common? A population-based study of incidence, clinicopathological features and outcomes in Alberta from 1975 to 1996. Ann Oncol 10:65–70 8. Cote TR, Manns A, Hardy CR, Yellin FJ, Hartge P (1996) Epidemiology of brain lymphoma among people with or without acquired immunodeficiency syndrome. AIDS/ Cancer Study Group. J Natl Cancer Inst 88:675–679 9. Coons SW, Ashby LS (1999) Pathology of intracranial neoplasms. Neuroimaging Clin North Am 9:615–649 10. Taiwo BO (2000) AIDS-related primary CNS lymphoma: a brief review. AIDS Read 10:486–491 11. Adams JH, Howatson AG (1990) Cerebral lymphomas: review of 70 cases. J Clin Pathol 43:544–547 12. Werner MH, Phuphanich S, Lyman GH (1995) Increasing incidence of primary brain tumors in the elderly in Florida. Cancer Control 2:309–314 13. Baumgartner JE, Rachlin JR, Beckstead JH et al (1990) Primary central nervous system lymphomas: natural history and response to radiation therapy in 55 patients with acquired immunodeficiency syndrome. J Neurosurg 73:206–211
J. Frühwald-Pallamar et al. 14. Zimmerman RA (1990) Central nervous system lymphoma. Radiol Clin North Am 28:697–721 15. Koeller KK, Smirniotopoulos JG, Jones RV (1997) Primary central nervous system lymphoma: radiologic-pathologic correlation. Radiographics 17:1497–1526 16. Grant JW, Isaacson PG (1992) Primary central nervous system lymphoma. Brain Pathol 2:97–109 17. Schwechheimer K, Braus DF, Schwarzkopf G, Feller AC, Volk B, Muller-Hermelink HK (1994) Polymorphous highgrade B cell lymphoma is the predominant type of spontaneous primary cerebral malignant lymphomas. Histological and immunomorphological evaluation of computed tomography-guided stereotactic brain biopsies. Am J Surg Pathol 18:931–937 18. Doerr M, Schumacher M, Mohadjer M (1987) Primary malignant lymphoma of the central nervous system – an increasingly more frequent tumor. Nervenarzt 58:538–542 19. Lee YY, Bruner JM, Van Tassel P, Libshitz HI (1986) Primary central nervous system lymphoma: CT and pathologic correlation. AJR Am J Roentgenol 147:747–752 20. Herrlinger U, Schabet M, Bitzer M, Petersen D, Krauseneck P (1999) Primary central nervous system lymphoma: from clinical presentation to diagnosis. J Neurooncol 43:219–226 21. Braus DF, Schwechheimer K, Muller-Hermelink HK, Schwarzkopf G, Volk B, Mundinger F (1992) Primary cerebral malignant non-Hodgkin’s lymphomas: a retrospective clinical study. J Neurol 239:117–124 22. Hochberg FH, Miller DC (1988) Primary central nervous system lymphoma. J Neurosurg 68:835–853 23. Ferreri AJ, Blay JY, Reni M et al (2003) Prognostic scoring system for primary CNS lymphomas: the International Extranodal Lymphoma Study Group experience. J Clin Oncol 21:266–272 24. Abrey LE, Ben-Porat L, Panageas KS et al (2006) Primary central nervous system lymphoma: the Memorial SloanKettering Cancer Center prognostic model. J Clin Oncol 24:5711–5715 25. Bataille B, Delwail V, Menet E et al (2000) Primary intracerebral malignant lymphoma: report of 248 cases. J Neurosurg 92:261–266 26. DeAngelis LM (1999) Primary CNS lymphoma: treatment with combined chemotherapy and radiotherapy. J Neurooncol 43:249–257 27. Thurnher MM, Thurnher SA, Schindler E (1997) CNS involvement in AIDS: spectrum of CT and MR findings. Eur Radiol 7:1091–1097 28. Roman-Goldstein SM, Goldman DL, Howieson J, Belkin R, Neuwelt EA (1992) MR of primary CNS lymphoma in immunologically normal patients. AJNR Am J Neuroradiol 13:1207–1213 29. Davenport C, Dillon WP, Sze G (1992) Neuroradiology of the immunosuppressed state. Radiol Clin North Am 30: 611–637 30. Reiche W, Deinzer M (1998) Neuroradiologic diagnosis of primary non-Hodgkin’s lymphoma of the brain. Radiologe 38:913–923 31. Vandermarcq P, Drapeau C, Ferrie JC (1997) Imaging aspects of primary cerebral lymphoma. Neurochirurgie 43:363–368 32. Reiche W, Hagen T, Schuchardt V, Billmann P (2007) Diffusion-weighted MR imaging improves diagnosis of CNS lymphomas. A report of four cases with common and
11 Central Nervous Lymphomas and Hemopoietic Neoplasms uncommon imaging features. Clin Neurol Neurosurg 109:92–101 33. Stadnik TW, Chaskis C, Michotte A et al (2001) Diffusionweighted MR imaging of intracerebral masses: comparison with conventional MR imaging and histologic findings. AJNR Am J Neuroradiol 22:969–976 34. Guo AC, Cummings TJ, Dash RC, Provencale JM (2002) Lymphomas and high-grade astrocytomas: comparison of water diffusibility and histologic characteristics. Radiology 224:177–183 35. Zacharia TT, Law M, Naidich TP, Leeds NE (2008) Central nervous system lymphoma characterization by diffusionweighted imaging and MR spectroscopy. J Neuroimaging 18:411–417 36. Toh CH, Castillo M, Wong AM, Wei KC, Wong HF, Ng SH, Wan YL (2008) Primary cerebral lymphoma and glioblastoma multiforme: differences in diffusion characteristics evaluated with diffusion tensor imaging. AJNR Am J Neuroradiol 29(3):471–475 37. Hourani R, Brant LJ, Rizk T, Weingart JD, Barker PB, Horska A (2008) Can proton MR spectroscopic and perfusion imaging differentiate between neoplastic and nonneoplastic brain lesions in adults? AJNR Am J Neuroradiol 29:366–372 38. Liao W, Liu Y, Wang X, Jiang X, Tang B, Fang J, Chen C, Hu Z (2009) Differentiation of primary central nervous system lymphoma and high-grade glioma with dynamic susceptibility contrast-enhanced perfusion magnetic resonance imaging. Acta Radiol 50(2):217–225 39. Lee IH, Kim ST, Kim HJ, Kim KH, Jeon P, Byun HS (2009) Analysis of perfusion weighted image of CNS lymphoma. Eur J Radiol 40. Williams RL, Meltzer CC, Smirniotopoulos JG, Fukui MB, Inman M (1998) Cerebral MR imaging in intravascular lymphomatosis. AJNR Am J Neuroradiol 19:427–431 41. Thurnher MM, Rieger A, Kleibl-Popov Ch et al (2001) Malignant lymphoma of the cranial vault in an HIV-positive patient: imaging findings. Eur Radiol 11:1506–1509 42. Goldstein JD, Zeifer B, Chao C et al (1991) CT appear ance of primary CNS lymphoma in patients with acquired immunodeficiency syndrome. J Comput Assist Tomogr 15: 39–44 43. Dina TS (1991) Primary central nervous system lymphoma versus toxoplasmosis in AIDS. Radiology 179:823–828 44. Chappell ET, Guthrie BL, Orenstein J (1992) The role of stereotactic biopsy in the management of HIV-related focal brain lesions. Neurosurgery 30:825–829 45. Ernst TM, Chang L, Witt MD et al (1998) Cerebral toxoplasmosis and lymphoma in AIDS: perfusion MR imaging experience in 13 patients. Radiology 208:663–669 46. Miller RF, ll-Craggs MA, Costa DC et al (1998) Magnetic resonance imaging, thallium-210 SPECT scanning, and laboratory analyses for discrimination of cerebral lymphoma and toxoplasmosis in AIDS. Sex Transm Infect 74:258–264 47. Chinn RJS, Wilkinson ID, Hall-Craggs MA et al (1995) Toxoplasmosis and primary central nervous system lymphoma in HIV infection: diagnosis with MR spectroscopy. Radiology 197:649–654 48. Ruiz A, Ganz WI, Post MJD et al (1994) Use of Thallium-201 brain SPECT to differentiate cerebral lymphoma from toxoplasma encephalitis in AIDS patients. AJNR Am J Neuroradiol 15:1885–1894
323 49. Licho R, Litofsky NS, Senitko M et al (2002) Inacuracy of TI-201 brain SPECT in distingushing cerebral infections from lymphoma in patients with AIDS. Clin Nucl Med 27:81–86 50. Villringer K, Jager H, Dichgans M et al (1995) Differential diagnosis of CNS lesions in AIDS patients by FDG-PET. J Comput Assist Tomogr 19:532–536 51. Heald AE, Hoffman JM, Bartlett JA et al (1996) Differentiation of central nervous system lesions in AIDS patients using positron emission tomography (PET). Int J STD AIDS 7:337–346 52. ÓDoherty MJ, Barrington SF, Campbell M et al (1997) PET scanning and the human immunodeficiency virus-positive patients. J Nucl Med 38:1575–1583 53. Grondin L, Auger R, Rioux E, Gould PV (1996) Multiple intracerebral granulocytic sarcomas in a patient with chronic myeloid leukemia. Can Assoc Radiol J 47:132–135 54. Liu PI, Ishimaru T, McGregor DH, Okada H, Steer A (1973) Autopsy study of granulocytic sarcoma (chloroma) in patients with myelogenous leukemia, Hiroshima-Nagasaki 1949–1969. Cancer 31:948–955 55. King A (1853) A case of chloroma. Monthly J Med 17:17 56. Rappaport H (1967) Tumors of the hematopoietic system. In: Atlas of tumor pathology, Sect. III, Fascicle 8. Armed Forces Institute of Pathology, Washington, pp 241–247 57. Muss HB, Moloney WC (1973) Chloroma and other myeloblastic tumors. Blood 42:721–728 58. Barnett MJ, Zussman WV (1986) Granulocytic sarcoma of the brain: a case report and review of the literature. Radiology 160:223–225 59. Pui MH, Fletcher BD, Langston JW (1994) Granulocytic sarcoma in childhood leukemia: imaging features. Radiology 190:698–702 60. Kao SC, Yuh WT, Sato Y, Barloon TJ (1987) Intracranial granulocytic sarcoma (chloroma): MR findings. J Comput Assist Tomogr 11:938–941 61. Leonard KJ, Mamourian AC (1989) MR appearance of intracranial chloromas. AJNR Am J Neuroradiol 10:S67–S68 62. Wright DH, Hise JH, Bauserman SC, Naul LG (1992) Intracranial granulocytic sarcoma: CT, MR, and angiography. J Comput Assist Tomogr 16:487–489 63. Velasco F, Ondarza R, Quiroz F, Arista J (1993) Meningiomalike intracranial granulocytic sarcoma (chloroma). Radiologic and surgical findings. Rev Invest Clin 45:473–478 64. Martínez-Galdámez M, Brea Alvarez B, Saura Lorente P (2007) Intracranial hemorrhages as first manifestation of acute myeloid leukemia. Neurologia 22(7):467–468 65. Chamberlain MC, Nolan C, Abrey LE (2005) Leukemic and lymphomatous meningitis: incidence, prognosis and treatment. J Neurooncol 75(1):71–83 66. Kaplan JG, DeSouza TG, Farkash A, Shafran B, Pack D, Rehman F, Fuks J, Portenoy R (1990) Leptomeningeal metastases: comparison of clinical features and laboratory data of solid tumors, lymphomas and leukemias. J Neuro Oncol 9:225–229 67. Chen CY, Zimmerman RA, Faro S, Bilaniuk LT, Chou TY, Molloy PT (1996) Childhood leukemia: central nervous system abnormalities during and after treatment. AJNR Am J Neuroradiol 17:295–310
Masses of the Sellar and Junxtasellar Region
12
Eric C. Bourekas, H. Wayne Slone, and Abhik Ray-Chaudhury
Contents 12.1 Introduction........................................................... 325 12.2 Anatomy................................................................. 325 12.3 Imaging.................................................................. 327 12.3.1 Pituitary Adenomas................................................ 327 12.3.2 Meningioma............................................................ 333 12.3.3 Craniopharyngioma................................................ 335 12.3.4 Chiasmatic and Hypothalamic Gliomas................. 342 12.3.5 Germinoma............................................................. 342 12.3.6 Epidermoids and Dermoids.................................... 348 12.3.7 Rathke’s Cleft Cyst................................................. 353 12.3.8 Arachnoid Cysts...................................................... 354 12.3.9 Neurosarcoidosis..................................................... 356 12.3.10 Hamartomas of the Tuber Cinereum....................... 357 12.3.11 Schwannomas and Neurofibromas.......................... 357 12.3.12 Aneurysms.............................................................. 359 12.3.13 Pituitary Hyperplasia.............................................. 360 12.3.14 Lymphocytic Hypophysitis..................................... 361 12.3.15 Pituicytoma............................................................. 361 12.3.16 Metastases............................................................... 361 12.3.17 Infundibular Masses................................................ 361 12.3.18 Miscellaneous Masses............................................ 364
12.1 Introduction Tumors of the sellar and junxtasellar regions are common, accounting for in excess of 10–15% of all intracranial tumors [1]. Although the majority of these are benign tumors, mostly pituitary adenomas, they can account for significant morbidity and a shortened life expectancy. Pituitary adenomas are the most common sellar and suprasellar masses in adults, with meningiomas being the second most common. In children, craniopharyngiomas and chiasmatic/hypothalamic gliomas account for the vast majority of lesions in this region. Adenomas, meningiomas, craniopharyngiomas, chiasmatic/hypothalamic gliomas, and aneurysms account for 75% of sellar and junxtasellar masses [2]. Although the focus of our discussion is tumors, some nonneoplastic entities are discussed, since these can be difficult to distinguish from tumors and must be differentiated, since the treatment approach can be very different.
References............................................................................ 369
12.2 Anatomy
E.C. Bourekas (*) and H.W. Slone Section of Interventional and Diagnostic Neuroradiology, Department of Radiology, The Ohio State University Medical Center, 623 Means Hall, 1654 Upham Drive, Columbus, OH 43210, USA e-mail:
[email protected] and A. Ray-Chaudhury Department of Pathology, The Ohio State University School of Medicine, Columbus, Ohio, USA
The pituitary gland lies in the sella turcica which is a depression in the skull base involving the sphenoid bone. The gland has functioning anterior and posterior lobes separated by the pars indermedia which is nonfunctioning. The anterior and posterior lobes are easily identifiable on sagittal T1-weighted images with the posterior lobe being very hyperintense (Fig. 12.1). In infants, the anterior pituitary is normally bright but quickly fades to its normal signal, isointense to cortex. On occasion, the posterior pituitary bright spot may not be in the sella and may be ectopic in the suprasellar cistern (Fig. 12.2). The posterior lobe, also known as the neurohypophysis, is connected to the hypothalamus via the infundibulum or
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_12, © Springer-Verlag Berlin Heidelberg 2011
325
326
a
Fig. 12.1 Normal anatomy. (a) Midline sagittal T1-weighted image with contrast and (b) coronal T1-weighted image without contrast through the sellar region. A optic chiasm; B infundibu-
a
E.C. Bourekas et al.
b
lum; C posterior pituitary; D anterior pituitary; E sphenoid sinus; F carotid artery; G cavernous sinus. Note the increased signal of the posterior pituitary
b
Fig. 12.2 Ectopic posterior pituitary. (a) Sagittal T1-weighted image without contrast and (b) coronal T1-weighted image without contrast demonstrate a hyperintense lesion in the supra-
sellar cistern with absence of the normal hyperintense posterior pituitary. If it were not for the absence of the posterior pituitary, the hyperintensity could also represent a lipoma or a dermoid
pituitary stalk. Superior to the pituitary and sella is the suprasellar cistern in which the infundibulum and optic chiasm are located in close proximity to each other. Inferior to the sella is the sphenoid sinus which provides surgical access to pituitary lesions. Laterally, the sella is
bordered by the cavernous sinus, anteriorly the tuberculum sella and anterior clinoid processes, and posteriorly the dorsum sella and posterior clinoids (Fig. 12.1). The superior surface of the pituitary is normally concave or sometimes flat. It can be convex when there is a mass or
327
12 Masses of the Sellar and Junxtasellar Region
during pregnancy, postpartum, and in children with precocious puberty. The normal pituitary is usually less than 10 mm in height but may normally be larger in girls during puberty and in women during pregnancy and immediately postpartum.
12.3 Imaging Imaging of the sellar region has improved dramatically over the last 30 years, progressing from skull radiographs, which are rarely used today, to the improved but not very sensitive or specific CT, to MR imaging which is the examination of choice in evaluating this region. Although CT is used at some institutions for evaluation of the sellar and parasellar region, soft tissue characterization is generally poor and evaluation of this region is hampered by beam hardening artifacts related to the dense bone of the skull base and by metallic artifacts from dental fillings, particularly in the coronal plane. CT is excellent in the evaluation of bony structures and calcifications, both of which can be important in the differential diagnosis. CT should be used only as the initial and primary mode of evaluation in patients in whom there is a contraindication to MR, such as, for example, having a pacemaker. In these patients, 1 mm axial and coronal sections are obtained postcontrast, with precontrast images in one plane being useful. MRI, because of its multiplanar capability and superior soft tissue characterization, is the examination of choice for evaluation of the pituitary, parasellar, and suprasellar regions. Involvement of the optic chiasm, cavernous sinus, sphenoid sinus, orbit, temporal lobes, and carotid arteries can all be best seen using MRI. A small field of view of 16–20, with thin sections of 3 mm or less and high resolution, with a matrix of 256 × 256 are essential. T1- and T2-weighted images are generally obtained, with T1-weighted images after contrast administration. Sagittal and coronal imaging is most useful for imaging the pituitary and cavernous sinus region. At our institution, we obtain a high-resolution 3-dimensional sequence in the coronal plane with contrast for optimal evaluation of the pituitary. Because the pituitary lacks a blood-brain barrier, it enhances intensely, early, and more so than tumors, so that tumors generally appear as areas of nonenhancement. Dynamic T1-weighted fast spin-echo imaging after a bolus of contrast can detect lesions not seen on
standard imaging in 11–14% of cases [3, 4]. However, in 12.5–17% of cases, lesions are better seen with standard postcontrast images and in 8–9% of cases, lesions are seen with standard imaging and not with dynamic imaging [4]. For these reasons, standard imaging is used in most cases, with dynamic imaging being performed in problematic cases. The increased sensitivity of MR creates the problem of incidental lesions. Autopsy studies reveal incidental pituitary pathology in up to 27% of cases, most being incidental microadenomas or pars intermedia cysts. Focal 2–3 mm lesions on imaging will prove to be incidental as often as they are endocrinologically significant [1]. In patients with incidental lesions, with normal testing and no symptoms, if the lesion is stable for a 2-year period on follow-up imaging, no further studies are necessary [5]. Newer imaging techniques such as diffusion tensor imaging and spectroscopy play no significant role in the evaluation of sellar region lesions. Venous sampling of the inferior petrosal or cavernous sinus is an invasive study that can be valuable in the diagnosis of a suspected adenoma and in particular Cushing’s disease. Catheters are placed from each of the femoral veins into the internal jugular veins and then advanced into each of inferior petrosal sinuses [6]. Blood samples are obtained from each side both before and after stimulation with corticotropin releasing hormone (CRH) and analyzed. This technique can reliably distinguish pituitary Cushing’s disease due to a microadenoma from ectopic ACTH syndrome, in patients with negative imaging, and can frequently lateralize the lesion [7, 8]. Although inferior petrosal sinus sampling is usually performed, it has been shown that bilateral, simultaneous cavernous sinus sampling, using corticotropin-releasing hormone, is as accurate as inferior petrosal sinus sampling in detecting Cushing’s disease and perhaps more accurate in lateralizing the abnormality within the pituitary gland [9].
12.3.1 Pituitary Adenomas These common tumors arise from adenohypophyseal cells. Approximately 91% of pituitary lesions are adenomas, with just over half of these representing hormone secreting tumors and the remainder nonfunctioning tumors [10]. Although many different classification schemes exist [11–13], pituitary adenomas are most
328
commonly classified according to size and function. Lesions smaller than 1 cm are classified as microadenomas and those larger than 1 cm are classified as macroadenomas. Microadenomas tend to be functioning lesions, whereas macroadenomas are usually nonfunctioning lesions. The most common functioning adenomas are prolactinomas. Other functioning adenomas include adrenocorticotropic hormone-secreting tumors, thyroidstimulating hormone-secreting tumors, and GH-secreting tumors. Carcinomas of the pituitary gland are rare [14]; in fact, metastases to the gland are more common, with incidental metastases noted at autopsy in 1–5% of cases [1]. Pituitary adenomas are usually seen in adults and are uncommon in children, representing less than 3% of all pediatric intracranial tumors [15]. When seen in childhood, they are usually seen in adolescent males; they are commonly macroadenomas and in particular prolactinomas and tend to be hemorrhagic [15]. Microadenomas most commonly present with a clinical picture reflecting the hormone excess. Nonsecreting macroadenomas go unrecognized until they produce visual compromise and headaches, as well as signs of hypopituitarism [16]. Modest increases in prolactin can be seen as a result of compression of the pituitary stalk. Although adenomas are overwhelmingly benign, they have been known to metastasize with seeding of the
a
Fig. 12.3 Microadenoma. (a) Conventional coronal spin-echo T1-weighted image with contrast through the pituitary, (b) highresolution 3D-SPGPR coronal image postcontrast through the sella. The patient is a 55-year-old female with hyperprolactine-
E.C. Bourekas et al.
CSF [17] and are associated with an increased morbidity and mortality due to an increased incidence of cardiovascular disease and cerebrovascular disease [18, 19]. Patients with nonfunctioning adenomas and acromegaly have also been shown to have a significantly higher incidence of malignancy than the general population [20]. Malignant pituitary adenomas are extremely rare and arise from the transformation of initially large, but benign, adenomas. The imaging appearance of pituitary adenomas is nonspecific, and no inference to histology can be made from the sellar patterns. However, additional clues may be present, related to other secondary endocrine changes. For instance, with GH-secreting tumors, acromegaly occurs, and one may visualize thickening of the scalp or enlargement of the mandible on imaging studies. Usually, Cushing’s adenomas are microadenomas, but compression vertebral fractures and a “buffalo hump” deformity may be clues. Prolactinomas are more variable in size; usually they are microadenomas, but may be macroadenomas. Nonfunctioning tumors tend to be macroadenomas. On MR, microadenomas appear as areas of nonenhancement (Fig. 12.3). The gland is generally not enlarged, although they may result in a change in the contour of the gland, with the upper margin becoming
b
mia. Both images demonstrate a small hypointense lesion of the right pituitary consistent with a microadenoma. Note that there is a partially empty sella
329
12 Masses of the Sellar and Junxtasellar Region
convex, the floor of the sella demonstrating a downsloping, or with deviation of the infundibulum. Macroadenomas cause expansion of the sella or erosion of the floor of the sella. The MR appearance of macroadenomas can be variable, although typically they follow gray matter signal on all sequences (Fig. 12.4). Cystic changes can be seen in up to 18% of cases [21] (Fig. 12.5). Enhancement is also variable
a
although generally intense and somewhat heterogeneous. Calcification is generally not seen [22]. A typical figure of eight configuration can be seen due to compression of the tumor at the diaphragma sella (Fig. 12.6). Knowledge of involvement of the cavernous sinus is important in surgical planning, with encasement of the carotid artery being the most specific sign of cavernous sinus invasion [23]. Invasion of
b
c
Fig. 12.4 Macroadenoma: (a) Coronal T1-weighted image precontrast. (b) Coronal T1-weighted image postcontrast. (c) Sagittal T1-weighted image postcontrast. Images demonstrate a large mass lesion of the sella and suprasellar region,
compressing the optic chiasm and with involvement of the right cavernous sinus. The unenhanced lesion has signal similar to gray matter on T1-weighted images and demonstrates moderate contrast enhancement
330
E.C. Bourekas et al.
a
b
c
d
Fig. 12.5 Macroadenoma: (a) Coronal CT after contrast. (b) Sagittal T1-weighted image precontrast. (c) Coronal T2-weighted image and (d) coronal T1-weighted image postcontrast. Images reveal a peripherally enhancing macroadenoma in a 64-year-old
male, causing expansion of the sella and compression of the optic chiasm that is not clearly seen. This case illustrates that macroadenomas can have a variable appearance and be cystic as in this case, solid, or solid and cystic
adjacent structures is seen in up to 35% of cases and is not indicative of malignancy [24] (Figs. 12.7 and 12.8). Malignant prolactinomas can be identified neither from the histopathologic nor neuroimaging aspects,
but only retrospectively from the presence of distant metastases (Fig. 12.9). Prolactinomas and growth-hormone secreting tumors can be treated medically [25]. Octreotide has
12 Masses of the Sellar and Junxtasellar Region
Fig. 12.6 Macroadenoma. Coronal T1-weighted image postcontrast. The patient is a 53-year-old male who is HIV-positive and is being evaluated for mental status changes. This incidental lesion has a classic “figure-of-eight” configuration seen with macroadenomas. The waist of the lesion is caused by the diaphragma sella through which the lesion has passed
331
Fig. 12.8 Invasive macroadenoma. Coronal 3D T1-weighted image with contrast in a 45-year-old male with headaches, no libido, night sweats, and nervousness. The patient has hyperprolactinemia and is being treated with bromocriptine. In the right aspect of the sella, there is a small nonenhancing macroadenoma, which has eroded the floor of the sella and extends into the sphenoid sinus, with probable involvement of the right cavernous sinus
revolutionized the management of patients with acromegaly [26]. Bromocriptine is commonly used in prolactinomas, with MRI used to evaluate the patient’s response to therapy [27]. A decrease in tumor size can be seen as early as 1 week after the start of therapy. Additionally, MRI can detect posttherapy hemorrhage into macroadenomas and mass effect or inferior herniation of the chiasm as a result of a decrease in the tumor size [28]. Bromocriptine has been associated
Fig. 12.7 Invasive macroadenoma. This coronal T1-weighted image postcontrast in a 46-year-old male demonstrates features of an invasive macroadenoma which include: complete involvement of the cavernous sinus with encasement of the carotid artery and tumor spreading through the foramen ovale of the skull base into the masticator space. The infundibulum is deviated to the right. The carotid, although encased, does not appear narrowed, a feature also typical of macroadenomas
332
a
c
E.C. Bourekas et al.
b
d
Fig. 12.9 Malignant adenoma. (a) CT shows a pituitary mass causing erosion of the sella and of the adjacent bones. (b) Axial noncontrast T1-weighted image shows a large isointense mass extending outside the sella, encasing the carotids. (c) Coronal
postcontrast T1-weighted image demonstrates intense and homogeneous enhancement. (d) Coronal postcontrast T1-weighted image of the same patient, more posteriorly, shows multiple enhanced metastatic lesions
with an increased incidence of intratumoral hemorrhage, also known as pituitary apoplexy [29] (Fig. 12.10). The clinical syndrome is characterized by sudden headache, vomiting, visual impairment, and meningismus, caused by rapid enlargement of an adenoma due to hemorrhagic infarction [30]. Subarachnoid
hemorrhage and vasospasm have been reported [31]. Transsphenoidal surgery is the preferred approach to the resection of pituitary adenomas because it is more associated with lower morbidity and mortality than the transcranial approach, which is generally the preferred approach for large tumors [32]. Imaging is essential
333
12 Masses of the Sellar and Junxtasellar Region
a
b
Fig. 12.10 Apoplexy. The patient is a 34-year-old female with a prolactinoma being treated with bromocriptine, who presented with complaints of a severe acute onset headache. (a) Sagittal T1-weighted image without contrast demonstrates an area of increased signal in the superior aspect of the pituitary consistent
with hemorrhage. (b) Coronal T1-weighted image postcontrast through the pituitary demonstrates an area of increased signal with a rim of decreased signal in the left aspect of the pituitary. This represents hemorrhage within a microadenoma
prior to surgery in order to define the extent of the lesion and also to be aware of any unusual anatomy such as medially coursing carotid arteries through the sella (Fig. 12.11). Grossly, pituitary adenomas are well-circumscribed benign tumors of softer consistency (Fig. 12.12a). Histologically, these tumors consist of monomorphic cells having round to ovoid nuclei and copious amounts of granular cytoplasm. The cells display a patternless sheet-like growth (Fig. 12.12b). Mitotic figures are usually not encountered, though bone invasion may be present. Collapse of the typical reticulin architecture, that surrounds nests of cells, is observed in the normal adenohypophysis, a feature which is diagnostic.
Meningiomas can arise from the diaphragma sella (Fig. 12.13), tuberculum sella, medial sphenoid ridge, cavernous sinus (Fig. 12.14), optic nerve sheaths, clinoids or clivus (Fig. 12.15), as well as the planum sphenoidale (Fig. 12.16). Rarely, meningiomas can be intrasellar, probably arising from the diaphragma or tuberculum sella and growing downward. Clinically, patients usually present with visual disturbances because of involvement of the cavernous sinus and without endocrine dysfunction. On CT, calcification is common, as is hyperostosis. Contrast enhancement is typically intense. Although CT is frequently useful in the diagnosis of meningiomas, MR is the examination of choice because of the multiplanar capability, and particularly with small lesions which may be difficult to see on CT. On MR, they are typically isointense to gray matter on T1 and isointense or mildly to moderately hyperintense on T2-weighted images and demonstrate homogeneous and relatively intense contrast enhancement (Figs. 12.14 and 12.15) in contrast to macroadenomas, where enhancement is not as intense and somewhat heterogeneous [33]. A dural tail is helpful although not diagnostic (Fig. 12.14b). The epicenter of the lesion is not in the sella, with the sella almost always being normal
12.3.2 Meningioma Meningiomas are the second most common primary brain tumors and the most common nonglial primary brain tumors. In the parasellar region in adults, they are the second most common tumors after pituitary macroadenomas [21]. Fifteen to twenty-five percent of all meningiomas occur in the parasellar region [21].
334
a
E.C. Bourekas et al.
c
b
Fig. 12.11 “Kissing carotids.” (a) Axial source images from an intracranial CT angiogram. (b) 3D reconstruction of a CT angiogram. (c) Axial FLAIR image of the head through the sella. All three images demonstrate ectatic carotids coursing medially
through the sella. This is critical information to the surgeon who is planning transsphenoidal surgery of the pituitary. Lack of this information can result in catastrophic and life-threatening hemorrhage at the time of surgery
and the pituitary gland easily identified (Figs. 12.13– 12.15). This is the major distinguishing factor from macroadenomas [33]. Another distinguishing factor is the fact that meningiomas typically encase and constrict the carotids (Fig. 12.14a), whereas macroadenomas may encase but typically do not constrict vessels. Invasion of the carotid artery is frequently seen, even when there is no evidence of narrowing of the artery [34]. Making the diagnosis preoperatively is important since meningiomas are treated via craniotomy rather than by a transphenoidal approach.
Meningiomas are mostly benign adult tumors that histologically arise from the meningothelial cells of the arachnoid mater and subsequently get attached with the adjacent dura mater. Grossly, they are usually circumscribed masses that have a soft to firm consistency and on cut sections a fish flesh appearance. Microscopically, meningiomas are composed of cells having mostly ovoid to spindled nuclei containing fine chromatin and scant cytoplasm. Meningiomas display several growth patterns, e.g., syncytial, transitional, fibroblastic etc, though most of them carry little prognostic significance.
335
12 Masses of the Sellar and Junxtasellar Region
a
b
Fig. 12.12 Pituitary adenoma pathology. (a) Gross specimen. Pituitary adenoma occupying the sellar region. (b) Sheets of monomorphic cells with eosinophilic cytoplasm showing loss of typical nested architecture of the adenohypophysis (H&E @ 400X)
The neoplasms display concentric cellular whorls and also psammoma bodies that are calcified degenerative forms of the cellular whorls (Fig. 12.17a, b).
12.3.3 Craniopharyngioma Craniopharyngiomas are the most common intracranial tumor of nonglial origin in children, comprising up to 10% of pediatric brain tumors [25, 35, 36]. They are formed from ectodermal remnants of Rathke’s pouch and are composed of a squamous epithelium. Cranio pharyngiomas can occur anywhere from the floor of the third ventricle (hypothalamus) to the pharyngeal tonsils, with 67% being found in the suprasellar region. These tumors have a bimodal incidence, with peaks in the first and fifth decades. Although they are histologically benign, they are behaviorally aggressive, invading adjacent structures and thus making resection difficult.
