Chronic oral VP-16 for recurrent medulloblastoma
M C Chamberlain 1, P A Kormanik
Chronic oral VP-16 (Etoposide) is a chemotherapy regi-men with wide application in oncology and documented efficacy against germ cell tumors, lymphomas, Kaposi sarcoma, and gliai brain tumors. Eight patients ranging in age from 4 to 36 years (median 7.5 years) with locally recurrent medulloblastoma were treated with VP-16. No patient displayed evidence of cerebrospinal fluid dissem-ination, distant brain or spine parenchymal metastases, or extraneural metastatic disease. All patients had previ-ously been treated with surgery (gross total resection, 5; subtotal resection, 3), craniospinal radiotherapy, and platinum-based chemotherapy (adjuvant, 3; salvage, 8). Each cyde of therapy consisted of 21 days of VP-16 (50 mg/m2/day) followed by a 7 to 14 day rest followed by an additional 21 days of VP-16 (50 mg/m2/day). Complete blood counts were obtained weekly. Neurologic examina-tion and brain magnetic resonance imaging scan with contrast were performed prior to each cycle of therapy. Treatment-related complications included: partial alope-cia (5 patients); diarrhea (4); weight loss (3); anemia (2); neutropenia (4); and thrombocytopenia (4). Two patients required transfusion and 1 patient received antibiotics for nentropenic fever. All patients were evaluable for re-sponse: 3 demonstrated progressive disease after the first cyde of VP-16, 3 had stable disease (range 4 to 6 months) and 2 had partial neuroradiographic responses (8 and 10 months). Median duration of response and stable disease was 6 months (range: 4 to 10 months) in 5 of 8 (62.5%) patients. Chronic oral VP-16 is a well-tolerated and relatively non-toxic chemotherapeutic agent with demon-strated activity in locally recurrent meduiloblastoma. © 1997 by Elsevier Science Inc. AH rights reserved.

Chamberlain MC, Kormanik PA. Chronic oral VP-16 for recurrent medulloblastoma. Pediatr Neurol 1997;17: 230-234.

From the Department of Neurosciences; University of California San Diego; San Diego, California.

Introduction

Brain tumors are the second most common malignancy of childhood, accounting for approximately 20% of all malignant neoplasms in children, and are exceeded in number only by acute leukemias [1-5]. Of every 100,000 children under the age of 16 years in the United States, 2 or 3 will manifest a new brain tumor each year [1-5]. Approximately 20% of all brain tumors of childhood are medulloblastomas, in contrast to adults where this tumor accounts for 1 to 2% of all primary brain tumors [1-6]. Of all childhood brain tumors, the most dramatic change in survival has occurred with medulloblastomas [3-5,7,8]. In Cushing’s series reported in 1930, 1 of 61 children (1.6%) was alive at the end of 3 years. Recently, Packer reported

a 5-year survival of >70% in children with high-risk medulloblastomas treated with adjuvant multi-agent che-motherapy [8]. Notwithstanding these figures, medullo-blastomas recur in approximately one third of patients, and in the majority of these patients, salvage therapies are palliative. New therapies for the treatment of recurrent medulloblastoma are desirable given the modest results of contemporary salvage therapies. This prospective Phase II study of 8 patients with previously treated locally recur-rent medulloblastomas examines the toxicity and efficacy of chronic oral VP-16 salvage chemotherapy.

Patients and Methods

Eight patients (5 boys; 3 girls), ranging in age from 4 to 36 years (median 7.5 years) were treated for recurrent medulloblastomas. The study was approved by the Institutional Review Board and consent was obtained from all adult participants and juveniles’ parents. Recurrent medulloblastomas were defined by objective neuroradiographic progres-sion ( > 2 5 % increase in tumor size) compared with prior images. No patient displayed evidence of cerebrospinal fluid (CSF) dissemination, distant brain or spine parenchymal metastases, or extraneural metastatic disease, All patients with locally recurrent medulloblastomas were evaluated with contrast-enhanced cranial magnetic resonance imaging

Communications should be addressed to:
Dr. Chamberlain; Department of Neurosciences; University of Califoruia, San Diego; 220 Dickinson Street; San Diego, CA 92103-8421.
Received February 24, 1997; accepted May 21, 1997.

