The Brain and Spinal Cord 33

shiao Y. Woo

TOLERANCE OF THE CENTRAL NERVOUS Craniopharyngioma SYSTEM TO IRRADIATION Primitive Neuroectodermal Tumors Postirradiation Cerebral Necrosis Tumors in Children Younger than 3 Years Postirradiation Myelopathy TUMORS COMMON IN ADULTS Other Late Effects Malignant Gliomas: Anaplastic CENTRAL NERVOUS SYSTEM Astrocytoma and Glioblastoma NEOPLASMS Multiforme TUMORS COMMON IN CHILDREN Oligodendroglioma Medulloblastoma Meningioma Astrocytoma Primary Malignant Lymphoma of the Brainstem Gliomas Central Nervous System Ependymomas Carcinoma Metastatic to the Brain Pineal Region and Germ Cell Tumors SPINAL CORD TUMORS

Tolerance of the Central with calcifications by 26 weeks; healing of the degenerative Nervous System to Irradiation foci along with increased microcalcifications was seen by 52 weeks. The radiation-induced changes were quantita- The central nervous system (CNS) is a relatively sensitive, tively greater in younger animals. After 80 Gy delivered often dose-limiting tissue in curative radiation therapy for in 40 fractions, confluent areas of necrosis appeared in the CNS and adjacent neoplasms.1 Radiation injury has been cerebral hemispheres of younger animals by 32 weeks, pro- documented on the basis of clinical signs, imaging findings, gressive necrosis was observed at 52 weeks, and coalescent and histopathologic correlations.2-6 CNS effects are more necrotic regions and atrophy were seen by 78 weeks. pronounced in children and seem to be inversely related Caveness’ findings at lower dose levels are similar to the to age, with the youngest children having the greatest CNS mineralizing microangiopathy seen in children with acute sensitivity. Of further concern is that numerous chemother- leukemia treated with cranial irradiation (20 to 24 Gy) and apy agents are associated with unique or enhanced sequelae methotrexate.12 Small, coalescent foci of necrosis related when given in conjunction with radiation therapy.7 to microvascular changes may account for the clinically Studies of laboratory models of cerebral and spinal silent lacunar lesions evident in magnetic resonance imag- cord changes have helped to clarify the pathophysiology ing (MRI) studies of long-term survivors of childhood brain of radiation-induced neural damage and to associate the tumors.13 extent of damage with time, dose, delivery, and volume.8-10 Postirradiation CNS effects are typically classified tem- For example, Caveness11 reported a detailed investiga- porally as acute, subacute, and late reactions. Fractionated tion of the clinical, histologic, and physiologic findings in radiation rarely produces clinically apparent acute changes, primates after irradiation comparable with that used clini- even when the total dose approaches 70 to 80 Gy.14 Anec- cally. In that study, a single fraction of 20 Gy was uniformly dotal reports of transient exaggeration of neurologic signs fatal by week 26; this dose produced increased intracranial have been related theoretically to perilesional or intralesion- pressure and ventriculomegaly with scattered necrotic foci al edema that develops at the beginning of irradiation.15-17 most dramatically affecting the brainstem. A single frac- Young and colleagues18 found abrupt neurologic deterioration of 15 Gy produced discrete microfoci of white matter tion in 50% of patients who were given large (7.5- to 10- necrosis by 26 weeks, with coalescing areas of white matter Gy) fractions for the treatment of symptomatic cerebral necrosis evident by 56 weeks. Pathologic findings were less metastases; the investigators attributed this effect to the pronounced in animals killed 78 weeks after irradiation, radiation therapy. Although no similar phenomena were suggesting that some recovery had taken place. No clini- reported in studies by the Radiation Therapy Oncology cal or histologic abnormalities were observed after whole- Group (RTOG) in which 6-Gy fractions were adminis- brain irradiation with a single fraction of 10 Gy or with tered, this finding might have been related to the more 40 Gy delivered in 20 fractions over 4 weeks. consistent use of corticosteroids by the RTOG.16 In experiments more relevant to clinical radiation ther- Subacute CNS effects are rather common. Jones19 apy, delivering 60 Gy to the whole brain in fractions of observed transient, self-limited paresthesias with neck 2 Gy over a period of 6 weeks caused papilledema and flexion (Lhermitte’s sign) that lasted for 1 to 2 months anorexia in pubertal monkeys, but no overt clinical findings after spinal cord irradiation. These paresthesias, which were apparent in adult primates. Histologically, the adult occur in 15% to 25% of patients after mantle irradiation brains showed discrete microfoci of white matter necrosis for Hodgkin disease, are thought to result from transient 835 836 PART 10 Central Nervous System

demyelination. Comparable transitory neurologic changes, therapy.15,16,30 Cerebral changes often mimic residual or called the somnolent syndrome, occur after cranial irradia- recurrent tumor, with mass effect, enhancement, or cyst tion such as that given for acute leukemia; mild encepha- formation evident on computed tomography (CT) scans; lopathy or focal neurologic changes can appear after the these changes occur at or near the site of the primary tumor treatment of intracranial tumors.20,21 The cranial syndrome (i.e., the high-dose volume for intracranial neoplasms).31 is thought to stem from a transient demyelinating event Cerebral necrosis unrelated to intraparenchymal neo- and is the result of the effects of radiation on the replicat- plasms has been described in reviews by Kramer and ing oligodendrocytes.8 MRI can detect specific dose-related colleagues32 and Sheline and coworkers.16 Necrosis has changes in gray and white matter within several weeks of been observed after therapy for extracranial tumors (pre- conventionally fractionated irradiation.22 The potential for dominantly those of the skin and paranasal sinuses) and self-limited neurologic deterioration, most often occurring for noninvasive pituitary tumors or craniopharyngiomas. during the second month after irradiation, is important to Postirradiation necrosis is a dose-related phenomenon bear in mind because clinical and radiologic changes can with a relative threshold radiation dose of 50 to 55 Gy if sometimes mimic tumor progression.23,24 treatment is given in a conventional fractionation schedule Late reactions of the normal brain are the dose-limiting (1.8- to 2-Gy fractions given once daily).16,30,33,34 Marks toxicities for radiation treatment. Clinical manifestations and colleagues30 reported the occurrence of necrosis in 5% of focal neurologic deficits, encephalopathy, or neuropsy- of 139 patients with intracranial tumors treated with con- chological dysfunction accompany various morphologic ventional fractionation to doses higher than 45 Gy; 6 of findings, such as atrophy, calcifications, diffuse white mat- the 7 patients who had necrosis were given total doses in ter degeneration, and focal necrosis. Permanent or progres- excess of 63 Gy. sive myelopathy can result from spinal cord irradiation. The dose-time relationship is critical in determining Late changes in hypothalamic-pituitary function manifest CNS tolerance, with fraction size the dominant factor as specific radiation-induced endocrine deficits. among patients receiving doses of 60 to 63 Gy. Necrosis in patients who receive less than 60 Gy usually has been Postirradiation Cerebral Necrosis associated with fractions of greater than 2.5 Gy.16 Aristiza- Postirradiation cerebral necrosis, the most direct effect of bal and colleagues33 documented necrosis after treatment CNS irradiation, is thought to result from direct effects of for pituitary tumors in 3% of 106 patients given less than radiation on the replicating glial cell compartments and 2.2 Gy per day (total dose of about 50 Gy) compared with the capillary endothelial cells.8,9 The theoretical relation- 15% of 13 patients given doses of greater than 2.2 Gy daily. ship between primary glial cells (i.e., type II astrocytes and The influence of fraction size was made apparent in the oligodendrocytes) and vascular endothelial components isoeffect analyses of brain necrosis reported by Sheline and is shown in Figure 33-1.8,25-29 Clinical signs of postir- colleagues.16 radiation cerebral necrosis are seen 6 to 36 months after Sheline and coworkers16 contend that the brain can tolerate a total radiation dose of 50 to 54 Gy given in fractions of 2 Gy once daily. At that dose, cerebral necrosis is estimated to occur in 0.04% to 0.4% of patients. The likelihood Vascular Astrocyte I Astrocyte II Oligodendrocyte of necrosis is higher among patients with primary brain tu- endothelium mors, although the incidence is still estimated to be less than 5% among patients receiving a dose of 55 Gy at 1.8 to 2 Gy Myelinated nerve fiber per fraction.34 Tolerance is affected by age; experimental Blood-brain barrier (BBB) and clinical data indicate that children younger than 2 years Paranodal Inter- nodal can tolerate approximately 10% to 20% less radiation to the brain than those older than 2 years.11,35,36 1–2 Transient edema ? Early (high dose) months demyelination Postirradiation Myelopathy Postirradiation myelopathy is the spinal cord equivalent Progressive 3–6 months Progressive of cerebral necrosis. The myelopathy manifests as sensory BBB demyelination and motor signs referable to a single cord level, signs that breakdown ? are often clinically indicative of hemisection of the cord 37,38 Capillaries (Brown-Séquard syndrome). Symptoms occur as early White matter 39 venules necrosis as 6 months after treatment (median, 20 months). The pathophysiology of spinal cord myelopathy is similar to that of cerebral necrosis, and there is continued debate concerning the relative importance of vascular endothelial Glial Telangiectasia Late delayed effects changes (particularly in the spinal cord with end-arterial infarcts atrophy blood supply to portions of the thoracic cord) and direct effects of the radiation on glial elements.28,29,40 FIGURE 33-1. The direct effects of radiation on glial cells (as- The time-dose relationship and tolerance of the spinal trocyte II and oligodendrocyte) and the indirect effects resulting from vascular changes are shown in a scheme developed by cord to irradiation have been analyzed in the context of van der Kogel. (From van der Kogel AJ. Central nervous system extraspinal tumors. The thoracic spinal cord seems to be the radiation injury in small animal models. In Gutin PH, Leibel SA, most sensitive portion; few instances of lumbar myelopa- Sheline GE [eds]: Radiation Injury to the Nervous System. New thy have been reported. Phillips and Buschke41 suggested York: Raven Press, 1991, pp 91-112.) that the thoracic spine could tolerate a dose as high as 33 The Brain and Spinal Cord 837

50 Gy given in fractions of 2 Gy once daily. Abbatucci and energy transfer (LET) radiation for pituitary or parasellar colleagues42 observed cervical cord myelopathy after doses tumors.63 The optic chiasm is often considered a dose-lim- of 55 Gy or more. The incidence of myelopathy rises rapidly iting structure when irradiation for suprasellar or adjacent with increasing doses, approaching 50% at 65 Gy.29 Many temporal lobe lesions is being planned, and special attention clinical trials of large-fraction irradiation for lung cancer is given to limiting the risk of dose-related visual complica- have confirmed that the frequency of reactions increases tions.64 Optic nerve or chiasm injury has been documented as the fraction size exceeds 2.5 Gy.43-46 Fraction size also in almost 20% of patients treated with fraction sizes of 2.5 seems to correlate inversely with the latent interval before Gy or greater (to total doses of 45 to 50 Gy), compared the development of myelopathy42; however, no experimen- with 0 of 27 patients and 0 of 500 patients treated with tal evidence exists to suggest improved tolerance to frac- fraction sizes of 1.7 to 2 Gy to the same level.16,65 Late tions less than 2 Gy.47 The Continuous Hyperfractionated, injury to the brachial plexus observed after the treatment Accelerated Radiation Therapy (CHART) regimen for lung of breast cancer similarly seems to occur largely after high- cancer was originally based on a presumed cord tolerance dose fractions.66 of a low dose per fraction, but that regimen initially was associated with an unacceptable risk of myelopathy.48,49 Tolerance decreases as the volume or length of cord irradi- Central Nervous System 41,42 ated increases. Neoplasms Other Late Effects Primary tumors of the CNS represent 2% of all cases of A unique late effect of cranial irradiation combined with cancer in the United States. Approximately 20,500 new chemotherapy, called leukoencephalopathy, has been cases are diagnosed annually.67 The incidence of these observed in children with acute lymphoblastic leukemia, tumors exceeds that of Hodgkin disease in adults, and they adults with small cell carcinoma of the lung, and patients account for 20% of all cases of childhood cancer. with primary CNS tumors.12,50,51 Leukoencephalopathy is The natural history of tumors of the brain and spinal a profoundly demyelinating, necrotizing reaction that usu- cord is relatively unusual because aggressive malignant ally occurs 4 to 12 months after combined treatment with lesions rarely disseminate beyond the neuraxis. Categoriz- methotrexate and radiation.52 Dementia and dysarthria ing CNS tumors as benign or malignant is often difficult may progress to seizures, ataxia, focal long-tract signs, or because histologically low-grade or benign neoplasms are death. Although most of these symptoms regress after sometimes associated with a poorer prognosis than certain systemic, intrathecal, or intraventricular methotrexate histologically malignant tumors. It is instead the potential therapy is discontinued, patients are left with permanent for local invasiveness or metastasis along the cerebrospinal neurologic deficits. The occurrence of leukoencephalopathy fluid (CSF) pathways that determines the “malignancy” of in patients treated with systemic and intrathecal methotrex- CNS tumors. The challenge for neurosurgeons and radia- ate without radiation and a dose-effect relationship that is tion oncologists is to remove or devitalize neoplastic tis- more apparent for methotrexate than for radiation suggest sue in the functionally vital, anatomically confined nervous that the clinical leukoencephalopathy is more significantly system. related to the methotrexate than to the radiation.7 How- ever, white matter changes interpreted as leukoencepha- Tumors Common in Children lopathy (and so coded in the NCI Common Terminology Criteria for Adverse Events) must be distinguished from Brain tumors in children account for one in five cases of the clinical syndrome previously described. childhood cancer. The incidence of primary CNS tumors, Decline in intellectual function has been related to cra- particularly the astrocytic lesions, increased in children nial irradiation of adults and particularly of children.53-57 during the first half of the 1990s; since then, the incidence Neurocognitive changes in children correlate significantly has leveled off, perhaps because of the added diagnostic with cranial irradiation, young age at diagnosis and ther sensitivity of MRI.68 Pediatric brain tumors arise most apy, and functional level at the time the irradiation was often in the posterior fossa. Posterior fossa tumors often begun.53-55,58,59 The occurrence of dose-related neuroco produce symptoms and signs of increased intracranial gnitive alterations suggests a role for continued, if cautious, pressure because the growing tumor obstructs the flow reduction in the full cranial radiation dose for patients with of CSF through the fourth ventricle. Ataxia often accom- medulloblastoma, for example, who exhibit favorable pre- panies pressure-related headaches and morning vomiting. senting signs.60 Changes in white matter volume obtained Brainstem gliomas cause cranial nerve palsies, ataxia, and by MRI segmentation constitute an objective variable that long-tract signs, including lateralizing weakness and sen- is closely tied to intellectual function and that may permit sory deficits. more detailed studies of the effects of radiation therapy and Supratentorial tumors occur in the cerebral hemispheres chemotherapy in the current generation of clinical trials.61 and the central or deep-seated areas of the thalamus, Intellectual impairment in adults has occurred primarily hypothalamus, suprasellar area, and pineal–third ventricular after intensive chemotherapy and radiation for the preven region. Symptoms are local and include lateralizing neuro- tion of metastasis from small cell carcinoma of the lung and logic deficits and seizures, which are associated with hemi- for treatment of primary CNS lymphomas.62 spheric and thalamic tumors; visual and endocrine changes; Peripheral nervous system tissue is relatively resistant to and specific oculomotor findings, which are associated with irradiation. Acute and subacute symptoms or signs occur tumors of the pineal region. infrequently; transient cranial nerve neuropathies are Primary tumors of the spinal cord account for only 5% observed only with high-dose, single-fraction, high–linear of tumors in children. Metastasis to the CNS is uncommon 838 PART 10 Central Nervous System in childhood cancer. Contiguous extradural compression Medulloblastoma within the spinal canal or at the base of the skull (e.g., in Medulloblastoma represents 20% of childhood brain neuroblastoma, rhabdomyosarcoma, primary bone tumors) tumors. Common throughout the pediatric age range, the is more common than hematogenous metastasis to the median age at diagnosis is 6 years, and 25% of tumors occur brain. in children younger than 3 years. Approximately 10% to Common primary intracranial tumors diagnosed in chil- 20% of medulloblastomas occur in adults, predominantly 25,26 dren are listed in Table 33-1. The histologic classification in the third to fifth decades. The classic presenting triad of of pediatric brain tumors has been relatively standardized by posterior fossa tumors such as medulloblastomas consists 69 the World Health Organization (WHO). The most com- of headaches and vomiting, which are signs of increased mon tumors are the astrocytic lesions (i.e., astrocytomas, intracranial pressure and ataxia. malignant gliomas, and brainstem gliomas). Medulloblasto- Medulloblastoma is by definition a posterior fossa tumor, ma, which accounts for 20% of all childhood CNS tumors, is most often arising in the cerebellar vermis, within the cere- part of the broader category of supratentorial primitive neu- bellar hemispheres (typically in adolescents and frequently roectodermal tumors (PNETs) and other embryonal CNS as the desmoplastic variant) or along the cerebellopontine tumors (i.e., cerebral neuroblastoma, ependymoblastoma, peduncle. The tumor grows to fill the fifth ventricle and 69-71 and pineoblastomas). often attaches to the brainstem. Medulloblastoma is the Biologic findings increasingly define the nature of CNS characteristic “seeding” CNS tumor and is associated with tumors in children, providing some associations with tumor subarachnoid dissemination through the CSF; metastasis aggressiveness and outcome. Medulloblastoma, for exam- may be apparent from cytologic findings or from images ple, is often associated with gene deletions in chromosomal of radiographic deposits within the intracranial subarach- region p17; expression of the EGFR gene (formerly desig- noid space (i.e., along the convexities or within the basal nated ERBB1), which encodes the epidermal growth fac- cisterns, sometimes limited to the region of the pituitary tor receptor, and NTRK3 (formerly called TRKC), which stalk) or within the spinal subarachnoid space. Subarach- encodes the neurotrophic tyrosine kinase receptor type noid disease is apparent at diagnosis in 20% to 25% of 3, are negatively and positively correlated with outcome, affected children.13,80 Extraneural disease, usually involv- 72,73 respectively. Several other molecular markers, such as ing the bone or bone marrow, is an uncommon finding at cyclin-dependent kinase 6, MYC, apurinic or apyrimidinic diagnosis or disease recurrence.81 endonuclease activity, and members of the WNT pathway, The embryonal CNS tumors represent a group of his- 74-77 also have prognostic significance. Several investigators tologically related neoplasms that occur predominantly have proposed combining information on gene expression in children (Box 33-1).69 The WHO classification defines patterns and clinical stage for stratifying risk groups in the as round, blue, small cell tumors that 78,79 embryonal tumors medulloblastoma. tend to seed the neuraxis and respond to radiation and chemotherapy, although the prognosis varies widely. Less common tumors in this category include the supratentorial TABLE 33-1. Relative Incidence of Brain Tumors in PNETs; the rare, primitive medulloepithelioma; and atypical Children or rhabdoid tumor (see “Tumors in Children Younger than 3 Years”). Pineoblastoma is classified separately (see “Pineal Tumor Type Incidence (%) Region and Germ Cell Tumors”). The term posterior fossa Infratentorial 55 PNET is synonymous with medulloblastoma.69 Medulloblastoma 25 Astrocytoma (cerebellum) 15 Brainstem glioma 10 BOX 33-1. Embryonal Central Nervous Ependymoma (fourth ventricular region) 6 System Tumors* Supratentorial 45 Medulloblastoma Astrocytoma 23 Central neuroblastoma Ependymoblastoma Malignant glioma (anaplastic astrocytoma, 6 Primitive neuroectodermal tumor (PNET)† glioblastoma multiforme) Additional Tumors with a Distinct Histology and Embryonal (primitive neuroectodermal) 4 Different Genetic Pathway of Evolution tumors Medulloepithelioma Ependymoma 3 Atypical teratoid or rhabdoid tumor Pineal Parenchymal Tumor Classified Clinically as an Craniopharyngioma 6 Embryonal Tumor Pineal region tumors 4 Pineoblastoma Oligodendroglioma 2 *In the definition proposed by Kleihues and Cavenee,69 embryonal Modified from Duffner PK, Cohen ME, Myers MH, et al. Survival of central nervous system tumors are “histologically identifiable entities children with brain tumors: SEER program, 1973-1980. Neurology [that] all develop on the background of an undifferentiated round-cell 1986;36:597-601; Childhood Brain Tumor Consortium. A study of tumor but show a variety of divergent patterns of differentiation.” childhood brain tumors based on surgical biopsies from 10 North †PNET refers to supratentorial tumors; tumors that occur in the American institutions: sample description. J Neurooncol 1988;6:9-23. posterior fossa are known as medulloblastomas. 33 The Brain and Spinal Cord 839

Medulloblastoma is thought to originate from the prim- The presence of only local residual disease is uncommon in itive external granular layer of the developing cerebellum. children with high-risk disease; more often, these children The tumor is characterized by undifferentiated small, have neuraxis dissemination. round, blue cells, often including neuroblastic rosettes and with various degrees of neuronal and glial differentiation. Radiation Therapy The desmoplastic tumors are classically identified by their Postoperative management consists of radiation therapy prominent nodularity and dense areas of reticular fibers.69 and chemotherapy.86 Infants and young children (i.e., Other histopathologic subsets include medulloblastoma with those younger than 3 years) are treated in accordance with extensive nodularity or advanced neuronal differentiation and developing investigational approaches (see “Tumors in large cell or anaplastic medulloblastoma.73 Children Younger than 3 Years”). The standard sequence of therapy for children 3 years old or older is to begin radia- Surgery and Staging tion therapy within 28 to 31 days of surgery, followed by Surgery is the initial therapy, with the goal of maximal chemotherapy.87 Although studies of preirradiation che- resection for the sake of therapy and reestablishing CSF motherapy have added significantly to our knowledge of flow. Intraoperative ventriculostomy has replaced the chemosensitivity, randomized trials have shown no benefit routine initial placement of a ventriculoperitoneal shunt; from administering chemotherapy before the radiation, and with current surgical approaches, less than 25% of children in some instances, results have suggested that this sequence require postoperative conversion of the temporary ventric- is ultimately detrimental to disease control.88-91 ulostomy to a permanent ventriculoperitoneal shunt. Gross Radiation therapy for medulloblastoma is technically total resection is limited only by substantial invasion of the demanding and biologically unforgiving. It requires the tumor into the brainstem or infiltration into the peduncle. systematic inclusion of the entire subarachnoid space (i.e., Gross total or near-total resection (i.e., more than 90% craniospinal irradiation [CSI]) followed by a boost to the of the tumor removed, with less than 1.5 cm2 of residual posterior fossa. One standard technique involving immo- disease) is achieved in almost 90% of children older than bilization of patients in the prone position is illustrated in 3 years in the United States.82 The operative mortality rate Figure 33-2. Lateral craniocervical fields encompass the should be less than 1% or 2% for children treated with cur- intracranial and upper cervical spine regions, with particu- rent approaches, and the postoperative morbidity is typical- lar attention to including the subfrontal and infratemporal ly acceptable, with a small proportion of children left with regions (at the cribriform plate in the midline subfrontal permanent new cranial nerve deficits or marked ataxia.82,83 area and along the infratemporal regions) (Fig. 33-3). The posterior fossa syndrome, marked by severe truncal The conventional dose to the neuraxis in patients ataxia, difficulties with speech and swallowing stemming given surgery and radiation is 36 Gy (delivered in 1.8-Gy from bulbar effects, and sometimes mutism, seems to be associated with extensive dissection along the lower brainstem; symptoms usually abate considerably over several BOX 33-2. Chang Staging System for months after the surgery, although late residual deficits of Medulloblastoma various degrees of severity have been seen.84 Postoperative neuraxis staging is standard in defining Primary Tumor (T)* risk groups and subsequent therapy for medulloblastoma T1 Tumor less than 3 cm in diameter and the other embryonal tumors. Immediate postopera- T2 Tumor at least 3 cm in diameter tive cranial MRI or CT (preferably done within 1 to 3 T3a Tumor larger than 3 cm in diameter with days of surgery) defines the postoperative residual effect, extension into the aqueduct of Sylvius and/ which is a more reliable prognostic factor than the opera- or into foramen of Luschka tive coding of the extent of resection.82 Neuraxis staging T3b Tumor larger than 3 cm in diameter with unequivocal extension into the brainstem begins with a contrast-enhanced spinal MRI (preferably T4 Tumor larger than 3 cm in diameter with more than 10 days after surgery to eliminate the confu- extension up past the aqueduct of Sylvius sion of findings with postoperative debris) and subsequent and/or down past the foramen magnum lumbar CSF cytologic studies; routine bone scanning and (i.e., beyond the posterior fossa) bone marrow assessments are standard only in investiga- Distant Metastasis (M) tional settings.13,80 Staging is based on the extent of re- M0 No evidence of gross subarachnoid or sidual tumor (average risk is defined as the presence of 2 hematogenous metastasis ≤1.5 cm of residual tumor) and the presence or absence M1 Microscopic tumor cells found in cerebrospinal of subarachnoid or extraneural metastasis (defined accord- fluid ing to a modification of the Chang system85) (Box 33-2). M2 Gross nodule seedings demonstrating in the Most important from the standpoint of planning therapy cerebellar, cerebral subarachnoid space, is determining whether isolated CSF cytologic evidence of or in the third or lateral ventricles metastatic disease (M1 disease), imaging evidence of intra- M3 Gross nodular seeding in spinal subarachnoid cranial disease outside the posterior fossa (M2 disease), or space imaging evidence of spinal metastasis (M3 disease) is pres- M4 Metastasis outside the cerebrospinal axis 85 ent; extranodal disease is classified as M4. Standard or *The T stage is no longer thought to be prognostically related to average-risk disease is defined as that which has been totally outcome. 2 removed or has residual tumor volume of less than 1.5 cm Modified from Chang CH, Housepian EM, Herbert C Jr. An operative with M0 disease; almost 60% to 65% of children enrolled staging system and a megavoltage radiotherapeutic technic for in U.S. studies at the turn of the century were so classified. cerebellar medulloblastoma. Radiology 1969;93:1351-1359. 840 PART 10 Central Nervous System

Posterior spine

A B FIGURE 33-2. A, Classic craniospinal irradiation configurations, with a divergent posterior spinal field matching opposed lateral craniocervical fields. The collimator rotation enabling the craniocervical fields to be matched to the divergence of the posterior spinal field is typically modified at least every 5 fractions. Field markings for medulloblastoma are determined by the cribriform plate at the orbital level (see Fig. 32-3). B, The parallel inferior border from the craniocervical fields is achieved by couch rotation, calculated to achieve a three-dimensional match in the depicted plane.

