Validation of the alternative fistula risk score for pancreatoduodenectomy by a single high volume center in the United States

Abstract

Introduction: Postoperative pancreatic fistula (POPF) is the most common cause of postoperative morbidity after pancreatoduodenectomy (PD). The alternative fistula risk score (a-FRS), a modified version of the fistula risk score (FRS), has been proposed to predict the risk of POPF after PD. The aFRS eliminates requirement of estimated blood loss for prediction. The aim of this study was to validate the performance of the a-FRS and compare it to that of FRS. Methods: Patients undergoing PD at Johns Hopkins Hospital between 2010 and 2015 were identified and clinicopathological data were extracted from institutional database. Missing data were estimated using multiple imputations. Performance of a-FRS and FRS was assessed by evaluating area under receiver-operating curve (AUC), and DeLong test was used to compare difference between AUCs. Observed mean risk was calculated for risk groups. Results: A total of 1406 patients were included and the mean age was 62.9±11.7 years. A majority was female(53.5%), and white (79.2%). The median BMI was 25.1(IQR:22.3-28.5), median duct size was 3.4 (IQR:1.9-5.0)mm, and a majority (53.1%) had hard gland. 164(11.7%) patients developed grade B/C fistula. Upon external validation an AUC of 0.82(95%CI:0.79-0.86) was observed for a-FRS, as compared to AUC of 0.81(95%CI:0.78-0.84) for FRS (p=0.447)(Figure-1). The mean observed risk for low-, intermediate-, and high-risk groups as defined by a-FRS was 3.5%, 10.9%, 28.4% respectively (p< 0.001). Conclusion: In this external validation, the a-FRS performed well and was comparable to the FRS. Furthermore, based on the suggested cut-off, a significant and clinically relevant risk stratification was achieved.