Sage Journals: Discover world-class research

Abstract

Atrial fibrillation (AF) is common and warrants consideration of oral anticoagulant (OAC) medication. Usually, the decision is straightforward, following the pathway outlined in the European Society of Cardiology’s guideline; however, certain situations fall outside of this evidence base – such as a diagnosis of subclinical AF made via implanted devices or wearable electrocardiogram monitors, or alternatively diagnosis of ‘secondary AF’ following a major stressor. Subclinical AF is associated with stroke, though not to the extent of clinical AF, and the benefits of anticoagulation appear to be lower. Longer episodes are more clinically meaningful, and recent randomised controlled trials have demonstrated that some patients derive benefit from OAC. Similarly, when AF is triggered by sepsis or non-cardiac surgery, specific evidence supporting OAC initiation is lacking and clinician behaviour is variable. Observational data demonstrate poorer outcomes in these patients, implying that the perception of a transient, reversible phenomenon may not be correct. Contrastingly, cardiac surgery very frequently induces AF, and the benefits of anticoagulation rarely outweigh the risks of bleeding. Following ischaemic stroke, recent evidence suggests that early (re-)initiation of OAC should be considered as this does not increase the risk of haemorrhagic transformation as previously hypothesised. This narrative review summarises the available literature and outlines, where possible, practical advice for clinicians facing these common clinical dilemmas.

Keywords

atrial fibrillation atrial high-rate episodes oral anticoagulation sepsis stroke surgery

Introduction

Atrial fibrillation (AF) is the most prevalent sustained cardiac arrhythmia, with a lifetime risk of approximately one in three.¹ AF is strongly associated with ischaemic stroke such that stroke prevention with anticoagulation is at the core of current AF guidelines.^2,3 Oral anticoagulant therapy (OAC) is indicated in patients with AF and a CHA₂DS₂-VASc score of ⩾1 in men and ⩾2 in women. Generally, decision-making to initiate anticoagulation or not is straightforward, but some clinical situations exist for which the evidence in favour of OAC is less clear-cut and real-world practice is more variable.

Common reasons for such clinical dilemmas often fall into one (or more) of three groups. First, the diagnosis of AF may be based on recordings of the heart’s electrical activity from sources other than a standard 12-lead electrocardiogram (ECG). Examples may include pacemaker devices, implantable cardiac monitors, or more recently, commercial wearable technology such as smartwatches.

Second, a diagnosis of new-onset AF may have been made in the presence of a significant physiologic stressor, such as sepsis, critical illness, or cardiac or non-cardiac surgery. If a patient reverts to sinus rhythm with treatment of the underlying stressor and no further episodes of AF are observed, there may be uncertainty regarding the long-term risks of recurrence and stroke.

Third, the benefits of anticoagulation may be felt to roughly balance with the risks of bleeding, for example, in patients with a history of a major bleeding event.

All three scenarios involve difficulties in accurately predicting stroke and/or bleeding risk in very diverse populations. Despite decades of research interest in AF, many challenges persist in relating data to individual patients.

Of these groups, the third is the most clinically variable and complex, as such generalisations may be less applicable, and a highly individualised approach is required. Hence, this review will focus on the first and second groups, providing a summary of the available evidence in a fast-moving field. In so doing, we aim to offer clinicians practical approaches to these challenging scenarios.

Finally, we include a section related to challenging decision-making following a recent stroke. This includes the timing of recommencement of OAC after ischaemic stroke or intracerebral haemorrhage (ICH), as well as advice for managing patients who have a ‘breakthrough’ stroke despite OAC therapy. A visual summary of the reviewed topics is presented in Figure 1.

Figure 1.

A summary of the key messages surrounding the reviewed dilemmas of subclinical AF, secondary AF and post-stroke AF.

Subclinical AF and atrial high-rate episodes

Nomenclature

Diagnosis of AF on a standard 12-lead ECG continues to be recommended as the gold standard; however, difficulty obtaining recordings during symptomatic periods for patients with paroxysmal arrhythmias remains a challenge.² Alternative means of assessing the cardiac rhythm have historically included cardiac implantable electronic devices (CIEDs) such as pacemakers and implantable cardioverter defibrillators (ICDs), plus implantable cardiac monitors (ICMs) which may be implanted for arrhythmia diagnosis. More recently, a proliferation of consumer-grade wearable technology capable of recording single or multi-lead ECG traces offers an additional route to a diagnosis of AF.⁴

This multiplicity has necessitated the creation of new diagnostic entities, and the variability of terminology used in the literature may be confusing. Clinical AF – demonstrated on a 12-lead ECG or alternatively at least 30 s of AF on a single lead ECG recording is generally distinguished from subclinical AF (SCAF) – asymptomatic episodes demonstrated on other recordings (e.g. Holter monitors, CIEDs, wearable monitors). Within SCAF, atrial high-rate episodes (AHREs) refer to arrhythmias detected only by CIEDs: defined by the European Heart Rhythm Association consensus as episodes with an atrial rate greater than 190 bpm.⁵ SCAF tends to progress to clinical AF, at a rate of around 10% per year.⁶

The majority of the published literature uses data from CIED patients and hence refers to AHREs and SCAF synonymously. It is possible that the findings can be generalised to SCAF diagnosed on direct-to-consumer wearables that serve the same purpose (extended ECG monitoring), though as yet large trials in this clinical setting are awaited.

