Atrial Fibrillation, With ACS and PCI: Walking a Tightrope

Roxana Mehran; Deborah N. Kalkman; Dominick J. Angiolillo


Eur Heart J. 2019;40(19):1563-1566. 

"'Do one thing every day that scares you.'"
Eleanor Roosevelt

One in three patients with atrial fibrillation (AF) has concomitant coronary artery disease, and ~5–8% of patients undergoing percutaneous coronary intervention (PCI) have AF.[1–3] The optimal antithrombotic therapy after PCI for patients with AF remains the subject of debate.[4] The current 2018 ESC guideline on myocardial revascularization recommends that triple therapy with aspirin, clopidogrel, and an oral anticoagulant (OAC) should be considered for at least 1 month and up to 6 months in patients in whom the ischaemic risk outweighs the bleeding risk, while dual therapy with clopidogrel and an OAC should be considered as an alternative to 1 month triple therapy in patients in whom the bleeding risk outweighs the ischaemic risk.[5] In contrast, the North American Consensus group have recommended as a default strategy to use single antiplatelet therapy (SAPT) in combination with an OAC.[6] However, both recommendations from both sides of the Atlantic agree that, in the absence of contraindications, a non-vitamin K antagonist OAC (NOAC) should be preferred over a vitamin K antagonist (VKA).[7] Nevertheless, there are a number of treatment options that have been proposed based on a patient's underlying risk, and evidence for the best treatment strategy in the NOAC era is now beginning to surface.

Several randomized studies have evaluated the use of OACs combined with antiplatelet therapy for the treatment of patients with AF undergoing PCI.[8–10] The WOEST trial randomized patients treated with a VKA and PCI to clopidogrel alone (double therapy) or clopidogrel plus aspirin (triple therapy).[8] This was followed by studies using an NOAC. The RE-DUAL PCI trial evaluated the effects of dabigatran (110 or 150 mg b.i.d.) with a P2Y12 inhibitor (double therapy) or a VKA with a P2Y12inhibitor and aspirin (triple therapy) in PCI patients with non-valvular AF.[9] The PIONEER-AF investigated treatment of AF patients with stent placement to either low-dose rivaroxaban (15 mg q.d.) with a P2Y12 inhibitor (double therapy), very low-dose rivaroxaban (2.5 mg b.i.d.) with DAPT, or therapy with a VKA plus DAPT (triple therapy).[10] All three trials showed that it is feasible to drop aspirin in AF patients after PCI who are already on an OAC + P2Y12 inhibitor, with a major reduction in bleeding and no signal of important ischaemic harm.

Although the strategy of dropping aspirin has indeed emerged as an approach to reduce the risk of bleeding complications in the presence of alternative antithrombotic drugs, whether the safety and efficacy of dropping aspirin is preserved in patients with AF who present with acute coronary syndrome (ACS) requiring PCI, known to be at increased risk of bleeding and ischaemic complications, remains unclear.[11] In addition, the comparison between choice of P2Y12 inhibitor in these treatment strategies has not been fully evaluated.

In this issue of the European Heart Journal, Oldgren et al. provide two pre-specified patient subgroups from the RE-DUAL PCI trial including patients with PCI due to ACS or undergoing elective PCI, and those treated with ticagrelor or clopidogrel.[12] These in-depth analyses include all 2725 patients from the main RE-DUAL PCI trial. Patients with a creatinine clearance <30 mL/min or other major comorbidities were excluded from this trial. Choice of P2Y12 inhibitor (clopidogrel or ticagrelor) was at the discretion of the investigator, and the protocol specified that this decision was to be taken prior to randomization. The primary endpoint consisted of the International Society on Thrombosis and Haemostasis (ISTH) major or clinically relevant non-major bleeding events.[13,14] As a secondary exploratory outcome, the composite of death, thrombo-embolic events, or unplanned revascularization was evaluated.

