Machine learning models are superior to noninvasive tests in identifying clinically significant stages of NAFLD and NAFLD-related cirrhosis.

BACKGROUND AND AIMS: We assessed the performance of machine learning (ML) models in identifying clinically significant NAFLD-associated liver fibrosis and cirrhosis. APPROACH AND RESULTS: We implemented ML models including logistic regression (LR), random forest (RF), and artificial neural network to predict histological stages of fibrosis using 17 demographic/clinical features in 1370 patients with NAFLD who underwent liver biopsy, FibroScan, and labs within a 6-month period at multiple U.S. centers. Histological stages of fibrosis (≥F2, ≥F3, and F4) were predicted using ML, FibroScan liver stiffness measurements, and Fibrosis-4 index (FIB-4). NASH with significant fibrosis (NAS ≥ 4 + ≥F2) was assessed using ML, FibroScan-AST (FAST) score, FIB-4, and NAFLD fibrosis score (NFS). We used 80% of the cohort to train and 20% to test the ML models. For ≥F2, ≥F3, F4, and NASH + NAS ≥ 4 + ≥F2, all ML models, especially RF, had primarily higher accuracy and AUC compared with FibroScan, FIB-4, FAST, and NFS. AUC for RF versus FibroScan and FIB-4 for ≥F2, ≥F3, and F4 were (0.86 vs. 0.81, 0.78), (0.89 vs. 0.83, 0.82), and (0.89 vs. 0.86, 0.85), respectively. AUC for RF versus FAST, FIB-4, and NFS for NASH + NAS ≥ 4 + ≥F2 were (0.80 vs. 0.77, 0.66, 0.63). For NASH + NAS ≥ 4 + ≥F2, all ML models had lower/similar percentages within the indeterminate zone compared with FIB-4 and NFS. Overall, ML models performed better in sensitivity, specificity, positive predictive value, and negative predictive value compared with traditional noninvasive tests. CONCLUSIONS: ML models performed better overall than FibroScan, FIB-4, FAST, and NFS. ML could be an effective tool for identifying clinically significant liver fibrosis and cirrhosis in patients with NAFLD.

Fatty liver in hepatitis C patients post-sustained virological response with direct-acting antivirals.

AIM: To determine steatosis and fibrosis prevalence in hepatitis C patients after a sustained virological response achieved with direct-acting antivirals. METHODS: Transient elastography with controlled attenuation parameter (CAP) was used to assess hepatic steatosis post-sustained virological response (SVR); the CAP technology was not available in the United States at study initiation. Liver stiffness/fibrosis was measured before and 47 wk after treatment completion. Patients with genotype 3 and patients with cirrhosis were excluded. RESULTS: One hundred and one patients were included in the study. Post-SVR there were decreases from baseline in alanine aminotransferase (ALT) (63.1 to 17.8 U/L), aspartate aminotransferase (51.8 to 21.5 U/L) and fibrosis score (7.4 to 6.1 kPa) (P < 0.05). Post-SVR, 48 patients (47.5%) had steatosis on CAP; of these, 6.25% had advanced fibrosis. Patients with steatosis had higher body mass index (29.0 vs 26.1 kg/m2), glucose (107.8 vs 96.6 mg/dL), ALT (20.4 vs 15.3 mg/dL), CAP score (296.3 vs 212.4 dB/m) and fibrosis score (7.0 vs 5.3 kPa); P < 0.05. Interestingly, compared to baseline, both patients with and without steatosis had change in fibrosis score post-SVR (7.7 kPa vs 7.0 kPa and 7.0 kPa vs 5.3 kPa); alternatively, (P < 0.05) and therefore patients with steatosis continued to have clinically significant stiffness (≥ 7 kPa). CONCLUSION: Fatty liver is very common in hepatitis C virus (HCV) patients post-SVR. These patients continue to have elevated mean fibrosis score (≥ 7 kPa) compared to those without fatty liver; some have advanced fibrosis. Long term follow up is needed to assess steatosis and fibrosis in HCV patients post-SVR.