2025 年 32 巻 9 号 p. 1164-1175
Aim: To evaluate the risk factors, location, treatment, and outcomes of stroke due to arterial dissection, we examined these characteristics in a substantial, long-standing, nationwide stroke cohort.
Methods: The study participants were patients with acute stroke who were registered in the Japan Stroke Data Bank between January 1999 and December 2020. We focused on patients with stroke caused by extracranial or intracranial artery dissection and examined their clinical characteristics, treatments, and outcomes. In addition, we compared the results between clinical subtypes with and without dissection.
Results: Among the 218,799 registered patients with acute stroke, 1,353 (0.62%) were attributed to artery dissection. Of these, 880 patients had ischemic stroke, 16 had intracerebral hemorrhage, and 457 had subarachnoid hemorrhage (SAH). Dissection cases were most prevalent among individuals in their 40s and 50s, with intracranial vertebral artery dissection being the primary cause of ischemic stroke and SAH. Male sex, dyslipidemia, diabetes mellitus, and a history of smoking were associated with a higher likelihood of ischemic stroke than SAH. Unfavorable outcomes, defined as a modified Rankin score ≥ 4 at discharge, were observed in 18.9% of ischemic stroke cases and 42.6% of SAH cases with dissection. Neurological severity and older age at admission are associated with unfavorable outcomes in patients with ischemic stroke and SAH.
Conclusions: Ischemic stroke was the most frequent subtype of stroke in patients with arterial dissection, followed by SAH. Patients with stroke due to dissection were younger than those without. Neurological severity and older age at admission are substantial risk factors for unfavorable stroke outcomes due to artery dissection.
Intracranial and extracranial artery dissections are associated with nonspecific symptoms, such as dizziness, vertigo, headache, and neck pain; however, they can lead to stroke, including ischemic stroke, intracerebral hemorrhage, and subarachnoid hemorrhage (SAH). Studies discussing the characteristics of arterial dissection and its association with stroke have been conducted1-3); however, owing to its rarity, the number of patients involved in these studies was typically small3). Most studies with ＞100 patients have been reported in East Asia, often focusing on patients with carotid or vertebrobasilar artery dissection or those presenting with SAH3). Therefore, a comprehensive understanding of stroke caused by arterial dissection has not been sufficiently obtained.
AimTo determine the distribution, clinical characteristics, dissection location, risk factors, and outcomes of ischemic and hemorrhagic stroke caused by artery dissection in a large cohort, we examined patients from the Japan Stroke Data Bank (JSDB)4). Furthermore, we compared results among stroke subtypes and between stroke subtypes caused by artery dissection and those that were not.
The JSDB is an ongoing hospital-based, multicenter, prospective registry of patients with acute stroke in Japan. Detailed information on JSDB has been described previously4, 5). The JSDB collects clinical diagnoses of acute stroke based on evaluations by stroke specialists from academic hospitals and regional stroke centers. Patients with acute stroke, including ischemic stroke, transient ischemic attacks, intracranial hemorrhage, and SAH, who were transferred to a hospital within seven days of onset were included. A total of 218,799 patients from 133 academic or stroke centers were enrolled in the database between January 1999 and December 2020. From these, we selected patients whose strokes were caused by arterial dissections.
Standard Protocol Approvals, Registrations, and Patient ConsentThe study protocol was approved by the Institutional Ethics Board. Due to the anonymous nature of the data, individual consent for entry into the database was waived by the institution. Instead, the opt-out consent method was used.
Data CollectionAcute stroke admission data were recorded using a standardized form in the computerized registry database (registration-URL: https://strokedatabank.ncvc.go.jp) by study physicians or clinical report coordinators at each institute. The database includes clinical information on demographics, stroke subtypes, onset style of neurological symptoms, disease severity (determined using the National Institutes of Health Stroke Scale [NIHSS]), treatment, vascular risk factors (hypertension, dyslipidemia, diabetes mellitus), arteries involved, and functional outcomes at discharge (measured using the modified Rankin Scale [mRS]). Unfavorable functional and vital outcomes were defined as an mRS score of 4–6. Patients with subarachnoid hemorrhage had their Glasgow Coma Scale (GCS)6) and World Federation of Neurosurgical Societies (WFNS)7) scores evaluated upon admission. Heavy alcohol drinkers were defined as those who consumed ＞160 g of ethanol per week.
