Abstract
Background
Whether carotid artery disease could improve stroke risk stratification tools in patients with atrial fibrillation (AF) remains uncertain.
Aim
To investigate the risk of ischemic stroke associated with occlusive and non-occlusive carotid disease in patients with AF in the prospective population-based Cardiovascular Health Study.
Methods
We included participants aged ≥65 years with AF. We used multivariable cox regression analysis to explore the risk of ischemic stroke associated between the percentage of carotid stenosis, plaque irregularity, echogenicity, and vulnerability (markedly irregular, ulcerated, or hypoechoic plaques).
Results
A total of 1398 participants were included, 55.2% females and 61.7% aged 65–74 years. The maximum carotid stenosis was <50%, 50–99%, and 100% in 94.5%, 5%, and 0.5% of participants, respectively. High-risk plaques based on echogenicity and plaque irregularity were found in 25.6% and 8.9% of participants, respectively. After a median follow-up of 10.9 years (IQR 7.5–15.6), 298 ischemic strokes occurred. There was no difference in the incidence of ischemic stroke according to the degree of carotid artery stenosis (p=0.44), plaque echogenicity (low versus high risk, p=0.68), plaque irregularity (low versus high risk, p=0.55), and plaque vulnerability (p=0.86). The CHA₂DS₂-VASc score was associated with an increased risk of ischemic stroke (adjusted hazard ratio 1.28, 95% CI: 1.18–1.40, p<0.001). Both maximum grade of stenosis and plaque vulnerability were not associated with incident ischemic stroke (all p>0.05).
Conclusion
Neither the degree of carotid stenosis nor the presence of vulnerable plaques was associated with incident ischemic stroke in this cohort with AF. This suggests that carotid disease was probably not a significant contributor to ischemic stroke in this population.
Keywords: atrial fibrillation, stroke, carotid artery stenosis, carotid plaque
INTRODUCTION
Stroke prevention is one of the pillars of atrial fibrillation (AF) management. Currently, this is mainly achieved through the use of oral anticoagulation (OAC) with either vitamin K antagonists, direct oral anticoagulants, or percutaneous left atrial appendage occlusion (1). The decision to start OAC is guided by an assessment of the risk of stroke and the risk of bleeding. Current stroke risk stratification schemes have limited performance for predicting thromboembolism, especially in patients with AF and low risk (1, 2). This is partly explained by the fact that these stroke risk prediction tools do not appropriately account for non-cardioembolic causes of stroke such as carotid atherosclerosis (1, 3). A recent meta-analysis showed that AF and carotid stenosis frequently coexist, with about one in ten patients with AF having carotid stenosis, and vice versa. Furthermore, about half of patients with AF have non-stenotic carotid disease (4). Interestingly, severe carotid stenosis accounted for about one fourth of ischemic strokes in a cohort of 103 patients with AF (5). The most popular stroke risk prediction tool, the CHA₂DS₂-VASc score, does not include carotid disease as a risk factor (2). Although carotid occlusive disease is included in the new GARFIELD-AF risk calculator, its weight in the risk calculation remains unclear (6).
There is some evidence suggesting that carotid plaque predicts stroke or transient ischemic attack in patients with AF (7–9). The inclusion of carotid plaque and carotid intima-media thickness (cIMT) into stroke risk stratification tools for patients with AF could significantly improve their performance. However, it remains unclear whether there are specific features in carotid plaque that are more predictive of cerebrovascular events in patients with AF. In fact, carotid plaques with high-risk features such as echolucency, lipid-rich necrotic core, neovascularization, intraplaque hemorrhage or ulcerations, are associated with an increased risk of ipsilateral ischemic stroke in asymptomatic carotid atherosclerosis (10).
To the best of our knowledge, no previous study has evaluated whether high-risk carotid plaques could improve stroke risk prediction in patients with AF. Using data from the Cardiovascular Health Study, we aimed to determine 1) the prevalence of occlusive and non-occlusive carotid disease in patients with and without AF; 2) the distribution of non-occlusive carotid disease and high-risk features (plaque irregularity, ulceration, and echolucency) in patients with and without AF and; 3) the risk of stroke associated with occlusive and non-occlusive carotid disease in patients with and without AF; 4) the added value of high-risk carotid plaques on stroke risk stratification. We hypothesized that carotid plaques with high-risk features such as plaque irregularity, ulceration, and echolucency, are associated with an increased risk of stroke in patients with AF, and that high-risk carotid plaques could improve stroke risk prediction in patients with AF.
