Validity of HELT-E2S2 Score in Predicting Ischemic Stroke in Japanese Patients With Non-Valvular Atrial Fibrillation With Cardiac Implantable Electronic Devices

Miyo Nakano; Yusuke Kondo; Yuki Shiko; Masahiro Nakano; Takatsugu Kajiyama; Ryo Ito; Toshinori Chiba; Yutaka Yoshino; Satoko Ryuzaki; Yukiko Takanashi; Yuya Komai; Shoko Narumi; Yoshio Kobayashi

doi:10.1253/circj.CJ-24-0715

Abstract

Background: The HELT-E₂S₂ score, a novel risk stratification system, was developed to determine the incidence of ischemic stroke in Japanese patients with non-valvular atrial fibrillation (NVAF). It has been suggested that the HELT-E₂S₂ score is more useful than the CHADS₂ and CHA₂DS₂-VASc scores for Japanese patients with NVAF. This study determined the incidence of ischemic stroke in patients with NVAF and cardiac implantable electronic devices (CIEDs) and assessed the validity of the HELT-E₂S₂ score in this population.

Methods and Results: We retrospectively analyzed the database of the CIED clinic of Chiba University Hospital and investigated the incidence of ischemic stroke according to the HELT-E₂S₂ score. Of the 730 consecutive patients who were followed-up at the CIED clinic, those without NVAF were excluded, leaving 362 patients in this study (mean [±SD] follow-up period, 64±48 months; mean age, 73±16 years; 65% male). The mean CHADS₂ and CHA₂DS₂-VASc scores were 1.8±1.2 and 2.8±1.6 points, respectively. During follow-up, 31 (8.6%) patients experienced ischemic stroke. The c-statistic for the HELT-E₂S₂ score was 0.719 (95% confidence interval [CI] 0.657–0.795), which was higher than the c-statistics for the CHADS₂ (0.704; 95% CI 0.647–0.768; P=0.025) and CHA₂DS₂-VASc (0.700; 95% CI:0.621–0.747; P=0.0097) scores.

Conclusions: Risk stratification for ischemic stroke using the HELT-E₂S₂ score is valid in Japanese patients with NVAF and CIEDs.

Atrial fibrillation (AF) is the most prevalent form of arrhythmia worldwide, affecting 0.5% of the global population. The incidence and prevalence of AF have increased over the past 20 years.¹ In patients with non-valvular AF (NVAF), the accumulation of risk factors for thromboembolism increases the incidence of ischemic stroke.²^,³ Therefore, determining appropriate anticoagulation therapy through a risk assessment for thromboembolism is recommended.⁴^–⁷

The HELT-E₂S₂ score was developed as a novel risk stratification strategy for ischemic stroke in Japanese patients with NVAF based on data from 5 major Japanese registries, namely the J-RHYTHM Registry, Fushimi AF Registry, Shinken Database, Keio Interhospital Cardiovascular Studies, and Hokuriku-Plus AF Registry.⁸ The factors of the HELT-E₂S₂ score have been weighted according to their relevance to the Japanese population, which comprises a significant proportion of older persons.⁸^,⁹ It has been proposed that the HELT-E₂S₂ score is a more valuable tool than the CHADS₂ and CHA₂DS₂-VASc scores for thromboembolic risk stratification in Japanese patients with NVAF.⁸

In March 2024, the Japanese Circulation Society (JCS)/Japanese Heart Rhythm Society (JHRS) 2024 guideline focused update on the management of cardiac arrhythmias was released.¹⁰ The guideline introduced the HELT-E₂S₂ score as an ischemic stroke assessment tool unique to Japan and recommended its use (Class IIa recommendation) for predicting the risk of ischemic stroke. Yamauchi et al. reported on the validation of the HELT-E₂S₂ score in Japanese patients with AF based on data from the RAFFINE and SAKURA studies.¹¹ However, few other reports have validated the HELT-E₂S₂ score as a risk assessment tool for ischemic stroke in patients with NVAF, and further external validations are needed in Japanese cohorts and cohorts in other Asian countries to confirm its efficacy.¹² In addition, little is known about the assessment of the HELT-E₂S₂ score in Japanese patients with NVAF who have undergone long-term follow-up analysis with a cardiac implantable electronic device (CIED). Therefore, the aims of this study were to determine the incidence of ischemic stroke in Japanese NVAF patients with CIEDs and to assess the validity of the HELT-E₂S₂ score in this population. Recent CIEDs exhibit improved early detection of AF episodes, especially in patients who are asymptomatic.¹³^,¹⁴ Moreover, patients with a CIED regularly attend hospital for device checks. This enables a relatively accurate evaluation of the validity of the HELT-E₂S₂ score in patients with a CIED over a long period of time.

Methods

Study Participants and Implanted Devices

We retrospectively analyzed the database of the CIED clinic of Chiba University Hospital. Patients who received pacemakers and implantable cardioverter defibrillators (ICDs) between May 1980 and May 2017 were initially enrolled in the study. All patients received single- or dual-chamber CIEDs from Medtronic (Minneapolis, MN, USA), Abbott (St. Paul, MN, USA; now Abbot), or Biotronik (Berlin, Germany), based on a Class I or IIa indication according to the JCS criteria.¹⁵ The devices were programmed to the nominal setting of each device, and atrial high-rate episodes (AHREs) were detected when the atrial rate reached 175, 190, and 200 beats/min according to the Medtronic, Abbott, and Biotronik devices, respectively. Atrial sensitivity was programmed to the nominal setting. For pacemakers, the atrial sensitivity was set at ≤0.5 mV, whereas in ICDs it was automatically controlled to adjust the atrial sensing thresholds. However, the setting of a CIED could be changed by the primary doctor depending on a patient’s condition.