36 Gy compared with 23.4 Gy for CSI and the benchmark outcome of more than 90% tumor removal with less than 1.5 cm2 of residual disease.92 Later findings suggest that the combination of reduced-dose CSI (23.4 Gy) and chemotherapy produces disease control rates comparable with or superior to those for conventional-dose CSI alone.97 Large-scale prospective studies are documenting the efficacy of what is now considered the standard treatment for localized medulloblastoma, which is radiation (CSI dose of 23.4 Gy) followed by cisplatin-containing chemotherapy. Reports of outcome from nonrandomized, single-arm trials, further supported by neuropsychological findings from the POG-CCG study previously mentioned, suggest that cognitive function in children with average-risk disease after therapy is superior among those given CSI to 23.4 Gy instead of 36 Gy, especially in children 3 to 8.5 years old.60 The current Children’s Oncology Group (COG) trial is testing, in children between 3 and 7 years old with average- risk disease, whether the CSI dose can be further reduced from 23.4 Gy to 18 Gy. Children with metastatic disease at presentation (i.e., M1 to M3 disease) require a dose of 36 to 39 Gy to the neuraxis. The use of proton therapy for CSI has a theoretical advantage over photon therapy because of the lack of exit dose from proton therapy. Miralbell and colleagues estimated that the risk of second neoplasms FIGURE 33-3. A conventional fluoroscopic image used for could be reduced with proton therapy relative to photon simulating the lateral craniocervical fields used in craniospi- therapy.98 nal irradiation for medulloblastoma (see Fig. 33-2). The long arrow indicates the margins along the cribriform plate. The Irradiation includes a boost to the posterior fossa region, radiopaque dot and the small arrows indicate the anterior mar- classically defined as the entire infratentorial volume. The gins of the lateral orbital rims. Divergence rather than a collima- use of opposed lateral posterior fossa fields has been stan- tor angle is needed when only the subarachnoid space is to be dard, but radiation oncologists are increasingly using a targeted. three-dimensional conformal radiation therapy approach99 to spare the auditory apparatus and, when possible, the pituitary-hypothalamic region (Fig. 33-4). Proton therapy fractions once daily). The progression-free survival rate for for pediatric medulloblastoma is also being considered patients with localized disease after maximal resection and as a way of maximizing target dose coverage and sparing CSI (36 Gy) is approximately 65% at 5 to 10 years (Table normal tissues (Fig. 33-5). The cumulative dose deliv- 33-2).36,92-96 A randomized study of patients with average- ered to the posterior fossa is typically 54 to 55.8 Gy. A risk medulloblastoma conducted by the Pediatric Oncol- few institutional summaries of patterns of treatment fail- ogy Group (POG) and Children’s Cancer Group (CCG) in ure for patients with medulloblastoma and descriptions the late 1980s established the superiority of a total dose of of preliminary experience with target volumes confined 33 The Brain and Spinal Cord 841

TABLE 33-2. Survival Rates after Craniospinal Irradiation for Medulloblastoma Study and Reference Period Number of Patients Survival Rate at 5 Years (%) Survival Rate at 10 Years (%) Bloom et al, 196936 1950-1964 71 40 30 Berry et al, 198193 1958-1978 122 56 43 Hughes et al, 198694 1977-1985 53 64 42 Bloom et al, 199095 1970-1981 53 53 45 Jenkin et al, 199096 1977-1987 72 71 63 Thomas et al, 200092 1978-1991 42 (CSI = 36 Gy) 67* 67* 46 (CSI = 23 Gy) 52* 52* *Event-free survival rates at 5 and 8 years. CSI, craniospinal irradiation.

FIGURE 33-4. A, Treatment plans for a three-dimensional conformal technique demonstrating the ideal dose distribution for a six- field, non-coplanar approach targeting the entire posterior fossa. Anatomically, this approach includes the tentorium as it extends superiorly in the posterior clinoid and then inferiorly to define the anterior and superior margins of the posterior fossa. The inferior and posterior margins are defined by the calvarium and the lower margin of foramen magnum. (Continued) to the operative or tumor bed highlight investigational a multi-institutional phase II trial of this approach showed approaches that are being studied prospectively in a COG 5-year event-free survival rates of 83% for children with study.100,101 average-risk disease and 70% for children with high-risk disease.103 Chemotherapeutic Agents Although posterior fossa recurrence previously was Several categories of chemotherapeutic agents have regarded as the dominant pattern of failure, later stud- shown efficacy in medulloblastoma: alkylating agents (e.g., ies have indicated that more patients exhibit neuraxis or cyclophosphamide, lomustine), platinating agents (e.g., cis- combined local and disseminated failure. Approaches to platin, carboplatin), topoisomerase II inhibitors (e.g., eto- the management of disease recurrence are not standardized poside), and the vinca alkaloids (e.g., vincristine). The most but often include chemotherapy, further surgery for local- common adjuvant treatment for standard- or average-risk ized disease, and the judicious use of additional irradiation. disease was described by Packer and colleagues87 and con- Secondary or salvage therapies are rarely successful. sists of concurrent vincristine during irradiation and postirradiation cycles of lomustine, cisplatin, and vincristine. Astrocytoma The addition of concurrent carboplatin or etoposide and Astrocytomas are the most common glial tumors in chil- high-dose, peripheral stem cell–supporting chemotherapy dren, developing in the cerebellum, brainstem, cerebral regimens has been studied in two separate protocols for hemispheres, diencephalon (i.e., thalamus, hypothalamus, the treatment of high-risk disease.102 The first report of and third ventricular region, including the optic chiasm), 842 PART 10 Central Nervous System

FIGURE 33-4.—cont’d B, Three-dimensional conformal dose distribution demonstrates coverage of the primary tumor site, which in this case is the operative bed after surgical resection of a medulloblastoma. A multifield, non-coplanar, 6-MeV photon approach uses a gross tumor volume that outlines the operative bed, expanded by 2 cm (as modified by anatomic boundaries) to identify the clinical target volume. A 5-mm planning target volume geometrically surrounds the clinical target volume. or spinal cord.104-106 In adults, low-grade tumors occur in the cerebellum in children; if they do, they are quite primarily in the cerebral hemispheres and thalamus; how- uncommon.116 ever, malignant astrocytomas far exceed the number of his- Therapy for cerebellar astrocytomas is largely surgical. tologically benign lesions beyond the age of 20 years. Reports from institutions indicate that gross total resection The WHO grades astrocytomas as grade I (juve- is done in 65% to 90% of cases, although this rate has risen nile pilocytic astrocytoma [JPA]) or grade II (fibrillary to approximately 85% to 90% in recent series.100,112,114,115 astrocytoma or astrocytoma not otherwise specified); high- Resection is more often incomplete in patients with tumors grade lesions are deemed grade III (anaplastic [malignant] involving the cerebellar peduncles or the brainstem.112-115 astrocytoma) or grade IV (glioblastoma multiforme).69 The operative mortality rate should be less than 4%, unlike Grossly, the tumors are solid with discrete or infiltrat- the rate for patients with medulloblastomas and ependy- ing margins or cystic. Pilocytic astrocytomas are biologi- momas that may invade the adjacent brainstem. Morbidity cally nonaggressive lesions that are often cystic, occurring after radical resection of cerebellar astrocytomas usually is predominantly in the cerebellum and diencephalon in limited. children (Fig. 33-6). Recurrence of tumor after radioimaging confirmation of The natural history of astrocytomas depends on the age complete surgical removal is uncommon; in the most recent of the patient and the site of origin and histologic classifi- series, the recurrence rate is less than 10%.112,114,117,118 cation of the tumor. Pilocytic tumors rarely progress to a Prognosis most closely parallels the extent of resection; more malignant histology; however, fibrillary astrocytomas multivariate analyses have also identified solid (rather than progress to a more malignant histology in 30% to 50% of cystic) lesions, transitional lesions (versus those confined cases.107-109 The proliferative index, assessed by staining to the cerebellum), and lesions showing a fibrillary or dif- for Ki-67 with the murine monoclonal antibody MIB1, has fuse histology (versus JPA) as potentially significant prog- been reported to influence prognosis.110 nostic factors.112,114,115,119 A direct correlation between the volume of residual tumor and outcome has also been Cerebellar Astrocytomas reported.112 Cerebellar astrocytomas involve the vermis or hemispheres Indications for radiation therapy are limited in patients in 75% of patients; in 25% of children, tumors manifest as with cerebellar astrocytomas. Although radiation therapy transitional neoplasms that involve the cerebellum and the has been suggested to be effective for incompletely resected cerebellar peduncles or adjacent brainstem.111-114 Cerebel- solid lesions, enthusiasm is generally limited for postopera- lar astrocytomas are classically cystic lesions; the textbook tive irradiation for any cerebellar astrocytomas.117,120 A appearance consisting of a large cyst associated with a mural small subset of tumors, often those that primarily involve nodule occurs in about one fourth of cases. Grossly, most the cerebellar peduncle, will recur after initial surgery; our tumors are mixed cystic and solid lesions or apparently approach to the treatment of such tumors is to use irradia- solid tumors with small cystic components. Histologically, tion only after tumor progression after a second attempt 75% of astrocytomas in this location are JPAs.111,112,114,115 at resection. MRI (T2-weighted or fluid attenuation inver- Debate continues about whether high-grade gliomas occur sion recovery [FLAIR] sequences) and MRI/CT fusion are 33 The Brain and Spinal Cord 843

A D

B E

C F

FIGURE 33-5. Medulloblastoma treatment plans. Posterior fossa axial slices are shown for three-dimensional conformal radiation therapy (3D CRT) (A), intensity-modulated radiation therapy (IMRT) (B), and protons (C). Isodose lines are designated for 3D CRT and IMRT: blue, 30.6 Gy; red, 15 Gy; yellow, 10 Gy; purple, 5 Gy. Isodose lines are designated for protons: red, 30.6 Gy; green, 15 Gy. For craniospinal irradiation, axial slices are shown for 3D CRT electrons (D), 3D CRT photons (E), and protons (F). Isodose lines are designated for 3D CRT and IMRT: blue, 23.4 Gy; red, 20 Gy; yellow, 10 Gy; purple, 5 Gy. Isodose lines are designated for protons: aqua/green, 23.4 Gy; light blue, 20 Gy; darker blue, 10 Gy; indigo, 5 Gy. (Continued) useful for target delineation, and a margin of 0.5 to 1 cm gliomas are histologically malignant (i.e., WHO grade seems adequate. The total dose is 50 to 54 Gy in conven- III to IV, anaplastic astrocytoma or glioblastoma).122-124 tional fractionation. Biopsy alone has been the standard surgery for most thalamic astrocytomas because knowledge of the tumor his- Supratentorial Astrocytomas tology is important in guiding therapy.95,125,126 However, Supratentorial astrocytomas include diencephalic or Bernstein and colleagues127 reported that major resections central lesions and lesions of the cerebral hemispheres. In were performed in up to 33% of patients with circum- children, total resection can be achieved for up to 90% scribed thalamic tumors, and the incidence of major opera- of astrocytomas in the cerebral hemisphere.104,108,121,122 tive morbidity was only 7%. Further experience confirmed Diffuse cerebral lesions, including those labeled gliomatosis that limited or aggressive resection might each have a role cerebri, can arise in several cerebral lobes with an often low- in the treatment of low-grade thalamic gliomas.128 grade histologic appearance (i.e., fibrillary, WHO grade II) The role of radiation therapy in childhood astrocytomas but clinically malignant potential. Similarly, thalamic tumors is evolving. It is thought to depend on tumor histology and represent a mix of circumscribed JPAs and diffuse, some- site, age of the patient, and presence or absence of clinical times bithalamic fibrillary tumors; up to 40% of thalamic signs. Several reviews129 have shown that radiation therapy 844 PART 10 Central Nervous System

G H

FIGURE 33-5.—cont’d Sagittal slices are shown for 3D CRT electrons (G), 3D CRT photons (H), and protons (I). Isodose lines are designated for 3D CRT and IMRT: green, 54 Gy; orange, 35 Gy; blue, 20 Gy; red, 15 Gy; yellow, 10 Gy; purple, 5 Gy. Isodose lines are designated for protons: orange, 54 Gy; green, 35 Gy; light blue, 20 Gy; aqua/green, 15 Gy; violet, 10 Gy; dark purple, 5 Gy. (From Lee CT, Bilton SD, Famiglietti RM, et al. Treatment planning with protons for pediatric retinoblastoma, medulloblastoma, and pelvic sarcoma: how do protons compare with other conformal techniques? Int J Radiat Oncol Biol Phys 2005;63:362-372.)

is effective in terms of tumor response, prolonged progres- completely) resected tumors.131-134 A trial by the Euro- sion-free interval, and long-term disease control. A joint pean Organisation for Research and Treatment of Cancer CCG-POG prospective trial of postoperative irradiation to evaluate postoperative observation versus irradiation for for incompletely resected astrocytomas was abandoned WHO grade I and II hemispheric lesions in adults showed in the early 1990s because the proportion of children in no convincing advantage for early irradiation or for low- whom gross total or near-total resection could be achieved dose (45 Gy) versus high-dose (60 Gy) therapy.135 had increased substantially and because the wisdom of The effect of radiation therapy is well documented performing irradiation before symptomatic or radioimag- for central supratentorial lesions. Neurologic improve- ing-documented progression of disease precluded uniform ment has been observed in 80% of patients after irradia- acceptance of irradiation as a randomized option. Prelimi- tion for diencephalic tumors. Objective reduction in the nary experience in that trial suggested that observation size of supratentorial astrocytomas has been confirmed in might be appropriate after substantial resection because approximately one half of the patients studied by prospec- disease seems to progress in only a minority of such cases tive CT assessment. Long-term survival of patients with 3 to 5 years after surgery126 (J. Wisoff, personal communica- biopsy-proven diencephalic lesions usually has required tion, 2001). For some patients with central low-grade glio- irradiation to therapeutic dose levels.125 Limited-volume, mas (e.g., tectal plate or midbrain; less often, hypothalamic fractionated irradiation has been reported to control most or thalamic), cautious observation can identify indolent, such lesions, with toxicity apparently being more limited in stable tumors that may not require immediate intervention the pediatric age group.136 if the child is neurologically stable.106,130 However, this approach implies a commitment to intervene with sur- Optic Pathway Glioma gery or radiation, or both, when clinical or radiographic Tumors of the optic nerves and chiasm are low-grade evidence of tumor progression appears. astrocytomas that occur predominantly in young children, The systematic use of postoperative irradiation has also often younger than 3 to 5 years. In 25% to 30% of cases, the been questioned in adults.125,131-133 Adults more often tumors are associated with neurofibromatosis.137 Tumors have sizable peripheral cerebral hemispheric tumors. Out- of the optic nerve often behave as hamartomatous lesions, come at 5 and 10 years seems to be improved for patients although up to 50% may extend proximally toward the given radiation therapy for incompletely (as opposed to chiasm. Chiasmatic gliomas manifest with visual and often 33 The Brain and Spinal Cord 845

of patients receiving radiation therapy, and neuroimaging shows a reduction in the size of the lesion in these patients.145,146 Survival rates after irradiation for chiasmatic gliomas reportedly range between 75% and 100% in large series of patients followed for 5 to 10 years.138,143-145 Controversy regarding management appropriately surrounds the often-benign clinical course of the disease in h.t. relation to the potential for treatment-related morbidity with surgery or irradiation. Exophytic JPAs arising in the chiasmatic or hypothalamic region are amenable to incomplete surgical resection147; the potential advantage of partial resection lies in delaying more definitive irradiation. Particularly among younger children, a balance between disease progression and the potential risk of adverse endocrine and intellectual effects from irradiation has favored a conservative approach.138,143 Packer and colleagues148-150 described their experience with chemotherapy, particularly the carboplatin and vincristine regimen. In most young children, previously progressive or symptomatic disease A was stabilized with chemotherapy; objective responses were documented in up to 50% of cases, and the median time to disease progression was in excess of 3 years. Initial management with relatively toxicity-free chemotherapy approaches is logical for children with progressive symptomatic chiasmatic or hypothalamic tumors, certainly for those younger than 3 to 5 years and possibly for all children before puberty. The frequency and durability of disease control are not clear for children for whom irradiation is begun when their disease progresses during or after chemotherapy. Outside the investigational setting, the treatment h.t. of choice remains primary irradiation for children between 5 and 10 years old because long-term disease control can be achieved in up to 90% of these cases.138,143-145 Radiation Therapy Technique Radiation therapy for low-grade astrocytomas uses local target volumes to encompass the typically discrete primary tumor quite narrowly.136,151,152 Uncommonly, low-grade B tumors diffusely infiltrate one or more cerebral lobes, requiring more wide-field or total cranial irradiation (Fig. 33-7). FIGURE 33-6. Midline sagittal T1-weighted MRI scans depict a hypothalamic–optic chiasmatic juvenile pilocytic astrocytoma Low-grade, often pilocytic, astrocytomas are associated that filled much of the sella and suprasellar region at diagnosis with multifocal disease or subarachnoid involvement at (A) and at 2 years after completion of radiation therapy (54 Gy diagnosis (or, less commonly, at recurrence) in 15% to 20% delivered in 30 fractions in a multifield, three-dimensional con- of patients with childhood astrocytomas.153-155 Multifocal formal approach) (B). The involution is typical of juvenile pilo- or metastatic disease is rarely confined to the spinal sub- cytic astrocytomas at this site, including the fairly long interval arachnoid region; multiple disease sites are almost always to reduction. h.t., hypothalamic tumor. apparent on cranial MRI studies, obviating any requirement for spinal imaging or staging in patients with this tumor. Specific histologic subtypes of tumors (i.e., pleomorphic with endocrine signs; it is often difficult to distinguish chi- xanthoastrocytoma and meningocerebral astrocytoma) can asmatic from hypothalamic tumors. In collected reports of directly infiltrate the contiguous meninges, but this seems 500 cases, 25% of patients with chiasmatic involvement to be associated with an indolent, benign course after sur- died of tumor-related events.138,139 In an often-quoted gery alone.156-158 series of 36 children, Hoyt and Baghdassarian140 indicated Establishing the target volume requires detailed MRI. that chiasmatic tumors were indolent and nonthreatening. A 1-cm margin beyond the abnormalities shown on T1- or However, follow-up at a median of 20 years documented a T2-weighted images seems adequate on the basis of early 25% tumor-related mortality rate in a population weighted follow-up data from patients for whom three-dimensionally toward the apparently more benign tumor associated with planned and delivered radiation techniques were neurofibromatosis.141,142 used.136,151,152,159 A phase II trial by the COG is using MRI/ The efficacy of radiation therapy for tumors of the CT fusion for target delineation, a 0.5-cm margin for clinical optic chiasm with or without hypothalamic involvement target volume, and a 0.3- to 0.5-cm margin for planning has been documented in numerous series.138,143,144 Visual target volume. The planning target volume is treated by us- acuity or visual fields are improved in approximately 25% ing a conformal technique to 54 Gy in 30 fractions. 846 PART 10 Central Nervous System

a role for radiation therapy was clearly indicated for most patients with astrocytomas, with 5- and 10-year survival rates of 68% and 39%, respectively, after at least 53 Gy. Patients treated with less than 53 Gy or no irradiation had a 5-year survival rate of 47% (both groups) and 10-year survival rates of 27% and 11%, respectively. The long-term potential for radiation therapy to control disease specifically in children with hemispheric astro cytomas has been documented in reviews by Bloom95 and Jenkin96 and their colleagues, who independently reported that the survival rate reached a plateau between 5 and 15 years after postoperative irradiation. The benefit of add- ing irradiation after incomplete resection for children with supratentorial hemispheric tumors was made apparent in the Pittsburgh series reported by Pollack and colleagues.161 Site-specific survival rates suggest that patients with hypothalamic and chiasmatic astrocytomas enjoy longer survival times than do patients with most other types of supratentorial astrocytomas. Bloom and colleagues95 cited 5-year survival rates of 60% to 70% for patients with thalamic or hypothalamic tumors after primary irradiation. Albright and colleagues125 reported a mean survival duration of 5.1 years after radiation therapy for patients with FIGURE 33-7. MRI scan demonstrates the diffuse, multilobar diencephalic gliomas compared with a mean survival time involvement of a biopsy-proven fibrillary astrocytoma. Although of less than 1 year for patients who did not undergo irradia- predominantly in the left temporal lobe, the lesion also extends tion. The virtual lack of tumor progression beyond 5 years throughout the left parietal, right parietal, and mesencephalic after radiation therapy further suggests a role for radiation regions. Objective improvement in response to irradiation is therapy as a curative modality in the primary management often transitory. This type of diffuse cerebral tumor, sometimes of hypothalamic or chiasmatic tumors.145,162 called gliomatosis cerebri, is a supratentorial lesion analogous to the more typical diffusely infiltrating pontine glioma. Brainstem Gliomas Tumors of the brainstem most often involve the pons A dose-response relationship for astrocytoma has (70%), followed by the medulla (15% to 20%) and the not been well defined. In most series, a radiation dose of midbrain (10% to 15%).163,164 Brainstem gliomas are char- approximately 50 Gy delivered in 1.8-Gy, once-daily frac- acterized as diffusely infiltrating or focal, depending on the tions has proved adequate.129,138 However, in statistical pattern of involvement; focal tumors are well demarcated, analyses of rather sizable series of patients, Albright125 and confined to one anatomic segment of the brainstem, and Laws108 and their associates observed an apparent bene- occupy less than one half of the brainstem.165 Most focal fit conferred by doses above 40 Gy. Justifying dose levels tumors are JPAs histologically; they occur more often in above 54 to 55 Gy in children is difficult in view of the the midbrain or medulla but can arise in the pons.163,166- increase in risk at radiation levels approaching cerebral tol- 168 Pontine gliomas are largely diffuse (Fig. 33-8).164,165 erance. For adults with supratentorial astrocytomas, Shaw Dorsally exophytic lesions are relatively benign tumors and colleagues,134 having shown a dose-response relation- that usually arise at the pontomedullary junction or from ship indicating improved survival after at least 53 Gy, rec- the medulla; histologically, they are JPAs.166-169 In the past, ommend levels of 55 to 60 Gy. However, two randomized brainstem tumors have been treated conservatively, with studies of adults with low-grade gliomas testing 50.4 Gy histologic confirmation limited to focal or cystic lesions for versus 64.8 Gy or 45 Gy versus 59.6 Gy showed no signifi- which knowledge of the histology or drainage might facili- cant difference in outcome between the lower doses and tate treatment.163,164 Findings from biopsied cases indicate the higher doses.135,160 that malignant glioma exists in 40% of patients at diagno- sis164; the remaining tumors are largely fibrillary astrocy- Prognosis tomas.163,170 At autopsy, however, malignant gliomas are Laws and colleagues108 reviewed the Mayo Clinic expe- found in more than 80% of cases.164,170 rience with 461 patients of all ages with supratentorial Radiation therapy is primarily used for diffusely infil- astrocytomas treated between 1915 and 1975. The most trating brainstem gliomas. Substantial neurologic improve- important correlates of 10- and 15-year survival were ment occurs in 60% to 70% of patients with diffusely age (86% long-term survival for children and adolescents infiltrating pontine gliomas, although long-term survival is younger than 20 years versus 33% for those 20 years old limited to 10% to 20% of patients with these gliomas.164 or older) and the degree of tumor involvement, reflect- Dorsally exophytic tumors are usually identified during ed by the neurologic deficit and the extent of resection surgery for a tumor that otherwise fills much of the fourth (37% survival rate for patients after biopsy or limited re- ventricle. The lesions in more than one half of the patients moval compared with 55% for those after subtotal or total remain indolent after incomplete resection; the 40% or resection). In Shaw’s review134 of the same Mayo experi- more of tumors that do progress are most often controlled ence with adult patients with supratentorial astrocytomas, with radiation (Fig. 33-9).167,168 33 The Brain and Spinal Cord 847

A B

FIGURE 33-8. A, Sagittal MRI scan depicts a diffusely infiltrating brainstem glioma involving much of the pons and extending toward the midbrain and the medulla. B, A typical two-dimensional treatment approach is shown. Use of three-dimensional conformal techniques also may be advantageous.

