Periductal iron-corrected T1 is a predictor of adverse outcomes in large-duct primary sclerosing cholangitis

Concentric periductal fibrosis, commonly referred to as “onion skin” fibrosis, is a histological hallmark of primary sclerosing cholangitis (PSC) and associated with the development of biliary-type cirrhosis and end-stage liver disease. Our aim was to investigate whether non-contrast enhanced quantitative MRI can provide a noninvasive biomarker of periductal fibro-inflammation in large-duct PSC. This prospective cohort study included adults with diagnosis of large-duct PSC, at baseline and 1-year follow-up, and healthy volunteers (HV) at baseline. MRCP + and LiverMultiScan (both Perspectum, UK) were used to align 3D biliary tree models with axial iron-corrected T1 (cT1) maps to quantify cT1 around the bile ducts (periductal-cT1; Pd-cT1). Regions of interest (ROIs) were defined as rings of tissue at increasing distance from the bile duct (1.5–3.5 mm (1), 3.5–5.5 mm (2), 5.5–7.5 mm (3) or 7.5–9.5 mm (4), whole liver (5)). Pd-cT1 was compared between high-risk or low-risk groups using existing markers of risk in PSC and Cox proportional hazard model was used for prediction of outcomes by elevated (> 800ms) Pd-cT1. From 80 recruited participants with PSC, n = 72 had available Pd-cT1 at baseline (median age 44 years; 65% male) and n = 20 HV were included (median age 35 years; 65% male). In PSC, Pd-cT1 in ROIs 1 and 2 was higher than ROIs 3–4 and whole liver cT1 (p < 0.0001) at baseline and follow-up, while there were no differences between ROIs in HV. Participants with high risk stratified by liver stiffness (LS) > 9.6 kPa, enhanced liver fibrosis (ELF) test > 9.8 and Amsterdam-Oxford Model (AOM) > 2 had significantly higher Pd-cT1 than those with low risk classification. In a median follow-up of 49 months (range: 14–66), elevated baseline Pd-cT1 could predict adverse events (cholangitis, hepatic decompensation, liver transplantation and all-cause mortality) in patients with large-duct PSC. Pd-cT1, a novel quantitative MRI method combining biliary and parenchymal technologies, detected periductal liver abnormalities in PSC and was associated with increased risk of clinical outcomes. This work proposes a non-invasive, objective biomarker of fibro-inflammation in the periductal parenchyma in PSC which, once validated, may have utility as a monitoring biomarker in clinical trials for large-duct PSC.