Weekly Journal Scan: Risk factors for recurrent ischaemic stroke in patients with atrial fibrillation on oral anticoagulation

Comment on the article ‘Recurrent Ischemic Stroke in Patients With Atrial Fibrillation While Receiving Oral Anticoagulants’ published in JAMA Neurology, https://doi.org/10.1001/jamaneurol.2024.1892. Patients with atrial fibrillation (AF) face a substantial residual risk of ischaemic stroke (IS) despite treatment with oral anticoagulants (OACs). The present study aimed to evaluate the incidence of recurrent IS in such patients and to assess the risk associated with OAC discontinuation.1 This is a nationwide registry-based cohort analysis conducted in Denmark, involving patients aged 50 or older who had non-valvular AF and experienced an IS between 2014 and 2021. These patients were treated with OACs at the time or within 30 days after their stroke (entry IS) and followed until June 2022 for any recurrent IS. A nested case–control analysis was performed within this cohort to compare patients who experienced recurrent IS with those who did not, focusing on the status of OAC use. The primary outcome was recurrent IS. The study included 8119 patients (mean age, 78.4 years; 54.1% male; median CHA2DS2-VASc score, 4.0; 81% with hypertension; 20% with diabetes; 27% with ischaemic heart disease; 16% with HF; 13% with prior IS) who initiated or restarted OAC, both vitamin K antagonists (VKAs) and direct OACs (DOACs), for their entry IS. Over an average follow-up of 2.9 years, 663 patients experienced recurrent IS, with 80% of these patients still on OACs at the time of the second stroke. The cumulative incidence of recurrent IS was 4.3% [95% confidence interval (CI), 3.8%–4.8%] at 1 yea, and 6.5% (95% CI, 5.9%–7.1%) at 2 years. The cumulative incidence of all-cause mortality was 15.4% (95% CI, 14.7%–16.3%) at 1 year and 24.7% (95% CI, 23.7%–25.7%) at 2 years. Recurrent IS incidence rates (IRs) were ∼70% higher in patients older than 85 than in those younger than 75 years. A history of IS before the entry IS was associated with a higher incidence of recurrent IS compared with no prior IS (1-year cumulative incidence, 6.5% vs. 4.0%) and all-cause mortality (19.1% vs. 14.9%). Analyses adjusted for age, sex, and accounting for the competing risk of death did not change the overall results. Patients who restarted OACs within 30 days after discharge of their entry IS had a significantly higher risk of recurrent IS compared with OAC initiators (IR ratio, 1.55; 95% CI, 1.34–1.79). In the nested case–control analysis, comparing 663 cases with recurrent IS with 2652 matched controls without recurrent IS, OAC discontinuation was associated with a 2-fold higher risk of recurrent IS compared with current use [89 cases (13.4%) vs. 180 controls (6.8%); adjusted OR 2.13 (95% CI, 1.57–2.89)]. The association was also observed in analyses restricted to discontinuation of the most commonly used DOACs in Denmark, apixaban and rivaroxaban. Among cases who discontinued OACs, only ∼10% had hospital contacts with codes for major bleeding events or surgical procedures in the month before OAC discontinuation. Oral anticoagulants (OACs) are highly effective and safe in preventing stroke in atrial fibrillation (AF) patients with and without previous stroke or transient ischaemic attack,2–5 and their use is universally recommended by treatment guidelines.6 This large nationwide observational study analysed five different registries recording clinical diagnoses, drug prescriptions, and causes of death on all Danish patients with ischaemic stroke (IS) and AF who initiated or restarted OACs within 30 days after their entry IS and found that recurrent IS was common and mortality rates were high.1 These findings are largely consistent with the results of a recent combined analysis of individual participant data (IPD) from five pivotal randomized controlled trials (RCTs) of direct oral anticoagulant (DOAC) therapy in AF designed to assess the outcomes of patients with an IS while on study medication during trial follow-up.7 The primary analysis included 1163 patients with a first post-randomization IS while on study medication. During follow-up, the cumulative incidence of recurrent IS at 1 year was 7.0% (95% CI, 5.2%–8.7%), and the cumulative incidence of mortality at 3 months after stroke was 12.4% (95% CI, 10.5%–14.4%).7 The results of this