RESEARCH—HUMAN—CLINICAL STUDIES

The Subarachnoid Hemorrhage Early Brain Edema

Score Predicts Delayed Cerebral Ischemia and
Clinical Outcomes
Sung-Ho Ahn, MD∗ ‡ BACKGROUND: Early brain injury (EBI) after subarachnoid hemorrhage (SAH) is an
Jude P. Savarraj, PhD∗ important determinant of clinical outcomes. However, a major hindrance to studies of EBI
Mubashir Pervez, MD∗ is the lack of radiographic surrogate marker.
OBJECTIVE: To propose a scoring system based on early changes in clinically obtained
Wesley Jones, MD§
computed tomography (CT), called the Subarachnoid hemorrhage Early Brain Edema
Jin Park, MD∗
Score (SEBES).
Sang-Beom Jeon, MD, PhD‡ METHODS: Patients with spontaneous aneurysmal SAH and a CT within 24 h of ictus were
Sun U. Kwon, MD, PhD‡ included. We defined SEBES as a scale of 0 to 4 points according to the (1) absence of visible
Tiffany R. Chang, MD∗ sulci caused by effacement of sulci or (2) absence of visible sulci with disruption of the gray–
Kiwon Lee, MD, PhD∗ white matter junction at 2 predetermined levels in each hemisphere. Prognostic value of
the SEBES grade for the prediction of delayed cerebral ischemia (DCI) and unfavorable
Dong H. Kim, MD, PhD§
outcomes was assessed. A separate cohort of patients was used as a validation cohort.
Arthur L. Day, MD, PhD§
RESULTS: Of the 164 subjects in our study, high-grade SEBES (3 or 4 points) was identified
H. Alex Choi, MD∗ in 48 patients (29.3%). CT interobserver reliability of SEBES grades was high with a Kappa
∗
value of 0.89. After adjusting for covariables, SEBES was identified as an independent
Department of Neurology, University of
Texas Health Science Center, Houston, predictor of DCI (OR = 2.24, 95% CI: 1.58–3.17) and unfavorable outcome (OR = 3.45, 95% CI:
Texas; ‡ Department of Neurology, Asan 1.95–6.07). In our validation cohort, 84 subjects showed similar predictive power of SEBES
Medical Center, University of Ulsan
for a prediction of DCI and unfavorable long-term outcome.
College of Medicine, Seoul, Korea; § De-
partment of Neurosurgery, Division of CONCLUSION: SEBES may be a surrogate marker of EBI and predicts DCI and clinical
Neurocritical care, University of Texas outcomes after SAH.
Health Science Center, Houston, Texas
KEY WORDS: Subarachnoid hemorrhage, Prognosis, CT, Early brain injury, Delayed cerebral ischemia
Correspondence:
H. Alex Choi, MD, Neurosurgery 0:1–9, 2017 DOI:10.1093/neuros/nyx364 www.neurosurgery-online.com
6431 Fannin,
MSB 7.154,
focus.3,4 Appropriately, there has been increasing

Houston, TX 77030.
E-mail: Huimahn.A.Choi@uth.tmc.edu
Received, November 16, 2016.
Accepted, June 5, 2017.
Copyright 
C 2017 by the
Congress of Neurological Surgeons
U
ntil recently, subarachnoid hemorrhage
(SAH) research has focused primarily
on preventing vasospasm and delayed
cerebral ischemia (DCI).1 The failure of recent
trials that focused on prevention and treatment
of vasospasm2 has necessitated a shift in research
ABBREVIATIONS: ASPECT, Alberta Stroke Program
Early CT; AUC, area under the curve; CT, computed
tomography; DCI, delayed cerebral ischemia; EBI,
Early brain injury; GCE, global cerebral edema; HH,
Hunt–Hess; IVH, intraventricular hematoma; LOC,
loss of consciousness; mFS, modified Fisher scale;
mRS, modified Rankin Scale; OR, odds ratio; SEBES,
Subarachnoid hemorrhage Early Brain Edema Score;
WFNS, World Federation Neurosurgical Score
Supplemental digital content is available for this article at
www.neurosurgery-online.com.
attention focused on early brain injury (EBI),
defined as injury in the first 72 h after hemor-
rhage.5-8 The concept of EBI is important as
it reflects what is clinically apparent, specifically
that the initial clinical presentation, whether
measured by the Hunt–Hess (HH) score or the
World Federation Neurosurgical Score (WFNS),
is the most important predictor of outcomes.9
Despite EBI’s importance, its pathophysiology
has not been extensively studied. EBI’s mecha-
nisms are likely multifactorial; they begin within
the first few minutes to hours after aneurysmal
rupture and continue for the next several
days.10,11 A significant limitation of EBI research
is the absence of a reliable radiographic surrogate
marker. The previously described characteristic,
called global cerebral edema (GCE), is the
only available radiographic marker of EBI.12,13

