OBJECTIVES: Nonalcoholic steatohepatitis (NASH) is an advanced stage of nonalcoholic fatty liver disease (NAFLD) and a common comorbidity in type 2 diabetes (T2D). Glucagon-like peptide-1 receptor agonists (like semaglutide) are recommended for treating patients with T2D and NASH by the American Association of Clinical Endocrinology. Liver biopsy is recommended for diagnosis of NASH but is invasive. Multiparametric magnetic resonance imaging (mpMRI) is a noninvasive alternative. We aimed to model and evaluate the life-time - liver mpMRI vs biopsy in the diagnosis of NASH among patients with T2D from the perspective of the US health system.

METHODS: We developed a multi-state Markov model for NAFLD progression in T2D comparing mpMRI to liver biopsy for the diagnosis of NASH. We considered three effects for semaglutide: slowing fibrosis progression, regression of fibrosis and reducing mortality from cardiovascular and liver related causes. Model input parameters were obtained from literature. We assumed a hypothetical cohort of 1000 patients with T2D aged 50 years suspected of having NASH in the US who are referred for semaglutide treatment on diagnosis of NASH. The outcomes were costs (in US dollar, year 2021 values) and quality-adjusted life-years (QALYs), discounted at 3% annually over time horizon.

RESULTS: Identifying patients with NASH for treatment with semaglutide using mpMRI results in direct cost savings of $3250 per patient compared with biopsy, and a gain of 0.162 QALYs over the patient lifetime. These findings were robust across deterministic and probabilistic sensitivity analyses.