Quantitative magnetic resonance cholangiopancreatography metrics improve prognostication in primary sclerosing
Tim E Middelburg1, Laura Cristoferi2, Willemijn Ponsioen1
1Department of Gastroenterology & Hepatology, Amsterdam UMC, Amsterdam, The Netherlands; Amsterdam Gastroenterology, Endocrinology & Metabolism, Amsterdam, The Netherlands.
Background & Aims:
Quantitative magnetic resonance cholangiopancreatography (MRCP) metrics showed prognostic potential in primary sclerosing cholangitis (PSC). This international multicentre study aimed to provide the first reference range for quantitative metrics in PSC and assess the additional role of quantitative metrics in prognostication.
Methods:
Quantified MRCP (MRCP+) and clinical data were collected from six European tertiary referral and transplantation centres. A healthy cohort was available for comparison. Least Absolute Shrinkage and Selection Operator analysis was used to derive variables for a prognostic model in a derivation cohort, referred to as the quantitative MRCP augmented Amsterdam-Oxford Model (qmAOM). Composite endpoint included PSC-related mortality, transplantation, or hepatic decompensation, without hepatic decompensation as secondary endpoint. Performance was assessed in the validation cohort, expressed as C-statistic (95% CI), and compared with the Amsterdam-Oxford Model (AOM), Mayo Risk Score and MRCP+, bilirubin, and aspartate aminotransferase (M+BA) scores.
Results:
In total, 457 participants with PSC were included for the reference range, which differed significantly in comparison with the healthy controls. Prognostic analysis included 320 participants. qmAOM included: age at PSC diagnosis, inflammatory bowel disease presence, serum alkaline phosphatase, aspartate aminotransferase, albumin, bilirubin, platelets, number of strictures, number of ducts with strictures or dilatations and proportion of 3-5-mm ducts. qmAOM, AOM, and M+BA demonstrated 0.83 (0.77-0.93), 0.79 (0.68-0.91), and 0.65 (0.64-0.84), respectively. In secondary endpoint analysis, qmAOM and Mayo Risk Score demonstrated 0.84 (0.76-0.86) and 0.74 (0.68-0.81), respectively. In bootstrap analysis using the entire prognostic cohort, qmAOM performed significantly better than AOM and M+BA (0.82; 0.75; 0.70, respectively; p <0.001).
Conclusions:
This study provides the first reference range of quantitative metrics in PSC and indicates that quantitative biliary metrics improve discriminative performance in event-free survival in PSC.
Impact And Implications:
This study establishes a reference range for quantitative MRCP metrics in primary sclerosing cholangitis and demonstrates their additive value predicting event-free survival. These findings support the incorporation of multidomain approaches in prognostic modelling and potential of quantitative imaging as surrogate endpoint in clinical trials.