This study included patients who had been diagnosed with NVAF at the time of CIED implantation and those with an episode of AF confirmed by a CIED. Patients who had no history of AF or no episode of AF detected by CIEDs were excluded. Patients with valvular AF were also excluded.

Diagnosis of AF and Definition of Early Diagnosed AF Based on AHREs

A history of AF was defined as AF documented by 12-lead electrocardiography (ECG) or 24-h Holter ECG monitoring. AHREs were checked by the CIEDs every 6 months after implantation. AHREs captured by the CIEDs were reviewed by at least 2 experienced cardiac electrophysiologists. We confirmed that each identified AHRE was an episode of AF, ruling out lead noise or regular supraventricular tachycardia. According to the JCS/JHRS 2021 guideline focused update, several studies indicate a close association between CIED-detected AF and an increased risk of stroke; however, the cut-off value for AF duration remains unclear.¹⁶^,¹⁷ Nonetheless, we have previously reported that a 30-s duration was the optimal cut-off time for AF detected by CIEDs in relation to ischemic stroke.¹⁸ The cut-off value for the frequency of AF detected by CIED was <0.1%, and analysis by frequency was considered statistically intolerable. Moreover, a systematic review and meta-analysis conducted in 2021 to identify the thromboembolic risk associated with AHRE thresholds suggested that AHRE detected by CIEDs with a duration ≥30 s were associated with an increased risk of ischemic stroke.¹⁹ Therefore, a duration of 30 s was deemed the optimal cut-off time for AF detected by CIEDs in relation to ischemic stroke. Thus, in the present study, an episode of AF was defined as AF detected by CIEDs lasting for ≥30 s.

Data Collection

Clinical data were retrieved from patients’ medical records and, when necessary, through interviews with the patients. For patients who had a history of AF when the CIED was implanted, we collected baseline data at the time of CIED implantation. For patients who did not have a history of AF when the CIED was implanted, we collected baseline data when the first AF was detected by the CIED. The baseline data collected was age, sex, body mass index (BMI), underlying disease, risk factors (including congestive heart failure, hypertension, diabetes), previous stroke/transient ischemic attack (TIA), vascular disease, type of AF (paroxysmal [lasting ≤7 days] or non-paroxysmal AF), chronic obstructive pulmonary disease, blood examination data, echocardiographic measurements, type and setting of a CIED (pacemaker or ICD), antiplatelet therapy, and anticoagulation therapy.

In this study, heart failure was defined as recent heart failure based on the criteria of the CHADS₂ score.⁴ Hypertension was defined as blood pressure ≥140/90 mmHg or the use of antihypertensive drugs. Diabetes was defined as an HbA1c ≥6.5% or receiving treatment for diabetes.

We examined patient characteristics and investigated the incidence of ischemic stroke according to the HELT-E₂S₂ score. The HELT-E₂S₂ score was calculated as follows: 1 point was assigned to each of hypertension (H), elderly (E; age 75–84 years), low BMI (L; BMI <18.5 kg/m²), and type of AF (T; persistent/permanent), and assigning 2 points were assigned to each of extreme elderly (EE; age ≥85 years) and previous stroke (S; ischemic stroke plus TIA).⁸ The incidence rates of ischemic stroke, stratified by HELT-E₂S₂ score, were analyzed separately in patients with and without oral anticoagulants (OACs). The CHADS₂ and CHA₂DS₂-VASc scores were also calculated in all patients to compare with the HELT-E₂S₂ score.⁴^,²⁰

Definition of Study Endpoint

The endpoint of the study was the occurrence of ischemic stroke, including TIA. In the present study, ischemic stroke/TIA was diagnosed by an experienced neurologist based on the patient’s history, symptoms, and imaging findings.

This study was approved by the Ethics Committee of Chiba University, Graduate School of Medicine (Approval no. 2423) and was conducted in accordance with the Declaration of Helsinki and the ethical guidelines for medical and health research involving human subjects in Japan.

Statistical Analysis

We examined patient characteristics and investigated the incidence of ischemic stroke according to the HELT-E₂S₂ score. The HELT-E₂S₂, CHADS₂, and CHA₂DS₂-VASc scores were treated as continuous variables, and concordance statistics (c-statistics) were calculated for each of them from the sensitivity and specificity corresponding to the incidence of ischemic stroke. Bootstrapping with 200 iterations was used to provide a 95% confidence interval (CI) for the c-statistics, which were compared between the HELT-E₂S₂ score and the other 2 scoring systems. The follow-up period in each patient was defined as the period from either the date of implantation for those who had a history of AF at the time of CIED implantation or the date of identification of AF for those who did not to the date of event onset or the end of follow-up.

