Abstract
Background
Although anticoagulants are indicated for many elderly patients with non-valvular atrial fibrillation (NVAF), some patients do not receive anticoagulant therapy, whose characteristics and outcomes are diverse.
Methods and results
In this sub-analysis of the All Nippon AF In the Elderly (ANAFIE) Registry, the phenotypes of patients who were not receiving anticoagulants at baseline were evaluated by cluster analysis using Ward’s linkage hierarchical algorithm. Of 32,275 enrolled patients, 2445 (7.6%) were not receiving anticoagulants. Two clusters were identified: (1) elderly paroxysmal AF (PAF) patients with a high proportion of catheter ablation history (57%) and (2) very elderly patients with a high prevalence of previous major bleeding (43%). Respective mean ages were 80.9 and 84.2 years, mean CHA2DS2-VASc scores were 3.8 and 4.9, PAF prevalences were 100.0% and 31.4%, proportions of patients with catheter ablation history were 21.0% and 7.9%, and proportions of patients with a history of major bleeding were 4.0% and 10.8%. Annual incidence rates were 2.72% and 8.81% for all-cause death, 1.66% and 5.85% for major adverse cardiovascular or neurological events, 1.08% and 3.30% for stroke or systemic embolism, and 0.69% and 1.19% for major bleeding, respectively.
Conclusions
In this cohort of elderly NVAF patients from the ANAFIE Registry who were not receiving anticoagulants, over half had PAF with a high proportion of catheter ablation history and a low incidence of adverse outcomes; for them, non-prescription of anticoagulants may be partially understandable, but they should be carefully monitored regarding AF burden or atrial cardiomyopathy and be adequately anticoagulated when adverse findings are detected. The remaining were very elderly patients with a high prevalence of previous major bleeding and a high incidence of adverse outcomes; for them, non-prescription of anticoagulants is inappropriate because of the high thromboembolic risk.
Trial registration
Registration: http://www.umin.ac.jp/; Unique identifier: UMIN000024006.
Introduction
Atrial fibrillation (AF) is one of the most common arrhythmias, and it is associated with a substantially increased risk of mortality and morbidity, especially ischemic stroke and heart failure [1]. Advancing age is a critical risk factor for AF, and because of population aging, the incidence and prevalence of AF are currently increasing worldwide [2]. Anticoagulation therapy can reduce the risk of ischemic stroke in AF patients by two-thirds but is associated with an increased risk of bleeding [3]. Of note, advanced age increases the risk of both ischemic stroke and bleeding in patients receiving anticoagulation therapy [4, 5], with the increased bleeding risk being one of the reasons for the underuse of anticoagulation therapy in elderly patients [6].
Before the introduction of direct oral anticoagulants (DOACs), the prescription rate of anticoagulants for patients with AF was approximately 50% in a Japanese clinical practice registry (Fushimi-AF) [6]. In a pooled analysis of three AF registries in Japan (Shinken, J-RHYTHM and Fushimi-AF), the incidence of ischemic stroke in patients not treated with anticoagulants despite having a high thromboembolic risk (CHADS2 score ≥2) was 2.47% per year [7]. In the main analysis of the nationwide All Nippon AF In the Elderly (ANAFIE) Registry, which enrolled elderly (age ≥75 years) patients with AF (n = 32,275), the anticoagulant prescription rate was 92.4% [8], which was remarkably higher than that in the pre-DOAC era [6]. In this registry, the incidence of ischemic stroke in elderly patients with AF was 1.74%, 1.50%, and 1.09% per year in patients who received no anticoagulation therapy, warfarin, or DOACs, respectively [8].
Although a high anticoagulant prescription rate was reported in the ANAFIE Registry, it is unclear whether this should be further improved. The decision to withhold anticoagulants from elderly patients with AF who potentially have a high risk of thromboembolism should be based on robust clinical grounds. One reason not to prescribe anticoagulants would be the physician’s concern about higher bleeding risk, irrespective of the thromboembolic risk [9]. Another reason may be the lower risk of AF-associated thromboembolism compared with bleeding risk [10]. Thus, it is likely that there may be two phenotypes of elderly patients with AF who are not treated with anticoagulation therapy: those with a very low thromboembolic risk and those with a very high bleeding risk.
Several cluster analyses have reportedly identified up to seven clinical phenotypes in patients with AF, with many of these analyses indicating that rates of mortality or major adverse cardiovascular or neurological events (MACNE) could be stratified by cluster [11–16]. For example, in an analysis of patients enrolled in the ORBIT-AF Registry, the incidence rates of MACNE were 2.58%, 3.97%, 5.10%, and 6.12% per year for clusters defined by low comorbidity rates, younger age or comorbid behavioral disorders, device implantation, or atherosclerotic comorbidities, respectively [11].