230 PEDIATRIC NEUROLOGY Vol. 17 No. 3 © 1997 by Elsevier Science Inc. All rights reserved.
PII S0887-8994(97)00098-2 • 0887-8994/97/$17.00

(MR1), contrast-enhanced spine MRI, and CSF analysis. At the time of tumor recurrence, patients presented with the following signs and symptoms: ataxia ( 6 : 5 with gait and 4 with limb ataxia); headache (5); cranial nerve VI paresis (2, 1 with bilateral and 1 with unilateral involvement); and evidence of raised intracranial pressure as evidenced by nausea and vomiting in association with headache (2). Patient performance status (either Karnofsky status or modified Lansky [9]) ranged from 60 to 100% with a median of 80% at time of documented tumor recurrence. Tumor location was as follows: in 5 patients tumors were midline with early involvement of the brainstem; in 3 patients tumors were hemispheric with extension into the middle cerebellar peduncle. All patients had prior surgery, either gross total resection (5) or subtotal resection (3), at time of initial diagnosis. In addition to surgery, 2 patients required placement of a ventriculoperitoneal shunt for man-agement of symptomatic hydrocephalus. At presentation no patient had evidence of disseminated medulloblastoma,

All patients had been previously treated with adjuvant craniospinal radiotherapy. Radiation therapy was administered according to conven-tional fractionation schedules in which 180 cGy was given daily for 8 to 8.5 weeks to a total dose of 54 Gy to posterior fossa, and 36 Gy to whole brain and spinal cord.

Three patients, all of whom had initial subtotal resections, were treated with adjuvant chemotherapy including cisplatin, lomustine (CCNU), and vincristine according to a previously published protocol [8,10-13]. These same 3 children had residual measurable disease following subtotal resection at the time of initial surgery. All received 8 cycles of this regimen. At the time of tumor recurrence, these 3 patients were treated with carboplatin (450 mg/mZ/day given once every 4 weeks) for a median of 5 cycles of therapy (range: 3 to 6) [14-16]. In addition, the 5 patienls initially treated with gross total resection who were not treated with adjuvant chemotherapy, at time of tumor recurrence received cisplatin, CCNU, and vincristine [8,10-13]. This regimen was administered in the same manner as that given adjuvantly without the inclusion of weekly vincristine. These 5 patients received between 2 to 6 cycles of this therapy with a median of 4 cycles. No patient had received prior VP-16.

Imaging. Cranial MRI examinations were performed using a 1.5 Tesla superconducting magnet (Signa: General Electric, Milwaukee, W1). Using a spin-echo pulse sequence, axial T2-weighted (relaxation time, TR: 3,00l) ms; echo time, TE: 80 ms) proton density-weighted (TR: 3,000 ms; TE: 30 ms) images were initially acquired. Subsequently, both sagittal and axial or coronal T~-weighted (TR: 600 ms; TE: 25 ms) images were acquired. Slice thickness was 5 mm, with a 5 m m interval between successive slices; a 250 × 250 matrix was used. After intravenous administration of 0.1 mmol/kg gadolinium-pentetic acid dimeglumine (Berlex Laboratories, Cedar Knolls, NJ), coronal, axial, and sagittal T~-weighted sequences (TR: 600 ms; TE: 25 ms) were obtained. All post-contrast images were obtained within 30 minutes of gadolinium infusion. All MRI were independently reviewed by a neuroradiologist in addition to the neuro-oncology staff,

Drug Schedule. VP-16 (Etoposide, Bristol Myers Squibb Co., Prince-ton, N J) was administered orally 50 mg/mZ/day for 21 consecutive days (A subcycle), tbllowed by a 14-day break, then an additional 21 days (B subcycle) of oral VP-16 at 50 mg/mZ/day [17-23]. All doses were given as a single daily drug administration in the morning. In small children, the dose administered was either 25 or 50 mg per day.

Dexamethasone was used concurrently in 6 children treated with VP-16 and was maintained as either a stable (4 children) or tapering dose (2 children) as patient clinical status permitted. VP-16 dose modifications are represented in Table 1.

Toxicity was assessed using the National Cancer Institute common toxicity scale, and dose modifications were made at the conclusion of each subcycle of VP- 16 [ 18-23 I. In children with stable disease, partial or complete response, an additional cycle of VP-16 was initiated, after which the children were assessed again as described below. The children were continued on VP-16 until documentation of progressive disease at which time they were removed from the study.

