T of high-grade astrocytomas are disappointing. The. Accelerated Fractionation for High-Grade Cerebral Astrocyfomas Preliminary Treatment Results

Transcription

1 Accelerated Fractionation for High-Grade Cerebral Astrocyfomas Preliminary Treatment Results George Shenouda, MBBCh, PhD, FRCP(C),* Luis Souhami, MD," Carolyn R. Freeman, MBBS, FRCP(C),* Joseph Hazel, MD, FRCP(C),* Shirley Lehnert, PhD,* and Lawrence Joseph, PhDt A Phase 1/11 accelerated fractionation study for high-grade cerebral astrocytomas began in October Forty-two patients, 25 men and 17 women, were entered in the study. Median age was 58.5 years of age (range, 32 to 78 years). Performance status was 0, 1, 2, and 3 on Eastern Cooperative Oncology Group (ECOG) scale for 13, 19, 9, and 1 patients, respectively. Thirty-six patients had undergone partial resection, and six had stereotactic biopsy only. All patients had histologically proven astrocytomas (6 Grade 3, and 36 Grade 4). Treatment consisted of radiation therapy doses of 4400 cgy in 22 daily fractions to the whole brain plus a boost of 1600 cgy in 8 fractions given concomitantly with the last 8 whole-brain treatments using a twice daily schedule with an interfraction interval of 8 hours. Median survival time was 57 weeks from the date of starting irradiation. Survival was 50% and 28% at 1 and 2 years, respectively. Alopecia and scalp erythema were seen in all patients; nine patients had localized areas of moist desquamation in the retroauricular region. Decreased hearing and serous otitis media were seen in five patients within 1 to 2 months from the end of treatment. Increasing somnolence was marked in eight patients with progressive deterioration of performance status; computerized axial tomography (CAT) scan studies in all eight patients showed evidence of disease recurrence with associated brain edema and mass effect. Three patients had a second resection for a recurrent tumor with no evidence of brain necrosis at craniotomy. To date, the accelerated fractionation schedule appears to be well tolerated with valuable shortening of overall treatment time. The preliminary results are encouraging, and longer follow-up time is required to evaluate tumor control and toxicity. Cancer 67: ,1991. HE RESULTS OF surgery alone for the management T of high-grade astrocytomas are disappointing. The highly invasive nature of these neoplasms and their localization usually preclude radical surgical excision, and the median survival time after tumor resection is usually only 14 Postoperative irradiation has been shown to prolong the median survival time to 37 weeks From the Divisions of *Radiation Oncology and?clinical Epidemiology, McGill University, Montrtal, Qutbec, Canada. The authors thank Ms. Janet Smith, Ms. Heather Allen, and Ms. Elaine Walker for their assistance in preparing this manuscript. Address for reprints: Dr. George Shenouda, Montrkl General Hospital, Department of Radiation Oncology, 1650 Cedar Avenue, Room 5700, Montrial, Qukbec, H3G 1A4, Canada. Accepted for publication February 6, in retrospective and prospective randomized studies. However, long-term survival rates are not affected by such Conventional postoperative radiation therapy consists of 6000 cgy delivered in 30 daily fractions. Experimental data suggest that the response of malignant gliomas to conventional photon irradiation may be limited by the rapid turnover rate of the clonogenic cells. Because tumor clonogen regeneration reduces the efficacy of treatment if it begins before therapy is completed, it is desirable to deliver the whole radiation course within the shortest practical overall treatment time. Currently, there is considerable interest in modified radiation therapy schedules8-12 such as hyperfractionation and accelerated fractionation. Both strategies use multiple fractions per day, 2247