To our knowledge, no cluster analyses have been performed in elderly patients with AF who were not receiving anticoagulation therapy. Therefore, we performed a cluster analysis for patients enrolled in the ANAFIE Registry who were not receiving treatment with anticoagulants at baseline to identify a better clinical management for this population.
Methods
Study design and participants
The ANAFIE Registry (UMIN Clinical Trials Registry UMIN000024006) was a multicenter, prospective, observational study of elderly (age ≥75 years) Japanese patients with non-valvular AF (NVAF). The registry was conducted in accordance with the Declaration of Helsinki and local regulatory requirements and ethical guidelines for clinical studies in Japan. All participants provided written informed consent prior to enrollment. The study was approved by the Ethics Committees of The Cardiovascular Research Institute (Tokyo, Japan) and is registered with the University hospital Medical Information Network with the identifier UMIN000024006. The study began recruitment in October 2016, and participants were followed up for a minimum of 2 years. Details of the study design and eligibility criteria have been published previously [17].
Outcomes
The outcomes of interest in this analysis were stroke or systemic embolic events (SEE), major bleeding (as defined by the International Society on Thrombosis and Haemostasis), intracranial hemorrhage (ICH), all bleeding, death from cardiovascular disease, all-cause death, heart failure requiring hospitalization, and MACNE. MACNE is a composite of cardiovascular death, stroke, SEE, and myocardial infarction [11, 13]. In this analysis, the lack of data precluded the inclusion of transient ischemic attack (TIA) in MACNE.
Patient categorization
The study participants were categorized via the following steps.
1) Data preparation
For this analysis, predicted rather than measured values were used for continuous variables to avoid the impact of outliers and missing data (data were missing for the following continuous variables: body mass index [14.1%], systolic blood pressure [10.3%], hemoglobin [18.7%], glycated hemoglobin [48.4%], estimated creatinine clearance [CrCl, 24.5%], and number of medications [4.6%]). Multiple regression models were developed using age, sex, and other categorical variables, and values for each continuous variable were predicted. For the proportion of categorical variables (no missing data, except for patients with a fall within 1 year before enrollment [missing: 11.3%]), predictive probabilities were calculated by multivariate logistic regression analysis using all categorical variables (other than the predicted variable) and age. These were used in the cluster analysis as an alternative to the categorical variable, providing a gradual distribution instead of a dichotomized category. The predicted value of a continuous variable can be interpreted to represent the patient characteristics that are related to the variable rather than the value itself.
2) Hierarchical cluster analysis
A hierarchical cluster analysis was performed using Ward’s linkage hierarchical algorithm on predicted values for continuous variables and predictive probabilities for categorical variables [11].
Statistical methods
Baseline characteristics are described using summary statistics, with mean ± standard deviation for continuous variables and n (%) for categorical variables. The risk of ischemic stroke was determined for each patient at baseline using the CHADS2 score (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, and prior stroke, TIA, or thromboembolism), the CHA2DS2-VASc score (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke, TIA, or thromboembolism, vascular disease, age 65–74 years, and sex category), and the HELT-E2S2 score (hypertension, elderly [75–84 years], low body mass index, type of AF, extremely elderly [≥85 years], and previous stroke) [18]. The risk of major bleeding was determined for each patient at baseline using the HAS-BLED score (hypertension, abnormal renal and liver function, stroke, bleeding, labile international normalized ratio, elderly, and drugs or alcohol). Statistical differences between the two clusters were tested using the unpaired t-test for continuous variables and Fisher’s exact test for categorical variables. The cumulative incidences of stroke/SEE, all bleeding, major bleeding, ICH, cardiovascular death, all-cause death, and MACNE during the 2-year follow-up were depicted using the Kaplan–Meier method, and the statistical differences between the two clusters were tested using the log-rank test. All statistical analyses were conducted using SPSS version 28.0 (IBM Corp., Armonk, NY, USA) and STATA 13.0 (STATA Corp., College Station, Texas, USA).
Results
Patients and clusters
Among the 32,275 patients enrolled in the ANAFIE Registry, 2445 (7.6%) patients who were not receiving anticoagulation therapy at baseline were included in this analysis. The baseline characteristics of this subgroup of patients are summarized in Table 1. The average age was 82.3 years, and 1273 (52.1%) were male. Two clusters of patients were identified (Fig 1): (1) elderly patients with paroxysmal AF (PAF) and a high proportion of catheter ablation history; and (2) very elderly patients with a high prevalence of previous major bleeding.