NEUROSURGERY VOLUME 0 | NUMBER 0 | 2017 | 1

AHN ET AL
However, its dichotomous nature, the small proportion of
patients with GCE, and qualitative adjudication limit its utility.
Similar to the utility of the Alberta Stroke Program Early CT
(ASPECT) score,14 diffusion weighted image volumes,15 and
computed tomography (CT) perfusion16 in ischemic stroke
research, a readily available and clinically useful surrogate marker
in SAH is needed.
Current radiographic rating scales, the most widely used being
the Fisher scale and the modified Fisher scale (mFS), are mainly
focused on quantifying the amount of blood for the purpose of
predicting vasospasm and DCI.17 However, as we have learned
from recent studies, DCI itself is a complex multifactorial process
that cannot be explained completely by the amount of blood in
the cisternal space and large vessel vasospasm.2,7 Previously, we
have described that sulcal volume quantified using a semiauto-
mated process was associated with clinical outcomes, suggesting
that sulcal volume may be an important marker of EBI.18 The
limitation of this approach is the tedious nature and necessity of
additional radiographic software to calculate sulcal volumes.
In this study, we aim to develop an easily obtainable radio-
graphic marker of EBI that represents early parenchymal changes
and is associated with established clinical markers of EBI, namely
the HH and WFNS, and which can predict DCI and functional
outcomes. We propose a novel radiographic marker of EBI
called the Subarachnoid hemorrhage Early Brain Edema Score
(SEBES). The SEBES is a semiquantitative CT grading scale
loosely modeled after the ASPECT grade for ischemic stroke. We
hypothesize that the SEBES score will reflect the clinical grade
of patients after SAH and additionally will predict occurrence of
DCI as well as clinical outcomes.
METHODS
Study Population
Patients with acute SAH admitted to the Neuroscience Intensive Care
Unit were prospectively enrolled into an observational study over 2 yr,
from March 2010 to March 2012. The study was approved by the Insti-
tutional Review Board, and in all cases, written informed consent was
obtained from the subjects or a surrogate.
The diagnosis of SAH was established by the admission CT scan.
Patients with spontaneous aneurysmal SAH and a CT within 24 h
of ictus were included. Patients with SAH due to trauma, arteri-
ovenous malformation rupture, and patients with isolated cortical SAH
or previous history of brain injury with associated chronic changes on
CT were excluded.
Clinical Variables
Clinical variables included baseline demographic data (age, gender,
past medical and social history), and clinical features at onset (HH
scale,19 WFNS grade,20 loss of consciousness, and symptomatic hydro-
cephalus on presentation). DCI was defined per consensus definition as
“clinical deterioration attributable to vasospasm (clinical vasospasm), or
a new infarct on brain CT related to vasospasm that was not visible
on the admission or immediate postoperative scan (new infarction
attributable to vasospasm), or both.”21 We identified patients with DCI
2 | VOLUME 0 | NUMBER 0 | 2017
only after rigorously excluding other possible causes. Patients were treated
according to available guidelines.22 Clinical outcomes at discharge were
assessed by the modified Rankin Scale (mRS), and unfavorable outcome
was defined as an mRS score of 4 to 6.23
Radiographic Variables
Admission and follow-up CT scans were independently evaluated by
a study neurologist for the amount and location of blood as categorized
by the mFS, the presence and degree of intraventricular hemorrhage as
the IVH score,24 the presence of hydrocephalus,25 and the presence of
cerebral infarction. Our clinical protocol includes performing follow-
up angiograms 7 d after SAH to examine for angiographic vasospasm.
Angiographic vasospasm was defined as narrowing of the arterial vessel
lumen by visual inspection on follow-up cerebral angiography. Severe
vasospasm was defined as narrowing of the arterial vessel lumen exceeding
50% of the normal caliber.
Radiographic Marker of EBI
We defined SEBES as a scale from 0 to 4 points. One point was
assigned for the (1) absence of visible sulci caused by effacement of sulci
or (2) absence of visible sulci with disruption of the gray–white matter
junction at 2 predetermined levels in each hemisphere12 : (a) at the level
of the insular cortex showing the thalamus and basal ganglion above the
basal cistern and (b) at the level of the centrum semiovale above the level
of the lateral ventricle (Figure 1). Each section was scored as a 0 for no
effacement or a 1 for effacement of sulci. Two slices with 2 sides resulted
in a maximum score of 4. We assessed interobserver reliability of the
scoring by comparing the ratings of 2 independent blinded examiners.
For the adjudication of the SEBES when there was a discrepancy between
the 2 independent blinded examiners, an independent third examiner
was used.
Validation Cohort
Our validation cohort comprised patients admitted with SAH after
the initial cohort. We applied the same inclusion criteria. SEBES was
graded routinely and prospectively at weekly consensus conferences,
blinded to clinical outcomes (as CTs were graded prospectively) since
March 2015. We included patients admitted between March 2015 and
June 2016. We identified the prognostic value of the SEBES grade for
prediction of DCI and unfavorable long-term outcome, defined as an
mRS score of 4 to 6 at 3 mo.
Statistical Analysis
Continuous variables were compared using unpaired Student t-tests,
and categorical variables were compared using chi-square or Fisher’s
exact tests. Multivariate logistic regression analyses were performed to
identify independent predictors of high-grade SEBES and to determine
the predictive value of SEBES grade for DCI and unfavorable outcome.
Univariate analysis was performed to select variables for inclusion into
the multivariate analysis using demographic data, clinical features, and
radiographic findings on admission, with the exception of radiographic
hydrocephalus and IVH score, due to their high collinearity to the
presence of symptomatic hydrocephalus and IVH (Table, Supplemental
Digital Content 1). All variables with P < .10 in the univariate analysis
were included in the multivariate logistic regression model. Due to high
collinearity between the HH scale and the WFNS grade, we implemented
2 separate models to identify the predictors of high-grade SEBES and
to test the predictive value of SEBES grade for DCI and unfavorable
www.neurosurgery-online.com
 SUBARACHNOID HEMORRHAGE EARLY BRAIN EDEMA SCORE (SEBES)

