Create a model of exposure-response (PK/PD) relationships for target engagement and tremor reduction in Essential Tremor (ET) for brexanolone injection and SAGE-217 to predict biomarker and efficacy PK/PD responses for SAGE-324.  

A PK/PD model for neuroactive steroid, GABA-A receptor positive allosteric modulators (PAMs) was generated using ET clinical trial data from brexanolone injection and SAGE-217 to inform SAGE-324 development. We tested PK/PD model based hypotheses in three SAGE-324 cohorts:  an 8:8 randomized placebo controlled qEEG study that tested the hypothesized PK/PD relationship for qEEG in healthy volunteers; and two open-label Essential Tremor cohorts (n=6 at 45 mg, n=5 at 60 mg) that tested the hypothesized PK/PD relationship for tremor reduction.

We found that qEEG and tremor reduction was well described by PK/PD modeling across all three compounds and that the clinically observed PK/PD responses for SAGE-324 were within confidence bounds of the predicted response.  The Essential Tremor Rating Scale (TETRAS) scores were modulated in exposure ranges where qEEG signals were observed, with reductions in tremor occurring at less than doubling of qEEG power and at well-tolerated exposures across all three compounds.

Our mechanistic modeling platform describing the exposure response-relationship for target engagement (by qEEG) and efficacy (through TETRAS) offers a unique opportunity to establish and utilize a common mechanism of action for neuroactive steroid-based GABA_A PAMs in ET to predict responses for novel compounds in development.  The results from our SAGE-324 Phase 1 signaling finding study in ET patients were consistent with the PK/PD model predictions. This confirmation supports our mechanistic PK/PD modeling approach and de-risks and informs Phase 2 trial planning.