Efficacy and safety of combination versus single-agent immunotherapy for BCG-unresponsive non-muscle-invasive bladder cancer.

Bacillus Calmette-Guérin (BCG) is the standard adjuvant therapy for high-risk non-muscle-invasive bladder cancer (NMIBC); however, a substantial proportion of patients develop BCG-unresponsive disease with limited bladder-preserving options. The objective of this study was to determine whether combination immunotherapy provides superior clinical efficacy and safety compared with single-agent immunotherapy for patients with BCG-unresponsive NMIBC. We conducted a systematic analysis of studies retrieved from PubMed, the Cochrane Library, Web of Science, Embase, Google Scholar, and MEDLINE searched up to December 2025. Eligible studies included single-arm trials and randomized controlled trials (RCTs) enrolling patients with pathologically confirmed BCG-unresponsive NMIBC treated with single-agent or combination immunotherapy. The intervention featured eight core drugs, including immune checkpoint inhibitors (PD-1 inhibitors), adenovirus vectors, and IL-15 superagonist Nogapendekin alfa inbakicept (N-803), administered primarily every 3 weeks. Primary outcomes were complete response rate (CRR) and progression-free survival (PFS). Secondary outcomes included high-grade recurrence, bladder preservation rate (BPR), and adverse events (AEs). Pooled analysis utilized fixed or random-effects models based on I2 heterogeneity. Ten studies (7 single-arm, 3 RCTs) involving 768 patients were included. For combination versus single-agent immunotherapy, the 3-month CRRs were 68% (95% CI: 63-75%) and 47% (95% CI: 43-51%), respectively. At 12 months, CRRs were 50% (95% CI: 43-59%) for combination and 28% (95% CI: 23-34%) for monotherapy. The 12-month PFS rate was significantly higher with combination therapy (90%; 95% CI: 85-94%) than with monotherapy (66%; 95% CI: 60-71%). Combination therapy achieved a BPR of 92.5% (95% CI: 87-98%). Regarding safety, any AEs occurred in 29.6% (95% CI: 14-46%) of the combination group compared with 87.4% (95% CI: 85-90%) in the monotherapy group. Limitations include the use of single-arm trial data, which lacks blinded evaluation, and moderate-to-severe heterogeneity in efficacy outcomes. Combination immunotherapy is associated with improved short and intermediate-term oncologic outcomes compared with single-agent immunotherapy in BCG-unresponsive NMIBC, with acceptable safety profiles. These findings support further investigation of combination strategies as bladder-preserving treatments; however, well-designed randomized trials are required before routine clinical adoption. https://www.crd.york.ac.uk/prospero/, identifier CRD420251269272.