Statistical AnalysesContinuous data were presented as the mean (standard deviation [SD]) or median (interquartile range), and categorical data were presented as numbers (%). The Mann–Whitney U test or Wilcoxon rank-sum test for continuous variables and chi-square test for categorical variables were used to test the significance of differences between the two groups. Univariate and multivariate mixed-effect logistic regression, with institutes as random intercepts, was performed to calculate the odds ratio (OR) and 95% confidence interval (CI) to assess the association of these factors with unfavorable prognosis. p＜0.05. All analyses were performed using the STATA software program, version 16 (StataCorp LP, College Station, TX, USA).
Among the 218,799 patients registered on the JSDB between January 1999 and December 2020, 165,585 (99,758 men and 65,827 women) had ischemic stroke, 40,280 (23,130 men and 17,150 women) had intracerebral hemorrhage, and 12,936 (4,240 men and 8,696 women) had SAH. Of these, 1,353 underwent arterial dissection, among whom 880 (644 men and 236 women) had ischemic stroke, 16 (11 men and 5 women) had intracerebral hemorrhage, and 457 (252 men and 205 women) had SAH. The ratios of each stroke subtype in the patients with and without arterial dissection are shown in Fig.1. Ischemic stroke due to arterial dissection accounted for 0.53% of all ischemic stroke cases, intracerebral hemorrhage due to dissection accounted for 0.04% of all intracerebral hemorrhage cases, and SAH due to dissection accounted for 3.53% of all SAH cases.
Ratio of each stroke subtype among patients with arterial dissection and the corresponding stroke subtype without dissection
Owing to the small number of patients with intracerebral hemorrhage due to arterial dissection, we omitted patients with this subtype from the subsequent analyses.
Sex DifferencesIn patients with arterial dissection, the male/total ratio was 67.0% compared with 58.0% in those with stroke without dissection (p＜0.001). The male ratio in patients with arterial dissection was significantly higher than that in those without dissection: 73.2% vs. 60.2% (p＜0.001) for ischemic stroke and 55.1% vs. 32.0% (p＜0.001) for SAH (Table 1).
| Ischemic stroke | SAH | |||
|---|---|---|---|---|
| With dissection (n = 880) | Without dissection (n = 164705) | With dissection (n = 457) | Without dissection (n = 12479) | |
| Mean ages ±SD | 54.4±14.7‡ | 73.3±12.4 | 54.9±12.9＊ | 64.3±14.9 |
| Men (%) | 73.2 (n = 644)＊,＊＊ | 60.2 (n = 99114) | 55.1 (n = 252)＊ | 32.0 (n = 3988) |
| NIHSS (median [IQR]) | 2 [1, 5] (n = 818)＊,＊＊ | 4 [1, 9] (n = 140828) | 8 [0, 38] (n = 277)＊ | 3 [0, 34] (n = 6430) |
| GCS (median [IQR]) | 15 [15,15] (n = 415)＊,＊＊ | 15 [14,15] (n = 99389) | 13 [4, 15] (n = 373)＊ | 14 [6, 15] (n = 9266) |
| WFNS (median [IQR]) | 3 [2, 5] (n = 450)＊ | 2 [1, 4] (n = 11810) | ||
| Complications | involved patients/total reports (%) | involved patients/total reports (%) | ||
| Heart disease | 79/873 (9.0)＊ | 53991/161538 (33.4) | 33/437 (7.6) | 1013/12107 (8.4) |
| Hypertension | 416/860 (48.4)＊ | 111886/159126 (70.3) | 195/418 (46.7) | 5948/11613 (51.2) |
| Dyslipidemia | 197/823 (23.9)＊,＊＊ | 52416/146938 (35.7) | 66/382 (17.3) | 1768/10012 (17.7) |
| Diabetes mellitus | 99/869 (11.4)＊,＊＊ | 43258/158078 (27.4) | 30/423 (7.1) | 938/11707 (8.0) |
| CKD | 17/723 (2.4)＊ | 9389/113906 (8.2) | 6/318 (1.9) | 270/8170 (3.3) |
| Liver disfunction | 10/873 (1.1) | 1530/161538 (0.9) | 2/437 (0.5) | 118/12107 (1.0) |
| Migraine | 17/40 (40.3)＊ | 108/8986 (1.2) | 2/10 (20.0) | 23/266 (8.6) |
| Heavy drinker | 75/527 (14.2)＊ | 10064/102876 (9.8) | 34/282 (12.1) | 731/7160 (10.2) |
| Smoker | 275 /567 (48.5)＊,＊＊ | 43968/108474 (40.5) | 130 /295 (44.1)＊ | 2832/7579 (37.4) |
In patients with stroke due to arterial dissection, the male/total ratio was 67.0% compared to 58.0% in those with stroke without dissection (p＜0.001).
