Neoadjuvant chemotherapy for muscle-invasive bladder cancer at the University of Southern California.

Abstract

301 Background: There is a paucity of data on neoadjuvant GC (gemcitabine, cisplatin) chemotherapy in patients with muscle-invasive bladder cancer (MIBC). Our aim was to compare pathologic and survival outcomes of neoadjuvant GC and M-VAC (methotrexate, vinblastine, doxorubicin, cisplatin) chemotherapy in patients with MIBC. Methods: A retrospective analysis of prospectively collected data from the University of Southern California (USC) Bladder Cancer Database was performed. Between 1985 and 2011, 116 patients received neoadjuvant chemotherapy prior to radical cystectomy and extended pelvic lymph node dissection for clinical stage T2-T4N0M0 bladder cancer. The outcomes were pathologic complete response (pT0N0), pathologic tumor downstaging (pT0N0, pTaN0, pT1N0, or pTisN0), overall survival (OS), and recurrence-free survival (RFS). The Kaplan-Meier method and Cox proportional regression models were used to analyze survival data. Results: The median follow-up duration was 4.5 years (range, 0 to 19.8 years). Fifty-eight patients each received GC and M-VAC chemotherapy. There were no statistically significant differences between the GC and M-VAC groups with regard to pathologic complete response (27.3% versus 17.1%, p=0.419) or pathologic tumor downstaging (45.5% versus 37.1%, p=0.498). The predicted 5-year OS (29% versus 38%, Log rank p=0.634) and RFS (36% versus 35%, Log-rank p=0.891) rates did not differ between the GC and M-VAC groups. However, in a subset of 37 patients with pathologic lymph node positive disease, the predicted 1-year RFS rate differed between the GC and M-VAC groups (0% versus 32%, log rank p=0.019). Multivariable analysis showed a trend toward an independent association between type of neoadjuvant chemotherapy and RFS (GC versus M-VAC: HR 1.64, 95% CI 0.93 to 2.89, p=0.089). Conclusions: Pathologic complete response, pathologic tumor downstaging, and survival did not differ in patients who received neoadjuvant GC and M-VAC chemotherapy. However, GC was associated with poorer RFS in a subset of patients with pathologic lymph node positive disease. Randomized controlled trials comparing neoadjuvant chemotherapy regimens are urgently needed.