2 2248 CANCER May I Vol. 67 compared with single daily dose in conventional fractionation. Hyperfractionation allows the delivery of higher total doses with late toxicity equivalent to conventional schedules as long as the dose per fraction is lower than the conventional fraction and an interval of at least 6 to 8 hours is allowed between fractions for the repair of sublethal damage. However, accelerated fractionation aims at improving the radiation-induced cell killing in fast proliferating tumors by shortening the overall treatment time using multiple daily fractions with an interfraction interval of 6 to 8 hours. Recently, FowlerI3 showed data to support that increasing the number of fractions each day to two and three results in a greater log cell kill by one and two logs, respectively. He also emphasized the importance of the tumor doubling times and compared overall treatment times for single, two, and three daily fractions given 5 days per week for tumors with different doubling times. The optimal overall times were found to depend more on the proliferation rate of the tumor than on the number of fractions per day. For tumors with doubling times of 3 days and less, the overall optimal time to achieve the best log cell kill was 2 to 3 weeks. For 5-day doubling times, overall times of 4 weeks were optimal, whether the number of fractions per day was one, two, or three. Recent experimental data suggest that the fraction size is the predominant factor in determining the incidence of late toxicity of neural tissue. The overall treatment time has a negligible influence on the incidence of late effects as long as the dose per fraction is kept constant and the interfraction interval allows for repair of radiation damage in normal neural cells."'.'4 In 1982, the European Organization for Research on Treatment of Cancer (EORTC) reported the results of a Phase I1 study using misonidazole (MISO) as a hypoxic cell sensitizer with an altered fractionation regimen consisting of delivering three daily fractions of 200 cgy each.15 The total dose was 6000 cgy delivered in 30 fractions over 4 weeks, including a 2-week rest period. The early results showed good tolerance of such an accelerated fractionation schedule and MISO in 122 patients treated between I979 and 198 I. However, the long-term results of the above-mentioned study were never published and the late radiation-induced toxicity of such a treatment is unknown. Other authors investigated other treatment regimens for high-grade astrocytomas using accelerated fractionation, I6,l7 hyperfractionation with or without systemic chemotherapy and hypoxic cell sensiti~er,~." and superfractionated radiation therapy. ' 9,20 Because of the poor prognosis of patients with malignant gliomas, and based on experimental and clinical data, a Phase 1/11 trial evaluating an accelerated fractionation schedule was initiated at McGill University in adult patients with high-grade astrocytomas. Accelerated radiation therapy was chosen because of the known biologic data regarding high-grade astrocytomas cell cycle kinetics with reported potential doubling times of less than 1 week,l',ll the potential advantage of valuable shortening of overall treatment time, and the encouraging results using accelerated fractionation in other sites, i.e., in head and neck cancer^.'^.'^ The rationale was to shorten the overall treatment time to avoid tumor cell repopulation that might be expected to occur in a longer conventional treatment time of 6 weeks. The main objectives of the study were to determine the feasibility and to assess the toxicity of such treatment in adults with high-grade cerebral astrocytomas. Eligibility Criteria Patients and Methods Patients included in the study were adults with a histologic diagnosis of Grade 3 or 4 cerebral astrocytomas with no history of brain irradiation or systemic chemotherapy treatment. Patients with Eastern Cooperative Oncology Group (ECOG) performance status scores of less than or equal to 3 were accepted in the study. All patients gave informed consent. Putient Population Between October 1987 and November I989,42 patients with high-grade astrocytomas were treated according to this Phase 1/11 study of accelerated radiation therapy. Patient characteristics are shown in Table l. The median duration of symptoms before diagnosis was 3 months (range, 2 days to 12 months). Two patients had a history of malignancy; one had an endometrial adenocarcinoma, and one had a lower extremity chondrosarcoma. Both patients had a disease-free survival longer than 5 years at the time of diagnosis of brain neoplasm. The initial workup included history and physical examination, baseline complete blood count and serum TABLE I. Patient Characteristics Characteristics No. of patients Sex Male 25 Female 17 Age (yr) Older than 70 5 Performance status (ECOG) Presenting symptoms Headaches 18 Motor weakness 17 Neuropsychiatric changes 17 Grand ma1 seizures 8 ECOG: Eastern Cooperative Oncology Group.

