Diagnostic accuracy of noninvasive markers of liver fibrosis in patients with psoriasis taking methotrexate: a systematic review and meta-analysis

Abstract

People with psoriasis taking methotrexate may be at increased risk of developing liver fibrosis compared with the general population. Noninvasive methods of detecting fibrosis have been widely adopted but their clinical utility is uncertain. To evaluate the diagnostic accuracy of noninvasive methods to detect fibrosis compared with liver biopsy (reference standard) in people with psoriasis taking methotrexate. A systematic search using Ovid/Medline, Embase, Cumulative Index to Nursing and Allied Health Literature, the Cochrane Library and Clinical Trials Register was performed. Diagnostic cohorts or case-control studies of adults taking or being considered for methotrexate therapy were considered. Study quality was evaluated using the Quality Assessment tool for Diagnostic Accuracy Studies (QUADAS-2). Pooled data analysis was performed using RevMan 5.1. Bayesian meta-analysis was conducted using Markov chain Monte Carlo simulation. Seventeen studies were included. Sensitivity and specificity were 38% and 83% for standard liver function tests (LFTs), 74% and 77% for procollagen-3 N-terminal peptide (P3NP), 60% and 80% for Fibroscan(®) (Echosens, France, www.echosens.com), and 55% and 49% for ultrasound. Confidence in these results is limited owing to low-quality data; old, small studies displayed significant selection bias and significant variation in the prevalence of fibrosis. No studies were identified evaluating recently developed markers. The clinical utility of LFTs, P3NP and liver ultrasound is poor. Therefore if these tests are used in isolation, a significant proportion of patients with liver fibrosis may remain unidentified. Larger prospective studies are required in this population to validate newer non-invasive methods.