2025 WSO科学声明：房颤和近期缺血性脑卒中、TIA或颅内出血患者缺血性中风的预防

Abstract

Background: Secondary stroke prevention in patients with atrial fibrillation (AF) is one of the fastest growing areas in the field of cerebrovascular diseases. This scientific statement from the World Stroke Organization Brain & Heart Task Force provides a critical analysis of the strength of current evidence on this topic, highlights areas of current controversy, identifies knowledge gaps, and proposes priorities for future research.

Methods: We select topics with the highest clinical relevance and perform a systematic search to answer specific practical questions. Based on the strength of available evidence and knowledge gaps, we identify topics that need to be prioritized in future research. For this purpose, we adopt a novel classification of evidence strength based on the availability of publications in which the primary population is patients with recent ( ${<}6$ months) cerebrovascular events, the primary study endpoint is a recurrent ischemic stroke, and the quality of the studies (e.g. observational versus randomized controlled trial).

Summary: Priority areas include AF screening, molecular biomarkers, AF subtype classification, anticoagulation in device-detected AF, timing of anticoagulation initiation, effective management of breakthrough strokes on existing anticoagulant therapy, the role of left atrial appendage closure, novel approaches, and antithrombotic therapy post-intracranial hemorrhage. Strength of currently available evidence varies across the selected topics, with early anticoagulation being the one showing more consistent data.

Conclusion: Several knowledge gaps persist in most areas related to secondary stroke prevention in AF. Prioritizing research in this field is crucial to advance current knowledge and improve clinical care.

Keywords

Atrial fibrillation, stroke, transient ischemic attack, anticoagulants, prevention, detection, recommendation

Received: 12 October 2024; accepted: 20 December 2024

Introduction

Atrial fibrillation (AF) is a cardiac arrhythmia affecting approximately 59.7 million individuals globally as of 2019, which represents a $111\%$ increase from 1990. 1 Populationbased projections estimate a two- to threefold increase in the global prevalence of AF by 2050–2060 due to population growth, aging, and advanced AF detection methods. 2 AF is associated with a fivefold risk for ischemic stroke (IS), 3 is present in $18–30\%$ of acute IS cases, 4–6 and its prevalence in IS hospitalizations has increased to $22\%$ in North America in recent decades. 4,6 Several aspects of AF diagnosis and management have advanced significantly in the last decade. This position statement aims to review current evidence, classify its strength, and identify priority areas for future research.

Methods

The writing group selected relevant topics with clinical impact to be addressed in this document. We performed a systematic search for each topic (Supplemental Table S1). Statements were organized in sections focused on the diagnosis and management of AF patients with a recent IS, intracranial hemorrhage (ICH), or transient ischemic attack (TIA). Sections for which newer evidence was available or was more controversial were discussed more extensively than others. The aim was to evaluate the strength of current

evidence and identify knowledge gaps for future research instead of providing clinical recommendations. We implemented a novel classification of evidence focused on clinical needs for physicians managing patients with acute IS. As such, we classified levels of evidence based on whether data addressed patients with a recent cerebrovascular event defined as ${\leqslant}6$ months (as opposed to remote cerebrovascular events) before inclusion in randomized controlled trials (RCTs) or observational studies (Figure 1). The classification of strength of evidence also prioritized studies in which recurrent IS was the primary endpoint or a prespecified secondary endpoint. Members of the Writing Group and the World Stroke Organization Brain & Heart Task Force reviewed each statement and their level of evidence. If a co-author disagreed with a statement, the wording and level of evidence adjudication were revised until reaching a consensus. All authors approved the final version of each statement and level of evidence adjudication.

AF screening

AF is associated with AF recurrence and IS risk, and thus, prolonged cardiac monitoring (PCM) is used to screen for subclinical AF. In patients with IS and TIA, RCTs have shown significantly increased AF detection using external devices and implantable cardiac monitors (ICM) (Supplemental Table S2) than standard-of-care diagnostics. None of the RCTs on PCM was designed to test whether

Figure 1. Classification of levels of evidence. RCT: randomized controlled trial; IS/TIA/ICH: ischemic stroke, transient ischemic attack or intracranial hemorrhage.

|Thestatement'soutcomeisaprimarystudyendpoint| Primary population is patients with recent (≤6 m) IS/TIA/CH | Subgroup analysis for patients with recent (s6m)IS/TIA/CH | Primary population or subgroup analysis for remoteIS/TIA/ICH | |---|---|---|---| |Consistent fata from≥2RCTs ormeta-analyses of RCTs|01|A3|81| |Data from a single RCT|2||B2| | Consistent data from meta-analyses of observational studies, largeprospectiveobservationalstudies,orstrokeregistries |B1|B2|B3| |Data from ≥1 retrospective observational studies||02|3| |Thestatement'soutcome is a secondarystudyendpoint|||| |ConsistentDatafrom22RCTsormeta-analysesof RCTs|A2|A4|B4| |Data from a single RCT|A3|B2|| | Consistent data from meta-analyses of observational studies, largeprospectiveobservationalstudies,orstrokeregistries |B2|B3|| |Data from ≥1retrospective observational studies|02|C3|C4| |Insufficientdata|||| | Inconsistent,underpowered,orconflictingdata ORnostudies orsubgroupanalysesinpatientswithIS/TIA/ICH andAF |||| |Expert opinion|||| | Moredataareneeded,andnosolidexpertopinion Level A: strong.Level B:moderately strong. Level C:weaker.Levels D,E,F:conflicting or uncertain. ||||

PCM reduces IS recurrence, and all were underpowered to show a significant effect. A study-level meta-analysis of six clinical trials with 68,556 patient-months of follow-up showed no association between PCM use and IS recurrence (incidence rate ratio (IRR): 0.90; $95\%$ confidence interval (CI): 0.71–1.15), recurrent IS or TIA (IRR: 0.97; $95\%$ CI: 0.80–1.18), or recurrent IS/ICH/TIA (IRR: 0.99; $95\%$ CI: 0.80–1.20). 7 It must be noted that the type of cardiac monitoring (e.g. external vs implantable), duration (7days to $\cong\!3$ years), and timing of initiation (3days to 6months) were heterogeneous across studies. Two RCTs are currently evaluating whether different intensities of PCM reduce stroke risk in patients with a recent IS or TIA (NCT04371055, NCT05134454).

## Blood biomarkers for improving AF screening

Measuring blood biomarkers capable of identifying patients more likely to have PCM-detected AF could potentially

streamline AF screening. Blood biomarkers can be classified into cardiac, thrombotic, and inflammatory. 8

Cardiac biomarkers

Elevated cardiac troponin has been associated with increased AF detection (area under the curve (AUC): 0.660–0.697) in several observational studies (Supple mental Table S3). 9 Natriuretic peptides are released from the cardiac atria or ventricles under strain. 8,10 Although both N-terminal pro b-type natriuretic peptide (NT-proBNP) and midregional pro-atrial natriuretic peptide (MR-proANP) are associated with AF diagnosis post-stroke, 11–15 NT-proBNP is less atrial-specific than MR-proANP. In the BIOSIGNAL (Biomarker Signature of Stroke Aetiology) study, which prospectively measured MR-proANP in 1759 patients within 24 h of acute IS onset, $\mathrm{Log}_{10}\mathrm{MR}$ -proANP levels were strongly associated with new AF diagnosis (adjusted odds ratio (aOR): 35.3, $95\%$ CI: 17.6–71.0). 15 A simple model with age and

 Figure 2. AF risk based on the timing of detection and intensity of cardiac monitoring. AF: atrial fibrillation; ECG: 12-lead electrocardiogram; AFDAS: atrial fibrillation detected after stroke.

MR-proANP showed good discrimination (AUC: 0.810) and higher net benefit than existing clinical AF risk scores.

