Mean Platelet Volume, Red Cell Distribution Width to Platelet Count Index, Globulin Platelet Index, and Twelve Other Simple Non-invasive Fibrosis Scores: How Much Do Routine Blood Tests Tell About Liver Fibrosis?: ACG Governors Award for Excellence in Clinical Research

Abstract

Introduction: There are many simple non-invasive scores to predict liver fibrosis that can be calculated from routine blood parameters. We aim to compare these simple non-invasive scores for evaluation of liver fibrosis with liver biopsy in patients with chronic hepatitis C (HCV). Methods: In 1,602 HCV patients who underwent liver biopsy, we compared the liver biopsy (Scheuer classification) fibrosis scores with APRI (AST/Platelet ratio), Fibrosis-4 (FIB-4), Lok score, GUCI score, Fibro-alpha score, Forns’ score, King score, AAR (AST/ALT ratio), Fibrosis index (FI), Pohl score, Fibro-Q score, FCI (Fibrosis cirrhosis index), and 3 new scores: mean platelet volume (MPV), RPR (red cell distribution width/platelet count ratio), and globulin/platelet count index (GPI). Results: Mean age of patients was 41.8±9.6 years (1,365 males); genotype 4 (65.6%) was the most common, followed by genotype 1 (10.9%). Liver biopsy showed stage 0 fibrosis (F0) in 1.9%, stage 1(F1) in 32.9%, stage 2(F2) in 39.5%, stage 3(F3) in 19%, and stage 4(F4) in 6.6% of patients. Of the baseline parameters, AST (adjusted odds ratio [OR] 1.015; confidence interval [CI] 1.008-1.022; p=0.001), albumin (adjusted OR 0.842; CI 0.742-0.915; p=0.001) and platelet count (adjusted OR 0.981; CI 0.974-0.989; p=0.001) were independent predictors of cirrhosis. We derived a study score (8.5 -0.2 [albumin, g/dl] + 0.01[AST, IU/l] - 0.02 [platelet count, 109/l]) at a cut-off of >4.7; it had high predictive accuracy (AUROC=0.868; CI 0.833-0.904) for cirrhosis. All the scores except AAR and Pohl score showed high predictive accuracy for cirrhosis (Table 1). Of the new scores, MPV and RPR had relatively low, and GPI had high, predictive accuracy for cirrhosis. GUCI score (cut-off >1.1) had the highest predictive accuracy for cirrhosis, followed by APRI (cut-off >1.5). King’s score had highest predictive accuracy for significant (F2-F4) and advanced fibrosis (F3-4), AUROC of 0.771 and 0.860, respectively.Table 1: Performance of Non-invasive Scores Compared to Cirrhosis(F4-Fibrosis) in Liver BiopsyConclusion: Excluding AAR, Pohl score, MPV, and RPR, all the non-invasive scores including the study score we derived, and GPI showed high predictive accuracy for cirrhosis. GUCI score and APRI for cirrhosis and King’s score for significant and advanced fibrosis seem to have the highest predictive accuracy.