Radiation Therapy Technique Brainstem gliomas tend to infiltrate longitudinally, usually extending beyond the pons to involve the midbrain or medulla and extending peripherally into the cerebellum or medial temporal lobes (see Fig. 33-8). Irradiation is limited to the brainstem and immediately adjacent neural tracts. The infiltrative nature of these tumors can be better appreciated on T2-weighted MRI studies. Clinical target volumes are defined with 2-cm margins along the tracts of infiltration (i.e., into the midbrain or medulla, through the peduncles, and into the adjacent cerebellum).171 Given the poor overall outcome for patients with pontine lesions, many centers continue to use two-dimensional, opposed lateral fields in this setting, providing a margin around the known tumor extent. The standard dose for brainstem gliomas is 54 to 56 Gy, given in a conventional fractionation scheme. In a series of trials, hyperfractionated doses given to total doses of 66, 70 to 72, and 76 to 78 Gy indicated no real benefit from the higher doses.172-174 Most long-term survivors have been adults or children with focal brainstem lesions often located outside the pons (i.e., in the midbrain or medulla).172 The decade-long experience with hyperfractionated therapy for these tumors has encouraged neurosurgeons, FIGURE 33-9. Sagittal MRI scan shows a dorsally exophytic radiation oncologists, and pediatric oncologists to enroll brainstem glioma. This discrete, briskly enhancing tumor fills patients in multi-institutional protocols. In a randomized much of the fourth ventricle, attaching along the posterior aspect trial conducted by the POG, conventional fractionation to of the brainstem at the pontomedullary junction. Invasion may 54 Gy was documented to be as effective as 70 Gy given in not be apparent on imaging, and at surgery, lesions often are twice-daily 1.2-Gy fractions.175 This trial also documented found to blend imperceptibly with the underlying brainstem. the apparent negative effect of irradiation in combination with concurrent cisplatin in this setting.175,176

Prognosis now understood that these tumors are usually JPAs168; sur- The 5-year overall survival rates are 10% to 20% for vival after surgery, with or without radiation, approaches children with brainstem gliomas and 20% to 30% for 90% in this selected group.166-168 Intrinsic pontine tumors, adults.172-174 Median survival after irradiation is only 8 constituting most of the classically identified brainstem months for patients with classic intrinsic tumors in the gliomas, are usually associated with multiple cranial nerve pons. Hoffman and colleagues166,167 initially identified as palsies, diffusely hypodense lesions on CT or MRI stud- highly favorable a subset of children in whom exophytic ies, and a high-grade histologic appearance indicative of an tumors extended into the fourth ventricles. However, it is infiltrating, aggressive neoplasm. Long-term survival rates 848 PART 10 Central Nervous System

for the latter group have consistently been less than 15%. Ependymomas arise from the ependymal cells lining Less common intrinsic tumors that are focal and confined the ventricular system and the central spinal canal. The to a segment of the pons or midbrain are associated with tumor characteristically includes ependymal rosettes and slightly better long-term survival rates.165 perivascular pseudorosettes. Ependymomas are differen- Chemotherapy regimens have little effect on brainstem tiated or low-grade lesions; variants include cellular, clear gliomas. Adjuvant therapy did not affect outcome in the cell, tanycytic, and papillary types.69 In one series, 15% to only published randomized trial.177 30% of ependymomas were classified as anaplastic tumors (i.e., WHO grade III), marked by increased cellularity and Ependymomas mitotic rate. The tumors usually are well demarcated but Intracranial ependymomas occur predominantly in young can show parenchymal invasion.69,115 Although ependy- children, but they are also seen in older children and adults. momas were regarded by leading neuropathologists in the Ependymomas also occur as primary spinal cord neo- 1980s as being of little clinical significance, an association plasms, but this presentation is more common in adults. has since been recognized between an anaplastic histology Posterior fossa ependymomas typically manifest with the and a less favorable outcome.186-192 Ependymoblastomas same symptom triad as other posterior fossa tumors in chil- are rare embryonal tumors and are discussed with the other dren—headache, vomiting, and ataxia. Associated symp- supratentorial PNETs. toms include torticollis and lower cranial nerve dysfunction. Surgery is the initial intervention for most intracranial The median age of children with these tumors is 4 years; ependymomas. Even extensive tumors of the fourth ventri- however, almost 30% of infant brain tumors (i.e., those in cle can often be grossly resected; total resection of tumors children younger than 3 years) are ependymomas.178,179 extending into or arising in the cerebellopontine angle has Infratentorial lesions in younger children may occur more been achieved.180,190 The advantage of near-total or total often in girls than in boys. Supratentorial lesions more often resection is apparent, with considerable improvement in manifest with seizures or focal neurologic abnormalities disease control in children in whom complete resection and are most common in adolescents and adults. was achieved.189-191 The price of this control is a significant Tumors of the posterior fossa occur predominantly in incidence of new postoperative neurologic deficits, includ- children; these tumors show a predilection for the fourth ing the posterior fossa syndrome and major new bulbar ventricle in young children and often involve or origi- signs (e.g., palsies of cranial nerves IX and X resulting in nate within the region of the foramen of Luschka (Fig. difficulty with speech and airway protection), sometimes 33-10).179,180 Direct extension through the foramen mag- requiring postoperative tracheostomy or the placement num into the upper cervical spinal canal is a characteristic of a feeding tube.88 Institutional experience shows that pattern of growth. Ependymomas disseminate into the sub- gross total resection can be accomplished in 50% to 60% of arachnoid space in approximately 12% of children.95,181-184 patients with infratentorial ependymoma; however, a total However, this manifestation is more often seen at diagno- resection rate as high as 90% has been reported.190 sis or as a primary site of recurrence in very young chil- Postoperative irradiation is essentially routine for patients dren and may be associated with an anaplastic histology; with ependymomas, almost regardless of age, because ad- in older children, this pattern is more common after local juvant radiation has been shown to increase disease control recurrence than earlier in the disease.183,184 Supratentorial rates in several major series.190,192 However, a single expe- tumors often arise in regions adjacent to the ventricular rience suggests that disease could be controlled in a certain system, and growth occurs largely through intraparenchy- subgroup of children with supratentorial ependymomas mal extension.185 who are treated with total resection alone; in this ongoing

A B

FIGURE 33-10. A typical ependymoma involving the fourth ventricle. A, Axial CT scan reveals a lesion that fills much of the fourth ventricle and extends through the right foramen of Luschka. B, Sagittal MRI scan shows the tumor extending caudally along the medulla and below the foramen magnum into the upper cervical spinal canal. 33 The Brain and Spinal Cord 849

COG prospective trial, only a small proportion of patients potential value of dose escalation is being explored in trials with such ependymomas is being treated with observation testing cumulative doses approaching 60 Gy.159 only.193 For all lesions of the fourth ventricle or cerebellopontine angle, high rates of tumor control have been Pineal Region and Germ Cell Tumors achieved only for patients treated with local postoperative A unique group of tumors occur in the pineal region: radiation and radical resection.189-191,194 Long-term disease germ cell tumors, pineal parenchymal tumors, and glial control was accomplished in approximately 70% of such and ependymal tumors (Table 33-3).69,116,201 Intracranial children, compared with 20% to 30% of those treated with germ cell tumors occur predominantly in the pineal region, incomplete resection and similar radiation. although they can also arise in the anterior third ventricu- Ependymomas respond to a variety of chemotherapeu- lar (suprasellar) region and, less commonly, in thalamic or tic agents, including cyclophosphamide and vincristine, deep cerebral hemispheric areas. cisplatin and carboplatin, and etoposide. A randomized This unique group of tumors manifests with clinically and trial of adjuvant lomustine, vincristine, and prednisone histologically linked syndromes. Pineal region germinomas conducted by the CCG showed no benefit from this occur predominantly in adolescent boys and are associated chemotherapy.195 The initial experience in infants with with multifocal or metastatic disease in the anterior ependymomas indicated an inverse relationship between third ventricular region in 10% to 40% of cases.202-206 the duration of chemotherapy (equivalent to the time of Suprasellar (anterior third ventricular) germ cell tumors initiating radiation therapy) and disease control, encour- occur more broadly throughout the pediatric age group aging current institutional and cooperative group trials to and about equally in girls and boys; these tumors are often limit the use of chemotherapy for this tumor; only for the germinomas but also include other germ cell types. Pineal youngest children (<18 months old) has a brief (3 to 4 parenchymal tumors in children are overwhelmingly month) course of chemotherapy been used before radiation malignant pineoblastomas, with a considerable proportion (or second resection and radiation if measurable disease is occurring in infants and children younger than 2 to 3 years apparent at the completion of chemotherapy).194 For as highly aggressive, widely metastasizing, small, round, infants, interest in chemotherapy dose escalation contin- blue cell tumors. Pineocytomas are decidedly less common ues because of positive preliminary results from the second in children occurring more often in adults.69,207-209 POG study of infant brain tumors, results that suggested Regardless of histology, pineal region tumors typically that dose-intensified chemotherapy produced better dis- manifest with Parinaud’s syndrome, consisting of near-light ease control rates than did standard chemotherapy.196 In dissociation (i.e., pupillary responses are diminished to older children, the usefulness of limited-duration chemo- responsiveness in facilitating second operative resections before radiation is being examined in a COG trial. Most clinical experience with ependymomas is based on treatment volumes for infratentorial lesions that include the entire posterior fossa and often the upper cervical spinal canal to one to two vertebral levels below the caudal tumor extent. Because recurrent disease is almost always localized to the initial tumor bed and often the site of attachment and because minimal residual disease is left adjacent to the brainstem or peduncle, current prospective investigations are targeting only the local tumor bed (with margins as tight as 1 cm) using three-dimensional conformal or intensity-modulated radiation therapy (Fig. 33-11).159 However, findings during an as yet insufficient follow-up period indicate that the limited treatment volumes should be used with caution outside a controlled setting. For supratentorial lesions, a 1- to 2-cm margin around the initial tumor extent is indicated, with potential modifications when the normal brain is reconfigured by tumor resection. Treatment volumes for differentiated and anaplastic ependymomas usually are not different.197,198 Pre- vious interest in CSI for anaplastic infratentorial lesions has not continued into contemporary protocols or treatment standards, except in cases in which neuraxis involvement is documented at diagnosis. Classically, radiation doses of 45 to 54 Gy have been used.182,189,192 However, doses of 54 to 55.8 Gy have been reported to achieve superior results and are indicated even FIGURE 33-11. The treatment plan shows the isodose volumes 199,200 for a three-dimensional conformal approach, using multiple after gross total resection. Despite apparently encour- non-coplanar 6-MeV photon fields to encompass the tumor aging results from a small POG trial of hyperfractionated within the lower aspect of the posterior fossa and the upper- radiation (70 Gy given twice daily in 1.2-Gy fractions) for most spinal canal. The target volume in contemporary protocols patients with posterior fossa ependymomas, no substanti- more often addresses the tumor bed or surgical bed than the ated role has been established for altered fractionation. The full posterior fossa. 850 PART 10 Central Nervous System

TABLE 33-3. Histologic Classification and Relative An open craniotomy is usually required to biopsy su- 213 Incidence of Pineal Region Tumors prasellar tumors. Biopsy has been possible during third ventriculoscopy, often at the time of ventriculoperitoneal Histologic Type* Relative Frequency† shunt insertion or third ventriculostomy. Pineal region Tumors of Germ Cell Origin tumors can be approached stereotactically, but their chal- lenging location adjacent to the major draining intracra- Germinoma 60% to 80% nial veins has led many neurosurgeons to prefer an open Teratoma (differentiated, immature, 65% approach. Tumor resection is feasible in the pineal region, malignant) although done in only a few instances because of the high Endodermal sinus tumor 18% rate of potential complications. Although insertion of a ventriculoperitoneal shunt is often necessary, the notable Embryonal carcinoma 7% incidence of shunt metastasis (with tumor arising through- Choriocarcinoma 5% out the peritoneal cavity),212,213 although low in absolute Tumors of Pineal Parenchymal 14% terms, is relatively more common for this group of tumors Origin than for any other intracranial lesions. Performance of a third ventriculostomy, rather than placement of a ventricu- Pineoblastoma loperitoneal shunt, is encouraged for accomplishing inter- Pineocytoma nal decompression. Tumors of Glial Origin 15% Radiation Therapy Technique Astrocytoma Radiation therapy is highly curative in patients with germinomas in the pineal region, anterior third ventricle, Malignant glioma or multiple midline locations.205,211,214-220 Considerable Ganglioglioma (ganglioneuroma) debate, however, surrounds the appropriate radiation volume for these tumors.206 Given the recognized incidence Ependymoma of intraventricular and subarachnoid dissemination, craniospinal (or less often, full cranial) irradiation has been *Modified from Burger PC, Scheithauer BW, Vogel FS (eds). Surgical Pathology of the Nervous System and its Coverings, 3rd ed. New recommended for pineal region and suprasellar germino- 205,214,221 York: Churchill Livingstone, 1991. mas. The outcome in many series indicates that †Modified from Burger PC, Scheithauer BW, Vogel FS (eds). Surgical disease is controlled in up to 95% of patients treated with Pathology of the Nervous System and its Coverings, 3rd ed. New full neuraxis radiation as a component of their therapy York: Churchill Livingstone, 1991; Kleihues P, Cavenee WK (eds). (Table 33-4).203,206,215,218-225 Local irradiation, variably Tumours of the Nervous System. Lyon, France: IARC Press, 2000. defined as encompassing the primary tumor site with a margin or the entire third ventricular region, has been associated with long-term disease-free survival rates of light but are maintained with attempted accommodation), more than 80% (see Table 33-4).203,206,215,218-225 The inci- lack of upward gaze, and incomplete ability to accommo- dence of neuraxis dissemination varies from less than 10% date. Obstructive hydrocephalus often occurs when the to 35%.206,214,219,221,226 No consensus has been reached tumors are near the sylvian aqueduct. Suprasellar lesions regarding the appropriate radiation volume because cure can also manifest as hydrocephalus when the growing rates are relatively high for local and full-neuraxis irradia- tumor obstructs the foramen of Monro; diabetes insipidus tion, and the toxic effects associated with the relatively and visual field deficits are often apparent. low doses necessary for this disease are exquisitely age Although sometimes debated in the past, it is important related.218,220 to histologically establish the diagnosis of pineal region Evidence is increasing that 45 Gy delivered to the pri- tumors. Suprasellar lesions have usually been biopsied be- mary tumor site is adequate to ensure virtually uniform cause of their relative accessibility and the broad differential control of pineal region and suprasellar germinomas. diagnosis that includes astrocytomas of the hypothalamic Wider-field components (craniospinal or, as done at some or chiasmatic region and craniopharyngioma. Germ cell institutions, full cranial irradiation) can be limited to doses tumors can occasionally be definitively diagnosed on the of 25 Gy.216,218,219 CSI doses as low as 18 to 19.8 Gy in basis of chemical markers. Any significant elevation of the children with no detectable leptomeningeal metastasis alpha-fetoprotein level in the serum or CSF is diagnostic of from germinomas at diagnosis have been reported to be a malignant germ cell histiotype (typically embryonal car- effective in preventing leptomeningeal recurrence.227 cinoma, yolk sac tumor, or teratocarcinoma), and marked Considerable interest has been expressed in chemo- elevation of β-human chorionic gonadotrophin level (β- therapy for intracranial germinomas. Although a high rate HCG; typically more than 100 to 1000 mIU/mL) is diag- of response has been documented, durable control of dis- nostic of a choriocarcinoma or malignant mixed germ cell ease has been achieved in only about one half of the pa- tumor (with choriocarcinomatous elements).210,211 Mod- tients treated with chemotherapy only, including some of erate elevation of the β-HCG level (50 to 100 mIU/mL) is the more rigorous treatment regimens that carry a 5% to compatible with a germinoma, but cytologic or histologic 10% rate of toxicity-related death.222,224,225,228 CSI is often evidence is required to confirm the diagnosis.212 After a a successful salvage treatment for germinomas that recur diagnosis is established, neuraxis staging (e.g., spinal MRI, after such therapy.218 lumbar CSF cytology) is mandatory for any of the germ Early studies conducted by Allen and colleagues223,229 cell tumors and pineoblastoma. established that disease control was excellent after combined 33 The Brain and Spinal Cord 851

TABLE 33-4. Therapy and Outcome in Intracranial Germinomas Radiation Therapy Therapy CSI Local (Gy) Chemotherapy 5-Year NED Rate (%) Study and Reference RT only + 50-55 85 Shibamoto et al, 1988221 + 45-55 86 Rich et al, 1985203 + 45-59 97 Hardenbergh et al, 1997219* + 45-50 91 Bamberg et al, 1999222 CrI, VtI 50-54 91 Wolden et al, 1995206 0 40-50 87 Calaminus et al, 1994220 0 44 90 Haddock et al, 1997215 + 50 100 Merchant et al, 2000218 Chemotherapy and 0 30 Cyclophosphamide 91 Allen et al, 1987223 reduced RT 0 40 VP-16/CBCDA; VP-16/ 96† Bouffet et al, 1999224 ifosfamide 0 30 (CBCDA or CDDP); 92 Calaminus et al, 1994220 vinblastine; bleomycin Chemotherapy only 0 0 CDDP, VP-16, bleomycin 42 Balmaceda et al, 1996225 *Includes cases from Rich et al, 1985203 (CSI or CrI, dose of 15 to 44 Gy). †Event-free survival at 3 years. CBCDA, carboplatin; CDDP, cisplatin; CrI, full cranial irradiation; CSI, craniospinal irradiation (dose levels 24 to 30+ Gy for localized tumors); NED, no evidence of disease; RT, radiation therapy; VP-16, etoposide; VtI, full ventricular volume irradiation (typically to doses of 20 to 36 Gy), +, used; 0, not used.

chemotherapy (with cyclophosphamide or carboplatin) agents or nitrosoureas in conjunction with vincristine).232 followed by limited-volume, reduced-dose radiation. Strat- Disease has been controlled in up to 60% of patients, egies incorporating moderately aggressive chemotherapy compared with the 0% to 10% survival rates for infants and local radiation (to 30 to 36 Gy for patients showing treated with prolonged chemotherapy with or without a good partial or complete response, as shown on imag- radiation.178 ing studies and by normalization of any elevated markers) or limited-dose CSI (for those with disease dissemination) Craniopharyngioma result in disease control rates exceeding 90%, with limited Craniopharyngiomas are histologically benign tumors de- toxic effects.220,222,224,228 Whether such an approach of- rived from squamous cell rests in the region of the pituitary fers any advantage over radiation therapy alone is unclear; stalk. More than one half of these tumors occur in patients prospective trials of this approach are being conducted in younger than 18 years, and craniopharyngiomas represent Europe and by the COG. It is difficult to improve on the 3% to 9% of the intracranial tumors of children. The tumor disease control rates produced by primary irradiation, al- arises as an extra-axial lesion, typically a densely calcified, though prepubertal patients and those who require CSI solid tumor in the suprasellar region with large cystic com- before completion of skeletal growth may benefit from a ponents (Fig. 33-12). These histologically classic adaman- combined-modality approach. tine tumors occur in children and adults.233 Less common In patients with malignant germ cell tumors (i.e., em- are the noncalcified, squamous papillary type, which usu- bryonal or yolk sac carcinoma, choriocarcinoma), treat- ally occur in adults.69,115 ment with radiation therapy alone has rarely controlled Total surgical resection is usually curative in patients with disease in more than 25% to 40% of cases.213,220 Combined craniopharyngiomas.233-240 The decision about whether to radiation and chemotherapy seems to be beneficial in such use surgery or radiation therapy depends on the site of cases, but the chemotherapy must be relatively aggressive, origin and growth of the tumor, which often damages the and the radiation dose to the craniospinal axis is usually hypothalamus or optic chiasm and sometimes damages the 24 to 36 Gy, followed by a boost dose to the tumor bed adjacent major vessels. Although encapsulated, the lesion of 9 to 25 Gy, depending on the response to prior chemo- usually adheres to these vital structures, even though it therapy.230,231 does not infiltrate to the level of the tuber cinereum.234-236 Modern microsurgical techniques allow total resection in Prognosis up to 75% of cases, with operative mortality rates of 1% to A fairly favorable outcome has been observed for patients 3%.236,241-243 The incidence of long-term functional deficits with pineoblastomas treated with radiation therapy (in- may be higher among patients who undergo radical resec- cluding CSI) and chemotherapy (including alkylating tion than among those treated with radiation.237,244-247 852 PART 10 Central Nervous System

Prognosis Excellent long-term survival rates have been attained for patients who undergo total resection or planned limited surgery and radiation. The proportion of patients in recent series who underwent successful resection approximates 75%.233-241,243,258 However, even the most vociferous ad- vocates of surgery acknowledge the existence of residual calcifications or apparently viable, radiographically obvious tumor in up to 50% of postoperative CT studies of patients who have undergone gross total resection.234-236 The rate of disease recurrence after complete resection ranges from 10% to 30%.233-235,241-243,259 Despite the benign nature of craniopharyngioma, tumor progresses in 70% to 90% of patients after incomplete resection and without radiation at a median of 2 to 3 years.254,259,260 Primary radiation after cyst decompression or partial resection has been associated with progression-free survival rates of 80% to 95% at 5 to 20 years.95,181,238, FIGURE 33-12. Midline sagittal non-enhanced MRI scan dem- 251-255,258,259,261 Survival rates of 90% to 100% are common onstrates a craniopharyngioma involving the upper aspect of among children and adults. the sella, with a solid component extending into the suprasellar The relative toxicity of primary surgery and radiation region along the pituitary-hypothalamic stalk. The cystic com- for craniopharyngioma are becoming increasingly appar- ponent is visible superiorly. ent. Diabetes insipidus occurs in 75% to 100% of patients after complete surgical resection.233,241,246,247 Endocrine dysfunction involving other pituitary-hypothalamic hor- Kramer and colleagues248,249 established the effectiveness mones has occurred in 40% to 80% of patients after sur- of high-dose radiation after limited cyst aspiration, initially gery.246,247 Hypothalamic obesity is apparent in more than reporting 5-year survival for 9 of 10 patients and with later 50% of long-term survivors; it is less common among those follow-up showing 13- to 15-year survival for 6 of 6 chil- who undergo limited surgery and radiation.246,247 Undesir- dren.250 Subsequent series showed similarly impressive long- able personality changes are associated with resections that term survival data.239,251-254 Bloom and colleagues95 updat- damage the hypothalamus. ed Kramer’s initial findings, reporting a 10-year survival rate An increased endocrine deficit has been identified in of 85% for the children and a 10-year relapse-free survival patients whose primary treatment was surgery compared rate of 80%.95 Disease control is even better in young adults with those who underwent limited resection and radia- (92% at 20 years for patients 16 to 39 years old at the time tion.246 Up to 25% of children show significant decreases of diagnosis).254 These latter figures are equal to the best in vision after resection241,243; reports of visual compli- results observed in primary surgical series when corrected to cations from radiation therapy have been anecdotal.255 account for cases with incomplete resection.235,243 Overall performance, intellectual function, and memory seem to be equal or better in the population treated by Radiation Therapy Technique conservative surgery and radiation than in those treated Radiation therapy for craniopharyngioma is designed to surgically.237,245,253,255 narrowly encompass the solid and cystic components of a well-delineated, central neoplasm. Much of the published Primitive Neuroectodermal Tumors information involves treatment using coronal arcs of 180 In 1973, Hart and Earle262 described the cerebral PNET to 220 degrees; three-dimensional conformal or fraction- as a malignant embryonal tumor that affects children and ated stereotactic radiation therapy is being used to improve young adults. The tumor accounts for approximately 3% dose conformity.239,251-254 of supratentorial neoplasms. PNETs are large, often cystic Craniopharyngiomas show a radiation dose-response hemispheric tumors characterized histologically by a highly relationship. Bloom and colleagues95,181 reported a supe- cellular, undifferentiated infiltrate that includes vascular rior outcome in response to 50 to 55 Gy compared with endothelial hyperplasia and necrosis.69,262,263 In classic the response to less than 50 Gy. Doses of 50 to 54 Gy at descriptions, focal areas of glial or neuronal differentiation 1.8 Gy per fraction are recommended for children. It is not constitute less than 5% to 10% of the tumor. The clini- clear that higher doses are necessary or desirable for adults. cally similar cerebral neuroblastoma is distinguished from Doses in excess of 60 Gy are associated with significantly the Hart/Earle PNET by its degree of neuronal differen- greater toxicity, especially in the visual system, with little tiation.264-266 The specific cerebral PNET encountered in improvement in survival.95,181,255 a patient, however, must be distinguished from the broad Single-fraction radiosurgery has been used more broadly concept of PNETs advanced by Rorke and colleagues, as in Europe than in the United States.255,256 Radiosurgery discussed earlier in this chapter.69,70,162,267 may be appropriate for children with small foci of residual Cerebral PNETs infiltrate widely despite their cir- disease, particularly disease that is limited to intrasellar de- cumscribed gross appearance. The tumor can also arise posits or foci of disease remote from the hypothalamus and multifocally. About one third of cases involve ventricu- chiasm.255,257 lar or spinal subarachnoid dissemination.261 The more 33 The Brain and Spinal Cord 853

