Increases in Cerebral Atherosclerosis According to CHADS2 Scores in Patients With Stroke With Nonvalvular Atrial Fibrillation
Background and Purpose—The CHADS2 score is used for risk stratification of ischemic stroke in patients with nonvalvular atrial fibrillation and high CHADS2 scores are associated with increased risk of stroke. Most components of the CHADS2 score are also risk factors for atherosclerosis. Therefore, high CHADS2 scores can be associated with concomitant cerebral atherosclerosis and subsequently atherothrombotic stroke. The aim of this study was to determine whether there are differences in the presence and burden of concomitant cerebral atherosclerosis according to CHADS2 scores in patients with stroke with nonvalvular atrial fibrillation.
Methods—We included 780 consecutive patients with nonvalvular atrial fibrillation who had undergone angiographic studies at index stroke between August 1994 and March 2010 in the present study. We investigated the relationships between the CHADS2 score and the presence, severity, and pattern of cerebral atherosclerosis and stroke mechanism.
Results—Of the 780 patients, concomitant arterial stenosis (≥50%) was found in 231 patients (29.6%). The number of arteries with atherosclerosis increased as the CHADS2 score increased (P<0.001) as did the proportion of combined extracranial and intracranial atherosclerosis (P<0.001). Multivariate analyses showed that high risk based on the CHADS2 score was an independent predictor of concomitant cerebral atherosclerosis (OR, 3.121; 95% CI, 1.770 to 5.504) and the presence of proximal stenosis at the symptomatic artery (OR, 3.043; 95% CI, 1.458 to 6.350).
Conclusions—The CHADS2 score can predict the presence of concomitant cerebral artery atherosclerosis. Increased risk of stroke in patients with high CHADS2 scores may be partly explained by increased frequency and burden of cerebral atherosclerosis.
Nonvalvular atrial fibrillation (NVAF) is a well-known risk factor of stroke, and maintaining an optimum intensity of anticoagulation is highly effective for stroke prevention.1 However, serious bleeding complications, including intracerebral hemorrhage, are related to warfarin use,2,3 which limits its routine use in patients with NVAF. Risk stratification schemes have been developed to identify patients with high risks for ischemic stroke or systemic embolic events and also to identify patients eligible for anticoagulation.4,5
Currently, the CHADS2 index is widely used to stratify patients with NVAF who are eligible for anticoagulation. Increases in CHADS2 scores are associated with worsening intracardiac thrombogenic conditions and increased risk of embolic events.6,7 However, it is also possible that increased risk of stroke with high CHADS2 scores is caused by atherothrombotic mechanisms as well as cardioembolic mechanisms, because most components of the CHADS2 score are also risk factors for atherosclerosis. Although previous studies indicated that approximately 20% to 50% of patients with NVAF have severe stenotic lesions at the carotid artery,8–10 few data are available regarding the relationship between the presence of cerebral atherosclerosis in the entire cerebral arterial beds and the CHADS2 score.
The aim of this study was to investigate whether there are differences in atherosclerotic burden in cerebral arterial beds according to risk levels based on CHADS2 scores and to assess the relationship between CHADS2 scores and stroke mechanisms.
Between August 1994 and March 2010, a total of 6059 consecutive patients with acute brain infarction or transient ischemic attack within 7 days after symptom onset were admitted to our hospital and registered to the Yonsei Stroke Registry. During admission, all patients were thoroughly investigated for demographic data, medical history, vascular risk factors, and clinical manifestations.11 They all underwent brain CT and/or MRI, 12-lead electrocardiography, and standard blood tests. The presence of cerebral atherosclerosis was investigated by digital subtraction angiography, MR angiography (MRA), or CT angiography (CTA) during hospitalization. Stroke subtypes and the presence of any abnormalities on angiographic studies were determined prospectively at a weekly stroke conference based on neuroradiologist reports and consensus of stroke specialists and were entered into the stroke registry.11 This study was approved by the Institutional Review Board of Severance Hospital, Yonsei University Health System, and the requirement for informed consent was waived.
Among 6059 patients, we excluded patients without atrial fibrillation (n=4995) or those with atrial fibrillation and valvular heart disease (n=144). For analyses of concomitant cerebral atherosclerosis, 140 patients who had not undergone angiographic studies were further excluded. The final study sample consisted of a total of 780 patients with NVAF.