Thrombosis biomarkers

Anti-thrombin III, D-dimer, and the MOCHA profile (markers of coagulation and hemostatic activation, including serum d-dimer, prothrombin fragment 1.2, thrombin-antithrombin complex, and fibrin monomer) have been associated with new AF detection, underlying malignancy, and stroke recurrence, with a good predictive ability when associated with left atrial volume index (AUC: 0.800). 14,16 The AUC of thrombotic markers for AF detection was 0.700 in another study and appeared to be a stronger association with underlying malignancy and venous thromboembolism. 17

Inflammatory and novel biomarkers

In a larger systematic review and meta-analysis, there was only a non-significant trend toward association with AF detection among people with higher levels of C-reactive protein. 10 Novel biomarkers, including Bone morphogenic protein 10, 18 symmetric dimethylarginine, 19 and soluble suppression of tumorigenicity-2, 20 have been associated with AF detection in stroke patients but more evidence is needed. Cytokines (IL-4, IL-6, IL-10, tumor necrosis factor, interferon-gamma, etc.) have been associated with AF relative to sinus rhythm. IL-17 has been implicated in the pathogenesis of AF, 21 and IL-6 is associated with increased AF incidence in patients undergoing cardiac surgery 22,23 and with AF recurrence after electrical cardioversion. 24 We did not identify any studies evaluating the role of cytokines for predicting AF detection in patients with a recent IS or TIA.

Classification of AF subtypes

Stroke recurrence rates according to the timing of AF diagnosis

The timing of AF detection relative to stroke onset and the intensity of monitoring determine the characteristics of the detected AF, with a gradient of stroke risk ranging from very high in patients with AF known before stroke onset to significantly lower risk in PCM-detected AF. 25 AF known before stroke occurrence is detected incidentally on 12-lead electrocardiograms (ECGs) performed during routine physical examination or when patients become symptomatic before they experience a stroke. Therefore, by the time it is diagnosed on an ECG, it has matured long enough to become a symptomatic high-burden arrhythmia. In contrast, AF detected on opportunistic PCM pursued post-stroke is generally an earlier and lower-burden arrhythmia. 25

Based on meta-analyses of RCTs and observational studies, AF in patients with a recent IS or TIA has been categorized into three main subtypes based on the timing of AF diagnosis: AF known before stroke onset or “Known AF” (KAF), AF newly-detected post-stroke on 12-lead ECG, and AF detected after stroke (AFDAS) on PCM, ranging from short (24h or 48h Holter) to long term $\scriptstyle{\left\langle{\geqslant7}\right\rangle}$ days). 26 The rationale behind this categorization is that KAF has a higher prevalence of risk factors and vascular comorbidities, more severe left atrial substrate, greater AF burden, and higher risk of stroke recurrence than AFDAS. 7,27 AF newly detected on 12-ECGs post-stroke has a fivefold higher risk of stroke recurrence than PCM-detected AF and is considered a preexisting AF that remained undiagnosed until stroke occurrence despite being high-burden, with a risk profile similar to KAF. 28 Therefore, newly 12-lead ECG-detected AF at any time-point post-stroke should not be considered AFDAS and has a similar long-term risk of stroke as KAF (Figure 2). 26 AFDAS is always PCM-detected.

Prevention of recurrent ischemic stroke

The pillars of IS prevention in AF are the management of risk factors, anticoagulation, rate/rhythm control, and minimizing the risk of bleeding. Patients with a recent IS or TIA usually undergo PCM, which adds a layer of complexity due to the wide range of AF burden found in this population.

Management of risk factors

Strong evidence from RCTs supports that optimizing the control of risk factors is crucial for IS prevention, regardless of the presence of AF. 29 In patients with AF, the strongest evidence from RCTs has shown that physical activity, reducing alcohol intake, and treating hypertension, sleepdisordered breathing, obesity, and diabetes can reduce AF incidence and recurrence. 30,31 No specific RCT has assessed the effect of risk factor management on recurrent IS in AF patients with a recent IS or TIA.

RCTs of anticoagulants in ECG-detected AF

Robust evidence from multiple large RCTs and meta-analyses of RCTs demonstrates that vitamin K antagonists (VKA) reduce IS risk by approximately $67\%$ compared to placebo or no therapy, and the risk of stroke (ischemic and hemorrhagic) by $38\%$ relative to Aspirin. 32 In more recent RCTs, direct oral anticoagulants (DOACs) were at least as effective as VKAs for the prevention of IS (relative risk (RR): 0.92, $95\%$ CI: 0.83–1.02), resulting in a $52\%$ lower risk of ICH (RR: 0.48; $95\%$ CI: 0.39–0.59) and $19\%$ lower risk of stroke/SE (RR: 0.81; $95\%$ CI: 0.73–0.91) in patients with and without a remote stroke/TIA. 33

Secondary analyses of RCTs of anticoagulants in device-detected AF

NOAH-AFNET 6 (Non-vitamin K Antagonist Oral Anticoagulants in Patients With Atrial High Rate Episodes) reported neutral findings in a randomized trial comparing edoxaban $60\,\mathrm{mg}$ daily versus placebo or aspirin for the prevention of stroke, systemic embolism (SE), or cardiovascular death in patients 65years of age or older with subclinical device-detected AF lasting ⩾ 6min and at least one risk factor. 34 It was stopped early due to excess major bleeding with edoxaban and had a low number of stroke events, potentially limiting the trial’s power to detect differences in the primary efficacy outcome. In contrast, the ARTESiA (Apixaban for Stroke Prevention in Subclinical Atrial Fibrillation) trial reported superior prevention of stroke or SE with random assignment to apixaban 5mg twice daily compared with aspirin 81mg daily in patients 55years of age or older with subclinical device-detected AF lasting

6min to $24\,\mathrm{h}$ . 35 An aggregate meta-analysis of the two trials demonstrated that oral anticoagulation with these agents reduced IS risk (RR: 0.68, $95\%$ CI: 0.50–0.92) and reported consistent estimates of treatment effect between the two trials $(\mathrm{I}^{2}\!=\!0\%)$ . 36 However, less than $10\%$ of participants in these trials had a history of IS or TIA. Subanalyses from NOAH-AFNET 6 and ARTESiA comparing the effect of DOACs versus aspirin or placebo on IS recurrence risk in patients with remote IS or TIA were conflicting (Supplemental Table S4). In both trials, DOACs significantly increased major bleeding risk.

Concerns have been raised by experts 37–39 and recent guidelines 31 regarding a one-size-fits-all approach for anticoagulation in IS or TIA patients with device-detected subclinical AF lasting ${<}24\,\mathrm{~h~}$ . A more comprehensive and personalized approach considering the interplay of AF burden, atrial substrate, and time between stroke occurrence and AF diagnosis has been proposed for patients with AFDAS. For instance, the B 2 AD-RISK scheme, which comprises the longitudinal measurement of biomarkers (B), AF burden (B), atrial substrate (A), age and sex demographics (D), and risk factors (R), is currently being tested in a pilot study (NCT0658970). 26

Early rhythm control

The EAST-AFNET 4 (Early Treatment of Atrial Fibrillation for Stroke Prevention Trial) trial randomized 2789 patients with AF diagnosed within the previous 12months to early rhythm control (ERC) with antiarrhythmic drugs or ablation versus standard of care. The primary composite efficacy outcome of cardiovascular death, stroke, or hospitalization with worsening of heart failure or acute coronary syndrome was less frequent in the ERC group (hazard ratio: HR: 0.79, $95\%$ CI: 0.66–0.94). Patients receiving ERC had a lower risk of stroke than the control group (HR: 0.65, $95\%$ CI: 0.44–0.97). Several observational studies and a subanalysis of EAST-AFNET 4 in patients with prior IS or TIA have shown similar results. 40 A small open-label, randomized, multicenter RCT including 300 patients with acute IS and AF within 2months of stroke onset found lower recurrent IS rates in patients undergoing ERC than in those receiving usual care (HR: 0.251; $95\%$ CI: 0.063–1.003). 41 EAST-STROKE (Early Treatment of Atrial Fibrillation for Stroke Prevention Trial in Acute STROKE) will test a similar approach in patients with recent ischemic cerebrovascular events (NCT05293080).

Timing of initiation of anticoagulation

Clinicians considering early initiation of anticoagulation therapy must balance the potential benefit of improved recurrent stroke prevention on the one hand and the potential harm of symptomatic ICH on the other hand.

 Figure 3. Etiological investigation of breakthrough strokes. OAC: oral anticoagulant; AF: atrial fibrillation; LAA: left atrial appendage.

