The effect of cystectomy, and perioperative methotrexate, vinblastine, doxorubicin and cisplatin chemotherapy on the risk and pattern of relapse in patients with muscle invasive bladder cancer

Purpose: Trials have demonstrated decreased relapse with perioperative meth otrexate, vinblastine, doxorubicin and cisplatin (M-VAC) chemotherapy in pa tients with muscle invasive bladder cancer. We evaluated whether the benefi t of chemotherapy correlates with its effects on distant or pelvic relapse. Materials and Methods: We retrospectively evaluated the records of all 107 patients who underwent cystectomy for muscle invasive bladder cancer at our institution between 1988 and 1994. Factors predicting relapse were identif ied and used to group patients at high or low risk. The outcome in each gro up with and without M-VAC chemotherapy was then analyzed in terms of overal l, metastatic and pelvic relapse. Univariate analysis was performed using t he Kaplan-Meier method and log rank statistic, and multivariate analysis wa s done using the Cox proportional hazards model. Median survival was 29 mon ths for patients free of disease. Results: Pathological stage T3 or greater according to the American Joint C ommittee on Cancer, tumor greater than 3 cm. and creatinine greater than 1. 5-fold normal were independent poor prognostic factors in patients treated with cystectomy only. Patients with any of these factors or metastatic invo lvement of the pelvic lymph nodes were considered at high risk. All 35 low risk patients were treated with cystectomy only and had an excellent outcom e with a 3-year relapse-free survival plus or minus standard error of 93% /- 5%. The 3-year rates in 52 and 20 high risk patients treated without and with chemotherapy, respectively, were 42% +/- 8% versus 57% +/- 13% for re lapse-free survival (p = 0.17), 38% +/- 9% versus 8% +/- 8% for pelvic fail ure (p = 0.02) and 39% +/- 9% versus 38% +/- 13% for distant metastases (no t significant). Multivariate analysis of patients who underwent pelvic lymp hadenectomy revealed that perioperative chemotherapy improved relapse-free survival and pelvic control but not metastatic control (p = 0.03, 0.02 and 0.31, respectively). Conclusions: Low risk patients have excellent disease control when treated with cystectomy only. Those with high risk features are at substantial risk for pelvic failure (38% at 3 years) after cystectomy only. Perioperative M -VAC chemotherapy has a profound impact on pelvic but not on metastatic fai lure.