Stroke Prevention in Atrial Fibrillation

Atrial fibrillation is the most common sustained cardiac arrhythmia and is associated with substantial increases in the risk of stroke and systemic thromboembolism. With the successful introduction of the first non-vitamin K antagonist direct oral anticoagulant (NOAC) in 2009, the role of vitamin K antagonists has been replaced in most clinical settings except in a few conditions when NOACs are contraindicated. Data for the use of NOACs in different clinical scenarios have been accumulating in the recent decade, and a more sophisticated strategy for atrial fibrillation patients is now warranted. JACC: Asia recently appointed a working group to summarize the most updated information regarding stroke prevention in AF. This statement aimed to provide possible treatment option in daily practice. Local availability, cost, and patient comorbidities should also be considered. Final decisions may still need to be individualized and based on clinicians’ discretion. This is the part 1 of the whole statement.

stroke prevention in AF, with a focus on Asia.
Details of working group members are shown in the Supplemental Table 1. The working group carefully reviewed the most recent data and formulated an updated statement for stroke prevention in AF patients, especially focusing on prioritizing specific NOACs in different clinical settings.

PREVALENCE OF AF IN ASIA
In a recent meta-analysis of 58 articles from 8 countries in Asia, the community-and hospital-based AF prevalence ranged from 0.37% to 3.56% and 2.8% to 15.8%, respectively. 6 The prevalence rate of AF is continuously increasing in Asia as in Western countries; for example, the prevalence rate of AF in Taiwan was around 1.5% in year 2020 and that will reach 4.0% in year 2050. 7 Recently, 2 schemes have been developed for the prediction of incident AF for patients in Asia and Taiwan, namely, the C 2 HEST (coronary artery disease or chronic obstructive pulmonary disease [ AF score, respectively. 8,9 The C 2 HEST score was derived from 471,446 Chinese subjects with 921 incident AF cases, and validated in the Korean nationwide data, whereas the Taiwan AF score was derived from 7,220,654 Taiwan patients with 438,930 incident AF cases. 8,9 The area under the curve for the prediction of AF was 0.749 (95% CI: 0.729-0.769) for the simple C 2 HEST score and 0.756 (95% CI: 0.755-0.757) for the Taiwan AF score. 8,9 The calculation rules of these 2 scoring schemes and the risk of incident AF in different score strata are shown in Figure 1.

CONSENSUS STATEMENT.
The prevalence of AF is increasing in Asia, and several scoring schemes (such as C 2 HEST and Taiwan AF scores) can be used to predict the risk of incident AF for Asian patients.

ASSESSMENT/RE-ASSESSMENT OF STROKE AND BLEEDING RISKS
For AF patients in Asia, the annual risk of ischemic stroke was approximately 3.0% (range: 1.60%-4.95%) based on a pooled analysis of 8 studies. 6 Most international guidelines recommends use of the CHA 2 DS 2score and the HAS-BLED score to assess the stroke and bleeding risks of AF patients respectively. 5,[10][11][12] The HAS-BLED score has also been validated in European and Asian AF patients taking NOACs, and draws attention to the modifiable bleeding risk factors (unlike other scores) and facilitates identification of high bleeding risk patients for early review and follow-up. 13 The usefulness of the CHA 2 DS 2 -VASc and HAS-BLED scores in the prediction of ischemic and bleeding events has been well validated for AF patients in Asia. 14 It is important to understand that the stroke and bleeding risks of AF patients were not static, as patients will become older and acquire incident comorbidities. 15 For example, in a study from Taiwan that enrolled 14,606 incident AF patients with a CHA 2 DS 2 -VASc score of 0 (males) or 1 (females) at baseline, approximately 16.1% of men and 16.2% of women had CHA 2 DS 2 -VASc score of at least 1 (men) or 2 (women) at 1 year after incident AF. 16 Both followup CHA 2 DS 2 -VASc scores and delta-CHA 2 DS 2 -VASc scores (the difference between the baseline and follow-up scores) were associated with a higher risk of ischemic stroke, and had better predictive values compared to the baseline CHA 2 DS 2 -VASc score ( Figure 2). 16 Importantly, the prescription of oral anticoagulants (OACs) when patients' CHA 2 DS 2 -VASc scores increased was associated with lower risk of clinical events. 17 More recently, machine-learning models that accounted for dynamic changes in risk including newly acquired risk factors improved the prediction of AF-associated stroke. 18 The details of machine learning will be discussed in part 2 of the statement.
Bleeding risks as assessed by the HAS-BLED score are also dynamic, altered by aging and incident comorbidities (and the mitigation of modifiable bleeding risk factors). The accuracy of the follow-up HAS-BLED score or delta-HAS-BLED scores in the prediction of major bleeding was significantly higher than that of the baseline HAS-BLED score ( Figure 2). 15 In 24,990 anticoagulated AF patients with initial HAS-BLED score #2, 5,229 (20.9%) patients had an increment of their HAS-BLED scores to $3 at the end of 1 year, mainly due to newly diagnosed hypertension, stroke, bleeding, and concomitant drug therapies. 19 Given the stroke and bleeding risks of AF patients change over time, the CHA 2 DS 2 -VASc and HAS-BLED scores should be re-assessed regularly, ideally at every patient-physician contact.
There were only few recommendations from guidelines regarding the frequency of risk re-assessment. 12 For AF patients who acquired incident comorbidities and experienced ischemic stroke, a recent report showed that the interval from the acquirement of incident comorbidities to the occurrence of ischemic stroke was only 4.4 months in 90% of patients. 17 Accordingly, the 2021 Asia Pacific Heart