Frequency of SCAF

Asymptomatic AF is more frequent than symptomatic AF^7,8 and is not benign – associated with higher mortality, likely because asymptomatic patients are more likely to be older and have a higher comorbidity burden.⁹ Due to the overlapping risk factor profile, SCAF is common amongst patients with indications for cardiac pacing. Gillis and Morck reported a 68% incidence of AF in patients who received a device for sinus node disease and a 37% incidence among patients with AV block. Also notably, the burden of AF increased throughout the study period (of approximately 2 years).¹⁰

The prospective ASSERT study followed 2580 patients aged above 64, without known clinical AF, who had recently undergone pacemaker or ICD implantation.¹¹ At 3 months, SCAF was detectable in 10.1%, and SCAF was strongly associated with the development of clinical AF (hazard ratio (HR): 5.56, p < 0.001). Another retrospective study of patients with pacemakers for sinus node dysfunction found that AHRE (here defined as >220 bpm for >10 beats, patients only included with at least one AHRE >5 min) was similarly associated with clinical AF (HR: 5.93, p = 0.0001), as well as mortality.¹²

Stroke risk in subclinical versus clinical AF

Numerous large studies have assessed the impact of AHREs on the risk of future stroke. The aforementioned ASSERT study reported an increase in the rate of stroke or systemic thromboembolism by a factor of 2.5, though the overall event rate was fairly low, with only 51 events in their population of 2580 – an annualised rate of 0.8% (the median CHADS₂ score of 2.3 would predict a rate of >4% per year).¹¹ Similarly, a 2018 meta-analysis of three studies described an annual stroke rate of 2.76%, with a mean CHADS₂ score of 2.1.¹³ Lau et al. established that whilst SCAF increased stroke risk, the temporal relationship between recorded episodes of AF and subsequent stroke events is inconsistent.

More recently, two large randomised trials of direct oral anticoagulant (DOAC) medications in patients with CIED-detected SCAF have been published. NOAH-AFNET6 (2536 participants) and ARTESIA (4012) compared DOAC therapy (edoxaban or apixaban, respectively) with placebo or aspirin.^14,15 NOAH-AFNET6 reported a stroke incidence of 0.9% per year in the edoxaban group and 1.1% in the placebo arm. Given the median CHA₂DS₂-VASc score of 4 in both groups – the expected annual stroke rate for a reference population with clinical AF would be 4.8%.¹⁶ Likewise, the stroke rates in ARTESIA were 0.78% in the apixaban arm and 1.24% in the aspirin arm, with an overall mean CHA₂DS₂-VASc score of 3.9.

In summary, whilst SCAF/AHREs are clinically meaningful with regard to ongoing stroke risk (with approximately a 2.5-fold increase), their effect is not as significant as a diagnosis of clinical AF when adjusting for other stroke risk factors.

Effect of event duration on stroke risk in SCAF

Given the stroke risk is lower than expected for the SCAF population as a whole, investigators have attempted to further stratify by the duration of the detected episodes, to identify candidates who may benefit from anticoagulation.

Firstly, very short AHREs do not appear to be clinically meaningful. Data from the RATE registry of 5379 patients show that episodes of AF less than 15–20 s did not confer any increase in the stroke risk over a 2-year follow-up,¹⁷ a finding that was later corroborated by Mahajan et al.’s meta-analysis.¹³

Beyond this, the threshold of effect has been difficult to pinpoint, and there exists variability between different analyses. In ASSERT, event durations ⩾6 min were predictive of stroke (HR: 1.77, p = 0.047); however, when stratified into quartiles by duration, only the group with the longest durations (⩾17.72 h) demonstrated significantly higher event rates (HR: 4.89). The discrepancy may be partially explained, as described above, by the low event rates in the study (only 51 events). Importantly, increasing AHRE duration maintained its predictive value across the spectrum of baseline stroke risk (as assessed by CHADS₂ score), with the highest event rates observed in CHADS₂ scores of >2.¹¹

Other studies report different thresholds: TRENDS, a 2009 prospective observational study found a doubling of thromboembolic risk in the group with AHREs above the median duration (5.5 h).¹⁸ Here thromboembolism was inclusive of transient ischaemic attack (TIA) in addition to stroke and systemic thromboembolism and total event rates were again low (40 events from a population of >2400 patients, with an annualised event rate of 1.2%). Contrastingly, Boriani and colleagues, in their analysis of pooled data totalling 10,016 patients, identified the highest risk at AHREs of 1 h duration, with slightly lower rates beyond this.¹⁹ Other meta-analysed data do not conclusively demonstrate any association at all between SCAF burden and stroke.²⁰

In another study, AHREs >5 min were predictive of major adverse cardiovascular events – though it is not clear to what extent this was driven by stroke.²¹

The risk-enhancing effect of increasing AHRE duration appears to be synergistic with other stroke risk factors. Botto et al. examined the relationship between both AHRE duration and CHADS₂ score with thromboembolic risk. Splitting their cohort into three groups by duration: <5 min, ⩾5 min but <24 h and ⩾24 h, they showed that at lower baseline stroke risk (CHADS₂ of 1), only patients with AHRE durations above 24 h were at elevated risk (4% annually). Contrastingly, at CHADS₂ of ⩾2, the same 4% risk was observed with all AHRE >5 min.²²

Whilst there is significant variability in the reported effect thresholds, data generally demonstrate that increasing durations of AHRE/SCAF portend a higher stroke risk. All these studies are limited by the low event rates which is a major challenge in this patient population.

Anticoagulation in SCAF

Given the observed increase in stroke rates in patients with SCAF, the natural question is whether a net clinical benefit could be obtained with anticoagulation. The argument for starting anticoagulant therapy, therefore, is stronger for patients with long durations of AHREs with additional stroke risk factors, especially when multiple factors are present. The 2020 ESC guidelines advocate consideration of anticoagulation in patients with event durations ⩾24 h, high CHA₂DS₂-VASc scores (⩾2 in males or ⩾3 in females), where there is no diagnostic doubt of AF on the examined device tracings, and when bleeding risks are judged to be low.² Nevertheless, these guidelines noted a lack of data to make specific recommendations and called for large randomised trials in this arena.