To investigate the ischemic stroke risk for each HELT-E₂S₂ score, it was treated as a categorical variable (scores of 0, 1, 2, 3, 4, and ≥5). The incidence rate of ischemic stroke in each category was calculated by person-years ×100. Kaplan-Meier curves for each category were created by plotting the cumulative incidence of ischemic stroke during follow-up. A Cox proportional hazards model was used to compute hazard ratios (HRs) and P values for HELT-E₂S₂ scores of 2, 3, 4, and ≥5 vs. a score of 1, because patients in our cohort with a score of 0 had no instances of ischemic stroke. In addition, the incidence rate for each score was examined according to the use or not of OACs. Univariate and multivariate Cox regression analyses for the components of the HELT-E₂S₂ score were performed to specify independent risk factors for ischemic stroke.

Continuous variables are presented as the mean±SD or as the median with interquartile range (IQR). Categorical variables are presented as numbers and percentages. We defined the statistical threshold as P<0.05. SPSS Statistics version 25.0 (IBM Corp., Armonk, NY, USA) was used for all statistical analyses.

Results

Clinical Characteristics of the Study Patients

We initially enrolled 730 consecutive patients who were followed up at the CIED clinic of Chiba University Hospital. After excluding 343 patients who had no history of AF or had no episode of AF detected by CIEDs, 13 patients with valvular AF, and 12 patients lacking data, 362 patients were included in this study. Table 1 presents the baseline characteristics of the patients included in this study. The mean follow-up period was 64±48 months (median 56 months; IQR 42–78) months. The mean age of patients was 73±16 years, and 127 (35%) were female. The mean BMI was 23.2±3.9 kg/m²; 188 (52%) patients had hypertension, 47 (13%) had had a previous stroke, and 206 (56%) had non-paroxysmal AF. In all, 206 (57%) patients received an ICD. The mean CHADS₂, CHA₂DS₂-VASc, and HELT-E₂S₂ scores were 1.8±1.2, 2.8±1.6, and 2.3±1.3 points, respectively. In this study, the most prevalent HELT-E₂S₂ score was 2, recorded for 33% of the study population. During follow-up period, 196 (54%) patients underwent catheter ablation therapy and 6 (1.7%) underwent left atrial appendage closure.

Table 1.

Baseline Characteristics of Study Patients (n=362)

Age (years)	73±16
<75	210 (58)
75–84	120 (33)
≥85	32 (8.8)
Follow-up period (months)	64±48
Female sex	127 (35)
Body mass index (kg/m²)	23.2±3.9
BMI <18.5 kg/m²	21 (5.8)
Congestive heart failure	173 (48)
Hypertension	188 (52)
Diabetes	65 (18)
Previous stroke	47 (13)
Vascular disease	93 (26)
Non-paroxysmal AF	206 (56)
COPD	22 (6.1)
eGFR ≤60 mL/min/1.73 m²	138 (38)
BNP (pg/mL)	78±56
TSH (μIU/mL)	1.7±1.0
FT₄ (ng/mL)	1.2±0.3
Left atrial diameter (mm)	42±12
LVEF (%)	52±17
ICD	206 (57)
Sick sinus syndrome	152 (42)
Atrioventricular block	145 (40)
Medications
Antiplatelet	91 (25)
Oral anticoagulants	216 (60)
Patients diagnosed with AF using CIEDs	122 (34)
CHADS₂ score	1.8±1.2
CHA₂DS₂-VASc score	2.8±1.6
HELT-E₂S₂ score	2.3±1.3
0	26 (7.2)
1	78 (22)
2	120 (33)
3	72 (20)
4	38 (10)
≥5	28 (10)
HAS-BLED score	1.7±1.2

Data are presented as the mean±SD or n (%). AF, atrial fibrillation; BNP, B-type natriuretic peptide; CIED, cardiac implantable electronic device; COPD, chronic obstructive pulmonary disease; eGFR, estimated glomerular filtration rate; FT₄, free thyroxine 4; ICD, implantable cardioverter defibrillator; LVEF, left ventricular ejection fraction; TSH, thyroid-stimulating hormone.

Of the 362 patients in this study, 31 (8.6%) experienced ischemic stroke during the follow-up period. The relationship between the duration of AF detected by CIEDs and the occurrence of ischemic stroke is shown in Figure 1.

Figure 1.

Duration of atrial fibrillation (AF) detected by cardiac implantable electronic devices and ischemic stroke events.

Comparison of CHADS₂, CHA₂DS₂-Vasc, and HELT-E₂S₂ Scores for the Prediction of Ischemic Stroke

To compare the risk stratification schemes for predicting ischemic stroke, c-statistics were evaluated using the following classifications: scores 0–7 for the HELT-E₂S₂ score; scores 0–6 for the CHADS₂ score; and scores 0–9 for the CHA₂DS₂-VASc score. The c-statistic for the HELT-E₂S₂ score was 0.719 (95% CI 0.657–0.795; Table 2), and the c-statistics for the CHADS₂ and CHA₂DS₂-VASc scores were 0.704 (95% CI 0.647– 0.768) and 0.700 (95% CI 0.621–0.747), respectively (Table 2). The c-statistic for the HELT-E₂S₂ score was significantly higher than that for the CHADS₂ (P=0.025 for comparison) and CHA₂DS₂-VASc (P=0.0097 for comparison) scores (Table 2).

Table 2.

