Search

Platelets are believed to promote tumor growth and metastasis in several tumor types. The prognostic role of blood platelets in pancreatic ductal adenocarcinoma (PDAC) remains controversial, and the prognostic value of tumor-infiltrating platelets (TIPs) remains unknown.A total of 303 patients who underwent curative pancreatectomy for PDAC were enrolled from two independent centers in China and divided into three cohorts. Paired preoperative blood samples and surgical specimens from all patients were analyzed. The correlations between patient outcomes and preoperative blood platelet counts and the presence of TIPs, respectively, were analyzed. TIPs were identified by immunohistochemical staining of CD42b. Prognostic accuracy was estimated by concordance index (C-index) and Akaike information criterion (AIC).TIPs, but not preoperative blood platelet counts, were associated with overall survival (OS; all P 0.001) and recurrence-free survival (RFS; all P 0.001) in the training, testing, and validation sets. Positive CD42b expression predicted poor postsurgical survival. Incorporation of TIPs improved the predictive accuracy of the 8th edition American Joint Committee on Cancer (AJCC) tumor-node-metastasis (TNM) staging system for OS in each of the three cohorts (C-index: 0.7164, 0.7569, and 0.7050, respectively; AIC: 472, 386, and 1019, respectively). The new predictor system was validated by incorporating TIPs with the 7th edition AJCC TNM staging system (C-index: 0.7052, 0.7623, and 0.7157; AIC: 476, 386, and 1015).TIPs were an independent prognostic factor that could be incorporated into the AJCC TNM staging system to refine risk stratification and predict surgical outcomes of patients with PDAC.

The role of N classification is controversial in several prognostication systems proposed for pancreatic neuroendocrine neoplasms (pNENs). The widely accepted modified European Neuroendocrine Tumor Society (mENETS) system suggests this contradiction may be related to T classification.Data were collected retrospectively from 981 patients in the Surveillance, Epidemiology, and End Results (SEER) database (1973-2012; cohort 1) and 140 patients from the Pancreatic Cancer Institute of Fudan University (2006-2016; cohort 2). All patients had resected well- to moderately differentiated locoregional pNENs, whereby the mENETS system was adopted. Factors related to N1 classification and the association between N and T classifications were analyzed, and N classification prognosis based on T classification was assessed.In cohorts 1 and 2, tumor size (2-4 cm: p 0.001 and p = 0.037, respectively; 4 cm: p 0.001 and p = 0.012, respectively) and tumors extending beyond the pancreas (p 0.001 and p = 0.016, respectively), which are factors for T classification, affected N1 classification. For tumors limited to the pancreas, the N1 classification was associated with tumor size (p 0.001 and p = 0.046, respectively) and predicted poor disease-specific survival (DSS), while for tumors extending beyond the pancreas, the N1 classification did not affect patient outcomes. Findings obtained with data from the SEER database were reproducible with our institutional data.N classification is associated with T classification, limiting the value of N1 classification for the pNENs tumor-node-metastasis system. A new risk model is necessary to predict patient outcomes and guide clinical practice for the prognosis of pNENs.

Tumor-infiltrating neutrophils (TINs) indicate poor prognosis for patients with pancreatic ductal adenocarcinoma (PDAC). Activated neutrophils can generate neutrophil extracellular traps (NETs). Little is known about the presence and prognostic significance of tumor-infiltrating NETs in PDAC. This study enrolled 317 patients, in two independent sets (training and validation), who underwent curative pancreatectomy for PDAC in Shanghai Cancer Center. TINs and NETs were identified by immunohistochemical staining for CD15 and citrullinated histone H3, respectively. The relationship between clinicopathological features and outcomes was analyzed. Accuracy of prognostic prediction models was evaluated using concordance index (C-index) and Akaike information criterion (AIC). NETs were associated with OS (both, P