Distribution of ischemic strokes in patients with atrial fibrillation

Abstract

Background

We aimed to determine the relative frequency of affected cerebrovascular territories in patients with atrial fibrillation (AF) suffering an ischemic stroke.

Methods

Altogether, 1,976 patients who suffered their first-ever ischemic stroke during 2003–2012 and were diagnosed with AF either before or within 30 days after the event were included in this retrospective multicenter cohort study. Strokes were classified radiographically to be located either within the anterior or the posterior cerebrovascular territory, and the effect of the CHA₂DS₂-VASc score, oral anticoagulant (OAC) use, and timing of AF diagnosis on lesion localization was determined.

Results

The median age of the patients was 78.4 (interquartile range: 71.7–84.2) years, 1,137 (57.5%) of them were women, their mean CHA₂DS₂-VASc score was 3.5 (95% confidence interval: 3.4–3.5), 656 (33.2%) were receiving OAC drugs, and altogether, 1,450 (73%) had a previous AF diagnosis. The localization of ischemic lesions between the anterior and the posterior cerebrovascular territories was not affected by the timing of AF diagnosis (p = 0.46), use of OACs (p = 0.70), or the CHA₂DS₂-VASc score (p = 0.10). Within the anterior territory, altogether 774 strokes (53.2%) were located in the left hemisphere and 3 (0.2%) were bilateral. The timing of AF diagnosis (p = 0.84), use of OACs (p = 0.90), or the CHA₂DS₂-VASc score (p = 0.21) did not affect the location of the ischemic lesion between the hemispheres.

Conclusions

The timing of AF diagnosis, use of OAC drugs, or the CHA₂DS₂-VASc score did not affect the distribution of ischemic strokes. Anterior territory strokes were slightly more often located within the left hemisphere.

Atrial fibrillation (AF) is an exceedingly common arrhythmia and an important causative factor of ischemic stroke that affects 25%–38% of all patients suffering an ischemic stroke.^1–3 The localization of ischemic lesions in all stroke patients has been described in several stroke registries during the 1980s–2000s.^4–7 However, the cerebrovascular distribution of strokes specifically in patients with AF has not been reported previously. The aim of the current study was to determine the relative frequency of affected cerebrovascular territories in a large multicenter series of consecutive patients with AF suffering their first-ever ischemic stroke.

Methods

In the multicenter FibStroke Study (ClinicalTrials.gov Identifier: NCT02146040), all patients older than 18 years, who had received a diagnosis of AF at any time and who had suffered an ischemic stroke, a TIA, or an intracranial bleed during 2003–2012, were retrospectively identified from the institutional discharge registries of 2 university hospitals (Turku University Hospital, Turku; Kuopio University Hospital, Kuopio) and 2 central hospitals (Central Finland Central Hospital, Jyväskylä [years 2006–2012]; Satakunta Central Hospital, Pori) in Finland.^8–10 All lacunar strokes were excluded from the study. For the current analysis, we chose patients who were treated for their first-ever ischemic stroke during the study period and were diagnosed with AF either before or within 30 days after the event. A flow diagram depicting the selection process of patients is presented in the figure.

Figure. Flow diagram of the selection process of study patients.

The clinical notes of each patient were reviewed according to a standardized data collection protocol. Information was recorded on patient characteristics, risk factors of stroke, use of anticoagulant medications, and cardiac rhythm at the time of the ischemic event. In addition, the presumed etiology of stroke was documented according to the 10th revision of the International Statistical Classification of Diseases and Related Health Problems diagnosis code at the time of discharge.

An ischemic stroke was defined as a permanent focal neurologic deficit adjudicated by a neurologist and confirmed by CT or MRI. The localization and cerebrovascular territory affected by the ischemic lesions was assessed by a radiologist in each case. Strokes were classified to be located either within the anterior or the posterior cerebrovascular territory. Strokes within the anterior cerebrovascular territory were classified as affecting either the middle cerebral artery (MCA) territory or the anterior cerebral artery (ACA) territory, whereas strokes within the posterior cerebrovascular territory were classified as affecting the posterior cerebral artery (PCA) territory, brain stem, or the cerebellum. It was also recorded in which hemisphere the lesions were located and whether multiple territories were affected. The localization of watershed lesions was documented as unknown, and they were thus excluded from the analyses.

Statistical analyses were performed with SPSS software (version 22.0; SPSS, Inc., Chicago, IL). Continuous data are presented as mean (95% confidence interval [CI]) or median [interquartile range] and categorical variables as absolute number and percentage. The χ² test and the Fisher exact test were used to compare differences between proportions. The Kruskal-Wallis test was used for analysis of continuous variables. The binominal test was used to analyze whether there was propensity for the ischemic lesions to be located in either cerebral hemisphere.