differentiated cerebral neuroblastoma has been associated 1 with CSF metastasis in an equal proportion of patients at 0.9 0.8 autopsy, although isolated involvement of the subarach- 0.7 noid space is uncommon.261,264,266,268 0.6 (51 ± 10%) Survival Studies conducted by the CCG have documented the 0.5 efficacy of combining chemotherapy (the tested regi- 0.4 Probability 0.3 (21 ± 8%) Progression-free survival men included lomustine, prednisone, and vincristine) 0.2 with CSI for children with cerebral PNETs who are more 0.1 than 3 or 4 years old.232,263,267,269-271 Long-term survival 0 approached 60% in this group. For infants and young chil- 0 2 4 6 8 10 1 dren, therapeutic approaches are comparable with those A Years postdiagnosis used for other embryonal CNS lesions in this age group (see “Tumors in Children Younger than 3 Years”). Despite 120 anecdotal reports of disease control after cranial irradiation, CSI is considered standard therapy for supratentorial 90 PNETs, including cerebral neuroblastoma and pineoblasto- 232,268 ma. The doses and techniques used for supratentorial 60 PNETs are similar to those described for medulloblastoma, with typically a 2-cm margin planned around the tumor Intelligence quotient 30 bed to define the clinical target volume. 0 4 8 Tumors in Children Younger than 3 Years B Time from diagnosis in years Between 15% and 20% of pediatric brain tumors occur FIGURE 33-13. A, Survival and progression-free survival for in infants and children younger than 3 years. The most 29 infants and children younger than 4 years with medulloblasto- common tumors in this age group are astrocytomas (prima. During the study period (1984-1995), prolonged chemother- marily optic chiasmatic or hypothalamic tumors; cere- apy was used in the hope of delaying or obviating the substantial bellar astrocytomas are uncommon), embryonal tumors neurotoxicity associated with irradiation in young children. The (medulloblastoma, PNET, and pineoblastoma), ependy- stable progression-free survival rate after 18 months (21%) moma, and malignant gliomas.272-274 Tumors occurring reflects the use of irradiation as adjuvant therapy (for those with resected or stable disease during 12 to 24 months of chemo- almost uniquely in this age group include atypical tera- therapy); the 51% overall survival rate at 5 years reflects a high toid or rhabdoid tumors and choroid plexus neoplasms salvage rate obtained from the use of craniospinal irradiation 275-279 (papillomas and carcinomas). Desmoplastic in- for persistent or progressive disease. B, Among the 19 patients fantile astrocytoma and ganglioglioma are uncommon, who survived for more than 24 months, the high survival rate and when they do occur, they are typically large tumors came at the cost of a steady decline in full-scale intelligence that are biologically benign.280 The operative mortality quotient scores, from a pretreatment level approximating 90 is relatively high in this age group because of tumor pre- to a level just above 60 at 7 years. (From Walter AW, Mulhern sentations (often sizable, midline or multilobed supra- RK, Gajjar A, et al. Survival and neurodevelopmental outcome tentorial lesions) and the relative lack of elasticity of the of young children with medulloblastoma at St. Jude Children’s still-myelinating brain. Research Hospital. J Clin Oncol 1999;17:3720-3728.) The infant brain is particularly vulnerable to the effects of radiation therapy, particularly with regard to neurocog- intravenous and intraventricular chemotherapy without nitive development.281 Survival rates for patients in this radiation therapy was used for infants with medulloblas- age group with ependymoma or medulloblastoma are sig- toma showed 5-year progression-free survival rates of 82%, nificantly lower than those for other age groups because 50%, and 33% for infants who had complete resection, the tumor in these very young patients is often advanced at residual tumor, and macroscopic metastases, respective- presentation and because tumoricidal radiation doses can- ly.294 Because these results seem to be better than those not be used.95,177,181,271,282-284 in any previous report, they will need confirmation. For The use of prolonged chemotherapy after surgery infants with ependymoma, earlier intervention with to delay or obviate the use of radiation therapy for very radiation after prolonged chemotherapy may help control young children has been studied for almost 20 years (Fig. disease; however, this strategy too seems to work in only 33-13).178,285-291 Trials of treatment for medulloblastoma a minority of cases.194 Chemotherapy alone has been an in infants conducted by the POG, CCG, and International unsuccessful postoperative intervention.295 Reports of sur- Society of Paediatric Oncology (SIOP) have documented vival of patients with pineoblastoma and atypical teratoid that chemotherapy is effective in only a minority of or rhabdoid tumors treated with this approach have been patients; durable disease control has been achieved in 20% anecdotal.178,194,232,296 to 30% of patients with chemotherapy alone and in only The current generation of infant brain tumor trials foc 35% of those treated with response-defined consolidative uses on the judicious use of irradiation. Based on preliminary radiation.178,288-292 Salvage radiation has achieved a surpris- data for children with recurrent medulloblastoma, the ing degree of secondary control after progression during frontline approach in the Pediatric Brain Tumor Consortium chemotherapy, but the dose required to overcome chemo- and COG trials is to introduce early, limited-volume, three- therapy resistance produces late changes in intelligence dimensional conformal radiation therapy to the posterior quotient that make this an ethically unacceptable alterna- fossa and primary tumor site alone for patients with M0 tive.291,293 A trial from Germany in which postoperative medulloblastoma and to the local tumor bed only for those 854 PART 10 Central Nervous System

with supratentorial PNETs.292 For patients with atypical Malignant gliomas typically infiltrate the adjacent nor- teratoid or rhabdoid tumors, targeted radiation volumes mal brain. Detailed comparisons of imaging and histologic are used after just 2 months of chemotherapy for children findings have established the anatomy of malignant glio- 12 months of age and older.296 For those with ependymomas. Histologically, the central low-density region seen on mas, the trend is toward immediate postoperative radia- neuroimaging studies is a necrotic and hypocellular area. tion, at least for patients with tumors that can be grossly The surrounding ring of enhancing tissue is composed of resected.159 densely cellular tumor. The hypodense area beyond the enhancing tissue seen on CT scans is hypocellular and edema- Tumors Common in Adults tous but infiltrated by small anaplastic T cells.300 In 50% of cases, scattered tumor cells immediately surrounding Malignant Gliomas: Anaplastic the hypodense volume are evident histologically.301 The Astrocytoma and Glioblastoma often sizable area of increased signal on T2-weighted MRI Multiforme similarly contains infiltrates of viable T cells.302 Tumor cells The dominant primary intracranial tumors that occur characteristically extend along neural tracts and across the in adults in terms of frequency and mortality are the corpus callosum.301 Distant subarachnoid or subependy- malignant gliomas. These tumors occur in the cerebral mal extension occurs in 5% to 9% of patients with malig- hemispheres as sizable, rapidly growing lesions with a nant gliomas.303,304 characteristic ringlike, enhancing appearance on CT or Surgery for malignant gliomas is usually limited to MRI, with central necrosis, infiltrating margins, and sur- subtotal resection. Macroscopically, complete removal is rounding low-density changes (Fig. 33-14). Malignant achieved in only 10% to 20% of cases.305-308 The extent of gliomas occur in all age groups but predominate in the residual tumor visualized on postresection CT scans seems fifth and sixth decades, and they account for 35% to 45% to correlate with outcome, with increased time to progres- of all adult brain tumors. sion and survival being associated with smaller residual tu- Histologically, malignant gliomas are heterogeneous mor volumes among adults and children.297,309 neoplasms composed of fibrillary astrocytes; gemistocytes; Radiation therapy has been the most effective treat- large, bizarre glial cells; and small anaplastic cells.69,115 In ment for these aggressive lesions, but it usually can only Kernohan and Sayre’s classic categorization297 of astrocytic delay disease progression or recurrence. Early clinical neoplasms, which defined these tumors as grade III and trials conducted by the Brain Tumor Study Group con- grade IV astrocytomas, malignant gliomas were thought to firmed the clinical efficacy of radiation. Median survival represent progressively dedifferentiated astrocytic tumors increased from 14 weeks after surgery to 36 weeks for marked by a certain degree of pleomorphism, hyperchro- patients given postoperative radiation.306 The survival rate maticism, and mitosis. In contemporary neuropathology, at 1 year was 24% for patients also treated with radiation malignant gliomas are classified as anaplastic astrocytoma therapy, compared with only 3% for patients who had or the more malignant glioblastoma multiforme, with surgery only.306 One fourth to one half of patients who the latter identified by the presence of tumor necrosis.298 receive radiation therapy show clinical improvement. Up Anaplastic astrocytomas account for 10% to 15% of malig- to 60% of patients are able to return to work and achieve nant gliomas; more than 85% are glioblastomas.299 full function.310-312

A B

FIGURE 33-14. MRI scans show a glioblastoma multiforme arising in the left thalamic region and extending just beyond midline. A, Gadolinium-enhanced T1-weighted MR image shows a ring-enhancing periphery with a dark, necrotic center. B, T2-weighted MR image shows surrounding edematous changes. 33 The Brain and Spinal Cord 855

Despite documented improvement in time to progres- constituting 20% to 40% of cases. The criteria for implanta- sion and survival, the response to radiation illustrated by tion include Karnofsky performance status scores above 70, imaging studies is complex and rarely indicative of rapid or unifocal supratentorial lesions less than 6 cm in the great- significant disease reduction. Essentially all glioblastomas est diameter, and no involvement of the midline, corpus and 65% to 80% of anaplastic astrocytomas recur within callosum, or subependyma.319,321 A retrospective review of 2 to 5 years and result in rapid death, causing malignant unselected patients treated with standard external beam ra- gliomas to live up to their reputation of being one of the diation therapy to the brain (EBRT) revealed that outcome most aggressive and lethal tumors.310 Although the anat- was significantly better among implant-eligible patients omy of malignant gliomas suggests that disease extension with glioblastoma (median survival of 14 months) than or metastasis is the source of recurrence, recurrent disease among those who were ineligible (median survival of less almost always arises from within the central or enhancing than 6 months),341 indicating that any perceived improve- portion of the tumor.313-317 The central tumor is relatively ment in outcome in patients attributed to the brachyther- radioresistant, perhaps because of an inherent lack of ra- apy may be artificial. However, median survival duration diosensitivity or a hypoxic, relatively radioinsensitive ne- and 2- and 3-year survival rates for patients with glioblas crotic core.316,317 This pattern of intralesional recurrence toma treated with 125I brachytherapy seem to be superior has stimulated several lines of clinical investigation to over- in single-institution experiences and in comparison with come hypoxic cell resistance; considerable interest has also the previously described findings.319,321,341 Comparable been expressed in exploiting the time-dose relationship to data for anaplastic astrocytomas are less convincing.321,341 identify an altered fractionation regimen of value in treating Symptomatic local necrosis occurs in 40% to 50% this tumor.299,312 Several trials have tested modifications in of patients with interstitial implants, and this often radiation delivery, including physical modalities other than requires surgical intervention.320,321,326,327,332 Ultimate photons and local boost techniques incorporating intersti- survival has been reported to be superior in patients who tial brachytherapy or stereotactic radiosurgery.318-329 require a second operation, which presumably reflects the Early trials of radiosensitizers focused on the nitro- combination-treatment effect.320 imidazoles misonidazole and etanidazole.330 Later stud- Stereotactic radiosurgery has been used to achieve a high- ies have investigated carbogen and nicotinamide. Despite dose boost. Eligibility criteria are selective, typically includ- some improvement realized from use of these agents with ing single lesions less than 4 to 5 cm in diameter without suboptimal radiation, clinically meaningful benefit was not subependymal extension and more than 5 mm away from achieved.331,332 Long-standing interest in the halogenated the optic chiasm or brainstem. In the initial U.S. series, pyrimidines bromodeoxyuridine and iododeoxyuridine has which included a high proportion of patients who had un- resulted in several trials of their use with conventional and dergone limited surgical resection, the median survival time altered fractionation, but only a marginal benefit has been was at least equal to that among patients with implants: suggested by the results.333-336 26 months for those with glioblastoma and more than The use of radiation with high-LET particles has been 36 months for those with anaplastic astrocytoma.342,343 In examined in clinical trials as a possible way of overcoming a multi-institutional trial of 189 patients, the median sur- the apparent oxygen-related radioresistance of these tu- vival time was 86 weeks for those with glioblastoma who mors. However, the use of fast–neutron beam radiation has met all criteria for gamma knife intervention, compared produced no significant improvement over conventional with 40 weeks for those with more extensive tumors or a photon radiation.318,322,337 The rationale for using proton less favorable performance status score.323,328,341 The inci- radiation stems more from the ability to restrict the vol- dence of symptomatic necrosis may be less than that for ume of tissue irradiated than from its being less affected patients with interstitial implants.271 by oxygenation; trials of doses as high as 90 Gy suggest a Several trials in the United States, Europe, and Japan potential benefit in terms of central tumor control, but no have assessed modifications in the radiation time-dose rel early demonstration of improvement in outcome has been ationship.299,312,338,344-347 Despite an early suggestion that seen.338 hyperfractionated delivery to higher doses might be ben- Stereotactic interstitial implantation of radioactive eficial, subsequent studies conducted by the RTOG have sources has been done at several European centers since shown no substantial benefit from such therapy. A large the 1950s. Over the past 2 decades, prospective studies of randomized dose-escalating trial conducted by the RTOG temporary stereotactic brain implants using high-activity tested hyperfractionated doses of 64.8 to 81.6 Gy, given in iodine 125 (125I) have evaluated the toxicity and efficacy of 1.2-Gy fractions twice daily with an interfractional interval volume-limited, high-dose radiation delivery.319-321 Among of 4 to 8 hours. The results indicated that the ideal dose selected patients with limited-volume recurrent malignant response occurred at 72 Gy, although the superiority of gliomas, geometrically planned implantations of radiation this schedule over 60 Gy delivered in conventionally frac- sources delivering an average of 60 Gy to the enhancing tionated radiation therapy was inapparent.344 Patients who lesion over 6 days resulted in survival durations of approxi- received doses of 75 to 81.6 Gy fared less well.344 Without mately 1 year for those with glioblastomas and 1.5 years documentation of any reduction in neurotoxicity from hy- for those with anaplastic astrocytomas.319-321,326,339 Results perfractionated delivery, few objective data indicate that of the Brain Tumor Cooperative Group’s prospective ran- the therapeutic ratio and effect of hyperfractionation are domized trial suggested that interstitial boost radiation greater than those for conventional fractionation.344,348,349 could be beneficial in appropriately selected cases.340 Alternative time-dose regimens have included accelerat- Patients with malignant gliomas who are potentially ed hyperfractionation (1.5- or 1.6-Gy fractions given twice suitable for local types of therapy (e.g., brachytherapy, ra- daily), accelerated fractionation (1.9- or 2.0-Gy fractions diosurgery) are those with a relatively favorable prognosis, given three times daily), and hypofractionation (fractions 856 PART 10 Central Nervous System of 2.5 to 6 Gy).332,334-347 Although the hypofractionation 70 regimens may expedite palliation in patients with advanced 25th percentile or unfavorable conditions, these regimens have provided 60 no overall benefit. Chemotherapy has been investigated extensively for the 50 treatment of glioblastomas. As sole adjuvants to surgery, 306 Median nitrosoureas have been ineffective. When combined 0.004 with radiation, nitrosoureas or combinations of agents (e.g., 40 carmustine-hydroxyurea plus procarbazine-teniposide, 0.110 carmustine-hydroxyurea, vincristine-procarbazine) have 30 Survival in weeks been associated with only marginally improved survival 0.174 306,311,350 75th percentile times for patients enrolled in prospective studies. 20 Despite one report of encouraging early results with penciclovir,351 a randomized trial conducted by the Medical 10 Research Council in the United Kingdom of more than 4500 5000 5500 6000 670 patients showed no survival benefit for patients given penciclovir in addition to standard radiation therapy.352 Two Rads randomized trials testing concurrent temozolomide plus FIGURE 33-15. Dose-response analysis for survival of adults radiation therapy with radiation therapy alone demonstrated with malignant gliomas. (Data from Walker MD, Green SB, Byar a significant survival benefit for the combination.353-354 DP, et al. Randomized comparisons of radiotherapy and nitro- In a trial by the European Organisation for Research and soureas for the treatment of malignant glioma after surgery. Treatment of Cancer, the survival benefit from temo- N Engl J Med 1980;303:1323-1329.) zolomide was more apparent when the MGMT gene was epigenetically silenced.355 The current RTOG trial uses MGMT status in the tumor as one of the criteria for stratifi- and colleagues306 showed that median survival time im- cation. Many biological agents are available for the treatment proved from 28 weeks for patients given 50 Gy in 25 to of malignant gliomas, including compounds with specific 28 fractions to 42 weeks for patients given 60 Gy in the biological targets (e.g., epidermal growth factor, platelet- first Brain Tumor Study Group studies (Fig. 33-15).306 derived growth factor) and antiangiogenesis agents.356 One Three-dimensional conformal and proton-based regimens ongoing trial is using a vaccine against a truncated form of the have made it possible to escalate doses to 70 to 90 Gy; epidermal growth factor receptor (EGFRvIII) for patients however, although central tumor control may be improved, whose glioblastomas have undergone gross total resection no increase in survival has been found.338,358 As summa- and demonstrate overexpression of EGFRvIII. Studies of lo- rized earlier, none of the altered fractionation regimens has cal chemotherapy, specifically carmustine-impregnated wa- been shown to improve outcome over that achieved with fers (Gliadel), have yielded inconclusive results in patients conventionally fractionated radiation to 60 to 65 Gy (in with malignant gliomas, producing an apparent response 1.8- to 2-Gy fractions given once daily).345 No data exist to but being of uncertain value as an adjuvant therapy.357 suggest that these recommendations for radiation therapy for adults be adjusted for children; however, the require- Radiation Therapy Technique ment for high-dose radiation to reasonably large volumes Radiation therapy remains the most effective single agent limits the use of this modality in young children with ma- in prolonging survival in patients with malignant gliomas. It lignant gliomas. should consist of standard EBRT with wide local coverage of the neoplasm.310,314 Anatomic studies of tumor extent, Prognosis imaging analyses of patterns of recurrence, and clinical tri- Reports of survival beyond 2 years for adults with glio- als have shown that the appropriate initial target volume blastoma remain anecdotal.359 However, approximately should extend 2 to 3 cm beyond the low-density periphery 15% to 25% of patients with anaplastic astrocytoma sur- shown on CT scans or 2 cm beyond the abnormal signal vive 5 years, with gradually decreasing survival beyond shown on T2-weighted MRI studies.301,302,314,315,350 Mar- that time documented in most series. Levin and col- gins of less than 2 cm are theoretically inadequate because leagues351 reported that adjuvant penciclovir chemother- of the histologic nature of the tumor301 and the way in apy is associated with a 35% survival rate at 3 to 5 years which the initial sites of disease progress.315 for patients with anaplastic lesions; however, these and The use of limited high-dose treatment volumes deliv- similar results seen in other small series were not sub- ered by three-dimensional conformal or proton radiation stantiated in a Medical Research Council trial.314,352,360 may be associated with a relative increase in the peripheral Other factors influencing prognosis include age (survival failure rate.338,358 Central or intralesional disease progres- is progressively better among younger patients, with the sion is less common in patients with interstitial implants. best among 18 to 39 year olds, the next best among 40 to Studies of patterns of failure in series of patients treated 59 year olds, and the worst among those 60 years old or with brachytherapy have shown that disease recurs primar- older) and initial performance status.305,325 Among chil- ily in the tumor margin, often within 2 cm of the high-dose dren with malignant gliomas, the overall progression-free implant; the incidence of distant intracranial recurrence is survival rates at 3 to 5 years range from 15% to 20%; also relatively high.316,317 for the small subset with completely resected lesions, Dose-response data were established early in clinical corresponding survival rates of up to 75% have been trials involving patients with malignant gliomas. Walker reported.309,361,362 33 The Brain and Spinal Cord 857

after complete resection occurs in 10% to 20% of cases, re- Oligodendroglioma lated in part to the extent of dural and sinus excision.376-379 Oligodendroglioma is a relatively uncommon tumor that Symptomatic tumor recurrence or progression takes place occurs in all age groups. The tumor arises most often in in approximately 50% of patients with subtotally excised the frontal lobes and occasionally involves the infratento- lesions, most often in patients with meningiomas of the rial or spinal regions. The most common symptom is sei- sphenoid ridge and parasellar areas, which represent almost zures. The relatively nonaggressive nature of this tumor is 30% of all meningiomas (Fig. 33-16).376,379-381 The efficacy indicated by the median survival time, which is more than of radiation therapy for incompletely resected meningio- 5 years from the onset of symptoms and diagnosis.363 The mas has been well documented. Wara and colleagues382 co-deletion of the short arm of chromosome 1 (1p) and reported local tumor recurrence in 74% of patients after the long arm of chromosome 19 (19q) has been associ- incomplete resection and in 29% of patients given post- ated with less aggressive behavior of the tumor and better operative radiation. Progression-free survival rates beyond response to therapy.364,365 5 years average 70% to 90% for patients undergoing incom- Primary surgical resection is the initial therapy for low- plete resection and radiation; disease control rates in excess grade oligodendrogliomas.366,367 The role of postoperative of 70% to 80% at 8 to 10 years are associated with survival radiation has been poorly defined. In a small, retrospective rates exceeding 90%.383-385 Similar outcomes have been study, Sheline and colleagues368 found an improvement in observed among patients who receive radiosurgery alone 5-year survival rates for patients who underwent irradia- or as an adjunct to incomplete surgery.386-389 Neurologic tion compared with those who did not (85% versus 55%), improvement has occurred in 44% of patients after radia- but the survival rate at 10 years was not statistically im- tion for unresected primary lesions.376 Serial CT studies proved. Chin and colleagues369 reported a progression-free have confirmed the regression of neoplasms in response to survival rate of 100% at 5 years for 24 patients complet- radiation therapy (see Fig. 33-16).381 In sites such as the ing postoperative radiation. Lindegaard and colleagues370 optic nerve sheath, operative intervention may be associ- observed a benefit from radiation only among those who ated with a significant functional deficit. Primary irradia- had undergone incomplete surgical resection. In a study of tion can improve tumor-related defects in visual acuity and a similar group of patients treated during roughly the same visual fields in these patients.390 period, Reedy and colleagues371 found no survival advantage from radiation therapy. Radiation Therapy Technique In the absence of a prospective trial assessing the effi- Because meningiomas are locally infiltrating tumors that cacy of radiation therapy, it seems justified and prudent to produce local tissue changes by means of direct tumor ex- use postoperative radiation for subtotally resected lesions. tension or have subclinical, limited multicentricity,391 local Local fields to doses of 50 to 54 Gy approximate the treat- irradiation should include a reasonable margin despite the ment recommended for other low-grade tumors.106,372 circumscribed appearance of these tumors on CT scans. Although 5-year survival rates in excess of 65% to Large tumors of the sphenoid ridge may extend beyond 90% have been observed in major clinical series, the the cranial vault into the orbit or the facial or cervical re- ultimate survival rate beyond 15 or 20 years was only gions. Optimal radiologic assessment of tumor extent and 30% in two large reviews from an earlier era that exam- the use of wide inferior treatment margins are important ined the pathologic behavior of tumors from more than in these cases.381 Three-dimensional conformal techniques, 500 patients.363,373 Late tumor progression and death take intensity-modulated radiation therapy, and proton therapy place more than 5 to 10 years after diagnosis. Histologic are suitable for the treatment of meningiomas. grading is of increasing importance because a role for Carella and colleagues383 were unable to identify a dose- penciclovir chemotherapy in the treatment of anaplas- response relationship for meningiomas between 50 and tic oligodendrogliomas has been established.374 Long- 75 Gy. However, excellent local tumor control has been term survival is higher for patients younger than 40 or achieved at doses approaching CNS tolerance, and the rec- 50 years.373 ommended treatment is a total dose of 50 to 55 Gy given in 25 to 30 fractions.376,381,384 The use of radiosurgery as Meningioma the primary treatment for selected lesions has produced Meningiomas are common CNS tumors, most often occur- equally positive results; at many centers, radiosurgery is the ring as intracranial tumors along the convexities or in para- treatment of first choice for limited-volume lesions that are sagittal regions. The usually benign meningioma constitutes difficult to approach surgically or are residual or recurrent 15% of adult CNS tumors, with a female-to-male ratio of after initial resection.386,388,389,392,393 2.5 to 1. The natural history of this tumor was elegantly described in Cushing and Eisenhardt’s 1938 monograph,375 Prognosis in which they identified the broad dural attachment (or The benign nature of meningiomas is confirmed by the local infiltration), growth along or into the venous sinuses, limited number of recurrences seen in patients after to- and osseous invasion or overlying reactive bone formation tal surgical resection. After incomplete resection, however, that typify these lesions. Angioblastic meningiomas have the median time to clinically apparent tumor progression a less favorable outcome than the other so-called benign is 4 years after surgery. The median intervals to tumor pro- tumors.376 Less than 10% of meningiomas are malignant, gression in patients given radiation therapy exceed 5 to and these lesions seem to be associated with an unfavorable 10 years.381,383,384 outcome.376-378 Initial postoperative irradiation for incompletely Surgery is the primary treatment for most peripherally resected or biopsied meningiomas has achieved excellent located intracranial meningiomas. Local tumor recurrence local tumor control, as described in the previous paragraphs. 858 PART 10 Central Nervous System

FIGURE 33-16. CT scans of an orbital sphenoidal meningioma demonstrate residual disease after surgery (top row) and 2 years after irradiation (bottom row). (From Petty AM, Kun LE, Meyer GA. Radiation therapy for incompletely resected meningiomas. J Neurosurg 1985;62:502-507.)