METHODS
Study design
This study was performed using data from the Cardiovascular Health Study (CHS). The CHS prospectively enrolled and continues to follow a community-dwelling cohort of men and women aged ≥65 years randomly selected from Medicare eligibility lists in four counties in the USA, California, Maryland, North Carolina, and Pennsylvania. An initial cohort of 5,201 participants was recruited in 1989–1990 and a supplemental cohort of 687 predominantly African American participants was recruited in 1992–1993. Participants returned for annual in-person study visits until 1998–1999 and again in 2005–2006. Throughout follow-up, participants were contacted via semiannual telephone calls, and data were linked with Medicare claims. Carotid ultrasound scans were performed at baseline, year 5, and year 11.
Participants
For this study, we included all participants with AF and available carotid ultrasound data. Atrial fibrillation was identified from 3 sources: i) participant’s self-report; ii) outpatient ECGs obtained yearly at study examinations; iii) hospital discharge primary diagnoses with ICD-9 codes for AF (427.31), excluding diagnosis assigned during the same hospitalization for a cardiac surgery (ICD-9 codes of 35.x, 36.x, or 37.x). Eligible participants were divided in 2 groups depending on when they had AF: Group 1 included participants with AF at baseline; Group 2 included participants without AF at baseline but who developed AF by year 5 (Figure 1. Panels A and B). Year 5 was selected because participants had follow-up carotid ultrasound at this time, allowing us to have baseline ultrasound data for participants who developed AF by year 5. The follow-up started at baseline for Group 1 and at year 5 for Group 2 (Figure 1. Panel B). Baseline characteristics of each group correspond to their characteristics at the beginning of their specific follow-up period.
Figure 1.
Participant selection and follow-up.
Legend: Panel A. Participant selection; Panel B. Follow-up
Exposure variables
The main exposure variables were carotid ultrasound features including the percentage of stenosis, plaque irregularity, and plaque echogenicity. Duplex ultrasonography of both carotid arteries was performed at baseline and at Year 5 with a Toshiba SSA-270A ultrasound device (Toshiba American Medical Systems, Tustin, CA) equipped with 5.0 MHz transducer. We considered carotid ultrasound measurements done at baseline for Group 1 and at Year 5 for Group 2.
Percentage stenosis was classified as: 0% or Normal, 1–24%, 25–49%, 50–74%, 75–99%, and 100% (11).
Plaque irregularity was classified as: smooth, mildly irregular, markedly irregular, and ulcerated (11).
Plaque irregularly was further dichotomized as low-risk (smooth and mildly irregular) and high-risk (markedly irregular and ulcerated).
Plaque echogenicity was coded as: no lesion, hypoechoic, isoechoic, hyperechoic, and calcified (11).
Plaque echogenicity was further dichotomized as low-risk (no lesion, isoechoic, hyperechoic, and calcified) and high-risk (hypoechoic).
Vulnerable plaque was defined as the presence of plaques in either of the carotid arteries with at least one of the following high-risk features: hypoechoic, markedly irregular, or ulcerated.
Plaques that are hypoechoic, markedly irregular, or ulcerated are considered high-risk based on previous studies showing that these ultrasound features are associated with an increased risk of stroke (10, 12).
Other variables of interest included sociodemographic characteristics; cardiovascular risk factors such as hypertension, diabetes mellitus, obesity, dyslipidemia (Supplementary Table 1), heart failure, coronary artery disease, peripheral artery disease, prior stroke or transient ischemic attack (TIA), chronic kidney disease, smoking, alcohol consumption, and use of antithrombotic medication. The CHA₂DS₂-VASc score was calculated for each participant (Supplementary Table 2).
Outcome variable
The outcome was incident ischemic stroke. Stroke was defined as the rapid onset of focal neurological deficit lasting more than 24 hours or until death, a lesion on computed tomography or magnetic resonance imaging, with the exclusion of brain concussion, tumor, or infection as the cause of the deficit. Ischemic stroke was further defined as: i) a focal neurological deficit without evidence of intracranial hemorrhage on computed tomography, magnetic resonance imaging, or cerebrospinal fluid analysis or ii) imaging evidence of brain ischemia in a location compatible with the presenting symptoms (13). We consider the start of follow-up as baseline for Group 1, and Year 5 for Group 2.
Statistical analysis
Categorical variables were expressed as frequencies and percentages while continuous variables were expressed as mean with standard deviation (SD) or median with interquartile range (IQR) as appropriate. We assessed differences between groups using the Student’s t-test for continuous variables and the Pearson Chi square test for categorical variables. Time to incident ischemic stroke was assessed using survival analysis. Participants were censored at the time of ischemic stroke, last follow-up, or death. Factors associated with incident ischemic stroke were assessed using univariable and multivariable cox regression analysis, with risk estimates reported as hazard ratio (HR) along with 95% confidence interval (CI). A two-tailed p-value of less than 0.05 was considered to indicate statistical significance. All analyses were performed using Stata 16.1 statistical package (Stata Corp, College Station, TX).
