- Email: [email protected]
Two different clinical entities of small vessel occlusion in TOAST classification
Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Two different clinical entities of small vessel occlusion in TOAST classification Dong-Eun Kim a , Min-Ji Choi a,b , Joon-Tae Kim a,b,∗ , Jane Chang a , Man-Seok Park a , Kang-Ho Choi a , Dong-Seok Oh a , Seung-Han Lee a , Ki-Hyun Cho a a b
Department of Neurology, Cerebrovascular Center, Chonnam National University Hospital, Gwangju, Republic of Korea Research Institute of Medical Sciences, Chonnam National University, Gwangju, Republic of Korea
a r t i c l e
i n f o
Article history: Received 16 January 2013 Received in revised form 15 February 2013 Accepted 22 March 2013 Available online 20 April 2013 Keywords: Small vessel occlusion Lacunar infarction Early neurological deterioration Branch-atheromatous infarct Small deep infarcts
a b s t r a c t Background: Small deep infarcts might be classified into 2 types: lacunar and branchatheromatous infarcts. However, since their initial description, small deep infarcts were still regarded as the same category of the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification, small vessel occlusion (SVO). We hypothesized that the 2 types of small deep infarcts would be distinct clinical entities. This study was conducted to investigate the clinical characteristics in the 2 groups of patients according to lesion pattern and combined atherosclerotic diseases. Methods: We included patients with small deep infarcts in the subcortical area. The patients were divided into 2 groups: (1) island lesions and (2) linear lesions on coronal diffusion weighted imaging. The status of the relevant artery was categorized as no stenosis, non-significant (<50% of luminal narrowing) and significant (≥50% of luminal narrowing). We compared the clinical and imaging characteristics of two lesion types according to various arterial status. Results: This study analyzed a total of 248 patients. Independent factors for island lesions on coronal DWI were male, severe leukoaraiosis, microbleeds, abnormal glycated hemoglobin (HbA1C ), and abnormal estimated glomerular filtration ratio (eGFR) adjusted by age, sex, and inititial National Institutes of Health Stroke Scale. In addition, in patients without significant relevant arterial stenosis, island lesion patterns were more frequently associated with severe periventricular white matter hyperintensity, diabetes mellitus, abnormal eGFR and abnormal HbA1C than linear lesion patterns. Conclusion: This study demonstrated that SVO of TOAST classifications had different imaging and clinical characteristics according to the lesion patterns of coronal imaging. It suggests that two types of SVO should be regarded as the different categories of stroke classification. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Whereas the term lacunar infarction is associated with uncertain cause or mechanism, small deep infarct can be used as a general term for perforating artery disease [1]. Fisher distinguished 2 causes of this type of stroke: (1) lipohyalinosis, which mainly occurs in the distal or middle perforating artery; and (2) microatheromatous disease, which mainly occurs in patients with larger, usually single symptomatic lacunae [2]. In addition, Caplan drew his attention to the branch-atheromatous cause of small-vessel obstruction [3]. The differentiation between these 2 entities is of clinical importance, because they might involve distinct etiologies, consequences
∗ Corresponding author at: Department of Neurology, Chonnam National University Medical School, 8 Hak-dong, Dong-ku, Gwangju 501-757, Republic of Korea. Tel.: +82 62 220 5623; fax: +82 62 228 3461. E-mail address: [email protected] (J.-T. Kim). 0303-8467/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.clineuro.2013.03.005
and management methods. However, since their initial description, knowledge of the distinctions between these 2 types of small deep infarcts is still limited. Furthermore, small deep infarcts may be regarded as the same category, small vessel occlusion (SVO) of stroke classification. Recently, Nah et al. [4] reported that small subcortical infarctions have pathogeneses according to lesion location in axial images and the presence of parent artery disease. However, axial plane images cannot elucidate the mechanisms of small deep infarcts in the subcortical area. Instead, coronal MRI can be used to determine whether ischemic lesions in the subcortical area are isolated in the parenchyma or are linear and extend to the basal surface (Fig. 1) [3,5]. In general, the prognosis of patients with small deep infarcts is variable [2]. In previous studies, a significant proportion (27–62%) of small deep infarcts showed progressive motor deficits (PMDs), which lead to significant disabilities and a high recurrence rate, within several days after first admission [2,6,7]. It is important to
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
1687
Fig. 1. MRI findings of isolated and linear lesion patterns. (I-A) Island lesion patterns. Axial DWI (A) shows a lesion with maximum diameters <20 mm; the lesion on coronal DWI (B) is isolated in the parenchyma. Fluid attenuated inversion recovery (FLAIR) images (C) show severe white matter hyperintensities. MRA (D) shows normal intracranial vasculatures. (I-B) Linear lesion patterns. Axial DWI (A) shows a lesion with maximum diameters <20 mm, the lesion has a linear pattern extending to the basal surface on coronal DWI (B). FLAIR images (C) shows no white matter hyperintensity and MRA (D) shows also normal vasculature.
predict whether small deep infarcts are progressive at the initial diagnosis so that the appropriate treatment can be provided. We hypothesized that small deep infarcts as classified by underlying pathologies would belong to different clinical entities and show different early outcomes. We expected that linear lesion patterns on coronal DWI would more reflect the branch-atheromatous infarcts and be associated with neurological deterioration. Therefore, this study was conducted to investigate the clinical characteristics and early neurological deterioration (END) in the 2 groups of patients according to lesion pattern and combined atherosclerotic diseases.
2. Materials and methods 2.1. Subjects This study was a retrospective study of ischemic stroke patients admitted to our tertiary stroke center between October 2007 and
October 2011. We previously reported a study with a portion of the patients included in this study [5]. Thus, the study methods are similar to those of the previous study. Briefly, we consecutively collected data from a set of patients who were diagnosed with small deep infarcts by axial magnetic resonance (MR) imaging. Patients who met the following criteria were included: (1) those with acute stroke within 48 h of symptom onset, (2) those with small deep infarcts in the subcortical area (lesions with a maximum diameter of <20 mm on axial diffusion-weighted imaging [DWI]) but not in brainstem, and (3) those undergoing MR imaging studies. The following patients were excluded from the study: (1) those with other etiologies of the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classifications, such as vasculitis, Moyamoya disease and cancer-related stroke; (2) those with multiple lesions on DWI; and (3) those who underwent incomplete imaging evaluations. This study was approved by the Institutional Review Board of Chonnam National University Hospital. Written informed consent was not obtained because of the retrospective study.
1688
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
2.2. MR imaging assessment According to our stroke imaging protocol, patients underwent emergency MR imaging immediately after admission. The MR imaging protocol included axial and coronal DWI, gradient-echo (GRE) imaging, FLAIR and time-of-flight MR angiography. MR imaging parameters were previously described in detail [5]. The MR images were analyzed by 2 neurologists (J.-T.K. and D.-E.K.) who were blind to clinical data. Discrepancies were resolved by consensus meeting. Lesion patterns on DWI were analyzed in terms of shape, extension to the surface, and continuity. The patients were divided into 2 groups: (1) those with island lesions, (Fig. 1) which were defined as lesions of round or oval shape, restricted to the parenchyma (basal ganglia or corona radiata), and measuring <20 mm on coronal DWI and (2) those with linear lesions, (Fig. 1) which were defined as linear hyperintense lesions extending to the basal surface on coronal DWI. The stenosis status of the relevant artery was categorized as normal, non-significant (<50% of luminal narrowing) and significant (≥50% of luminal narrowing). In cases with carotid stenosis, the degree of stenosis was calculated by NASCET criteria. Relevant arterial stenosis (RAS) represented a clinical condition with any stenosis (non-significant and significant) relevant to an ischemic lesion. Any extracranial and intracranial cerebral artery atherosclerotic stenoses except for those related to current lesions on DWI were also assessed. The presence of previous lesions and lacunae were assessed as well. Microbleeds on GRE images were defined as rounded areas of marked and homogeneous signal losses which were not located in the sulcal area. Based on FLAIR imaging results, leukoaraiosis was graded from 0 to 3 on Fazeka’s scale for white matter changes, with scores of 2 and 3 considered to be severe leukoaraiosis [8]. 2.3. Demographic and clinical assessment Demographic, clinical and laboratory data were collected from a prospectively collected stroke registry. The following stroke risk factors were identified: age, sex, current cigarette smoking defined as cigarette smoking within the last 5 years, hypertension, diabetes mellitus, dyslipidemia, a history of previous stroke, and homocysteinemia. All patients underwent laboratory tests during admission, including high sensitivity C-reactive protein, glycated hemoglobin and estimated glomerular filtration rate (eGFR). Abnormal glycated hemoglobin (HbA1C ) was defined as an increase in HbA1C >6.4 mg/dL. The eGFR was estimated by the recommended MDRD Study equation according to the NKF’s KDOQI clinical practice guidelines [9]. Abnormal eGFR was defined as a moderate-to-severe decrease in GFR (<60 ml/min/1.73 m2 ). We assessed neurological status at admission and on each hospital day by using the NIHSS score. Any early neurological deterioration (END) was defined as any deterioration in neurological functions from baseline. END was defined as an increase in the NIHSS score by 2 or more points between hospital days 0 and 5. We used the combined definitions of ENDs based on previous studies because we chose subjects who were within 48 h of symptom onset [10,11]. 2.4. Statistics analysis The data is presented as means ± SD or as the frequencies of categorical variables. The 2 test or Fisher’s exact test was used for categorical variables, and the Mann–Whitney U test was used for continuous variables in univariate analysis. We compared clinical and radiological characteristics between the island lesion and linear lesion groups according to the status of arterial diseases. Multiple logistic regression analysis was performed to evaluate the independent factors of island lesion patterns and END. The variables tested in a multivariate logistic regression model were
Table 1 Clinical and imaging characteristics of the patients.