Two trials, both published in late 2023, set out to answer this specific question. The first, NOAH-AFNET 6, recruited over 2500 patients aged ⩾65 years with relatively high predicted stroke risk (average CHA₂DS₂-VASc score of 4), randomising them 1:1 to the DOAC edoxaban or placebo. The mean duration of AHREs was 2.8 h. The trial was prematurely terminated due to higher rates of major bleeding in the edoxaban arm without any clinical benefit detected, possibly limiting its ability to detect a small benefit of edoxaban.¹⁴

Secondly, the ARTESIA trial randomised around 4000 patients to either apixaban or aspirin therapy. The majority of the included patients had AHREs lasting 6 min to 6 h, and the mean CHA₂DS₂-VASc score was 3.9. In contrast to NOAH-AFNET 6, this study did register a reduction in stroke risk in the apixaban arm (0.78% vs 1.24% per patient-year, HR: 0.63, p = 0.007); however, this came at the expense of a concurrent increase in major bleeding (1.71% vs 0.94%, HR: 1.80, p = 0.001).¹⁵ Furthermore, the difference in bleeding is likely to have been lessened by the choice of aspirin as the control arm, which is known to increase bleeding (around half of patients in NOAH-AFNET 6 were also taking aspirin).²³

Despite these two studies being seemingly at odds with one another, a subsequent meta-analysis found the results to be consistent (I² heterogeneity = 0%). McIntyre et al. reported that when taken together, DOAC therapy reduced the risk of ischaemic stroke by 32% and increased major bleeding by 62%.²⁴

At first glance, the figures would suggest net clinical harm associated with anticoagulation in AHREs. However, the impact of the events may not be equivalent. As Svennberg points out in her editorial, DOAC therapy prevented severe strokes (likely avoiding high levels of disability) and caused mostly non-fatal bleeding.²⁵ Indeed, in ARTESIA, there was no difference in fatal bleeding, and the vast majority of major bleeding events required supportive or conservative treatment only.

Taken together, these data would support a clinical practice similar to that recommended by contemporary guidelines, justifying the initiation of anticoagulant medications in patients with high CHA₂DS₂-VASc scores with relatively low bleeding risk, if observed durations of SCAF are longer.

AF triggered by a significant stressful event

Secondary AF – triggered or unmasked?

Commonly in clinical practice, the first diagnosis of AF is made incidentally during admission to secondary care for another reversible pathology, such as sepsis, surgery, thyrotoxicosis or acute pulmonary embolism.²⁶ Clinicians often consider AF in this scenario to be a secondary phenomenon with a high chance of resolution after treatment of the precipitating condition, but long-term stroke risk is increased after first-diagnosed AF in the context of sepsis,^26,27 non-cardiac²⁸ and cardiac surgery.²⁹

These data would imply an alternative framing of the situation – that patients with a predisposition towards AF (or pre-existing SCAF) have their condition unmasked by the stimulus and would likely have later developed the condition regardless.

Nevertheless, the evidence base for anticoagulation decision-making for this patient group is not robust, and this is reflected in the wide variation in clinical practice for in-hospital and post-discharge anticoagulation initiation. A UK-wide survey of predominantly intensive care physicians indicated that around two-thirds do not routinely anticoagulate critically ill patients with new-onset AF, whereas one-third anticoagulate within 72 h.³⁰

Of the secondary triggers, the commonest with the largest evidence base are sepsis and surgical procedures which are reviewed below.

New-onset AF associated with sepsis

Infection is one of the most common triggers of new-onset AF, accounting for over 20% of cases associated with a secondary precipitant.²⁶ Pathophysiologic hypotheses for the association include electrolyte derangement, atrial stretch due to dysregulated volume status and increased myocyte automaticity in the setting of severe inflammation.³¹ Indeed, the degree of inflammation, as measured by serum C-reactive protein, is correlated with the risk of developing AF during infection.³²

Inflammation induces a hypercoagulable state³³; thus, it is unsurprising that new-onset AF during sepsis is associated with higher rates of in-hospital thromboembolic events, including stroke. Walkey et al., in their retrospective analysis of almost 50,000 patients with severe sepsis, found that new-onset AF increased the likelihood of in-hospital stroke to 2.7% compared with 0.6% in patients without AF, as well as in-hospital mortality.³⁴ However, initiation of anticoagulation during the index admission is not without risk: two studies have shown higher rates of bleeding complications in the short term.^35,36

Studies that examine longer-term risks associated with sepsis-driven AF suggest that the deleterious effects persist many years after hospital discharge. Firstly, new-onset AF during admission for infection increases the likelihood of subsequent AF. Reported effect size varies with HRs as high as 26.0 in one Danish registry study³⁷ and 5.4 in another analysis on US Medicare recipients.²⁷ The latter also noted a 54.9% rate of post-discharge AF diagnosis at 5 years, correlating with findings from the Framingham Heart Study (5-, 10- and 15-year recurrence rates for secondary AF of 42%, 56% and 62%, respectively).

Additionally, rates of ischaemic stroke and other thromboembolic events remain high following sepsis complicated by AF. Gundlund et al. described a doubling in long-term thromboembolic risk in sepsis survivors with AF compared to patients with sepsis but without AF,³⁷ whereas Walkey et al. showed a 22% increased risk of ischaemic stroke at 5-year follow-up.²⁷ A retrospective chart review of patients in Taiwan found new-onset AF during septicaemia to increase the stroke risk at a mean 4.6-year follow-up (HR: 1.88, 95% CI: 1.33–2.65).³⁸ Finally, a meta-analysis of 12 studies performed by Arero et al. found new-onset AF to be a significant risk factor for stroke following sepsis hospitalisation (adjusted HR: 1.80, 95% CI: 1.42–2.28).³⁹

We have seen that around half of cases of sepsis-driven AF recur and that long-term outcomes, especially stroke, are poor. However, all the aforementioned studies are observational and there is, to date, no evidence from randomised trials for the use of OAC in sepsis-driven AF. Ultimately, this leads to the perception of sepsis-driven AF as a marker of disease severity rather than a pathogenic mediator. Furthermore, there are likely to be short-term net harms from early anticoagulation during sepsis. As such, there is no clear guidance issued by major societies for this patient group.

A reasonable strategy is set out by Induruwa et al. in their review.³¹ Individuals should undergo stroke risk assessment via the CHA₂DS₂-VASc score as well as an assessment of bleeding risk factors. Where there are additional stroke risk factors, patients should in most cases be offered anticoagulation following a frank discussion of the risks and benefits.⁴⁰ In borderline cases, extended cardiac monitoring is valuable given the high chance of recurrence.

