Andrew Rh Shepherd, Emily Shepherd, Nicholas R Brook

Despite local therapies, commonly transurethral resection (TUR) followed by adjuvant treatments, non-muscle-invasive bladder cancer (NMIBC) has a high rate of recurrence and progression. Intravesical Bacillus Calmette-Guérin (BCG) has been shown to reduce recurrence and progression in people with NMIBC following TUR, however many people do not respond to treatment, have recurrence shortly after, or cannot tolerate standard-dose therapy. The potential for synergistic antitumour activity of interferon (IFN)-alpha (α) and BCG provides some rationale for combination therapy for people who do not tolerate or respond to standard-dose BCG therapy. To assess the effects of intravesically administered BCG plus IFN-α compared with BCG alone for treating non-muscle-invasive bladder cancer. We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (Issue 8, 2016), MEDLINE (OvidSP) (1946 to 2016), Embase (OvidSP) (1974 to 2016), ClinicalTrials.gov, the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) as well as reference lists of retrieved articles and handsearched abstract proceedings of relevant conferences for the past three years. We applied no language restrictions. The date of last search of all databases was 25 August 2016. We included randomised controlled trials (RCTs) and pseudo-randomised trials assessing intravesically administered BCG plus IFN-α versus BCG alone in adults of either gender with histologically confirmed Ta and T1 superficial bladder cancer, with or without carcinoma in situ, treated with TUR. Two review authors independently assessed study eligibility, extracted data, and assessed the risk of bias of included studies. We used Review Manager 5 for data synthesis and employed the random-effects model for meta-analyses. For prespecified outcomes, where we were unable to derive time-to-event information (e.g. time-to-recurrence), we assessed dichotomous outcomes (e.g. recurrence) instead. We assessed the quality of the evidence for the main comparisons using the GRADE approach. We included five RCTs involving a total of 1231 participants with NMIBC in this review. Due to poor reporting, the risk of bias in the included studies was often unclear. We assessed the studies under two main comparisons: intravesical BCG plus IFN-α versus intravesical BCG alone (four RCTs), and intravesical BCG alternating with IFN-α versus intravesical BCG alone (one RCT). Intravesical BCG plus IFN-α versus intravesical BCG alone (four RCTs): We observed no clear difference between BCG plus IFN-α and BCG alone for recurrence (average risk ratio (RR) 0.76, 95% confidence interval (CI) 0.44 to 1.32; 4 RCTs; 925 participants; very low-quality evidence) or progression (average RR 0.26, 95% CI 0.04 to 1.87; 2 RCTs; 219 participants; low-quality evidence). The included RCTs did not report on the other primary outcome of this review, discontinuation of therapy due to adverse events. Regarding secondary outcomes, we observed no clear difference for disease-specific mortality (RR 0.38, 95% CI 0.05 to 3.05; 1 RCT; 99 participants; very low-quality evidence). Two RCTs reporting contradictory findings for adverse events could not be pooled due to variation in definitions. There were no data from the included RCTs on time-to-death or disease-specific quality of life. Intravesical BCG alternating with IFN-α versus intravesical BCG alone (one RCT): We observed shorter time-to-recurrence for participants in the BCG alternating with IFN-α group compared with the BCG alone group (hazard ratio (HR) 2.86, 95% CI 1.98 to 4.13; 1 RCT; 205 participants; low-quality evidence), but no clear differences in time-to-progression (HR 2.39, 95% CI 0.92 to 6.21; 1 RCT; 205 participants; low-quality evidence) and discontinuation of therapy due to adverse events (RR 2.97, 95% CI 0.31 to 28.09; 1 RCT; 205 participants; low-quality evidence). Regarding secondary outcomes, there were no clear differences between the BCG alternating with IFN-α and BCG alone groups for disease-specific mortality (HR 2.74, 95% CI 0.73 to 10.28; 1 RCT; 205 participants; low-quality evidence), time-to-death (overall survival) (HR 1.00, 95% CI 0.68 to 1.47; 1 RCT; 205 participants; low-quality evidence), or systemic or local adverse events (RR 1.65, 95% CI 0.41 to 6.73; 1 RCT; 205 participants; low-quality evidence). There were no data on disease-specific quality of life. We found low- to very low-quality evidence suggesting no clear differences in recurrence or progression with BCG plus IFN-α compared with BCG alone for people with NMIBC; there was no information to determine the effect on discontinuation of therapy due to adverse events. Low-quality evidence suggests BCG alternating with IFN-α compared with BCG alone may increase time-to-recurrence, however low-quality evidence also suggests no clear differences for time-to-progression or discontinuation of therapy due to adverse events.Additional high-quality, adequately powered trials using standardised instillation regimens and doses of both BCG and IFN-α, reporting outcomes in subgroups stratified by patient and tumour characteristics, and on long-term outcomes related not only to recurrence but also to progression, discontinuation due to adverse events, and mortality may help to clarify the ideal treatment strategy and provide a more definitive result.