3 No. 9 HIGH-GRADE CEREBRAL ASTROCYTOMAS - Shenouda et al biochemistry, and chest radiographs. A11 patients had plain and contrast-infused computerized axial tomography (CAT) scans. The anatomic location of the cerebral lesions on CAT scan is shown in Table 2. The median size of the enhancing lesion was 4 cm (range, 3 to 7 cm). Surgical Management and Pathology Thirty-six patients had partial resection. Six patients had stereotactic biopsy only because of the anatomic location of the tumors and the potential morbidity of surgical resection. Histologic diagnosis was obtained in all patients; 6 had Grade 3, and 36 had Grade 4 astrocytomas according to Kernohan's pathologic classification system.25 Radiation Therapy Patients started radiation therapy after a median delay of 20 days (range, 1 to 37 days) from the date of surgery. All patients were treated with photons in the megavoltage range (cobalt-60 gamma rays or 4 to 6 megavolt x-rays). The whole brain was treated using a pair of parallel opposed fields with appropriate eye shielding. The boost volume was determined from the preoperative and postoperative CAT scans and consisted of the enhancing tumor rim plus a 2-cm margin. The radiation treatment schedule consisted of 4400 cgy in 22 fractions given to the whole brain and a boost of 1600 cgy in 8 fractions to the tumor bed given concomitantly, with the final 8 days of whole-brain irradiation using a twice daily schedule with an interfraction interval of 8 hours (Fig. 1). Corticosteroid administration was continued at the pretreatment dosage and tapered off slowly toward the end or shortly after completion of irradiation. Anticonvulsant medication was used in patients with a history of grand ma1 seizures as their presenting symptoms. Follow- Up After radiation therapy, patients were examined once a month for the first 2 months, then on a bimonthly basis. Neurotoxicity was evaluated by detailed neurologic examination and assessment of performance status at each visit. Brain CAT scan was repeated 3 months after com- Site Frontal Parietal Temporal Occipital Hypothalamus, thalamus More than one lobe Total TABLE 2. Tumor Location No. of oatients CO"TI0NAL FRACTIONATION SCHEDULE Field Weeks of treatment Whole brain xxxxx xxxxx xxxxx xxxxx xx BOOSt xxx xxxxx ACCELKRATKD FRACITONATION S(IIIEDuLE Field Weeks of treatment Whole brain xxxxx xxxxx xxxxx xxxxx xx Boost x xxxxx xx FIG. 1. Comparison between conventional and accelerated fractionation schedule in the current study. pletion of radiation therapy and thereafter on a 6-month basis or as clinically indicated. One patient was lost to follow-up and was considered dead at the date of his last follow-up visit. The cases of all other patients were followed to death or January 31, Statistical Considerations Survival time was calculated from the first day of radiation therapy. This decision was based on the fact that the interval between surgery and radiation therapy vaned from 1 to 37 days. Other confounding factors, such as the postoperative clinical course, could have affected the duration of survival should the latter have been calculated from the day of surgery. Survival curves were calculated using the Kaplan-Meier method.?' The interaction of each potential prognostic factor and their effect on survival were analyzed using the Cox proportional hazard model~.~' The Wilcoxon test of significance2* was used to assess the degree of correlation between different prognostic factors. Feasibility of Treatment Results The concomitant boost fractionation schedule was well tolerated by all patients, and all 42 patients completed treatment as planned. The whole course of irradiation was delivered with median duration of 3 1 days (range, 28 to 36 days) from the first day of radiation therapy. Interruptions occurred during the course of therapy in only three patients, all ofwhom lived out of town and, because of transportation difficulties, had to be admitted to the hospital for the eight twice daily treatments. The duration of the treatment interruption was 2 days in one patient and 3 days in two patients until hospital beds became available. Siirvival ilnal.vsis With a median follow-up time of 46.5 weeks (range, 10 to 109 weeks) for surviving patients, the median sur-

