Abstract

Aims

The CHA2DS2-VASc score is a modification of the CHADS2 score that aims to improve stroke risk prediction in patients with atrial fibrillation (AF) by adding three risk factors: age 65–74, female sex, and history of vascular disease. Whereas previous evaluations of the CHA2DS2-VASc score included all AF patients, the aim of this analysis was to evaluate its discriminative ability only in those patients for whom recommendations on antithrombotic treatment are uncertain (i.e. CHADS2 score of 1).

Methods and results

We selected all patients with a CHADS2 score of 1 from the AVERROES and ACTIVE trials who were treated with acetylsalicylic acid with or without clopidogrel and calculated the incidences of ischaemic or unspecified stroke or systemic embolus (SSE) according to their CHA2DS2-VASc score. Of 4670 patients with a baseline CHADS2 score of 1, 26% had a CHA2DS2-VASc score of 1 and 74% had a score of ≥2. After 11 414 patient-years of follow-up, the annual incidence of SSE was 0.9% (95% CI: 0.6–1.3) and 2.1% (95% CI: 1.8–2.5) for patients with a CHA2DS2-VASc score of 1 and ≥2, respectively. The c-statistic of the CHA2DS2-VASc score was 0.587 (95% CI: 0.550–0.624). Age 65 to <75 years was the strongest of the three new risk factors in the CHA2DS2-VASc score.

Conclusion

The CHA2DS2-VASc score reclassifies 26% of patients with a CHADS2 score of 1 to a low annual risk of SSE of 1%. This risk seems low enough to consider withholding anticoagulant treatment.

See page 168 for the editorial comment on this article (doi:10.1093/eurheartj/ehs378)

Introduction

Atrial fibrillation (AF) is a common cardiac arrhythmia that increases the risk of stroke five-fold.1 Dose-adjusted vitamin K antagonists (VKAs) and acetylsalicylic acid (ASA) reduce the risk of stroke by 64 and 19%, respectively.2 Although VKA therapy is more effective than ASA at preventing ischaemic stroke, its benefit is offset by an increased haemorrhage risk.2 Therefore, the key to deciding to initiate VKA therapy requires identifying those patients in whom the risk of ischaemic stroke without anticoagulants is sufficiently high to outweigh the increased risks of intracranial and major extracranial haemorrhage associated with VKA therapy. Several stroke risk stratification schemes have been proposed for patients with AF.3 Of these, the CHADS2 score is most widely used as it does not require costly additional tests, is easily applied and remembered by physicians, and has been validated to provide significant risk discrimination.3 The CHADS2 score assigns 1 point for heart failure, hypertension, age ≥75 years, and diabetes mellitus and 2 points for prior stroke or transient ischaemic attack. All guidelines recommend that patients with a CHADS2 score of 2 or higher should be treated with VKA therapy because the risk of ischaemic stroke outweighs the increased risk of bleeding induced by VKA therapy.4–6 However, guidelines are less firm in their recommendations concerning patients with a CHADS2 score of 1, reflecting uncertainty about the benefits of VKA therapy in this population.4–6

The CHA2DS2-VASc score has been proposed as an improvement to the CHADS2 score specifically for risk discrimination of lower risk patients.7 Compared with the CHADS2 score, the CHA2DS2-VASc score includes three additional risk factors for ischaemic stroke: age 65–74 years, female sex, and vascular disease, the latter defined as previous myocardial infarction or peripheral arterial disease.7 The CHA2DS2-VASc score has been validated in several cohorts of patients with AF and was shown to provide a modest, but significant discrimination of stroke risk.7–10 However, as most of the patients in those validation cohorts had a CHADS2 score of 2 or higher, treatment recommendations were unchanged irrespective of the CHA2DS2-VASc score. The important unresolved issue is whether the CHA2DS2-VASc score improves risk discrimination in patients in whom it is unclear if treatment with anticoagulants is beneficial (i.e. those with a CHADS2 score of 1). Furthermore, in some of the previous validation cohorts, patients were receiving anticoagulant therapy with a VKA.7,8 This is problematic because risk discrimination in patients already receiving a VKA is not relevant to the decision whether or not to treat with a VKA. So it is ideal to validate a risk score for stroke in patients treated with no therapy or, more practically, with antiplatelet therapy because it is deemed unethical to withhold all antithrombotic therapies from patients with AF who have an additional risk factor for stroke.

The aim of the present study was to determine the ability of the CHA2DS2-VASc score to discriminate stroke risk in AF patients with a CHADS2 score of 1 and thereby identify those patients for whom anticoagulant therapy may not be of benefit.

Methods

Patients

For the present analyses, we selected patients with AF and a CHADS2 score of 1, who were treated either with ASA only or with ASA and clopidogrel from three previously published trials: AVERROES, ACTIVE-W, and ACTIVE-A.11–13 The AVERROES trial (Apixaban vs. Acetylsalicylic Acid to Prevent Stroke in Atrial Fibrillation Patients Who Have Failed or Are Unsuitable for Vitamin K Antagonist Treatment) was designed to determine the efficacy and safety of apixaban compared with ASA for the treatment of patients with AF for whom VKA therapy was considered unsuitable.11 The ACTIVE trials (Atrial Fibrillation Clopidogrel Trial with Irbesartan for Prevention of Vascular Events) were initiated to evaluate the role of clopidogrel plus ASA for the prevention of stroke and other vascular events in patients with AF. The ACTIVE-W trial compared clopidogrel plus ASA with VKA therapy, and the ACTIVE-A trial compared clopidogrel plus ASA with ASA alone in patients for whom therapy with a VKA was considered unsuitable.12,13

The inclusion criteria for the AVERROES and ACTIVE trials were similar and are described in detail elsewhere.11–13 In short, patients were eligible if they had documented AF (in the 6 months before enrolment or at baseline) and one of the following risk factors for stroke: prior stroke or transient ischaemic attack, an age of 75 years or older, hypertension, diabetes mellitus, documented peripheral arterial disease, or a left ventricle ejection fraction (LVEF) of 35–45% or less. Patients with congestive heart failure (New York Heart Association class 2 or higher) and a LVEF > 45% were eligible for the AVERROES trials, but not for the ACTIVE trial. Patients with non-central nervous system (CNS) systemic emboli, or with coronary artery disease (CAD) and an age of 55–74 years as the only risk factor for stroke were eligible for the ACTIVE trials, but not for the AVERROES trial.

