33O PRIMA/ENGOT-OV26/GOG-3012 study: Long-term conditional PFS

Abstract

Niraparib (nir) showed a blinded independent central review–assessed PFS benefit as a first-line (1L) maintenance therapy (MT) in the primary analysis of PRIMA (data cut 17 May 2019) across biomarker subgroups, including a substantial benefit in patients (pts) with homologous recombination–deficient (HRd) tumours. These results were the basis for approval of nir as MT after response to 1L platinum-based chemotherapy (CT). Here we report investigator-assessed (IA) cPFS (the probability of remaining alive and progression free beyond a specified landmark) in PRIMA. This double-blind, placebo (PBO)-controlled phase 3 trial evaluated nir in pts with newly diagnosed, advanced high-grade serous or endometrioid ovarian, primary peritoneal, or fallopian tube cancer (OC) at high risk for relapse after a complete or partial response (CR or PR) to 1L platinum-based CT. Stratification factors were best response to 1L CT regimen (CR/PR), receipt of neoadjuvant CT (yes/no), and homologous recombination deficiency status (HRd or HRp/HRnd) per the Myriad myChoice® CDx PLUS assay. Pts received nir or PBO once daily (2:1 ratio). IA cPFS was analysed for the HRd and intention-to-treat (ITT) populations, using the 17 Nov 2021 data cut. The median follow-up time was 3.5 y. The estimated PFS rate at 4 y was 38% for nir-treated pts and 17% for PBO-treated pts in the HRd population and 24% (nir) vs 14% (PBO) in the ITT population. The 2-y cPFS probabilities beyond the 1- and 2-y landmarks were higher in the nir arm than in the PBO arm (HRd: 1 y: 62% vs 50%, 2 y: 74% vs 60%; ITT: 1 y: 54% vs 46%, 2 y: 67% vs 64%). Safety was previously reported (González-Martín, et al. Ann Oncol. 2022;33[suppl 7]:S789). A durable PFS benefit (nir vs PBO) was observed up to 4 y after randomisation in the ITT and HRd populations, as determined by IA. Pts free from disease progression or death at the 2-y landmark had a high probability of remaining free from progression or death at 4 y, supporting the use of nir as a 1L MT.Table: 33OHRdNirPBOLandmark time from randomizationEvents/total pts2-y probability from landmark, % (95% CI)Events/total pts2-y probability from landmark, % (95% CI)0137/24751 (44–57)98/12629 (21–37)1 y69/15962 (54–70)33/5750 (36–62)2 y26/11074 (64–82)11/3460aITTLandmark time from randomizationNirPBOEvents/total pts2-y probability from landmark, % (95% CI)Events/total pts2-y probability from landmark, % (95% CI)0332/48736 (31–40)199/24622 (17–28)1 y124/24454 (47–60)54/9246 (36–56)2 y42/15267 (57–76)15/5164aa95% CI were not calculated at time points with <10 pts. Open table in a new tab