Journal of Clinical Neuroscience 22 (2015) 554–560

Contents lists available at ScienceDirect

Journal of Clinical Neuroscience journal homepage:

Clinical Study

Predictors of malignant brain edema in middle cerebral artery infarction observed on CT angiography Hoon Kim a, Seon Tak Jin b, Young Woo Kim a, Seong Rim Kim a, Ik Seong Park a, Kwang Wook Jo a,⇑ a b

Department of Neurosurgery, Bucheon St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Republic of Korea Department of Neurosurgery, Pohang Stroke and Spine Hospital, Pochang, Republic of Korea

a r t i c l e

i n f o

Article history: Received 12 May 2014 Accepted 3 August 2014

Keywords: ASPECTS CT angiography Edema Ischemic stroke

a b s t r a c t Patients with middle cerebral artery (MCA) infarction accompanied by MCA occlusion with or without internal carotid artery (ICA) occlusion have a poor prognosis, as a result of brain cell damage caused by both the infarction and by space-occupying and life-threatening edema formation. Multiple treatments can reduce the likelihood of edema formation, but tend to show limited efficacy. Decompressive hemicraniectomy with duroplasty has been promising for improving functional outcomes and reducing mortality, particularly improved functional outcomes can be achieved with early decompressive surgery. Therefore, identifying patients at risk for developing fatal edema is important and should be performed as early as possible. Sixty-four patients diagnosed with major MCA infarction with MCA occlusion within 8 hours of symptom onset were retrospectively reviewed. Early clinical, laboratory, and computed tomography angiography (CTA) parameters were analyzed for malignant brain edema (MBE). Twenty of the 64 patients (31%) had MBE, and the clinical outcome was poor (3 month modified Rankin Scale >2) in 95% of them. The National Institutes of Health Stroke Scale (NIHSS) score, Alberta Stroke Program Early Computed Tomography Score, Clot Burden Score, and Collateral Score (CS) showed statically significant differences in both groups. Multivariable analyses adjusted for age and sex identified the independent predictors of MBE: NIHSS score >18 (odds ratio [OR]: 4.4, 95% confidence interval [CI]: 1.2–16.0, p = 0.023) and CS on CTA 50%, but 18 was higher in the MBE group compared to in the non-MBE group (60% versus 22.7%, OR: 5.1, 95% CI: 1.63–15.93, p = 0.04). The mean time from symptom onset to CTA imaging and medical comorbidities, including diabetes mellitus, hypertension, and hyperlipidemia were similar in both groups. Body temperature and laboratory parameters on admission were balanced across the groups. Electrocardiogram results showing

H. Kim et al. / Journal of Clinical Neuroscience 22 (2015) 554–560


Fig. 2. Computed tomography angiography images were reconstructed in the axial plane at 5 mm thick maximum intensity projections. Examples of the different collateral scores are shown as follows. (A) 0 = absent collateral supply to the occluded middle cerebral artery (MCA) territory. (B) 1 = collateral supply filling 650% of the occluded MCA territory. (C) 2 = collateral supply filling >50%, but 18 3 month mRS 62 (good outcome) >2 (poor outcome)

Non-MBE (n = 44)

MBE (n = 20)

p value

69.2 (12.2) 24 (54.5%)

68.8 (15.8) 14 (70%)

0.911 0.283

13 (29.5%) 27 (61.3%) 9 (20.4%) 6 (13.6%)

5 (25%) 15 (75%) 8 (40%) 7 (35%)

0.773 0.396 0.13 0.09

13.1 (1.8) 9.22 (5.85) 239.86 (76.56) 129.8 (47.3) 5.7 (9) 36.4 (0.3) 153 (137) 12.8 (7.2) 10 (22.7%)

13.3 (1.4) 9.82 (4.59) 207.15 (66.76) 121.2 (45.4) 4.1 (5.5) 36.4 (0.4) 169 (174) 18.5 (6.3) 12 (60%)

0.654 0.659 0.09 0.497 0.454 0.53 0.712 0.003 0.04

20 (45.5%) 24 (54.5%)

1 (5%) 19 (95%)


MBE = malignant brain edema, min = minutes, mRS = modified Rankin Scale, NIHSS= National Institutes of Health Stroke Scale, SD = standard deviation.