＊p＜0.05 when compared with the patients with the corresponding stroke subtype without dissection
＊＊p＜0.05 when compared with the patients with SAH due to arterial dissection
SAH, Subarachnoid hemorrhage; NIHSS, National Institutes of Health Stroke Scale; GCS, Glasgow Coma Scale; WFNS, World Federation of Neurosurgical Soc; CKD, chronic kidney disease
The median NIHSS score for patients with ischemic stroke due to arterial dissection was 2, whereas that for patients without dissection was 4 (p＜0.001; Table 1). The median NIHSS score for patients with SAH was 8 and 3 for those with and without arterial dissection, respectively (p = 0.002; Table 1).
Comparison of Age DistributionThe mean age±SD of patients with total arterial dissection was 54.7±14.2 years (men: 53.4±13.3; women: 57.4±15.5, p＜0.001), 54.4±14.7 for patients with ischemic stroke due to arterial dissection (men: 53.5±13.9; women: 57.0±16.4, p = 0.02), and 54.9±12.9 for patients with SAH due to arterial dissection (men: 52.7±11.2; women: 57.7±14.4, p＜0.001) (Table 1).
The age distribution of ischemic stroke and SAH due to arterial dissection is shown in Fig.2, along with the corresponding stroke subtypes without dissection for comparison. The frequency of ischemic stroke and SAH due to arterial dissection peaked in patients in their 40s and 50s, respectively (Fig.2). In contrast, the frequency of ischemic stroke without dissection increased with age and was most frequent in patients ≥ 80 years old (Fig.2, top). Similarly, the frequency of SAH without dissection peaked in the 70s (Fig.2, bottom). When comparing the age distribution of each stroke subtype between patients with and without arterial dissection, those with arterial dissection were significantly younger (p＜0.001) for both stroke subtypes.
Age distribution of patients with ischemic stroke due to arterial dissection and that of patients without dissection (top) and SAH due to arterial dissection and those without dissection (bottom) are shown. The age distribution was significantly different between the two groups (p＜0.05). The numbers on each bar represent the number of eligible patients. SAH, subarachnoid hemorrhage.
The frequencies of ischemic stroke and SAH due to arterial dissection among all patients with ischemic stroke and SAH were 0.5% and 5.0%, respectively. The frequencies of ischemic stroke and SAH due to arterial dissection among all patients with ischemic stroke and SAH were determined by the age group (Fig.3). The frequency of ischemic stroke due to arterial dissection was between 3.4% and 7.1% in the age groups of teens, 20s, 30s, and 40s, and＜1% in those ≥ 60 years old. The frequency of SAH due to arterial dissection was between 4.3% and 7.5% between 10 and 59 years old in our stroke cohort.
Number on each bar represents the number of eligible patients. SAH, subarachnoid hemorrhage.
Among younger patients 10–49 years old, ischemic stroke due to arterial dissection accounted for 5.0% of all patients with ischemic stroke, whereas SAH due to arterial dissection accounted for 7.1% of all patients with SAH.
Location of DissectionThe location of dissection was reported in only 24% of patients with arterial dissection. The dissected arteries of the patients with ischemic stroke or SAH are shown in Supplementary Table 1. Intracranial vertebral artery dissection was the primary cause, accounting for 44.1% of ischemic strokes and 46.8% of SAH cases among all arterial dissections that caused stroke. Dissection of the anterior cerebral and extracranial vertebral arteries is the second most common cause of ischemic stroke. Extracranial carotid artery dissection was found in 7.8% of patients with ischemic stroke due to arterial dissection. The anterior cerebral artery is the second-most frequently involved artery in patients with SAH due to arterial dissection. No marked difference was observed in the frequency of ischemic stroke or SAH between the intracranial carotid and vertebral arteries.