Age (year, mean ± SD) Sex (male, n, %) Onset to visit time (hours, mean ± SD) Risk factors (n, %) Hypertension Diabetes mellitus Atrial fibrillation Dyslipidemia Smoking Previous stroke Homocysteinemia Initial NIHSS scores (med, IQR) Laboratory findings (mean, SD) High sensitivity CRP (mg/dl) Total cholesterol (mg/dl) LDL-cholesterol (mg/dl) BUN (mg/dl) Creatinine (mg/dl) Estimated GFR (ml/min/1.73 m2 ) eGFR < 60 ml/min/1.73 m2 ) (n, %) HbA1C (>6.4 mg/dl) (n, %) Initial blood glucose (mg/dl) White matter hyperintensity (n, %) PVWMH grade > 1 DWMH grade > 1 Lacunae (n, %) Previous lesions (n, %) Microbleeds (n, %) Axial lesion diameter (mm, mean ± SD) Relevant arterial stenosis (RAS) (n, %) Significant RAS Non-significant RAS Any END END (NIHSS > 2)
Island lesion (n = 124)
Linear lesion (n = 124)
p
66.01 ± 11.78 82 (66.9) 13.7 ± 15.42
65.58 ± 12.35 58 (46.8) 14.1 ± 16.80
0.765 0.002 0.844
87 (70.2) 40 (32.3) 7 (5.6) 51 (41.1) 38 (30.6) 16 (12.9) 6 (4.8) 2.0 (3.0)
75 (60.5) 29 (23.4) 7 (5.6) 49 (39.5) 32 (25.8) 12 (9.7) 4 (3.2) 2.0 (2.0)
0.142 0.158 >0.999 0.897 0.481 0.548 0.749 0.474
0.49 ± 2.09 194.3 ± 42.8 125.9 ± 37.3 16.0 ± 7.3 0.97 ± 0.74 82.9 ± 23.2
0.56 ± 1.72 200.7 ± 36.8 129.2 ± 34.8 13.9 ± 5.0 0.75 ± 0.28 90.4 ± 17.5
0.768 0.211 0.483 0.011 0.002 0.005
19 (15.3)
7 (5.6)
0.011
39 (31.5) 141.2 ± 50.3
24 (19.4) 135.9 ± 44.7
0.041 0.387
68 (54.8) 40 (32.3) 38 (30.6) 26 (21.0) 32 (25.8) 11.37 ± 3.73
39 (31.5) 31 (25.0) 39 (31.5) 21 (16.9) 17 (13.7) 14.09 ± 3.76
<0.001 0.261 >0.999 0.517 0.025 <0.001 0.006
13 (10.5) 29 (23.4) 25 (20.2) 6 (4.8)
16 (12.9) 47 (37.9) 45 (36.3) 21 (16.9)
0.007 0.004
CRP, C-reactive protein; GFR, glomerular filtration ratio; HbA1C, glycated hemoglobin; PVWMH, periventricular white matter hyperintensity; DWMH, deep white matter hyperintensity; END, early neurological deterioration.
age, sex and initial NIHSS scores or those with clinical significance and p < 0.2 by univariate analysis. A value of p < 0.05 was considered statistically significant. SPSS for Windows (Version 17.0; SPSS Inc, Chicago, IL, USA) was used for all statistical analyses. 3. Results 3.1. General characteristics A total of 328 patients with subcortical infarction were screened. Of these patients, 80 were excluded: 38 due to the presence of axial DWI lesions exceeding 20 mm, 12 due to incomplete imaging workups, 15 due to other etiologies and concomitant intracranial diseases such as status of aneurismal clipping and hydrocephalus, and 15 due to additional lesions beyond the subcortical area. Ultimately, 248 (140 men, 108 women) patients were included in this study. Of these patients, 86 had been included in our previous study [5]. The mean age of the patients was 65.79 ± 12.04 years. Table 1 shows the general characteristics of the island lesion and linear lesion groups. The number of males was higher, the eGFR values were lower, and the frequency of patients with abnormal HbA1C and eGFR was higher in the island lesion group than in the linear lesion group. Although the frequencies of hypertension and diabetes were higher in the island lesion group than in the linear lesion group, these differences were not statistically significant.
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
1689
4. Discussion
Fig. 2. The comparisons of the values of estimated GFR (eGFR) and the grades of periventricula white matter hyperintensities (PVWMH). PVWMH, periventricular white matter hyperintensity; Post-hoc analysis was performed by the Scheffe test.
In imaging characteristics, severe periventricular white matter hyperintensity (PVWMH) (Fazeka’s grade > 1) and the presence of microbleeds were more frequently detected in the island lesion groups than in the linear lesion groups. After adjustment for age, sex and initial NIHSS scores, factors for island lesion patterns were no END, abnormal HbA1C , abnormal eGFR, severe PVWMH, microbleeds, no RAS, and short axial lesion diameter on DWI (Table 2). 3.2. Different characteristics of two lesion types according to various arterial status In patients without significant RAS (N = 219), male, severe PVWMH and the presence of microbleeds were more frequent, and the frequency of abnormal HbA1c and eGFR, were higher in the island lesion group than in the linear lesion group. However, END was more frequent in the linear lesion group than in the island lesion group. Results were similar for patients with no stenosis (N = 143). However, in patients with RAS (N = 105) and nonsignificant RAS (N = 76), the differences between island lesion and linear lesion were reduced (Table 3). In patients with significant RAS (N = 29), there were no significant differences between two groups, but it may be due to small sample. END was more frequent in the linear lesion group than in the island lesion group, without statistical significance (3/16 vs 0/13, p = 0.232). 3.3. Factors for END and severe PVWMH The linear lesion patterns (OR 4.012; 95% CI 1.550–10.382; p = 0.004 in model 1, 4.055; 95% CI, 1.507–10.909; p = 0.004 in model 2) were independently associated with END after adjustment for age and initial NIHSS scores or variable with clinical significance (age, initial NIHSS scores, diabetes, microbleeds, non-significant RAS and coronal diameter). In addition, lesion diameter on coronal DWI also was independently associated with END after adjustment for models 1 and 2 (Table 4). Supplemental Table e-1 shows the general characteristics of patients with END and those without. In addition, severe PVWMH was independently associated with island lesion (adjusted OR, 3.750; 95% CI, 1.903–7.388; p < 0.001, supplemental table e-2). The grades of PVWMH were related with the values of eGFR (Fig. 2).
