Survival Benefit Is Unchanged With Addition of Pembrolizumab in G/GEJ Adenocarcinoma
Final analysis from KEYNOTE-585 revealed that event-free survival (EFS) was not significantly improved.
Gastric stomach cancer cells, closeup view, 3D illustration: © LASZLO - stock.adobe.com
Final analysis from KEYNOTE-585 (NCT03221426) revealed that event-free survival (EFS) was not significantly improved, although the pathological complete response (pathCR) rate favored the treatment arm of perioperative pembrolizumab (Keytruda) and chemotherapy compared with the control arm of placebo and chemotherapy in patients with locally advanced gastric/gastroesophageal junction (G/GEJ) adenocarcinoma. Kohei Shitara, MD, presented results during the European Society for Medical Oncology Gastrointestinal Cancers Congress, June 26 to 29, 2024, in Munich, Germany.
At a median follow-up of 59.9 months, the median EFS for the pembrolizumab arm was 44.4 months vs 25.7 months for the control arm (HR, 0.81; 95% CI, 0.67–0.98). Further, the median overall survival (OS) was 71.8 months with pembrolizumab vs 55.7 months with placebo (HR 0.86; 95% CI 0.71–1.06).
“The pathCR rate by central assessment was 13.4% in the pembolizumab arm and 2% in the placebo arm, so there was an improvement of 11.4 [percentage points; 95% CI, 8.0-15.3],” Shitara, director, Department of Gastrointestinal Oncology, National Cancer Center Hospital East, in Japan, said during the presentation.
A total of 1007 patients with localized G/GEJ cancer were randomly assigned to receive 200 mg of pembrolizumab every 3 weeks and chemotherapy (XP/FP; capecitabine and cisplatin/5-fluorouracil/cisplatin) vs placebo and chemotherapy (main cohort; n = 804). A subcohort (n = 203) used the FLOT (fluorouracil, leucovorin, oxaliplatin, and docetaxel) regimen instead of the XP/FP backbone.
Patients who received FLOT had similar results. The median EFS was 47.0 months in the pembrolizumab arm compared with 26.9 months in the control arm (HR, 0.80; 95% CI, 0.67-0.95). OS was not reached in the pembrolizumab arm (NR; 95% CI, 59.2-NR) compared with 55.7 months in the control arm (95% CI, 42.8-NR) with an HR of 0.86 (95% CI, 0.71-1.03). PathCR was 14.2% in the pembrolizumab group vs 2.8% in the placebo group showing an 11.4 percentage point difference (95% CI, 8.1-15.0).
Patients were stratified by geographic region (Asia vs non-Asia), tumor staging (II vs III vs IVa), and chemotherapy backbone (XP/FP vs FLOT). The primary end points were pathCR rate, EFS, OS for the main cohort, and safety for the FLOT cohort. Secondary end points were safety in the main cohort and safety, OS, and EFS for both cohorts.
“The patient characteristics were well-balanced between the 2 arms in the main cohort and nearly half of the enrolled patients were from Asia,” Shitara said. PD-L1 status was also similar between the treatment vs control group with the majority of patients having a combined positive score (CPS) of 1 or greater (73% vs 76%), followed by CPS of less than 10 (65% vs 65%), and CPS score of 10 and greater (26% vs 29%). Microsatellite instability high, tumor stage, and tumor location were similar for the treatment and control groups, according to investigators.
Adverse events (AEs) for the treatment and control group were the same, with 99% of patients reporting any grade AEs for the pembrolizumab group (n = 399) and placebo group (n = 400) each. Treatment-related (TR) AEs were also similar for both groups: grade 3-4 were reported in 64% of the pembrolizumab group vs 63% in the placebo group with 4 patients (1%) in the pembrolizumab group and 2 patients (<1%) in the placebo group reporting grade 5 TRAEs. In addition, investigators reported zero grade 5 immune-related (IR) AEs for both groups. Grade 3-4 IRAEs was 11% vs 3%, respectively.
“Safety and efficacy outcomes at final analysis were consistent with prior studies,” Shitara said. “These results suggest that we need additional research to study checkpoint inhibitors in the perioperative setting for gastric cancer,” he concluded.
