- Email: [email protected]
Etiologic Subtypes of Watershed Infarcts
Etiologic Subtypes of Watershed Infarcts Mine Hayriye Sorgun, MD, Sefer Rzayev, MD, Volkan Yilmaz, MD, and Canan Togay Isıkay, MD
Background: Two types of watershed infarcts (WI) are recognized. Internal WI are usually attributed to either severe stenosis in large arteries or acute hypotensive events, whereas external WI are thought to be caused by embolism. The aim of this study was to determine the etiologic background and prognosis of external and internal WI in our patients. Methods: We reviewed the medical records and diffusion-weighted images of the patients who were admitted to our stroke unit with acute ischemic stroke between January 2012 and November 2014. The demographics, clinical features, radiologic investigations, and other etiologic tests of the patients with internal or external WI were recorded. We determined etiologic stroke subtypes according to the automated Causative Classification System. Results: Fiftythree patients with WI were detected in our registry. Twenty-two (41.5%) of them were women. The mean age was 69 6 12.8 (33-98) years. Twenty-one (39.6%) patients had external WI: 7 (33.3%) of them had large-artery atherosclerosis (LAA), 8 (38.1%) patients had cardioembolism, 3 (14.3%) patients had stroke due to other causes (vasculitis; n 5 3), and etiologic subtype was undetermined in 3 patients (14.3%). Thirty-two (60.4%) patients had internal WI: 21 (65.6.%) of them had LAA, 5 (15.6%) patients had cardioembolism, 3 (9.4%) patients had stroke due to other causes (aneurysm; n 5 1, hypercoagulability due to chronic myeloid leukemia; n 5 1, vasculitis; n 5 1), and etiologic subtype of 3 (9.4%) patients remained cryptogenic. LAA was significantly associated with internal WI (P 5 .024). Hypertension was more common in patients with internal WI (P 5 .035). Conclusions: In this series, cardioembolism was the most common etiologic subtype in the patients with external WI, whereas internal WI were significantly associated with LAA. Uncommon causes should also be investigated in cryptogenic patients. Key Words: Ischemic stroke—watershed infarct—etiology—prognosis. Ó 2015 by National Stroke Association
Border-zone or watershed infarcts (WI) are ischemic lesions located at the junction between 2 main arterial territories.1 These lesions constitute approximately 10% of all brain infarcts.2 Border-zone distribution of infarc-
From the Department of Neurology, Ankara University School of _ Medicine, Ibni Sina Hospital, Samanpazarı, Ankara, Turkey. Received January 13, 2015; revision received March 14, 2015; accepted June 8, 2015. Address correspondence to Mine Hayriye Sorgun, MD, Depart_ ment of Neurology, Ankara University School of Medicine, Ibni Sina Hospital, Samanpazarı, Ankara, 06200 Turkey. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2015 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2015.06.002
tion has traditionally been attributed to hypoperfusion related to reduced blood flow in areas between major hemispheric vascular territories. Two types of WI are recognized. Internal (subcortical) WI usually appear as multiple small infarcts in a rosary-like pattern parallel to the lateral ventricle in the centrum semiovale or corona radiata.3 The external (cortical) WI are usually wedgeshaped and located in the cerebral cortex between the territories of anterior, middle, and posterior cerebral arteries.4,5 Internal WI are believed to be caused mainly by hemodynamic compromise secondary to severe stenosis or occlusion of paroxysmal craniocervical arteries and/or hypoperfusion under the circumstance of cardiac arrest or systemic hypotension, whereas external WI are usually attributed to embolization from the heart or
Journal of Stroke and Cerebrovascular Diseases, Vol. -, No. - (---), 2015: pp 1-6
1
M.H. SORGUN ET AL.
2
atherosclerotic plaques in large arteries. Severe stenosis or occlusion of major cerebral arteries was found to be substantially fewer in external WI.6 In the patients with both external and internal WI, the most probable mechanism of stroke seems to be hemodynamic impairment.3,6 Patients with external WI have a more benign clinical course and a better prognosis than those with internal border-zone infarcts.4 In previous studies, WI usually have been reported in stroke patients with carotid occlusive disease, with a cardioembolic source, or with severe cerebral hypoperfusion during carotid endarterectomy, systemic hypotension, cardiac arrest, or surgery.2,6-14 However, several cases with acute WI due to idiopathic hypereosinophilic syndrome, parasitic infections (eg, schistosomiasis), toxicity due to cyclosporine treatment, cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), and malignancyrelated coagulopathy have also been reported previously.15-23 However, etiologic subtypes and prognosis of WI have not been studied systematically in the patients with acute ischemic stroke. The aim of this study was to determine the etiologic subtypes and prognosis of external and internal WI in our stroke registry.
Methods The medical records of 488 consecutive patients who were admitted to the stroke unit with acute ischemic stroke from January 2012 to November 2014 were retrospectively evaluated. Magnetic resonance imaging scans of all patients were reviewed by 2 investigators (S.R. and V.Y.). Patients with external and internal WI were selected based on their lesion location on diffusionweighted imaging (DWI). The subcortical infarcts that appear in a rosary-like pattern arranged in a linear
Figure 1.
fashion parallel to the centrum semiovale or corona radiata have been recorded as internal WI. The patients with a single infarct at the same location were not included in the study. External WI are defined as wedge-shaped or ovoid infarcts located between the anterior, middle, or posterior cerebral arteries. The patients were divided into 2 groups: group 1 included patients with external WI and group 2 included patients with internal WI (Fig 1). The patients with internal WI in association with external WI were also included into group 2 based on the previous studies suggesting a similar mechanism for both patient groups. Patient demographics and medical risk factors, including history of hypertension, diabetes, hyperlipidemia, atrial fibrillation, congestive heart failure, coronary artery disease, previous transient ischemic attack, previous stroke, and the National Institutes of Health Stroke Scale (NIHSS) scores at admission, were collected using a standard data collection form and entered into an institutional database. Hypertension was defined as blood pressure 140/90 mm Hg or more on repeated measurements or prior use of antihypertensive medication, diabetes mellitus as fasting blood glucose level 126 mg/dL or more on repeated measurements or the use of medications to lower blood glucose, and atrial fibrillation by previous history or if detected on ECG or Holter. Coronary artery disease included any history of angina, myocardial infarction, or coronary revascularization. A single rater (M.H.S.) determined etiologic stroke subtypes using the automated Causative Classification System (CCS, available at https://ccs.mgh.harvard. edu).24 The CCS subtypes included supra-aortic largeartery atherosclerosis, cardioaortic embolism, smallartery occlusion, other causes, and undetermined causes. Etiologic workup included vascular imaging studies, such as carotid Doppler ultrasonography, computerized tomography angiography, magnetic resonance angiography, or digital subtraction angiography, transthoracic
Axial diffusion-weighted images show internal watershed infarct (A) and external watershed infarct (B).