Key exclusion criteria for the three trials were an indication for VKAs other than AF (e.g. mechanical heart valves), or an indication for clopidogrel (ACTIVE trials only, e.g. recent coronary artery stents), valvular disease requiring surgery, or a high risk of bleeding (defined slightly different between the AVERROES and ACTIVE trials, but including active peptic ulcer disease or serious bleeding in the preceding 6 months, a history of intracranial haemorrhage, ongoing alcohol or drug abuse, thrombocytopenia of <50–100 × 109/L, or known documented haemorrhagic tendencies). The AVERROES trial also excluded patients with severe renal failure or liver transaminases levels greater than two times the upper limit of normal.

The daily dose of ASA in the ACTIVE and AVERROES trials ranged from 75 to 324 mg. Clopidogrel was given at a fixed dose of 75 mg once daily in the ACTIVE trials.

The AVERROES and ACTIVE trials complied with the Declaration of Helsinki. The study protocols were approved by institutional ethics boards and all participants provided written informed consent.

Outcomes and statistical analysis

The outcome of this analysis was the composite of ischaemic or non-specified stroke and non-CNS systemic embolus. Stroke was a clinical diagnosis that was made on the basis of typical symptoms lasting at least 24 h. Brain imaging, which was available in the vast majority of patients, was not required but was recommended for the general diagnosis of stroke. In the three trials, strokes were sub-classified into those that were ischaemic, primary haemorrhagic, or of uncertain type. In this analysis, primary haemorrhagic stroke was disregarded because the focus was on the prediction of ischaemic stroke. Non-CNS systemic embolus was defined as an acute loss of blood flow to a peripheral artery supported by evidence of embolism from surgical specimens, angiography, or other objective testing. All study outcomes were adjudicated by adjudication committees blinded to assigned treatments. Information on documented cardiovascular events prior to enrolment in the study was collected at the beginning of the study.

The CHA2DS2-VASc score was calculated with the relevant variables collected at baseline. No information on prior CAD or myocardial infarction was collected for patients in the AVERROES trial. Proportions of participants with CHA2DS2-VASc scores of 1, 2, 3, and 4 were calculated. Incidence rates were calculated by dividing the number of events by the number of patient-years of follow-up. The time to first event was time from enrolment until earliest occurrence of ischaemic or non-specified stroke or non-CNS systemic embolus. Patients were censored at either death, loss to follow-up, or end of study, whichever occurred first. Incidence rates were calculated for patients with CHA2DS2-VASc scores of 1, 2, 3, 4, and 2–4. Incidence rates were calculated for all patients (i.e. patients treated either with ASA only or with combined ASA and clopidogrel) and for patients treated with ASA only. Kaplan–Meier cumulative hazard rates over time were plotted for all patients and patients with the CHA2DS2-VASc scores of 1 and 2–4. Cox proportional hazard regression models were used to assess the relative increase in hazard associated with the CHA2DS2-VASc scores of 2, 3, or 4, and 2–4 vs. 1.

The ability of the CHA2DS2-VASc score to discriminate between high and low risk patients was assessed with the Harrell's c-statistic and net reclassification improvement (NRI) for time-to-event data as described by Pencina and colleagues.14–16 In calculation of the NRI, Kaplan–Meier estimates of 1-year risk were used. Patients with a CHA2DS2-VASc score of 1 were regarded as ‘reclassified down’, and patients with a CHA2DS2-VASc score of 2–4 were regarded as ‘reclassified up’. The 95% confidence interval (CI) limits for NRI were the 2.5th and 97.5th percentiles obtained with 1000 bootstrap samples.

In order to assess the relative contributions of the new CHA2DS2-VASc risk factors (i.e. age category 65 to <75 years, the extra point for age over 75 years, peripheral arterial disease or prior myocardial infarction, and female sex), univariate Cox regression models and a multivariate model to adjust the factors for each other were fitted. The analyses were done using Statistical Analysis Software, version 9.2 of the SAS System for SunOS (SAS Institute Inc., Cary, NC, USA).

Results

Of the 13 673 patients who were randomized to antiplatelet therapy in the three trials, 4670 had a CHADS2 score of 1 and were used in the present analysis. Of those, 48% (2240 patients) was randomized to receive ASA only, and 52% (2430 patients) was randomized to receive the combined ASA and clopidogrel treatment. Mean follow-up time was 2.5 years [standard deviation (SD) 1.4 years]. Patients from the two ACTIVE trials had a longer follow-up time than patients from the AVERROES trial [means of 2.9 years (SD 1.4 years) and 1.1 years (0.5 years), respectively]. Baseline characteristics are shown in Table 1. A total of 1924 patients (41%) were younger than 65 years, 2241 patients (48%) were aged 65 to <75 years, and 505 patients (11%) were 75 years or older. Most patients had permanent AF (52%), 19% had persistent AF, and 28% had paroxysmal AF. Apart from age, hypertension was the most common risk factor for stroke (present in 79% of patients, Table 1).

