Pharmacological augmentation strategies in treatment resistant obsessive-compulsive disorder: a double-blind placebo-controlled trial

Obsessive-compulsive disorder (OCD) manifests often as a chronic illness and is characterized by the presence of obsessions and compulsions. Firstline treatment options, which include selective serotonin reuptake inhibitors (SSRI) and cognitive-behavior therapy with exposure and response prevention techniques, fail to achieve a satisfactory response in up to 40% of patients. Current evidence supports the augmentation of SSRI with antipsychotics, such as quetiapine. However, anti-psychotics are effective for only one-third of the patients and have been associated with severe long term side effects. This study aimed to compare clomipramine and quetiapine augmentation of the SSRI fluoxetine. Previously to the beginning of this trial all patients had to: report OCD as they primary diagnosis, be taking the highest tolerated or recommended dose of fluoxetine for at least eight weeks, have a current Yale Brown Obsessive-Compulsive Scale (YBOCS) total of at least 16, and have had a reduction of less than 35% of the initial total YBOCS score with fluoxetine treatment. Fifty-four patients were allocated trough a minimization procedure in one of three arms: quetiapine (up to 200 mg/day) plus fluoxetine (up to 40 mg/day) (QTP/FLX) (N=18), clomipramine (up to 75 mg/day) plus fluoxetine (up to 40 mg/day) (CMI/FLX) (N=18) and 18 placebo plus sustained maximum dose fluoxetine (up to 80 mg/day) (PLC/FLX) (N=18). Blinded raters collected YBOCS scores at weeks 0 and 12. Analyses were made with intention-to-treat and hot-deck imputation of missing data. Wald statistics from non-parametric ANCOVA for ordinal categorical repeated measures were used to evaluate group, time and interaction effects for YBOCS scores and secondary outcome measures considering initial measures as covariates. Clinical Global Impression scores of improvement (CGI-I) were used to classify individuals in responders or non-responders. Chi-square was used to evaluate frequency of responders in each group...