Abstract

PURPOSE Radiation therapy for CNS germ cell tumors (GCT) is commonly associated with neurologic sequelae. We designed a therapeutic trial to determine whether irradiation could be avoided. PATIENTS AND METHODS Patients received four cycles of carboplatin, etoposide, and bleomycin. Those with a complete response (CR) received two further cycles; others received two cycles intensified by cyclophosphamide. RESULTS Seventy-one patients were enrolled (45 with germinoma and 26 with nongerminomatous GCT [NGGCT]). Sixty-eight were assessable for response. Thirty-nine of 68 (57%) achieved a CR within four cycles. Of 29 patients with less than a CR, 16 achieved CR with intensified chemotherapy or second surgery. Overall, 55 of 71 (78%) achieved a CR without irradiation. The CR rate was 84% for germinomas and 78% for NGGCT. With a median follow-up duration of 31 months, 28 of 71 patients were alive without relapse or progression. Thirty-five showed tumor recurrence (n = 28) or progression (n = 7) at a median of 13 months. Twenty-six of 28 patients (93%) who recurred following remission underwent successful salvage therapy. Pathology was the only variable predictive of survival. The probability of surviving 2 years was .84 for germinoma patients and .62 for NGGCT. Seven of 71 patients died of toxicity associated with study chemotherapy. CONCLUSION Forty-one percent of surviving patients and 50% of all patients were treated successfully with chemotherapy only without irradiation. Chemotherapy-only regimens for CNS GCT, although encouraging, should continue to be used only in the setting of formal clinical trials.