WATERSHED INFARCTS
3
or transesophageal echocardiogram, 24-hour cardiac rhythm monitoring, and laboratory tests for hypercoagulability and vasculitis. Computerized tomography angiography or magnetic resonance angiography were performed in the patients with a suspicion of largeartery stenosis according to the transcranial Doppler (TCD) or carotid Doppler ultrasonography findings. The modified Rankin Scale scores were recorded according to follow-up visits of the patients.
Statistics Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) 16.0 version (SPSS Inc., Chicago, IL). Descriptive statistics were expressed as means/standard deviations for normally distributed continuous variables and median/interquartile range for non-normally distributed variables. Statistical analysis was performed to determine the etiologic background and prognosis of the patients with external and internal WI. The group rates were compared using a chi-squared test, and the means were compared using Student t test; P , .05 was considered to be statistically significant.
Results A total of 53 patients with WI (31 men [58.5%] and 22 women [41.5%]; mean age, 69 6 12.8 [33-98] years) were included in the study. Patient demographics and medical risk factors, including history of hypertension (n 5 32,
60.4%), diabetes (n 5 20, 37.7%), hyperlipidemia (n 5 26, 49.1%), atrial fibrillation (n 5 7, 13.2%), congestive heart failure (n 5 6, 11.3%), coronary artery disease (n 5 13, 24.5%), previous transient ischemic attack (n 5 7, 13.2%), and previous stroke (n 5 9, 16.9%), are shown in Table 1. No between-group differences in history of diabetes, hyperlipidemia, atrial fibrillation, congestive heart failure, coronary artery disease, previous transient ischemic attack, or previous stroke were observed (P . .05). However, hypertension was significantly more common in the group 2 compared to group 1 (P 5 .035; Table 1). Twenty-one (39.6%) patients had external WI (group 1): 7 (33.3%) of them had large-artery atherosclerosis (LAA), 8 (38.1%) patients had cardioembolism, 3 (14.3%) patients had stroke due to other causes (vasculitis; n 5 3), and etiologic subtype was undetermined in 3 (14.3%) patients (Table 2). Among the patients with LAA, 3 patients had total occlusion, 2 patients had 70%-90% stenosis, and 2 patients had 50%-69% stenosis in internal carotid artery (ICA) or middle cerebral artery (MCA). Thirty-two (60.4%) patients had internal WI (group 2): 21 (65.6%) of them had LAA, 5 (15.6%) patients had cardioembolism, 3 (9.4%) patients had stroke due to other causes (aneurysm, n 5 1; hypercoagulability due to chronic myeloid leukemia, n 5 1; vasculitis, n 5 1), and etiologic subtype of 3 (9.4%) patients remained undetermined. LAA was significantly associated with internal WI, whereas cardioembolism was found more frequently
Table 1. Epidemiologic and clinical characteristics of patients with WI
Age (y), mean 6 SD (Min-max) Sex, n (%) Female Male Medical history Hypertension, n (%) Diabetes mellitus, n (%) Atrial fibrillation, n (%) Hyperlipidemia, n (%) CAD, n (%) CHF, n (%) Previous TIA history, n (%) Previous stroke history, n (%) Admission NIHSS, mean 6 SD Median (min-max) Follow-up mRS, median (min-max) Mean 6 SD Follow-up (mo), median (min-max) Mean 6 SD
Internal WI (N 5 32)
External WI (N 5 21)
P value
72 6 10.7 (43-98)
65 6 14.8 (33-84)
.092
12 (37.5%) 20 (62.5%)
10 (47.6%) 11 (52.4%)
.465
23 (71.9) 13 (40.6) 3 (9.4) 16 (50) 10 (31.2) 4 (12.5) 5 (23.8) 8 (25) 6.94 6 5.48 6 (0-15) 4 (0-6) 2.76 6 2.14 5.5 (1-36) 8.07 6 9.0
9 (42.9) 7 (33.3) 4 (19) 10 (47.6) 3 (14.3) 2 (9.5) 2 (6.2) 1 (4.8) 3.24 6 3.19 2 (0-12) 1 (0-4) 1.57 6 1.55 8 (1-29) 12.46 6 9.73
.035 .592 .309 .865 .160 .738 .65 .055 .095 .067 .216
Abbreviations: CAD, coronary artery disease; CHF, congestive heart failure; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; TIA, transient ischemic attack; WI, watershed infarcts. Bold value indicates statistical significance.
M.H. SORGUN ET AL.
4
Table 2. Etiologic stroke subtypes of external and internal WI Internal WI External WI P value N (%) 32 (60.4) CCS classification, (%) Large-artery 21 (65.6) atherosclerosis Cardioembolism 5 (15.6) Other rare causes 3 (9.4) Undetermined 3 (9.4) causes
21 (39.6) 7 (33.3)
.024
8 (38.1) 3 (14.3) 3 (14.3)
.06 .58 .58
Abbreviations: CCS, Causative Classification System; WI, watershed infarcts. Bold value indicates statistical significance.
among the patients with external WI (P 5 .024; Table 2). Six of the patients with LAA had total occlusion, 7 patients had 70%-95% stenosis, and 8 patients had 50%-69% stenosis in (ICA) or MCA. The median NIHSS score at admission was 4 (0-28). The NIHSS score was higher in the patients with internal WI than the patients with external WI, but not significant (P 5 .095; Table 1). The mean follow-up time of the patients was 7.5 (1-36) months, and the median modified Rankin Scale score was 2 (0-6) in follow-up period. Although the patients with internal WI showed less favorable outcome than the patients with external WI, there was no significant difference between the 2 groups (P . .05; Table 1).
