Platinum-based chemotherapy plus bevacizumab and paclitaxel is the standard first-line therapy for persistent, recurrent, or metastatic cervical cancer. The phase 2 KEYNOTE-158 trial found evidence of a response to pembrolizumab among patients with programmed death ligand-1 (PD-L1) positive cervical cancer.This double-blind phase 3 trial (KEYNOTE-826) evaluated the efficacy and safety of pembrolizumab or placebo in addition to platinum-based chemotherapy with or without bevacizumab in patients with persistent, recurrent, or metastatic cervical cancer. Patients older than 18 years who had not been treated with systemic chemotherapy and had not received radiotherapy within the 2 weeks before randomization were eligible. PD-L1 expression was measured according to combined positive score, defined as the number of PD-L1-staining cells in a tumor sample divided by the total number of viable tumor cells, multiplied by 100. Patients were enrolled at 151 sites in 19 countries and randomized in a 1:1 ratio to pembrolizumab (200 mg) or placebo every 3 weeks for up to 35 cycles. All patients received paclitaxel and the investigators' choice of cisplatin or carboplatin every 3 weeks, whereas bevacizumab was given according to local practice at the investigator's discretion. The primary outcomes were overall survival and progression-free survival. Superiority of pembrolizumab to placebo was tested in all patients with PD-L1 combined positive score of 1 or more, combined positive score of 10 or more, and in the intention-to-treat population.A total of 617 patients were randomized between November 2018 and January 2020, with 308 in the pembrolizumab group and 309 in the placebo group. A total of 548 patients had a PD-L1 combined positive score of 1 or more (273 in pembrolizumab group, 275 in placebo group) and 317 patients with a score of 10 or more (158 in pembrolizumab group, 159 in placebo group). The median follow-up time at first interim analysis was 22.0 months (15.1 to 29.4). The median treatment duration was 10.0 months in the pembrolizumab group and 7.7 months in the placebo group. Progression-free survival was significantly prolonged in the pembrolizumab group compared with the placebo group in patients with a PD-L1 score of 1 or more (median, 10.4 months [95% confidence interval (CI), 9.7-12.3] vs 8.2 months [95% CI, 6.3-8.5 months]), in patients with a PD-L1 score of 10 or more (median, 10.4 months [95% CI, 8.9-15.1] vs 8.1 months [95% CI, 6.2-8.8]), and in the intention-to-treat population (median, 10.4 months [95% CI, 9.1-12.1] vs 8.2 months [95% CI,). Overall survival was likewise significantly longer in the pembrolizumab group compared with the placebo group among patients with a PD-L1 score of 1 or more (24-month estimate of patients alive, 53.0%; 95% CI, 34.9-48.2), a score of 10 or more (24-month estimate, 54.4%; 95% CI, 45.5-62.4), and in the intention-to-treat population (24-month estimate, 50.4%; 95% CI, 43.8-56.6). Serious adverse events occurred in 49.8% of the patients in the pembrolizumab group and 42.4...