Dual therapy with direct oral anticoagulants significantly increases the risk of stent thrombosis compared to triple therapy

Several randomized controlled trials (RCTs), in patients with non-valvular atrial fibrillation and acute coronary syndromes or undergoing percutaneous coronary intervention (PCI), have shown superior safety of dual vs. triple antithrombotic therapy (ATT) without significant differences in efficacy outcomes.1–5 Each of these RCTs was individually powered to assess safety but not ischaemic endpoints. The most recent, two-by-two, factorial AUGUSTUS trial5 showed superior safety of a direct oral anticoagulant (DOAC), apixaban, over a vitamin K antagonist (VKA) within both dual or triple ATT regimens. Nevertheless, with respect to efficacy, all three recent RCTs with DOACs vs. VKA3–5 showed a trend towards increased rates of stent thrombosis with early dual (i.e. dropping aspirin) vs. early triple ATT. We meta-analysed the data from the three RCTs that compared a DOAC plus a single antiplatelet agent vs. a DOAC or a VKA plus dual antiplatelet therapy (DAPT). A total of 8670 patients were involved. Given the low degree of heterogeneity calculated by the Q Cochran test and Higgins I2 statistics and the overall low event rates, a fixed-effect model with Mantel–Haenszel weighting was used. Treatment effect was reported as odds ratio (OR) and 95% confidence interval (CI). Statistical analysis was performed using RevMan software version 5.3. Meta-analysis for the individual efficacy outcome of stent thrombosis showed that dual ATT was associated with a 66%, significantly higher risk of events compared to triple ATT (OR 1.66, 95% CI 1.01–2.75; P = 0.05) (Figure 1); the absolute rates of stent thrombosis were 1% with dual therapy and 0.58% with triple therapy. Sensitivity analysis including only patients actually undergoing coronary interventions did not change the overall results. Meta-analysis of the three RCTs that compared a DOAC plus a P2Y12 inhibitor versus a DOAC or a VKA plus DAPT for the individual outcome of stent thrombosis. Anticoagulation is crucial to reduce the risk of stroke in patients with atrial fibrillation, but withdrawing DAPT in patients undergoing PCI can have important clinical implications.6–8 In 1996, Schömig et al.9 showed increased risk of coronary stent thrombosis at 30 days among 517 PCI patients treated with oral anticoagulant plus a single antiplatelet agent compared to DAPT, reporting 13 (5%) events in the oral anticoagulant arm and 0 events in the DAPT arm (P < 0.001). These results were confirmed 2 years later by Leon et al.,10 who randomly assigned 1653 patients who underwent PCI to receive aspirin alone, aspirin+warfarin or aspirin+ticlopidine (DAPT group). A significant reduction of angiographically evident thrombosis at 30 days was found both with DAPT compared to aspirin alone (relative risk (RR) 0.19, 95% CI 0.06–0.57; P = 0.001) and with DAPT compared to aspirin+warfarin (RR 0.20, 95% CI 0.07–0.61; P = 0.01). Absolute rates of stent thrombosis were 2.9% with aspirin alone, 2.7% with aspirin+warfarin and 0.55% with DAPT. More than 20 years later, our analysis provides further evidence that DAPT is superior to a single antiplatelet regimen for the prevention of coronary stent thrombosis in the presence of concomitant anticoagulant therapy. The use of stents to treat complex proximal coronary lesions, e.g. involving the left main stem, is increasing, as is the prevalence of patients at high ischaemic risk (e.g. with high total length of implanted stents, double stenting at bifurcations, prior stent thrombosis). Thus, in patients with atrial fibrillation undergoing complex coronary interventions, the significant increased rate of stent thrombosis with dual ATT with DOACs as compared to triple ATT should be taken into account before opting for very early dual therapy, unless the patient’s bleeding risk is prohibitively high. Conflict of interest: F.A. reports receiving fees from Amgen, Bayer, B-I, BMS/Pfizer and Daiichi Sankyo.