Table 1. Patient characteristics.
| Total | Elderly PAF with | Very elderly with | p value | |
|---|---|---|---|---|
| high prevalence of | high prevalence of | |||
| catheter ablation | bleeding history | |||
| history | ||||
| n = 2445 | n = 1388 | n = 1057 | ||
| Men | 1273 (52.1) | 709 (51.1) | 564 (53.4) | 0.270 |
| Age, years | 82.3 ± 5.5 | 80.9 ± 4.8 | 84.2 ± 5.8 | <0.001 |
| 75 to <80 | 893 (36.5) | 631 (45.5) | 262 (24.8) | <0.001 |
| 80 to <85 | 725 (29.7) | 422 (30.4) | 303 (28.7) | |
| 85 to <90 | 547 (22.4) | 256 (18.4) | 291 (27.5) | |
| 90 to <95 | 224 (9.2) | 71 (5.1) | 153 (14.5) | |
| 95 to <100 | 51 (2.1) | 8 (0.6) | 43 (4.1) | |
| > = 100 | 5 (0.2) | 0 (0) | 5 (0.5) | |
| > = 85 | 827 (33.8) | 335 (24.1) | 492 (46.5) | <0.001 |
| BMI, kg/m2 | 22.7 ± 3.5 | 22.7 ± 3.4 | 22.6 ± 3.7 | 0.674 |
| SBP, mmHg | 130 ± 17.5 | 131.6 ± 16.4 | 127.9 ± 18.6 | <0.001 |
| DBP, mmHg | 70.4 ± 11.5 | 70.9 ± 10.8 | 69.8 ± 12.3 | 0.037 |
| HbA1c, % | 6 ± 0.8 | 5.9 ± 0.6 | 6.2 ± 1 | <0.001 |
| Creatinine clearance, mL/min | 45.6 ± 18.2 | 49.8 ± 16.9 | 40.4 ± 18.5 | <0.001 |
| <15 or dialysis | 64 (2.6) | 15 (1.1) | 49 (4.6) | <0.001 |
| >15 to <30 | 307 (12.6) | 97 (7) | 210 (19.9) | |
| >30 to <50 | 737 (30.1) | 402 (29) | 335 (31.7) | |
| >50 to <80 | 675 (27.6) | 466 (33.6) | 209 (19.8) | |
| >80 | 64 (2.6) | 41 (3) | 23 (2.2) | |
| CHADS2 score | 2.6 ± 1.2 | 2.3 ± 0.9 | 3.2 ± 1.2 | <0.001 |
| CHA2DS2-VASc score | 4.3 ± 1.4 | 3.8 ± 1.1 | 4.9 ± 1.5 | <0.001 |
| HELT-E2S2 score | 3.0 ± 1.3 | 2.4 ± 1.0 | 3.8 ± 1.2 | <0.001 |
| HAS-BLED score | 2.0 ± 0.9 | 1.7 ± 0.8 | 2.2 ± 1.0 | <0.001 |
| History of major bleeding | 170 (7.0) | 56 (4.0) | 114 (10.8) | <0.001 |
| AF type | <0.001 | |||
| Paroxysmal | 1720 (70.3) | 1388 (100) | 332 (31.4) | |
| Persistent/long persistent | 255 (10.4) | 0 (0.0) | 255 (24.1) | |
| Permanent | 180 (7.4) | 0 (0.0) | 180 (17.0) | |
| Unknown | 290 (11.9) | 0 (0.0) | 290 (27.4) | |
| Non-pharmacological AF therapy | 580 (23.7) | 390 (28.1) | 190 (18) | <0.001 |
| Catheter ablation | 376 (15.4) | 292 (21.0) | 84 (7.9) | <0.001 |
| Electrical defibrillation | 40 (1.6) | 22 (1.6) | 18 (1.7) | 0.872 |
| ICD | 17 (0.7) | 10 (0.7) | 7 (0.7) | 0.999 |
| Pacemaker | 189 (7.7) | 93 (6.7) | 96 (9.1) | 0.032 |
| Others | 13 (0.5) | 5 (0.4) | 8 (0.8) | 0.261 |
| Comorbidities | ||||
| Heart failure | 752 (30.8) | 238 (17.1) | 514 (48.6) | <0.001 |
| Myocardial infarction | 138 (5.6) | 28 (2.0) | 110 (10.4) | <0.001 |
| Hypertension | 1821 (74.5) | 1015 (73.1) | 806 (76.3) | 0.083 |
| Diabetes mellitus | 580 (23.7) | 201 (14.5) | 379 (35.9) | <0.001 |
| Chronic kidney disease | 441 (18.0) | 179 (12.9) | 262 (24.8) | <0.001 |
| Dyslipidemia | 1066 (43.6) | 598 (43.1) | 468 (44.3) | 0.564 |
| Cerebrovascular disease | 468 (19.1) | 153 (11.0) | 315 (29.8) | <0.001 |
| Gastrointestinal disease | 805 (32.9) | 412 (29.7) | 393 (37.2) | <0.001 |
| Severe liver disease | 22 (0.9) | 14 (1.0) | 8 (0.8) | 0.666 |
| Active cancer | 284 (11.6) | 164 (11.8) | 120 (11.4) | 0.750 |
| Dementia | 248 (10.1) | 80 (5.8) | 168 (15.9) | <0.001 |
| Fall within 1 year | 167 (6.8) | 67 (4.8) | 100 (9.5) | <0.001 |
