Discovery and Model‐Informed Drug Development of a Controlled‐Release Formulation of Nonracemic Amisulpride that Reduces Plasma Exposure but Achieves Pharmacodynamic Bioequivalence in the Brain

Abstract

Nonracemic amisulpride (SEP‐4199) is an investigational 85:15 ratio of aramisulpride to esamisulpride and currently in clinical trials for the treatment of bipolar depression. During testing of SEP‐4199, a pharmacokinetic/pharmacodynamic (PK/PD) disconnect was discovered that prompted the development of a controlled‐release (CR) formulation with improved therapeutic index for QT prolongation. Observations that supported the development of a CR formulation included (i) plasma concentrations of amisulpride enantiomers were cleared within 24‐hours, but brain dopamine D2 receptor (D2R) occupancies, although achieving stable levels during this time, required 5 days to return to baseline; (ii) nonracemic amisulpride administered to non‐human primates produced significantly greater D2R occupancies during a gradual 6‐hour administration compared with a single bolus; (iii) concentration–occupancy curves were left‐shifted in humans when nonracemic amisulpride was gradually administered over 3 and 6 hours compared with immediate delivery; (iv) CR solid oral dose formulations of nonracemic amisulpride were able to slow drug dissolution in vitro and reduce peak plasma exposures in vivo in human subjects. By mathematically solving for a drug distribution step into an effect compartment, and for binding to target receptors, the discovery of a novel PK/PD model (termed here as Distribution Model) accounted for hysteresis between plasma and brain, a lack of receptor saturation, and an absence of accumulation of drug occupancy with daily doses. The PK/PD disconnect solved by the Distribution Model provided model‐informed drug development to continue in Phase III using the non‐bioequivalent CR formulation with diminished QT prolongation as dose‐equivalent to the immediate release (IR) formulation utilized in Phase II.