Almost 30% of patients with atrial fibrillation (AF) have co-existing coronary artery disease (CAD) [1, 2]. Anticoagulants are prescribed to AF patients to reduce the risk of stroke or systemic embolism. However, there is still a risk of the patient developing acute coronary syndrome (ACS), requiring percutaneous coronary intervention (PCI). Following PCI, antiplatelet drugs such as aspirin and P2Y12 inhibitors (clopidogrel, prasugrel and ticagrelor) play a predominant role in the prevention of in-stent thrombosis and cardiovascular events [3]. Antiplatelet therapy (APT) alone is inferior to oral anticoagulation for reducing the risk of thromboembolic events in AF patients [4]. According to the present European Society of Cardiology (ESC) guidelines, AF patients with co-existing ACS and those who undergo a coronary intervention are eligible for both anticoagulation and (dual) antiplatelet therapy ((D)APT) [5, 6]. However, a combination of warfarin with (D)APT carries a more than threefold higher risk for non-fatal and fatal bleeding compared with warfarin monotherapy [7]. An optimal balance has to be found to reduce the thromboembolic risk (i.e. stroke, systemic embolism and myocardial infarction) and to minimise the increased risk of bleeding resulting from concomitant use of an anticoagulant and (D)APT. Owing to the lack of evidence, the 2016 ESC guideline on the treatment of ACS in AF patients and the 2017 ESC focused update on DAPT in CAD are predominantly based on expert opinion [3, 6, 8]. This review summarises and discusses the most recent key developments and future studies with regard to combining anticoagulation (non-vitamin K oral anticoagulants (NOACs) or vitamin K antagonists (VKA)) with APT in AF patients undergoing PCI.

The 2014 AF guideline for the management of AF patients of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the Heart Rhythm Society (AHA/ACC/HRS) recommends warfarin rather than NOACs as the first-choice therapy in AF patients with ACS. This guideline mentions considering dual therapy consisting of an OAC plus clopidogrel 75 mg once daily (q.d.) as an alternative to initial triple therapy [5].

The 2015 ESC non-ST segment elevation myocardial infarction (NSTEMI) guideline recommends that the use of prasugrel or ticagrelor as a part of triple therapy should be avoided in the absence of safety and efficacy data (class C recommendation) [3]. Because it is known that the addition of (D)APT to oral coagulation increases the bleeding risk, limited data suggest that clopidogrel is probably the safest of the available P2Y12 inhibitors because of its lowest bleeding risk [9]. If the OAC is a VKA, the ESC NSTEMI guideline recommends a target international normalisation ratio (INR) of 2.0–2.5 with the exception of patients with a mechanical prosthetic valve in the mitral position [3].

The 2016 ESC guideline for AF recommends a short period of triple therapy (OAC + aspirin + clopidogrel) followed by a period of dual therapy (OAC + APT, preferably up to 12 months after the event) in AF patients with co-existing ACS and those who undergo PCI [6]. The duration of the dual and triple therapy depends on the bleeding risk as calculated by the HAS-BLED score, the reason for the PCI (ACS or stable CAD) and the type of stent (bare-metal versus drug-eluting). Modifiable bleeding risk factors should be corrected to minimise the risk of bleeding. This guideline also mentions dual therapy with an OAC and clopidogrel as an emerging alternative to triple therapy based on the results of the WOEST study (What is the Optimal antiplatElet and anticoagulant therapy in patients with oral anticoagulation and coronary StenTing) [10] When a NOAC is preferred for anticoagulation, the lowest effective dose for stroke prevention should be used (apixaban 5 mg twice daily (b.i.d.), dabigatran 110 mg b.i.d., edoxaban 60 mg q.d., rivaroxaban 20 mg q.d.) or the appropriate reduced dose if indicated by dose-reduction criteria according to the drug labelling [2].

Fairly recently, a new ESC focused update was published on DAPT in CAD [8]. The guideline expresses a preference for clopidogrel as a part of triple antithrombotic therapy; prasugrel and ticagrelor should be avoided in combination with oral anticoagulation because of worrisome bleeding rates in registries (level C recommendation). In patients treated with a NOAC, the lowest effective dose for stroke prevention is also recommended in this guideline. Interestingly, the most innovative recommendation in this guideline is the use of rivaroxaban 15 mg q.d. as an alternative to rivaroxaban 20 mg q.d. when combined with aspirin and/or clopidogrel. The new guideline seems to base this level B recommendation on the results of the PIONEER AF-PCI (Open-Label, Randomized, Controlled, Multicenter Study Exploring Two Treatment Strategies of Rivaroxaban and a Dose-Adjusted Oral Vitamin K Antagonist Treatment Strategy in Subjects with Atrial Fibrillation who Undergo Percutaneous Coronary Intervention) study. It is noteworthy that this same guideline states that the PIONEER AF-PCI study is largely underpowered for the assessment of meaningful differences in the incidence of relevant ischaemic events, such as stent thrombosis or stroke rates. Dual therapy with clopidogrel 75 mg and oral anticoagulation is still recommended for consideration as an alternative to 1 month of triple therapy for patients in whom the bleeding risk outweighs the ischaemic risk [8].