| Dissected artery | Ischemic stroke (n = 220) | SAH (n = 109) | P | ||
|---|---|---|---|---|---|
| Number of patients | % | Number of Patients | % | ||
| Intracranial artery | 177 | 80.5 | 109 | 100 | ＜0.01 |
| Intracranial vertebral artery | 97 | 44.1 | 51 | 46.8 | ns |
| Anterior cerebral artery | 21 | 9.5 | 17 | 15.6 | ns |
| Middle cerebral artery | 16 | 7.3 | 15 | 13.8 | ns |
| Basilar artery | 15 | 6.8 | 0 | 0 | ＜0.05 |
| Posterior cerebral artery | 14 | 6.4 | 15 | 13.8 | |
| Intracranial carotid artery | 14 | 6.4 | 11 | 10.1 | ＜0.05 |
| Extracranial artery | 43 | 19.5 | 0 | 0 | ＜0.01 |
| Extracranial vertebral artery | 21 | 9.5 | 0 | 0 | 0 |
| Extracranial carotid artery | 16 | 7.3 | 0 | 0 | ＜0.05 |
| Common carotid artery | 4 | 1.8 | 0 | 0 | ns |
| Brachial artery | 1 | 0.5 | 0 | 0 | ns |
| Subclavian artery | 1 | 0.5 | 0 | 0 | ns |
SAH, subarachnoid hemorrhage
In patients with ischemic stroke, 177 (80.5%) and 43 (19.5%) arteries were intracranial and extracranial, respectively (Supplementary Table 1). In addition, 71 (32.3%) and 148 (67.3%) arteries were involved in the anterior and posterior circulation, respectively. In patients with SAH, 43 (39.4%) and 66 (60.6%) arteries were involved in anterior and posterior circulation, respectively.
Vascular Risk FactorsThe prevalence of vascular risk factors in patients with arterial dissection presenting with ischemic stroke or SAH is shown in Table 1. Hypertension was observed in approximately 50% of the patients with ischemic stroke and SAH. Diabetes and dyslipidemia were more prevalent in patients with ischemic stroke than in those with SAH.
Vascular risk factors for stroke such as hypertension, heart disease, dyslipidemia, diabetes mellitus, and chronic kidney disease are frequently reported in patients with ischemic stroke. However, the frequency was much lower than that in those without dissection (Table 1). In contrast, migraine, heavy drinking, and a history of smoking were more frequent in patients with ischemic stroke due to arterial dissection than in those without dissection. However, in patients with SAH, no significant difference was observed in the frequency of these risk factors between those with and without dissection, with the exception of smoking (Table 1).
When comparing vascular risk factors between patients with dissection presenting with ischemic stroke and those with SAH, patients with ischemic stroke exhibited significantly higher ratios of males, dyslipidemia, diabetes mellitus, and a history of smoking (Table 1).
Initial SymptomsThe initial symptoms of the patients presenting with ischemic stroke or SAH due to arterial dissection are shown in Supplementary Table 2. Patients with ischemic stroke frequently present with limb paralysis, vertigo, dysarthria, and headaches. More than 60% of patients who presented with SAH exhibited disturbed consciousness and/or headache.
| Ischemic stroke | SAH | |||
|---|---|---|---|---|
| With dissection | Without dissection | With dissection | Without dissection | |
| Number of reported patients | 753 (%) | 123800 (%) | 357 (%) | 9999 (%) |
| Disturbance of consciousness | 94 (12.5)＊,＊＊ | 23406 (18.9) | 216 (60.5)＊ | 5162 (51.6) |
| Paralysis | 247 (32.8)＊,＊＊ | 80714 (65.2) | 16 (4.5)＊ | 974 (9.7) |
| Dysarthria | 208 (27.6)＊,＊＊ | 54244 (43.8) | 8 (2.2) | 324 (3.2) |
| Facial palsy | 36 (4.8)＊＊ | 5333 (4.3) | 2 (0.6) | 27 (0.3) |
| Sensory disturbance | 148 (19.7)＊,＊＊ | 13048 (10.5) | 7 (2.0) | 86 (0.9) |
| Headache | 201 (26.9)＊,＊＊ | 3994 (3.2) | 237 (66.4)＊ | 5940 (59.4) |
| Vertigo | 224 (29.7)＊,＊＊ | 6850 (5.5) | 7 (2.0) | 236 (2.4) |
| Visual disturbance | 37 (4.9)＊＊ | 6472 (5.2) | 0 | 49 (0.5) |
| Neck pain | 40 (5.3)＊,＊＊ | 142 (0.1) | 5 (1.4) | 191 (1.8) |
| Dysphagia | 61 (8.1)＊,＊＊ | 3400 (2.7) | 0 | 31 (0.3) |
| Nausea, vomiting | 116 (15.4)＊,＊＊ | 4072 (3.3) | 87 (24.4) | 2692 (26.9) |
| Hearing disturbance | 3 (0.4)＊ | 102 (0.1) | 0 | 6 (0.1) |
| Gait disturbance | 87 (10.2)＊,＊＊ | 9965 (8.0) | 2 (0.6) | 1122 (1.2) |
＊p＜0.05 when compared with the patients with the corresponding stroke subtype without dissection
＊＊p＜0.05 when compared with the patients with SAH due to arterial dissection
SAH, subarachnoid hemorrhage
The frequency of initial symptoms was compared between the stroke subtypes with and without arterial dissection (Supplementary Table 2).
