Memantine to Treat Social Impairment in Youths With Autism Spectrum Disorder: A Randomized Clinical Trial.

Importance: Pharmacologic interventions for addressing social impairments in autism spectrum disorder (ASD) are lacking. Proton magnetic resonance spectroscopy (1H-MRS) studies in individuals with ASD have documented altered glutamate levels in the pregenual anterior cingulate cortex (pgACC). Objectives: To evaluate the safety and efficacy of memantine for treating social impairments in youths with ASD and to explore pgACC glutamate levels as a potential biomarker for treatment response. Design, Setting, and Participants: This 12-week, placebo-controlled, double-blind, parallel-design randomized clinical trial was conducted between January 20, 2015, and July 11, 2018. The study population comprised youths aged 8 to 17 years with ASD without intellectual disability (IQ≥85) recruited from ambulatory psychiatry clinics at an academic institution. Age- and sex-matched healthy control participants provided reference data for pgACC glutamate levels. Data analysis was conducted between January 7, 2020, and December 19, 2024. Interventions: Participants with ASD were randomized to memantine or placebo, with dose titration up to 20 mg/d. 1H-MRS scans were acquired to assess pgACC glutamate levels. Main Outcomes and Measures: Response was defined a priori as (1) a 25% or greater reduction in informant-rated Social Responsiveness Scale-Second Edition total scores and (2) a clinician-rated Clinical Global Impression-Improvement subscale (anchored for ASD) score of 2 or less. The association between pgACC glutamate levels and treatment response was explored using receiver operating characteristic (ROC) curve analysis. Results: This study included 42 youths with ASD who initiated treatment (mean [SD] age, 13.2 [2.6] years; 32 males [76.2%]). Of these youths, 35 were included in the intention-to-treat efficacy analysis (n = 16 treated with memantine and 19 with placebo), and 33 completed the trial (n = 16 treated with memantine and 17 with placebo). Significantly more memantine-treated participants met the response criteria compared with placebo-treated participants (9 of 16 [56.2%] vs 4 of 19 [21.0%]; odds ratio, 4.8 [95% CI, 1.1-21.2]; P = .03). Memantine was well tolerated and did not have significantly more adverse events compared with placebo. Mean (SD) pgACC glutamate levels were significantly higher in youths with ASD vs healthy control participants (95.5 [14.6] IU vs 76.6 [17.7] IU; standardized mean difference, -1.2 [95% CI, -1.8 to -0.6]; P < .001). Abnormally elevated pgACC glutamate levels (≥1 SD above that of healthy control participants) were observed in 20 of 37 participants (54.0%) with ASD and were associated with more treatment responders to memantine than placebo (8 of 10 [80.0%] vs 2 of 10 [20.0%]; odds ratio, 16.0 [95% CI, 1.8-143.2]; P = .007). ROC curve analysis indicated that pgACC glutamate levels were highly efficient at identifying treatment responders. Conclusions and Relevance: In this trial, memantine was well tolerated and significantly improved social impairments in youths with ASD. Elevated pgACC glutamate levels were associated with a favorable treatment response, supporting their potential as a biomarker for assessing memantine efficacy in individuals with ASD. Trial Registration: ClinicalTrials.gov Identifier: NCT01972074.