Salma Bahnasawy 1, Johanna Melin 1, Jitendar Reddy 2, Niklas Bergh 3, Alexis Hofherr 3, Helena Edlund 1
1 Clinical Pharmacology and Quantitative Pharmacology, Clinical Pharmacology & Safety Sciences, R&D, AstraZeneca (Gothenburg, Sweden), 2 Clinical Pharmacology and Quantitative Pharmacology, Clinical Pharmacology & Safety Sciences, R&D, AstraZeneca (Cambridge, UK ), 3 Translational Sciences and Clinical Development. Cardiovascular, Renal and Metabolism, BioPharmaceuticals R&D, AstraZeneca (Gothenburg, Sweden)
Background
AZD1705 is a N-acetylgalactosamine (GalNAc)–conjugated small interfering RNA (siRNA) targeting angiopoietin-like 3 (ANGPTL3), a regulator of lipoprotein metabolism [1]. While dose selection for siRNAs is typically guided by pharmacodynamic (PD) biomarkers due to temporal dissociation between PD and plasma pharmacokinetics (PK), plasma PK remains important for safety margins and assessments of non-linearity [2]. A review of nine GalNAc-siRNAs suggests that this drug class broadly exhibits similar, and dose-proportional, plasma PK [3]. A recent report, however, also describes non-linear behaviour within this class [4]. The aim of this analysis was to develop a population PK model of AZD1705 in participants with dyslipidaemia and explore any non-linearities as well as covariates influencing plasma exposure.
Methods
Data were from a Phase I trial in participants with dyslipidaemia, with or without T2D, (NCT06238466) comprising single and multiple ascending subcutaneous doses (15–600 mg AZD1705). Plasma sampling was rich through 48 hours post dose, then sparser in the terminal elimination phase (72 hours to 16 weeks). Non-compartmental analyses indicated more than dose-proportional increases in maximum concentration (Cmax) and area-under the curve for the first 48 h (AUC0-48h) at doses >150 mg. Different absorption (first-order vs sequential zero-/first-order) and disposition model structures (two- vs three-compartment) were explored. Candidate models for non-linearity included concentration-dependent saturation on clearance, intercompartmental clearance, or bioavailability. Given no intravenous data and prior evidence of rapid absorption [3], constraints were imposed to mitigate flip-flop kinetics (i.e., enforcing ka > ke). Empirical Bayes estimate (EBE) plots were evaluated to explore potential covariates. The modelling was conducted in NONMEM 7.6.
Results
The analysis included 1,568 plasma concentrations from 74 participants. AZD1705 plasma PK were best described by a three-compartment disposition model with first-order absorption and linear elimination. The observed supra-proportional exposure at higher doses was most parsimoniously captured by a saturable intercompartmental distribution process, implemented via a Michaelis–Menten function on intercompartmental clearance (Q). This aligns with reported saturable asialoglycoprotein receptor (ASGPR)–mediated hepatic uptake for GalNAc-siRNAs [4,5]. Parameter estimation was numerically stable and physiologically plausible under the ka > ke constraint, yielding typical values of ka ≈ 0.29 h⁻¹ and ke ≈ 0.11 h⁻¹. Allometric scaling (fixed components) with baseline body weight significantly improved fit (ΔOFV = −136), whereas no additional clinically meaningful covariates were identified. Goodness-of-fit diagnostics and predictive checks were acceptable overall, with a residual tendency to underpredict Cmax at the highest doses.
Conclusions
This population PK model characterises AZD1705 plasma PK across a wide dose range and quantifies the supra-proportional increases in early exposure (Cmax, AUC0–48h) at the highest doses. Modelling a saturable intercompartmental distribution provides a semi-mechanistic surrogate for potential saturation of ASGPR-mediated hepatic uptake, consistent with slower hepatic uptake at high plasma concentrations. This offers a pragmatic way to characterise target-mediated drug disposition when ASGPR receptor occupancy and liver concentrations are not measured in the clinical study. The limited reporting of non-linearity in the class may stem from practical constraints (e.g., limited sampling beyond 24–48 hours) and evaluation over narrower dose ranges. The developed model can be extended to assess any impact of high-dose non-linearity on ANGPTL3 reduction.
References:
[1] Adam RC, Mintah IJ, Alexa-Braun CA, et al. Angiopoietin-like protein 3 governs LDL-cholesterol levels through endothelial lipase-dependent VLDL clearance. Journal of Lipid Research 2020;61:1271–86.
[2] Boianelli A, Aoki Y, Ivanov M, et al. Cross-Species Translation of Biophase Half-Life and Potency of GalNAc-Conjugated siRNAs. Nucleic Acid Ther 2022;32:507–12.
[3] Sten S, Cardilin T, Antonsson M, et al. Plasma Pharmacokinetics of N-Acetylgalactosamine-Conjugated Small-Interfering Ribonucleic Acids (GalNAc-Conjugated siRNAs). Clin Pharmacokinet 2023;62:1661–72.
[4] Ogawa T, Goeyvaerts N, Kakuda TN, et al. Population Pharmacokinetics of siRNA JNJ-73763989 in Healthy Participants and Patients With Chronic Hepatitis B. Clinical Pharmacology & Therapeutics 2025;118:1451–62.
[5] Amaeze O, Isoherranen N, Shum S. The absorption, distribution, metabolism and elimination characteristics of small interfering RNA therapeutics and the opportunity to predict disposition in pregnant women. Drug Metabolism and Disposition 2025;53.
Reference: PAGE 34 (2026) Abstr 12202 [www.page-meeting.org/?abstract=12202]
Poster: Drug/Disease Modelling - Other Topics