TreatmentsThe treatments administered to patients with acute ischemic stroke due to arterial dissection upon admission are shown in Supplementary Table 3. Most patients (98%) with SAH due to arterial dissection received nonspecific conservative treatment.
| Treatment | Ischemic stroke (n = 880) | SAH (n = 457) |
|---|---|---|
| Number of patients (%) | Number of patients (%) | |
| Tissue plasminogen activator | 18 (2.0) | 1 (0.2) |
| Warfarin | 8 (0.9) | 1 (0.2) |
| Antithrombin | 165 (18.8) | 2 (0.4) |
| Edarabone | 227 (25.8) | 0 |
| Low molecular weight dextran | 30 (3.4) | 1 (0.2) |
| Hypertonus liquid | 41 (4.7) | 1 (0.2) |
| Thrombectomy | 1 (0.2) | 0 |
| Surgery | 6 (0.7) | 1 (0.2) |
| Aspirin | 66 (7.5) | 0 |
| Cilostazol | 20 (2.3) | 0 |
| Clopidogrel | 19 (2.2) | 0 |
| Ozagrel | 47 (5.3) | 1 (0.2) |
| Supportive treatment only | 358 (40.7) | 449 (98.2) |
The most commonly selected treatments at discharge for patients with ischemic stroke due to arterial dissection included antiplatelet therapy, antihypertensive drugs, and statins (Supplementary Table 4). For patients with SAH due to arterial dissection, the primary treatments at discharge were antihypertensive and antiplatelet drugs (Supplementary Table 4).
| Treatment | Ischemic stroke (n = 604) | SAH (n = 171) | p |
|---|---|---|---|
| Number of patients (%) | Number of patients (%) | ||
| Antiplatelets | 442 (73.2) | 65 (38.0) | ＜0.01 |
| Anticoagulant | 60 (9.9) | 5 (2.9) | ＜0.01 |
| Blood pressure medicines | 232 (38.4) | 83 (48.5) | ＜0.05 |
| Statins | 188 (30.1) | 30 (17.5) | ＜0.01 |
| Diabetes medicines | 32 (5.3) | 4 (2.3) | ns |
| No medication | 37 (6.1) | 33 (19.3) | ＜0.01 |
SAH, subarachnoid hemorrhage
In patients with ischemic stroke due to arterial dissection, 56.8% showed a favorable outcome (mRS ≤ 1), 18.9% had an unfavorable outcome (mRS ≥ 4), and the mortality rate was 3.7% at discharge. In contrast, 43.0%, 31.1%, and 4.3% of patients with stroke without dissection resulted in favorable outcomes, unfavorable outcomes, and death, respectively (p＜0.05 for favorable and unfavorable outcomes compared to those in patients with dissection) (Fig.4). Among patients with SAH due to arterial dissection, 38.2% showed favorable outcomes, 42.6% showed unfavorable outcomes, and 20.0% died at discharge (Fig.4). The proportion of patients with SAH without dissection presenting with favorable outcomes, unfavorable outcomes, and death was 40.7%, 41.5%, and 20.9%, respectively (no statistical significance for each compared to those for patients with dissection). The total mortality rate of stroke due to arterial dissection is 9.2%.
The ratio of each mRS score was significantly different between the patients with ischemic stroke with and without dissection (p＜0.05). No difference was observed between patients with SAH irrespective of dissection. mRS, modified Rankin Scale; SAH, subarachnoid hemorrhage.
