Prediction of trajectories and outcomes in early-stage metabolic dysfunction-associated steatotic liver disease: a narrative review

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most prevalent chronic liver disorder, with manifestations ranging from steatosis to steatohepatitis, advanced fibrosis, cirrhosis, and hepatocellular carcinoma. At all stages, MASLD is also associated with increased risks of cardiovascular disease, type 2 diabetes, and extrahepatic malignancies. Timely and accurate prediction of disease onset, progression, and complications remains an unmet need. Although hepatic fibrosis is a strong predictor of liver-related and all-cause mortality, it reflects relatively advanced disease. Growing evidence suggests that steatosis may mark early divergence of disease trajectories. Effective MASLD forecasting therefore requires early risk assessment and longitudinal evaluation. Emerging approaches combine genetic risk with routine clinical, behavioural, and social data, allowing machine learning methods to better identify MASLD subtypes and predict individual disease courses. However, cost and logistical barriers limit widespread adoption, and further research is needed to determine whether early forecasting can improve long-term outcomes and healthcare value.