Perioperative enfortumab vedotin-pembrolizumab, gemcitabine-cisplatin-durvalumab, and dose-dense MVAC in cisplatin-eligible muscle-invasive bladder cancer: A benefit-risk Bayesian network meta-analysis.

To compare the efficacy and safety of three perioperative regimens for cisplatin-eligible muscle-invasive bladder cancer (MIBC); enfortumab vedotin plus pembrolizumab (EV + P), gemcitabine-cisplatin plus durvalumab (GC + Durva), and dose-dense MVAC (dd-MVAC), in the absence of head-to-head trials, using a Bayesian network meta-analysis (NMA). A systematic search through March 2026 identified phase III trials comparing perioperative regimens against GC in cisplatin-eligible MIBC. The primary analysis was a Bayesian NMA of event-free survival (EFS), overall survival (OS), and pathological complete response (pCR), with SUCRA rankings; risk of bias used RoB-2. Three trials (EV-304, NIAGARA, VESPER; 2,364 patients) formed a star-shaped network vs. GC, all regimens improved EFS-EV + P (Hazard ratio [HR] 0.53, 95% confidence interval [CI] 0.41-0.70), GC + Durva (0.68, 0.56-0.82), dd-MVAC (0.74, 0.55-0.99), and OS, EV + P (0.65, 0.48-0.89), GC + Durva (0.75, 0.60-0.93), dd-MVAC (0.71, 0.52-0.97). EV + P ranked first across all endpoints (SUCRA: EFS 96%, OS 81%, pCR > 99%). Indirect comparisons between experimental regimens were non-significant for survival; only pCR favored EV + P over GC + Durva (odds ratio [OR] 1.7, P < 0.01) and dd-MVAC (OR 2.0, P < 0.01). GC + Durva showed no excess Grade ≥3 toxicity over GC (41% vs. 41%). This NMA provides the first quantitative benefit-risk framework integrating efficacy ranking, toxicity, and subgroup heterogeneity for these regimens. EV + P had the highest probability of improving EFS, OS, and pCR, with GC + Durva ranking second; although directionally consistent with each trial, the magnitude, ranking, and subgroup heterogeneity could not be quantified without indirect synthesis. GC + Durva offered the most favorable benefit-risk profile. Prospective head-to-head trials remain an unmet need.