C-Statistics and Their Comparison Among Risk Stratification Schemes

Risk stratification scheme	C-statistic (95% CI)	P value (vs. HELT-E₂S₂ score)
HELT-E₂S₂ score	0.719 (0.657–0.795)	–
CHADS₂ score	0.704 (0.647–0.768)	0.025
CHA₂DS₂-VASc score	0.700 (0.62–0.747)	0.0097

C-statistics were evaluated using the following classifications: scores 0–7 for the HELT-E₂S₂ score, scores 0–6 for the CHADS₂ score, and scores 0–9 for the CHA₂DS₂-VASc score. CI, confidence interval.

In addition, this study was a long-term study performed from May 1980 to May 2017; ischemic stroke events may vary over time, which could affect ischemic stroke risk stratification using the HELT-E₂S₂ score. Thus, we divided the patients into 2 groups (before and after 2000) and compared the c-statistics for the HELT-E₂S₂ score between these 2 groups. As indicated in Table 3, there were no significant differences between these 2 groups.

Table 3.

C-Statistics and Their Comparison Between Patients Before and After the Year 2000

HELT-E₂S₂ score	C-statistic (95% CI)	P value
Before 2000	0.715 (0.707–0.7223)	0.41
After 2000	0.728 (0.711–0.728)

C-statistics were evaluated using scores 0–7 for the HELT-E₂S₂ score. CI, confidence interval.

Validation of the HELT-E₂S₂ Score for Ischemic Stroke

When study patients were stratified by HELT-E₂S₂ score, the incidence rate of ischemic stroke per 100 person-years during the follow-up period was 0.00, 0.47, 1.06, 2.27, 3.32, and 4.29 for scores of 0, 1, 2, 3, 4, and ≥5, respectively (Table 4). Kaplan-Meier curves for each score are shown in Figure 2A, indicating that the cumulative incidence of ischemic stroke increased with an increase in the HELT-E₂S₂ score, consistent with the findings of a previous study.⁸ Compared with patients with a HELT-E₂S₂ score of 1, which served as the reference, patients with HELT-E₂S₂ scores of 3, 4, and ≥5 had a significantly higher incidence of stroke (with HRs of 4.92 [95% CI 1.10–22.8, P=0.040], 7.88 [95% CI 1.63–38.0, P=0.0083], and 8.85 [95% CI 1.78–44.0, P=0.0052], respectively; Figure 2B).

Table 4.

Ischemic Stroke Events Stratified by HELT-E₂S₂ Scores

	No. patients with ischemic stroke (incidence rate of ischemic stroke)
HELT-E₂S₂ score
0	0 (0.00)
1	2 (0.47)
2	7 (1.06)
3	9 (2.27)
4	7 (3.32)
≥5	6 (4.29)
Total no. patients	31

Values are presented as the number of patients (per 100 person-years).

Figure 2.

(A) Kaplan-Meier curves for the incidence of ischemic stroke stratified by HELT-E₂S₂ scores. (B) Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated. Patients with an HELT-E₂S₂ score of 0 had no instances of ischemic stroke; therefore, analyzing the data using score 0 as a reference was not possible in the present cohort.

Univariate and multivariate Cox proportional hazards models were used to assess the risk factors for the incidence of ischemic stroke, including components of the HELT-E₂S₂ score. The results of these analyses indicated that all the components of the HELT-E₂S₂ score were independently associated with the incidence of ischemic stroke events (Table 5). Elderly (age 75–84 years), extreme elderly (age ≥85 years), hypertension, previous stroke, type of AF (persistent/permanent), and low BMI were significant risk factors for ischemic stroke even in multivariate Cox proportional hazards models, with HRs of 2.54 (95% CI 1.10–5.88; P=0.028), 7.58 (95% CI 2.57–22.34; P<0.001), 2.95 (95% CI 1.31–6.44; P=0.0089), 3.96 (95% CI 1.97–7.83; P<0.001), 2.31 (95% CI 1.14–4.67; P=0.021), and 2.93 (95% CI 1.27–6.71; P=0.012), respectively (Table 5).

Table 5.

Univariate and Multivariate Analyses by the Cox Proportional Hazards Model

Characteristic	Univariate		Multivariate
Characteristic	HR (95% CI)	P value	HR (95% CI)	P value
Age (<75 years)	Reference		Reference
Age 75–84 years (elderly)	2.09 (1.04–4.21)	0.039	2.54 (1.10–5.88)	0.028
Age ≥85 years (extreme elderly)	4.96 (1.97–12.48)	<0.001	7.58 (2.57–22.34)	<0.001
Female sex	1.69 (0.92–3.08)	0.087	1.58 (0.77–3.27)	0.21
Congestive heart failure	1.46 (0.74–2.88)	0.27	1.27 (0.62–2.63)	0.50
Hypertension	3.64 (1.79–7.42)	<0.001	2.95 (1.31–6.44)	0.0089
Diabetes	1.76 (0.89–3.51)	0.10	1.39 (0.61–3.23)	0.43
Previous stroke	6.01 (3.28–10.99)	<0.001	3.96 (1.97–7.83)	<0.001
Vascular disease	2.17 (1.17–4.02)	0.015	1.51 (1.74–3.10)	0.26
Persistent/permanent AF	1.82 (1.00–3.32)	0.052	2.31 (1.14–4.67)	0.021
Low BMI (<18.5 kg/m²)	7.12 (3.67–13.81)	<0.001	2.93 (1.27–6.71)	0.012
No oral anticoagulant	3.79 (1.87–7.70)	<0.001	1.53 (0.68–3.48)	0.29

AF, atrial fibrillation; BMI, body mass index; CI, confidence interval; HR, hazard ratio.

