Phase II trial of high-dose carboplatin and etoposide with autologous bone marrow transplantation in first-line therapy for patients with poor-risk germ cell tumors

Background: Between 20% and 30% of patients with advanced germ cell tumors (GCTs) fail to have durable complete response to conventional cisplatin-based induction chemotherapy. However, third-line therapy with high-dose carboplatin and etoposide plus autologous bone marrow transplantation (AuBMT) has induced durable complete response in 10%-20% of patients with cisplatinresistant GCT. Purpose: We conducted a phase II trial of first-line therapy that included high-dose carboplatin and etoposide plus AuBMT in untreated men with advanced GCTs and unfavorable prognosis (i.e., "poor-risk" patients). Methods: Twenty-eight patients were treated with a conventional-dose, cisplatin-containing regimen (VAB-6 [cisplatin, vinblastine, bleomycin, cyclophosphamide, dactinomycin]) with or without high-dose carboplatin (1500 mg/m²) and etoposide (1200 mg/m²) plus AuBMT. Twenty-two of these patients were selected for treatment with two cycles of high-dose carboplatin and etoposide plus AuBMT when reduced clearance of serum tumor markers (alpha-fetoprotein [AFP] or human chorionic gonadotropin [HCG]), as evidenced by prolonged half-life (>7 days for AFP and >3 days for HCG), was observed after two cycles of conventional treatment. Results: Fifteen (56%) of 27 patients considered assessable for response achieved a complete response (12 treated with high-dose chemotherapy plus AuBMT). Sixteen (57%) are alive; 13 (46%) are free of disease at a median follow-up of 31.2 months. For 36 cycles of high-dose chemotherapy, the median duration from bone marrow infusion until a granulocyte count of 0.5/mm³ and a platelet count of 50000/mm³ was 16 days (range, 7-41 days and 8-30 days, respectively). Analysis showed a trend toward improved survival (P =.07) in patients treated with high-dose chemotherapy in this study, compared with 68 poor-risk patients with GCT treated with conventional-dose therapy alone in two earlier studies. Toxicity was not cumulative, and recovery of blood counts after AuBMT was generally rapid. Conclusions: Inclusion of high-dose carboplatincontaining chemotherapy in treatment of poor-risk GCT patients is feasible when serum tumor marker half-life is used to predict resistance to standard cisplatin-based therapy. High-dose therapy in this setting was well tolerated. Implications: Early use of a dose-intensiveregimen may increase survival compared with conventional-dose therapy alone. Further studies with standard induction therapy and intensive high-dose therapy using hematopoietic growth factor support are warranted, followed by a randomized trial comparing this strategy with standard therapy. [J Natl Cancer Inst 85: 1828-1835, 1993]