4 ~~~~~~ 2250 CANCER May Vol. 67 viva1 time was 57 weeks. Survival was 50% and 28% at 1 and 2 years, respectively (Fig. 2). At last follow-up visit, four patients were alive with no evidence of disease. Ten patients were alive with recurrent disease at the site of their initial tumor. Twenty-two patients died after a median interval time of 20 weeks (range, 4 to 52 weeks) from the time of radiation therapy. The cases of the last six patients were followed for periods shorter than 12 weeks, and their disease status could not be evaluated even though they were clinically stable. When the stratified survival curves were compared, patients who had resection did better than those who had stereotactic biopsy only (log-rank test P = 0.014) (Fig. 3). Analysis using the Cox proportional hazard model demonstrated a statistically significant effect of pretreatment performance status (PS) of patients on survival. Patients with ECOG PS of 2 and 3 fared worse than those with a PS of 0 or 1 (Fig. 4). These two variables were found to be correlated; patients with stereotactic biopsy only also had worse PS (Wilcoxon test P = 0.058). A recursive partition algorithm was used to determine the optimal cutpoints for age analysis and to define its effect on survival. Cut points of 39 and 64 years were selected. Patients from 39 to 64 years of age had a better prognosis when compared with to younger (less than 39 years of age, log-rank test P = 0.028) or older (more than 64 years of age, P = 0.019) patients. In the current series, the histologic grade was not found to have a statistically significant effect on patient survival; however, there were only six patients with Grade 3 tumors. Acute Toxicity All patients had alopecia. Hair regrowth occurred in three patients over 6 to 12 months from the end of treatment. Generalized scalp erythema was seen in all patients with localized areas (less than 2 X 2 cm2) of moist desquamation in the retro-auricular region in nine patients. These areas of desquamation were within the boost fields and were seen irrespective of the photon beam energy. They responded well to topical steroids and complete healing took place within 10 to 14 days after completion of therapy. Nausea and vomiting were reported by four patients 0.L z o 60 80,011 XURUlUElL (LIEEYS, FIG. 3. Probability of survival of patients according to type of surgery. Stereotactic biopsy only (-) (n. 6). Surgical resection (...) (n = 36). and were treated with antiemetic medication. Two patients had deep vein thrombosis, and two others had corticosteroid-induced upper gastrointestinal toxicity. Delayed Complications Serous otitis media was noticed in five patients and was associated with decreased hearing in all patients. Surgical drainage was required in two of these five patients with complete recovery. Somnolence was marked in ten patients after a median interval of 20 weeks from the completion of radiation therapy. Of these, two were transient and coincident with corticosteroid withdrawal. In the remaining eight patients, somnolence progressed to stupor and coma with CAT scan study results demonstrating disease progression associated with reappearance of contrast enhancement, brain edema, and mass effect in all patients. Twelve patients were tapered off steroids within 4 to 6 weeks of the completion of irradiation; 30 patients required steroid therapy longer than 12 weeks after treatment. All patients with prolonged steroid use showed progression of disease at subsequent follow-up visits. Deterioration of performance status was seen in 22 patients, including the 10 with progressive somnolence, after a median interval of 12 weeks from irradiation. The results of follow-up CAT scans done on these patients showed r f O. L L-7, ~ :... L... FIG. 2. Kaplan-Meier estimates of probability of survival (all patients). FIG. 4. Effect of initial performance status (ECOG) on survival. PS 1 (...) (n = 19); PS = 2 and 3 ( ) (n = lo) =O (--) (n. 13); PS.