Approximately half of patients presented with an ACS (n = 1375, 50.5%), while the other half consisted of patients with elective PCI (n = 1349, 49.5%). Baseline characteristics significantly differed between ACS and elective PCI patients, with ACS patients being more frequently without prior AF treatment. Moreover, bare metal stents were more frequently implanted in ACS patients. Median follow-up duration in this trial was 14.0 months. In this first subgroup analysis, no statistically significant differences in the primary endpoint were observed between ACS and elective PCI patients [both 20.9%, hazard ratio (HR) 0.97, 95% confidence interval (CI) 0.81–1.15]. In line with findings from the overall study cohort, the risk of ISTH major bleeding events alone and TIMI major bleeding events was also reduced in ACS and elective PCI with both doses of dabigatran 110 mg dual therapy and dabigatran 150 mg dual therapy regimens compared with warfarin triple therapy (Figure 1). Death, thrombo-embolic events, or unplanned revascularization was seen in 14.8% of patients with ACS and in 12.4% of patients undergoing elective PCI (multivariable adjusted HR 1.13, 95% CI 0.91–1.41). All interaction P-values were non-significant.

Figure 1.

Event rates of the primary endpoint in percentages for dual and triple warfarin therapy and hazard ratio (HR) [95% confidence interval]. The primary endpoint consists of the International Society of Thrombosis and Haemostasis major or clinically relevant non-major bleeding events. All interaction P-values >0.10. ACS, acute coronary syndrome; PCI, percutaneous coronary intervention.

The second subgroup analysis showed that the vast majority of all trial patients were treated with clopidogrel (n = 2398, 88%) as compared with ticagrelor (n = 327, 12%). Ticagrelor was prescribed more frequently in ACS patients, and a higher percentage of ticagrelor patients had PCI procedures due to ST-segment elevation myocardial infarctions. Patients treated with clopidogrel had slightly higher mean CHA2DS2-VASc and modified HAS-BLED scores. Patients treated with ticagrelor had more clinical complexity factors. The overall primary endpoint occurred in 26.3% of patients with ticagrelor and in 20.1% of patients with clopidogrel (multivariable adjusted HR 1.35, 95% CI 1.05–1.72). Although the absolute bleeding rate was higher with ticagrelor, all interaction P-values were non-significant. Consistent with the main publication and the ACS subgroup analysis, dabigatran dual therapy (both doses) resulted in fewer bleeding events than warfarin triple therapy (Figure 1). The incidence of death, thrombo-embolic events, or unplanned revascularization was 18.7% in those treated with ticagrelor and 12.9% in those treated with clopidogrel (multivariable adjusted HR 1.34, 95% CI 1.00–1.82), with non-significant interaction P-values. The authors conclude their article stating that in patients with or without ACS, and patients treated with ticagrelor or clopidogrel, both dabigatran 110 mg and 150 mg dual therapy compared with warfarin triple therapy reduces the bleeding risks in patients with AF after PCI.

This article addresses important issues in the complex AF–PCI puzzle. Triple therapy consistently results in higher bleeding events, and the presented results show that this is consistent in ACS or elective procedures and regardless of P2Y12 inhibitor (ticagrelor or clopidogrel). However, some limitations should be kept in mind when interpreting these results. First, the patients presented in these analyses are a trial population with selected inclusion and exclusion criteria, which should be taken into consideration when translating results to patients from daily clinical practice. Secondly, with regards to the thrombo-embolic events, the main RE-DUAL PCI study was not adequately powered for individual thrombo-embolic events; therefore, in this subgroup analysis, the secondary outcomes are only exploratory. Thirdly, the choice of P2Y12 inhibitor, albeit chosen prior to randomization, was at the discretion of the investigator. Therefore, this analysis can only be seen as observational because no randomization between ticagrelor and clopidogrel occurred, with possible patient selection contributing to any differences in outcomes. However, it is also important to acknowledge that this represents the largest data set on the use of a newer generation P2Y12 inhibitor in combination with an NOAC which has ultimately impacted updated expert consensus recommendations which indicate that among patients at high ischaemic/thrombotic (e.g. ACS patients) and low bleeding risk, ticagrelor may represent a reasonable treatment option.[6]

The AUGUSTUS (an open-label, 2 × 2 factorial, randomized controlled, clinical trial to evaluate the safety of apixaban vs. vitamin K antagonist and aspirin vs. aspirin placebo in patients with atrial fibrillation and acute coronary syndrome or percutaneous coronary intervention, NCT02415400) and ENTRUST-AF PCI (edoxaban treatment versus VKA inpatients with AF undergoing PCI), NCT02866175) will further explore the role of dual therapy with an NOAC and P2Y12 inhibitor in AF patients undergoing PCI.[15,16] Putting together the pieces of this highly complex puzzle will help provide the antithrombotic treatment regimen associated with the best balance between bleeding and ischaemic risk in the AF population undergoing PCI.