Ethical considerations
The CHS received approval from institutional review boards of all participating centers and all participants provided written informed consent. In addition, prior to receiving the CHS data, the current study was approved by the Human Research Ethics Committee of the Central Adelaide Local Health Network, Adelaide, Australia (CALHN Reference Number: 15839).
RESULTS
General characteristics
Out of 5888 participants in CHS, 1398 with AF were included in the analysis. There were 266 (19.0%) participants with AF at baseline (Group 1) and 1132 (81.0%) without AF at baseline but who developed AF by Year 5 (Group 2) (Figure 1. Panels A and B). The general characteristics of the study population are presented in Table 1. There were more females (55.2%) than males, and participants aged 65–74 years (61.7%) than those aged ≥75 years. The prevalence of cardiovascular risk factors was 81.4% for dyslipidemia, 62.3% for hypertension, 31.8% for vascular disease, 21.9% for chronic kidney disease, 16.4% for diabetes mellitus, 6.4% for heart failure and, 6.1% for previous stroke or TIA. The median CHA₂DS₂-VASc score was 3 (IQR 2–4).
Table 1.
General characteristics of the study population
| Variables | Group 1 (n = 266) |
Group 2 (n =1132) |
Total (n = 1398) |
|---|---|---|---|
| Age | |||
| 65–74 years | 145 (54.5%) | 718 (63.4%) | 863 (61.7%) |
| ≥75 years | 121 (45.5%) | 414 (36.6%) | 535 (38.3%) |
| Sex | |||
| Male | 135 (50.8%) | 492 (43.5%) | 627 (44.8%) |
| Female | 131 (49.2%) | 640 (56.5%) | 771 (55.2%) |
| Body mass index | |||
| <25 | 120 (45.1%) | 405 (35.8%) | 525 (37.6%) |
| 25–29.9 | 101 (38.0%) | 464 (41.0%) | 565 (40.4%) |
| ≥30 | 45 (16.9%) | 263 (23.2%) | 308 (22.0%) |
| Cardiovascular risk factors | |||
| Hypertension | 182 (68.4%) | 689 (60.9%) | 871 (62.3%) |
| Diabetes mellitus | 59 (22.2%) | 170 (15.0%) | 229 (16.4%) |
| Dyslipidemia | 221 (83.1%) | 917 (81.0%) | 1138 (81.4%) |
| Heart failure | 44 (16.5%) | 46 (4.1%) | 90 (6.4%) |
| Previous stroke or TIA | 28 (10.5%) | 57 (5.0%) | 85 (6.1%) |
| Current smoking | 20 (7.5%) | 14 (1.2%) | 34 (2.4%) |
| CAD | 114 (42.9%) | 307 (21.7%) | 421 (30.1%) |
| PAD | 15 (5.6%) | 29 (2.6%) | 44 (3.1%) |
| Vascular disease | 121 (45.5%) | 323 (28.5%) | 444 (31.8%) |
| Chronic kidney disease | 88 (33.1%) | 218 (19.3%) | 306 (21.9%) |
| CHA₂DS₂-VASc score | |||
| Median (IQR) | 4 (3–5) | 3 (2–4) | 3 (2–4) |
| Antithrombotic medication | |||
| Antiplatelet | 15 (5.6) | 30 (2.7) | 45 (3.2) |
| Oral anticoagulant | 30 (11.3) | 34 (3.0) | 64 (4.6) |
CAD: coronary artery disease; IQR: interquartile range; PAD: peripheral artery disease; TIA: transient ischemic attack
Carotid artery ultrasound features
Carotid artery stenosis
Most participants had a normal ultrasound examination or carotid artery stenosis <50% (96.6% on the right and 96.5% on the left). On either the left or the right carotid arteries, the maximum stenosis was <50% in 94.5% of participants, 50–99% in 5% of participants, and 0.5% had carotid occlusion (Table 2).
Table 2.