Standard protocol approvals, registrations, and patient consents

The FibStroke Study protocol (ClinicalTrials.gov Identifier: NCT02146040) was approved by the Medical Ethics Committee of the Hospital District of Southwest Finland and the ethics committee of the National Institute for Health and Welfare. Informed consent was not required because of the retrospective registry nature of the study. The study conforms to the Declaration of Helsinki.

Data availability

Data are available from the corresponding author on request.

Results

Among the 1,976 patients included in the study, the median age was 78.4 (71.7–84.2) years, 1,137 (57.5%) of them were women, their mean CHA₂DS₂-VASc (Congestive heart failure, Hypertension, Age ≥75, Diabetes mellitus, prior Stroke or transient ischemic attack or thromboembolism, Vascular disease, Age 65-74 years, Sex category [i.e., female sex]) score was 3.5 (95% CI: 3.4–3.5), and 656 (33.2%) were receiving oral anticoagulant (OAC) drugs before the onset of stroke (653 patients were receiving warfarin and 3 patients dabigatran). Among the patients receiving warfarin, 295 (45.2%) had an international normalized ratio of at least 2.0 at the time of stroke. The clinical characteristics of the patients are found in table 1.

Table 1.

Clinical characteristics of the patients

The distribution of ischemic lesions is presented in table 2 for all patients and according to the timing of AF diagnosis, the use of OACs, and the presumed etiology of stroke. The localization of the ischemic lesions between the anterior and the posterior cerebrovascular territories was not affected by the timing of AF diagnosis (p = 0.456), the use of OACs (p = 0.696), OAC therapy within treatment range (p = 0.761), or the CHA₂DS₂-VASc score (p = 0.101), whereas embolic and thrombotic strokes were more often located within the anterior territory compared with ischemic strokes of other or undetermined etiology (p < 0.001). Within the anterior territory, the timing of AF diagnosis (p = 0.400), the CHA₂DS₂-VASc score (p = 0.225), or the etiology of stroke (p = 0.075) did not affect the lesion distribution, whereas differences were yielded by the use of OACs and OAC therapy within treatment range (p = 0.006 and p = 0.044, respectively). Within the posterior territory, the timing of AF diagnosis (p = 0.177), the CHA₂DS₂-VASc score (p = 0.468), or OAC therapy within treatment range (p = 0.154) had no effect on the lesion distribution, whereas statistical differences were observed according to the use of OACs and etiology of stroke (p = 0.026 and p < 0.001).

Table 2.

Distribution of ischemic stroke lesions in all patients with AF and according to the timing of AF diagnosis and OAC use

Within the anterior territory, altogether 774 strokes (53.2%) were located in the left hemisphere and 3 (0.2%) were bilateral. After excluding the bilateral lesions, a small propensity for the ischemic lesions to be located within the left hemisphere was observed (p = 0.014). The timing of AF diagnosis (p = 0.836), the use of OACs (p = 0.902), OAC therapy within treatment range (p = 0.525), etiology of stroke (p = 0.104), or the CHA₂DS₂-VASc score (p = 0.210) had no effect on the localization of the ischemic lesions between the hemispheres within the anterior circulation.

Discussion

In the current study, the distribution of ischemic lesions was described in a large series of patients with AF suffering their first ischemic stroke. The timing of AF diagnosis or the CHA₂DS₂-VASc score did not affect the lesion localization. Although some differences in lesion localization were observed according to OAC use, the distinctions in absolute terms were small and do not seem meaningful. Strokes classified as embolic or thrombotic were more often located within the anterior cerebrovascular territory in comparison to strokes of other or undetermined etiology. Anterior territory strokes were slightly more often located within the left hemisphere, but the observed difference was so small that its clinical significance is questionable.

The Lausanne Stroke Registry, published 30 years ago, was the first study wherein the affected cerebrovascular territories were determined in detail among all patients with ischemic stroke irrespective of stroke etiology.⁶ Localization within the anterior circulation was reported in 70% of patients (MCA 67% and ACA 2%) and within the posterior circulation in 27% of patients (brainstem 13%, cerebellum 2%, PCA 10%, and multiple locations 2%), whereas 3% of strokes affected multiple territories.⁶ Similar distributions of ischemic lesions among the general ischemic stroke population have also been recounted in several later stroke registries corroborating the results of the Lausanne registry.^4,5,7 It may be observed that these figures are in a remarkably close agreement with those reported here among patients with AF.