This study showed that island lesion patterns of small deep infarcts have imaging characteristics closer to small vessel diseases compared to linear lesion patterns as well as less frequent END. Previously observed characteristics in small vessel diseases, such as severe PVWMH, numerous microbleeds, abnormal eGFR and abnormal HbA1C , were more frequently found in the island lesion group than in the linear lesion group. In addition, lesion diameter on coronal DWI, while there were small differences in axial images, was an independent predictor for END in small deep infarcts. These characteristics were more obvious when significant arterial stenosis and RAS were absent. Therefore, our results suggest that the pathogeneses of small deep infarcts may differ according to lesion patterns on coronal DWI. Fisher [2] divided small deep infarcts in the subcortical into lacunar infarctions and branch-atheromatous diseases. However, despite previously mentioned differences, small deep infarcts have been regarded as a homogeneous entity. According to the TOAST classification, SVO cannot distinguish between branchatheromatous infarction and lacunar infarction [12]. Although many efforts have been made to distinguish between various small deep infarcts, previous studies have only evaluated a limited number of patients and have not assessed the relationship between lacunar infarcts and branch-atheromatous diseases [13–15]. Therefore, depending on which methods are used to distinguish between various entities of small deep infarcts, they might be associated with different clinical characteristics and outcomes. In addition, combined atherosclerotic diseases, such as large artery disease and parent artery disease, may obscure the differences between 2 distinct entities of small deep infarcts, because of the common risk factors between lacunar infarcts and large artery atherosclerosis. Our results seem to agree with those of previous studies on pathology. Although linear lesion and island lesion patterns on coronal DWI without significant arterial stenosis have currently been considered to be SVO according to the TOAST classification, they had distinct clinical characteristics and different early clinical courses in our study. Patients with island lesion patterns had higher frequencies of microbleeds and more severe PVWMH than those with linear lesion patterns. Generally, if linear lesion patterns represent a more severe type of SVO than island lesion patterns, the results should have been the opposite of what they were, and if it were under the same mechanism but based simply on lesion size, these differences would not have arisen. These results support the hypothesis that island lesion patterns are closely related to PVWMH and microbleeds and that pathomechanisms may be different between island and liner lesion patterns. Furthermore, abnormal HbA1C and eGFR were more frequently observed in patients with island lesion patterns than those with linear lesion patterns. Previous studies demonstrated that abnormal eGFR is significantly associated with decreased arterial elasticity in smaller arteries than in larger arteries [16], periventricular hyperintensity [17], and cerebral small vessel diseases [18]. Wardlaw et al. [19] have shown that brain damage arising from small-vessel endothelial leakage is a potential cause of lacunar infarction and WMH. In addition, impaired kidney function is closely associated with the presence of cerebral microbleeds in ischemic stroke [20]. Furthermore, severe PVWMH was independently associated with island lesion type in our study. This finding also supported the hypothesis of distinct pathophysiology according to the lesion patterns in SVO. Pinto et al have also reported a significant association between diabetes and lacunar infarction: a better clinical outcome, and a higher prevalence of the lacunar subtype in diabetic patients [21]. Our study also found that in patients without significant arterial stenosis, island lesion patterns were more frequently associated
1690
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
Table 2 Factors for island lesions after adjustment of age, sex and initial NIHSS scores. Variables
Model 1 OR (95% CI)
p
Model 2 OR (95% CI)
p
Male END Abnormal HbA1 C Decreased eGFR PVWMH grade > 1 Microbleeds Relevant arterial stenosis Axial lesion diameter
2.583 (1.490–4.408) 0.258 (0.098–0.677) 2.097 (1.136–3.869) 2.930 (1.146–7.496) 3.625 (1.991–6.600) 2.515 (1.269–4.985) 0.484 (0.285–0.823) 0.812 (0.750–0.880)
0.001 0.006 0.018 0.025 <0.001 0.008 0.007 <0.001
3.335 (1.750–6.357) 0.295 (0.098–0.889) 1.910 (0.612–5.963) 2.619 (0.829–7.546) 3.471 (1.774–6.791) 1.142 (0.504–2.588) 0.425 (0.228–0.791) 0.821 (0.756–0.893)
<0.001 0.030 0.265 0.104 <0.001 0.750 0.007 <0.001
Model 1 was adjusted for age, male and initial NIHSS scores. Model 2 was adjusted for variables with clinical significance and p < 0.2 in univariate analysis. END, early neurological deterioration; PVWMH, periventricular white matter hyperintensity.
with diabetes mellitus and abnormal HbA1C . However, these characteristics disappear when comparing island lesion patterns and linear lesion patterns in patients with relevant arterial stenosis, suggesting that these characteristics related to small vessel disease may be obscured due to risk factors for more severe atherosclerosis. Furthermore, recent studies have demonstrated that diabetes is not associated with any ischemic stroke subtypes [22,23]. In
our study, we could not show that how much and how some risk factors contributed to the two types of SVO. Because DM is a common risk factor of large artery atherosclerosis and lacunar infarct, the risk factors of linear lesion with arterial stenosis might not be different from those with island lesion without arterial stenosis. Therefore, further investigations are needed to confirm this hypothesis.
Table 3 Characteristics of isolated and linear lesions according to arterial stenosis. No significant RAS (n = 219)
Age Male Risk factors Hypertension Diabetes mellitus Atrial fibrillation Dyslipidemia Smoking Previous stroke Initial NIHSS scores Laboratory findings Decreased eGFR Abnormal HbA1C WMH PVWMH grade > 1 DWMH grade > 1 Lacunes Previous lesions Microbleeds END
No stenosis (n = 143)
Island (n = 111)
Linear (n = 108)
p
Island (n = 82)
Linear (n = 61)
p
65.9 ± 11.5 79 (71.2)
65.6 ± 11.6 50 (46.3)
0.909 <0.001
65.7 ± 11.5 57 (69.5)
64.2 ± 11.9 26 (42.6)
0.451 0.002
78 (70.3) 40 (36.0) 7 (6.3) 43 (38.7) 35 (31.5) 15 (13.5) 2.0 (4.0)
67 (62.0) 25 (23.1) 7 (6.5) 41 (38.0) 28 (25.9) 9 (8.3) 2.0 (2.0)
0.203 0.040 >0.999 >0.999 0.374 0.281 0.726
59 (72.0) 24 (29.3) 3 (3.7) 31 (37.8) 28 (34.1) 11 (13.4) 2.0 (3.0)
39 (63.9) 10 (16.4) 6 (9.8) 23 (37.7) 19 (31.1) 4 (6.6) 2.0 (3.0)
0.364 0.079 0.171 >0.999 0.723 0.271 0.907
17 (15.3) 39 (35.1)
6 (5.6) 20 (18.5)
0.026 0.006
9 (11.0) 26 (31.7)
3 (4.9) 8 (13.1)
0.236 0.010
60 (54.1) 35 (31.5) 36 (32.4) 24 (21.6) 29 (26.1) 6 (5.4)
36 (33.3) 28 (25.9) 34 (31.5) 18 (16.7) 15 (13.9) 18 (16.7)
0.003 0.372 0.886 0.393 0.028 0.009
44 (53.7) 25 (30.5) 26 (31.7) 16 (19.5) 24 (29.3) 2 (2.4)
16 (26.2) 10 (16.4) 18 (29.5) 11 (18.0) 7 (11.5) 10 (16.4)
0.001 0.076 0.855 >0.999 0.013 0.004
RAS (n = 105)
Age Male Risk factors Hypertension Diabetes mellitus Atrial fibrillation Dyslipidemia Smoking Previous stroke Initial NIHSS scores Laboratory findings Decreased eGFR Abnormal HbA1C WMH PVWMH grade > 1 DWMH grade > 1 Lacunes Previous lesions Microbleeds END
Non-significant RAS (n = 76)
Island (n = 42)
Linear (n = 63)
P
Island (n = 29)
Linear (n = 47)
p
66.6 ± 12.4 26 (61.9)
66.9 ± 12.7 32 (50.8)
0.909 0.318
67.0 ± 11.7 22 (75.9)
67.4 ± 11.0 24 (51.1)
0.898 0.052
28 (66.7) 16 (38.1) 4 (9.5) 20 (47.6) 10 (23.8) 5 (11.9) 2.5 (5.0)
36 (57.1) 19 (30.2) 1 (1.6) 26 (41.3) 13 (20.6) 8 (12.7) 3.0 (2.0)
0.415 0.408 0.155 0.552 0.811 >0.999 0.382
19 (65.5) 16 (55.2) 4 (13.8) 12 (41.4) 7 (24.1) 4 (13.8) 3.0 (4.5)
28 (59.6) 15 (31.9) 1 (2.1) 18 (38.3) 9 (19.1) 5 (10.6) 3.0 (2.0)
0.638 0.057 0.067 0.813 0.773 0.725 0.854
10 (23.8) 13 (31.0)
4 (6.3) 16 (25.4)
0.017 0.656
8 (27.6) 13 (44.8)
3 (6.4) 12 (25.5)
0.017 0.131
24 (57.1) 15 (35.7) 12 (28.6) 10 (23.8) 8 (19.0) 4 (9.5)
23 (36.5) 21 (33.3) 21 (33.3) 10 (15.9) 10 (15.9) 11 (17.5)
0.046 0.836 0.671 0.323 0.793 0.394
16 (55.2) 10 (34.5) 10 (34.5) 8 (27.6) 5 (17.2) 4 (13.8)
20 (42.6) 18 (38.3) 16 (34.0) 7 (14.9) 8 (17.0) 8 (17.0)
0.347 0.810 >0.999 0.237 >0.999 >0.999
‘No significant RAS’ represented a clinical condition without any significant arterial stenosis (>50% of the luminal narrowing) relevant to an ischemic lesion. ‘No stenosis’ represents a clinical condition without significant and non-significant arterial stenosis. ‘RAS’ represented a clinical condition with any arterial stenosis relevant to an ischemic lesion. ‘Non-significant RAS’represented a clinical condition with less than 50% of luminal narrowing of relevant artery.
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
1691
Table 4 Independent factors for early neurological deterioration by multivariate logistic regression analysis.