Analysis of Patients According to OACs Use

Patients with (n=216) and without (n=146) OACs were analyzed separately. The clinical characteristics of patients with and without OACs are summarized in the Supplementary Table. Figure 3 shows the incidence rate of ischemic stroke calculated by person-years ×100, stratified by HELT-E₂S₂ score in patients with and without OACs. Consistent with the findings of a previous study,⁸ incremental increases in ischemic stroke incident rates were observed with increasing HELT-E₂S₂ scores in patients both with and without OACs. Furthermore, as the HELT-E₂S₂ score increased, the ratio of ischemic stroke events increased to a greater degree in patients without than with OACs.

Figure 3.

Incidence rates of ischemic stroke, stratified by HELT-E₂S₂ scores, in patients with and without oral anticoagulants (OACs).

Discussion

The major findings of this study in Japanese patients with NVAF and CIEDs were as follows: (1) 31 (8.6%) patients experienced ischemic stroke during the follow-up period (mean 64±48 months); (2) the c-statistic for the HELT-E₂S₂ score was 0.719, which was significantly higher than that for the CHADS₂ and CHA₂DS₂-VASc scores; and (3) risk stratification for ischemic stroke using the HELT-E₂S₂ score is valid in Japanese patients with NVAF and CIEDs.

Comparison of HELT-E₂S₂ With CHADS₂ and CHA₂DS₂-VASc Scores in Japanese NVAF Patients With CIEDs

The CHA₂DS₂-VASc score is widely used in guidelines or consensus statements as a risk assessment tool for thromboembolism.²¹^–²³ The primary scoring system in Japan is the CHADS₂ score, with the JCS/JHRS 2020 guidelines recommending its use (Class I recommendation) for predicting the risk of thromboembolism.⁶ According to a pooled analysis of 3 Japanese AF registries (J-RHYTHM Registry, Fushimi AF Registry, and Shinken Database), the factors added to the CHA₂DS₂-VASc score (vascular disease, age 65–74 years, and female sex) were not associated with an increased risk of ischemic stroke in Japanese patients with NVAF not receiving anticoagulation therapy.²⁴

The HELT-E₂S₂ score was developed as a novel risk stratification tool for Japanese patients with NVAF based on data from 5 major Japanese registries, known as the J-RISK study: J- RHYTHM Registry, Fushimi AF Registry, Shinken Database, Keio Interhospital Cardiovascular Studies, and Hokuriku-Plus AF Registry.⁸ In the J-RISK study, age, hypertension, and previous stroke were identified as common risk factors with the CHADS₂ and CHA₂DS₂-VASc scores, whereas diabetes, congestive heart failure, and vascular disease were not identified as such. Instead, the following new risk factors were identified: age ≥85 years, BMI <18.5 kg/m², and persistent/permanent AF.⁸ Overall, the HELT-E₂S₂ score incorporates novel region-specific risk factors, including extreme older age (≥85 years), persistent/permanent AF, and low BMI, in addition to the 3 risk factors (hypertension, age, and stroke) that are used for the CHADS₂ score.

The results of the present long-term follow-up study indicate that risk stratification for ischemic stroke using the HELT-E₂S₂ score is valid in Japanese patients with NVAF and CIEDs. The c-statistic for the HELT-E₂S₂ score was significantly higher than that obtained for the CHADS₂ and CHA₂DS₂-VASc scores (Table 2). Although both the CHADS₂ and CHA₂DS₂-VASc scores also demonstrated good performance in risk prediction, the findings of our study suggest that the HELT-E₂S₂ score may be more effective in identifying Japanese patients with NVAF at risk of ischemic stroke.

Components of the HELT-E₂S₂ Score

The HELT-E₂S₂ score comprises 6 factors; congestive heart failure and diabetes are factors included in the CHADS₂ score, but not in the HELT-E₂S₂ score. According to a pooled analysis of the 3 major domestic AF registries (J-RHYTHM Registry, Fushimi AF Registry, and Shinken Database), congestive heart failure and diabetes were not identified as independent risk factors for ischemic stroke in Japanese patients with NVAF, with HRs of 0.86 (95% CI 0.45–1.65) for congestive heart failure and 1.18 (95% CI 0.64–2.15) for diabetes.²⁴ In this study, congestive heart failure and diabetes were not independent risk factors for ischemic stroke. Possible reasons why they were not identified as significant risk factors for ischemic stroke may include differences in current standard management and therapeutic drugs compared with those used in the past. A further reason why the CHADS₂ score component was not identified as a significant risk for ischemic stroke is that the definition of heart failure is ambiguous. It is also influenced by a possible overestimation of AF, which is discussed further below. Furthermore, it should be considered that the CHADS₂ score was originally generated using data from patients from Western countries, not from Japan. As noted above, the definition of heart failure is ambiguous and varies between studies. In the original CHADS₂ score, congestive heart failure was defined as recent heart failure.⁴ Conversely, in the J-RHYTHM Registry, it was defined as recent heart failure or clinically diagnosed heart failure.²⁵ In the Fushimi AF Registry, it was defined as a history of hospitalization for heart failure, New York Heart Association Class II or higher, or a left ventricular ejection fraction of ≤40%.²⁶ Thus, variations in the definition of heart failure between studies have influenced the results, which should be taken into consideration.

