1535 Final Report of a Phase 1/11 Trial of Hyperfractionated and Accelerated Hyperfractionated Radiation Therapy with Carmustine for Adults with Supratentorial Malignant Gliomas Radiation Therapy Oncology Group Study 83-02 Maria Werner-Wasik, M.D.' Charles B. Scott, M.S? Diana F. Nelson, M.D? Laurie E. Gaspar, M.D." Kevin J. Murray, M.D? Jennifer A. Fischbach, M . D ~ James S. Nelson, M.D.' Alan S. Weinstein, M . D . ~ Walter J. Curran, Jr., M.D? ' Department of Radiation Oncology, Thomas Jefferson University, Philadelphia, Pennsylvania. Radiation Therapy Onlcology Group, Philadelphia, Pennsylvania. Department of Radiation Oncology, University of Rochester, R'ochester, New York. ' Department Of Radiation Oncology, Wayne State University Detroit, Michigan. Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, Wisconsin. Department of Radiation Oncology, LDS Hospital, Salt Lake City, Utah. Department of Pathology, Louisiana State University, New Orleans, Lousiana. Hematology/Oncology Associates, Moorestown, New Jersey. Presented at the American Radium Society Annual Meeting, Paris, France, A.pril 28-May 2, 1995. Address for reprints: Maria Werner-Wasik, M.D., Department of Radiation Oncology, Bodine Center for Caiicer Treatment, Thomas Jefferson University Hospital, 111 South 11th Street, Philadelphia, PA 19107. Received June 1, 1995; revision received August 28, 1995; accepted August 28, 1995. (01996 American Canct?r Society BACKGROUND. Efforts to improve local control and survival by increasing the dose of once-daily radiation therapy beyond 70 Gray (Gy) for patients with malignant gliomas have as yet been unsuccessful. Hyperfractionated radiation therapy (HF) should allow for delivery of a higher total dose without increasing normal tissue late effects, whereas accelerated hyperfractionated radiation therapy (AHF) may minimize tumor repopulation by shortening overall treatment time. The Radiation Therapy Oncology Group (RTOG) conducted a randomized Phase 1/11 study of escalating doses of HF and AHF with carmustine (bis-chloroethyl nitrosourea [BCNU]) for adults with supratentorial glioblastoma multiforme (GBM) or anaplastic astrocytoma (AA). Primary study endpoints were overall survival and acute and chronic treatment-related toxicity. METHODS. From 1983 to 1989, 786 patients with supratentorial gliomas (81% with GBM and 19% with AA) were stratified by histology, age, and performance status and randomized to receive partial brain irradiation, utilizing either HF (1.2 Gy twice daily to doses of 64.8, 72, 76.8, or 81.6 Gy) or AHF (1.6 Gy twice daily to doses of 48 or 54.4 Gy). All patients received carmustine. The distribution of prognostic factors was similar in all arms. RESULTS. There were 747 eligible and analyzable patients among 786 enrolled patients (95%).Two patients had a Grade 5 and 65 patients had a Grade 4 chemotherapy toxicity. Two patients in the 81.6 Gy arm experienced late Grade 4 radiation toxicity and there was 1 late radiation-associated death in the 54.4 Gy arm. The rate of Grade 3 or worse radiation toxicity at 5 years, calculated by the delivered dose level, was 3% in the lowest total dose arms (48 and 54.4 Gy), 4% in the intermediate dose arms (64.8 and 72 Gy), and 5% in the highest dose arms (76.8 and 81.6 Gy) (P = 0.54). Survival rates at 2 and 5 years were: 21% and 11%, respectively, for all patients; 62% and 41%, respectively, for AA patients; and 10% and 4%, respectively, for GBM patients. There were no significant differences between the treatment arms with regard to median survival time (MST),when analyzed by the originally assigned dose. The MST for all patients varied between 10.8 months and 12.7 months ( P = 0.59); between 9.6 months and 11 months lor patients with GBM (P = 0.43); and between 30.4 months and 85.8 months for patients with AA (P = 0.78). Analysis of the survival rates for all patients by dose received rather than by dose assigned revealed a 14% 5-year survival rate for the lower HF doses (64.8 and 72 Gy), 11% for the higher doses (76.8 and 81.6 Gy), and 10% for the AHF doses (48 and 54.4 Gy)(P = 0.1). The subgroup of AA patients had a better MST (49.9 months) in the lower received HF doses than in the higher HF doses (34.6 months)(P = 0.35). In contrast, GBM patients who received 1536 CANCER April 15,1996 / Volume 77 / Number 8 the higher HF doses had survival superior to the patients in the AHF arms (MST of 11.6 months and 10.2 months, respectively; P = 0.04). CONCLUSIONS. The use of HF with BCNU and dose escalation up to 81.6 Gy is both feasible and tolerable, although late toxicity increases slightly with increasing dose. The best MST with the least toxicity were observed for AA in the lower received HF doses (72 and 64.8 Gy). Accordingly, 72 Gy in two 1.2 Gy fractions was used as the investigational arm of a completed Phase 111 trial (RTOG 90-06). In contrast, for GBM patients, longer survival times were noted in the higher received HF doses (76.8 and 81.6 Gy), suggesting the role for further dose escalation. The low toxicity rate with AHF arms suggest that further dose escalation is possible and is currently occurring in RTOG 94-11. Cancer 1996; 721535-93. 