Authors :
Wei Yin, PhD – Takeda Development Center Americas, Inc., Cambridge, MA, USA
Hongxia Jia, MD – Takeda Development Center Americas, Inc., Cambridge, MA, USA
Wenping Wang, PhD – Takeda Development Center Americas, Inc., Cambridge, MA, USA
Axel Facius, PhD – thinkQ2, Baar, Switzerland
Gezim Lahu, PhD – thinkQ2, Baar, Switzerland
Thomas Wagner, MSc – thinkQ2, Baar, Switzerland
Venkatesha Murthy, MD – Takeda Development Center Americas, Inc., Cambridge, MA, USA
Presenting Author: Mahnaz Asgharnejad, PharmD – Takeda Development Center Americas, Inc., Cambridge, MA, USA
Rationale: Soticlestat (TAK-935) is a first-in-class potent, selective inhibitor of cholesterol 24-hydroxylase (CH24H), a brain-specific enzyme that catabolizes cholesterol to 24
S-hydroxycholesterol (24HC) resulting in reduced glutamatergic hyperexcitability. This analysis evaluated the pharmacokinetics (PK) and pharmacodynamics (PD) of soticlestat, focusing on CH24H enzyme occupancy and reduction in systemic 24HC.
Methods: Data were from 13 clinical studies: 8 phase 1 studies in healthy volunteers; 3 phase 2 studies of participants with developmental and/or epileptic encephalopathies (1 in adults; 2 in children); 2 phase 3 studies in Dravet syndrome and Lennox-Gastaut syndrome. PK/PD models were based on data from 9 studies. Three linked models were developed: a population PK model to characterize and quantify the PK of soticlestat; a PK/enzyme occupancy model to describe the binding kinetics of soticlestat with CH24H in the brain; and a PK/PD model to quantify exposure-induced inhibition of systemic 24HC concentrations. All models were grounded in previous work and refined to improve structural and stochastic aspects. Models considered covariate effects determined by statistical significance and clinical relevance, ensuring biological plausibility and precision of parameter estimates. The reference individual was defined as: 18 y; White; male; 60 kg; 20 µmol/L alpha-1-acid glycoprotein (AGP); 300 mg BID soticlestat; no antiseizure medications (ASMs). Goodness-of-fit was assessed through various numerical and graphical assessments; models were compared using statistical and clinical criteria.
Results:
The strongest covariate effects identified in the PK model were age and body weight, but these did not translate into observed exposure differences due to the weight-adjusted dosing algorithm applied in all pediatric studies. Women had a 44% higher dose-normalized exposure and Asian individuals a 13.4–23.7% lower exposure compared with the reference, suggesting possible sex-based and regional variations in drug absorption and elimination. Lower (8 µmol/L) and higher (40 µmol/L) AGP levels were linked to 34.8–47.4% lower and 37.7–66.3% greater dose-normalized exposures, respectively, highlighting the role of protein binding in drug elimination. ASMs that are strong CYP3A inducers had minimal impact on total PK exposure. The 3 soticlestat doses tested (100, 200, and 300 mg BID) attained the CH24H enzyme occupancy (averaging ≥ 85%) and 24HC inhibition (averaging ≥ 73%) required for efficacy. This PD effect extended to pediatric doses, as defined by the weight-adjusted dosing used in the phase 3 trials, ensuring pediatric and adult exposure levels are comparable.
Conclusions: Strong CYP3A inducers (ASMs) had minimal impact on soticlestat PK exposures so suggesting dose adjustments may not be required. All soticlestat doses tested achieved efficacious levels of occupancy and 24HC reductions. These robust models offer a solid foundation for guiding clinical decision-making and optimizing dosing strategies.
Funding:
Takeda Development Center Americas, Inc.