Article Text
Abstract
Introduction In this multicentre randomized, phase II/III trial, we sought to examine if the 1) combination of anti-programmed death ligand 1 (PD-L1) monoclonal antibody atezolizumab (ATEZO) with pegylated-liposomal- doxorubicin (PLD) [Arm 1] and/or 2) addition of ATEZO to PLD and bevacizumab (BEV) [Arm 2] result in an improvement in survival for patients with platinum resistant ovarian cancer (PROC) compared to the standard PLD/BEV [Arm 3].
Methods Patients were randomly assigned 1:1:1 to PLD/ATEZO, PLD/BEV/ATEZO or PLD/BEV (IV PLD 40 mg/m2 q4weeks; BEV 10 mg/kg q2weeks; ATEZO 800 mg q2weeks). Key eligibility: 1–2 prior lines of therapy (no PLD), ECOG 0–2, and RECIST measurable/evaluable PROC. No stratification by PD-L1 status. The phase II primary endpoint was PFS. The phase III coprimary endpoints were PFS/OS.
Results From 05/2017–10/2021 444 patients with PROC were enrolled. The median age was 63 yrs (35–86). All had received prior chemotherapy; 434 (97.7%) prior surgery and 9 (2%) prior biological therapy. At the phase III interim analysis Arm 1 (PLD/ATEZO) was discontinued for futility. The phase III OS/PFS analysis included accruals to Arms 2 (PLD/BEV/ATEZO) and 3 (PLD/BEV) from all phases. With median follow-up of 47 months, median PFS was 7.4 months and 5.6 months (HR 0.79 with 99.99% 1-sided CI 0.0–1.21), and median OS was 14.9 months and 12.3 months (HR 0.80; 98.78% 1-sided CI 0.00–1.06; 1-sided p=0.038) for Arms 2 and 3, respectively. Adverse events were as expected.
Conclusion/Implications The addition of ATEZO to PLD/BEV did not result in a statistically significant longer OS than PLD/BEV in PROC. Subset analysis are planned to evaluate survival outcomes with high PD-L1 expression (NRG-GY009/NCT02839707).