The CHADS2 score was calculated for each patient with 1 point assigned to patients with a history of congestive heart failure, hypertension, age ≥75 years, or diabetes mellitus and 2 points for a history of stroke or transient ischemic attack. For this study, the CHADS2 score was based on the previous diagnosis or clinical history before the index stroke. The study sample was divided into 3 groups: low risk (CHADS2 score of 0), moderate risk (CHADS2 score of 1 or 2), and high risk (CHADS2 score of 3 to 6) based on previous studies.4,5
The complete history of medication before admission was available for 697 patients. Information regarding prior use of antithrombotics, statins, and angiotensin-converting enzyme inhibitors and angiotensin receptor blockers was collected. Data for the international normalized ratio at admission, smoking history, history of ischemic heart disease, and hypercholesterolemia were also collected.
Concomitant Cerebral Atherosclerosis
The presence of steno-occlusive lesions of the cerebral artery was investigated using the results of digital subtraction angiography in 265 (34.0%) patients, contrast-enhanced MRA in 386 (49.5%) patients, and CTA in 129 (16.5%) patients among 780 patients enrolled. Arterial stenosis was identified when the degree of stenosis was ≥50% based on the North American Symptomatic Carotid Endarterectomy Trial12 or Warfarin versus Aspirin for Symptomatic Intracranial Disease method.13 The presence of arterial stenotic lesions was determined for each arterial segment, including the extracranial or intracranial internal carotid artery, the anterior cerebral artery, the middle cerebral artery, the extracranial or intracranial vertebral artery, and the posterior cerebral artery on each side as well as the basilar artery. When multiple stenotic lesions were observed in 1 arterial segment, data from the most severe stenotic lesion were used. In patients with data for both digital subtraction angiography and contrast-enhanced MRA or CTA, the results of digital subtraction angiography were used for analysis, and in patients with data for both contrast-enhanced MRA and CTA, the results of CTA were used for analysis.
For the analysis of concomitant cerebral atherosclerosis, we included only steno-occlusive lesions after excluding symptomatic arterial lesions to infracted areas. However, if a patient had a stenotic lesion at the proximal carotid bulb or the orifice of the vertebral artery, this lesion was considered the concomitant cerebral atherosclerotic lesion. Likewise, in cases with tandem stenotic lesions (distal and proximal steno-occlusive lesions), we considered only the proximal arterial lesion to be a concomitant cerebral atherosclerotic lesion. After exclusion of the symptomatic arterial lesions, the stenotic lesions were divided into isolated extracranial, isolated intracranial, and combined extracranial and intracranial lesions. The total number of stenotic arterial lesions was also calculated.
Statistical analyses were performed using the Windows SPSS software package (Version 18.0; Chicago, IL). Continuous variables were compared with independent sample t tests, Mann-Whitney U tests, Kruskal-Wallis tests, or analysis of variance, and categorical variables with χ2 tests, as appropriate. Post hoc analyses were conducted with the Scheffe test. The increasing proportions of clinical variables according to the CHADS2 score were compared using the Mantel-Haenszel test. The relationship between numbers of lesions and CHADS2 scores was determined by Spearman rank test. Multivariate logistic analysis was used to compute the ORs of covariates, including the CHADS2 score, when determining independent factors predicting concomitant cerebral atherosclerosis or the presence of proximal arterial stenosis.
The baseline characteristics of the 780 patients enrolled in this study are provided in Table 1. Among 697 patients for whom a complete medication history was obtainable, warfarin use with optimum intensity (international normalized ratio, 2 to 3) was observed in only 32 patients (4.6%). According to CHADS2 scores, there was an increasing trend of having the components of the CHADS2 index (P<0.001) and receiving prior cardiovascular medications, except for anticoagulants.
Concomitant Cerebral Atherosclerosis According to the CHADS2 Score
In total, 231 patients (29.6%) were found to have atherosclerosis (≥50%) in ≥1 arteries at any cerebral arterial bed. Among these, isolated intracranial stenosis (47.6% [110 of 231]) was the most commonly involved location followed by isolated extracranial stenosis (33.8% [78 of 231]) and combined extracranial and intracranial stenosis (18.6% [43 of 232]). If we included all steno-occlusive lesions with the baseline angiographic studies, involvement of the middle cerebral artery was most common. However, after exclusion of the symptomatic arterial lesion, the extracranial vertebral artery was the most common artery involved followed by the middle cerebral artery and the extracranial internal carotid artery (Table 2).
