Liver scoring system and hepatectomy outcomes: a veterans affairs corporate data warehouse analysis

Abstract

Presenter: Yazan Ashouri MD | University of Arizona Background: Several liver scoring systems have been introduced. The optimal predictive liver scoring system for hepatectomy outcomes is yet to be defined. Methods: The Veterans Affairs (VA) Corporate Data Warehouse (CDW) and VA Informatics and Computing Infrastructure (VINCI) were utilized to identify veterans who underwent hepatectomy between 2008-2018. Logistic regression was performed to evaluate the association between ALBI grade, FIB4 and APRI with 90-day mortality and cirrhosis decompensation after surgery. Receiver operating characteristic (ROC) curves and Youden index were performed to determine optimal cutoff values. Kaplan-Meier method and log-rank test were used for survival analysis. Results: A total of 1477 veterans underwent hepatectomy in VA hospitals during the study period. Most of the patients were white (60.2%), males (97%) and underwent surgery for hepatocellular carcinoma (62.8%), followed by colorectal liver metastases (20.3%) and biliary cancers (11.2%). Partial hepatectomy was the most common operation (73.5%) followed by right hepatectomy (12.3%). The 90-day mortality was 5.1% and the rate of cirrhosis decompensation, with development of ascites or encephalopathy within 90 days from surgery, was 9.8%. Multivariable logistic regression showed that FIB4 and APRI were associated with 90-day mortality: Odds Ratio (OR): 1.17, Confidence interval (CI) 1.07-1.28 (p 2 was not associated with 90-day mortality (OR: 1.60, CI: 0.86-2.97). FIB4 and APRI were also independently associated with cirrhosis decompensation within 90 days: OR: 1.18, CI: 1.09-1.28 (P < .0001) and OR: 1.29, CI: 1.11-1.51 (P < .001), respectively. ALBI grade was not associated with cirrhosis decompensation (OR:1.18, CI: 0.75-1.84). ROC curves evaluating the ability to predict cirrhosis decompensation showed areas under the curve (AUC) of 0.66 for FIB4 and 0.65 for APRI. Youden index analysis showed that the best cutoff values for FIB4 and APRI to be 2.56 and 0.70, respectively. Survival analysis revealed that patients with FIB4 ≥ 2.56 had worse overall survival (OS) than patients with FIB4 < 2.56, OS: 46.8 versus 62.4 months (Log-rank: P < .0001). Likewise, veterans with APRI ≥ 0.70 had worse survival when compared to veterans with APRI < 0.70, OS: 51.6 versus 60.8 months (Log-rank: P < .01). Conclusion: In this large cohort study, FIB4 and APRI were independently associated with 90-day mortality and post-hepatectomy cirrhosis decompensation. They were also shown to be associated with overall survival. These findings suggest that these non-invasive fibrosis markers could be useful in the preoperative selection criteria for hepatectomy.