Treatment options for patients with platinum-resistant ovarian cancer (PROC) are limited, and new therapeutic strategies are urgently needed. This phase II, randomized, multicentre trial evaluated the safety and activity of the anti-PD-L1 antibody atezolizumab (atezo) combined with the VEGF inhibitor bevacizumab (bev) and the irreversible cyclooxygenase 1/2 inhibitor aspirin [acetylsalicylic acid (ASA)] in PROC.
Patients were randomized to bev monotherapy 15 mg/kg (arm 1), atezo 1,200 mg plus placebo (pbo; arm 2), atezo 1,200 mg plus ASA 320 mg/daily (arm 3), bev 15 mg/kg plus atezo 1,200 mg plus pbo (arm 4), or bev 15 mg/kg plus atezo 1,200 mg plus ASA 320 mg/daily (arm 5). Primary endpoint was progression-free survival at 6 months (PFS-6) according to RECIST v1.1 assessed by a local investigator. Secondary objectives included overall survival, PFS, second PFS (PFS2), and tolerability. Time to first subsequent therapy (TFST) was evaluated in a post hoc analysis.
In arms 1 (bev), 4 (bev-atezo-pbo), and 5 (bev-atezo-ASA), there were 7/32 [21.9%, 70% confidence interval (CI), 14.0-32.0], 8/32 (25.0%, 70% CI, 16.6-35.3), and 8/32 (25.0%, 70% CI, 16.6-35.3) patients alive and progression-free at 6 months. The primary objective was not reached in any arm. Median PFS and response rates were 2.3 for bev monotherapy, 4.1 for bev-atezo-pbo, and 4.0 months for bev-atezo-ASA and 10%, 19%, and 15%, respectively. Two patients achieved an ongoing complete response lasting for more than 5 years from randomization (1 in bev-atezo-pbo and 1 in bev-atezo-ASA). A post hoc analysis of TFST suggested benefit of adding bev to atezo-ASA (P < 0.001). Tumor-infiltrating lymphocytes (TIL) increased in the atezo-containing arms after the first two cycles, and increased TIL were associated with a significantly longer TFST.
The addition of ASA to bev plus atezo was well-tolerated but did not improve efficacy in PROC. Relative to bev alone, the bev plus atezo combination numerically improved PFS. Exploratory translational analyses suggest clinical benefit in a subgroup of patients characterized by high TIL infiltration and PD-L1-positive tumors at baseline.