Comparison of two standard chemotherapy regimens for good-prognosis germ-cell tumours: a randomised trial

Background Most patients with metastatic germ-cell tumours are cured with chemotherapy. However, the optimum chemotherapy regimen is uncertain, and there is variation in international practice. We did a multicentre randomised trial to compare two standard chemotherapy regimens for men with good-prognosis germ-cell tumours. Methods Good prognosis was defined by modified Memorial Sloan-Kettering criteria. The first regimen (regimen A) was based on treatment recommendations from Indiana University and comprised three cycles of 20 mg/m2 cisplatin on days 1–5, 100 mg/m2 etoposide on days 1–5, and 30 kU bleomycin on days 1, 8, and 15, repeated every 21 days. The second regimen (regimen B) was based on the control regimen of a published randomised clinical trial and comprised four cycles of 100 mg/m2 cisplatin on day 1, 120 mg/m2 etoposide on days 1–3, and 30 kU bleomycin on day 1, repeated every 21 days. The primary outcome measure was overall survival. Analysis was by intention to treat. Findings 166 patients were randomised, 83 to each regimen. The trial was stopped when the second planned interim analysis met predefined stopping rules. The median follow-up was 33 months. Overall survival was substantially better with regimen A (three vs 13 deaths, hazard ratio 0·22 [95% CI 0·06–0·77], p=0·008). This difference was due to deaths from cancer (one vs nine), and not deaths from treatment (two vs two) and remained significant after adjustment for other prognostic factors (0·25 [0·07–0·88], p=0·03). Interpretation In men with good-prognosis germ-cell tumours, the regimen developed at Indiana University is superior to the alternative regimen studied in this trial. The lower total dose and dose-intensity of bleomycin and the lower dose-intensity of etoposide in regimen B could be responsible for the worse outcome.