Indications for initial or delayed radiation therapy remain the acquired immunodeficiency syndrome (AIDS) epi- controversial, although increasing information suggests that demic.396-398 The tumor occurs primarily in the supraten- incompletely resected lesions along the cavernous sinus and torial region, most often as an isodense, diffusely enhancing those with malignant histologic features may warrant im- lesion involving the basal ganglia or thalamus, the periven- mediate postoperative irradiation.377,385,389,392 No appar- tricular white matter, or the corpus callosum.399 Up to 15% ent decrease in tumor control has been observed in most to 40% of tumors are multifocal intracranial lesions.400,401 series of patients who undergo repeat surgery and radia- The CSF contains tumor cells in 10% to 20% of cases; in- tion therapy for disease that recurs after initial incomplete volvement of the ocular vitreous at diagnosis or metachro- resection.376,380,384 A more rapid rate of tumor regrowth nously occurs in 10% to 25% of cases.400,402 after subtotal excision has been indicated in children, rais- The lesion is usually classified as an immunoblastic or ing some concern about the advisability of delaying radio- lymphoblastic diffuse, large cell, malignant lymphoma, therapeutic intervention in this age group.394,395 most often of a B-cell immunophenotype. Systematic Identifying the histologic subtype of meningioma69,378 evaluation is necessary to distinguish primary lymphoma is important because the local recurrence rate is higher for of the brain from secondary CNS manifestations of malig- malignant (anaplastic) meningiomas and because postoper- nant lymphoma, although intraparenchymal deposits oc- ative radiation therapy can effectively control these tumors. cur more often with primary tumors than with secondary For example, Dziuk and colleagues377 and others378,380 have tumors. reported 15% to 80% differences in 5-year progression- Surgery is limited, often to biopsy alone, but radiation free survival rates related to radiation. A long-term disease therapy produces a high rate of objective response. Neu- control rate of only 25% at 5 years has been observed for rologic improvement can be achieved in up to 70% of pa- patients who receive single-fraction radiosurgery, in most tients, but median survival has been only 7 to 8 months.403 cases for recurrent malignant meningiomas.393 Long-term survival after radiation alone has been limited.398,400,401,404,405 Chemotherapy has shown increas- Primary Malignant Lymphoma of the ing efficacy in patients with CNS malignant lymphoma, Central Nervous System with rates of objective response often exceeding 50% to Primary malignant lymphoma of the CNS was previously 70% in those treated with regimens including high-dose a rare neoplasm, occurring spontaneously in conjunction methotrexate or, less often, with conventional systemic with immunosuppression. However, the frequency with chemotherapy regimens (e.g., cyclophosphamide, doxoru- which it occurs has increased greatly in conjunction with bicin, vincristine, and prednisone).406-410 The combination 33 The Brain and Spinal Cord 859 of preirradiation chemotherapy and cranial irradiation re- 1.00 sulted in median survival times as long as 60 months and 4-year progression-free and overall survival rates of ap- 0.9 proximately 50% in a moderate-size series from Memorial Sloan-Kettering Cancer Center.408,411 One drawback to the use of combined methotrexate-radiation regimens has 0.8 been their considerable neurotoxicity, which sometimes produces frank dementia, particularly in patients older 0.7 than 60 years.407,408,412

Radiation Therapy Technique 0.6 Given the size and often the multiplicity of most primary CNS lymphomas, the minimal target volume is often 0.5 full cranial irradiation.401-404 Isolated subarachnoid spinal recurrence after cranial radiation occurs in 4% to 25% of patients.106,404 The inability to control the primary tumor survival Fraction 0.4 ³50 Gy within the cranium in a high proportion of patients and the need to coordinate radiation with chemotherapy have 0.3 P<.05 led clinicians away from considering CSI. This raises a valid argument for avoiding spinal irradiation in deference to the 0.2 potential benefits conferred by chemotherapy. A radiation dose-response relationship has been <50 Gy suggested in the treatment of primary malignant lym- 0.1 phoma of the CNS. Murray and colleagues401 reviewed 198 cases reported in the literature and found a statistically significant improvement in the 5-year survival rate for 0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60 patients given doses of more than 50 Gy (42% versus 13%) Survival (months) (Fig. 33-17).401 However, an analysis at Memorial Sloan- Kettering Cancer Center did not find additional benefit FIGURE 33-17. Dose-response analysis for survival of patients from doses higher than 45 Gy. Recurrence of tumor at the with primary cerebral malignant lymphoma after irradiation. primary site even in patients treated with doses of 50 to 55 (From Murray K, Kun L, Cox J. Primary malignant lymphoma of the central nervous system. J Neurosurg 1986;65:600-607.) Gy led to a major RTOG trial, the results of which failed to demonstrate a significant added benefit from high-dose cranial irradiation.403 In small series of patients, however, irradiation achieves more prolonged palliation, with objec- the addition of radiation therapy seems to have prolonged tive improvement in neurologic function in 50% to 70% of disease control and improved survival relative to the out- patients treated in this manner.319 come for patients given chemotherapy only.398,408,411 Earlier time-dose studies showed the relative equiva- lence of the following treatment regimens: 30 to 36 Gy Prognosis given in 10 to 12 fractions; 20 Gy given in 5 fractions; and Overall survival in patients with primary malignant lym- 40 Gy given in 15 fractions.417a The current standard ther- phoma seems to have increased among those treated with apy is 30 Gy given in 10 fractions; however, randomized combined, sequential chemoradiotherapy; median survival trials to assess the benefit of surgery or radiosurgery in se- times have increased several-fold to approximately 4 to lected settings favor 37.5 Gy given in 15 fractions.418,419 5 years. However, most patients do die of the disease, some- For patients with a solitary metastasis, the addition of times as long as 5 years after diagnosis.408,411 surgical resection or radiosurgery has significantly improved survival time (from a median of 15 weeks to a median of Carcinoma Metastatic to the Brain 40 weeks in a randomized trial reported by Patchell and col- Intracranial metastasis, primarily in the cerebrum, occurs in leagues420) and the rate of local control at metastatic sites 10% to 15% of patients with cancer.413 Of the types of can- (from less than 50% for patients treated with cranial irradi- cer that metastasize preferentially to the brain, cancer of the ation alone to 80% for patients undergoing resection). This lung and breast predominate, and gastrointestinal and renal increase in survival also has been associated with improved carcinomas are less common. Brain metastasis often occurs functional outcome.420 In several series, the outcome from at multiple sites. In a classic autopsy series, solitary metasta- radiosurgery has been similar to that from operative removal; ses were found in approximately 40% of cases, two to three in all of these series, the patients with favorable factors (e.g., lesions in 25% of cases, and four or more foci in 35%.414 age, performance status, absence of disease at other sites) Treatment of brain metastases is palliative; long-term are the ones who benefit the most from the addition of local survival can be achieved for up to 20% of patients with fa- therapy.388,418,419,421,422 Surgery and radiosurgery seem to vorable presenting characteristics, such as age younger than achieve equivalent results in most reports.419,421 Detailed 60 to 65 years, favorable performance status (typically, a analyses have shown that the addition of full cranial irradia- Karnofsky performance status score of at least 70), control tion improves survival and local tumor control and may be of the primary tumor, and absent or controlled extracra- associated with long-term survival of a small but meaning- nial metastasis.415-417 Corticosteroids alone can transiently ful number of patients undergoing aggressive local thera- reduce symptoms in 60% of patients.347 However, cranial py.423 One randomized trial comparing radiosurgery alone 860 PART 10 Central Nervous System with the combination of radiosurgery and whole-brain irra- patients with ependymomas despite frequent neurologic diation showed significantly fewer intracranial recurrences improvement.429,431 The disease-free survival rate at 5 to with the combination therapy, although survival times were 10 years ranges from 25% to 60% among patients with spi- similar in both groups.424 Results for patients with malig- nal astrocytomas.429,438,439 However, the histologic grade nant melanoma that has metastasized to the brain have affects outcome, in that patients with low-grade astrocyto- generally paralleled those seen for patients with other car- mas enjoy a short-term (5-year) survival rate of as high as cinomas; however, local therapy may be particularly advan- 89%, compared with 0% among patients with higher-grade tageous for patients with few metastatic sites and favorable tumors (e.g., glioblastoma).438 performance status.421,425 Spinal cord tolerance is an important consideration when The treatment selected for patients with brain metastasis reviewing results of patients with intramedullary tumors depends on the absence or presence of the clinical features treated with radiation.440 Doses of approximately 50 Gy consistently associated with improved outcome (i.e., age are recommended for most spinal cord tumors; however, younger than 60 to 65 years, Karnofsky performance status higher doses and the suggestion of a dose-response effect score of 70 or higher, and the lack of apparent or symptom- at or above 50 Gy have been reported for localized spinal atic disease at the primary site or other extracranial loca- cord lesions.441,442 tions). However, there does seem to be a definite advantage Extradural metastasis is the most common oncologic di- to surgery or radiosurgery for patients with only one meta- agnosis for patients with spinal cord disease. An estimated static focus; a similar advantage seems to exist for those with 5% of patients who die of cancer have clinical signs of epi- only two or three metastatic foci.388,418,419,421,422,426 In all of dural metastases.443 Extradural involvement usually results these patients and in those with more advanced cranial or ex- from the contiguous extension of a vertebral metastasis, traneural disease, the addition of full cranial irradiation is an most often in patients with primary cancer of the breast, important and worthwhile palliative intervention.415,418,423 lung, or prostate. Less common are vertebral metastases from malignant lymphomas and myeloma or carcinomas Spinal Cord Tumors of the kidney and gastrointestinal tract. Early evaluation with spinal MRI has been advocated for Primary tumors of the spinal cord represent 10% to patients with pain and radiographic evidence of vertebral 15% of CNS neoplasms. Extramedullary tumors, including metastasis without objective neurologic signs.443,444 The schwannomas and meningiomas, account for almost one treatment of epidural metastases is controversial, largely half of the lesions in the spinal canal. Primary intramedul- with respect to the use of surgery before radiation. Con- lary neoplasms are most often gliomas or vascular tumors. ventional procedures have included decompressive lami- Intramedullary gliomas include ependymomas (60%) nectomy, at least for patients with rapidly evolving signs or and astrocytomas (30%). Ependymomas occur more often complete block shown by myelography; patients for whom in the lumbar region, affecting the conus and cauda equina. the diagnosis is uncertain or the diagnosis of malignancy is Spinal ependymomas usually are low-grade neoplasms; not established require surgical intervention.339 those occurring in the cauda equina usually are well- Retrospective and prospective studies comparing lami- differentiated myxopapillary tumors.427 The ependymo- nectomy and postoperative radiation with radiation alone mas tend to develop as discrete lesions, facilitating subtotal confirm an apparent advantage to surgical intervention or gross total resection. only for patients with nonambulatory symptoms; however, Gross total resection led to systemic disease control in there has been little difference in outcome for most pa- recent series with a 5-year follow-up. Aggressive surgery, tients.443,445-447 Experience with vertebral body resection however, is associated with postoperative neurologic dete- in selected cases indicates improved neurologic outcome rioration in 15% of patients.428 Response to radiation is well after a more aggressive operative approach.448-450 Primary documented, with 75% of patients showing improvement irradiation is often the treatment of choice for certain his- in neurologic status and progression-free survival rates of tologic types of tumors known to respond relatively rapidly 70% to 100% at 10 years.429-431 A local tumor control rate and for patients with residual functional deficits. of 80% with no identifiable long-term neurologic toxicity Irradiation for spinal cord compression is considered despite doses in excess of 45 Gy indicates that radiation a radiotherapeutic emergency. Prompt initiation of dexa- has a role in the treatment of patients whose spinal tumors methasone and high-dose (2- to 4-Gy) fractions are indi- are incompletely resected.431 Patients with tumors of the cated. After two or three treatments, the dose per fraction cauda equina region and conus have particularly high sur- and total dose depend on the type of malignancy and clini- vival rates after radiation or surgical resection.432,433 cal status of the patient. Spinal astrocytomas are evenly distributed along the Carcinomatous meningitis or diffuse leptomeningeal length of the spinal cord. These lesions often contain infiltration is a relatively uncommon late manifestation of cystic components and most often are low-grade fibril- systemic cancer. Primary tumors are usually carcinomas of lary tumors. Astrocytomas are more infiltrating than ep- the breast or lung; small cell lung cancer in particular is endymomas, limiting surgery to cyst decompression and often associated with intracranial metastasis. Symptoms partial resection.432 Epstein and colleagues,434-437 how- and signs referable to the spine predominate in 25% of pa- ever, described operative techniques that can be used to tients; signs related to spinal root infiltration are often ap- completely resect many low-grade astrocytomas in chil- parent.451 Treatment consists of cranial irradiation or CSI dren and adults. Aggressive operative intervention is not (depending on the histologic findings and coordinated use advantageous for high-grade spinal astrocytomas.429,438,439 of chemotherapy), with or without intrathecal or systemic Long-term control in patients with spinal astrocytomas chemotherapy. Responses to treatment are common, but treated with radiation is less predictable than that in long-term survival results are largely anecdotal. 33 The Brain and Spinal Cord 861

REFERENCES 23. Hoffman WF, Levin VA, Wilson CB. Evaluation of malignant glioma patients during the postirradiation period. J Neurosurg 1. Schultheiss TE , Kun L E , Ang K K , et al. Radiation response 1979;30:624-628. of the central nervous system. Int J Radiat Oncol Biol Phys 24. Longee DC, Fiedman HS, Djang NW, et al. Transient late mag- 1995;31:1093-1112. netic resonance imaging changes suggesting progression of brain 2. Fajardo LF, Berthrong M, Anderson RE. Central nervous system stem glioma: implications for entry criteria for phase II trials. and spinal cord. In Radiation Pathology. New York: Oxford Uni- Neurosurgery 1988;23:248-253. versity Press, 2001, pp 351-361. 25. Duffner PK, Cohen ME, Myers MH, et al. Survival of children 3. Burger PC, Mahaley MS Jr, Dudka L, et al. The morphologic with brain tumors: SEER program, 1973-1980. Neurology effects of radiation administered therapeutically for intracra- 1986;36:597-601. nial gliomas: a postmortem study of 25 cases. Cancer 1979;44: 26. Childhood Brain Tumor Consortium. A study of childhood brain 1256-1272. tumors based on surgical biopsies from 10 North American 4. Constine LS, Konski A, Ekholm S, et al. Adverse effects of brain institutions: sample description. J Neurooncol 1988;6:9-23. irradiation correlated with MR and CT imaging. Int J Radiat 27. Groothuis DR, Wright DC, Ostertag CB. The effect of 125I Oncol Biol Phys 1988;15:319-330. interstitial radiotherapy on blood-brain barrier function in 5. Valk PE, Dillon WP. Radiation injury of the brain. AJNR Am J normal canine brain. J Neurosurg 1987;67:895-902. Neuroradiol 1991;12:45-62. 28. Schultheiss TE, Stephens LC. The pathogenesis of radiation 6. Gutin PH. Treatment of radiation necrosis of the brain. In Gutin myelopathy: widening the circle. Int J Radiat Oncol Biol Phys PH, Leibel SA, Sheline GE (eds): Radiation Injury to the Nerv- 1992;23:1089-1091. ous System. New York: Raven Press, 1991, pp 271-282. 29. Schultheiss TE, Stephens LC. Invited review: permanent radia- 7. DeAngelis LM, Shapiro WR. Drug/radiation interactions and tion myelopathy. Br J Radiol 1992;65:737-753. central nervous system injury. In Gutin PH, Leibel SA, Sheline 30. Marks JE, Baglan RJ, Prassad SC, et al. Cerebral radionecrosis: GE (eds): Radiation Injury to the Nervous System. New York: incidence and risk in relation to dose, time, fractionation and Raven Press, 1991, pp 361-382. volume. Int J Radiat Oncol Biol Phys 1981;7:242-252. 8. Van der Kogel AJ. Central nervous system radiation injury in 31. Mikhael MA. Dosimetric considerations in the diagnosis of small animal models. In Gutin PH, Leibel SA, Sheline GE (eds): radiation necrosis of the brain. In Gilbert HA, Kagen AR (eds): Radiation Injury to the Nervous System. New York: Raven Press, Radiation Damage to the Nervous System. New York: Raven 1991, pp 91-112. Press, 1980, pp 59-92. 9. Fike JR, Gobbe GT. Central nervous system radiation injury in 32. Kramer S, Southard ME, Mansfield CM. Radiation effect and large animal models. In Gutin PH, Leibel SA, Sheline GE (eds): tolerance of the central nervous system. Front Radiat Ther On- Radiation Injury to the Nervous System. New York: Raven Press, col 1972;6:332-345. 1991, pp 113-136. 33. Aristizabal S, Caldwell WL, Avila J. The relationship of time- 10. Ostertag CB. Experimental central nervous system injury from dose fractionation factors to complications in the treatment of implanted isotopes. In Gutin PH, Leibel SA, Sheline GE (eds): pituitary tumors by irradiation. Int J Radiat Oncol Biol Phys Radiation Injury to the Nervous System. New York: Raven Press, 1977;2:667-673. 1991, pp 183-190. 34. Leibel SA, Sheline GE. Tolerance of the brain and spinal cord to 11. Caveness WF. Experimental observations: delayed necrosis in conventional irradiation. In Gutin PH, Leibel SA, Sheline GE normal monkey brain. In Gilbert HA, Kagen AR (eds): Radia- (eds): Radiation Injury to the Nervous System. New York: Raven tion Damage to the Nervous System. New York: Raven Press, Press, 1991, pp 239-256. 1980, pp 1-38. 35. Clemente CD, Yamazaki JN, Bennett LR, et al. Brain radiation in 12. Price RA, Jamieson PA. The central nervous system in childhood newborn rats and differential effects of increased age. Neurology leukemia. Subacute leukoencephalopathy. Cancer 1975;35 (Pt 1960;10:669-675. 2):306-318. 36. Bloom HJG, Wallace ENK, Henk JM. The treatment and prog- 13. Fouladi M, Langston J, Mulhern R, et al. Silent lacunar lesions de- nosis of medulloblastoma in children. AJR Am J Roentgenol tected by magnetic resonance imaging of children with brain tu- 1969;105:43-62. mors: a late sequela of therapy. J Clin Oncol 2000;18:824-831. 37. Boden G. Radiation myelitis of the brain stem. J Fac Radiol 14. Salazar OM, Rubin P, Feldstein ML, et al. High dose radiation (Great Britain) 1950;2:79-94. therapy in the treatment of malignant gliomas: final report. Int J 38. Pallis CA, Louis S, Morgan RL. Radiation myelopathy. Brain Radiat Oncol Biol Phys 1979;5:1733-1740. 1961;84:460-479. 15. Kramer S, Lee KF. Complications of radiation therapy: the cen- 39. Schultheiss TE, Higgins EM, el-Mahdi AM. The latent period tral nervous system. Semin Roentgenol 1974;9:72-83. in clinical radiation myelopathy. Int J Radiat Oncol Biol Phys 16. Sheline GE, Wara WM, Smith V. Therapeutic irradiation and 1984;10:1109-1115. brain injury. Int J Radiat Oncol Biol Phys 1980;6:1215-1228. 40. Myers R, Rogers MA, Hornsey S. A reappraisal of the role of 17. Sheline GE. Irradiation injury of the human brain: a review of glial and vascular elements in the development of white mat- clinical experience. In Gilbert HA, Kagen AR (eds): Radiation ter necrosis in irradiated rat spinal cord. Br J Cancer 1986;7: Damage to the Nervous System. New York: Raven Press, 1980, 221-223. pp 39-58. 41. Phillips TL, Buschke F. Radiation tolerance of the thoracic spinal 18. Young DF, Posner JB, Chu F, et al. Rapid-course radiation ther- cord. Am J Roentgenol Rad Ther Nucl Med 1969;105:659-664. apy of cerebral metastases: results and complications. Cancer 42. Abbatucci JS, Delozier T, Quint R, et al. Radiation myelopathy 1974;34:1069-1076. of the cervical spinal cord: time, dose and volume factors. Int J 19. Jones A. Transient radiation myelopathy (with reference to Radiat Oncol Biol Phys 1978;4:239-248. Lhermitte’s sign of electrical paresthesia). Br J Radiol 1964;37: 43. Hatievoll R, Host H, Kaalhus O. Myelopathy following radio- 727-744. therapy of bronchial carcinoma with large single fractions: a ret- 20. Freeman JE, Johnston PGB, Voke JM. Somnolence after prophy- rospective study. Int J Radiat Oncol Biol Phys 1983;9:41-44. lactic cranial irradiation in children with acute lymphoblastic 44. Wara WM, Phillips TL, Sheline GE, et al. Radiation tolerance of leukaemia. Br Med J 1973;4:523-525. the spinal cord. Cancer 1975;35:1558-1562. 21. Boldrey E, Sheline G. Delayed transitory clinical manifestations 45. Cohen L, Creditor M. An isoeffect table for radiation tolerance after radiation treatment of intracranial tumors. Acta Radiol of the human spinal cord. Int J Radiat Oncol Biol Phys 1981;7: Ther Phys Biol 1966;5:5-10. 961-966. 22. Steen RG, Koury M, Granja CI, et al. Effect of ionizing radia- 46. Van der Kogel AJ. Radiation tolerance of the rat spinal cord: tion on the human brain: white matter and gray matter T1 in time-dose relationships. Radiology 1977;122:505-509. pediatric brain tumor patients treated with conformal radiation therapy. Int J Radiat Oncol Biol Phys 2001;49:79-91. 862 PART 10 Central Nervous System