0 1996 American Cancer Society. KEYWORDS: glioma, glioblastoma multiforme, anaplastic astrocytoma, radiation therapy, hyperfractionated radiation therapy, accelerated hyperfractionated radiation therapy, bis-chlorethyl nitrosourea. M alignant gliomas constitute 3 3 4 5 % of primary brain tumors in adults’ and carry a very poor prognosis. Two main histologies are glioblastoma multiforme (GBM) and anaplastic astrocytoma (AA),with 5% and 18%, respectively, of patients surviving 5 years.’ Treatment modalities demonstrated to improve the survival of patients with malignant gliomas over surgery alone include external beam radiation therapy (RT),’,j and bis-chloroethyl (carmustine) (BCNU) or methylchloroethyl-cyclohexylnitrosourea (semustine) (MeCCNU) chemotherapy combined with external beam radiation therapy.’“ More recently, interstitial brain i m p l a n t ~either ,~ temporary-’ or permanent,8 and stereotactic radiosurgery9-” or stereotactic radiation therapy” have been used successfully in selected patients with malignant gliomas. However, despite major advances in treatment techniques, local tumor recurrence in the brain remains the overwhelming pattern of failure and is cause of death because no uniformly effective salvage therapy exists for those recurrent tumors. This suggests that the delivery of a higher dose of radiation to the tumor may potentially result in improved local control and therefore influence survival, because the length of survival of patients with GBM has been reported to correlate with dose in most series.l3,I4Walker et al.l5 described increasing median survival time (MST) from 18 weeks without RT to 28 weeks with 50 Gray (Gy), 36 weeks with 55 Gy, and 42 weeks with 60 Gy, in their secondary analysis of 3 successive Brain Tumor Study Group protocols. No additional improvement was seen in the Radiation Therapy Oncology Group 7401/Eastern Cooperative Oncology Group (RTOG 7401/ECOG) study4 for patients receiving a total dose to the tumor of 70 Gy versus those treated with 60 Gy. Treatment-related brain necrosis is a radiation dose-limiting factor, as seen in external beam radiation therapy combined with interstitial brain implants, in which a total dose to the tumor of 110-120 Gy was associated with a 40% incidence of reoperation necessary to alleviate mass effect, which in turn was related in 71% of patients to the presence of brain necrosis, with or without viable tumor cells.16The incidence of brain necrosis with external beam RT has been described by Marks et al.” as 0% with doses below 57 Gy to 17.9% for doses between 64.8 Gy and 75.6 Gy (with 1.8-2 Gy fraction sizes). As demonstrated in a classic study by Sheline et al.,IR the incidence of brain necrosis increases with an increasing radiation fraction size. This was further supported by Withers,I9 who observed that there is greater sparing of late effects for slowly proliferating tissues (such as brain tissue) with smaller fraction size. Therefore, delivery of a larger number of smaller fractions, or hyperfractionation (HF), may potentially allow a higher total dose delivered to the brain tumor and improved local control with the same probability of late effects. Several studies of multiple daily fractionated radiation therapy in gliomas reported conflicting result^.^"^'^ A dose of 61.41 Gy in 3 0.89 Gp daily fractions resulted in significantly improved survival (MST of 45 weeks) in patients with gliomas, when compared with a dose of 58 Gy given in 2 Gy, once-daily fractions (MST of 29 weeks).” Shin et al. reported a survival benefit to 40 Gy in 3 daily fractions to the whole brain, followed by a once-daily boost to a total of 50 Gy over 50 Gy given in a once-daily 2 Gy fraction.*’Another study (Brain Tumor Study Group [BTSG] 77-02) randomized patients to 1 of 4 arms: whole brain hyperfractionated radiation, 1.1 Gy twice daily to a dose of 66 Gy with BCNU, versus once-daily whole brain RT (2 Gy to 60 Gy), with either BCNU or BCNU and misonidazole, or streptozocin.” There was no significant difference between the 4 treatment arms (MST of 10.4 months for the hyperfractionated arm vs. 9.5 months for the once-daily RT with BCNU). Finally, Payne et al. found no improvement in survival with 36-40 Gy in 3 1 Gy fractions over 50 Gy in 2 Gy once-daily fraction^.'^ Radiation Therapy for Malignant GliomasMlerner-Wasik et al. Another fractionation scheme, in the form of accelerated hyperfractionated RT (AHF), (i.e., delivery of more than 1 standard size [ 1.6-2 Gy] fractions daily), may minimize tumor cell repopulation by shortening the overall treatment 1 ime, therefore increasing the probability of tumor control for a given dose 1 e ~ e l . l ~ In 198.3, the Radiation Therapy Oncology Group (RTOG)initiated a Phase 1/11RT dose escalation trial with BCNU for patients with supratentorial malignant gliomas (RTOG 83-02). The objective of the study was to determine the survival rate and the disease free survival rate for patients with malignant gliomas treated with HF or AHF to the progressively escalating doses, as well as to determine normal tissue toxicity. The preliminary results of the trial were reported p r e v i o u ~ l y . ~This ~ ~ ‘is~ the final report of that