Stroke Prevention in Atrial Fibrillation
A U G U S T 2 0 2 2 : 3 9 5  for the baseline, follow-up, and delta-HAS-BLED scores in predicting major bleeding. Adapted with permission from Chao et al. 15

Stroke Prevention in Atrial Fibrillation
A U G U S T 2 0 2 2 : 3 9 5

DOSE-REDUCTION CRITERIA OF NOACS FOR AF
The doses of NOACs must be adjusted in certain conditions when the risk of bleeding is presumably high. Table 2 shows the ABCD rule of the dosereduction criteria of NOACs that were described in the drug labels and in major AF guidelines. 12 For rivaroxaban, apixaban, and edoxaban, the dosereduction criteria were adopted in their clinical trials, respectively. [36][37][38] In a recent meta-analysis of these 3 trials, patients eligible for reduced-dose NOACs were at elevated risk of thromboembolic and hemorrhagic complications when treated with The ABCD rule for dose reduction of NOACs should be followed to obtain best efficacy and safety results in patients with AF.

MAJOR NOAC TRIALS
The efficacy and safety of NOACs have been confirmed in 4 major clinical trials, and also by a recent patient-level meta-analysis. 4,[36][37][38]45 The J-ROCKET trial was the only NOAC trial dedicated specifically to Asian patients, but is underpowered for the efficacy outcomes. 46  For Asian patients, NOACs are more effective and safer than warfarin in stroke prevention for AF.