Lesion diameters on coronal DWI Age Initial NIHSS scores Diabetes mellitus Non-significant RAS Microbleeds
Model 1 OR (95% CI)
p
Model 2 OR (95% CI)
P
1.082 (1.018–1.068) 1.028 (0.989–1.068) 1.031 (0.875–1.215) NA NA NA
0.012 0.164 0.715 NA NA NA
1.079 (1.012–1.151) 1.032 (0.991–1.074) 1.038 (0.873–1.234) 2.073 (0.773–5.370) 1.587 (0.629–4.001) 0.367 (0.078–1.729)
0.021 0.133 0.673 0.150 0.328 0.205
A linear lesion was also an independent factor for END irrespective of the model type: model 1 adjusted for age and initial NIHSS scores (OR, 4.012; 95% CI, 1.550–10.382; p = 0.004) and model 2 adjusted for variables with p < 0.2 in univariate analysis (OR 4.055; 95% CI, 1.507–10.909; p = 0.006). Model 1 was adjusted for age and initial NIHSS scores. Model 2 was adjusted for variables with clinical significance and p < 0.2 in univariate analysis (age, initial NIHSS scores, diabetes, microbleeds, non-significant arterial stenosis and lesion diameters on coronal DWI).
Our results showed that END was observed more frequently in the linear lesion group than in the island lesion group. In addition, lesion diameter on coronal DWI was independently associated with END by multivariate analysis. Our previous study also showed the similar results and Bang et al described that striatocapsular infarction is frequently associated with PMDs and that such patients have more frequent MCA stenosis [24,25]. These results were similar to ours which reported that patients with linear lesion patterns had higher frequencies of RAS and END than those with island lesion patterns.5 Linear lesion patterns may be more similar to large artery atherosclerosis of the stroke subtypes in early clinical courses. However, in our study, no laboratory data support the consideration that linear lesions are similar to large artery atherosclerosis. Further prospective study will be needed to confirm our hypothesis. Furthermore, this suggests that early treatment strategies for patients with linear lesion patterns may differ from those for patients with island lesion patterns. We hypothesized that statins may be more efficacious in patients with linear lesion than in those with isolated lesion. Angiotensin converting enzyme inhibitor may be more important for protection of renal function, and due to bleeding tendency (associated with microbleeds and WMH) antithrombotics may have to be used more cautiously in patients with isolated lesion than in those with linear lesion. This study had several limitations, including the retrospective design and the modest sample size from a single tertiary center. The retrospective design may have introduced bias toward END as these patients would require more medical attention. In addition, this study only included patients from a single tertiary stroke center. Since some patients with lacunar stroke do not visit the hospital if the stroke is mild, this factor could also introduce bias in the study. In addition, the differences in long-term outcomes, recurrence rates and cognitive functions need to be investigated for a complete understanding of patients with island lesion or linear lesion patterns. Therefore, prospective observational studies are needed. In conclusion, the results of this study suggest that island lesion patterns of small deep infarcts may have similar characteristics to lacunar infarcts, such as severe PVWMH, numerous microbleeds, abnormal eGFR and abnormal HbA1C, and that linear lesion patterns may have a higher rate of END than island lesion patterns. It suggests that two types of SVO should be regarded as the different categories of stroke classification. Therefore, lesion patterns on coronal DWI in small deep infarcts may need different treatment strategies due to different pathomechanisms of each pattern. Competing interests and funding None. Disclosure None.
Author contributions D.E. Kim and M.J. Choi were equally contributed as the first author in this study. Study concept and design: J.T. Kim, D.E. Kim, M.J. Choi. Acquisition of data: J.T. Kim, D.E. Kim, M.J. Choi. Analysis and interpretation of data: J.T. Kim, D.E. Kim, M.S. Park, D.S. Oh, K.H. Choi. Drafting of the manuscript: J.T. Kim, M.J. Choi, D.E. Kim, J. Chang. Critical revision of the manuscript for important intellectual content: M.S. Park, J. Chang, K.H. Cho, D.S. Oh, S.H. Lee, K.H. Choi. Statistical analysis: J.T. Kim, D.E. Kim. Administrative, technical, or material support: M.S. Park, K.H. Cho, S.H. Lee, K.H. Choi, D.S. Oh. Acknowledgements This study was supported by grants of the Korea Health Care Technology R&D project, Ministry of Health and Welfare, Republic of Korea (A102065). This work was supported by a research grant from the Research Institute of Medical Sciences, Chonnam National University (2011CURIMS-DR008). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.clineuro. 2013.03.005. References [1] Caplan LR. Small deep brain infarcts. Stroke 2003;34:653–9. [2] Fisher CM. Lacunar strokes and infarcts: a review. Neurology 1982;32:871–6. [3] Caplan LR. Intracranial branch atheromatous disease: a neglected, understudied, and underused concept. Neurology 1989;39:1246–50. [4] Nah HW, Kang DW, Kwon SU, Kim JS. Diversity of single small subcortical infarctions according to infarct location and parent artery disease: analysis of indicators for small vessel disease and atherosclerosis. Stroke 2010;41:2822–7. [5] Kim JT, Yoon GJ, Park MS, Nam TS, Choi SM, Lee SH, et al. Lesion patterns of small deep infarcts have different clinical and imaging characteristics. European Neurology 2010;63:343–9. [6] Lodder J, Gorsselink EL. Progressive stroke caused by CT-verified small deep infarcts; relation with the size of the infarct and clinical outcome. Acta Neurologica Scandinavica 1985;71:328–30. [7] Nakamura K, Saku Y, Ibayashi S, Fujishima M. Progressive motor deficits in lacunar infarction. Neurology 1999;52:29–33. [8] Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MR signal abnormalities at 1.5 T in Alzheimer’s dementia and normal aging. American Journal of Roentgenology 1987;149:351–6. [9] K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. American Journal of Kidney Diseases 2002;39:S1–266. [10] Kwan J, Hand P. Early neurological deterioration in acute stroke: clinical characteristics and impact on outcome. Quarterly Journal of Medicine 2006;99:625–33. [11] Ois A, Martinez-Rodriguez JE, Munteis E, Gomis M, Rodriguez-Campello A, Jimenez-Conde J, et al. Steno-occlusive arterial disease and early neurological deterioration in acute ischemic stroke. Cerebrovascular Diseases 2008;25:151–6. [12] Adams Jr HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke, definitions for use in a multicenter
1692
[13] [14]
[15]
[16]
[17]
[18]
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993;24:35–41. de Jong G, Kessels F, Lodder J. Two types of lacunar infarcts: further arguments from a study on prognosis. Stroke 2002;33:2072–6. Adachi T, Kobayashi S, Yamaguchi S, Okada K. MRI findings of small subcortical “lacunar-like” infarction resulting from large vessel disease. Journal of Neurology 2000;247:280–5. Arauz A, Murillo L, Cantu C, Barinagarrementeria F, Higuera J. Prospective study of single and multiple lacunar infarcts using magnetic resonance imaging: risk factors, recurrence, and outcome in 175 consecutive cases. Stroke 2003;34:2453–8. Peralta CA, Katz R, Madero M, Sarnak M, Kramer H, Criqui MH, et al. The differential association of kidney dysfunction with small and large arterial elasticity: the multiethnic study of atherosclerosis. American Journal of Epidemiology 2009;169:740–8. Shima H, Ishimura E, Naganuma T, Ichii M, Yamasaki T, Mori K, et al. Decreased kidney function is a significant factor associated with silent cerebral infarction and periventricular hyperintensities. Kidney and Blood Pressure Research 2011;34:430–8. Wada M, Nagasawa H, Iseki C, Takahashi Y, Sato H, Arawaka S, et al. Cerebral small vessel disease and chronic kidney disease (CKD): results of a cross-sectional study in community-based Japanese elderly. Journal of the Neurological Sciences 2008;272:36–42.
[19] Wardlaw JM, Sandercock PA, Dennis MS, Starr J. Is breakdown of the blood-brain barrier responsible for lacunar stroke, leukoaraiosis, and dementia? Stroke 2003;34:806–12. [20] Cho AH, Lee SB, Han SJ, Shon YM, Yang DW, Kim BS. Impaired kidney function and cerebral microbleeds in patients with acute ischemic stroke. Neurology 2009;73:1645–8. [21] Pinto A, Tuttolomondo A, Di Raimondo D, Fernandez P, Licata G. Cerebrovascular risk factors and clinical classification of strokes. Seminars in Vascular Medicine 2004;4:287–303. [22] Jackson C, Sudlow C. Are lacunar strokes really different? A systematic review of differences in risk factor profiles between lacunar and nonlacunar infarcts. Stroke 2005;36:891–901. [23] Jackson CA, Hutchison A, Dennis MS, Wardlaw JM, Lindgren A, Norrving B, et al. Differing risk factor profiles of ischemic stroke subtypes: evidence for a distinct lacunar arteriopathy? Stroke 2010;41:624–9. [24] Bang OY, Joo SY, Lee PH, Joo US, Lee JH, Joo IS, et al. The course of patients with lacunar infarcts and a parent arterial lesion: similarities to large artery vs small artery disease. Archives of Neurology 2004;61: 514–9. [25] Bang OY, Heo JH, Kim JY, Park JH, Huh K. Middle cerebral artery stenosis is a major clinical determinant in striatocapsular small, deep infarction. Archives of Neurology 2002;59:259–63.