| Medication | ||||
| Antiarrhythmic drugs for AF rhythm control | 646 (26.4) | 488 (35.2) | 158 (14.9) | <0.001 |
| Antiarrhythmic drugs for AF rate control | 714 (29.2) | 324 (23.3) | 390 (36.9) | <0.001 |
| Antiplatelet | 840 (34.4) | 386 (27.8) | 454 (43.0) | <0.001 |
| Proton pump inhibitor | 780 (31.9) | 375 (27.0) | 405 (38.3) | <0.001 |
| P-glycoprotein inhibitor | 48 (2.0) | 20 (1.4) | 28 (2.6) | 0.038 |
| Number of medications | 5.9 ± 3.3 | 5.3 ± 3.1 | 6.6 ± 3.4 | <0.001 |
Data are presented as n (%) or mean ± standard deviation.
Abbreviations: AF, atrial fibrillation; BMI, body mass index; DBP, diastolic blood pressure; ICD, implantable cardioverter-defibrillator; SBP, systolic blood pressure.
Fig 1. Dendrogram of the hierarchical cluster analysis.
Abbreviation: PAF, paroxysmal atrial fibrillation.
Cluster 1: Elderly patients with PAF and a high proportion of catheter ablation history (n = 1388, 57%)
Cluster 1 accounted for more than half of the study population. The average age was 80.9 years, and 51.1% were male. PAF accounted for 100.0% of patients in this cluster, and 21.0% had undergone catheter ablation at baseline. Regarding thromboembolic risk, mean scores were 2.3 for CHADS2, 3.8 for CHA2DS2-VASc, and 2.4 for HELT-E2S2. The prevalence of comorbidities was relatively low and included heart failure (17.1%), myocardial infarction (2.0%), cerebrovascular disease (11.0%), diabetes mellitus (14.5%), and history of major bleeding (4.0%). Antiplatelets were prescribed for 27.8% of patients, and the average number of medications prescribed was 5.3.
Cluster 2: Very elderly patients with a high prevalence of previous major bleeding (n = 1057, 43%)
This cluster accounted for less than half of the study population. The average age was 84.2 years, and 53.4% were male. PAF accounted for 31.4% of these patients, and 7.9% had undergone catheter ablation at baseline. Regarding thromboembolic risk, mean risk scores (3.2 for CHADS2, 4.9 for CHA2DS2-VASc, and 3.8 for HELT-E2S2 score) were higher in cluster 2 compared with cluster 1. The prevalence of comorbidities was higher than in cluster 1 and included heart failure (48.6%), myocardial infarction (10.4%), cerebrovascular disease (29.8%), diabetes mellitus (35.9%), and history of major bleeding (10.8%). Antiplatelets were prescribed for 43.0% of these patients, and the average number of medications prescribed was 6.6.
Outcomes
Annual incidence rates for clusters 1 and 2 were 2.72% and 8.81% for all-cause death, 1.66% and 5.85% for MACNE, 1.08% and 3.30% for stroke/SEE, 0.69% and 1.19% for major bleeding, and 2.14% and 7.14% for heart failure requiring hospitalization, respectively (Table 2). Kaplan–Meier curves for outcome events are shown in Fig 2. For all outcomes analyzed other than bleeding events, incidence rates for cluster 2 were significantly higher than those for cluster 1 (log-rank test, p <0.001). Incidence rates of major bleeding and ICH were numerically higher in cluster 2 than in cluster 1, but the difference did not reach statistical significance (p = 0.087 and 0.273, respectively).