NOACS FOR ASIAN PATIENTS
There was no head-to-head

Stroke Prevention in Atrial Fibrillation
A U G U S T 2 0 2 2 : 3 9 5 -4 1 1 ( Table 4). The PIONEER AF-PCI trial was the only trial in that the doses of rivaroxaban (2.5 mg twice daily and 15 mg once daily) used in the trial have not been proven to be effective in AF patients for stroke prevention. 36,54 The AUGUSTUS trial is the largest one, the only one testing aspirin in a placebo-controlled fashion, and the only trial that showed lower bleeding with the NOAC (apixaban) vs VKA in a direct comparison using a factorial design. 49 The AUGUSTUS trial also enrolled patients with chronic coronary syndrome who received medical therapy alone. 49 These patients were excluded from other trials. These 4 trials focused on bleeding as the primary endpoint, with coronary events and stroke as important secondary endpoints. To summarize, these trials showed that dual therapy with a NOAC plus a P2Y 12 inhibitor reduced bleeding risk compared to triple therapy of warfarin, aspirin, and a P2Y 12 inhibitor. The reduction in the bleeding risk appeared to be driven by both receiving a NOAC instead of warfarin as well as by omitting aspirin. 41 The benefit was also observed in patients with ACS treated with medical therapy. 55 One should be reminded that all 4 trials were underpowered for thrombotic coronary event analyses; however, NOAC-based dual therapy seems to be safe in terms of coronary ischemic events. Several recent meta-analyses including the 4 NOAC trials have shown that there might be small but significant increases in the risk of coronary events and stent thrombosis when omitting aspirin. 56,57 We recommend the use of aspirin during the peri-PCI period up to 1 week after PCI in patients with high bleeding risk (HAS-BLED score $3) (Figure 4). Given the irreversible inhibition with aspirin on platelets, residual inhibition may persist for the lifespan of platelets (7 to 10 days). 58 For patients with low bleeding risk (HAS-BLED score <3), it is reasonable to continue aspirin for up to 1 month after PCI (Figure 4) as the thrombosis risk is highest in the first month after ACS. 59 In the AUGUSTUS trial, the use as aspirin immediately and for up to 30 days resulted in an equal tradeoff between an increase in severe bleeding and a reduction in severe ischemic events. 60 Extending aspirin therapy beyond 1 month after PCI is not recommended. The dual therapy consisting of a NOAC plus a P2Y 12 inhibitor should continue after the triple therapy, and persists for 6 to 12 months, depending on the bleeding risk ( Figure 4).  The Japanese AFIRE trial showed that continuing NOAC monotherapy beyond 1 year after a revascularization procedure in AF patients not only decreased the risk of major bleeding but also demonstrated noninferiority for the primary composite endpoint of cardiovascular events compared with the combination of NOAC and antiplatelet therapy. 61 It is generally accepted that most AF patients with chronic coronary syndrome should be transitioned to NOAC monotherapy without an antiplatelet agent as recommended in recent guidelines or consensus. 12,41 CONSENSUS STATEMENTS.
For patients with AF and ACS undergoing PCI, triple therapy with a NOAC, P2Y 12 inhibitor (clopidogrel preferred), and aspirin should be limited to 1 week after PCI when their bleeding risk is high (HAS-BLED score $3), followed by double therapy (a NOAC plus clopidogrel) for 6 months, and monotherapy with NOAC alone after 6 months.
For patients with AF and ACS undergoing PCI, triple therapy can be used up to 1 month after PCI when their bleeding risk is not high (HAS-BLED score <3), followed by double therapy (a NOAC plus clopidogrel) for 12 months, and monotherapy with NOAC alone after 12 months.
For patients with AF and ACS who receive medical therapy, triple therapy should be limited to 1 week after PCI, followed by double therapy (a NOAC plus clopidogrel) for 6 months, and monotherapy with NOAC alone after 6 months.

HEART DISEASES
Patients with moderate to severe mitral stenosis were excluded from major NOAC trials based on the previous Framingham study which showed that these patients might have a significantly increased risk of NOAC can be (re)initiated 1 day following a transient ischemic attack, but this is best prolonged to 14 days following a severe stroke.
NOACs are preferred over VKA in the secondary prevention of AF-related stroke.

MANAGEMENT OF ACUTE ISCHEMIC STROKE WHILE
RECEIVING OACs. Ischemic stroke in a patient while receiving NOAC therapy was milder compared to those without OAC. 81 The 2021 European Practical Guide recommended that thrombolytic therapy should not be given within 48 hours after the last dose of NOAC. Prolonged activated partial thromboplastin time for dabigatran or prothrombin time for factor Xa inhibitors indicated the anticoagulation effect, and thrombolytic agent should not be administered immediately after acute ischemic stroke. If NOAC plasma levels are below the lower limit of detection, thrombolysis may be proceeded. 41 In cases when the plasma level is unavailable and last NOAC intake is more than 48 hours in patients with normal renal function, thrombolysis may proceed. 41

Stroke Prevention in Atrial Fibrillation
A U G U S T 2 0 2 2 : 3 9 5

NOACS IN ELDERLY PATIENTS
Clinical guidelines recommended NOACs for stroke prevention in elderly patients. 11,12 In the landmark