study. MATERIALS AND METHODS Patient Characteristics From 1983 to 1989, 786 patients with histologically confirmed unifcical supratentorial malignant gliomas, who were 18-70 years of age, and had a Karnofsky Performance Status (KPS) of at least 60 and an estimated survival of at least 8 weeks, were enrolled on the study. The patients were required to have normal hematologic parameters as well as normal renal and hepatic function. A normal chest X-ray or a diffusion capacity of at least 75% was required Written consent was obtained from the patient or the responsible family member. The informed consent form was in keeping with guidelines established by a local Human Investigations Committee upon recommendations of the RTOG Institutional Review Board, in accordance with assurances filed with and approved by the Department of Health and Human Services. Therapy had to start within 4 weeks from surgery. Patients with prior chemotherapy, head and neck RT, acquired immune deficiency syndrome (AIDS), or malignant neoplasms (except for skin cancers not on the head and neck or in situ carcinoma of the uterine cervix) were not eligible for the study. Eligible patients were stratified by histology (GBM vs. AA), age (18-44 years vs. 45-55 years vs. 56-70 years), and performance status (KPS > 80 vs. KPS < 80) and randomized to 1 of the 6 treatment arms (4 HF arms or 2 AHF arms)(Table 1). The randomization scheme as described by Zelen et al. to achieve institutional balance of treatment assignments was used with the three patient-related stratification variables.LG Pretreatment evaluation included a complete history and physical examination with a detailed neurologic examination and KPS assignment, as well as a complete blood count, blood chemistries, and a chest X-ray. Preoperative and postoperative brain computed tomography (CT) scans were mandatory. I537 Radiation Therapy Eligible patients were randomized to receive partial brain irradiation, utilizing either HF (1.2 Gy twice daily to doses of 64.8, 72, 76.8, or 81.6 Gy) or AHF (1.6 Gy twice daily to doses of 48 or 54.4 Gy). Within the HF arms, the initial randomization was to 3 arms, 64.8 Gy, 72 Gy, and 76.8 Gy. Because no Grade 3 or higher acute or late toxicity was found, the 81.6 Gy arm was opened in October, 1985, and the 64.8 Gy arm was closed at the same time. Subsequently, the 76.8 Gy arm was closed in February, 1986, and a 1:2 randomization favoring the 81.6 Gy arm (vs. the 72.0 Gy arm) was initiated, and is referred to as a “final randomization.” Within the AHF arms, initial plans calYed for dose escalation to 60.8 Gy. However, because 1 death was seen during the first year of accrual with the 54.4 Gy dose and was thought attributable to the treatmentrelated toxic effects, the study accrual was closed after completing accrual to the initial 2 treatment arms. Megavoltage machines with energies of 1.25-10 mv were used with source to axis distance of at least 80 cm. The target volumes were based on the preoperative CT scan. The initial target volume included the contrast-enhancing lesion and surrounding edema with a 2-cm margin and was carried to a dose of 57.6 Gy on the HF arms and 35.2 Gy on the AHF arms. The cone-down volume included the contrast-enhancing lesion only plus a 2.5-cm margin and was treated to the assigned total dose. The maxiniuin dose to the optic chiasm was recommended to be 60 Gy, to the retina, 50 Gy, and to the brainstem, 60 Gy. The dose was specified on the central ray at midseparation of beams for the parallel opposed beams and at the intersection of the central ray of the beams for an arrangement of two or more intersecting beams. RT was delivered with two daily fractions, separated by 4 to 8 hours, 5 days per week. A minimum dose of 4 mg of dexamethasone was recommended during the RT course. Chemotherapy All patients received BCNU, 80 mg/m’ intravenously, on Days 1, 2, and 3 of the first week of RT and subsequently on 3 consecutive days every 8 weeks for a period of 1 year, to a maximum dose of 1440 mglm‘. For patients older than 60 years, the BCNU dose during the first cycle was decreased to 75% of a calculated full dose and increased again to the full dose during next cycles, if no Grade 2 or higher hematologic toxicity was observed. BCNU doses were modified according to the criteria presented in a report from a prior RTOG trial.28 Histopathology A central pathology review process was used in the study with all slides reviewed by one pathologist (J. S. N.) according to the RTOG/ECOG criteria. For a diagnosis of GBM, a marked hypercellularity, foci of coagulation ne- 1538 CANCER April 15,1996 / Volume 77 / Number 8 TABLE 1 Distributlon of Patient Characteristicsby Treatment % Male Age % < 40 % 40-59 % 60t Mean Range KPS % 80-100 % 60-70 % < 50 Neurologic %Work % Home % Hospital Previous surgery % Biopsy only % Partial resection % Total resection % Shunt & partial Other Tumor size (Institution) X<5cm % 5-10 cm ?’