Although the distribution of ischemic strokes specifically in patients with AF has not been described in previous literature, the prevalence of AF as a risk factor in ischemic strokes of the different vascular territories has been reported in 1 previous study, the Besançon Stroke Registry.⁴ After excluding watershed and lacunar infarcts from the earlier work, it can retrospectively be calculated that among the 395 comparable patients with AF included in the registry, 72% of ischemic lesions were located within the anterior circulation, 14% within the posterior circulation, and 13% within multiple territories.⁴

Most patients with AF possess risk factors for strokes of other etiologies. Consequently, not all strokes suffered by patients with AF are cardioembolic in origin. In a previous work, it was reported that the share of embolic strokes in patients with AF was slightly below 70%, a figure in line with the share of presumed embolic strokes included in the current study.¹¹ The distribution of ischemic lesions in patients with presumed cardioembolic strokes has been described in several previous works, although differences in reporting renders comparing the results of these studies to those of the current one difficult.^4–7 However, the lesion localization observed here among the patients with presumed embolic stroke etiology was largely congruent with the distribution among all patients with AF.

The left common carotid artery branches directly off the aortic arch and is in line with the blood flow in the ascending aorta, whereas the right common artery generally is situated at an angle to the flow of the brachiocephalic trunk. Subsequently, it may be speculated that cardiogenic emboli could more easily be siphoned toward the left hemisphere. However, conflicting results regarding the predilection of ischemic strokes to be located in either hemisphere have been reported in previous literature. A right-sided propensity of cardiogenic emboli has been reported in 2 previous studies,^12,13 whereas a left-sided predominance of embolic or atherosclerotic strokes has been observed in others.^14–17 Although there was a small left-sided propensity in the current study, no difference between embolic and thrombotic strokes was observed.

Reflecting the results presented here on the previous literature, it seems reasonable to infer that among patients with AF, there is no difference in the distribution of ischemic lesions, whether cardioembolic or not in origin, as compared to the general stroke population. However, some limitations pertaining to the current study must also be acknowledged. A notable limitation is that a third of all strokes included in the study cohort were excluded because of missing data on lesion localization. Regrettably, the reason for missing data was not recorded, but the most likely explanation is that there was no identifiable lesion on the initially conducted imaging study. The exclusion of watershed and lacunar strokes may also be seen as a limitation. A notable limitation is that no central analysis of source imaging, which would have allowed the most precise and uniform classification of stroke localization, was performed, but rather site of stroke was determined from the radiology reports included in the patients' clinical notes. Moreover, it was not recorded, which imaging modalities were used in each case, which is a limitation. Fetal-type PCA is a common variant of vascular anatomy, which entails that the distal PCA territory is perfused by a branch of the internal carotid artery and is therefore part of the anterior circulation. Unfortunately, individual vascular anatomy could not be accounted for in the current study, as many patients did not have angiographic imaging studies performed on them, and additionally, this was not consistently declared in the radiology reports. One important limitation is also that data on the status of carotid arteries were available for too few patients to be included in the current analysis.

In conclusion, the timing of AF diagnosis, use of OACs, or the CHA₂DS₂-VASc score did not affect the distribution of lesions between the cerebrovascular territories. Anterior territory strokes were slightly more often located within the left hemisphere.

Appendix. Authors

Study funding

This study was funded by the Finnish Foundation for Cardiovascular Research (Helsinki, Finland) and the Clinical Research Fund (EVO) of Turku University Hospital (Turku, Finland).

Disclosure

J. Jaakkola and P. Hartikainen report no disclosures. T.O. Kiviniemi reports grants from the Finnish Medical Foundation, the Finnish Foundation for Cardiovascular Research, the State Research Fund, BMS-Pfizer, and Atricure Ltd and lecture fees from AstraZeneca, Bayer, Boehringer Ingelheim, BMS-Pfizer, MSD, and Orion Pharma. I. Nuotio and A. Palomäki report grants from the Finnish Foundation for Cardiovascular Research and the Finnish Medical Foundation and lecture fees from Bayer, MSD, Bristol-Meyers Squibb, Pfizer, and Sanofi. J.E.K. Hartikainen and A. Ylitalo report no disclosures. P. Mustonen reports lecture fees from Boehringer Ingelheim, Bayer, Bristol-Myers Squibb, Novartis, Pfizer, Sanofi-Aventis, and Leo Pharma. K.E.J. Airaksinen reports grants from the Finnish Foundation for Cardiovascular Research and lecture fees from Bayer, AstraZeneca, and BMS. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/cp.

TAKE-HOME POINTS

• → Previous data on the localization of ischemic strokes in patients with AF are scarce.

• → The relative frequency of affected cerebrovascular territories was reported in a large multicenter series of 1,976 consecutive patients with AF suffering their first-ever ischemic stroke.

• → Altogether, 72.0% of all strokes were located within the anterior cerebrovascular territory, 25.3% within the posterior territory, and 2.7% within multiple territories.

• → The timing of AF diagnosis, use of OAC drugs, or the CHA₂DS₂-VASc score did not affect stroke location between the cerebrovascular territories.

• → Anterior territory strokes were slightly more often located within the left hemisphere.