5 No. 9 HIGH-GRADE CEREBRAL ASTROCYTOMAS - Shenouda et al radiologic findings suggestive of disease progression. Of these 22 patients, 3 had a second craniotomy and resection of recurrent astrocytoma at 10 and 12 months from the end of their radiation therapy. There was no evidence of brain necrosis in any of these three patients. In another patient, positron emission tomography results were suggestive of early tumor recurrence, and this was confirmed by subsequent follow-up CAT scans. The remaining 18 patients had CAT scan changes highly suggestive of tumor recurrence. The reappearance of contrast enhancement in the original tumor bed, associated mass effect, and brain edema favored the diagnosis of disease recurrence rather than brain necrosis, although the differentiation between the two conditions was difficult. Treatment at Recurrence Thirty-two patients had radiologic evidence of disease progression after a median interval of 28 weeks from the time of starting radiation therapy, with biopsy-proven diagnosis of recurrence in 3 of 32 patients. Twenty-four patients with recurrence received supportive care and corticosteroids with a median survival time of 6 weeks from the time of recurrence. Five patients received systemic chemotherapy only at recurrence with a median survival of 2 months (range, 1 to 6 months) after chemotherapy. Three patients had a second craniotomy and resection of recurrent astrocytoma. Of these three patients, one received systemic chemotherapy after his second surgery and died at 9 months with disease. Another patient with recurrent disease had only a second craniotomy and resection with a postoperative survival of 5 months. The third patient had a second resection 1 year after her irradiation and was found to have a recurrent high-grade astrocytoma. She received postoperative systemic chemotherapy, had cerebrospinal fluid (CSF) seeding as demonstrated by myelographic and cytologic study 3 months later, and received palliative irradiation to the spine. Discussion The accelerated irradiation schedule used in this study was well tolerated by all patients. Hospitalization was required for three patients, all of whom lived out of town, and for a duration of 8 to 10 days. Analysis showed that shortening of overall treatment time from 6 to 4.5 weeks was feasible with an acceptable level of toxicity. The median survival time of 57 weeks is comparable with other reports of conventional postoperative radiation therapy,4,29-30 in patients with high-grade cerebral astrocytomas, as well as other approaches using different experimental treatment strategies of radiation therapy with or without systemic chemotherapy and/or hypoxic cell ~ensitizer.~.,~ We compared our study population with that reported in the literature regarding the main prognostic factors such as age, performance status, extent of surgery, and histo- logic grade. In general, our patients were older (median age, 58.5 years of age) as compared with other series (median age, 52 years of age).8316 ECOG performance status was less than 2 in 97% and 93% of our patients and those reported in other series, respectively. l5 Fourteen percent of patients in the current study underwent biopsy only, compared with 7% reported by the European Organization on Research and Treatment of Cancer (EORTC)15 and 19% by Shin et a1.i Fourteen percent of our patients had Grade 3 astrocytoma compared with a range of 25%3 to 60%16 reported in other series. In the present study, survival time was calculated from the first day of radiation therapy, which is a disadvantage when compared with other where survival time was calculated from the date of surgery. Early toxicity is comparable to our previous experience and that reported in the literature for conventional radiation therapy as well as for altered fractionation schedules with or without sensitizer^.^,^^,^^,'^ Somnolence was reported to occur during and after therapy in other st~dies. ~,~ In the current study, somnolence was seen in ten patients after completion of therapy, associated with disease progression in all patients (except two) in whom the somnolence was coincident with the time of steroid withdrawal, and reversed by the reinstitution of medication. Prolonged steroid use has been reported by Freeman et al.34 in 43% of pediatric patients after the use of hyperfractionated radiation therapy in children with brain stem gliomas. In our patients, prolonged steroid therapy was invariably associated with disease progression. Radiologically, it is difficult to differentiate between radiation-induced brain necrosis and tumor recurrence. Three patients had a second craniotomy, and in all there was histologic diagnosis of recurrence without evidence of brain necrosis. None of the patients who died underwent an autopsy. The survival of these patients was short for brain necrosis to have become clinically overt; the true incidence of radiation necrosis in this group of patients is, therefore, difficult to estimate. The question of the optimum volume for treatment of high-grade astrocytomas is still a subject of debate. Despite the improved accuracy of tumor localization achieved by CAT scans and especially by magnetic resonance imaging, the true microscopic extension of the tumor remains difficult to identify. Most studies demonstrate that 80% to 90% of all recurrences occur within a margin of 2 cm from the primary site,35*36 and recent studies using partial brain irradiati~n **~~~ show results comparable with those obtained with whole-brain irradiation. Because the risk of toxicity is greater with the use of whole-brain fields, partial brain ports are now accepted as standard treatment of high-grade astrocytomas. Additional studies in our department will be performed using partial-brain irradiation in a trial to improve the radiation tolerance of normal brain tissue.

