ALBAN (GETUG-AFU 37): a phase III, randomized, open-label international trial of intravenous atezolizumab and intravesical Bacillus Calmette–Guérin (BCG) versus BCG alone in BCG-naive high-risk, non-muscle-invasive bladder cancer (NMIBC)☆

Abstract

Background: Standard treatment of high-risk non-muscle-invasive bladder cancer (NMIBC) after transurethral resection of bladder tumor is intravesical instillation with Bacillus Calmette–Guérin (BCG) (induction and maintenance regimens); however, BCG therapy still fails in 30%-40% of patients. Prior studies suggest that programmed death-ligand 1 (PD-L1) expression and alterations in immune infiltration might be associated with BCG failure. Patients and methods: The ALBAN trial is an international, 1 : 1 randomized, open-label phase III trial comparing the combination of atezolizumab (an anti-PD-L1 antibody) and BCG (arm B) versus BCG alone (arm A) for BCG-naive patients with high-risk NMIBC. The primary endpoint was investigator-assessed event-free survival (EFS). Key secondary endpoints included high-grade recurrence-free survival, complete response rate, and duration of response in patients with carcinoma in situ disease, as well as overall survival. Results: In total, 255 patients were randomly assigned to arm A and 262 to arm B. The trial did not meet its primary endpoint; there was no significant difference in EFS between arms, with a hazard ratio of 0.98 (95% confidence interval 0.71-1.36, P = 0.9106). EFS results were consistent across all prespecified subgroups. The safety profile of the treatment combination was consistent with that of the individual agents but showed higher rates of treatment-related adverse events (TRAEs) and grade ≥3 TRAEs than BCG alone. Conclusion: This phase III trial in BCG-naive patients with high-risk NMIBC did not demonstrate an EFS benefit from the addition of atezolizumab to 1-year BCG therapy, in contrast to positive EFS results reported with another PD-(L)1 agent, suggesting that any benefit from checkpoint–BCG therapy may be context- and agent-specific rather than a class effect. Future research should focus on biomarker-driven patient selection and optimization of the timing, duration, and route of checkpoint delivery relative to BCG.