Applying LASSO logistic regression for prediction of biliary complications after ex-vivo liver resection and autotransplantation in end-stage hepatic alveolar echinococcosis

Abstract

Abstract Background: The purpose of this study was to explore the relevant risk factors for biliary complications (BC) in patients with end-stage hepatic alveolar echinococcosis (HAE) following ex-vivo liver resection and autotransplantation (ELRA) and to establish and visualize a nomogram based on LASSO logistic regression. Methods: From January 2010 to May 2022, clinical data were collected from patients diagnosed with end-stage hepatic alveolar echinococcosis (HAE) who underwent ELRA treatment at the First Affiliated Hospital of Xinjiang Medical University. LASSO logistic regression analysis was performed using R software version 4.2.2. In the variable selection stage, we applied the lasso regression to select the most valuable variables as candidates for further logistic regression analysis. Afterward, we utilized the variables obtained from the previous stage to construct a binary logistic regression analysis. The aim of this analysis was to elucidate the relationship between perioperative clinical data and postoperative biliary complications. Bootstrapping resampling was used to validate the resulting nomogram, and decision curve analysis (DCA) was performed to evaluate the clinical effectiveness of the model. Results:The study included a total of 118 patients with end-stage HAE after screening. Among them, 48 patients (40.68%) belonged to the biliary complication group and 70 patients (59.32%) belonged to the non-biliary complication group. Through the above analysis, it was found that age (OR=1.04, 95%CI, 1.0-1.09), biliary dilatation (OR=6.71, 95%CI, 2.62-17.19), and portal hypertension (OR=3.62, 95%CI, 1.22-10.76) were independent risk factors for BC after ELRA, while regular liver resection (OR=0.13, 95%CI, 0.02-0.75) was a protective factor. The area under the receiver operating characteristic curve (AUROC) was 0.8188 (95%CI: 0.7417-0.8958) and the calibration curve for this prediction model were satisfactory. Decision Curve Analysis (DCA) showed that within the 9%-85% range, the nomogram had a greater net benefit than uniform intervention for the entire population. Conclusion: Age, biliary dilatation and portal hypertension are independent risk factors for biliary complications after ELRA. Postoperative biliary complications can be reduced while preserving liver volume by intraoperative regular liver resection. Appropriate interventions based on the nomogram should be used for patients at high risk of postoperative biliary complications.