Until recently, there has been little evidence regarding the safety of triple therapy (TAT) in AF patients after PCI or DES. To investigate whether TAT is the optimum antithrombotic strategy in these patients, a number of randomized clinical trials (RCTs) have been conducted. The WOEST trial (What is the Optimal antiplatelet and anticoagulant therapy in patients with oral anticoagulation and coronary stenting) was the first RCT to highlight this issue and it was a tiebreaker between interventionalist and electrophysiologist. While interventionalists advocate for antiplatelet therapy to avoid stents thrombosis, electrophysiologists point to the importance of anticoagulants for stroke prevention in AF, with neither therapy adequate on its own [20]. WOEST compared DAT (warfarin plus clopidogrel 75 mg/day) to TAT (warfarin plus clopidogrel 75 mg/day plus aspirin 80 mg/day) in patients who underwent PCI with stenting. A bleeding event (primary outcome) at 12 months, occurred in 19.4% of the DAT group versus 44.4% of the TAT group (p < 0.001). The secondary outcomes [death, MI, stroke, stent thrombosis, or target vessel revascularization (TVR)] appeared in 11.1% of the DAT cohort versus 17.6% of the TAT group (p = 0.025) [21]. However, it is worth mentioning that this trial did not compare with NOACs, and it was underpowered to detect differences in stent thrombosis and mortality. The bleeding reported was any bleeding, even if it may not be clinically relevant, which may explain the high bleeding rate in this study. Moreover, there was a low use of proton pump inhibitors (PPIs) at around 30%, which could be another explanation for the high bleeding rate.

Four RCTs compared TAT (with Vit K antagonist) versus DAT (with a P2Y12 inhibitor, mainly clopidogrel) plus multiple adjusted doses of NOACs, rivaroxaban 15 mg o.d. (PIONEER AF-PCI), dabigatran 110 mg or 150 mg b.i.d. (RE-DUAL PCI), apixaban 5 mg b.i.d. (AUGUSTUS), and edoxaban 60 mg o.d. (ENTRUST-AF PCI) in AF patients undergoing PCI (see the summary of these trials in Table 2). The PIONEER AF-PCI randomized 2214 individuals with non-valvular AF who underwent PCI with stent implantation to three treatment groups (approximately 700 each). The study aimed to assess the safety of two different strategies of rivaroxaban (rivaroxaban 10–15 mg daily with P2Y12 inhibitor for 1 year or rivaroxaban 2.5 mg twice daily plus DAPT for 1, 6, or 12 months) and a Vit K antagonist (VKA) strategy (TAT, warfarin with DAPT for 1, 6, or 12 months). Various DAPTs were used, mainly clopidogrel (also prasugrel or ticagrelor) and low-dose aspirin (75–100 mg/day). The primary outcome of clinically relevant bleeding was less in the rivaroxaban population (16.8%, 18.0%; p < 0.001) than in the warfarin population (26.7%; p < 0.001) at the 1-year follow-up. The diversity of ethnicity was better than in previous trials.; however, the study still lacked good numbers from the East Asian population. This may be important due to the increased incidence of clopidogrel resistance, especially in the Asian population due to the prevalence of genetic polymorphisms [22]. Moreover, this trial lacked robustness in assessing efficacy and secondary outcomes (CV mortality, stent thrombosis, MI, or stroke), and there is inadequacy in the presentation of the clinical adverse events that led to therapy cessation.