Table 1

Baseline characteristics of patients with a CHADS2 score of 1 treated with ASA only or with combined ASA and clopidogrel

Characteristic All (N = 4670) CHA2DS2-VASc score 0 or 1 (N = 1224) CHA2DS2-VASc score ≥2 (N = 3446) 
Mean age, years (SD) 65.5 (9.0) 56.5 (6.4) 68.9 (7.3) 
Female sex, n (%) 1589 (34) 0 (0) 1589 (46) 
Antiplatelet treatment, n (%) 
 ASA 2240 (48) 602 (49) 1638 (48) 
 ASA and clopidogrel 2430 (52) 622 (51) 1808 (53) 
Classification of atrial fibrillation, n (%)a 
 Permanent 2439 (52) 567 (46) 1872 (54) 
 Paroxysmal 1324 (28) 374 (31) 950 (28) 
 Persistent 899 (19) 280 (23) 619 (18) 
Presence of CHA2DS2-VASc variables, n (%) 
 Heart failure 282 (6) 98 (8) 184 (5) 
 Hypertension, receiving treatment 3710 (79) 1063 (87) 2647 (77) 
 Age <65 years 1924 (41) 1224 (100) 700 (20) 
 Age 65 to <74 years 2241 (48) 0 (0) 2241 (65) 
 Age ≥75 years 505 (11) 0 (0) 505 (15) 
 Diabetes mellitus, receiving treatment 119 (3) 45 (4) 74 (2) 
 Peripheral arterial disease 88 (2) 0 (0) 88 (3) 
 Myocardial infarctionb 337 (9) 0 (0) 337 (12) 
 Coronary artery diseaseb 646 (18) 43 (5) 603 (22) 
 Ischaemic stroke or transient ischaemic attack 0 (0) 0 (0) 0 (0) 
Characteristic All (N = 4670) CHA2DS2-VASc score 0 or 1 (N = 1224) CHA2DS2-VASc score ≥2 (N = 3446) 
Mean age, years (SD) 65.5 (9.0) 56.5 (6.4) 68.9 (7.3) 
Female sex, n (%) 1589 (34) 0 (0) 1589 (46) 
Antiplatelet treatment, n (%) 
 ASA 2240 (48) 602 (49) 1638 (48) 
 ASA and clopidogrel 2430 (52) 622 (51) 1808 (53) 
Classification of atrial fibrillation, n (%)a 
 Permanent 2439 (52) 567 (46) 1872 (54) 
 Paroxysmal 1324 (28) 374 (31) 950 (28) 
 Persistent 899 (19) 280 (23) 619 (18) 
Presence of CHA2DS2-VASc variables, n (%) 
 Heart failure 282 (6) 98 (8) 184 (5) 
 Hypertension, receiving treatment 3710 (79) 1063 (87) 2647 (77) 
 Age <65 years 1924 (41) 1224 (100) 700 (20) 
 Age 65 to <74 years 2241 (48) 0 (0) 2241 (65) 
 Age ≥75 years 505 (11) 0 (0) 505 (15) 
 Diabetes mellitus, receiving treatment 119 (3) 45 (4) 74 (2) 
 Peripheral arterial disease 88 (2) 0 (0) 88 (3) 
 Myocardial infarctionb 337 (9) 0 (0) 337 (12) 
 Coronary artery diseaseb 646 (18) 43 (5) 603 (22) 
 Ischaemic stroke or transient ischaemic attack 0 (0) 0 (0) 0 (0) 

ASA, acetyl salicylic acid; SD, standard deviation.

aAvailable for 4662 patients.

bAvailable only for 3658 patients from ACTIVE A and ACTIVE W.12,13

Of the 4670 patients with a CHADS2 score of 1, the CHA2DS2-VASc score was 1 in 26% of patients and 42%, 26%, and 3% of patients had CHA2DS2-VASc scores of 2, 3, and 4, respectively (Table 2).

Table 2

Distribution of the CHA2DS2-VASc score in patients with a CHADS2 score of 1

CHA2DS2-VASc score N (%) 
All 4670 (100) 
1224 (26) 
1984 (42) 
1338 (29) 
124 (3) 
CHA2DS2-VASc score N (%) 
All 4670 (100) 
1224 (26) 
1984 (42) 
1338 (29) 
124 (3) 

Table 3 shows the incidence rates of the composite outcome of ischaemic or non-specified stroke or non-CNS systemic embolus. Of the 4670 patients with 11 414 patient-years, 205 patients had experienced an outcome event, amounting to an incidence of 1.8 per 100 patient-years (95% CI: 1.6–2.1). Patients with a CHA2DS2-VASc score of 1 had an incidence rate of 0.9 per 100 patient-years (95% CI: 0.6–1.3) and patients with a CHA2DS2-VASc score of ≥2 had a more than two-fold increased rate of 2.1 per 100 patient-years (95% CI: 1.8–2.5; hazard ratio: 2.45, 95% CI: 1.66–3.75). The incidence rates in patients with CHA2DS2-VASc scores of 2 and 3 or 4 were 2.0 per 100 patient-years (95% CI: 1.6–2.4) and 2.4 per 100 patient-years (95% CI: 1.9–2.9), respectively (Table 3). The incidence for patients treated with ASA only was higher than for patients treated with ASA and clopidogrel combined (2.3 vs. 1.3 per 100 patient-years, Table 4). The relative risk increase with increasing CHA2DS2-VASc score was similar regardless of the type of antiplatelet treatment.

