135: Population Kinetic-Pharmacodynamic Modeling of Multiple Lipid Biomarker Responses to Zodasiran in Patients With Mixed Dyslipidemia

Statement of Purpose, Innovation or Hypothesis: Zodasiran is a small interfering RNA (siRNA)-based therapeutic targeting hepatic angiopoietin-like protein 3 (ANGPTL3) in development for cardiometabolic indications. Subcutaneously (SC) administered zodasiran undergoes plasma elimination within days,¹ but the beneficial pharmacodynamic (PD) effects on lipid biomarkers, including reductions in triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), non-high-density lipoprotein cholesterol (non-HDL-C), and apolipoprotein B (ApoB), persist for months after one dose. Here, we develop a population kinetic-pharmacodynamic (K-PD) model to describe the dose-dependent PD responses to zodasiran in mixed dyslipidemia (MD). Model-based simulations are used to inform Phase 3 dose regimen selection in the MD population.

Description of Methods & Materials: Dosing records, lipid measurements, and covariate data from a Phase 2b study in patients (N=204) with MD (NCT04832971) administered either 50, 100, or 200 mg SC zodasiran every 12 weeks for up to 32 months were used for model development. A nonlinear mixed-effects modeling (NONMEM v7.5.1) approach was taken. Model selection was based on the zodasiran mechanism of action and guided by Objective Function Value (OFV), model stability, and evaluation of diagnostic plots. Covariate inclusion was based on the statistical strength of covariate-parameter relationships and resulting improvement in OFV. ² Parameter uncertainty was explored with sampling-importance-resampling and the predictive ability of the model was verified via visual predictive check (VPC). Simulations of zodasiran dose regimens were conducted in R (v4.3.2) with the mrgsolve package (v1.3.0).

Data & Results: Zodasiran pharmacokinetics (PK) in the virtual effect compartment, or biophase, were represented by a one-compartment model with constant first-order clearance. The silencing of ANGPTL3 mRNA expression by zodasiran was described by a sigmoidal I_max inhibitory function, and cascading turnover models captured the downstream reductions of TG, LDL-C, and non-HDL-C. A reduction of serum ANGPTL3 eases inhibition of TG elimination (I_max) and attenuates synthesis of LDL-C and non-HDL-C via sigmoid maximum effect (E_max) functions. ApoB, which is co-transported with non-HDL-C, was modeled via a direct-effect proportional relationship to circulating non-HDL-C. The estimated biophase half-life of zodasiran was 61 days, providing insight into the long-lasting PD effects despite the rapid plasma clearance. Body weight influenced the zodasiran dose needed to achieve a 50% reduction of serum ANGPTL3 protein. Model diagnostic plots and VPCs indicated the K-PD model adequately describes the dynamics of ANGPTL3, TG, LDL-C, non-HDL-C, and ApoB. Simulations from posterior Bayes estimates revealed those patients with greater levels of baseline TG and/or non-HDL-C experience larger percentage reductions from baseline of key atherogenic lipid biomarkers. Monte-Carlo simulations of zodasiran administered every 3 months (Q3M) project the median time-averaged decreases of LDL-C, non-HDL-C, and ApoB over the steady-state dosing interval to be -9.9%, -24.1%, and -14.7%, respectively, at 50 mg, -12.9%, -28.8%, and -17.8%, respectively, at 100 mg, and -15.2%, -31.5%, and -19.5%, respectively, at 200 mg.

Interpretation, Conclusion or Significance: We present a K-PD model which adequately characterizes lipid biomarker dynamics after multiple doses of zodasiran in MD patients. Simulations support a dosing regimen of 200 mg zodasiran Q3M in a Phase 3 trial for prevention of atherosclerotic cardiovascular disease (ASCVD). We identify intrinsic patient factors, including hyperlipidemia severity, that may associate with treatment benefit.

Disclosures: This work was supported by Arrowhead Pharmaceuticals. All authors were employees of Arrowhead Pharmaceuticals at the time of the analysis and may own stock and/or stock options in the company.

Citations/References: ¹Watts et al. RNA interference targeting ANGPTL3 for triglycerides and cholesterol lowering: phase 1 basket trial cohorts. Nature Medicine. (2023) ²Ayral et al. A novel method based on unbiased correlations tests for covariate selection in nonlinear mixed effects models: The COSSAC approach. CPT: Pharmacometrics and Systems Pharmacology. (2021)

Additional Information/Authors: N/A