Recent genomic and molecular characterization of gastric adenocarcinoma (GAC), such as Cancer Genome Atlas (TCGA), have shed light on the PD-L1 positivity and four different subtypes. Several PD-L1 IHC assays of scoring-criteria are developed in parallel. This study aimed to evaluate PD-L1 staining patterns and molecular profiling by TMA in GAC.
In total, 331 consecutive patients (pts) with GAC underwent curative surgery at Aichi Cancer Center Hospital between Jan 2009 and Dec 2010. Pts with no systemic chemotherapy before surgery and sufficient tumor content on the formalin-fixed and paraffin-embedded (FFPE) slides were eligible. EBV-encoded RNA in situ hybridization, IHC for deficient mismatch repair proteins (dMMR, MLH1/MSH2/MSH6/PMS2), HER2 and PD-L1 (22C3, 28-8, E1L3), and HER2 DISH were evaluated in TMA. Four cores of tumor tissue were punched out from individual FFPE tumor blocks and arrayed in a new TMA block. PD-L1 tumor positivity was categorized to ≥ 1%/≥25%/≥50% tumor cell membrane staining. Disease free survival (DFS) and overall survival (OS) were estimated by the Kaplan-Meier method.
Among the cases, 226 pts were included in this study, excluding pts with prior chemotherapy (n = 26) and insufficient tumor samples (n = 79). Patient characteristics were as follows: male/female, 162/64; pStage (AJCC 7th ed.) I/II/III/IV, 100/39/58/29; EBV +/-, 13/213; proficient MMR/dMMR, 197/29, HER2 +/-, 24/202, and PD-L1 ≥1%/≥25%/≥50% (22C3 vs. 28-8 vs. E1L3), 11/4/2 vs. 7/6/6 vs. 12/10/8. Among 22C3, 28-8, and E1L3 assays, 94–97% of the pairs were concordant. PD-L1 positivity by all three assays was associated with EBV + (p
The three tested PD-L1 assays did not show comparable staining patterns in all cases. Studies that correlate not only PD-L1 staining patterns but also EBV+/dMMR/HER2+ and response to immunotherapy are required to evaluate.Table:
|EBV+ N = 13||EBV− N = 213||P||pMMR N = 197||dMMR N = 29||P||HER2+ N = 24||HER2− N = 202||P|
|E1L3||+ −||3 10||9 204||0.02||8 191||4 25||0.06||0 24||12 190||0.62|
|28-8||+ −||3 10||4 209|
Clinical trial identification
Legal entity responsible for the study
All authors have declared no conflicts of interest.