Pooled multivariate analysis of the RAFFINE and SAKURA registries revealed that the presence of diabetes was a significant factor for ischemic stroke (HR 1.40; 95% CI 1.09–1.80).¹¹ These results suggest that appropriate control of heart failure and diabetes may be beneficial in preventing thromboembolism in Japanese patients with NVAF.

Regarding the type of AF included in the HELT-E₂S₂ score but not in the CHADS₂ score, some cohort studies have reported that the risk of thromboembolism in patients with persistent or permanent AF is higher than that in those with paroxysmal AF.²⁷^–³⁰ In the JCS/JHRS 2020 guideline, persistent or permanent AF and low body weight are newly listed as “other risks” to be considered when prescribing anticoagulation therapy.⁶ In addition, the risk of adverse events was transiently elevated during the progression period from paroxysmal to sustained AF.³¹ Therefore, when using the HELT-E₂S₂ score, it is better to re-evaluate the type of AF periodically by auscultation and ECG. Furthermore, in NVAF patients with CIEDs, device checks clearly reveal the frequency and duration of AF, which is key for the management of AF. As Figure 1 shows, patients with short AF also have ischemic stroke events. Thus, careful monitoring of CIED information and patient risk management are crucial.

CIED-Detected AF and Stroke Risk

This study included many patients from nearby clinics or affiliated hospitals for CIED implantation, which may have influenced the results. In addition, the study included patients with asymptomatic AF lasting ≥30 s detected by a CIED, which may include many cases of early-stage AF, which may also have influenced the results. Indeed, 122 (34%) patients diagnosed with AF by a CIED were included in this study. Our definition of AF with a CIED may have overestimated patients with AF, which may have influenced the ischemic stroke event rate and risk assessment ability of the HELT-E₂S₂ score. Although AF may have been overestimated, the ischemic stroke event rate was higher than reported in previous studies. This could be attributed to the possibility that patients with a CIED are basically at a higher risk of stroke, and the small number of patients in this study compared with previous studies.⁸^,¹¹

Here, the incidence of acute ischemic stroke among Asian populations has reportedly decreased significantly over a 12-year period from 2005 to 2016.³³ In addition, the Fushimi AF Registry documented a decline in the incidence of stroke/systemic embolism over a 10-year period from 2011 to 2021.³³ Although risk factors and treatments for ischemic stroke may have evolved over time, in this long-term study we found no significant differences in the c-statistics for the HELT-E₂S₂ score comparing data before and after the year 2000 (Table 3). Nevertheless, it remains essential to continuously re-evaluate risk scores for ischemic stroke over time.

The independent risk factors for ischemic stroke identified in this study were comparable to those in the J-RISK study (Table 5). Notably, age ≥85 years, persistent or permanent AF, and low BMI (<18.5 kg/m²), which were specifically identified in Japanese patients in the J-RISK study and form the novel aspects of the HELT-E₂S₂ score, were similarly identified as independent risk factor in the present study (Table 5). Most patients with a CIED regularly attend hospital for device checks and use remote monitoring systems, which allow for early intervention for NVAF. It is important to consider appropriate treatment (OACs, catheter ablation, and left atrial appendage closure among others) for NVAF patients with CIEDs based on the appropriate risk score.

Using the HELT-E₂S₂ Score to Guide OAC Initiation

In this study, the incidence rate of ischemic stroke, stratified by the HELT-E₂S₂ score, increased progressively from a score of 0 to scores of 1, 2, 3, 4, and ≥5 (Table 4). Even when patients with and without OACs were analyzed separately, the results were similar, with the risk increasing as scores increased (Figure 3). In this study, because the HRs of scores ≥3 were significantly higher than that of a score of 1 (reference), Japanese patients with NVAF and HELT-E₂S₂ scores ≥3 are considered to be at a particularly increased risk of ischemic stroke (Figure 2B). However, there was no instance of ischemic stroke among patients with a HELT-E₂S₂ score of 0; therefore, analyzing the data using a score of 0 as a reference was not possible in the current cohort. In the original study, patients with HELT-E₂S₂ scores ≥2 had significantly higher HRs than those with a score of 0, which was used as the reference.⁸ These results suggest that patients with HELT-E₂S₂ scores ≥2 may be considered candidates for OACs. However, further studies are required to establish the optimal cut-off value of the HELT-E₂S₂ score for ischemic stroke. This discussion should be grounded in an analysis of the net clinical benefit, considering the effects of anticoagulation therapy in balancing thromboembolic and bleeding events.

