(C-183) Risk of gastrointestinal hemorrhage associated with SSRI and SNRI antidepressant products: a review and meta-analysis

Background: Selective serotonin reuptake inhibitors (SSRIs) and serotonin and norepinephrine reuptake inhibitors (SNRI) medications block the uptake of serotonin into platelets and thereby increase the risk of bleeding, including gastrointestinal bleeding (GIB). While many analyses have been conducted on the relationship between SSRI/SNRI and bleeding, the risk of bleeding is not necessarily the same across the different antidepressants agents due to the degree of inhibition of serotonin reuptake.

Objectives: To conduct a systematic review and meta-analysis to examine the relationship between unique SSRI/SNRI compounds and risk of GIB.

Methods: A literature search using PubMed was conducted by using the chemical name for approved SSRI/SNRI antidepressants (i.e. SSRIs (citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, and sertraline), SNRIs (i.e. desvenlafaxine/venlafaxine, and duloxetine) combined with search terms for GIB and Mesh term “Gastrointestinal Hemorrhage.” Studies were required to report results specific to GIB. Data was extracted from those publications reporting relative risk, odds ratio, hazard ratio, and/or number of events/observations. Extracted data were used to conduct a random effects meta-analysis.

Results: A total of 15 observational studies were identified containing data at a product level. Study designs included 11 case control studies (73.3%), 2 cohort studies (13.3%), 1 case (6.7%) crossover and 1 using both case-control and cross-over designs. Sample sizes ranged from to over a thousand (N=1,579) to over half million (N= 666,235). Fluoxetine was evaluated in 14 studies, with evidence for duloxetine coming from only four studies. All SSRI/SNRI products evaluated had significant elevated risk of GIB, although there was some variation between the evaluated products. Venlafaxine had the highest estimated risk (OR =1.50, CI 95% 1.28-1.75), followed by citalopram (OR= 1.37 IC 95% 1.13-1.65). Escitalopram and fluoxetine had slightly lower risks (OR 1.34 IC 95% 1.10-1.62 and OR=1.34, 95% CI 1.19-1.51, respectively). Paroxetine and sertraline had the lowest GIB risks (OR =1.28, CI 95% 1.06-1.56 and OR =1.29, CI 95% 1.09-1.52, respectively).

Conclusions: SSRI and SNRI products all have elevated risk of GIB, but there is no significant difference across the agents. The increase in risk ranged from 28% for paroxetine to 50% for venlafaxine.