Project Epsilon
CNS-Sparing Oxo-Hydantoin AR Antagonists
Overview
Prostate cancer remains AR-dependent across disease stages. Enzalutamide-class AR antagonists are effective but cause CNS AEs (fatigue, cognitive effects, seizures) and strong CYP/PXR-mediated DDIs.
Unmet Medical Need
Clinical unmet need remains for potent AR antagonists that reduce CNS adverse events and minimize drug-drug interaction liability, which are well-recognized with enzalutamide.
Therapeutic Hypothesis
Redesign enzalutamide scaffold: Thiohydantoin → oxo-hydantoin (C=S→C=O) to increase polarity and reduce BBB penetration. Block N-dealkylation at the benzylic/tertiary amide nitrogen using N-CH₂F (C-F metabolic shield) or N-CD₃ to exploit primary KIE and reduce CYP-mediated clearance/induction.
Computational Work Performed
- DiffDock (generative diffusion docking) for pose generation & confidence modeling
- GROMACS 2020/2021 MD for complex stability sampling
- ADMET-AI (Chemprop-RDKit ensemble) property prediction
- ADMETlab 3.0 for orthogonal, large-panel property inference
- CNS MPO and PSA/logP property tuning toward non-CNS space
Next Milestones (Success Criteria)
- 1AR potency (biochemical + cellular) within 1-2× enzalutamide
- 2MDR1 efflux ratio ≥~3 (brain-sparing proxy)
- 3HLM metabolite ID showing N-dealkylation suppression
- 4PXR TR-FRET: ≤~50% of enzalutamide's activity
- 5Rat Kp,uu,brain ≤0.5 by brain microdialysis
Competitive Landscape
Enzalutamide/apalutamide: potent but CNS AEs and CYP induction. Darolutamide: low BBB penetration, fewer CNS AEs, different scaffold. Emerging AR NTD/DBD ligands and PROTACs seek to overcome resistance.
Key Differentiation
- 'Darolutamide-like CNS safety' in an 'enzalutamide-class target profile'
- Dual advantage: brain-sparing + metabolic/DDI de-risking via N-CH₂F / N-CD₃
- Property-led design with mechanism-plus-DMPK differentiation
Interested in This Asset?
Contact our licensing team to discuss partnership opportunities, access detailed data packages, and explore collaboration terms.