P-109 ALKALINE PHOSPHATASE AND CIRRHOSIS AT DIAGNOSIS ARE ASSOCIATED WITH DEEP RESPONSE TO URSODEOXYCHOLIC ACID IN PRIMARY BILIARY CHOLANGITIS

Abstract

Conflict of interest: No Introduction and Objectives: Primary biliary cholangitis is a chronic and progressive autoimmune liver disease, whose prognosis can be improved by normalizing alkaline phosphatase and bilirubin. While ursodeoxycholic acid (UDCA) is first line standard of care, approximately 40% of patients exhibit incomplete response. We aimed to identify prognostic markers for deep response to UDCA therapy at presentation. Patients / Materials and Methods: Data from the Brazilian Cholestasis Study Group cohort were analyzed retrospectively. Patients were assessed for deep response (defined as normalization of alkaline phosphatase and bilirubin) after 1 year of UDCA treatment. With the purpose of selecting the set of relevant variables related to the deep response for a parsimonious multivariate model, we applied the Varrank algorithm. Additionally, the performance of the UDCA response score in predicting deep response was evaluated. Results and Discussion: A total of 297 patients were analyzed, with 57.2% achieving an adequate response according to the Toronto criteria, while 22.9% reached deep response. Cirrhosis (OR 0.460; 95% CI 0.225-0.942; p=0.034) and elevated baseline alkaline phosphatase levels (OR 0.629; 95% CI 0.513-0.770; p<0.001) were associated with reduced odds of deep response. The UDCA response score exhibited moderate discrimination power (AUROC=0.769) but lacked calibration. Conclusions: Baseline ALP, and cirrhosis at diagnosis emerge as the most important prognostic factors to predict normalization of alkaline phosphatase and bilirubin after UDCA. The UDCA response score is inadequate for predicting deep response in the Brazilian PBC population.