Observational studies found that DOAC therapy is initiated early (median within 4days) after a recent stroke in clinical routine, even in the absence of RCT data and despite more conservative historical guideline recommendations. 42–47 Several registry-based observational studies were conducted to answer the “timing question” in various populations (Supplemental Table S5). 42–47 These studies found no strong evidence of a heightened ICH risk in patients with early initiation of anticoagulation. However, most studies artificially split timing to make a comparison. The most relevant limitations are that early and late start were defined differently across studies, with thresholds ranging from ${\leqslant}2$ days to ${\leqslant}7$ days; they had retrospective designs with risk of confounding by indication, and had no standardized procedures for early or later treatment selection.

Three RCTs specifically addressed the topic of early or late initiation of DOACs. The TIMING (Timing of Oral Anticoagulant Therapy in Acute IS With Atrial Fibrillation) was a registry-based noninferiority RCT that randomized 888 IS patients (median NIHSS 4) with AF admitted within 72 h of symptom onset to either early ( ${\widetilde{\mathbf{\Gamma}}}{=}4$ days) or delayed (5–10days) start of DOAC treatment. 48 Early DOAC initiation was non-inferior to delayed start (IS rates: $3.1\%$ vs

$4.6\%$ ), and no patient in either group experienced a symptomatic ICH. ELAN (Early versus Late Initiation of Direct Oral Anticoagulants in Post-IS Patients with Atrial Fibrillation) is the largest RCT comparing early versus later initiation of DOAC treatment in AF-related IS. 49 The time frame for early or late start of DOAC treatment was defined according to the infarct size on neuroimaging. 49 Patients with minor or moderate stroke randomized to early initiation of DOACs were started within $48\;\mathrm{h}$ , and patients with major stroke on day 6 or 7 $\mathbf{\tilde{n}}\!=\!1006$ . The primary outcome, a composite of recurrent IS, SE, major extracranial bleeding, symptomatic ICH, or vascular death within 30days, occurred in $2.9\%$ versus $4.1\%$ in the early and late groups, respectively. Numerically, fewer patients in the early group had recurrent IS within 30days ( $1.4\%$ vs $2.5\%$ ). Two patients in each group had symptomatic ICH. The OPTIMAS (Optimal Timing of Anticoagulation After Acute Ischaemic Stroke) trial was a phase 4, multicenter, parallel-group, RCT applying an open-label intervention and blinded endpoint adjudication. 50 It used a hierarchical non-inferiority-superiority gatekeeper design (sequentially assessing a non-inferiority margin of 2 percentage points and then proceeding to test for superiority) to compare

early initiation of DOACs (within 4days after stroke onset) versus delayed initiation (7–14days following stroke onset) in 3621 patients with AF and IS. 50 The primary endpoint was a composite of recurrent IS, symptomatic ICH, stroke of unknown type, or systemic embolism at 90days in a modified intention-to-treat analysis. Early DOAC initiation within 4days post-IS was noninferior to delayed initiation for the composite primary endpoint. Early initiation was not superior to late initiation.

CATALYST (CollAboration on the optimal Timing of anticoagulation after ischaemic stroke and Atrial fibrillation: prospective individuaL participant data meta-analYsiS of randomized controlled Trials) is an individual participant data meta-analysis of RCTs investigating the optimal timing of DOAC initiation after acute IS in patients with AF. CATALYST included data from 5411 patients from TIMING, ELAN, OPTIMAS, and START. 51 The primary endpoint was a composite of recurrent IS, symptomatic intracerebral hemorrhage, or unclassified stroke at 30days. Early DOAC initiation (within 4days) was superior to later initiation ( $\mathord{\leftr{\vphantom{\left|\mathcal{D}_{\infty}\right|}}\rightleftharpoons}5$ days) for the primary endpoint at 30days $(2.12\%$ vs $3.02\%$ , odds ratio (OR): 0.70, $95\%$ CI: 0.50–0.98). Symptomatic intracerebral hemorrhage rates were low in both the early and late groups: $0.45\%$ and $0.40\%$ , respectively. At 90days, primary endpoint events were numerically lower in the earlier than the later group, but without reaching statistical significance. The CATALYST meta-analysis supports the initiation of DOACs early after acute IS in patients with atrial fibrillation.

competing mechanisms are more frequent in AF patients on anticoagulants at the time of the event than among those off anticoagulation, 57 others have shown a similar prevalence on and off anticoagulation. 58 Among competing causes, the most frequently reported are large $(18–61\%)$ and small $(25{-}26\%)$ artery disease. 56–58 Cancer-related coagulopathy is a potential competing mechanism. Approximately $7\%$ of patients with AF have cancer and this is associated with an increased risk of IS (e.g. breast). 59 If cancer-related coagulopathy is suspected, further targeted investigations should be undertaken if the results are likely to change treatment.

Anticoagulant adherence or dosing failure

Adherence or dosing issues represent $32\%$ of all breakthrough strokes. 56 Poor adherence and persistence are the leading causes of inefficient anticoagulation in patients with AF. In a meta-analysis of 48 observational studies including 594,784 AF patients, the pooled proportion of good adherence to oral anticoagulants at 12months, defined as ${>}80\%$ of days covered or medication possession ratio, was only $68\%$ . 54 Similarly, the pooled proportion of persistence on anticoagulation at 12months was $62\%$ . 54 Both non-persistence (HR: 4.6; $95\%$ CI: 2.8–7.4) and poor adherence (HR: 1.4; $95\%$ CI: 1.06–1.8) were associated with increased stroke risk. 54 Other causes of inefficient anticoagulation beyond the scope of this work but still of clinical importance include poor absorption, underdosing, drug–drug interactions, and inappropriate interruption surrounding surgical procedures. 60

Management of breakthrough strokes on anticoagulation

Data from RCTs and population-based studies show that approximately $1\%$ of patients on DOACs experience a breakthrough IS annually. 52 Recurrent IS risk is particularly high in patients with breakthrough stroke ranging between $5\%$ and $9\%$ annually. 53 While suboptimal adherence to anticoagulants is still common, 54 breakthrough strokes can occur even with the best medication compliance and prescribing practices. 53 Several aspects must be considered before labeling a breakthrough event as DOAC failurerelated. The specific cause of breakthrough strokes can be identified by applying a comprehensive and systematic investigation, which in turn can help tailor secondary prevention strategies (Figure 3). 55

Identification of competing stroke mechanisms

The proportion of breakthrough strokes explained by competing mechanisms other than AF ranges between $24\%$ and $35\%$ . 56,57 While some series have shown that

Poor management of risk factors

As discussed previously, the management of risk factors is an essential component of stroke prevention, which is sometimes suboptimal, and explains a proportion of stroke recurrences. 29 Although not a stroke mechanism, part of the risk of stroke recurrence can be explained by poor control of risk factors.

AF-related residual risk

The most frequent cause of breakthrough strokes in patients on optimal anticoagulation and no competing mechanisms is cardioembolism from AF-related residual risk $(44\%)$ . 56 This risk is explained by AF-specific structural and functional factors, including left atrial appendage (LAA) morphology (e.g. LAA shape, 61 bend angle, 62 and orifice size) 63 and flow. 64 Approximately $90\%$ of cardiac thrombi in patients with AF originate in the LAA. 65 The prevalence of LAA thrombus among individuals receiving DOACs is approximately $2.3\%$ . 66 Among AF patients on DOACs, LAA thrombi seem to be more

frequent in those with a prior stroke than among those without 67 and the general population. 66 These LAA features can be investigated with transesophageal echocardiography, cardiac computed tomography, and cardiac magnetic resonance imaging.

Secondary stroke prevention in patients with breakthrough strokes

Given the high risk of early recurrence, secondary stroke prevention in patients with breakthrough strokes is essential. There are no data from RCTs evaluating whether switching a DOAC to a different DOAC or a VKA at the time of experiencing a breakthrough stroke reduces recurrent stroke risk. A study-level meta-analysis of six retrospective observational studies comprising 12,159 patients suggests that remaining on a DOAC instead of switching from DOACs to VKAs is associated with lower risk of IS recurrence (RR: 0.55; $95\%$ CI: 0.43 -0.70) and ICH (RR: 0.37, $95\%$ CI: 0.25 -0.55). Additionally, switching from a DOAC to a VKA was associated with an $85\%$ increased risk of death compared to remaining on a DOAC (RR: 1.85; $95\%$ CI: 1.06 -3.24). 68 This analysis is subject to the limitations of retrospective observational studies. In observational studies, adding an antiplatelet agent to anticoagulants was not associated with lower IS risk reduction. 56 In a subanalysis of RCTs, adding an antiplatelet agent was linked to increased ICH risk, 69 and a meta-analysis of RCTs and observational studies showed overall increased bleeding risk. 70 Evidence is missing for short-term addition of antiplatelet agents in patients with a competing large-artery stroke mechanism. The Frail Atrial Fibrillation (FRAIL AF) trial randomized frail individuals $'\!\geqslant\!75$ years of age and a Groningen Frailty Indicator score ${\geqslant}3$ ) with AF who were receiving VKAs to continue VKA therapy versus switching to a DOAC. 71 The primary outcome of major or clinically relevant nonmajor bleeding complication was more frequent in the DOAC group (HR: 1.69; $95\%$ CI: 1.23-2.32), without differences in the risk of thromboembolic events at 12months of follow-up. The proportion of patients with a previous stroke and the risk of IS were not reported.