47. Ang KK, van der Kogel AJ, van der Schueren E. Lack of evidence 71. Hart MN, Earle KM. Primitive neuroectodermal tumors of the for increased tolerance of rat spinal cord with decreasing fraction brain in children. Cancer 1973;32:890-898. doses below 2 Gy. Int J Radiat Oncol Biol Phys 1985;11:105-110. 72. Grotzer MA, Janss AJ, Fung JA, et al. TrkC expression predicts 48. Dische S, Saunders MI. Continuous, hyperfractionated, acceler- good clinical outcome in primitive neuroectodermal brain ated radiotherapy (CHART): an interim report upon late mor- tumors. J Clin Oncol 2000;18:1027-1035. bidity. Radiother Oncol 1989;16:67-74. 73. McLendon RE, Friedman HS, Fuchs HE, et al. Diagnostic mark- 49. Dische S. Accelerated treatment and radiation myelitis. Radi- ers in paediatric medulloblastoma: a Pediatric Oncology Group other Oncol 1991;20:1-2. study. Histopathology 1999;34:154-162. 50. Norrell H, Wilson CB, Slagel DE, et al. Leukoencephalopathy 74. Mendrzyk F, Radlwimmer B, Joos S, et al. Genomic and protein following the administration of methotrexate into the cerebro- expression profiling identifies CDK6 as novel independent spinal fluid in the treatment of primary brain tumors. Cancer prognostic marker in medulloblastoma. J Clin Oncol 2005;34: 1974;33:923-932. 8853-8862. 51. Lee JS, Umsawasdi T, Lee Y-Y, et al. Neurotoxicity in long-term 75. Aldosari N, Bigner SH, Burger PC, et al. MYCC and MYCN survivors of small cell lung cancer. Int J Radiat Oncol Biol Phys oncogene amplification in medulloblastoma. A fluorescence in 1986;12:313-321. situ hybridization study on paraffin sections from the Children’s 52. Bleyer WA. Neurologic sequelae of methotrexate and ioniz- Oncology Group. Arch Pathol Lab Med 2002;126:540-544. ing radiation: a new classification. Cancer Treatment Reports 76. Bobola MS, Finn LS, Ellenbogen RG, et al. Apurinic/apyrimidinic 1981;65:89-98. endonuclease activity is associated with response to radiation 53. Mulhern RK, Ochs J, Kun LE. Changes in intellect associated and chemotherapy in medulloblastoma and primitive neuroec- with cranial radiation therapy. In Gutin PH, Leibel SA, Sheline todermal tumors. Clin Cancer Res 2005;11:7405-7414. GE (eds): Radiation Injury to the Nervous System. New York: 77. Clifford SC, Lusher ME, Lindsey JC, et al. Wnt/Wingless path- Raven Press, 1991, pp 325-340. way activation and chromosome 6 loss characterize a distinct 54. Mulhern RK, Kovnar E, Langston J, et al. Long-term survivors of molecular sub-group of medulloblastomas associated with a leukemia treated in infancy: factors associated with neuropsy- favorable prognosis. Cell Cycle 2006;5:2666-2670. chologic status. J Clin Oncol 1992;10:1095-1102. 78. Lamont JM, McManamy CS, Pearson AD, et al. Combined his- 55. Jannoun L, Bloom HJG. Long-term psychological effects in chil- topathological and molecular cytogenetic stratification of medul- dren treated for intracranial tumors. Int J Radiat Oncol Biol Phys loblastoma patients. Clin Cancer Res 2004;10:5482-5493. 1990;18:747-753. 79. Gajjar A, Hernan R, Kocak M, et al. Clinical, histopathologic and 56. Hochberg FH, Slotnick B. Neuropsychologic impairment in molecular markers of prognosis: toward a new disease risk strati- astrocytoma survivors. Neurology 1980;30:172-177. fication system for medulloblastoma. J Clin Oncol 2004;22: 57. Maire J, Coudin B, Guerin J, et al. Neuropsychologic impairment 971-974. in adults with brain tumors. Am J Clin Oncol 1987;10: 80. Gajjar A, Fouladi M, Walter AW, et al. Comparison of lumbar 156-162. and shunt cerebrospinal fluid specimens for cytologic detection 58. Kun LE, Mulhern RK, Crisco JJ. Quality of life in children treated of leptomeningeal disease in pediatric patients with brain for brain tumors: intellectual, emotional, and academic function. tumors. J Clin Oncol 1999;17:1825-1828. J Neurosurg 1983;58:1-6. 81. Kleinman GM, Hockberg FH, Richardson EP Jr. Systemic 59. Radcliffe J, Packer RJ, Atkins TE, et al. Three- and four-year cog- metastases from medulloblastoma: report of two cases and nitive outcome in children with noncortical brain tumors treated review of the literature. Cancer 1981;48:2296-2309. with whole-brain radiotherapy. Ann Neurol 1992;32:551-554. 82. Albright AL, Wisoff JH, Zeltzer PM, et al. Effects of medullob- 60. Mulhern RK, Kepner JL, Thomas PRM, et al. Neuropsycho- lastoma resections on outcome in children: a report from the logical functioning of survivors of childhood medulloblastoma Children’s Cancer Group. Neurosurgery 1996;38:265-271. randomized to receive conventional (3,600 cGy/20) or reduced 83. Cochrane DD, Gustavsson B, Poskitt KP, et al. The surgical and (2,340/13) dose craniospinal irradiation: a Pediatric Oncology natural morbidity of aggressive resection for posterior fossa Group study. J Clin Oncol 1998;16:1723-1728. tumors in childhood. Pediatr Neurosurg 1994;20:19-29. 61. Mulhern RK, Reddick WE, Palmer SL, et al. Neurocognitive defi- 84. Aguiar PH, Plese JP, Ciquini O, et al. Transient mutism follow- cits in medulloblastoma survivors and white matter loss. Ann ing a posterior fossa approach to cerebellar tumors in children: Neurol 1999;46:834-841. a critical review of the literature. Childs Nerv Syst 1995;11: 62. Craig JB, Jackson DV, Moody D, et al. Prospective evaluation of 306-310. changes in computed cranial tomography in patients with small 85. Chang CH, Housepian EM, Herbert C Jr. An operative staging cell lung carcinoma treated with chemotherapy and prophylac- system and a megavoltage radiotherapeutic technic for cerebel- tic cranial irradiation. J Clin Oncol 1984;2:1151-1156. lar medulloblastoma. Radiology 1969;93:1351-1359. 63. Urie MM, Fullerton B, Tatsuzaki H, et al. A dose response analysis 86. Kun LE, Constine LS. Medulloblastoma: caution regarding new of injury to cranial nerves and/or nuclei following proton beam treatment approaches. Int J Radiat Oncol Biol Phys 1991;20: radiation therapy. Int J Radiat Oncol Biol Phys 1992;23:27-29. 897-899. 64. Parsons JT, Bova FJ, Fitzgerald CR, et al. Radiation optic neu- 87. Packer RJ, Sutton LN, Elterman R, et al. Outcome for children ropathy after megavoltage external-beam irradiation: analysis with medulloblastoma treated with radiation and cisplatin, of time-dose factors. Int J Radiat Oncol Biol Phys 1994;30: CCNU, and vincristine chemotherapy. J Neurosurg 1994;81: 755-763. 690-698. 65. Harris JR, Levene MB. Visual complications following irradia- 88. Heideman RL, Kovnar EH, Kellie SJ, et al. Preirradiation chemo- tion for pituitary adenomas and craniopharyngiomas. Radiology therapy with carboplatin and etoposide in newly diagnosed 1976;120:167-171. embryonal pediatric CNS tumors. J Clin Oncol 1995;13: 66. Svensson H, Westling P, Larsson LG. Radiation-induced lesions 2247-2254. of the bronchial plexus correlated to the dose-time-fraction 89. Kovnar EH, Kellie SJ, Horowitz ME, et al. Pre-irradiation cispla- schedule. Acta Radiol 1975;14:228-238. tin and etoposide in the treatment of high-risk medulloblastoma 67. Jemal A, Siegel R, Ward E, et al. Cancer statistics, 2007. CA and other malignant embryonal tumors of the central nervous Cancer J Clin 2007;57:43-66. system: a phase II study. J Clin Oncol 1990;8:330-336. 68. Linabery AM, Ross JA. Trends in cancer incidence among chil- 90. Bailey CC, Gnekow A, Wellek S, et al. Prospective randomized dren in the U.S. (1992-2004). Cancer 2008;112:416-432. trial of chemotherapy given before radiotherapy in childhood 69. Kleihues P, Cavenee WK (eds). Tumours of the Nervous System. medulloblastoma. International Society of Paediatric Oncol- Lyon, France: IARC Press, 2000. ogy (SIOP) and the (German) Society of Paediatric Oncology 70. Rorke L. The cerebellar medulloblastoma and its relationship to (GPO): SIOP II. Med Pediatr Oncol 1995;25:166-178. primitive neuroectodermal tumors. J Neuropathol Exp Neurol 1983;42:1-15. 33 The Brain and Spinal Cord 863

91. Hartsell WF, Gajjar A, Heideman RL, et al. Patterns of failure in 111. Ilgren EB, Stiller CA. Cerebellar astrocytomas. Clinical char- children with medulloblastoma: effects of pre-irradiation chem- acteristics and prognostic indices. J Neurooncol 1987;4: otherapy. Int J Radiat Oncol Biol Phys 1997;39:15-24. 293-308. 92. Thomas PR, Deutsch M, Kepner JL, et al. Low-stage medul- 112. Smoots DW, Geyer JR, Lieberman DM, et al. Predicting disease loblastoma: final analysis of trial comparing standard-dose progression in childhood cerebellar astrocytoma. Childs Nerv with reduced-dose neuraxis irradiation. J Clin Oncol 2000;18: Syst 1998;14:636-648. 3004-3011. 113. Tomita T, Chou P, Reyes-Mugica M. IV ventricle astrocytomas 93. Berry MP, Jenkin RDT, Keen CW, et al. Radiation treatment in childhood: clinicopathological features in 21 cases. Childs for medulloblastoma: a 21-year review. J Neurosurg 1981;55: Nerv Syst 1998;14:537-546. 43-51. 114. Pencalet P, Maixner W, Sainte-Rose C, et al. Benign cerebellar 94. Hughes EN, Winston K, Cassady JR, et al. Medulloblastoma: astrocytomas in children. J Neurosurg 1999;90:265-273. results of surgery and radical radiation therapy at the JCRT 115. Sgouros S, Fineron PW, Hockley AD. Cerebellar astrocytoma 1968-1984. Int J Radiat Oncol Biol Phys 1986;12:179. of childhood: long-term follow-up. Childs Nerv Syst 1995;11: 95. Bloom HJG, Glees J, Bell J. The treatment and long-term prog- 89-96. nosis of children with intracranial tumors: a study of 610 cases, 116. Burger PC, Scheithauer BW, Vogel FS (eds). Surgical Pathology 1950-1981. Int J Radiat Oncol Biol Phys 1990;18:723-745. of the Nervous System and its Coverings, 3rd ed. New York: 96. Jenkin D, Goddard K, Armstrong D, et al. Posterior fossa medul- Churchill Livingstone, 1991. loblastoma in childhood: treatment results and a proposal for 117. Garcia DM, Marks JE, Latifi HR, et al. Childhood cerebellar a new staging system. Int J Radiat Oncol Biol Phys 1990;19: astrocytomas: is there a role for postoperative irradiation? Int J 265-274. Radiat Oncol Biol Phys 1990;18:815-818. 97. Packer RJ, Goldwein J, Nicholson HS, et al. Treatment of chil- 118. Schneider JH, Raffel C, McComb JG. Benign cerebellar astrocy- dren with medulloblastomas with reduced-dose craniospinal ra- tomas of childhood. Neurosurgery 1992;30:58-63. diation therapy and adjuvant chemotherapy: a Children’s Can- 119. Ilgren EB, Stiller CA. Cerebellar astrocytomas: therapeutic man- cer Group study. J Clin Oncol 1999;17:2127-2136. agement. Acta Neurochir 1986;81:11-26. 98. Miralbell R, Lomax A, Cella L, et al. Potential reduction of the 120. Larson DA, Wara WM, Edwards MSB. Management of child- incidence of radiation-induced second cancers by using proton hood cerebellar astrocytoma. Int J Radiat Oncol Biol Phys beams in the treatment of pediatric tumors. Int J Radiat Oncol 1980;18:971-973. Biol Phys 2002;54:824-829. 121. Hirsch J-F, Rose CS, Pierre-Kahn A, et al. Benign astrocytic and 99. Fukunaga-Johnson N, Sandler HM, Marsh R, et al. The use of oligodendrocytic tumors of the cerebral hemispheres in chil- 3D conformal radiotherapy (3D CRT) to spare the cochlea in dren. J Neurosurg 1989;70:568-572. patients with medulloblastoma. Int J Radiat Oncol Biol Phys 122. Hoffman HJ, Soloniuk DS, Humphreys RP, et al. Management 1998;41:77-82. and outcome of low-grade astrocytomas of the midline in chil- 100. Merchant TE, Happersett L, Finlay J, et al. Preliminary results of dren: a retrospective review. Neurosurgery 1993;33:964-971. conformal radiation therapy for medulloblastoma. Neurooncol- 123. Reardon D, Gajjar A, Walter AW, et al. Bithalamic involvement ogy 1999;1:177-187. predicts poor outcome among children with thalamic glial 101. Fukunaga-Johnson N, Lee JH, Sandler HM, et al. Patterns of fail- tumors. Pediatr Neurosurg 1998;29:29-35. ure following treatment for medulloblastoma: is it necessary to 124. Krouwer HGJ, Prados MD. Infiltrative astrocytomas of the tha- treat the entire posterior fossa? Int J Radiat Oncol Biol Phys lamus. J Neurosurg 1995;82:548-557. 1998;42:143-146. 125. Albright AL, Price RA, Guthkelch AN. Diencephalic gliomas of 102. Strother D, Ashley D, Kellie SJ, et al. Feasibility of four consecu- children: a clinicopathologic study. Cancer 1985;55:2789-2793. tive high-dose chemotherapy cycles with stem-cell rescue for 126. Cairncross JG, Laperriere NJ. Low-grade glioma: to treat or not patients with newly diagnosed medulloblastoma or supratentorial to treat? Arch Neurol 1989;46:1238-1239. primitive neuroectodermal tumor after craniospinal radio- 127. Bernstein M, Hoffman HJ, Halliday WC, et al. Thalamic tumors therapy: results of a collaborative study. J Clin Oncol 2001;19: in children: long-term follow-up and treatment guidelines. 2696-2704. J Neurosurg 1984;61:649-656. 103. Gajjar A, Chintagumpala M, Ashley D, et al. Risk-adapted 128. Kelly PJ. Stereotactic biopsy and resection of thalamic astrocyto- craniospinal radiotherapy followed by high-dose chemotherapy mas. Neurosurgery 1989;25:185-194. and stem-cell rescue in children with newly diagnosed medul- 129. Leibel SA, Sheline GE. Review article: radiation therapy for neo- loblastoma (St Jude Medulloblastoma-96): long-term results plasms of the brain. J Neurosurg 1987;66:1-22. from a prospective, multicentre trial. Lancet Oncol 2006;7: 130. Boydston WR, Sanford RA, Muhlbauer MS, et al. Gliomas of 813-820. the tectum and periaqueductal region of the mesencephalon. 104. Gajjar A, Sanford RA, Heideman R, et al. Low-grade astrocy- Pediatr Neurosurg 1991;17:234-238. toma: a decade of experience at St. Jude Children’s Research 131. Medbery CA, Straus KL, Steinberg SM, et al. Low-grade astro- Hospital. J Clin Oncol 1997;15:2792-2799. cytomas: treatment results and prognostic variables. Int J Radiat 105. Leibel SA, Sheline GE, Wara WM, et al. The role of radiation Oncol Biol Phys 1988;15:837-841. therapy in the treatment of astrocytomas. Cancer 1975;35: 132. Whitton AC, Bloom HJG. Low-grade glioma of the cerebral 1551-1557. hemispheres in adults: a retrospective analysis of 88 cases. Int J 106. Morantz RA. Radiation therapy in the treatment of cerebral Radiat Oncol Biol Phys 1990;18:783-786. astrocytoma. Neurosurgery 1987;20:975-982. 133. Leighton C, Fisher B, Bauman G, et al. Supratentorial low-grade 107. Wallner KE, Gonzales MF, Edwards MSB, et al. Treatment glioma in adults: an analysis of prognostic factors and timing of results of juvenile pilocytic astrocytoma. J Neurosurg 1988;69: radiation. J Clin Oncol 1997;15:1294-1301. 171-176. 134. Shaw EG, Daumas-Duport C, Scheithauer BW, et al. Radiation 108. Laws ER, Taylor WF, Clifton MB, et al. Neurosurgical manage- therapy in the management of low-grade supratentorial astrocy- ment of low-grade astrocytoma of the cerebral hemispheres. tomas. J Neurosurg 1989;70:853-861. J Neurosurg 1984;61:665-673. 135. Karim AB, Maat B, Hatlevoll R, et al. A randomized trial of 109. Krieger MD, Gonzalez-Gomez I, Levy ML, et al. Recurrence dose-response in radiation therapy of low-grade cerebral glioma: patterns and anaplastic change in a long-term study of pilocytic EORTC study 22844. Int J Radiat Oncol Biol Phys 1996;36: astrocytomas. Pediatr Neurosurg 1997;27:1-11. 549-556. 110. Roessler K, Bertalanffy A, Jezan H, et al. Proliferative activity 136. Dunbar SF, Tarbell NJ, Kooy HM, et al. Stereotactic radiotherapy as measured by MIB-1 labeling index and long-term outcome for pediatric and adult brain tumors: preliminary report. Int J of cerebellar juvenile pilocytic astrocytomas. J Neurooncol Radiat Oncol Biol Phys 1994;30:531-539. 2002;58:141-146. 864 PART 10 Central Nervous System

137. Listernick R, Louis DN, Packer RJ, et al. Optic pathway glio- 160. Shaw E, Arusell R, Scheithauer B, et al. Prospective randomized mas in children with neurofibromatosis 1: consensus statement trial of low- versus high-dose radiation therapy in adults with from the NF1 Optic Pathway Gliomas Task Force. Ann Neurol supratentorial low-grade glioma: initial report of a North Cen- 1997;41:143-149. tral Cancer Treatment Group/Radiation Therapy Oncology 138. Horwich A, Bloom HJG. Optic gliomas: radiation therapy and Group/Eastern Cooperative Oncology Group Study. J Clin On- prognosis. Int J Radiat Oncol Biol Phys 1985;11:1067-1079. col 2002;20:2367-2376. 139. Alvord EC, Lofton S. Gliomas of the optic nerve or chiasm: out- 161. Pollack IF, Claassen D, Al-Shboul Q, et al. Low grade gliomas come by patient’s age, tumor site, and treatment. J Neurosurg of the cerebral hemispheres in children: an analysis of 71 cases. 1988;68:85-98. J Neurosurg 1995;82:536-547. 140. Hoyt WF, Baghdassarian SA. Optic glioma of childhood: natural 162. McLaurin RL, Breneman J, Aron B. Hypothalamic gliomas: re- history and rationale for conservative management. Br J Oph- view of 18 cases. Concepts Pediatr Neurosurg 1987;7:19. thalmol 1969;53:793-798. 163. Albright AL, Price RA, Guthkelch AN. Brainstem gliomas of 141. Imes RK, Hoyt WF. Childhood chiasmal gliomas: update on the children: a clinicopathologic study. Cancer 1983;52:2313-2319. fate of patients in the 1969 San Francisco study. Br J Ophthal- 164. Albright AL, Guthkelch AN, Packer RJ, et al. Prognostic factors mol 1986;70:179-182. in pediatric brainstem gliomas. J Neurosurg 1986;65:751-755. 142. Wong JYC, Wara W, Sheline GE. Optic gliomas: a reanalysis of 165. Barkovich AJ, Krischer J, Kun LE, et al. Brain stem gliomas: a the University of California, San Francisco, experience. Cancer classification system based on magnetic resonance imaging. Pedi- 1987;60:1847-1855. atr Neurosurg 1990;16:73-83. 143. Jenkin D, Angyalfi S, Becker L, et al. Optic glioma in children: 166. Hoffman HJ, Becker I, Craven MA. A clinically and pathologi- surveillance, resection, or irradiation? Int Radiat Oncol Biol Phys cally distinct group of benign brainstem gliomas. Neurosurgery 1993;25:215-225. 1980;7:243-248. 144. Bataini JP, Delanian S, Ponvert D. Chiasmal gliomas: results of 167. Stroink AR, Hoffman HJ, Hendrick EB, et al. Diagnosis and irradiation management in 57 patients and review of literature. management of pediatric brainstem gliomas. J Neurosurg Int J Radiat Oncol Biol Phys 1991;21:615-623. 1986;65:745-750. 145. Tao ML, Barnes PD, Billett AL, et al. Childhood optic chiasm 168. Khatib ZA, Heideman RL, Kovnar EH, et al. Predominance gliomas: radiographic response following radiotherapy and long- of pilocytic histology in dorsally exophytic brain stem tumors. term clinical outcome. Int J Radiat Oncol Biol Phys 1997;39: Pediatr Neurosurg 1994;20:2-10. 579-587. 169. Fisher PG, Breiter SN, Carson BS, et al. A clinicopathologic 146. Fletcher WA, Imes RK, Hoyt WF. Chiasmal gliomas: appearance reappraisal of brain stem tumor classification. Identification of and long-term changes demonstrated by computerized tomog- pilocytic astrocytoma and fibrillary astrocytoma as distinct enti- raphy. J Neurosurg 1986;65:154-159. ties. Cancer 2000;89:1569-1576. 147. Wisoff JH, Abbott R, Epstein F. Surgical management of exo- 170. Mantravadi RVP, Phatak R, Bellur S, et al. Brainstem gliomas: an phytic chiasmatic-hypothalamic tumors of childhood. J Neuro- autopsy study of 25 cases. Cancer 1982;49:1294-1296. surg 1990;73:661-667. 171. Schulz-Ertner D, Debus J, Lohr F, et al. Fractionated stereotactic 148. Packer RJ, Lange B, Ater J, et al. Carboplatin and vincristine for conformal radiation therapy of brain stem gliomas: outcome and recurrent and newly diagnosed low-grade gliomas of childhood. prognostic factors. Radiother Oncol 2000;57:215-223. J Clin Oncol 1993;11:850-856. 172. Shrieve DC, Wara WM, Edwards MSB, et al. Hyperfractionated 149. Packer RJ, Ater J, Allen J, et al. Carboplatin and vincristine radiation therapy for gliomas of the brainstem in children and in chemotherapy for children with newly diagnosed progressive adults. Int J Radiat Oncol Biol Phys 1992;24:599-610. low-grade gliomas. J Neurosurg 1997;86:747-754. 173. Freeman CR, Krischer JP, Sanford RA, et al. Final results of a 150. Mahoney DH, Cohen ME, Friedman HS, et al. Carboplatin is ef- study of escalating doses of hyperfractionated radiotherapy in fective therapy for young children with progressive optic path- brain stem tumors in children: a Pediatric Oncology Group way tumors: a Pediatric Oncology Group phase II study. Neu- study. Int J Radiat Oncol Biol Phys 1993;27:197-206. rooncology 2000;2:213-220. 174. Freeman CR, Bourgouin PM, Sanford RA, et al. Long term sur- 151. Somaza SC, Kondziolka D, Lunsford LD, et al. Early outcomes vivors of childhood brain stem gliomas treated with hyperfrac- after stereotactic radiosurgery for growing pilocytic astrocyto- tionated radiotherapy, clinical characteristics and treatment re- mas in children. Pediatr Neurosurg 1996;25:109-115. lated toxicities. Cancer 1996;77:555-562. 152. Grabb PA, Lunsford LD, Albright AL, et al. Stereotactic 175. Mandell LR, Kadota R, Freeman C, et al. There is no role for radiosurgery for glial neoplasms of childhood. Neurosurgery hyperfractionated radiotherapy in the management of children 1996;38:696-702. with newly diagnosed diffuse intrinsic brainstem tumors: results 153. Pollack IF, Hurtt M, Pang D, et al. Dissemination of low grade of a Pediatric Oncology Group phase III trial comparing conven- astrocytomas in children. Cancer 1994;73:2869-2878. tional vs. hyperfractionated radiotherapy. Int J Radiat Oncol Biol 154. Prados M, Mamelak AN. Metastasizing low grade gliomas in chil- Phys 1999;43:959-964. dren. Redefining an old disease. Cancer 1994;73:2671-2673. 176. Freeman CR, Kepner J, Kun LE, et al. A detrimental effect of 155. Gajjar A, Bhargava R, Jenkins JJ, et al. Low-grade astrocy- a combined chemotherapy-radiotherapy approach in children toma with neuraxis dissemination at diagnosis. J Neurosurg with diffuse intrinsic brain stem gliomas? Int J Radiat Oncol 1995;83:67-71. Biol Phys 2000;47:561-564. 156. Kepes JJ, Rubinstein LJ, Eng LF. Pleomorphic xanthoastrocy- 177. Tait DM, Thornton-Jones H, Bloom HJG, et al. Adjuvant chem- toma: a distinctive meningocerebral glioma of young subjects otherapy for medulloblastoma: the first multicenter control trial with relatively favorable prognosis: a study of 12 cases. Cancer of the International Society of Paediatric Oncology (SIOP I). 1979;44:1839-1852. Eur J Cancer Clin Oncol 1990;26:464-469. 157. Taratuto AL, Monges J, Lylyk P, et al. Superficial cerebral astro- 178. Duffner PK, Horowitz ME, Krischer JP, et al. Postoperative cytoma attached to dura. Cancer 1984;54:2505-2512. chemotherapy and delayed radiation in children less than 158. Fouladi M, Jenkins J, Burger P, et al. Pleomorphic xanthoastro- three years of age with malignant brain tumors. N Engl J Med cytoma (PXA): favorable outcome following complete surgical 1993;328:1725-1731. resection. Neuro Oncol 2001;3:184-192. 179. Ilgren EB, Stiller CA, Hughes JT, et al. Ependymomas: a clinical 159. Merchant TE, Zhu Y, Thompson SJ, et al. Preliminary and pathologic study, part 1: biologic features. Clin Neuropathol results from a phase II trial of conformal radiation therapy 1984;3:113-121. for pediatric patients with localized low-grade astrocytomas 180. Sanford RA, Kun LE, Heideman RL, et al. Cerebellar pontine and ependymoma. Int J Radiat Oncol Biol Phys 2002;52: angle ependymoma in infants. Pediatr Neurosurg 1997;27: 325-332. 84-91. 33 The Brain and Spinal Cord 865