FIGURE 1. Grade 4 of SEBES with the effacement of sulci at 2 predetermined levels in each hemisphere (A and B), and grade 2 of SEBES with the effacement sulci
on the left hemisphere (C and D).

TABLE 1. Baseline Characteristics

SEBES grade

Variable High grade (n = 48) Low grade (n = 116) P-value†

Age (yr) 50.9 ± 14.4 52.4 ± 13.5 .55

Gender, male 9 (18.8) 39 (33.6) .06
Risk factors
Hypertension 31 (64.6) 70 (60.3) .61
Diabetes mellitus 3 (6.3) 20 (17.2) .08
Hyperlipidemia 1 (2.1) 11 (9.5) .18
Current smoking 20 (41.7) 64 (55.2) .12
Alcohol 15 (31.3) 48 (41.4) .23
Drug 7 (14.6) 13 (11.2) .55
Clinical features
LOC at ictus 30 (62.5) 25 (21.6) <.01
Symptomatic hydrocephalus 17 (35.4) 34 (29.3) .44
HH grade 4.5 [3.5–5] 2 [2–3] <.01
WFNS grade 4 [2–4.5] 2 [1–3] <.01
Radiographic findings
mFS grade 3 [3–4] 3 [3–3] .02
IVH 34 (70.8) 70 (60.3) .21
Radiographic hydrocephalus 28 (58.3) 49 (42.2) .06
Size of aneurysm 7.2 ± 4.4 6.1 ± 3.7 .10
Therapeutic intervention .07
Clipping 17 (35.4) 44 (37.9)
Coiling 24 (50.0) 67 (57.8)

SEBES, Subarachnoid hemorrhage Early Brain Edema Score; LOC, loss of consciousness; HH, Hunt-Hess; WFNS, World Federation of Neurosurgical Societies; mFS, modified Fisher
Score; IVH, intraventricular hemorrhage.
Variables are presented as mean ± SD, median [interquartile range], or number (%).
†P-values are calculated by Pearson chi-square test or Fisher’s exact test, or Student’s t-test as appropriate.

outcome. The results of the multivariate logistic regression analysis unfavorable outcome, we constructed an ROC curve and calculated the
are reported as odds ratio (OR) at a 95% CI. In addition to testing area under the curve (AUC). A P-value ≤ .05 was considered statisti-
the diagnostic accuracy of SEBES and the traditional grading system cally significant (two-tailed). All statistical analyses were performed using
comprising HH, WFNS, and mFS grade for prediction of DCI and SPSS 17.0 (IBM, Armonk, New York).