Discussion Until recently, border-zone infarcts were traditionally attributed to reduced blood flow caused by severe stenosis or occlusion of paroxysmal craniocervical arteries or hypoperfusion under the circumstance of cardiac arrest or severe systemic hypotension.3,9,11-13 Then, microembolism from either the heart or paroxysmal stenotic arteries has been postulated to be an alternative cause of WI.6 Caplan and Hennerici8 reported a case of transient neurologic deficits who had subcortical border-zone infarcts on DWI and severe carotid stenosis by vascular imaging. In contrast to traditional hypothesis of WI, TCD of MCA showed microembolic signals in this patient. This case and several others have demonstrated that embolization as well as reduced perfusion or a combination of both conditions may contribute to the genesis of WI. Microembolism of the vessels in arterial boundary zones has also been shown in autopsy series and experimental studies.25,26 Caplan and Hennerici8 suggested that border-zone infarcts are best explained by impaired clearance of emboli in distal arteries due to decreased perfusion. Later on, some studies focusing on a specific subtype of border-zone infarcts were reported. In one of these studies, severe carotid occlusive disease and transiently
impaired cardiac output due to ischemic heart disease, cardiomyopathy, arrhythmias, or a combination of these conditions were found significantly associated with internal WI.7 The relation between carotid occlusive disease and internal WI has been shown in subsequent studies, also.3,9,10 Lee et al10 evaluated 720 consecutive ischemic stroke patients, retrospectively. Twenty-nine patients had internal WI and 54 patients had superficial perforator infarct. They found that larger infarct size and carotid stenosis or occlusion in vascular imaging studies were significantly more frequent in the patients with internal WI. Cerebral blood flow studies using positron emission tomography, single photon emission computed tomography, and TCD in patients with carotid occlusive disease and internal WI have demonstrated reduced blood flow, reduced perfusion reserve, elevated oxygen extraction, and impaired vasomotor reactivitiy.3,11,13,27,28 In one study carried out with positron emission tomography for the measurement of cerebral oxygen extraction fraction in the patients with carotid occlusion, it has been shown that the presence of severe chronic hemodynamic compromise was significantly associated with internal WI.3 Bilateral WI usually arise secondary to systemic hypoperfusion. Gottesman et al14 investigated 98 patients with ischemic stroke after cardiac surgery and they observed that perioperative bilateral border-zone infarcts (n 5 47) were related to a decrease in intraoperative blood pressure of at least 10 mm Hg. The susceptibility of internal border-zone areas to hypoperfusion is probably caused by little collateral supply of the deep perforating lenticulostriate arteries and relatively lower perfusion pressure in the perforating medullary arteries, which are the most distal branches of the ICA. Perfusion studies revealed also that paraventricular white matter is the most vulnerable location to hemodynamic hypoperfusion. Yong et al carried out the largest study concerning the pathogenesis of different types of border-zone infarcts and they reviewed 45 patients with internal WI and 75 patients with external WI. Their results revealed an association between the presence of internal border-zone infarcts and hemodynamic compromise, whereas this relationship was less obvious in patients with external borderzone infarcts. Severe stenosis or occlusion of MCA or ICA accounted for 76% of internal border-zone infarcts without a potential source of cardiac embolism. They also observed that internal border-zone infarcts were more frequently associated with a rosary-like pattern, which is believed to be indicative of hemodynamic failure. Almost 20% of their patients with external WI had less than 50% stenosis in MCA or ICA, whereas this ratio was only 4% in the patients with internal WI. Hypertension was more common in the patients with internal WI.6 In the present study, we retrospectively investigated the etiologic subtypes of internal and external WI in a
WATERSHED INFARCTS
relatively large group of patients. In this registry, internal WI were caused most frequently by LAA, whereas cardioaortic embolism was the most common mechanism for the genesis of external WI (P 5 .024). A higher proportion of normal or less severe stenosis (,50%) in craniocervical arteries were detected in the patients with external WI (66.7%) than those with internal WI (34.4%). The rate of hypertension was significantly higher in the patients with internal WI. In previous studies, hypertension was not correlated with any type of border-zone infarcts.7,9,10,12,14 Only the results of Yong et al showing an association between hypertension and internal WI were in concordance with our findings.6 Other causes of border-zone infarcts were reported in several cases previously. There are case reports demonstrating the relation of border-zone infarcts with idiopathic hypereosinophilic syndrome, parasitic infection (eg, schistosomiasis), and toxic effects of cyclosporine therapy.23-28 In cases with idiopathic hypereosinophilic syndrome, it has been proposed that WI develop due to either thromboembolism from endomyocardial fibrosis or vascular endothelial toxic effects of eosinophilic cells. Thromboembolism may occur in conjunction with cardiac involvement throughout the course of hypereosinophilic syndrome. In a few cases, infarcts have been attributed to local thrombus formation instead of a thromboembolic cause.16-20 Lee and Hong reported a 78-year-old woman with both internal and external WI due to malignancy-related hypercoagulation (pancreatic tumor with liver metastasis).21 Malignancyrelated hypercoagulation should be considered in the patients with WI, especially if other mechanism are excluded and if the patient has impaired coagulation factors. A 48-year-old woman with bihemispheric acute internal WI caused by CADASIL was reported by Gordhan and Hudson. Microcirculatory hypoperfusion– related WI in the context of global hypoperfusion may be a causative factor in symptomatic CADASIL.22 In our study, other causes of border-zone infarcts were primary central nervous system vasculitis, cerebral aneurysm, and malignancy-related hypercoagulation (due to chronic myeloid leukemia). There was no significant difference in terms of other causes between the 2 groups. Microcirculatory hypoperfusion may also be the causative mechanism in our patients with primary central nervous system vasculitis. Cerebral aneurysm could be embolic source. Malignancy-related infarcts have been attributed to arterial thrombus formation due to coagulopathy instead of a thromboembolic source, because cardiac thrombus was excluded by echocardiography. The patients with internal WI may have clinical deterioration and prolonged hospitalization, and they have an increased tendency to remain in a disabled state during clinical follow-up. They also show an increased risk of stroke during the first days after infarction. Moriwaki et al4 reported that external WI had benign clinical course
5
and a better prognosis than those with internal borderzone infarcts. The patients with internal WI revealed poor outcome in follow-up at 3 months after stroke in the study reported by Yong et al.6 In our study, the patients with internal WI showed less favorable outcome than the patients with external WI, but the difference was not statistically significant. There are several limitations of our study, including its retrospective design, unavailable data of clinical findings and hemodynamic parameters on admission, and a relatively short follow-up duration. In addition, this is a restriction of the cohort to a single medical center. However, all of our patients underwent a detailed workup for determining the causative mechanism of ischemic stroke. Acute ischemic lesions were identified with DWI in all patients. In contrast to the previous studies, etiologic subtypes of WI have been determined systematically according to the CCS in our patients. In conclusion, our study supported the hypothesis of distinct causative mechanisms for internal and cortical border-zone infarcts. There was a consistent association between severe stenosis of craniocervical arteries and internal WI, whereas the association between cardioaortic embolism and external WI was more prominent. The patients with internal WI have a tendency to show a less favorable outcome than the patients with external WI. In addition to common etiologies of stroke, uncommon causes should also be evaluated in the patients with border-zone infarcts, especially if they remain cryptogenic. Uncommon causes in our study were primary central nervous system vasculitis, aneurysm, and malignancy-related hypercoagulation. Future studies with larger patient groups are needed to confirm the etiologic background and treatment strategies in different types of border-zone infarcts.
