Pharmacological therapies and major gastrointestinal bleeding in patients with atrial fibrillation

To the Editor: We were surprised that, in the systematic review and meta-analysis on the effect of pharmacological therapies for stroke prevention on major gastrointestinal bleeding in patients with atrial fibrillation, the authors considered triflusal in the same group as indobufen and aspirin, as they also inhibit the cyclooxygenase (COX)-1 enzyme (1). Although this is true, some pharmacological differences between these agents and some important evidence-based clinical differences between triflusal and aspirin should be considered. With respect to pharmacological differences, it should be mentioned that the COX-1 inhibition by aspirin is irreversible (2) whereas the inhibition by indobufen is reversible (3). Concerning triflusal, it should be taken into account that this agent is an irreversible inhibitor of COX-1, whereas its active metabolite 2-hydroxy-4-(trifluo-romethyl) benzoic acid (HTB) is a reversible inhibitor of the enzyme (4). Triflusal has little influence on vascular endothelial prostacyclin production, thus potentially preserving the favourable vasodilatory and platelet anti-aggregatory effects of prostacyclin. Conversely, aspirin reduces both thromboxane and prostacyclin release, thus reducing the favourable vascular effects of prostacyclin (4). Lastly, triflusal, but not aspirin, inhibits cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP) phosphodiesterase, and therefore cAMP and cGMP breakdown in platelets, thus inhibiting calcium mobilisation and calcium-dependent platelet aggregation (4). Triflusal has not been compared with indobufen, but the available evidence from randomised clinical trials (5-7) demonstrates that triflusal has an efficacy similar to aspirin to prevent serious vascular events, but with a lower haemorrhagic risk, in patients with acute myocardial infarction (5) and ischaemic stroke or transient ischaemic attack (TIA) (6, 7). These results have been confirmed in a meta-analysis (8). Moreover, in The Triflusal vs. Aspirin in Cerebral Infarction Prevention (TACIP) Study (6), comparing triflusal against aspirin in 2107 patients with ischaemic stroke or TIA, the incidence of gastrointestinal haemorrhage was 7.6% in the aspirin group and 4.6% in the triflusal group (p = 0.005). A meta-analysis in 5169 patients with acute myocardial infarction or ischaemic stroke or TIA (8) showed that the risk of gastrointestinal haemorrhage is lower with triflusal than with aspirin (OR for aspirin vs. triflusal: 1.83; 95% CI: 1.35–2.48), and in two case-control studies it was reported that the risk of upper gastrointestinal bleeding associated with triflusal is negligible, whereas the risk associated with the use of aspirin is clear (9, 10). Concerning patients with atrial fibrillation, the National Study for Prevention of Embolism in Atrial Fibrillation (NASPEAF) (10) showed that the addition of triflusal to reduced-intensity anticoagulation in patients with atrial fibrillation, stratified for risk of thromboembolism, significantly decreases subsequent vascular events compared with standard anticoagulation, and does so without increasing neither systemic nor gastrointestinal bleeding risk. The clinical benefit of the association of triflusal with reduced-intensity anticoagulation in patients with atrial fibrillation has not been shown neither with aspirin (12-14) nor with another other antiplatelet drug. In our opinion, all the above are important reasons to distinguish triflusal from aspirin and indobufen in a meta-analysis on gastrointestinal bleeding in patients with atrial fibrillation. Javier Borja and Alejandro Doménech are employees of Vifor Pharma Spain, S.L., the Company that markets triflusal in Spain. Julián García-Rafanell is employee of J. Uriach y Compañia, S.A., the Company that markets triflusal in other countries.