Table 2. Annualized incidence rate for clinical outcomes by clusters.
| Total | Elderly PAF with high prevalence of catheter ablation history | Very elderly with high prevalence of bleeding history | |
|---|---|---|---|
| n = 2445 | n = 1388 | n = 1057 | |
| Stroke/SEE | 2.00 (1.58, 2.41) | 1.08 (0.68, 1.48) | 3.30 (2.47, 4.14) |
| Ischemic stroke | 1.74 (1.35, 2.13) | 0.93 (0.56, 1.30) | 2.91 (2.13, 3.69) |
| Hemorrhagic stroke | 0.22 (0.09, 0.36) | 0.19 (0.02, 0.36) | 0.27 (0.03, 0.50) |
| SEE | 0.04 (0.00, 0.11) | 0.00 (0.00, 0.00) | 0.11 (0.00, 0.26) |
| All bleeding | 2.65 (2.17, 3.14) | 2.31 (1.72, 2.89) | 3.14 (2.33, 3.96) |
| Major bleeding | 0.90 (0.62, 1.18) | 0.69 (0.37, 1.01) | 1.19 (0.69, 1.69) |
| ICH | 0.61 (0.38, 0.83) | 0.50 (0.23, 0.77) | 0.75 (0.36, 1.15) |
| GI bleeding | 1.34 (1.00, 1.68) | 1.24 (0.81, 1.67) | 1.47 (0.92, 2.02) |
| All-cause death | 5.26 (4.58, 5.93) | 2.72 (2.09, 3.35) | 8.81 (7.46, 10.16) |
| Cardiovascular death | 1.63 (1.26, 2.01) | 0.65 (0.34, 0.96) | 3.01 (2.22, 3.80) |
| Heart failure admission | 4.19 (3.58, 4.80) | 2.14 (1.57, 2.70) | 7.14 (5.90, 8.38) |
| MACNE | 3.38 (2.84, 3.93) | 1.66 (1.16, 2.16) | 5.85 (4.73, 6.96) |
Notes: Data in the table are the % per patient-year (95% confidence interval). a MACNE was defined as the composite of stroke/SEE, myocardial infarction, and cardiovascular death.
Abbreviations: GI = gastrointestinal; ICH = intracranial hemorrhage; SEE = systemic embolic events
MACNE = major adverse cardiovascular or neurological event
Fig 2. Kaplan–Meier curves for stroke/SEE, all bleeding, major bleeding, intracranial hemorrhage, all-cause death, and major adverse cardiovascular or neurological event.
Abbreviations: PAF, paroxysmal atrial fibrillation; SEE, systemic embolic events; MACNE, major adverse cardiovascular or neurological events.
Discussion
Main findings
In this analysis from the ANAFIE Registry, two distinct clinical phenotypes were identified in a subgroup of patients not receiving anticoagulation therapy at baseline by hierarchical cluster analysis: (1) elderly patients with PAF and a high proportion of catheter ablation history (57%), and (2) very elderly patients with a high prevalence of previous major bleeding (43%). Cluster 1 was characterized by a high prevalence of PAF and a high proportion of patients who underwent catheter ablation at baseline. Patients in cluster 2 were older and had a high comorbidity burden, with a particularly high prevalence of heart failure (48.6%) and a history of major bleeding (10.8%). The incidence of adverse outcomes was significantly higher in cluster 2 than in cluster 1, except for bleeding events.
Cluster 1: Elderly PAF with high proportion of catheter ablation history
This cluster accounted for more than half of the study population. In this cluster, PAF accounted for 100% of the patients, and catheter ablation had already been performed in approximately 20%. As a result, the AF burden seemed to be low, and this may be one of the reasons for the lack of anticoagulant use in this cluster. This speculation could be supported by the lower incidence rate of ischemic stroke (1.08% per year) in these patients, despite the lack of anticoagulation, which was almost identical to that in patients prescribed DOACs in the ANAFIE Registry (1.09%) [8]. A low AF burden, as determined by an implanted device, has been shown to be associated with a low incidence of stroke [10]. However, given that asymptomatic AF is not uncommon in older adults, in clinical practice it can be challenging to identify a low AF burden by symptoms alone [19]. For this cluster, the physicians might have appropriately determined the AF burden as low and consequently selected no anticoagulation therapy, but it is unknown how the AF burden was estimated.
Based on the mean values of conventional risk scores for thromboembolic risk, such as CHADS2 and CHA2DS2-VASc scores of 2.3 and 3.8, respectively, anticoagulation would have been strongly recommended for patients in this cluster. Interestingly, the recently developed risk score of thromboembolic risk, HELT-E2S2 score, indicated modest risk (mean, 2.4) in this cluster. In the J-RISK study [18], the incidence of ischemic stroke was 2.8%/year for CHADS2 ≥2 and 2.3%/year for CHA2DS2-VASc ≥2, but 1.4%/year for a HELT-E2S2 score of 2 points. Based on the risk assessment for ischemic stroke by HELT- E2S2 score, in which 2 points are assigned for age ≥85 years and 1 point for non-PAF, the lack of oral anticoagulant (OAC) use for this cluster may be partially understandable. However, it is essential that elderly AF patients be carefully monitored regarding AF burden or atrial cardiomyopathy, and they should be adequately anticoagulated when adverse findings are detected.
