Fluoxetine for stroke recovery: Meta-analysis of randomized controlled trials

Abstract

Objective: To determine whether fluoxetine, at any dose, given within the first year after stroke to patients who did not have to have mood disorders at randomisation led to a reduction in disability, dependency, neurological deficits and fatigue; improved motor function, mood, and cognition at the end of treatment and follow-up, with the same number
or fewer adverse effects.
Methods: Searches in July 2018 included several databases, trials registers, reference lists, contact with experts. We excluded RCTs requiring patients to have mood disorder at randomisation. Co-primary outcomes were dependence and disability. Dichotomous data were synthesised using risk ratios (RR) and continuous data using standardised mean
differences (SMD). Quality was appraised using Cochrane risk of bias methods. Sensitivity analyses explored influence of study quality.
Results: The searches identified 3412 references of which 491 full texts were assessed for eligibility. Six new completed RCTs (n=3710) were eligible, making a total of 13 trials (n=4145). There was no difference in the proportion independent at the end of treatment (3 trials, n=3249, 36·6% fluoxetine vs 36·7% control; RR 1·00, 95% confidence interval 0·91 to
1.09, p=0·99, I2 78%) and no difference in disability (7 trials n=3404, SMD 0·05, -0·02 to 0·12 p=0·15, I2=81%). Fluoxetine was associated with better neurological scores and less depression but more seizures. Among the four (n=3283) high quality RCTs, the only difference between groups was lower depression scores with fluoxetine.
Conclusion: Fluoxetine does not reduce disability and dependency after stroke. It improves depression scores but increases seizures. Ongoing RCTs will determine its effects in stroke vary depending on ethnicity, background treatment and other factors.
Classification of evidence: meta-analysis