Although current recommendations suggest that immune checkpoint inhibitors (ICIs) may be used for all patients with advanced gastric, gastro-oesophageal junction, or oesophageal adenocarcinoma in the first-line setting, the analyses of previously unpublished data from pivotal phase III randomised studies fail to demonstrate a survival advantage when adding ICIs to conventional chemotherapy in patients with low PD-L1–expressing tumours. According to Dr. Raghav Sundar and colleagues of the National University Cancer Institute and National University of Singapore, who published their findings on 3 December 2021 in the Journal of Clinical Oncology, these findings provide the basis to initiate a more detailed reassessment in the selection of treatment options for first-line metastatic gastric, gastro-oesophageal junction, or oesophageal adenocarcinoma.

The authors wrote in the background that ICIs have recently demonstrated significant survival benefit in combination with chemotherapy in the first-line treatment of advanced gastric, gastro-oesophageal junction, or oesophageal adenocarcinoma in CheckMate-649 and KEYNOTE-590 studies. Based on these results, the US Food and Drug Administration (FDA) has granted regulatory approval for the use of nivolumab in the first-line treatment of all advanced gastro-oesophageal adenocarcinomas and pembrolizumab for oesophageal and gastro-oesophageal junction carcinoma. More recently, the FDA has also granted regulatory approval for the addition of pembrolizumab to chemotherapy and trastuzumab for HER2–positive gastric cancer. 

PD-L1 expression measured by calculating the combined positive score (CPS) from immunohistochemistry was studied as a predictive biomarker in these studies. Regulatory approval for nivolumab, however, was granted regardless of PD-L1 status. The efficacy of the nivolumab-chemotherapy combination compared with standard chemotherapy regimens is evident in patients with high PD-L1–expressing tumours, but the efficacy in patients with low PD-L1–expressing tumours is unclear. Data presented thus far from these studies focus on the presentation of all randomly assigned populations and PD-L1–positive (CPS ≥ 1 or ≥ 5 or 10) populations. Forest plots have not informed several clinically relevant PD-L1 subgroups. Therefore, it remains unclear if the survival benefit in all randomly assigned populations is driven largely by the benefit in PD-L1 CPS–positive patients.

In their analysis, the researchers from Singapore tried to elucidate unreported treatment outcomes in low PD-L1–expressing tumours by deriving individual patient survival data, reconstructed from the reported Kaplan-Meier (KM) plots. A graphical reconstructive algorithm was adopted to estimate time-to-event outcomes from reported overall survival (OS) and progression-free survival (PFS) KM plots describing overall or subgroup cohorts. Using reconstructed time-to-event outcomes, KMSubtraction conducts bipartite matching of patients from the reported subgroup among the overall cohort. By excluding matched patients, KM plots and survival analyses of the unreported subgroups were retrieved.

CheckMate-649, KEYNOTE-062, and KEYNOTE-590 were included in the analysis. Two PD-L1 subgroups were identified with data unreported in the primary manuscripts: PD-L1 CPS 1-4 from CheckMate-649 and PD-L1 CPS 1-9 from KEYNOTE-062. No significant differences in OS and PFS were demonstrated in ICI-chemotherapy combinations when compared with chemotherapy among CheckMate-649 PD-L1 CPS 1-4 (OS hazard ratio [HR] 0.950, 95% confidence interval [CI] 0.747 to 1.209, p = 0.678; PFS HR 0.958, 95% CI 0.743 to 1.236, p = 0.743) and KEYNOTE-062 PD-L1 CPS 1-9 subgroups. In the KEYNOTE-062 PD-L1 CPS 1-9 subgroup, patients treated with pembrolizumab had an increased hazard of tumour progression (HR 2.092, 95% CI 1.661 to 2.635, p < 0.001).

By using a novel workflow, KMSubtraction, data of PD-L1 subgroups previously unreported by primary manuscripts were retrieved for pivotal randomised phase III studies. The authors concluded that the results suggest the lack of benefit when adding ICIs to chemotherapy in low PD-L1–expressing gastric, gastro-oesophageal junction, or oesophageal adenocarcinoma.

Dr Sundar was supported by the National Medical Research Council; as part of the Singapore Gastric Cancer Consortium, the study was also partially funded by the NMRC Open Fund-Large Collaborative Grant.


Zhao JJ, Yap DWT, Chan YH, et al. Low Programmed Death-Ligand 1–Expressing Subgroup Outcomes of First-Line Immune Checkpoint Inhibitors in Gastric or Esophageal Adenocarcinoma. JCO; Published 3 December 2021. DOI: 10.1200/JCO.21.01862