Validity of routine biochemical and ultrasound scores for prediction of hepatic fibrosis and steatosis in NAFLD

Background We evaluated the validity of some non-invasive scores and ultrasound findings to predict fibrosis and steatosis in a cohort of NAFLD patients who underwent liver biopsy. Ninety-seven NAFLD patients were enrolled and classified into NASH (66) and simple steatosis groups (31) based on liver biopsy. ROC curves were constructed for Fibrosis-4 index (FIB4), aspartate aminotransferase to platelet ratio index (APRI), and NAFLD fibrosis score (NFS) in fibrosis prediction, also for (hepatic steatosis index; HSI, fatty liver index; FLI) and ultrasonographic subcutaneous and visceral adipose tissue measurements (SAT and VAT) for steatosis prediction. Results FIB4 had AUC of 0.6, APRI and NFS at cutoffs of 0.3 and -.2.4 had AUC of 0.64 and 0.63 in detecting the presence of any grade of fibrosis, and of (0.52, 0.55, and 0.58) for significant fibrosis. FIB4 at a cut-off of (0.76) had the highest AUC in detecting any grade of fibrosis in the simple steatosis group (0.81). SAT (at cutoff of 2.1 and 2.5) was superior to VAT. HSI (at cutoff 45.35 and 45.7) was superior to FLI in detecting moderate or marked steatosis. Conclusion FIB4 and NFS can be used in screening for silent liver disease with ongoing fibrosis in simple steatosis. They are unsatisfactory predictors for significant fibrosis in NAFLD. SAT is better than VAT in predicting moderate steatosis and is slightly better than biochemical HSI.