PD-L1 expression and GEP have been associated with response to anti–PD-1/PD-L1 therapy. We studied PD-L1 expression and GEP in pts receiving standard therapy for GC in South Korea and the US in an observational study.
Tumor samples were collected from 2003-2017 from the Yonsei Cancer Center (South Korea), MD Anderson Cancer Center (US), and Memorial Sloan Kettering Cancer Center (US). PD-L1 expression was assessed by PD-L1 IHC 22C3 pharmDx assay (Agilent Technologies); PD-L1+ was defined as combined positive score (CPS) ≥1. GEP was analyzed only in Korean pts. GEP score was derived from an 18-gene signature measured using extracted tumor RNA analyzed by NanoString nCounter platform. High/low GEP was based on a cutoff of –0.318. PD-L1 expression was analyzed using chi-squared and multiple logistic regression; survival was estimated using the Kaplan-Meier method, log-rank test, and Cox proportional hazards models.
574 tumor samples (73% biopsy, 27% surgical resection) from pts with GC (median [range] age 61 [21-90] years) were analyzed. Of the 574 pts, 49% (277/567) had stage IV disease; 43% (249/574) had operable disease (half of whom received adjuvant chemotherapy), and 57% (325/574) had inoperable disease (74% received chemotherapy, 18% received chemoradiation). The prevalence of PD-L1+ expression was 67% overall, 69% in Korean pts, 66% in US pts, 68% in biopsy samples, and 67% in surgical resection samples. Greater PD-L1+ expression was seen in older (≥65 years) vs younger pts (P = 0.011) and in pts with inoperable vs operable disease (P < 0.0001). PD-L1 expression was not associated with overall survival (OS) (log-rank test P = 0.25). 85% (243/286) of Korean pts had high GEP. High GEP was associated with tumor grade (P < 0.001) and clinical stage (P < 0.0001). High GEP may be associated with longer OS (adjusted hazard ratio 0.46 [95% CI 0.32-0.66]). GEP was mildly correlated with PD-L1 CPS (Spearman r = 0.24 [P < 0.0001]).
PD-L1 expression was similar in Korean and US pts with GC and in biopsy and surgical resection samples. GEP was associated with OS while PD-L1 expression was not. Further investigation of GEP with anti–PD-1/PD-L1 therapy is warranted.
Clinical trial identification
Legal entity responsible for the study
Merck & Co., Inc.
Merck & Co., Inc.
Medical writing and/or editorial assistance was provided by Sarita S. Shaevitz, PhD, of the ApotheCom pembrolizumab team (Yardley, PA, USA). This assistance was funded by Merck & Co., Inc., Kenilworth, NJ, USA.
S.Y. Rha: Advisory board member: Eli Lilly, MSD, Celltrion, Ipsen; Speakers\' bureau: Eli Lilly; Institution research funding: MSD, BMS. G.Y. Ku: Advisory board member: Merck & Co., Inc., Eli Lilly, AstraZeneca; Institution research funding: Merck & Co. Inc., BMS, Arog Pharmaceuticals, AstraZeneca. H.C. Chung: Advisory board member: Taiho, Celltrion, MSD, Eli Lilly, Quintiles, BMS, Merck-Serono; Speakers\' bureau: Merck-Serono, Lilly, Foundation Medicine; Institution research funding: Lilly, GSK, MSD, Merck-Serono, BMS-Ono, Taiho. C. Fein: Employee of Novant Health; Travel expenses, accommodations: Aduro Biotech. M.J. Savage, W. Zhou, S.A. Peter: Employee, Stock owner: Merck & Co., Inc. T. Wu: Employee of Merck & Co., Inc. at the time the analysis was conducted for this abstract. J.A. Ajani: Institution research funding: BMS, Merck & Co., Inc., Taiho, DeltaFly, Gilead. All other authors have declared no conflicts of interest.