A retrospective analysis of efficacy and toxicity of intravesical Bacillus Calmette-Guerin (BCG) and intravesical gemcitabine (GEM) in the treatment of non muscle invasive urothelial bladder cancer (NMIBC).
Data collected from medical records, from local hospitals in Canberra, Australia, between 2010 and 2015, including patients with carcinoma in situ (CIS), Ta, T1 lesion, both with recurrent and new diagnosis. After initial trans-urethral resection of bladder tumour (TURBT) and subsequently confirmed cancer free with second cystoscopy before commencing intravesical therapy. All were treated with weekly induction therapy of either BCG or GEM for 6 weeks and then went on to maintenance treatment depending on their risk profile. Primary end point was disease-free survival (DFS), defined as time from the commencement of induction treatment to last documented cystoscopy with negative biopsy. Secondary analyses included, toxicity and DFS rate. Statistical analysis was performed with SPSS package.
104 patients were included in the study with 53 and 51 receiving BCG and GEM, respectively. Most were male and median age was 77 and 78, respectively. Patients with high grade cancer, CIS and frequent recurrence received maintenence therapy. Median DFS was 20.5 months for BCG, where as median DFS was not reached with GEM, HR of 0.2 (95% CI of 0.08-0.46, p = < 0.005). Relapse rate was better with GEM compared with BCG at 6, 12 and 24 months (3%, 7%, 17% and 24%, 31%, 53%, respectively). 3 patients developed muscle invasive cancer in BCG arm and underwent cystectomy.Toxicity was significantly higher in BCG. More importantly 4 patients developed systemic BCG infection while on BCG, necessitating anti-tuberculous therapy.
Our study shows that intravesical GEM is superior to BCG in the treatment of NMIBC with statistically significant DFS and low relapse rate. Further GEM was better tolerated by patients and certainly much safer than BCG. There for GEM is a reasonable alternative to BCG in cases where BCG is contraindicated or patients who are intolerant to BCG. However, this needs to be further evaluated by randomized controlled phase III trials.
Clinical trial identification
All authors have declared no conflicts of interest.