The European Heart Rhythm Association conducted a survey in 2014 to provide an insight into current practice in Europe regarding the management of AF patients presenting with ACS. This survey showed that, in 91.1% of cases, combining warfarin with aspirin and clopidogrel was preferable in AF patients after PCI [11]. Most of the responding centres treated patients with triple therapy (OAC + aspirin + clopidogrel) for 3 months followed by 12 months of dual therapy (OAC + aspirin), and the majority used warfarin [11, 12]. Of the NOACs, rivaroxaban 15 mg o.d. and apixaban 2.5 mg b.i.d. were most frequently used.

All of the ongoing trials are primarily focused on the safety endpoint in terms of major bleeding or clinically relevant non-major bleeding. All of these trials are underpowered to draw firm conclusions about the efficacy for stroke prevention in AF. This is of importance, since some of the trials (i.e. PIONEER AF-PCI and RT-AF) use a suboptimal NOAC dose for stroke prevention in AF. The short follow-up period of all the studies limits their ability to assess the effect on stroke and venous thromboembolic events, especially in the long term. Furthermore, these studies are also underpowered for the clinical endpoints death, myocardial infarction and in-stent thrombosis in patients with ACS. Most of the studies do not include patients who have had a stroke or bleeding, although these events do not contraindicate NOAC use. When determining the optimal treatment for an individual patient, the clinician should always take into account bleeding risk factors. Modifiable bleeding risk factors are, for example: adding a proton pump inhibitor with DAPT to reduce the risk of gastrointestinal bleeding, hypertension (<160 mm Hg target), alcohol use and the use of predisposing co-medication such as non-steroidal anti-inflammatory drugs.

Despite the efforts of researchers to synthesise more evidence from the ongoing trials, there will still be challenges for clinicians to choose the optimal treatment strategies for individual patients on NOACs and (D)APT with AF undergoing PCI. The ESC guidelines and the AHA/ACC/HRS recommendation to use the lowest effective NOAC dose in combination with (D)APT in patients with AF and co-existing ACS or PCI were based on limited evidence concerning the optimal combination of NOACs and APT (level of evidence C; expert opinion) [3, 5, 8]. In this respect, the term “lowest effective dose” could be confusing and the term “lowest suitable dose for AF” would be more appropriate. Current evidence indicates that the combination of NOAC and clopidogrel is safer than VKA plus DAPT, which increases the risk of bleeding, without clear advantages in efficacy.

As to whether (N)OAC should be combined with single APT rather than dual APT, evidence from the WOEST, PIONEER AF-PCI and RE-DUAL PCI trials seems to favour a combination with clopidogrel only, thus omitting aspirin [21]. In this respect, a very interesting trial would involve a treatment arm with only an OAC in an adequate dose for stroke prevention in AF without APT. Unfortunately, this has not been incorporated into any of the studies. A direct comparison of clopidogrel versus aspirin as APT after PCI has not been studied. Evidence from a randomised controlled trial in patients with atherosclerotic vascular disease (manifested as either ischaemic stroke, myocardial infarction or symptomatic peripheral artery disease) showed that clopidogrel was more effective than aspirin in reducing the combined risk of ischaemic stroke/myocardial infarction and vascular death [22]. A nationwide registry study confirmed that there was a marginally significant reduction of MI and coronary death with OAC plus clopidogrel compared to triple therapy in AF patients after myocardial infarction or PCI. This reduction was not significant for OAC plus aspirin [23]. Also, OAC plus aspirin was associated with a significantly increased risk of death from all causes. The choice of clopidogrel over aspirin in ongoing trials seems to be based mainly on the WOEST trial results. None of the studies thus far has made a direct comparison of a VKA plus single APT versus a NOAC plus single APT. The upcoming results of the AUGUSTUS trial will therefore be interesting owing to its 2 × 2 factorial design (see Tab. 1). The reduced-dose NOACs should only be given if the dose reduction is indicated according to the patients’ characteristics as listed in the drug labelling, irrespective of the APT used. A reduced-dose NOAC in “healthy” patients, with the exception of dabigatran 110 mg q.d., has not been shown to be effective for stroke prevention in AF. A common mistake is to give apixaban 2.5 mg combined with APT instead of the full dose. However, this would only be appropriate in the presence of dose-reduction criteria (creatinine clearance 10–30 ml/min or two out of three following: serum creatinine ≥133 µmol/l, body weight ≤60 kg and/or age ≥80 years). The recommendation of the 2017 ESC guideline on DAPT to consider rivaroxaban 15 mg q.d. instead of 20 mg when combined with aspirin and/or clopidogrel should therefore be followed with caution because this dose has not been proved to be effective for stroke prevention in patients with AF. Consequently, rivaroxaban 2.5 mg is not an option for stroke prevention in ACS patients with AF.

The majority of the above-mentioned NOAC trials are still recruiting at present, and the results will become available in the coming years. For now, triple therapy has proved to be too much in terms of safety outcomes. However, it can be expected that guidelines will continue to change in the near future, and what seems wise today may become obsolete tomorrow.