Table 3

Incidence rates of ischaemic or unspecified stroke or systemic non-CNS embolus in patients with a CHADS2 score of 1, treated with ASA only or combined ASA and clopidogrel

 Number of eventsa/patients Patient-years of follow-up Incidence rate, per 100 patient-years Hazard ratio (95%CI)b Harrell's c-statistic (95%CI)c NRI (95% CI)d 
All 205/4670 11414 1.8 (1.6–2.1)    
CHA2DS2-VASc 
 1 27/1224 3074 0.9 (0.6–1.3) 0.587 (0.550–0.624)  
 2 92/1984 4729 2.0 (1.6–2.4) 2.2 (1.5–3.5)   
 3–4 86/1462 3610 2.4 (1.9–2.9) 2.7 (1.8–4.3)   
 1 27/1224 3074 0.9 (0.6–1.3) 0.567 (0.541–0.592) 0.27 (0.11–0.41) 
 2–4 178/3446 8340 2.1 (1.8–2.5) 2.5 (1.7–3.8)   
 Number of eventsa/patients Patient-years of follow-up Incidence rate, per 100 patient-years Hazard ratio (95%CI)b Harrell's c-statistic (95%CI)c NRI (95% CI)d 
All 205/4670 11414 1.8 (1.6–2.1)    
CHA2DS2-VASc 
 1 27/1224 3074 0.9 (0.6–1.3) 0.587 (0.550–0.624)  
 2 92/1984 4729 2.0 (1.6–2.4) 2.2 (1.5–3.5)   
 3–4 86/1462 3610 2.4 (1.9–2.9) 2.7 (1.8–4.3)   
 1 27/1224 3074 0.9 (0.6–1.3) 0.567 (0.541–0.592) 0.27 (0.11–0.41) 
 2–4 178/3446 8340 2.1 (1.8–2.5) 2.5 (1.7–3.8)   

ASA, acetylsalicylic acid; CI, confidence interval; NRI, net reclassification improvement.

aAn event is the first occurrence of ischaemic or unspecified stroke or non-CNS systemic embolus. Time to event is the time between randomization and event first occurrence.

bCox proportional hazards regression model. 95% CI limits for hazard ratio are profile likelihood limits.

cHarrell's c-statistic and its 95% CI were estimated using the SAS macro %survcstd.15

dNRI for time-to-event data using Kaplan–Meier estimated of 1-year risk.16 Patients with a CHA2DS2-VASc score of 1 were regarded as ‘reclassified down’ and patients with a score of ≥2 were regarded as ‘reclassified down’. The 95% CI limits for NRI were the 2.5th and 97.5th percentiles obtained with 1000 bootstrap samples.

Table 4

Incidence rates of ischaemic or unspecified stroke or systemic non-CNS embolus according to type of antiplatelet treatment

 Number of eventsa/patients Patient-years of follow-up Incidence rate, per 100 patient-years Hazard ratio (95%CI)b 
Combined ASA and clopidogrel 
All 79/2430 5950 1.3 (1.1–1.7)  
CHA2DS2-VASc 
 1 11/622 1609 0.7 (0.3–1.2) 
 2 38/1020 2421 1.6 (1.1–2.2) 2.4 (1.2–4.8) 
 3–4 30/788 1921 1.6 (1.1–2.2) 2.3 (1.2–4.8) 
 1 11/622 1609 0.7 (0.3–1.2) 
 2–4 68/1808 4341 1.6 (1.2–2.0) 2.3 (1.3–4.6) 
ASA only 
 All 126/2240 5463 2.3 (1.9–2.8)  
CHA2DS2-VASc 
 1 16/602 1465 1.1 (0.6–1.8) 
 2 54/964 2309 2.3 (1.8–3.1) 2.2 (1.3–3.9) 
 3–4 56/674 1689 3.3 (2.5–4.3) 3.0 (1.8–5.5) 
 1 16/602 1465 1.1 (0.6–1.8) 
 2–4 110/1638 3998 2.8 (2.3–3.3) 2.5 (1.5–4.4) 
 Number of eventsa/patients Patient-years of follow-up Incidence rate, per 100 patient-years Hazard ratio (95%CI)b 
Combined ASA and clopidogrel 
All 79/2430 5950 1.3 (1.1–1.7)  
CHA2DS2-VASc 
 1 11/622 1609 0.7 (0.3–1.2) 
 2 38/1020 2421 1.6 (1.1–2.2) 2.4 (1.2–4.8) 
 3–4 30/788 1921 1.6 (1.1–2.2) 2.3 (1.2–4.8) 
 1 11/622 1609 0.7 (0.3–1.2) 
 2–4 68/1808 4341 1.6 (1.2–2.0) 2.3 (1.3–4.6) 
ASA only 
 All 126/2240 5463 2.3 (1.9–2.8)  
CHA2DS2-VASc 
 1 16/602 1465 1.1 (0.6–1.8) 
 2 54/964 2309 2.3 (1.8–3.1) 2.2 (1.3–3.9) 
 3–4 56/674 1689 3.3 (2.5–4.3) 3.0 (1.8–5.5) 
 1 16/602 1465 1.1 (0.6–1.8) 
 2–4 110/1638 3998 2.8 (2.3–3.3) 2.5 (1.5–4.4) 

ASA, acetylsalicylic acid; CI, confidence interval.

aAn event is the first occurrence of ischaemic or unspecified stroke or non-CNS systemic embolus. Time to event is the time between randomization and event first occurrence.

bCox proportional hazards regression model. 95% CI limits for hazard ratio are profile likelihood limits.

In this group of patients with a CHADS2 score of 1 treated with ASA only or combined ASA and clopidogrel, the Harrell's c-statistic was 0.587 (95% CI: 0.550–0.624; Table 3). The NRI for 1-year risk prediction calculated assuming the patients with the CHA2DS2-VASc score of 1 were reclassified down and patients with a score of 2 or higher were reclassified up was 0.74 (95% CI: 0.58–0.88) for events and −0.47 (95% CI: −0.50 to −0.45) for non-events, leading to an overall NRI of 0.27 (95% CI: 0.11–0.41; Table 3). When regarded as a dichotomized risk score (CHA2DS2-VASc score of 1 is test-negative and 2–4 is test-positive), the positive likelihood ratio of the CHA2DS2-VASc score was 1.18 (95% CI: 1.08–1.28) and the negative likelihood ratio was 0.49 (95% CI: 0.23–0.79).