NEUROSURGERY VOLUME 0 | NUMBER 0 | 2017 | 3

AHN ET AL
TABLE 2. Multivariable Logistic Regression Analysis Identifying
Predictors of High Grade SEBES
Odds ratio (95% CI)
Clinical variables Model 1
Age, per 1-yr increase 0.99 (0.96–1.02)
Gender, male 0.58 (0.23–1.47)
Diabetes 0.37 (0.09–1.49)
LOC at ictus 3.04 (1.20–7.70)
HH grade, per 1-point increase 1.60 (1.02–2.50)
WFNS grade, per 1-point increase N/A
mFS grade, per 1-point increase 1.17 (0.62–2.21)
Radiographic hydrocephalus 1.58 (0.70–3.57)
SEBES, Subarachnoid hemorrhage Early Brain Edema Score; CI, confidence interval;
LOC, loss of consciousness; HH, Hunt–Hess; WFNS, World Federation of Neurosurgical
Societies; mFS, modified Fisher Score; N/A, not applicable.
Age and variables predictive in the univariate analysis (P < .10) were included in a multi-
variable logistic regression analysis.
RESULTS
Baseline Characteristics
A total of 164 subjects were included in our analysis. The
mean age was 51.9 ± 13.7 yr (range 21–83 yr), and 116 patients
(70.7%) were women. Seven patients (4.3%) had no aneurysms
identified on angiography. Of the 157 patients with an identified
aneurysm, 59 patients had their aneurysm clipped, 93 patients
had embolization with coils, and 5 patients were untreated.
Predictors of High-Grade SEBES
CT interobserver reliability of SEBES grades was high, with a
Kappa value of 0.89. High-grade SEBES (either 3 or 4 points)
was identified in 48 patients (29.3%). Subjects with high-grade
SEBES were more often female (P = .06), had lower diabetes (P =
.08), higher HH grade (P < .01), higher WFNS grade (P < .01),
more frequent history of loss of consciousness (LOC) at ictus (P <
.01), and higher mFS grade on admission CT (P = .02) and radio-
graphic hydrocephalus (P = .06) than low-grade SEBES subjects
(Table 1).
Multivariate logistic regression analyses were performed to
identify independent predictors of high-grade SEBES. In multi-
variate model 1, LOC at ictus (OR = 3.04, 95% CI 1.20–
7.70) and high-grade HH (OR = 1.60, 95% CI 1.02–2.50) were
identified as predictors of high-grade SEBES. In model 2, LOC
at ictus (OR = 3.54, 95% CI 1.50–8.37) and high-grade WFNS
(OR = 1.36, 95% CI 1.02–1.80) were additionally identified as
predictors of high-grade SEBES (Table 2).
Prognostic Value of SEBES for DCI and Unfavorable
Outcomes
DCI occurred in 42 patients (25.6%) mostly within 7 d after
SAH (n = 33 of 42). Death or severe disability (mRS of 4 to 6
4 | VOLUME 0 | NUMBER 0 | 2017
score) at discharge was identified in 53 patients (32.3%). Multi-
variate logistic regression analyses showed that for each point
increase in SEBES, there was a 2.24 (95% CI = 1.58–3.17)
increase in the odds of developing DCI and a 3.45 (95% CI
= 1.95–6.07) or 3.25 (95% CI = 1.88–5.59) increase in the
Model 2 odds of an unfavorable outcomes after adjustment for covariates,
0.99 (0.96–1.02) including the HH or the WFNS grade, respectively (Table 3).
0.59 (0.23–1.49) In addition, old age (OR = 1.14, 95% CI = 1.08–1.21 or OR
0.40 (0.10–1.62) = 1.14, 95% CI = 1.07–1.21) and increasing the HH (OR =
3.54 (1.50–8.37) OR = 3.61, 95% CI = 1.83–7.12) or WFNS grade (OR = 2.18,
N/A 95% CI = 1.43–3.35) were identified as additional predictors for
1.36 (1.02–1.80) unfavorable outcomes.
1.24 (0.66–2.34)
1.39 (0.61–3.19)
For each clinical or radiographic scale, the distribution of
patients in each scale and the ROC curve of each scale for the
diagnosis of DCI and unfavorable outcome are shown in Figure 2.
SEBES and WFNS grading showed a balanced distribution of
patients. The HH score and mFS grade showed a skewed distri-
bution of patients. The AUC of the ROC curve of SEBES for the
prediction of unfavorable outcome and DCI was 0.79 (95% CI
0.72–0.86) and 0.78 (95% CI 0.69–0.86), respectively.
Characteristics and Incidence of DCI According to the
Status of Vasospasm
Angiographic vasospasm occurred in 88 patients (53.7%),
of whom 54 were treated by either intra-arterial injection of
vasodilators or balloon angioplasty, or both, within 7 d after
SAH. Of the 25 subjects with DCI in the high-grade SEBES
group, 16 developed both clinical symptoms and radiographic
lesion within 7 d after SAH. Five of the 17 DCI subjects
in the low-grade SEBES group developed DCI after 7 d of
SAH (Figure 3A). The incidence of DCI increased linearly with
increased severity of the SEBES and vasospasm, with rates being
higher in patients with both high-grade SEBES and presence of
vasospasm (P < .01, Figure 3B). The AUC of the ROC curve
of SEBES for the prediction of angiographic vasospasm was 0.62
(95% CI 0.53–0.70) and similar to the AUC of the ROC curve
of the WFNS, HH, and mFS grades (Figure 3C). Furthermore,
prognostic value of SEBES grade for prediction of DCI and
unfavorable outcome remained significant regardless of treatment
modalities, ie, clipping and coiling (Table, Supplemental Digital
Content 2).
Internal Validation for Prognostic Value of SEBES Grade
in Prospectively Enrolled Patients
A validation cohort was included in the study. Subjects in the
validation cohort were not included in the initial cohort. A total
of 84 subjects were included in our validation analysis. High-
grade SEBES (either 3 or 4 points) was identified in 30 patients
(35.7%). Subjects with high-grade SEBES had a higher HH grade
(P < .01), higher WFNS grade (P < .01), more frequent history
of LOC at ictus (P < .01), and a higher mFS grade on admission
CT (P = .05) and IVH (P < .01), and were more often treated by
www.neurosurgery-online.com