The univariate model of the multilevel mixed-effect regression analysis detected several risk factors for an mRS ≥ 4 at discharge in patients with ischemic stroke and SAH due to arterial dissection (Table 2). Neurological severity upon admission was scored using the NIHSS and GCS, and older age was a risk factor for unfavorable outcomes for both ischemic stroke and SAH (p＜0.05). Disease severity, as scored by the WFNS, was also a risk factor for SAH (p＜0.05). Hypertension and diabetes mellitus were risk factors for unfavorable outcomes in patients with ischemic stroke (p＜0.05) but not in patients with SAH.
| Risk factor | Ischemic stroke (n = 880) | SAH (n = 457) | ||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p | OR | 95%CI | p | |
| Total patients | 1 | 1 | ||||
| NIHSS score on admission＊ | 1.19 | 1.15–1.23 | ＜0.001 | 1.05 | 1.04–1.07 | ＜0.001 |
| GCS score on admission＊ | 0.31 | 0.21–0.45 | ＜0.001 | 0.83 | 0.80–0.87 | ＜0.001 |
| WFNS score on admission＊ | 1.96 | 1.70–2.26 | ＜0.001 | |||
| Hypertension | 1.62 | 1.14–2.29 | 0.006 | 1.26 | 0.84–1.87 | 0.266 |
| Dyslipidemia | 1.07 | 0.72–1.61 | 0.73 | 1.07 | 0.61–1.85 | 0.822 |
| Diabetes mellitus | 1.66 | 1.02–2.70 | 0.04 | 1.61 | 0.75–3.47 | 0.059 |
| Female sex | 1.24 | 0.85–1.79 | 0.263 | 1.42 | 0.96–2.09 | 0.077 |
| Age ≤ 64 years | 1 | 1 | ||||
| Age 65–74 years | 2.01 | 1.41–2.86 | ＜0.001 | 2.32 | 1.25–4.32 | 0.054 |
| Age ≥ 75 years | 5.63 | 3.54–8.98 | ＜0.001 | 5.62 | 2.70–4.32 | 0.001 |
＊Increase in odds ratio per score
SAH, Subarachnoid hemorrhage; OR, odds ratio; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; GCS, Glasgow Coma Scale; WFNS, World Federation of Neurosurgical Soc
Regarding death at discharge, disease severity and an older age (≥ 75 years old) at disease onset were potent risk factors for both ischemic stroke and SAH due to arterial dissection (Supplementary Table 5). Hypertension was a risk factor for mortality in patients with ischemic stroke (p＜0.05). The OR of death at discharge in patients with SAH due to arterial dissection was 4.05 when the group of patients with ischemic stroke due to arterial dissection was set as a reference, and the OR for older age (≥ 75 years old) was 2.92 when the group of patients ＜65 years old was set as the reference (Supplementary Table 5).
| Risk factor | Univariate model | Multivariate model | ||||
|---|---|---|---|---|---|---|
| OR | 95% CI | P | OR | 95%CI | p | |
| Ischemic stroke | 1 | 1 | ||||
| SAH | 7.24 | 4.58–11.45 | ＜0.001 | 4.05 | 2.01–8.51 | ＜0.001 |
| NIHSS score on admission＊ | 1.08 | 1.07–1.10 | ＜0.001 | 1.04 | 0.98–1.09 | 0.166 |
| GCS score on admission＊ | 0.80 | 0.76–0.84 | ＜0.001 | 0.96 | 0.81–1.14 | 0.654 |
| Female sex | 1.75 | 1.20–2.56 | 0.004 | 0.85 | 0.48–1.53 | 0.591 |
| Hypertension | 1.16 | 0.78–1.72 | 0.468 | |||
| Dyslipidemia | 0.72 | 0.42–1.23 | 0.234 | |||
| Diabetes mellitus | 0.87 | 0.44–1.71 | 0.68 | |||
| Age ≤ 64 years | 1 | 1 | ||||
| Age 65–74 years | 1.59 | 0.94–2.67 | 0.081 | 1.95 | 1.02–1.06 | 0.083 |
| Age ≥ 75 years | 2.53 | 1.50–2.67 | ＜0.001 | 2.92 | 1.35–6.33 | 0.007 |
＊Increase in odds ratio per score
SAH, subarachnoid hemorrhage; NIHSS, National Institutes of Health Stroke Scale; GCS, Glasgow Coma Scale
Arterial dissection is a rare cause of strokes. Thus, a previous report3) discussing the characteristics of stroke due to arterial dissection included a relatively small number of patients (＜100), and the institutes involved were often limited to either neurology or neurosurgery departments. In contrast, our study included over 1000 patients with stroke due to arterial dissection registered in both neurology and neurosurgery departments in Japan. Therefore, the results of the present study better represent the actual characteristics of stroke caused by arterial dissection than previous reports. We found that the most common subtype of stroke due to arterial dissection was ischemic stroke, followed by SAH. The ratio of stroke subtypes associated with arterial dissection differed between patients with and without arterial dissection. Debette et al.3) reviewed 4 reports involving more than 40 patients with intracranial and extracranial arterial dissection, indicating SAH predominance over ischemic stroke as a clinical subtype of stroke. Among the reports involving over 40 patients with intracranial vertebrobasilar artery dissection and stroke, 5 indicated SAH predominance, and 2 indicated ischemic stroke predominance3). In the present study, both intracranial carotid artery and vertebral artery dissection-induced ischemic stroke occurred more frequently than SAH did. Most authors reporting SAH predominance were from neurosurgery departments, suggesting an institutional bias. Our institute includes both the neurology and neurosurgery departments, potentially offering results that are more representative of real-world outcomes. In other words, ischemic stroke is more prevalent than SAH, in association with arterial dissection.
