Shruti Sonni, MD
Mortensen JK, Larsson H, Johnsen SP, and Andersen G. Post-stroke use of selective serotonin reuptake inhibitors andclinical outcome among patients with ischemic stroke: A nationwidepropensity-score matched follow-up study. Stroke. 2013; 44: 420-426.
SSRIs are indicated for treating post-stroke depression because they decrease symptoms, improve recovery and reduce mortality. Studies show that they also impair platelet aggregation because they decrease platelet serotonin uptake, and they have been associated with vasoconstriction, including the reversible cerebral vasoconstriction syndrome. This large nationwide Danish study aims to investigate whether these effects of SSRIs on platelets and vasospastic activity have a clinical significance in ischemic stroke patients. The authors compared 5833 patients on SSRIs post-stroke with non SSRIs users, using propensity matching while controlling for major confounders, and estimated the hazards ratios for acute myocardial infarction (MI), recurrent stroke, major bleeding and death. At follow-up, 2.9% patients had an MI, 8.1% had recurrent ischemic stroke, 20.2% had major bleeding (1.4% had intracranial bleeding), and 34.4% died during follow-up. SSRI users had a lower risk of MI and recurrent stroke, but a higher risk of major bleeding (including a non-significantly higher risk of intracranial bleeding). Mortality was higher in the SSRI group.
Prior studies found that among new SSRI users, the risk of major hemorrhage is significantly higher with intermediate and high degree of serotonin reuptake inhibition. It is unclear how much higher mortality is in stroke patients compared to all SSRI users. The authors attribute the higher mortality in SSRI users to possible “confounding by indication,” suggesting underlying depression in this group could be the reason.
It is unclear whether including patients who reached the primary endpoint of MI, stroke, hemorrhage or death within 30 days would have altered the results of this study, given high risk of recurrent strokes in the 90 day period post-stroke. Also, dosage of SSRIs was not reported in this study. There may be an ideal dose which has a protective antithrombotic effect and has lesser bleeding propensity, which could be best demonstrated by a prospective study. Since it is difficult to quantify the degree of platelet inhibition by SSRIs, it makes reliable utilization of their anti-platelet property challenging. Though this study stimulates an interesting discussion regarding the various aspects of SSRI effects in post-stroke patients, the clinical implications remain confusing.