TABLE 3. Unadjusted and Adjusted Odds Ratios for SEBES Grade to Predict DCI and Unfavorable Outcome
Odds ratio (95% CI)
Clinical variables Unadjusted
DCI
SEBES grade, per 1-point increase 2.28 (1.66–3.13)
Unfavorable outcome
SEBES grade, per 1-point increase 2.47 (1.81–3.38)
SEBES, Subarachnoid hemorrhage Early Brain Edema Score; DCI, delayed cerebral ischemia; CI, confidence interval; N/A, not applicable; HH, Hunt-Hess; WFNS, World Federation of
Neurosurgical Societies; mFS, modified Fisher Score; LOC, loss of consciousness.
Age and variables predictive in the univariate analysis (P < .10) were included in a multivariate logistic regression analysis.
a
Adjustments were made for age, WFNS and mFS grade, and radiographic hydrocephalus.
b
Adjustments were made for age, current smoking, alcohol, LOC at ictus, HH and mFS grade, IVH, radiographic hydrocephalus, and size of aneurysm.
c
Adjustments were made for age, current smoking, alcohol, LOC at ictus, WFNS and mFS grade, IVH, radiographic hydrocephalus, and size of aneurysm.
coiling (P < .01) than low-grade SEBES subjects (Table, Supple-
mental Digital Content 3).
DCI occurred in 23 patients (27.4%) within 7 d after SAH.
Death or severe disability (mRS of 4 to 6 score) at 3 mo after
SAH was identified in 36 patients (42.9%). Multivariate logistic
regression analyses showed that for each point increase in SEBES,
there was a 2.35 (95% CI 1.33–4.18) increase in the odds of
developing DCI and a 2.79 (95% CI 1.48–5.25) or 2.33 (95%
CI 1.26–4.30) increase in the odds of an unfavorable long-term
outcome after adjustment for covariates including the HH or
the WFNS grade, respectively (Table, Supplemental Digital
Content 4).
DISCUSSION
Our main finding is that the SEBES is associated with other
clinical markers of EBI, namely the HH and WFNS, and is an
independent predictor of DCI and poor short-term and long-
term clinical outcomes. Modeled loosely on the ASPECT score,
we examined the absence of visible sulci caused by either (1)
effacement of sulci or (2) absence of sulci with disruption of the
gray–white matter junction at 2 different levels (at the insular
cortex and over the convexity beyond centrum semiovale) and
both hemispheres in the CT (Figure 1). At admission, high-grade
SEBES was present in 29.3% of our patients.
Predictors of SEBES were poor HH or WFNS grade and LOC
at ictus on admission (Table 2). This is consistent with earlier
reports for predictors of GCE.12 Similar to the hypothesized
pathophysiology of GCE, the SEBES grade may be a reflection
of tissue and microvascular damage caused by intracranial circu-
latory arrest immediately after aneurysm rupture.13 Insufficient
cerebral perfusion caused by a surge in intracranial pressure
leads to a state of transient diffuse ischemic encephalopathy.11,26
Furthermore, patients with poor WFNS grade at admission are
prone to develop excessive elevated ICP, which can exacerbate
factors for microvascular and cellular damage, leading to higher
a SEBES grade.7 The fact that poor WFNS and LOC at ictus
NEUROSURGERY
SUBARACHNOID HEMORRHAGE EARLY BRAIN EDEMA SCORE (SEBES)
Adjusteda Adjustedb Adjustedc
2.24 (1.58–3.17) N/A N/A
N/A 3.45 (1.95–6.07) 3.25 (1.88-5–59)
were associated with worse a SEBES grade suggest that the SEBES
grade is a radiographic reflection of the pathophysiology of EBI.
SEBES showed a similar and better performance for prediction
of unfavorable outcome and DCI compared to the traditional
grading system composed of HH, WFNS, and mFS grades
(Figure 2). Furthermore, subjects with high-grade SEBES were
at an increased risk of developing early DCI with radiographic
evidence of ischemic damage within 7 d after hemorrhage
and DCI, regardless of prophylactic treatment or angiographic
vasospasm (Figure 3A and 3B). This may be attributed to the
incorporation of early brain tissue disruption, possibly at the
microcirculatory level, after SAH.27 Although the Fisher scale
is a reliable and well-accepted radiographic grading system for
prediction of DCI, this scale focuses on the amount of blood
in basal cisterns and the intraventricular system for prediction
of large vascular vasospasm. Our finding that the SEBES has a
high sensitivity for prediction of DCI suggests that the SEBES
grade reflects additional changes in brain physiology, possibly
representing microcirculatory changes, which have been hypoth-
esized to increase the risk for DCI.28 Our results also showed
that prediction of angiographic vasospasm was not significantly
different between grading systems (Figure 3C). We interpret our
findings to suggest that patients with EBI are more prone to DCI,
and the occurrence of delayed arterial vasospasm works synergis-
tically to increase the risk of DCI.
SEBES is an independent predictor of clinical outcomes even
after adjustments for well-known predictors HH grade and
WFNS grade (Table 3). Traditionally, clinical scores, such as HH,
Glasgow Coma Scale, and WFNS classification, have been used
as an indicator of clinical severity and a marker of prognosis.
The severity of initial hemorrhage (as measured by the presence
of IVH or high-grade mFS score), age, and re-bleeding have
been consistently shown to impact mortality in SAH.29 SEBES
incorporates a clinically important but often ignored radiographic
finding, namely that of edematous changes seen on CT. Thus, the
SEBES grade may be used to classify patients for more intensive
monitoring or an earlier start of specific preventive treatments.30
VOLUME 0 | NUMBER 0 | 2017 | 5
AHN ET AL

FIGURE 2. Distribution of patients A, B, and ROC curve for prediction of unfavorable outcome and DCI according to the clinical
and radiographic grades C. DCI, delayed cerebral ischemia; HH, Hunt-Hess; WFNS, World Federation of Neurosurgical Societies; mFS,
modified Fisher Score; SEBES, Subarachnoid hemorrhage Early Brain Edema Score.

