Background
Phase 3, open-label, randomized (1:1:1). N=695 patients with muscle-invasive bladder cancer ineligible for or declining cisplatin-based chemotherapy. Dual primary endpoints: EFS in each durvalumab group vs control (the two durvalumab arms were not compared with each other).
Results
Interventions and follow up: Arm A: neoadjuvant durvalumab + enfortumab vedotin (EV) q3w x3, radical cystectomy, adjuvant durvalumab q4w x9 (n=233)
Arm B: neoadjuvant durvalumab + tremelimumab (2 cycles) + EV, cystectomy, adjuvant durvalumab + tremelimumab (1 cycle) (n=233)
Arm C: radical cystectomy alone (n=229)
Primary endpoints: EFS (Arm A vs C; Arm B vs C)
Key secondary: OS
mFollow up: NR
Results: EFS (24mo), Arm A vs C: 63.5% (95% CI 56.5–69.7) vs 45.1% (38.2–51.8), HR 0.51 (95% CI 0.38–0.67), P<.001
EFS (24mo), Arm B vs C: 60.1% (53.2–66.2), HR 0.61 (95% CI 0.47–0.80), P<.001
OS (24mo), Arm A vs C: 78.4% vs 69.4%, HR 0.58 (95% CI 0.40–0.83), P=.003
OS, Arm B vs C: NR
pCR: NR in abstract
Arm B: neoadjuvant durvalumab + tremelimumab (2 cycles) + EV, cystectomy, adjuvant durvalumab + tremelimumab (1 cycle) (n=233)
Arm C: radical cystectomy alone (n=229)
Primary endpoints: EFS (Arm A vs C; Arm B vs C)
Key secondary: OS
mFollow up: NR
Results: EFS (24mo), Arm A vs C: 63.5% (95% CI 56.5–69.7) vs 45.1% (38.2–51.8), HR 0.51 (95% CI 0.38–0.67), P<.001
EFS (24mo), Arm B vs C: 60.1% (53.2–66.2), HR 0.61 (95% CI 0.47–0.80), P<.001
OS (24mo), Arm A vs C: 78.4% vs 69.4%, HR 0.58 (95% CI 0.40–0.83), P=.003
OS, Arm B vs C: NR
pCR: NR in abstract
Adverse events
Grade ≥3 (any): 71.1% (Arm A) vs 79.5% (Arm B) vs 50.2% (control)
Treatment-related grade ≥3: 30.7% (Arm A) vs 44.5% (Arm B)
Specific toxicities / discontinuation: NR in abstract
Treatment-related grade ≥3: 30.7% (Arm A) vs 44.5% (Arm B)
Specific toxicities / discontinuation: NR in abstract
Conclusions
Perioperative durvalumab-EV significantly improved EFS and OS vs cystectomy alone in cisplatin-ineligible or cisplatin-declining MIBC; adding tremelimumab improved EFS but with more toxicity. Establishes an EV-based perioperative option for this population.
Key Limitations
Open-label; control is surgery alone (no perioperative therapy), which is not an active comparator such as adjuvant nivolumab or EV-pembrolizumab. Arms A and B not compared; tremelimumab adds toxicity without a compared benefit. OS in the tremelimumab arm NR in abstract. Follow-up short (24-mo estimates). More than half of durvalumab-treated patients had grade ≥3 events.
Clinical Context
Context: cisplatin-ineligible MIBC has relied on upfront cystectomy ± adjuvant nivolumab. VOLGA is the first phase 3 of perioperative PD-L1 + EV in this group; EV-pembrolizumab (KEYNOTE-905/EV-303) has also shown benefit in cisplatin-ineligible disease. Regulatory status of durvalumab-EV in this setting is not claimed here.