Contents lists available at ScienceDirect
Clinical Neurology and Neurosurgery journal homepage: www.elsevier.com/locate/clineuro
Two different clinical entities of small vessel occlusion in TOAST classification Dong-Eun Kim a , Min-Ji Choi a,b , Joon-Tae Kim a,b,∗ , Jane Chang a , Man-Seok Park a , Kang-Ho Choi a , Dong-Seok Oh a , Seung-Han Lee a , Ki-Hyun Cho a a b
Department of Neurology, Cerebrovascular Center, Chonnam National University Hospital, Gwangju, Republic of Korea Research Institute of Medical Sciences, Chonnam National University, Gwangju, Republic of Korea
a r t i c l e
i n f o
Article history: Received 16 January 2013 Received in revised form 15 February 2013 Accepted 22 March 2013 Available online 20 April 2013 Keywords: Small vessel occlusion Lacunar infarction Early neurological deterioration Branch-atheromatous infarct Small deep infarcts
a b s t r a c t Background: Small deep infarcts might be classified into 2 types: lacunar and branchatheromatous infarcts. However, since their initial description, small deep infarcts were still regarded as the same category of the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification, small vessel occlusion (SVO). We hypothesized that the 2 types of small deep infarcts would be distinct clinical entities. This study was conducted to investigate the clinical characteristics in the 2 groups of patients according to lesion pattern and combined atherosclerotic diseases. Methods: We included patients with small deep infarcts in the subcortical area. The patients were divided into 2 groups: (1) island lesions and (2) linear lesions on coronal diffusion weighted imaging. The status of the relevant artery was categorized as no stenosis, non-significant (<50% of luminal narrowing) and significant (≥50% of luminal narrowing). We compared the clinical and imaging characteristics of two lesion types according to various arterial status. Results: This study analyzed a total of 248 patients. Independent factors for island lesions on coronal DWI were male, severe leukoaraiosis, microbleeds, abnormal glycated hemoglobin (HbA1C ), and abnormal estimated glomerular filtration ratio (eGFR) adjusted by age, sex, and inititial National Institutes of Health Stroke Scale. In addition, in patients without significant relevant arterial stenosis, island lesion patterns were more frequently associated with severe periventricular white matter hyperintensity, diabetes mellitus, abnormal eGFR and abnormal HbA1C than linear lesion patterns. Conclusion: This study demonstrated that SVO of TOAST classifications had different imaging and clinical characteristics according to the lesion patterns of coronal imaging. It suggests that two types of SVO should be regarded as the different categories of stroke classification. © 2013 Elsevier B.V. All rights reserved.
1. Introduction Whereas the term lacunar infarction is associated with uncertain cause or mechanism, small deep infarct can be used as a general term for perforating artery disease [1]. Fisher distinguished 2 causes of this type of stroke: (1) lipohyalinosis, which mainly occurs in the distal or middle perforating artery; and (2) microatheromatous disease, which mainly occurs in patients with larger, usually single symptomatic lacunae [2]. In addition, Caplan drew his attention to the branch-atheromatous cause of small-vessel obstruction [3]. The differentiation between these 2 entities is of clinical importance, because they might involve distinct etiologies, consequences
∗ Corresponding author at: Department of Neurology, Chonnam National University Medical School, 8 Hak-dong, Dong-ku, Gwangju 501-757, Republic of Korea. Tel.: +82 62 220 5623; fax: +82 62 228 3461. E-mail address: [email protected] (J.-T. Kim). 0303-8467/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.clineuro.2013.03.005
and management methods. However, since their initial description, knowledge of the distinctions between these 2 types of small deep infarcts is still limited. Furthermore, small deep infarcts may be regarded as the same category, small vessel occlusion (SVO) of stroke classification. Recently, Nah et al. [4] reported that small subcortical infarctions have pathogeneses according to lesion location in axial images and the presence of parent artery disease. However, axial plane images cannot elucidate the mechanisms of small deep infarcts in the subcortical area. Instead, coronal MRI can be used to determine whether ischemic lesions in the subcortical area are isolated in the parenchyma or are linear and extend to the basal surface (Fig. 1) [3,5]. In general, the prognosis of patients with small deep infarcts is variable [2]. In previous studies, a significant proportion (27–62%) of small deep infarcts showed progressive motor deficits (PMDs), which lead to significant disabilities and a high recurrence rate, within several days after first admission [2,6,7]. It is important to
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
1687
Fig. 1. MRI findings of isolated and linear lesion patterns. (I-A) Island lesion patterns. Axial DWI (A) shows a lesion with maximum diameters <20 mm; the lesion on coronal DWI (B) is isolated in the parenchyma. Fluid attenuated inversion recovery (FLAIR) images (C) show severe white matter hyperintensities. MRA (D) shows normal intracranial vasculatures. (I-B) Linear lesion patterns. Axial DWI (A) shows a lesion with maximum diameters <20 mm, the lesion has a linear pattern extending to the basal surface on coronal DWI (B). FLAIR images (C) shows no white matter hyperintensity and MRA (D) shows also normal vasculature.
predict whether small deep infarcts are progressive at the initial diagnosis so that the appropriate treatment can be provided. We hypothesized that small deep infarcts as classified by underlying pathologies would belong to different clinical entities and show different early outcomes. We expected that linear lesion patterns on coronal DWI would more reflect the branch-atheromatous infarcts and be associated with neurological deterioration. Therefore, this study was conducted to investigate the clinical characteristics and early neurological deterioration (END) in the 2 groups of patients according to lesion pattern and combined atherosclerotic diseases.
2. Materials and methods 2.1. Subjects This study was a retrospective study of ischemic stroke patients admitted to our tertiary stroke center between October 2007 and
October 2011. We previously reported a study with a portion of the patients included in this study [5]. Thus, the study methods are similar to those of the previous study. Briefly, we consecutively collected data from a set of patients who were diagnosed with small deep infarcts by axial magnetic resonance (MR) imaging. Patients who met the following criteria were included: (1) those with acute stroke within 48 h of symptom onset, (2) those with small deep infarcts in the subcortical area (lesions with a maximum diameter of <20 mm on axial diffusion-weighted imaging [DWI]) but not in brainstem, and (3) those undergoing MR imaging studies. The following patients were excluded from the study: (1) those with other etiologies of the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classifications, such as vasculitis, Moyamoya disease and cancer-related stroke; (2) those with multiple lesions on DWI; and (3) those who underwent incomplete imaging evaluations. This study was approved by the Institutional Review Board of Chonnam National University Hospital. Written informed consent was not obtained because of the retrospective study.
1688
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
2.2. MR imaging assessment According to our stroke imaging protocol, patients underwent emergency MR imaging immediately after admission. The MR imaging protocol included axial and coronal DWI, gradient-echo (GRE) imaging, FLAIR and time-of-flight MR angiography. MR imaging parameters were previously described in detail [5]. The MR images were analyzed by 2 neurologists (J.-T.K. and D.-E.K.) who were blind to clinical data. Discrepancies were resolved by consensus meeting. Lesion patterns on DWI were analyzed in terms of shape, extension to the surface, and continuity. The patients were divided into 2 groups: (1) those with island lesions, (Fig. 1) which were defined as lesions of round or oval shape, restricted to the parenchyma (basal ganglia or corona radiata), and measuring <20 mm on coronal DWI and (2) those with linear lesions, (Fig. 1) which were defined as linear hyperintense lesions extending to the basal surface on coronal DWI. The stenosis status of the relevant artery was categorized as normal, non-significant (<50% of luminal narrowing) and significant (≥50% of luminal narrowing). In cases with carotid stenosis, the degree of stenosis was calculated by NASCET criteria. Relevant arterial stenosis (RAS) represented a clinical condition with any stenosis (non-significant and significant) relevant to an ischemic lesion. Any extracranial and intracranial cerebral artery atherosclerotic stenoses except for those related to current lesions on DWI were also assessed. The presence of previous lesions and lacunae were assessed as well. Microbleeds on GRE images were defined as rounded areas of marked and homogeneous signal losses which were not located in the sulcal area. Based on FLAIR imaging results, leukoaraiosis was graded from 0 to 3 on Fazeka’s scale for white matter changes, with scores of 2 and 3 considered to be severe leukoaraiosis [8]. 2.3. Demographic and clinical assessment Demographic, clinical and laboratory data were collected from a prospectively collected stroke registry. The following stroke risk factors were identified: age, sex, current cigarette smoking defined as cigarette smoking within the last 5 years, hypertension, diabetes mellitus, dyslipidemia, a history of previous stroke, and homocysteinemia. All patients underwent laboratory tests during admission, including high sensitivity C-reactive protein, glycated hemoglobin and estimated glomerular filtration rate (eGFR). Abnormal glycated hemoglobin (HbA1C ) was defined as an increase in HbA1C >6.4 mg/dL. The eGFR was estimated by the recommended MDRD Study equation according to the NKF’s KDOQI clinical practice guidelines [9]. Abnormal eGFR was defined as a moderate-to-severe decrease in GFR (<60 ml/min/1.73 m2 ). We assessed neurological status at admission and on each hospital day by using the NIHSS score. Any early neurological deterioration (END) was defined as any deterioration in neurological functions from baseline. END was defined as an increase in the NIHSS score by 2 or more points between hospital days 0 and 5. We used the combined definitions of ENDs based on previous studies because we chose subjects who were within 48 h of symptom onset [10,11]. 2.4. Statistics analysis The data is presented as means ± SD or as the frequencies of categorical variables. The 2 test or Fisher’s exact test was used for categorical variables, and the Mann–Whitney U test was used for continuous variables in univariate analysis. We compared clinical and radiological characteristics between the island lesion and linear lesion groups according to the status of arterial diseases. Multiple logistic regression analysis was performed to evaluate the independent factors of island lesion patterns and END. The variables tested in a multivariate logistic regression model were
Table 1 Clinical and imaging characteristics of the patients.