Limitations study was based on a qualitative assessment of edematous changes

Several limitations of the present study deserve mention. First,
our study is limited by the bounds of a single-center observa-
tional study and all the potential biases associated. To address
this, CT grades were performed completely blinded to all clinical
information. In addition, SEBES was examined by independent
examiners who were also blinded to clinical information. Our
on CTs performed clinically. Although we demonstrated good
inter-rater reliability, it may only reflect the fact that our radio-
graphic criteria were intended to identify only very severe cases.
Despite this limitation, the rapid qualitative assessment of the
scores and ability to use clinically obtained imaging makes the
SEBES grade clinically practical to use. The SEBES grade has

6 | VOLUME 0 | NUMBER 0 | 2017 www.neurosurgery-online.com

 SUBARACHNOID HEMORRHAGE EARLY BRAIN EDEMA SCORE (SEBES)

FIGURE 3. Characteristics of DCI and incidence of DCI according to the status of vasospasm in patients with high- and low-grade SEBES
A, B, and ROC curve for prediction of angiographic vasospasm according to the clinical and radiographic grades C. DCI, delayed cerebral
ischemia; SEBES, Subarachnoid hemorrhage Early Brain Edema Score; HH, Hunt–Hess; WFNS, World Federation of Neurosurgical Societies;
mFS, modified Fisher Score. ∗ P < .05 by linear trend test.