A potential novel option for patients with breakthrough stroke is LAA occlusion as a matched observational cohort study found a lower risk of recurrent stroke compared to standard–of-care DOAC therapy alone in patients with breakthrough stroke (HR: 0.33, $95\%$ CI: 0.19–0.58). 72 Although promising, a major limitation of these data is that for the LAA occlusion patients, follow-up started from the moment of the LAA occlusion procedure bypassing the high-risk early post-stroke time period, while for the nonLAA occlusion cohort, follow-up started immediately after the index event (inclusive of the high-risk early post-stroke time period) introducing substantial bias in favor of LAA occlusion. As such, uncertainty remains and this approach

is currently being investigated in RCTs (NCT05976685; NCT05963698).

Left atrial appendage closure and other interventions

The left atrial appendage (LAA) is the primary cardioembolic structural source in AF patients. 65 As such, LAA closure (LAAC) has been tested in several RCTs as a potential strategy for stroke prevention in patients with AF.

Studies of percutaneous LAAC versus VKAs

The PROTECT-AF (Watchman Left Atrial Appendage System for Embolic Protection in Patients with Atrial Fibrillation) trial compared VKAs versus percutaneous LAAC in 707 anticoagulant-naïve AF with a $\mathrm{CHAPSD}_{2}$ score $\geqslant\!1$ . 73 LAAC met prespecified criteria for noninferiority and superiority (rate ratio (RR): 0.60, $95\%$ CI: 0.41– 1.05) for its primary efficacy endpoint (composite of stroke, SE, and cardiovascular/unexplained death). 73 The PREVAIL (Prospective Randomized Evaluation of the Watchman LAA Closure Device In Patients With Atrial Fibrillation Versus Long Term Warfarin Therapy) trial compared VKAs versus LAAC in 407 anticoagulant-naïve AF patients with a $\mathrm{CHAPSD}_{2}$ score ${\geqslant}2$ or 1 and another risk factor. 74 Percutaneous LAAC was non-inferior to warfarin for IS prevention or SE ${>}7$ days post-closure but did not achieve the prespecified noninferiority threshold for the composite endpoint of stroke, SE, and cardiovascular or unexplained death. 74 A prospective registry found no differences in outcomes in patients with and without a prior stroke. 75

Studies of percutaneous LAAC versus DOACs

The PRAGUE-17 (Left Atrial Appendage Closure vs Novel Anticoagulation Agents in Atrial Fibrillation) trial included 402 AF patients with at least one of the following: bleeding requiring intervention or hospitalization; breakthrough stroke while on anticoagulants or $\mathrm{CHA}_{2}\mathrm{DS}_{2}$ -VASc score ⩾ 3 + HAS-BLED score ${\geqslant}2$ . Patients were randomized to percutaneous LAAC versus DOACs. 76 LAAC was noninferior to DOAC in preventing the composite outcome of stroke, TIA, SE, cardiovascular death, major or nonmajor clinically relevant bleeding, or procedure-/device-related complications. There were no significant differences between groups in the risk of IS or TIA (HR: 1.13, $95\%$ CI: 0.44–2.93) or major/non-major bleeding (HR: 0.81, $95\%$ CI: 0.44–1.52). A propensity-matched analysis comparing percutaneous LAAC versus DOACs in 587 patients with AF and a prior stroke (median time between stroke and LAAC of 7.6months) showed no differences in the rates of IS and ICH, but a lower risk of the primary composite outcome of IS, major bleeding, and all-cause death. 77

Trials of surgical LAAC

Anticoagulation

The ATLAS (AtriClip Left Atrial Appendage Exclusion Concomitant to Structural Heart Procedures) pilot trial included patients undergoing nonmechanical valve and/or coronary artery bypass grafting without preoperative AF or needing anticoagulation, $\mathrm{CHA}_{2}\mathrm{DS}_{2}$ -VASc score ${\geqslant}2$ , and HAS-BLED score ⩾ 2. 78 A total of 562 patients were randomized to surgical LAA exclusion (LAAE) versus no exclusion. 78 The proportion of patients with postoperative AF was $44.3\%$ . The proportion of thromboembolic events was $3.4\%$ in LAAE patients and $5.6\%$ in the noLAAE group. 78 LAAOS III (Left Atrial Appendage Occlusion Study) randomized 4770 AF patients with a $\mathrm{CHA}_{2}\mathrm{DS}_{2}$ -VASc score ${\geqslant}2$ undergoing cardiac surgery to surgical LAAE versus no-LAAE. 79 Surgical LAAE reduced the risk of stroke or SE compared to no-LAAE (HR: 0.67; $95\%$ CI: 0.53–0.85) in a population where $80\%$ of patients continued to receive oral anticoagulation. 79 Results were consistent in interaction analysis for patients with and without prior IS, TIA or SE. 79 The results from LAAOS III have catalyzed several ongoing trials testing the combination of mechanical therapy (percutaneous LAAC or carotid filter devices) combined with oral anticoagulation for improved stroke prevention in AF patients who remain at high risk of stroke despite anticoagulation. 53

Percutaneous carotid filters

Vine TM is a novel permanent common carotid filter system that is implanted percutaneously under ultrasound guidance. It has been designed to prevent emboli ${>}1.4\,\mathrm{mm}$ that result in large vessel occlusions from reaching the anterior circulation, which is affected by the majority of AF-related ischemic strokes. 80,81 On the basis of a promising phase 2 program establishing the feasibility and safety of the device, in which there were no strokes due to large vessel occlusion following carotid filter implantation in over 268 patient-years of follow-up (106 participants), the INTERCEPT (Carotid Implants for PreveNtion of STrokE ReCurrEnce From Large Vessel Occlusion in Atrial Fibrillation Patients Treated With Oral Anticoagulation) RCT (NCT05723926) will be testing the superiority of bilateral carotid filter implantation + DOAC versus DOAC alone in patients with AF and stroke within the past year. 82,83

Secondary prevention in AF and previous ICH

Patients with a previous ICH, particularly intracerebral hemorrhage, have an inherently high risk of recurrent ICH. Therefore, the decision to start or reinstate antithrombotic therapy in these patients is challenging.

Observational studies suggest that resumption of anticoagulation after ICH may be associated with reduced thromboembolic events without an offsetting increase in the risk of ICH recurrence. 84 In an individual patient-level meta-analysis combining information from three small early-phase RCTs and subgroup data of a single-phase III RCT totalling 412 patients with an ICH and AF, oral anticoagulation did not lead to a significant reduction in the primary outcome of any stroke or cardiovascular death. 85 Patients who reinitiated anticoagulation had a lower frequency of recurrent IS ( $(4\%$ vs $19\%$ ) and major ischemic cardiovascular events, including IS, SE, pulmonary embolism, and myocardial infarction ( $(4\%$ vs $19\%$ ). However, anticoagulation had numerically higher ICH recurrence events ( $(6\%$ vs $3\%$ ). 85 Ongoing phase 3 clinical trials testing the safety and efficacy of anticoagulation in ICH survivors with AF will provide more evidence to inform clinical decision-making in the future (Supplemental Table S6).