181. Bloom HJG. Intracranial tumors: response and resistance to 204. Ho DM, Lieu H-C. Primary intracranial germ cell tumor. Patho- therapeutic endeavors. 1970-1980. Int J Radiat Oncol Biol Phys logic study of 51 patients. Cancer 1992;70:1577-1584. 1982;8:1083-1113. 205. Dernaley DP, A’Hern RP, Whittaker S, et al. Pineal and CNS 182. Vanuystel LJ, Bessell EM, Ashley SE, et al. Intracranial ependy- germ cell tumors: Royal Marsden Hospital experience 1962- moma: long-term results of a policy of surgery and radiotherapy. 1987. Int J Radiat Oncol Biol Phys 1990;18:773-781. Int J Radiat Oncol Biol Phys 1982;23:313-319. 206. Wolden SL, Wara WM, Larson DA, et al. Radiation therapy for 183. Goldwein JW, Glauser TA, Packet RJ, et al. Recurrent intracra- primary intracranial germ-cell tumors. Int J Radiat Oncol Biol nial ependymomas in children: survival, patterns of failure, and Phys 1995;32:943-949. prognostic factors. Cancer 1990;66:557-563. 207. Borit A, Blackwood W, Mair WGP. The separation of pineocy- 184. Rezai AR, Woo HH, Lee M, et al. Disseminated ependymo- toma from pineoblastoma. Cancer 1980;45:1408-1418. mas of the central nervous system. J Neurosurg 1996;85: 208. DiSclafani A, Hudgins RJ, Edwards MSB, et al. Pineocytomas. 618-624. Cancer 1989;63:302-304. 185. Centeno RS, Lee AA, Winter J, et al. Supratentorial ependymo- 209. Schild SE, Scheithauer BW, Schomberg PJ, et al. Pineal paren- mas. Neuroimaging and clinicopathological correlation. J Neu- chymal tumors. Clinical, pathologic, and therapeutic aspects. rosurg 1986;64:209-215. Cancer 1993;72:870-880. 186. Rawlings CE III, Giangaspero F, Burger PC, et al. Ependymomas: 210. Schild SE, Haddock MG, Scheithauer BW, et al. Nongermino- a clinicopathologic study. Surg Neurol 1988;29:271-281. matous germ cell tumors of the brain. Int J Radiat Oncol Biol 187. Rorke LB. Relationship of morphology of ependymoma in chil- Phys 1996;36:557-563. dren to prognosis. Prog Exp Tumor Res 1987;30:170-174. 211. Shibamoto Y, Takahashi M, Sasai K. Prognosis of intracranial ger- 188. Ross GW, Rubinstein LJ. Lack of histopathological correlation minoma with syncytiotrophoblastic giant cells treated by radia- of malignant ependymomas with postoperative survival. J Neu- tion therapy. Int J Radiat Oncol Biol Phys 1997;37:505-510. rosurg 1989;70:31-36. 212. Jennings MT, Gelman R, Hochberg F. Intracranial germ-cell 189. Rousseau P, Habrand JL, Sarrazin D, et al. Treatment of intracra- tumors: natural history and pathogenesis. J Neurosurg 1985;63: nial ependymomas of children: review of a 15-year experience. 155-167. Int J Radiat Oncol Biol Phys 1994;28:381-386. 213. Jenkin D, Berry M, Chan H, et al. Pineal region germinomas 190. Pollack IF, Gerszten PC, Martinez AJ, et al. Intracranial ependy- in childhood: treatment considerations. Int J Radiat Oncol Biol momas of childhood: long-term outcome and prognostic factors. Phys 1990;18:541-545. Neurosurgery 1995;37:655-667. 214. Sung D, Harisiadis L, Chang CG. Midline pineal tumors and 191. Healey EA, Braners PD, Kupsky WJ, et al. The prognostic signifi- suprasellar germinomas: highly curable by irradiation. Radiology cance of postoperative residual tumor in ependymoma. Neuro- 1978;128:745-751. surgery 1991;28:666-672. 215. Haddock MG, Schild SE, Scheithauer BE, et al. Radiation ther- 192. Nazar GB, Hoffman HJ, Becker LE, et al. Infratentorial ependy- apy for histologically confirmed primary central nervous system momas in childhood: prognostic factors and treatment. J Neuro- germinoma. Int J Radiat Oncol Biol Phys 1997;38:915-923. surg 1990;72:408-417. 216. Shibamoto Y, Takahashi M, Abe M. Reduction of the radiation 193. Hukin J, Epstein F, Lefton D, et al. Treatment of intracranial dose for intracranial germinoma: a prospective study. Br J Can- ependymoma by surgery alone. Pediatr Neurosurg 1998;29: cer 1994;70:984-989. 40-45. 217. Shirato H, Nishio M, Sawamura Y, et al. Analysis of long-term 194. Duffner PK, Krischer JP, Sanford RA, et al. Prognostic factors in treatment of intracranial germinoma. Int J Radiat Oncol Biol infants and very young children with intracranial ependymomas. Phys 1997;37:511-515. Pediatr Neurosurg 1998;28:215-222. 218. Merchant TE, Sherwood SH, Mulhern RK, et al. CNS germi- 195. Lefkowitz I, Evans A, Sposto R, et al. Adjuvant chemotherapy noma: disease control and long-term functional outcome for of childhood posterior fossa (PF) ependymoma: craniospinal ir- 12 children treated with craniospinal irradiation. Int J Radiat radiation with or without CCNU, vincristine (VCR) and pred- Oncol Biol Phys 2000;46:1171-1176. nisone (P) [abstract]. Proc Am Soc Clin Oncol 1989;8:87a. 219. Hardenbergh PH, Golden J, Billet A, et al. Intracranial germi- 196. Strother D, Kepner J, Aronin P, et al. Dose-intensive (DI) chem- noma: the case for lower dose radiation therapy. Int J Radiat otherapy (CT) prolongs event-free survival (EFS) for very young Oncol Biol Phys 1997;39:419-426. children with ependymoma (EP). Results of Pediatric Oncology 220. Calaminus G, Bamberg M, Baranzelli MC, et al. Intracranial Group (POG) Study 9233 [abstract]. Proc Am Soc Clin Oncol germ cell tumors: a comprehensive update of the European 2000;19. 585a, 2302. data. Neuropediatrics 1994;25:26-32. 197. Goldwein JW, Corn BW, Finlay JL, et al. Is craniospinal irradia- 221. Shibamoto Y, Abe M, Yamashita J, et al. Treatment results of tion required to cure children with malignant (anaplastic) in- intracranial germinoma as a function of the irradiation volume. tracranial ependymomas? Cancer 1991;67:2766-2771. Int J Radiat Oncol Biol Phys 1988;15:285-290. 198. Merchant TE, Haida T, Wang M-H., et al. Anaplastic ependymo- 222. Bamberg M, Kortmann R-D, Calaminus G, et al. Radiation ma: treatment of pediatric patients with or without craniospinal therapy for intracranial germinoma: results of the German Co- radiation therapy. J Neurosurg 1997;86:943-949. operative Prospective Trials MAKEI 83/86/89. J Clin Oncol 199. Kovnar EH, Curran W, Tomita T, et al. Hyperfractionated 1999;17:2585-2592. irradiation for childhood ependymoma: improved local control 223. Allen JC, Kim JH, Packer RJ. Neoadjuvant chemotherapy for in subtotally resected tumors [abstract]. Paper presented at the newly diagnosed germ cell tumors of the central nervous system. Eighth International Symposium on Pediatric Neuro-Oncology, J Neurosurg 1987;67:65-70. May 6-9, 1998, Rome, Italy. 224. Bouffet E, Baranzelli MC, Patte C, et al. Combined treatment 200. Aggarwal R, Yeung D, Muhlbauer M, et al. Efficacy and feasibil- modality for intracranial germinomas: results of a multicentre ity of stereotactic radiosurgery in the primary management of SFOP experience. Br J Cancer 1999;79:1199-1204. unfavorable pediatric ependymoma. Radiother Oncol 1997;43: 225. Balmaceda C, Heller G, Rosenblum M, et al. Chemotherapy 269-273. without irradiation. A novel approach for newly diagnosed CNS 201. Herrick MK. Pathology of pineal tumors. In Neuwelt EA (ed): germ cell tumors: results of an international cooperative trial. Diagnosis and Treatment of Pineal Region Tumors. Baltimore: J Clin Oncol 1996;14:2908-2915. Williams & Wilkins, 1984, pp 31-60. 226. Linstadt D, Wara WM, Edwards MSB, et al. Radiotherapy of pri- 202. Glenn OA, Barkovich AJ. Intracranial germ cell tumors: a com- mary intracranial germinomas: the case against routine craniospi- prehensive review of proposed embryonic deprivation. Pediatr nal irradiation. Int J Radiat Oncol Biol Phys 1988;15:291-297. Neurosurg 1996;24:242-251. 227. Maity A, Shu HK, Janss A, et al. Craniospinal radiation in the 203. Rich TA, Cassady JR, Strand RD, et al. Radiation therapy for treatment of biopsy-proven intracranial germinomas: twenty- pineal and suprasellar germ cell tumors. Cancer 1985;55: five years’ experience in a single center. Int J Radiat Oncol Biol 932-940. Phys 2004;58:1165-1170. 866 PART 10 Central Nervous System

228. Buckner JC, Peethambaram PP, Smithson WA, et al. Phase II 252. Brada M, Thomas DGT. Craniopharyngioma revisited. Int J trial of primary chemotherapy followed by reduced-dose radia- Radiat Oncol Biol Phys 1993;27:471-475. tion for CNS germ cell tumors. J Clin Oncol 1999;17:933-940. 253. Hetelididis S, Barnes PD, Tao ML, et al. Twenty year experience 229. Allen JC, DaRosso RC, Donahue B, et al. A phase II trial of preir- in childhood craniopharyngioma. Int J Radiat Oncol Biol Phys radiation carboplatin in newly diagnosed germinoma of the cen- 1993;27:189-195. tral nervous system. Cancer 1994;74:940-944. 254. Rajan B, Ashley S, Gorman C, et al. Craniopharyngioma—long 230. Robertson PL, DaRosso RC, Allen JC. Improved prognosis of term results following limited surgery and radiotherapy. Radi- intracranial non-germinoma germ cell tumors with multimodal- other Oncol 1993;26:1-10. ity therapy. J Neurooncol 1997;32:71-80. 255. Flickinger JC, Lunsford LD, Singer J, et al. Megavoltage external 231. Aoyama H, Shirato H, Yoshida H, et al. Retrospective multi- beam irradiation of craniopharyngiomas: analysis of tumor institutional study of radiotherapy for intracranial non-germino- control and morbidity. Int J Radiat Oncol Biol Phys 1990;19: matous germ cell tumors. Radiother Oncol 1998;49:55-59. 117-122. 232. Jakacki RI, Zeltzer PM, Boyett JM, et al. Survival and prognostic 256. Lunsford LD, Pollack BE, Kondziolka DS, et al. Stereotactic op- factors following radiation and/or chemotherapy for primitive tions in the management of craniopharyngioma. Pediatr Neuro- neuroectodermal tumors of the pineal region in infants and chil- surg 1994;21:90-97. dren: a report of the Children’s Cancer Group. J Clin Oncol 257. Backlund EO, Axelsson B, Bergstrand CG, et al. Treatment of 1995;13:1377-1383. craniopharyngiomas—the stereotactic approach in a ten to 233. Adamson TE, Wiestler OD, Kleihues P, et al. Correlation of clin- twenty-three years’ perspective. Surgical, radiological and oph- ical and pathological features in surgically treated craniopharyn- thalmological aspects. Acta Neurochir 1989;99(Pt 1):11-19. giomas. J Neurosurg 1990;73:12-17. 258. Baskin DS, Wilson CB. Surgical management of craniopharyngi- 234. Hoffman HJ, De Silva M, Humphreys RP, et al. Aggressive surgi- omas: a review of 74 cases. J Neurosurg 1986;65:22-27. cal management of craniopharyngiomas in children. J Neurosurg 259. Merchant TE, Kiehna EN, Sanford RA, et al. Craniopharyngi- 1992;76:47-52. oma: the St. Jude Children’s Research Hospital experience. Int J 235. Hoffman HJ. Surgical management of craniopharyngioma. Pedi- Radiat Oncol Biol Phys 2002;53:533-542. atr Neurosurg 1994;21:44-49. 260. Hoogenhout J, Otten BJ, Kazem I, et al. Surgery and radiation 236. Yarsargil MG, Curcic M, Kis S, et al. Total removal of craniophar- therapy in the management of craniopharyngiomas. Int J Radiat yngiomas: approaches and long-term results in 144 patients. Oncol Biol Phys 1984;10:2293-2297. J Neurosurg 1990;73:3-11. 261. Sung DI, Chang CH, Harisiadis L, et al. Treatment results in 237. Sanford RA. Craniopharyngioma: results of survey of the craniopharyngiomas. Cancer 1981;47:847-852. American Society of Pediatric Neurosurgery. Pediatr Neurosurg 262. Hart MN, Earle KM. Primitive neuroectodermal tumors of the 1994;21:39-43. brain in children. Cancer 1973;32:890-897. 238. Scott RM, Hetelekidis S, Barnes PD, et al. Surgery, radiation, and 263. Kosnik EJ, Boesel CP, Bau J, et al. Primitive neuroectodermal combination therapy in the treatment of childhood craniophar- tumors of the central nervous system in children. J Neurosurg yngioma—a 20-year experience. Pediatr Neurosurg 1994;21: 1978;48:741-746. 75-81. 264. Bennett JP, Rubinstein LJ. The biological behavior of primary 239. Wen B-C., Hussey DH, Staples J, et al. A comparison of the roles cerebral neuroblastoma: a reappraisal of the clinical course in a of surgery and radiation therapy in the management of crani- series of 70 cases. Ann Neurol 1984;16:21-27. opharyngiomas. Int J Radiat Oncol Biol Phys 1989;16:17-24. 265. Cruz-Sanchez FF, Rossi ML, Hughes JT, et al. Differentiation 240. Sweet WH. History of surgery for craniopharyngiomas. Pediatr in embryonal neuroepithelial tumors of the central nervous sys- Neurosurg 1994;21:28-38. tem. Cancer 1991;67:965-976. 241. Tomita T, McLone DG. Radical resections of childhood crani- 266. Berger MS, Edwards MSB, Wara WM, et al. Primary cerebral opharyngiomas. Pediatr Neurosurg 1993;19:6-14. neuroblastoma: long-term follow-up review and therapeutic 242. Yasargil MC, Curcic M, Kis S, et al. Total removal of craniophar- guidelines. J Neurosurg 1983;59:418-423. yngiomas: approaches and long-term results in 144 patients. 267. Gaffney CC, Sloane JP, Bradley NJ, et al. Primitive neuroec- J Neurosurg 1990;73:3-11. todermal tumors of the cerebrum: pathology and treatment. 243. Tomita T, McLone DG. Radical resections of childhood crani- J Neurooncol 1985;3:23-33. opharyngiomas. Pediatr Neurosurg 1993;19:6-14. 268. Berger MS, Edwards MSB, LaMasters D, et al. Pediatric brain- 244. Fischer EG, Welch K, Shillito J Jr, et al. Craniopharyngiomas in stem tumors: radiographic, pathological, and clinical correla- children: long-term effects of conservative surgical procedures tions. Neurosurgery 1983;12:298-302. combined with radiation therapy. J Neurosurg 1990;73: 269. Horten BC, Rubinstein LJ. Primary cerebral neuroblastoma: a 534-540. clinicopathological study of 35 cases. Brain 1976;99:735-756. 245. Fischer EG, Welch K, Shilito J Jr, et al. Craniopharyngiomas 270. Ashwal S, Hinshaw DB Jr, Bedros A. CNS primitive neuroec- in children. Long-term effects of conservative surgical proce- todermal tumors of childhood. Med Pediatr Oncol 1984;12: dures combined with radiation therapy. J Neurosurg 1990;73: 180-188. 534-540. 271. Zeltzer PM, Boyett JM, Finlaye JL, et al. Metastasis stage, ad- 246. Sklar CA. Craniopharyngioma: endocrine sequelae of treatment. juvant treatment, and residual tumor are prognostic factors for Pediatr Neurosurg 1994;21:120-123. medulloblastoma in children: conclusions from the Children’s 247. Curtis J, Daneman D, Hoffman HJ, et al. The endocrine outcome Cancer Group 921 randomized phase III study. J Clin Oncol after surgical removal of craniopharyngiomas. Pediatr Neurosurg 1999;17:832-845. 1994;21:24-27. 272. Stiller CA, Bunch KJ. Brain and spinal tumors in children aged 248. Kramer S, Southard M, Mansfield CM. Radiotherapy in the under two years: incidence and survival in Britain, 1971-1985. management of craniopharyngiomas. AJR Am J Roentgenol Br J Cancer 1992;66:S50-S53. 1968;103:44-52. 273. Cohen BH, Packer RJ, Siegel KR, et al. Brain tumors in children 249. Kramer S, McKissock W, Concannon JP. Craniopharyngiomas: under 2 years: treatment, survival and long-term prognosis. Pedi- treatment by combined surgery and radiation therapy. J Neuro- atr Neurosurg 1993;19:171-179. surg 1961;18:217-226. 274. Brown K, Mapstone TB, Oakes WJ. A modern analysis of intrac- 250. Kramer S, Southard M, Mansfield CM. Radiotherapy in the ranial tumors of infancy. Pediatr Neurosurg 1997;26:25-32. management of craniopharyngiomas: further experience and 275. Gilles FH, Sobel EL, Tavare CJ, et al. Age-related changes in late results. Am J Roentgenol Radium Ther Nucl Med 1968;103: diagnoses, histological features, and survival in children with 44-52. brain tumors: 1930-1979. Neurosurgery 1995;37:1056-1068. 251. Regine WF, Kramer S. Pediatric craniopharyngiomas: long-term results of combined treatment with surgery and radiation. Int J Radiat Oncol Biol Phys 1992;24:611-617. 33 The Brain and Spinal Cord 867

276. Rorke LB, Packer RJ, Biegel JA, et al. Central nervous system 296. Hilden JM, Watterson J, Longee DC, et al. Central nervous atypical teratoid/rhabdoid tumors of infancy and childhood: system atypical teratoid tumor/rhabdoid tumor: response to definition of an entity. J Neurosurg 1996;85:56-65. intensive therapy and review of the literature. J Neurooncol 277. Burger PC, Yu I-T, Tihan T, et al. Atypical teratoid/rhabdoid tu- 1998;40:265-275. mor of the central nervous system: a highly malignant tumor 297. Kernohan JW, Sayre GP. Tumors of the Central Nervous Sys- of infancy and childhood frequently mistaken for medullob- tem. Washington, DC: American Registry of Pathology, Armed lastoma. A Pediatric Oncology Group study. Am J Surg Pathol Forces Institute of Pathology, 1952. 1998;22:1083-1092. 298. Nelson JS, Taukada Y, Schoenfeld D, et al. Necrosis as a prog- 278. Pancelet P, Sainte-Rose C, Lellough-Tubiana A, et al. Papillomas nostic criterion in malignant supratentorial, astrocytic gliomas. and carcinomas of the choroid plexus in children. J Neurosurg Cancer 1983;52:550-554. 1998;88:521-528. 299. Werner-Wasik M, Scott CB, Nelson DF, et al. Final report of 279. Weiss E, Behring B, Behnke J, et al. Treatment of primary malig a phase I/II trial of hyperfractionated and accelerated hyper- nant rhabdoid tumor of the brain: report of three cases and re- fractionated radiation therapy with carmustine for adults with view of the literature. Int J Radiat Oncol Biol Phys 1998;41: supratentorial malignant gliomas. Radiation Therapy Oncology 1013-1019. Group study 83-02. Cancer 1996;77:1535-1543. 280. Taratuto AL, Monges J, Lylyk P, et al. Superficial cerebral astro- 300. Burger PC, Heinz ER, Shibata T, et al. Topographic anatomy cytoma attached to dura: report of six cases in infants. Cancer and CT correlations in the untreated glioblastoma multiforme. 1984;54:2505-2512. J Neurosurg 1988;68:698-704. 281. Mulhern RK, Hancock J, Fairclough D, et al. Neuropsychologi- 301. Halperin EC, Bentel G, Heinz ER, et al. Radiation therapy treat- cal status of children treated for brain tumors: a critical review ment planning in supratentorial glioblastoma multiforme: an and integrative analysis. Med Pediatr Oncol 1992;20:181-191. analysis based on postmortem topographic anatomy with CT 282. Evans AE, Jenkin RDT, Sposto R, et al. The treatment of medul- correlations. Int J Radiat Oncol Biol Phys 1989;17:1347-1350. loblastoma. Results of a prospective randomized trial of radiation 302. Kelly PJ, Daumas-Duport C, Kispert DB, et al. Imaging-based therapy with and without CCNU, vincristine, and prednisone. stereotaxic serial biopsies in untreated intracranial glial neo- J Neurosurg 1990;72:572-582. plasms. J Neurosurg 1987;66:865-874. 283. Deutsch M. Radiotherapy for primary brain tumors in very 303. Choucair AK, Levin VA, Gutin PH, et al. Development of mul- young children. Cancer 1982;50:2785-2789. tiple lesions during radiation therapy and chemotherapy in pa- 284. Jenkin D, Greenberg M, Hoffman H, et al. Brain tumors in chil- tients with gliomas. J Neurosurg 1986;65:654-658. dren: long-term survival after radiation therapy. Int J Radiat 304. Kopelson G, Linggood R. Infratentorial glioblastoma: the role Oncol Biol Phys 1995;31:445-451. of neuraxis irradiation. Int J Radiat Oncol Biol Phys 1982;8: 285. Allen JC, Helson L, Jereb B. Preradiation chemotherapy for just 999-1003. this minute diagnosed childhood brain tumors: a modified phase II 305. Nelson DF, Nelson JS, Davis DR, et al. Survival and prognosis of trial. Cancer 1983;52:2001-2006. patients with astrocytoma with atypical or anaplastic features. 286. Baram TZ, van Eys J, Dowell RE, et al. Survival and neurologic J Neurooncol 1985;3:99-103. outcome in infants with medulloblastoma treated with sur- 306. Walker MD, Green SB, Byar DP, et al. Randomized comparisons gery and MOPP chemotherapy: a preliminary report. Cancer of radiotherapy and nitrosoureas for the treatment of malignant 1987;60:173-177. glioma after surgery. N Engl J Med 1980;303:1323-1329. 287. Horowitz ME, Mulhern RK, Kun LE, et al. Brain tumors in the 307. Prados MD, Gutin PH, Phillips TL, et al. Highly anaplastic very young child: postoperative chemotherapy in combined- astrocytoma: a review of 357 patients treated between 1977 and modality treatment. Cancer 1988;61:428-434. 1989. Int J Radiat Oncol Biol Phys 1992;23:3-8. 288. Duffner PK, Horowitz ME, Krischer JP, et al. The treatment of 308. Wood JR, Green SB, Shapiro WR. The prognostic importance of malignant brain tumors in infants and very young children: an tumor size in malignant gliomas: a CT scan study by the Brain update of the Pediatric Oncology Group experience. Neuroon- Tumor Cooperative Group. J Clin Oncol 1988;6:338-343. cology 1999;1:152-161. 309. Wisoff JH, Boyett JM, Berger MS, et al. Current neurosurgi- 289. Geyer JR, Zeltzer PM, Boyett JM, et al. Survival of infants with cal management and the impact of the extent of resection in primitive neuroectodermal tumors or malignant ependymomas the treatment of malignant gliomas of childhood: a report of of the CNS treated with eight drugs in 1 day: a report from the the Children’s Cancer Group Trial no. CCG-945. J Neurosurg Children’s Cancer Group. J Clin Oncol 1994;12:1607-1615. 1998;89:52-59. 290. Kuhl J, Beck J, Bode U, et al. Delayed radiation therapy (RT) 310. Garden AS, Maor MH, Yung WKA, et al. Outcome and patterns after postoperative chemotherapy (PCH) in children less than of failure following limited-volume irradiation for malignant as- 3 years of age with medulloblastoma. Results of the trial HIT- trocytomas. Radiother Oncol 1991;20:99-110. SKK’87, and preliminary results of the pilot trial HIT-SKK’92 311. Fine HA, Dear KB, Loeffler JS, et al. Meta-analysis of radiation [abstract]. Med Pediatr Oncol 1995;25:250. therapy with and without adjuvant chemotherapy for malignant 291. Walter AW, Mulhern RK, Gajjar A, et al. Survival and neu- gliomas in adults. Cancer 1993;71:2585-2597. rodevelopmental outcome of young children with medullob- 312. Murray KJ, Nelson DF, Scott C, et al. Quality-adjusted survival lastoma at St. Jude Children’s Research Hospital. J Clin Oncol analysis of malignant glioma. Patients treated with twice-daily 1999;17:3720-3728. radiation (RT) and carmustine: a report of Radiation Therapy 292. Dupuis-Girod S, Hartmann O, Benhamou E, et al. Will high Oncology Group study 83-02. Int J Radiat Oncol Biol Phys dose chemotherapy followed by autologous bone marrow trans- 1995;31:453-459. plantation supplant craniospinal irradiation in young children 313. Sneed PK, Gutin PH, Larson DA, et al. Patterns of recurrence treated for medulloblastoma? J Neurooncol 1996;27:87-98. of glioblastoma multiforme after external irradiation fol- 293. Gajjar A, Mulhern RK, Heideman RL, et al. Medulloblastoma lowed by implant boost. Int J Radiat Oncol Biol Phys 1994;29: in very young children: outcome of definitive craniospinal irra- 719-727. diation following incomplete response to chemotherapy. J Clin 314. Hochberg FH, Pruitt A. Assumptions in the radiotherapy of Oncol 1994;12:1212-1216. glioblastoma. Neurology 1980;30:907-911. 294. Rutkowski S, Bode U, Deinlein F, et al. Treatment of early child- 315. Wallner KE, Galicich JH, Krol G, et al. Patterns of failure fol- hood medulloblastoma by postoperative chemotherapy alone. lowing treatment for glioblastoma multiforme and anaplastic N Engl J Med 2005;352:978-986. astrocytoma. Int J Radiat Oncol Biol Phys 1989;16:1405-1409. 295. Grill J, Le Deley M-C., Gambarelli D, et al. Postoperative chem- 316. Loeffler JS, Alexander E III, Hochberg FH, et al. Clinical pat- otherapy without irradiation for ependymoma in children under terns of failure following stereotactic interstitial irradiation 5 years of age: a multicenter trial of the French Society of Pedi- for malignant gliomas. Int J Radiat Oncol Biol Phys 1990;19: atric Oncology. J Clin Oncol 2001;19:1288-1296. 1455-1462. 868 PART 10 Central Nervous System