In the RE-DUAL PCI trial (randomized Evaluation of dual antithrombotic therapy plus dabigatran versus TAT with warfarin in Patients with non-valvular AF following PCI), a total of 2725 daily patients with AF who had PCI were randomly allocated (1:1:1) to obtain TAT (warfarin plus a P2Y12 inhibitor and aspirin for a period of 1–3 months), dual antithrombotic therapy (DAT) which includes dabigatran (dose of 110 mg twice daily) plus a P2Y12 inhibitor, or DAT with dabigatran (150 mg twice daily) plus a P2Y12 inhibitor. The most important findings of this study were a significant decrease (11.5%) of major bleeding (clinically meaningful) in DAPT (110 mg dabigatran) when compared with TAT (p < 0.001), and 5.5% in DAPT (150 mg dabigatran) when compared with TAT (p = 0.002). Furthermore, the safety regarding stent thrombosis between the regimens was almost similar, with just a 1.1% increase in stent thrombosis in the DAT group [23]. Nevertheless, the study had some patients (12%) who received ticagrelor (thienopyridine) regardless of their age or risk of bleeding, which may potentially increase the risk of bleeding associated with ticagrelor, unlike clopidogrel, which is known from other studies to cause less bleeding. Similar to the WOEST, this trial was insufficiently powered to assess the thrombotic event (within a stent), as both DAT regimens showed no significant difference in stent thrombosis when compared to TAT. The definite stent thrombosis in the 110 mg dabigatran group was 1.5% versus 1.3% in the TAT group (HR 1.30 (95% CI 0.63–2.67); p = 0.15), while in the 150 mg dabigatran cohort, it was exactly the same as in the TAT group (0.9% vs. 0.9%) (HR = 0.99 (95% CI 0.35–2.81), p = 0.98) [23].

The ENTRUST-AF PCI trial enrolled 1506 individuals with AF and recent PCI who were randomized into two groups including TAT (warfarin and clopidogrel 75 mg daily for 1 year and aspirin 100 mg q.d. for 1–12 months) and edoxaban 60 mg daily plus clopidogrel 75 mg daily for 12 months. This trial adds to accumulating evidence that TAT is linked to higher major bleeding rates (20%), compared to 17% in the edoxaban arm (p = 0·001 only for non-inferiority), without any significant difference in the ischemic outcomes. This was the first trial to compare DAT with edoxaban against TAT (warfarin-based) in AF patients following PCI [24]. However, this was the only trial that failed to show superiority (p = 0.12) of NOAC but only non-inferiority (p = 0.001) for bleeding events, compared with a warfarin-based strategy.

More recently, the AUGUSTUS trial included 4614 patients to compare two antithrombotic regimens (apixaban vs. VKA/warfarin). In addition, this study compared aspirin against a placebo. This broad objective came after the uncertainty of whether the low rate of bleeding observed on DAT in previous trials was the result of NOAC usage or due to withholding aspirin. Major or clinically significant bleeding occurred in only 10.5% of the individuals who received apixaban, while in 14.7% of those who had warfarin (p < 0.001 for non-inferiority and superiority test). Patients who received a placebo had lower bleeding event rates with only 9% as compared with 16.1% of those who received aspirin (p < 0.001). For the secondary outcomes, the apixaban group had lower mortalities and hospitalizations than the warfarin group (23.5% vs. 27.4%; p = 0.0002), and a comparable ischemic event (including stent thrombosis) in both groups. Regarding the aspirin versus placebo groups, there were similar numbers of deaths, hospitalizations, and ischemic events between the two groups [25]. Interestingly, the AUGUSTUS trial found that most stent thrombosis events occur early within the first month of implantation (Fig. 3), which means that aspirin can still be considered for up to one month in patients at prevailing risk of a thrombotic event [26]. The so-called thrombotic risk is not only driven by the CHADSVASC score but also by other clinical aspects such as the complexity of the performed PCI (number of stents, bifurcation, stent length, etc.) [27]. The trial may be considered a powerful study with reasonable population size to test both primary and secondary outcomes. Since this study is an open-label design, this may lead to selection and observation bias; nevertheless, this seems to be mitigated by blinded outcome assessment.

Stent Thrombosis in Patients With Atrial Fibrillation Undergoing Coronary Stenting in the AUGUSTUS Trial (2020): As seen in this figure that the incidence of stent thrombosis is largely confined to the first 30 days after stent implantation [26]. https://doi.org/10.1161/CIRCULATIONAHA.119.044584

The accumulating evidence of antithrombotic therapy following PCI in AF patients has been reflected in the guidelines for atrial fibrillation, which resulted in a significant shift from the TAT strategy toward DAT over time (Fig. 4) [28]. Before the publication of the WOEST (2013) and ISAR-TRIPLE (2015) trials, clinical guidelines suggested that these patients should be on TAT (warfarin and DAPT) following PCI. Nowadays, the current ACC/AHA and ESC guidelines recommend the discontinuation of aspirin before the first month (preferably 1 week, depending on the bleeding and thrombotic risk profile) and while continuing NOACs and P2Y12 inhibitors. Due to the bleeding safety profile, the preferred antiplatelet (P2Y12 inhibitor) is clopidogrel, especially in low thrombotic risk patients, which can be discontinued 6–12 months post DES implantation depending on the thrombotic and bleeding risk [29, 30] (Figs. 5, 6, and 7).

