Usefulness of the abdominal fat area-based risk score for postoperative pancreatic fistula after pancreatoduodenectomy

Abstract

BackgroundPancreatic fistulas remain a significant concern after pancreatectomy, and it is not possible to perform preoperative risk stratification for all patients. This study aimed to evaluate the usefulness of a unique risk model, based on the abdominal fat area (AFA) calculated by computed tomography, for pancreatic fistula development after pancreatoduodenectomy and compare it with models based on the body mass index (BMI) or abdominal thickness.Material and Methods Patient characteristics, preoperative laboratory data, radiographic findings, and their association with pancreatic fistula development after pancreaticoduodenectomy were analysed for 158 patients who underwent resection between 2011 and 2017. Clinically relevant postoperative pancreatic fistulas (CR-POPF) were defined as Grade B or C fistulas based on the International Study Group of Pancreatic Surgery (ISGPS) 2016 consensus.ResultsCR-POPF developed in 38 patients (24.2%). Multivariate logistic analysis indicated that the AFA, BMI, and intra-abdominal thickness were potential candidates for predictive models for pancreatic fistula development, small pancreatic duct diameter, diabetes mellitus development, and the pathology of non-pancreatic cancers. When comparing the three risk models (AFA-, BMI-, and intra-abdominal thickness-based), the AFA-derived risk model was superior to the BMI-based and intra-abdominal thickness-based risk stratification models (area under the curve 0.836 vs 0.824 vs 0.826).Conclusions The risk model based on AFA calculation was superior to that based on BMI or intra-abdominal thickness measurements. The model must be validated further to elucidate the efficacy of the risk scoring system in more detail.