1216 Radiation Oncology 0 Biology 0 Physics

25 CEREBELLAR MEDULLOBLASTOMA

fkptember 1981, Volume 7, Number 9

Craig Silverman, M.D., Joseph R. Simpson, M.D.

Washington University School of Medicine Ciallinckrodt Institute of Radioloqy

Division of Radiation Oncology St. Louis, Missouri 63108

The records of 46 patients with cerebellar medulloblastoma treated at our institution between 1954 and 1977 were analyzed for prognostic features, tumor control, patterns of failure and NED survival followina suraical resection and radiation therapy. Three-year and five-year NED s&viva1 was 48% and 42%. Tumor was controlled in the posterior fossa in 52% of all patients. Eighty-five percent of local failures occurred within three years of treatment. All spinal axis failures were associated with concomitant posterior fossa failure whereas all extra-nervous system metastases, i.e., to bones and lymph nodes occurred in patients whose primary tumor was controlled. There were seven (7) patients with cerebral failures of which five (5) failed solely in the brain, three of five in periventricular locations, two of five in the frontal lobes. These were usually late failures.

Those patients receiving at least 4500 rad to the posterior fossa had a local control rate of 63% while those receiving less than 4500 rad had a con- 'trol rate of 37%. Five-year local control for those receiving greater than 5000 rad to the posterior fossa was 82% (14/17). T staging, as proposed by Harisiadis and Chang, was not found to be of any prognostic significance.

Analysis with regard to technique demonstrated that when the posterior fossa was adequately covered with lateral fields that extended below the level of foramen magnum, the five-year survival was 52% versus 26% when the fields were considered inadequate, and local control was 63% versus 45%.

MEDULLOBLASTOMA:

Identification and Implications of Prognostic Subgroups

Gene Kopelson'MD, Rita M. LinggoodlMD, George M. Kleinman2MD

'Department of Radiation Medicine and 2Neuropathology Laboratory of the Department of Pathology, Massachusetts General Hospital,

Boston, MA 02114

Forty-three medulloblastoma patients seen from 1962-1979 were evaluated for prognostic factors and post-irradiation local control rates as a function of a new histopathologic subclassification scheme (TS=Tumor Score, a function of necrosis, desmoplasia, mitoses, and cytoplasmic processes), TM operative staging, as well as classical clinical parameters.

An overall actuarial 5-and lo-year survival rate of 55% was achieved. Factors from the literature which previously have had prognostic

importance (age. sex) did not affect prognosis in this series. -The most siqnificant parameters wern staqe and TS. Five-year survival rates were better for: low T-stage (82% Tl,2_vs 46% T3,4, P <5.02), low M-stage (63% MO.1 vs. 0% M2,3, P < 0.03), and favorable histopathologic subtype (81% TS < 5 vs 41% TS > 6, P ~0.05). Posterior fossa local control rates post- irrzdiation were a function of T-stage (90% T1.2 vs 38% T3,4), and histology (83% TS < 5 vs 38% TS > 6). Combining T-stage and TS resulted in 3 major prognostic and local-cZntro1 groups:

Small (Tl,2) tumors of favorable histology (TS < 5) had a 5-and lo-year survival of 92% with 100% (8/8) local control, No change from present management is suggested.

An intermediate prognosis (survival rates of 67% and 70%, respectively) was found for small tumors (Tl,2) of unfavorable histology (TS > 6) and for large tumors (T3,4) of favorable histology (TS < 5). These patTents showed evidence of an irradiation dose-response; local-control was achieved in O/3 receiving < 5000 rad vs 3/3 receiving > 5000 rad. In this group, by only increasing the posterior fossa dose; Improved survival may-result..

The poor proanosis arouo (T3.4-TS > 6) had a 5-vear survival of onlv 42% and increasing-the poiterior fossa doze to > 5000 rad did not increase- local control. It is suggested that patients for future adjuvant chemotherapy or radiosensitizer trials be selected from this poor risk group of patients with large and/or necrotic tumors.

Proceedings of the 23rd ANNA ASTR Meeting 1217

This paper identifies prognostic subgroups based upon histology and TM staging in medulloblastoma patients, which potentially may be utilized to improve therapeutic results.

27 SUPERFRACTIONATION RADIATION THERAPY IN THE TREATMENT OF MALIGNANT ASTROCYTOMA

Kyu H. Shin, M.D., F.R.C.P.(C) Paul J. Muller, M.D., F.R.C.S.(C)

Peter H.S. Geggie, M.D., F.R.C.P.(C)

Southern Alberta Cancer Centre and University of Calgary 2104 - 2nd Street, S.W., Calgary, Alberta, Canada

To evaluate the benefit of the superfractionation radiation therapy (SF) over conventional fractionation (CF) in the treatment of malignant astrocytoma, a randomized prospective clinical trial was started in July 1978 and closed in December 1980. Thirty- five patients were randomized to SF and received 5,000 rads/four weeks/50 fractions. Thirty-five patients were randomized to CF and received 5,000 rads/five weeks/25 fractions. Following radi- ation therapy, all patients were assigned to CCNU chemotherapy.

Chi squares were done to assess the possible differences in key variables in the two groups and there were no differences in sex, age, performance stuatus, location of tumour, size of the tumour, degree of surgical removal and pathology.

The median survival time is ten months for SF group and six months for CF group. The one and two year actuarial survival rate is 58% and 32% for SF group while 39% and 9% for CF group.

A trend suggesting the superiority of SF has been identified and merits further study with a higher radiation dose.

28 MALIGNANT GLIOMA: HYPERFRACTIONATED AND STANDARD RADIOTHERAPY AND CHEMOTHERAPY IN A RANDOMIZED

PROSPECTIVE CLINICAL TRAIL

Dr. D.G. Payne, Dr. W. J. Simpson, Dr. C. Keen Dr. M.E. Platts

The Princess Margaret Hospital, Toronto, Ontario, Canada

A prospective randomized controlled trail of 157 patients with malignant astrocytoma (grade III or IV) was carried out at a single institution. The minimization technique ensured balanced distribution of prognostic factors between the treatment groups. All receivgloral lomustine (CCNU, 80 mg/m2) six weekly and hydroxyurea (HU, 3.5 gm/m* over 5 days) three weekly, for one year or until recurrence, with doses adjusted for myelosuppression. Patient5 were randomized to Daily (5,000 rads in 25 fractions (fr) in 5 weeks)or Q3h (3,600- 4,000 rads in 36-40 fr of 100 rads each, given 4 fr per day at 3 hourly inter- vals over 2 weeks) Cobalt-60 irradiation. Partial brain fields (120 cm2 were used for most patients. Steroid therapy (up to 16 mg/day Dexamethasone) was permitted.

Complications were moderate in both groups. There was one death on treatment due to brain edema. Chemotherapy dose reductions were required by myelosuppression in 18 patients. No significant survival or toxicity differences were seen between the two groups. Age, initial performance status, and extent of surgical resection were found to be significant (ptO.01) proq- nostic factors. Median survival of the whole group was 48 weeks, with a'mini- mum follow-up of 1 year. Thirteen patients (8%) are alive 2 to 4 years. There was no advantage to large radiation fields.

The hyperfractionation and daily regimes had similar efficacy and toxcity. Hyperfractionation with chemotherapy offers a useful alternative approach in the management of this disease.