Cross-Sectional and Longitudinal Performance of Non-Invasive Tests of Liver Fibrosis in Patients with Non-Alcoholic Fatty Liver Disease

Abstract

Background and aims: Non-invasive tests (NITs) are needed in clinical practice to replace histology for the identification of liver fibrosis and prognostication in Non-Alcoholic Fatty Liver Disease (NAFLD). Novel collagen-derived fibrogenesis markers including N-terminal type III collagen pro-peptide (PRO-C3) are among the most promising tools in this field. The aim of this study was to assess the diagnostic accuracy of PRO-C3, the derivative ADAPT score, and other NITs for the identification of advanced fibrosis (stages 3–4) and changes over 12 months of follow-up. Methods: In this longitudinal study, 96 patients with biopsy-proven NAFLD were evaluated at baseline, of which 50 underwent a follow-up visit after 12 months. Clinical-biochemical parameters, liver stiffness (LS) by transient elastography, PRO-C3, and other NITs (ADAPT, FIB-4, NFS, APRI) were collected at baseline and follow-up. Results: LS showed the best accuracy for the identification of advanced fibrosis, with Area under the Receiving Operator Curve (AUROC) 0.82 (0.73–0.89) for a cut-off value of 9.4 kPa. Among the other NITs, the ADAPT score showed the best accuracy, with AUROC 0.80 (0.71–0.88) for a cut-off of 5.02 (Se 62%, Sp 89%, PPV 74%, NPV 83%). The comparison between the AUROC of LS with that of ADAPT was not statistically different (DeLong test p value 0.348). At follow-up, LS was slightly reduced, whilst PRO-C3 displayed a significant increase from baseline median 11.2 ng/mL to 13.9 ng/mL at follow-up (p = 0.017). Accordingly, ADAPT score increased from median 5.3 to 6.1 (p = 0.019). The other NITs did not significantly change over 12 months. Conclusions: The ADAPT score shows the best performance among non-invasive scores for the identification of advanced fibrosis, not different from LS. Collagen-derived biomarker PRO-C3 and the derivative score ADAPT display significant changes over time, and may be useful tools for monitoring the progression of liver disease or assessing responses to treatments.