& >I0 cm % Unknown Tumor size (central review) %<5cm % 5-10 cm Histology (Institutional review) 96 Astrocytoma % Glioblastoma W Other % Unknown Histology (Central review) % Astrocytoma % Gfiobfastorna % Other 64.8 Gy (N = 78) 72 Gy (N = 158) 76.8 Gy IN = 86) 81.6 Gy (N = 120) 48 (N = 168) 54.4 Gy (N = 137) Total (N = 747) 67 62 5a 64 62 65 63 14 46 40 53.2 26-70 20 46 34 51.2 21-70 23f 43 34 51.7 22.69 20 47 33 51.1 21-69 16 46 38 53.2 19-70 ia 39 43 53.1 19-70 19 44 37 52.3 19-70 64 30 6 69 27 4 70 27 3 73 27 0 70 30 0 70 30 0 70 2a 2 50 40 10 51 42 7 41 51 a 56 40 4 55 42 4 51 44 5 51 43 6 15 6 ia 1 0 23 56 21 0 0 17 69 14 25 54 20 0 0 1 0 31 51 17 0 1 27 72 22 0 0 24 56 19 <l <I 33 53 3 11 37 49 3 11 30 52 3 14 41 50 4 5 46 45 2 7 4a 4a 2 2 41 49 2 8 34 66 (N = 56) 42 5a (N = 132) 32 68 48 52 47 53 41 59 (N = 69) 37 63 (N = 100) (N = 144) (N = 109) (N = 610) 26 72 2 0 32 68 0 0 28 69 3 0 32 68 0 0 27 72 <1 <1 29 70 1 0 29 70 12 21 79 0 (N = 1481 19 a1 0 (N = 84) 18 81 1 (N = 112) 22 78 0 20 a0 0 (N = 127) 19 a1 <I (N = 690) aa 0 (N = 75) crosis, and vascular mural cell proliferation had to be present. Mixed glioblastoma and fibrosarcoma (gliosarcoma) were included in the GBM group. For a diagnosis of AA, one or more of the following features had to be present: (1) increased cellularity; (2) pleomorphic nuclei or cell bodies; (3) mitotic figures; (4) increase in blood vessels with mild endothelial proliferation; and ( 5 ) spongioblastic or other incompletely differentiated glial cell^.'^ Patient Follow-Up Patients were observed weekly during RT and monthly thereafter. CT scans of the brain were obtained 4 months (N = 144) 1 <1 after initiation of therapy, then every 3 months, and at the time of neurologic deterioration. Disease free survival and overall survival were measured from the day of randomization until documented deterioration by neurologic exams and CT scans, or until death. Treatment-related effects on normal tissue were evaluated using RTOGlECOG toxicity criteria (Table 2) and defined as acute if occurring within 90 days of the start of treatment, or late if occurring or persisting beyond 90 days. Major effects included Grades 3 and higher, with Grade 5 scored as fatal effects. All suspected major toxic effects were reviewed by one of the study chairpersons Radiation Therapy for Malignant GliomasMTerner-Wasik et al. TABLE 2 Toxicity and Late Effects Criteria Organ/tissue Grade 3 Grade 4 Moist desquamation Marked atrophy Ulceration, necrosis Ulceration Alteration affecting function <50oio of time of function Function 2 50% affected Function 2 50% affected Not controlled by medication Comatose Skin Acute Chronic Neurologic Mental status Motor paresis Cerebellar function Seizures Paralysis Confined to bed Status epilepticus Note: Anv toxic effect causing death was eraded as 5 (D. E. N. or W. J. C.) and only ascribed to the treatment received, if there was no evidence of recurrent or progressive tumor on the CT scan or pathologic review. Statistical Methods Differences in pretreatment characteristics between the randomized treatment arms were evaluated with the Pearson chi-square test for differences in proportions. The estimated probability for a late toxicity was calculated using a cumulative incidence that adjusts for patients dying without toxi~ity.~’ Differences in late toxicities were compared using the test statistic proposed by Gray.3’All time events, including survival, were calculated from the date of randomization. Survival was estimated using the product-limit m e t h ~ d . Differences ~’ in survival distributions were compared using the log rank statistiC.29,33 Patients receiving less than 44.8 Gy were not included in the dose received analysis. All P values presented are two-sided. RESULTS Patient Characteristics and Protocol Violations Distribution of patient characteristics by treatment arm is presented in Table 1. Distribution of prognostic factors was similar in all arms. The mean patient age was 52.3 years and 63% of the patients were males. Median follow-up was 11.7 months (range, 0.01- 126 months). Distribution of prognostic factors was similar in all arms. On the average, 70% of patients had KPS of 80% or greater, 19% of patients underwent total resection, 56% underwent partial resection, and 51% had tumors of larger than 5 cm in diameter on central review (W. J. C.) (Table 1). There was a discrepancy between the institutional and central histology assignment, which has been documented e l ~ e w h e r e Seventy .~~ percent of patients were diagnosed as having a GBM in their institutions, but 81%had GBM on the central review. 1539 Twenty-nine percent had AA on institutional review versus 19% on central review. The total number of patients entered on trial was 786. No On-Study form was available for 11 patients (l%),and 28 patients (3.5%) canceled or were ineligible. Overall, 747 of 786 patients (95%) were eligible and analyzable. Major unacceptable protocol violations were found in 5.3% of patients with regard to radiotherapy delivery. A radiation therapy major unacceptable deviation was scored when the total dose deviated by 13% or greater from the randomized dose at the central axis, when there was interruption of the hyperfractionated schedule by 8 or more days, or when either the initial or the cone-down field did not include the entire tumor volume. Radiation therapy dose was delivered as per protocol guidelines in 650 of 745 of patients (87%).The proportions of patients whose delivered dose was within 5 2.4 Gy from the assigned dose (or ? 