Methods ~1~Evaluation. Blood counts were obtained weekly, neuro-logic examination was performed monthly, and contrast-enhanced cranial

Table 1. VP-16 dose modifications
Granulocytes/l~l Platelets/l~l VP-16 (%)
> 1,500 > 100,000 100
1,000 - 1,499 99,999 - 75,000 50
< 1,000 < 75,000 0

MRI was performed every 8 weeks following a cycle of VP- 16 and prior to initiating the next cycle of chemotherapy.

Neuroradiographic response using the criteria of MacDonald [241 were as follows: a complete response required disappearance of all lesion(s). A partial response required a > 5 0 % decrease in the product of greatest orthagonal diameter(s) of each measurable lesion on cranial-enhanced MRI. Stable disease required a < 5 0 % decrease in the product of greatest orthagonal diameter(s) of each measurable lesion, or demonstration of no significant change on enhanced cranial MRI, Additionally, the dose of steroids, if administered, must have been stable or decreased, and the neurologic examination must have been stable or improved to warrant a complete/partial response or stable disease. Progressive disease required a 25% increase in the greatest product of orthagonal diameters of any measurable lesion, the appearance of a new lesion, or deterioration of the neurologic examination not explained by other causes. The time to tumor progression was measured as the interval from entry into the study until documentation of progressive disease.

Results

All patients were assessable for toxicity. Treatment-related complications included: partial alopecia in 5 chil-dren (62.5%); non-bloody diarrhea easily managed by oral anti-diarrheal medications (i.e., Immodium) in 4 (15%); >10% baseline weight loss in 3 (37.5%); Grade III to IV neutropenia (leukocyte count <1,500/1~1) in 4 patients (50%), in whom 1 required antibiotic treatment of neutro-penic fever; Grade IIl to IV thrombocytopenia (platelet count <40,000/~tl) in 4 patients (50%) in whom 1 required platelet transfusion; and Grade III to IV anemia (hemo-globin <8 mg/dl) in 2 patients (25%), 1 of whom required packed erythrocyte transfusion.

A total of 38 cycles of VP- 16 were administered, 2 (5%) of which were complicated by a transfusion requirement (erythrocyte transfusion in 1 and platelet transfusion in 1). A median of 2 cycles of therapy (range l to 5) was administered. Four cycles of therapy (11%) were compli-cated by neutropenic fever without bacteriologic docu-mentation. Because of myelosuppression, 8 cycles of therapy (21%) in 4 patients required a delay in initiation of VP-16. Ten cycles of therapy (46%) required VP-16 dose reduction because of myelosuppression. There were no treatment-related deaths and no patients were removed from the study because of drug-related toxicity.

All patients were assessable for response (Table 2). After 1 cycle of VP-16, 3 patients (37.5%) demonstrated progressive disease and were offered alternative or sup-portive therapy. Five patients either responded to therapy or manifested stable disease; in 2 patients (25%) a partial response was observed. Three patients (37.5%) had stable disease. Thus, there was a total response plus stable

Chamberlain and Kormanik: VP-16 for Medulloblastorna 231

Table 2. Locally recurrent meduiloblastomas: Salvage therapy with VP-16

Adjuvant Therapy Salvage Therapy VP-16 Salvage Therapy
Radiation Response/
Sex/Age (Gray) No. Duration Survival
Patient (years) Surgery PF/WB, SC Chemotherapy Chemotherapy Cycles in months* in months
1 M/4 GTR 54/36 CDDP, CCNU, VCR [ PD 3
2 M/5 GTR 54/36 CDDP, CCNU, VCR 2 SD/5 6
3 F/6 GTR 54/36 CDDP, CCNU, VCR 3 SD/6 7
4 M/7 STR 54/36 CDDP, CCNU, VCR CBDCA 4 PR/9 13
5 M/8 STR 54/36 CDDP, CCNU, VCR CBDCA I PD 2
6 F/I 1 STR 54/36 CDDP, CCNU, VCR CBDCA 5 P ~ l l 12
7 F/16 GTR 54/36 CDDP, CCNU, VCR 2 SDN 5
8 M/36 GTR 54/36 CDDP, CCNU, VCR I PD 2
* Duration refers to time to tumor progression
Abbreviations:
CBDCA = Carboplatinum
CDDP, CCNU, VCR = Cis-platinum, Lomustine, Vincristine
F - Female
M = Male
PD = Progressive disease
PF/WB, SC - Posterior fossa/whole brain and spinal cord
PR - Partial response
SD = Stable disease