References 1. Schneider M. Durchblutung und sauerstoffversorgung des gehirns. Verh Dtsch Tes Kreislaufforsch 1953;19:3-12. 2. Torvik A. The pathogenesis of watershed infarcts in the brain. Stroke 1984;15:221-223. 3. Derdeyn CP, Khosla A, Videen TO, et al. Severe hemodynamic impairment and border-zone region infarction. Radiology 2001;220:195-201. 4. Moriwaki H, Matsumoto M, Hashikawa K, et al. Hemodynamic aspect of cerebral watershed infarction: assessment of perfusion reserve using iodine-123-iodoamphetamine SPECT. J Nucl Med 1997;38:1556-1562. 5. Krapf H, Widder B, Skalej M. Small rosarylike infarctions in the centrum ovale suggest hemodynamic failure. AJNR Am J Neuroradiol 1998;19:1479-1484. 6. Yong SW, Bang OY, Lee PH, et al. Internal and cortical border-zone infarction: clinical and diffusion-weighted imaging features. Stroke 2006;37:841-846. 7. Bladin CF, Chambers BR. Clinical features, pathogenesis, and computed tomographic characteristics of internal watershed infarction. Stroke 1993;24:1925-1932.
6 8. Caplan LR, Hennerici M. Impaired clearance of emboli (washout) is an important link between hypoperfusion, embolism, and ischemic stroke. Arch Neurol 1998; 55:1475-1482. 9. Del Sette M, Eliasziw M, Streifler JY, et al. Internal borderzone infarction: a marker for severe stenosis in patients with symptomatic internal carotid artery disease. For the North American Symptomatic Carotid Endarterectomy (NASCET) Group. Stroke 2000;31:631-636. 10. Lee PH, Bang OY, Oh SH, Joo IS, Huh K. Subcortical white matter infarcts: comparison of superficial perforating artery and internal border-zone infarcts using diffusion-weighted magnetic resonance imaging. Stroke 2003;34:2630-2635. 11. Moustafa RR, Izquierdo-Garcia D, Jones PS, et al. Watershed infarcts in transient ischemic attack/minor stroke with .50% carotid stenosis hemodynamic or embolic? Stroke 2010;41:1410-1416. 12. Li HF, Zhang X, Zhang Y, et al. Clinical and neuroradiological features of internal watershed infarction and the occlusive diseases of carotid artery system. Neurol Res 2010;32:1090-1096. 13. Moustafa RR, Momjian-Mayor I, Jones PS, et al. Microembolism versus hemodynamic impairment in rosarylike deep watershed infarcts. A combined positron emission tomography and transcranial Doppler study. Stroke 2011;42:3138-3143. 14. Gottesman RF, Sherman PM, Grega MA, et al. Watershed strokes after cardiac surgery: diagnosis, etiology, and outcome. Stroke 2006;37:2306-2311. 15. Dejode JM, Antonini F, Lagier P, et al. Capgras syndrome: a clinical manifestation of watershed cerebral infarct complicating the use of extracorporeal membrane oxygenation. Crit Care 2001;5:232-235. 16. Sarazin M, Caumes E, Cohen A, et al. Multiple microembolic border-zone brain infarctions and endomyocardial fibrosis in idiopathic hypereosinophilic syndrome and in Schistosoma mansoni infestation. J Neurol Neurosurg Psychiatry 2004;75:305-307.
M.H. SORGUN ET AL. 17. McMillan HJ, Johnston DL, Doja A. Watershed infarction due to acute hypereosinophilia. Neurology 2008;70:80-82. 18. Perini GF, Kassab C, Bley C, et al. Acute cerebral infarction in watershed distribution in a patient with hypereosinophilic syndrome. Arq Neuropsiquiatr 2009; 67:510-512. 19. Grandiere-Perez L, Caumes E. Corticosteroids for watershed infarction in acute schistosomiasis. Clin Infect Dis 2013;57:918-919. 20. Aida L, Parkhutik V, Tembl JI, et al. Embolism and impaired washout: a possible explanation of borderzone strokes in hypereosinophilic syndrome. J Neurol Sci 2013;325:162-164. 21. Lee SP, Hong CT. Widespread watershed infarct in patient with malignancy-related hypercoagulation. Acta Neurol Taiwan 2012;21:49-50. 22. Gordhan A, Hudson BK. Acute watershed infarcts with global cerebral hypoperfusion in symptomatic CADASIL. Radiol Case 2013;7:8-15. 23. Bartynski WS, Grabb BC, Zeigler Z, et al. Watershed imaging features and clinical vascular injury in cyclosporin A neurotoxicity. J Comput Assist Tomogr 1997; 21:872-880. 24. Ay H, Furie KL, Singhal A, et al. An evidence-based causative classification system for acute ischemic stroke. Ann Neurol 2005;58:688-697. 25. Pollanen MS, Deck JH. The mechanism of embolic watershed infarction: experimental studies. Can J Neurol Sci 1990;17:395-398. 26. Masuda J, Yutani C, Ogata J, et al. Atheromatous embolism in the brain: a clinicopathologic analysis of 15 autopsy cases. Neurology 1994;44:1231-1237. 27. Arakawa S, Minematsu K, Hirano T, et al. Topographic distribution of misery perfusion in relation to internal and superficial borderzones. AJNR Am J Neuroradiol 2003;24:427-435. 28. Mangla R, Kolar B, Almast J, et al. Border zone infarcts: pathophysiologic and imaging characteristics. Radiographics 2011;31:1201-1214.