3.2 Gy for AHF arms only) were as follows: 90% for the 64.8 Gy arm; 90% for the 72 Gy arm, 82% for the 76.8 Gy arm; 79% for the 81.6 Gy arm, 94% for 48 Gy arm, and 93% for the 54.4 Gy arm. Four percent of all patients received a dose of less than 44.8 Gy and less than 0.01% of patients were given more than 84 Gy. Major unacceptable protocol violations were found in 3% of patients with regard to chemotherapy delivery. A chemotherapy major unacceptable deviation was scored when failure to modify the BCNU dose resulted in termination of therapy, when protocol therapy was not given, or when it was interrupted for more than one cycle unrelated to toxicity or disease progression. Treatment-Related Toxicity Two patients died because of chemotherapy-related fatal cytopenia andlor infection (one in the 76.8 Gy arm and one in the 54.4 Gy arm) and 68 patients experienced a Grade 4 chemotherapy toxicity. The most common acute chemotherapy toxicities were nausea and vomiting (89% of patients), leukopenia (78%),thrombocytopenia (72%), and gastrointestinal symptoms (39%). The most common acute radiation toxicities were skin reactions (74% of patients), followed by somnolence (13%),ototoxicity (lo%), and mental status change (8%). Overall, 14 patients (0.02%)experienced Grade 3 or worse acute toxicity. There were no early radiation-associated deaths and there was 1 late death in the 54.4 Gy arm. Two patients experienced late Grade 4 (1 hearing loss and 1 increased intracranial pressure) radiation toxicity, both in the 81.6 Gy arm, and 14 patients experienced Grade 3 toxicity (4 in the 76.8 Gy arm). Overall, 27% patients suffered chronic skin changes, 13% memory change, 14% somnolence, 12% mental status change, and 6% each hearing loss and increased intracranial pressure. At 5 years, 3% patients in either AHF arm had Grade 1540 CANCER April 15,1996 / Volume 77 / Number 8 TABLE 3 RTOG 83-02 Survival by Assigned Treatment AHF Low Dose HF RTOG 83-02:SURVIVAL FOR AA PATIENTS BY DOSE LEVEL RECEIVED High Dose HF Arm CY 48 54.4 64.8 72 76.8 81.6 Pvalue All patients GBM 11.9 10.2 35.0 10.8 10.4 40.6 11.4 9.6 85.8 12.7 11 50 12.0 10.9 30.4 11.7 10.2 35.4 0.598 0.43 R AA E m T 0.78 ; RTOC 83-02 Radiation Therapy Oncology Group Study 83-02. AHF: accelerated hyperfractionated radiation therapy; HF: hyperfractionated radiation therapy; Gy: Gray: G B M glioblastoma multiforme; AA: anaplastic astrocytoma. 40 I V E 20 ~ 0 1 RTOG 83-02: SURVIVAL FOR ALL PATIENTS BY ASSIGNED TREATMENT 2 ~~~~~~ 6 4 3 YEARS FROM ONSTUDY FIGURE 2. Radiation Therapy Oncology Group Study 83-02 survival for patients with anaplastic astrocytoma by dose level received. TABLE 4 RTOG 83-02 Survival by Dose Level Received Dose level Patient group 0 All patients ? 0 1 2 3 4 5 YEARS FROM ONSTUDY FIGURE 1. Radiation Therapy Oncology Group Study 83-02 survival for all patients by assigned treatment. 3 or worse toxicity versus 3% in the 64.8 Gy arm, 4% in the 72 Gy arm, 6% in the 76.8 Gy and 5% in the 81.6 Gy arm. The actuarial rate of Grade 3 or worse radiation toxicity at 5 years, calculated by the delivered dose, was 3% in the lowest total dose arms (48 and 54.4 Gy), 4% in the intermediate dose arms (64.8 and 72 Gy), and 5% in the highest dose arms (76.8 and 81.6 Gy) ( P = 0.54). Two percent of patients whose delivered dose was lowest (in 48 and 54.4 Gy arms) had this type of toxicity at 5 years, versus 5% patients whose delivered dose was highest (76.8 and 81.6 Gy) (P = 0.14). Survival Eighty-nine percent (663 of 747) patients died over the study period. Overall survival rates at 2 and 5 years were 21% and 11%,respectively, for all patients; 62% and 41%, respectively, for AA patients; and 10% and 4%, respectiveiy, for GBM patients. MST was 11.8 months for all patients, 40.3 months for AA patients, and 10.6 months for GBM patients. MST by the assigned treatment arm AA GBM AHF (48 and 54.4) Low dose HF (64 and 72) High dose HF (76 and 81) P value 11.7 40.3 10.2 12.6 49.93 11.02 12.4 34.6 11.6 0.10 0.35 0.04 RTOG: 83-02: Radiation Therapy Oncology Group Study 83-02; AHF: accelerated hypetftactionated radiation therapy: HF. hyperfractionated radiation therapy; AA: anaplastic astrocytoma; GBM: glioblastoma multiforme. for all patients was as follows: on the HF arms, 11.4 months for 64.8 Gy, 12.7 months for 72 Gy, 12 months for 76.8 Gy, and 11.7 months for 81.6 Gy (P= 0.598). For the A H F arms, MST were 11.9 months for 48 Gy and 10.8 months for 54.4 Gy (Table 3 and Fig 1). MST for GBM and AA patients by the assigned treatment arm are also depicted in Table 3. There were no significant differences between the treatment arms with regard to MST (Table 3) or 5-year survival rates, when analyzed by the originally assigned dose. Analysis of survival performed by dose actually received on any single treatment arm did not reveal any advantage of one particular treatment arm over another. When analysis of sunrival was performed by the received dose levels (rather than single arms), defined as “lower HF dose” (64.8 Gy and 72 Gy) versus “higher HF dose” (76.8 Gy and 81.6 Gy) versus “AHF dose” (48 Gy and 54.4 Gy), the following observations became manifest. For all patients, MST was 11.7 months in the AHF dose level, 12.6 months in the low dose HF level, Radiation Therapy for Malignant GliomasMlerner-Wasik et al. 1541 TABLE 5 RTOG 83-OZSURVIVM FOR GEM PATIENTS W DOSE LEML RECEIVED RTOG 83-02 Survival by Radiation Therapy Fractionation (HF vs AHP) .I b 6 4 4 (11 64 IL 72 GY Patient group AHF HF P value AA GBM 40.3 42.30 10.2 10.80 0.67 0.08 7 6 6 L 1 1 1 GI . 