ABSTRACT
BACKGROUND: Ursodeoxycholic acid (UDCA) is the standard treatment for primary biliary cholangitis (PBC), but a significant proportion of patients do not respond adequately, leading to increased risk of adverse outcomes. This study aims to develop a new and straightforward predictive score to identify PBC patients likely to achieve a complete response to UDCA. METHODS: A logistic regression analysis was conducted using a derivation cohort of PBC patients to identify pre-treatment variables associated with response to UDCA. This analysis led to the development of the ALP-A score, calculated as: Age at diagnosis divided by (alkaline phosphatase at diagnosis/upper limit of normal). ALP-A score accuracy was evaluated using the area under the ROC curve, validated with a large external cohort from Brazil. Additionally, the correlation between the ALP-A score and the previously validated UDCA response score (URS) was assessed. RESULTS: ALP-A score had good predictive power for adequate (AUC 0.794; 95% CI, 0.737-0.852) and deep (0.76; 95% CI, 0.69-0.83) UDCA response at 1 year of treatment. A cutoff score of 17 and 23 points was determined to be the optimal threshold for distinguishing adequate and deep responders, respectively, from non-responders. ALP-A score demonstrated a sensitivity of 73%, specificity of 71%, positive predictive value of 65%, negative predictive value of 78%, and overall accuracy of 72% for biochemical response. The URS displayed similar discriminative ability (AUC 0.798; 95% CI, 0.741-0.855). CONCLUSION: ALP-A score performs comparably to URS but offers the great advantage of simplicity for routine clinical use. It serves as a valuable tool to identify PBC patients less likely to respond to UDCA treatment, facilitating early consideration of alternative therapeutic approaches.
