Jennifer Lang (1), C. Steven Ernest II (2,3), Xiaosu Ma (3) and Parag Garhyan (3)
(1) Eli Lilly & Company, Bracknell, UK; (2) Currently, Metrum Research Group, Connecticut, US; (3) Eli Lilly & Company, Indianapolis, US
Introduction: Lipoprotein(a) (Lp(a)) is a low-density lipoprotein-like particle attached to apolipoprotein(a), and elevated Lp(a) levels are associated with increased risk of cardiovascular events such as myocardial infarction. Lepodisiran is a tris N-acetylgalactosamine conjugated small interfering ribonucleic acid (GalNAc-siRNA) which prevents the transcription of apolipoprotein(a) and leads to reduction of Lp(a) [1]. Following subcutaneous (SC) administration, GalNAc-siRNAs are quickly internalised by hepatocytes through specific binding to the asialoglycoprotein receptor. Consequently, GalNAc-siRNAs typically have very short plasma half-lives (≤10 hrs) [2]. GalNAc-siRNAs are then cleaved and released as unconjugated siRNA in the cytoplasm, where they bind to the RISC complex leading to mRNA degradation [3]. In preclinical species, higher liver exposure of several GalNAc-siRNAs compared to plasma was reported and liver half-lives are considerably prolonged (i.e., several weeks to months) [4]. The extended liver exposure is highly correlated with the sustained effect of siRNAs on target proteins. In this study, we developed a novel PK/PD model by first describing the short plasma PK of lepodisiran, and then linking the plasma PK to the prolonged Lp(a) lowering effect.
Methods: Lepodisiran was investigated in a phase 1 single-ascending dose study in healthy adults with elevated Lp(a) levels [1]. Forty-eight participants received either a single SC dose of lepodisiran (at 4, 12, 32, 96, 304 or 608 mg) or placebo. Serial plasma PK samples were collected up to 24 hours and on days 2, 3, 8, 15 and 29. Measurement of Lp(a) was scheduled (first weekly then bi-weekly/monthly) to evaluate the long-term drug effect up to 49 weeks. A compartmental PK model was developed to describe lepodisiran disposition and different absorption models were tested (i.e., zero/first-order, combined). A liver compartment (surrogate for RISC-bound concentration) was added to connect plasma lepodisiran PK and its effect on Lp(a), and accounts for the temporal difference between plasma concentrations and drug response. Inhibition of Lp(a) production was assumed to be driven by liver compartment concentrations using an indirect response model with Emax relationship. The liver volume was calculated based on individual body surface area [5]. Modelling was done using NONMEM v7.5.0 using FOCE-I.
Results: The one-compartment PK model adequately described lepodisiran plasma PK. The body weight was included as a covariate on the volume of distribution using the standard allometric scaling coefficient (= 1). Maximal concentrations (Cmax) and time to achieve Cmax (Tmax) were highly variable between subjects (observed Tmax [min-max] = [0.5-16] hours, observed Cmax CV% [min-max] =[22-54] %). Therefore, a combined zero- and first-order absorption was used to best describe the large variability in the absorption phase. Inter-individual variabilities were estimated for absorption rate (Ka), clearance (CL) and volume of distribution (V) parameters, and proportional residual error was used. Lepodisiran was rapidly absorbed into the liver but slowly eliminated (CL=0.000409 L/hr, RSE%=10.4%). A clear exposure-response relationship was established in the studied dose range in healthy participants with elevated Lp(a). The indirect response model estimated the maximal effect (Emax) at 97.6% (RSE%=0.7%), suggesting almost complete Lp(a) knockdown (KD). Significant Lp(a) KD (> 50% KD at nadir) was estimated at 12 mg dose. For doses greater than 96 mg, the maximum mean KD was above 90% and durability of protein KD increased with doses. Model-based simulations of patients with higher Lp(a) levels at baseline in the potential treated population predicted durable and substantial mean Lp(a) reduction, which may potentially support bi-yearly and yearly dosing.
Conclusion: A pop-PK/PD model was developed using phase 1 data from healthy adults with elevated Lp(a) which robustly explained the time course of lepodisiran plasma concentrations and Lp(a) levels. The inclusion of the liver compartment based on the mechanistic understanding of GalNAc-siRNAs enabled characterization of the temporal difference between short-lived plasma exposure and prolonged effect on Lp(a) lowering. Lepodisiran demonstrated durable and significant Lp(a) KD at doses greater than 96 mg. Results from this modelling suggest the potential to dose once or twice per year; these dosing regimens are being assessed in the lepodisiran ALPACA phase 2 trial (NCT05565742).
References:
[1] S.E. Nissen, H. Linnebjerg, X. Shen, K. Wolski, X. Ma, S. Lim, L.F. Michael, G. Ruotolo, G. Gribble, A.M. Navar, S.J. Nicholls, JAMA (2023).
[2] G. An, The Journal of Clinical Pharmacology n/a (2023).
[3] V.S. Ayyar, D. Song, S. Zheng, T. Carpenter, D.L. Heald, J Pharmacol Exp Ther 379 (2021) 134–146.
[4] A. Boianelli, Y. Aoki, M. Ivanov, A. Dahlén, P. Gennemark, Nucleic Acid Therapeutics 32 (2022) 507–512.
[5] K. Urata, S. Kawasaki, S. Ishizone, A. Komiyama, Hepatology 21 (1995) 1317–1321.
Reference: PAGE 32 (2024) Abstr 10904 [www.page-meeting.org/?abstract=10904]
Poster: Drug/Disease Modelling - Other Topics