Triple therapy: A review of antithrombotic treatment for patients with atrial fibrillation undergoing percutaneous coronary intervention; this diagram shows the history of guidelines and how the use of TAT is becoming less endorsed [28]. DOI: https://doi.org/10.1016/j.jjcc.2018.09.001

WOEST trial (2013): Primary outcomes of any bleeding, Significant high bleeding in the TAT group compared to DAT [21]. DOI: https://doi.org/10.1016/S0140-6736(12)62177-1

PIONEER AF-PCI (2016) trial: Primary outcomes of major bleeding, Significant high bleeding rates in the TAT group compared to DAT. Group 1 = rivaroxaban 10–15 mg daily + P2Y12 inhibitor, Group 2 = Riva. 2.5 mg BID + DAP, Group 3 = VKA + DAPT [33]. https://doi.org/10.1056/NEJMoa1611594

RE-DUAL PCI (2017) trial: Primary outcomes of major bleeding were significantly high bleeding in the TAT group compared to both DAT (dabi. 110 and 150 mg). dabi. 110 mg had a better safety profile (lower bleeding) than TAT and DAT (dabi. 150 mg) [23]. https://doi.org/10.1056/NEJMoa1708454

The clinical guidelines and trials are extremely important for identifying the optimal therapy, but they do not tell us everything that happens in the real context of clinical practice. Despite the importance of assessing the safety of triple therapy in these trials, there remain unanswered questions about whether we should consider lower doses of NOACs in addition to antiplatelet therapy, when aspirin is needed due to the high risk of ischemia. Another question is whether the new biodegradable stents would offer better clinical outcomes for these patients. As seen from the AUGUSTUS trial, stent thrombosis is characterized by short-term occurrence (in the first 4 weeks). However, this observation does not fully explain why it is higher during the first month. This leads to the consideration of other possible significant factors such as stent under-expansion and whether intracoronary imaging (e.g., IVUS, OCT) was done to ensure stent optimization would add more clarity to the result. Additionally, some of these trials had a mixed population of ACS patients who were treated by PCI and others who had medical therapy alone. Therefore, further work is required to explore the mechanisms behind early stent thrombosis and to highlight other important clinical questions (Figs. 8 and 9).

ENTRUST-AF-PCI (2019) trial: Primary outcomes of major bleeding were significantly high in the TAT group compared to both DAT (edoxapan 60 mg). dabi. *failed to show any superiority but only non-inferiority [24]. DOI: https://doi.org/10.1016/S0140-6736(19)31872-0

AUGUSTUS trial (2019): Primary outcomes of major/clinically relevant bleeding, showing significant high bleeding in the TAT group compared to DAT. *Note: all patients received P2Y12 inhibitor besides the given medications [25]. https://doi.org/10.1056/NEJMoa1817083

There are some strategies cardiologists can use to mitigate bleeding risk when managing such patients. First comes an appropriate assessment of the bleeding and ischemic risk using validated risk scores, such as the CHA2DS2-VASc and the HAS-BLED score. Second is the use of DAT with NOACs and clopidogrel instead of using triple therapy whenever possible [31]. Because of the increased risk of bleeding with ticagrelor and prasugrel, clopidogrel is chosen over the other two antiplatelet agents in stable patients who are already on anticoagulants [32, 33]. Third, in cases of high ischemic risk, clinicians should keep triple therapy as short as possible, and use a low dose of aspirin (< 100 mg daily) [31]. Another important strategy is to consider the routine use of proton pump inhibitors (PPIs) with anticoagulants, which can reduce upper GI bleeding and hospitalization in high-risk patients [32]. Lastly, in critical patients with contraindications to both VKA and NOACs due to high bleeding risk, left atrial appendage occlusion can be an alternative nonpharmacological option to maintain safety and reduce stroke events [4].