One‐year Follow‐up Results of the Optimal Thromboprophylaxis in Elderly Chinese Patients with Atrial Fibrillation (ChiOTEAF) registry

Abstract Background The high prevalence of atrial fibrillation (AF) in the very elderly population (aged >80 years) might be underestimated. The elderly are at increased risk of both fatal stroke and bleeding. The Optimal Thromboprophylaxis in Elderly Chinese Patients with Atrial Fibrillation (ChiOTEAF) registry provides contemporary management strategies among the elderly Chinese patients in the new era of non‐vitamin K antagonists. Objective To present the 1‐year follow‐up data from the ChiOTEAF registry, focusing on the use of antithrombotic therapy, rate vs. rhythm control strategies, and determinants of mortality and stroke. Methods The ChiOTEAF registry analyzed consecutive AF patients presenting in 44 centers from 20 Chinese provinces from October 2014 to December 2018. Endpoints of interest were mortality, thromboembolism, major bleedings, cardiovascular comorbidities, and hospital re‐admissions. Results Of the 7077 patients enrolled at baseline, 657 patients (9.3%) were lost to the follow‐up and 435 deaths (6.8%) occurred. The overall use of anticoagulants remains low, approximately 38% of the entire cohort at follow‐up, with similar proportions of vitamin K antagonists (VKA) and non‐vitamin K antagonists (NOACs). Antiplatelet therapy was used in 38% of the entire cohort at follow‐up, and more commonly among high‐risk patients (41%). Among those on a NOAC at baseline, 22.4% switched to antiplatelet therapy alone after one year. Independent predictors of stroke/transient ischemic attack/peripheral embolism and/or mortality were age, heart failure, chronic kidney disease, prior ischemic stroke, dementia, and chronic obstructive pulmonary disease. Conclusions The ChiOTEAF registry provides contemporary data on AF management, including stroke prevention. The poor adherence of NOACs and common use of antiplatelet in these high‐risk elderly population calls for multiple comorbidities management.


| INTRODUC TI ON
Atrial fibrillation (AF) incidence is increasing over the past decade 1 ; patients with AF are older and overburdened with multimorbidity. 2,3 The prevalence of AF in the elderly (aged >80 years) ranges from 10% to 17% 4 ; however, the exact numbers might be underestimated due to asymptomatic AF. Likewise, age is an independent risk factor for adverse outcomes in patients with AF, 5 and the elderly are at a high risk of fatal ischemic stroke and major bleeding. 6 Recent data demonstrate the beneficial effect of oral anticoagulants (OACs) for stroke prevention in elderly patients with AF. [7][8][9][10][11][12] One meta-analysis showed a significant reduction in the risk of stroke and systemic embolism without increasing major bleeding events among the elderly treated with non-vitamin K antagonist OACs (NOACs) compared with vitamin K antagonists (VKAs). 13 A Taiwanese cohort study of extreme elderly (aged >90) AF patients showed superior effectiveness and safety of NOACs compared with VKAs. 11 Furthermore, the use of NOACs was associated with a reduction in adverse events, especially the risk of intracranial hemorrhage. 11 Despite the clear benefit of OAC therapy is maintained in elderly patients with AF, "real-world" data showed that OACs are substantially underused [14][15][16][17][18] due to a fear of bleeding, especially among those with frailty or dementia. 19 Of note, the NOACs showed better efficacy and safety among Asian patients compared with non-Asians. 20 The introduction of the NOACs has led to a major change in the landscape of stroke prevention in AF, but limited contemporary nationwide data are evident from China. Thus, the prospective, nationwide Optimal Thromboprophylaxis in Elderly Chinese Patients with Atrial Fibrillation (ChiOTEAF) registry aimed to explore contemporary regional management strategies, including antithrombotic therapy among the high-risk AF population of the elderly Chinese patients, in the new era of the NOACs.
In this analysis, we present the 1-year follow-up data from the ChiOTEAF registry, focusing on the use of antithrombotic therapy, rate vs. rhythm control strategies, and determinants of mortality and stroke.