Recurrence is local, although meningeal seeding has been described [37]. Clinicopathologically, two distinct subtypes are recognized: the adamantinous, which tend to occur in children and the squamous-papillary variants, which tend to occur in adults [38]. Craniopharyngiomas are thought to be part of a continuum of ectodermally derived cystic epithelial lesions which include arachnoid cysts, Rathke’s cleft cysts, epidermoids, and dermoids [39]. They typically present because of mass effect on the chiasm and hydrocephalus, with visual disturbances, headaches, and pituitary and hypothalamic dysfunction. Endocrine deficiency is seen in 80%, with growth hormone deficiency noted in 75% [40]. On CT, craniopharyngiomas can be cystic, mixed cystic and solid, or solid and exhibit enhancement of the more solid portions. Hemorrhage is not an uncommon finding, particularly within cystic portions of the tumors. Craniopharyngiomas may grow to compress the optic chiasm superiorly, to displace the normal pituitary gland and stalk, invade the cavernous sinuses, and
336
E.C. Bourekas et al.
a
b
Fig. 12.13 Meningioma of the diaphragma sella: (a) Axial T2-weighted image in an elderly female reveals a large incidental lesion of the suprasellar cistern, which is isointense to gray matter. (b) Sagittal T1-weighted image demonstrates a large suprasellar mass, which extends into the sella and compresses the pituitary.
a
Fig. 12.14 Meningioma of the cavernous sinus: the patient is a 55-year-old male being evaluated for hearing loss. (a) Coronal T1-weighted image demonstrates a large mass lesion of the cavernous sinus, encasing and narrowing the carotid and invading the
The signal of the lesion is almost equal to that of the cortex. The imaging characteristics and the fact that the pituitary can clearly be identified and is separate from the lesion are consistent with a meningioma and allow for differentiation from a macroadenoma, which would be the other major differential diagnosis
b
sella. (b) High-resolution axial T1-weighted image demonstrates intense and uniform enhancement of the meningioma, which extends posteriorly along the tentorium and medially along the clivus in a “dural tail” fashion, indicative of a meningioma
337
12 Masses of the Sellar and Junxtasellar Region
a
b
c
Fig. 12.15 Meningioma of the clivus and petrous apex: the patient is a 65-year-old with ataxia. (a) Sagittal T1-weighted image without contrast shows a large hypointense mass lesion along the clivus causing compression of the brain stem and erosion of the clivus. (b) Axial T2-weighted image shows the lesion
involving the petrous apex and clivus and compressing the pons. The lesion is isointense to gray matter. (c) Axial T1-weighted image with contrast demonstrates uniform and moderately intense enhancement, characteristic of a meningioma
even to encase or occlude the carotid arteries. CT is the examination of choice for the evaluation of calcification which is seen in 87% of cases [41] (Figs. 12.18, 12.19a and 12.21a). The imaging characteristics of craniopharyngiomas on MRI are variable, reflecting the wide range of components histologically composing these tumors. The tumors may be cystic, mixed cystic, and solid or
primarily solid (Figs. 12.19–12.22). High signal intensity on T1- and T2-weighted images is seen in cysts with high cholesterol content or with subacute hemorrhage (Figs. 12.15 and 12.17). Craniophar yngiomas can also be of low signal intensity on T1-weighted images if the cyst contains a large amount of keratin [42]. Fluid levels can be seen in cystic regions. Adamantinous craniopharyngiomas tend to be
338
a
E.C. Bourekas et al.
b
c
Fig. 12.16 Meningioma of the planum sphenoidale: (a) axial T2-weighted image in a 66-year-old male demonstrates a mass lesion in the anterior cranial fossa along the midline, which is essentially isointense to gray matter. (b) Sagittal T1-weighted image without contrast shows an isointense lesion arising from
the region of the planum sphenoidale. (c) Axial T1-weighted image postcontrast shows intense and uniform enhancement. The imaging findings are classic for a meningioma of the planum sphenoidale
339
12 Masses of the Sellar and Junxtasellar Region
a
b
Fig. 12.17 Meningioma pathology. (a) Concentric cellular whorls formed by the meningothelial cells (H&E @ 600×). (b) Calcified psammoma body (H&E @ 400×)
a
Fig. 12.18 Craniopharyngioma. (a) Axial CT of the head without contrast with (b) bone windowing. This 20-year-old male presented with complaints of visual problems. He was initially treated with steroids for presumed optic neuritis with improvement of
b
symptoms. The CT reveals a large hyperdense mass lesion of the suprasellar cistern with extension into the anterior and middle cranial fossa. There is coarse calcification associated with the lesion, which is fairly characteristic for a craniopharyngioma on CT
340
a
c
Fig. 12.19 Craniopharyngioma. This is a 10-year-old male with acute onset ataxia, agitation, dysconjugate gaze, lethargy, nausea, and vomiting. (a) Axial CT without contrast demonstrates a large hyperdense mass of the suprasellar cistern with calcifications, which results in hydrocephalus. (b) Axial T1-weighted image without contrast shows the mass with areas of increased signal likely representing cysts with proteinaceous or fatty material. (c) Axial T2-weighted image shows the mass to be very heteroge-
E.C. Bourekas et al.
b
d
neous with areas of high signal representing cysts and areas of very low signal representing calcifications. (d) Coronal T1- and (e) sagittal T1-weighted images both after contrast demonstrate a very heterogeneous mass with solid and cystic areas and variable enhancement, expanding the sella and extending into the suprasellar cistern and interpeduncular cistern, causing compression of the third ventricle and resulting in hydrocephalus. All of the above findings are characteristic of a craniopharyngioma
341
12 Masses of the Sellar and Junxtasellar Region
Fig. 12.19 (continued)
primarily cystic or mixed cystic-solid lesions that tend to occur in children, whereas squamous-papillary subtypes tend to be predominately solid or mixed solid-
a
Fig. 12.20 Craniopharyngioma: This is a 69-year-old male who presented with visual problems. (a) Coronal T1-weighted image without contrast demonstrates a primary cystic lesion of the sella and suprasellar region with no involvement of the parasellar region. (b) Coronal T1-weighted image after contrast shows
cystic and occur in adults [38]. Distinguishing between the two has a prognostic significance since adamantinous tumors tend to recur. MRI can be helpful in distinguishing between the two, with encasement of vessels, a lobulated shape, and the presence of hyperintense cysts favoring adamantinous tumors and a round shape, presence of hypointense cysts, and a predominately solid appearance being seen with squamous-papillary tumors [38]. Calcification, encasement of vessels, and recurrence favor adamantinous tumors [38]. Pre operative differentiation from arachnoid cysts, Rathke’s cleft cysts, epidermoids, and dermoids can be difficult. Calcification and solid components are features more commonly seen with craniopharyngiomas [41]. In adults, calcified aneurysms must be part of the differential diagnosis of calcified lesions in the sellar, parasellar, and suprasellar region. Treatment is primarily surgical, with the efficacy of radiotherapy being well documented. Gamma knife radiosurgery may also play a role, although there are no specific guidelines [43]. Recurrence-free survival after total resection is 86.9% at 5 years, but falls to 48.8% with subtotal resection [44].
b
mostly peripheral enhancement of the cystic lesion. (c) Sagittal T1-weighted image postcontrast demonstrates that the lesion is solid and cystic without expansion of the sella. The optic chiasm is not visualized on any of the images and is probably involved/ compressed by tumor accounting for the patient’s symptoms
342
E.C. Bourekas et al.
c
the site of origin in most cases, since both the chiasm and the hypothalamus are involved irrelevant of the site of origin [49]. For this reason, both are discussed as one entity. Chiasmatic gliomas may occasionally demonstrate extension along the optic tracts or optic nerves thus indicating their site of origin (Fig. 12.26c). Mean age of presentation is 5 years [47]. Patients usually present with visual symptoms, headaches, and endocrine abnormalities. Endocrine abnormalities occur in 42% with growth hormone deficiency being the most common [25, 47]. MR is the examination of choice. The lesions tend to be solid with microcyst formation. They are usually isoor hypointense on T1-weighted images, hyperintense on T2-weighted images, and demonstrate enhancement with contrast (Figs. 12.24–12.26). The sella is usually normal [25, 47] (Fig. 12.25). Significant morbidity and mortality is associated with treatment which may involve surgery, chemotherapy, and/or radiation [50]. Chemotherapy may be used to postpone treatment until after the age of 5 years, which may reduce neurological morbidity [45]. Chiasmatic/ hypothalamic gliomas are more aggressive in the very young and in adults [45, 50, 51]. In NF-I the course is indolent [45]. Overall, sixty percent eventually relapse [45]. Survival is 93% at 5 years and 74% at 10 years with chiasmatic lesions having a 19 year 44% survival [46]. The prognosis of chiasmatic gliomas is worse than with optic nerve gliomas [46]. As previously stated, gliomas of the hypothalamus and optic chiasm are usually pilocytic astrocytomas which are low-grade, WHO grade I tumors. Grossly, the neoplasms are commonly well circumscribed and often cystic. Microscopically, the tumors are biphasic, showing loose reticulated as well as compact and dense regions. The compact regions display elongated cells having bipolar, hair-like processes (piloid cells) and Rosenthal fibers that are beaded, corkscrew-shaped hyaline bodies (Fig. 12.27). The loose areas often show eosinophilic granular bodies or protein droplets.
Fig. 12.20 (continued)
Microscopically, the adamantinomatous craniopharyngioma consists of nests and cords of squamous cells that assume a more columnar appearance in the peripheral layers. The inner cells display a loose, spongy reticulum. The cysts are filled with so-called “wet” keratin that is shed by the squamous cells. A cholesterol-rich fluid that resembles motor oil is also encountered within the cystic cavities. Dystrophic calcification is common (Fig. 12.23). The rare papillary variant, on the other hand, lacks the keratin, calcification, and the liquid cyst content. This is usually a solid tumor that demonstrates a papillary architecture.
12.3.4 Chiasmatic and Hypothalamic Gliomas Gliomas of the hypothalamus and optic pathways represent 5% of pediatric intracranial tumors [45] with 60% involving the optic chiasm and hypothalamus [46]. These represent 25–30% of pediatric neoplasms of the suprasellar region [47]. The vast majority represent slow growing pilocytic astrocytomas, although malignant gliomas and in particular glioblastoma multiforme may occur especially in adults [48]. Thirty-three percent of patients with chiasmatic/hypothalamic gliomas have neurofibromatosis type I (NF-I) [45, 47]. It is almost impossible to determine
12.3.5 Germinoma The sellar and suprasellar region is the second most common location of these germ cell tumors after the pineal region. Other germ cell tumors such as teratomas, embryonal carcinomas, choriocarcinomas, and mixed tumors are much less common intracranially, particularly in the
343
12 Masses of the Sellar and Junxtasellar Region
a
b
d
c
Fig. 12.21 Craniopharyngioma: the patient is a 27-year-old male. (a) Axial CT without contrast and (b) Coronal CT postcontrast through the sella demonstrate a large cystic mass of the sella and primarily suprasellar region with calcifications at the margins and peripheral enhancement, resulting in hydrocephalus. (c) Sagittal T1-weighted image without contrast demonstrates a large cystic mass, of increased signal, in the suprasellar cistern and sella, compressing the hypothalamus and third ven-
tricle and resulting in hydrocephalus. The high signal within the cyst suggests highly proteinaceous fluid or fluid with high lipid content. (d) Axial T2-weighted image shows the cystic lesion with areas of low signal at the periphery on the left representing calcification. (e) Axial T1-weighted image postcontrast demonstrates no significant enhancement of this cystic lesion. This however is difficult to determine given the increased signal of the cyst
344
E.C. Bourekas et al.
Fig. 12.23 Craniopharyngioma pathology. Squamous epithelium-lined cystic structures containing amorphous “wet” keratin (H&E @ 100×)
Fig. 12.21 (continued)
sellar and suprasellar region. Synchronous pineal and suprasellar lesions occur in 6–12% of germinomas and are considered diagnostic [52]. Primary suprasellar a
germinomas have no sexual predilection in contrast to pineal germinomas which show a male predominance [25]. The clinical presentation often includes hypopituitarism, diabetes insipidus, and visual disturbances. On MR, the mass appears as a well-marginated, round or lobulated, in- or homogeneous tumor with prolonged T1 and T2 relaxation times, which strongly enhances after gadolinium administration. The presence of these imaging findings along with the presence of diabetes insipidus and a suprasellar mass is a strong clue b
Fig. 12.22 Craniopharyngioma. (a) Sagittal T1-weighted image shows a solid, isointense suprasellar mass. (b) Coronal postcontrast T1-weighted image shows intense and homogeneous enhancement
345
12 Masses of the Sellar and Junxtasellar Region
a
b
c
d
Fig. 12.24 Chiasmatic/hypothalamic glioma. The patient is a 6-year-old female with visual problems. (a) Axial T1-weighted image without contrast reveals a large hypointense lesion of the suprasellar and interpeduncular cistern, with resultant hydrocephalus. (b) Axial T2-weighted image demonstrates the hyperintense lesion causing hydrocephalus. (c) Axial T1-weighted image postcontrast shows intense enhancement with some areas
of nonenhancement. (d) Sagittal T1-weighted image postcontrast demonstrates intense contrast enhancement with areas of nonenhancement representing necrosis or cyst formation. The sella, which is typically normal with optic/chiasmatic gliomas, is involved in this case and expanded. (e) 3D time-of-flight MRA shows the close relationship to the cerebral vasculature which appears normal
346
Fig. 12.24 (continued)
to the diagnosis of germinoma [53] (Figs. 12.28 and 12.29). Biopsy is necessary prior to treatment except in the case of synchronous suprasellar and pineal lesions.
a
Fig. 12.25 Chiasmatic/hypothalamic glioma: this 12-year-old female presented with visual complaints. (a) Axial T1-weighted image postcontrast demonstrates a large, enhancing mass lesion of the suprasellar cistern with linear enhancement along the left aspect of the mass which may represent enhancement of the optic chiasm
E.C. Bourekas et al.
The lesions are extremely radiosensitive, with over 90% of patients being effectively treated with radiation therapy alone. Germinomas are however also chemosensitive, with recent reports suggesting that the dose and volume of radiation required can be lessened with the addition of adjuvant chemotherapy [54]. Radical resection offers no benefit over biopsy, making the preoperative diagnosis very important in decision making [55]. Germinomas may secrete beta-human chorionic gonadotropin or alpha-fetoprotein, which can be detected in the CSF or blood and can aid in the preoperative diagnosis. Prognosis is good with 10- and 20-year survival rates of 92.7 and 80.6% respectively [56]. Germinomas are believed to arise from the developmentally derived ectopic rests of germ cells. The tumors are histologically similar to the seminoma arising in the testis. Germinomas can extend into the CSF and disseminate widely. Grossly, the tumor is circumscribed with a pale, fleshy, and homogenous stroma. Microscopically, the tumor is composed of large polyhedral cells having a large central nucleus with a
b
thus indicating that this is arising from the chiasm and not the hypothalamus. (b) Sagittal T1-weighted image postcontrast shows the large somewhat heterogeneously enhancing mass with a cystic area anteriorly and inferiorly. The pituitary is normal in the sella. The findings are rather characteristic for a chiasmatic glioma
347
12 Masses of the Sellar and Junxtasellar Region
a
b
c
Fig. 12.26 Chiasmatic-hypothalamic pilocytic astrocytoma in an 11-year-old patient with NF1. (a) Axial T1-weighted image shows an isointense suprasellar mass. (b) Coronal postcontrast
T1-weighted image shows intense and homogeneous enhancement of the mass. (c) On axial FLAIR image, the tumor appears hyperintense with extension along the optic pathways
348
prominent nucleolus. The cells are arranged in lobules demarcated by fibrous septae. Within this fibrous septae, variable numbers of inflammatory cells, predominantly lymphocytes, are present (Fig. 12.30).
Fig. 12.27 Chiasmatic and hypothalamic glioma pathology. The tumor is biphasic, showing loose reticulated as well as compact and dense regions. The compact regions display elongated cells having bipolar, hair-like processes (piloid cells) and Rosenthal fibers that are beaded, corkscrew-shaped hyaline bodies. The loose areas often show eosinophilic granular bodies or protein droplets (H&E @400×)
a
Fig. 12.28 Germinoma: (a) sagittal T1-weighted image in a young male with diabetes insipidus demonstrates a large mass lesion of the sella and suprasellar region. (b) Coronal T1-weighted
E.C. Bourekas et al.
12.3.6 Epidermoids and Dermoids Epidermoids and dermoids are uncommon, slow-growing masses that account for 1% of all intracranial neoplasms [57]. These lesions are similar in their development, histology, behavior, and imaging and for this reason are discussed together. As previously discussed, epidermoids and dermoids along with Rathke’s cleft cysts, arachnoid cysts, and craniopharyngiomas are considered by some to represent a continuum of ectodermally derived cystic epithelial lesions [39]. Both lesions are generally considered developmental/congenital masses rather than neoplastic, arising from ectodermal heterotopia. Both cysts are lined with stratified squamous epithelium, with dermoids adding mesodermal elements such as hair, sebaceous, and sweat glands. Epidermoids are slightly more common than dermoids intracranially. They typically spread along the basal surfaces, with the cerebellopontine angles being the most common location, followed by the parasellar [58–60]. They are extraaxial lesions with only 1.5% being intracerebral [61]. They are overwhelmingly benign, although they can rarely be malignant. Average age of presentation is 37.3, with a male to female ratio of 3:2 [57]. The symptomatic onset is generally slow, lasting for 2 years or more, although for suprasellar
b
image postcontrast shows moderately intense contrast enhancement, with evidence of left cavernous sinus involvement
349
12 Masses of the Sellar and Junxtasellar Region
a
b
c
Fig. 12.29 Malignant germ cell tumor in a 19-year-old patient with increased b-chorionic gonadotropin. (a) Sagittal T1-weighted image shows a low intensity suprasellar mass. (b) On axial
T2-weighted image the mass demonstrates high signal intensity. (c) On sagittal postcontrast T1-weighted image the mass shows intense and inhomogeneous enhancement
lesions it is much shorter [61]. Presenting symptoms may include headaches, visual problems, cranial nerve symptoms, and seizures, which typically indicate rupture. Rupture can produce aseptic meningitis, which can be lethal although not necessarily so. Epidermoids on CT appear as hypodense masses, with irregular borders and rare contrast enhancement. Dense lesions have been reported [62] and calcification
is occasionally seen [60]. On MR, they are typically of low signal on T1- and of increased signal on T2-weighted images, following that of CSF on all pulsing sequences [57]. They can demonstrate increased signal on T1-weighted images, which is due to high lipid content [60]. A classic differentiating feature from arachnoid cysts is the presence of restricted diffusion on diffusion weighted images (Fig. 12.31).
350
Dermoids are midline lesions, occurring in the parasellar, frontobasal region or posterior fossa [63]. Average age of presentation is 36.2 with a male to female ratio of 3:1 [57]. The complications of dermoids are similar to epidermoids. They can present with headaches, seizures, meningeal signs, and TIAs
Fig. 12.30 Germinoma. Large polyhedral cells with big nucleus and prominent nucleolus. Lymphocytic infiltration is present separating groups of cells (H&E @ 400×)
a
Fig. 12.31 Epidermoid. (a) Axial T1-weighted image without contrast demonstrates a hypointense lesion of the suprasellar cistern that cannot be demarcated from the CSF. (b) Axial T2-weighted image shows that the lesion is hyperintense, similar to CSF. (c) Axial T1-weighted image with contrast demon-
E.C. Bourekas et al.
[63, 64]. Most of these symptoms are indicative of rupture, which produces a chemical or aseptic meningitis and which can be lethal [63, 65]. The CT appearance of dermoids is characterized by the presence of the fatty component or calcification (Fig. 12.32a). Their MR appearance depends on the amount of fat present, although generally, they are of increased signal on both T1- and T2-weighted images [57] (Fig. 12.32). CT or MR can make the diagnosis of rupture although MR is the preferred preoperative study [64]. Treatment is surgery, with 86% being in good or excellent condition postoperatively. The 20-year survival of epidermoids is 92.8%, with good survival even with recurrence [57]. Epidermoids have a classic mother-of-pearl appearance at surgery. Dermoid cysts are uncommon and rarer than epidermoids. Grossly, the tumors are well-demarcated, cystic lesions with varying wall thickness. Papillary projections and tangled masses of hairs are usually observed within the cyst cavity. Microscopically, varying amounts of cellular elements from the three germ lines are seen, most commonly, stratified squamous epithelium, hair follicles, sebaceous glands, neural tissue, glandular structures, and rarely, bone and cartilage. Epidermoid cysts, on the other
b
strates no enhancement. At this point, the differential diagnosis for this lesion includes arachnoid cyst, Rathke’s cleft cyst, and epidermoid. (d) Axial diffusion-weighted image demonstrates restricted diffusion (hyperintensity) consistent with an epidermoid
351
12 Masses of the Sellar and Junxtasellar Region
c
d
Fig. 12.31 (continued)
a
Fig. 12.32 Dermoid. (a) Axial CT without contrast demonstrates a mass of the suprasellar cistern, posterior to the optic chiasm, with calcifications and fat suggesting a dermoid. (b) Midline sagittal T1-weighted image shows that the lesion is
b
hyperintense consistent with fat. (c) On this axial T2-weighted image the lesion is not well seen. (d) Axial diffusion-weighted image demonstrates no restricted diffusion and (e) axial T1-weighted image with contrast demonstrates no enhancement
352
c
Fig. 12.32 (continued)
E.C. Bourekas et al.
d
353
12 Masses of the Sellar and Junxtasellar Region
hand, are not uncommon. These are well-encapsulated lesions of variable size and smooth and lobulated with a mother-of-pearl sheen. Microscopically, the cyst lining consists of stratified squamous epithelium and the cyst contains lamellar keratin flakes that fills the cyst cavity (Fig. 12.33).
Fig. 12.33 Epidermoid cyst pathology. Squamous epithelium-lined cystic structure containing lamellar keratin material (H&E @ 200×)
a
Fig. 12.34 Rathke’s cleft cyst. (a) Sagittal T1-weighted image without contrast demonstrates a cystic lesion of the sella and suprasellar region. (b) Axial T2-weighted image confirms that the lesion is cystic. (c) Axial FLAIR image shows the lesion to
12.3.7 Rathke’s Cleft Cyst Rathke’s cleft cysts are nonneoplastic epithelial cysts that arise from remnants of Rathke’s pouch from the pars intermedia. They usually occur in the midline of the anterior or superior sella, causing anterior displacement of the infundibulum. Rathke’s cleft cysts are usually intrasellar cysts (Fig. 12.34) that may be asymptomatic, although they frequently present with endocrine abnormalities, headaches, and visual field defects, especially when there is suprasellar extension, which is usually seen [41]. Mean age of presentation is 38 years, with a female predominance of 2:1 [25]. Histologically, they have a single row of cuboidal or columnar epithelial lining in contrast to craniopharyngiomas, which have a squamous epithelial lining. Frequently, these cystic lesions are difficult to differentiate by imaging, and even histologically from other epithelial cystic lesions of the sellar and suprasellar region as previously noted. On CT, the sella may be normal or slightly expanded. The cysts can usually be seen as an area of lower attenuation that is similar to CSF. Calcification is seen in
b
be hypointense and (d) Axial T1-weighted image with contrast demonstrates some peripheral enhancement along the posterior margin confirming that this is a Rathke’s cleft cyst and distinguishing it from an arachnoid cyst which would not enhance
354
c
E.C. Bourekas et al.
d
Fig. 12.34 (continued)
approximately 13% of cases usually peripherally, in contrast to craniopharyngiomas in which calcification is seen in 87% of cases [41]. Peripheral rim enhancement may be seen [21] On MRI, the cysts are usually isointense with CSF on all pulse sequences. Occasionally, they may have a more unusual signal due to varying cyst fluid composition. Cyst wall enhancement can be seen [25] (Fig. 12.34). Cyst wall biopsy and aspiration are considered to be curative [39]. Histologically, the cysts are filled with mucoid material and are lined by simple ciliated columnar epithelium interspersed by mucin producing goblet cells. The epithelial lining maybe attenuated and may also not be continuous due to stretching. Sometimes, adenohypophyseal cells are seen attached to the walls due to close proximity of the adenohypophysis (Fig. 12.35).
12.3.8 Arachnoid Cysts These nonneoplastic cysts are a rare cause of cystic lesions in the sellar region. As noted previously, they are thought to be part of a continuum of epithelial cysts at the more benign end of the spectrum behaviorally. They tend to present at an older age, usually the fifth decade, with headaches, visual field defects, and
Fig. 12.35 Rathke’s cleft cyst pathology. Benign epitheliumlined cyst wall. Small groups of adenohypophyseal cells are present within the wall (H&E @ 400×)
impotence [41]. On MR, these sellar and/or suprasellar cysts follow CSF signal on all pulsing sequences. They are well defined, with no calcification and no enhancement [25, 41] (Fig. 12.36). Grossly, the cyst wall is thin and translucent. The cysts contain clear, colorless fluid that occasionally may be xanthochromic. Microscopically, the cyst wall is composed of an inner layer of arachnoid cells and an outer collagenous layer (Fig. 12.37).
355
12 Masses of the Sellar and Junxtasellar Region
a
b
c
Fig. 12.36 Arachnoid cyst in a young patient being evaluated for headaches. (a) Axial CT through the suprasellar cistern shows enlargement of the suprasellar cistern raising a question of a cystic mass. (b) Sagittal T1-weighted image without contrast demonstrates a large cystic mass of the suprasellar cistern, compressing the pituitary inferiorly, the third ventricle superiorly, and extending inferiorly into the prepontine cistern with displacement of the basilar artery posteriorly. (c) Coronal
d
T2-weighted image demonstrates the large hyperintense cystic lesion compressing the third ventricle and resulting in dilatation of the lateral ventricles with extension of the lesion into the left middle cranial fossa. (d) Coronal T1-weighted image postcontrast shows no enhancement of the lesion ruling out a Rathke’s cleft cyst. The diffusion-weighted image (not shown) demonstrated no restricted diffusion, thus excluding an epidermoid and making arachnoid cyst the likely diagnosis
356
Fig. 12.37 Arachnoid cyst pathology. The image shows a cyst wall partially lined by meningothelial cells (H&E @ 400×)
12.3.9 Neurosarcoidosis Sarcoid is a chronic, granulomatous disease which involves the CNS in 5–10% of cases [66]. It has a wide
a
Fig. 12.38 Sarcoidosis: This young black female with known sarcoid, presented complaining of headaches and visual complaints. (a) Coronal T1-weighted image postcontrast demonstrates a small mass lesion of the suprasellar cistern, adjacent to
E.C. Bourekas et al.
variety of manifestations in the CNS, with a propensity to involve the basilar cisterns and the suprasellar region. The hypothalamic–pituitary region is involved in 25% of cases [67]. On MR, a basal or diffuse leptomeningitis may be seen with secondary involvement of the optic chiasm, hypothalamus, floor of the third ventricle, pituitary, infundibulum, or underlying brain parenchyma (Fig. 12.38). Sarcoidosis can mimic almost any other lesion [68, 69], even presenting as a large suprasellar mass simulating a neoplasm [70]. CNS involvement without systemic disease is unusual. Macroscopically, the lesions range from small nodular thickenings to large globular masses. Microscopically, the lesions are circumscribed and composed of large number of epithelioid histiocytes (transformed macrophages) and scattered multinucleated giant cells. Mature reactive lymphocytes are also seen surrounding the lesions (Fig. 12.39). The finding of minute foci of necrosis is not uncommon, though large areas of necrosis should raise the suspicion of tuberculosis. Older, healed lesions display less inflammatory cells as they are replaced by collagen.
b
the optic chiasm. (b) Coronal T1-weighted image postcontrast just anterior to (a) shows evidence of meningeal and leptomeningeal disease consistent with sarcoid
357
12 Masses of the Sellar and Junxtasellar Region
hypothalamus and is exophytic. The nodular mass (<1 cm) hangs into the suprasellar cistern adjacent to the mammillary bodies. On T1-weighted images, the signal is isointense with normal brain and on T2-weighted images there is mild hyperintensity or isointensity. These lesions usually do not enhance with contrast administration (Fig. 12.40). Microscopically, they are composed of mature neurons arranged in clusters within abundant neuropil. The glial cells within the neuropil also appear normal.
Fig. 12.39 Neurosarcoidosis. Brain parenchyma with welldemarcated, nonnecrotizing granulomas (H&E @ 200×)
12.3.10 Hamartomas of the Tuber Cinereum Hypothalamic hamartomas of the tuber cinereum usually presents with precocious puberty or gelastic seizures in a young child [71]. It is important to differentiate this lesion from a hypothalamic glioma, because the prognosis for hamartoma is much more favorable. Imaging is best with thin-section coronal and sagittal MRI. The findings are usually characteristic: The mass arises from the undersurface of the a
Fig. 12.40 Hamartoma of the tuber cinereum. Sagittal (a) and coronal (b) T1-weighted images show an isointense pedunculated mass in the region of the tuber cinereum (arrows). On
12.3.11 Schwawnnomas and Neurofibromas Schwannomas, which are tumors derived from the myelin sheath of peripheral nerves, can be found involving the cranial nerves within the cavernous sinus and parasellar regions. In general, pituitary function is not affected; however, often the cranial nerves III, IV, V, and VI are affected within the cavernous sinuses or in the suprasellar and prepontine cisterns. Schwannomas may remodel the foramina of the skull where the individual nerves exit. When multiple lesions are seen, neurofibromatosis should be considered. On CT, schwannomas are usually hyperdense lesions with homogeneous enhancement, and they may be hard to differentiate from meningiomas. b
T2WI (c) the mass remains isointense to the brain parenchyma (arrow). (d) After the administration of contrast material the lesion does not enhance
358
c
E.C. Bourekas et al.
d
Fig. 12.40 (continued)
Fig. 12.41 Schwannoma. Neoplastic cells with elongated, spindled nuclei showing nuclear palisades (H&E @ 200×)
Fig. 12.42 Neurofibroma. Spindle cells with thin, wavy nuclei in a loose, myxoid background. Eosinophilic collagen bundles are present in between the cells (H&E @ 200×)
On MRI, they may be isointense or hyperintense to gray matter on T1-weighted images, and they enhance homogeneously. Histologically, they are WHO grade I lesions. Grossly, they are encapsulated, lobulated masses and may be partially cystic. Microscopically, they are composed of hypo and hypercellular regions (Antoni A and Antoni B areas respectively). Individual tumor cells display an elongated, spindled nucleus with tapered ends. Nuclear palisades created by alignment of nuclei alternated by nuclear free zones are a histological hallmark of schwannomas (Fig. 12.41).
Neurofibromas are rare intracranially. Solitary tumors are circumscribed, globoid masses that diffusely enlarge the nerve they are arising from. Microscopically, the tumors are composed of spindle cells with thin, wavy nuclei that are arranged in loose bundles. The background stroma is rich in collagen and mucopolysaccharide. The cellular elements consist of Schwann cells and varying proportions of fibroblasts and perineurial cells. Presence of remnants of myelinated nerve fibers within the tumor, unlike schwannomas, is a common finding (Fig. 12.42).
359
12 Masses of the Sellar and Junxtasellar Region
12.3.12 Aneurysms Aneurysms can occur in the sella and more commonly in the parasellar and suprasellar region arising from the internal carotid artery in the cavernous or paraclinoid region or from the anterior cerebral artery arising from the anterior communicating artery. Their significance lies in that these “do not touch” lesions
must be distinguished from other parasellar lesions in order to avoid the potentially catastrophic consequences of a biopsy. On MRI, aneurysms usually demonstrate signal voids on both T1 and T2-weighted images although they can have variable signal depending on flow characteristics, turbulence, and the presence and stage of thrombus noted within the aneurysm (Fig. 12.43).
a
b
c
d
Fig. 12.43 Aneurysm: The “DO NOT BIOPSY” lesion. (a) Cor onal T1-weighted image without contrast demonstrates a lesion of the cavernous/parasellar region that encroaches on the sella medially and that on (b) axial and (c) coronal T2-weighted images reveals flow voids within the lesion indicating that this is
an aneurysm. (d) Coronal T1-weighted image postcontrast shows vascular enhancement. Aneurysms can have variable signal characteristics depending on rapidity of flow, turbulence, and the stage of thrombus that may be within the aneurysm
360
12.3.13 Pituitary Hyperplasia Pituitary hyperplasia represents an asymptomatic, physiologic enlargement of the pituitary gland that can be seen at birth, during puberty, pregnancy, and with hypothyroidism. Instead of the normal
a
c
Fig. 12.44 Pituitary hyperplasia. This young female was being imaged for headaches and noted to have an enlarged pituitary which led to evaluation for an adenoma, including dedicated pituitary imaging. (a) Sagittal T1- and (b) coronal T2-weighted images confirm an enlarged pituitary with a convex upper margin that is in close proximity to the optic chiasm. After the
E.C. Bourekas et al.
concave upper margin, the pituitary has a convex upper margin. On CT, no calcifications are noted and there is normal contrast enhancement of the enlarged gland. MRI demonstrates an enlarged gland with normal imaging characteristics and enhancement (Fig. 12.44).
b
d
administration of contrast, (c) sagittal T1- and (d) coronal T1-weighted images demonstrate uniform enhancement of the enlarged pituitary. Hormonal analysis revealed no abnormalities that would suggest an active adenoma and the pituitary has remained unchanged for several years
361
12 Masses of the Sellar and Junxtasellar Region
12.3.14 Lymphocytic Hypophysitis Lymphocytic hypophysitis (LH) is an uncommon intrasellar lesion characterized by lymphocytic infiltration of the adenohypophysis. Evidence suggests that the cause is autoimmune, and the symptoms are usually related to either the mass effect or endocrine dysfunction. Lymphocytic hypophysitis has been rarely described in the setting of other simultaneous pathological processes that involve the pituitary and sella turcica, and is postulated to arise from an intrinsic inflammatory response. It is predominantly identified in women in the third and fourth decades of life. It often has a marked temporal relation to late pregnancy or the postpartum period. The nature of this relationship is not understood, but may stem from the marked hyperestrogenemia and increased pituitary blood flow that occurs during pregnancy. The most common clinical symptoms are headache and visual impairment. The diagnosis can be challenging in many cases, because distinction from pituitary adenomas and other sellar masses is not obvious. CT and MRI demonstrate a mass in 75–90% of patients, in case series presenting as LH. Even in cases where the initial scan may be normal, repeated imaging evaluations months later might evidence a mass-like image [32]. A marked contrast enhancement is considered to be common to other inflammatory processes of the pituitary gland as well [55]. However, a triangular enhancement of the enlarged anterior pituitary(reflecting extension of the process towards the pituitary stalk) with thickened nontapered infudibulum seem to be particularly specific to LH (Fig. 12.45). In the majority of LH loss of the posterior pituitary bright spot may be seen. The differential diagnosis includes pituitary hyperplasia, macroadenoma, sarcoid, and metastases. In pituitary hyperplasia the infudibulum is usually normal. Diabetes insipidus is common in LH and rare with adenomas. Pituitary involvement in sarcoidosis represents part of systemic disease. The history of a known primary tumor may be helpful in pituitary metastases [72–74].
12.3.15 Pituicytoma An uncommon tumor arising from pituicytes, a specialized glial cell population in neurohypophysis and
pituitary stalk, pituicytoma is a benign WHO grade I glioma. It has a peak incidence in the fifth decade of life. The most common clinical symptoms are visual disturbances or/and endocrine dysfunction. In some cases pituicytomas may be asymptomatic. On MRI, they are solid tumors arising from neurohypophysis or infundibulum, which appear iso-to hypointense on T1-weighted images, iso- to hyperintense on T2-weighted images, and demonstrate diffuse enhancement. The “bright spot” of the posterior lobe is usually absent (Fig. 12.46). The differential diagnosis includes pituitary adenoma, lymphocytic hypophysitis, pituitary hyperplasia, and metastases [75, 76].
12.3.16 Metastases Metastasis to the sellar, suprasellar, or parasellar regions may arise in the sphenoid bone or sinus, cavernous sinus, pituitary gland, hypothalamus, or surrounding soft tissues. Endocrine symptoms are uncommon with pituitary metastasis, but are often seen when the hypothalamus is involved. It may be difficult to distinguish a metastasis from a primary pituitary abnormality on the basis of imaging alone; however, the presence of bony destruction or the history of a known primary tumor may be helpful (Fig. 12.47). Lung (adenocarcinoma, small cell carcinoma) and breast (adenocarcinoma) are the two most common sources of metastasis. Grossly, these are usually circumscribed lesions that are soft to firm in consistency. Microscopically, they mimic the architecture of the primary neoplasm. Adenocarcinomas from breast show complex glandular architecture. Secondary lymphomas and leukemias may involve the region though usually they are meningeal based.
12.3.17 Infundibular Masses The thickness of the normal pituitary stalk averages 3.5 mm at the median eminence and 2.8 mm near its midpoint. The normal stalk enhances markedly on CT with contrast and on MRI with gadolinium. The most common clinical problem associated with disease of
362
E.C. Bourekas et al.
a
b
c
d
Fig. 12.45 Lymphocytic hypophysitis. Coronal (a) and sagittal (b) non contrast T1-weighted images show a pituitary mass with thickened nontapered infudibulum. Post contrast coronal (c) and sagittal (d) images show intense and homogeneous enhancement
the pituitary stalk is diabetes insipidus. When this is present, there is usually absence of the normal hyperintensity of the posterior pituitary noted on T1-weighted MRI. Diabetes insipidus may also occur as a result of transection of the pituitary stalk. The differential diagnosis of a thickened stalk includes: sarcoidosis, tuberculosis, histiocytosis X, germinoma, lymphoma, leukemia, metastases (Figs. 12.48– 12.50), and ectopic posterior pituitary. A thickened stalk
can also be due to an extension of a glioma within the hypothalamus. In patients with neurosarcoidosis and tuberculous infiltration of the stalk, the chest radiograph is generally abnormal and may be helpful in the differentiation from histiocytosis X. Clinically, patients with histiocytosis X may have skin lesions, otitis media, or bone lesions, in addition to interstitial lung disease [77].
