Subclinical Atrial Fibrillation: A review of 11 studies

You might ask; "Why's is subclinical Atrial fibrillation important?" We know that patients in atrial fibrillation(AF) have a 5 x risk of stroke compared to those not in AF and that even short bursts of AF predispose to stroke. These can be avoided with anticoagulant use.  (25 minutes reading)

Does subclinical atrial fibrillation also predispose to stroke and do we know if patients who present with AF, have had recent subclinical AF?

The Bottom Line is:

1. Subclinical Atrial Fibrillation is common and ranges in incidence for 9% to 30%
2. Atrial tachyarrhythmias  represent a 2.5 factor increase in the risk of ischaemic stroke or systemic embolism
3. Although the burden of AF needed for stroke has been postulated to be about 6 minutes, finding episodes of < 30 seconds in patients with stroke is not uncommon.
4. There appears to be no temporal relationship in patients with subclinical atrial fibrillation and stroke.
5. There is a temporal relationship between cardioversion and stroke, with the majority of strokes and TIAs occurring with the median time from cardioversion to event of 2 days and over 80% of events occurring during the first week after the cardioversion.
6. Earlier cardioversion (<12 hours) is associated with lower risk of thromboembolism
7. Heart failure and diabetes are associated with a higher risk of thromboembolism.

What is my practice?

This is what I do in terms of anticoagulation. Please tell if you do differently. This is based on my interpretation of the literature.

• If the patient presents at <12 hours post symptom (AF) onset and they have no risk factors, I cardiovert and don't anticoagulate.
• If the patient presents between 12 and 48 hours and no risk factors I may anticoagulate peri and after cardioversion for a minimum of 1 week and up to 4 weeks (shared decision making). We have to balance the risk of doing nothing versus anticoagulation..... enter the GARFIELD AF risk Score
  • If they have a CHA2DS2-VASc > 1 I anticoagulate until seen in outpatient cardiology clinic and a minimum of 4 weeks.
• If they are > 60 yo and have diabetes and heart failure history, I symptom control and anticoagulate, but may not always cardiovert as these are high risk patients.

Here is some of that evidence.

What is subclinical AF?

It is simply episodes of AF detected by monitoring without clinical symptoms being present.

How prevalent is subclinical AF?

The CRYSTAL AF Trial (1) was a prospective multicentre randomised trial that included patients > 40yo with a cryptogenic stroke in the last 90 day.

The aim of the trial was to determine the incidence of AF in patients randomised to implantable cardiac monitor versus standard monitoring.

They detected AF (average age of patients with AF was 61 + 11 years) in:

• 9% at 6 months
• 4% at 12 months
• 30% at 36 months

Cotter et al (2) performed a cohort study in 51 patients who had an implantable recording device and cryptogenic stroke and who had continuous monitoring for 50 days. AF was detected in 25.5% of patients with a median duration of 6 minutes.

The ASSERT Trial (3) where 2580 patients with implanted devices were monitored for 3 months, found that in patients with a history of hypertension, that 10.1% had subclinical atrial tachyarrhythmias, which were independently associated with a 2.5 factor increase in the risk of ischaemic stroke or systemic embolism.

The risk of stroke with atrial tachyarrhythmia was modulated by the patient's risk profile. A CHADS₂ score of >2, increased the risk of ischemic stroke or systemic embolism by nearly 4% per year. 

The IMPACT Study (4), a multicenter, blinded randomised trial, found that 20% of patients had asymptomatic or paroxysmal atrial fibrillation.

How long does AF need to be present to predispose to stroke?

Boriani et al (5) in a study of over 10,000 patients with implantable devices, looked at the duration and frequency of AF that was needed to confer stroke risk. They found that:

• AF was commonly asymptomatic
• Paroxysmal AF adds to stroke burden
• Threshold for AF burden associated with highest risk of stroke was 1 hour
• Relative risk of stroke increases by 3% for every additional hour of AF daily burden

This is concerning as we again see the fact that AF can be asymptomatic, but now we see that the burden of AF, in order to increase thromboembolic events is approximately 1 hour of the arrhythmia.

In a systematic review and meta-analysis(6) that looked at the frequency of stroke and TIA in patients with no history of AF, they found: 

• 6.5% had an episode of AF > 30 seoconds
• 9.0% had episodes of AF < 30 seconds.

The EMBRACE Study(7) assigned 572 patients to an event triggering recorder versus conventional 24 hour monitoring, with a primary endpoint of detection of AF for > 30 seconds within 90 days.

They found that in patients with a recent cryptogenic stroke or TIA, Paroxysmal AF was common:

• AF lasting > 30 seconds was found in 16.1% with the event recorder vs 3.2% with cardiac monitoring.