Cluster 2: Very elderly with a high prevalence of previous major bleeding
This cluster accounted for nearly half of the elderly patients not anticoagulated in the ANAFIE Registry. These patients had a higher mean CHADS2 score (3.2), CHA2DS2-VASc score (4.9), and HELT-E2S2 score (3.8), and higher annual incidence of ischemic stroke (2.91%). Thromboembolic risks for patients in this cluster appeared to be similar to those observed in other Japanese cohorts with similar thromboembolic risk scores. For example, in a pooled analysis of three Japanese registries, the annual incidence rate of ischemic stroke in patients with AF not receiving anticoagulation therapy was 2.66% for those with a CHADS2 score of 3 and 4.43% for those with a CHA2DS2-VASc score of 5 [7]. Similarly, in the J-RISK study, the annual incidence rate of ischemic stroke in patients with AF and a HELT-E2S2 score of 4 and without anticoagulation therapy was 3.96% [18].
Despite having a mean HAS-BLED score of <3 (i.e., not considered to be at high risk of bleeding), the patients in this cluster did have high bleeding risk characteristics, such as advanced age (mean 84.2 years), mean CrCl of 40.4 mL/min, and a relatively high prevalence of previous major bleeding (10.8%) and dementia (15.9%). These characteristics are similar to those reported in the ELDERCARE-AF trial [9], which recruited very elderly Japanese patients with high thromboembolic and bleeding risks who were deemed inappropriate candidates for standard-dose anticoagulation by physicians. In that trial, the mean age was 86.6 years, 22.6% had a history of major bleeding, 16.3% had dementia, and the mean CrCl was 36.3 mL/min. The participants also had a mean HAS-BLED score comparable to patients in cluster 2 (2.3 vs 2.2). In the placebo arm of ELDERCARE-AF, the annual incidence rates of major bleeding and ICH were 1.8% and 0.6%, respectively, which were comparable to those in cluster 2 (1.19% and 0.75%, respectively), although the incidence of all bleeding was higher in ELDERCARE-AF compared with this cluster (45.0% vs 3.14%). Although the annual incidence rate of stroke or SEE in cluster 2 (stroke or SEE: 3.30%; ischemic stroke: 2.91%) was lower than that in the placebo arm of ELDERCARE-AF (stroke or SEE: 6.7%; ischemic stroke: 5.9%), the incidence rate was three times higher than that in cluster 1. Therefore, non-prescription of OACs is inappropriate in this cluster because of the high thromboembolic risk.
It seems likely that the decision not to prescribe anticoagulants for patients in this cluster was driven by a high bleeding risk irrespective of thromboembolic risk. In the ELDERCARE-AF trial, very low-dose edoxaban (15 mg) was superior to placebo in preventing stroke or SEE and did not result in a significantly higher incidence of major bleeding than placebo [9]. Of note, very low-dose edoxaban did not increase the risk of life-threatening bleeding, including ICH, but significantly increased gastrointestinal bleeding risk [9]. Given the very similar profiles of patients in this cluster compared with those enrolled in ELDERCARE-AF, very low-dose anticoagulation, such as edoxaban 15 mg, could be carefully considered for the cohort of very elderly patients with a significant bleeding history that was identified in this analysis. Moreover, in such high-risk AF patients, left atrial appendage closure may also be considered for effective stroke prevention with minimized bleeding events [20].
Comparison to previous cluster analyses
Previous Japanese analyses in patients with AF have identified clusters of patients with PAF that show some similarities to cluster 1 in this analysis. For example, an analysis of the KiCS-AF Registry identified a cluster of younger patients with PAF that accounted for almost half (48.4%) of the total patients enrolled in the registry [11]. Patients in this cluster had a mean age of 65.9 years, and almost all had PAF (96.5%). The cumulative incidence of MACNE at 1 year for this cluster was approximately 2% based on Kaplan–Meier analysis. Similarly, a cluster of younger patients (4.9% of all patients) was identified in the Fushimi-AF Registry [13]. Patients in this cluster had a mean age of 48.3 years, and 76.6% had PAF. The incidence of MACNE in this cluster was only 0.5% per year. Cluster 1 in the present analysis had a mean age of 80.9 years and the prevalence of PAF was 100.0%. Despite their higher age, the incidence of MACNE in this cluster was 1.66% per year, which is similar to that in the younger paroxysmal AF cluster in the KiCS-AF Registry [11] but higher than that in the younger age cluster in the Fushimi-AF Registry [13].