6 2252 CANCER Mav Vol. 67 We conclude that the accelerated fractionation radiation therapy regimen used was well tolerated with an acceptable risk to benefit ratio and the advantage of shortening the overall treatment time by 25%. The preliminary results are encouraging. Although acute toxicity is acceptable, additional follow-up is necessary to assess delayed toxicity and efficacy. Additional studies are planned to use higher total doses of accelerated fractionation and partial-brain irradiation with the goal of improving local control without increasing toxicity. REFERENCES 1. Walker MD, Strike TA, Sheline GE. An analysis of dose-effect relationship in the radiotherapy of malignant gliomas. Int J Radial Oncol Biol Phjjs 1979: 5 1: Walker MD, Alexander E Jr, Hunt WE et 01. Evaluation of BCNU and/or radiotherapy in the treatment of anaplastic gliomas: A cooperative clinical trial. J Neurosurg 1978; 49: Kristiansen K, Hagen S, Kollevold T et al. Combined modality therapy of operated astrocytomas grade I11 and 1V: Confirmation of the value of postoperative irradiation and lack of potentiation of bleomycin on survival time. A prospective multicenter trial of the Scandinavian Glioblastoma Group. Cancer 1981; 47: Salazar OM, Rubin P, Feldstein ML, Pizzutiello R. High dose radiation therapy in the treatment of malignant gliomas: Final report. Jnt J Rudiat Oncol Biol Phys 1979: 5: Green SB, Byar DP, Walker MD et a/. Comparison of carmustine, procarbazine, and high-dose methyl prednisolone as additions to surgery and radiotherapy for the treatment of malignant glioma. Cancer Treatment Rep 1983; 67: Nelson DF, Schoenfeld D, Weinstein AS et a/. A randomized comparison of misonidazole sensitized radiotherapy plus BCNU and radiotherapy plus BCNU for treatment of malignant glioma after surgery: Preliminary results of an RTOG study. Int J Radiat Oncol Biol Phys 1983; 9: Deutsch M, Green SB, Strike TA et a/. Results of a randomized trial comparing BCNU plus radiotherapy, streptozotocin plus radiotherapy, BCNU plus hyperfractionated radiotherapy, and BCNU following misonidazole plus radiotherapy in the postoperative treatment ofmalignant glioma. Int JRadiat Oncol Biol Phys 1989; 16: Withers HR. Biologic basis for altered fractionation schemes. Cancer 1985; 55: Thames HD, Withers HR, Peters LJ, Fletcher GH. Changes in early and late radiation responses with altered dose fractionation: Implications for dose-survival relationships. Int J Radiat Oncol Biol Phys 1982; 8: Thames HD, Peters U, Withers HR, Fletcher GH. Accelerated fractionation versus hyperfractionation: Rationales for several treatments per day. Int JRadiat Oncol Biol Phys 1983; 9: I. Fowler JF. Intervals between multiple fractions per day: Differences between late and early reactions. Acta Oncologica 1988; 27: Peschel RE, Fischer JJ. Multiple daily fractionation schedules. Int JRadiat Oncol Biol Phys 1982; &I Fowler JF. How worthwhile are short schedules in radiotherapy?: A series of exploratory calculations. Radiother Oncoll990; 18: I. 14. Van Der Kogel AJ. Radiation tolerance of the rat spinal cord: Time-dose relationships. Radiolog), 1977; Ang K, Van Der Schueren E, Notter G et al. Split course multiple daily fractionated radiotherapy schedule combined with misonidazole for the