Background: GOG-0218 established the addition of BEV 15 mg/kg every 3 weeks (q3w) for 15 months to standard front-line chemotherapy for advanced ovarian cancer, but the optimal BEV duration remained unknown. We report primary results from a randomized phase 3 trial designed to address this question. Methods: Eligible pts with FIGO stage IIB–IV EOC, FTC, or PPC and ECOG PS ≤2 underwent primary cytoreductive surgery followed by 6 cycles of chemotherapy (paclitaxel 175 mg/m² + carboplatin AUC 5 q3w) and BEV (15 mg/kg q3w). Pts were randomized to receive BEV for either 15 months (standard arm BEV15) or 30 months (experimental arm BEV30), stratified by FIGO stage/residual tumor (stage IIB–IIIC/no residual tumor vs stage IIB–IIIC/residual tumor or stage IV). The primary endpoint was investigator-assessed progression-free survival (PFS) according to RECIST v1.1. Secondary endpoints were overall survival (OS), objective response rate, quality of life, safety, and tolerability. The trial was designed with 80.2% power to detect a hazard ratio (HR) of 0.66 favoring BEV30 (2-sided log-rank test, 5% significance level, 10% dropout rate) after 697 PFS events. The trial was funded by F. Hoffmann-La Roche Ltd., performed according to ENGOT model A. Results: From Nov 2011 to Aug 2013, 927 women (83% with EOC) from 161 centers were randomized. Baseline characteristics were balanced between arms; median age was 61 years, 96% had ECOG PS 0/1, 58% had no residual tumor, and 77% had high-grade serous histology. Serious adverse events of special interest for BEV occurred in 51/448 pts (11%) vs 61/442 pts (14%) receiving BEV15 vs BEV30, respectively: hypertension (2.7%/4.5%), thromboembolic event (2.2%/3.2%), fistula (3.1%/1.1%), gastrointestinal perforation (0.2%/0.9%), proteinuria (0.7%/1.4%), hemorrhage (0.2%/0.9%), and myocardial infarction (0%/1.1%). Efficacy is shown in the table. Conclusions: Longer treatment with BEV for up to 30 months improves neither PFS nor OS in pts with primary EOC, FTC, or PPC. Therefore BEV treatment duration of 15 months remains standard of care. Clinical trial information: NCT01462890

*Performed because of evidence of nonproportional hazards, restricted at the time point of the last observed event (BEV15/BEV30).