Similar to cluster 2, a very elderly patient cluster was also identified in the Fushimi-AF Registry (6.1% of the total patients) [13]. Patients in this cluster had a similar mean age to those in cluster 2 (83.4 vs 84.2 years), but the proportion of patients with a history of major bleeding was considerably lower (1.9% vs 10.8%). The prescription rate for OACs was 48.5% in the very elderly cluster in Fushimi-AF compared with 0% in cluster 2. The annual incidence rates for all-cause death and major bleeding were higher in this cluster in Fushimi-AF compared with cluster 2 (all-cause death: 15.9% vs 8.81%; major bleeding 3.8% vs 1.19%). Lower annual incidence rates were also reported for cluster 2 in this analysis vs the very elderly cluster in Fushimi-AF for MACNE (5.85% vs 11.35%) and stroke/SEE (3.30% vs 7.04%).
Limitations
Some limitations of this study warrant mention, including those inherent to the observational study design. The incidence of major bleeding in the overall ANAFIE Registry was lower than expected, which may have also affected the results of this sub-analysis. In addition, data were missing for some continuous variables; however, these were replaced with predicted values. Finally, our findings are limited to Japanese patients eligible for outpatient management and cannot be generalized to other ethnic populations.
Conclusions
Elderly patients with NVAF who were not receiving OACs at baseline in the ANAFIE Registry were phenotypically heterogeneous. Two clusters of patients were identified; one had a low incidence of adverse outcomes during follow-up and primarily comprised patients with PAF. This cluster accounted for over half of the overall population of unanticoagulated patients in ANAFIE. For this cluster, non-prescription of anticoagulants may be partially understandable, but they should be carefully monitored regarding AF burden or atrial cardiomyopathy and be adequately anticoagulated when adverse findings are detected. The other cluster comprised the remaining patients (i.e., very elderly patients with a high prevalence of major bleeding) and had a high incidence of adverse outcomes. For this cluster, non-prescription of anticoagulants is inappropriate because of the high thromboembolic risk.
Supporting information
(DOCX)
(PDF)
(PDF)
(PDF)
(PDF)
(PDF)
(PDF)
(PDF)
Acknowledgments
The authors wish to thank Michelle Belanger, MD, and Stephanie Carter of Edanz (www.edanz.com), for providing medical writing support in accordance with Good Publication Practice (GPP3) guidelines (http://www.ismpp.org/gpp3). In addition, the authors thank Daisuke Chiba, of Daiichi Sankyo Co., Ltd., for supporting preparation of the manuscript.
Institutional review board information
Ethical approval was obtained from all relevant institutional review boards, and all patients provided written informed consent and were free to withdraw from the registry at any time. The name of principal ethics committee was The Ethics Committees of The Cardiovascular Institute (Tokyo, Japan) and the number was 299.
Abbreviations and Acronyms
- AF
atrial fibrillation
- ANAFIE
All Nippon AF In the Elderly
- CrCl
creatinine clearance
- DOAC
direct oral anticoagulant
- ICH
intracranial hemorrhage
- MACNE
mortality or major adverse cardiovascular or neurological event
- NVAF
non-valvular atrial fibrillation
- OAC
oral anticoagulant
- PAF
paroxysmal atrial fibrillation
- SEE
systemic embolic events
- TIA
transient ischemic attack
Data Availability
All relevant data are within the paper and its Supporting Information files.
Funding Statement
This study was supported by Daiichi Sankyo Co., Ltd.
