Predictors of survival and positive surgical margins in pancreatic adenocarcinoma

A single centre retrospective study

Abstract

Introduction: Pancreatic cancer is one of the most common causes of cancer-related death. The overall survival is short even in the small proportion of patients who are eligible for surgical resection. The aims of our study were to assess the value of preoperative CT scan in the prediction of the surgical margin status and to identify imaging characteristics that predict the survival of patients with resectable pancreatic adenocarcinoma. Methods: Our study included 62 patients with histologically confirmed pancreatic adenocarcinoma, who underwent surgery between 1^st January 2010 and 30^th June 2019. Medical records and preoperative CT images were reviewed in order to collect clinical and imaging data. Data was analyzed with the Qui-squared test and with binary logistic regression. Kaplan-Meier estimates, Log-Rank test, and the multivariate Cox proportional hazards regression were used for the survival analysis.

Results: From the 62 patients who underwent surgery, negative surgical margins were achieved in 35 of them and 27 patients had positive surgical margins. In multivariate analysis, vascular involvement was a predictor of positive surgical margins. The mean survival was 789 days, being significantly different between the two groups of surgical margins. In survival analysis, tumor density on CT (HR=0,985, p-value=0,035) and imaging signs of extra-pancreatic perineural involvement (HR=2,324, p-value=0,048) were identified as predictors.

Conclusion: Preoperative CT is a useful tool to predict positive surgical margins and survival. It helps to identify patients with resectable tumors but worse prognosis who can benefit from different therapeutical strategies.

Keywords: pancreas; survival analysis; computer tomography.