Memantine Augmentation in Treatment-Resistant Obsessive-Compulsive Disorder An Open-Label Trial

Background: Data from the fields of genetics, neuroimaging, and animal studies, along with case reports and small clinical trials. point to a role for glutamatergic dysfunction in the pathophysiology of obsessive-compulsive disorder (OCD). We report on the first open-label study to test the hypothesis that memantine, a noncompetitive glutamate antagonist, will result in a clinically meaningful reduction in OCD symptoms in adults with treatment-resistant OCD, Methods: We recruited 15 adult Subjects with Diagnostic and Statistical manual of Mental Disorders, Fourth Edition-defined OCD and a baseline Yale-Brown Obsessive-Compulsive Scale (Y-BOCS) of IS or higher, who had failed to respond to treatment with a serotonin reuptake inhibitor (SRI), given at an adequate and stable dose for at least 12 weeks. The duration of memantine treatment was 12 weeks, and the dose was gradually increased to a target of 20 mg/d. Response was defined as a 25% or greater reduction in the Y-BOCS score at study end and a Clinical Global Impression-Improvement scale rating Of "much" or "very much" improved. Results: Data from 14 subjects were analyzable. Mean baseline Y-BOCS score was 27.4 (SD, 5.0). Subjects had failed an average of 2.8 (SD, 1.8) SRI trials: 6 subjects had failed augmentation with atypical antipsychotics. At study end, 6 subjects (42.9%) were responders, and response was achieved by EOW4. Responders had significantly lower baseline Y-BOCS scores (2-tailed t tests, P < 0.05, t = 2.2) and had failed fewer SRIs (P <= 0.05, t = 2.2). Side effects to memantine were mild and transient, and no subject withdrew from the study for an adverse event. Summary: In this open-label augmentation trial of memantine in treatment-resistant OCD. almost half the subjects had a meaningful improvement in symptoms. Our Study was limited by its small size, presence of comorbidities. and lack of control. Large double-blind placebo-controlled trials are needed to further test Our findings.