disease rate of 62.5%. In responding or stable patients, improvement or resolution was seen in neurologic status with particular improvement noted in signs of raised intracranial pressure, i.e., headache, nausea, vomiting and ataxia. No change in Karnofsky or modified Lansky performance was observed. Median response plus stable disease duration was 6 months, with a range of 4 to l0 months. Patients who experienced failure of VP-16 after initial response or disease stabilization were offered alter-native or supportive therapy. Median survival following initiation of oral VP- 16 was 5.5 months in the entire study group with a range of 2 to 13 months. In the subgroup of patients with either a neuroradiographic response or stable disease, median survival following initiation of oral VP-16 was 7 months with a range of 5 to 13 months as contrasted to nonresponding patients in whom the median survival was 2 months with a range of 2 to 3 months. This was statistically significant (P < .001) as assessed by log rank analysis (contingency table chi-square test).

Discussion

Notwithstanding the advances in the treatment of pedi-atric brain tumors and in particular medulloblastoma, in most series 30 to 40% of patients develop recurrences which result in death due to tumor progression [1-6]. A variety of approaches have been used in patients with recurrent medulloblastoma including reoperation, re-irra-diation and an assortment of chemotherapy regimens [1-6,8,10-13,17,25-34]. In the vast majority of patients these salvage therapies are palliative resulting in modest prolongation of life. Patients with locally recurrent medul-loblastomas not involving brainstem and without evidence

of central nervous system dissemination are candidates for reoperation and dose-intensive chemotherapy with autol-ogous bone marrow transplantation. In particular, Finlay, Friedman, Kalifa and colleagues have been advocates of such an approach using combinations of high dose carbo-platin, melphalan, busulfan, VP-16 and thio-triethylene/ thiophosphoramide (thio-TEPA) [35-37]. Response rates of 75% and disease-free survival of 50% have been reported. This approach remains investigational and is best conducted at participating neuro-oncology treatment cen-ters.

The majority of patients are not candidates for reopera-tion; for example, patients with both intra- and extraneural metastatic medulloblastoma. These patients may poten-tially be candidates for dose-intensive chemotherapies, however, inclusion criteria for this therapy is evolving. Determining patient suitability for transplantation remains problematic. Poor outcomes have been reported by trans-plantation groups in patients with recurrent medulloblas-tomas complicated by bulky disease or leptomeningeal spread of disease [35-371. Again, the application of this therapy is best performed at investigational centers.

Because the majority of patients are not candidates lbr reoperation or dose-intensive chemotherapy with autolo-gous transplantation, re-irradiation and standard dose sal-vage chemotherapies are used in the majority of patients treated for recurrent medulloblastoma [1-6,8,10-13,17, 25-29,33,341. Re-irradiation provides modestly effective palliative therapy with reported median patient survival of 13 months, though with marked acute toxicity and poten-tially late neurotoxicity [34]. Whether re-irradiation can be used as neoadjuvant or consolidative therapy in conjunc-

232 PEDIATRIC NEUROLOGY Vol. 17 No. 3

tion with systemic chemotherapy is as yet unanswered and awaiting formal study.

Lastly, a variety of salvage chemotherapies, many with high response rates, have been used in treating patients with recurrent medulloblastomas. Based on patient re-sponse, platinum-based regimens such as cis-platinum, CCNU, and vincristine or carboplatin, are most commonly chosen [1,8,10-16,33]. In the event of failure following platinum-based chemotherapy, fewer treatment options are available including: PCV (procarbazine, CCNU, and vin-cristine), dibromodulcitol (DBD), MOPP (nitrogen mus-tard, vincristine, prednisone, procarbazine), cyclophos-phamide with or without vincristine, and melphalan [25-31]. We detail the University of California at San Diego Neuro-Oncology Service experience treating pa-tients with surgically nonresectable and locally recurrent medulloblastoma with chronic oral VP-16 following fail-ure of platinum-based therapies.