analysis also suggested that the rate of stroke recurrence despite OACs is particularly high in patients treated with OACs at the time of the index event.7 However, both the data set used for this analysis and the registry data used for the observational study did not include detailed information on adherence to oral anticoagulation, adequacy of risk factor control during follow-up (e.g. hypertension and diabetes management), and stroke mechanism, thereby limiting any insight into the cause(s) of OAC treatment failures. Although cardioembolism is the prevailing mechanism of IS in patients with AF, other causes such as stroke due to small vessel disease or a lacunar stroke may be responsible for OAC treatment failures, given their limited susceptibility to antithrombotic therapy alone. Many patients with AF have other risk factors (e.g. hypertension, diabetes mellitus, obesity, or dyslipidaemia) that are associated with an increased risk of stroke. Current guidelines for the treatment of AF patients emphasize the importance of lifestyle intervention and adequate management of risk factors.8 The observed higher risk of recurrent IS in OAC restarters (38% of patients in the present study) compared with OAC initiators is consistent with a previous IPD pooled analysis of seven prospective cohort studies comparing the risk of recurrent IS in patients with AF with and without OAC treatment at the time of their index IS.9 This analysis yielded a 4.7% annual risk for recurrent IS and a 10.2% risk of death. Prior OAC treatment was associated with a 60% higher risk of recurrent IS.9 The other main finding of the present study is that the risk of recurrent IS was 2-fold higher in patients discontinuing OACs compared with patients who continued therapy, regardless of the OAC subgroup.1 Reasons underlying OAC discontinuation are poorly understood. The vast majority of cases who discontinued OACs did not have hospital contacts with codes for major bleeding events or surgical procedures in the month before OAC discontinuation.1 Prior research showed that OAC discontinuation was linked to higher IS risk, with a similar increase to that reported in the present study, but it did not focus specifically on recurrent IS and it did not report short-term outcomes and stroke severity associated with OAC status.10 The current study benefits from a large nationwide cohort and use of registries characterized by a high degree of completeness, minimizing selection, and providing statistically robust data.1 Moreover, the study was conducted in a setting where healthcare is free of charge and drug costs are partially covered, which reduces the influence of socioeconomic status on health care access. The use of both cohort and nested case–control analyses allowed for a detailed examination of risk factors associated with recurrent IS. On the other hand, this study is limited by its observational nature; thus, it cannot establish causality of the association between OAC discontinuation and recurrent IS, though a cause–effect relationship is biologically and pharmacologically plausible.1 Furthermore, the study relies on registry data, which may not capture all relevant clinical details, such as reasons for OAC discontinuation and adequacy of cardiovascular risk factor management, and reports data from just one country of largely European ancestry, thus limiting generalizability of its findings. The current results align with prior research, confirming the persistent risk of IS despite OAC treatment of AF patients and emphasizing the fundamental importance of uninterrupted OAC therapy to minimize recurrence risk.1 There is an unmet therapeutic need for these patients, requiring further research possibly based on neuroimaging and biomarkers, including a multi-ethnic and socially heterogeneous patient population, specifically focused on stroke mechanisms, response to OACs, reasons for discontinuation, and information on nutrition and physical exercise, that may lead to precision therapy and better patient outcomes. L.G. reports no conflicts of interest. C.P. reports personal fees from AbbVie, Eli Lilly, and Tremeau and past grant support (to the Institution) for investigator-initiated research from AIFA (Italian Drug Agency), Bayer, Cancer Research UK, and European Commission; he chaired the Scientific Advisory Board of the International Aspirin Foundation.