REIMAGINE: A central nervous system basket trial showing safety and efficacy of vafidemstat on aggression in different psychiatric disorders

Abstract

Aim: Vafidemstat is a brain-penetrant, orally bioavailable, small molecule irreversible inhibitor of the histone lysine-specific demethylase KDM1A (also known as LSD1), which corrects memory deficits and behavior alterations including aggression and social interaction deficits in preclinical models.

Methods: Here, we report the results of REIMAGINE, a phase IIa, single-center, open-label, one-arm basket trial that evaluated the safety and efficacy of vafidemstat on aggression in adult patients with borderline personality disorder (BPD), attention-deficit/hyperactivity disorder (ADHD), and autistic spectrum disorder (ASD). Participants received 1.2 mg/day of vafidemstat for 8 weeks.

Results: Vafidemstat was shown to be safe and well tolerated, and no drug-related clinically significant adverse events were observed. Furthermore, all neuropsychiatric scales assessed showed notable efficacy signals, whether assessing agitation/aggression (Clinical Global Impression for Severity [CGI-S] and Clinical Global Impression for Improvement [CGI-I] and Neuropsychiatric Inventory [NPI] questionnaire for Agitation-Aggression [NPI-AA]), overall patient functioning (total NPI), or disease-specific features (Attention-Deficit/Hyperactivity Disorder Rating Scale [ADHD-RS] and Borderline Personality Disorder Checklist [BPDCL]). Statistically significant improvements were observed in the aggregated data (all participants) and for each of the three disease groups independently. Changes were evident within the first 2 weeks of treatment.

Conclusion: In summary, the REIMAGINE study supports that vafidemstat is safe, well tolerated, and causes a significant and consistent reduction in agitation/aggression and nonaggression features in BPD, ADHD, and ASD. These data support continuing the development of vafidemstat as a new treatment option for these psychiatric disorders.

Keywords: aggression; attention‐deficit/hyperactivity disorder; autistic spectrum disorder; borderline personality disorder; vafidemstat.

Fig. 1

Effect of vafidemstat on suicidal ideation/behavior in patients with borderline personality disorder (BPD). Effect of vafidemstat on suicidal ideation/behavior in participants with BPD assessed by the Columbia‐Suicide Severity Rating Scale (C‐SSRS). Results are presented as Tukey whisker plots with outliers. Wilcoxon signed rank test was used for statistical comparisons.

Fig. 2

Vafidemstat efficacy on aggression. End of treatment. Effect of vafidemstat on aggression was assessed by the Clinical Global Impression for Severity (CGI‐S) (a), Clinical Global Impression for Improvement (CGI‐I) (b), and Neuropsychiatric Inventory questionnaire for Agitation‐Aggression (NPI‐A/A) (c) scales in all participants. Baseline and end‐of‐treatment (week 8) data are presented as Tukey whisker plots with dots representing outliers. Wilcoxon signed rank test was used for statistical comparisons.

Fig. 3

Vafidemstat efficacy on aggression. Change over time. Effect of vafidemstat on aggression was assessed by the Clinical Global Impression for Severity (CGI‐S) (a), Clinical Global Impression for Improvement (CGI‐I) (b), and Neuropsychiatric Inventory questionnaire for Agitation‐Aggression (NPI‐A/A) (c) scales in all participants. Change over time in the respective scores is presented as Tukey whisker plots with dots representing outliers. Friedman and Dunn multiple comparison post hoc tests were used for statistical comparisons (***P < 0.001, ****P < 0.0001).

Fig. 4

Efficacy of vafidemstat on overall patient functioning. Effect of vafidemstat on overall patient functioning, including total Neuropsychiatric Inventory (NPI) scores (a,b) for all participants and disease‐specific scales for participants with borderline personality disorder (BPD) assessed by the Borderline Personality Disorder Checklist (BPDCL) (c) and participants with attention‐deficit/hyperactivity disorder (ADHD) assessed by the Attention‐Deficit/Hyperactivity Disorder Rating Scale (ADHD‐RS) (d). Results are presented as Tukey whisker plots with dots representing outliers. Wilcoxon signed rank (a), Friedman and Dunn multiple comparison post hoc (b), or paired t test (c,d) were used for statistical comparisons (****P < 0.0001).

Fig. 5

Correlation between clinical scores. Pearson correlation coefficient (r) and P‐value between aggression scores: Clinical Global Impression for Severity (CGI‐S) and Neuropsychiatric Inventory questionnaire for Agitation‐Aggression (NPI‐A/A) (a) and Clinical Global Impression for Improvement (CGI‐I) and NPI‐A/A (b) for all participants, and between overall functioning scores for participants with borderline personality disorder (BPD): Borderline Personality Disorder Checklist (BPDCL) and total Neuropsychiatric Inventory (NPI) (c).

Fig. 6

Pharmacokinetics and KDM1A target engagement (TE) of vafidemstat. Plasma levels of vafidemstat (upper panels) and KDM1A TE in peripheral blood mononuclear cells (bottom panels) at pre‐dose (