The ENRICH-AF (Edoxaban for Intracranial Haemorr hage Survivors with Atrial Fibrillation trial) trial is comparing standard dosing edoxaban with non-anticoagulant medical treatment for stroke prevention in intracranial hemorrhage survivors with atrial fibrillation. 86 Following an initial safety review of the first 699 patients—where 174 $(25\%)$ presented with lobar intracranial hemorrhage and 34 $(5\%)$ with isolated convexity subarachnoid hemorrhage—the trial’s Data Safety Monitoring Board advised halting the enrollment of those with these two hemorrhage subtypes, which are typically caused by underlying cerebral amyloid angiopathy. They additionally recommended discontinuing the study drug immediately in this subgroup of patients. 86 The data leading to this recommendation have yet to be published, and no treatment interactions were identified in patients with lobar or isolated convexity subarachnoid hemorrhage in the abovementioned metaanalysis. 85 Further data from ongoing RCTs where patients with CAA-related intracranial hemorrhage remain eligible, and repeated meta-analyses will be important to clarify the net-benefit in these high-risk patients.

Antiplatelet therapy

Antiplatelet monotherapy, while inferior to anticoagulation, offers a modest $23\%$ reduction in the risk of thromboembolic events in patients with AF relative to placebo. 32 RESTART (REstart or STop Antithrombotics Randomised Trial) included 537 participants with ICH and a prior history of ischemic vascular disease. 87 Antiplatelet therapy did not increase the risk of recurrent ICH at a median follow-up of 2years. Still, it significantly reduced the risk of major ischemic vascular disease, although it must be noted that only $25\%$ (134/537) of the patients had AF. 87 The effect of Aspirin compared to no treatment or placebo in a

population exclusively comprising AF patients with a prior ICH has not been tested in an RCT.

LAAC in patients with ICH

Meta-analyses have indicated that LAAC may have similar efficacy to warfarin in lowering the risk of IS, but the risk of ICH may be significantly lower with LAA closure. 88 Data on head-to-head comparisons between LAAC and DOACs in patients with prior ICH are lacking. In addition, given that studies evaluating LAAC excluded patients with ICH, it is unclear if these results can be extrapolated to patients with ICH.

Timing of antithrombotic therapy post-ICH

Clinical equipoise exists on the optimal timing of antithrombotic therapy after ICH. Literature-based estimates on the optimal timing of OAC following intracranial hemorrhage range broadly from 3days to 30weeks. 89,90 While antiplatelet medications were started at a median of 76days (interquartile range (IQR): 29–146) in the RESTART trial, 87 observational data are inconclusive. 91

One of the most pressing uncertainties, due to its potential impact on health care costs and clinical outcomes, is the ideal duration of monitoring for AF detection in patients with a recent ischemic cerebrovascular event. Whether a single device approach or is better than stepwise combination of short-term followed by longer-term cardiac monitoring in selected patients is unknown.

Blood biomarkers

Evidence supporting the role of blood biomarkers, mainly natriuretic peptides, suggests they could be incorporated into clinical practice to select patients who may benefit from PCM. In addition, due to the association with AF-related outcomes, using natriuretic peptides may also help identify patients who could benefit from anticoagulation if AF is detected. RCTs specifically addressing these questions are needed. The results of the MOSES (MidregiOnal Proatrial Natriuretic Peptide to Guide SEcondary Stroke Prevention) trial are awaited (NCT03961334). Other blood biomarkers are at earlier stages of investigation or are less specific for AF detection.

Knowledge gaps and future directions

Despite significant advances in stroke prevention in patients with AF, we identified several knowledge gaps in this position statement. Relevant ongoing RCTs and observational studies addressing these gaps are listed in the supplementary file (Supplemental Table S6). This group has identified several research questions of clinical relevance that should be addressed in future studies (classified with a level of evidence C, D, E, or F in Table 1).

AF screening and classification of AF subtypes

Differences between subtypes of AF and their specific risk of IS outcomes are well-established. However, the clinical impact of this classification has not yet been demonstrated. Future studies are required to assess how the interplay between timing of AF detection, intensity of monitoring, AF burden, risk factors, and blood biomarkers affects stroke recurrence risk. It is also crucial to understand how AF burden and left atrial substrate progress over time. Given the increased risk of ICH resulting from the addition of antiplatelet agents to DOACs, there are concerns about potential harm when screening for AF in patients with an established competing cause (e.g. severe carotid artery stenosis). Therefore, the clinical implications of detecting AF in patients with a defined cause of stroke remains to be determined and adding oral anticoagulants to antiplatelet therapy in this population should be investigated in RCTs.

Secondary stroke prevention in AF and management of breakthrough strokes

Given the increasing use of PCM, one of the main uncertainties is whether oral anticoagulation can reduce stroke recurrence risk in patients with a recent IS or TIA and device-detected AF lasting < 24 h and not confirmed on 12-lead ECG are lacking, relative to antiplatelet therapy. Early rhythm control therapy has proven to be effective in patients with remote cerebrovascular events. However, no large, multicenter RCT has demonstrated its benefit in patients with a recent IS or TIA. EAST-STROKE study will address this question. There is no clear strategy for the management of patients with breakthrough strokes. Several strategies are being tested, including switching from a DOAC to a VKA versus staying on a DOAC, carotid filter implantation + DOAC, and LAAC versus LAAC + DOAC therapy.

Resuming or starting anticoagulation post-ICH

Whether resuming or starting anticoagulation in patients with a previous ICH is safe, improves survival, or effectively provides net benefit remains unknown. The results of PRESTIGE-AF (PREvention of STroke in Intracerebral haemorrhaGE Survivors With Atrial Fibrillation, NCT03996772) and ENRICH-AF, which have completed recruitment, are awaited, and the ASPIRE (Anticoagulation in ICH Survivors for Stroke Prevention and Recovery, NCT03907046) and A3ICH (Avoiding Anticoagulation

Table 1. WSO Brain & Heart Task Force Statements on AF in Patients With Recent IS, TIA and ICH.

| Level of Section ¥70 Statement evidence | |---| |AF Screening PCM increases AF detection compared to usual care. AI| |N Longer monitoring is associated with higher AF detection rates. AI| |The ideal duration of PCM in patients with recent IS/TIA is unknown. E| | A No strong evidence supports that PCM in recent IS/TIA patients reduces stroke A2 recurrence. | | Biomarkers S Cardiac troponin and natriuretic peptides NT-proBNP and MR-proANP are BI for guiding AF associated with AF detection. Other novel molecular biomarkers for AF prediction screening require further testing. | | Classification of 6 Based on differences in burden, risk factors, left atrial substrate, and embolic risk, B2 AF subtypes AF in patients with recent IS/TIA can be classified into known AF, ECG-detected AF, and AFDAS, which is always detected on PCM. | | Secondary stroke Optimizing the management of risk factors, including hypertension, diabetes, prevention obesity, sleep-disordered breathing, and excessive alcohol use in patients with recent IS/TIA with AF can reduce AF burden progression and recurrence post- ablation but there is no direct evidence supporting a reduction of IS recurrence risk in patients with AF. This evidence is unlikely to be generated, as improved risk factors control has been shown to reduce the risk of stroke in patients with and without AF. | | 8 Anticoagulation significantly reduces IS risk in patients with previously known AF. A4 Subanalyses in patients with remote IS/TIA show no significant interaction. | | DOACs are at least equally effective as VKAs for the prevention of recurrent IS A4 in patients with a remote cerebrovascular event, with a significantly lower risk of intracranial hemorrhage, ischemic stroke and systemic embolism. | | Data on the efficacy of anticoagulation in reducing stroke recurrence risk in patients with a remote IS/TIA and device-detected AF lasting |