There was a positive correlation between the CHADS2 score and the number of arteries affected by concomitant cerebral atherosclerosis (r=0.187, P<0.001). The burden of the concomitant cerebral atherosclerosis increased in an incremental fashion according to the CHADS2 score when the number of stenotic arteries was divided into 4 groups such as 0, 1, 2, 3, or ≥4 (P<0.001; Figure A). An increasing tendency for concomitant cerebral atherosclerosis was observed in all cerebral arterial beds as the CHADS2 score increased (Figure B). Concomitant cerebral atherosclerosis was observed in 20 patients (19.2%) in the low-risk group, 128 (27.0%) in the moderate-risk group, and 83 (41.1%) in the high-risk group. In particular, the proportion of combined extracranial and intracranial stenoses was relatively larger in the high-risk group than in the low- or moderate-risk group (Figure B).
The factors related to concomitant cerebral atherosclerosis were determined. Univariate analysis identified age (71.7±10.4 years with atherosclerosis versus 68.6±10.5 years without, P<0.001), hypertension (P<0.001), diabetes mellitus (P=0.017), and previous stroke or transient ischemic attack (P=0.019) along with belonging to the high-risk group (P<0.001) were associated with the presence of concomitant cerebral atherosclerosis. Multivariate logistic regression analysis revealed that belonging to the high-risk group with CHADS2 scores of 3 to 6 was a significant and independent predictor for concomitant cerebral atherosclerosis (Table 3). Among components of the CHADS2 score, hypertension, age ≥75 years, and previous ischemic stroke were determinants of cerebral atherosclerosis in patients with NVAF. When we further analyzed the data of digital subtraction angiography only, the high-risk group was also an independent and significant predictor for concomitant cerebral atherosclerosis (Supplemental Table I, http://stroke.ahajournals.org).
Stroke Mechanism According to the CHADS2 Score
Of 780 patients, 21 patients had transient ischemic attack without a relevant acute ischemic lesion on brain imaging studies. The stroke mechanism was investigated in 759 patients who had neurological symptoms and clinically relevant ischemic lesions on brain imaging.
In this sample, 128 patients (16.9%) had a proximal stenotic lesion at the symptomatic artery, which suggests that the atherothrombotic mechanism could be a possible cause of ischemic stroke. Proximal atherosclerotic lesions were most frequently found in the high-risk group (24.0% [46 of 192]) followed by the moderate-risk (15.5% [72 of 465]) and low-risk groups (9.8% [10 of 102]), and these differences were significant (P=0.001). Among components of the CHADS2 score, hypertension (P<0.001) and previous ischemic stroke (P=0.001) were associated with the presence of proximal atherosclerotic lesions. Other variables including age ≥75 years, diabetes mellitus, smoking, hyperlipidemia, previous ischemic heart disease, and prior medication before the stroke were not significant determinants in univariate analysis. Multivariate analysis revealed that high risk based on the CHADS2 score was associated with the presence of proximal arterial stenosis compared with the low-risk group. Among components of the CHADS2 score, hypertension and previous ischemic stroke were significant predictors of proximal atherosclerotic lesions (Table 4).
We have demonstrated that cerebral atherosclerosis is more common in patients with higher CHADS2 scores and that these patients also have higher risks of atherothrombotic stroke as well as cardioembolic stroke.
In our study sample, approximately one third of the patients demonstrated significant atherosclerotic lesions in the cerebral artery other than the steno-occlusive lesion in the symptomatic arteries. The frequency and burden of cerebral atherosclerosis increased with increases in the CHADS2 score. Among components of the CHADS2 score, older age, hypertension, diabetes, and previous ischemic stroke are well-known vascular risk factors for atherosclerosis and first-ever or recurrent ischemic stroke. The frequency or burden of cerebral atherosclerosis escalates with increasing numbers of vascular risk factors,14 and such increases may also occur in patients with NVAF. Along with the known contributory roles of the components of the CHADS2 score on atherosclerosis, our findings indicate that the risk of cerebral atherosclerosis increases as the CHADS2 score increases.
A previous study showed that the frequency of the left atrial or left atrial appendage thrombus increased with the ascending CHADS2 score, which suggests that the components of the CHADS2 score may enhance the thrombogenic milieu in the intracardiac chamber.6,7 Such changes in the intracardiac chamber may increase the risk of cardioembolism. However, the risk of atherothrombotic stroke in patients with NVAF may also increase in patients with higher CHADS2 scores. In patients with NVAF, not only cardioembolic, but also noncardioembolic ischemic stroke can occur.9,15 In our study, patients included in the high-risk group more frequently had combined extracranial and intracranial stenosis and proximal stenotic lesions of the symptomatic artery. The incidence of ischemic and total cerebrovascular events was greater in patients with combined extracranial and intracranial stenosis than in patients with only isolated extracranial or intracranial stenosis.14 Given that artery-to-artery embolism from the proximal atherosclerotic plaques is the main stroke mechanism in atherothrombotic stroke,9 the proximal arterial lesions may actually cause stroke in some patients with NVAF. In this regard, increased risk of stroke in patients with higher CHADS2 scores may be associated with increased risk of atherothrombotic stroke in addition to preexisting atrial fibrillation-related risk of cardioembolic stroke.
