Abstract
Background: Cisplatin-based combination chemotherapy before surgery is the standard of care for muscle-invasive bladder cancer. However, the optimal chemotherapy modalities have not been precisely defined to date.Patients and methods: In the VESPER trial, patients received after randomization either gemcitabine and cisplatin (GC, 4 cycles) or methotrexate, vinblastine, doxorubicin and cisplatin (dose dense [dd]-MVAC, 6 cycles). Creatinine clearance (CrCl) was calculated before each cycle according to the Cockroft and Gault formula. Definition criteria for local control after neoadjuvant chemotherapy included pathological complete response (ypT0N0), pathological downstaging (<ypT2N0) and organ-confined disease (<ypT3N0) at cystectomy. Baseline and treatment characteristics were studied in multivariate logistic models to determine their potential role for each type of pathological responses.Results: A total of 2128 cycles of chemotherapy were delivered, including 2120 (99.6%) with cisplatin. Full doses of cisplatin were given in 1866 (88%) cycles. Twenty-three (4.7%) patients had to stop chemotherapy (12 GC, 11 dd-MVAC) because of renal failure. No difference in CrCl median values was observed between the 2 regimens during the first 4 cycles. A mild decrease occurred thereafter in patients treated with 2 additional cycles of dd-MVAC. A minimum total dose of 270 mg/m2 for cisplatin was mandatory to optimize pathological complete responses.Conclusion: At least 4 cycles of cisplatin-based chemotherapy should be delivered before cystectomy. Increasing the number of cycles beyond 4 cycles does not lead to a clinically significant deterioration in renal function but without obvious gain on local control.