Figure 1 shows the Kaplan–Meier cumulative hazard curves for all patients, and for those with CHA2DS2-VASc scores of 1 and 2 or higher. The rates of stroke or non-CNS systemic embolus are fairly constant over the first 5 years (Figure 1).

Figure 1

Kaplan–Meier cumulative hazard rates of the composite outcome of ischaemic or unspecified stroke and non-CNS systemic embolus in patients treated with acetylsalicylic acid (ASA) only or with combined ASA and clopidogrel. Black line represents all patients with a CHADS2 score of 1; blue line represents patients with a CHA2DS2-VASc score of 1; green line represents patients with a CHA2DS2-VASc score of 2–4.

Figure 1

Kaplan–Meier cumulative hazard rates of the composite outcome of ischaemic or unspecified stroke and non-CNS systemic embolus in patients treated with acetylsalicylic acid (ASA) only or with combined ASA and clopidogrel. Black line represents all patients with a CHADS2 score of 1; blue line represents patients with a CHA2DS2-VASc score of 1; green line represents patients with a CHA2DS2-VASc score of 2–4.

Table 5 shows the hazard ratios for the three new risk factors that are introduced by the CHA2DS2-VASc score. Age 65 to <74 years and age ≥75 years were associated with a two-fold increased risk of the composite outcome. Female sex was a weaker risk factor (adjusted hazard ratio: 1.32) and a personal history of peripheral arterial disease or myocardial infarction was not a risk factor in this cohort (hazard ratio: 0.97).

Table 5

Relative risk of the composite outcome of ischaemic or unspecified stroke and non-CNS systemic embolus associated with the additional risk factors introduced by the CHA2DS2-VASc score

Characteristic Hazard ratio (95%CI)a Adjusted hazard ratio (95%CI)b 
Age 
 <65 years 
 65 to <74 years 1.97 (1.44–2.74) 1.90 (1.38–2.64) 
 ≥75 years 2.31 (1.47–3.58) 2.24 (1.42–3.48) 
Previous peripheral arterial disease or myocardial infarction 
 No 
 Yes 0.99 (0.66–1.42) 0.97 (0.65–1.41) 
Sex 
 Male 
 Female 1.45 (1.10–1.91) 1.32 (1.00–1.75) 
Characteristic Hazard ratio (95%CI)a Adjusted hazard ratio (95%CI)b 
Age 
 <65 years 
 65 to <74 years 1.97 (1.44–2.74) 1.90 (1.38–2.64) 
 ≥75 years 2.31 (1.47–3.58) 2.24 (1.42–3.48) 
Previous peripheral arterial disease or myocardial infarction 
 No 
 Yes 0.99 (0.66–1.42) 0.97 (0.65–1.41) 
Sex 
 Male 
 Female 1.45 (1.10–1.91) 1.32 (1.00–1.75) 

CI, confidence interval.

aUnivariate Cox proportional hazard models fitted to all patients with a CHADS2 score of 1 treated with acetylsalicylic acid (ASA) only or with combined ASA and clopidogrel. 95% CI limits for hazard ratio are profile likelihood limits.

bAge, previous peripheral arterial disease or myocardial infarction, and sex were included in the multivariate Cox model.

Discussion

This is the first study to show that in patients with AF in whom there is a real clinical question about the decision to anticoagulate, the CHA2DS2-VASc score adds potentially valuable information. Previous evaluations of the CHA2DS2-VASc score have studied AF patients across the whole range of stroke risk and have thereby included many patients in whom there is no real potential for improved risk stratification because they will anyway benefit from anticoagulation (i.e. CHADS2 score ≥2).7–10 Unlike most previous analyses, all patients in this cohort had a CHADS2 score of 1 and were treated only with antiplatelet therapy. The results show that the CHA2DS2-VASc score separates a very low risk group (i.e. 1% per year stroke risk) from the other patients. The very low risk patients with a CHA2DS2-VASc score of 1 were men aged <65 years with either hypertension (87%), heart failure (8%), or diabetes mellitus (4%). Given the risks of bleeding associated with VKA treatment and the 1% per year stroke risk in patients with a CHA2DS2-VASc score of 1, we think it is unlikely that these patients will benefit from treatment with VKA therapy. This is partially in contrast with the 2010 guidelines of the European Society of Cardiology that recommends either ASA or oral anticoagulant therapy for patients with a CHA2DS2-VASc score of 1.4

Within this group of patients with a CHADS2 score of 1, the CHA2DS2-VASc score significantly improves risk stratification. However, the improvement by the CHA2DS2-VASc score is modest, as expressed by the c-statistic of 0.587 (95% CI: 0.550–0.624). This is likely due to the selection of patients with a CHADS2 score of 1 (all at moderate risk of stroke) and probably also due to the fact that the CHA2DS2-VASc shares four risk factors with the CHADS2 score. Despite the modest degree of the risk discrimination improvement, the clinical consequence of this reclassification is substantial. Whereas most guidelines suggest that oral anticoagulant therapy preferred in patients with a CHADS2 score of 1, reclassification using the CHA2DS2-VASc score means that 26% of patients (1224 of 4670 patients, Table 2) are reclassified as low risk and can be treated with ASA rather than VKA. Conversely, if patients with a CHADS2 score of 1 would normally be treated with ASA only, the CHA2DS2-VASc score reclassifies 74% into a higher risk category that may benefit from VKA therapy.