363
12 Masses of the Sellar and Junxtasellar Region
a
b
c
d
Fig. 12.46 Pituicytoma. The patient is a 36-year-old female with headaches and visual disturbance. (a) Sagittal T1-weighted image without contrast, (b) sagittal T2-weighted, and (c) axial FLAIR images all demonstrate a mass of the infundibulum. (d) Sagittal
T1-weighted image with contrast shows thickening and enhancement of the mass, which after histologic evaluation proved to be a rare pituicytoma
364 Fig. 12.47 Metastatic ependymoma: this 3.5-yearold male child presented with lethargy and difficulty in walking. (a) Axial T1-weighted image without contrast demonstrates a large mass lesion of the interpeduncular and suprasellar cistern with hydrocephalus. (b) Coronal T1-weighted image postcontrast shows a large, irregular lesion involving the sella and suprasellar region and invading the hypothalamus. (c) Sagittal T1-weighted image of the lower spine reveals a large mass lesion of the conus, which proved to be an ependymoma. This is likely the primary tumor, with the suprasellar lesion representing metastases, although ependymomas arising in the suprasellar region can rarely occur, in which case, the spinal lesion is a drop metastasis
E.C. Bourekas et al.
a
c
b
12.3.18 Miscellaneous Masses Other entities that should be considered in the differential diagnosis of lesions of the sellar and junxtasellar region or those that have been reported in this region
are: aneurysms, lymphoma (Fig. 12.51), leukemia, teratoma, histiocytosis, chordoma, melanoma, nasopharyngeal carcinoma, mucoceles, lipoma (Fig. 12.52) hemangioma [78] (Fig. 12.53), pituitary astrocytoma [79], and xanthogranuloma [80]
365
12 Masses of the Sellar and Junxtasellar Region
a
b
c
d
Fig. 12.48 Sarcoid. The patient is a 30-year-old African American female with known sarcoidosis who presented complaining of weakness, fatigue, and weight loss. (a) Sagittal T1-weighted image without contrast shows a slightly thickened infundibulum which after the administration of contrast shows uniform enhancement and enlargement of the infundibulum on
(b) the coronal T1-weighted image postcontrast. (c) Note the enhancement of the infundibulum and the right optic nerve on this axial T1-weighted image postcontrast and with fat suppression. (d) Coronal T1-weighted image postcontrast and with fat suppression confirms the enhancement of the right optic nerve which can be seen with sarcoid
366
E.C. Bourekas et al.
a
b
c
d
Fig. 12.49 Histiocytosis X. This 3-year-old child has diabetes insipidus and growth hormone deficiency. Sagittal (a) and coronal (b) T1-weighted image without contrast shows no evidence of the bright posterior pituitary. After contrast, sagittal (c) and coronal (d) T1-weighted images show thickening and enhance-
ment of the pituitary stalk. The imaging findings of loss of the posterior pituitary bright spot on T1W and thickening and enhancement of the infundibulum along with the clinical presentation are classic for Langerhans histiocytosis X
367
12 Masses of the Sellar and Junxtasellar Region
a
a
b b
Fig. 12.50 Infundibular metastasis. (a) Coronal T1-weighted image through the sellar region demonstrates a small enhancing lesion of the infundibulum in a patient with lung carcinoma representing metastatic disease. (b) Axial T1-weighted image postcontrast higher through the brain reveals an enhancing dural lesion which could be a metastatic lesion or a meningioma
Fig. 12.51 Lymphoma. (a) Axial T1-weighted image shows a hypointense lesion of the suprasellar and interpenduncular cistern displacing the optic chiasm anteriorly. (b) On axial T2-weighted image the lesion appears isointense to the gray matter. (c, d) Postcontrast axial and coronal T1-weighted images show intense homogenous enhancement of the lesion. Biopsy proved a hypothalamic non-Hodgkin B-cell lymphoma
368
E.C. Bourekas et al.
c
d
Fig. 12.52 Suprasellar lipoma: midline sagittal T1-weighted image without contrast in a 16-year-old female being evaluated for headaches reveals a hyperintense lesion of the suprasellar cistern/hypothalamus, just posterior to the optic chiasm, representing an incidental lipoma
a
Fig. 12.51 (continued)
Fig. 12.53 Chiasmatic hemangioma: this 38-year-old male presented after an episode of vision loss. The week prior he had been experiencing headaches and visual disturbances. (a) Axial CT without contrast reveals a dense, rounded lesion of the suprasellar cistern, thought to represent an aneurysm. An angiogram was however negative. (b) Sagittal T1-weighted image without contrast demonstrates the lesion, involving the optic chiasm, with a focus of increased signal probably representing hemorrhage. (c) Coronal T1-weighted image after contrast shows minimal contrast enhancement of the chiasmatic lesion. Histologic evaluation revealed a cavernous hemangioma. The episode of vision loss was due to hemorrhage within the lesion representing chiasmatic apoplexy
12 Masses of the Sellar and Junxtasellar Region
b
c
Fig. 12.53 (continued)
References 1. Elster AD (1993) Modern imaging of the pituitary. Radiology 187:1–14 2. Johnsen DE, Ww W, Allen IS et al (1991) MR imaging of the sellar and junxtasellar regions. Radiographics 11:727–758 3. Miki Y, Matsuo M, Nishizawa S et al (1990) Pituitary adenomas and normal pituitary tissue: enhancement patterns on Gadopentetate-enhanced MR imaging. Radiology 177:36 4. Bartynski WS, Lin L (1997) Dynamic and conventional spinecho MR of pituitary microlesions. AJNR Am J Neuroradiol 18(5):965–972
369 5. Naidich MJ, Russell EJ (1999) Current approaches to imaging of the sellar region and pituitary. Endocrinol Metab Clin North Am 28(1):45–79 6. Miller DL, Doppman JL, Nieman LK et al (1990) Petrosal sinus sampling: discordant lateralization of ACTHsecreting pituitary microadenomas before and after stimulation with corticotropin-releasing hormone. Radiology 176:429 7. Graham KE, Samuel MH, Nesbit GM et al (1999) Cavernous sinus sampling is highly accurate in distinguishing cushing’s disease from the ectopic adrenocorticotropin syndrome and in predicting intrapituitary tumor location. J Clin Endocrinol Metab 84(5):1602–1610 8. Mamelak AN, Dowd CF, Tyrrell JB, McDonald JF, Wilson CB (1996) Venous angiography is needed to interpret inferior petrosal sinus and cavernous sinus sampling data for lateralizing adrenocorticotropin-secreting adenomas. J Clin Endocrinol Metab 81(2):475–481 9. Oliverio PJ, Monsein LH, Wand GS, Debrun GM (1996) Bilateral simultaneous cavernous sinus sampling using corticotropin-releasing hormone in the evaluation of cushing disease. AJNR 17:1669–1674 10. Freda PU, Wardlaw SL (1999) Clinical review 110: diagnosis and treatment of pituitary tumors. J Clin Endocrinol Metab 84(11):3859–3866 11. Asa SL (1999) The pathology of pituitary tumors. Clin North Am 28(1):13–43 12. Kovacs K, Scheithauer B, Horvath E, Lloyd R (1996) The world health organization classification of adenohypophysial neoplasms. Cancer 78(3):502–510 13. Mindermann T (1997) Letter to the editor: classification of pituitary adenomas. Acta Neurochir (Wien) 139:267–270 14. Pernicone PJ, Scheithauer BW, Sebo TJ (1997) Pituitary carcinoma. Cancer 79:804–812 15. Poussaint TY, Barnes PD, Anthony DC, Spack N, Scott RM, Tarbell NJ (1996) Hemorrhagic pituitary adenomas of adolescence. AJNR 17(10):1907–1912 16. Liuzzi A, Vittorio T, Pirro MT et al (1996) Nonfunctioning adenomas of the pituitary. Metabolism 45(9 suppl 1):80–82 17. De Boucaud L, Dousset V, Caillaud P, Viaud B, Guerin J, Caille JM (1999) Metastases from a pituitary ademonma: MRI. Neuroradiology 41(10):785–787 18. Brada M, Burchell L, Ashley S, Traish D (1999) The incidence of cerebrovascular accidents in patients with pituitary adenoma. Int J Radiation Oncology Biol Phys 45(3):693–698 19. Nilsson B, Gustavsson-Kadaka E, Bengtsson B, Jonsson B (2000) Pituitary adenomas in Sweden between 1958 and 1991: incidence, survival and mortality. J Clin Endocrinol Metab 85(4):1420–1425 20. Popovic V, Damjanovic S, Micic D et al (1998) Increased incidence of neoplasia in patients with pituitary adenomas. Clin Endocrinol 49:441–445 21. Simonetta AB (1999) Imaging of suprasellar and parasellar tumors. Neuroimaging Clin N Am 9(4):717–732 22. Majos C, Coll S, Aguilera C, Acebes JJ, Pons LC (1998) Imaging of giant pituitary adenomas. Neuroradiology 40: 651–655 23. Scotti G, Yu C, Dillon WP et al (1988) MR Imaging of cavernous sinus involvement by pituitary adenomas. AJR 151: 799–806
370 24. Levy RA, Quint DJ (1998) Giant pituitary adenoma with unusual orbital and skull base extension. AJR Am J Roentgenol 170(1):194–196 25. Freda PU, Post KD (1999) Differential diagnosis of sellar masses. Endocrinol Metab Clin North Am 28(1):81–117 26. Melmed S (1996) Acromegaly. Metabolism 45(8):51–52 27. Pinzone JJ, Katznelson L, Danila DC, Pauler DK, Miller CS, Klibanski A (2000) Primary medical therapy of micro- and macroprolactinomas in men. J Clin Endocrinol Metab 85(9):3053–3057 28. Lundin P, Bergstrom K, Nyman R et al (1992) Macro prolactinomas: serial MR imaging in long-term bromocriptine therapy. AJNR 13:1287 29. Yousem DM, Arrington JA, Zinreich JS, Kumar AJ, Bryan RN (1989) Pituitary adenomas: possible role of bromocriptine in intratumoral hemorrhage. Radiology 170(1):239–243 30. Randeva H, Schoebel J, Byrne J, Esiri M, Adams CBT, Wass JAH (1999) Classical pituitary apoplexy: clinical features, management and outcome. Clin Endocrinol 51:181–188 31. Sanno N, Ishii Y, Sugiyama M et al (1999) Subarachnoid hemorrhage and vasospasm due to pituitary apoplexy after pituitary function tests. Acta Neurochir 141:1009–1010 32. Giovanelli M, Losa M, Mortini P (1996) Surgical therapy of pituitary adenomas. Metabolism 45(8):115–116 33. Cappabianca P, Cirillo S, Alfieri A et al (1999) Pituitary macroadenoma and diaphragma sellae meningioma: differential diagnosis on MRI. Neuroradiology 41:22–26 34. Sen C, Hague K (1997) Meningiomas involving the cavernous sinus: histological factors affecting the degree of resection. J Neurosurg 87:535–543 35. Young SC, Zimmerman REA, Nowell MA et al (1987) Giant cystic craniopharyngiomas. Neuroradiology 29:468 36. Khafaga Y, Jenkin D, Kanaan I, Hassounah M, Shabanah MA, Gray A (1998) Craniopharyngioma in children. Int J Radiation Oncology Biol Phys 42(3):601–606 37. Gupta k, Kuhn MJ, Shevlin DW, Wacaser LE (1999) Metastatic craniopharyngioma. AJNR Am J Neuroradiol 20:1059–1060 38. Sartoretti-Schefer S, Wichmann W, Aguzzi A, Valavanis A (1997) MR differentiation of adamantinous and squamouspapillary craniopharyngiomas. AJNR Am J Neuroradiol 18(1):77–87 39. Harrison MJ, Morgello S, Post KD (1994) Epithelial cystic lesions of the sellar and parasellar region: a continuum of ectodermal derivatives? J Neurosurg 80:1018–1025 40. Lafferty AR, Chrousos GP (1999) Pituitary tumors in children and adolescents. J Clin Endocrinol Metab 84(12):4317–4323 41. Shin JL, Asa SL, Woodhouse LJ, Smyth HS, Ezzat S (1999) Cystic lesions of the pituitary: clinicopathological features distinguishing craniopharyngioma, Rathke’s cleft cyst, and arachnoid cyst. J Clin Endocrinol Metab 84(11):3972–3982 42. Pusey E, Kortman KE, Flannigan BD et al (1987) MR of craniopharyngiomas: tumor delineation and characterization. AJNR 8:443 43. Laws ER, Vance ML (1999) Radiosurgery for pituitary tumors and caniopharyngiomas. Neurosurg Clin N Am 10(2):327–336 44. Fahlbusch R, Honegger J, Paulus W, Huk W, Buchfelder M (1999) Surgical treatment of craniopharyngiomas: experience with 168 patients. J Neurosurg 90:237–250
E.C. Bourekas et al. 45. Janss AJ, Grundy R, Cnaan A et al (1995) Optic pathway and hypothalamic/chiasmatic gliomas in children younger than age 5 years with a 6-year follow-up. Cancer 75(4):1051–1059 46. Rodriguez LA, Edwards MS, Levin VA (1990) Management of hypothalamic gliomas in children: an analysis of 33 cases. Neurosurgery 26(2):242–246 47. Collet-Solberg PF, Sernyak H, Satin-Smith M, Katz LL, Sutton L, Molloy P, Moshang T Jr (1997) Endocrine outcome in long-term survivors of low-grade hypothalamic/ chiasmatic glioma. Clin Endocrinol (Oxf) 47(1):79–85 48. Barbaro NM, Rosenblum ML, Maitland CG, Hoyt WF, Davis RL (1982) Malignant optic glioma presenting radiologically as a “cystic” suprasellar mass: case report and review of the literature. Neurosurgery 11(6):787–789 49. Albert A, Lee BC, Saint-Louis L, Deck MD (1986) MRI of optic chiasm and optic pathways. AJNR Am J Neuroradiol 7(2):255–258 50. Alshail E, Rutka JT, Becker LE, Hoffman HJ (1997) Optic chiasmatic-hypothalamic glioma. Brain Pathol 7(2):799–806 51. Nishio S, Takeshita I, Fujiwara S, Fukui M (1993) Opticohypothalamic glioma: an analysis of 16 cases. Childs Nerv Syst 9(6):334–338 52. Sugiyama K, Uozumi T, Kiya K et al (1992) Intracranial germ cell tumor with synchronous lesions in the pineal and suprasellar regions: six cases and review of the literature. Surg Neurol 38:114–120 53. Kollias SS, Barkovich AJ, Edwards MS (1991) Magnetic resonance analysis of suprasellar tumors of childhood. Pediatr Neurosurg 17(6):284–303 54. Packer RJ, Cohen BH, Consy K (2000) Intracranial germ cell tumors. Oncologist 5(4):312–320 55. Sawamura Y, de Tribolet N, Ishii N, Abe H (1997) Management of primary intracranial germinomas: diagnostic surgery or radical resection. J Neurosurg 87(2):262–266 56. Matsutani M, Sano K, Takakura K et al (1997) Primary intracranial germ cell tumors: a clinical analysis of 153 histologically verified cases. J Neurosurg 863:446–455 57. Yaşargil MG, Abernathey CD, Sarioglu AÇ (1989) Micro surgical treatment of intracranial dermoid and epidermoid tumors. Neurosugery 23(4):561–567 58. Mori K, Handa H, Moritake K, Takeuchi J, Nakano Y (1982) Suprasellar epidermoid. Neurochirurgia (Stuttg) 25(4): 138–142 59. Yamakawa K, Shitara N, Genka S, Manaka S, Takakura K (1989) Clinical course and surgical prognosis of 33 cases of intracranial epidermoid tumors. Neurosurgery 24(4):568–573 60. Horowitz BL, Chari MV, James R, Bryan RN (1990) MR of intracranial epidermoid tumors: correlation of in vivo imaging with in vitro 13C spectroscopy. AJNR Am J Neuroradiol 11(2):299–302 61. Netsky MG (1988) Epidermoid tumors. Review of the literature. Surg Neurol 29(6):477–483 62. Braun IF, Naidich TP, Leeds NE, Koslow M, Zimmerman HM, Chase NE (1977) Dense intracranial epidermoid tumors. Computed tomographic observations. Radiology 122(3):717–719 63. Wilms G, Casselman J, Demaerel P, Plets C, De Haene I, Baert AL (1991) CT and MRI of ruptured intracranial dermoids. Neuroradiology 33(2):149–151
12 Masses of the Sellar and Junxtasellar Region 64. Smith AS, Benson JE, Blaser SI, Mizushima A, Tarr RW, Bellon EM (1991) Diagnosis of ruptured intracranial dermoid cyst: value MR over CT. AJNR Am J Neuroradiol 12(1):175–180 65. Cohen JE, Abdallah JA, Garrote M (1997) Massive rupture of suprasellar dermoid cyst into ventricles. Case illustration. J Neurosurg 87(6):963 66. Kumpe DA, Rao KCVG, Garcia JIH, Hechk AF (1979) Intracranial neurosarcoidosis. J Comp Assist Tomogr 3:324–330 67. Khalil MK, Arthus BP, Burnier MN (1996) Sarcoidosis of the sella turcica in association with bilateral sarcoidosis of the lacrimal glands. Can J Ophthalmol 31(1): 32–35 68. Hayes WS, Sherman JL, Stern BJ et al (1987) Magnetic resonance and CT evaluation of intracranial sarcoidosis. AJR 8:1043 69. Lexa FJ, Grossman RI (1994) MR of sarcoidosis in the head and spine: spectrum of manifestations and radiographic response to steroid therapy. AJNR 15:973 70. Bakshi R, Fenstermaker A, Bates V, Ravidhandran TP, Goodloe S, Kinkel WR (1998) Neurosarcoidosis presenting as a large suprasellar mass: magnetic resonance imaging findings. Clin Imaging 22(5):323–326 71. Hahn FJ, Leinbrock LG, Huseman CA, Makos MM (1988) The MR appearance of hypothalamic hamartoma. Neuro radiology 30:67
371 72. Lim S, Elston MS, Swarbrick MJ, Conaglen JV (2009) Lymphocytic hypophysitis with associated thyroiditis in a man with aseptic meningitis. Pituitary 12:375–379 73. Gutenberg A, Hans V, Puchner MJA et al (2006) Primary hypophysitis: clinical-pathological correlations. Eur J Endo crinol 155(1):101–107 74. Caturegli P, Newschaffer C, Olivi A, Pomper MG, Burger PC, Rose NR (2005) Autoimmune hypophysitis. Endocr Rev 26:599–614 75. Schultz AB, Brat DJ, Oyesiku NM, Hunter SB (2000) Intrasellar pituicytoma in a patient with other endocrine neoplasms. Arch Pathol Lab Med 125(4):527–530 76. Wolfe SQ, Bruce J, Morcos JJ (2008) Pituicytoma: case report. Neurosurgery 63(1):E173–E174 77. Tien RD, Newton TH, McDermott MW et al (1990) Thickened pituitary stalk on MR images in patients with diabetes insipidus and Langerhans cell histiocytosis. AJNR 11:707 78. Bourekas EC, Tzalonikou M, Christoforidis GA (2000) Cavernous hemangioma of the optic chiasm. AJR 175(3): 888–891 79. Nishizawa S, Yokoyama T, Hinokuma K et al (1997) Pituitary astrocytoma: magnetic resonance and hormonal characteristics. J Neurosurg 87:131 80. Moskowitz SI, Hamrahian A, Prayson RA, Pineyro M, Lorenz RR, Weil RJ (2006) Concurrent lymphocytic hypophysitis and pituitary adenoma. Case report and review of the literature. J Neurosurg 105(2):309–314
Brain Metastasis
13
Nicholas J. Patronas
Contents 13.1 Background................................................................ 373 13.2 Pathogenesis................................................................ 374 13.3 Pathology.................................................................... 375 13.4 Clinical Symptoms..................................................... 376 13.5 Imaging Studies.......................................................... 377 13.6 Differential Diagnosis................................................ 391 13.7 The Role of Imaging in the Posttherapy Period.......................................... 394 References............................................................................ 396
Intracranial metastasis is a complication of cancer with formidable consequences. Brain imaging is extensively used to screen patients with newly diagnosed malignant tumors and to evaluate patients with known malignancies who develop neurologic deficits. Recent advances in the management of patients with brain metastases have made accurate diagnosis and localization of these tumors of paramount importance. Brain imaging is also used in accessing responses to and complications of therapy. In this chapter, we examine the pathogenesis of metastatic brain tumors and their clinical presentation, review the available imaging methods and their relative values in diagnosis, and address diagnostic issues that arise after treatment.
13.1 Background
N.J. Patronas Professor of Radiology, National Institute of Health, Marryland, USA e-mail:
[email protected]
The incidence of metastasis in the central nervous system (CNS) during the course of systemic malignancies varies widely in published reports. These discrepancies are best explained by a) the different biases that influence the inclusion of patients in these studies, b) differences in the methods of diagnosis (i.e., autopsy, surgery, or imaging), and c) differences in demographic composition and tumor type [1–4]. It is currently believed that brain metastasis occurs in about 25–30% of patients with cancer [5, 6]. The propensity of primary tumors to metastasize to the CNS varies among different malignancies with lung and breast cancers known to be the tumors that most commonly cause this complication [1]. Metastatic CNS tumors primarily affect the brain parenchyma. Metastasis in the meninges or in the epidural intracranial space can also occur producing
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_13, © Springer-Verlag Berlin Heidelberg 2011
373
374
symptoms indistinguishable from intraparenchymal brain tumors. It has been estimated that 80–85% of brain metastases are located in the cerebrum and 15–20% in the cerebellum and brain stem. The frontal and the parietal lobes of the cerebral hemispheres are most commonly affected, with the corticomedullary junction being the earliest site of involvement
13.2 Pathogenesis The departure of tumor cells from the primary tumor site, the implantation of these cells in a remote organ, and the development of a metastatic colony follows a specific pattern and occurs in stages. The first barrier that neoplastic cells cross at the site of their origin is the basement membrane. An impermeable basement membrane that demarcates a mass and clearly separates it from its surroundings is a constant feature of benign tumors. On the other hand, the basement membrane is poorly formed or absent in malignant tumors, and when identifiable, is always breached by neoplastic cells. The degree of maturation of the basement membrane and the extent to which the neoplastic cells permeate it represents an index of the malignant potential of a neoplasm. The invasion of the basement membrane by tumor cells follows a three-step process. Initially, special receptors on the surface of these cells recognize a glycoprotein (laminin) of the basement membrane to which they attach. This is followed by proteolysis of type IV collagen of the basement membrane by a specific collagenase found in tumor cells. Once an area of the membrane is dissolved, locomotion follows, during which neoplastic cells exhibit increased mobility allowing them to cross the defective membrane and to move between cells into the interstitial space, where they may start replicating. Using similar mechanisms of invasion, tumor cells may penetrate the basal membrane and enter the lumen of capillaries and lymphatics [7–11]. Entrance of neoplastic cells into the vascular lumen is also greatly facilitated by angiogenesis. It has been shown that tumors at the primary site of development can grow up to 1 mm in size, since at this stage, the necessary nutrients can reach the tumor cells by diffusion from existing vessels. The formation of new vessels within the tumor is a mandatory process initiated by the tumor cells, so that further growth can
N.J. Patronas
continue unimpeded. The process of angiogenesis is a complicated biological phenomenon during which there is change in the local equilibrium of the proangiogenic and antiangiogenic molecules in favor of the former. The newly formed tumor vessels are structurally different from the vessels in normal tissue. They are characterized by larger lumen, tortuous course, and permeable walls through which neoplastic cells can gain access into the circulating blood [12–14]. Once within the intravascular compartment, neoplastic cells can escape the primary tumor site and may establish colonies in different organs. As a rule, tumor cells are arrested in the first capillary or lymphatic bed encountered. Thus, the first organ of metastasis to a large degree can be predicted from the anatomic routes that tumor cells are obliged to follow. For example, tumor cells originating from the mucosal glands of the large bowel after they have invaded the vascular capillaries at the primary site enter the local venous channels that permit them to eventually lodge into the liver parenchyma. Similarly, a skin melanoma in one of the extremities after invading the wall of the local lymphatic vessels will eventually end up in the regional lymph nodes. At the new site, by further invasion of the capillaries or the small venules, tumor cells may again enter the systemic circulation and gain access to a wider spread. Most of the freely circulating tumor cells are destroyed by the defense mechanisms of the immune system. Perhaps more importantly, the mechanical shearing stress forces they encounter from the site of origin to site of their destination can cause them to die. Cells that survive may be implanted and produce metastatic colonies in different organs at a rate that is roughly proportional to the blood flow of these organs. Once a neoplastic cell is attached to the wall of a capillary, the endothelial cells retract, allowing direct contact of the tumor cells with the basal membrane. At this point, the same mechanisms used by the tumor cells to penetrate the basement membrane at the primary site are repeated so that tumor cells eventually enter the interstitial space of the host organs. Some primary tumors metastasize preferentially to certain organs, suggesting that factors other than circulatory considerations play a significant role. A good example of this phenomenon is the known affinity of the eye melanomas to metastasize in the liver. It is believed that the development of metastatic colonies is regulated by the target organs. The regulatory mechanisms of this process are the topic of active research. It
375
13 Brain Metastasis
has been suggested that the target organ either uses chemical signals to attract neoplastic cells or that the endothelial cells of the vasculature of these organs possess a special affinity for certain types of tumor cells. Also, it is recognized that the presence of neoplastic cells in the interstitial space of an organ is not always synonymous with the development of a new tumor colony, because other autocrine growth mechanisms or factors from the local tissues must be mobilized to promote the development of tumor [15]. In the case of CNS metastasis, tumor cells arrive in the intracranial cavity by the arterial route. Because 15–20% of the cardiac output enters the cerebral circulation, it is not surprising that the brain is a common site of metastatic deposits. Tumor cells of any primary neoplasm can enter the arterial circulation after passing through the pulmonary capillary bed. The proximity of lung carcinoma to the pulmonary vascular bed provides a rationale for the high incidence of brain metastases from this primary site. Clusters of tumor cells are found lodged in small arteries at the gray–white matter junction, where the luminal diameter changes abruptly from 100–200 to 50–150 mm. This embolic, size filtration mechanism is a plausible explanation for the high frequency of tumor nodules found at the gray–white matter interfaces. The venous route has also been implicated as a possible mode of access to posterior fossa. The venous system of the posterior fossa is in direct communication with the venous plexus of the spinal canal (Batson’s plexus). It has been theorized that tumor cells from metastatic deposits in the spinal column enter this plexus and are taken into the intracranial circulation by retrograde venous flow. Carcinomas of the genitourinary and gastrointestinal tracts are said to utilize this route although this is not unequivocally proven. On rare occasions, tumor cells may pass from the venous to the arterial circulation through an atrial septal defect that remains patent in a small percentage of individuals, potentially permitting bi-directional flow. Tumor metastasis to the leptomeninges takes place primarily by the arterial route. Alternatively, neoplastic cells may implant on the leptomeninges after escaping the ependymal surface or the choroid plexuses and circulating through the cerebrospinal fluid (CSF). The dura may be seeded by neoplastic cells of the arterial circulation, or be invaded by metastatic calvarial lesions either directly or from circulating cells in the small venules shared by both [7, 16–18].
The involvement of the epidural space in the intracranial cavity or in the spinal canal usually occurs as a result of direct extension of metastatic tumor in the skull or vertebrae. Tumors of the nasopharynx commonly invade the intracranial space though direct extension, via the foramina of the skull base. Head and neck tumors can also extend intracranially by perineural spread with high propensity for such a spread found in the adenoid cystic carcinomas.
13.3 Pathology On macroscopic examination, metastatic brain tumors often appear discolored due to alteration of tumor circulation and microscopic hemorrhages within the tumor parenchyma. These tumors are rounded, firm, and well demarcated. When large in size they frequently undergo central necrosis. Certain metastatic tumors such as melanomas tend to bleed forming frank hematoma in the tumor bed (Fig. 13.1). Calcifications have been described only rarely in metastatic brain tumors and are most commonly seen in metastatic osteosarcomas. Edema is commonly found in the brain parenchyma adjacent to the tumor. The amount of edema is often proportional to the size of the tumor. This edema is vasogenic produced by extravasation of
Fig. 13.1 Gross specimen of metastatic melanoma shows a large hemorrhagic lesion (arrow). Additional small metastatic lesions at the corticomedullary junction are also seen (arrowheads)
376
fluid from the intravascular compartment in the interstitial space. The exit of fluid is facilitated by the defective walls of the tumor vessels. Metastatic lesions to leptomeninges appear as focal areas of abnormal meningeal thickening on the surface of the brain. Nodular projections of these tumors often invaginate from the meninges into the adjacent brain (Fig. 13.2). The ependymal walls of the ventricles are occasionally infiltrated by tumor and present with increased thickness or with nodular formations and often occur concomitantly with meningeal carcinomatosis. The ventricular system may be deformed by metastatic tumors adjacent to the ventricles. Enlargement of the ventricles can occur when a metastatic tumor obstructs the foramina of Monroe, the aqueduct of Sylvius, or the outlets of the fourth ventricle. Ventriculomegaly can also occur as a result of meningeal carcinomatosis caused by infiltration of the arachnoid villi by the tumor inhibiting CSF resorption. Microscopically, metastatic tumors in the brain parenchyma when properly differentiated exhibit histologic features similar to those in their primary sites (Fig. 13.3). In poorly differentiated tumors, specific immunohistochemical markers are utilized to characterize these tumors (Fig. 13.4). Regardless of the cell type, neovascularity is evident within the tumor parenchyma characterized by immature vessels with endo thelial cells containing abundant vesicles and defective tight junctions [1, 7].
N.J. Patronas
Fig. 13.3 Metastatic lung adenocarcinoma in brain (hematoxylin-eosin, original magnification ×400)
Fig. 13.4 Metastatic malignant melanoma in brain (Immunoreac tivity to HMB-45, original magnification ×400)
13.4 Clinical Symptoms
Fig. 13.2 Dural metastases of breast carcinoma. Macroscopic image shows a thickened dura with nodular projections (arrows)
During the early stages of the disease, patients with metastatic brain tumors are usually asymptomatic. As the tumor enlarges edema develops and symptoms appear. Headaches reported in as high as 88% of cases is the most common symptom appearing insidiously and becoming progressively worse with time. Head aches develop as a result of increased intracranial pressure caused by the tumor and edema or hydrocephalus. Confusion or behavioral changes are also symptoms, which at the onset of the metastatic process are vague and may not be appreciated during the early stages of the disease. Focal motor weakness can develop
13 Brain Metastasis
377
gradually becoming progressively worse. Motor deficits have been elicited in as many as 66% of patients with metastatic brain tumors. In some patients there is abrupt onset of neurologic deficit mimicking acute stroke. Seizures occur commonly in patients with intracerebral or leptomeningeal metastasis and occasionally may be the initial clinical presentation. Most of the remaining symptoms are specific to the location of the metastatic lesion. Thus, metastatic tumors can produce visual disturbances, vertigo, aphasia, and imbalance. Endocrine disorders develop if the tumor involves the hypothalamus, the pituitary gland or its stalk. Meningeal carcinomatosis is often manifested by cranial neuropathies [3, 19].
13.5 Imaging Studies Computed tomography (CT). For the last 25 years, CT has played a major role in the diagnosis of brain tumors. The diagnosis of these tumors depends on differences in X-ray absorption by the tissues through which the X-ray beam passes [20, 21]. In living tissues, the differences in X-ray absorption are relatively small, so that most metastatic brain tumors have density similar to normal brain ranging between 20 and 40 Hounsfield units (HU) and escape detection. In such cases, the presence of an intracerebral tumor can be suspected indirectly from the mass effect that it produces. Features of mass effect include effacement of the adjacent cortical sulci, compression of a ventricular wall, displacement of the midline structures, and distortion of the subarachnoid cisterns. Occasionally, some tumors exhibit densities greater than normal brain. This is a feature of highly cellular tumors with relatively small interstitial spaces and high nuclear--cytoplasmic ratio. Such tumors include lymphoma, small cell carcinoma of the lung, and melanoma. Hemorrhagic tumors exhibit even higher densities than normal brain measuring 60–90 HU and can be easily detected (Fig. 13.5). Calcified metastatic tumors demonstrate high densities (usually over 110 HU) and become even more apparent by CT. Both calcified and hemorrhagic tumors are however rare. If necrosis develops in a metastatic tumor, the necrotic region presents with decreased density that can be visually appreciated on CT images. Peritumoral edema, when present, also shows low density in the brain parenchyma near the tumor. The detection of brain
Fig. 13.5 Precontrast CT scan of the brain. Metastatic malignant bronchial carcinoid in both cerebral hemispheres. Multiple tumors are identified. They are hyperdense with respect to normal brain due to hemorrhagic elements in the tumor parenchyma
tumors by CT was greatly improved with the use of intravenous contrast agents. These are salts of iodinated acids whose molecules enter into the interstitial space of the tumor increasing the radiographic density of the tumors. The diffusion of the contrast agents into the tumor parenchyma occurs because the blood–brain barrier (BBB) of the metastatic tumor is disrupted. On postcontrast images, small metastatic tumors appear as solid nodular lesions and enhance homogeneously. Larger tumors with necrosis demonstrate ring-like enhancement (Fig. 13.6). The thickness of the enhancing ring is usually not uniform and its inner border is often irregular [22–25]. Meningeal carcinomatosis can be diagnosed only in postcontrast studies. The meninges infiltrated by the tumor are abnormally thickened and show increased enhancement within the cortical sulci, the fissures, or the subarachnoid cisterns [26] (Fig. 13.7). In order to improve the diagnostic accuracy of CT, various investigators have doubled the dose of injected iodine from 40 to 80 g and reported increased diagnostic yield. The poor contrast resolution of CT
378 Fig. 13.6 (a) Postcontrast CT scan of the brain. Metastatic Ewing’s sarcoma in the left occipital lobe. Two faintly enhancing cavitary tumors are identified (arrows). A third metastatic tumor was present in the right parietal lobe that is not clearly visible in this section. (b). Precontrast scan of the brain after whole brain irradiation. There is complete resolution of the tumors. Dystrophic calcifications are present in the tumor bed
N.J. Patronas
a
Fig. 13.7 Postcontrast CT scan of the brain. Meningeal carcinomatosis in a patient with lymphoma. There is abnormal enhancement of the leptomeninges on the medial aspect of both cerebral hemispheres (arrows)
and the partial volume artifact found in every tomographic imaging method represent drawbacks in the detection of small metastatic tumors. Thus, such tumors may not enhance sufficiently to be visualized even with the increased dose of contrast [27, 28].
