Neoadjuvant Pembrolizumab in Locally Advanced Gastric Cancer

Immune checkpoint inhibitors improve response, progression-free survival, and overall survival when added to first-line chemotherapy in patients with esophagogastric adenocarcinoma. Investigators now report results of KEYNOTE-585, an industry-sponsored, international, phase 3 trial evaluating the addition of pembrolizumab to pre- and postoperative chemotherapy in patients with clinical stage T3–4 or node-positive resectable gastric cancer.

A main cohort of 804 patients were randomized to receive either pembrolizumab (200 mg) or placebo combined with chemotherapy — investigator's choice of capecitabine plus cisplatin or a 5-day infusion of fluorouracil (FU) plus cisplatin — for 3 cycles before surgery and 3 cycles after, followed by 11 cycles of placebo or pembrolizumab. Another cohort of 203 patients were randomized to receive pembrolizumab or placebo combined with an alternative chemotherapy regimen of 5-FU, oxaliplatin, and docetaxel (FLOT) for 4 cycles pre- and 4 cycles postsurgery. In the main cohort, 47% of patients were treated in Asia, 79% had gastric primary tumors, 75% had PD-L1 combined positive scores ≥1%, 9% had microsatellite instability (MSI) high status, and 76% had clinical stage III disease.

The primary endpoint of pathologic complete response was improved with pembrolizumab compared with placebo in the main cohort (12.9% vs. 2.0%; P<0.00001) and in the combined main plus FLOT cohort (13.0% vs. 2.4%). A coprimary endpoint of event-free survival was not significantly improved with pembrolizumab compared with placebo (hazard ratio, 0.81), despite numeric superiority for pembrolizumab (median, 44.4 vs. 25.3 months); results were similar in the combined cohort (45.8 vs 25.7 months). There was no difference in overall survival between the pembrolizumab and placebo arms in the main cohort (median, 60.7 vs. 58.0 months; HR, 0.90; P=0.174) or the combined cohort (median, 60.7 vs. 45.7 months). No new safety signals were observed.