Validation and epidemiologic definition of the novel steatotic liver disease nomenclature in a national United States cohort with cirrhosis.

Abstract

Background and aims: Novel steatotic liver disease (SLD) definitions were introduced in 2023. Accurate and meaningful classifications using clinical data are needed to study interventions and outcomes.

Methods: In a national cohort of Veterans with cirrhosis and imaging-confirmed steatosis, seven algorithms differentially emphasizing cardiometabolic risk factors (CMRFs) and alcohol exposure were developed to define alcohol-associated liver disease (ALD), metabolic dysfunction associated SLD (MASLD), and MASLD with increased alcohol intake (MetALD). The primary outcome was classification of SLD, which was validated using hospitalizations for major acute cardiac events (MACE) and alcohol use disorder (AUD). Secondary outcomes included longitudinal CTP class, incident hepatocellular carcinoma, and all-cause mortality.

Results: In all, 31,514 patients with cirrhosis (median age 64 years) were included. CMRFs (98.8% ≥1) and hazardous alcohol use (65.3%) were highly prevalent. The distributions of MASLD, MetALD, and ALD varied substantially across classification methods with varying CMRF and alcohol criteria. For example, MetALD ranged from 4.7% to 47.2%. Using method 4, incidence rates of MACE hospitalizations in MASLD, MetALD, and ALD were 16.7, 14.5, and 8.4 per 100 person-years, respectively and incidence rates of AUD hospitalizations were 2.0, 26.1, and 47.6 per 100 person-years, respectively. Hypertriglyceridemia and low HDL were common across SLD subtypes, including ALD (67.3% hypertriglyceridemia, 48.2% low HDL). Patients with ALD (HR 1.41, 95% CI 1.34-1.48) had significantly higher hazards of mortality relative to MASLD.

Conclusion: Classification of SLD is highly sensitive to relative weighting of CMRFs and alcohol use. Clinically relevant definitions should consider data availability on alcohol and the limitations of lipid measurements in distinguishing SLD subtypes.