Chronic oral VP-16 has been used in salvage chemo-therapy in a variety of systemic tumors including recurrent germ cell tumors, non-Hodgkin’s lymphoma and Kaposi’s sarcoma, in addition to recurrent brainstem gliomas, chi-asmatic-hypothalamic gliomas, cerebellar astrocytomas, and supratentorial hemispheric gliomas in children [ 18-23, 38]. The rationale for investigating the use of VP-16 beyond the 3 to 5 day standard intravenous dosage schedule is based on three major considerations: the mechanism of action, preclinical studies, and clinical data. VP-16 appears to be a relatively phase-specific agent that inhibits cells in the GoS phase of the cell cycle. VP-16 has been shown to damage DNA by interacting with the enzyme topoisomerase II by stabilizing the DNA-topo-isomerase II complex. VP-16 prevents DNA strands from rejoining, resulting in double-strand DNA breaks. It ap-pears that interaction of VP-16 with topoisomerase II is reversible once the VP-16 concentration falls below a critical level. It would follow that prolonged exposure to a critical VP-16 concentration would enhance the antineo-plastic activity of the drug both by the cell’s cycle-specific mechanism of action and by prolonging its interaction with topoisomerase lI. However, there is limited experi-ence in the use of topoisomerase inhibitors in the treatment of malignam brain tumors.

In a prior study of recurrent brainstem gliomas treated with chronic oral VP-16, 12 patients were evaluable, of whom 6 demonstrated a radiographic response (1 com-plete; 3 partial; 2 stable disease) with a median duration of response of 8 months [18]. In another study of oral VP-[6, 14 children with recurrent supratentorial malignant glio-mas were treated of whom 50% demonstrated either a radiographic partial response (3 patients) or stable disease (4 patients) with a median duration of response of 8 months [191. In a third study of oral VP-16, 12 children with recurrent cerebellar juvenile pilocytic astrocytomas were treated 120]. Six children (50%) demonstrated either a radiographic response or stable disease with a median duration of response of 7 months. In the fourth study of

children treated with oral VP-16, 14 patients with recur-rent chiasmatic-hypothalamic gliomas were treated [21]. Five children demonstrated a radiographic response (1 complete and 4 partial) and 3 children demonstrated stable disease with a median duration of response of 8 months. Similar to the findings we report, these four small studies suggest that oral VP-16 has apparent activity in children with recurrent brain tumors regardless of topography and is well tolerated with only modest toxicity.

In a comparable study, Ashley et al. reported on experience with chronic oral VP-16 in 7 patients with recurrent medulloblastoma [17]. The study by Ashley et al. differed from this report in the following ways. Their patients were heavily pretreated prior to initiating oral VP-16 with 1 to 8 (median 3) prior chemotherapy regi-mens including parenteral VP-16 as compared to our 1 to 2 (median 1) prior chemotherapy treatments. Importantly, the Ashley et al. study enrolled patients with recurrent medulloblastoma regardless of pattern of disease recur-rence, and 5 of their 7 patients had leptomeningeal disease. All of our patients had locally recurrent parenchymal disease with no evidence of cerebrospinal fluid dissemi-nation or metastases. Except that we observed no complete responders to oral VP-16, our results are similar to those of the Duke University group [17]. We have, in additiom reported on durability of response to oral VP- 16, data not presented in the report by Ashley.

In conclusion, chronic oral VP- 16 appears active against locally recurrent medulloblastomas which are resistant to platinum chemotherapy and warrants further investigation in a larger, cooperative group setting. The activity of chronic oral VP-I.6 against recurrent medulloblastomas, however, is tempered by its relatively short response durations, probably due to the emergence of VP-16 resis-tant tumor clonogens. Incorporating oral VP- 16 with other active agents such as high dose parental cyclophospha-mide or carboplatin may result in further improvements in survival in patients with recurrent medulloblastomas.

References

[1] Packer RJ, Finlay JI. Medulloblastoma: Presentation. diagnosis and management. Oncology 1988:2:35-44.

[2] Tomlinson FH, Scheithauer BW, Meyer EB, Smithson WA, Shaw EG, Miller GM, Groover RV. Medulloblastoma: I. clinical, diagnostic, and therapeutic overview. J Child Neurol 1992:7: 142-55.

