Characterization of six clinical drugs and dietary intervention in the non-obese CDAA-HFD mouse model of MASH and progressive fibrosis

The choline-deficient L-amino acid defined-high fat diet (CDAA-HFD) mouse model is widely used in preclinical metabolic dysfunction-associated steatohepatitis (MASH) research. To validate the CDAA-HFD mouse, we evaluated disease progression and responsiveness to dietary and pharmacological interventions with semaglutide, lanifibranor, elafibranor, obeticholic acid (OCA), firsocostat and resmetirom.Disease phenotyping was performed in C57BL/6J mice fed CDAA-HFD for 3-20 weeks and ranked using the MASLD Human Proximity Score (MHPS). Semaglutide, lanifibranor, elafibranor, OCA, firsocostat or resmetirom were profiled as treatment intervention for 8 weeks, starting after 6 weeks of CDAA-HFD feeding. Semaglutide and lanifibranor were further evaluated as early (preventive) therapy for 9 weeks, starting 3 weeks after CDAA-HFD diet feeding. Additionally, benefits of dietary intervention (chow reversal) for 8 weeks were characterized following 6 weeks of CDAA-HFD feeding. CDAA-HFD mice demonstrated a non-obese phenotype with fast onset and progression of MASH and fibrosis, high similarity to human MASH-fibrosis, and tumor development after 20 weeks of diet-induction. Semaglutide and lanifibranor partially reversed fibrosis when administered as prevention, but not as treatment intervention. Elafibranor was the only interventional drug to improve fibrosis. In comparison, chow-reversal resulted in complete steatosis regression with improved liver inflammation and fibrosis in CDAA-HFD mice. CDAA-HFD mice recapitulate histological hallmarks of advanced MASH with progressive severe fibrosis, however, in the absence of a clinical translational obese dysmetabolic phenotype. CDAA-HFD mice are suitable for profiling drug candidates directly targeting hepatic lipid metabolism, inflammation, and fibrosis. The timing of pharmacological intervention is critical for determining antifibrotic drug efficacy in the model.