b
Another major problem with CT is the presence of artifacts caused by hardening of the X-ray beam as it passes through the skull. Such artifacts usually obscure small tumors located on the surface of the brain and may obscure even larger lesions in the posterior fossa where these artifacts are more prominent. The diagnosis of meningeal carcinomatosis over the convexity of the cerebral hemispheres also remains problematic since the overwhelmingly increased density of the calvarium obscures the abnormally enhancing meninges in those areas. A variety of problems have been encountered after intravenous administration of iodinated contrast agents limiting their use. The incidence of such side effects has been reported to be as high as 5–12%. Severe anaphylactic reactions associated with acute cardiopulmonary failure occur in one of 1,000 injected patients, whereas, the incidence of death is between 1:12,000 and 1:75,000. Patients with a prior history of allergy to iodine may not be eligible for contrast-enhanced CT examination or require premedication prior to such study. The development of a major reaction is unpredictable and cannot be excluded by prior testing. Iodinated contrast agents are also known for their nephrotoxicity. Contrast-induced acute renal failure has been reported in 15–42% of patients with azothemia and diabetes mellitus. Therefore, patients with serum creatinine greater than 1.6–1.8 mg/100 mL may not be safely studied with iodinated contrast agents. Other risk factors include cardiovascular diseases, severe
379
13 Brain Metastasis
pulmonary decompensation, bronchial asthma, pheochromocytoma, sickle cell disease, multiple myeloma, diabetes mellitus, dehydration, advanced age, and severe debilitation [29–32]. The utilization of CT as a CNS imaging modality has gradually subsided over the last several years and has been largely displaced by magnetic resonance imaging (MRI). Today, CT is mainly used in claustrophobic patients or patients with known contraindication to MRI examinations. Additionally, CT continues to play a role in the examination of very ill patients who cannot be examined safely within an MRI scanner or in patients with skull lesions offering superior resolution of bone anatomy. Finally, CT may be preferred for its lower cost and greater speed. Magnetic resonance imaging. This method has become the diagnostic modality of choice for the detection of metastatic brain tumors. The method images hydrogen nuclei, which are abundant in the body tissues, and exploits the differences of the relaxation times T1 and T2 of these nuclei in the various tissues under the influences of radiofrequency pulses [33–36]. On precontrast T1-weighted MR studies, metastatic brain tumors are usually isointense with respect to normal gray matter and are not clearly visible. Tumors with areas of necrosis can be seen on precontrast studies showing decreased signal intensity in the necrotic region (Fig. 13.8). Low signal intensity is also present in regions of peritumoral edema [37, 38]. In hemorrhagic tumors the signal intensity can be altered depending on the age of hemorrhage. During the acute
Fig. 13.8 (a) Precontrast T1-weighted MRI scan of the brain. A cavitary mass is demonstrated in the left parietal lobe representing metastatic breast carcinoma (arrow). (b) Postcontrast MRI scan shows abnormal ring-like enhancement in the viable part of the tumor (arrow). Note the nonuniform thickness of the cavity. In addition, several other solid but smaller metastatic tumors are seen scattered in both cerebral hemispheres. These smaller tumors were not appreciated on the precontrast scan
a
phase of hemorrhage, that is, the initial 24 h, the hemorrhagic metastasis contains deoxyhemoglobin which is not discernible on the T1-weighted technique. As deoxyhemoglobin is converted to methemoglobin, the signal intensity of the lesion increases and the hemorrhagic tumor becomes hyperintense due to the paramagnetic properties of this product (Fig. 13.9). In melanotic melanomas or other melanotic tumors, the melanine possesses similar paramagnetic properties increasing the signal intensity of these tumors on the precontrast T1-weighted technique. Regardless of the signal characteristics of the tumor, features of mass effect as those described on CT studies are equally well shown on T1-weighted images, but are largely dependent on the size and the location of the tumor. On T2-weighted MR studies, brain tumors are hyperintense with respect to normal brain due to increase in the T2 value of the overhydrated tumor cells [39] (Fig. 13.10). Conventional spin echo (SE), fast spin echo (FSE), and fluid-attenuated inversion recovery (FLAIR) sequences are used. A major disadvantage of the T2-weighted technique is encountered in the case of small tumors located near the surface of the brain adjacent to CSF spaces. The overwhelmingly high signal intensity of the CSF commonly obscures small tumors in those areas decreasing the sensitivity of this method. This phenomenon does not exist with the FLAIR technique, which presents CSF with low signal intensity providing good contrast between tumor and CSF or normal brain (Fig. 13.11). Peritumoral edema and tumor necrosis show increased signal
b
380 Fig. 13.9 (a) Precontrast T1-weighted MRI scan in a patient with melanoma. A large hemorrhagic metastatic mass is noted in the right basal ganglia compressing the lateral ventricle (arrows). The mass is hyperintense due to methemoglobin formation. (b) On the postcontrast MRI, the signal intensity of the mass is increased particularly in its periphery (arrows). This is due to enhancement in the nonhemorrhagic part of the tumor. (c) On FLAIR MRI scan a zone of edema is identified around the tumor demonstrating increased signal intensity (arrows)
N.J. Patronas
a
b
c
intensity with T2-weighted techniques. Oftentimes, there is an apparent difference in the signal intensity between the necrotic region and the solid part of the tumor that can be further accentuated by appropriate selection of the TR and the TE used in a T2-weighted technique. The signal difference between tumor and peritumoral edema is less obvious making clear separation of the viable tumor from its surrounding more problematic. Some tumors, especially adenocarcinomas, retain the signal characteristics of the primary tumor, with signal intensities similar to or lower than normal gray matter. In these types of tumors, T2-weighted techniques can provide good demarcation of the tumor mass not only from the necrotic
portion, but also from the peritumoral edema [40] (Fig. 13.12). The signal abnormalities encountered in hemorrhagic tumors depend on the age of the hemorrhage. During the acute stage, when deoxyhemoglobin is the dominant product of hemoglobin degradation, the signal intensity of the tumor is decreased on the conventional T2-weighted SE technique but less prominently decreased on the FSE technique. In this stage, T2-weighted gradient echo technique provides the best opportunity to demonstrate even minute elements of hemorrhage in the tumor parenchyma presenting with markedly decreased signal due to T2* effect. In the next phase which takes place about 24–48 h after the
381
13 Brain Metastasis Fig. 13.10 (a) Postcontrast T1-weighted MRI scan of the brain in a patient with metastatic breast carcinoma. There is an enhancing mass on the medial aspect of the left occipital lobe abutting the posterior falx (white arrow). The flat surface of the tumor against the falx suggests extraaxial tumor such as meningioma. (b) On the T2-weighted scan the tumor is hyperintense (black arrow). Prominent edema is seen in the adjacent brain parenchyma exhibiting similar signal characteristics. The edema obscures the tumor margins underestimating its size
a
b
a
b
c
Fig. 13.11 (a) Postcontrast MRI scan in a patient with metastatic renal cell carcinoma. An enhancing tumor is noted in the right occipital lobe (arrow). (b) On the T2-weighted scan, the tumor cannot be distinguished from the cerebral spinal fluid of the adjacent sulcus. (c) On FLAIR technique, the tumor is clearly demonstrated as an abnormal area of increased signal intensity (arrow)
382 Fig. 13.12 (a) Postcontrast T1-weighted MRI scan of the brain. Metastatic adenocarcinoma in the right occipital lobe (white arrow). There is heterogeneous enhancement of the tumor with the central hyperenhancing region representing an area of necrosis. (b) On the T2-weighted scan a broad zone of edema is present around the tumor. Note the relative hypointensity of the tumor with respect to the necrotic center and the peritumoral edema (black arrow)
N.J. Patronas
a
extravasation of blood, deoxyhemoglobin is converted to methemoglobin from the periphery to the center increasing the signal intensity in the T2-weighted techniques making the hemorrhagic parts of the tumor indistinguishable from the nonhemorrhagic parts. During the final stage of evolution of the hemorrhagic event, blood products are removed via phagocytosis and hemosiderin if formed in the tumor bed producing a T2* effect causing decreased signal intensity on the T2-weighted and the gradient echo techniques. This latter observation is usually made in successfully treated tumors. Similar to CT, MRI studies for intracranial tumor detection are performed after intravenous administration of contrast. Gadolinium (Gd), an element of the lanthanide series, has been found to be most suitable for MR imaging. This element with its seven peripheral electrons shortens the T1 value of the tissues in which it concentrates resulting in increased signal intensity. Different chelating compounds in ionic or nonionic formulations are available for intravenous injection (0.1 mmol/kg body weight). Gadolinium contrast agents pass through the disrupted BBB into the interstitial space of the metastatic tumors distinguishing the tumor from the adjacent brain. Tumors as small as 2 mm in diameter can be seen (Fig. 13.13). Thus, postcontrast T1-weighted MRI studies represent the best technique available to date for the detection of brain tumors [37, 38, 41]. Various investigators have sought to improve the sensitivity of MRI in the
b
detection of metastatic brain tumors by increasing the administered dose of gadolinium. Several studies have shown that double or triple dose of gadolinium results in an increase in the number of detected brain metastases. The improved sensitivity in tumor detection ranges from 13 to 43%, which represents an important gain considering the impact of this diagnosis on the prognosis and management of these patients [42–47]. The value of triple dose of gadolinium is even greater when scanning is performed in low Tesla open magnets. Akerson et al. [48], demonstrated that with single dose scanning at 1.5 T, the contrast between lesion and brain is superior to that obtained at 0.3 T, but this difference is eliminated with triple dose. These findings suggest that triple dose of gadolinium can overcome the inherent reduced contrast effect of the low Tesla magnets [49]. False positive results may be occasionally encountered by enhancing peripheral cortical veins that mimic tumors. This problem is more commonly observed when a double or triple dose of gadolinium is used or when a 3D T1-weighted gradient echo technique is applied instead of the conventional SE. The multiplanar capability of MRI is helpful to separate these enhancing vessels from small tumors. Furthermore, postcontrast FLAIR imaging is also offered in demonstrating better than postcontrast T1-weighted SE images the difference of the enhancing tumors from the surface veins [50, 51]. Other MR pulse sequences available to improve the diagnostic yield in patients with suspected metastatic
383
13 Brain Metastasis Fig. 13.13 Postcontrast T1-weighted MRI scans of the brain. Two consecutive sections (a, b) are presented. Multiple enhancing metastatic tumors from primary lung carcinoma are shown in both hemispheres. Some of the metastatic tumors measure two millimeters in diameter. Note the absence of significant mass effect
a
a
b
b
Fig. 13.14 (a) Postcontrast T1-weighted MRI scan of the brain in a patient with breast carcinoma. Three metastatic tumors are seen in the left cerebral hemisphere and two in the right. (b) Postcontrast T1-weighted scan with magnetization transfer demonstrate the same tumors but the contrast between the tumor and the normal brain is improved with this technique
tumors include the magnetization transfer (MT) technique. MT pulses are applied prior to initiation of a standard SE sequence. This method effectively reduces the signal intensity of the normal brain parenchyma. After intravenous administration of gadolinium, the shortening of the T1 value of the tissues in which it concentrates produces increased signal in these tissues, which is not affected by the MT pulses. The end result of this is to achieve an improved contrast to noise ratio between enhancing and nonenhancing tissues. This method has been tried in a variety of brain lesions
including metastasis and was proven to be superior to conventional SE T1-weighted techniques in terms of sensitivity and conspicuity of the detected tumors (Fig. 13.14). Increasing the dose of gadolinium while using MT technique was not accompanied by increased sensitivity of tumor detection [52–55]. The value of delayed imaging after contrast has been examined for the evaluation of primary and metastatic tumors. These studies showed that the maximal enhancement in brain tumors occurs between 3.5 and 20 min after injection of contrast. Therefore, scanning
384
should begin 2–5 min after injection and further delay of scanning does not improve sensitivity [56, 57]. The safety of gadolinium-based contrast agents has been addressed in the literature. The experience gained in the past 20 years has shown that MR contrast agents are safer than iodine-based agent used in CT scanning. Certain limitations in the usage of gadolinium-based agent have been recently identified. Patients with advanced renal failure run a high risk of developing nephrogenic systemic fibrosis /nephrogenic fibrosing dermopathy, a debilitating disease which in severe cases can even lead to death. It is now recommended that patients with estimated glomerular filtration rates less than 30 mL/min/1.73 m2 (stage 4 and 5) should not be given these agents. Also gadolinium-based agents should not be given to patients on dialysis and to those with reduced renal function who have had or are awaiting liver transplantation. European guidelines, Food and Drug Administration guidelines, and a number of publications have become available in the last few years and address issues related to toxicity of gadoliniumbased agents. These publications assess other additional risks factors as well as the relative danger of the various commercially available MRI contrast agents [58–60]. Diffusion and perfusion imaging. Various authors have examined the value of diffusion-weighted imaging (DWI) in assessing the histologic type of metastatic tumors. Thus, Hayashida et al. [61] found that welldifferentiated metastatic tumors exhibit low signal intensity relative to normal gray matter, whereas poorly differentiated tumors were hyperintense on DWI (Fig. 13.15c). The hyperintensity of the poorly differentiated tumors is attributed to synergistic effect of restricted diffusion and T2-shine through. On the contrary, apparent diffusion coefficient (ADC) was found to be lower in poorly differentiated tumors than in well-differentiated tumors. The ADC value was inversely correlated with tumor cellularity. Thus, poorly differentiated hypercellular tumors have lower signal intensity than less cellular well-differentiated tumors (Fig. 13.15d). Such correlation between cellularity and ADC has also been observed in gliomas and meningiomas [62]. Perfusion studies have also been performed in patients with brain tumors to evaluate the tumor microcirculation and shown to provide information that is useful in grading primary brain tumors or in distinguishing primary from metastatic tumors. Dynamic susceptibility-weighted perfusion technique is most
N.J. Patronas
commonly used. In this method, acquisition of data takes place during the first pass of gadolinium-based bolus through the capillary bed of an organ. The passing of gadolinium through the capillaries of any tissue causes a transient drop of signal intensity as compared to the baseline which is followed by recovery. The diminished signal intensity is proportional to the capillary density or the blood volume of the tissues. Full recovery of the signal occurs in normal brain after the first pass of gadolinium. In pathological tissues, if there is disruption of the BBB, the signal recovery is partial. This technique allows measurement of relative cerebral blood volume (rCBV) within the tumor parenchyma and has been used to assess tumor grading and differentiate primary from metastatic tumors. Both are highly vascular tumors and demonstrate increased rCBV (Fig. 13.15e). Thus, Bulakbasi et al. [63] found that this method provides high sensitivity (95.46%) and specificity (91.67%) in grading glial tumors. They also found higher rCBV (i.e., capillary density) in primary tumors as compared to metastatic and that the difference is statistically significant. Other investigators though, have observed that rCBV measurements vary widely in metastatic brain lesions of different primary tumors. Law et al. [64] found no statistical difference in rCBV between primary and metastatic tumors, while Kremer et al. [65] reported that certain metastatic tumors such as melanomas or renal cell carcinomas have greater rCBV values than high-grade gliomas. The peritumoral region which exhibit high signal intensity on conventional T2-weighted scans and demonstrate no abnormal enhancement on the postcontrast has attracted the attention of many investigators (Fig. 13.15a, b). Dynamic susceptibility-weighted scans have been used to measure rCBV in these regions and assess possible differences between metastatic and primary brain tumors. Such studies have shown higher rCBV values in the peritumoral regions of primary high-grade gliomas as compared to metastatic tumors. This is because the primary brain tumors infiltrate adjacent normal brain at the microscopic level and are accompanied by vascular proliferation causing increased blood volume. In metastases, on the other hand, the peritumoral hyperintensity is caused only by vasogenic edema and the rCBV is not increased [63, 64] (Fig. 13.15f). The degree of signal recovery after the first pass of gadolinium through the capillary bed has been found to
385
13 Brain Metastasis Fig. 13.15 (a) FLAIR image shows an isointense metastatic lesion in the right frontal lobe, from a poorly differentiated lung adenocarcinoma, with peritumoral edema exhibiting high signal intensity. (b) Postcontrast T1-weighted image shows intense enhancement of the tumor. The peritumoral hypointense edema shows no abnormal enhancement. (c) On diffusion-weighted image the tumor shows high signal. (d) On ADC map the tumor demonstrates low signal intensity. (e) On perfusion color map, the rCBV is increased within the metastatic lesion and decreased in the peritumoral vasogenic edema. (f) Long TE (135 ms) spectrum obtained from the enhancing lesion shows significant elevation of the Cho and decrease NAA. (g) Normal Cho/Cr and NAA is noted in the peritumoral edema
a
c
b
d
e
be inversely proportional to the degree of disruption of the BBB. The more disrupted is the BBB the more the leakage of gadolinium in the interstitial space the less signal recovery. This parameter measures capillary permeability which is often proportional to tumor aggressiveness and can also be used to assess treatment
responses. Thus, it has been found that the recovery of the signal is greater in low grade as compared to gliomas of higher grade. Cha et al. [66] used dynamic susceptibility-weighted contrast enhanced perfusion MRI to evaluate patients with primary high-grade gliomas and metastatic tumors. They found that in
386 Fig. 13.15 (continued)
N.J. Patronas
f
g
glioblastomas, the average signal recovery during the first pass was 80.9%, whereas the average signal recovery in metastases was 62.5%. It was concluded that when the average percentage of signal intensity recovery in the contrast enhancing lesion is more than 82% and less than 66%, the prediction of GBM and single brain metastasis, respectively, had a specificity of 100%. Therefore, these data support the notion that the capillaries of the metastatic tumors are leakier than those of GBM. Histologic studies using electron microscopy have shown that the metastatic brain tumors completely lack BBB, whereas the newly formed capillaries of GBM with hyperplastic endothelial cells surrounded by pericytes retain some aspects of the BBB and are less permeable [67]. In the case of meningeal carcinomatosis, postcontrast T1-weighted and FLAIR studies offer the best imaging method to demonstrate such lesions. Two patterns of meningeal enhancement are recognized. The first is that of neoplastic involvement of the lepto
meninges presenting with increased enhancement in the subarachnoid spaces of the cortical sulci, the fissures, and the cisterns. Nodular formations can also develop in the subarachnoid spaces that may invaginate within the adjacent brain parenchyma. The pattern of leptomeningeal infiltration is identical to that described on the CT, but the intensity of enhancement and thus the sensitivity is greater in the postcontrast MR studies (Figs. 13.16 and 13.17). The second pattern of involvement is increased thickness and enhancement of the dura. Since the inner table of the skull is hypointense on MRI, the abnormally enhancing dura clearly stands out between the CSF and the bone of the skull [68–71] (Fig. 13.18). A major drawback in the diagnosis of meningeal carcinomatosis by either CT or MRI is that the changes described above lack specificity. Indeed, abnormal enhancement of the leptomeninges identical to carcinomatosis occurs in a variety of infections including viral, bacterial, or fungal diseases [72]. Inflammatory processes such as
13 Brain Metastasis
387
Fig. 13.16 Postcontrast T1-weighted MRI scan of the brain in a patient with metastatic breast carcinoma. There is evidence of meningeal carcinomatosis manifested by abnormal enhancement of the leptomeninges over the convexity of the right cerebral hemisphere
Fig. 13.17 Postcontrast T1-weighted MRI scans of the brain in a patient with melanoma. Enhancing masses are noted in both cerebello-pontine angles mimicking acoustic schwannomas (arrows). This abnormality represents part of the spectrum of meningeal carcinomatosis
sarcoidosis or Langerhans histiocytosis also present with abnormal enhancement of the leptomeninges. In the case of dural involvement the differential diagnosis is even broader. Besides the infectious or the specific inflammatory processes already discussed, the dura enhances in response to inflammatory reaction that occurs following previous subarachnoid hemorrhage. Furthermore, the dura enhances in response to a variety of neoplastic or nonneoplastic diseases that involve the calvarium. In the case of metastatic or primary tumors of the calvarium, inflammatory reaction often develops in the dura even in the absence of dural involvement by the neoplasm. Previous published reports suggest that the sensi tivity of contrast enhanced MRI is rather poor since positive results consistent with meningeal carcinomatosis are elicited in only 36–66% of patients with positive CSF cytology. These findings are hardly surprising considering the fact that only a few malignant cells in the CSF are sufficient for establishing this diagnosis by cytology. On the other hand, the superior sensitivity of CSF cytology is secured only if multiple spinal taps are performed.
Fig. 13.18 Coronal postcontrast T1-weighted MRI scan of the brain in a patient with metastatic prostate carcinoma to the skull. Three destructive metastatic lesions are seen in the skull. Associated soft tissue tumors are also present extending into the epidural space and compressing the brain parenchyma. The abnormal enhancement of the dura (arrows) represents either tumor invasion or inflammatory reaction
388
Metastatic tumors in the epidural space are easily diagnosed by contrast-enhanced MRI presenting with an enhancing extraaxial mass that may compress the adjacent brain parenchyma. Since these tumors represent an extension of a metastatic lesion of the calvarium, the abnormal enhancement of the epidural tumor can be traced in the bone from where it originated (Fig. 13.18). On fat suppression T2-weighted images, the intraosseous component of the mass can also be seen as an area of increased signal sharply contrasted against the normally hypointense calvarium. Single-photon emission computed tomography (SPECT). Brain tumors have been studied with a number of radiotracers. The bulk of this research, aimed at evaluating tumor biology, has been with primary brain tumors but also applies to metastatic brain tumors. The difference between SPECT and positron emission tomography (PET) is that SPECT uses isotopes that decay by single photon emission while PET uses isotopes that decay by positron emission. With either technique, the radioactive element is coupled with another compound to improve specificity [73, 74]. Technetium-99m tagged to hexamethylpropylene amine oxime (HMPAO) is a perfusion agent that has been used to study regional cerebral blood flow. Intravenous injection with dynamic data acquisition has shown increased tracer concentration in vascular primary and metastatic tumors. Static delayed images have revealed persistent abnormal uptake in tumors as well as a definite correlation between degree of uptake and tumor grade. Therefore, SPECT imaging with perfusion agents is a relative noninvasive predictor of tumor aggressiveness. However, the tracer uptake is not always specific for malignancy, since nonmalignant tumors including meningiomas and angiomas may demonstrate increased activity. Conversely, cavitary metastatic lesions with a thin rim of viable tumor suffer from the limited special resolution of the SPECT instrument and may be photopenic [75, 76]. Thallium-201 is another single-photon emitter having chemical properties similar to potassium that can pass through the cellular membrane. The accumulation of this tracer into the tumor depends on various factors that include blood flow, degree of disruption of the BBB, cellular membrane permeability, and tumor histologic type [74, 77, 78]. Ueda et al. [79] showed that rapid initial uptake is related to tumor vascularity and BBB disruption, whereas delayed retention reflects greater degree of malignancy. The value of SPECT
N.J. Patronas
with thallium-201 was assessed by Dierckx et al. [80] in metastatic brain tumors. They found increased uptake in 11 of 14 patients with metastatic brain tumors but in three of them the tumor was missed. Since most of the tumors in this study were large, it could be concluded that this method should not be used for tumor detection. Thallium-201 studies have been performed to assess the results of treatment and distinguish radiation necrosis from residual or recurrent tumor and good correlation with histology was noted when a very low or high uptake was observed [81, 82]. Other investigators used SPECT with thallium-201 to evaluate the clinical outcome in patients with a variety of brain tumors including metastasis. They found that high tracer uptake and retention correlated with poor prognosis [83–86]. SPECT imaging has been used to study amino acid uptake by primary and metastatic brain tumors. Iodine-123 labeled methyl tyrosine (IMT) is transported across the intact BBB but is not incorporated into cerebral proteins. SPECT studies with IMT differentiated high from low-grade gliomas and can distinguish high-grade tumors from nonneoplastic lesions [87, 88]. SPECT studies have also been performed with technetium 99m-labeled methoxyisobutylisonitrile (MIBI) to evaluate primary and secondary brain tumors. It has been shown that early uptake and rapid wash out occurs in high-grade gliomas and metastatic tumors. It has also been suggested that the rate of wash out provides valuable information to predict tumor response to chemotherapy [89, 90]. Positron emission tomography: PET is performed using isotopes that decay by positron emission. The most commonly used isotope is fluorine-18 attached to an analog of glucose, the deoxyglucose (DG) [91]. PET with fluorodeoxyglucose (FDG) has been widely used in primary brain neoplasms. These studies have demonstrated that the rate of glucose utilization directly relates to the biological behavior of these tumors. Thus, it has been shown that high-grade gliomas have increased metabolic activity whereas low-grade gliomas are hypometabolic [92]. FDG-PET studies have been found valuable as a predictor of prognosis in patients with brain tumors and have also been used to assess the results of therapy [93–96]. However, FDG-PET studies have not been shown to be useful for screening patients with suspected brain metastasis. The poor spatial resolution of
13 Brain Metastasis
PET compared to CT or MRI does not permit detection of small tumors or tumors with thin viable rim and large necrotic center. Furthermore, the gray matter normally demonstrates high rates of metabolic activity on PET, and hypermetabolic tumors located in the gray matter are not always discernible (Fig. 13.19). Also, hemorrhagic metastatic tumors exhibit low activity, since the overwhelming presence of the extravasated blood with zero metabolism masks the presence of such tumor. Griffeth et al. [97] evaluated 31 metastatic brain tumors with FDG-PET and were able to demonstrate only 21 of them. In addition, these authors found that the metabolic activity in metastatic tumors is highly variable and thus the method is not useful for determining tumor origin. Lassen et al. [98] studied patients with brain metastasis from small cell lung carcinoma before and after irradiation therapy. They showed that the rate of glucose metabolism in the posttreatment scans was decreased compared to baseline, but the changes were not statistically significant. Other investigators have shown that when the metabolic activity in metastatic tumors was decreased following radiation treatment, the median survival was longer in patients with hypometabolic tumors [97, 99]. Carbon-11 is another positron emitter that has been used in PET scanning. A number of carbon-11 labeled amino acids have been evaluated which include methionine, thymidine, and tyrosine. In theory, such studies were to assess DNA and protein synthesis and
Fig. 13.19 Metastatic melanoma to brain imaged by MRI and FDG-PET. On the left, a T1-weighted image of the brain shows a small enhancing mass in the left frontal lobe (short arrow). On the right, an FDG-PET scan presented in color shows the same tumor imbedded within the cortical ribbon (long arrow). The increased metabolic activity of the metastatic tumor is superior to that of the normal gray matter in this case
389
indirectly cellular proliferation in primary brain tumors. Targeting primary brain neoplasms, these studies reve aled increased activity in the tumor parenchyma as compared to normal brain, but there was no clear distinction between tumors of different aggressiveness. Furthermore, it is debated whether the concentration of activity reflects increased utilization of amino acids by the tumor cells or whether this phenomenon is due to simple diffusion of the radioactive compound in the interstitial space of the tumor [100–102]. Such studies have not been performed in patients with metastatic brain tumors, and in these neoplasms, it is even more questionable. Oxygen-15, a positron emitter has also been used in studies of brain tumors. A number of methods were developed to measure regional cerebral blood volume, cerebral blood flow, and metabolic rate of oxygen. Bolus injection of radioactive water (H215O) is the most commonly used method to measure cerebral blood flow. Inhalation techniques using oxygen-15 mixed with air and inhalation of carbon dioxide or traces of carbon monoxide labeled with oxygen-15 are used to measure blood flow and volume [103–105]. Again, the bulk of this work involves primary brain tumors and a variety of other diseases of the CNS including neurodegenerative and psychiatric disorders. In brain tumors, the focus of this research effort has been in the assessment of regional circulation before and after treatment. It is currently believed that these methods of measuring
390
blood volume and flow play no major role in tumor diagnosis, but if the local tumor circulation is diminished after treatment, this represents a favorable prognostic feature. Furthermore, such methods can be used to study the side effects that various treatments may have on normal brain [106–108]. Magnetic resonance spectroscopy (MRS). MRS is a method used to identify certain metabolites in brain tissues. The technical parameters of MRS vary widely with investigators using different echo times, single voxel vs. multivoxel spectroscopic imaging and scanning in magnets of different magnetic field. The various metabolites observed in MRS represent a different signature of the tissue under study. This method has the potential of showing metabolic changes prior to the development of structural abnormalities in brain [109, 110]. Proton spectroscopy has emerged as the dominant method of MRS in clinical practice although spectroscopic studies of phosphorus and sodium have also been used in experimental settings. In proton spectroscopy, metabolites such as choline, creatine, N-acetyl aspartate, lipids, and lactate are easily identifiable and their concentration can be assessed in various disease states. Choline reflects membrane synthesis and turnover, creatine is important in cellular energetics, and N-acetylaspartate (NAA) is a neuronal marker. Lactate reflects anaerobic metabolism and lipids are observed in regions of cellular breakdown [111–114]. Although, the results of MRS in patients with the same anatomic brain abnormality may vary and are often difficult to interpret, certain observations have been verified and many of the spectroscopic findings were shown to have a definite diagnostic value. Thus, a number of investigators have tried to correlate MRS findings with histologic grade of gliomas (low vs. high grade) and reported sensitivity and specificity ranging from 73 to 96% and from 63 to 88%, respectively [115, 116]. MRS in metastatic lesions show elevated choline/ creatine ratio, low or absence of NAA, and elevated lactate and lipids (Fig. 13.15f). Jijens et al. [117, 118] evaluated 66 patients with metastatic brain tumors and correlated the metabolic changes with tumor morphology and size. They found that small solid tumors have increased choline, larger tumors with heterogeneous enhancement showed elevated lipids and even higher choline, while larger tumors with prominent central necrosis had increased lactate and lower choline. Other investigators have also found increased lipids in metastatic brain tumors [119, 120].
N.J. Patronas
Discrimination between brain metastasis and GBM using MRS has been attempted by various investigators since the imaging presentation of these tumors can be very similar by conventional MRI techniques whereas the management is different. The results of these studies have shown that discrimination of these tumors can be difficult since NAA is depressed and choline is elevated in both. Furthermore, lactate and lipid peaks are elevated in both, since anaerobic glycolysis and tissue necrosis is a common feature of both tumors [112–117, 121, 122]. Opstad et al. [123] used single voxel proton MRS with short echo time (TE = 30 ms) and was able to observe lipid and macromolecule signals (LM) of two types: Highly mobile presenting with very sharp clearly defined peaks called L1 and less mobile with broad, overlapping resonances called L2. The ratio L1/L2 was found to be 2.6 in GBM and 3.8 in metastasis (p < 0.0001) documenting differences in metabolite or lipid concentration in these tumor groups allowing separation between the two with sensitivity and specificity of 80%. MRS can also be applied to assess metabolic abnormalities in regions surrounding brain tumors where conventional T2-weighted MRI scans show increased signal due to peritumoral edema. These studies have demonstrated that choline/creatine ratio was increased in the peritumoral T2 hyperintensity of gliomas but not in peritumoral region of metastatic tumors (Fig. 13.15g). The known microscopic infiltration of the surrounding tissues by primary brain tumors provides the only probable explanation for this observation [115, 124]. MRS has been found to be helpful in discriminating brain tumors from various other focal brain lesions such as abscess, infarction, or demyelination. MRS has been applied in patients with brain tumors treated with irradiation to differentiate recurrent tumors from radiation necrosis. Choline/creatine and choline/ NAA ratios were used and sensitivity and specificity of 87–89 and 83–89%, respectively, were reported [125–127]. Other investigators have pointed out that moderate increase in choline (Cho/Cr ratio less than 2.5) can occur after irradiation therapy due to cell destruction, whereas marked increase in choline (Cho/Cr ratio greater than 2.5) is found in recurrent disease caused by rapid division of tumor cells [128]. MRS may also be used in the evaluation of various other therapies since decreasing choline/NAA ratios after treatment should herald a favorable response.
13 Brain Metastasis
Finally, Warren et al. [129] examined the value of magnetic resonance spectroscopic imaging in assessing the biological behavior of recurrent primary brain tumors in children. They found the median survival to be decreased if the Choline/NAA ratio was greater than 4.5 and significantly longer if the Choline/NAA ratio was less than 4.5. From the experience gained thus far with MRS in brain tumors, it can be concluded that this method is not appropriate as a screening test for tumor detection. Also, in the case of metastatic tumors, it cannot be used to identify the origin of the primary tumor.
13.6 Differential Diagnosis The available imaging methods and particularly MRI have been proven highly accurate in the detection of metastatic brain tumors as small as 2 mm in diameter. Although tumor detection is relatively easy, specificity remains a diagnostic issue. Various brain lesions present with features that are indistinguishable from those of tumor metastasis. In many instances the patient’s history provides an essential guide in diagnosis. Primary brain gliomas are among the brain lesions that can mimic metastasis since both exhibit abnormal enhancement on the postcontrast scans, mass effect, and edema in the surrounding brain [130, 131]. In the majority of cases, primary brain tumors present with an infiltrative pattern and demonstrate ill-defined boarders unlike metastatic tumors that are rounded and have well-defined margins. Exceptionally, hemangioblastomas, low-grade pilocytic astrocytomas, primary brain lymphomas, and rarely high-grade gliomas present on MR imaging with fairly well-circumscribed solid or cavitary masses and can be indistinguishable from metastasis. Often times, the family history, the age of the patient, the location of the tumor, and a number of other clinical considerations may provide guidance for the correct diagnosis. In some cases though, stereotactic biopsy remains the last resort in establishing the diagnosis of solitary metastasis and rule out these rare tumors or infectious processes which may also present with similar characteristics on the postcontrast scans. Meningiomas, among the most common of intracranial tumors, present with a flat surface of dural attachment and a convex surface compressing the
391
adjacent brain. Metastatic tumors on the surface of the brain can become flattened by the firm calvarian bones or by the falx and present with a configuration similar to meningiomas. In such cases, the search for other smaller brain lesions becomes imperative in establishing the correct diagnosis since metastatic tumors are so often multiple. Also metastatic brain tumors are often cavitary whereas meningiomas develop cavities rarely and usually when large. Neuromas arising from the cranial nerves can also simulate metastases. These tumors are found along the paths of the cranial nerves in the cerebellopontine angle cisterns, the parasellar regions, and the jugular foramina. Neuromas present with signal intensity characteristics and enhancement similar to metastatic tumors. In addition, they produce smooth bone erosions, a characteristic feature of their benign nature. Metastatic skull tumors, on the other hand, invade adjacent bone in a permeative fashion, causing irregular erosions. Sometimes, patients with meningeal carcinomatosis present with infiltration of the leptomeninges in the internal auditory canals and mimic acoustic schwannomas. In these cases though, usually, there is meningeal carcinomatosis evident in other areas of the intracranial cavity that establishes the correct diagnosis. Germ cell tumors or tumors of the pineal body are extraaxial tumors that can be mistaken for metastasis. The location of these tumors in the suprasellar or the quadrigeminal cisterns and the usually young age of the patients provide clues for the diagnosis [132]. Metastatic tumors can rarely develop in the pituitary gland or in the stalk. Nonmalignant lesions in these locations include pituitary adenomas, sarcoidosis, histiocytosis x, and a variety of infections. These lesions share common imaging features with metastasis and in the absence of clinical information, the differentiation from them can be very difficult [133–135]. Primary tumors from the nasopharynx, the paranasal cavities, the orbits, and the skull can invade the bone and may produce an intracranial mass. These masses are usually confined in the epidural space or may invade an adjacent dural sinus. Metastatic neoplasms in these locations demonstrate imaging features similar to primary tumors and distinction between the two is often problematic. In such cases, the patient’s history and other clinical and laboratory findings are essential for diagnosis. When in doubt, biopsy procedures of the extracranial portion of the tumor provide the ultimate tool to verify the nature of these tumors.