Moreover, given the paucity of data for anticoagulation, the authors suggest a greater focus on cardiovascular risk factor management – as described in the Atrial Fibrillation Better Care (ABC) pathway, which represents the current holistic or integrated care approach to AF management that has been associated with improved clinical outcomes.^41,42 This is important given the clinical complexity associated with AF, as these patients are often frail and have multimorbidity and pharmacy, with implications for treatments and clinical outcomes.^41,43,44

Consistently, a 2023 consensus statement by the American Heart Association advocates a higher than usual threshold for anticoagulation initiation following acute AF (a term substituted by the authors for secondary AF) – with CHA₂DS₂-VASc ⩾2 for men and ⩾3 for women.⁴⁵ This is reflective of the trajectory of the field – recognition of sepsis-driven AF as an important clinical entity but acknowledgement that randomised trial data are so far insufficient to justify more aggressive anticoagulation strategies.

Post-operative AF

Surgery and the peri-operative period involve a major physiologic stress and frequently it is complicated by new-onset post-operative AF (POAF). Typically POAF occurs between post-operative days 2 and 4⁴⁶ and is particularly common following cardiac surgery, which involves manipulation and frequent cannulation of atrial anatomic structures.

Previously, incidence has been reported as 30%–50% for cardiac surgery (with variation depending on the type of operation)^46,47 and 5%–10% for major non-cardiac surgery.⁴⁸ Recent studies report more conservative figures, with incidences of 18.8% for cardiac surgery and 0.8% for non-cardiac surgery,⁴⁹ which may reflect increased use of prophylactic measures as described in contemporary guidelines, or equally under-detection of shorter, asymptomatic episodes.²

POAF following cardiac surgery

AF is the most common complication of cardiac surgical procedures and frequently presents acute challenges involving haemodynamic instability and heart failure.⁴⁶ Though incompletely understood, putative contributory factors include inflammation of the pericardium, intraoperative volume shifts and alterations in the neurohormonal milieu post-operatively.

POAF is strongly associated with myriad post-operative complications, such as reoperation, development of sepsis, need for prolonged ventilatory weaning or dialysis and mortality. It is unclear whether this represents a different risk profile or whether AF plays a causal role in these negative outcomes.⁵⁰

The relationship with ischaemic stroke, however, is less clear, and it appears that POAF does not carry the same long-term stroke risk as non-surgical AF.⁵¹ Estimates of the effect size vary significantly. A meta-analysis of four studies including approximately 9000 patients reported a four-fold increase in the risk of ischaemic stroke associated with POAF.⁵²

In a large cohort study of over 1,700,000 patients, Gialdini et al. found a more modest increase in stroke risk (HR: 1.3, 95% CI: 1.1–1.6) with relatively low overall rates (0.99% in the AF group), corresponding with findings from Lin et al. who demonstrated a 20% increased risk.^53,54 Contrastingly, Bianco et al. found no significant difference in stroke (2.76% vs 2.32% with and without POAF, respectively, p = 0.191). The explanation for the variability is uncertain, but it seems likely that at least a small difference does exist, and some investigators believe this reflects a subset of POAF patients who may more closely resemble ‘non-operative AF’ and benefit from conventional anticoagulation strategies.

On the flip side, anticoagulation is a high-risk intervention in this patient group, particularly if initiated early. It is associated with an increased risk of significant complications like cardiac tamponade, as well as other bleeding events.⁵⁵ Additionally, the efficacy of OAC in reducing stroke has not been demonstrated. A large observational study of OAC for POAF following coronary artery bypass grafting noted an associated increase in bleeding (HR: 1.60, 95% CI: 1.38–1.85) and mortality (HR: 1.16, 95% CI: 1.06–1.26) without any accompanying reduction in stroke rates (HR: 0.97, 95% CI: 0.82–1.15), even in patients with high CHA₂DS₂-VASc scores (>5).⁵⁶ These findings are limited by the observational nature of the study and are hence subject to bias – given the lack of evidence, clinician behaviour may lead OAC medications to be commenced only in patients with very perceived risks, who are likely to be older and more comorbid.

The 2020 ESC guidelines recommend that OAC may be considered for POAF after cardiac surgery, after considering the stroke and bleeding risk profiles of the patient (IIb recommendation, level of evidence B). In practice, given the lack of randomised trials, a conservative approach is common, reserving anticoagulation for patients with a significant expected net benefit.²

POAF following non-cardiac surgery:

Non-cardiac operations are also frequently complicated by new-onset AF, though not as commonly as after cardiac surgery.⁴⁹ The difference may be explained by a lower burden of cardiovascular comorbidities, as well as less direct impact of surgery on atrial myocardium. Proposed arrhythmic triggers in the peri-operative period include volume loss, inflammatory cytokine release and high concentrations of circulating catecholamines.⁴⁵ POAF is typically short-lived and around 40% of cases revert to sinus rhythm without any treatment within the admission,⁵⁷ but similarly to other forms of secondary AF there is increasing recognition that the long-term trajectory is not benign, associated with incident heart failure, myocardial infarction, stroke and mortality.^49,58

The largest synthesis in this area has been carried out by Lin et al.; in their 2019 meta-analysis totalling 2.5 million patients, the authors found POAF after non-cardiac surgery to be more strongly associated with long-term stroke risk (HR: 2.0, 95% CI: 1.70–2.35) than following cardiac surgery (HR: 1.2, 95% CI: 1.07–1.34). The only recent data not included in this analysis come from a cohort study by Butt et al.⁵⁹ The authors found comparable stroke risk with POAF and non-surgical AF around 3% risk per year, though the high event rate and very low incidence of POAF (0.4%) relative to other studies are notable, and may suggest under-detection of short-lived, lower-risk cases.

As to whether anticoagulation is of benefit to these patients, there are very limited data to guide decision-making. A relatively small meta-analysis of 29,566 patients with POAF related to non-cardiac surgery was recently published by Neves et al.⁶⁰ The key findings were that anticoagulation use was associated with a non-significant reduction in thromboembolic risk (odds ratio (OR): 0.71, 95% CI: 0.33–1.15) but a significant increase in bleeding (OR: 1.2, 95% CI: 1.10–1.32), and no effect on mortality. It is very difficult to make robust inferences from observational data, and there is a need for randomised trials of this patient group.

As seen with the examples of ‘secondary AF’ thus far, there is a perceived spectrum of intensity for any particular triggering stimulus for AF. The more intense the stressor, the less likely that the AF is a result of pre-existing arrhythmogenic substrate and so lesser is the long-term-associated risk. Cardiac surgery seems to exist at one end of the spectrum such that the outcome profile of POAF following cardiac procedures does not seem to resemble that of non-secondary AF.