References
- 1.Benjamin E.J., et al. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation 98 (1998) 946–952. doi: 10.1161/01.cir.98.10.946 [DOI] [PubMed] [Google Scholar]
- 2.Kornej J., Borschel C.S., Benjamin E.J. & Schnabel R.B. Epidemiology of Atrial Fibrillation in the 21st Century: Novel Methods and New Insights. Circ Res 127 (2020) 4–20. doi: 10.1161/CIRCRESAHA.120.316340 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Hart R.G., Pearce L.A. & Aguilar M.I. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med 146 (2007) 857–867. doi: 10.7326/0003-4819-146-12-200706190-00007 [DOI] [PubMed] [Google Scholar]
- 4.Hindricks G., 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): The Task Force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) Developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J 42 (2021) 373–498. doi: 10.1093/eurheartj/ehaa612 [DOI] [PubMed] [Google Scholar]
- 5.Otto C.M., et al. 2020 ACC/AHA Guideline for the Management of Patients With Valvular Heart Disease: Executive Summary: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation 143 (2021) e35–e71. doi: 10.1161/CIR.0000000000000932 [DOI] [PubMed] [Google Scholar]
- 6.Akao M., et al. Inappropriate use of oral anticoagulants for patients with atrial fibrillation. Circ J 78 (2014) 2166–2172. doi: 10.1253/circj.cj-14-0344 [DOI] [PubMed] [Google Scholar]
- 7.Suzuki S., et al. Incidence of ischemic stroke in Japanese patients with atrial fibrillation not receiving anticoagulation therapy—pooled analysis of the Shinken Database, J-RHYTHM Registry, and Fushimi AF Registry. Circ J 79 (2015) 432–438. doi: 10.1253/circj.CJ-14-1131 [DOI] [PubMed] [Google Scholar]
- 8.Yamashita T., et al. Two-year outcomes of more than 30 000 elderly patients with atrial fibrillation: results from the All Nippon AF In the Elderly (ANAFIE) Registry. Eur Heart J Qual Care Clin Outcomes (2021) doi: 10.1093/ehjqcco/qcab025 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Okumura K., et al. Low-Dose Edoxaban in Very Elderly Patients with Atrial Fibrillation. N Engl J Med 383 (2020) 1735–1745. doi: 10.1056/NEJMoa2012883 [DOI] [PubMed] [Google Scholar]
- 10.Kaplan R.M., et al. Stroke Risk as a Function of Atrial Fibrillation Duration and CHA2DS2-VASc Score. Circulation 140 (2019) 1639–1646. doi: 10.1161/CIRCULATIONAHA.119.041303 [DOI] [PubMed] [Google Scholar]
- 11.Inohara T., et al. Association of Atrial Fibrillation Clinical Phenotypes With Treatment Patterns and Outcomes: A Multicenter Registry Study. JAMA Cardiol 3 (2018) 54–63. doi: 10.1001/jamacardio.2017.4665 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Inohara T., et al. A Cluster Analysis of the Japanese Multicenter Outpatient Registry of Patients With Atrial Fibrillation. Am J Cardiol 124 (2019) 871–878. doi: 10.1016/j.amjcard.2019.05.071 [DOI] [PubMed] [Google Scholar]
- 13.Ogawa H., et al. Characteristics and clinical outcomes in atrial fibrillation patients classified using cluster analysis: the Fushimi AF Registry. Europace 23 (2021) 1369–1379. doi: 10.1093/europace/euab079 [DOI] [PubMed] [Google Scholar]
- 14.Watanabe E., et al. Clinical phenotypes of patients with non-valvular atrial fibrillation as defined by a cluster analysis: A report from the J-RHYTHM registry. Int J Cardiol Heart Vasc 37 (2021) 100885. doi: 10.1016/j.ijcha.2021.100885 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Vitolo M., Proietti M., Shantsila A., Boriani G. & Lip G.Y.H. Clinical Phenotype Classification of Atrial Fibrillation Patients Using Cluster Analysis and Associations with Trial-Adjudicated Outcomes. Biomedicines 9 (2021) doi: 10.3390/biomedicines9070843 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Suzuki S., et al. Identifying risk patterns in older adults with atrial fibrillation by hierarchical cluster analysis: A retrospective approach based on the risk probability for clinical events. IJC Heart & Vasculature 37 (2021) 100883. 10.1016/j.ijcha.2021.100883. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Inoue H., et al. Prospective observational study in elderly patients with non-valvular atrial fibrillation: Rationale and design of the All Nippon AF In the Elderly (ANAFIE) Registry. J Cardiol 72 (2018) 300–306. doi: 10.1016/j.jjcc.2018.02.018 [DOI] [PubMed] [Google Scholar]
- 18.Okumura K., et al. A Novel Risk Stratification System for Ischemic Stroke in Japanese Patients With Non-Valvular Atrial Fibrillation. Circ J 85 (2021) 1254–1262. doi: 10.1253/circj.CJ-20-1075 [DOI] [PubMed] [Google Scholar]
- 19.Boriani G., et al. Asymptomatic atrial fibrillation: clinical correlates, management, and outcomes in the EORP-AF Pilot General Registry. Am J Med 128 (2015) 509–518.e502. doi: 10.1016/j.amjmed.2014.11.026 [DOI] [PubMed] [Google Scholar]
- 20.Kar S., et al. Primary Outcome Evaluation of a Next-Generation Left Atrial Appendage Closure Device: Results From the PINNACLE FLX Trial. Circulation 143 (2021) 1754–1762. doi: 10.1161/CIRCULATIONAHA.120.050117 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(DOCX)
(PDF)
(PDF)
(PDF)
(PDF)
(PDF)
(PDF)
(PDF)
Data Availability Statement
All relevant data are within the paper and its Supporting Information files.