Performance of the Mayo Risk Score in Predicting Transplant and Mortality in a Single-Center U.S. Cohort of Primary Sclerosing Cholangitis.

Abstract

Background: The Mayo Risk Score (MRS) predicts short-term mortality in primary sclerosing cholangitis (PSC) using the age, bilirubin, albumin, aspartate aminotransferase (AST), and variceal bleeding history. While the MRS has been validated in end-stage PSC, its ability to predict liver transplantation (LT) and outcomes in newly diagnosed patients without advanced disease remains unclear. This study evaluated the effectiveness of the MRS in predicting LT and mortality in this patient population. Methods: We analyzed data from 109 adults with PSC enrolled in a prospective registry (2018-2024) with ≥4 years of follow-up. Logistic regression identified the predictors of LT or death, and the model performance was assessed using the area under the receiver operating characteristic curve (AUROC). Multicollinearity was evaluated using the variance inflation factor (VIF). Results: Among the 109 patients (mean age 45 ± 15 years, 51% female), 85% remained alive without LT, 12% underwent LT, and 3% died over a median follow-up of 4.63 years. The MRS was significantly associated with LT or death (OR 3.08, p < 0.001) and demonstrated excellent predictive performance (AUROC 0.99, p < 0.001). The model achieved 95.45% sensitivity, 98.85% specificity, and a correct classification rate of 98.17%, supporting its clinical utility. Conclusion: The MRS is a robust tool for risk stratification in PSC, predicting LT and mortality. These findings highlight its broader applicability beyond end-stage PSC and underscore its potential for guiding clinical management and early intervention strategies.