On the other hand, POAF after non-cardiac operations appears to confer long-term risks more similar to those of non-secondary AF. However, history tells us that many logical concepts do not survive first contact with a randomised trial, and inferring a net benefit of OAC to negate the stroke risks in these populations is risky without such data. Currently, guidelines recommend that OAC should be considered for POAF after non-cardiac surgery, after considering the stroke and bleeding risk profiles of the patient (IIb recommendation, level of evidence B).²

Anticoagulation conundrums in patients with recent stroke

Anticoagulation timing after acute ischaemic stroke in patients with AF

The optimal timing for (re-)initiating OAC following an acute ischaemic stroke in patients with AF remains uncertain, and balancing the risk of recurrent stroke against the competing risk of secondary haemorrhagic transformation is crucial. In the study by Chang et al., patients with severe stroke were at higher risk of haemorrhagic transformation following early OAC initiation compared with late, whereas no increased risk was observed in patients with mild-moderate stroke.⁶¹

Current European guidelines recommend the (re-)initiation of OAC 1–3 days after a TIA, depending on brain imaging findings, or at ⩾3, ⩾6–8 or ⩾12–14 days after a mild, moderate or severe ischaemic stroke, respectively, in the absence of haemorrhagic transformation.⁶²

The Early Versus Delayed Non-Vitamin K Antagonist Oral Anticoagulant Therapy After Acute Ischaemic Stroke in Atrial Fibrillation (TIMING) trial assessed the noninferiority of early versus late initiation of DOACs.⁶³ Patients with acute stroke were randomised within 72 h of symptom onset to early (⩽4 days) or delayed (5–10 days) DOAC initiation. The primary outcome – a composite of recurrent ischaemic stroke, symptomatic ICH or all-cause death within 90 days – showed early initiation was non-inferior to delayed initiation. Notably, early initiation was associated with numerically lower rates of ischaemic stroke and death, with no symptomatic ICH recorded in either group during the 90-day follow-up. However, the study was underpowered and did not stratify patients according to stroke lesion severity.

The Early versus Late Initiation of Direct Oral Anticoagulants in Post-ischaemic Stroke Patients with Atrial Fibrillation (ELAN) trial, an international multicentre study, compared early (within 48 h after a minor or moderate stroke or on days 6 or 7 after a major stroke) versus late (days 3 or 4 after a minor stroke, days 6 or 7 after a moderate stroke or days 12, 13 or 14 after a major stroke) initiation of DOAC in patients with AF.⁶⁴ No statistically significant difference was observed between the groups regarding the primary outcome – a composite of recurrent ischaemic stroke, systemic embolism, major extracranial bleeding, symptomatic ICH or vascular death within 30 days after randomisation. Similarly, no difference was found in the risk of recurrent ischaemic stroke, ICH or vascular death at 30 and 90 days. These data suggest that early DOAC use after ischaemic stroke does not increase harm, potentially indicating an important shift in clinical practice and guidelines.

The Optimal Timing of Anticoagulation after Acute Ischemic Stroke (OPTIMAS) trial, registered on ClinicalTrials.gov (NCT03759938), is underway to provide more evidence on the topic above.

Anticoagulation decision-making in patients with AF and recent ICH

The evidence base to support OAC decision-making for ischaemic stroke prevention in patients with AF and recent ICH is limited. One meta-analysis published by Ivany et al. indicated that OAC effectively reduces the risk of ischaemic stroke in patients with AF post-ICH (relative risk (RR): 0.51, 95% CI: 0.30–0.86, p = 0.01), without significant elevating the risk of recurrent ICH (RR: 1.44, 95% CI: 0.38–5.46, p = 0.59).⁶⁵ However, the studies underlying these findings exhibit methodological heterogeneity, including differences in participant inclusion/exclusion criteria, follow-up durations and types of OAC utilised.

The principal obstacle to OAC therapy for stroke prevention in patients with AF and ICH is the concern and uncertainty over major bleeding risks.⁶⁶ A survey among clinicians revealed that fears of recurrent ICH were the predominant reason for hesitance to initiate or resume OAC in patients with AF post-ICH, with 63% of clinicians expressing apprehension about major bleeding.^67,68 This anxiety is partially mirrored by patients with AF who have recently experienced ICH.⁶⁹

This variability in clinical practice regarding OAC use for ischaemic stroke prevention in patients with AF post-ICH primarily stems from the scarcity of robust evidence on the topic.⁶⁷

Breakthrough ischaemic stroke in patients with AF already on oral anticoagulation

Despite advancements in stroke prevention, patients with AF maintain an annual ischaemic stroke risk of 1%–3% even when on effective OAC treatment.⁷⁰ A pooled analysis of seven prospective cohort studies revealed that AF patients experiencing stroke while on OAC had a higher risk of stroke recurrence than OAC-naïve patients, despite comparable CHA₂DS₂-VASc and HAS-BLED scores.⁷¹ The authors also found no benefit to changing the OAC type post-index event on the risk of subsequent strokes. This highlights the challenge of managing breakthrough strokes, where the optimal antithrombotic strategy remains uncertain.

Ip et al. investigated antithrombotic strategies for AF patients on DOACs at the time of an ischaemic stroke.⁷² The study evaluated four strategies: continuation of the same DOAC (DOAC-same), switching from DOAC-to-warfarin, switching between DOACs (DOAC-switch) and adding antiplatelet agents. The study found that continuing the same DOAC was associated with the lowest annual risk of recurrent stroke (8.7%). By contrast, both DOAC-switch and DOAC-to-warfarin switch strategies were linked to a higher risk of recurrent stroke compared to the DOAC-same strategy (adjusted hazard ratio (aHR): 1.96, 95% CI: 1.29–3.02, p = 0.002, and aHR: 1.62, 95% CI: 1.25–2.11, p < 0.001, respectively). Adding antiplatelet treatment to the DOAC-same strategy did not significantly reduce the risk of recurrent ischaemic stroke (aHR: 1.28, 95% CI: 0.88–1.84, p = 0.188), ICH (aHR: 1.20, 95% CI: 0.54–2.68, p = 0.654) or death (aHR: 1.09, 95% CI: 0.84–1.41, p = 0.512). Furthermore, there were no significant differences in the risk of ICH and death among the groups.

