Longitudinal Comparison of Currently Used Risk Scores for Prognostication of Primary Sclerosing Cholangitis (PSC) in...

Background/Objectives: Primary sclerosing cholangitis (PSC) is a chronic cholestatic liver disease with limited epidemiological data from Central-Eastern Europe. This study characterized a Hungarian PSC cohort, comparing patients with and without inflammatory bowel disease (IBD), and longitudinally evaluated the predictive efficacy of established prognostic scores (Mayo Risk Score, Amsterdam-Oxford Model [AOM], UK-PSC short/long). Methods: Data from 135 PSC patients (median diagnosis age 31 years, 57.7% male) were collected yearly at two Hungarian centers, with a median follow-up of 8.8 years. Outcomes included liver transplantation (LT) and liver-related death. Prognostic value of baseline clinical scores was assessed for 2-, 5-, 8-, and 10-year composite outcome. Results: PSC-IBD patients (54.1%) were younger with higher baseline Mayo and AOM scores, and had increased rates of colorectal carcinoma (8.22% vs. 0.00%) and liver transplantation (26.03% vs. 9.68%) within 10 years than PSC-only patients. There were no differences in liver-related mortality or composite outcomes between the groups. All prognostic scores showed good short-term predictive ability for poor outcomes (AUROC at 2 years: 0.858-0.958), which diminished over time (AUROC at 10 years: 0.708-0.756). The AOM demonstrated the most consistent performance. Persistent alkaline phosphatase (ALP) elevation (≥2.2×ULN) 2 years post-diagnosis, despite ursodeoxycholic acid therapy, strongly predicted 10-year adverse outcomes (HR: 3.927, p < 0.001), outperforming formal scoring systems (HR: 2.688-1.522). Conclusions: While PSC-IBD patients had more CRC and liver transplantation, overall transplantation-free survival was similar to PSC-only patients. Prognostic utility of current scores declines with longer follow-up; AOM was most stable. Sustained ALP elevation is a robust long-term prognostic indicator.