Abstracts

Rationale: Dravet syndrome (DS) is a severe and progressive genetic developmental and epileptic encephalopathy that typically begins in the first year of life. Approximately 85% of cases are caused by heterozygous, loss of function, de novo mutations in the SCN1A gene, which encodes the voltage-gated sodium channel type 1 α subunit (Na_v1.1) protein. DS is characterized by high seizure frequency (SF) and severity, intellectual disability, ataxia/motor abnormalities, and a high risk of sudden unexplained death in epilepsy. STK-001 is an investigational ASO treatment designed to upregulate Na_v1.1 protein expression in the brain by leveraging the wild-type (non-mutant) copy of SCN1A to restore physiological Na_v1.1 levels, thereby potentially reducing both SF and non-seizure comorbidities. Non-clinical PK/pharmacodynamic studies identified the STK-001 levels in brain that elicit a 2-fold increase in Na_v1.1 levels, which is the desired pharmacology in patients. The study objective was to provide simulations in the pediatric population with DS to predict dose and frequency that would lead to pharmacologically active brain levels of STK-001.

Methods: A semi-mechanistic population PK model was developed based on STK-001 concentration profiles collected in 95 non-human primates (NHP) after single or multiple intrathecal administrations. The NHP model was then scaled to predict exposure in patients with DS aged 2 to 18 years based on published data considering: (1) scaling for interspecies extrapolation; (2) age differences in brain/ cerebrospinal fluid (CSF) volumes and in body weight; and (3) reduced height and weight growth trend in patients with DS. Monte Carlo simulations were performed at single dose and two repeat dosing regimens: (1) 3 doses to be administered every 4 weeks (Q4W) and (2) 3 doses at Days 1, 57, and 85.

Results: Observed time-concentration profiles in plasma, CSF, and brain compartments were adequately described using a 14 compartment NHP model with parallel linear and saturable elimination from the plasma compartment and a lag time for the dose input. The model was used to simulate exposures in plasma, CSF and brain of patients treated with doses and frequency used in ongoing Phase 1/2a clinical studies in US and UK. Validation of the population PK model was performed by over-lapping observed STK-001 concentrations in plasma and CSF from approximately 40 patients in these studies with the 95% confidence intervals of the predicted concentrations (2.5 - 97.5th percentiles) in patients with DS.

Conclusions: The PK model developed with NHP data along with published data for animal to human scaling was leveraged to predict the pharmacologically active doses of STK-001 in pediatric patients with DS. The model will be further validated or adjusted using data from ongoing non-clinical studies and phase 1/2 and open-label extension studies in patients with DS receiving STK-001.

Funding: Stoke Therapeutics