Ideally, a risk score aimed to guide treatment of patients with AF should take into account both the risk of stroke without anticoagulants and the risk of bleeding with anticoagulants. The aim of our analyses was, however, to explore the discriminative ability of the CHA2DS2-VASc score for stroke risk prediction in patients with a CHADS2 score of 1 and we have not examined bleeding risks in an anticoagulated population.

The major strength of the present study is that the analysis is restricted to the group of patients for whom current guidelines provide conflicting recommendations.5,6 Despite this restriction, this study included a large number of patients (n = 4670) enrolled in large randomized trials with rigorous patient follow-up and all events were carefully adjudicated.

Several points of this study merit discussion. First, information on prior myocardial infarction or CAD, one of the components of the CHA2DS2-VASc score, was not collected in the 1012 patients from the AVERROES trial (22% of patients). However, when the results of the AVERROES and ACTIVE trials were analysed separately, the results were consistent (results not shown). Second, half of the patients in this cohort were treated with clopidogrel in addition to ASA. The incidence of the primary outcome was higher in patients treated with ASA only compared with patients treated with combined ASA and clopidogrel (2.3 and 1.3% per year, respectively). However, even in the group treated with ASA only, the risk of patients with a CHA2DS2-VASc score of 1 was still sufficiently low (1.1% per year, Table 4) to argue that they may not benefit from VKA therapy.17 This threshold incidence to prefer anticoagulants over ASA may be different with the new oral anticoagulants. Dabigatran, apixaban, and rivaroxaban had a 30–70% lower risk of intracranial haemorrhage compared with VKA therapy and apixaban did not increase major or intracranial bleeding compared with aspirin.11,18–20 These data suggest that the threshold for the use of the new anticoagulants, although likely subject to heavy debate, may be as low as 1% per year.17,19

The stroke risk of 0.9% per year for patients with a CHA2DS2-VASc score of 1 without VKA therapy in this study is in line with the risk of 0.6–1.5% per year found in other evaluations of the CHA2DS2-VASc score.7,9,10 The risk found in these studies, however, was derived from registries with less accurate outcome adjudication and varying proportions of patients treated with antiplatelet drugs.

The added points for age (1 additional point for age ≥75 years and 1 point for age 65–74 years) were the most important of the three new factors in the CHA2DS2-VASc score. Female sex was a weaker risk factor and previous vascular disease was not a significant predictor of stroke in our cohort. Of note, only 2% of patients in this cohort had documented peripheral arterial disease, which is substantially lower than a cohort that did show an effect of previous vascular disease on stroke risk.21 This can be regarded as an argument to generate a new stroke risk scheme without vascular disease, but this would require new validations and given the acceptance of the CHA2DS2-VASc score, it would be less helpful to propose yet another clinical score.4

In conclusion, the CHA2DS2-VASc score reclassifies 26% of patients with a CHADS2 score of 1 to a low risk of stroke of around 1% per year. This risk may be considered sufficiently low to refrain from oral anticoagulant therapy.

Funding

The ACTIVE and AVERROES trials were supported by Sanofi-Aventis; Bristol-Myers Squibb; and Pfizer.

Conflict of interest: none declared.

References

1
Wolf
PA
Abbott
RD
Kannel
WB
Atrial fibrillation as an independent risk factor for stroke: the Framingham Study
Stroke
 , 
1991
, vol. 
22
 (pg. 
983
-
988
)
2
Hart
RG
Pearce
LA
Aguilar
MI
Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation
Ann Intern Med
 , 
2007
, vol. 
146
 (pg. 
857
-
867
)
3
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
, vol. 
285
 (pg. 
2864
-
2870
)
4
Camm
AJ
Kirchhof
P
Lip
GY
Schotten
U
Savelieva
I
Ernst
S
Van Gelder
IC
Al-Attar
N
Hindricks
G
Prendergast
B
Heidbuchel
H
Alfieri
O
Angelini
A
Atar
D
Colonna
P
De
CR
De
SJ
Goette
A
Gorenek
B
Heldal
M
Hohloser
SH
Kolh
P
Le Heuzey
JY
Ponikowski
P
Rutten
FH
Guidelines for the management of atrial fibrillation: the Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC)
Eur Heart J
 , 
2010
, vol. 
31
 (pg. 
2369
-
2429
)
5
Fuster
V
Ryden
LE
Cannom
DS
Crijns
HJ
Curtis
AB
Ellenbogen
KA
Halperin
JL
Le Heuzey
JY
Kay
GN
Lowe
JE
Olsson
SB
Prystowsky
EN
Tamargo
JL
Wann
S
Smith
SC
Jr.
Jacobs
AK
Adams
CD
Anderson
JL
Antman
EM
Halperin
JL
Hunt
SA
Nishimura
R
Ornato
JP
Page
RL
Riegel
B
Priori
SG
Blanc
JJ
Budaj
A
Camm
AJ
Dean
V
Deckers
JW
Despres
C
Dickstein
K
Lekakis
J
McGregor
K
Metra
M
Morais
J
Osterspey
A
Tamargo
JL
Zamorano
JL
ACC/AHA/ESC 2006 Guidelines for the Management of Patients with Atrial Fibrillation: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the European Society of Cardiology Committee for Practice Guidelines (Writing Committee to Revise the 2001 Guidelines for the Management of Patients With Atrial Fibrillation): developed in collaboration with the European Heart Rhythm Association and the Heart Rhythm Society
Circulation
 , 
2006
, vol. 
114
 (pg. 
e257
-
e354
)
6
You
JJ
Singer
DE
Howard
PA
Lane
DA
Eckman
MH
Fang
MC
Hylek
EM
Schulman
S
Go
AS
Hughes
M
Spencer
FA
Manning
WJ
Halperin
JL
Lip
GY
Antithrombotic therapy for atrial fibrillation: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines
Chest
 , 
2012
, vol. 
141
 (pg. 
e531S
-
e575S
)
7
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
, vol. 
137
 (pg. 
263
-
272
)
8
Lip
GY
Frison
L
Halperin
JL
Lane
DA
Identifying patients at high risk for stroke despite anticoagulation: a comparison of contemporary stroke risk stratification schemes in an anticoagulated atrial fibrillation cohort
Stroke
 , 
2010
, vol. 
41
 (pg. 
2731
-
2738
)
9
Olesen
JB
Lip
GY
Hansen
ML
Hansen
PR
Tolstrup
JS
Lindhardsen
J
Selmer
C
Ahlehoff
O
Olsen
AM
Gislason
GH
Torp-Pedersen
C
Validation of risk stratification schemes for predicting stroke and thromboembolism in patients with atrial fibrillation: nationwide cohort study
BMJ
 , 
2011
, vol. 
342
 pg. 
d124
 
