Expression Patterns and Prognostic Value of DNA Damage Repair Proteins in Resected Pancreatic Neuroendocrine Neoplasms.

OBJECTIVE: This study aimed to examine the expression profiles and prognostic value of multiple DDR proteins in resected PanNENs. BACKGROUND: DDR proteins play important roles in various cancers, including pancreatic ductal adenocarcinoma. However, the expression patterns and prognostic value of DDR proteins in PanNENs remain unclear. METHODS: This retrospective analysis included PanNEN patients who underwent resection at the Fudan University Shanghai Cancer Center from 2012 to 2018. Immunohistochemical staining was performed for 12 DDR proteins in tissue microarrays. The associations of DDR protein expression and clinicopathological features with recurrence-free survival (RFS) were examined via a Cox regression model and random survival forest. A recurrence signature was constructed using recursive partitioning analysis. RESULTS: In total, 131 PanNEN patients were included, with 32 (24.4%) cases of recurrence. Among the 12 DDR proteins, low checkpoint kinase 2 (CHK2) expression (P = 0.020) and loss of ataxia-telangiectasia-mutated (ATM) (P = 0.0007) significantly correlated with recurrence. Multivariable Cox regression analysis identified tumor size >/=3 cm, lymph node (LN) metastasis, high tumor grade, low CHK2 expression, and ATM loss as independent risk factors for recurrence. A recurrence signature was established based on the importance of recurrence-specific risk factors; patients with the LNnegTumorSize<3cm signature had a 5-year RFS rate of 96.8%, whereas patients with the LNposCHK2low signature had the worst 5-year RFS rate (0%). Discrimination (concordance index: 0.770) and calibration plots indicated that the recurrence signature had a good ability to identify patients at risk for recurrence. CONCLUSIONS: By analyzing large-scale tissue microarrays of PanNENs, we evaluated 12 DDR protein expression profiles. We developed a recurrence signature that can identify distinct subpopulations according to RFS, which may help refine individual follow-up.