Serum levels of alpha-fetoprotein (AFP) were reported to increase in patients with significant or advanced hepatic fibrosis. Combination of non-invasive tests decreases the use of liver biopsy in large proportion of chronic HCV patients. The aim of the study was to compare and combine AFP with commonly used non-invasive fibrosis tests in novel scores for prediction of different stages of hepatic fibrosis. Six hundred and fifty two treatment naïve chronic hepatitis C patients were enrolled. Demographic data, basic pre-treatment laboratory tests including complete blood count (CBC), liver biochemical profile and renal functions test, international normalized ratio (INR) in addition to AFP, liver stiffness measurement (LSM) by Fibroscan and liver biopsies were retrospectively analyzed. AST to Platelet Ratio Index (APRI) and FIB-4 scores were calculated. Different predictive models using multivariate logistic regression analysis were generated and presented in equations (scores) composed of a combination of AFP, LSM plus FIB-4/APRI scores. AFP was correlating significantly with LSM, FIB-4, and APRI scores. Areas under receiver operating characteristic curves (AUROCs) for predicting significant hepatic fibrosis, advanced hepatic fibrosis, and cirrhosis were 0.897, 0.931, and 0.955, respectively, for equations (scores) containing AFP, LSM, and FIB-4. AUROCs for predicting significant hepatic fibrosis, advanced hepatic fibrosis and cirrhosis were 0.897, 0.929, and 0.959, respectively, for equations (scores) containing AFP, LSM, and APRI. The study shows that combining AFP to serum biomarkers and LSM increases their diagnostic performance for prediction of different stages of liver fibrosis.
Keywords: AFP; HCV; fibrosis; novel scores.
© 2018 Wiley Periodicals, Inc.