| ME THODS
The protocol of the ChiOTEAF registry has previously been published. 21 The study was approved by the Central Medical Ethics Committee of Chinese PLA General Hospital, Beijing, China (approval no S2014-065-01) and local institutional review boards.
The registry was conducted between October 2014 and December 2018. Briefly, the registry population comprises consecutive in-and outpatients presenting with AF to cardiologists (mainly), neurologist, and surgeons, enrolled in 44 sites from 20 Chinese provinces. The main inclusions criteria were age ≥65 years (for the extended analysis, AF patients aged >50 years were included) and the qualifying AF event in the 12 months prior to enrolment (recorded by a 12-lead ECG or 24 hours ECG Holter).
Data were collected at the moment of enrolment and during the follow-up visits (including patient visit and/or chart review and/or telephone follow-up) by any investigator and reported into an electronic case report form. Follow-up was performed by the local investigators, initially at 6 and 12 months in the first year and annually for the next 2 years. Endpoints of interest were mortality, thromboembolism, major bleedings, cardiovascular comorbidities, and hospital re-admissions. For this analysis, we focused on 1-year outcomes.
The ChiOTEAF registry has common definitions and protocol for the EURObservational Research Programme Atrial Fibrillation (EORP-AF) General Registry. 22 Based on the ESC guidelines, 23 thromboembolic risk was categorized using the CHA 2 DS 2 -VASc score. 5 "Low-risk" patients were defined as males with a CHA 2 DS 2 -VASc 0 or females with a CHA 2 DS 2 -VASc 1; "moderate risk" was defined as male patients with a CHA 2 DS 2 -VASc score 1 or females with a CHA 2 DS 2 -VASc 2; and "high risk" was defined as CHA 2 DS 2 -VASc score ≥2.
Bleeding risk was assessed based on the HAS-BLED bleeding score. 23

| Statistical analyses
Univariate analysis was applied to continuous and categorical variables. Continuous variables were reported as mean+SD and/or as median and inter-quartile range (IQR). Among-group comparisons were Independent predictors of stroke/transient ischemic attack/peripheral embolism and/or mortality were age, heart failure, chronic kidney disease, prior ischemic stroke, dementia, and chronic obstructive pulmonary disease.

Conclusions:
The ChiOTEAF registry provides contemporary data on AF management, including stroke prevention. The poor adherence of NOACs and common use of antiplatelet in these high-risk elderly population calls for multiple comorbidities management.

K E Y W O R D S
atrial fibrillation, mortality, prognosis, registry, stroke prevention, thromboprophylaxis made using a non-parametric test (Kruskal-Wallis test). Categorical variables were reported as percentages, and the χ 2 test or Fisher's exact test (if required) was used for among-group comparisons.
All the statistically significant variables at univariate analysis and variables considered of relevant clinical interests were included in the multivariable model to distinguish the independent predictors of all-cause death and/or stroke/transient ischemic attack (TIA)/peripheral embolism during the 1-year follow-up period.
A Cox proportional hazard model was performed by adjusting for the following covariates: sex, hypertension, coronary artery disease, liver dysfunction, and prior major bleeding. All Cox regression analyses were reported as hazard ratio (HR) and 95% confidence interval [CI]. A two-sided P-value of <.05 was considered statistically significant.

| RE SULTS
Available data on patient demography and baseline characteristics in relation to clinical AF subtype are summarized in Table 1, and the patient disposition is shown in Figure 1 Table 1). Analysis of AF subtypes showed that those patients with permanent AF were older, but no statistically significant difference was found in a gender ratio between groups. Differences in the risk of stroke and bleeding (HAS-BLED score ≥3) strata were evident, with more high-risk patients in the subgroups of permanent and long-standing persistent AF (Table 1). Patients were overburdened with multi-morbidity (particularly patients with long-standing persistent and permanent AF), including hypertension (62.6%), coronary artery disease (44.8%), heart failure (32.1%), diabetes mellitus (25.8%), prior ischemic stroke (20.1%) chronic kidney disease (10.7%), chronic obstructive pulmonary disease (6.8%), and dementia (2.1%).

| Antithrombotic therapy
Overall, 5350 patients had available data on antithrombotic drugs at baseline and 1-year follow-up, in relation to AF type. The use of antithrombotic therapy at a 1-year follow-up visit, concerning antithrombotic therapy used at the baseline visit is shown in Figure 2.
Of those on a vitamin K antagonist (VKA), 75.1% remained on a VKA, and 9.9% had switched to a NOAC during the follow-up. Among those on a NOAC at baseline, 2.2% had changed to a VKA and 22.4% to antiplatelet therapy alone. Of those on antiplatelet therapy at the baseline, 14.8% had switched to OAC, and 4.5% had dual therapy (OAC and antiplatelet).
Drug therapies prescribed at follow-up are shown in Table 2