Background: Two types of watershed infarcts (WI) are recognized. Internal WI are usually attributed to either severe stenosis in large arteries or acute hypotensive events, whereas external WI are thought to be caused by embolism. The aim of this study was to determine the etiologic background and prognosis of external and internal WI in our patients. Methods: We reviewed the medical records and diffusion-weighted images of the patients who were admitted to our stroke unit with acute ischemic stroke between January 2012 and November 2014. The demographics, clinical features, radiologic investigations, and other etiologic tests of the patients with internal or external WI were recorded. We determined etiologic stroke subtypes according to the automated Causative Classification System. Results: Fiftythree patients with WI were detected in our registry. Twenty-two (41.5%) of them were women. The mean age was 69 6 12.8 (33-98) years. Twenty-one (39.6%) patients had external WI: 7 (33.3%) of them had large-artery atherosclerosis (LAA), 8 (38.1%) patients had cardioembolism, 3 (14.3%) patients had stroke due to other causes (vasculitis; n 5 3), and etiologic subtype was undetermined in 3 patients (14.3%). Thirty-two (60.4%) patients had internal WI: 21 (65.6.%) of them had LAA, 5 (15.6%) patients had cardioembolism, 3 (9.4%) patients had stroke due to other causes (aneurysm; n 5 1, hypercoagulability due to chronic myeloid leukemia; n 5 1, vasculitis; n 5 1), and etiologic subtype of 3 (9.4%) patients remained cryptogenic. LAA was significantly associated with internal WI (P 5 .024). Hypertension was more common in patients with internal WI (P 5 .035). Conclusions: In this series, cardioembolism was the most common etiologic subtype in the patients with external WI, whereas internal WI were significantly associated with LAA. Uncommon causes should also be investigated in cryptogenic patients. Key Words: Ischemic stroke—watershed infarct—etiology—prognosis. Ó 2015 by National Stroke Association
Border-zone or watershed infarcts (WI) are ischemic lesions located at the junction between 2 main arterial territories.1 These lesions constitute approximately 10% of all brain infarcts.2 Border-zone distribution of infarc-
From the Department of Neurology, Ankara University School of _ Medicine, Ibni Sina Hospital, Samanpazarı, Ankara, Turkey. Received January 13, 2015; revision received March 14, 2015; accepted June 8, 2015. Address correspondence to Mine Hayriye Sorgun, MD, Depart_ ment of Neurology, Ankara University School of Medicine, Ibni Sina Hospital, Samanpazarı, Ankara, 06200 Turkey. E-mail: [email protected]. 1052-3057/$ - see front matter Ó 2015 by National Stroke Association http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2015.06.002
tion has traditionally been attributed to hypoperfusion related to reduced blood flow in areas between major hemispheric vascular territories. Two types of WI are recognized. Internal (subcortical) WI usually appear as multiple small infarcts in a rosary-like pattern parallel to the lateral ventricle in the centrum semiovale or corona radiata.3 The external (cortical) WI are usually wedgeshaped and located in the cerebral cortex between the territories of anterior, middle, and posterior cerebral arteries.4,5 Internal WI are believed to be caused mainly by hemodynamic compromise secondary to severe stenosis or occlusion of paroxysmal craniocervical arteries and/or hypoperfusion under the circumstance of cardiac arrest or systemic hypotension, whereas external WI are usually attributed to embolization from the heart or
Journal of Stroke and Cerebrovascular Diseases, Vol. -, No. - (---), 2015: pp 1-6
1
M.H. SORGUN ET AL.
2
atherosclerotic plaques in large arteries. Severe stenosis or occlusion of major cerebral arteries was found to be substantially fewer in external WI.6 In the patients with both external and internal WI, the most probable mechanism of stroke seems to be hemodynamic impairment.3,6 Patients with external WI have a more benign clinical course and a better prognosis than those with internal border-zone infarcts.4 In previous studies, WI usually have been reported in stroke patients with carotid occlusive disease, with a cardioembolic source, or with severe cerebral hypoperfusion during carotid endarterectomy, systemic hypotension, cardiac arrest, or surgery.2,6-14 However, several cases with acute WI due to idiopathic hypereosinophilic syndrome, parasitic infections (eg, schistosomiasis), toxicity due to cyclosporine treatment, cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), and malignancyrelated coagulopathy have also been reported previously.15-23 However, etiologic subtypes and prognosis of WI have not been studied systematically in the patients with acute ischemic stroke. The aim of this study was to determine the etiologic subtypes and prognosis of external and internal WI in our stroke registry.
Methods The medical records of 488 consecutive patients who were admitted to the stroke unit with acute ischemic stroke from January 2012 to November 2014 were retrospectively evaluated. Magnetic resonance imaging scans of all patients were reviewed by 2 investigators (S.R. and V.Y.). Patients with external and internal WI were selected based on their lesion location on diffusionweighted imaging (DWI). The subcortical infarcts that appear in a rosary-like pattern arranged in a linear
Figure 1.
fashion parallel to the centrum semiovale or corona radiata have been recorded as internal WI. The patients with a single infarct at the same location were not included in the study. External WI are defined as wedge-shaped or ovoid infarcts located between the anterior, middle, or posterior cerebral arteries. The patients were divided into 2 groups: group 1 included patients with external WI and group 2 included patients with internal WI (Fig 1). The patients with internal WI in association with external WI were also included into group 2 based on the previous studies suggesting a similar mechanism for both patient groups. Patient demographics and medical risk factors, including history of hypertension, diabetes, hyperlipidemia, atrial fibrillation, congestive heart failure, coronary artery disease, previous transient ischemic attack, previous stroke, and the National Institutes of Health Stroke Scale (NIHSS) scores at admission, were collected using a standard data collection form and entered into an institutional database. Hypertension was defined as blood pressure 140/90 mm Hg or more on repeated measurements or prior use of antihypertensive medication, diabetes mellitus as fasting blood glucose level 126 mg/dL or more on repeated measurements or the use of medications to lower blood glucose, and atrial fibrillation by previous history or if detected on ECG or Holter. Coronary artery disease included any history of angina, myocardial infarction, or coronary revascularization. A single rater (M.H.S.) determined etiologic stroke subtypes using the automated Causative Classification System (CCS, available at https://ccs.mgh.harvard. edu).24 The CCS subtypes included supra-aortic largeartery atherosclerosis, cardioaortic embolism, smallartery occlusion, other causes, and undetermined causes. Etiologic workup included vascular imaging studies, such as carotid Doppler ultrasonography, computerized tomography angiography, magnetic resonance angiography, or digital subtraction angiography, transthoracic
Axial diffusion-weighted images show internal watershed infarct (A) and external watershed infarct (B).