392
A large variety of infectious diseases (bacterial, fungal, protozoan, parasitic, and viral) can involve the brain. Such brain lesions have CT and MRI features similar to tumors and since their treatment is so different, distinction between the two is imperative. Similar to tumors, infections in the brain may present as nodular or cavitary enhancing lesions with edema in the surrounding brain parenchyma and mass effect. Experience has shown that temporal evolution of CT and MRI findings are the most reliable signs in distinguishing infectious from neoplastic lesions. Early in its development, the infectious process presents as cerebritis with ill-defined margins, low density on CT, hypointensity on T1-, and hyperintensity on T2-weighted MRI sequences. At this stage, postcontrast studies demonstrate either no enhancement or mild heterogeneous enhancement. As the process continues to evolve, there is tissue destruction, breakdown of the BBB, and cavity formation. Conversely, metastatic tumors begin as a solid lesion, which become larger with time and eventually cavitate. At this late stage, both infectious and metastatic lesions demonstrate ring-like enhancement on the postcontrast T1-weighted studies. It has been pointed out that the cavitary lesions of the infectious processes have smooth wall, whereas the cavities of the neoplasms are irregular. Furthermore, it has been observed that on the T2-weighted scans, the walls of abscesses are relatively hypointense due to the presence of paramagnetic species within the inflammatory tissue. Although both these observations may be helpful, neither provides an infallible criterion for the correct diagnosis, since it has been shown that cavitary tumors can have thin, smooth walls, and certain neoplasms such as metastatic adenocarcinomas demonstrate low signal on the T2-weighted studies [136]. Certain infectious brain lesions and particularly granulomatous diseases often present with a cavitary or nodular enhancing mass that remains unchanged for an indefinite period of time. Unlike metastatic tumors that show well-demarcated borders on the postcontrast studies, the periphery of the granulomas tends to be less well-defined. Similar to tumors, granulomas may be associated with edema in the adjacent brain, which make these lesions more prone to develop seizures. In long-standing granulomas, calcifications may be found on CT but typically missed on MRI. Since this finding presents a strong supporting feature of the granulomatous nature of the lesion, a CT examination of the brain is imperative every time granuloma is considered as a possible diagnosis of a focal abnormality on an MRI
N.J. Patronas
scan [137, 138]. It should be noted though, that some malignant neoplasms may also calcify when metastasizing to various organs including brain. This has most often been observed in osteosarcomas. Since there is considerable overlap in the imaging presentation of the neoplasms and infectious processes, various investigators have sought other methods of establishing diagnosis short of biopsy. PET and SPECT studies have shown that in the vast majority of cases, the neoplastic lesions demonstrate increased concentration of the radiotracer where the infectious diseases are, as a rule, photopenic. Furthermore, MRS studies can be used to support the diagnosis of metastasis if, in the lesion under question, there is increased choline or lipids. When the meninges are involved by neoplasm or infection the distinction between the two is problematic. In both instances the meninges are abnormally thickened and enhance intensely. When the meninges are infiltrated by either of these two conditions, there may be an accompanied involvement of the ependymal lining of the ventricular walls. Sometimes, nodular formations from the meninges or from the ventricular wall invaginate into the adjacent brain parenchyma. Such finding is strongly suggestive of neoplasm although sarcoidosis or other granulomatous diseases may also present with a similar fashion. The overlapping imaging features of these entities and their potential coexistence calls for analysis of CSF for definitive diagnosis. Vascular occlusive disease of the cerebral arteries leading to infarctions and spontaneous intracerebral hemorrhage could at times present with a diagnostic dilemma. Such events have abrupt rather that insidious onset and in the history of these patients, one can find vascular risk factors rather that a neoplastic disease. From the imaging point of view during the acute phase of a cerebral ischemic infarction, there is noted decreased blood perfusion in the territory of the occluded vessel that can be demonstrated by PET, SPECT, or MRI. Diffusion-weighted MRI studies demonstrate restricted diffusion in the infarcted area that is usually smaller than the perfusion defect. Both perfusion and diffusion abnormalities are detected in the absence of any abnormality on conventional T1 and T2-weighted techniques or on CT. Within a few hours, edema will develop in the infarcted area, the extent of which depends on the size of occluded vessel. Edema is best appreciated by FLAIR technique and is accompanied by mass effect that may be very subtle. Abnormal enhancement appears within days post ictus
13 Brain Metastasis
when neovascularity with disrupted BBB develops [139–142]. During healing, the edema and the abnormal enhancement gradually disappear. In the last stage of healing, a gliotic scar will be formed in the infarcted area characterized by complete lack of enhancement and loss of cerebral mantle producing signs of focal atrophy. This sequence of events is not encountered in malignant brain tumors, which demonstrate increased perfusion due to prominent vascularity and variable diffusion depending on the cellularity with the most cellular tumors having restricted diffusion [118]. Furthermore, there is a striking difference between the enhancement of the infarction, which is heterogeneous and has a curvilinear or a gyral pattern, and that of the tumor that is nodular and well circumscribed. Finally, the abnormal enhancement in the metastatic brain tumors is observed at the onset of the neurologic deficit. Spontaneous brain hemorrhage cannot be easily distinguished from a hemorrhagic metastatic tumor. Both lesions can present abruptly with a neurologic deficit of sudden onset and show similar imaging features on both CT and MRI. In such cases, it should be noted that there is a predilection of spontaneous hematomas in the basal ganglia, whereas the majority of the metastatic tumors occur in the corticomedullary junction. Besides this anatomic consideration, the postcontrast studies may reveal abnormal enhancement in a nonhemorrhagic part of the lesion favoring neoplasm rather than hemorrhage. Definite clue for the correct diagnosis may also be provided in the postcontrast scans by the presence of another nonhemorrhagic but enhancing lesion in a different region of the brain. Finally, the natural history of the two types of hemorrhagic lesions is apparently different since one eventually resolves leaving behind a gliotic scar or a porencephalic cavity and the other relentlessly progresses with time. Vascular brain abnormalities only rarely mimic metastatic tumors. The most common diagnostic dilemma occurs with cavernous hemangiomas that present as small, rounded, space-occupying lesions and abnormal enhancement on the postcontrast CT and MRI studies. These congenital vascular lesions appear slightly hyperdense on precontrast CT and hypointense on the precontrast T1-weighted MRI scans. The most characteristic appearance of hemangiomas is shown on the T2-weighted images, where they usually demonstrate mixed signal intensities caused by the stagnating blood pool in the cavernous component of the lesion and by hemosiderin deposition, a product of hemoglobin
393
degradation. Also, characteristic feature of cavernous hemangiomas is the absence of mass effect and the lack of edema in the surrounding brain (Fig. 13.20). Arteriovenous malformation (AVM) is another congenital brain vascular lesion that is less likely to be mistaken for tumor. These vascular anomalies consist of a nidus with multiple dilated and tortuous abnormal vessels through which there is rapid blood flow due to anomalous arteriovenous shunting. On precontrast CT scans, arteriovenous malformations appear slightly hyperdense with respect to normal brain and on the postcontrast images they enhance intensely. The rapid blood flow through the dilated vessels of the AVM produces signal voids on both T1 and T2-weighted MRI studies representing the cardinal feature of AVMs. Because the flow pattern of the AVMs is uneven, areas of relatively slow flow will show increased enhancement on the postcontrast MRI scans, while in areas with more rapid flow signal voids prevail. Similar to cavernous hemangiomas uncomplicated AVMs present without mass effect or edema. In case of hemorrhage, mass effect and edema develop and the hematoma partially obscures these lesions. Overall, the imaging features described above is so characteristic that separation of AVMs and hemangiomas from tumors can be achieved quite easily by either CT or MRI. Giant cerebral aneurysms are vascular space-occupying lesions that, as a rule, are not mistaken for tumors. These lesions, similar to AVMs, produce signal voids on MRI and appear slightly hyperdense on the precontrast CT. Unlike AVMs, giant aneurysms have well-circumscribed walls and a single cavity that is often partially filled with clotted blood. On the wall of the aneurysm, there may be calcium deposition presented best on CT as a curvilinear density. The flowrelated phenomena, the partially thrombosed lumen, and the relationship of this lesion to a parent cerebral artery provide enough diagnostic features which distinguish this lesion from metastasis quite easily [143]. Multiple sclerosis (MS) is another brain abnormality that can occasionally mimic metastatic brain disease. The younger age of onset and female preponderance found in MS patients may provide initial diagnostic clues. On MRI and CT studies, the new MS lesions present with either ring-like or solid enhancement very similar to metastatic tumors. The main diagnostic feature that assists the observer in the correct diagnosis is the finding that the T2-weighted scans show many more plaques than the enhancing active lesions shown on the postcontrast T1-weighted
394 Fig. 13.20 (a) Precontrast CT scan of the brain in a patient with cavernous hemangioma mimicking brain tumor. Note the relatively hyperdense lesion in the right frontal lobe (arrow). The increased density is due to the blood pool of the hemangioma. (b) On the precontrast T1-weighted MRI scan the lesion is hypointense (arrow). (c) On the postcontrast scan there is evident abnormal enhancement (arrow). (d) On the T2-weighted scan the lesion is hyperintense. Note the absence of edema in the surrounding brain and the lack of mass effect
N.J. Patronas
a
c
studies. This represents a distinct feature of MS not found in metastatic brain tumors. Furthermore, most MS lesions are within periventricular rather than subcortical white matter – an unusual distribution for metastatic tumors (Fig. 13.21).
13.7 The Role of Imaging in the Posttherapy Period In the posttreatment period, imaging plays an important role for assessing the results of therapy and detecting complications. The primary focus of attention during
b
d
this period is the change in the size of the metastatic lesion before and after treatment. To this goal, on the postcontrast CT or MRI studies, the two greatest perpendicular diameters of the treated tumor is measured or a segmentation technique is applied to evaluate tumor volume. For tumors treated by conventional surgery or radiosurgery, it is important that the imaging study be obtained in the immediate posttherapy period to serve as a baseline. This is particularly important for surgically treated tumors since abnormal enhancement in the tumor bed detected within 24 h postsurgery usually implies residual tumor. In the days or weeks following the resection, granulation tissue develops in the surgical bed that
395
13 Brain Metastasis Fig. 13.21 (a) Postcontrast T1-weighted MRI scan in a patient with multiple sclerosis mimicking tumor metastasis. Three lesions are identified; one with ring-like enhancement and two smaller with solid enhancement (arrows). All three lesions represent demyelinating plaques in acute stage and are indistinguishable from metastatic tumors. (b) On the T2-weighted scan disproportionally larger numbers of plaques are present providing a clue to the correct diagnosis
a
enhances on postcontrast studies and is not distinguishable from recurrent tumor by CT or MRI. In patients in whom the resection had been complete, the granulation tissue will gradually subside and the enhancement will fade out, while in patients with residual tumor, the abnormal enhancement will increase with time. In patients with solitary metastatic brain tumors not accessible to surgical resection, radiosurgery remains an alternative option. The results of this type of treatment are not immediately apparent. In fact, if a postcontrast MRI scan is performed within the first 2 months after radiosurgery, the area of enhancement may be increased as compared to the pretreatment examination. This observation is well recognized and it should not be interpreted as evidence of disease progression since it is due to inflammatory reaction or necrosis of the radiated tissues. Eventually, the patients with favorable response to treatment will demonstrate decrease or complete disappearance of the abnormal enhancement. In assessing treatment response of cavitary tumors following radiation therapy it is not enough to demonstrate the decrease in the thickness of the enhancing rim. The overall size of the treated tumor must also become smaller for an accurate documentation of successful treatment. Increasing size of the cavity represents a feature of disease progression even if the thickness of the enhancing rim is smaller or unchanged. In metastatic tumors treated with irradiation, dystrophic calcifications often develop at the site of the tumor – a sign of good response to therapy. This finding can be appreciated only on CT scans. Radiation necrosis is a well-known complication in the treatment of brain tumors. In such cases, imaging
b
studies with either CT or MRI will demonstrate increasing area of abnormal enhancement and greater mass effect, both features of tumor progression. PET with FDG can assist in solving the diagnostic dilemma. If the questionable area appears hypermetabolic on FDG-PET recurrent tumor is the most probable diagnosis. If it is hypometabolic, either radiation necrosis alone or necrosis with quiescent tumor is present. The absence of increased metabolic activity does not entirely exclude residual tumor [144, 145]. Brain images obtained after treatment of intracranial tumors often demonstrate changes in the brain parenchyma distant to the tumor bed. Patients receiving whole-brain irradiation for an intracranial malignancy are at risk of developing abnormalities in the white matter. Children and elderly patients are more vulnerable to such a complication – children because of the injury to immature white matter, and elderly because of preexisting vascular arteriosclerosis presenting an additional burden of treatment. Chemotherapy, particularly when administered intrathecally or intraventricularly, can damage the white matter and produce symptoms of CNS dysfunction. The combined effect of cranial irradiation and chemotherapy increases the incidence and the severity of this complication. MRI using T2-weighted or FLAIR techniques provides the most sensitive method in demonstrating these lesions. They appear as multiple focal areas of increased signal that eventually become confluent and extend in large regions of the white matter. It is currently believed that these abnormalities in the early stage represent edema or acute, partial demyelination that can be transient if detected
396
early and the causative agent is discontinued. Often times though, the white matter abnormalities become permanent areas of leukomalacia with irreversible loss of myelin and superimposed gliosis. Brain irradiation and chemotherapy can also lead to diffuse brain atrophy manifested by widening of the cortical sulci and ventricular enlargement. Leukomalacia and brain atrophy, alone or in combination, are commonly associated with mental decline that is clinically obvious or, in mild forms, can be diagnosed only with psychometric testing. MRS has been proposed to provide additional information for the detection of metabolic changes in the white matter after such treatments [146]. It is still unclear though whether this methodology will play a useful role in prevention, early detection, or management of these treatment-related abnormalities. Vascular complications can also occur during either irradiation or chemotherapy. During the induction period, patients treated with systemic chemotherapy pass through a hypercoagulable state and may develop intravascular thrombosis that affects primarily the venous side of the circulation. This leads to the development of focal edema of both gray and white matter and neurologic complications that include seizures, cortical blindness, confusion, or motor deficits. On other occasions, chemotherapy drugs and cyclosporine have been incriminated in causing loss of autoregulation of the cerebral microcirculation, which results in brain edema, accompanied by a variety of clinical manifestations that mimic those of patients with eclampsia. The presence of edema is best appreciated on the T2-weighted and on the FLAIR techniques, while abnormal enhancement is not a feature of these lesions. Both clinical and imaging abnormalities can be reversible although chances of complete recovery are inversely proportional to their duration. Vasculopathy is another complication that can result from whole-brain irradiation, which damages the wall of the small vessels with the endothelial cells being most vulnerable. This type of injury can be diagnosed on the T2-weighted MRI scans showing scattered hypointense lesions in the brain parenchyma that are due to hemosiderin deposits secondary to microscopic hemorrhage. On rare occasions the damaged vessels may actually rupture and produce a frank hematoma. Vasculopathy can also manifest itself after healing of the acute postirradiation vascular injury by excessive calcium deposition on the wall of the damaged vessels. This abnormality affects preferentially the small
N.J. Patronas
vessels in the basal ganglia and the corticomedullary junction and is best appreciated by precontrast CT scan examinations. In conclusion, a number of imaging modalities are available to evaluate intracranial metastatic lesions. Experience has shown that postcontrast T1-weighted MRI technique is the most appropriate method to establish such a diagnosis. This technique in conjunction with FLAIR and T2-weighted technique provide accurate information vital in managing these patients as to the size, location, and number of such lesions. Other imaging studies such as CT, PET, MRS, and SPECT are also available. These modalities are appropriate only to address specific problems related to brain metastasis and evaluate response to treatment or possible treatment-related complications.
References 1. Henson RA, Urich H (1973) Cancer and the nervous system. The neurological manifestation of systematic malignant disease. Blackwell, Oxford, pp 7–58 2. Willis RA (1973) The spread of the tumors of the human body, 3rd edn. Butterworths, London, pp 251–258 3. Cairncross G, Posner J (1983) The management of brain metastasis. In: Walker MD (ed) Oncology of the nervous system. Martinus Nijhoff, Boston, pp 341–377 4. Posner JB, Chernik NL (1978) Intracranial metastasis from systemic cancer. Adv Neuro l18:579–592 5. Patchell RA (1991) Brain Metastases. Neurol Clin 9:817–824 6. Black MP (1991) Brain tumors. N Engl J Med 324:1555–1564 7. Russell DS, Rubinstein LJ (1989) Pathology of tumors of the nervous system, 5th edn. Williams & Wilkins, Baltimore, pp 825–842 8. Liotta LA, Rao NC, Terranova VP, Barsky S, Thorgeirsson U (1984) Tumor cell attachment and degradation of basement membranes. In: Nicolson GL, Milas L (eds) Cancer invasion and metastasis: biologic and therapeutic aspects. Raven, New York, pp 169–176 9. Katz DA, Liotta LA (1986) Tumor invasion and metastasis in the central nervous system. Prog Neuropathol 6:119–131 10. Nicolson GL, Irimura T, Nakajima M, Estrada J (1984) Metastatic cell attachment to and invasion of vascular endothelium and its underlying basal lamina using endothelial cell monolayers. In: Nocolson GL, Milas L (eds) Cancer invasion and metastasis biologic and therapeutic aspects. Raven, New York, pp 145–167 11. Liotta LA, Kohn E (1990) Cancer invasion and metastasis. JAMA 23:135–150 12. Folkman J (1986) How is blood vessels growth regulated in normal and neoplastic tissues? Twenty-sixth G.H.A. Clowes Memorial award lecture. Cancer Res 46:467–473 13. Folkman J (1992) The role of angiogenesis in tumor growth. Semin Cancer Biol 3:65–71
13 Brain Metastasis 14. Fidler IJ (2002) The organ microenvironment and cancer metastasis. Differentiation 70:498–505 15. Fidler IJ, Yano S, Zhang RD, Fujimaki T, Bucana CD (2002) The seed and soil hypothesis: vascularisation and brain metastasis. Lancet Oncol 3:53–57 16. Olson ME, Chernick NL, Posner JB (1974) Infiltration of the leptomeninges by systemic cancer: a clinical and pathological study. Arch Neurol 30:122–137 17. Gonzalez-Vitale JC, Garcia-Benuel R (1976) Meningeal carcinomatosis. Cancer 37:2906–2911 18. Kokkoris PC (1983) Leptomeningeal carcinomatosis: how does cancer reach the pia-arachnoid. Cancer 51:154–160 19. O’Neil BP, Buckner JC, Coffey RJ et al (1994) Brain metastatic lesions. Mayo Clin 69:1062–1068 20. Hounsfield GN (1973) Computerized transverse axial scanning tomography. Part I. Description of the system. Br J Radiol 46:1016–1022 21. Boyd DP, Parker DL (1983) Basic principles of computed tomography. In: Moss AA, Gamsu GE, Genant HK (eds) Computed tomography of the body. WB Saunders, Philadelphia, pp 1–21 22. Deck MDF, Messima AV, Sackett JF (1976) Computed tomography in metastatic disease of the brain. Radiology 119:115–120 23. Kane RC (1978) Brain scans for metastasis. JAMA 239:2115–2116 24. Potts DG, Abbott GF, von Sneidern JV (1980) National cancer institute study: evaluation of computed tomography in the diagnosis of intracranial neoplasms: III. Metastatic tumors. Radiology 136:657–664 25. Dupont MG, Baleriaux-Waha D, Kuhn G, Bollaert a, Jeanmart I (1981) Computerized axial tomography in the diagnosis of cerebral metastasis. Comput Tomogr 5:103–113 26. Lee YY, Glass JP, Geoffray A, Wallace S (1984) Cranial computed tomographic abnormalities in leptomeningeal metastasis. AJNR Am J Neuroradiol 5:559–563 27. Davis PC, Hudgins PA, Peterman SB, Hoffman JC Jr (1991) Diagnosis of cerebral metastases: double-dose delayed CT vs contrast-enhanced MR imaging. AJNR Am J Neuroradiol 12:293–300 28. Hayman LA, Evans RA, Hink VC (1980) Delayed high iodine dose contrast computed tomography in cranial neoplasms. Radiology 136:677–684 29. Shehadi WH (1982) Contrast media adverse reaction: occurance, recurrence and distribution patterns. Radiology 143:11–17 30. Lasser EC, Berry CC, Talner LB et al (1987) Pretreatment with corticosteroids to alleviate reactions to intravenous contrast material. N Engl J Med 317:845–849 31. Katayama H, Yamaguchi K, Kozuka T, Takashima T (1990) Adverse reactions to ionic and nonionic contrast media: a report from the Japanese Committee of Safety of Contrast Media. Radiology 175:621–628 32. McClennan BL (1990) Ionic and nonionic iodinated contrast media: evolution and strategies for use. AJR Am J Roentgenol 155:255–263 33. Damadian R (1971) Tumor detection by nuclear magnetic resonance. Science 171:1151–1153 34. Lauterbur PC (1973) Image formation by induced local interactions: examples employing nuclear magnetic resonance. Nature 242:190–191
397 35. Mansfield P, Maudsley AA (1977) Medical imaging by NMR. Br J Radiol 50:188–194 36. Pickett IL, Newhouse JH, Buonanno FS et al (1982) Principles of nuclear magnetic resonance imaging. Radiology 143:157–168 37. Brant-Zawadski M, Berry I, Osaki L et al (1986) Gd-DTPA in clinical MR brain. I. Intra-axial lesions. AJNR Am J Neuroradiol 7:781–788 38. Russel EJ, Geremia GK, Johnson CE et al (1987) Multiple cerebral metastases: detectability with Gd-DTPA-enhanced MR imaging. Radiology 165:609–617 39. Tice HM, Jones KM, Mulkern RV et al (1993) Fast spinecho imaging of intracranial neoplasms. J Comput Assist Tomogr 17:425–431 40. Carrier DA, Mawad ME, Kirkpatrick JB, Schmid MF (1994) Metastatic adenocarcinoma to the brain: MR with pathologic correlation. AJNR Am J Neuroradiol 15:155–159 41. Abdullah ND, Mathews VP (1999) Contrast issues in brain tumor imaging. Neuroimaging Clin North Am 9(4):733–749 42. Runge VM, Wells JW, Nelson KL, Linville PM (1994) MR imaging detection of cerebral metastases with a single injection of high-dose gadoteridol. J Mang Reson Imaging 4:669–673 43. Van Dijk P, Sijens PE, Schmitz PI, Oudkerk M (1997) Gd-enhanced MR imaging of brain metastases: contrast as a function of dose and lesion size. Magn Reson Imaging 15(5):535–541 44. Niendorf HP, Laniado M, Semmler W, Schorner W, Felix R (1987) Dose administration of gadolinium-DTPA in MR imaging of intracranial tumors. AJNR Am J Neuroradiol 8:803–815 45. Yuh WTC, Engelken JD, Muhonen MG, Mayr NA, Fisher DJ, Ehrhardt JC (1992) Experience with high-dose gadolinium MR imaging in the evaluation of brain metastasis. AJNR Am J Neuroradiol 13:335–345 46. Yuh WTC, Fisher DJ, Runge VM et al (1994) Phase III multicenter trial of high-dose gadoteridol in MR evaluation of brain metastasis. AJNR Am J Neuroradiol 15:1037–1051 47. Kuhn MJ, Hammer GM, Swenson LC, Youssef HT, Gleason TJ (1994) MRI evaluation of solitary brain metastases with triple-dose gadoteridol: comparison with contrast enhanced CT and conventional-dose gadopentetate MRI studies in the same patients. Comput Med Imaging Graph 18:391–399 48. Akeson P, Vikhoff B, Stahlberg F et al (1977) Brain lesion contrast in MR imaging dependence on field strength and concentration of gadodiamide injection in patients and phantoms. Acta Radiol 38:14–18 49. Chang KH, Ra DR, Han MH et al (1994) Contrast of brain tumors at different MR field strength: Comparison of 0.5 and 2.0 T. AJNR Am J Neuroradiol 15:1413–1419 50. MathewsVP CKS, Lowe MJ et al (1999) Gadolinium-enhanced fast FLAIR imaging of the brain. Radiology 211:257–263 51. Melhem ER, Bert RJ, Walker RE (1998) Usefulness of optimized gadolinium-enhanced fast fluid attenuated inversion recovery MR imaging in revealing lesions of the brain. AJR Am J Roentgenol 171:803–807 52. Grossman RI, Gomori JM, Ramer KN, Lexa FJ, Schnall MD (1994) Magnetization transfer: theory and clinical applications in neuroradiology. Radiographics 14:279–290 53. Boorstein JM, Wong KT, Grossman RI, Bolinger L, McGowan JC (1994) Metastatic lesions of the brain imaging with magnetization transfer. Radiology 191:799–803
398 54. Kurki TLI, Niemi PT, Lundbom N (1992) Gadoliniumenhanced magnetization transfer contrast imaging of intracranial tumors. J Magn Reson Imaging 2:401–406 55. Finelli DA, Hurst GC, Gullapali RP, Bellon EM (1994) Improved contrast enhancing brain lesions on post gadolinium, T1-weighted spin-echo images with use of magnetization transfer. Radiology 190:553–559 56. Schorner W, Laniado M, Niendorf HP, Schuber C, Felix R (1986) Time dependent changes in image contrast in brain tumors after gadolinium-DTPA. AJNR Am J Neuroradiol 7:1013–1020 57. Akeson P, Nordstrom CH, Holtas S (1997) Time-dependency in brain lesion enhancement with gadodiamide injection. Acta Radiol 38:19–24 58. Sadowski EA, Bennett LK, Chan MR et al (2007) Nephrogenic systemic fibrosis: risk factors and incidence estimation. Radiology 243:148–157 59. Thomsen HS (2007) ESUR guideline: gadolinium-bases contrast media and nephrogenic systemic fibrosis. Eur Radiol 17:2692–2696 60. U.S. Food and Drug Administration (2006) Public health advisory: gadolinium-containing contrast agents for magnetic resonance imaging (MRI) – Omniscan, OptiMARK, Magnevist, Prohance, and MultiHance. www.fda.gov/cder/ drug/advisory/gadolinium-agents.htm (Published 8 June 2006. Update 22 Dec 2006. Accessed 7 Dec 2006) 61. Hayashida Y, Hirai T, Morishita S et al (2006) Diffusionweighted imaging of metastatic brain tumors: comparison with histologic type and tumor cellularity. AJNR 27:1419–1425 62. Kono K, Inoue Y, Nakayama K et al (2001) The role of diffusion-weighted imaging in patients with brain tumors. AJNR 22:1081–1088 63. Bulakbasi N, Kocaoglu M, Farzaliyev A et al (2005) Assessment of diagnostic accuracy of perfusion MR imaging in the primary and metastatic solitary malignant tumors. AJNR 26:2187–2199 64. Law M, Soonmee C, Knopp EA et al (2002) High-grade gliomas and solitary metastases: differentiation by using perfusion and proton spectroscopic MR imaging. Radiology 222:715–721 65. Kremer S, Grand S, Berger F et al (2003) Dynamic contrastenhanced MRI:differentiating melanomas and renal carcinoma metastases from high-grade gliomas and other metastases. Neuroradiology 45:44–49 66. Cha S, Lupo JM, Chen MH et al (2007) Differentiation of glioblastoma multiforme and single metastasis by peak heght and percentage of signal intensity recovery derived from susceptibility-weighted contrast enhanced perfusion MR imaging. AJNR 28:1078–1084 67. Long DM (1979) Capillary ultrastructure in human metastatic brain tumors. J Neurosurg 51:53–58 68. Davis PC, Friedman NC, Fry SM, Maldo JA, Hoffman JC, Braun IF (1987) Leptomeningeal metastasis: MR imaging. Radiology 163:449–454 69. Yousem DM, Patrone MP, Grossman RI (1990) Lepto meningeal metastasis: MR evaluation. J Comput Assist Tomogr 14:255–261 70. Sze G (1993) Diseases of the intracranial meninges: MR imaging features. AJR 160:727–733 71. Paako E, Patronas NJ, Schellinger D (1990) Meningeal Gd-DTPA enhancement in patients with malignancies. J Comput Assist Tomogr 14:542–546
N.J. Patronas 72. Phillips ME, Ryals TJ, Kambhu SA, Yuh WTC (1990) Neoplastic vs inflammatory meningeal enhancement with Gd-DTPA. J Comput Assist Tomogr 14:536–541 73. Hustinx R, Alavi A (1999) SPECT and PET imaging of brain tumors. Neuroimaging Clin North Am 9(4):751–765 74. Reba RC, Holman BL (1991) Brain perfusion radiotracers. In: Diksie M, Reba RC (eds) Radiopharmaceuticals and brain pathology studied with PET and SPECT. CRC, Boston, pp 35–65 75. Neurinckx RD, Canning LR, Piper IM et al (1987) Technitium-99m d, 1-HM-PAO: a new radiopharmaceutical for SPECT imaging of regional cerebral blood perfusion. J Nucl Med 27:191–202 76. Biersack HJ, Grunwald F, Kropp J (1991) Single photon emission computed tomograph imaging of the brain tumors. Semin Nucl Med 21:2–10 77. Kim KT, Black KL, Marciano D et al (1990) Thalium-201 SPECT imaging of brain tumors; methods and results. J Nucl Med 31:965–969 78. Black KL, Hawkins RA, Kim KT et al (1989) Use of thallium-201 SPECT to quantutate malignancy grade of gliomas. J Neurosurg 71:342–346 79. Ueda T, Kaji Y, Wakisaka S et al (1993) Time sequential single photon emission computed tomography studies in brain tumor using thallium-201. Eur J Nucl Med 20:138–145 80. Dierckx RA, Martin JJ, Dobbelieir A, Crols R, Neetens I, De Deyn PP (1994) Sensitivity and specificity of thalium-201 single-photon emission tomography in the functional detection and differential diagnosis of brain tumors. Eur J Nucl Med 21:621–633 81. Schwartz RB, Calvaho PA, Alexander ED et al (1991) Radiation necrosis vs high grade glioma: differentiation by using dual-isotope SPECT with 201Th and 99m Tc-HMPAO. AJNR Am J Neuroradiol 12:1187–1192 82. Lorberboym M, Mandell LR, Mosesson RE et al (1997) The role of thallium-201 uptake and retention in intracranial tumors after radiotherapy. J Nucl Med 38:223–226 83. Kosuda S, Fujii H, Aoki S et al (1994) Prediction of survival in patients with suspected recurrent cerebral tumors by quantitative thallium-201 single photon emission computed tomography. Int J Radiat Oncol Biol Phys 30:1201–1206 84. Vertosick FT Jr, Selker RG, Grossman SJ (1994) Correlation of thallium-201 single photon emission computed tomography and survival after treatment failure in patients with glioblastoma multiforme. Neurosurgery 34:396–401 85. Yoshii Y, Satou M, Yamamoto T et al (1993) The role of thalium-201 single photon emission tomography in the investigation and characterization of brain tumors in man and their response to treatment. Eur J Nucl Med 20:39–45 86. Hirano T, Otake H, Kazama K et al (1997) Technetium-99m (V)-DMSA and thallium-201 in brain tumor imaging: correlation with histology and malignant grade. J Nucl Med 38:1741–1749 87. Langen KJ, Coenen HH, Roosen N et al (1990) SPECT studies of brain tumors with L-3-(123I) iodo-alpha-methyl tyrosine: correlation with PET 124IMT and first clinical results. J Nucl Med 31:281–286 88. Kuwert T, Morgeroth C, Woesler B et al (1996) Uptake of iodine-123-alpha-methyl tyrosine by gliomas and non-neoplastic brain lesions. Eur J Nucl Med 23:1345–1353
13 Brain Metastasis 89. Andrews DW, Das R, Kim S et al (1997) Technitium- MIBI as a glioma imaging agent for the assessment of multi-drug resistance. Neurosurgery 40:1323–1332 90. Yokogami K, Kawano H, Moriyama T et al (1998) Applications of SPECT using technitium-99m sestamibi in brain tumors and comparison with expression of MDR-1 gene: Is it possible to predict the response to therapy in patients with gliomas by means of 99mTc-sestamibi SPECT? Eur J Nucl Med 25:401–409 91. Sokoloff L, Reivich M, Kennedy C et al (1977) The (14C) deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem 28:897–916 92. Di Chiro G, De LaPaz RL, Brooks RA et al (1982) Glucose utilization of cerebral gliomas measured by (18F) fluorodeoxyglucose and positron emission tomography. Neurology 32:1323–1329 93. Patronas NJ, Di Chiro G, Kufta C et al (1985) Prediction of survival in glioma patients by means of positron emission tomography. J Neurosurg 62:816–822 94. Alavi JB, Alavi A, Chawluk J et al (1988) Positron emission tomography in patients with glioma: a predictor of prognosis. Cancer 62:1074–1078 95. Barker FG II, Chang SM, Valk PE et al (1997) 18-Fluorodeoxyglucose uptake and survival of patients with suspected recurrent malignant glioma. Cancer 79:115–126 96. Holzer T, Herholz K, Jeske J et al (1993) FDG-PET as a prognostic indicator in radiochemotherapy of glioblastoma. J Comput Assist Tomogra 17:681–687 97. Griffeth JK, Rich KM, Dehdashti F et al (1993) Brain metastasis from non-central nervous system tumors: evaluation with PET. Radiology 186:37–44 98. Lassen U, Andersen P, Daugaard G et al (1998) Metabolic and hemodynamic evaluation of brain metastases from small cell lung cancer with positron emission tomography. Clin Cancer Res 4(11):2591–2597 99. Ericson K, Kihlstrom L, Morgard J et al (1996) Positron emission tomography using 18F-fluorodeoxyclucose in patients with steriotactically irradiated brain metastases. Stereotact Funct Neurosurg 66(suppl 1):214–224 100. Ogawa T, Shishido F, Kanno I et al (1993) Cerebral glioma; evaluation with methionine PET. Radiology 186:45–53 101. Borght TV, Pauwels S, Lambotte L et al (1994) Brain tumor imaging with PET and 2-(carbon-11) thymidine. J Nucl Med 35:974–982 102. Pruim J, Willemsren ATM, Molenaar WM et al (1995) Brain tumors: L-(1-C-11)tyrosine PET for visualization and qualification of protein synthesis rate. Radiology 197: 221–226 103. Lammertsma AA, Ito M, McKenzie CG et al (1981) Quantitative tomographic measurements of regional cerebral flow and oxygen utilization in patients with brain tumors using oxygen-15 and positron emission tomography. J Cereb Blood Flow Metab 1(suppl 1):S567–S568 104. Jones T, Chesler DA, Ter-Pogossian MM (1976) The continuous inhalation of oxygen-15 for assessing regional oxygen extraction in the brain of man. Br J Radiol 49:339–343 105. Frackoviak R, Lenzi G, Jones T et al (1980) Quantitative measurement of regional cerebral blood flow and oxygen metabolism in man using 15O and positron emission tomog-
399 raphy: theory, procedure and normal values. J Comp Assist Tomogr 4:727–736 106. Ito M, Lammertsma AA, Wise RJ et al (1982) Measurement of regional cerebral bloodflow and oxygen utilization in patients with cerebral tumors using 15O and positron emission tomography: analytical techniques and preliminary results. Neuroradiology 23:63–74 107. Ogawa T, Uemura K, Shishido F et al (1988) Changes of cerebral blood flow, and oxygen and glucose metabolism following radiochemotherary of gliomas: a PET study. J Comput Assist Tomogr 12:290–297 108. Mineura K, Sasajima T, Kowada M et al (1994) Perfusion and metabolism in predicting the survival of patients with malignant gliomas. Cancer 72:2386–2394 109. Jackson EF (1992) In vivo magnetic resonance spectroscopy in humans: a brief review. Am J Physiol Imaging 314:146–154 110. Meyerhoff DJ, MacKay s, Baker A, Schaefer S, Weiner MW (1992) Magnetic rasonance spectroscopy. In: Higgins CB, Hricak H, Helms CA (eds) Magnetic resonance imaging of the body, 2nd edn. Raven, New York, pp 287–302 111. Preul MC, Caramanos Z, Collins DL et al (1996) Acurate noninvasive diagnosis of human brain tumors by using proton magnetic resonance spectroscopy. Nat Med 2: 323–325 112. Fulham MJ, Bizzi A, Dietz MJ et al (1992) Mapping of brain metabolites with proton MR spectroscopic imaging; clinical relevance. Radiology 185:675–686 113. Negendank W (1992) Studies of human tumors by MRS: a review. NMR Biomed 5:303–324 114. Lee PL, Gonzalez RG (2000) Magnetic resonance spectroscopy of brain tumors. Curr Opin Oncol 12(3):199–204 115. Law M, Yang S, Wang H et al (2003) Glioma grading: sensitivity, specificity and predictive values of perfusion MR imaging and proton spectroscopic imaging compared with conventional MR imaging. AJNR 24:1989–1998 116. Astrakas LG, Zurakowski D, Tzika AA et al (2004) Noninvasive magnetic resonance spectroscopic imaging biomarkers to predict clinical grade of pediatric brain tumors. Clin Cancer Res 10:8220–8228 117. Sijens PE, Knop MV, Brunetti A et al (1995) 1H MR Spectroscopy in patients with metastatic brain tumors: a multicenter study. Magn Reson Med 33:818–826 118. Sijens PE, Levendag PC, Vecht CJ et al (1996) 1H MR spectrocopy detection of lipids and lactate in metastatic brain tumors. NMR Biomed 9(2):65–71 119. Poptani H, Rakesh K, Roy R et al (1995) Characterization of intracranial mass lesions with in vivo proton MR spectroscopy. AJNR Am J Neuroradiol 16:1593–1603 120. Kugel H, Heindel W, Ernestus RI, Bunke J, du Mesnil R, Friedman G (1992) Human brain tumors: spectral patterns detected with localized H-1 MR spectroscopy. Radiology 183:701–709 121. Tate AR, Majos C, Moreno A et al (2003) Automated classification of short echo time in vivo 1H brain tumor spectra: a multicenter study. Magn Reson Med 49:29–36 122. Devos A, Lukas L, Suykens JA et al (2004) Classification of brain tumors using short echo time 1H MR spectra. J Magn Reson 170:164–175 123. Opstad KS, Murphy MM, Wilkins PR et al (2004) Differentiation of metastases from high grade-gliomas using
400 short echo time 1H spectroscopy. J Magn Reson Imaging 20:187–192 124. Burtscher IM, Skagerberg G, Geijer B et al (2000) Proton MR spectroscopy and preoperative diagnostic accuracy: an evaluation of intracranial mass lesions characterized by stereotactic biopsy findings. AJNR 21:84–93 125. Lichy MP, Henze M, Plathow C et al (2004) Matabolic imaging to follow stereotactic radiation gliomas – the role of 1H spectroscopy in comparison to FDG-PET and IMTSPECT. ROFO 176:126–134 126. Plotkin M, Eisenacher J, Bruhn H et al (2004) 123I-IMT SPECT and 1H MR-spectroscopy at 3.0T in the differential diagnosis of recurrent or residual gliomas: a comparative study. J Neurooncol 70:49–58 127. Taylor JS, Langston JW, Reddick WE et al (1996) Clinical value of proton magnetic resonance spectroscopy for differentiating recurrent or residual brain tumor from delayed cerebral necrosis. Int J Radiat Oncol Biol Phys 36: 1251–1261 128. Chan YL, Yeung DK, Leung SF, Cao G (1999) Proton magnetic resonance spectroscopy of late delayed radiationinduced injury to the brain. J Magn Reson Imaging 19: 130–137 129. Warren KE, Frank JA, Black JL et al (2000) Proton magnetic resonance spectroscopic imaging in children with recurrent primary brain tumors. J Clin Oncol 18(8):1020–1026 130. Smirniotopoulos JG, Olmsted WW (1994) Primary and secondary neoplasms of the skull. In: Putman CE, Ravin CE (eds) Textbook of diagnostic imaging. WB Saunders, Philadelphia, pp 106–125 131. Atlas SW (1990) Adult supratentorial tumors. Semin Roentgenol 25:130–154 132. Smirniotopoulos JG, Rushing EJ, Mena H (1992) Pineal region masses: differential diagnosis. Radiographics 12: 577–596 133. Johnsen DE, Woodruff WW, Alen IS et al (1991) MR imaging of the sellar and juctasellar regions. Radiographics 11:727–758 134. Chong BW, Newton TH (1993) Hypothalamic and pituitary pathology. Radiol Clin North Am 31:1147–1183
N.J. Patronas 135. Seltzer S, Mark AS, Atlas SW (1991) CNS sarcoidosis: evaluation with contrast-enhanced MR imaging. AJNR Am J Neuroradiol 12:1227–1233 136. Zimmerman RD, Weingarten K (1991) Neuroimaging of cerebral abscesses. Neuroimaging Clin North Am 1:1–16 137. Bazan C III, Rinaldi MG, Rauch RR, Jinkins IR (1991) Fungal infections of the brain. Neuroimag Clin North Am 1:57–88 138. De Castro CC, Hesselink JR (1991) Tuberculosis. Neuro imaging Clin North Am 1:119–139 139. Bryan RN, Levy NM, Whitlow WD, Killian JM, Preziosi TJ, Rosario JA (1991) Diagnosis of acute cerebral infarction: comparison of CT and MR imaging. AJNR Am J Neuroradiol 12:611–620 140. Warach S, Gaa J, Siewert B, Wielopolski P, Delman RR (1995) Acute stroke studied by whole brain echo planar diffusion-weighted magnetic resonance imaging. Ann Neurol 37:231–241 141. Loubinoux I, Volk A, Borredon J et al (1997) Spreading of vasogenic edema and cytotoxic edema assessed by quantitative diffusion and T2 magnetic resonance imaging. Stroke 28:419–426 142. Castillo M, Smith JK, Kwock L, Wilber K (2001) Apparent diffusion coefficient in the evaluation if high grade gliomas. AJNR Am J Neuroradiol 22:60–64 143. Drake C, Peerless SJ (1997) Giant fusiform intracranial aneurysms; review of 120 patients treated surgically from 1965 to 1992. J Neurosurg 87:141–162 144. Patronas NJ, Di Chiro G, Brooks RA et al (1982) (18F) Fluorodeoxyglucose and positron emission tomography in the evaluation of radiation necrosis of the brain. Radiology 144:885–889 145. Mogard J, Kihlstrom L, Ericson K et al (1994) Recurrent tumor vs radiation effect after gamma knife radiosurgery of intracerebral metastases; diagnosis with PET-FDG. J Comp Assist Tomogr 18:177–181 146. Virta A, Patronas N, Raman R et al (2000) Spectroscopic imaging of radiation-induced effects in the white matter of glioma patients. Magn Reson Imaging 18:815–857
Scintigraphy for Brain Tumors
14
George N. Sfakianakis, Efrosyni Sfakianaki, and Hilton Gomes
Contents 14.1 Introduction............................................................. 401 14.2 Radiopharmaceuticals............................................ 402 14.2.1 General Information................................................. 402 14.2.2 Radiopharmaceuticals for the Evaluation of the Nervous System.............................................. 402 14.3 Instrumentation...................................................... 405 14.3.1 Single Photon Imaging (Collimated Anger Camera Principle) Planar/SPECT............................. 405 14.3.2 PET Imaging [Coincidence Principle and Time-of-Flight (TOF)]....................................... 405 14.4 Clinical Applications.............................................. 407 14.4.1 Preoperative Tissue Characterization....................... 407 14.4.2 Differentiation of Tumor from Infection.................. 412 14.4.3 Therapy Planning and Evaluation of Effectiveness of Treatment.............................................................. 412 14.4.4 Diagnosis of Recurrence (Recurrence vs. Radiation Necrosis)................................................................... 421 14.5 Therapy of Brain Tumors with Unsealed Radiopharmaceuticals............................................ 423 14.6 Conclusion............................................................... 423 References............................................................................ 424
G.N. Sfakianakis (*), E. Sfakianaki, and H. Gomes Department of Radiology, Division of Nuclear Medicine, University of Miami/Jackson Memorial Medical Center, 016960, Miami, FL 33101, USA e-mail:
[email protected]
14.1 Introduction Parmenides said that what cannot be thought, cannot be, therefore, what can be, can be thought. So it was that ancient Greek philosophers had thought of the atoms, and particularly, the radioactive atoms we use in Nuclear Medicine. Indeed, Democritos in the sixth century bc formulated the idea of the atoms as the indestructible smaller elements of the universe that combine among themselves to form the visible world; he thought of atoms on a philosophical basis as the explanation of the changes in the environment, which occur without the perishment of matter. Rearrangements of “atoms” could explain the changes around us and inside us. Two centuries later, Epicuros, as if anticipating the discovery of the radioactive atoms, introduced the idea of the “unstable” atom, which, after a period of instability, takes its final stable form. More than 2,000 years later, when science overtook these frontiers, John Dalton knew Democritos’ Atomic Theory of Matter and used it to explain chemical experiments. If the atom (=not possible to cut) can be cut and split into parts, it is not Democritos’ fault. Today we understand that by “atoms” Democritos actually meant the “quarks” or the “strings,” or perhaps some other, yet to be discovered, elemental particles. As for Henri Beckerel and Marie Curie, who were among the first to deal with radioactivity and the “unstable” or “radioactive atoms,” it is not known if they knew that the theoretical father of Nuclear Science was Epicuros. In Nuclear Medicine, we use the “radioactive atoms,” which are the “atoms” meant by Democritos (as applied by Dalton), in their unstable form, which was anticipated by Epicuros, for imaging of tissues or diseases and for therapy of malignant or benign diseases.
