Leukocyte extracellular vesicle concentration is inversely associated with liver fibrosis severity in NAFLD.

Abstract

The enhanced liver fibrosis (LFS) score and the nonalcoholic fatty liver disease (NAFLD) fibrosis score (NFS) are algorithmic-derived scores for diagnosing severe (F3/F4) liver fibrosis. In a pilot, substudy of the Wessex Evaluation of fatty Liver and Cardiovascular markers in NAFLD with OMacor thErapy (WELCOME) trial, we tested whether measurements of plasma platelet-, endothelial-, and leukocyte-derived extracellular vesicles (EVs) counts are (a) associated with, and predict, F3/F4 fibrosis and (b) able to improve risk prediction of F3/F4 fibrosis in NAFLD, building upon LFS or NFS algorithms. Twenty-six individuals with NAFLD had liver fibrosis severity determined by Kleiner scoring after liver biopsy. Plasma samples stained with CD41a, CD42b, CD31, CD105, CD14, CD16, and CD284 antibodies were analyzed using flow cytometry to measure platelet-, endothelial-, and leukocyte-derived EVs counts. The independence of associations between EVs and F3/F4 fibrosis were tested using logistic regression. Receiver operator characteristic (ROC) curves were used to evaluate F3/F4 fibrosis prediction models. LFS was more strongly associated with F3/F4 fibrosis than NFS (χ2=15.403, P<0.0001, and χ2=6.300, P=0.012, respectively). The association between LFS and F3/F4 fibrosis was further improved by addition of CD14+ EVs (χ2=20.847,P=0.016 vs. χ2=12.803,P=0.015, respectively) or CD16+ EVs (χ2=22.205,P=0.009 vs. χ2=17.559,P=0.001, respectively), and the area under the ROC for LFS (AUC=0.915, se=0.055, P=0.001) was increased by the addition of CD14+ or CD16+ EVs (AUC=0.948, se=0.042, and P<0.001 and AUC=0.967, se=0.055, P<0.001, respectively) as predictor variables. In this small preliminary study, CD14+ and CD16+ EV counts show potential to predict liver fibrosis severity with either marker improving the ability of the LFS to identify F3/F4 fibrosis in this small preliminary cohort study.