A narrative review of PD-L1 in esophageal cancer: biomarker uncertainty, surgical implications, and emerging pathways to personalization

Background and Objective: Esophageal cancer remains a deadly disease with high recurrence rates despite advances in multimodal therapy. Over the past decade, the introduction of immune checkpoint inhibitors (ICIs) has transformed treatment options for locally advanced disease. Among the key biomarkers in immunotherapy, programmed death-ligand 1 (PD-L1) has emerged as a central—though imperfect—tool for patient selection and treatment planning. XXXXXXXXXXXX.

Methods: A narrative review was conducted using PubMed and Embase to identify relevant studies published between January 2010 and March 2026. Studies evaluating PD-L1, immunotherapy, and related biomarkers in esophageal and gastroesophageal cancers were included.

Key Content and Findings: PD-L1 expression is frequently assessed to stratify patients in immunotherapy trials; however, its reliability remains inconsistent due to intratumoral heterogeneity, variability between primary and metastatic sites, assay differences, and temporal changes following treatment. While higher PD-L1 expression is generally associated with greater benefit from ICIs, responses are also observed in PD-L1-low populations. Emerging evidence suggests that PD-L1, particularly when integrated with other biomarkers such as microsatellite instability (MSI), tumor mutational burden (TMB), and circulating tumor DNA (ctDNA), may help refine treatment strategies across neoadjuvant, adjuvant, and surveillance settings. The use of PD-L1 to guide surgical decision-making, including potential omission of surgery in complete responders, remains investigational.

Conclusions: PD-L1 plays an important but limited role as a standalone biomarker in esophageal cancer. Its optimal use will likely depend on integration into a broader, multi-biomarker framework to support personalized, risk-adapted treatment strategies.