Table 1. (Continued)

| Level of Section ¥70 Statement evidence | |---| | Management of 6 A thorough and systematic investigation of competing stroke mechanisms and B2 breakthrough causes of AF-related residual risk is required in patients with breakthrough strokes stroke on anticoagulants. | | There is no definite evidence that switching anticoagulants makes a difference in B2 any direction regarding IS recurrence risk in patients with a breakthrough stroke on anticoagulation. | | 8 Adding long-term antiplatelet agents to anticoagulants in patients with a recent IS/ B2 TIA and AF can increase ICH risk and offers no additional protection against IS recurrence. | | 9 C Percutaneous LAAC ± DOACs was associated with lower IS recurrence rates than persisting on a DOAC after a breakthrough stroke in a single retrospective observational study. | | LAA closure 20 Percutaneous LAAC is non-inferior to DOACs for the prevention of the composite C3 of stroke, TIA, SE, cardiovascular death, major or nonmajor clinically relevant bleeding, or procedure-/device-related complications in selected patients with and without a remote stroke. No differences between treatments were found in the risk of IS/TIA or major bleeding. There are no RCTs or subgroup analyses from RCTs in patients with a recent or remote IS/TIA/ICH. | | 2I Surgical LAA exclusion reduced the risk of IS/SE in patients with and without B2 remote IS/TIA undergoing cardiac surgery for another reason. | | NN Studies are needed for assessing novel protection devices or the combination of different strategies versus standard-of-care management in patients with a prior IS/ TIA. | | Intracranial 23 No robust data from RCTs support that anticoagulation can be safely initiated hemorrhage or resumed in patients with AF and a prior ICH without increasing the risk of a recurrent ICH. An individual patient-level meta-analysis from four RCTs showed that DOACs post-ICH result in significantly less major ischemic cardiovascular events compared to no anticoagulation. The number of recurrent ICHs was higher among patients with anticoagulation. The composite of major vascular events and death was inconclusive due to small sample size. Results of ongoing RCTs are awaited. | | 24 In patients with AF and prior ICH, there is insufficient data regarding the timing of anticoagulation resumption. | | 25 Currently, there are no robust data on the safety and efficacy of other interventions, such as antiplatelet therapy and LAAC, for preventing IS in AF with a previous ICH. |

After IntraCerebral Haemorrhage, NCT03243175) trials are ongoing. The role of other approaches such as LAAC versus antiplatelet therapy or anticoagulation requires further investigation.

In summary, many uncertainties remain on how to screen for AF and how to prevent AF-related strokes in varying scenarios. Funding agencies should prioritize research in these fields. Academic–industry partnerships are also strongly encouraged to advance knowledge.

Acknowledgements

We thank Dr. Alba Hernandez Pinilla and Ioannis Farah for their contribution to the systematic search of current evidence and ongoing studies.

Declaration of conflicting interest

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: LAS: Speaker/consulting honoraria from Boehringer

Ingelheim, Pfizer, Bayer, AstraZeneca, Medtronic; ACC: Speaker honoraria from BMS, Pfizer, AstraZeneca, and Boehringer Ingelheim; MK: Speaker/Consulting honoraria from Astra Zeneca, BMS, Medtronic; NBS: Consulting honoraria from Medtronic; VC: Speaker/consulting honoraria from Boehringer Ingelheim, Pfizer, Bayer, EVER PHARMA, Daiichi Sankyo; C-YH: Speaker honoraria from Boehringer Ingelheim, Daiichi Sankyo, Pfizer, Bayer, Medtronic; MJH: Speaker honoraria from Sanofi; DJS: Speaker/consulting honoraria from Pfizer and AstraZeneca; MCB: Speaker/consulting honoraria from J&J, Anthos Therapeutics, Merck; JFS: Speaker honoraria from BMS, Pfizer, AstraZeneca; AS: Speaker/consulting honoraria from AstraZeneca, Bayer AG, Daiichi Sankyo Ltd, Javelin Inc; other authors: no disclosures relevant to this work.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: LAS: Kathleen and Dr Henry Barnett Chair in Stroke Research; ACC: David Cargill Trust Clinical Senior Research Fellow; JFS: Support from Deutsche Herzstiftung e.V; MK: Swiss National Science Foundation (NR 182267, NR 213471) and Swiss Heart Foundation; other authors: no funding relevant to this work.

ORCID iDs

Luciano A. Sposato $\circledast$ https://orcid.org/0000-0001-6425-9343 Alan C. Cameron $\circledast$ https://orcid.org/0000-0001-6965-1109 Valeria Caso $\circledast$ https://orcid.org/0000-0002-2020-8891 Cheng-Yang Hsieh https://orcid.org/0000-0002-8772-4073 David J. Seiffge $\circledast$ https://orcid.org/0000-0003-3890-3849 Sheila C. O. Martins https://orcid.org/0000-0002-8452-712X Ashkan Shoamanesh https://orcid.org/0000-0002-2802-1626

Supplemental material

Supplemental material for this article is available online.

