Clinicopathological significance of nerve density and perineural invasion in esophageal squamous cell carcinoma: A retrospective study with neuropeptide Y.

The clinicopathological and prognostic implications of neural involvement, including nerve density (ND) and perineural invasion (PNI), in esophageal squamous cell carcinoma (ESCC) remain incompletely understood. This retrospective study aimed to evaluate ND and PNI using neuropeptide Y (NPY) in 201 patients with ESCC. Scoring methods for ND and PNI were applied based on digital slicing, and Receiver Operating Characteristic (ROC), Kaplan-Meier curves, and Cox regression were used for survival analysis and to explore the association between ND/PNI and clinicopathological features. ROC analyses showed slightly higher AUCs for ND compared with PNI-related indicators in predicting overall survival (OS) and disease-free survival (DFS), without statistical significance. Kaplan-Meier curves and Cox model demonstrated that ND was a significant factor influencing survival prognosis [Hazard Ratio (HR) = 2.065, 95% confidence interval (CI) 1.384-3.080, P < 0.001 for OS; HR = 1.929, 95% CI 1.198-3.108, P = 0.007 for DFS]. Multivariate analysis indicated that ND was independently associated with OS (HR = 2.131, 95% CI 1.418-3.021, P < 0.001) and DFS (HR = 1.818, 95% CI 1.112-2.972, P = 0.017). Notably, the adverse prognostic impact of high-level ND was primarily observed among patients with lymph node metastasis (pN+) for both OS and DFS. In this pN+ subgroup, NII (HR = 2.203, P = 0.022) and PNI status (HR = 1.985, P = 0.048) were independently associated with DFS, whereas NISS (HR = 1.821, P = 0.086) did not reach statistical significance. In conclusion, NPY-based quantification of tumor-associated neural parameters-particularly ND-was associated with survival outcomes in ESCC. Overall, these results highlight the potential relevance of NPY-positive neural alterations in ESCC progression and support further investigation into their biological and prognostic roles within the tumor microenvironment.