These data indicate that continuation of the same OAC is preferable following a breakthrough ischaemic stroke.

Conclusion

AF is a very diverse condition with a wide spectrum of associated risks. The benefits of OAC in routine practice for patients with AF are well established; however, many grey areas exist where estimating the balance of risk and benefit is challenging. We have reviewed short runs of SCAF (detected by CIEDs or wearable monitors), AF secondary to sepsis or the peri-operative period, as well as common dilemmas relating to recent stroke or ICH.

SCAF carries a modestly enhanced risk of stroke when compared with clinical AF, and the benefits of anticoagulation are smaller. A tailored approach is therefore required – with OAC reserved for patients with longer episodes (approaching 24 h), with higher CHA₂DS₂-VASc scores, and lower estimated bleeding risk.

AF triggered by sepsis recurs in over 50% of patients and carries a greater risk of ischaemic stroke. The consensus opinion is that patients with particularly high stroke risk (CHA₂DS₂-VASc ⩾2 for men and ⩾3 for women) should warrant consideration of anticoagulation.

Patients with POAF seem to derive less benefit from anticoagulation, despite the increase in stroke. Data supporting the routine use of OAC is currently lacking, and guidelines advocate consideration of OAC initiation only if the balance of risks is clearly favourable. Cardiac surgery-related POAF appears to be a special case – it is induced very frequently and the associated risks appear to be much lower. OAC use in this population is rarely indicated.

The evidence base for secondary AF is plagued by a lack of randomised clinical trials and therefore subject to possible bias and confounding. Higher-risk interventions like OAC should therefore have a higher threshold for initiation, with more focus on other aspects of the holistic ‘ABC’ care pathway as set out in the European guidelines.

Recent evidence suggests that early (re-)initiation of OAC for secondary prevention in patients with AF does not increase the risk of recurrent ischaemic stroke or ICH. Breakthrough strokes in patients with AF who were already on DOAC pose an important clinical dilemma, due to multiple potential competing stroke aetiologies, but changing to an alternative OAC appears to be deleterious. There is a paucity of evidence to guide the decision for OAC in patients with AF and recent ICH.

Footnotes

Acknowledgements

None.

Declarations

ORCID iD

Michael Griffin

References

Elliott

Middeldorp

Van Gelder

, et al. Epidemiology and modifiable risk factors for atrial fibrillation. Nat Rev Cardiol 2023; 20: 404–417

Hindricks

Potpara

Dagres

, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J 2021; 42: 373–498

Chao

Joung

Takahashi

, et al. 2021 Focused update consensus guidelines of the Asia Pacific Heart Rhythm Society on stroke prevention in atrial fibrillation: executive summary. Thromb Haemost 2022; 122: 20–47

Sun

Freedman

Martinez

, et al. Atrial fibrillation detected by single time-point handheld electrocardiogram screening and the risk of ischemic stroke. Thromb Haemost 2022; 122: 286–294

Gorenek

Bax

Boriani

, et al. Device-detected subclinical atrial tachyarrhythmias: definition, implications and management-an European Heart Rhythm Association (EHRA) consensus document, endorsed by Heart Rhythm Society (HRS), Asia Pacific Heart Rhythm Society (APHRS) and Sociedad Latinoamericana de Estimulación Cardíaca y Electrofisiología (SOLEACE). Europace 2017; 19: 1556–1578

Noseworthy

Kaufman

Chen

, et al. Subclinical and device-detected atrial fibrillation: pondering the knowledge gap: a scientific statement from the American Heart Association. Circulation 2019; 140: E944–E963.

Israel

Grönefeld

Ehrlich

, et al. Long-term risk of recurrent atrial fibrillation as documented by an implantable monitoring device: implications for optimal patient care. J Am Coll Cardiol 2004; 43: 47–52

Page

Wilkinson

Clair

, et al. Asymptomatic arrhythmias in patients with symptomatic paroxysmal atrial fibrillation and paroxysmal supraventricular tachycardia. Circulation 1994; 89: 224–227

Boriani

Laroche

Diemberger

, et al. Asymptomatic atrial fibrillation: clinical correlates, management, and outcomes in the EORP-AF Pilot General Registry. Am J Med 2015; 128: 509–518.e2.

10.

Gillis

Morck

Atrial fibrillation after DDDR pacemaker implantation. J Cardiovasc Electrophysiol 2002; 13: 542–547.

11.

Healey

Connolly

Gold

, et al. Subclinical atrial fibrillation and the risk of stroke. N Engl J Med 2012; 366: 120–129.

12.

Glotzer

Hellkamp

Zimmerman

, et al. (2003) Atrial high rate episodes detected by pacemaker diagnostics predict death and stroke: report of the Atrial Diagnostics Ancillary Study of the MOde Selection Trial (MOST). Circulation 2003; 107: 1614–1619.

13.

Mahajan

Perera

Elliott

, et al. (2018) Subclinical device-detected atrial fibrillation and stroke risk: a systematic review and meta-analysis. Eur Heart J 2018; 39: 1407–1415.

14.

Kirchhof

Toennis

Goette

, et al. Anticoagulation with edoxaban in patients with atrial high-rate episodes. N Engl J Med 389:1167–1179.

15.

Healey

Lopes

Granger

, et al. Apixaban for stroke prevention in subclinical atrial fibrillation. N Engl J Med 2023; 390(2): 107–117.

16.

Friberg

Rosenqvist

Lip

GYH

(2012) Evaluation of risk stratification schemes for ischaemic stroke and bleeding in 182 678 patients with atrial fibrillation: the Swedish Atrial Fibrillation cohort study. Eur Heart J 2012; 33: 1500–1510.

17.

Swiryn

Orlov

Benditt

, et al. Clinical implications of brief device-detected atrial tachyarrhythmias in a cardiac rhythm management device population: results from the Registry of Atrial Tachycardia and Atrial Fibrillation Episodes. Circulation 2016; 134: 1130–1140.

18.