10
Van Staa
TP
Setakis
E
Di Tanna
GL
Lane
DA
Lip
GY
A comparison of risk stratification schemes for stroke in 79,884 atrial fibrillation patients in general practice
J Thromb Haemost
 , 
2011
, vol. 
9
 (pg. 
39
-
48
)
11
Connolly
SJ
Eikelboom
JW
Joyner
C
Diener
HC
Hart
R
Golitsyn
S
Flaker
G
Avezum
A
Hohnloser
SH
Diaz
R
Talajic
M
Zhu
J
Pais
P
Budaj
A
Parkhomenko
A
Jansky
P
Commerford
P
Tan
RS
Sim
KH
Lewis
BS
Van
MW
Lip
GY
Kim
JH
Lanas-Zanetti
F
Gonzalez-Hermosillo
A
Dans
AL
Munawar
M
O'Donnell
M
Lawrence
J
Lewis
G
Afzal
R
Yusuf
S
Apixaban in patients with atrial fibrillation
N Engl J Med
 , 
2011
, vol. 
364
 (pg. 
806
-
817
)
12
Connolly
SJ
Pogue
J
Hart
R
Pfeffer
M
Hohnloser
S
Chrolavicius
S
Pfeffer
M
Hohnloser
S
Yusuf
S
Clopidogrel plus aspirin versus oral anticoagulation for atrial fibrillation in the Atrial fibrillation Clopidogrel Trial with Irbesartan for prevention of Vascular Events (ACTIVE W): a randomised controlled trial
Lancet
 , 
2006
, vol. 
367
 (pg. 
1903
-
1912
)
13
Connolly
SJ
Pogue
J
Hart
RG
Hohnloser
SH
Pfeffer
M
Chrolavicius
S
Yusuf
S
Effect of clopidogrel added to aspirin in patients with atrial fibrillation
N Engl J Med
 , 
2009
, vol. 
360
 (pg. 
2066
-
2078
)
14
Harrell
FE
Jr.
Lee
KL
Mark
DB
Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors
Stat Med
 , 
1996
, vol. 
15
 (pg. 
361
-
387
)
15
Kremers
WK
SAS macro %survcstd
Mayo Clinic SAS macros
 , 
2008
 
16
Pencina
MJ
D'Agostino
RB
Jr
Steyerberg
EW
Extensions of net reclassification improvement calculations to measure usefulness of new biomarkers
Stat Med
 , 
2011
, vol. 
30
 (pg. 
11
-
21
)
17
Eckman
MH
Singer
DE
Rosand
J
Greenberg
SM
Moving the tipping point: the decision to anticoagulate patients with atrial fibrillation
Circ Cardiovasc Qual Outcomes
 , 
2011
, vol. 
4
 (pg. 
14
-
21
)
18
Granger
CB
Alexander
JH
McMurray
JJ
Lopes
RD
Hylek
EM
Hanna
M
Al-Khalidi
HR
Ansell
J
Atar
D
Avezum
A
Bahit
MC
Diaz
R
Easton
JD
Ezekowitz
JA
Flaker
G
Garcia
D
Geraldes
M
Gersh
BJ
Golitsyn
S
Goto
S
Hermosillo
AG
Hohnloser
SH
Horowitz
J
Mohan
P
Jansky
P
Lewis
BS
Lopez-Sendon
JL
Pais
P
Parkhomenko
A
Verheugt
FW
Zhu
J
Wallentin
L
Apixaban versus warfarin in patients with atrial fibrillation
N Engl J Med
 , 
2011
, vol. 
365
 (pg. 
981
-
992
)
19
Oldgren
J
Alings
M
Darius
H
Diener
HC
Eikelboom
J
Ezekowitz
MD
Kamensky
G
Reilly
PA
Yang
S
Yusuf
S
Wallentin
L
Connolly
SJ
Risks for stroke, bleeding, and death in patients with atrial fibrillation receiving dabigatran or warfarin in relation to the CHADS2 score: a subgroup analysis of the RE-LY trial
Ann Intern Med
 , 
2011
, vol. 
155
 (pg. 
660
-
667
W204
20
Patel
MR
Mahaffey
KW
Garg
J
Pan
G
Singer
DE
Hacke
W
Breithardt
G
Halperin
JL
Hankey
GJ
Piccini
JP
Becker
RC
Nessel
CC
Paolini
JF
Berkowitz
SD
Fox
KA
Califf
RM
Rivaroxaban versus warfarin in nonvalvular atrial fibrillation
N Engl J Med
 , 
2011
, vol. 
365
 (pg. 
883
-
891
)
21
Olesen
JB
Fauchier
L
Lane
DA
Taillandier
S
Lip
GY
Risk factors for stroke and thromboembolism in relation to age among patients with atrial fibrillation: the Loire Valley Atrial Fibrillation Project
Chest
 , 
2012
, vol. 
141
 (pg. 
147
-
153
)

