Background: The outlook for patients with esophageal and gastric (EG) cancer remains poor. Hence, there is a compelling need to identify novel treatment strategies and complementary biomarkers. Programmed death ligand 1 (PD-L1) and mismatch repair deficiency (dMMR) are putative biomarkers of response to immune-checkpoint blockade, but their prognostic value and interrelationship in EG cancer have been sparsely investigated. Methods: Immunohistochemical expression of PD-L1 on tumour cells (TC) and tumour-infiltrating immune cells (TIC), and of PD-1 (programmed death receptor 1) on TIC was assessed using tissue microarrays with primary tumours and a subset of paired lymph node metastases from a consecutive, retrospective cohort of 174 patients with chemoradiotherapy-naïve EG adenocarcinoma. MMR proteins MLH1, PMS2, MSH2, and MSH6 were assessed by immunohistochemistry. The total number (intratumoural, tumour-adjacent, and stromal) of CD8+ T cells in each core was calculated by automated analysis. Results: High PD-L1 expression on both TC and TIC, but not PD-1 expression, was significantly associated with dMMR. PD-L1 expression on TIC was significantly higher in lymph node metastases than in primary tumours. High expression of PD-L1 or PD-1 on TIC was significantly associated with a prolonged survival, the former independently of established prognostic factors. A significant stepwise positive association was found between CD8+ T cells and categories of PD-L1 expression on TIC. Conclusion: PD-L1 expression on TIC is higher in lymph node metastases compared to primary tumours, correlates with dMMR, and is an independent factor of prolonged survival in patients with chemoradiotherapy-naïve EG adenocarcinoma. These findings suggest that PD-L1 expression on TIC may be a useful biomarker for identifying patients who may not need additional chemo- or chemoradiotherapy, and who may benefit from PD-1/PD-L1 immune-checkpoint blockade.
Keywords: MMR status; MSI status; PD-1; PD-L1; esophageal cancer; gastric cancer; the cancer genome atlas.