João A. Abrantes (1), Daniel A. Norris (2), Jayashree Sahni (1), Michael McCaleb (2), Yanfeng Wang (2), Nicolas Frey (1)
(1) Roche Innovation Center, Basel, Switzerland, (2) Ionis Pharmaceuticals, Carlsbad, CA, USA
Introduction: IONIS-FB-LRx is a triantennary N-acetyl galactosamine (GalNAc)-conjugated antisense oligonucleotide (ASO) in development as a potential treatment for geographic atrophy associated with age-related macular degeneration (AMD). The ASO portion of IONIS-FB-LRx targets complement factor B (FB) mRNA thus preventing the production of FB protein, which increased levels are associated with AMD [1,2].
Objectives: Characterize the pharmacokinetics and pharmacodynamics (PK-PD) of IONIS-FB-LRx, in particular, the relationship between PK and FB as well as the alternative complement pathway hemolytic activity (AH50) levels using a population model-based approach.
Methods: Data from 41 healthy volunteers (HV) were available from two Phase 1 clinical studies (single ascending dose (SAD), N=17; multiple ascending dose (MAD), N=24). The drug was administered subcutaneously and the dose levels were 10, 20 and 40 mg single-dose (SAD) and 10 and 20 mg administered in 8 doses during a 6-week treatment period (2 times per week during the first 2 weeks and then Q1W for 4 weeks). In total, 796 plasma concentrations, 437 FB observations and 288 AH50 observations (measured with the AH50 hemolytic assay) were included in the analysis. The median subject was a 45-year male weighing 84.2 kg. The PK and PD models were built simultaneously. Modelling was carried out using NONMEM version 7.4.1, and the dataset and graphical analyses were generated using R version 3.5.1. Models were selected and qualified based on the likelihood ratio test, prediction- and simulation-based diagnostic plots, parameter uncertainty and reliability of parameter estimates.
Results: A two-compartment PK model with first-order absorption and linear elimination and two indirect response PD models were found to describe well the longitudinal PK, FB and AH50 data collected during the SAD and MAD studies. Body weight was implemented on the PK disposition parameters using theory-based allometric exponents. The estimated ka was 1.3 h-1, and the derived disposition half-lives t1/2 alpha and t1/2 beta were 4.7 hours and 35 days, respectively. The inter-individual variability associated with the estimated primary parameters ranged between 15% and 50% CV, and the proportional residual error was 27% CV. The decreases in FB and AH50 levels following the administration of the drug were described using two structurally independent indirect response models with an Imax inhibitory drug effect affecting the input. Full inhibition was assumed to be possible. A body weight effect was identified on the estimated baseline of FB and AH50 (p<0.001; positive relation). For FB, the IC50 was 0.28 ng/mL and the kout 0.0024 h-1, and for AH50 the IC50 was 0.50 ng/mL and the kout 0.0016 h-1, which described the larger decrease observed in FB compared to AH50. In addition, covariance elements between inter-individual variability on similar PD parameters (i.e.: baseline, IC50, kout) of FB and AH50 were included, with an estimated correlation of 64%, 77% and 58%, respectively. The proportional residual error for FB was 14% CV, and the additive error for AH50 was 8.4 IU/mL. All PK and PD parameters were estimated with reasonable precision (RSE<12% for typical parameters and <31% for inter-individual variability parameters). The model predictive performance was good as shown by prediction-corrected visual predictive checks. Simulations using the PK-PD model indicated that to reach an appropriate level of inhibition in FB in the median subject (Q2Wx2 + Q4W dosing at steady-state), at least 20 mg would have to be administered. To reach an appropriate level of inhibition in AH50 in the median subject, at least 40 mg would have to be administered.
Conclusion: A population PK-PD model to characterize the PK of IONIS-FB-LRx, and its inhibitory effect on FB production and AH50 in HV was successfully developed. This model was used to inform the selection of doses and dosing frequency in the clinical development studies of IONIS-FB-LRx in AMD patients.
References:
[1] Boyer et al. The pathophysiology of geographic atrophy secondary to age-related macular degeneration and the complement pathway as a therapeutic target. Retina. 2017 May;37(5):819-835.
[2] Reynolds et al. Plasma complement components and activation fragments: associations with age-related macular degeneration genotypes and phenotypes. Invest Ophthalmol Vis Sci. 2009 Dec;50(12):5818-27.
Reference: PAGE () Abstr 9442 [www.page-meeting.org/?abstract=9442]
Poster: Drug/Disease Modelling - Other Topics