Adiponectin changes in HCV-Genotype 4: relation to liver histology and response to treatment

Abstract

Summary. Recently, attention has been focussed on adiponectin and its changes in different types of chronic liver disease. Its relation to hepatic fibrosis and insulin resistance in post-hepatitis liver disease is not clear. The aim of this study was to clarify the adiponectin changes in genotype 4 hepatitis C virus (HCV)-infected patient in relation to liver histology and insulin resistance, and its usefulness as a predictor of hepatic fibrosis and response to treatment. Total adiponectin and its high molecular weight (HMW) form as well as insulin levels were studied in 92 chronic HCV, genotype 4 and 66 healthy control volunteers. Neither total adiponectin (r = 0.101, P = 0.220) nor HMW adiponectin (r = 0.081, P = 0.328) correlated with viral load. Total and not HMW adiponectin was significantly correlated with hepatic fibrosis and inflammation (r = 0.267, P = 0.002, r = 0.278, P < 0.001, respectively).In addition, total adiponectin (r = 0.224, P = 0.002) and HMW adiponectin (r = 0.266, P < 0.0006) significantly correlated with insulin resistance. As fibrosis did not correlate with insulin resistance (r = 0.081, P = 0.204), the correlation between total adiponectin and fibrosis was not mediated by insulin resistance. Multivariable regression analysis, (including pretreatment cases and controls) revealed that total adiponectin was significantly associated with gender, being lower among male subjects (X2 = 13.04, P = 0.0001). The multivariable regression model supported the lack of association between insulin resistance and total adiponectin levels (X2 = 1.88, P = 0.171), while non cirrhotics had significantly lower total adiponectin levels than cirrhotics (X2 = 10.90, P = 0.004) and lower level of inflammation significantly lower total adiponectin levels than more severe inflammation (X2 = 8.95, P = 0.003). Total or HMW adiponectin did not yield receiver operating characteristic (ROC) curves with area under the curve (AUC) >75%, thus the cutoff points have poor sensitivity/specificity as predictors of fibrosis. However, as a predictor of end-of-treatment response, the ROC curve of adiponectin index gave yield an AUC = 81.4%. We can conclude that total adiponectin level, in HCV genotype 4 patients, increases with progression of hepatic fibrosis regardless of insulin resistance. Its high molecular form does not have such correlation. The adiponectin changes are not related to viral load, insulin resistance or other demographic data suggesting that this change is histologically related. In spite of this, no adiponectin cutoff level had reasonable sensitivity/specificity for predicting hepatic fibrosis stage, while this may be used as a predictor for antiviral response possibly reflecting improvement in hepatic inflammation post treatment.