Comments

2 Comments
Prognostic value of atrial fibrillation in patients with recurrent ischemic stroke
4 January 2013
Gulcin Benbir (with Birsen Ince)

Prognostic value of atrial fibrillation in patients with recurrent ischemic stroke

A recent paper by Coppens et al published in European Heart Journal (1) showed that CHA2 DS2-VASc scale was a predictive tool to improve stroke risk prediction in patients with atrial fibrillation (AF). It is known that AF is a frequent cause of cerebral infarction with high risk of first-ever and recurrent cerebral infarction (2). A better knowledge of the influence of AF on stroke prognosis independent of other factors is of crucial importance because the effective use of anticoagulant therapy is achieved only in half of the patients (3). On this context, we analyzed all patients registered in our Stroke Clinic between years 1996 and 2011, with a mean follow-up of 50.2+42.7 months. Of 2798 patients, 2529 patients (90.3%) had ischemic stroke, and 309 patients (12.2%) had AF. Patients with AF were older in compared to patients without AF (69.8+10.2 vs 61.0+13.2 years, p<0.001); and most of them were females (p=0.001). Hypertension was more common in patients with AF (p<0.001); while other risk factors were not different between two groups. The percentage of recurrent stroke was higher in patients with AF in compared to those without AF, but not significant (p=0.461). The mean CHADS2 scores were also not different between two groups (3.4+0.7 vs 3.3+0.7, p=0.530). On the other hand, CHA2-DS2-VASc scores were significantly higher in patients with recurrent stroke (4.9+1.1 vs 4.4+1.2, p<0.001). Many risk stratification schemas have been developed to categorize the risk of stroke, and to aid decisions regarding the most appropriate therapy (4). While CHADS2 scale is more commonly used in cardiology clinics, CHA2DS2-VASc analysis seems to a potential tool in predicting the risk of recurrent stroke in patients with AF, and needs to be utilized more commonly by neurologists.

References

1. Coppens M, Eikelboom JW, Hart RG, Yusuf S, Lip GY, Dorian P, Shestakovska O, Connolly SJ. The CHA2DS2-VASc score identifies those patients with atrial fibrillation and a CHADS2 score of 1 who are unlikely to benefit from oral anticoagulant therapy. Eur Heart J 2012; [Epub ahead of print].

2. Hohnloser SH, Pajitnev D, Pogue J, Healey JS, Pfeffer MA, Yusuf S, Connolly SJ; ACTIVE W Investigators. Incidence of stroke in paroxysmal versus sustained atrial fibrillation in patients taking oral anticoagulation or combined antiplatelet therapy. An ACTIVE W Substudy. J Am Coll Cardiol 2007;22:2156-2161.

3. Ince B, Benbir G, Bolukbasi F. Underuse of oral anticoagulants in ischemic stroke patients with atrial fibrillation. J Thromb Haemost 2011;9:633(abstract).

4. Li SY, Zhao XQ, Wang CX, Liu LP, Liu GF, Wang YL, Wang YJ. One-year Clinical Prediction in Chinese Ischemic Stroke Patients Using the CHADS2 and CHA2DS2-VASc Scores: The China National Stroke Registry. CNS Neurosci Ther 2012;18:988-993.

Conflict of Interest:

None declared

Birsen Ince, Prof. Dr., Gulcin Benbir, Asist. Prof. Dr.

Istanbul University Cerrahpasa Faculty of Medicine, Department of Neurology, Stroke Clinic, Istanbul, Turkey

Corresponding author: Gulcin Benbir Address: Istanbul University Cerrahpasa Faculty of Medicine, Department of Neurology, Istanbul, 34098, Turkey Phone: +905332263797 Fax: +902126329696 E-mail: drgulcinbenbir@yahoo.com

Submitted on 04/01/2013 7:00 PM GMT
does aspirin significantly reduce the risk of systemic embolism in NVAF subjects with low CHA2DS2-Vasc score?
13 February 2013
oscar,m jolobe

Now that it has been shown that the CHA2DS2-Vasc score is superior to the CHADS2 for identifying which of the patients with nonvalvular atrial fibrillation(NVAF) are unlikely to benefit from a vitamin K antagonist(VKA)(1),what remains is to evaluate whether or not aspirin is superior to placebo in further reducing the already low risk of systemic embolism in such subjects. This is an issue which needs to be adressed with some urgency, given the fact that there is already a point of view(almost amounting to an axiom)that, in a patient with "effectively zero active CHA2DS2VASC risk factors", clinicians should "feel comfortable using aspirin for stroke prevention"(2). Trials which adress the issue of aspirin use in low-risk NVAF subjects should enrol NVAF subjects with CHAD2DS2-Vasc score of 1, and those subjects should be randomised to aspirin vs placebo. What might be contentious is the choice of aspirin dose. Accordingly trial subjects might have to be stratified according to aspirin dose. Finally, in order to eliminate the confounding effect of coexisting carotid artery disease, high-risk carotid artery disease(evaluated by imaging techniques) should be an exclusion factor for such a trial.

References (1) Coppens M., Eikelboom JW., Hart RG et al The CHA2DS2-Vasc score identifies those patients with atrial fibrillation and a CHADS2 score of 1 who are unlikely to benefit from oral anticoagulany therapy Eur Heart J 2013;34:170-176 (2) Chinitz JS., Vaishnava P., Narayan RL., Fuster V Atrial fibrialltion through the years. Contemporary evaluation and management Circulation 2013;127:408-416

Conflict of Interest:

None declared

Submitted on 13/02/2013 7:00 PM GMT