Fig. 3. (A) National Institutes of Health Stroke Scale (NIHSS) scores and (B) collateral scores (CS) are represented in a box and whisker plot for statistical comparison between the non-malignant brain edema (MBE) and MBE groups.

Table 2 Comparisons of the radiological features and treatment modalities between the malignant brain edema (MBE) and non-MBE groups

Vessel occlusion ICA + MCA MCA Occlusion side Left ASPECTS (SD) CBS (SD) CS (SD) ASPECTS 67 CBS 66 CS 18 (OR: 4.1) have modest sensitivity and specificity, as well as a high negative predictive value. The ASPECTS was originally developed to semi-quantitatively assess infarct magnitude on CT images and is useful for assessing the extent of ischemic changes within the MCA territory [7]. The baseline ASPECTS correlates inversely with the severity of the NIHSS score and with functional outcome. When the ASPECTS is dichotomized to scores 67 or scores >7, scores 67 indicate more extensive MCA involvement and are correlated with poor functional outcome [14]. In our study, the mean ASPECTS were statistically different between the MBE and non-MBE groups (p = 0.002). The mean ASPECTS of the MBE group was close to 7 points. Therefore, we expected dichotomization of the ASPECTS to 67 and >7 would be predictive of MBE. A larger percentage of patients in the MBE group had ASPECTS of 67 compared to the percentage in the non-MBE group, but this difference was not statistically significant (p = 0.095). We believe these results were due to the ASPECTS being validated for assessing parenchymal ischemic changes including hypoattenuation and swelling. However, Na et al. [15] suggested that hypoattenuation and focal swelling do not accurately predict the extent of the final infarction. Butcher et al. [16] also reported that while focal swelling has similar cerebral blood flow, and cerebral blood volume was increased relative to the normal CT scan area, infarction occurred in only 32% of the regions with focal swelling, while it occurred in all of the hypoattenuated regions. Consequently, it is possible that the ASPECTS overestimates the likelihood of future MBE. In addition, artifact can affect CT images. If a patient is uncooperative, motion artifacts are very likely to occur. Other potential artifacts include streak artifact at the temporal skull base area and age-related periventricular changes. These artifacts can lead to incorrect ASPECT scoring. Additionally, ASPECTS results may be similar for both proximal and distal occlusion sites. The site of occlusion influences treatment results. Therefore, the CBS, which assigns up to 10 points for the presence of contrast opacification in the anterior circulation on CTA images and provides information about clot extent and location, was introduced. Tan et al. [9] and Sillanpaa et al. [17] showed that low CBS were associated with higher admission NIHSS scores and larger infarct volumes. The CBS is also a significant independent predictor of clinical and radiological outcomes in acute MCA ischemic stroke. In our study, as with previous studies, the mean CBS was significantly different between the MBE and non-MBE groups (p = 0.005). However, when the scores were dichotomized to 66 or >6, CBS 66 were predictive of poor clinical outcomes in the


H. Kim et al. / Journal of Clinical Neuroscience 22 (2015) 554–560

above studies. In our study, patients with CBS 66 were more likely to have MBE, but there was no statistical significance. This result suggests that while useful, the CBS is limited because it does not consider residual flow at the site of arterial occlusion nor does it consider pial vessel retrograde filling due to collaterals from the anterior or posterior cerebral arteries. Several studies have established the importance of collateral blood supply in the prediction of stroke outcome. Since irreversible neuronal damage may occur within minutes, collateral flow in patients with acute arterial occlusion infarction can prevent the extension of the infarction. However, it has been difficult to simply and quantifiably measure collateral flow despite its importance. Therefore, evaluating collateral flow by CTA may provide information for assessing permanent ischemic injury in an emergency setting. In our study, a lower baseline CS and a CS

Predictors of malignant brain edema in middle cerebral artery infarction observed on CT angiography.

Patients with middle cerebral artery (MCA) infarction accompanied by MCA occlusion with or without internal carotid artery (ICA) occlusion have a poor...
1MB Sizes 0 Downloads 9 Views