The study randomized 161 participants (102 adolescents, 59 adults), with prespecified analyses planned by age group and baseline irritability severity.

As an exploratory signal-finding Phase 2 study, IRIS was explicitly designed to test multiple clinical endpoints based on preclinical findings, including social communication and irritability, and to identify the most appropriate development path forward. The study did not meet its primary endpoint of change from baseline to Week 12 in the caregiver-reported Autism Behavioral Inventory (ABI)–Social Communication Domain score. Social communication is a domain for which no approved pharmacologic therapies exist, and for which validated, treatment-sensitive outcome measures remain an area of active scientific investigation.

However, in a prespecified analysis of adolescents (age 12–17) with moderate or greater baseline irritability (double-blind baseline ABC-I score >16, N=20), ML-004 demonstrated a clinically meaningful improvement in irritability over placebo as measured by change from baseline in the care-partner reported ABC-I subscale (LS mean difference vs. Placebo −9.58, ES=1.33, nominal p value=0.013). Consistent with this finding, clinically meaningful improvement over placebo was observed on the Clinician Global Impression-Improvement (CGI-I)-Irritability domain in the adolescent population randomized with moderate or greater baseline irritability (LS mean difference −0.63; ES=1.08, nominal p value=0.036). The treatment effects on the ABC-I and CGI-I irritability domain were more pronounced among adolescents with greater baseline irritability. In the total population of participants (age 12-45) with baseline ABC-I score >16 (N=26), ML-004 demonstrated an effect size of 0.64 (nominal p-value =0.13) on the key secondary endpoint of change from baseline in the ABC-I score at week 12.

"We are very encouraged by the robust improvements observed in adolescents with clinically significant irritability," said Erin Pennock Foff, Chief Medical Officer. "These results are consistent with our compelling pre-clinical evidence for reduction in aggression/irritability in animal models. Given that there is an established regulatory path in this indication using the ABC-I, and given the magnitude of the effect on this measure observed in this study, we look forward to engaging with the FDA to discuss a possible path forward. We are grateful to the participants, families, and investigators whose commitment made this trial possible."

"Irritability is a pressing clinical problem in adolescents with autism, and the only approved pharmacologic options are antipsychotics, which carry substantial metabolic and neurological burdens," said Matthew State, M.D. Ph.D. (Chair of Psychiatry and Behavioral Sciences at the University of California, San Francisco (UCSF) and member of MapLight’s Scientific Advisory Board). "An effect size of this magnitude, particularly in those most severely affected, points to a clinically meaningful improvement and warrants further investigation of ML-004 in this population."

Safety and Tolerability

ML-004 was generally well-tolerated, with treatment-emergent adverse events (TEAEs) that were all mild to moderate in severity. Adolescents experienced fewer TEAEs than adults (Adolescent TEAEs: 62.7% for ML-004 versus 41.2% for placebo; Adult TEAEs: 86.7% for ML-004 versus 72.4% for placebo).

  • There were no SAEs or severe AEs reported in the ML-004 treated participants; among placebo-treated participants, two experienced a severe TEAE and one experienced a serious adverse event
  • In the randomized population, the most common TEAEs (>5% in ML-004 arm and >placebo) were headache, nausea, somnolence, vomiting, fatigue, and dizziness
  • No events of extrapyramidal TEAEs were observed with active treatment. The mean weight gain over the course of the study was lower for ML-004 than placebo
  • In adolescents, the most common adverse events (occurring in ≥5% of ML-004-treated adolescents and at least twice the rate of placebo) were headache (13.7% vs. 3.9%), somnolence (11.8% vs. 0%), nausea (9.8% vs. 0%), and vomiting (5.9% vs. 0%)
  • Two (3.9%) adolescents in the active arm discontinued the study due to an adverse event (0% for placebo)