317. Agbi CB, Bernstein M, Laperriere N, et al. Patterns of recur- 337. Saroja KR, Mansell J, Hendrickson FR, et al. Failure of acceler- rence of malignant astrocytoma following stereotactic intersti- ated neutron therapy to control high grade astrocytomas. Int J tial brachytherapy with 125I implants. Int J Radiat Oncol Biol Radiat Oncol Biol Phys 1989;17:1295-1297. Phys 1992;23:321-326. 338. Fitzek MM, Thornton AF, Rabinov JD, et al. Accelerated proton/ 318. Laramore GE, Diener-West M, Griffin TW, et al. Randomized photon irradiation to 90 cobalt gray equivalent for glioblastoma neutron dose searching study for malignant gliomas of the multiforme: results of a phase II prospective trial. J Neurosurg brain: results of an RTOG study. Int J Radiat Oncol Biol Phys 1999;91:251-260. 1988;14:1093-1102. 339. Leibel SA, Gutin PH, Wara WM, et al. Survival and quality of life 319. Loeffler JS, Alexander E III, Wen PY, et al. Results of stereotactic after interstitial implantation of removable high-activity iodine brachytherapy used in the initial management of patients with 125 sources for treatment of patients with recurrent malignant glioblastoma. J Natl Cancer Inst 1990;82:1918-1921. gliomas. Int J Radiat Oncol Biol Phys 1989;17:1129-1139. 320. Scharfen CO, Sneed PK, Wara WM, et al. High activity iodine- 340. Selker RG, Shapiro WR, Green S, et al. A randomized trial of 125 interstitial implant for gliomas. Int J Radiat Oncol Biol Phys interstitial radiotherapy (IRT) boost for the treatment of newly 1992;24:583-591. diagnosed malignant glioma (glioblastoma multiforme, ana- 321. Prados MD, Gutin PH, Phillips TL, et al. Interstitial brachytherapy plastic astrocytoma, anaplastic oligodendroglioma, malignant for newly diagnosed patients with malignant gliomas: the UCSF mixed glioma). Brain Tumor Cooperative Group study 87-01 experience. Int J Radiat Oncol Biol Phys 1992;24:593-597. [abstract]. Paper presented at the 45th Meeting of the Congress 322. Stelzer KJ, Laramore GE, Griffin TW, et al. Fast neutron radio- of Neurological Surgeons. San Francisco: CA, October 14-18, therapy: the University of Washington experience. Acta Oncol 1995. 1994;33. 275-280. 341. Florell RC, MacDonald DR, Irish W, et al. Selection bias, sur- 323. Larson DA, Gutin PH, McDermott M, et al. Gamma knife for vival, and brachytherapy for glioma. J Neurosurg 1992;76: glioma: selection factors and survival. Int J Radiat Oncol Biol 179-183. Phys 1996;36:1045-1053. 342. Loeffler JSI, Alexander E III, Shea WM, et al. Radiosurgery as 324. Pickles T, Goodman GB, Rheaume DE, et al. Pion radiation for part of the initial management of patients with malignant glio- high grade astrocytoma: results of a randomized study. Int J Ra- mas. J Clin Oncol 1992;10:1379-1385. diat Oncol Biol Phys 1997;37:491-497. 343. Sankaria JN, Mehta MP, Loeffler JS, et al. Radiosurgery in the 325. Sneed PK, Prados MD, McDermott MW, et al. Large effect of age initial management of malignant gliomas: survival comparison on survival of patients with glioblastoma treated with radiothera- with the Radiation Therapy Oncology Group recursive parti- py and brachytherapy boost. Neurosurgery 1995;36:898-904. tioning analysis. Int J Radiat Oncol Biol Phys 1995;32:931-942. 326. Wen PY, Alexander E III, Black PM, et al. Long-term results of 344. Nelson DF, Curran WJ Jr, Scott C, et al. Hyperfractionated stereotactic brachytherapy used in the initial treatment of pa- radiation therapy and BIS-chlorethyl nitrosourea in the treat- tients with glioblastomas. Cancer 1994;73:3029-3036. ment of malignant glioma: possible advantage observed at 327. Shrieve DC, Alexander E III, Wen PY, et al. Comparison of stere- 72 Gy in 1.2 Gy bid fractions: report of the Radiation Therapy otactic radiosurgery and brachytherapy in the treatment of recur- Oncology Group protocol 8302. Int J Radiat Oncol Biol Phys rent glioblastoma multiforme. Neurosurgery 1995;36:275-284. 1993;25:193-207. 328. Mehta MP, Masciopinto J, Rozental J, et al. Stereotactic radio- 345. Fallia C, Olmi P. Hyperfractionated and accelerated radiation surgery for glioblastoma multiforme: report of a prospective therapy in central nervous system tumors (malignant glio- study evaluating prognostic factors and analyzing long-term sur- mas, pediatric tumors, and brain metastases). Radiother Oncol vival advantage. Int J Radiat Oncol Biol Phys 1994;30:541-549. 1997;43:235-246. 329. Buatti JM, Friedman WA, Bova FJ, et al. Linac radiosurgery for 346. Brada M, Sharpe G, Rajan B, et al. Modifying radical radiothera- high grade gliomas: the University of Florida experience. Int J py in high grade gliomas: shortening the treatment through ac- Radiat Oncol Biol Phys 1995;32:205-210. celeration. Int J Radiat Oncol Biol Phys 1999;43:287-292. 330. Nelson DF, Diener-West M, Weinstein AS, et al. A randomized 347. Levin VA, Maor MH, Thall PF, et al. Phase II study of acceler- comparison of misonidazole sensitized radiotherapy plus BCNU ated fractionation radiation therapy with carboplatin followed and radiotherapy plus BCNU for treatment of malignant glioma by vincristine chemotherapy for the treatment of glioblastoma after surgery: final report of an RTOG study. Int J Radiat Oncol multiforme. Int J Radiat Oncol Biol Phys 1995;33:357-364. Biol Phys 1986;12:1793-1800. 348. Halperin EC. Multiple-fraction-per-day external beam radio- 331. Urtusan R, Band P, Chapman JD, et al. Radiation and high-dose therapy for adults with supratentorial malignant gliomas. J Neu- metronidazole in supratentorial glioblastomas. N Engl J Med rooncol 1992;14:255-262. 1976;294:1364-1367. 349. Fulton DS, Urtasun RC, Scott-Brown I, et al. Increasing radia- 332. Miralbell R, Mornex F, Greiner R, et al. Accelerated radiother- tion dose intensity using hyperfractionation on patients with apy, carbogen, and nicotinamide in glioblastoma multiforme: malignant glioma: final report of a prospective phase I-II dose report of European Organization for Research and Treatment of response study. J Neurooncol 1992;14:63-72. Cancer Trial 22933. J Clin Oncol 1999;17:3143-3149. 350. Shapiro WR, Green SB, Burger PC, et al. Randomized trial of 333. Hegarty TJ, Thornton AF, Diaz RF, et al. Intraarterial bromode- three chemotherapy regimens and two radiotherapy regimens oxyuridine radiosensitization of malignant gliomas. Int J Radiat in postoperative treatment of malignant glioma: Brain Tumor Oncol Biol Phys 1990;19:421-428. Cooperative Group Trial no. 8001. J Neurosurg 1989;71:1-9. 334. Sullivan FJ, Herscher LL, Cook JA, et al. National Cancer In- 351. Levin VA, Silver P, Hannigan J, et al. Superiority of post- stitute (phase II) study of high-grade glioma treated with ac- radiotherapy adjuvant chemotherapy with CCNU, procar- celerated hyperfractionated radiation and iododeoxyuridine: bazine, and vincristine (PCV) over BCNU for anaplastic gliomas: results in anaplastic astrocytoma. Int J Radiat Oncol Biol Phys NCOG 6G61 final report. Int J Radiat Oncol Biol Phys 1990;18: 1994;30:583-590. 321-324. 335. Groves MD, Maor MH, Meyers C, et al. A phase II trial of 352. Medical Research Council Brain Tumour Working Party. Ran- high-dose bromodeoxyuridine with accelerated fractionation domized trial of procarbazine, lomustine, and vincristine in radiotherapy followed by procarbazine, lomustine, and vincris- the adjuvant treatment of high-grade astrocytoma: a Medical tine for glioblastoma multiforme. Int J Radiat Oncol Biol Phys Research Council Trial. J Clin Oncol 2001;19:509-518. 1999;45:127-135. 353. Stupp R, Mason WP, van den Bent MJ, et al. Radiotherapy plus 336. Prados MD, Scott C, Sandler H, et al. A phase 3 randomized concomitant and adjuvant temozolomide for glioblastoma. study of radiotherapy plus procarbazine, CCNU, vincristine N Engl J Med 2005;352:987-996. (PCV) with or without BrdU for the treatment of anaplastic 354. Athanassiou H, Synodinou M, Maragoudakis E, et al. Rand- astrocytoma: a preliminary report of the Radiation Therapy omized phase II study of temozolomide and radiotherapy com- Oncology Group study 9404. Int J Radiat Oncol Biol Phys pared with radiotherapy alone in newly diagnosed glioblastoma 1999;45:1109-1115. multiforme. J Clin Oncol 2005;23:2372-2377. 33 The Brain and Spinal Cord 869

355. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing 377. Dziuk T, Woo S, Butler B, et al. Malignant meningioma: an and benefit from temozolomide in glioblastoma. N Engl J Med indication for initial aggressive surgery and adjuvant radiotherapy. 2005;352:997-1003. J Neurooncol 1998;37:177-188. 356. del Rowe J, Scott C, Werner-Wasik M, et al. Single-arm, open- 378. Jaaskelainen J, Haltia M, Servo A. Atypical and anaplastic men- label phase II study of intravenously administered tirapazamine ingiomas: radiology, surgery, radiotherapy, and outcome. Surg and radiation therapy for glioblastoma multiforme. J Clin Oncol Neurol 1986;25:233-242. 2000;18:1254-1259. 379. Mirimanoff RO, Rosoretz DE, Linggood RM, et al. Meningioma: 357. Valtonen S, Timonen U, Toivanan P, et al. Interstitial chem- analysis of recurrence and progression following neurosurgical otherapy with carmustine-loaded polymers for high-grade resection. J Neurosurg 1985;62:18-24. gliomas: a randomized double-blind study. Neurosurgery 380. Goldsmith BJ, Wara WM, Wilson CB, et al. Postoperative 1997;41:44-49. irradiation for subtotally resected meningiomas. A retrospective 358. Nakagawa K, Aoki Y, Fujimaki T, et al. High-dose conformal ra- analysis of 140 patients treated from 1967 to 1990. J Neurosurg diotherapy influenced the pattern of failure but did not improve 1994;80:195-201. survival in glioblastoma multiforme. Int J Radiat Oncol Biol 381. Petty AM, Kun LE, Meyer GA. Radiation therapy for incom- Phys 1998;40:1141-1149. pletely resected meningiomas. J Neurosurg 1985;62:502-507. 359. Sheline GE. Radiotherapy for high grade gliomas. Int J Radiat 382. Wara WM, Sheline GE, Newman H, et al. Radiation therapy of Oncol Biol Phys 1990;18:793-803. meningiomas. AJR Am J Roentgenol 1975;123:453-458. 360. Yung WKA, Janus TJ, Maor M, et al. Adjuvant chemotherapy 383. Carelia RJ, Ransohoff J, Newall J. Role of radiation therapy in with carmustine and cisplatin for patients with malignant glio- the management of meningioma. Neurosurgery 1982;10:332- mas. J Neurooncol 1992;12:131-135. 339. 361. Heideman RL, Kuttesch J, Gajjar AJ, et al. Supratentorial malig- 384. Forbes AR, Goldberg ID. Radiation therapy in the treatment of nant gliomas in childhood: a single institution perspective. Can- meningioma: the Joint Center for Radiation Therapy experience, cer 1997;80:497-504. 1970 to 1982. J Clin Oncol 1984;2:1139-1143. 362. Finlay JL, Boyett JM, Yates AJ, et al. Randomized phase III trial 385. Maguire PD, Clough R, Friedman AH, et al. Fractionated exter- in childhood high-grade astrocytoma comparing vincristine, nal-beam radiation therapy for meningiomas of the cavernous lomustine, and prednisone with the eight-drugs-in-1 regimen. sinus. Int J Radiat Oncol Biol Phys 1999;44:75-79. J Clin Oncol 1995;13:112-123. 386. Lunsford L. Contemporary management of meningiomas: radia- 363. Mork SJ, Lindegaard K-F., Halvorsen TB, et al. Oligodendro tion therapy as an adjuvant and radiosurgery as an alternative to glioma: incidence and biological behavior in a defined popula- surgical removal? J Neurosurg 1994;80:187-190. tion. J Neurosurg 1985;63:881-889. 387. Kondziolka D, Flickinger JC, Perez B. Judicious resection and/ 364. Jenkins RB, Blair H, Ballman KV, et al. A t(1;19)(q10;p10) me- or radiosurgery for parasagittal meningiomas: outcomes from a diates the combined deletions of 1p and 19q and predicts a bet- multicenter review. Neurosurgery 1998;43:405-413. ter prognosis of patients with oligodendroglioma. Cancer Res 388. Kondziolka D, Levy EI, Niranjan A, et al. Long-term outcomes 2006;66:9852-9861. after meningioma radiosurgery: physician and patient perspec- 365. Brandes AA, Tosoni A, Cavallo G, et al. Correlations between tives. J Neurosurg 1999;91:44-50. O6-methylguanine DNA methyltransferase promoter methyla- 389. Shafron DH, Friedman WA, Buatti JM, et al. Linac radiosur- tion status, 1p and 19q deletions, and response to temozolomide gery for benign meningiomas. Int J Radiat Oncol Biol Phys in anaplastic and recurrent oligodendroglioma: a prospective 1999;43:321-327. GICNO study. J Clin Oncol 2006;24:4746-4753. 390. Smith JL, Vuksanovic MM, Yates BM, et al. Radiation therapy 366. Shaw EG, Scheithauer BW, O’Fallon JR, et al. Oligodendro- for primary optic nerve meningiomas. J Clin Neuroophthalmol gliomas: the Mayo Clinic experience. J Neurosurg 1992;76: 1981;1:85-89. 428-434. 391. Borovich B, Doron Y, Braun J, et al. Recurrence of intracra- 367. Shaw EG, Scheithauer BW, O’Fallon JR. Management of nial meningiomas: the role played by regional multicentricity. supratentorial low-grade gliomas. Oncology 1993;7:97-107. Clinical and radiological aspects. J Neurosurg 1986;65(Pt 2): 368. Sheline GE, Boldney E, Karisburg P, et al. Therapeutic consider- 168-171. ations in tumors affecting the central nervous system: oligoden- 392. Roche P-H, Regis J, Dufour H, et al. Gamma knife radiosurgery drogliomas. Radiology 1964;82:44-49. in the management of cavernous sinus meningiomas. J Neuro- 369. Chin HW, Hazel JJ, Kim TH, et al. Oligodendrogliomas. I. A surg 2000;93:68-73. clinical study of cerebral oligodendrogliomas. Cancer 1980;45: 393. Ojemann SG, Sneed PK, Larson DA, et al. Radiosurgery for 1458-1466. malignant meningioma: results in 22 patients. J Neurosurg 370. Lindegaard K-F., Mork SJ, Eide GE, et al. Statistical analysis of 2000;93:62-67. clinicopathological features, radiotherapy, and survival in 270 394. Deen HC Jr, Scheithauer BW, Ebersold MJ. Clinical and path- cases of oligodendroglioma. J Neurosurg 1987;67:224-230. ologic study of meningiomas of the first two decades of life. 371. Reedy PD, Bay JW, Hahn JF. Role of radiation therapy in the J Neurosurg 1982;56:317-322. treatment of cerebral oligodendroglioma: an analysis of 57 cases 395. Drake JM, Hendrick EB, Becker LE, et al. Intracranial meningi- and a literature review. Neurosurgery 1983;67:224-230. omas in children. Pediatr Neurosci 1986;12:134-139. 372. Allison RR, Schulsinger A, Vongtama V, et al. Radiation and 396. Snow RB, Lavyne MH. Intracranial space-occupying lesions in chemotherapy improve outcome in oligodendroglioma. Int J AIDS patients. Neurosurgery 1985;16:148-153. Radiat Oncol Biol Phys 1997;37:399-403. 397. So YT, Beckstead JH, Davis RL. Primary central nervous system 373. Ludwig CL, Smith MT, Godfrey AD, et al. A clinicopathologi- lymphoma in AIDS: a clinical and pathologic study. Ann Neurol cal study of 323 patients with oligodendrogliomas. Ann Neurol 1986;20: 566-S72. 1986;19:15-21. 398. Reni M, Ferreri AJ, Garancini MP, et al. Therapeutic manage- 374. Cairncross JG, MacDonald DR. Successful chemotherapy for ment of primary central nervous system lymphoma in immuno- recurrent malignant oligodendroglioma. Ann Neurol 1988;23: competent patients: results of a critical review of the literature. 460-464. Ann Oncol 1997;8:227-234. 375. Cushing H, Eisenhardt L. Meningiomas: Their Classification, 399. Jack CR, Reese DF, Scheithauer BW. Radiographic findings in Regional Behavior, Life History, and Surgical End Results. 32 cases of primary CNS lymphoma. AJR Am J Roentgenol Springfield, IL: Charles C Thomas, 1938. 1986;146:271-276. 376. Glaholm J, Bloom HJG, Crow JH. The role of radiotherapy in the 400. Hochberg FH, Miller DC. Primary central nervous system lym- management of intracranial meningiomas: the Royal Marsden phoma. J Neurosurg 1988;68:835-853. Hospital experience with 186 patients. Int J Radiat Oncol Biol 401. Murray K, Kun L, Cox J. Primary malignant lymphoma of the Phys 1990;18:755-761. central nervous system. J Neurosurg 1986;65:600-607. 870 PART 10 Central Nervous System

402. Socie G, Piprot-Chauffat C, Schlienger M, et al. Primary lym- 422. Flickinger JC, Kondziolka D, Lunsford LD, et al. A multi- phoma of the central nervous system: an unresolved therapeutic institutional experience with stereotactic radiosurgery for problem. Cancer 1990;65:322-326. resectable single brain metastasis. Int J Radiat Oncol Biol Phys 403. Nelson DF, Martz KL, Bonner H, et al. Non-Hodgkin’s lympho- 1994;28:797-802. ma of the brain: can high-dose, large-volume radiation therapy 423. Smalley SR, Laws ER Jr, O’Fallon JR, et al. Resection for solitary improve survival? Report on a prospective trial by the Radiation brain metastasis: role of adjuvant radiation and prognostic vari- Therapy Oncology Group: RTOG 8315. Int J Radiat Oncol Biol ables in 229 patients. J Neurosurg 1992;77:531-540. Phys 1992;23:9-17. 424. Aoyama H, Shirato H, Tago M, et al. Stereotactic radiosurgery 404. Loeffler JS, Ervin TJ, Mauch P, et al. Primary lymphomas of the plus whole-brain radiation therapy vs stereotactic radiosurgery central nervous system: patterns of failure and factors that influ- alone for treatment of brain metastases: a randomized controlled ence survival. J Clin Oncol 1985;3:490-494. trial. JAMA 2006;295:2483-2491. 405. Laperriere NJ, Cerezo L, Milosevic MF, et al. Primary lympho- 425. Choi KN, Withers HR, Rotman M. Intracranial metastases from ma of brain: results of management of a modern cohort with melanoma: clinical features and treatment by accelerated frac- radiation therapy. Radiother Oncol 1997;43:247-252. tionation. Cancer 1985;56:1-9. 406. Glass J, Gruber ML, Cher L, et al. Preirradiation methotrexate 426. Weltman E, Salvajoli JV, Brandt RA, et al. Radiosurgery for brain chemotherapy of primary central nervous system lymphoma: metastases: a score index for predicting prognosis. Int J Radiat long-term outcome. J Neurosurg 1994;81:188-195. Oncol Biol Phys 2000;46:1155-1161. 407. Abrey LE, DeAngelis LM, Yahalom J. Long-term survival in pri- 427. Mork SJ, Loken AC. Ependymoma: a follow-up study of 101 mary CNS lymphoma. J Clin Oncol 1998;16:859-863. cases. Cancer 1977;40:907-915. 408. Abrey LE, Yahalom J, DeAngelis LM. Treatment for pri- 428. McCormick PC, Torres R, Post KD, et al. Intramedullary epend- mary CNS lymphoma: the next step. J Clin Oncol 2000;18: ymoma of the spinal cord. J Neurosurg 1990;72:523-532. 3144-3150. 429. Reimer R, Onofrio BM. Astrocytomas of the spinal cord in chil- 409. Schultz C, Scott C, Sherman W, et al. Preirradiation chemo- dren and adolescents. J Neurosurg 1985;63:669-675. therapy with cyclophosphamide, doxorubicin, vincristine, and 430. Kopelson G, Linggood RM, Leinman GM, et al. Management dexamethasone for primary CNS lymphomas: initial report of of intramedullary spinal cord tumors. Radiology 1980;135: Radiation Therapy Oncology Group Protocol 88-06. J Clin On- 473-479. col 1996;14:556-564. 431. Chun HC, Schmidt-Ullrich RK, Wolfson A, et al. External beam 410. O’Neill BP, Wang C-H., O’Fallon JR, et al. Primary central nerv- radiotherapy for primary spinal cord tumors. J Neurooncol ous system non-Hodgkin’s lymphoma (PCNSL): Survival ad- 1990;9:211-217. vantages with combined initial therapy? A final report of the 432. Guidetti B, Mercuri S, Vagnozzi R. Long-term results of the sur- North Central Cancer Treatment Group study 86-72-52. Int J gical treatment of 129 intramedullary spinal gliomas. J Neuro- Radiat Oncol Biol Phys 1999;43:559-563. surg 1981;54:323-330. 411. Ferreri AJM, Reni M, Villa E. Therapeutic management of pri- 433. Cooper PR, Epstein F. Radical resection of intramedullary spinal mary central nervous system lymphoma: lessons from prospec- cord tumors in adults: recent experience in 29 patients. J Neuro- tive trials. Ann Oncol 2000;11:927-937. surg 1985;63:492-499. 412. DeAngelis LM, Seiferheld W, Schold SC, et al. Combination 434. Epstein J. Spinal cord astrocytomas of childhood. Prog Exp chemotherapy and radiotherapy for primary central nervous Tumor Res 1987;30:135-153. system lymphoma: Radiation Therapy Oncology Group Study 435. Epstein FJ, Farmer J-P., Freed D. Adult intramedullary astrocyto- 93-10. J Clin Oncol 2002;20:4643-4648. mas of the spinal cord. J Neurosurg 1992;77:355-359. 413. Posner JB. Management of central nervous system metastases. 436. Epstein F, McCleary EL. Intrinsic brain-stem tumors of child- Semin Oncol 1977;4:81-91. hood: surgical indications. J Neurosurg 1986;64:11-15. 414. Pickren JW, Lopez G, Tsukada Y, et al. Brain metastases: an autopsy 437. Epstein F. Intraaxial tumors of the cervicomedullary junction in study. Cancer Treatment Symposia 1983;2:295-313. children. Concepts Pediatr Neurosurg 1987;7:117. 415. Gaspar L, Scott C, Rotman M, et al. Recursive partitioning 438. Kopelson G, Linggood RM, Kleinman GM, et al. Management analysis (RPA) of prognostic factors in three Radiation Therapy of intramedullary spinal cord tumors. Radiology 1980;135: Oncology Group brain metastases trials. Int J Radiat Oncol Biol 473-479. Phys 1997;37:745-751. 439. O’Sullivan C, Jenkin RD, Doherty MA, et al. Spinal cord tumors 416. Swift PS, Phillips T, Martz K, et al. CT characteristics of patients in children: long-term results of combined surgical and radiation with brain metastases treated in Radiation Therapy Oncology treatment. J Neurosurg 1994;81:507-512. Group study 79-16. Int J Radiat Oncol Biol Phys 1993;25: 440. Kopelson G. Radiation tolerance of the spinal cord previously 209-214. damaged by tumor and operation: long term neurological 417. Smalley SR, Laws ER Jr, O’Fallon JR, et al. Resection for solitary improvement and time-dose-volume relationships after irradia- brain metastasis: role of adjuvant radiation and prognostic vari- tion of intraspinal gliomas. Int J Radiat Oncol Biol Phys 1982;8: ables in 229 patients. J Neurosurg 1992;77:531-540. 925-929. 417a. Borgelt G, Gelber R, Kramer S, et al. The palliation of brain 441. Shaw EG, Evans RG, Scheithauer BW, et al. Radiotherapeutic metastases: final results of the first two studies by the Radia- management of adult intraspinal ependymomas. Int J Radiat tion theraphy Oncology Group. Int J Radiat Oncol Biol Phys. Oncol Biol Phys 1986;12:323-327. 1980;6:11-9. 442. Garcia DM. Primary spinal cord tumors treated with surgery 418. Shaw EG. Radiotherapeutic management of multiple brain me- and postoperative irradiation. Int J Radiat Oncol Biol Phys tastases: “3000 in 10” whole brain radiation is no longer a “no 1985;11:1933-1939. brainer.” Int J Radiat Oncol Biol Phys 1999;45:253-254. 443. Bryne TN. Spinal cord compression from epidural metastases. 419. Auchter RM, Lamond JP, Alexander E III, et al. A multi- N Engl J Med 1992;327:614-619. institutional outcome and prognostic factor analysis of radiosur- 444. Rodichok LD, Harper GR, Ruckdeschel JC, et al. Early diagnosis gery for resectable single brain metastasis. Int J Radiat Oncol of spinal epidural metastases. Am J Med 1981;70:1181-1188. Biol Phys 1996;35:27-35. 445. Young RF, Post EM, King GA. Treatment of spinal epidural 420. Patchell RA, Tibbs PA, Walsh JW, et al. A randomized trial of metastases. J Neurosurg 1980;53:741-748. surgery in the treatment of single metastases to the brain. N Engl 446. Sorensen PS, Borgesen SE, Rohde K, et al. Metastatic epidural J Med 1990;322:494-500. spinal cord compression: results of treatment and survival. Can- 421. Alexander E III, Moriarty TM, Davis RB, et al. Stereotactic radio- cer 1990;65:1502-1508. surgery for the definitive, noninvasive treatment of brain metas- 447. Latini P, Maranzano E, Ricci S, et al. Role of radiotherapy in tases. J Natl Cancer Inst 1995;87:34-40. metastatic spinal cord compression: preliminary results from a prospective trial. Radiother Oncol 1989;15:227-233. 33 The Brain and Spinal Cord 871

448. Sundaresan N, Galicich JH, Lane JM, et al. Treatment of neo- 450. Herbert SH, Solin LJ, Rate WR, et al. The effect of palliative plastic epidural cord compression by vertebral body resection radiation therapy on epidural compression due to metastatic and stabilization. J Neurosurg 1985;63:676-684. malignant melanoma. Cancer 1991;67:2472-2476. 449. Siegal T, Seigal T. Surgical decompression of anterior and pos- 451. Olsen ME, Chernik NL, Posner JB. Infiltration of the leptome- terior malignant epidural tumors compressing the spinal cord: a ninges by systemic cancer. Arch Neurol 1974;30:122-137. prospective study. Neurosurgery 1985;17:424-432.