Several reports indicated a male predisposition (53%–86%) for inducing stroke due to arterial dissection2, 8, 9). In the present study, the proportions of male patients with ischemic stroke and SAH due to arterial dissection were significantly higher than those of the corresponding stroke subtypes without dissection. Among the patients with SAH due to arterial dissection, more than half were men, which was only one-third of the patients with SAH without dissection. de Rooij et al.9) reported that the incidence of total SAH was 1.24 times higher in women than in men, and this difference in sex was observed in patients ≥ 55 years old. The higher prevalence of SAH and arterial dissection in male patients may be partly attributed to the predominance of individuals ＜55 years old.
In our cohort, the number of patients with ischemic stroke and SAH due to arterial dissection peaked in their 40s and 50s, respectively. The frequencies of ischemic stroke and SAH due to arterial dissection were 0.5% and 4.0% of all ischemic stroke and SAH cases, respectively. Among patients 10–49 years old, these frequencies increased to 5.0% and 7.1%, respectively. The reported frequency of ischemic stroke due to arterial dissection accounts for 10%–20% of all ischemic strokes in young adults 10–49 or 10–59 years old10-14). In contrast, SAH due to arterial dissection accounts for 4%–8% of total SAH in the young adult population. The frequency of ischemic stroke due to arterial dissection in our cohort was lower, and that of SAH due to arterial dissection was similar to that previously reported.
Dissection of the intracranial vertebral artery was the most frequent (44.1% ischemic stroke and 46.8% SAH) in patients with stroke due to arterial dissection. In the European population, the extracranial carotid artery is commonly involved in ischemic stroke due to arterial dissection3, 15, 16), whereas in the Asian population, the intracranial vertebral artery is predominantly involved3, 17-19). Our results from Japan, which is located in East Asia, showed an apparent predominance of intracranial arteries to induce ischemic stroke, which is consistent with the study results from an Asian population.
Male sex, hypertension, diabetes, dyslipidemia, and history of smoking are risk factors for cryptogenic and total ischemic stroke20, 21). Hypertension, high fasting plasma glucose, and smoking are also risk factors for SAH21). Among these, we found frequent complications such as hypertension, smoking, and dyslipidemia in patients with ischemic stroke or SAH due to arterial dissection. Patients with ischemic stroke are frequently complicated with diabetes mellitus. Blum and Yahgi1) reviewed articles on risk factors for the development of cervical artery dissection published between 1990 and 2015 and reported the risks of hypertension, hypercholesterolemia, migraine, and hyperhomocysteinemia. Risk factors for cryptogenic stroke may also predispose patients with arterial dissection to an ischemic stroke. However, the involvement of these factors, except for migraine, heavy drinking, and smoking habit, is less frequent than with ischemic stroke without dissection.
Among patients with stroke due to arterial dissection, male sex, dyslipidemia, diabetes mellitus, and a history of smoking predispose ischemic stroke patients to SAH. These vascular risks may induce ischemic stroke rather than rupture of the dissected arteries.
Migraine can be a risk factor for both ischemic stroke and cerebral artery dissection22). In the present cohort, patients with migraine presented with ischemic stroke due to arterial dissection more frequently than those without dissection. Migraine is a cardinal risk factor for ischemic stroke, comparable to smoking23), and may be a more potent risk factor for arterial dissection. However, the number of reported patients was small.