WATERSHED INFARCTS
3
or transesophageal echocardiogram, 24-hour cardiac rhythm monitoring, and laboratory tests for hypercoagulability and vasculitis. Computerized tomography angiography or magnetic resonance angiography were performed in the patients with a suspicion of largeartery stenosis according to the transcranial Doppler (TCD) or carotid Doppler ultrasonography findings. The modified Rankin Scale scores were recorded according to follow-up visits of the patients.
Statistics Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) 16.0 version (SPSS Inc., Chicago, IL). Descriptive statistics were expressed as means/standard deviations for normally distributed continuous variables and median/interquartile range for non-normally distributed variables. Statistical analysis was performed to determine the etiologic background and prognosis of the patients with external and internal WI. The group rates were compared using a chi-squared test, and the means were compared using Student t test; P , .05 was considered to be statistically significant.
Results A total of 53 patients with WI (31 men [58.5%] and 22 women [41.5%]; mean age, 69 6 12.8 [33-98] years) were included in the study. Patient demographics and medical risk factors, including history of hypertension (n 5 32,
60.4%), diabetes (n 5 20, 37.7%), hyperlipidemia (n 5 26, 49.1%), atrial fibrillation (n 5 7, 13.2%), congestive heart failure (n 5 6, 11.3%), coronary artery disease (n 5 13, 24.5%), previous transient ischemic attack (n 5 7, 13.2%), and previous stroke (n 5 9, 16.9%), are shown in Table 1. No between-group differences in history of diabetes, hyperlipidemia, atrial fibrillation, congestive heart failure, coronary artery disease, previous transient ischemic attack, or previous stroke were observed (P . .05). However, hypertension was significantly more common in the group 2 compared to group 1 (P 5 .035; Table 1). Twenty-one (39.6%) patients had external WI (group 1): 7 (33.3%) of them had large-artery atherosclerosis (LAA), 8 (38.1%) patients had cardioembolism, 3 (14.3%) patients had stroke due to other causes (vasculitis; n 5 3), and etiologic subtype was undetermined in 3 (14.3%) patients (Table 2). Among the patients with LAA, 3 patients had total occlusion, 2 patients had 70%-90% stenosis, and 2 patients had 50%-69% stenosis in internal carotid artery (ICA) or middle cerebral artery (MCA). Thirty-two (60.4%) patients had internal WI (group 2): 21 (65.6%) of them had LAA, 5 (15.6%) patients had cardioembolism, 3 (9.4%) patients had stroke due to other causes (aneurysm, n 5 1; hypercoagulability due to chronic myeloid leukemia, n 5 1; vasculitis, n 5 1), and etiologic subtype of 3 (9.4%) patients remained undetermined. LAA was significantly associated with internal WI, whereas cardioembolism was found more frequently
Table 1. Epidemiologic and clinical characteristics of patients with WI
Age (y), mean 6 SD (Min-max) Sex, n (%) Female Male Medical history Hypertension, n (%) Diabetes mellitus, n (%) Atrial fibrillation, n (%) Hyperlipidemia, n (%) CAD, n (%) CHF, n (%) Previous TIA history, n (%) Previous stroke history, n (%) Admission NIHSS, mean 6 SD Median (min-max) Follow-up mRS, median (min-max) Mean 6 SD Follow-up (mo), median (min-max) Mean 6 SD
Internal WI (N 5 32)
External WI (N 5 21)
P value
72 6 10.7 (43-98)
65 6 14.8 (33-84)
.092
12 (37.5%) 20 (62.5%)
10 (47.6%) 11 (52.4%)
.465
23 (71.9) 13 (40.6) 3 (9.4) 16 (50) 10 (31.2) 4 (12.5) 5 (23.8) 8 (25) 6.94 6 5.48 6 (0-15) 4 (0-6) 2.76 6 2.14 5.5 (1-36) 8.07 6 9.0
9 (42.9) 7 (33.3) 4 (19) 10 (47.6) 3 (14.3) 2 (9.5) 2 (6.2) 1 (4.8) 3.24 6 3.19 2 (0-12) 1 (0-4) 1.57 6 1.55 8 (1-29) 12.46 6 9.73
.035 .592 .309 .865 .160 .738 .65 .055 .095 .067 .216
Abbreviations: CAD, coronary artery disease; CHF, congestive heart failure; mRS, modified Rankin Scale; NIHSS, National Institutes of Health Stroke Scale; TIA, transient ischemic attack; WI, watershed infarcts. Bold value indicates statistical significance.
M.H. SORGUN ET AL.
4
Table 2. Etiologic stroke subtypes of external and internal WI Internal WI External WI P value N (%) 32 (60.4) CCS classification, (%) Large-artery 21 (65.6) atherosclerosis Cardioembolism 5 (15.6) Other rare causes 3 (9.4) Undetermined 3 (9.4) causes
21 (39.6) 7 (33.3)
.024
8 (38.1) 3 (14.3) 3 (14.3)
.06 .58 .58
Abbreviations: CCS, Causative Classification System; WI, watershed infarcts. Bold value indicates statistical significance.
among the patients with external WI (P 5 .024; Table 2). Six of the patients with LAA had total occlusion, 7 patients had 70%-95% stenosis, and 8 patients had 50%-69% stenosis in (ICA) or MCA. The median NIHSS score at admission was 4 (0-28). The NIHSS score was higher in the patients with internal WI than the patients with external WI, but not significant (P 5 .095; Table 1). The mean follow-up time of the patients was 7.5 (1-36) months, and the median modified Rankin Scale score was 2 (0-6) in follow-up period. Although the patients with internal WI showed less favorable outcome than the patients with external WI, there was no significant difference between the 2 groups (P . .05; Table 1).
