Atezolizumab (atezo) with or without Bacille Calmette-Guérin (BCG) in patients (pts) with high-risk nonmuscle-invasive bladder cancer (NMIBC): Results from a phase Ib/II study

493 Background: Standard treatment (tx) for high-risk NMIBC is transurethral resection of bladder tumor (TURBT) followed by BCG induction and maintenance. However, ≈50% of pts experience recurrence and/or progression after tx and may be ineligible for or refuse cystectomy. The PD-L1/PD-1 pathway may be involved with immune escape in NMIBC following BCG exposure. Here, we report results of atezo (anti–PD-L1) ± BCG in BCG-unresponsive, high-risk NMIBC. Methods: This multicenter study (NCT02792192) enrolled pts with BCG-unresponsive NMIBC with carcinoma in situ who had repeat TURBT. Cohort 1A and 1B pts received atezo 1200 mg IV q3w for ≤96 wk. Cohort 1B pts also received standard BCG induction (qw × 6 doses) and maintenance (qw × 3 doses at 3 mo), with optional maintenance courses at 6, 12, 18, 24, and 30 mo. For cohort 1B only, de-escalation was allowed for ≤3 BCG dose levels (full dose 50 mg, 66% and 33% of full dose). Co-primary outcomes were safety and complete response (CR) rate at 6 mo (6-mo bladder biopsy required). Duration of CR and 3-mo CR rate (key secondary outcomes) and 12-mo CR rate (exploratory) were also shown. Results: Cohorts 1A and 1B enrolled 12 pts each. Median age was 74 y; most pts had ECOG PS 0 (n = 7 [58%] in each cohort). At data cutoff (Sep 29, 2020), median atezo tx duration was 22.7 wk in cohort 1A and 31.6 wk in 1B. Following dose de-escalation in cohort 1B, the recommended BCG dose was 50 mg. BCG dose modification/interruption occurred in 4 pts (33%) due to an AE. The most common reason for tx discontinuation was disease recurrence or progression in both cohorts. Three pts (25%) in cohort 1A had atezo-related Gr 3 AEs (most common: maculopapular rash, n = 2); no atezo- or BCG-related Gr ≥3 AEs were seen in cohort 1B. Three dose-limiting toxicities occurred (1 [8%] in cohort 1A and 2 [17%] in cohort 1B), all reported as AEs of special interest. No Gr 4/5 AEs were reported. CRs, which appeared durable, were seen in both cohorts (Table). Conclusions: In this first report of atezo + BCG in NMIBC, atezo as mono- and combination therapy was well tolerated, with no new safety signals or tx-related deaths. Preliminary data suggested clinically meaningful activity, especially with atezo + BCG, requiring confirmation in a larger setting. Clinical trial information: NCT02792192. [Table: see text]