Carotid artery ultrasound features
| Variables | Group 1 (n = 266) |
Group 2 (n =1132) |
Total (n = 1398) |
|---|---|---|---|
| Percentage stenosis, RIGHT | |||
| Normal | 79 (29.7%) | 366 (32.3%) | 445 (31.8%) |
| 1–24% | 73 (27.4%) | 360 (31.8%) | 433 (31.0%) |
| 25–49% | 107 (40.2%) | 366 (32.3%) | 473 (33.8%) |
| 50–74% | 5 (1.9%) | 30 (2.7%) | 35 (2.5%) |
| 75–99% | 1 (0.4%) | 8 (0.7%) | 9 (0.6%) |
| 100% | 1 (0.4%) | 2 (0.2%) | 3 (0.2%) |
| Percentage stenosis, LEFT | |||
| Normal | 68 (25.6%) | 398 (35.2%) | 466 (33.3%) |
| 1–24% | 81 (30.5%) | 334 (29.5%) | 415 (29.7%) |
| 25–49% | 111 (41.7%) | 358 (31.6%) | 469 (33.5%) |
| 50–74% | 5 (1.9%) | 30 (2.7%) | 35 (2.5%) |
| 75–99% | 0 (0.0%) | 8 (0.7%) | 8 (0.6%) |
| 100% | 1 (0.4%) | 4 (0.4%) | 5 (0.4%) |
| Percentage stenosis, MAX | |||
| Normal | 37 (13.9%) | 220 (19.4%) | 257 (18.4%) |
| 1–24% | 81 (30.5%) | 357 (31.5%) | 438 (31.3%) |
| 25–49% | 137 (51.5%) | 489 (43.2%) | 626 (44.8%) |
| 50–74% | 9 (3.4%) | 44 (3.9%) | 53 (3.8%) |
| 75–99% | 1 (0.4%) | 16 (1.4%) | 17 (1.2%) |
| 100% | 1 (0.4%) | 6 (0.5%) | 7 (0.5%) |
| Plaque irregularity, RIGHT | |||
| No lesion | 79 (29.7%) | 366 (32.3%) | 445 (31.8%) |
| Smooth | 127 (47.7%) | 397 (35.1%) | 524 (37.5%) |
| Mildly irregular | 55 (20.7%) | 306 (27.0%) | 361 (25.8%) |
| Markedly irregular | 5 (1.9%) | 57 (5.0%) | 62 (4.4%) |
| Ulcerated | 0 (0.0%) | 6 (0.5%) | 6 (0.4%) |
| Plaque irregularity, LEFT | |||
| No lesion | 68 (25.6%) | 398 (35.2%) | 466 (33.3%) |
| Smooth | 143 (53.8%) | 365 (32.2%) | 508 (36.4%) |
| Mildly irregular | 54 (20.3%) | 297 (26.2%) | 351 (25.1%) |
| Markedly irregular | 1 (0.4%) | 68 (6.0%) | 69 (4.9%) |
| Ulcerated | 0 (0.0%) | 4 (0.4%) | 4 (0.3%) |
| Plaque echogenicity, RIGHT | |||
| No lesion | 80 (30.1%) | 369 (32.6%) | 449 (32.1%) |
| Hypoechoic | 44 (16.5%) | 167 (14.8%) | 211 (15.1%) |
| Isoechoic | 94 (35.3%) | 394 (34.8%) | 488 (34.9%) |
| Hyperechoic | 42 (15.8%) | 161 (14.2%) | 203 (14.5%) |
| Calcified | 6 (2.3%) | 41 (3.6%) | 47 (3.4%) |
| Plaque echogenicity, LEFT | |||
| No lesion | 68 (25.6%) | 398 (35.2%) | 466 (33.3%) |
| Hypoechoic | 52 (19.5%) | 152 (13.4%) | 204 (14.6%) |
| Isoechoic | 99 (37.2%) | 333 (29.4%) | 432 (30.9%) |
| Hyperechoic | 45 (16.9%) | 195 (17.2%) | 240 (17.2%) |
| Calcified | 2 (0.8%) | 54 (4.8%) | 56 (4.0%) |
| Plaque irregularity | |||
| Low risk | 260 (97.7%) | 1013 (89.5%) | 1273 (91.1%) |
| High risk | 6 (2.3%) | 119 (10.5%) | 125 (8.9%) |
| Plaque echogenicity | |||
| Low risk | 185 (69.5%) | 855 (75.5%) | 1040 (74.4%) |
| High risk | 81 (30.5%) | 277 (24.5%) | 358 (25.6%) |
| Vulnerable plaque | |||
| No | 180 (67.7%) | 748 (66.1%) | 928 (66.4%) |
| Yes | 86 (32.3%) | 384 (33.9%) | 470 (33.6%) |
Carotid plaques
Regarding plaque irregularity, most participants had either smooth carotid arteries (69.3% on the right and 69.6% on the left) or mildly irregular plaques (25.8% and 25.2% on the right side and left side, respectively). Markedly irregular (4.4% and 4.9% on the right side and left side, respectively) and ulcerated plaques (0.4% and 0.3% on the right and left sides, respectively) were much less frequent (Table 2). Regarding echogenicity, most participants had either no lesion (32.0% and 33.3% on the right side and left side, respectively) or isoechoic lesion (35.0% and 30.9% on the right side and left side, respectively) (Table 2).
The proportion of participants with high-risk plaques was 25.6% based on echogenicity and 8.9% based on plaque irregularity. When considering both irregularity and echogenicity, 33.6% of participants had vulnerable plaques (Table 2).