NEUROSURGERY VOLUME 0 | NUMBER 0 | 2017 | 7

AHN ET AL
some advantages over GCE scoring. First, while GCE is a quali-
tative and dichotomized grading system, SEBES is a semiquan-
titative grading system, allowing for a gradient quantification of
injury severity at clinical presentation. Second, SEBES grade is
evenly distributed compared to other clinical and radiographic
grading scores (HH and mFS), allowing for a more discrim-
inative value. Another limitation is that the SEBES grade is
semiquantitative. Previously, we have reported that quantitative
measurements of sulcal volume predicted function outcomes.18
The disadvantage of a completely quantitative approach is its
tedious nature. The SEBES can be performed quickly without
any additional time or software. Third, our study included a
rather modest number of SAH patients and could not show the
long-term prognostic value of the SEBES grade. Therefore, future
study is needed to reconfirm the practical value of the SEBES
grade and whether it can be also useful to predict the long-term
outcomes based on the large number of SAH patients.
CONCLUSION
Our findings indicate that EBI, characterized by the absence
of visible sulci in 2 slices of the initial CT and calculated using
the SEBES grade, is an important predictor of DCI and poor
outcome after SAH. Further research is needed to elucidate the
pathophysiology of EBI. The application of imaging techniques
to characterize EBI will help to develop strategies to halt the
process of EBI after SAH and improve outcomes.
Disclosures
Dr Choi receives funding from the CCTS Scholar Program, University of
Texas Health Science at Houston, McGovern Medical School. The authors have
no personal, financial, or institutional interest in any of the drugs, materials, or
devices described in this article.
REFERENCES
1. Kassell NF, Torner JC, Haley EC, Jr, Jane JA, Adams HP, Kongable GL. The
International cooperative study on the timing of aneurysm surgery. part 1: overall
management results. J Neurosurg. 1990;73(1):18-36.
2. Macdonald RL, Higashida RT, Keller E, et al. Clazosentan, an endothelin
receptor antagonist, in patients with aneurysmal subarachnoid haemorrhage under-
going surgical clipping: a randomised, double-blind, placebo-controlled phase 3
trial (CONSCIOUS-2). Lancet Neurol. 2011;10(7):618-625.
3. Budohoski KP, Guilfoyle M, Helmy A, et al. The pathophysiology and
treatment of delayed cerebral ischaemia following subarachnoid haemorrhage. J
Neurol Neurosurg Psychiatry. 2014;85(12):1343-1353.
4. Vergouwen MD, Ilodigwe D, Macdonald RL. Cerebral infarction after
subarachnoid hemorrhage contributes to poor outcome by vasospasm-dependent
and -independent effects. Stroke. 2011;42(4):924-929.
5. Frontera JA, Ahmed W, Zach V, et al. Acute ischaemia after subarachnoid
haemorrhage, relationship with early brain injury and impact on outcome: a
prospective quantitative MRI study. J Neurol Neurosurg Psychiatry. 2015;86(1):71-
78.
6. Rowland MJ, Hadjipavlou G, Kelly M, Westbrook J, Pattinson KT. Delayed
cerebral ischaemia after subarachnoid haemorrhage: looking beyond vasospasm. Br
J Anaesth. 2012;109(3):315-329.
7. Naidech AM, Drescher J, Tamul P, Shaibani A, Batjer HH, Alberts MJ. Acute
physiological derangement is associated with early radiographic cerebral infarction
8 | VOLUME 0 | NUMBER 0 | 2017
after subarachnoid haemorrhage. J Neurol Neurosurg Psychiatry. 2006;77(12):1340-
1344.
8. Cahill J, Calvert JW, Zhang JH. Mechanisms of early brain injury after
subarachnoid hemorrhage. J Cereb Blood Flow Metab. 2006;26(11):1341-1353.
9. Brisman JL, Song JK, Newell DW. Cerebral aneurysms. N Engl J Med.
2006;355(9):928-939.
10. Sehba FA, Pluta RM, Zhang JH. Metamorphosis of subarachnoid hemor-
rhage research: from delayed vasospasm to early brain injury. Mol Neurobiol
2011;43(1):27-40.
11. Grote E, Hassler W. The critical first minutes after subarachnoid hemorrhage.
Neurosurgery. 1988;22(4):654-661.
12. Claassen J, Carhuapoma JR, Kreiter KT, Du EY, Connolly ES, Mayer SA.
Global cerebral edema after subarachnoid hemorrhage: frequency, predictors, and
impact on outcome. Stroke. 2002;33(5):1225-1232.
13. Helbok R, Ko SB, Schmidt JM, et al. Global cerebral edema and brain metabolism
after subarachnoid hemorrhage. Stroke. 2011;42(6):1534-1539.
14. Pexman JH, Barber PA, Hill MD, et al. Use of the Alberta Stroke Program Early
CT Score (ASPECTS) for assessing CT scans in patients with acute stroke. AJNR
Am J Neuroradiol. 2001;22(8):1534-1542.
15. Bang OY. Multimodal MRI for ischemic stroke: from acute therapy to preventive
strategies. J Clin Neurol (Seoul, Korea). 2009;5(3):107-119.
16. Lin L, Bivard A, Parsons MW. Perfusion patterns of ischemic stroke on computed
tomography perfusion. J Stroke. 2013;15(3):164-173.
17. Frontera JA, Claassen J, Schmidt JM, et al. Prediction of symptomatic
vasospasm after subarachnoid hemorrhage: the modified fisher scale. Neurosurgery.
2006;59(1):21-27; discussion 21–27.
18. Choi HA, Bajgur SS, Jones WH, et al. Quantification of cerebral edema after
subarachnoid hemorrhage. Neurocrit Care. 2016;25(1):64-70
19. Hunt WE, Hess RM. Surgical risk as related to time of intervention in the repair
of intracranial aneurysms. J Neurosurg. 1968;28(1):14-20.
20. Rosen DS, Macdonald RL. Grading of subarachnoid hemorrhage: modification of
the world World Federation of Neurosurgical Societies scale on the basis of data for
a large series of patients. Neurosurgery. 2004;54(3):566-575; discussion 575-566.
21. Claassen J, Bernardini GL, Kreiter K, et al. Effect of cisternal and ventricular
blood on risk of delayed cerebral ischemia after subarachnoid hemorrhage: the
Fisher scale revisited. Stroke. 2001;32(9):2012-2020.
22. Connolly ES, Jr, Rabinstein AA, Carhuapoma JR, et al. Guidelines for the
management of aneurysmal subarachnoid hemorrhage: a guideline for healthcare
professionals from the American Heart Association/american Stroke Association.
Stroke. 2012;43(6):1711-1737.
23. van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J. Inter-
observer agreement for the assessment of handicap in stroke patients. Stroke.
1988;19(5):604-607.
24. Hallevi H, Dar NS, Barreto AD, et al. The IVH score: a novel tool for estimating
intraventricular hemorrhage volume: clinical and research implications. Crit Care
Med. 2009;37(3):969-974, e961.
25. van Gijn J, Hijdra A, Wijdicks EF, Vermeulen M, van Crevel H. Acute hydro-
cephalus after aneurysmal subarachnoid hemorrhage. J Neurosurg. 1985;63(3):355-
362.
26. Choi HA, Ko SB, Chen H, et al. Acute effects of nimodipine on cerebral vascu-
lature and brain metabolism in high grade subarachnoid hemorrhage patients.
Neurocrit Care. 2012;16(3):363-367.
27. Ko SB, Choi HA, Carpenter AM, et al. Quantitative analysis of hemorrhage
volume for predicting delayed cerebral ischemia after subarachnoid hemorrhage.
Stroke. 2011;42(3):669-674.
28. Charidimou A, Meegahage R, Fox Z, et al. Enlarged perivascular spaces as
a marker of underlying arteriopathy in intracerebral haemorrhage: a multicentre
MRI cohort study. Neurol Neurosurg Psychiatry. 2013;84(6):624-629.
29. Broderick JP, Brott TG, Duldner JE, Tomsick T, Leach A. Initial and recurrent
bleeding are the major causes of death following subarachnoid hemorrhage. Stroke.
1994;25(7):1342-1347.
30. Laskowitz DT, Kolls BJ. Neuroprotection in subarachnoid hemorrhage. Stroke.
2010;41(10 Suppl):S79-84.
Supplemental digital content is available for this article at www.neurosurgery-
online.com.
www.neurosurgery-online.com