Glotzer

Daoud

Wyse

, et al. The relationship between daily atrial tachyarrhythmia burden from implantable device diagnostics and stroke risk: the TRENDS study. Circ Arrhythm Electrophysiol 2009; 2: 474–480.

19.

Boriani

Glotzer

Santini

, 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). Eur Heart J 2014; 35: 508–516

20.

Proietti

Labos

AlTurki

, et al. Asymptomatic atrial fibrillation burden and thromboembolic events: piecing evidence together. Expert Rev Cardiovasc Ther 2016; 14: 761–769.

21.

Pastori

Miyazawa

, et al. Atrial high-rate episodes and risk of major adverse cardiovascular events in patients with cardiac implantable electronic devices. Clin Res Cardiol 2020; 109: 96–102

22.

Botto

Padeletti

Santini

, et al. Presence and duration of atrial fibrillation detected by continuous monitoring: crucial implications for the risk of thromboembolic events. J Cardiovasc Electrophysiol 2009; 20: 241–248

23.

Rodríguez

LAG

Martín-Pérez

Hennekens

, et al. Bleeding risk with long-term low-dose aspirin: a systematic review of observational studies. PLoS One 2016; 11: e0160046.

24.

McIntyre

Benz

Becher

, et al. (2023) Direct oral anticoagulants for stroke prevention in patients with device-detected atrial fibrillation: a study-level meta-analysis of the NOAH-AFNET 6 and ARTESiA trials. Circulation 2024; 149: 981–988.

25.

Svennberg

What lies beneath the surface – treatment of subclinical atrial fibrillation. N Engl J Med 2023; 390: 175–177.

26.

Lubitz

Yin

Rienstra

, et al. Long-term outcomes of secondary atrial fibrillation in the community: the Framingham Heart Study. Circulation 2015; 131: 1648–1655.

27.

Walkey

Hammill

Curtis

, et al. Long-term outcomes following development of new-onset atrial fibrillation during sepsis. Chest 2014; 146: 1187–1195

28.

Albini

Malavasi

Vitolo

, et al. Long-term outcomes of postoperative atrial fibrillation following non cardiac surgery: a systematic review and metanalysis. Eur J Intern Med 2021; 85: 27–33

29.

Wang

Meyre

Heo

, et al. (2022) Short-term and long-term risk of stroke in patients with perioperative atrial fibrillation after cardiac surgery: systematic review and meta-analysis. CJC Open 2022; 4: 85.

30.

Chean

McAuley

Gordon

, et al. Current practice in the management of new-onset atrial fibrillation in critically ill patients: a UK-wide survey. PeerJ 2017; 5: e3716.

31.

Induruwa

Hennebry

, et al. Sepsis-driven atrial fibrillation and ischaemic stroke. Is there enough evidence to recommend anticoagulation? Eur J Intern Med 2022; 98: 32

32.

Klein Klouwenberg

PMC

Frencken

Kuipers

, et al. Incidence, predictors, and outcomes of new-onset atrial fibrillation in critically ill patients with sepsis: a cohort study. Am J Respir Crit Care Med 2017; 195: 205–211

33.

Semeraro

Ammollo

Semeraro

, et al. Coagulopathy of acute sepsis. Semin Thromb Hemost 2015; 41: 650–658

34.

Walkey

Wiener

Ghobrial

, et al. Incident stroke and mortality associated with new-onset atrial fibrillation in patients hospitalized with severe sepsis. JAMA 2011; 306: 2248–2255

35.

Darwish

Strube

Nguyen

, et al. (2013) Challenges of anticoagulation for atrial fibrillation in patients with severe sepsis. Ann Pharmacother 2013; 47(10): 1266–1271.

36.

Walkey

Quinn

Winter

, et al. Practice patterns and outcomes associated with use of anticoagulation among patients with atrial fibrillation during sepsis. JAMA Cardiol 2016; 1: 682–690.

37.

Gundlund

Olesen

Butt

, et al. One-year outcomes in atrial fibrillation presenting during infections: a nationwide registry-based study. Eur Heart J 2020; 41: 1112–1119.

38.

Cheng

Lee

, et al. An analysis of long-term ischemic stroke risk in survivors of septicemia. J Stroke Cerebrovasc Dis 2017; 26: 2893–2900

39.

Arero

Vasheghani-Farahani

Tigabu

, et al. Long-term risk and predictors of cerebrovascular events following sepsis hospitalization: a systematic review and meta-analysis. Front Med (Lausanne) 2022; 9: 1065476.

40.

Nelson

Johnston

Waite

AAC

, et al. A systematic review of anticoagulation strategies for patients with atrial fibrillation in critical care. Thromb Haemost 2021; 121: 1599–1609

41.

Romiti

Pastori

Rivera-Caravaca

, et al. Adherence to the ‘Atrial Fibrillation Better Care’ pathway in patients with atrial fibrillation: impact on clinical outcomes: a systematic review and meta-analysis of 285,000 patients. Thromb Haemost 2022; 122: 406–414.

42.

Romiti

Guo

Corica

, et al. Mobile health-technology-integrated care for atrial fibrillation: a win ratio analysis from the mAFA-II randomized clinical trial. Thromb Haemost 2023; 123: 1042–1048.

43.

Grymonprez

Petrovic

De Backer

, et al. (2024) The impact of polypharmacy on the effectiveness and safety of non-vitamin K antagonist oral anticoagulants in patients with atrial fibrillation. Thromb Haemost 2024; 124: 135–148.

44.

Zheng

Liu

, et al. (2023) Effect of oral anticoagulants in atrial fibrillation patients with polypharmacy: a meta-analysis. Thromb Haemost. Epub ahead of print July 2023. DOI: 10.1055/S-0043-1770724.

45.

Chyou

Barkoudah

Dukes

, et al. Atrial fibrillation occurring during acute hospitalization: a scientific statement from the American Heart Association. Circulation 2023; 147: E676–E698.

46.

Echahidi

Pibarot

O’Hara

, et al. (2008) Mechanisms, prevention, and treatment of atrial fibrillation after cardiac surgery. J Am Coll Cardiol 2008; 51: 793–801.

47.

Gillinov

Bagiella

Moskowitz

, et al. Rate control versus rhythm control for atrial fibrillation after cardiac surgery. N Engl J Med 2016; 374: 1911–1921.