Study Limitations

The present study has some limitations. First, this study had a single-center, retrospective, observational design, with a relatively small number of patients with CIEDs. This limited sample size may have affected the analysis and should be carefully considered when interpreting the results. Second, we focused on patients with CIEDs, and it remains uncertain whether the HELT-E₂S₂ score can be applied to the general population without a CIED or patients from other ethnicities, countries, or regions. Third, only patients with AF detected by CIEDs lasting ≥30 s were included in the present study. The results depend on this definition of AF. Fourth, the risk scoring system should be ideally evaluated among patients without OACs. However, if only patients without OACs had been analyzed, the sample size and number of ischemic stroke events would have been too small to adequately assess the validity of the HELT-E₂S₂ score. Fifth, catheter ablation may also contribute to the incidence of ischemic stroke. In our study, 196 (54%) patients underwent catheter ablation during their clinical course. Performing an appropriate analysis was challenging owing to the variations in the timing of treatment. Sixth, we collected clinical data at the time of CIED implantation for patients who had a history of AF at the time of CIED implantation; therefore, data at follow-up were not considered. Finally, 206 patients with implanted ICDs were included in this study and were not excluded; however, the impact of low left ventricular function on ischemic stroke was not analyzed.

Conclusions

Ischemic stroke events are common among Japanese patients with NVAF and CIEDs. Risk stratification for ischemic stroke using the HELT-E₂S₂ score is valid in this population. This study demonstrated that the HELT-E₂S₂ score may be more effective than the CHADS₂ and CHA₂DS₂-VASc scores in identifying Japanese patients with NVAF at risk of ischemic stroke. When detecting AF in patients with CIEDs, physicians should consider a comprehensive assessment of the risk and benefit of prescribing an anticoagulant using the HELT-E₂S₂ score.

Acknowledgments

None.

Sources of Funding

This study did not receive any specific funding.

Disclosures

Y. Kondo has received lecture fees from Daiichi-Sankyo, Bayer, Abbott Medical Japan, Biotronik Japan, Boston Scientific, Japan Lifeline, and Medtronic, as well as research funds from Daiichi-Sankyo and Boston Scientific. Ma. Nakano has received lecture fees from Medtronic. Ma. Nakano, T.C., and S.N. belong to a department endowed by Abbott Medical Japan, Biotronik Japan, and Fukuda Denshi. T.K. and S.R. belong to a department endowed by Medtronic Japan. Y. Kobayashi has received lecture fees from Abbott Medical Japan, Bayer Japan, Bristol-Myers Squibb, Boehringer Ingelheim, and Daiichi-Sankyo, as well as scholarship funds from Takeda Pharmaceutical, Abbott Medical Japan, Terumo, Otsuka Pharmaceutical, Boehringer Ingelheim, Astellas, Daiichi-Sankyo, Win International, Japan Lifeline, and Nipro. Y. Kobayashi is a member of Circulation Journal’s Editorial Team.

IRB Information

This study was approved by the Ethics Committee at Chiba University, Graduate School of Medicine (Approval no. 2423).

Data Availability

The data generated in this study will not be shared.

