Exposure-Response Analysis of Centanafadine Sustained Release Tablets in Adults with Attention-Deficit/Hyperactivity Disorder
Yanlin Wang1, Asma Baliouze2, Mélanie Wilbaux2, Viki Cotero-Lefevre1, Susan Shoaf1, Osman Turkoglu1, Joga Gobburu3, Xiaofeng Wang1
(1) Otsuka Pharmaceutical Development & Commercialization Inc., Princeton, NJ, USA (2) Pumas-AI, Dover, DE, USA (3) School of Pharmacy, University of Maryland, Baltimore, MD, USA
Objectives:
Centanafadine is a small-molecule with inhibitory activity at norepinephrine, dopamine, and serotonin reuptake transporters and is under development for the treatment of attention-deficit/hyperactivity disorder (ADHD) in adults and children.
The purpose of this analysis is to:
- Quantify drug effect and to characterize the impact of relevant covariates on adult ADHD investigator symptom rating scale (AISRS) total score upon centanafadine sustained release (SR) tablets treatment.
- Explore the relationship between centanafadine exposure and the Selected Safety Endpoints (SSEs): abuse potential, rash and skin eruptions, somnolence, insomnia, and suicidality.
Methods:
The analysis dataset included data from two randomized, double-blind, multicenter parallel-group, placebo-controlled phase 3 trials. These trials evaluated the efficacy, safety, and tolerability of centanafadine SR tablets administered twice daily, at total daily doses (TDD) of 200 and 400 mg, in adults with ADHD.
The exposure-efficacy model of longitudinal AISRS total score was developed sequentially: first, a suitable structural model was developed for the placebo response, which was followed by the incorporation of drug effect. Drug effect was accounted for by modifying the parameters of the placebo structural model either by incorporating a treatment effect or drug exposure (measured as average concentration [Cavg] or minimum concentration [Cmin]) effect. The effect of covariates, including age, weight, sex, ethnicity, race, trial, and baseline AISRS total score, on model parameters was also evaluated. Modeling was conducted in Pumas (version 2.4.0, Pumas- AI, Inc., DE, USA) [1]) using first-order conditional estimation with interaction.
For the exposure-safety analysis, logistic regression analyses were conducted to assess the relationship between the centanafadine exposure (Cavg, Cmin, and maximum concentration [Cmax]) and the probability of occurrence of each SSE. Subsequently, time-to-event data were modeled using Cox proportional hazards models.
Results:
The longitudinal changes in AISRS total score were adequately characterized by a placebo-response model with a binary on/off treatment effect to describe the improvement in AISRS total score. A linear function was used to describe the worsening (increase) of AISRS.
Based on data collected from these two trials with 6-week treatment duration, even though a flat relationship between centanafadine exposure and AISRS total score was observed, a significant treatment effect was demonstrated. The model parameters effectively captured the data by estimating a mean of 22% decrease in AISRS total score from baseline to week 6 in the placebo group and a mean of 29% decrease in the treatment group, along with a positive slope of 0.04 per day to account for the increase in the AISRS total score over time. The model's estimated mean change from baseline closely matched the observed change from baseline in the data for both the placebo and treatment groups. None of the covariates evaluated were found to have statistically significant impact.
Regarding the exposure-safety analyses, both logistic regression and survival analysis suggested that Cmax is a statistically significant predictor (p < 0.05) for Rash and Skin Eruption, and for Insomnia. However, the overall probabilities of events were low. For patients treated with placebo (Cmax=0 ng/mL), centanafadine SR tablets 200 mg TDD (median Cmax=546 ng/mL), or 400 mg TDD (median Cmax=1127 ng/mL), based on the logistic regressions, the predicted probabilities of Rash and Skin Eruptions are 1.29%, 1.94%, and 2.98%, respectively; the predicted probabilities of Insomnia are 3.47%, 4.64%, and 6.27%, respectively. No statistically significant relationship were identified between exposure and the other SSEs.
Conclusion:
The longitudinal trend in AISRS total score was adequately characterized in ADHD adults by a disease-drug model where a linear function was used to capture the worsening of AISRS total score, and a binary on/off treatment effect was used to describe the improvement in AISRS total score. No significant covariates were identified.
Logistic regressions and time-to-event analyses showed statistically significant relationships between centanafadine Cmax and Rash and Skin Eruption, and Insomnia. However, the overall probabilities of events were low. No statistically significant association between centanafadine exposure and other SSEs were identified.
References:
[1] Rackauckas, Chris, Yingbo Ma, Andreas Noack, Vaibhav Dixit, Patrick Kofod Mogensen, Simon Byrne, Shubham Maddhashiya, et al. 2020. “Accelerated Predictive Healthcare Analytics with Pumas, a High Performance Pharmaceutical Modeling and Simulation Platform