References

1. Li X, Liu Z, Jiang X, etal. Global, regional, and national burdens of atrial fibrillation/flutter from1990 to 2019: an age-period-cohort analysis using the global burden of disease 2019 study. J Glob Health 2023; 13: 4154. 2. Lippi G, Sanchis-Gomar F and Cervellin G. Global epidemiology of atrial fibrillation: an increasing epidemic and public health challenge. Int J Stroke 2021; 16: 217–221. 3. Wolf PA, Abbott RD and Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham study. Stroke 1991; 22: 983–988. 4. Jewett GA, Lindsay MP, Goia C, etal. National trends in hospital admission, case fatality, and sex differences in atrial fibrillation-related strokes. Int J Stroke 2020; 15: 521–527. 5. Sur NB, Wang K, Di Tullio MR, etal. Disparities and temporal trends in the use of anticoagulation in patients with ischemic stroke and atrial fibrillation. Stroke 2019; 50: 1452–1459. 6. Alkhouli M, Alqahtani F, Aljohani S, Alvi M and Holmes DR. Burden of atrial fibrillation-associated ischemic stroke in the United States. JACC Clin Electrophysiol 2018; 4: 618–625. 7. Sposato LA, Chaturvedi S, Hsieh CY, Morillo CA and Kamel H. Atrial fibrillation detected after stroke and transient ischemic attack: a novel clinical concept challenging current views. Stroke 2022; 53: e94–e103. 8. Ward K, Vail A, Cameron A, etal. Molecular biomarkers predicting newly detected atrial fibrillation after ischaemic stroke or TIA: a systematic review. Eur Stroke J 2023; 8: 125–131. 9. Scheitz JF, Erdur H, Haeusler KG, etal. Insular cortex lesions, cardiac troponin, and detection of previously unknown atrial fibrillation in acute ischemic stroke: insights from the troponin elevation in acute ischemic stroke study. Stroke 2015; 46: 1196–1201. 10. Cameron A, Cheng HK, Lee RP, etal. Biomarkers for atrial fibrillation detection after stroke: systematic review and metaanalysis. Neurology 2021; 97: e1775–e1789. 11. De Marchis GM, Schneider J, Weck A, etal. Midregional proatrial natriuretic peptide improves risk stratification after ischemic stroke. Neurology 2018; 90: e455–e465. 12. Arnold M, Nakas C, Luft A, Christ-Crain M, Leichtle A and Katan M. Independent prognostic value of MRproANP (Midregional Proatrial Natriuretic Peptide) levels in patients with stroke is unaltered over time. Stroke 2020; 51: 1873– 1875. 13. Cameron AC, Arnold M, Katsas G, etal. Natriuretic peptides to classify risk of atrial fibrillation detection after stroke: analysis of the BIOSIGNAL and PRECISE cohort studies. Neurology 2024; 103: e209625. 14. Tancin Lambert A, Ratajczak-Tretel B, Al-Ani R, etal. Biomarkers predictive of atrial fibrillation in patients with cryptogenic stroke. Insights from the nordic atrial fibrillation and stroke (NOR-FIB) study. Eur J Neurol 2023; 30: 1352– 1363. 15. Schweizer J, Arnold M, König IR, etal. Measurement of midregional pro-atrial natriuretic peptide to discover atrial fibrillation in patients with ischemic stroke. JACC 2022; 79: 1369–1381. 16. Ellis D, Rangaraju S, Duncan A, etal. Coagulation markers and echocardiography predict atrial fibrillation, malignancy or recurrent stroke after cryptogenic stroke. Medicine (Baltimore) 2018; 97: e13830. 17. Nahab F, Sharashidze V, Liu M, etal. Markers of coagulation and hemostatic activation aid in identifying causes of cryptogenic stroke. Neurology 2020; 94: e1892–e1899. 18. Hijazi Z, Benz AP, Lindback J, etal. Bone morphogenetic protein 10: a novel risk marker of ischaemic stroke in patients with atrial fibrillation. Eur Heart J 2023; 44: 208–218. 19. Hannemann J, Wasser K, Mileva Y, etal. Symmetric dimethylarginine predicts previously undetected atrial fibrillation in patients with ischemic stroke. JAHA 2024; e034994. 20. Hou W, Li Y, Wang J, etal. Soluble suppression of tumorigenicity 2 associated with atrial fibrillation detected after stroke: a retrospective study. Heliyon 2023; 9: e21778. 21. Yue H, Gu J, Zhao X, Liang W and Wu Z. Role of the interleukin-17 pathway in the pathogenesis of atrial fibrillation associated with inflammation. Arch Med Sci 2021; 17: 262–265. 22. Tao H, Shen X, Zou L, etal. Left atrial volume index and interleukin-6 as predictors for postoperative atrial fibrillation. J Cardiothorac Surg 2024; 19: 325. 23. Kaireviciute D, Blann AD, Balakrishnan B, etal. Characteri sation and validity of inflammatory biomarkers in the prediction of post-operative atrial fibrillation in coronary artery disease patients. Thromb Haemost 2010; 104: 122– 127. 24. Henningsen KM, Therkelsen SK, Bruunsgaard H, Krabbe KS, Pedersen BK and Svendsen JH. Prognostic impact of hs-CRP and IL-6 in patients with persistent atrial fibrillation treated with electrical cardioversion. Scand J Clin Lab Invest 2009; 69: 425–432. 25. Sposato LA, Lip GYH and Haeusler KG. Atrial fibrillation first detected after stroke: is timing and detection intensity relevant for stroke risk? Eur Heart J 2024; 45: 396–398. 26. Sposato LA, Field TS, Schnabel RB, etal. Towards a new classification of atrial fibrillation detected after a stroke or a transient ischaemic attack. Lancet Neurol 2024; 23: 110– 122. 27. Fridman S, Vargas Gonzalez JC, Jimenez-Ruiz A, etal. Clinical and prognostic differences between atrial fibrillation detected before and after stroke and TIA. Cerebrovas Dis 2022; 51: 152–157. 28. Alvarado-Bolaños A, Ayan D, Khaw AV, etal. Differences in stroke recurrence risk between atrial fibrillation detected on ECG and 14-Day cardiac monitoring. Stroke 2023; 54: 2022– 2030. 29. Mead GE, Sposato LA, Silva GS, etal. Systematic review and synthesis of global stroke guidelines on behalf of the world stroke organization. Int J Stroke 2023; 18: 499-531. 30. Hendriks JM, Gallagher C, Middeldorp ME, etal. Risk factor management and atrial fibrillation. EP Europace 2021; 23: ii52–ii60. 31. Van Gelder IC, Rienstra M, Bunting KV, etal. 2024 ESC guidelines for the management of atrial fibrillation. Eur Heart J 2024; 00: 1–101. 32. Hart RG, Pearce LA and Aguilar MI. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med 2007; 146: 857–867. 33. Ruff CT, Giugliano RP, Braunwald E, etal. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: a meta-analysis of randomised trials. Lancet 2014; 383: 955–962. 34. Kirchhof P, Toennis T, Goette A, etal. Anticoagulation with edoxaban in patients with atrial high-rate episodes. N Engl J Med 2023; 389: 1167–1179. 35. Healey JS, Lopes RD, Granger CB, etal. Apixaban for stroke prevention in subclinical atrial fibrillation. N Engl J Med 2024; 390: 107–117. 36. McIntyre WF, Benz AP, Becher N, etal. Direct oral anticoagulants for stroke prevention in patients with device-detected atrial fibrillation: a study-level meta-analysis of the NOAHAFNET 6 and ARTESiA trials. Circulation 2023; 149: 981– 988. 37. Sanders P, Svennberg E, Diederichsen SZ, etal. Great debate: device-detected subclinical atrial fibrillation should be treated like clinical atrial fibrillation. Eur Heart J 2024; 45: 2594– 2603. 38. Siegler JE, Sposato LA and Yaghi S. Toward more personalized management of device-detected atrial fibrillation. JAMA Neurol 2024; 81: 573–574. 39. Siegler JE and Brorson JR. Device-detected atrial fibrillation and the impact of prior stroke in stroke prevention. JAHA 2024; 13: e037124. 40. Jensen M, Suling A, Metzner A, etal. Early rhythm-control therapy for atrial fibrillation in patients with a history of stroke: a subgroup analysis of the EAST-AFNET 4 trial. Lancet Neurol 2023; 22: 45–54. 41. Park J, Shim J, Lee JM, etal. Risks and benefits of early rhythm control in patients with acute strokes and atrial fibrillation: a multicenter, prospective, randomized study (the RAFAS trial). J Am Heart Assoc 2022; 11: e023391. 42. Seiffge DJ, Traenka C, Polymeris A, etal. Early start of DOAC after ischemic stroke: risk of intracranial hemorrhage and recurrent events. Neurology 2016; 87: 1856–1862. 43. Yaghi S, Trivedi T, Henninger N, etal. Anticoagulation timing in cardioembolic stroke and recurrent event risk. Ann Neurol 2020; 88: 807–816. 44. De Marchis GM, Seiffge DJ, Schaedelin S, etal. Early versus late start of direct oral anticoagulants after acute ischaemic stroke linked to atrial fibrillation: an observational study and individual patient data pooled analysis. J Neurol Neurosurg Psychiatry 2022; 93: 119–125. 45. Kimura S, Toyoda K, Yoshimura S, etal. Practical “1-2-3-4- day” rule for starting direct oral anticoagulants after ischemic stroke with atrial fibrillation: combined hospital-based cohort study. Stroke 2022; 53: 1540–1549. 46. Grosse GM, Hüsing A, Stang A, etal. Early or late initiation of dabigatran versus vitamin-K-antagonists in acute ischemic stroke or TIA: the PRODAST study. Int J Stroke 2023; 18: 1169–1177. 47. Sharobeam A, Lin L, Lam C, etal. Early anticoagulation in patients with stroke and atrial fibrillation is associated with fewer ischaemic lesions at 1 month: the ATTUNE study. Stroke Vasc Neurol 2024; 9: 30–37. 48. Oldgren J, Åsberg S, Hijazi Z, etal. Early versus delayed nonvitamin k antagonist oral anticoagulant therapy after acute ischemic stroke in atrial fibrillation (TIMING): a registrybased randomized controlled noninferiority study. Circulation 2022; 146: 1056–1066. 