A. Drevelegas (ed.), Imaging of Brain Tumors with Histological Correlations, DOI: 10.1007/978-3-540-87650-2_14, © Springer-Verlag Berlin Heidelberg 2011
401
402
14.2 Radiopharmaceuticals 14.2.1 General Information In Nuclear Medicine, the patient is the source of the electromagnetic radiation, which is used for imaging, and the patient remains radioactive until decay or excretion of the source of radiation. All the other imaging modalities use external sources of electromagnetic radiation and the patient’s body is temporarily interacting with them. Indeed, for Nuclear Medicine imaging (or scintigraphy), the patient is injected with (or ingests) a radiolabeled substance, a radiopharmaceutical, an organic or inorganic molecule, labeled with suitable radioactive atoms with special properties, physical (appropriate for imaging) and chemical/biochemical (able to label without changing the biological recognition of the molecule) [1]. The physical properties include: (a) a relatively short, but long enough for the duration of the study, half life (T½) (few seconds to several days), (b) a decay by eventual emission of either photons that are within the range that allows imaging with the available instruments (>60 and <600 keV), or appropriate particles for therapy, (c) easy production and availability, (d) a “reasonable” cost, and (e) a decay to a daughter element, which does not contaminate the body of the patient or the environment. As clinically applied, four categories of radioisotopes, with some overlap, are used either in elemental form or as labeling tracers on larger molecules for imaging, therapy, or in vitro studies: Single photon emitters; they emit gamma or (after reaction with their peripheral electrons) X-rays suitable for the “traditional” single photon nuclear imaging, dynamic or static, planar or tomographic (single photon emission computed tomography or SPECT). Examples are 99mTc, 131I, 123I, 67Ga, 111In, 201Tl. Positron emitters; they decay by emission of a positron (=positively charged electron or e+) which soon annihilates when it encounters an electron (or e−) and as such provides two 511 keV photons in opposing directions on a straight line. These photons are simultaneously detected by opposing detectors (coincidence principle (CP) of detection and imaging). Examples are 18 F, 11C, 13N, 15O, 68Ga, 82Rb, 124I, most with very low T½ necessitating the presence of a cyclotron on site to produce them or a generator to make their clinical use possible. 18F with a T½ = 108 min can be transported from out of town. More recently, the “time of flight” was
G.N. Sfakianakis et al.
introduced in positron emission tomography (PET) imaging which, based on the time difference of arrival to opposing detectors, improves imaging efficiency. Other particle emitters; they decay by beta (or e−) 131 ( I), alpha (or He++), or Auger electron (125I) decay and can be used for internal therapy. Some low energy photon or particle emitters; they are used for in vitro studies, scintillation counting, liquid scintillation counting, autoradiography, etc., like 125I, 3H. Radioisotopes for clinical use are produced in atomic reactors or dedicated cyclotrons and are brought to the nuclear medicine laboratories, either in the form of generators (99mTc, 82Rb) or in their final form (131I), ready to be used or to label stable chemicals in kits. They are kept until usage in appropriately shielded secure hoods. Radioactive remnants are handled appropriately to protect the environment. It should be stated that, due to appropriate selection and dosage, the human applications of radioisotopes in imaging are absolutely safe and no scientific study has ever proven any untoward effects nor are any anticipated. Therapeutic applications, due to much higher doses, have side effects from mild, local, to serious, general, complications.
14.2.2 Radiopharmaceuticals for the Evaluation of the Nervous System These radiolabeled molecules (Table 14.1) are injected intravenously and either accumulate in the normal brain or parts thereof or in abnormal tissues such as tumors, infection, etc [1, 2]. Radiopharmaceuticals labeled with single photon emitting radioisotopes enable imaging with conventional gamma cameras, planar or tomographic studies, (SPECT). These agents may or may not cross the normal blood–brain barrier (BBB). Older conventional brain imaging tracers (or BBB agents) like 99mTc-pertechnetate (PTC), 99mTc-diethylene triamino pentacetic acid (DTPA), and 99mTc-glucoheptonate (GH), do not cross the normal BBB, and have been utilized in the past for delayed static planar imaging or SPECT for brain pathology. Dynamic first pass cerebral blood flow studies are currently performed with these agents (and also with 99mTc-MAG3 or other agents) [1]. Newer brain imaging agents, labeled with single photon emitters (99mTc or 123I) cross the BBB and are
403
14 Scintigraphy for Brain Tumors Table 14.1 Radiopharmaceuticals for brain studies Single photon emitters (g or X) Radioisotope T1/2 (h)
Dose (mCi)
Molecule
Clinical use
Agents, which do not cross the BBB Tc
99m
Tl
201
6
73
30
PTC, DTPA, GHa
Break down of BBB
20
MIBI
Tumor imaging (mitochondria)
5
Thallous chloride
Tumor imaging (Na/K pump)
Agents, which cross the BBB Tc
6
20
HMPAO, ECD
rCBF
I
13
1
IBZM
Acetylcholine receptors
I
13
1
IQNB
Muscarine receptors
I
13
1
IMT
Tumor protein synthesis
I/131I
13/8 days
1
IDOU
Tumor proliferation
99m 123 123 123 123
Receptor imaging In (99mTc)
111
67(6)
4(20)
Octreotide
Tumor (somatostatin receptors)
Cisternography In
111
67
0.5
In-DTPA
Cisternography
5–10
2-Deoxy-d-glucose
Metabolism/tumors
Haloperidol
Neuroreceptor
Ammonia
Perfusion
Amino acids
Tumors
Carbon monoxide
Blood volume
Carbon dioxide
Tissue pH
Glucose
Metabolism/tumors
Dopamine
Neuroreceptors
N-methyl-spiperone
Neuroreceptors
MET/TYR/CHOL
AA metabolism/tumors
Thymidine
DNA synthesis/tumors
Oxygen
Metabolism/tumors
Water
Blood flow
Carbon monoxide
Blood volume
Carbon dioxide
Blood flow
Positron emitters (e+) F
18
N
13
C
11
O
15
110
10
20
2
30
30
100
See text for meaning of abbreviations
a
utilized to perform SPECT for regional cerebral blood flow (rCBF) studies and for receptor distribution. Agents that are localized in the brain in a pattern proportional to rCBF, the amphetamine analog 123IMP, the
diamine H 123 IPDM, and the more recently introduced, 99m Tc-HMPAO (hexa methyl propylene amine oxime) and 99mTc-ECD (ethyl cysteinate dimer), have effectively replaced the radioactive noble gas 133Xenon,
404
which was utilized for decades to study rCBF. Other single photon agents have been developed for receptor distribution imaging with SPECT, 123IBZM (iodobenzamide), and 123IQNB (3-quinuclidinyl-4-Iodobenzilate), but they are still in the experimental stage. The tumor imaging radiopharmaceuticals, 201Thallium (Thallous) chloride (Tl) or 99mTc-Sestamibi (MIBI) do not accumulate in the normal brain, but accumulate in tumors through leaks of the broken BBB and generate marked contrast with the nonradioactive brain parenchyma on SPECT imaging. Once in the extracellular space of the tumor, Tl, a potassium analog, is transferred across the cellular membrane of the cells and kept intracellular by the Na/K ATP pump. MIBI crosses the membrane freely and accumulates in the mitochondria; it is transported outside the cell by the P-glycoprotein. Both agents may reflect tumor cell proliferation. Tl and MIBI are utilized, especially at centers where PET facilities are not available, to differentiate low-grade from high-grade tumors, to confirm the neoplastic nature of unspecified brain lesions identified by CT or MRI (for example, in AIDS, to differentiate lymphoma from toxoplasmosis), to monitor effects of therapy, and to identify recurrence. Radiolabeled amino acids (AA) marked with single photon radioisotopes may be used for tumor imaging, because many tumors have enhanced protein synthesis.123 I labeled iodo-methyl-tyrosine (IMT) has been used for tumor imaging and therapy planning with some success. IMT is transmitted across the intact BBB, is not incorporated into the normal cerebral proteins, but rather localizes in high-grade tumors. IMT may assess the rate of AA uptake by tumors and visualize them depending on their rate of protein metabolism. Radiolabeled deoxy-uridine (DOU) with 123I or 131I has been used to study brain tumor proliferative activity with potential clinical applications. However, low count rate images and high background from radiolabeled metabolites did not produce optimal results. Radiolabeled receptor ligand analogs are utilized, because they were proven useful in the evaluation of tumors, which express the specific receptors. A growing number of such receptor ligand analogs (Somatostatin or SST, epidermal growth factor, etc.) are being introduced into clinical practice and the most widely used has been the SST analog Octreotide (Octreoscan). These agents do not cross the BBB and do not accumulate in the normal brain. Instead, they accumulate in tumor sites with, and sometimes without, SST receptors, and therefore generate clear images that display high contrast against a nonradioactive background.
G.N. Sfakianakis et al.
Other miscellaneous agents, such as 111In-bleomy cine, 123I methoxybenzamin, 123I or 131I meta-iodo- benzyl-guanidine (MIBG), and 123INa or 131INa have been used as general brain tumor, tumor receptor, or metastatic thyroid brain tumor imaging agents, some with clinical applications. Radiopharmaceuticals for cisternography, like 111 In-DTPA, when injected intrathecally, do not cross the ependyma of the subarachnoid space and are utilized clinically to evaluate cerebrospinal fluid (CSF) kinetics, shunt function, or CSF leaks [1] Radiopharmaceuticals labeled with positron emitters (18F, 11C, 13N, 15O) in general cross the BBB, accumulate in the brain depending on the biological properties of the carrier molecule, and enable the study by PET of rCBF, regional cerebral blood volume (rCBV), regional cerebral glucose, AA, etc. metabolism (rCGLM, etc.), and the distribution of several types of neuroreceptors [1]. The positron emitters possess unique chemical and physical characteristics. They can be used to either synthesize or label biologically active molecules (glucose, AAs, etc), and due to their small size, they do not alter the biochemical properties of these agents and thus do not disturb their biological recognition by the cells and their enzymes. On the other hand, they decay by emitting positrons (and not photons which are used for the routine scintigraphic tests of single photon imaging) and require special detectors, the PET cameras. PET studies visualize tumors (most of which have increased glucose, AA, etc., metabolism) and other abnormalities based on their glucose, amino-acid, or receptor uptake characteristics, and are utilized clinically to help confirm or exclude the neoplastic nature of lesions identified by CT or MRI, which the latter imaging modalities cannot always stratify or grade, or differentiate from infections or radiation necrosis [2]. A limitation for this modality is the higher cost of the positron cameras and the need for close proximity of a cyclotron, which produces the short-lived positron emitters, except for 18F, which can be transported from out of town (T½ = 108 h). 18 F-Fluoro-deoxy-glucose (FDG) crosses the BBB and the cellular membrane of normal neurons and abnormal cells (tumor, inflammatory cells, etc.), is phosporylated, but cannot be metabolized (since it is not glucose but deoxyglucose), and accumulates within the cells, reflecting their glucose uptake. When the background clears, imaging of both normal and abnormal tissues is feasible in a quantitative way. 18 F or 11C labeled AAs have been utilized in the same general way as AA labeled with single photon
14 Scintigraphy for Brain Tumors
radioisotopes. Individual agents studied in this capacity include: (a) 18F-methyl-tyrosine; (b) 18F dihydroxyphenylalanine (FDOPA); (c) 11C or 18F-1-amino-3Fcyclobutane-1-carboxylic acid, which was proposed but not yet used in clinical practice; (d) 11C-methionine (MET) now used routinely in some centers for brain tumor studies reflecting cell membrane transport of AA; (e) 11C-l-tyrosine (TYR), which reflects protein synthesis by tumor cells more accurately; and (f) 11 C-choline (CHOL), whose use began with interesting initial results. PET-labeled agents for DNA Synthesis have been studied for brain tumor evaluation with some success. 11 C-thymidine and 124I-deoxyuridine are in the investigational stage. The intensity of their uptake reflects the proliferative rate of tissues and promises to help in defining tumor type and recurrence. 11 C-labeled receptor binding agents (deprenyl, raclopride, and methyl-spiperone) are taken up only by pituitary tissue and have been tried in an effort to analyze and differentiate certain brain tumors.
14.3 Instrumentation Single photon studies (planar or SPECT) are less expensive and are more widely available; however, there are physical, clinical, and biochemical needs for PET studies. Glucose and AAs either cannot be labeled with available single photon emitters, or when this becomes chemically possible, there is a loss of biochemical recognition of these molecules that are radiolabeled with single photon emitters. Cells do not recognize such labeled glucose or AAs and the purpose of alternative labeling is thus defeated. Those positron emitters, 18F, 11 C, 13N, and 15O, have irreplaceable chemical and biochemical characteristics after all! For these and for other reasons, it is necessary to use both single photon and positron imaging. On the other hand, PET provides better resolution and allows reasonable quantification.
405
laboratories. Since its introduction in the late 1960s, the gamma camera (Anger camera) has become the instrument of choice for scintigraphy in clinical practice. This camera detects and processes gamma or X-rays emitted from within the bodies of patients injected with single photon emitting radiopharmaceuticals. It employs a sodium iodide thallium activated crystal for photon detection and metal collimators to localize the point of origin of the photons, enabling reconstruction of their two dimensional distribution in the field of view (FOV). Gamma cameras of this type are utilized for dynamic or planar static imaging such as brain flow, CSF kinetics, and delayed imaging of the brain, or brain pathology. They can be either stationary or portable, the latter enabling the performance of studies in the emergency room, intensive care unit, or at the bedside. In the 1980s SPECT was introduced into clinical practice. This methodology utilizes the same type of single photon gamma cameras, which are used for planar imaging; they are, however, appropriately attached to a rotating gantry. SPECT cameras acquire multiple images usually through a 360o rotation around the patient and apply reprojection and reconstruction algorithms to generate semiquantitatively, the distribution of the radiotracers within the body in a three dimensional mode. The volume of information is reviewed either in multiprojectional slices (transaxial, coronal, sagittal) or in 3D transparent rotating volume format. SPECT can be performed with single or dual detector instruments, but it is more appropriate for the evaluation of the central nervous system (CNS) to use instruments with three detectors or dedicated systems (multidetector type). Two or more detector systems are capable of greater spatial resolution, which is essential for the study of rCBF or metabolism of the brain. Single or dual head cameras are quite sufficient for the study of the brain with Tl or MIBI or for performing conventional studies with agents injected intravenously or intrathecally. Resolution for most pathology is in the range of 1–2 cm. Quantification with single photon cameras for planar or SPECT studies remain a problem. Accuracy for semiquantitative results remains low.
14.3.1 Single Photon Imaging (Collimated Anger Camera Principle) Planar/SPECT
14.3.2 PET Imaging [Coincidence Principle and Time-of-Flight (TOF)]
Rectilinear scanners and scintillation counters are no longer utilized for brain imaging in most modern
PET is especially important for the study of the function of the CNS and for tissue characterization of
406
G.N. Sfakianakis et al.
CT or MRI positive studies. This methodology has been developed before SPECT and utilizes positron emitting radioisotopes (see radiopharmaceuticals above) and special cameras in a fundamentally tomographic mode using not lead collimators but “electronic collimation” (CP traditionally and TOF more recently). The studies are acquired tomographically and processed in a mode that provides resolution in a fraction of 1 cm, and the results can be quantitative. PET agents decay by emission of a “positron,” a positively charged electron, which is the antiparticle of the common negatively charged electron. After its emission, the positron quickly loses kinetic energy and encounters an electron in the vicinity of the decayed atom and within a few millimeters from its origin. As it takes place when antiparticles interact, the positron and electron “annihilate” to produce two 511 keV photons, which travel along the same straight line in opposite directions (180°) from the point of annihilation. “Electronic collimation” is based on the “coincidence” principle, the electronic identification of the line of travel in space of the two photons of the pair. It is achieved through sensing and recording in space, the
loci of the essentially simultaneous reactions of these two photons with the crystal of two opposing detector elements in the PET camera. This principle enables the tomographic imaging of the radioactive object with high efficiency and resolution as well as accurate quantification. More recently, TOF applies the difference in the time of arrival of the two photons into the opposing detectors to generate images. There are two types of PET cameras, those with ring detectors (dedicated PET cameras) and those with dual opposing detectors (hybrid PET/single photon cameras); the latter allow either coincidence detection (PET) without collimators or single photon imaging, planar or SPECT after placing the appropriate collimators. Changing the mode of imaging, however, is time consuming and complicated to a certain extent. PET cameras have crystals in their detectors made of either Sodium Iodide or Bismuth Germanate. It should be mentioned that the 511 keV photons of annihilation can be detected using very high energy lead collimators and SPECT or just a planar single photon imaging approach, but with very limited efficiency. Table 14.2 outlines the differences between SPECT and PET with
Table 14.2 SPECT, PET with dedicated ring detector and hybrid SPECT–PET cameras SPECT
Dedicated PET
Hybrid
Single photon emitters
Positron emitters
Either
Longer half-life
Short half -life
Both
Generators or long shelf life
Require cyclotron (most) generator or fast transport (18F)
Both
Detectors
Gamma camera (single-2 or 3)
Ring detectors
Two detectors g camera
Crystals
Sodium iodide crystal
Sodium iodide or bismuth germanate
Sodium iodide
Principle
Collimation of photons
Coincidence
Either
Resolution
In centimeter
In millimeter
In centimeter
Quantitation
Nonaccurate
Accurate
Nonaccurate
Operation
Easy
Easy
Difficult
Radiopharmaceuticals
Clinical applications – studies rCBF
Semiquantitative
Quantitative
Semiquantitative
Metabolic studies (including tumor)
Semiquantitative
Quantitative
Semiquantitative
Brain receptor studies
Semiquantitative
Quantitative
Semiquantitative
Thallium-MIBIsomatostatin
Semiquantitative
–
Semiquantitative
Cost
<$500,000
>$1,000,000
<$1,000,000
407
14 Scintigraphy for Brain Tumors
dedicated ring detectors or hybrid cameras, and their relative merits and costs. Despite the quantitative ability of PET, it has been found that in clinical practice, interpretation of studies for brain tumors is equally correct by visual analysis as it may be by sophisticated ways, at least in the case of FDG studies [3]. Computers, of course, play an essential role in the acquisition and reconstruction of data for both SPECT and PET; they enable quantification and aid in the interpretation of the images. Attenuation and motion correction, and coregistration of different types of data display are based on computer programs and enjoy great clinical recognition as very useful additions to imaging.
14.4 Clinical Applications Metabolic imaging with PET currently plays a significant role in evaluation of brain tumors. Although CT and MRI provide excellent information about brain tumors, it is clinically challenging to evaluate disease status using only these imaging modalities in patients who have been already treated. First, necrosis induced by radiation therapy or stereotactic radiosurgeries can be difficult to distinguish from recurrent tumor [4, 5]. Second, steroid-induced reduction in tumor size may falsely suggest response to therapy when using imaging modalities, which provide mostly anatomical information. Last, it is challenging with MRI to evaluate recurrent low-grade tumors without anaplastic transformation as changes on MRI can often be indistinct from treatment-induced changes. More recently, PET/CT and PET/MRI imaging have been introduced in clinical practice and provide the ability to correlate findings of these important imaging modalities. Exciting progress has been made in recent years in PET imaging with 18F-fluorodeoxyglucose (FDG) as well as alternative PET tracers, such as AA tracers, nucleotide analog tracers, and hypoxia agents used in evaluation of primary brain tumors. In summary, the main clinical applications of this imaging modality focus on distinguishing tumor recurrence from radiation necrosis, PET-guided diagnosis, treatment planning, and predicting treatment response. By using radiopharmaceuticals and scintigraphic methods, a wide variety of cellular functional characteristics can be studied, such as glucose metabolism, protein synthesis rate, AA uptake, DNA synthesis, and
expression of SST or D2 dopamine receptors. With appropriate interpretation, correlation with CT and MRI (PET/CT and PET/MRI), and full appreciation of the individual clinical conditions, nuclear studies can dramatically change patient management.
14.4.1 Preoperative Tissue Characterization 14.4.1.1 High vs. Low-Grade Gliomas (a) FDG PET For therapy planning and prognosis, it is important to differentiate high vs. low-grade gliomas. Although imaging with CT and MRI provide some discriminating characteristics, nuclear studies are especially helpful. High-grade tumors usually exhibit higher FDG uptake than the normal brain gray matter and can be differentiated from low-grade gliomas. Low-grade gliomas appear as photon deficient regions on FDG studies when the tumors are located in the cortex (Fig. 14.1) [6–9]. In previously known low-grade tumors, the high FDG uptake establishes the diagnosis of anaplastic transformation and is a well established prognostic factor [10, 11]. More recently, studies have demonstrated some diagnostic limitations of FDG-PET [12, 13]. High physiologic metabolic rate of normal brain tissue can make the detection of tumors difficult with only modest increases in glucose metabolism such as lowgrade tumors located in the white matter. Similarly, high-grade tumors in the gray matter can have equal or less metabolic rate than the surrounding grey matter, therefore, decreasing the sensitivity of FDG-PET in detecting the lesion. Furthermore, some tumors may demonstrate high FDG accumulation, but not equally increased true glucose uptake [14]. Correlation of FDG-PET with MRI images greatly improves the performance of this modality in evaluating tumor grade as well as recurrence, and some authors consider it critical to have MRI images available while interpreting PET studies [15, 16]. Recently, Spence et al. studying 19 patients with gliomas showed that FDG imaging 3–8 h after injection could improve the distinction between tumor and normal gray matter [17]. The authors suggested
408 Fig. 14.1 Tissue characterization with 11C-methionine (MET). Upper row: MRI, MET, and 18FDG studies in the same patient with a Grade II astrocytoma. The lesion is hypermetabolic on the MET scan but hypometabolic (relative to the brain) on the FDG scan. Note the very low uptake of MET by the normal brain parenchyma. Lower row: same studies in a patient with a Grade IV glioblastoma. The lesion is hypermetabolic on both PET scans (slightly modified From Kaschten et al. [7] with permission). This approach can be used by centers with PET and cyclotron on site facilities
G.N. Sfakianakis et al.
MRI
THALLIUM EARLY
THALLIUM LATE
that higher FDG-6-phosphate degradation at delayed times in normal brain tissue when compared to tumor may be responsible for higher excretion of FDG from normal brain, and consequently, relatively greater tumor uptake in delayed imaging. (b) Amino acid PET tracers Amino acid PET traces are particularly attractive for imaging brain tumors because of the high uptake in tumor tissue and low uptake in normal brain tissue and, thus, higher tumor-to-normal-tissue contrast [18]. Among these, extensive research has been performed utilizing MET (11C-methyl-methionine) to differentiate low from high-grade tumors (Fig. 14.1). Brain MET uptake is partially related to passive diffusion in tumors with significant breakdown of the BBB. This radiopharmaceutical constitutes an imperfect measure of protein synthesis and is incorporated into the cell phospholipids. The advantage of MET is that it visualizes both high and low-grade gliomas and provides better tumor delineation when compared to FDG [19–21], therefore may also help in planning therapy. Ogawa et al. in a study of 50 glioma
patients with MET PET, observed adequate uptake in nearly all high-grade gliomas (31/32) [22]. In addition, Herholz et al. in a large series of 196 patients demonstrated that MET had an accuracy of 79% in differentiating low-grade gliomas from nontumoral lesions [23]. Both MET and FDG are currently used concomitantly for evaluation of low-grade oligodendrogliomas and astrocytomas [24] and for diagnostic or therapeutic stereotactic processes [24–26]. The major drawback of 11C-MET is the short half-life of carbon-11, and the fact that as opposed to FDG, this agent does not provide prognostic information and cannot predict survival rates [20, 21, 27]. More recently, F-DOPA has shown excellent results in both low and high-grade brain tumor imaging, with greater sensitivity than FDG [28]. The higher tumor to normal tissue ratio proved useful in detecting low-grade gliomas. Standard visual analysis of the F-DOPA PET showed high sensitivity in identifying tumor. Specificity was increased by using thresholds of tumor to normal brain tissue [29]. FET (O-2-[18F]fluoroethyl-l-tyrosine) is a promising tracer for PET that has demonstrated convincing
14 Scintigraphy for Brain Tumors
409
results in the diagnostics of brain tumors and although not incorporated into proteins, it shows high uptake in cerebral gliomas and was able to differentiate between low and high-grade tumors with a sensitivity of 94% and specificity of 100% [30]. Initial studies with 11C-thymidine suggested that brain uptake reflected proliferative rates of the tissue and could be used as an index of metabolic activity [31–34]. However, later work showed it to be a less sensitive brain imaging agent since BBB disruptions appear to account for significant uptake when the label is attached to methyl carbon [35]. Other AAs, such as 11C-tyrosine and 11C-leucine have been proposed as protein synthesis rate imaging agents [36–39], but the clinical experience with these radiotracers is still limited. Carbon-11 labeling is a limiting factor for routine clinical use and for regional distribution. Tyrosine can also be labeled with 18F, with high yields and specific activity [40, 41]. l-[2-18F] Fluorotyrosine holds promise to replace MET and complement FDG in tumor diagnosis.
A number of D2 Dopamine receptor ligands (11C-raclopride, 11C-methylspiperone, 18F-ethylspiperone), and 11C-deprenyl (which binds to the monoamine oxidase B) have been utilized for detection of pituitary adenomas, which are characterized by positive studies due to their specific receptors, and for their differentiation from meningiomas and other parasellar tumors, which do not exhibit such receptors [42]. (c) SPECT tracers An alternative to positron tracers is the potassium analog 201Tl. The uptake of 201Tl in brain tumors is considered to be related to the Na+/K+ ATP-ase activity, to BBB integrity, blood flow, and malignant cell density. In general, Tl visualizes well only the high-grade and not the low-grade gliomas. When the tumor has a size of greater than 1–2 cm, Tl has a high specificity in characterizing it as lowor high-grade [43–45]. The sensitivity and specificity are 72% and 87%, respectively [44] (Fig. 14.2). The accuracy of this differentiation has been
Fig. 14.2 Tissue characterization with 201Thallium. The patient had a history of resected malignant meningioma and returned with central nervous system (CNS) symptoms and signs. He was found to have focal lung lesions by chest CT. A brain MRI showed a brain lesion compatible with either recurrence of the brain tumor or metastasis from a new lung tumor (upper row,
arrows). An early (middle row) and late (lower row) 201Tl study was positive for high-grade tumor, suggesting recurrence of the primary malignant tumor (arrows). Biopsy showed recurrent malignant meningioma with lung metastasis. Thallium is positive in high-grade brain tumors but negative in low-grade, including many metastatic tumors
410
reported as being up to 89% [43]. The highest sensitivities have been reported in glioblastoma mutliforme and metastatic lesions. False positives in SPECT Tl are noted with other causes of disruption of the normal BBB barrier such as skull metastasis, hemorrhagic strokes, and epidural hematomas; or in cases of angiomas [44]. False negatives are found in posterior fossa and temporal gliomas, as well as in small tumors. Although data in the pediatric population is also available, results are inferior to those obtained in adults [46, 47]. Tumor necrosis, considered a marker of high-grade activity, can be a common cause of underestimation of tumor grade in thallium imaging [48]. In an effort to further improve the specificity of Tl brain imaging, early and late imaging can be performed, with a reported specificity as high as 100% for high-grade gliomas [49–51] (Fig. 14.2); however, these results were not uniformly good. A study of Hirano et al. comparing Tl with 99m Tc (V)-DMSA found the latter to be more effective in differentiating high-grade tumors from low-grade ones [52]. Different authors have suggested the sequential use of TI and HMPAO for more accurate results [53]. Comparisons have been made between Tl and FDG in differentiating high from low-grade gliomas [51]. Both were found to be good, as both visualize the highgrade and not the low-grade gliomas; however, there was an overlap between the two methods and there was no absolute correlation between their results, which reflects the different uptake mechanisms. This suggests a complementary role of TI and FDG imaging in characterizing high-grade gliomas [54]. MIBI is a cationic compound that passively diffuses into the mitochondria and cytoplasm, and the uptake, although nonspecific, is driven by metabolic demand and normally does not cross the BBB. This radiopharmaceutical can be given in higher doses than Tl, provides better anatomic delineation of the tumor boundaries, and has also been used to differentiate high from low-grade gliomas. Results similar to Tl have been reported with MIBI for both primary tumors, such as gliomas, and also for some metastatic tumors [55–59]. This approach, however, is plagued by visualization by MIBI of the choroid plexus. SPECT studies with 123 I-IMT (123I-alpha-methyl tyrosine) have been used for differentiation of high from low-grade gliomas with sensitivity and specificity of 71 and 83%, respectively, and an accuracy
G.N. Sfakianakis et al.
of 83% [60, 61]. IMT has the property of normally crossing the BBB and brain cell membranes and its intracellular accumulation is proportional to cellular proliferation, but this compound is not incorporated into cellular proteins [62]. Woesler et al., in a study of 23 histologically proven brain tumors found that IMT and FDG were equally good at differentiating low-grade and high-grade tumors with accuracies of 83 and 91%, respectively [63]. A new SPECT tracer that has been studied for glioma grade differentiation is IPA (p-[I-123]iodol-phenylalanine). This radioiodinated aminoacid appears to be similar to IMT in many respects; however, it may be somewhat more specific for brain tumors and has a longer retention time, allowing for more flexibility in imaging [64].
