Vincent Madelain
Translational Pharmacometrics - Quantitative Pharmacology, Servier, Suresnes, France
Objectives:
Nucleic acid therapeutics, designed to hybridize with target RNAs to modulate gene expression, are promising therapeutic approaches for neurological diseases affecting central nervous system (CNS). However, due to the difficulty to directly observe drug distribution and pharmacological effect in CNS compartments, translational approaches aiming to link preclinical activity and distribution of a new therapeutic modality to clinics remain challenging [1-3]. Here we present a model-based approach to guide the FIH study design of a candidate nucleic acid therapeutic with improved distribution targeting a gene coding for a protein involved in a rare neurological disease.
Methods:
Pharmacological data used in the model were generated in mice expressing human gene of interest instead of murin one, with no phenotypic expression of the disease. The mice received a single administration of candidate nucleic acid therapeutic S by intra cerebro-ventricular route. Study design included dose levels ranging from 1 to 100 nmol and follow up up to 20 weeks with sequential sacrifices performed at 2, 4, 6, 12 weeks post administration and at termination. Target engagement measurements in brain structures consisted in mRNA inhibition quantified by RT-qPCR and human protein level expression quantified by ELISA compared to control group and were performed in parallel to the measurement of candidate S tissue concentrations.
Sequential PKPD modeling analysis using population approach was followed aiming i) to describe nucleic acid therapeutic concentration kinetic in mouse brain structures, ii) to characterize concentration effect relationship on mRNA level and iii) to capture the kinetic of targeted protein conditioning on mRNA depletion. PK and PKPD model selection was performed based on the value of BIC and diagnostics plots as NPDE and VPC. Finally, simulations using clinical PK profiles of a compound of same class with more classical chemistry and predicted PK profiles for candidate S from preliminary NHP data were performed to investigate the expected kinetic of protein decrease in patient.
Model estimations and simulations were performed using the Monolix Suite 2019R2.
Results:
PK profile of nucleic acid therapeutic candidate S was described in three brain structures, cortex, cerebellum and striatum, using a one compartment model with Michaelis Menten elimination, allowing to compute half-lives of 41, 61 and 34 days respectively in the linear phase of the kinetic. Distribution volumes were fixed based on literature values [4-5], allowing to estimate bioavailability of 2.5%, 3.3% and 1.4% respectively after ICV administration.
Targeted mRNA kinetic was described with a classical turnover model, assuming the nucleic acid therapeutic effect increasing mRNA kout, consistently to its mechanism of action. Power function with gamma parameter estimated to 1.3 and 3-transit compartments delay were selected to describe the effect of nucleic acid therapeutic concentrations on mRNA inhibition in the 3 brain structures.
Targeted protein kinetic was also described with a turnover model, with a production rate directly proportional to mRNA level in the 3 brain structures. Protein turnover was estimated with apparent half-life of 15 days, allowing to capture delay between PK and protein nadir.
Exploratory simulations performed with generic PK profile associated with classical chemistry suggest that usual clinical dose could not be sufficient to significantly decrease targeted protein, letting room for differentiation with an improved distribution profile.
Conclusions:
The developed PKPD model allowed to improve the understanding of the kinetic of action of the candidate nucleic acid therapeutic in brain structures and open the way to guide the dosing regimen in phase I study once full PK data package in NHPs would be available. Beyond this project, the translational strategy can be re-used for other candidates sharing similar chemistry developed in other neurological indications although adaptation may be needed depending on the target and disease.
[1] Suzan M Hammond, Annemieke Aartsma-Rus, Sandra Alves, Sven E Borgos, Ronald A M Buijsen, Rob W J Collin, Giuseppina Covello, Michela A Denti, Lourdes R Desviat, Lucía Echevarría, Camilla Foged, Gisela Gaina, Alejandro Garanto, Aurelie T Goyenvalle, Magdalena Guzowska, Irina Holodnuka, David R Jones, Sabine Krause, Taavi Lehto, Marisol Montolio, Willeke Van Roon-Mom, Virginia Arechavala-Gomeza, Delivery of oligonucleotide-based therapeutics: challenges and opportunities, EMBO Mol Med. 2021 Apr 9;13(4):e13243.
[2] Kiara Fairman, Miao Li, Baitang Ning, Annie Lumen, Physiologically based pharmacokinetic (PBPK) modeling of RNAi therapeutics: Opportunities and challenges Biochem Pharmacol. 2021 Jul;189:114468. doi: 10.1016/j.bcp.2021.114468. Epub 2021 Feb 10.
[3] Wei Yin, Mark Rogge, Targeting RNA: A Transformative Therapeutic, Strategy Clin Transl Sci . 2019 Mar;12(2):98-112.
[4] A Badea, A A Ali-Sharief, G A Johnson, Morphometric analysis of the C57BL/6J mouse brain, Neuroimage. 2007 Sep 1;37(3):683-93.
[5] Xiaoyun Liu, Benjamin R Miller, George V Rebec, David E Clemmer, Protein expression in the striatum and cortex regions of the brain for a mouse model of Huntington’s disease, J Proteome Res. 2007 Aug;6(8):3134-42.
Reference: PAGE 30 (2022) Abstr 10003 [www.page-meeting.org/?abstract=10003]
Poster: Drug/Disease Modelling - CNS