49. Fischer U, Koga M, Strbian D, etal. Early versus later anticoagulation for stroke with atrial fibrillation. N Engl J Med 2023; 388: 2411–2421. 50. Werring DJ, Dehbi HM, Ahmed N, etal. Optimal timing of anticoagulation after acute ischaemic stroke with atrial fibrillation (OPTIMAS): a multicentre, blinded-endpoint, phase 4, randomised controlled trial. Lancet . Epub ahead of print 23 October 2024. DOI: 10.1016/S0140-6736(24)02197-4. 51. Asberg Sea. CollAboration on the optimal Timing of anticoagulation after ischaemic stroke and Atrial fibrillation: prospective individuaL participant data meta-analYsiS of randomized controlled Trials (CATALYST). Presented at the World Stroke Congress. 23–26 October 2024, Abu Dhabi, United Arab Emirates, WSC, 2024. 52. Sposato LA, Sur NB, Katan M, etal. Embolic stroke of undetermined source: new data and new controversies on cardiac monitoring and anticoagulation. Neurology 2024; 103: e209535. 53. Seiffge DJ, Cancelloni V, Räber L, etal. Secondary stroke prevention in people with atrial fibrillation: treatments and trials. Lancet Neurol 2024; 23: 404–417. 54. Ozaki AF, Choi AS, Le QT, etal. Real-world adherence and persistence to direct oral anticoagulants in patients with atrial fibrillation: a systematic review and meta-analysis. Circ Cardiovasc Qual Outcomes 2020; 13: e005969. 55. Johansen MC. The future of ischemic stroke diagnosis and a review of underrecognized ischemic stroke etiologies. Neurotherapeutics 2023; 20: 613–623. 56. Polymeris AA, Meinel TR, Oehler H, etal. Aetiology, secondary prevention strategies and outcomes of ischaemic stroke despite oral anticoagulant therapy in patients with atrial fibrillation. J Neurol Neurosurg Psychiatry 2022; 93: 588–598. 57. Herlekar R, Sur Roy A, Hajiev S, etal. The contribution of competing mechanisms in stroke despite anticoagulation in patients with atrial fibrillation. Eur Stroke J 2023; 8: 541–548. 58. Yaghi S, Henninger N, Giles JA, etal. Ischaemic stroke on anticoagulation therapy and early recurrence in acute cardioembolic stroke: the IAC study. J Neurol Neurosurg Psychiatry 2021; 92: 1062–1067. 59. Ajabnoor AM, Parisi R, Zghebi SS, etal. Common cancer types and risk of stroke and bleeding in patients with nonvalvular atrial fibrillation: a population-based study in England. J Am Heart Assoc 2023; 12: e029423. 60. Guenoun M, Cohen S, Villaceque M, etal. Characteristics of patients with atrial fibrillation treated with direct oral anticoagulants and new insights into inappropriate dosing: results from the French National Prospective Registry: PAFF. Europace 2023; 25: euad302. 61. Di Biase L, Santangeli P, Anselmino M, etal. Does the left atrial appendage morphology correlate with the risk of stroke in patients with atrial fibrillation? Results from a multicenter study. JACC 2012; 60: 531–538. 62. Yaghi S, Chang A, Ignacio G, etal. Left atrial appendage morphology improves prediction of stagnant flow and stroke risk in atrial fibrillation. Circ Arrhythm Electrophysiol 2020; 13: e008074. 63. Takaya Y, Nakayama R, Yokohama F, etal. Left atrial appendage morphology with the progression of atrial fibrillation. PLoS ONE 2022; 17: e0278172. 64. Nio SS, Rinkel L, Cramer O, etal. Left atrial appendage opacification on cardiac CT in acute ischemic stroke: the clinical implications of slow-flow. JAHA 2024; 13: e034106. 65. Cresti A, García-Fernández MA, Sievert H, etal. Prevalence of extra-appendage thrombosis in non-valvular atrial fibrillation and atrial flutter in patients undergoing cardioversion: a large transoesophageal echo study. EuroIntervention 2019; 15: e225–e230. 66. Alqarawi W, Grose E, Ramirez FD, etal. Prevalence of left atrial appendage thrombus in patients anticoagulated with direct oral anticoagulants: systematic review and meta-analysis. CJC Open 2021; 3: 658–665. 67. Durmaz E, Karpuz MH, Bilgehan K, etal. Left atrial thrombus in patients with atrial fibrillation and under oral anticoagulant therapy; 3-D transesophageal echocardiographic study. Int J Cardiovasc Imaging 2020; 36: 1097–1103. 68. Mota Telles JP, Cenci GI, Marinheiro G, etal. Anticoagulation strategy for patients presenting with ischemic strokes while using a direct oral anticoagulant: a systematic review and meta-analysis. Int J Stroke 2025; 20(1): 42–52. 69. Lopes RD, Guimarães PO, Kolls BJ, etal. Intracranial hemorrhage in patients with atrial fibrillation receiving anticoagulation therapy. Blood 2017; 129: 2980–2987. 70. Ghule P, Panic J and Malone DC. Risk of bleeding with concomitant use of oral anticoagulants and aspirin: a systematic review and meta-analysis. Am J Health Syst Pharm 2024; 81: 494–508. 71. Joosten LPT, van Doorn S, van de Ven PM, etal. Safety of switching from a vitamin K antagonist to a non-vitamin K antagonist oral anticoagulant in frail older patients with atrial fibrillation: results of the FRAIL-AF randomized controlled trial. Circulation 2024; 149: 279–289. 72. Maarse M, Seiffge D, Werring DJ, etal. Left atrial appendage occlusion vs standard of care after ischemic stroke despite anticoagulation. JAMA Neurol 2024; 81: 1150–1158. DOI: 10.1001/jamaneurol.2024.2882. 73. Reddy VY, Sievert H, Halperin J, etal. Percutaneous left atrial appendage closure vs warfarin for atrial fibrillation: a randomized clinical trial. JAMA 2014; 312: 1988–1998. 74. Holmes DR, Jr, Kar S, Price MJ, etal. Prospective randomized evaluation of the watchman left atrial appendage closure device in patients with atrial fibrillation versus long-term warfarin therapy: the PREVAIL trial. J Am Coll Cardiol 2014; 64: 1–12. 75. Ansari U, Brachmann J, Lewalter T, etal. LAA occlusion is effective and safe in very high-risk atrial fibrillation patients with prior stroke: results from the multicentre German LAARGE registry. Clin Res Cardiol 2024; 113: 1451–1462. 76. Osmancik P, Herman D, Neuzil P, etal. Left atrial appendage closure versus direct oral anticoagulants in high-risk patients with atrial fibrillation. JACC 2020; 75: 3122–3135. 77. Korsholm K, Valentin JB, Damgaard D, etal. Clinical outcomes of left atrial appendage occlusion versus direct oral anticoagulation in patients with atrial fibrillation and prior ischemic stroke: a propensity-score matched study. Int J Cardiol 2022; 363: 56–63. 78. Gerdisch MW, Garrett HE, Jr, Mumtaz MA, etal. Prophylactic left atrial appendage exclusion in cardiac surgery patients with elevated CHA(2)DS(2)-VASc score: results of the randomized ATLAS trial. Innovations (Phila) 2022; 17: 463–470. 79. Whitlock RP, Belley-Cote EP, Paparella D, etal. Left atrial appendage occlusion during cardiac surgery to prevent stroke. N Engl J Med 2021; 384: 2081–2091. 80. Jaakkola J, Hartikainen P, Kiviniemi TO, etal. Distribution of ischemic strokes in patients with atrial fibrillation: the FibStroke study. Neurol Clin Pract 2019; 9: 330–336. 81. De Potter T, Yodfat O, Shinar G, etal. Permanent bilateral carotid filters for stroke prevention in atrial fibrillation. Curr Cardiol Rep 2020; 22: 144. 82. Reddy VY, Neuzil P, de Potter T, etal. Permanent percutaneous carotid artery filter to prevent stroke in atrial fibrillation patients: the CAPTURE trial. J Am Coll Cardiol 2019; 74: 829–839. 83. Shoamanesh A, Reddy V, Benz AP, etal. Permanent bilateral carotid artery implants to prevent large vessel occlusion stroke in atrial fibrillation: CAPTURE 2. Eur Stroke J 2023; 8: 97. 84. Murthy SB, Gupta A, Merkler AE, etal. Restarting anticoagulant therapy after intracranial hemorrhage: a systematic review and meta-analysis. Stroke 2017; 48: 1594–1600. 85. Al-Shahi Salman R, Stephen J, Tierney JF, etal. Effects of oral anticoagulation in people with atrial fibrillation after spontaneous intracranial haemorrhage (COCROACH): prospective, individual participant data meta-analysis of randomised trials. Lancet Neurol 2023; 22: 1140–1149. 86. Shoamanesh A. Anticoagulation in patients with cerebral amyloid angiopathy. Lancet 2023; 402: 1418–1419. 87. Collaboration R. Effects of antiplatelet therapy after stroke due to intracerebral haemorrhage (RESTART): a randomised, open-label trial. Lancet 2019; 393: 2613–2623. 88. Reddy VY, Doshi SK, Kar S, etal. 5-year outcomes after left atrial appendage closure: from the PREVAIL and PROTECT AF trials. J Am Coll Cardiol 2017; 70: 2964–2975. 89. Hawryluk GW, Austin JW, Furlan JC, Lee JB, O’Kelly C and Fehlings MG. Management of anticoagulation following central nervous system hemorrhage in patients with high thromboembolic risk. J Thromb Haemost 2010; 8: 1500–1508. 90. Majeed A, Kim YK, Roberts RS, Holmström M and Schulman S. Optimal timing of resumption of warfarin after intracranial hemorrhage. Stroke 2010; 41: 2860–2866. 91. Murthy SB, Biffi A, Falcone GJ, etal. Antiplatelet therapy after spontaneous intracerebral hemorrhage and functional outcomes. Stroke 2019; 50: 3057–3063.

相关指南

• 2020 JCS／JHRS指南：心律失常的药物治疗 · 心血管内科 2020
• 心房颤动：目前的认识和治疗建议（2021） · 心血管内科 2021
• 2023年心房颤动诊断和治疗中国指南 · 心血管内科 2023
• 老年心房颤动诊治中国专家共识（2024） · 心血管内科 2024
• 2024 ACC专家共识决策路径：卒中后心律失常监测的实用方法 · 心血管内科 2024
• 中国心血管健康与疾病报告2022概要 · 心血管内科 2023
• 2024 通心络防治冠心病临床应用中国专家共识 · 心血管内科 2024
• 泛血管疾病抗栓治疗中国专家共识（2024版） · 心血管内科 2024