Incidence of ischemic stroke according to carotid plaque characteristics
After a median follow-up of 10.9 years (IQR 7.5–15.6), 298 ischemic strokes occurred. There was no difference in the incidence of ischemic stroke according to the degree of carotid artery stenosis (p=0.44, Figure 2. Panel A), the echogenicity of plaque (low risk versus high risk, p=0.68, Figure 2. Panel B), the irregularity of plaque (low risk versus high risk, p=0.55, Figure 2. Panel C), and the vulnerability of the plaque (p=0.86, Figure 2. Panel D).
Figure 2.
Incidence of ischemic stroke according to carotid ultrasound features
Legend: Panel A. Incidence of ischemic stroke according to maximum stenosis; Panel B. Incidence of ischemic stroke according to plaque echogenicity; Panel C. Incidence of ischemic stroke according to plaque irregularity; Panel D. Incidence of ischemic stroke according to plaque vulnerability
Predictors of ischemic stroke
Factors associated with incident ischemic stroke in univariable cox regression analysis included aged ≥75 years, hypertension, diabetes mellitus, vascular disease, previous stroke or TIA, and chronic kidney disease. In multivariable analysis, predictors of ischemic stroke included aged ≥75 years (adjusted hazard ratio [aHR] 1.35, 95% CI: 1.05–1.71, p=0.015), hypertension (aHR 1.37, 95% CI: 1.06–1.76, p=0.016), diabetes mellitus (aHR 1.95, 95% CI: 1.46–2.62, p<0.001), previous stroke or TIA (aHR 1.59, 95% CI: 1.01–2.51, p=0.045), and chronic kidney disease (aHR 1.33, 95% CI: 1.01–1.76, p=0.043) (Table 3). The CHA₂DS₂-VASc score was associated with an increased risk of ischemic stroke (aHR 1.28, 95% CI: 1.18–1.40, p<0.001) (Table 4). Both maximum grade of stenosis and plaque vulnerability were not associated with incident ischemic stroke (all p>0.05) (Tables 3 and 4).
Table 3.
Association between clinical and carotid ultrasound features and incident ischemic stroke
| Variables | Univariable | Multivariable | ||||
|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | |
| Age | ||||||
| 65–74 years | Ref | - | - | Ref | - | - |
| ≥75 years | 1.40 | 1.11–1.78 | 0.005 | 1.35 | 1.05–1.71 | 0.015 |
| Sex | ||||||
| Female | Ref | - | - | Ref | - | - |
| Male | 0.92 | 0.72–1.16 | 0.457 | 0.91 | 0.71–1.16 | 0.430 |
| Current smoker | ||||||
| No | Ref | - | - | Ref | - | - |
| Yes | 1.05 | 0.39–2.83 | 0.922 | 1.22 | 0.45–3.32 | 0.697 |
| Body mass index | ||||||
| <25 | Ref | - | - | Ref | - | - |
| 25–29.9 | 0.92 | 0.71–1.19 | 0.503 | 0.89 | 0.68–1.17 | 0.414 |
| ≥30 | 0.96 | 0.71–130 | 0.791 | 0.79 | 0.58–1.09 | 0.154 |
| Hypertension | ||||||
| No | Ref | - | - | Ref | - | - |
| Yes | 1.55 | 1.22–198 | <0.001 | 1.37 | 1.06–1.76 | 0.016 |
| Diabetes mellitus | ||||||
| No | Ref | - | - | Ref | - | - |
| Yes | 1.96 | 1.48–2.58 | <0.001 | 1.95 | 1.46–2.62 | <0.001 |
| Dyslipidemia | ||||||
| No | Ref | - | - | Ref | - | - |
| Yes | 1.20 | 0.87–1.65 | 0.261 | 1.18 | 0.86–1.63 | 0.301 |
| Heart failure | ||||||
| No | Ref | - | - | Ref | - | - |
| Yes | 1.58 | 0.94–2.67 | 0.085 | 1.27 | 0.74–2.17 | 0.390 |
| Vascular disease | ||||||
| No | Ref | - | - | Ref | - | - |
| Yes | 1.30 | 1.01–1.67 | 0.044 | 1.07 | 0.82–1.40 | 0.616 |
| Previous stroke or TIA | ||||||
| No | Ref | - | - | Ref | - | - |
| Yes | 1.81 | 1.16–2.82 | 0.009 | 1.59 | 1.01–2.51 | 0.045 |
| Chronic kidney disease | ||||||
| No | Ref | - | - | Ref | - | - |
| Yes | 1.42 | 1.09–1.86 | 0.011 | 1.33 | 1.01–1.76 | 0.043 |
| Percentage stenosis | ||||||
| Normal | Ref | - | - | Ref | - | - |
| 1–49% | 1.21 | 0.90–1.61 | 0.202 | 1.22 | 0.89–1.67 | 0.210 |
| 50–100% | 1.14 | 0.61–2.13 | 0.685 | 1.02 | 0.53–1.96 | 0.956 |
| Vulnerable plaque | ||||||
| No | Ref | - | - | Ref | - | - |
| Yes | 1.02 | 0.80–1.30 | 0.867 | 0.94 | 0.72–1.22 | 0.619 |
95% CI: 95% confidence interval; HR: hazard ratio; Ref: reference category; TIA: transient ischemic attack
Table 4.
