Estimation of the Eligible Population For Resmetirom Among Adults in the United States for Treatment of Non-Cirrhotic NASH with Moderate-to-Advanced Liver Fibrosis

Abstract

Introduction

As of March 2024, resmetirom is the first and only therapy approved in the United States (US) for the treatment of adults with non-cirrhotic nonalcoholic steatohepatitis (NASH) with moderate-to-advanced liver fibrosis (MALF) consistent with stages F2/F3 fibrosis. Estimates of the diagnosed, treatment-eligible NASH population are poorly understood due to diagnostic variability. This study provides a contemporary estimate of the size of the US resmetirom treatment-eligible population.

Methods

A dynamic population calculator was developed combining literature, screening guidelines, resmetirom study criteria, and analyses of the NHANES 2017–March 2020 cycle. It computes NASH prevalence, proportion non-cirrhotic NASH with MALF, Year 1 diagnosis, and new diagnoses in Years 2 and 3. NASH prevalence was estimated by applying the American Association of Clinical Endocrinology screening algorithm and recommended NIT cut-offs in the NHANES dataset. The proportion of non-cirrhotic NASH with MALF was informed by analyses of the Forian US integrated medical claims database using NASH and cirrhosis-specific ICD-10-CM codes and FIB-4 scores. NASH diagnosis rates were obtained from published estimates and NHANES responses. Treatment-eligible population growth was projected using published incidence data. Estimates were compared to a NASH budget-impact-analysis (BIA) from the Institute for Clinical and Economic Review (ICER).

Results

In the base case, a NASH prevalence of 4.6% was modeled (range 1.3–14.2%). This value was multiplied by the proportion estimated to have non-cirrhotic MALF (i.e., 35%). Published analyses suggest a diagnosis rate of ~ 10% (range 3.3–14.3%) and ~ 16% year-over-year growth in the treatment-eligible population. Assuming a 1-million commercial-member population, the resmetirom treatment-eligible population was estimated as 1255–1699 in Years 1–3 following approval. Sensitivity analyses were conducted and comparison to the ICER BIA was influenced by different diagnosis rates.

Conclusion

Estimation of the treatment-eligible population for resmetirom depends importantly on NASH diagnosis rates, which are predicted to be < 15% in the 3 years after drug approval.

Plain Language Summary

Nonalcoholic steatohepatitis (NASH) is an advanced form of nonalcoholic fatty liver disease. Previously there were no treatments for NASH in the United States (US), but as of March 2024, the US Food and Drug Administration (FDA) approved resmetirom (REZDIFFRA™), a once-daily, oral therapy, in conjunction with diet and exercise, under accelerated approval for the treatment of adults (aged 18 years or older) with non-cirrhotic NASH with moderate-to-advanced liver fibrosis (MALF), consistent with stages F2–F3. It is not well understood how many diagnosed patients with NASH would be eligible for treatment with resmetirom; thus, this study aimed to estimate the size of the US resmetirom treatment-eligible population. To do so, we created a flexible population calculator that considers how many people have NASH, what proportion would be eligible for resmetirom treatment—i.e., have non-cirrhotic NASH with MALF—and of those how many people would be diagnosed. We used published literature, screening guidelines, resmetirom study criteria, and analyses of national surveys to inform our range of estimates. In the main analysis, we modeled a NASH prevalence of 4.6% (range 1.3–14.2%), which was then limited to the proportion estimated to have non-cirrhotic NASH with MALF (i.e., 35%) and diagnosed (i.e., 10%, range 3.3–14.3%). A year-over-year growth of approximately 16% in the treatment-eligible population was modeled in years following approval. Assuming a population of 1 million commercial insurance enrollees, the resmetirom treatment-eligible population was estimated to be 1255–1699 in Years 1–3 following approval. We assessed alternative scenarios and have compared our results to existing models.