48.

Philip

Berroëta

Leblanc

Perioperative challenges of atrial fibrillation. Curr Opin Anaesthesiol 2014; 27: 344–352.

49.

Goyal

Kim

Krishnan

, et al. Post-operative atrial fibrillation and risk of heart failure hospitalization. Eur Heart J 2022; 43: 2971–2980.

50.

Bianco

Kilic

Yousef

, et al. The long-term impact of postoperative atrial fibrillation after cardiac surgery. J Thorac Cardiovasc Surg 2023; 166: 1073–1083.e10.

51.

Butt

Xian

Peterson

, et al. Long-term thromboembolic risk in patients with postoperative atrial fibrillation after coronary artery bypass graft surgery and patients with nonvalvular atrial fibrillation. JAMA Cardiol 2018; 3: 417–424.

52.

Eikelboom

Sanjanwala

, et al. Postoperative atrial fibrillation after cardiac surgery: a systematic review and meta-analysis. Ann Thorac Surg 2021; 111: 544–554.

53.

Gialdini

Nearing

Bhave

, et al. Perioperative atrial fibrillation and the long-term risk of ischemic stroke. JAMA 2014; 312: 616–622.

54.

Lin

Kamel

Singer

, et al. (2019) Perioperative/postoperative atrial fibrillation and risk of subsequent stroke and/or mortality. Stroke 2019; 50: 1364–1371.

55.

Meurin

Weber

Renaud

, et al. Evolution of the postoperative pericardial effusion after day 15: the problem of the late tamponade. Chest 2004; 125: 2182–2187

56.

Riad

Grau-Sepulveda

Jawitz

, et al. Anticoagulation in new-onset postoperative atrial fibrillation: An analysis from the Society of Thoracic Surgeons Adult Cardiac Surgery Database. Heart Rhythm O2 2022; 3:325

57.

Goldman

Supraventricular tachyarrhythmias in hospitalized adults after surgery. Clinical correlates in patients over 40 years of age after major noncardiac surgery. Chest 1978; 73: 450–454

58.

Alturki

Marafi

Proietti

, et al (2020) Major adverse cardiovascular events associated with postoperative atrial fibrillation after noncardiac surgery: a systematic review and meta-analysis. Circ Arrhythm Electrophysiol 2020; 13: e007437.

59.

Butt

Olesen

Havers-Borgersen

, et al. Risk of thromboembolism associated with atrial fibrillation following noncardiac surgery. J Am Coll Cardiol 2018; 72: 2027–2036.

60.

Neves

Magalhães

Lima da Silva

, et al. Anticoagulation therapy in patients with post-operative atrial fibrillation: systematic review with meta-analysis. Vascul Pharmacol 2022; 142: 106929.

61.

Chang

Wang

, et al. Oral anticoagulation timing in patients with acute ischemic stroke and atrial fibrillation. Thromb Haemost 2022; 122: 939–950.

62.

Steffel

Collins

Antz

, et al. 2021 European Heart Rhythm Association practical guide on the use of non-vitamin K antagonist oral anticoagulants in patients with atrial fibrillation. EP Europace 2021; 23: 1612–1676.

63.

Oldgren

Åsberg

Hijazi

, et al. Early versus delayed non-vitamin K antagonist oral anticoagulant therapy after acute ischemic stroke in atrial fibrillation (TIMING): a registry-based randomized controlled noninferiority study. Circulation 2022; 146: 1056–1066.

64.

Fischer

Koga

Strbian

, et al. Early versus later anticoagulation for stroke with atrial fibrillation. N Engl J Med 2023; 388: 2411–2421.

65.

Ivany

Ritchie

Lip

GYH

, et al. Effectiveness and safety of antithrombotic medication in patients with atrial fibrillation and intracranial hemorrhage: systematic review and meta-analysis. Stroke 2022; 53: 3035–3046.

66.

Gorog

Gue

Chao

, et al. Assessment and Mitigation of bleeding risk in atrial fibrillation and venous thromboembolism: executive summary of a European and Asia-Pacific expert consensus paper. Thromb Haemost 2022; 122: 1625–1652.

67.

Ivany

Lotto

Lip

GYH

, et al. Managing uncertainty: physicians’ decision making for stroke prevention for patients with atrial fibrillation and intracerebral hemorrhage. Thromb Haemost 2022; 122: 1603–1611.

68.

Ivany

Lane

Dan

, et al. Antithrombotic therapy for stroke prevention in patients with atrial fibrillation who survive an intracerebral haemorrhage: results of an EHRA survey. Europace 2021; 23: 806–814.

69.

Ivany

Lane

DA.

Patient satisfaction: a key component in increasing treatment adherence and persistence. Thromb Haemost 2021; 121: 255–257.

70.

Seiffge

Hooff

Nolte

, et al. Recanalization therapies in acute ischemic stroke patients: impact of prior treatment with novel oral anticoagulants on bleeding complications and outcome. Circulation 2015; 132: 1261–1269.

71.

Seiffge

De Marchis

Koga

, et al. Ischemic stroke despite oral anticoagulant therapy in patients with atrial fibrillation. Ann Neurol 2020; 87: 677.

72.

YMB

Lau

, et al. Association of alternative anticoagulation strategies and outcomes in patients with ischemic stroke while taking a direct oral anticoagulant. Neurology 2023; 101: E358–E369.

Challenging anticoagulation decisions in atrial fibrillation: a narrative review

Abstract

Keywords

Introduction

Subclinical AF and atrial high-rate episodes

Nomenclature

Frequency of SCAF

Stroke risk in subclinical versus clinical AF

Effect of event duration on stroke risk in SCAF

Anticoagulation in SCAF

AF triggered by a significant stressful event

Secondary AF – triggered or unmasked?

New-onset AF associated with sepsis

Post-operative AF

POAF following cardiac surgery

POAF following non-cardiac surgery:

Anticoagulation conundrums in patients with recent stroke

Anticoagulation timing after acute ischaemic stroke in patients with AF

Anticoagulation decision-making in patients with AF and recent ICH

Breakthrough ischaemic stroke in patients with AF already on oral anticoagulation

Conclusion

Footnotes

Acknowledgements

Declarations

ORCID iD

References