Acknowledgments
We would like to acknowledge the physicians and nurses at Memorial
Hermann Hospital Neuroscience Intensive Care Unit.
COMMENTS
T his paper explored potential prognostic factors in patients with
subarachnoid hemorrhage. The authors focused on developing an
accessible radiographic surrogate of early brain injury and assessed the
performance of their proposed scale using their institutional database.
Their Subarachnoid Hemorrhage Early Brain Edema Scale (SEBES),
with a maximum of 4 points, is based upon effacement of sulci at 2
different axial levels (thalamus/basal ganglia and lateral ventricle), and a
total of 164 prospectively enrolled patients were utilized as the validation
dataset. Multivariable logistic regression was performed to determine
factors associated with high SEBES score, as well as the prognostic value
of SEBES adjusted for potential confounding factors. The authors found
that SEBES can be reliably assessed by independent scorers, and high
SEBES is associated with high Hunt and Hess (HH) or World Feder-
ation of Neurosurgical Societies (WFNS) grade and loss of consciousness
at ictus. SEBES was also found to be an independent prognostic factor
of delayed cerebral ischemia (DCI) and unfavorable outcome (mRS
4–6). Based on these results, the authors concluded that the SEBES
grade is an important prognostic factor of DCI and poor outcome after
subarachnoid hemorrhage. One of the major questions is the utility of
the new SEBES score compared to existing well-established clinical scales
for evaluating patient prognosis. According to the ROC analysis, SEBES
has better prognostic value for DCI but not for unfavorable outcome
compared to HH or WFNS, and the 4 scores (SEBES, HH, WFNS, and
mFS) are similarly suboptimal for predicting angiographic vasospasm.
Therefore, its utility remains to be determined but its ease of use (similar
to ASPECTS for ischemic stroke) may be advantageous.
Judy Huang
Baltimore, Maryland
T he term “early brain injury” refers to global brain injury that occurs
within the first 72 hours after subarachnoid hemorrhage (SAH)1 .
The pathophysiology of early brain injury includes global cerebral
ischemia, oxidative stress, cytotoxic and vasogenic brain edema, neuroin-
flammation, leading to brain cell death including necrosis, apoptosis, and
autophagy2 . Early brain injury seems associated with early mortality after
SAH, and may be a precursor associated with delayed cerebral arterial
vasospasm and delayed cerebral ischemia.
NEUROSURGERY
SUBARACHNOID HEMORRHAGE EARLY BRAIN EDEMA SCORE (SEBES)
One of the key features of early brain injury is the global cerebral
edema, which may occur immediately or develop quickly within a few
days after the ictus of SAH. The nature of the global cerebral edema
may be a mixture of both cytotoxic and vasogenic brain edema, as well
as brain swelling, either due to direct impact of the initial bleeding or
to secondary brain injury. Global cerebral edema has been observed in
both experimental animal models and in clinical studies. The pathophys-
iology of global brain edema is also multiple factors including abnormal
autoregulation of cerebral arterial system, rupture of blood-brain barrier,
neuronal cell death, and may be in some cases related to the poor cerebral
venous returns.
However, clinical measures of early brain injury especially global
cerebral edema have been rather limited.
In this study, the authors proposed a novel radiographic marker of
early brain injury named the Subarachnoid Hemorrhage Early Brain
Edema Score (SEBES), which was scored according to the change of
sulci in computed tomography within 24 hours of ictus. The results
demonstrated SEBES was well associated with other clinical markers of
early brain injury, such as Hunt and Hess, World Federation of Neuro-
logic Surgeons, and SEBES was identified as an independent predictor
of delayed cerebral ischemia (odds ratio [OR] = 2.24, 95% confi-
dence intervals [CI] 1.58-3.17) and unfavorable outcome (OR = 3.45,
95% CI 1.95-6.07). This is apparently a landmark paper for clinical
neuroimaging recognition of early brain injury and has potential to
reshape clinical diagnostic tools of early brain injury and is beneficial
to the clinical management of SAH.
There are some limitations as the authors have discussed, that a
prospective, randomized, double-blinded, and multicenter studies of
SEBES are needed. Additionally, gender bias may be discussed since
70.7% subjects were females in the study. Prior imaging study demon-
strated that SAH induced more severe brain injury in females when
compared to males in the acute phase3 . Overall, the application of
imaging tools definitely has potential to detect and characterize early
brain injury, to benefit developing therapeutic strategies, and to improve
outcomes in SAH patients.
Sheng Chen
John H. Zhang
Loma Linda, California
1. Chen S, Feng H, Sherchan P, et al. Controversies and evolving new mechanisms in
subarachnoid hemorrhage. Prog Neurobiol. 2014;115:64-91.
2. Suzuki H. What is early brain injury? Transl Stroke Res. 2015;6(1):1-3.
3. Guo D, Wilkinson DA, Thompson BG, et al. MRI Characterization in the Acute
Phase of Experimental Subarachnoid Hemorrhage. Transl Stroke Res. 2016.
VOLUME 0 | NUMBER 0 | 2017 | 9