Association between PD-1/PD-L1 Expression and Clinicopathological Parameters as well as Prognosis in Neuroendocrine Neoplasms of the Digestive System: A Systematic Review and Meta-Analysis

Abstract

Abstract Neuroendocrine neoplasms (NENs) are a highly heterogeneous group of tumors originating from the neuroendocrine system, predominantly occurring in the gastrointestinal tract and lungs. This meta-analysis aimed to clarify the association of programmed death 1 (PD-1)/programmed death-ligand 1 (PD-L1) expression with clinicopathological parameters and prognosis in patients with digestive system NENs. A systematic literature search was performed in PubMed, Web of Science, Embase, the Cochrane Library, China National Knowledge Infrastructure (CNKI), Wanfang Database, and China BioMedical Literature Database (CBM) from each database’s inception to February 2026. Eligible studies were analyzed using Stata 17.0 software, with 29 studies involving 2,633 patients included.Results showed no significant association between PD-1 expression and clinicopathological parameters or prognosis. In contrast, PD-L1 expression was closely correlated with high tumor grade (OR=2.73, 95%CI: 1.33-5.60; p=0.006) and pancreatic primary tumor site (OR=2.58, 95%CI: 1.21-5.52; p=0.014). Regarding prognosis, positive PD-L1 expression was associated with inferior overall survival (OS) in gastroenteropancreatic NENs (GEP-NENs) (HR=1.77, 95%CI: 1.25-2.51; p=0.001) but favorable OS in esophageal NENs (HR=0.60, 95%CI: 0.37-0.99; p=0.047). In conclusion, PD-L1 expression correlates with high tumor grade and pancreatic origin. It serves as a predictive biomarker for poor OS in GEP-NENs but favorable prognosis in esophageal NENs, highlighting the divergent prognostic value of the PD-1/PD-L1 axis across digestive system NEN subtypes. These findings provide hypothesis-generating evidence for the potential rationale of immune checkpoint inhibitors in high-grade or poorly differentiated GEP-NENs, although their clinical application requires further prospective validation.