Antibodies against programmed death 1 (PD-1) and its ligand, programmed death-ligand 1 (PD-L1) have recently shown promising results in gastric cancer (GC). However, clinicians still lack predictive biomarkers for the efficacy of anti-PD-1 therapy; thus, we investigated the expression of PD-L1 in GC and further assessed its clinical relevance with other clinicopathological features.
We retrospectively collected clinical data on 968 consecutive GC cases from Nanfang Hospital between November 2018 and August 2021. Discrepancy in the combined positive score (CPS) of PD-L1 protein expression between gastric mucosa biopsy and postoperative pathology were investigated. Correlations between CPS and clinicopathological parameters were determined using chi-squared test, multiple logistic aggression analysis, and linear regression analysis.
Among the 968 consecutive GC patients, 199 who did not receive preoperative chemotherapy or immunotherapy were tested for CPS both in gastric mucosa biopsy and postoperative pathology, and the results showed that the CPS of gastric mucosa biopsy was significantly lower than that of postoperative pathology [mean ± SD: 5.5 ± 9.4 vs. 13.3 ± 17.4; M(IQR): 2(5) vs. 5(12), p<0.001)]. 62.3% of patients (579/930) had CPS≥ 1, 49.2% of patients (458/930) had CPS≥5, and 33.3% of patients (310/930) had CPS≥10. Mismatch repair deficiency (dMMR) status was seen in 6.1% of patients (56 of 919). Positive Epstein–Barr virus (EBV) status was detected in 4.4% of patients (38 of 854). The patients with CPS≥1/CPS≥5/CPS≥10 were significantly independently correlated with age, Lauren classification, Ki-67 index, and EBV status. According to linear regression analysis, PD-L1 expression was correlated with age (p<0.001), Ki-67 index (p<0.001), EBV (p<0.001), and Lauren classification (p=0.002).
Our results confirmed that PD-L1 expression has Intratumoral heterogeneity in GC. Furthermore, the variables of age, Ki-67 index, and Lauren classification, which are common and accessible in most hospitals, are worth exploring as potential biomarkers for anti-PD-1 therapy in GC.