Unruptured intracerebral and extracerebral artery dissections do not usually have an aggressive clinical course. However, the optimal treatment strategy remains controversial. The American Heart Association guidelines suggest conservative treatment involving anticoagulation24). In the present study, 36.1% of the patients received either anticoagulants or antiplatelet agents with or without edaravone. Edaravone was selected for 28.8% of patients with arterial dissection presenting with ischemic stroke. As a free radical scavenger, it has potential benefits in acute ischemic stroke25). Given its nonbleeding properties and ease of selection, clinicians often opt for edaravone to manage acute ischemic stroke due to arterial dissection.
Convalescence of arterial dissection is beneficial for ischemic stroke26). In the present study, more than half of the patients with ischemic stroke due to arterial dissection showed favorable outcomes (mRS ≤ 1) at discharge, and 18.7% showed unfavorable outcomes, including 3.7% deaths. Patients with ischemic stroke secondary to dissection may have better prognosis than those without dissection. A younger age at admission for ischemic stroke and dissection may contribute to a more favorable prognosis. The NIHSS score of patients with ischemic stroke upon admission was lower in those with arterial dissection than in those without. In contrast, SAH associated with arterial dissection results in a poor prognosis, with 20.0% mortality, as well as SAH without SAH (Fig.4).
As shown in Table 2, a univariate analysis revealed disease severity on admission, including disturbed consciousness and older age, as potent risk factors for unfavorable outcomes in both ischemic stroke and SAH in patients with artery dissection. Older age, consciousness level, and the scale of motor disability on admission have previously been implicated as patient factors associated with a poor prognosis in general stroke27). Our findings are consistent with these results. Hypertension and diabetes mellitus may also be risk factors for unfavorable outcomes, including death (Supplementary Table 5), in patients with dissection and ischemic stroke but not in patients with SAH. These are the known risk factors for cryptogenic ischemic stroke21). Vascular disorders induced by these risk factors may also predispose patients to ischemic stroke owing to arterial dissection and poor outcomes.
Study Strengths and LimitationsThe primary advantage of this study is that a large number of patients from both the neurology and neurosurgery departments were involved. The primary limitation was the registry-based nature of the study, which lacked a comprehensive patient history, thus preventing adjusting for confounders, such as the family history, medication, and history of stroke and/or cerebral artery dissection. Second, dissections with local symptoms were not considered. Therefore, the precise risk of developing either hemorrhagic or non-hemorrhagic stroke cannot be determined in patients with cerebral artery dissection without stroke. Third, the stroke population was recruited from an academic hospital and a regional stroke center and was not representative of all stroke patients. Fourth, the registry dataset often has incomplete data entries. Although age, sex, and stroke subtype were mandatory, other information was not consistently provided. For example, the location of dissection has only been reported in 24% of patients with arterial dissection, and the presence of migraine has rarely been reported.
Among the 218,799 patients with acute stroke registered in the JSDB, we identified 880 with ischemic stroke and 457 with SAH associated with arterial dissection. Contrary to many previous reports, arterial dissection induces cerebral infarction more commonly than SAH. The age of the patients presenting with dissection and these stroke subtypes peaked in the 40s and 50s. The most frequent location of dissection in patients with ischemic stroke and SAH was the intracranial vertebral artery. Male sex, dyslipidemia, and diabetes mellitus were more likely to be associated with ischemic stroke than with SAH due to arterial dissection. Among patients with ischemic stroke and SAH due to artery dissection, 56.8% and 38.2%, respectively, achieved a favorable outcome with no significant disability (mRS ≤ 1), whereas 18.9% and 42.6%, respectively, experienced unfavorable outcomes (mRS ≥ 4). Neurological severity and older age at admission were identified as the risk factors for poor outcomes. In the present study, we investigated the characteristics of patients with stroke due to arterial dissection. The effects of treatments, including endovascular thrombectomy, on outcomes should be further clarified in future studies.
This study was supported by JSPS KAKENHI (JP21K07472 and JP23H02831) and SoftBank.
Dr. Kashihara reported honoraria from Takeda, Sumitomo Pharma, and Kyowa Kirin.
Dr. Koga reported honoraria from Daiichi Sankyo and research funds from Daiichi Sankyo and Nippon Boehringer Ingelheim.
Drs. Nakai, Handa, Kobayashi, and Usumoto report no conflicts of interest.
Dr. Yoshimura received research funding from SoftBank Corp. and Astra Zeneca Plc.
Dr. Toyoda received personal funding from Daiichi Sankyo, Bayer, Bristol Myers Squibb, Otsuka and Janssen.