Association between CHA₂DS₂−VASc score and carotid ultrasound features and incident ischemic stroke
| Variables | HR | 95% CI | P value |
|---|---|---|---|
| Model 1 | |||
| CHA₂DS₂−VASc score (per 1 unit) | 1.29 | 1.18–141 | <0.001 |
| Maximum percentage stenosis | |||
| • Normal | Ref | − | − |
| • 1–49% | 1.24 | 0.93–1.66 | 0.151 |
| • 50–100% | 1.04 | 0.56–1.95 | 0.898 |
| Model 2 | |||
| CHA₂DS₂−VASc score (per 1 unit) | 1.28 | 1.18–1.40 | <0.001 |
| Vulnerable plaque | |||
| • No | Ref | − | − |
| • Yes | 1.00 | 0.79–1.28 | 0.972 |
95% CI: 95% confidence interval; Ref: reference category; HR: hazard ratio
Model 1: CHA₂DS₂−VASc score and maximum percentage stenosis; Model 2: CHA₂DS₂-VASc score and plaque vulnerability
DISCUSSION
This study aimed to determine the pattern of occlusive and non-occlusive carotid artery disease in patients with AF and assess whether prediction of their stroke risk could be improved by considering the presence of carotid stenosis or vulnerable carotid plaque. We observed that 1) moderate to severe carotid stenosis (≥ 50%) was uncommon, affecting only 5.5% of participants; 2) a third of patients had carotid plaques considered vulnerable or high-risk; 3) the degree of carotid stenosis and the presence of vulnerable plaques were not associated with incident ischemic stroke and therefore, could not have any added value in stroke risk stratification in this cohort of patients.
The prevalence of carotid stenosis in this cohort of patients with AF is much lower than was reported in several previous study. In fact, a meta-analysis of prevalence rates of carotid stenosis in patients with AF showed a pooled prevalence of 13.9% (4), much higher than the 5.5% prevalence in our study. There are few potential explanations for this difference. There was a wide variation in the prevalence of carotid stenosis across the studies included in the meta-analysis, ranging from 4.4% to 24.3% (4). Subgroup analysis by geographic region showed that the pooled prevalence was 8.2% in studies done in Northern America, an estimate closer to our observation in this cohort from the US, compared with 14.2% and 12.9% in studies done in Asia and Europe, respectively. Hence, variations in the prevalence of carotid stenosis in patients with AF might be partly related to geographic regions. These geographic differences in prevalence rates might be due to variations in the genetic predisposition to risk factors of atherosclerosis, to differences in primary prevention or disease detection. Furthermore, our cohort is quite old, with the first participants enrolled in 1989–1990 (three decades ago). It is possible that the burden of carotid artery disease has increased over time, with recent studies reporting higher prevalence rates. Finally, our cohort consists of people recruited from the general population who likely had a better cardiovascular risk profile than that of patients with AF recruited in hospital settings, hence a lower burden of atherosclerosis including in the carotid arteries. Indeed, the proportions of people with cardiovascular risk factors such as hypertension, diabetes mellitus, previous stroke, or smoking were lower in our cohort compared to other studies on carotid artery disease in patients with AF (4).
Contrary to our hypothesis, the degree of carotid stenosis was not associated with incident ischemic stroke. Studies on the association of carotid stenosis with thromboembolism in patients with AF have shown inconsistent results (3). Carotid stenosis was found to be associated with stroke or TIA in one retrospective study (aHR 1.49, 95% CI 1.30–1.71, mean follow-up 5.3 years) (14), and with recurrent stroke in another retrospective cohort of patients with AF and a previous stroke or TIA (aHR 2.02, 95% CI 1.37–3.01) (15). Dissimilarly, a prospective cohort study in Italy showed no association between carotid stenosis and ischemic stroke or TIA (adjusted HR 1.03, 95% CI 0.30–3.45), and a composite of stroke, TIA, or systemic embolism (adjusted HR 1.31, 95% CI 0.45–3.81, mean follow-up of 3.4 years) (16). Some inconsistencies have also been reported in studies on the association of carotid plaque with thromboembolism in patients with AF, although most studies suggested an association (3).
