Purpose: To examine the influence of various prognostic factors on local control of medulloblastoma.
Materials and methods: Sixty-five patients who had been treated between 1980 and 1990 at our six hospitals were retrospectively studied. Factors included in the Cox's multivariate analysis were sex, age (log [y.o. + 1]), performance status, pretreatment T-stage (judged from the findings of CT and/or MRI; T 1: 3 patients, T2: 17, T3 [with hydrocephalus]: 44, T4: 1), extent of surgical resection (total: 30 patients, less than total: 35), total dose (21-87 Gy, median: 55) and overall treatment time (19-163 days, median: 54) of local irradiation, and use of adjuvant chemotherapy or immunotherapy.
Results: Recurrence occurred in one T2 patient and in 17 T3 patients. The multivariate analysis showed that local tumor control decreased with advance in T-stage (
p=0.04) and with prolonged overall treatment times (p=0.003), and that it increased with higher total doses (
p=0.01). When analysis was limited to T3 patients, usefulness of the chemotherapy was also suggested (
p=0.03). However, the influence of the extent of resection on local tumor control was not statistically significant, because resectability depended on T-stage.
Conclusion: This analysis showed that local control of medulloblastoma was influenced by pretreatment T-stage, total dose and overall irradiation treatment time, and probably by the adjuvant chemotherapy used. Some of the observed losses by prolongations in radiotherapy may reflect proliferation of tumor cells during radiotherapy.
View full abstract