Evaluation of prognostic factors in chemotherapy of recurrent brain tumors.

We have evaluated the prognostic significance of several easily obtainable and commonly used factors in patients with primary brain tumors (recurrent after irradiation) undergoing chemotherapy. Age, sex, tumor grade, on-study performance score, time from initial diagnosis to initial postirradiation progression, and prior chemotherapy or no were evaluated for affects on tumor response. None of the six factors significantly correlated with response to therapy. The above six factors plus response to chemotherapy were evaluated for their affects on time to progression and survival. Tumor regression following chemotherapy most strongly correlated with prolonged time to progression and survival. A good on-study performance score (not, or only minimally, symptomatic) significantly correlated with prolonged survival, but not with prolonged time to progression. Surprisingly, tumor grade did not significantly correlate with response, time to progression, or survival. Further study is required, but initial data analyses would suggest that essentially all patients with all grades of recurrent tumor could be used in chemotherapeutic trials with proper pretreatment stratification.