In a meta-analysis of 64 studies that enrolled 20751 participants, high-risk plaques were common in patients with asymptomatic carotid stenosis, and the associated annual incidence of ipsilateral ischemic events was higher than the currently accepted estimates (10). These findings suggested that extending the assessment of asymptomatic carotid stenosis beyond the grade of stenosis is needed in routine practice to improve risk stratification (10). In the current study, we hypothesized that carotid plaques with high-risk features such as plaque irregularity, ulceration, and echolucency, are associated with an increased risk of stroke in patients with AF, and that high-risk carotid plaques could improve stroke risk prediction in patients with AF. Contrary to our hypothesis, vulnerable plaques defined as the presence of hypoechoic, markedly irregular, or ulcerated plaques were not associated with incident ischemic stroke. Yet, a previous analysis of the CHS showed that risk of incident stroke was associated with two ultrasound features: presence of a hypoechoic plaque in the internal carotid artery and an estimated grade of stenosis ≥50% (12). Importantly, this previous analysis of the CHS was not restricted to patients with AF, and it excluded all strokes of cardiac origin (12). Therefore, the absence of association between grade of carotid stenosis or presence of high-risk carotid plaques and incident stroke suggests that carotid disease was probably not an important cause of the stroke events in our cohort of patients with AF. These findings indicate that vulnerable carotid plaques might not improve stroke risk stratification in patients with AF.
This study has some limitations. First, our study population included only patients aged ≥65 years and mostly whites, hence limiting its generalizability. Second, data was not available for systemic embolism which is a relevant endpoint for our study hypothesis. Third, several parameters could not be considered in our analysis, notably changes in medical therapy over time (antiplatelet, antihypertensive, and lipid-lowering drugs), vascular interventions (endarterectomy because of TIA or amaurosis fugax), and localization of the ischemic stroke that might not be in the relevant territory. Fourth, there was a small proportion of participants with AF affected by moderate to severe carotid stenosis. Hence, the study might have limited power to detect an association between moderate to severe carotid stenosis and incident stroke in this population. Strengths of our study include a well-characterized population and larger than those from previous studies on the association of carotid disease with thromboembolism in patients with AF (3), as well as a meticulous adjudication of the outcome of stroke (13).
CONCLUSION
Moderate to severe carotid stenosis (≥50%) was uncommon, affecting only 5.5% of participants with AF, whereas a third of patients with carotid atherosclerosis had plaques considered vulnerable or high-risk. Neither the degree of carotid stenosis nor the presence of vulnerable plaques was associated with incident ischemic stroke. This suggests that carotid disease was probably not a significant contributor to ischemic stroke in this population of patients with AF and, therefore, vulnerable carotid plaques might not improve stroke risk stratification in patients with AF.
Supplementary Material
Funding:
There was no targeted funding for this study. The Cardiovascular Health Study (CHS) was supported by contracts HHSN268201200036C, HHSN268200800007C, HHSN268201800001C, N01HC45133, N01HC55222, N01HC85079, N01HC85080, N01HC85081, N01HC85082, N01HC85083, N01HC85085, N01HC85086, 75N92021D00006, and grants U01HL080295 and U01HL130114 from the National Heart, Lung, and Blood Institute (NHLBI), with additional contribution from the National Institute of Neurological Disorders and Stroke (NINDS). Additional support was provided by R01AG023629 from the National Institute on Aging (NIA).
Financial disclosures:
Dr Noubiap is supported by a Postgraduate Scholarship from the University of Adelaide. Dr Thomas is supported by a Postdoctoral Fellowship from the University of Adelaide. Dr Kamtchum-Tatuene is supported by the Faculty of Medicine and Dentistry Motyl Graduate Studentship in Cardiac Sciences, an Alberta Innovates Graduate Student Scholarship, the Ballermann Translational Research Fellowship, the Izaak Walton Killam Memorial Scholarship, and the Andrew Stewart Memorial Graduate Prize. Dr Middeldorp is supported by a Postdoctoral Fellowship from the Cedar-Sinai Medical Centre. Dr Sanders is supported by a Practitioner Fellowships from the National Health and Medical Research Council of Australia and by the National Heart Foundation of Australia.
Footnotes
Availability of data and material:
This manuscript was prepared using CHS data obtained from the National Heart, Lung, and Blood Institute (NHLBI) following an application through its Biologic Specimen and Data Repository Information Coordinating Center (BioLINCC). The terms of the Research Materials Distribution Agreement explicitly prohibit the release or distribution of research material in any form to any third party unless required by NHLBI policies and approved by the ad hoc regulatory authorities.
Conflict of Interest Disclosures:
Dr Sanders reports having served on the advisory board of Medtronic, Abbott Medical, Boston Scientific, CathRx and PaceMate. Dr Sanders reports that the University of Adelaide has received on his behalf research funding, lecture and/or consulting fees from Medtronic, Abbott Medical, Boston Scientific and Microport. All other authors report no disclosures.
