Abstract
This study assessed the outcome of patients with locally advanced bladder cancer randomized to treatment with neoadjuvant gemcitabine/cisplatin or radical cystectomy alone. A total of 114 patients were randomized between the 2 arms. Overall survival and disease-free survival benefit could not be obtained from the use of gemcitabine/cisplatin chemotherapy in the neoadjuvant setting in this group of patients, but bladder preservation could be achieved in 11 of 49 patients who received chemotherapy.
Despite aggressive local therapy, patients with locally advanced bladder cancer have a significant risk of distant metastases. This study evaluated the role of neoadjuvant combination chemotherapy with gemcitabine/cisplatin (GC) in improving the outcome of this group of patients over radical cystectomy alone.
A total of 114 patients with newly diagnosed bladder cancer (T3-4, N0-2, M0) were randomized to radical cystectomy alone or initial 3 cycles of GC, then managed according to response. Patients who achieved complete response completed 6 cycles of GC followed by local radiation therapy (RT) only. If tumors were downstaged to T1, complete transurethral resection was done, followed by 3 cycles of GC and then RT. Patients with partial response underwent radical cystectomy followed by 3 cycles of GC. Patients with stable disease or disease progression underwent radical cystectomy.
The overall response rate to GC was 55.1%, and complete response was achieved in 28.6%. The 3-year overall survival (OS) was 51.9% versus 51.2% in the chemotherapy and surgery arms, respectively (P = .399). The 3-year disease-free survival was 31.8% in the chemotherapy arm and 45.1% in the surgery arm (P = .06). Bladder preservation was achieved in 22.5% of patients in the neoadjuvant arm. OS was 78% in responding patients and 100% in patients with complete response.
Neoadjuvant GC did not improve survival in locally advanced bladder cancer over radical cystectomy alone. However, bladder preservation was feasible, and OS in responding patients was impressive. Therefore, predictive models to select patients are needed. This is the largest prospective study of squamous cell carcinoma and transitional cell carcinoma using neoadjuvant GC.