Age (year, mean ± SD) Sex (male, n, %) Onset to visit time (hours, mean ± SD) Risk factors (n, %) Hypertension Diabetes mellitus Atrial fibrillation Dyslipidemia Smoking Previous stroke Homocysteinemia Initial NIHSS scores (med, IQR) Laboratory findings (mean, SD) High sensitivity CRP (mg/dl) Total cholesterol (mg/dl) LDL-cholesterol (mg/dl) BUN (mg/dl) Creatinine (mg/dl) Estimated GFR (ml/min/1.73 m2 ) eGFR < 60 ml/min/1.73 m2 ) (n, %) HbA1C (>6.4 mg/dl) (n, %) Initial blood glucose (mg/dl) White matter hyperintensity (n, %) PVWMH grade > 1 DWMH grade > 1 Lacunae (n, %) Previous lesions (n, %) Microbleeds (n, %) Axial lesion diameter (mm, mean ± SD) Relevant arterial stenosis (RAS) (n, %) Significant RAS Non-significant RAS Any END END (NIHSS > 2)
Island lesion (n = 124)
Linear lesion (n = 124)
p
66.01 ± 11.78 82 (66.9) 13.7 ± 15.42
65.58 ± 12.35 58 (46.8) 14.1 ± 16.80
0.765 0.002 0.844
87 (70.2) 40 (32.3) 7 (5.6) 51 (41.1) 38 (30.6) 16 (12.9) 6 (4.8) 2.0 (3.0)
75 (60.5) 29 (23.4) 7 (5.6) 49 (39.5) 32 (25.8) 12 (9.7) 4 (3.2) 2.0 (2.0)
0.142 0.158 >0.999 0.897 0.481 0.548 0.749 0.474
0.49 ± 2.09 194.3 ± 42.8 125.9 ± 37.3 16.0 ± 7.3 0.97 ± 0.74 82.9 ± 23.2
0.56 ± 1.72 200.7 ± 36.8 129.2 ± 34.8 13.9 ± 5.0 0.75 ± 0.28 90.4 ± 17.5
0.768 0.211 0.483 0.011 0.002 0.005
19 (15.3)
7 (5.6)
0.011
39 (31.5) 141.2 ± 50.3
24 (19.4) 135.9 ± 44.7
0.041 0.387
68 (54.8) 40 (32.3) 38 (30.6) 26 (21.0) 32 (25.8) 11.37 ± 3.73
39 (31.5) 31 (25.0) 39 (31.5) 21 (16.9) 17 (13.7) 14.09 ± 3.76
<0.001 0.261 >0.999 0.517 0.025 <0.001 0.006
13 (10.5) 29 (23.4) 25 (20.2) 6 (4.8)
16 (12.9) 47 (37.9) 45 (36.3) 21 (16.9)
0.007 0.004
CRP, C-reactive protein; GFR, glomerular filtration ratio; HbA1C, glycated hemoglobin; PVWMH, periventricular white matter hyperintensity; DWMH, deep white matter hyperintensity; END, early neurological deterioration.
age, sex and initial NIHSS scores or those with clinical significance and p < 0.2 by univariate analysis. A value of p < 0.05 was considered statistically significant. SPSS for Windows (Version 17.0; SPSS Inc, Chicago, IL, USA) was used for all statistical analyses. 3. Results 3.1. General characteristics A total of 328 patients with subcortical infarction were screened. Of these patients, 80 were excluded: 38 due to the presence of axial DWI lesions exceeding 20 mm, 12 due to incomplete imaging workups, 15 due to other etiologies and concomitant intracranial diseases such as status of aneurismal clipping and hydrocephalus, and 15 due to additional lesions beyond the subcortical area. Ultimately, 248 (140 men, 108 women) patients were included in this study. Of these patients, 86 had been included in our previous study [5]. The mean age of the patients was 65.79 ± 12.04 years. Table 1 shows the general characteristics of the island lesion and linear lesion groups. The number of males was higher, the eGFR values were lower, and the frequency of patients with abnormal HbA1C and eGFR was higher in the island lesion group than in the linear lesion group. Although the frequencies of hypertension and diabetes were higher in the island lesion group than in the linear lesion group, these differences were not statistically significant.
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
1689
4. Discussion
Fig. 2. The comparisons of the values of estimated GFR (eGFR) and the grades of periventricula white matter hyperintensities (PVWMH). PVWMH, periventricular white matter hyperintensity; Post-hoc analysis was performed by the Scheffe test.
In imaging characteristics, severe periventricular white matter hyperintensity (PVWMH) (Fazeka’s grade > 1) and the presence of microbleeds were more frequently detected in the island lesion groups than in the linear lesion groups. After adjustment for age, sex and initial NIHSS scores, factors for island lesion patterns were no END, abnormal HbA1C , abnormal eGFR, severe PVWMH, microbleeds, no RAS, and short axial lesion diameter on DWI (Table 2). 3.2. Different characteristics of two lesion types according to various arterial status In patients without significant RAS (N = 219), male, severe PVWMH and the presence of microbleeds were more frequent, and the frequency of abnormal HbA1c and eGFR, were higher in the island lesion group than in the linear lesion group. However, END was more frequent in the linear lesion group than in the island lesion group. Results were similar for patients with no stenosis (N = 143). However, in patients with RAS (N = 105) and nonsignificant RAS (N = 76), the differences between island lesion and linear lesion were reduced (Table 3). In patients with significant RAS (N = 29), there were no significant differences between two groups, but it may be due to small sample. END was more frequent in the linear lesion group than in the island lesion group, without statistical significance (3/16 vs 0/13, p = 0.232). 3.3. Factors for END and severe PVWMH The linear lesion patterns (OR 4.012; 95% CI 1.550–10.382; p = 0.004 in model 1, 4.055; 95% CI, 1.507–10.909; p = 0.004 in model 2) were independently associated with END after adjustment for age and initial NIHSS scores or variable with clinical significance (age, initial NIHSS scores, diabetes, microbleeds, non-significant RAS and coronal diameter). In addition, lesion diameter on coronal DWI also was independently associated with END after adjustment for models 1 and 2 (Table 4). Supplemental Table e-1 shows the general characteristics of patients with END and those without. In addition, severe PVWMH was independently associated with island lesion (adjusted OR, 3.750; 95% CI, 1.903–7.388; p < 0.001, supplemental table e-2). The grades of PVWMH were related with the values of eGFR (Fig. 2).