Patients with higher CHADS2 score may require additional measures for prevention of ischemic stroke because concomitant cerebral atherosclerosis was frequent in those groups. The measures may include combined use of an antiplatelet drug and warfarin as well as strict risk factor controls. However, in case of combined use of an anticoagulant and antiplatelet drugs, increased risks of hemorrhagic complications also should be considered, in particular in Asian patients who are vulnerable to intracranial hemorrhage. Alternatively, combined use of antiplatelet agents and a direct thrombin inhibitor, 110 mg Dabigatran, which had less hemorrhagic complications than warfarin,16 or closure of the left atrial appendage may be considered.
There are some limitations in this study. First, this is a retrospective study and angiographic results with various modalities were used for the investigation of cerebral arterial stenosis. Although contrast-enhanced MRA had higher sensitivity than other imaging modalities such as CTA or MRA,17,18 Contrast-enhanced MRA also might produce an artifactual overestimation of arterial stenosis in some patients.19 Second, in our study, stenotic lesions were more common in the intracranial artery than in the extracranial artery, possibly due to ethnic characteristics. Intracranial atherosclerosis, especially isolated intracranial artery stenosis, is more common in East Asians than in Westerners.20–23 Stenosis or occlusion of the symptomatic intracranial artery can be either due to embolism from the heart or in situ thrombosis/arterial embolism from the atherosclerotic lesion. These lesions often cannot be differentiated on angiographic studies. Therefore, steno-occlusion of the symptomatic intracranial artery was not considered as an atherosclerotic lesion in our analysis. This definition in our study might lead to an underestimation of the frequency of intracranial artery atherosclerosis and the atherothrombotic mechanism attributed to intracranial artery atherosclerosis.
The results of our study suggest that the presence and burden of cerebral atherosclerosis can be predicted by CHADS2 scores. Our findings also suggest that atherothrombosis should be considered to be a potential stroke mechanism in patients with high CHADS2 scores. Approximately 40% of high-risk patients (CHADS2 score ≥3) have significant stenosis of the large artery that may necessitate interventional treatment in some patients for improved stroke prevention. These patients may also require additional antithrombotic treatment. In this regard, strict evaluation of cerebral atherosclerosis should be considered in patients with higher CHADS2 scores.
Sources of Funding
This work was supported by a grant from the Korea Health 21 R&D Project, Ministry of Health & Welfare, Republic of Korea (A060171, 085136).
The online-only Data Supplement is available at http://stroke.ahajournals.org/cgi/content/full/STROKEAHA.110.602987/DC1.
- Received September 16, 2010.
- Revision received November 11, 2010.
- Accepted November 12, 2010.
- © 2011 American Heart Association, Inc.
- Hart RG,
- Boop BS,
- Anderson DC
- Flibotte JJ,
- Hagan N,
- O'Donnell J,
- Greenberg SM,
- Rosand J
Stroke Risk in Atrial Fibrillation Working Group. Comparison of 12 risk stratification schemes to predict stroke in patients with nonvalvular atrial fibrillation. Stroke. 2008;39:1901–1910.
- Puwanant S,
- Varr BC,
- Shrestha K,
- Hussain SK,
- Tang WH,
- Gabriel RS,
- Wazni OM,
- Bhargava M,
- Saliba WI,
- Thomas JD,
- Lindsay BD,
- Klein AL
- Kanter MC,
- Tegeler CH,
- Pearce LA,
- Weinberger J,
- Feinberg WM,
- Anderson DC,
- Gomez CR,
- Rothrock JF,
- Helgason CM,
- Hart RG
- Connolly SJ,
- Ezekowitz MD,
- Yusuf S,
- Eikelboom J,
- Oldgren J,
- Parekh A,
- Pogue J,
- Reilly PA,
- Themeles E,
- Varrone J,
- Wang S,
- Alings M,
- Xavier D,
- Zhu J,
- Diaz R,
- Lewis BS,
- Darius H,
- Diener HC,
- Joyner CD,
- Wallentin L
- Khan S,
- Rich P,
- Clifton A,
- Markus HS
- Wong KS,
- Ng PW,
- Tang A,
- Liu R,
- Yeung V,
- Tomlinson B
- Suh DC,
- Lee SH,
- Kim KR,
- Park ST,
- Lim SM,
- Kim SJ,
- Choi CG,
- Lee HK