Abstract

Treatment options for nonalcoholic steatohepatitis (NASH) are limited. One approach targets hepatic acetyl-CoA carboxylase (ACC), which influences de novo lipogenesis (DNL) and fatty acid oxidation. An oral, liver-targeted ACC inhibitor, GS-0976, has demonstrated reductions in hepatic fat, liver biochemistry, and markers of fibrosis in pre-clinical and clinical studies. However, increased plasma triglycerides (TG) have been observed in some patients, warranting mechanistic investigation. Quantitative systems pharmacology (QSP) approaches can help evaluate hypotheses by simulating alternative mechanistic pharmacodynamic (PD) effects. The QSP mathematical model, NAFLDsym, simulates inter-patient pathophysiologic variability in pathways that contribute to steatosis and lipotoxicity in NAFLD and NASH patient populations (SimPops). Mechanistically, NAFLDsym v1A includes hepatic steatosis, lipotoxicity, plasma TG, liver injury and regeneration, and biomarkers (e.g., ALT). NAFLDsym v1A was used to predict reductions in liver TG and plasma ALT in addition to plasma TG increases following treatment with the ACC inhibitor GS-0976. Following simulations of previous clinical studies (Stiede 2017 [1], Lawitz 2018 [2]), additional predictions of GS- 0976 treatment (5 or 20 mg q.d.) for 12 weeks (ClinicalTrials.gov NCT02856555 [3]) were performed in the SimPops; the simulations included the following PD mechanisms in various combinations: A) DNL inhibition, B) increased VLDL-TG secretion, C) lipoprotein lipase (LPL) inhibition. Significant reductions in liver fat and plasma ALT were predicted for both GS-0976 doses when PD mechanisms A, B, and C were all employed. Plasma TG were predicted to increase in several simulated patients. These simulation results were comparable with observed clinical patient data from a phase 2 trial (NCT02856555) of GS- 0976 [3]. However, simulation results absent any one or two of the PD mechanisms did not agree with the clinical data, suggesting the clinical response to ACC inhibition requires DNL inhibition, increased VLDL-TG secretion, and LPL inhibition. Simulated patients with exaggerated predicted plasma TG increases included notable LPL inhibition. The NAFLDsym predictions suggest that liver fat and plasma ALT reductions in patients treated with GS-0976 are due to DNL inhibition and increased VLDL-TG secretion, while plasma TG increases are due to increased VLDL-TG secretion and LPL inhibition.

AASLD Liver Meeting, November 9-13, 2018, San Francisco, CA

By Scott Q Siler, Grant T Generaux, Brett A Howell, Adrian Ray, G Mani Subramanian, Robert P Myers, and Paul B Watkins