0 1 2 3 4 . . 5 E A R S FROM ONSNDY FIGURE 3. Radiation Therapy Oncology Group Study 83-02 survival for patients with supratentorial glioblastorna rnultiforrne by dose level received. and 12.4 months in the high dose HF level (P = 0.10). AA patients had better MST (49.9 months) in the lower HF dose level than in the higher HF dose level (34.6 months) and in the AHF dose level (40.3 months) ( P = 0.35) (Table 3 and Fig. 2 ) . In contrast, GBM patients who received the higher HF doses had superior survival (11.6 months), when compared with the lower HF dose level (11.0 months) and AHF dose level (10.2 months) ( P = 0.04) (Table 4 and Fig. 3). Patients' survival was further analyzed by the type of assigned fractionation regimen used (AHF vs. HF). Only 3% of GBM patients treated on AHF arms were alive at 5 years versus 5% of patients on the HF arms, with MST of 10.2 months and 10.8 months, respectively ( P = 0.08). Corresponding values for AA patients were 40% 5-year survival on the AHF arms versus 41% on the HF arms, with MST of 40.3 months and 42.3 months, respectively ( P = 0.67) (Table 5). DISCUSSION RTOG 83-02 was a Phase 1/11 study that was designed for the purpose of identifying a potential RT dose, delivered with either a tWF or HF regimen, that might be associated with improved survival of patients with malignant gliomas. At the same time, toxicity of each treatment arm was prospectively assessed. A trend (of increasing Grade 3 or higher late treatment-related toxicity with the increasing total dose level was identified, although the difference between the toxicities of the lowest dose level (48 and 54.4 Gy) and the highest dose level (76.8 and 81.6 Gy) was statistically not significant (f' = 0.54). Overall, the incidence of severe toxic effects attributable to the delivered treatment was low (3-5%), which justifies the conclusion that further dose escalation may be feasible in both the AHF and HF groups. No difference in adverse effects was seen between the AHF and HF subgroups, but this might have been due to the lower equivalent doses used in the AHF arms. The originally planned dose escalation above 54.4 Gy in the AHF subgroup did not occur because of 1 death observed during the study, believed to be attributable to the treatment, and subsequently confirmed as such. The incidence of Grade 6 white matter changes, as reported by Corn et al.,35observed in patients on the AHF a r m was lowest (1.6%) when compared with the intermediaie HF dose range (64.8 Gy and 72 Gy) (4.6%) or the high HF dose range (76.8 Gy and 81.6 Gy) (19.2%), further supporting a feasibility of dose escalation for AHF. Higher AHF doses (64 Gy and 70.4 Gy) are currently employed in a Phase I1 dose escalation trial for radiosurgery-ineligible GBM patients (RTOG 94-11). There are several difficulties in making a precise evaluation of treatment-related brain injury and distinguishing it from recurrent or progressive tumor. Standard radiographic techniques, such as CT scans or magnetic resonance imaging (MRI) scans are unable to differentiate those two entities with absolute certainty. Even two promising modalities, 18F-FDG (fluorine-18-2-deoxy-2!fluoro-D-glucose) scans and thallium-201 single photon emission computed tomography, although valuable in distinguishing necrosis from tumor, '' will require more time before their role is fully assessed. Consequently, the scoring of treatment toxicity is imperfect. In addition, because the survival of patients with gliomas is relatively short, the full extent of late sequelae may not become manifest before their death. The analysis of white matter changes on follow-up MRI and CT scans available for 177 patients treated on the RTOG 83-02 study confirmed the significant correlation of the incidence of severe white matter changes with the total dose of HF RT, although no correlation between Grade 2 or worse clinically observed toxicity and those changes was found.'5 Due to these constraints, survival has been used as a measure of both treatment efficacy and treatment toxicity. There was no dramatic survival advantage observed in this study for any of the treatment arms, whether analyzed by the assigned or delivered dose, because the study was not specifically designed to achieve significant overall 1542 CANCER April 15,1996 / Volume 77 / Number 8 differences i n survival, but rather to choose t h e best of a number of treatment options. In fact, two different “best” arms were identified for different histologies of malignant gliomas. Patients with GBM, who received t h e highest doses (76.8 Gy and 81.6 Gy) survived significantly longer ( P = 0.041, when compared with patients receiving 48 Gy and 54.4 Gy. Although the