14.4.1.2 Tumors with Somatostatin Receptors In-pentetreotide scintigraphy has been used to detect brain metastasis of all other tumors exhibiting SST receptors in case a confirmation of the nature of brain metastasis is needed in decision-making for therapy (Fig. 14.3). It has also been successful for the localization of extrapituitary ACTH- and CRH-secreting tumors and their metastases, especially in those difficult cases in which conventional radiological studies had initially failed to localize the tumors [65, 66]. Finally, 111In-pentetreotide scintigraphy allows postsurgical differentiation between scar tissue, radionecrosis, and residual pituitary tumor or tumor recurrence [67, 68]. 111 In-pentetreotide scintigraphy has been used to evaluate patients with cellar and supracellar tumors; however, it is important to realize that the nonpathologic anterior pituitary gland also takes up this radiotracer [54, 68]. Thus, the role of somatostatin receptor scintigraphy in the diagnosis and differential diagnosis of pituitary tumors is limited, since a large variety of lesions in and around the pituitary region express SST 2, 3, and 5 receptors, and therefore, can be visualized by 111In-pentetreotide scintigraphy. The majority of meningiomas, class III and IV gliomas, some metastases from breast carcinomas and other adenocarcinomas, osteosarcomas, Hodgkin and nonHodgkin lymphomas, abscesses and granulomatous lesions, angioleiomyomas, chordomas and hemangiopericytomas as well as fibrous dysplasia can be visualized by 111In-pentetreotide scintigraphy [69]. 111
411
14 Scintigraphy for Brain Tumors
a
b
d
c
e
Fig. 14.3 Tissue characterization with somatostatin receptor imaging (Octreoscan). (a–c) Planar imaging of the head in anterior and lateral projections in patients with meningioma, pituitary tumor, and paraganglioma respectively (arrows). (d) A total body planar scan in anterior (left) and posterior (right) projections. The patient had a metastatic insulinoma (arrow). (e) Preoperative (left
column) and postoperative (right column) SPECT images (transaxial: upper; coronal: middle; sagittal: lower). The patient with a glomus jugularis tumor has been operated but part of the tumor was not resected. This imaging application can be used in centers with the simplest nuclear imaging (planar single photon scans) or, better, with SPECT capabilities
14.4.1.3 Metastatic Tumors
evaluation of lesions shown by CT or MR (see Sect. 14.4.2.2). The same is true for lymphomas, colon cancer, melanomas, and ovarian and other tumors. Metastatic thyroid tumors have been traditionally studied and treated with radioactive Iodine. Such tumors occasionally metastasize to the brain and it is essential to identify their presence by FDG or Iodine imaging before administering radioactive Iodine for treatment. Tumor edema resulting from effective treatment with radioactive Iodine may be a severe problem if the patient is not concomitantly treated with steroids.
FDG is used for imaging of many types of body tumors, some of which metastasize to the brain. Staging lung cancer with FDG for example, allows the visualization of its brain metastasis. Brain metastases in the cortex appear as cold foci for most tumors; however, lung tumors may accumulate higher FDG than the normal gray matter. Although the sensitivity of imaging metastatic tumors is lower with FDG as compared to CT and MR, there are clinical conditions where staging for lung and other tumors includes FDG brain imaging for
412
14.4.2 Differentiation of Tumor from Infection 14.4.2.1 Primary CNS Lymphoma vs. Toxoplasmosis in AIDS CT and MRI cannot distinguish with certainty between the two common lesions in AIDS patients, primary CNS lymphoma and toxoplasmosis. They both appear as “ring enhancing lesions.” Although newer applications including measurements of chemical indices in the CSF constitute a sort of additional help, in many patients with AIDS, who present with central nervous system symptoms and signs (CNSSS) and a CT or MRI, which shows a ring-enhancing lesion, additional information is necessary to avoid biopsy in deciding the appropriate therapeutic approach. This is an indication for either Tl or FDG imaging. The sensitivity of Tl imaging in visualizing CNS lymphoma (thus excluding toxoplasmosis) has been reported to be in the range of 75–100% with a specificity, which ranges around 90% [70, 71] (Fig. 14.4). Experience in our center indicates good but not ideal results in applying Tl imaging to differentiate lymphoma from toxoplasmosis [70, 71]. Accurate results have also been reported with the use of FDG for the same purpose [72, 73]. However, FDG-PET may visualize not only lymphoma but also untreated toxoplasmosis and it is advised to employ this test after 10–15 days of therapy for this infection; at that time, toxoplasmosis appears as an area with less than the normal cortex activity (Fig. 14.5), whereas lymphoma remains hyperactive.
14.4.2.2 Cavitating Lung Lesions with Associated Brain Lesions In some cases of body tumors, especially lung tumors, a central necrosis of the tumor associated with a clinical presentation with fever raises the possibility, before biopsy, that the lesion represents an infection rather than a tumor. If such patients have brain lesions, FDG or Tl total body imaging including the brain may provide additional information clarifying the identity of the lesions. In general, tumors accumulate a higher intensity of Tl or FDG activity compared to infectious or inflammatory lesions. Therefore, the coexistence of ring-like positive body lesions and solid brain lesions of similar intensity suggest the same identity of the
G.N. Sfakianakis et al.
lesions (Fig. 14.6). However, it should be mentioned that Tl might be positive in abscess [74], in candidiasis [75], in CMV encephalitis [76], and even, sometimes, in infarcts [77]. PET studies may also be positive in abscesses of the brain because of activation of adjacent cortex, and the differentiation from tumor is again difficult (see Sect. 14.4.2.3).
14.4.2.3 Cystic Malignant Glioma vs. Brain Abscess This is not an unusual situation and falls, in general, within the clinical issues discussed above (Sect. 14.4.2.2), but involves primary brain lesions. At least one report indicated that Tl was a good radiopharmaceutical for differentiating malignant glioma, which is visualized as a ring lesion, from brain abscess, which may not be visualized [78]. However, the results are always questionable, unless Tl is absolutely negative, thus excluding high-grade tumor. The same holds true for FDG studies. A totally negative FDG scan provides much more useful information by excluding high-grade tumor and is more reliable than a positive study. Indeed, in a positive FDGPET study, if the hyperactive focus coincides exactly with CT or MRI lesion, then it is most probably a tumor, if not, it may be due to activation of brain tissue adjacent to an abscess, as was the case of Fig. 14.7.
14.4.3 Therapy Planning and Evaluation of Effectiveness of Treatment CT and MR imaging provide the necessary information for therapy planning in many conditions. However, additional information is needed in a number of cases, when considering the treatable volume, the boost volume, and prognosis. Nuclear imaging may provide such information. In addition, CT and MRI do not provide information about early response to therapy, and again, nuclear imaging may be helpful to that effect. Functional brain mapping may be performed in some instances to identify adverse effects of treatment. Furthermore, when evaluation of completeness or surgery is required, nuclear studies are indicated. CT-guided stereotactic biopsies can be helped by nuclear imaging as well. In all the above conditions, PET is utilized with a higher accuracy and better resolution than SPECT imaging.
413
14 Scintigraphy for Brain Tumors Fig. 14.4 Differentiation between primary CNS lymphoma and toxoplasmosis in patients with AIDS; primary CNS lymphoma. The patient, HIV-positive with AIDS, was admitted with CNS symptoms and had a CT, which showed a ring enhancing lesion in the basal ganglia on the left (arrows), lymphoma or toxoplasmosis. Early and late 201Tl SPECT studies showed high intensity uptake in the area of the CT lesion (arrows) compatible with lymphoma. A biopsy confirmed the diagnosis of primary CNS lymphoma and the patient was treated accordingly
Coregistration of nuclear studies and CT or MRI is needed for such therapy planning and evaluation of effectiveness. Although coevaluation of the different imaging modalities is possible by visual appreciation, computer-achieved coregistration appears to be a more reliable, reproducible, and easier to apply approach.
14.4.3.1 Radiation Therapy Planning PET has been investigated to delineate tumor volumes for radiation therapy with promising results. Douglas et al. evaluated radiation dose escalation using FDG-PET defined volume in 27 glioblastoma patients treated initially with standard conformal fractionated radiotherapy
414 Fig. 14.5 Differentiation between primary CNS lymphoma and toxoplasmosis in patients with AIDS; toxoplasmosis. Corresponding transaxial slices of CT, Thallium SPECT, and FDG-PET scans in a patient with AIDS admitted with CNS symptoms. On admission, the CT showed a ring-enhancing lesion in the right thalamus (arrows), but the Thallium study was negative for lymphoma and the patient was treated for toxoplasmosis. Fifteen days later, an FDG-PET study confirmed the diagnosis. The ring-enhancing lesion of the CT (arrows) appears as a defect of the right thalamic activity on the FDG-PET scan (arrows). FDG-PET studies in toxoplasmosis and other infections may be false positive early but, performed 10–15 days into treatment, become negative for infection while they remain positive for lymphoma. The patient improved and was discharged home. FDG-PET studies may be performed in centers with any type of PET facilities (see text), even without a cyclotron, because FDG can be shipped from out-of-town cyclotrons
G.N. Sfakianakis et al.
415
14 Scintigraphy for Brain Tumors
with volumes defined by MRI. Multivariate analysis demonstrated that positive FDG-PET uptake was the only parameter significant for predicting survival and time to tumor progression [79].
a
TRANSAXIAL
Fig. 14.6 Cavitating lung lesions with associated brain lesions. (a) Selected slices from a total body FDG-PET study in a patient with CNS symptoms and a cavitary lesion in the lung; the arrows show a ring lesion in the left upper lung field corresponding to the X-ray cavitating lesion. (b) The brain FDG study was performed to determine if a brain lesion on the MRI (arrows) was of the same metabolic quality as the lung lesion, which was found to be nonsmall cell lung cancer (NSCLC) by biopsy. The FDG-PET study indeed showed that the brain lesion was of the same metabolic activity. Final diagnosis was metastatic NSCLC
SAGITTAL
CORONAL
MET and 123I-alphamethyl-tyrosine fusion with CT and MRI were studied in 44 patients with recurrent glioblastoma after surgery and postoperative conventional radiotherapy [80]. The PET(SPECT)/CT/MRI-delineated
416 Fig. 14.6 (continued)
G.N. Sfakianakis et al.
b
MRI
FDG-PET
TRANSAXIAL
SAGITTAL
CORONAL
volume for fractional stereotactic radiotherapy treatment planning was compared with volume using CT/MRI alone. A significant survival advantage was found, 9 months median survival in patients whose treatment volume was based on PET(SPECT)/CT/MRI vs. 5 months median survival in patients whose treatment planning was based on CT/MRI. Hypoxia in tumors is a pathophysiologic consequence of structurally and functionally disturbed angiogenesis along with deterioration in the ability of oxygen to diffuse through tissues and is associated with progression and resistance to radiotherapy. 18F-fluoromisonidazole (FMISO), a nitroimidazole derivative whose metabolites are trapped in hypoxic cells, was developed as a PET imaging agent to image hypoxia. FMISO uptake is
independent of perfusion and BBB disruption and has been found to be present in both high-grade and lowgrade gliomas. These properties added to the significant relationship found between FMISO uptake and expression of the angiogenesis marker VEGF-R1 and suggest that this agent may have a role in directing and monitoring targeted hypoxic therapy [81, 82].
14.4.3.2 Functional Brain Mapping FDG, 15O labeled water (15OH2) and 11C-MET in combination can provide functional brain mapping, which is very useful for neurosurgical planning. Brain maps are used to characterize the relationship between potentially
18
417
14 Scintigraphy for Brain Tumors Fig. 14.7 Differentiation of tumor from abscess; brain abscess. The patient was admitted with CNS symptoms, seizures, and fever. There was a history of left sinus surgery 6 years ago. The MRI study showed a ring-enhancing lesion in the left frontal lobe (upper row, white arrows). The FDG-PET study showed decreased activity in the region of the MRI lesion (black arrows). There was, however, high intensity activity in the medial cortex of the right frontal lobe (red arrows) corresponding to an area of cortical activation (seizure activity). It is absolutely important to correlate anatomic and functional studies for the most accurate interpretation of the images. Functional changes have the same appearance as anatomic lesions and may be confusing. In this case (contra lateral hemisphere adjacent to the lesion), cortical activation (seizure focus) may be misinterpreted as tumor
resectable tumors and functionally eloquent brain areas. PET is used to characterize the anatomical relationships of the tumors to functional cortex. The cortical activation maps are obtained during control periods and during behavioral tasks and are used to document motor, visual, and speech and language organizational areas. Wada tests (injection of ECD or HMPAO IV, while balloon occluding the artery to be resected) are also performed in some patients. Language and speech activation is concordant with the results of Wada testing. Functional brain mapping using PET scans and coregistered MR images provide the neurosurgeon with precise definitions of structural and functional cortical areas; this alters surgical management in some cases and/or is used to predict outcome.
14.4.3.3 Early Response to Therapy Both CT and MRI do not show early response to treatment and, in some instances, such information is extremely important. Response to chemotherapy can be shown early in lymphomas, and gliomas in adults, or medulloblastomas in children utilizing FDG [27, 83, 84]. At our center, FDG imaging is used routinely to evaluate early response to chemotherapy in patients with lung cancer and lymphomas, some with CNS lesions; this is more useful when options of different therapies exist. The results have been encouraging with FDG; however, C-11 Methionine has also been utilized with substantial success [20]. The situation is different when brain tumors are treated with radiation,
418
because radiation initially increases the FDG uptake by the tumor before it reduces it [83]. Increased FDG uptake was found to be present for months after treatment with intracavitary antibodies [56, 85]. Therefore FDG uptake is not a good indicator for immediate response to radiation therapy, but may be useful after 4 months. Thallium has also been utilized to evaluate early response to chemotherapy in patients with primary
Fig. 14.8 Evaluation of early response to therapy. Regression of primary CNS lymphoma in patients treated with AZT, GCV, and IL-2. Examples of three patients, two with 201Tl SPECT and one with CT scans (slightly modified from Raez et al. [86], with permission)
G.N. Sfakianakis et al.
brain lymphomas treated with anti-AIDS agents (zidovudine, ganciclovir, and interleukin-2) [86]. In such patients, Tl uptake was found to be reduced from the pretreatment levels very early (within days) after initiation of treatment, reflecting and predicting good results, verified weeks or months later by CT or MRI [86] (Fig. 14.8). Additional experience at our center verified the preliminary encouraging results of Thallium imaging for this purpose.
419
14 Scintigraphy for Brain Tumors Fig. 14.9 Evaluation of response to therapy with FLT. A case of recurrent glioblastoma in a 61-year-old woman during treatment. Upper panel shows FLT (a) before and (b) 1 week after starting treatment. Lower panel shows MRI images (c) before and (d) 3 months after starting treatment (Chen et al. [91], p 4718)
a
b
c
d
The thymidine analog 3¢-deoxy-3¢-[18F]-fluoro thymidine (FLT)-PET was developed as a noninvasive method to evaluate tumor cell proliferation [87] 18 F-FLT uptake is related to the rate of DNA synthesis and increases with higher proliferation rates in many types of cancers. Uptake of FLT correlates with thymidine kinase-1 (TK1) activity, an enzyme expressed during the DNA synthesis phase of the cell cycle [88]; and phosphorylation of FLT intracellularly by TK1 results in trapping of the negatively charged FLT monophosphate. FLT uptake has been investigated in brain tumors and correlations with proliferation index Ki-67 have been observed [89, 90]. Thus, FLT as a prognostic marker has potential for monitoring treatment response (Fig. 14.9). A study of 19 patients with malignant gliomas treated with angiogenesis inhibitors showed that both early (1–2 week) and late (6 weeks)
decrease in FLT uptake from baseline were more significant predictors of overall survival than were MRI responses [16, 91].
14.4.3.4 Evaluation of Completeness of Surgery or Final Effectiveness of Chemotherapy Depending on the tumor involved, and immediately after an attempt to resect the tumor, scintigraphy with the appropriate radiopharmaceutical can help identify if there is any substantial tumor remnant (Fig. 14.3e). Similarly, after completion of chemotherapy, a nuclear study may indicate if the tumor completely regressed or if it is still functional (Fig. 14.10). One should be aware of the limitations of this approach since the resolution of the nuclear methods is not good for lesions
420
G.N. Sfakianakis et al.
Fig. 14.10 Evaluation for residual tumor after gamma knife therapy. Patient with known ovarian cancer post radiation therapy and chemotherapy for metastatic brain tumor 5 months ago, admitted with CNS symptoms. MRI showed a ring-enhancing lesion with central necrosis (arrow). FDG-PET study showed an extensive area of lack of brain function (necrosis and suppression) and a small focus of increased activity corresponding to the lesion of the MRI compatible with residual, regrowing tumor. Due to radiation effect on the normal brain tissue, which results in photon-deficient regions surrounding the lesion, even low-grade primary and metastatic tumors are visualized by FDG-PET. A Thallium study was falsely negative due to the small size of the lesion and the relatively lower resolution of Tl-SPECT as compared to FDG-PET
less than 1–2 cm in diameter and many tumors do not visualize anyway; as such, only positive studies have real clinical significance.
14.4.3.5 PET-Guided Stereotactic Biopsies Brain tumors may consist of different parts, which are heterogeneous with respect to tumor grading; thus, low and high-grade areas may be present within the same tumor. Stereotactic biopsy aims at the tumor sites with the highest tumor grade. Therefore, suitable targets for biopsy will have hypermetabolism on 18F-FDG-PET and high accumulation of MET [26]. Since AA tracers have shown greater sensitivity in imaging tumors that are either hypo- or iso-metabolic to normal cortex with FDG, combining FDG and AA tracer to guide biopsy has been studied. Pirotte et al. evaluated FDG and MET
to guide stereotactic biopsy in 32 patients with unresectable gliomas. The double tracer approach was proposed for these patients because they presented with a tumor considered unresectable and located in the cortical or subcortical gray matter, hence likely lower sensitivity with FDG. FDG was used for target selection when the uptake was higher in tumor than gray matter and MET was used when tumor FDG uptake was less or equal to the surrounding gray matter. All MET-positive trajectories yielded a diagnosis of tumor and all MET-negative trajectories were nondiagnostic. FET-PET has also shown promising results in biopsies of patients with suspected gliomas. The diagnostic performance of neuronavigated tissue biopsies with MRI alone or MRI combined with FET was compared [92]. MRI yielded a sensitivity of 96% for the detection of tumor tissue but a specificity of 53%. Combined use of MRI and FET-PET yielded a sensitivity of 93%
421
14 Scintigraphy for Brain Tumors
and a specificity of 94%. The authors concluded that combined use of MRI and FET-PET significantly improves the identification of tumor tissue.
14.4.4 Diagnosis of Recurrence (Recurrence vs. Radiation Necrosis) 14.4.4.1 FDG-PET F-FDG PET performs well in identifying growing high-grade gliomas; however, in lesions that are equivocal on MRI and in cases of recurrent low-grade tumors without anaplastic transformation, this imaging modality shows limited sensitivity [16]. Dexamethasone has no effect on tumor FDG uptake, and FDG-PET imaging can be successfully performed in patients receiving dexamethasone therapy [93]. High FDG uptake in a previously diagnosed lowgrade glioma with low FDG uptake is diagnostic of anaplastic transformation and is considered a strong prognostic factor. De Witte et al. studying 28 patients with low-grade gliomas for a mean period of 27 months found that all 19 patients with hypometabolic lesions were alive at the end of the period, and that 6 of 9 patients with hypermetabolic lesions had died [11]. Correlation of FDG-PET with MRI can also greatly increase the performance of this modality in identifying tumor recurrence [15, 16]. Recurrence should be considered if the radiotracer uptake in the region of interest, identified on MRI, is greater than the expected background, even if the uptake is equal or less to the adjacent normal brain (Fig. 14.11).
in most low- and high-grade tumors; therefore, they might be preferable for evaluating recurrent tumors. Initial research with 45 brain lesions that did not show increased uptake on FDG-PET was evaluated with MET and this tracer demonstrated a sensitivity of 92% in gliomas. All ten benign brain lesions (cysticercosis, radiation necrosis, tuberculous granuloma, hemangioma, organized infarction, and benign cyst) showed normal or decreased MET uptake (100% specificity). MET was false negative in cases of intermediate oligodendroglioma, metastatic tumor, chordoma, and cystic ganglioma [23].
18
14.4.4.2 Aminoacid PET In contrast to 18F-FDG uptake, AA uptake has been shown to be increased relative to normal brain tissue in most low- and high-grade tumors, and radiolabeled AAs might therefore be preferable for evaluating recurrent tumors. MET-PET C-labeled AAs, particularly Methionine (MET), have been shown to be increased relative to normal brain tissue 11
FET-PET Using FET in a series of 53 patients with clinically suspected recurrent glioma and setting a threshold value of 2.0 for the maximum SUV to background ratio or a threshold of 2.2 for the absolute maximum SUV value, clinicians were able to reliably distinguish between recurrent tumor and therapy-induced benign changes with 100% accuracy [94]. Thus, focal and high FET uptake was considered suspicious for tumor recurrence, whereas low and homogenous uptake around the resection cavity was considered benign, posttreatment changes from disrupted BBB. Authors have suggested MRI to be used for screening first as it has high sensitivity but poor specificity. In the event of suspected tumor recurrence, additional FET-PET investigation seems to differentiate between posttreatment changes and tumor recurrence and avoids both under and overtreatment.
F-DOPA PET For many years, 18F-FDOPA imaging of the integrity of the striatal dopamine pathway has been used to evaluate patients with movement disorders. More recently, this aminoacid analog has been studied in brain tumor imaging and the results have been promising. 18 F-FDOPA demonstrates excellent visualization of high and low-grade tumors and was more sensitive and specific than 18F-FDG for evaluating recurrent tumors [29] (Fig. 14.12). 18F-FDOPA may prove particularly valuable for examining recurrent low-grade gliomas, because these tumors are difficult to evaluate by MRI and are usually not visible on 18F-FDG PET. Standard visual analysis of 18F-DOPA PET seemed adequate in
422
G.N. Sfakianakis et al.
Fig. 14.11 Differentiation between recurrent metastasis and gamma-knife radiation necrosis. The patient has been treated 7 months ago with gamma-knife radiation for a brain metastasis from NSCLC. She is admitted for CNS symptoms. MRI studies were interpreted as recurrence v/s radiation necrosis (arrows). FDG-PET study shows an area of intermediate intensity (corresponding to MRI focus) in the middle of an extensive area of brain hypofunction due to radiation necrosis (arrows). Due to the radiation effect on the brain, even metastatic recurrent lesions of lower intensity can be visualized by FDG-PET
that it provided a high sensitivity in identifying tumor and its specificity for brain tumor imaging could be increased by using thresholds of tumor to striatum ratio T/S of 0.75 or 1.0, tumor to normal hemi- spheric brain ratio T/N of 1.3 or tumor to normal white matter ratio T/W of 1.6 [29].
SPECT Tracers Reports of a small number of patients indicated that TI might identify recurrence of high-grade gliomas with low sensitivity but with a specificity of nearly 100% [95] (Fig. 14.2). However, other reports indicate lower
423
14 Scintigraphy for Brain Tumors Fig. 14.12 MRI (left), 18 F-FDG PET (middle), and 18 F-FDOPA PET (right) for evaluating recurrent tumors. (a) Recurrent glioblastoma. (b) Recurrent grade II oliogodendroglioma (Chen et al. [28], p 907)
specificity, although higher sensitivity [96–98]. The sensitivity is, in general, low because of the known resolution limitations of Tl imaging. 123 I-alpha-methyl-l-tyrosine (IMT) has been utilized for the same purpose with very good results. 123I-IMT may assess AA uptake by tumors. In gliomas, 123I-IMT uptake correlates with the proliferative activity of the tumor, but not with its cellular density [60, 61]. More data is needed to establish its clinical use. 123I-IMT imaging has shown high sensitivity and specificity in differentiating high-grade gliomas from necrosis, infarcts, and inflammatory lesions (sensitivity/specificity of 82%/100%). Recurrent or residual [99–101] meningiomas have been identified successfully using 111In-Octreo scan (Fig. 14.3). As stated above, this compound accumulates in high-grade gliomas not expressing SST receptors [99, 101].
14.5 Therapy of Brain Tumors with Unsealed Radiopharmaceuticals Unsealed radiopharmaceuticals for internal use, either intravenous or local, are applied for therapy of brain tumors. 131INa has been utilized for decades to treat
functioning thyroid tumor metastasis including brain metastasis. Imaging before therapeutic applications is necessary to avoid complications from the early effects of the radioactive Iodine on the tumor (edema, etc.). In such cases pretreatment of the patient with steroids is indicated. Systemic administration and intracavitary installation of 131INa and 90Y labeled antibodies are currently applied to treat lymphomas or high-grade gliomas respectively. The results are encouraging, but the method has not been established in a large number of patients. Certainly this application has potential merits. Interstitial brachytherapy is also applied. Therapeutic applications are beyond the scope of this book.
14.6 Conclusion The information provided by CT and MRI in the evaluation of patients with brain tumors is invaluable; however, PET adds functional information that has important prognostic value and brain scintigraphy may help in tissue characterization when other modalities are inconclusive [102]. In previously treated patients, 18F-FDG PET can be helpful in differentiating recurrent tumor from radiation necrosis. AA tracers are promising in
424
that they are more sensitive in imaging brain tumors. The use of PET to plan and monitor treatment is an active area of investigation. With the development of targeted therapies, PET biomarkers might be used to select patients who are likely to respond to treatment, as well as to monitor treatment response. In addition, it is mandatory that the nuclear studies be interpreted side by side with the anatomical studies, MRI or CT, or better, after computer-assisted coregistration, or in the newly applied PET/CT, SEPCT/CT or PET/MRI and SPECT/MRI concurrent imaging.
References 1. Sfakianakis GN, Mallin W (1999) Scintigraphic neuroimaging in paediatrics. In: Panteliadis CP, Darris BT (eds) Encyclopedia of pediatric neurology theory and practice, Thessaloniki, Greece. 2nd edn:pp 164–195 2. Hustinx R, Alavi A (1999) SPECT and PET imaging of brain tumors. Neuroimaging Clin N Am 9(4):751–766 3. Alavi JB et al (1988) Positron emission tomography in patients with glioma. A predictor of prognosis. Cancer 62(6): 1074–1078 4. Del Sole A et al (2001) Anatomical and biochemical investigation of primary brain tumours. Eur J Nucl Med 28(12): 1851–1872 5. Levivier M et al (1995) Diagnostic yield of stereotactic brain biopsy guided by positron emission tomography with [18F] fluorodeoxyglucose. J Neurosurg 82(3):445–452 6. Tyler JL et al (1987) Metabolic and hemodynamic evaluation of gliomas using positron emission tomography. J Nucl Med 28(7):1123–1133 7. Kaschten B et al (1998) Preoperative evaluation of 54 gliomas by PET with fluorine-18-fluorodeoxyglucose and/or carbon-11-methionine. J Nucl Med 39(5):778–785 8. Delbeke D et al (1995) Optimal cutoff levels of F-18 fluorodeoxyglucose uptake in the differentiation of low-grade from high-grade brain tumors with PET. Radiology 195(1):47–52 9. Di Chiro G et al (1987) Glucose utilization by intracranial meningiomas as an index of tumor aggressivity and probability of recurrence: a PET study. Radiology 164(2):521–526 10. Padma MV et al (2003) Prediction of pathology and survival by FDG PET in gliomas. J Neurooncol 64(3):227–237 11. De Witte O et al (1996) Prognostic value positron emission tomography with [18F]fluoro-2-deoxy-D-glucose in the low-grade glioma. Neurosurgery 39(3):470–476, discussion 476–477 12. Olivero WC, Dulebohn SC, Lister JR (1995) The use of PET in evaluating patients with primary brain tumours: is it useful? J Neurol Neurosurg Psychiatry 58(2):250–252 13. Ricci PE et al (1998) Differentiating recurrent tumor from radiation necrosis: time for re-evaluation of positron emission tomography? AJNR Am J Neuroradiol 19(3):407–413
G.N. Sfakianakis et al. 14. Krohn KA et al (2005) True tracers: comparing FDG with glucose and FLT with thymidine. Nucl Med Biol 32(7):663–671 15. Wong TZ et al (2004) PET and brain tumor image fusion. Cancer J 10(4):234–242 16. Chen W (2007) Clinical applications of PET in brain tumors. J Nucl Med 48(9):1468–1481 17. Spence AM et al (2004) 18F-FDG PET of gliomas at delayed intervals: improved distinction between tumor and normal gray matter. J Nucl Med 45(10):1653–1659 18. Kubota K et al (1984) Tumor detection with carbon-11-labelled amino acids. Eur J Nucl Med 9(3):136–140 19. Fulham MJ et al (1993) Neuroimaging of juvenile pilocytic astrocytomas: an enigma. Radiology 189(1):221–225 20. Ogawa T et al (1993) Cerebral glioma: evaluation with methionine PET. Radiology 186(1):45–53 21. Ogawa T et al (1991) Clinical value of PET with 18F-fluorodeoxyglucose and L-methyl-11C-methionine for diagnosis of recurrent brain tumor and radiation injury. Acta Radiol 32(3):197–202 22. Ogawa T et al (1994) Methionine PET for follow-up of radiation therapy of primary lymphoma of the brain. Radiographics 14(1):101–110 23. Herholz K et al (1998) 11C-methionine PET for differential diagnosis of low-grade gliomas. Neurology 50(5):1316–1322 24. Derlon JM et al (1997) The in vivo metabolic pattern of lowgrade brain gliomas: a positron emission tomographic study using 18F-fluorodeoxyglucose and 11C-L-methylmethionine. Neurosurgery 40(2):276–287, discussion 287–288 25. Goldman S et al (1997) Regional methionine and glucose uptake in high-grade gliomas: a comparative study on PETguided stereotactic biopsy. J Nucl Med 38(9):1459–1462 26. Pirotte B et al (2004) Combined use of 18F-fluorodeoxyglucose and 11C-methionine in 45 positron emission tomographyguided stereotactic brain biopsies. J Neurosurg 101(3):476–483 27. De Witte O et al (1994) Acute effect of carmustine on glucose metabolism in brain and glioblastoma. Cancer 74(10): 2836–2842 28. Chen W et al (2006) 18F-FDOPA PET imaging of brain tumors: comparison study with 18F-FDG PET and evaluation of diagnostic accuracy. J Nucl Med 47(6):904–911 29. Seibyl JP, Chen W, Silverman DH (2007) 3, 4-dihydroxy-6[18f]-fluoro-L-phenylalanine positron emission tomography in patients with central motor disorders and in evaluation of brain and other tumors. Semin Nucl Med 37(6):440–450 30. Popperl G et al (2007) FET PET for the evaluation of untreated gliomas: correlation of FET uptake and uptake kinetics with tumour grading. Eur J Nucl Med Mol Imaging 34(12):1933–1942 31. Vander Borght T et al (1991) Noninvasive measurement of liver regeneration with positron emission tomography and [2-11C]thymidine. Gastroenterology 101(3):794–799 32. Vander Borght T et al (1994) Brain tumor imaging with PET and 2-[carbon-11]thymidine. J Nucl Med 35(6):974–982 33. Goethals P et al (1996) [Methyl-carbon-11] thymidine for in vivo measurement of cell proliferation. J Nucl Med 37(6): 1048–1052 34. Mankoff DA et al (1998) Kinetic analysis of 2-[carbon-11] thymidine PET imaging studies: compartmental model and mathematical analysis. J Nucl Med 39(6):1043–1055