This study showed that island lesion patterns of small deep infarcts have imaging characteristics closer to small vessel diseases compared to linear lesion patterns as well as less frequent END. Previously observed characteristics in small vessel diseases, such as severe PVWMH, numerous microbleeds, abnormal eGFR and abnormal HbA1C , were more frequently found in the island lesion group than in the linear lesion group. In addition, lesion diameter on coronal DWI, while there were small differences in axial images, was an independent predictor for END in small deep infarcts. These characteristics were more obvious when significant arterial stenosis and RAS were absent. Therefore, our results suggest that the pathogeneses of small deep infarcts may differ according to lesion patterns on coronal DWI. Fisher [2] divided small deep infarcts in the subcortical into lacunar infarctions and branch-atheromatous diseases. However, despite previously mentioned differences, small deep infarcts have been regarded as a homogeneous entity. According to the TOAST classification, SVO cannot distinguish between branchatheromatous infarction and lacunar infarction [12]. Although many efforts have been made to distinguish between various small deep infarcts, previous studies have only evaluated a limited number of patients and have not assessed the relationship between lacunar infarcts and branch-atheromatous diseases [13–15]. Therefore, depending on which methods are used to distinguish between various entities of small deep infarcts, they might be associated with different clinical characteristics and outcomes. In addition, combined atherosclerotic diseases, such as large artery disease and parent artery disease, may obscure the differences between 2 distinct entities of small deep infarcts, because of the common risk factors between lacunar infarcts and large artery atherosclerosis. Our results seem to agree with those of previous studies on pathology. Although linear lesion and island lesion patterns on coronal DWI without significant arterial stenosis have currently been considered to be SVO according to the TOAST classification, they had distinct clinical characteristics and different early clinical courses in our study. Patients with island lesion patterns had higher frequencies of microbleeds and more severe PVWMH than those with linear lesion patterns. Generally, if linear lesion patterns represent a more severe type of SVO than island lesion patterns, the results should have been the opposite of what they were, and if it were under the same mechanism but based simply on lesion size, these differences would not have arisen. These results support the hypothesis that island lesion patterns are closely related to PVWMH and microbleeds and that pathomechanisms may be different between island and liner lesion patterns. Furthermore, abnormal HbA1C and eGFR were more frequently observed in patients with island lesion patterns than those with linear lesion patterns. Previous studies demonstrated that abnormal eGFR is significantly associated with decreased arterial elasticity in smaller arteries than in larger arteries [16], periventricular hyperintensity [17], and cerebral small vessel diseases [18]. Wardlaw et al. [19] have shown that brain damage arising from small-vessel endothelial leakage is a potential cause of lacunar infarction and WMH. In addition, impaired kidney function is closely associated with the presence of cerebral microbleeds in ischemic stroke [20]. Furthermore, severe PVWMH was independently associated with island lesion type in our study. This finding also supported the hypothesis of distinct pathophysiology according to the lesion patterns in SVO. Pinto et al have also reported a significant association between diabetes and lacunar infarction: a better clinical outcome, and a higher prevalence of the lacunar subtype in diabetic patients [21]. Our study also found that in patients without significant arterial stenosis, island lesion patterns were more frequently associated
1690
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
Table 2 Factors for island lesions after adjustment of age, sex and initial NIHSS scores. Variables
Model 1 OR (95% CI)
p
Model 2 OR (95% CI)
p
Male END Abnormal HbA1 C Decreased eGFR PVWMH grade > 1 Microbleeds Relevant arterial stenosis Axial lesion diameter
2.583 (1.490–4.408) 0.258 (0.098–0.677) 2.097 (1.136–3.869) 2.930 (1.146–7.496) 3.625 (1.991–6.600) 2.515 (1.269–4.985) 0.484 (0.285–0.823) 0.812 (0.750–0.880)
0.001 0.006 0.018 0.025 <0.001 0.008 0.007 <0.001
3.335 (1.750–6.357) 0.295 (0.098–0.889) 1.910 (0.612–5.963) 2.619 (0.829–7.546) 3.471 (1.774–6.791) 1.142 (0.504–2.588) 0.425 (0.228–0.791) 0.821 (0.756–0.893)
<0.001 0.030 0.265 0.104 <0.001 0.750 0.007 <0.001
Model 1 was adjusted for age, male and initial NIHSS scores. Model 2 was adjusted for variables with clinical significance and p < 0.2 in univariate analysis. END, early neurological deterioration; PVWMH, periventricular white matter hyperintensity.
with diabetes mellitus and abnormal HbA1C . However, these characteristics disappear when comparing island lesion patterns and linear lesion patterns in patients with relevant arterial stenosis, suggesting that these characteristics related to small vessel disease may be obscured due to risk factors for more severe atherosclerosis. Furthermore, recent studies have demonstrated that diabetes is not associated with any ischemic stroke subtypes [22,23]. In
our study, we could not show that how much and how some risk factors contributed to the two types of SVO. Because DM is a common risk factor of large artery atherosclerosis and lacunar infarct, the risk factors of linear lesion with arterial stenosis might not be different from those with island lesion without arterial stenosis. Therefore, further investigations are needed to confirm this hypothesis.
Table 3 Characteristics of isolated and linear lesions according to arterial stenosis. No significant RAS (n = 219)
Age Male Risk factors Hypertension Diabetes mellitus Atrial fibrillation Dyslipidemia Smoking Previous stroke Initial NIHSS scores Laboratory findings Decreased eGFR Abnormal HbA1C WMH PVWMH grade > 1 DWMH grade > 1 Lacunes Previous lesions Microbleeds END
No stenosis (n = 143)
Island (n = 111)
Linear (n = 108)
p
Island (n = 82)
Linear (n = 61)
p
65.9 ± 11.5 79 (71.2)
65.6 ± 11.6 50 (46.3)
0.909 <0.001
65.7 ± 11.5 57 (69.5)
64.2 ± 11.9 26 (42.6)
0.451 0.002
78 (70.3) 40 (36.0) 7 (6.3) 43 (38.7) 35 (31.5) 15 (13.5) 2.0 (4.0)
67 (62.0) 25 (23.1) 7 (6.5) 41 (38.0) 28 (25.9) 9 (8.3) 2.0 (2.0)
0.203 0.040 >0.999 >0.999 0.374 0.281 0.726
59 (72.0) 24 (29.3) 3 (3.7) 31 (37.8) 28 (34.1) 11 (13.4) 2.0 (3.0)
39 (63.9) 10 (16.4) 6 (9.8) 23 (37.7) 19 (31.1) 4 (6.6) 2.0 (3.0)
0.364 0.079 0.171 >0.999 0.723 0.271 0.907
17 (15.3) 39 (35.1)
6 (5.6) 20 (18.5)
0.026 0.006
9 (11.0) 26 (31.7)
3 (4.9) 8 (13.1)
0.236 0.010
60 (54.1) 35 (31.5) 36 (32.4) 24 (21.6) 29 (26.1) 6 (5.4)
36 (33.3) 28 (25.9) 34 (31.5) 18 (16.7) 15 (13.9) 18 (16.7)
0.003 0.372 0.886 0.393 0.028 0.009
44 (53.7) 25 (30.5) 26 (31.7) 16 (19.5) 24 (29.3) 2 (2.4)
16 (26.2) 10 (16.4) 18 (29.5) 11 (18.0) 7 (11.5) 10 (16.4)
0.001 0.076 0.855 >0.999 0.013 0.004
RAS (n = 105)
Age Male Risk factors Hypertension Diabetes mellitus Atrial fibrillation Dyslipidemia Smoking Previous stroke Initial NIHSS scores Laboratory findings Decreased eGFR Abnormal HbA1C WMH PVWMH grade > 1 DWMH grade > 1 Lacunes Previous lesions Microbleeds END
Non-significant RAS (n = 76)
Island (n = 42)
Linear (n = 63)
P
Island (n = 29)
Linear (n = 47)
p
66.6 ± 12.4 26 (61.9)
66.9 ± 12.7 32 (50.8)
0.909 0.318
67.0 ± 11.7 22 (75.9)
67.4 ± 11.0 24 (51.1)
0.898 0.052
28 (66.7) 16 (38.1) 4 (9.5) 20 (47.6) 10 (23.8) 5 (11.9) 2.5 (5.0)
36 (57.1) 19 (30.2) 1 (1.6) 26 (41.3) 13 (20.6) 8 (12.7) 3.0 (2.0)
0.415 0.408 0.155 0.552 0.811 >0.999 0.382
19 (65.5) 16 (55.2) 4 (13.8) 12 (41.4) 7 (24.1) 4 (13.8) 3.0 (4.5)
28 (59.6) 15 (31.9) 1 (2.1) 18 (38.3) 9 (19.1) 5 (10.6) 3.0 (2.0)
0.638 0.057 0.067 0.813 0.773 0.725 0.854
10 (23.8) 13 (31.0)
4 (6.3) 16 (25.4)
0.017 0.656
8 (27.6) 13 (44.8)
3 (6.4) 12 (25.5)
0.017 0.131
24 (57.1) 15 (35.7) 12 (28.6) 10 (23.8) 8 (19.0) 4 (9.5)
23 (36.5) 21 (33.3) 21 (33.3) 10 (15.9) 10 (15.9) 11 (17.5)
0.046 0.836 0.671 0.323 0.793 0.394
16 (55.2) 10 (34.5) 10 (34.5) 8 (27.6) 5 (17.2) 4 (13.8)
20 (42.6) 18 (38.3) 16 (34.0) 7 (14.9) 8 (17.0) 8 (17.0)
0.347 0.810 >0.999 0.237 >0.999 >0.999
‘No significant RAS’ represented a clinical condition without any significant arterial stenosis (>50% of the luminal narrowing) relevant to an ischemic lesion. ‘No stenosis’ represents a clinical condition without significant and non-significant arterial stenosis. ‘RAS’ represented a clinical condition with any arterial stenosis relevant to an ischemic lesion. ‘Non-significant RAS’represented a clinical condition with less than 50% of luminal narrowing of relevant artery.
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
1691
Table 4 Independent factors for early neurological deterioration by multivariate logistic regression analysis.