absolute gain of 1.4 m o n t h in MST may n o t be clinically meaningful, it clearly points t o t h e need for using t h e highest possible doses for t h e treatment of GBM. In addition, it suggests t h e role for further dose escalation, because an acceptable Grade 3 or higher toxicity incidence (5%) was seen even with those highest doses and no fatal toxicity was recorded. In contrast t o patients with GBM, patients with AA had best survival rates in t h e 64 Gy and 72 Gy arms ( P = 0.35) and n o t in t h e 76.8 Gy and 81.6 Gy arms, presumably because of excessive treatment toxicity. This observation was first reported in t h e initial analysis of t h e ~ t u d y , ‘ ~ b u t was applicable for b o t h histologies (AA and GBM). Quality-adjusted survival analysis of all glioma patients treated on RTOG 83-0237showed t h e longest quality-adjusted survival for patients treated on t h e 72 Gy and 64.8 Gy arms, with younger patients and patients with a high KPS benefiting m o s t from t h e assignment t o t h e 72 Gy arm. Accordingly, a recently completed Phase I11 RTOG trial (90-06) used a dose of 72 Gy in 1.2 Gy twice-daily fractions plus BCNU as an investigational arm, which was compared with 60 Gy delivered in 2 Gy fractions plus BCNU a n d will therefore enable a direct comparison of HF and once-daily RT. Larramore e t al. reviewed survival of patients with AA treated on four RTOG studies, two of which included n e u t r o n irradiation and one, misonidazole as radiation sensitizer, in addition t o photon external beam RT and BCNU.38The authors observed a decreasing survival as t h e “aggressiveness” of t h e treatment increased. The weakness of t h e study lies in comparing completely different treatment modalities. Because AA appear t o respond differently t o therapy t h a n GBM, these two entities may need to be studied i n separate trials in 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. t h e future. Although interstitial brachytherapy for gliomas6”6 or stereotactic radiosurgery“ seem promising, with impressive survivals reported i n subsets of patients wirh gliomas, a majority of patients with glioma do n o t qualify for brachytherapy or radiosurgery because of the size of their tumor~l”3Y For those patients, efforts t o optimize external beam radiation therapy m u s t continue, including identifymg the most effective and tolerable HF or AHF regimens. 15. REFERENCES 16. 1. Mahley SM, Mettlin C, Natarajan N, Laws ER, Peace BB. National survey of patterns of care for brain tumor patients. J Neurosurg 1989;71:826-36. 13. 14. Walker MD, Green SB, Byar DP, Alexander E, Batzdorf U, Brooks WH, et al. Randomized comparisons of radiotherapy and nitrosoureas for the treatment of malignant glioma after surgery. N Engl J Med 1980;303:1323-9. Kristiansen K, Hagen S, Kollevold T, Torvik A, Holme I, Stat M, et al. Combined modality therapy for operated astrocytomas grade 111and IV.Confirmation of the value of postoperative irradiation and lack of potentiation of bleomycin on survival time. A prospective multicenter trial of the Scandinavian glioblastoma study group. Cancer 1981;47:649-52. Chang CH, Horton J, Schoenfeld D, Salazer 0,Perez-Tamayo R, Kramer S, et al. Comparison of postoperative radiotherapy and combined postoperative radiotherapy and chemotherapy in the multidisciplinary management of malignant gliomas. Cancer 1983;52:997- 1007. Gutin PH, Prados MD, Phillips TL, Wara WM, Larson DA, Leibel SA, et al. External irradiation followed by an interstitial high activity iodine-125 implant “boost” in the initial treatment of malignant gliomas: NCOG study 6G-82-2. Int J Radiat Oncol Biol Phys 1991;21:601-6. Green SB, Shapiro WR, Burger PC, Selker RG, Van Gilder JC, Saris S, et al. A randomized trial of interstitial radiotherapy (RT) boost for newly diagnosed malignant glioma: brain tumor cooperative group (BTCG) Trial 8701. Proc. Am. SOC. Clin. Oncol. 1994;13:174. Gutin PH, Leibel SA,Wara WM, Choucair A, Levin VA, Philips TL, et al. Recurrent malignant gliomas: survival following interstitial brachytherapy with high-activity iodine-125 sources. J Neurosurg 1987:67:864-73. Gaspar L, Zamorano L, Garcia L, Shamsa F, Warmelink C, Yakar D. Malignant gliomas: permanent iodine-125 implants. Int J Radiut Oncol Biol Phys 1994;30 (1 Suppl):295 (1095). Loeffler JS, Alexander E, Shea WM, Wen PY, Fine HA, Kooy HM, et al. Radiosurgery as part of the initial management of patients with malignant glioma. J Clin Oncol 1992; 10:1379-85. Masciopinto JE, Levin AB, Mehta MP, Rhode BS. Stereotactic radiosurgery for glioblastoma: a final report of 31 patients. J Neurosurg 1995;82:530-5. Mehta MP, Masciopinto J, Rosental J, Levin A, Chappell R, Bastin K, et al. Stereotactic radiosurgery for glioblastoma multiforme: report of a prospective study evaluating prognostic factors and analyzing long-term survival advantage. Int J Radiat Oncol Biol Phys 1994;30(3):541-9. Brada M, Laing RW, Graham J, Warrington AP, Hines F. Fractionated stereotactic radiotherapy in the treatment of recurrent high grade glioma: dose escalation study. Acta Neurochir (Wien) 1993;122:151-5. Rutten EHJM, Kazem I, Slooff JL, Walder AHD. Postoperative radiation therapy