14 Scintigraphy for Brain Tumors 35. De Reuck J et al (1999) [Methyl-11C]thymidine positron emission tomography in tumoral and non-tumoral cerebral lesions. Acta Neurol Belg 99(2):118–125 36. Pruim J et al (1995) Brain tumors: L-[1-C-11]tyrosine PET for visualization and quantification of protein synthesis rate. Radiology 197(1):221–226 37. Willemsen AT et al (1995) In vivo protein synthesis rate determination in primary or recurrent brain tumors using L-[1-11C]-tyrosine and PET. J Nucl Med 36(3):411–419 38. de Wolde H et al (1997) Proliferative activity in human brain tumors: comparison of histopathology and L-[1-(11)C] tyrosine PET. J Nucl Med 38(9):1369–1374 39. Hara T et al (1997) PET imaging of brain tumor with [methyl-11C]choline. J Nucl Med 38(6):842–847 40. Hustinx R et al (2003) Whole-body tumor imaging using PET and 2-18F-fluoro-L-tyrosine: preliminary evaluation and comparison with 18F-FDG. J Nucl Med 44(4):533–539 41. Rutten I et al (2007) PET/CT of skull base meningiomas using 2-18F-fluoro-L-tyrosine: initial report. J Nucl Med 48(5):720–725 42. Lucignani G et al (1997) Differentiation of clinically nonfunctioning pituitary adenomas from meningiomas and craniopharyngiomas by positron emission tomography with [18F]fluoro-ethyl-spiperone. Eur J Nucl Med 24(9): 1149–1155 43. Black KL et al (1989) Use of thallium-201 SPECT to quantitate malignancy grade of gliomas. J Neurosurg 71(3):342–346 44. Dierckx RA et al (1994) Sensitivity and specificity of thallium-201 single-photon emission tomography in the functional detection and differential diagnosis of brain tumours. Eur J Nucl Med 21(7):621–633 45. Kaplan WD et al (1987) Thallium-201 brain tumor imaging: a comparative study with pathologic correlation. J Nucl Med 28(1):47–52 46. Rollins NK, Lowry PA, Shapiro KN (1995) Comparison of gadolinium-enhanced MR and thallium-201 single photon emission computed tomography in pediatric brain tumors. Pediatr Neurosurg 22(1):8–14 47. Maria BL et al (1997) Correlation between gadoliniumdiethylenetriaminepentaacetic acid contrast enhancement and thallium-201 chloride uptake in pediatric brainstem glioma. J Child Neurol 12(6):341–348 48. Ricci M et al (1996) Relationship between thallium-201 uptake by supratentorial glioblastomas and their morphological characteristics on magnetic resonance imaging. Eur J Nucl Med 23(5):524–529 49. Ishibashi M et al (1995) Thallium-201 in brain tumors: relationship between tumor cell activity in astrocytic tumor and proliferating cell nuclear antigen. J Nucl Med 36(12):2201–2206 50. Oriuchi N et al (1993) Clinical evaluation of thallium-201 SPECT in supratentorial gliomas: relationship to histologic grade, prognosis and proliferative activities. J Nucl Med 34(12):2085–2089 51. Oriuchi N et al (1996) Independent thallium-201 accumulation and fluorine-18-fluorodeoxyglucose metabolism in glioma. J Nucl Med 37(3):457–462 52. Hirano T et al (1997) Technetium-99m(V)-DMSA and thallium-201 in brain tumor imaging: correlation with histology and malignant grade. J Nucl Med 38(11):1741–1749
425 53. Zingale A et al (1995) Thallium-201-SPECT and 99Tc-HMPAO SPECT imaging to study functionally cerebral supratentorial neoplasms: the biological basis of the functional imaging interpretation. J Neurosurg Sci 39(4):227–235 54. Buchpiguel CA et al (1995) PET versus SPECT in distinguishing radiation necrosis from tumor recurrence in the brain. J Nucl Med 36(1):159–164 55. Bagni B et al (1995) SPET imaging of intracranial tumours with 99Tcm-sestamibi. Nucl Med Commun 16(4):258–264 56. Beauchesne P et al (1998) Is cerebral tomoscintigraphy with 99mTc-MIBI useful in the diagnosis of local recurrence in patients with malignant gliomas? Cancer Radiother 2(1): 42–48 57. Maffioli L et al (1996) Clinical role of technetium-99m sestamibi single-photon emission tomography in evaluating pretreated patients with brain tumours. Eur J Nucl Med 23(3):308–311 58. O’Tuama LA et al (1993) Thallium-201 versus technetium99m-MIBI SPECT in evaluation of childhood brain tumors: a within-subject comparison. J Nucl Med 34(7):1045–1051 59. Shih WJ et al (1993) Tc-99m sestamibi uptake by cerebellar metastasis from bronchogenic carcinoma. Clin Nucl Med 18(10):887–890 60. Kuwert T et al (1996) Uptake of iodine-123-alpha-methyl tyrosine by gliomas and non-neoplastic brain lesions. Eur J Nucl Med 23(10):1345–1353 61. Kuwert T et al (1997) Iodine-123-alpha-methyl tyrosine in gliomas: correlation with cellular density and proliferative activity. J Nucl Med 38(10):1551–1555 62. Langen KJ et al (2000) Transport mechanisms of 3-[123I] iodo-alpha-methyl-L-tyrosine in a human glioma cell line: comparison with [3H]methyl-L-methionine. J Nucl Med 41(7):1250–1255 63. Woesler B et al (1997) Non-invasive grading of primary brain tumours: results of a comparative study between SPET with 123I-alpha-methyl tyrosine and PET with 18F-deoxyglucose. Eur J Nucl Med 24(4):428–434 64. Hellwig D et al (2008) Intra-individual comparison of p-[123I]-iodo-L-phenylalanine and L-3-[123I]-iodo-alphamethyl-tyrosine for SPECT imaging of gliomas. Eur J Nucl Med Mol Imaging 35(1):24–31 65. Pacak K et al (2004) The role of [(18)F]fluorodeoxyglucose positron emission tomography and [(111)In]-diethylenetri aminepentaacetate-D-Phe-pentetreotide scintigraphy in the localization of ectopic adrenocorticotropin-secreting tumors causing Cushing’s syndrome. J Clin Endocrinol Metab 89(5):2214–2221 66. Tabarin A et al (1999) Usefulness of somatostatin receptor scintigraphy in patients with occult ectopic adrenocorticotropin syndrome. J Clin Endocrinol Metab 84(4):1193–1202 67. Acosta-Gomez MJ et al (2005) The role of somatostatin receptor scintigraphy in patients with pituitary adenoma or postsurgical recurrent tumours. Br J Radiol 78(926):110–115 68. Moulik PK et al (2002) The role of somatostatin receptor scintigraphy in the management of pituitary tumours. Nucl Med Commun 23(2):117–120 69. de Herder WW et al (2007) Diagnostic imaging of dopamine receptors in pituitary adenomas. Eur J Endocrinol 156(1): 53–56
426 70. Kessler LS et al (1998) Thallium-201 brain SPECT of lymphoma in AIDS patients: pitfalls and technique optimization. AJNR Am J Neuroradiol 19(6):1105–1109 71. Lorberboym M et al (1998) Thallium-201 retention in focal intracranial lesions for differential diagnosis of primary lymphoma and nonmalignant lesions in AIDS patients. J Nucl Med 39(8):1366–1369 72. O’Doherty MJ et al (1997) PET scanning and the human immunodeficiency virus-positive patient. J Nucl Med 38(10): 1575–1583 73. Villringer K et al (1995) Differential diagnosis of CNS lesions in AIDS patients by FDG-PET. J Comput Assist Tomogr 19(4):532–536 74. Krishna L et al (1992) Abnormal intracerebral thallium localization in a bacterial brain abscess. J Nucl Med 33(11): 2017–2019 75. Tonami N et al (1990) Thallium-201 accumulation in cerebral candidiasis: unexpected finding on SPECT. Clin Nucl Med 15(6):397–400 76. Gorniak RJ et al (1997) Thallium-201 uptake in cytomegalovirus encephalitis. J Nucl Med 38(9):1386–1388 77. Bernat I, Toth G, Kovacs L (1994) Tumour-like thallium-201 accumulation in brain infarcts, an unexpected finding on single-photon emission tomography. Eur J Nucl Med 21(3): 191–195 78. Kallen K et al (1997) Evaluation of malignancy in ring enhancing brain lesions on CT by thallium-201 SPECT. J Neurol Neurosurg Psychiatry 63(5):569–574 79. Douglas JG et al (2006) [F-18]-fluorodeoxyglucose positron emission tomography for targeting radiation dose escalation for patients with glioblastoma multiforme: clinical outcomes and patterns of failure. Int J Radiat Oncol Biol Phys 64(3): 886–891 80. Grosu AL et al (2005) Reirradiation of recurrent high-grade gliomas using amino acid PET (SPECT)/CT/MRI image fusion to determine gross tumor volume for stereotactic fractionated radiotherapy. Int J Radiat Oncol Biol Phys 63(2): 511–519 81. Cher LM et al (2006) 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 47(3):410–418 82. Bruehlmeier M et al (2004) Assessment of hypoxia and perfusion in human brain tumors using PET with 18F-fluoromi sonidazole and 15O-H2O. J Nucl Med 45(11):1851–1859 83. Rozental JM, Levine RL, Nickles RJ (1991) Changes in glucose uptake by malignant gliomas: preliminary study of prognostic significance. J Neurooncol 10(1):75–83 84. Rozental JM et al (1993) Acute changes in glucose uptake after treatment: the effects of carmustine (BCNU) on human glioblastoma multiforme. J Neurooncol 15(1):57–66 85. Barker FG II et al (1997) 18-Fluorodeoxyglucose uptake and survival of patients with suspected recurrent malignant glioma. Cancer 79(1):115–126 86. Raez L et al (1999) Treatment of AIDS-related primary central nervous system lymphoma with zidovudine, ganciclovir, and interleukin 2. AIDS Res Hum Retroviruses 15(8):713–719
G.N. Sfakianakis et al. 87. Shields AF et al (1998) Imaging proliferation in vivo with [F-18]FLT and positron emission tomography. Nat Med 4(11):1334–1336 88. Rasey JS et al (2002) Validation of FLT uptake as a measure of thymidine kinase-1 activity in A549 carcinoma cells. J Nucl Med 43(9):1210–1217 89. Saga T et al (2006) Evaluation of primary brain tumors with FLT-PET: usefulness and limitations. Clin Nucl Med 31(12):774–780 90. Yamamoto Y et al (2006) 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 8(6):340–347 91. Chen W et al (2007) Predicting treatment response of malignant gliomas to bevacizumab and irinotecan by imaging proliferation with [18F] fluorothymidine positron emission tomography: a pilot study. J Clin Oncol 25(30):4714–4721 92. Pauleit D et al (2005) O-(2-[18F]fluoroethyl)-L-tyrosine PET combined with MRI improves the diagnostic assessment of cerebral gliomas. Brain 128(pt 3):678–687 93. Roelcke U et al (1998) Dexamethasone treatment and plasma glucose levels: relevance for fluorine-18-fluorodeoxyglucose uptake measurements in gliomas. J Nucl Med 39(5):879–884 94. Rachinger W et al (2005) Positron emission tomography with O-(2-[18F]fluoroethyl)-l-tyrosine versus magnetic resonance imaging in the diagnosis of recurrent gliomas. Neurosurgery 57(3):505–511 95. Schwartz RB et al (1991) Radiation necrosis vs high-grade recurrent glioma: differentiation by using dual-isotope SPECT with 201TI and 99mTc-HMPAO. AJNR Am J Neuroradiol 12(6):1187–1192 96. Schwartz RB et al (1998) Dual-isotope single-photon emission computerized tomography scanning in patients with glioblastoma multiforme: association with patient survival and histopathological characteristics of tumor after highdose radiotherapy. J Neurosurg 89(1):60–68 97. Lorberboym M et al (1997) The role of thallium-201 uptake and retention in intracranial tumors after radiotherapy. J Nucl Med 38(2):223–226 98. Vertosick FT Jr et al (1994) Correlation of thallium-201 single photon emission computed tomography and survival after treatment failure in patients with glioblastoma multiforme. Neurosurgery 34(3):396–401 99. Haldemann AR et al (1995) Somatostatin receptor scintigraphy in central nervous system tumors: role of blood-brain barrier permeability. J Nucl Med 36(3):403–410 100. Schmidt M et al (1998) Somatostatin receptor imaging in intracranial tumours. Eur J Nucl Med 25(7):675–686 101. Klutmann S et al (1998) Somatostatin receptor scintigraphy in postsurgical follow-up examinations of meningioma. J Nucl Med 39(11):1913–1917 102. Nelson SJ, Vigneron DB, Dillon WP (1999) Serial evaluation of patients with brain tumors using volume MRI and 3D 1H MRSI. NMR Biomed 12(3):123–138
Index
A Abscess, 176 Acquired immuno deficiency syndrome (AIDS), 8, 304, 305, 308, 310, 404, 412–414, 418 ACTH, 327 Adamantinous craniopharyngiomas, 337 ADC. See Apparent diffusion coefficient Adult brain tumors, 2, 8, 10 Alpha-fetoprotein (AFP), 205 Aminoacid PET, 421–423 Anaplastic astrocytoma (WHO grade III), 36, 37, 157–161 Anaplastic ependymoma, 43, 44, 113, 118, 119 Anaplastic (malignant) meningiomas, 259, 263 Anaplastic neurocytoma, 144, 146 Anaplastic oligoastrocytoma, 43, 45 Anaplastic oligodendroglioma, 98, 104–106 Anatomy, 325–327, 333 Aneurysms, 325, 341, 359, 364 Angiocentric lymphomas, 304, 307, 315 Angiogenesis, 374 Angioglioma, 101 Angiography, 93, 101, 112, 127, 128, 142, 146, 147 Angiomatous meningiomas, 279 Anisotropy/Anisotropic, 21–24, 56–58 Antoni A type, 241, 242 Antoni type B, 241, 242 Apparent diffusion coeffcient (ADC), 21, 23, 24, 76, 127, 169, 171, 176, 178, 179, 183, 224, 263, 270, 273, 289, 297, 305, 306, 312, 384, 385 Arachnoid cysts, 21, 335, 341, 348, 349, 354–356 Arteriovenous malformation (AVM), 393 Astrocytic tumors, 2, 4, 5, 8, 10 Astrocytic tumour, 35–42, 45 Astrocytoma, 2, 4, 5, 8, 10, 42, 44, 45, 76, 78–98, 100, 105, 109, 119, 121, 127, 131, 132, 136, 141–144, 148 ATP, 64 Atypical meningiomas, 259, 263 Atypical teratoid/rhabdoid tumors, 221–225 B Basement membrane, 374 Basics of DTI, 56–60 Batson’s plexus, 375 Biophysical basis of DTI, 56–57 Blood−brain barrier (BBB), 166, 173, 377, 382, 384–386, 388, 392, 393, 402–404, 408–410, 416, 421
Blood oxygenation level dependent (BOLD), 13, 20, 26, 28, 50–52, 55, 56, 64, 65 Brain abscess, 176, 177, 180, 412, 417 Brain metastases, 373, 374, 382 Brainstem glioma (BSG), 2, 10, 11, 78, 118, 121 Bromocriptine, 331, 333 C Calvarial, 93, 98, 104, 107, 125, 141 Capillary/lymphatic bed, 374, 375, 384 Carbon-11, 389 Cavernous hamangiomas, 393, 394 CDK4, 37–38, 44 CDK6, 37–38 Cerebral blood flow (CBF), 50–52, 64, 173 Cerebral blood volume (CBV), 50, 64, 173 Cerebral blood volume (CBV) mapping, 74, 78, 96 Cerebral metabolic rate of oxygen (CMRO2), 50, 52 Cerebral neuroblastoma, 221–223 Chemotherapy, 395, 396 Chiari I, 139 Chiasmatic and hypothalamic gliomas, 325, 342 Childhood meningiomas, 256, 257 Chloroma, 311, 313 Cho/Cr, 174, 176, 178, 191, 195 Choline, 29–32, 74, 80, 89, 112, 128, 132, 146, 173, 390–392 Cho/NAA, 174, 176, 185, 188, 191, 195 Chordoid, clear cell, 259 Choriocarcinoma, 203, 205, 342 Choroid plexus carcinomas, 111, 112 Choroid plexus papillomas (CPP), 73, 111–117 Chromosome 7 and 1q, 45 17p, 38, 40, 45 6q, 9 and 13, 45 (i icl 13) Classic medulloblastomas, 216 Classification, 4, 5, 230, 235, 257, 259, 276, 289 Clear cell ependymoma, 117–118 Clinical course, brain tumors, 1–11 Clinical features, PCNSL 305 Clinical findings choroid plexus papilloma 111 neurocytoma 144 Clinical signs, cerebral neuroblastoma, 221 Clinical symptoms, brain metastasis, 376–377
427
428 Clinical syndrome, pituitary adenomas, 332 11 C-methyl-methionine, 78, 105, 109, 142, 408 CMRO2. See Cerebral metabolic rate of oxygen Computed tomography (CT), 13–15, 79, 80, 82, 85–87, 90–95, 97–99, 104, 105, 112, 116, 119, 123–125, 128–136, 138–141, 144, 146, 159, 160, 164–166, 189, 191, 192, 201, 203–205, 207–209, 217, 219, 221, 222, 231, 235, 237–240, 242, 243, 246, 248–250, 260, 263–266, 273, 276, 278, 279, 285, 287, 289, 293, 294, 296, 305, 306, 308–310, 312, 316, 322, 327, 330, 332–335, 337, 339, 340, 343, 349–351, 353, 355, 357, 360, 361, 368, 377–379, 382, 384, 386, 389, 392–396 Constructive interference of steady state (CISS), 18, 20 Contrast enhancement, 166, 173, 182, 191 Conventional spin echo (SE), 379, 380, 382, 383 CPP. See Choroid plexus papillomas Craniopharyngioma, 6, 10, 325, 335–344, 348, 354 Creatine (Cr), 29, 31, 74, 80, 92, 119, 127, 128, 132, 146, 173, 390 CSF cytology, 387 Cushing’s disease, 327, 328 Cystic meningioma, 274, 276, 278, 289 D DCG. See Dysplastic cerebellar gangliocytoma Deleted in malignant brain tumours 1 (DMBT1), 40 De novo/GBM, 159 Deoxyhemoglobin, 51, 52, 55, 379, 380, 382 Dermoids, 348–353 Desmoplastic infantile astrocytoma (DIA), 78, 141–144 Desmoplastic infantile ganglioglioma (DIG), 141–144 Desmoplastic medulloblastomas, 216 Differential diagnosis, 97, 117, 119, 131, 144, 176–178, 191, 220–222, 225, 232, 233, 238, 239, 242, 243, 246, 251–253, 263, 282–289, 294, 308, 310, 313, 391–394 Diffuse astrocytoma, 36 Diffuse infiltrative astrocytoma, 36–41, 78–80, 89 Diffusion, 20–24, 30 imaging, 127 perfusion, 20 and perfusion imaging, 384 tensor, 21, 24 Diffusion tensor imaging (DTI), 21, 24, 49–67, 170, 172–173, 263 Diffusion-weighted imaging (DWI), 20–23, 30, 76, 116, 166, 169, 176, 178, 179, 204, 219, 279, 294, 297, 305, 306, 312 DIG. See Desmoplastic infantile ganglioglioma Direct intra-operative cortical, 63 Dissemination, 159, 166, 168, 216, 222 DMBT1. See Deleted in malignant brain tumours 1 DNT. See Dysembryoplastic neuroepithelial tumor DTI. See Diffusion tensor imaging Dural metastasis, 287, 288 Dural tail, 333, 336 Dural tail sign, 266, 291 DWI. See Diffusion-weighted imaging Dynamic contrast enhanced perfusion (PWI), 166 Dynamic contrast enhancement methods (DCE-MRI), 76 Dysembryoplastic neuroepithelial tumor (DNT), 109, 125–127, 132
Index Dysplastic cerebellar gangliocytoma (DCG), 135, 138–141 Dysplastic gangliocytomas, 135, 138–141 E Echo-planar imaging (EPI), 16, 20, 28 Edema, 256, 267–270, 272–274, 276, 278, 283, 297, 375–377, 379–382, 384, 385, 390–396 EGF. See Epidermal growth factor EGFR. See Epidermal growth factor receptor EGFRvIII, 40 Eigenvectors, 57, 59 Eloquent cortical or subcortical area, 63 Embryonal carcinoma, 203, 205–207, 342 Embryonal cell carcinoma, 203, 205–207 Embryonal tumors, 215–226 Endodermal sinus or yolk-sac tumors, 203, 205 En plaque meningiomas, 257, 260, 280 Ependymal rosettes, 117 Ependymal tumour, 44–45 Ependymoma, 2, 4, 5, 10–11, 35, 43–45 Ependymomas, 73, 113–125, 142, 143 EPI. See Echo-planar imaging Epidemiology, 1–11 Epidermal growth factor (EGF), 40 Epidermal growth factor receptor (EGFR), 37, 40, 42–44 Epidermoid, 21, 23, 335, 341, 348–353 Esthesioneuroblastoma (Olfactory neuroblastoma), 229–232 Extensive nodular medulloblastomas, 216 Extra-axial, 2, 9 Extramedullary myeloid cell tumors (EMT), 311 F FA. See Fractional anisotropy Facial nerve schwannoma, 242, 246, 250, 251 Fast imaging with steady precession (FISP), 18 Fast spin echo (FSE), 16–18, 20, 379, 380 FDG-PET, 93, 112, 128, 132, 142, 310, 388, 389, 395, 407–408, 412–415, 417, 420–423 F-DOPA, 408 F-DOPA PET, 408, 421–422 FET-PET, 420, 421 18 F-flurodeoxyglucose (FDG), 78, 93, 105, 112, 127, 128, 132, 142, 310 FGF. See Fibroblast growth factor Fibrillary WHO grade II astrocytomas, 80 Fibroblast growth factor (FGF), 40, 41, 44 Fibrous displasia, 260 Fibrous (fibroblastic) meningiomas, 258 Fingertapping, 52–54, 61, 62, 64, 65 FISP. See Fast imaging with steady precession Fluid-attenuated inversion recovery (FLAIR), 16–19, 32, 379–382, 385, 386, 392, 395, 396 fMRI. See Functional MR imaging Focal, circumscribed astrocytomas, 79 Fractional anisotropy (FA), 58, 59, 61, 66, 67, 76, 171–172, 183, 263, 306 Frequency, 255, 256, 282 FSE. See Fast spin echo Functional brain mapping, 412, 416–417 Functional MR imaging (fMRI), 20, 24, 26, 31, 49–67
429
Index G Gadolinium (Gd), 15, 16, 382–385 Galactocerebroside (GC), 98, 101 Gangliocytomas, 109, 129, 131, 132, 135–140 Gangliogliomas, 93–95, 109, 127, 129–136, 138, 141–144 Ganglion cell tumors, 73, 79, 129–144 Gemistocytic astrocytoma, 98, 103, 158, 162 General limitations DTI, 59–60 fMRI, 50, 53–56 Genes MDM2, 37–39 p53, 37–40, 42, 44, 45 p14ARF, 38–39, 45 p21CIP1, 37, 39 p16 INK4A, 37–38, 40, 44, 45 p15 INK4B, 37, 38, 45 p27KIP1, 37, 39 Rb, 37–39, 41, 44 Germ cell tumors, 201–208 Germinoma, 203–206, 342–348, 350, 356, 361, 362, 365 GFAP. See Glial fibrillary acid protein Giant cerebral aneurysms, 393 Glial fibrillary acid protein (GFAP), 79, 93, 94, 98, 101, 103, 111, 118, 141, 144 Glioblastoma, 121 Glioblastoma multiforme (WHO grade IV), 2, 4, 8, 36, 37, 88, 157, 159–188, 191 Gliofibroma, 141 Gliomas, 2, 4, 8, 10, 11, 15, 25, 26, 29, 30, 35–45, 157–159, 166, 173, 178, 182, 411–412, 416, 417, 419–423 Gliomatosis cerebri, 158, 189–195 Gliosarcoma, 157, 188–190 Glomus jugulare, 246, 253 Glucose, 30, 50, 51, 64 Glutamate (Glx), 29 Glutamine, 29 Glycoprotein (laminin), 374 Gradient echo, 18 Gradient echo pulse sequences and echo planar imaging (EPI), 16 Grading, 76, 78, 93, 98, 101, 104, 105, 109, 127 Grading/Grading system, 3–5 Grading of astrocytomas, 157, 173 Grading system is the Smith (AFIP), 101 Granulocytic sarcoma (chloroma), 311–313 Granulomas, 392 GRASE, 20 Growth factors, 37, 39–41, 43, 44 Growth-hormone secreting tumors, 330 H Half-Fourier acquisition and single-shot turbo spin-echo (HASTE), 20 Hamartomas of the tuber cinereum, 357 Hemangioblastoma, 2, 6, 9, 292–296 Hemangiopericytoma, 282, 287, 289–291 Hemorrhage, 158, 159, 164, 166, 178, 189, 257, 267, 276, 279, 280, 294, 331–333, 335, 337, 368, 375, 379, 380, 392, 393, 396 Hexa methylpropylene amine oxime (HMPAO), 388, 403, 410, 417
High-grade astrocytic tumors, 8 High-grade astrocytomas, 78 High-grade gliomas, 76, 79–80, 128, 157–195 High-grade tumors, 404, 407–410, 412, 421 Histiocytosis X, 362, 366 Histological classification, 35, 36 Histologic classification, 1–11 Histopathology, 1, 3 Hodgkin’s lymphoma, 303 1H-proton magnetic resonance spectroscopy (1H-MRS), 28 Hydrogen proton magnetic resonance spectroscopy, 73–74 Hyperostosis, 260, 263–265, 271, 287, 289 I IAT. See Intracarotid amobarbital test I by WHO, 259 123 I labeled iodo-methyl-tyrosine (IMT), 403, 404, 410, 423 Imaging CT, 221 findings, 305–310, 338, 344, 366 studies, 377–391, 395, 396 Immunohistochemical marker, 98 Immunohistochemistry, 101, 118, 135 131 Ina, 404, 423 Incidence, 1, 2, 5–11 Infection, 386, 391, 392 Infectious brain, 392 Infratentorial, 2, 9–11, 114, 117, 119–121, 141 Infundibular masses, 361–362 111 In-pentetreotide, 410 Intra-axial, 2, 8, 9 Intracanalicular vestibular schwannoma, 242, 243, 252 Intracarotid amobarbital test (IAT), 63 Intralabyrinthine vestibular or cochlear schwannoma, 242 Intra-operative mapping, 60, 63 Intraosseous meningiomas, 256, 257, 280 Intraventricular, 113, 115, 119, 124, 127, 132, 137, 143, 144, 148 Intraventricular meningiomas, 280, 289 Inversion recovery (IR), 16–18 Iso-and anisotropic diffusion, 170 Isotropic diffusion, 56–58 J Jugular foramen schwannomas, 243 L Lactate (Lac), 64, 74, 75, 80, 88, 127, 128, 146, 173, 174, 176, 178, 180–182, 185, 188, 390 Language expression tasks, 52 Language functions, 52 Large cell medulloblastoma, 216, 217 Large vestibular schwannoma, 241–243, 246 Leptomeningeal dissemination, 88 Leptomeningeal primary lymphomas, 307 Leptomeninges, 93, 141, 375, 376, 378, 386, 387, 391 Leukemia, 311–322, 362, 364 Lhermitte-Duclos, 135, 138
430 Limitations DTI, 50, 56, 59–60 fMRI, 50, 53–56 presurgical DTI, 66–67 presurgical fMRI, 64–66 Lipids, 159, 174, 176, 180, 181, 183, 185, 187, 188, 191, 390, 392 Lipoblastic meningioma, 278 Location, 216, 221, 222, 256, 257, 261, 274, 276, 279, 284, 288, 292, 296 Low-grade astrocytomas, 8, 10, 76, 78–80, 82, 105, 119, 121, 132, 136 Low-grade gliomas, 25, 26, 73–148, 407–410, 421 Low-grade tumors, 73–77, 89, 104, 109, 110, 127, 129, 144, 169, 173 Lung carcinoma, 375, 383 Lymhocytic hypophysitis, 361, 362 Lymphoma, 2, 6, 8, 176, 178, 179, 181, 183, 266, 287, 303–322, 362, 364, 367, 404, 410–414, 417–418, 423 Lymphomatous and leukemic meningitis (LM), 316 M Macroadenomas, 328–331, 333, 334 Magnetic resonance (MR), 110, 111, 131, 166, 169, 170, 173, 174, 178, 181, 182, 186, 188, 189, 220, 222, 224, 225, 231, 233, 235, 237–240, 242, 246, 248, 249, 260, 261, 263, 265, 266, 268, 271, 273, 276, 279, 288, 305, 306, 308, 310, 313, 316, 327–329, 333, 342, 344, 349, 350, 354, 356 DTI (see Diffusion tensor imaging) spectroscopy, 29, 30 techniques, 16, 20 Magnetic resonance imaging (MRI), 13, 15–32, 75–82, 85, 89, 90, 93, 95–97, 99, 104, 107, 108, 111, 112, 114, 116, 118–121, 125–133, 135, 138–148, 159, 166, 170, 178, 182, 185, 191, 201, 203, 205, 208, 209, 217–221, 223, 224, 226, 231, 241, 260, 266, 270, 276, 278, 279, 289, 293, 305–308, 310, 322, 327, 331, 337, 341, 354, 357–361, 379–396 Magnetic resonance spectroscopy (MRS), 74, 80, 88, 89, 112, 119, 132, 146, 166, 173, 178, 182, 188, 191, 263, 306, 309, 310, 390–392, 396 Magnetization transfer (MT), 16, 17, 383 Malignant optic nerve glioma (MONG), 233–235 MALT lymphomas, 304 Max interacting protein 1 (MXI1), 40 MBP. See Myelin basic protein Mean diffusivity (MD), 58 Mean transit time (MTT), 173 Medulloblastoma, 2, 4, 6, 10, 11, 21, 215–221 Melanocytic neoplasms–Melanocytoma, 240–241 Melanoma, 374–377, 379, 380, 387, 389 Meningeal carcinomatosis, 376–378, 386, 387, 391 Meningioma, 2, 6, 9, 15, 21, 229, 230, 232, 233, 235, 237–239, 241, 242, 246, 247, 249, 252, 253, 255–289, 294, 325, 333–339, 357, 367 Meningothelial (syncytial) meningiomas, 258, 273 Metastases, 2, 8, 9, 166, 176, 178, 181, 189, 328, 330, 361, 362, 364, 373, 374, 376, 382, 384, 386, 391 Metastasis, 176, 373–396 Metastatic tumors, 375–379, 383–386, 388–395, 411
Index Methionine (MET), 109, 112, 403, 405, 407–409, 415–417, 420, 421 MET-PET, 408, 421 MIB-1 index, 79 Microadenomas, 328 Microcystic meningiomas, 258, 289 Miscelaneous mases, 364 Mixed germ cell tumor, 203 Mixed germ tumor, 203 Mixed tumors, 342 Molecular, 35–45 Monroe, 376 Motor, sensory, 50, 52, 60 MR. See Magnetic resonance MRI. See Magnetic resonance imaging MRS. See Magnetic resonance spectroscopy MT. See Magnetization transfer MTT. See Mean transit time 99 mTc-HMPAO, 78, 127, 403 99m Tc-Sestamibi (MIBI), 403–406, 410 Multicentric/GBM, 166, 169, 189 Multicentric oligodendrogliomas, 101 Multifocal/GBM, 166, 169 Multiple sclerosis (MS), 393–395 Myelin basic protein (MBP), 98, 101 Myeloperoxidase, 311–313 Myo-inositol (mI), 30 Myxopapillary ependymoma, 44 N NAA/Cr, 176, 178, 182, 191, 195 NAA/Cho, 176 N-acetylaspartate (NAA), 29, 74, 75, 80, 92, 119, 127, 128, 132, 146, 385, 390, 391 Necrosis, 74, 76, 78, 80, 89, 93, 101, 103, 118, 119, 125, 135, 144, 157–159, 164–167, 169, 173, 182, 185, 187–189, 404, 407, 410, 412, 420–423 Neuroblastoma, 148 Neurocytomas, 73, 97, 128, 144–148 Neurofibromas, 357–358 Neurofibromatosis, 232, 251–254 Neurofibromatosis type 1 (NF1), 41, 42, 232, 233, 251, 254 Neurofibromatosis type 2 (NF2), 43, 45, 251, 254 Neuronavigation, 50, 60, 63–64 Neurosarcoidosis, 356, 357, 362 Non-Hodgkin’s lymphoma, 303, 304 Non-infiltrative diffuse or circumscribed astrocytomas, 80–98 O Octreotide (Octreoscan), 403, 404 Oligoastrocytic tumour, 45 Oligoastrocytoma, 43, 45, 101, 103 Oligodendroglioma, 2–5, 8, 35, 42–45, 73, 77, 79, 94, 97–111, 127, 144, 148 Optic nerve arachnoid cyst, 239–240 Optic nerve glioma, 229, 232–235, 238–240 Optic nerve meningioma, 235–238, 253 Osteoma, 260 Oxygen-15, 389 Oxyhemoglobine/deoxyhemoglobin, 52
431
Index P Paget’s disease, 260 Papillary ependymoma, 117 Papillary meningiomas, 259 Parenchymal pineal tumors of intermediate differentiation (PPTID), 207–209 Pathogenesis of brain metastasis, 374–375 Pathology, 78–79, 84, 88, 93–104, 111–112, 116, 128, 131, 141, 144, 157–159, 166, 173, 182, 189, 191, 216, 221, 222, 231–235, 241, 254, 257–260, 271, 289, 292–293, 304–305, 375–376 Pediatric brain tumors, 1–2, 10–11 Perfusion, 384, 385, 388, 392, 393 Perfusion MR, 263, 306, 310 Perfusion MR imaging, 178, 182 Perfusion-weighted imaging (PWI), 25–32, 166, 173 Peritumoral edema, 76, 79, 93, 148 Perivascular pseudorosettes, 94, 117, 118 PET. See Positron emission tomography PET/CT, 407, 424 PET/MRI, 407, 424 Phosphatidylinositol 3-kinase pathway/PTEN, 40–41 Pilocytic, 2, 4, 8 Pilocytic astrocytoma, 4, 5, 8, 10, 41–42, 78, 84–92, 97, 342 Pilomyxoid astrocytoma (PMA), 42, 88–92 Pineal cell tumor, 201–213 Pineal gland, 201–205, 212 Pineal region tumors, 3, 5, 10 Pineal tumor of intermediate differentiation (PTID), 203, 207–209 Pineal tumors, 201–213 Pineoblastoma, 203, 207–211 Pineocytomas, 203, 208 Pituicytoma, 361, 363 Pituitary adenomas, 2, 8, 325, 327–333, 361 Pituitary apoplexy, 8 Pituitary gland, 325, 327, 328, 334, 360, 361 Pituitary hyperplasia, 360, 361 Plasmacytomas, 304 Plasmocytoma, 282, 288 Platelet-derived growth factor (PDGF) Platelet-derived growth factor-a (PDGF-a) Pleomorphic xanthoastrocytoma (PXA), 42, 78, 92–96, 127, 132, 136, 141, 142 Pleomorphism, 84, 93, 94, 98, 101, 111 PMA. See Pilomyxoid astrocytoma PMC. See Premotor cortex PNETs. See Primitive neuroectodermal tumor Positron emission tomography (PET), 388, 389, 392, 395, 396 guided stereotactic biopsies, 420–421 imaging, 402, 405–407, 416 Positron emitters, 402–406 Posterior parietal cortex (PPC), 65 Post-therapeutic MR, 182 Post therapy period, 394–396 Premotor cortex (PMC), 65 Presurgical DTI, 60–67 Presurgical fMRI, 60–67 Primary cerebral schwannoma, 251, 252 Primary extraneuraxial meningiomas (PEMs), 255, 256, 280, 282
Primary glioblastomas, 159 Primary neuraxial meningiomas (PNM), 255, 280 Primitive neuroectodermal tumors (PNETs), 4, 5, 6, 119, 121, 142, 209, 215, 221, 222, 224 Prognosis, 157–159, 163, 164, 166, 189, 194 Prolactinomas, 328, 330 Proteolysis, 374 Psammomatous meningioma, 258, 259, 263 PTID. See Pineal tumor of intermediate differentiation PWI. See Perfusion-weighted imaging PXA. See Pleomorphic xanthoastrocytoma R Radiation necrosis, 169, 182, 185, 187, 390, 395 Radiation therapy, 407, 413–416, 418, 420 Radioisotopes, 402–404, 406 Radiolabeled amino acids, 404, 421 Radiolabeled deoxy-uridine, 404 Radiopharmaceuticals, 402–408, 410, 412, 419, 423 Radiosurgery, 394, 395 Rathke cleft cyst, 335, 341, 348, 350, 353–355 Rathke’s pouch, 335, 353 Receptor, 37, 39–40, 43 Regional cerebral blood volume, 306, 307, 310 Relative cerebral blood volume (rCBV), 25, 28, 29, 76, 77, 79, 87, 96, 105, 127, 132, 171, 173, 174, 178, 181–183, 185, 187, 191, 193, 384, 385 Response to therapy, 407, 412, 417–419 Rhabdoid meningiomas, 259 Rosenthal fibers, 84, 88, 131 S Sarcoidosis, 356, 361, 362, 365 Schwannoma, 2, 6, 9, 229, 230, 241–254 Schwannomas, 357–358 Scintigraphy for brain tumors, 401–424 Scyllo-inositol (sI), 30 Secondary glioblastoma, 159 Secretory meningioma, 258, 278 Sensorimotor cortex, 54, 61, 62, 65, 66 SGCA. See Subependymal giant cell astrocytoma Short tau inversion recovery (STIR), 16, 17 Single-photon emission computed tomography (SPECT), 74, 78, 127, 129, 132, 142, 310, 312, 388, 392, 396 Tl, 410, 413, 418, 420 tracers, 409–410, 422–423 Single photon emitters, 402, 403, 405, 406 Single photon studies (planar/SPECT), 402, 405, 406, 411 SMA. See Supplementary motor area Somatostatin, 403, 404, 406, 410–411 Somatostatin receptors, 403, 410–411 SPECT. See Single-photon emission computed tomography Spectroscopy, 127, 166, 178, 182, 188, 191, 195 Spectroscopy 1H-MR, 128 Steady-state, 18 STIR. See Short tau inversion recovery Subcortical electro-stimulation (ICS and ISS), 63 Subependymal giant cell astrocytoma (SGCA), 36, 42, 78, 94–99 Subependymoma, 113, 127–129, 143 Subtypes, 79, 80, 98
432 Supplementary motor area (SMA), 55, 61–63 Suppressor gene, 217 Supratentorial, 2, 4, 8–11, 104, 114, 119, 122–123 Susceptibility-weighted sequences–T2*-weighted images, 52 Symptoms, 2, 3, 8–11, 78, 94, 111, 114, 116, 128, 131, 136, 144, 157–159, 166, 176, 191, 194, 201–203, 216, 256, 257, 268, 287, 292 Synaptophysin, 131, 141, 144 T T2*, 380, 382 Tanycytic ependymoma, 117 Target organs, 374, 375 Tc99m-HMPAO, 127 Technetium-99m, 388 Teratoid-rhabdoid tumors, 221–226 Teratoma, 203, 205, 207 Teratomas, 342 Thallium-205, 388 201 Thallium (Thallous) Chloride (Tl), 404, 410, 422 Therapy of brain tumors, 423 Therapy planning, 404, 407, 408, 412–416 The spectroscopic findings, 191 Time to peak (TTP), 173 Toxoplasmosis, 310, 312, 404, 412–414 Tractography, 24–26, 50, 59–62, 67 Transforming growth factor-b (TGF-b), 40 Transitional (mixed) meningiomas, 258, 260 Trigeminal nerve schwannoma, 242, 247 201 T1-SPECT, 132 Tuberculosis, 356, 362
Index Tumefactive multiple sclerosis (MS), 176 Tumor grading, 25, 30 Tumours of the cranial nerves, 229–254 T1-weighted MR, 379–383, 385–387, 389, 392–396 T2-weighted MR, 379, 390, 392, 393, 396 V Vascular complications, 396 Vascular endothelial growth factor (VEGF), 272, 273 Vascular occlusive disease, 392 Vasogenic edema, 159, 166, 167, 178, 181, 189 VEGFR, 78 Venous system, 375 von Hippel-Lindau (VHL), 292, 294, 296 W Well-differentiated diffuse astrocytomas, 78–80 WHO classification, 36, 42, 45, 78, 84, 92, 94, 111, 113, 189, 303 WHO grade I, 94, 125, 131, 141 WHO grade II, 131, 144, 259, 289 WHO grade III, 259 World Health Organization (WHO), 4, 5, 216 X Xanthoastrocytoma, 42, 78, 92–96, 127, 132, 136, 141, 142 Xanthogranuloma, 364 Y Y labeled, 423S
90