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-24-0715

References

1. Lippi G, Sanchis-Gomar F, Cervellin G. Global epidemiology of atrial fibrillation: An increasing epidemic and public health challenge. Int J Stroke 2021; 16: 217–221.
2. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: The Framingham Study. Stroke 1991; 22: 983–988.
3. Stewart S, Hart CL, Hole DJ, McMurray JJ. A population-based study of the long-term risks associated with atrial fibrillation: 20-year follow-up of the Renfrew/Paisley study. Am J Med 2002; 113: 359–364.
4. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: Results from the National Registry of Atrial Fibrillation. JAMA 2001; 285: 2864–2870.
5. Inoue H, Nozawa T, Hirai T, Iwasa A, Okumura K, Lee JD, et al. Accumulation of risk factors increases risk of thromboembolic events in patients with nonvalvular atrial fibrillation. Circ J 2006; 70: 651–656.
6. Ono K, Iwasaki YK, Akao M, Ikeda T, Ishii K, Inden Y, et al. JCS/JHRS 2020 guideline on pharmacotherapy of cardiac arrhythmias. Circ J 2022; 86: 1790–1924.
7. Hara T, Sata M. Current real-world status of oral anticoagulant management in Japanese patients. Circ J 2025; 89: 150–152.
8. Okumura K, Tomita H, Nakai M, Kodani E, Akao M, Suzuki S, et al. A novel risk stratification system for ischemic stroke in Japanese patients with non-valvular atrial fibrillation. Circ J 2021; 85: 1254–1262.
9. Okumura K, Tomita H, Nakai M, Kodani E, Akao M, Suzuki S, et al. Risk factors associated with ischemic stroke in Japanese patients with nonvalvular atrial fibrillation. JAMA Netw Open 2020; 3: e202881.
10. JCS/JHRS 2024 guideline focused update on management of cardiac arrhythmias [in Japanese]. https://www.j-circ.or.jp/cms/wp-content/uploads/2024/03/JCS2024_Iwasaki.pdf (accessed December 17, 2024).
11. Yamauchi T, Okumura Y, Nagashima K, Watanabe R, Saito Y, Yokoyama K, et al. External validation of the HELT-E₂S₂ score in Japanese patients with nonvalvular atrial fibrillation: A pooled analysis of the RAFFINE and SAKURA registries. Circ J 2023; 87: 1777–1787.
12. Suzuki S. Choosing between HELT-E₂S₂ and CHA₂DS₂-VASc in Japan: Moving beyond CHADS₂ reliance? Circ J 2023; 87: 1788–1789.
13. Pastori D, Miyazawa K, Li Y, Shahid F, Hado H, Lip GYH. Inflammation and the risk of atrial high-rate episodes (AHREs) in patients with cardiac implantable electronic devices. Clin Res Cardiol 2018; 107: 772–777.
14. Müller P, Ivanov V, Kara K, Klein-Wiele O, Forkmann M, Piorkowski C, et al. Total atrial conduction time to predict occult atrial fibrillation after cryptogenic stroke. Clin Res Cardiol 2017; 106: 113–119.
15. JCS Joint Working Group. Guidelines for non-pharmacotherapy of cardiac arrhythmias (JCS 2011). Circ J 2013; 77: 249–274.
16. Nogami A, Kurita T, Kusano K, Goya M, Shoda M, Tada H, et al. JCS/JHRS 2021 guideline focused update on non-pharmacotherapy of cardiac arrhythmias. Circ J 2022; 86: 337–363.
17. Takase B, Ikeda T, Shimizu W, Abe H, Aiba T, Chinushi M, et al. JCS/JHRS 2022 guideline on diagnosis and risk assessment of arrhythmia. Circ J 2024; 88: 1509–1595.
18. Nakano M, Kondo Y, Nakano M, Kajiyama T, Hayashi T, Ito R, et al. Impact of atrial high-rate episodes on the risk of future stroke. J Cardiol 2019; 74: 144–149.
19. Sagris D, Georgiopoulos G, Pateras K, Perlepe K, Korompoki E, Milionis H, et al. Atrial high-rate episode duration thresholds and thromboembolic risk: A systematic review and meta-analysis. J Am Heart Assoc 2021; 10: e022487.
20. Lip GY, Nieuwlaat R, Pisters R, Lane DA, Crijns HJ. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor- based approach: The euro heart survey on atrial fibrillation. Chest 2010; 137: 263–272.
21. January CT, Wann LS, Calkins H, Chen LY, Cigarroa JE, Cleveland JC, et al. 2019 AHA/ACC/HRS focused update of the 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Rhythm Society. Heart Rhythm 2019; 16: e66–e93.
22. Hindricks G, Potpara T, Dagres N, Arbelo E, Bax JJ, Blomström-Lundqvist C, et al. Corrigendum to: 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 2021; 42: 4194.
23. Chao TF, Joung B, Takahashi Y, Lim TW, Choi EK, Chan YH, et al. 2021 Focused update of the 2017 consensus guidelines of the Asia Pacific Heart Rhythm Society (APHRS) on stroke prevention in atrial fibrillation. J Arrhythm 2021; 37: 1389–1426.
24. Suzuki S, Yamashita T, Okumura K, Atarashi H, Akao M, Ogawa H, 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 2015; 79: 432–438.
25. Atarashi H, Inoue H, Okumura K, Yamashita T, Origasa H. Investigation of optimal anticoagulation strategy for stroke prevention in Japanese patients with atrial fibrillation-the J-RHYTHM Registry study design. J Cardiol 2011; 57: 95–99.
26. Akao M, Chun YH, Wada H, Esato M, Hashimoto T, Abe M, et al. Current status of clinical background of patients with atrial fibrillation in a community-based survey: The Fushimi AF Registry. J Cardiol 2013; 61: 260–266.
27. Al-Khatib SM, Thomas L, Wallentin L, Lopes RD, Gersh B, Garcia D, et al. Outcomes of apixaban vs. warfarin by type and duration of atrial fibrillation: Results from the ARISTOTLE trial. Eur Heart J 2013; 34: 2464–2471.
28. Steinberg BA, Hellkamp AS, Lokhnygina Y, Patel MR, Breithardt G, Hankey GJ, et al. Higher risk of death and stroke in patients with persistent vs. paroxysmal atrial fibrillation: Results from the ROCKET-AF Trial. Eur Heart J 2015; 36: 288–296.
29. Link MS, Giugliano RP, Ruff CT, Scirica BM, Huikuri H, Oto A, et al. Stroke and mortality risk in patients with various patterns of atrial fibrillation: Results from the ENGAGE AF-TIMI 48 trial (Effective Anticoagulation With Factor Xa Next Generation in Atrial Fibrillation-Thrombolysis in Myocardial Infarction 48). Circ Arrhythm Electrophysiol 2017; 10: e004267.
30. Takabayashi K, Hamatani Y, Yamashita Y, Takagi D, Unoki T, Ishii M, et al. Incidence of stroke or systemic embolism in paroxysmal versus sustained atrial fibrillation: The Fushimi Atrial Fibrillation Registry. Stroke 2015; 46: 3354–3361.
31. Ogawa H, An Y, Ikeda S, Aono Y, Doi K, Ishii M, et al. Progression from paroxysmal to sustained atrial fibrillation is associated with increased adverse events. Stroke 2018; 49: 2301–2308.
32. Tan BYQ, Tan JTC, Cheah D, Zheng H, Pek PP, De Silva DA, et al. Long-term trends in ischemic stroke incidence and risk factors: Perspectives from an Asian stroke registry. J Stroke 2020; 22: 396–399.
33. Akao M, Ogawa H, Masunaga N, Minami K, Ishigami K, Ikeda S, et al. 10-Year trends of antithrombotic therapy status and outcomes in japanese atrial fibrillation patients: The Fushimi AF Registry. Circ J 2022; 86: 726–736.

Corresponding author

Register with J-STAGE for free!