Pembrolizumab for HER2-Positive Gastric or Gastro-Esophageal Junction Cancer: More Interim Results from KEYNOTE-811

In a preplanned interim analysis of the industry-sponsored, placebo-controlled, randomized KEYNOTE-811 trial in patients with HER2-positive esophagogastric adenocarcinoma, adding pembrolizumab to trastuzumab and chemotherapy led to a marked increase in response to therapy — and to accelerated FDA approval of pembrolizumab (NEJM JW Oncol Hematol Sept 2020 and Nature 2021; 600:727). Now, investigators report updates in response, progression-free survival (PFS), and overall survival (OS).

All patients received pembrolizumab (200 mg) or placebo plus trastuzumab in three-week cycles, combined with either capecitabine/oxaliplatin or a 5-day infusion of 5-FU plus cisplatin. Of the 698 treated patients, 85% received capecitabine/oxaliplatin, 67% had gastric primaries, 78% were immunohistochemistry positive at 3+ for HER2, and 85% had a combined positive score (CPS) ≥1% for PD-L1.

At the second interim analysis (median follow-up, 28 months), the response rate was superior for pembrolizumab compared with placebo (72.6% vs. 59.8%). The coprimary endpoint of superior PFS for pembrolizumab compared with placebo at 28 months was met (median, 10.0 vs. 8.1 months; hazard ratio, 0.72), which was maintained at the third interim analysis (median follow-up, 38 months). In patients with CPS ≥1%, pembrolizumab compared with placebo was associated with significantly improved PFS (10.8 vs. 7.2 months; HR, 0.70). However, in those with CPS <1%, median PFS did not differ by treatment (about 9.6 months). Median OS, the other coprimary endpoint, trended nonsignificantly in favor of pembrolizumab over placebo (20.0 and 16.9 months). This possible benefit was seen only in patients with CPS ≥1% (20.5 and 15.6 months, respectively), with a survival detriment suggested in patients with CPS <1% (16.1 and 22.3 months, respectively). No new safety signals were observed.