MapLight's ML-004 Autism Trial Misses Social Goal, Improves Irritability in Adolescents

IRIS (ML-004-002) randomized 161 participants—102 adolescents and 59 adults—with prespecified analyses by age group and baseline irritability severity.
The trial’s primary endpoint was the change from baseline to Week 12 in a caregiver-reported measure: the Autism Behavioral Inventory (ABI) Social Communication Domain score.
In the prespecified adolescent subgroup with moderate-to-severe baseline irritability, ML-004’s irritability improvement was quantified using caregiver-reported ABC-I and clinician-rated CGI-I Irritability, with TipRanks noting effect sizes exceeding one (and greater effects at higher baseline irritability).
Beyond “generally favorable safety,” the company also highlighted tolerability details including low rates of sedation/somnolence and lower mean weight gain than placebo—positioned as potential differentiation versus D2-based atypical antipsychotics.
MapLight Therapeutics reported on June 22 that its Phase 2 IRIS trial of ML-004 failed its primary goal of improving social communication in autism spectrum disorder. The stock still rose about 6% after the announcement, as investors focused on a separate signal: a strong reduction in irritability in a subgroup of adolescents. Endpoints News reported the irritability effect size exceeded 1.0, a number analysts called competitive with existing drugs.
The company, which holds $395.2 million in cash, said it will seek FDA guidance on a development path focused solely on irritability. CEO Christopher Kroeger called the study an "exploratory signal-finding" trial and said MapLight is "encouraged" by the results, according to BioSpace.
The IRIS trial enrolled 161 participants — 102 adolescents and 59 adults — and ran from September 2022 through March 2026. The primary endpoint was a caregiver-reported measure called the Autism Behavioral Inventory Social Communication Domain score at 12 weeks. ML-004 did not beat placebo on that measure, according to Nasdaq.
But a prespecified subgroup told a different story. Among adolescents with moderate-to-severe baseline irritability — those with a starting score of 16 or higher on the ABC-I scale — ML-004 showed a 9.6-point drop versus placebo. Effect sizes were 1.33 on the caregiver-reported ABC-I (p=0.013) and 1.08 on the clinician-rated CGI-I (p=0.036), TipRanks reported. That subgroup included only 20 patients.
The FDA has approved only two drugs for irritability in autism: risperidone (Risperdal) and aripiprazole (Abilify). Both block dopamine D2 receptors and often cause weight gain, sedation, and movement disorders. ML-004 works differently. It is a 5-HT1B receptor agonist — a modified form of zolmitriptan, a migraine drug — designed to regulate emotion without those metabolic side effects.
MapLight leaned hard on that distinction. The trial showed no extrapyramidal events — meaning no movement side effects — and mean weight gain was actually lower in the ML-004 arm than in the placebo arm, according to MarketWatch. Low sedation rates were also highlighted as a potential edge over existing antipsychotics.
Analysts at Jefferies compared ML-004's results to approved drugs. Existing antipsychotics show placebo-adjusted ABC-I improvements of 6 to 11 points. ML-004's 9.6-point drop puts it squarely in that range. Endpoints News reported Jefferies described the signal as "competitive," suggesting the drug could carve out a meaningful share if approved.
The catch: the positive result came from just 20 patients. Skeptics noted that small subgroup analyses carry a high risk of noise and may not hold in a larger Phase 3 trial, according to Endpoints News. A registrational trial would need to show the same effect in a much bigger population.
MapLight will now pursue an End-of-Phase 2 meeting with the FDA to discuss a Phase 3 program built around irritability in adolescents — a full pivot from the original social communication goal. Social communication still has no approved pharmacologic treatment, leaving that unmet need unaddressed for now.
The company's financial position gives it room to maneuver. With $395.2 million in cash as of March 2026, MapLight has runway into 2027, according to Nasdaq. The next major test comes in mid-August, when topline results from the ZEPHYR Phase 2 trial of ML-007C-MA in schizophrenia are expected. That readout will be a broader test of MapLight's "circuit-based" drug discovery platform.
Publishers
12
Articles
18
Reach
30