Intravesical Bacillus Calmette-Guérin versus epirubicin for Ta and T1 bladder cancer

Bladder cancer accounts for approximately 4.4% of adult malignancies, and approximately 80% of bladder cancer presents initially as transitional cell carcinoma that is confined to the urothelium (stage Ta) or lamina propria (stage T1). Intravesical administration of Bacillus Calmette-Guérin (BCG) and epirubicin (EPI) has been proven to reduce tumour recurrence and prevent or delay progression to muscle invasion and metastases. However, comparison of the effectiveness and safety of intravesical BCG and EPI in bladder cancer has yet to be explored.To compare the effectiveness and safety of BCG with EPI in the treatment of Ta and T1 bladder cancer.A comprehensive search of MEDLINE (1966 to April 2010), EMBASE (1980 to April 2010), Health Services Technology, Administration, and Research (HealthSTAR), the Cochrane Central Register of Controlled Trials (CENTRAL), CancerLit, and Database of Abstracts of Reviews of Effectiveness (DARE), was performed, and handsearching of relevant journals was undertaken.All randomised or quasi-randomised trials (in which allocation was obtained by alternation - e.g., alternate medical records, date of birth, or other predictable methods) in patients with Ta or T1 bladder cancer that compared intravesical BCG with EPI were included. No language restrictions were applied.Trial eligibility, methodological quality and data extraction were assessed independently by two reviewers. We compared dichotomous outcomes (frequency of tumour recurrence, progressive disease by stage, mortality, distant metastases, local and systemic adverse effects, treatment delayed or stopped due to adverse effects) using risk ratios (RR) with 95% confidence intervals (CI).Five trials of 1111 participants were included in this review. For BCG, 549 patients were treated, and 562 with EPI. Of the evaluated patients, 35.5% (195/549) in the BCG group and 51.4% (289/562) in the EPI group had tumour recurrence (P < 0.05). For disease progression (BCG, 44/549; EPI, 58/562) and distant metastases (BCG, 23/487; EPI, 31/495), there were no significant differences (P = 0.19 and P = 0.29, respectively). Only two trials, including 769 patients, had sufficient data for us to analyze disease-specific (BCG, 22/383; EPI, 26/386) and overall mortality (BCG, 125/383; EPI, 147/386). Neither comparison was significant (P = 0.93 and P = 0.12, respectively). In four studies reporting toxicity, BCG was associated with significantly more drug-induced cystitis [BCG, 54.1% (232/429); EPI, 31.7% (140/441)] and haematuria [BCG, 30.8% (132/429); EPI, 16.1% (71/440)]. Similarly, in three studies reporting systemic toxicity, BCG had significantly higher toxicity than the EPI (34.8% (134/385) versus 1.3% (5/393), respectively). In a meta-analysis comparing 'treatment delayed or stopped' (BCG, 40/431; EPI, 33/441), there was no significant difference between BCG and EPI treatments (P = 0.82).The data from the present meta-analysis indicate that intravesical BCG treatment is more efficacious than EPI in reducing tumour recurrence for Ta and T1 bladder cancer. However, BCG appears to be associated with a higher incidence of adverse effects, such as drug-induced cystitis, haematuria and systemic toxicity, than EPI. The overall quality of the evidence is rather low. Well-designed, high quality randomised controlled trials with good allocation concealment are required.