Concordance of programmed death-ligand 1 (PD-L1) status in primary esophageal squamous cell carcinoma (ESCC) and matched visceral metastases.

435 Background: PD-L1 expression is a predictive biomarker for response to first-line immunochemotherapy (anti-PD-1 therapy plus platinum-based chemotherapy) in advanced ESCC. However, the concordance of PD-L1 expression between primary ESCC tumors and their matched metastatic tumors remains unclear. Methods: This study evaluated the concordance of PD-L1 expression between primary and metastatic ESCC tumors. Patients with archival tumor tissues from both primary and visceral metastatic sites were included. PD-L1 expression was assessed using the combined positive score (CPS) via immunohistochemistry with the PD-L1 (SP142) assay on formalin-fixed paraffin-embedded tumor tissues. PD-L1 CPS were analyzed both as a continuous variable and as a categorical variable using CPS cut-offs of 1, 5, and 10. Correlation and concordance between the PD-L1 expression levels of primary and metastatic tumors were assessed. Results: A total of 78 paired primary and metastatic ESCC tumor specimens were analyzed, including 44 lung, 17 liver, and 17 brain metastases from patients diagnosed between 2006 and 2016. The median interval between primary and metastatic sample collection was 290 days (range: 0-1,992 days). While a statistically significant correlation of the PD-L1 expression levels of primary and metastatic tumors was observed ( P = 0.019), the Pearson correlation coefficient was low (r = 0.266), indicating a weak correlation. PD-L1 CPS concordance rates were 67.9% at a CPS cut-off of 1, 56.4% at a CPS cut-off of 5, and 67.9% at a CPS cut-off of 10. Subgroup analysis based on time interval between samples collection showed concordance rates ranging from 52.7% to 73.9% across different CPS cut-offs. Conclusions: A weak correlation and a moderate-to-good agreement in the PD-L1 expression between primary and metastatic ESCC tumors suggest that PD-L1 assessment should not rely solely on either primary or metastatic specimens. Whenever feasible, obtaining representative tissue from multiple tumor sites may enhance the accuracy of PD-L1 evaluation, improving clinical decision-making in the context of immunochemotherapy.