1951 patients with ischemic stroke in the anterior circulation were identified from the PASS cohort. Out of these patients, 471 patients (24.1%), no diagnostic test was performed or reported to assess the presence of CAS. (Figure 1). In these patients AF was diagnosed more often than patients in which the diagnostic test for CAS was reported (26.7% versus 12.4%, p<0.001). In the remaining 1480 patients, 277 had CAS (18.7%, 95%CI 17.7-19.7%).
Patient characteristics
The patient characteristics are shown in Table 1. The mean age of patients with CAS was 72.5
|
ICA stenosis <50%
(n=1203)
|
ICA stenosis ≥ 50%
(n=277)
|
Age (mean with standard deviation)
|
70.70 (±12.5)
|
72.47 (±10.8)
|
Hypertension
|
52.99%
|
61.01%
|
Hypercholesterolemia
|
26.7%
|
32.4%
|
Smoking
|
60.1%
|
70.3%
|
Gender (male)
|
56.8%
|
69.0%
|
NIHSS (median, IQR)
|
4 (3-7)
|
6 (4-12)
|
Thrombolysis
|
36.7%
|
43.3%
|
Pre-stroke modified Rankin Scale (median, IQR)
|
0 (0-1)
|
0 (0-1)
|
Modified Rankin Scale after 90 days (median, IQR)
|
2 (1-3)
|
2 (1-4)
|
Chronic obstructive pulmonary disease
|
6.5%
|
11.9%
|
Atrial fibrillation
|
12.6%
|
12.0%
|
Myocardial infarction
|
12.0%
|
16.6%
|
Peripheral vascular disease
|
7.1%
|
11.6%
|
Table 1. Baseline characteristics.
(SD 10.8) and the median NIHSS was 6 (IQR 4-12). Most patients with a CAS suffered from hypertension (61%), had smoked prior to stroke (70%) and were male (69%). Patients with CAS received intravenous thrombolysis more frequent than patients without. Compared to the group without CAS, the median NIHSS is higher in the CAS group.
Logistic regression analyses
Univariate logistic regression analyses showed several associations with the known predictors of presence of CAS (Table 2).
|
p-value
|
Odds ratio (95% CI)
|
Age (year)
|
0.031
|
1.012 (1.001-1.023)
|
Hypertension
|
0.016
|
1.388 (1.063-1.812)
|
Hypercholesterolemia
|
0.060
|
1.312 (0.988-1.742)
|
Smoking
|
0.002
|
1.578 (1.182-2.107)
|
Gender (male)
|
<0.001
|
1.691 (1.279-2.235)
|
Table 2. Univariate logistic regression analyses with predictors of internal carotid artery (ICA) stenosis as independent variables and ICA stenosis ≥50% as the dependant variable.
Subsequently, a multivariable logistic regression was performed (Table 3).
|
p-value
|
Odds ratio (95% CI)
|
Age (year)
|
0.002
|
1.02 (1.007-1.033)
|
Hypertension
|
0.021
|
1.386 (1.05-1.83)
|
Smoking
|
<0.001
|
1.782 (1.305-2.433)
|
Gender (male)
|
<0.001
|
1.837 (1.38-2.447)
|
Table 3. Multivariate logistic regression analyses with associated predictors of CAS as independent variables and CAS ≥50% as the dependant variable. Backward selection was used with the univariate associated variables.
This analysis showed a best-fit model with age, hypertension, gender and smoking as variables for presence of CAS. The test for multi-collinearity indicated no multi-collinearity.
Survival and functional outcome 90 days after stroke
A median mRS after 90 days of 2 was found in both groups, with a different IQR (see Table 1). A multivariate ordinal regression analysis showed a significant shift towards a higher mRS in patients with CAS (Figure 2) (p <0.001). It showed an adjusted common odds ratio (OR) for CAS of 1.66 (95% CI 1.30-2.10), for male gender an OR of 0.76 (95% CI 0.63-0.91), for NIHSS of 1.18 (95% CI 1.16-1.21) and for thrombolysis an OR of 0.76 (95% CI 0.62-0.92). For pre-stroke mRS, only a significant shift was found for either mRS 0 and mRS 1, with OR for pre-stroke mRS of 0 of 0.29 (95% CI 0.16-0.52) and for pre-stroke mRS of 1 of 0.44 (95% CI 0.23-0.83). Age showed no significant relation with mRS (p=0.078).
A significant difference was seen for mRS 0-2 versus >2 between patients with a carotid stenosis compared to patients without, with a larger proportion of patients with a good functional outcome in the group without stenosis (72.0% versus 52.0%, p<0.001). Mortality was significantly increased in patient with carotid stenosis with 12.6% in the stenosis group versus 3.5% in the group without stenosis (p<0.001).