Tumor-Infiltrating Neutrophils Predict Poor Survival of Non-Functional Pancreatic Neuroendocrine Tumor.

Abstract

This study retrospectively characterized the immune infiltrating profile in nonfunctional pancreatic neuroendocrine tumors (NF-PanNETs). Tumor tissues from the 109-patient Fudan cohort and a 73-patient external validation set were evaluated by immunohistochemistry for 9 immune cell types: tumor-infiltrating neutrophils (TINs), tumor-associated macrophages (TAMs), CD11c+ dendritic cells, anti-NCR1+ natural killer (NK) cells, CD4+ and CD8+ T cells, CD45RO+ memory T cells, FOXP3+ regulatory T cells (Tregs), and CD20+ B cells. TINs were primarily distributed in the intratumoral area, dendritic cells and NK cells were scattered evenly in intratumoral and stromal areas, and Tregs were rarely detected. The remaining 5 cell types were primarily present in peritumoral stroma. Total TINs (P < .001) and TAMs (P = .002) increased as NF-PanNET grade rose. Kaplan-Meier analyses showed that high intratumoral TINs, total TAMs, and stromal CD4+ T-cell infiltration correlated with shorter recurrence-free survival (RFS, P = .010, P = .027, and P = .035, respectively) and overall survival (OS, P = .017, P = .029, and P = .045, respectively). Additionally, high intratumoral CD8+ T cell infiltration correlated with prolonged RFS (P = .039). Multivariate Cox regression demonstrated that intratumoral TINs, World Health Organization (WHO) classification, and eighth edition of the American Joint Committee on Cancer tumor-node-metastasis staging system (AJCC8th TNM) were independent factors for RFS (P = .043, P = .023, and P = .029, respectively), whereas intratumoral TINs and WHO classification were independent factors for OS (P = .010 and P = .007, respectively). Furthermore, the combination of TINs, WHO classification, and AJCC8th TNM remarkably improved prognostic accuracy for RFS. These results have been verified in the external validation set. Intratumoral TINs are an independent and unfavorable predictor of postoperative NF-PanNETs. A combination of TINs, WHO classification, and AJCC8th TNM could improve prognostic accuracy for RFS.