First-line treatments for BCG-naïve non-muscle invasive bladder cancer: a systematic review and meta-analysis.

PURPOSE: To evaluate novel intravesical and systemic combination therapies for improving outcomes in BCG-naïve patients with non-muscle invasive bladder cancer (NMIBC), this systematic review and meta-analysis assessed first-line treatment strategies, including combinations with systemic immune-checkpoint inhibitors (ICIs). METHODS: In this prospectively registered review (CRD420251163026), MEDLINE, Embase, Web of Science, and the ESMO 2025 abstract book were searched for randomized controlled trials (RCTs) evaluating first-line therapies in BCG-naïve NMIBC. Meta-analyses estimated HRs for recurrence-related time-to-event outcomes (disease-free and event-free survival). Grade ≥ 3 treatment-related adverse events were pooled as relative risks (RRs). Risk of bias was assessed using the Cochrane RoB 2 tool. RESULTS: Out of 5202 records screened, six RCTs including 3485 patients were eligible. Modifications to intravesical BCG alone (three trials, n = 895) did not significantly improve recurrence-related outcomes. Systemic ICIs given additional to intravesical BCG significantly reduced recurrence-related events compared to BCG alone (HR 0.77, 95%CI 0.6–0.97, n = 1899; number needed to treat at two years: 25), but increased grade ≥ 3 treatment-related adverse events by a statistically and clinically significant margin (RR 3.97, 95% CI 2.53–6.22, n = 1879; number needed to treat to harm one patient: 5). Sensitivity analysis using the ALBAN alternative endpoint of high-grade recurrence-free survival yielded HR 0.78 (95% CI 0.59–1.02), not reaching statistical significance. Limitations included heterogeneity in trial design and endpoint definitions. CONCLUSION: In BCG-naïve NMIBC, systemic ICIs combined with intravesical BCG improve recurrence-related outcomes but are offset by a substantially increased risk of severe treatment-related adverse events. These findings highlight the need for careful, risk-based, biomarker-guided patient selection to balance over- and undertreatment within a shared decision-making process.