[3] Duffner PK, Cohen ME, Myers MH. Heise HW. Survival of children with brain tumors: SEER prograln, 1973-1980. Neurology 1986:36:597-601.

[4] Duffner P, Horowitz M, Krischer J, Friedman HS, Burger PC, Cohen ME, Sanford RA, Mulhern RK, James HE, Freeman CR, Seidel FG, Kun LE. Postoperative chemotherapy and delayed radiation in children less than three years of age with malignant brain tumors. N Engl J Med 1993;328:1725-31.

[5] R u t k a JT, Hoffman HJ. Medulloblastoma: A historical perspec-tive and overview. J Neuro-Oncol 1996;29: I-7.

[6] Peterson K, Walker RW. Medulloblastoma/primitive neuroecto-dermal tumor in 45 adults. Neurology 1995;45:44(l-2.

[7] Karoly MD, Casey ATH, Hayward RD, Harkness WFJ, Phipps

Chamberlain and Kormanik: VP-16 for Medulloblastoma 233

K, Wase AM. Medulloblastoma: Is the 5-year survival rate improving? A review of 80 cases from a single institution. J Neurosurg 1997;86:13-2 I.

[8] Packer RJ, Sutton LN, EIterman R, Lange B, Goldwein J, Nicholson S, Mulne L, et al. Outcome for children with medulloblastoma treated with radiation and cisplatin, CCNU, and vincristine chemother-apy. J Neurosurg 1994;81:690-8.

[9] Lansky LL, List MA, Lansky SB, Cohen ME, Sinks LF. Toward the development of a play performance scale of children (PPSC). Cancer 1985;56:1837-40.

]10] Packer RJ, Siegel KR, Sutton LN, Evans AE, D’Angio G, Rorke LB, Bunin GR, Schut L. Efficacy of adjuvant chemotherapy for patients with poor-risk medalloblastoma: A preliminary report. Ann Neurol 1988;24:503-8.

[11] Packer RJ. Chemotherapy for medulloblastoma/primitive neu-roectodermal tumors of the posterior fossa. Ann Neurol 1990;28:823-8.

[12] Lefkowitz IB, Packer RJ, Siegal KR, Sutton LN, Schut L, Evans AE. Results of treatment of children with recurrent medulloblas-tomaJprimitive neuroectodermal tumors with Iomustine, cisplatin and vincristine. Cancer 1990;65:412-7.

[13] Packer RJ, Sutton LN, Goldwein JW, Perilongo G, Bunin G, Ryan J, Cohen BH, et al. Improved survival with the use of adjuvant chemotherapy in the treatment of medulloblastoma. J Neurosurg 1991; 74:433-40.

[14] Lewis IJ, Stevens MCG, Pearson A, Pinkerton R, Stevens R. Phase II study of carboplatin in children’s tumours. United Kingdom Children’s Cancer Study Group, University of Leicester, UK (abstract). Proc Am Soc Clin Oncol 1993;12:121.

[15] Kovnar E, Heideman R, Kellie S, Douglass E, Greenwald C, Kuttesch J, Ochs J, et al. Carboplatin and VP-16 in the treatment of high stage medulloblastoma, pineoblastoma and supratentorial PNET (ab-stract). Proc Am Soc Clin Oncol 1993;12:425.

[16] Lewis IJ, Stevens MCG, Pearson A, Pinkerton CR, Stevens R. Phase II study of carboplatin in children’s tumors (abstract). Proc Am Soc Clin Oncol 1993;12:412.

[17] Ashley DM, Meier L, Kerby T, Zalduondo M, Friedman HS, Gajjar A, Kun L, et al. Response of recurrent medulloblastoma to low-dose oral etoposide. J Clin Oncol 1996;14:1922-7.

[18] Chamberlain MC. Recurrent brainstem gliomas treated with oral VP-16. J Neuro-Oncol 1993;15:133-9.

[19] Chamberlain MC. Recurrent supratentorial gliomas: Salvage therapy with oral etoposide (abstract). Ann Neurol 1996;40:335.

[20] Chamberlain MC. Recurrent cerebellar gliomas: Salvage ther-apy with oral VP-16. Child Neurol 1997;12:200-4.

