A Double-Blind, Placebo-Controlled Trial of Memantine Vs. Placebo in Children with Autism Targeting Neurocognitive Outcomes

Background: Abnormalities in the modulation of the glutamate system and glutamatergic function have been reported in disorders affecting a variety of behavioral and neurological domains, such as cognitive flexibility, memory, and motor function. Impairments in memory and motor deficits in individuals with ASD are well documented. Whereas the neurobiology of such deficits is an area of active research, there is a paucity of intervention research for such deficits in autism. Memantine, an NMDA inhibitor, may provide improvements in memory and motor praxis/expressive language skills in children with autism.

Objectives: The study was initially designed as a large, multi-site, randomized trial to evaluate the tolerability of memantine and primary efficacy on memory and motor planning. Budgeting and recruitment changes resulted in narrowing the original trial objectives to a signal finding study at two sites.  Treatment targets include apraxia, expressive communication, memory, motor skills, and adaptive skills in children with ASD in a double-blind, placebo-controlled parallel study.

Methods: Children with ASD and phrase speech were recruited. Participants (n=23), ages 6-12 (M=9.45, SD=2.03) were randomized at a 1:1 ratio to treatment with memantine or placebo for 24 weeks.  Memantine and placebo were initiated at 3 mg and the dose was titrated up 3 mg a week for a maximum of dose of 12 mg depending on weight. Apriori defined outcomes included: apraxia and expressive language (primary); and memory and adaptive skills (secondary). Exploratory measures included additional neurocognitive tests of memory, fine & gross motor skills and apraxia; as well as functional outcomes measures (e.g. maladaptive behavior, social communication skills). All outcomes were collected at three time-points: baseline, Week 12, and Week 24. MANOVAs were conducted on change score composites for each primary outcome domain and individual subtests scores by domain (e.g. neurocognitive, functional skills, behavior).

Results: Memantine was well-tolerated with no significant adverse events reported during the trial.  Initial analyses were guided by the analytic protocol for the original, large clinical trial. No significant effects were found on composite change scores on apraxia subtests of the NEPSY2 (e.g. oromotor sequences, manual motor sequences, imitating hand postures) or expressive language measures (e.g. EVT2 and NEPSY2 repetition of nonsense words).  Exploratory analyses for the revised signal finding protocol included analyses of individual neurocognitive and functional outcome measures. Results suggest trends on measures of motor proficiency, apraxia, and expressive language. One representative test was selected for each domain and entered into a MANOVA: motor proficiency, expressive language, verbal memory, and visual-spatial memory. The MANOVA indicated an overall trend for facilitative effects of memantine on neurocognitive outcomes after 6 months of treatment (F=3.056, df=4, p=.050).  Effect sizes were moderate to strong (partial eta squared = .449). Future analyses include evaluation of composite scores, responder analyses, and covariates of treatment response to support the development of larger, randomized controlled trials targeting neurocognitive outcomes.

Conclusions: This 24-week randomized placebo-controlled pilot study suggests that memantine may improve understudied neurocognitive domains including motor proficiency and expressive speech in school-aged children with ASD.