Discussion Until recently, border-zone infarcts were traditionally attributed to reduced blood flow caused by severe stenosis or occlusion of paroxysmal craniocervical arteries or hypoperfusion under the circumstance of cardiac arrest or severe systemic hypotension.3,9,11-13 Then, microembolism from either the heart or paroxysmal stenotic arteries has been postulated to be an alternative cause of WI.6 Caplan and Hennerici8 reported a case of transient neurologic deficits who had subcortical border-zone infarcts on DWI and severe carotid stenosis by vascular imaging. In contrast to traditional hypothesis of WI, TCD of MCA showed microembolic signals in this patient. This case and several others have demonstrated that embolization as well as reduced perfusion or a combination of both conditions may contribute to the genesis of WI. Microembolism of the vessels in arterial boundary zones has also been shown in autopsy series and experimental studies.25,26 Caplan and Hennerici8 suggested that border-zone infarcts are best explained by impaired clearance of emboli in distal arteries due to decreased perfusion. Later on, some studies focusing on a specific subtype of border-zone infarcts were reported. In one of these studies, severe carotid occlusive disease and transiently
impaired cardiac output due to ischemic heart disease, cardiomyopathy, arrhythmias, or a combination of these conditions were found significantly associated with internal WI.7 The relation between carotid occlusive disease and internal WI has been shown in subsequent studies, also.3,9,10 Lee et al10 evaluated 720 consecutive ischemic stroke patients, retrospectively. Twenty-nine patients had internal WI and 54 patients had superficial perforator infarct. They found that larger infarct size and carotid stenosis or occlusion in vascular imaging studies were significantly more frequent in the patients with internal WI. Cerebral blood flow studies using positron emission tomography, single photon emission computed tomography, and TCD in patients with carotid occlusive disease and internal WI have demonstrated reduced blood flow, reduced perfusion reserve, elevated oxygen extraction, and impaired vasomotor reactivitiy.3,11,13,27,28 In one study carried out with positron emission tomography for the measurement of cerebral oxygen extraction fraction in the patients with carotid occlusion, it has been shown that the presence of severe chronic hemodynamic compromise was significantly associated with internal WI.3 Bilateral WI usually arise secondary to systemic hypoperfusion. Gottesman et al14 investigated 98 patients with ischemic stroke after cardiac surgery and they observed that perioperative bilateral border-zone infarcts (n 5 47) were related to a decrease in intraoperative blood pressure of at least 10 mm Hg. The susceptibility of internal border-zone areas to hypoperfusion is probably caused by little collateral supply of the deep perforating lenticulostriate arteries and relatively lower perfusion pressure in the perforating medullary arteries, which are the most distal branches of the ICA. Perfusion studies revealed also that paraventricular white matter is the most vulnerable location to hemodynamic hypoperfusion. Yong et al carried out the largest study concerning the pathogenesis of different types of border-zone infarcts and they reviewed 45 patients with internal WI and 75 patients with external WI. Their results revealed an association between the presence of internal border-zone infarcts and hemodynamic compromise, whereas this relationship was less obvious in patients with external borderzone infarcts. Severe stenosis or occlusion of MCA or ICA accounted for 76% of internal border-zone infarcts without a potential source of cardiac embolism. They also observed that internal border-zone infarcts were more frequently associated with a rosary-like pattern, which is believed to be indicative of hemodynamic failure. Almost 20% of their patients with external WI had less than 50% stenosis in MCA or ICA, whereas this ratio was only 4% in the patients with internal WI. Hypertension was more common in the patients with internal WI.6 In the present study, we retrospectively investigated the etiologic subtypes of internal and external WI in a
WATERSHED INFARCTS
relatively large group of patients. In this registry, internal WI were caused most frequently by LAA, whereas cardioaortic embolism was the most common mechanism for the genesis of external WI (P 5 .024). A higher proportion of normal or less severe stenosis (,50%) in craniocervical arteries were detected in the patients with external WI (66.7%) than those with internal WI (34.4%). The rate of hypertension was significantly higher in the patients with internal WI. In previous studies, hypertension was not correlated with any type of border-zone infarcts.7,9,10,12,14 Only the results of Yong et al showing an association between hypertension and internal WI were in concordance with our findings.6 Other causes of border-zone infarcts were reported in several cases previously. There are case reports demonstrating the relation of border-zone infarcts with idiopathic hypereosinophilic syndrome, parasitic infection (eg, schistosomiasis), and toxic effects of cyclosporine therapy.23-28 In cases with idiopathic hypereosinophilic syndrome, it has been proposed that WI develop due to either thromboembolism from endomyocardial fibrosis or vascular endothelial toxic effects of eosinophilic cells. Thromboembolism may occur in conjunction with cardiac involvement throughout the course of hypereosinophilic syndrome. In a few cases, infarcts have been attributed to local thrombus formation instead of a thromboembolic cause.16-20 Lee and Hong reported a 78-year-old woman with both internal and external WI due to malignancy-related hypercoagulation (pancreatic tumor with liver metastasis).21 Malignancyrelated hypercoagulation should be considered in the patients with WI, especially if other mechanism are excluded and if the patient has impaired coagulation factors. A 48-year-old woman with bihemispheric acute internal WI caused by CADASIL was reported by Gordhan and Hudson. Microcirculatory hypoperfusion– related WI in the context of global hypoperfusion may be a causative factor in symptomatic CADASIL.22 In our study, other causes of border-zone infarcts were primary central nervous system vasculitis, cerebral aneurysm, and malignancy-related hypercoagulation (due to chronic myeloid leukemia). There was no significant difference in terms of other causes between the 2 groups. Microcirculatory hypoperfusion may also be the causative mechanism in our patients with primary central nervous system vasculitis. Cerebral aneurysm could be embolic source. Malignancy-related infarcts have been attributed to arterial thrombus formation due to coagulopathy instead of a thromboembolic source, because cardiac thrombus was excluded by echocardiography. The patients with internal WI may have clinical deterioration and prolonged hospitalization, and they have an increased tendency to remain in a disabled state during clinical follow-up. They also show an increased risk of stroke during the first days after infarction. Moriwaki et al4 reported that external WI had benign clinical course
5
and a better prognosis than those with internal borderzone infarcts. The patients with internal WI revealed poor outcome in follow-up at 3 months after stroke in the study reported by Yong et al.6 In our study, the patients with internal WI showed less favorable outcome than the patients with external WI, but the difference was not statistically significant. There are several limitations of our study, including its retrospective design, unavailable data of clinical findings and hemodynamic parameters on admission, and a relatively short follow-up duration. In addition, this is a restriction of the cohort to a single medical center. However, all of our patients underwent a detailed workup for determining the causative mechanism of ischemic stroke. Acute ischemic lesions were identified with DWI in all patients. In contrast to the previous studies, etiologic subtypes of WI have been determined systematically according to the CCS in our patients. In conclusion, our study supported the hypothesis of distinct causative mechanisms for internal and cortical border-zone infarcts. There was a consistent association between severe stenosis of craniocervical arteries and internal WI, whereas the association between cardioaortic embolism and external WI was more prominent. The patients with internal WI have a tendency to show a less favorable outcome than the patients with external WI. In addition to common etiologies of stroke, uncommon causes should also be evaluated in the patients with border-zone infarcts, especially if they remain cryptogenic. Uncommon causes in our study were primary central nervous system vasculitis, aneurysm, and malignancy-related hypercoagulation. Future studies with larger patient groups are needed to confirm the etiologic background and treatment strategies in different types of border-zone infarcts.