[21] Chamberlain MC, Grafe MR. Recurrent chiasmatic-hypotha-lamic glioma treated with oral etoposide. J Clin Oncol 1995;13:2072-6.

[22] Hainsworth JD, Greco FA. Etoposide: Twenty years later. Ann Oncol 1995;6:325-41.

[23] Greco FA, Johnson DH, Hainsworth JD. Chronic oral etopo-side. Cancer 1991;67:303-9.

[24] MacDonald DR, Cascino TL, Schald SC Jr, Cairncross JG. Response criteria for phase II studies of supratentorial malignant glioma. J Clin Oncol 1990;8:1277-80.

[25] Levin VA, Edwards MSB, Gutin PH, Vestnys P. Fulton D, Seager M, Wilson CB. Phase II evaluation of dibromodulcitol in the treatment of recurrent medulloblastoma, ependymoma, and malignant astrocytoma. J Neurosurg 1984;61:1063-8.

[26] Crafts DC, Levin VA, Edwards MS, Pischer TL, Wilson CB. Chemotherapy of recurrent medullo-blastoma with combined procarba-zine, CCNU, and vincristine. J Neurosurg 1978;49:589-92.

[27] Van Eys J, Baram TZ, Cangir A, Brunet JM, Martinez-Prieto J. Salvage chemotherapy for recurrent primary brain tumors in children. J Pediatr 1988;113:601-6.

[28] Friedman HS, Schold SC Jr, Mahaley MS Jr, Calvin OM, Oakes WJ, Vick NA, Burger PC, et al. Phase II treatment of medullo-blastoma and pineoblastoma with melphalan: Clinical therapy based on experimental models of human medulloblastoma. J Clin Oncol 1989;7: 904-11.

[29] Baram TZ, Van Eys J, Dowell RE, Cangir A, Pack B, Bruner JM. Survival and neurologic outcome of infants with medulloblas-toma treated with surgery and MOPP chemotherapy. Cancer [987;60: 173-7.

[30J Allen JC, Helson L. High-dose cyclophosphamide chemother-apy for recurrent CNS tumors in children. J Neurosurg 1981 ;55:749-56.

[31] Van Eys J, Cangir A, Coody D, Smith B. MOPP regimen as primary chemotherapy for brain tumors in infants. J Neuro-Oncol 1985;3:237-43.

[32] Bertoione SJ, Baum ES, Krivit W, Hammond GD. A phase II study of cisplatin therapy in recurrent childhood brain tumors, J Neuro-Oncol 1989;7:5-11.

[33] Walker RW, Allen JC. Treatment of recurrent primary intra-cranial childhood tumors with cis-diamminedichloroplatinum (abstract). Ann Neurol 1983;14:371-2.

[34] Wara WM, Wallner KE, Levin VA, Liu H, Edwards MSB. Retreatment of pediatric brain tumors with radiation and misonidazole. Results of a CCSG/RTOG Phase I/II study. Cancer 1986;58:1636-40.

[35] Dupuis-Girod S, Hartmann O, Benhamou E, Doz F, Mechi-naud F, Bouffet E, Coze C, Kalifa C. Will high dose chemotherapy followed by autologous bone marrow transplantation supplant cranio-spinal irradiation in young children treated for medulloblastoma? J Neuro-Oncol 1996;27:87-98.

[36] Mahoney DH, Strother D, Camitta B, Bowen T, Ghim T, Pick T, Wall D, Yu L, Shuster J J, Friedman H. High-dose melphalan and cyclophosphamide with autologous bone marrow rescue for recurrent/ progressive malignant brain tumors in children: A pilot pediatric oncol-ogy group study. J Clin Oncol 1996;14:382-8.

[37] Finlay JL, Goldman S, Wong MC, Cairo M, Garvin J, August C, Cohen BH, et al. Pilot study of high-dose thiotepa and etoposide with autologous bone marrow rescue in children and young adults with recurrent CNS tumors. J Clin Oncol 1996;14:2495-503.

[38] Phillips PC, Carson BS, Calvin OM. Do topoisomerase inhib-itors have a role in brain tumor chemotherapy? Detection of topoisomer-ase I and I1 in pediatric brain tumors (abstract). Ann Neurol 1990;28:218.