| Rate and rhythm control strategy
For the analysis of rate and rhythm control strategies, 6022 patients with available data by 1-year follow-up were included. Drugs used for rhythm and rate control therapy at follow-up are summarized in Table 2C. Beta-blockers (54.5%) and digitalis (10.2%) remained the most common drugs used, especially in persistent and long-standing persistent AF; while Class Ic and III drugs were more often used in paroxysmal AF (5.9% and 12.7%, respectively).
Among patients managed with rate control at baseline, only 4.2% continued a rate control strategy, while rhythm control was considered in 43.1% ( Figure S1). Of those considered for a rhythm control at baseline, 23.9% continued the strategy, and 16.4% were eventually considered for a rate control therapy. Table 3 shows the interventions performed by the 1-year follow-up. Any rhythm control intervention was performed in 9% of the overall cohort-especially among persistent and long-standing persistent AF patients (12.8 and 17.6%, respectively). Catheter ablation was performed in 5.5% of the population, commonly among paroxysmal AF patients (8%); whereas pacemaker implantation was required in 6.9% of permanent AF patients.

| Mortality and morbidity
After one year, 6.8% (435/6420) of the patients enrolled in the study died between the enrolment and the 1-year follow-up visit (

| Multivariate analysis
A Cox proportional hazard model was compiled to establish clinical factors associated with the composite outcome of stroke/TIA/peripheral embolism and/or death (  Given that many elderly AF patients are asymptomatic, opportunistic screening is recommended for early AF detection in those aged ≥65 years. 24 Likewise, the rhythm control strategy should be recommended for symptomatic patients to mitigate their symptoms and improve the quality of life. 24 In the elderly, rate control is often the management of choice 25 ; while rhythm control may be a preferable strategy among younger AF patients (aged <65 years), resulting in a higher rate of sinus rhythm restoration and a lower risk of all-cause mortality than rate control strategy. 26 An increasingly common approach is to use catheter ablation as first-line treatment to reduce AF-related adverse clinical outcomes among patients with recently diagnosed AF, with superior results compared to antiarrhythmic drugs. [27][28][29] The ChiOTEAF registry showed that the overall use of OACs was relatively low (38% of patients at follow-up), with similar uptake of a VKA (18%) and NOACs (20%-21%) among Chinese elderly.  Despite guideline recommendations, we found that antiplatelet therapy (commonly aspirin) was still used in 23.7% of low-risk and 41% of high-risk patients. When a NOAC was discontinued, over a fifth of patients was started on antiplatelet therapy. However, the reasons for this antiplatelet "overuse" in Chinese patients are not evidence-based; indeed, OACs were found to have superior efficacy with similar safety than aspirin among the elderly with AF. 8,35 In the EORP-AF Long-Term Registry, antiplatelet therapy was prescribed in 20% of patients, while 6.4% had no antithrombotic treatment. 32 The Given that stroke prevention is central to AF management, better education and awareness are needed to improve outcomes in this AF population. 37 Indeed, guideline-driven anticoagulation is related to significantly better outcomes in the elderly (including

| Limitations
The primary limitation of the study is its observational nature, and given its modest size, it was not powered to detect differences in some endpoints. Patients were enrolled in 44 centers, which implies a potential variability in the therapeutic strategies for AF.
Moreover, the enrolment period was relatively long, which may affect the generalizability of the results. There was a moderate proportion of patients lost to follow-up (9.3%) consistent with large European registries. 49 Also, the causes of 67 deaths (15.4%) are unknown, and 919 patients have unknown (182 patients)/missing data (737 patients) of the AF type. Finally, data on anticoagulation control are not currently available for this cohort and cannot be considered in this analysis.

| CON CLUS IONS
The Optimal Thromboprophylaxis in Elderly Chinese Patients with Atrial Fibrillation registry provides contemporary data on AF management, including stroke prevention. The rate of OAC use was <40%, and antiplatelet therapy is still commonly prescribed among high-risk patients. Given that Chinese patients with AF are increasingly elderly and overburdened with multimorbidity, our large cohort data may help establish best practices to reduce morbidity and mortality.

| Clinical perspectives
Stroke prevention is central to AF management; better education and awareness are needed to improve outcomes in high-risk AF populations.
Given the substantial clinical impact and healthcare burden associated with AF, the collection of prospective data from local AF cohorts may help establish best practice to reduce AF-related morbidity and mortality.

ACK N OWLED G M ENTS
A list of ChiOTEAF investigators in the appendix.