management of grade 111 and IV gliomas. In1 J Radiat Oncol Biol Phys 1982; 8: Bignardi M, Bertoni F. Radiation treatment with twice a day fractionation versus conventional fractionation in high grade astrocytomas: A retrospective study. Acta Oncologica 1987; 26: Keim H, Potthoff PC, Schmidt K, Schiebusch M, Neiss A, Trott KR. Survival and quality of life after continuous accelerated radiotherapy of glioblastomas. Radiotlier Oncol 1987; 9: Shin KH, Urtasun RC, Fulton D et al. Multiple daily fractionated radiation therapy and misonidazole in the management of malignant astrocytoma: A preliminary report. Cancer 1985; 56: Ludgate CM, Douglas BG, Dixon DF, Steinbok P, Jackson SM, Goodman GB. Superfractionated radiotherapy in grade 111, IV intracranial gliomas. Int J Radiat Oncol Biol Phys 1988; 15: Shin KH, Muller PJ, Geggie PH. Superfractionation radiation therapy in the treatment of malignant astrocytomas. Cancer 1983; 52: Hoshino T, Baker M, Wilson CB, Boldrey EB, Fewer D. Cell kinetics of human gliomas. J Neurosurg 1972; 37: Hoshino T, Wilson CB. Cell kinetic analyses ofhuman malignant brain tumors (gliomas). Cancer 1979: 44: Gray AJ. Treatment of advanced head and neck cancer with accelerated fractionation. Int J Radiat Oncol Biol Phys 1986; 12: Wang CC. Suit HD, Blitzer PH. Twice-a-day radiation therapy for supraglottic carcinoma. Int J Radial Oncol Biol Phys 1986; 12: Kernohan JW, Sayre GP. Tumors of the central nervous system. In: Firminger HI, ed. Atlas of Tumor Pathology, section 10, fascicle 35. Washington, DC: Armed Forces Institute of Pathology, 1952: Kaplan EL, Meier P. Nonparametric estimation for incomplete observations. JAm Stat Assoc 1958: 53: Kalbfleisch JD, Prentice RL. The Statistical Analysis of Failure Time Data. New York: Wiley, 1980: Lehmann EL. Nonparametrics: Statistical Methods Based on Ranks. San Francisco: Holden-Day, 1975; Uihlen A, Colby MJ, Layton DD, Parsons WR, Cargent L. Comparison of surgery and surgery plus irradiation in the treatment of supratentorial gliomas. Acta Radiation Therapy Phvsics Biology 1966: 5: Stage WS, Stein JJ. Treatment of malignant astrocytomas. Am 1 Roentgenol 1974; I20: Marcial-Vega VA, Wharam MD, Leibel S, Clark A, Zweig R, Order SE. Treatment of supratentorial high grade gliomas with split course high fractional dose postoperative radiotherapy. Int J Radiat Oncol Biol PhJ)s 1989: 16: Urtasun R. Feldstein ML, Partington J et al. Radiation and mitroimidazoles in supratentorial high grade gliomas: A second clinical trial. Br J Cancer 1982; 46:lOl Payne DG, Simpson WJ, Keen C, Platts ME. Malignant astrocytoma: Hyperfractionated and standard radiotherapy with chemotherapy in a randomized prospective clinical trial. Cancer 1982; 50: Freeman CR, Krischer J, Sanford RA, Furgery PC, Cohen M, Noms D. Hyperfractionated radiotherapy in brain stem tumors: Results of a pediatric oncology group study. Int J Radiat Oncol Biol Phys 1988; 15: Hochberg FH, Pruitt A. Assumptions in the radiotherapy of glioblastoma. Neuro1og.y 1980; 30: Kelly PJ, Daumas-Duport C, Kispert DB, Rall BA, Scheithauer BW, lllig JJ. Imaging-based sterotaxic serial biopsies in untreated intracranial glial neoplasms. J Neurosurg 1987: 66: Ramsey RG, Brand WN. Radiotherapy of glioblastoma multiforme. J Nerrrosurg 1973; 39: