Accuracy of Liver Stiffness Measurement in assessing liver fibrosis in naïve patients with Primary Biliary Cholangitis

Background and Aims: Non-invasive evaluation of liver fibrosis in primary biliary cholangitis (PBC) with Liver Stiffness Measurements (LSM) by Transient Elastography (TE) is routinely undertaken at diagnosis for disease staging and risk stratification. However, evidence on correlation between LSM and liver fibrosis is based on cross-sectional studies in PBC-treated patients. Moreover, the impact of potential confounders, e.g. cholestasis and steatosis, is unclear. Aim of our study was to investigate the accuracy of LSM to predict moderate to severe fibrosis in newly diagnosed PBC patients naïve to therapy. Method: We collected data from 182 adult patients who underwent liver biopsy (LB) for PBC at diagnosis in five Italian liver centers from Jan 2006 through Aug 2019. TE examinations within 3 months from LB were included. LB were scored centrally by two expert pathologists, blinded to clinical data and disease stage, according to Ludwig staging system. In all patients Fibrosis-4 (FIB-4) and aspartate aminotransferase [AST]/platelet ratio (APRI) score have been calculated. Diagnostic accuracy of LSM, FIB-4 and APRI score was estimated using the area under the receiver operating characteristic curves (AUROCs) for fibrosis. The effects of liver biochemistry and histological parameters were appraised using multivariable logistic model. Results: 123 PBC patients had adequate LB (≥9 portal tracts) and valid LSM values. According to histological assessment, Ludwig stage distribution was as follows: stage I = 38 (30.9%), stage II = 46(37.4%), stage III = 27 (21.9%), stage IV = 12 (9.8%). TE identified patients with Ludwig stage III/IV with an AUROC of 0.83 (95% confidence interval [CI] (0.74, 0.90)) (Fig.1). The optimal threshold was identified at 6.75kPa (CI (6.55, 7.50)), with 85% of sensitivity, 75% of specificity, 78% of accuracy and 6 false negative. TE was superior to APRI and FIB-4 having AUROC of 0.638 (CI = (0.535, 0.740)) and 0.641 (CI = (0.516, 0.766)), respectively. At multivariable analysis only LSM was associated with liver fibrosis and no significant effect of biochemical measures and steatosis was found. Conclusion: LSM is accurate to predict moderate to severe liver fibrosis at diagnosis in PBC naïve patients using a threshold of 6.75kPa with an AUROC of 0.83. TE outperformed FIB-4 and APRI score and is not confounded by liver biochemistry and steatosis. These data can inform strategies for patient surveillance and trial design in PBC.