Validation of the APRI+ALBI smart phone application for preoperative risk assessment after liver resection in a routine clinical setting

Abstract

Presenter: David Pereyra MD | Mayo Clinic, Rochester Background: Liver surgery obtains the risk of development of postoperative liver dysfunction (LD) and associated increased morbidity and mortality. In order to mitigate the individual risk for adverse postoperative outcome, predictive markers and scores have been developed. In this context, APRI and ALBI, both based on routine laboratory parameters, and their summative conjunction were previously established as predictive markers for postoperative LD in patients undergoing liver resection after neoadjuvant chemotherapy, as well as in a set of more than 12,000 patients in the NSQUIP database. Here, we aimed to validate APRI+ALBI and our Smart phone application system in a routine clinical set of patients undergoing liver surgery. Methods: We identified 160 patients with available data on preoperative albumin, bilirubin, aspartate aminotransferase and platelet count and postoperative outcome. APRI and ALBI were calculated and prognostic value for postoperative outcome was evaluated. Results: Using receiver operating curve analysis, APRI+ALBI was found to be superior to APRI alone or ALBI alone in prediction of postoperative LD (area under the curve [AUC]=0.717, AUC=0.669, AUC=0.691, respectively). Further, APRI+ALBI was superior to parameters of indocyanine green clearance (plasma disappearance rate: AUC=0.628, retention rate at 15 minutes: AUC=0.669). A preoperative cut-off at -2.46 was able to stratify patients at risk for postoperative LD (6.1% vs 18.0%, p=0.017), morbidity (32.7% vs 63.9%, p<0.001), severe morbidity (11.2% vs 34.4%, p<0.001), and mortality within 90 postoperative days (1.0% vs 8.5%, p=0.028). Utilizing our previously established smart phones application, we were able to validate appropriate classification patients in individual risk groups according to their preoperative APRI/ALBI score. Conclusion: Here, we validate APRI+ALBI as a predictive marker for postoperative adverse outcome in a clinical routine set of patients undergoing liver resection. The detection of multiple with pathophysiological processes , such as for example chemotherapy associated liver injury but also alcoholic liver disease, by the APRI/ALBI score, allowed prediction of outcome after liver resection in an routine clinical setting. Applying this twice validated marker based on routine laboratory parameters allows estimation of individual patient risk prior to liver resection and might ultimately reduce occurrence of postoperative complications. A specific and gradual risk assessment seems to be possible using the APRI/ALBI smart phone application in a routine clinical setting. Until more specific markers are available, this application represents a valuable tool to assess individualized patients risk for postoperative outcome based on routine laboratory parameters.