REFERENCES
- 1.Alkhouli M, Friedman PA. Ischemic Stroke Risk in Patients With Nonvalvular Atrial Fibrillation: JACC Review Topic of the Week. Journal of the American College of Cardiology 2019. Dec 17;74(24):3050–65. [DOI] [PubMed] [Google Scholar]
- 2.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. Feb;137(2):263–72. [DOI] [PubMed] [Google Scholar]
- 3.Noubiap JJ, Kamtchum-Tatuene J, Fitzgerald JL, Sanders P. Stroke risk associated with carotid and aortic atherosclerosis in patients with atrial fibrillation: A systematic review. J Neurol Sci 2021. Jun 15;425:117444. [DOI] [PubMed] [Google Scholar]
- 4.Noubiap JJ, Agbaedeng TA, Tochie JN, Nkeck JR, Ndoadoumgue AL, Fitzgerald JL, et al. Meta-Analysis Comparing the Frequency of Carotid Artery Stenosis in Patients With Atrial Fibrillation and Vice Versa. Am J Cardiol 2021. Jan 1;138:72–9. [DOI] [PubMed] [Google Scholar]
- 5.Chang YJ, Ryu SJ, Lin SK. Carotid artery stenosis in ischemic stroke patients with nonvalvular atrial fibrillation. Cerebrovascular diseases (Basel, Switzerland) 2002;13(1):16–20. [DOI] [PubMed] [Google Scholar]
- 6.Fox KAA, Lucas JE, Pieper KS, Bassand JP, Camm AJ, Fitzmaurice DA, et al. Improved risk stratification of patients with atrial fibrillation: an integrated GARFIELD-AF tool for the prediction of mortality, stroke and bleed in patients with and without anticoagulation. BMJ open 2017. Dec 21;7(12):e017157. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bekwelem W, Jensen PN, Norby FL, Soliman EZ, Agarwal SK, Lip GY, et al. Carotid Atherosclerosis and Stroke in Atrial Fibrillation: The Atherosclerosis Risk in Communities Study. Stroke 2016. Jun;47(6):1643–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Basili S, Loffredo L, Pastori D, Proieti M, Farcomeni A, Vesti AR, et al. Carotid plaque detection improves the predictve value of CHA2DS2-VASc score in patients with non-valvular atrial fibrilation: The ARAPACIS Study. International journal of cardiology 2017;231:143–9. English. [DOI] [PubMed] [Google Scholar]
- 9.Cho DH, Choi JI, Choi J, Kim YG, Oh SK, Kook H, et al. Impact of carotid atherosclerosis in CHA2DS2-VASc-based risk score on predicting ischemic stroke in patients with atrial fibrillation. The Korean journal of internal medicine 2020. Feb 27. [DOI] [PMC free article] [PubMed]
- 10.Kamtchum-Tatuene J, Noubiap JJ, Wilman AH, Saqqur M, Shuaib A, Jickling GC. Prevalence of High-Risk Plaques and Risk of Stroke in Patients With Asymptomatic Carotid Stenosis: A Meta-analysis. JAMA Neurol 2020. Aug 3;77(12):1–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.O’Leary DH, Polak JF, Wolfson SK Jr., Bond MG, Bommer W, Sheth S, et al. Use of sonography to evaluate carotid atherosclerosis in the elderly. The Cardiovascular Health Study. CHS Collaborative Research Group. Stroke 1991. Sep;22(9):1155–63. [DOI] [PubMed] [Google Scholar]
- 12.Polak JF, Shemanski L, O’Leary DH, Lefkowitz D, Price TR, Savage PJ, et al. Hypoechoic plaque at US of the carotid artery: an independent risk factor for incident stroke in adults aged 65 years or older. Cardiovascular Health Study. Radiology 1998. Sep;208(3):649–54. [DOI] [PubMed] [Google Scholar]
- 13.Longstreth WT Jr., Bernick C, Fitzpatrick A, Cushman M, Knepper L, Lima J, et al. Frequency and predictors of stroke death in 5,888 participants in the Cardiovascular Health Study. Neurology 2001. Feb 13;56(3):368–75. [DOI] [PubMed] [Google Scholar]
- 14.Bunch TJ, Bair TL, Crandall BG, Cutler MJ, Day JD, Graves KG, et al. Stroke and dementia risk in patients with and without atrial fibrillation and carotid arterial disease. Heart rhythm 2020. Jan;17(1):20–6. [DOI] [PubMed] [Google Scholar]
- 15.Lehtola H, Airaksinen KEJ, Hartikainen P, Hartikainen JEK, Palomaki A, Nuotio I, et al. Stroke recurrence in patients with atrial fibrillation: concomitant carotid artery stenosis doubles the risk. European journal of neurology 2017. May;24(5):719–25. [DOI] [PubMed] [Google Scholar]
- 16.Becattini C, Dentali F, Camporese G, Sembolini A, Rancan E, Tonello C, et al. Carotid atherosclerosis and risk for ischemic stroke in patients with atrial fibrillation on oral anticoagulant treatment. Atherosclerosis 2018. Apr;271:177–81. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.