in the management of brain astrocytomata-retrospective study of 142 patients. IntJRadiat Oncol Biol Phys 1981;7:191-5. Salazar OM, Rubin P, Feldstein ML, Pizzutiello R. High dose radiation therapy in the treatment of malignant gliomas: final report. Int J Radiat Oncol Biol Phys 1979;5:1733-40. Walker MD, Strike TA, Sheline GE. An analysis of dose-effect relationship in the radiotherapy of malignant gliomas. Int J Radiat Oncol Biol Phys 1979;5:1725-31. Scharfen CO, Sneed PK, Wara WM, Larson DA, Phillips TL, Prados MD, et al. High activity iodine-125 interstitial implant for gliomas. Int J Radiut Oncol Biol Phys 1992;24:58391. Radiation Therapy for Malignant GliomasiWerner-Wasik et al. 17. Marks JE, Baglan RJ, Prassad SC, Blank WF. Cerebral radionecrosis: incidence and risk in relation to dose, time, fractionation and volume. Int / Radiat Oncol Biol Phys 1981;7~243-52. 18. Sheline G, Wara WM, Smith V. Therapeutic irradiation and brain surgery. Int/ Radiat Oncol Biol Phys 1980;6:1215-28. 19. Withers HR. Biologic basis for altered fractionation schemes. Cancer 1985;552086-95. 20. Fulton DS, Urtasun RC, Shin KH, Geggie PHS, Thomas H, Muller PJ, et al. Misonidazole combined with hyperfractionation in the management of malignant glioma. Int J Radiat Oncol Biol Phys 1984;10:1709-12. 21. Shin KH, Muller PH, Geggie PHS. Superfractionation radiation therapy in the treatment of malignant astrocytoma. Cancer 1983;52:2040-3. 22. Deutsch M, Green SB, Strike TA, Burger PC, Robertson JT, Selker RG, et al. Results of a randomized trial comparing BCNU plus radiotherapy, streptozocin plus radiotherapy, BCNU plus hyperfractionated radiotherapy, and BCNU following misonidazole plus radiotherapy in the postoperative treatment of malignant glioma. Int J Rudiut Oncol Biol Phys 1989;16:1389-96. 23. Payne D, Simpson WJ, Keen C, Platts ME. Malignant astrocytoma: hyperfractionated and standard radiotherapy with chemotherapy in a randomized prospective clinical trial. Cancer 1982;50:2301-6. 24. Curran WJ, Scott CB, Nelson JS, Weinstein AS, Phillips TL, Murray K, et al. A randomized trial of accelerated hyperfractionated radiation therapy and BCNU for malignant glioma. Cancer 1992;70:2909- 17. 25. Nelson DF, Curran WJ, Scott C, Nelson JS, Weinstein AS, Ahmad K, et al. Hyperfractionated radiation therapy and bis-chlorethyl nitrosourea in the treatment of malignant glioma-possible advantage observed at 72.0 Gy in 1.2 Gy BID fractions: report of the RTOG protocol 8302. Int / Radiat Oncol Biol Phys 1993;25:193-207. 26. Zelen M. The randomization and stratification of patients of clinical rials. Int J Radiat Oncol Biol Phys 1979;5:172531. 27. Nelson DF, Schoenfeld D, Weinstein AS, Nelson JS, Wasserman Goodman et aL A randomized Of misonidazole sensitized radiotherapy plus BCNU and radiotherapy plus BCNU for treatment of malignant glioma after surgery: preliminary results of an RTOG study. I n t J Radiat Oncol BiolPhys 1983;9:1143-51. Tp RLp 1543 28. Nelson JS, Schoenfeld D, Tsukada Y, Fulling K, Lamarche J, Peress N. Necrosis as a prognostic factor criterion in malignant, supratentorial astrocytic gliomas. Cancer 1983;525504. 29. Pet0 R, Pike MC, Armitage P, Breslow NE, Cox DR, Howard SV, et al. Design and analysis of randomized trials required prolonged observation of each patient: 11. analysis and examples. B r / Cancer 1977;35:1-39. 30. Caplan RJ, Pajak TF, Cox JD. Analysis of the probability and risk of cause-specific failure. Int J Radiat Oncol Biol Phys 1994;29:1183-6. 31. Gray RJ. A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat 1988;16:114154. 32. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J A m Star Assoc 1958;53:457-81. 33. Mantel N. Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 1966;50:163-70. 34. Scott CB, Nelson JS, Farnan NC, Curran WJ, Murray KJ, Fischbach AJ, et al. Central pathology review in clinical trials for patients with malignant glioma: a report of RTOG 83-02. Cancer 1995;76:307-13. 35. Corn BW, Yousem DM, Scott CB, Rotman M, Asbell S, Nelson DF, et al. White matter changes are correlated significantly with radiation dose. Cancer 1994;74:2828-35. 36. Kosuda S, Fuji H, Aoki S, Suzuki K, Tanaka Y, Nakamura 0, et al. Prediction of survival in patients with suspected recurrent cerebral tumors by quantitative thallium-201 single photon emission computed tomography. Inr / Radiat Oncol Bid Phys 1994;30:1201-6. 37. Murray KJ, Nelson DF, Issacson S, Scott C, Fischbach JA, Porter A, et al. Quality adjusted survival analysis of malignant glioma patients treated with twice-daily radiation and carmustine: a report of RTOG 83-02. Int J Radiat Oncol Biol Phys 1995;31:453-9. 38. Laramore GE, Martz KL, Nelson JS, Griffin TW, Chang CH, Horton J. Radiation therapy oncology group (RTOG! suMval data on anaplastic astrowomas of the brain: does a more aggressive form of treatment adversely impact survival? Int / Radiat Oncol Biol Phys 1989;17:1351-6. 39. Curran WJ, Scott CB, Weinstein AS, Martin L, Nelson JS, Phillips TL, et al. Survival comparison of radiosurgery-eligible and -ineligible malignant glioma patients treated with hyperfractionated radiation therapy and carmustine: a report of RTOG 83-02. J Clin Oncol 1993;11:857-62.