A RANDOMIZED TRIAL ON DOSE-RESPONSE IN RADIATION-THERAPY OF LOW-GRADECEREBRAL GLIOMA - EUROPEAN ORGANIZATION FOR RESEARCH AND TREATMENT OFCANCER (EORTC) STUDY-22844

Purpose: Cerebral low-grade gliomas (LGG) in adults are mostly compose d of astrocytomas, oligodendrogliomas, and mixed oligoastrocytomas. Th ere is at present no consensus in the policy of treatment of these tum ors. We sought to determine the efficacy of radiotherapy and the prese nce of a dose-response relationship for these tumors in two multicentr ic randomized trials conducted by the European Organization for Resear ch and Treatment of Cancer (EORTC). The dose-response study is the sub ject of this article. Methods and Materials: For the dose-response tri al, 379 adult patients with cerebral LGGs were randomized centrally at the EORTC Data Center to receive irradiation postoperatively (or post biopsy) with either 45 Gy in 5 weeks or 59.4 Gy in 6.6 weeks with qual ity-controlled radiation therapy. All known parameters with possible i nfluences on prognosis were prospectively recorded. Conventional treat ment techniques were recommended. Results: With 343 (91%) eligible and evaluable patients followed up for at least 50 months with a median o f 74 months, there is no significant difference in terms of survival ( 58% for the low-dose arm and 59% for the high-dose arm) or the progres sion free survival (47% and 50%) between the two arms of the trial How ever, this prospective trial has revealed some important facets about the prognostic parameters: The T of the TNM classifications as propose d in the protocol appears to be one of the most important prognostic f actors (p < 0.0001) on multivariate analysis. Other prognostic factors , most of which are known, have now been quantified and confirmed in t his prospective study. Conclusion: The EORTC trial 22844 has not revea led the presence of radiotherapeutic dose-response for patients with L GG for the two dose levels investigated with this conventional setup, but objective prognostic parameters are recognized. The tumor size or T parameter as used in this study appears to be a very important facto r. Copyright (C) 1996 Elsevier Science Inc.