References 1. Schneider M. Durchblutung und sauerstoffversorgung des gehirns. Verh Dtsch Tes Kreislaufforsch 1953;19:3-12. 2. Torvik A. The pathogenesis of watershed infarcts in the brain. Stroke 1984;15:221-223. 3. Derdeyn CP, Khosla A, Videen TO, et al. Severe hemodynamic impairment and border-zone region infarction. Radiology 2001;220:195-201. 4. Moriwaki H, Matsumoto M, Hashikawa K, et al. Hemodynamic aspect of cerebral watershed infarction: assessment of perfusion reserve using iodine-123-iodoamphetamine SPECT. J Nucl Med 1997;38:1556-1562. 5. Krapf H, Widder B, Skalej M. Small rosarylike infarctions in the centrum ovale suggest hemodynamic failure. AJNR Am J Neuroradiol 1998;19:1479-1484. 6. Yong SW, Bang OY, Lee PH, et al. Internal and cortical border-zone infarction: clinical and diffusion-weighted imaging features. Stroke 2006;37:841-846. 7. Bladin CF, Chambers BR. Clinical features, pathogenesis, and computed tomographic characteristics of internal watershed infarction. Stroke 1993;24:1925-1932.
6 8. Caplan LR, Hennerici M. Impaired clearance of emboli (washout) is an important link between hypoperfusion, embolism, and ischemic stroke. Arch Neurol 1998; 55:1475-1482. 9. Del Sette M, Eliasziw M, Streifler JY, et al. Internal borderzone infarction: a marker for severe stenosis in patients with symptomatic internal carotid artery disease. For the North American Symptomatic Carotid Endarterectomy (NASCET) Group. Stroke 2000;31:631-636. 10. Lee PH, Bang OY, Oh SH, Joo IS, Huh K. Subcortical white matter infarcts: comparison of superficial perforating artery and internal border-zone infarcts using diffusion-weighted magnetic resonance imaging. Stroke 2003;34:2630-2635. 11. Moustafa RR, Izquierdo-Garcia D, Jones PS, et al. Watershed infarcts in transient ischemic attack/minor stroke with .50% carotid stenosis hemodynamic or embolic? Stroke 2010;41:1410-1416. 12. Li HF, Zhang X, Zhang Y, et al. Clinical and neuroradiological features of internal watershed infarction and the occlusive diseases of carotid artery system. Neurol Res 2010;32:1090-1096. 13. Moustafa RR, Momjian-Mayor I, Jones PS, et al. Microembolism versus hemodynamic impairment in rosarylike deep watershed infarcts. A combined positron emission tomography and transcranial Doppler study. Stroke 2011;42:3138-3143. 14. Gottesman RF, Sherman PM, Grega MA, et al. Watershed strokes after cardiac surgery: diagnosis, etiology, and outcome. Stroke 2006;37:2306-2311. 15. Dejode JM, Antonini F, Lagier P, et al. Capgras syndrome: a clinical manifestation of watershed cerebral infarct complicating the use of extracorporeal membrane oxygenation. Crit Care 2001;5:232-235. 16. Sarazin M, Caumes E, Cohen A, et al. Multiple microembolic border-zone brain infarctions and endomyocardial fibrosis in idiopathic hypereosinophilic syndrome and in Schistosoma mansoni infestation. J Neurol Neurosurg Psychiatry 2004;75:305-307.
M.H. SORGUN ET AL. 17. McMillan HJ, Johnston DL, Doja A. Watershed infarction due to acute hypereosinophilia. Neurology 2008;70:80-82. 18. Perini GF, Kassab C, Bley C, et al. Acute cerebral infarction in watershed distribution in a patient with hypereosinophilic syndrome. Arq Neuropsiquiatr 2009; 67:510-512. 19. Grandiere-Perez L, Caumes E. Corticosteroids for watershed infarction in acute schistosomiasis. Clin Infect Dis 2013;57:918-919. 20. Aida L, Parkhutik V, Tembl JI, et al. Embolism and impaired washout: a possible explanation of borderzone strokes in hypereosinophilic syndrome. J Neurol Sci 2013;325:162-164. 21. Lee SP, Hong CT. Widespread watershed infarct in patient with malignancy-related hypercoagulation. Acta Neurol Taiwan 2012;21:49-50. 22. Gordhan A, Hudson BK. Acute watershed infarcts with global cerebral hypoperfusion in symptomatic CADASIL. Radiol Case 2013;7:8-15. 23. Bartynski WS, Grabb BC, Zeigler Z, et al. Watershed imaging features and clinical vascular injury in cyclosporin A neurotoxicity. J Comput Assist Tomogr 1997; 21:872-880. 24. Ay H, Furie KL, Singhal A, et al. An evidence-based causative classification system for acute ischemic stroke. Ann Neurol 2005;58:688-697. 25. Pollanen MS, Deck JH. The mechanism of embolic watershed infarction: experimental studies. Can J Neurol Sci 1990;17:395-398. 26. Masuda J, Yutani C, Ogata J, et al. Atheromatous embolism in the brain: a clinicopathologic analysis of 15 autopsy cases. Neurology 1994;44:1231-1237. 27. Arakawa S, Minematsu K, Hirano T, et al. Topographic distribution of misery perfusion in relation to internal and superficial borderzones. AJNR Am J Neuroradiol 2003;24:427-435. 28. Mangla R, Kolar B, Almast J, et al. Border zone infarcts: pathophysiologic and imaging characteristics. Radiographics 2011;31:1201-1214.