Maja Skataric 1, Susan Zhou 1, David Jaworowicz 2, Mitali Gaurav 2, Ramin Mehrani 1, Douglas G. Johns 1, Puja Banka 1, Islam R. Younis 1, Prajakti A. Kothare 1, Ferdous Gheyas 1
1 Merck & Co., Inc. (Rahway, United States), 2 Simulations Plus, Inc. (Research Triangle Park, United States)
Introduction: Despite the well-established cardiovascular benefit of LDL-cholesterol (LDL-C) lowering, real-world use of PCSK9-targeted therapies remains limited, in part due to reliance on injectable formulations. Enlicitide is a novel macrocyclic peptide PCSK9 inhibitor being developed as a once-daily oral LDL-C-lowering therapy to enable robust LDL-C reduction and reduction of cardiovascular risk, offering potential for improved accessibility and adherence [1]. Characterization of pharmacokinetics and exposure–response relationships is critical to support dose recommendation and consistent treatment effects across patient subgroups.
Objectives:
• Characterize the population pharmacokinetics (popPK) of enlicitide across Phase 1, Phase 2 and Phase 3 studies, and evaluate intrinsic and extrinsic factor effects.
• Characterize the relationship between enlicitide exposure and LDL-C reduction in participants with hypercholesterolemia, assess the impact of intrinsic and extrinsic factors on the exposure–response (ER) relationship, assess whether exposure differences translate into clinically meaningful effect on LDL-C reduction and whether the same dose can be recommended across all patient subgroups.
Methods: The popPK dataset pooled enlicitide plasma concentration data from 11 clinical studies, comprising 19,158 observations from 2713 participants, including healthy volunteers and hypercholesterolemia patients (including heterozygous familial hypercholesterolemia (HeFH)). The dataset integrated rich pharmacokinetic sampling from eight Phase 1 studies and sparse sampling from one Phase 2 and two Phase 3 trials, enabling robust characterization of enlicitide PK across a broad dose range (6-300 mg). The ER dataset pooled LDL-C data from two Phase 3 studies, comprising 15338 observations from 3183 participants, with supportive Phase 2b data (1120 LDL-C observations from 379 participants) informing the exposure-response relationship at lower exposure levels.
PopPK analysis evaluated one- and two-compartment models with first-order absorption and both linear and concentration-dependent elimination to account for PCSK9-mediated clearance [2].
Individual steady-state exposures (AUCss) were simulated from the final popPK model and used to inform exposure–response analyses. LDL-C reduction over time was parameterized in terms of a model-estimated baseline LDL-C, maximum pharmacological effect (Emax), AUCss corresponding to the half-maximal reduction in LDL-C (AUC50), and the Hill coefficient, with enlicitide AUCss as a predictor of percentage reduction in LDL-C from baseline. A time-dependent change in LDL-C values in both placebo and active-treated participants was added to the model, accounting for drug-free disease progression.
Candidate covariates explored in popPK and ER models were evaluated using a stepwise procedure, and included body weight, estimated glomerular filtration rate (eGFR), concomitant statin use, disease status, demographics, and diabetes mellitus status.
Model adequacy was assessed using goodness-of-fit diagnostics, visual predictive checks, parameter precision, and simulation-based evaluations.
Results: Enlicitide pharmacokinetics was best described by a two-compartment model with zeroeth-order infusion into the depot compartment and first-order absorption and concentration-dependent clearance characterized by a Hill function. Statistically significant covariate effects included body weight and eGFR on maximal apparent clearance (CLmax/F), body weight on apparent central volume of distribution Vc/F, and categorical covariate effects for statin use and disease status on CLmax/F. Diagnostic plots demonstrated good agreement across studies, dose levels, and populations.
Results from final popPK model demonstrated that inter-individual variability in exposure was moderate (<30% CV) in the Phase 3 population. Relative to non-HeFH, HeFH participants demonstrated slightly higher exposure (~12%). Concomitant statin use was associated with a small exposure increase (~8%). The magnitude of covariate effects of body weight and eGFR on exposure was also small, and not clinically meaningful.
LDL-C reduction increased with exposure in a saturable manner and was well described by a Hill function. Although several covariates in the ER model were statistically significant, their effects on maximal LDL-C reduction were small (<5–7%) and not clinically meaningful.
The 20-mg once-daily dose resulted in exposures associated with near maximal effect, with minimal incremental benefit predicted at higher exposures.
Conclusions: An integrated modeling framework translated pharmacokinetic variability into clinical response and demonstrated that therapeutic exposures are near maximal efficacy. These results support a 20-mg once-daily oral dosing of enlicitide without adjustment for specific patient subgroups and highlight the value of end-to-end model-informed drug development.
References:
[1] Navar, A.M., et al. A placebo-controlled trial of the oral PCSK9 inhibitor enlicitide. New England Journal of Medicine (2026) 394:529-539.
[2] Gibiansky, L., et al. Approximations of the target-mediated drug disposition model and identifiability of model parameters. J Pharmacokinet Pharmacodyn (2008) 35(5): 573–591.
Reference: PAGE 34 (2026) Abstr 12313 [www.page-meeting.org/?abstract=12313]
Poster: Drug/Disease Modelling - Other Topics