Lesion diameters on coronal DWI Age Initial NIHSS scores Diabetes mellitus Non-significant RAS Microbleeds
Model 1 OR (95% CI)
p
Model 2 OR (95% CI)
P
1.082 (1.018–1.068) 1.028 (0.989–1.068) 1.031 (0.875–1.215) NA NA NA
0.012 0.164 0.715 NA NA NA
1.079 (1.012–1.151) 1.032 (0.991–1.074) 1.038 (0.873–1.234) 2.073 (0.773–5.370) 1.587 (0.629–4.001) 0.367 (0.078–1.729)
0.021 0.133 0.673 0.150 0.328 0.205
A linear lesion was also an independent factor for END irrespective of the model type: model 1 adjusted for age and initial NIHSS scores (OR, 4.012; 95% CI, 1.550–10.382; p = 0.004) and model 2 adjusted for variables with p < 0.2 in univariate analysis (OR 4.055; 95% CI, 1.507–10.909; p = 0.006). Model 1 was adjusted for age and initial NIHSS scores. Model 2 was adjusted for variables with clinical significance and p < 0.2 in univariate analysis (age, initial NIHSS scores, diabetes, microbleeds, non-significant arterial stenosis and lesion diameters on coronal DWI).
Our results showed that END was observed more frequently in the linear lesion group than in the island lesion group. In addition, lesion diameter on coronal DWI was independently associated with END by multivariate analysis. Our previous study also showed the similar results and Bang et al described that striatocapsular infarction is frequently associated with PMDs and that such patients have more frequent MCA stenosis [24,25]. These results were similar to ours which reported that patients with linear lesion patterns had higher frequencies of RAS and END than those with island lesion patterns.5 Linear lesion patterns may be more similar to large artery atherosclerosis of the stroke subtypes in early clinical courses. However, in our study, no laboratory data support the consideration that linear lesions are similar to large artery atherosclerosis. Further prospective study will be needed to confirm our hypothesis. Furthermore, this suggests that early treatment strategies for patients with linear lesion patterns may differ from those for patients with island lesion patterns. We hypothesized that statins may be more efficacious in patients with linear lesion than in those with isolated lesion. Angiotensin converting enzyme inhibitor may be more important for protection of renal function, and due to bleeding tendency (associated with microbleeds and WMH) antithrombotics may have to be used more cautiously in patients with isolated lesion than in those with linear lesion. This study had several limitations, including the retrospective design and the modest sample size from a single tertiary center. The retrospective design may have introduced bias toward END as these patients would require more medical attention. In addition, this study only included patients from a single tertiary stroke center. Since some patients with lacunar stroke do not visit the hospital if the stroke is mild, this factor could also introduce bias in the study. In addition, the differences in long-term outcomes, recurrence rates and cognitive functions need to be investigated for a complete understanding of patients with island lesion or linear lesion patterns. Therefore, prospective observational studies are needed. In conclusion, the results of this study suggest that island lesion patterns of small deep infarcts may have similar characteristics to lacunar infarcts, such as severe PVWMH, numerous microbleeds, abnormal eGFR and abnormal HbA1C, and that linear lesion patterns may have a higher rate of END than island lesion patterns. It suggests that two types of SVO should be regarded as the different categories of stroke classification. Therefore, lesion patterns on coronal DWI in small deep infarcts may need different treatment strategies due to different pathomechanisms of each pattern. Competing interests and funding None. Disclosure None.
Author contributions D.E. Kim and M.J. Choi were equally contributed as the first author in this study. Study concept and design: J.T. Kim, D.E. Kim, M.J. Choi. Acquisition of data: J.T. Kim, D.E. Kim, M.J. Choi. Analysis and interpretation of data: J.T. Kim, D.E. Kim, M.S. Park, D.S. Oh, K.H. Choi. Drafting of the manuscript: J.T. Kim, M.J. Choi, D.E. Kim, J. Chang. Critical revision of the manuscript for important intellectual content: M.S. Park, J. Chang, K.H. Cho, D.S. Oh, S.H. Lee, K.H. Choi. Statistical analysis: J.T. Kim, D.E. Kim. Administrative, technical, or material support: M.S. Park, K.H. Cho, S.H. Lee, K.H. Choi, D.S. Oh. Acknowledgements This study was supported by grants of the Korea Health Care Technology R&D project, Ministry of Health and Welfare, Republic of Korea (A102065). This work was supported by a research grant from the Research Institute of Medical Sciences, Chonnam National University (2011CURIMS-DR008). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.clineuro. 2013.03.005. References [1] Caplan LR. Small deep brain infarcts. Stroke 2003;34:653–9. [2] Fisher CM. Lacunar strokes and infarcts: a review. Neurology 1982;32:871–6. [3] Caplan LR. Intracranial branch atheromatous disease: a neglected, understudied, and underused concept. Neurology 1989;39:1246–50. [4] Nah HW, Kang DW, Kwon SU, Kim JS. Diversity of single small subcortical infarctions according to infarct location and parent artery disease: analysis of indicators for small vessel disease and atherosclerosis. Stroke 2010;41:2822–7. [5] Kim JT, Yoon GJ, Park MS, Nam TS, Choi SM, Lee SH, et al. Lesion patterns of small deep infarcts have different clinical and imaging characteristics. European Neurology 2010;63:343–9. [6] Lodder J, Gorsselink EL. Progressive stroke caused by CT-verified small deep infarcts; relation with the size of the infarct and clinical outcome. Acta Neurologica Scandinavica 1985;71:328–30. [7] Nakamura K, Saku Y, Ibayashi S, Fujishima M. Progressive motor deficits in lacunar infarction. Neurology 1999;52:29–33. [8] Fazekas F, Chawluk JB, Alavi A, Hurtig HI, Zimmerman RA. MR signal abnormalities at 1.5 T in Alzheimer’s dementia and normal aging. American Journal of Roentgenology 1987;149:351–6. [9] K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. American Journal of Kidney Diseases 2002;39:S1–266. [10] Kwan J, Hand P. Early neurological deterioration in acute stroke: clinical characteristics and impact on outcome. Quarterly Journal of Medicine 2006;99:625–33. [11] Ois A, Martinez-Rodriguez JE, Munteis E, Gomis M, Rodriguez-Campello A, Jimenez-Conde J, et al. Steno-occlusive arterial disease and early neurological deterioration in acute ischemic stroke. Cerebrovascular Diseases 2008;25:151–6. [12] Adams Jr HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke, definitions for use in a multicenter
1692
[13] [14]
[15]
[16]
[17]
[18]
D.-E. Kim et al. / Clinical Neurology and Neurosurgery 115 (2013) 1686–1692
clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993;24:35–41. de Jong G, Kessels F, Lodder J. Two types of lacunar infarcts: further arguments from a study on prognosis. Stroke 2002;33:2072–6. Adachi T, Kobayashi S, Yamaguchi S, Okada K. MRI findings of small subcortical “lacunar-like” infarction resulting from large vessel disease. Journal of Neurology 2000;247:280–5. Arauz A, Murillo L, Cantu C, Barinagarrementeria F, Higuera J. Prospective study of single and multiple lacunar infarcts using magnetic resonance imaging: risk factors, recurrence, and outcome in 175 consecutive cases. Stroke 2003;34:2453–8. Peralta CA, Katz R, Madero M, Sarnak M, Kramer H, Criqui MH, et al. The differential association of kidney dysfunction with small and large arterial elasticity: the multiethnic study of atherosclerosis. American Journal of Epidemiology 2009;169:740–8. Shima H, Ishimura E, Naganuma T, Ichii M, Yamasaki T, Mori K, et al. Decreased kidney function is a significant factor associated with silent cerebral infarction and periventricular hyperintensities. Kidney and Blood Pressure Research 2011;34:430–8. Wada M, Nagasawa H, Iseki C, Takahashi Y, Sato H, Arawaka S, et al. Cerebral small vessel disease and chronic kidney disease (CKD): results of a cross-sectional study in community-based Japanese elderly. Journal of the Neurological Sciences 2008;272:36–42.
[19] Wardlaw JM, Sandercock PA, Dennis MS, Starr J. Is breakdown of the blood-brain barrier responsible for lacunar stroke, leukoaraiosis, and dementia? Stroke 2003;34:806–12. [20] Cho AH, Lee SB, Han SJ, Shon YM, Yang DW, Kim BS. Impaired kidney function and cerebral microbleeds in patients with acute ischemic stroke. Neurology 2009;73:1645–8. [21] Pinto A, Tuttolomondo A, Di Raimondo D, Fernandez P, Licata G. Cerebrovascular risk factors and clinical classification of strokes. Seminars in Vascular Medicine 2004;4:287–303. [22] Jackson C, Sudlow C. Are lacunar strokes really different? A systematic review of differences in risk factor profiles between lacunar and nonlacunar infarcts. Stroke 2005;36:891–901. [23] Jackson CA, Hutchison A, Dennis MS, Wardlaw JM, Lindgren A, Norrving B, et al. Differing risk factor profiles of ischemic stroke subtypes: evidence for a distinct lacunar arteriopathy? Stroke 2010;41:624–9. [24] Bang OY, Joo SY, Lee PH, Joo US, Lee JH, Joo IS, et al. The course of patients with lacunar infarcts and a parent arterial lesion: similarities to large artery vs small artery disease. Archives of Neurology 2004;61: 514–9. [25] Bang OY, Heo JH, Kim JY, Park JH, Huh K. Middle cerebral artery stenosis is a major clinical determinant in striatocapsular small, deep infarction. Archives of Neurology 2002;59:259–63.