Introduction/Background In KEYNOTE-826 (NCT03635567), pembrolizumab (pembro) + chemotherapy (chemo) ± bevacizumab (bev) provided statistically significant and clinically meaningful overall survival (OS) and progression-free survival (PFS) improvements in patients with persistent, recurrent, or metastatic cervical cancer. In the present exploratory analysis, we examined treatment outcomes in patient subgroups defined by bev use.

Methodology Eligible adult patients had persistent, recurrent, or metastatic squamous cell carcinoma, adenosquamous carcinoma, or adenocarcinoma of the cervix not previously treated with chemo and not amenable to curative treatment; measurable disease per RECIST v1.1; ECOG PS 0–1; and provided a tumor sample to determine PD-L1 status. Patients were randomized 1:1 to pembro 200 mg Q3W or placebo for up to 35 cycles + chemo (paclitaxel 175 mg/m² + cisplatin 50 mg/m² or carboplatin AUC 5) ± bev 15 mg/kg. Dual primary endpoints are OS and PFS by investigator assessment per RECIST v1.1 in the PD-L1 CPS ≥1, all comer, and CPS ≥10 populations. Treatment outcomes were assessed in patient subgroups defined by bev use (yes or no). Hazard ratios and 95% CIs were based on a stratified Cox regression model.

Results 617 patients were randomized (pembro + chemo, n = 308 [63.6% with bev]; placebo + chemo, n = 309 [62.5% with bev]). The most common reason for bev exclusion was medical contraindication (75.9%). Pembro + chemo prolonged OS and PFS vs placebo + chemo in patient subgroups defined by bev use in the CPS ≥1 and all-comer populations (table 1). In the pembro and placebo arms, respectively, the incidences of treatment-related grade ≥3 AEs were 74.0% vs 60.4% in the with bev subgroup, and 66.8% vs 62.1% in the without bev subgroup.

Conclusion Pembro + chemo prolonged OS and PFS vs placebo + chemo regardless of bev use and had a manageable safety profile.