• AF lasting > 2.5 minutes was found in 9% of the event recorder group versus 2.5% in the normal recording group.

Ambulatory monitoring for 30 days increased the detection of AF by a factor of 5.

What is the temporal association of AF to Stroke?

Daoud et al (8)  found that patients with accelerated heart rate episodes were at a higher risk of stroke. In the majority of patients there was no AF detected in the 30 days prior to the thromboembolic event.

In this study, it did not matter if the AF episode was proximal to the stroke event.

What is the temporal relationship of stroke to cardioversion?

The FibStroke Study (9), although a registry study and looking at all forms of AF including permanent, paroxysmal and persistent AF, found that the majority of strokes and TIAs occurred with the median time from cardioversion to event of 2 days and over 80% of events occur during the first week after the cardioversion. 

They occurred in patients not using oral anticoagulation before cardioversion and a significant proportion of strokes were observed in relatively young patients with a low CHA2DS2-VASc score.

Other studies (10) have also shown that regardless of whether cardioversion is electrical or pharmacological, in the non-anti coagulated patient, there is a 5-7% thromboembolic risk in the first 30 days. 

The FinCV study(11)

This deserves a more detailed review here. It was a retrospective registry study from 3 hospitals, with over 2481 patients.

They looked at the "risk factors of thromboembolic complications after successful cardioversion of acute atrial fibrillation when no anticoagulation is used." There was a 30 day followup.

Primary Outcome: Thromboembolic event within 30 days post cardioversion.

What did they find?

The rate of embolic events was 0.7% (38 patients) and occurred between 1 and 27 days after cardioversion (median 2 days, mean 4.6 days)

• 31 strokes
• 1 stroke plus systemic emboli
• 4 TIAs

A subgroup analysis found that if cardioversion occurred < 12 hours from onset the risk was 0.3%. This increased to 1.1% at > 12 hours.

What is the risk of bleeding on a DOAC?

If we look at all-comers, we know that DOACs have a lower bleeding rate than Vit K antagonists. Rivaroxaban appears to have the highest risk of bleeding (11). If we look at the ARTESiA trial (12) the annual rate of major bleeding was 1.7%. Given this rate vs the 1.1% chance of thromboembolic event if cardioverted within 12-48 hours, and also that most events occur early in the use of anticoagulants, I am very conservative in using them in patients with no risk factors.

References

1.  Botto G L et al. Presence and duration of atrial fibrillation detected by continuous monitoring: crucial implications for the risk of thromboembolic events. J Cardiov Electrophysiol 2009;20:241-8
2. Cotter P E et al. Incidence of Atrial Fibrillation detected by implantable loop recorders in unexplained stroke. Neurology 2013;80: 1546-50
3. Healey J S etal. Subclinical Atrial Fibrillation and the risk of Stroke. NEJM 2012;266:120-129

4. Ip J et al. Multicenter randomised study of anticoagulation guided remote rhythm monitoring in patients with implantable cardiaoverter-defibrillator and CRT-D devices: Rationale, design and clinical characteristics of the initially enrolled cohort. Am Heart J 2009;188:364-70.

5. Boriani G et al. Device-detected atrial fibrillation and risk for stroke: an analysis of >10,000 patients from the SOS AF project (Stroke prevention Strategies based on Atrial Fibrillation information from implanted devices. Europ Heart J 2014;35:508-516.
6. Sposato L et al. Very short paroxysm account for more than half of the cases of atrial fibrillation detected after stroke and TIA: A Systematic review and Meta- Analysis. Int J Stroke. 2015;10:801-7
7. Gladstone D J et al. Atrial fibrillation in patients with cryptogenic stroke. NEJM 2014; 26:2467-2477.
8. Daoud EG et al. Temporal relationship of atrial tachyarrhythmias, cardiovascular events and systemic emboli based on stored device data: a subgroup analysis of trends. Heart Rhythm 2011; 8:1416-23

9. Palomaki A et al. Strokes after cardioversion of atrial fibrillation- The FibStroke Study. Int J Card 2016;203:269-273

10. Johansson AK, et al. Is one month treatment with dabigatran before cardioversion of atrial fibrillation sufficient to prevent thromboembolism? Europace. 2015;17:1514-1517.

11.  Hendrika A van den Ham et al. Major bleeding in users of direct oral anticoagulants in atrial fibrillation: A pooled analysis of results from multiple population-based cohort studies. Pharmacoepidemiol Drug Saf. . 2021 Oct;30(10):1339-1352.

12.  Lopes RD, et al. Rationale and design of the Apixaban for the Reduction of Thrombo-Embolism in Patients With Device-Detected Sub-Clinical Atrial Fibrillation (ARTESiA) trial. Am Heart J. 2017;189:137–145