Do Selective Serotonin Reuptake Inhibitors (SSRIs) Promote Stroke Recovery within the First Year After Stroke? ‐ A Cochrane Review Summary with Commentary

The aim of this commentary is to discuss in a rehabilitation perspective the Cochrane Review “Selective serotonin reuptake inhibitors (SSRIs) for stroke recovery” produced by Legg LA, Tilney R, Hsieh CF,Wu S, Lundström E, Rudberg AS, KutlubaevMA, DennisM, Soleimani B, Barugh A, HackettML,HankeyGJ,MeadGE https://doi.org/10.1002/ 14651858.CD009, and published by the Cochrane Stroke Group. This Cochrane Corner is produced in agreementwith PM&R: The Journal of Injury, Function, and Rehabilitation by Cochrane Rehabilitation.


Background
Despite major achievements in both acute stroke care and rehabilitation, stroke is still the third leading cause of disability 2 and years lived with stroke-related disability are increasing in societies worldwide. 3 Although multiprofessional specialized stroke care including rehabilitative treatment has been shown to reduce disability, that is, health-related limitations in activities of daily living and restrictions in participation in societal activities, and hence to promote independence after stroke, 4 there is great interest in potential additional therapeutic options to promote recovery after stroke. One of these options are drugs that influence cerebral neurotransmitter levels and promote neuroplastic changes associated with functional recovery. Selective serotonin reuptake inhibitors (SSRIs) are a class of drugs that are widely used to treat mood disorders (depression, anxiety) and had been shown to have a potential to promote recovery after stroke in a first Cochrane Review on the subject in 2012. 5  year post stroke and whether the treatment is associated with side effects.

What Was Studied in the Cochrane Review?
The population addressed in this review was stroke survivors within 12 months post stroke. The interventions studied were SSRIs including fluvoxamine, fluoxetine, sertraline, citalopram, and paroxetine administered at any dose, duration, and delivery mode for any reason. The SSRI intervention was compared to placebo treatment or usual care. The outcomes studied were disability as measured mostly using Barthel index (BI) or Functional Independence Measure (FIM) or other tools and independence as measured using the modified Rankin Scale (primary outcomes) as well as neurological impairments, depression, anxiety, quality of life, fatigue, healthcare cost, death, adverse events, and leaving the trial early (secondary outcomes).

Search Methodology and Up-to-Dateness of the Cochrane Review?
The review authors extensively searched for studies that had been published up to July 2018.
What Are the Main Results of the Cochrane Review?
Of those, only 3 RCTs (all using fluoxetine) were at low risk of bias for all of the assessed domains. The metaanalyses were primarily based on these three placebocontrolled RCTs and indicated: • no effect of SSRIs compared to placebo on disability scores: standardized mean difference (SMD) −0. When all studies (regardless of risk of bias) were used for a sensitivity analysis, SSRIs improved disability scores slightly compared to placebo or standard care/practice: SMD 0.23 (95% CI, 0.18 to 0.29; P < .001; 26 studies, 5334 participants) with considerable heterogeneity between trials (I 2 = 92%).

How Did the Authors Conclude?
The authors concluded that they found no reliable evidence regarding routine use of the SSRIs to promote recovery after stroke based on the meta-analyses of three trials at low risk of bias, which indicated lack of improvement in recovery from stroke with SSRIs. They identified potential improvements in disability when trials at high risk of bias were included in the analyses and hence a high risk to overestimate beneficial effects.
What Are the Implications of the Cochrane Evidence for Practice in Rehabilitation?
The strength of this Cochrane Review 1 is that its primary analyses were based on data from trials at low risk of bias with a substantial number of participants providing a valid, clinically speaking clear estimate of no therapeutic effect on disability indicating that SSRIs prescribed for stroke survivors within the first year after stroke did not reduce disability or dependence.
The results of the meta-analyses have largely been driven by a single large trial; accordingly, future evidence from other large trials using SSRIs with different approaches (eg, specific drug and dosage used, time post stroke, duration of treatment, combination with training etc.) might modify "the overall picture." A slight clinical benefit could be corroborated for depression but not for other aspects of interest.
The SSRI treatment was found to be acceptable to patients (no increase in the number of participants leaving the study early), and the results of the adverse event analyses were not worrisome (increased gastrointestinal side effects with SSRI prescription as expected, no increase in seizures or death rates).
In conclusion, the overall benefit-risk analysis speaks against the routine use of SSRIs as an adjunct treatment and means to reduce disability and dependence among stroke survivors within the first year after stroke.