radiol.2422051698

䡲 NEURORADIOLOGY
Note: This copy is for your personal non-commercial use only. To order presentation-ready
copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.
ORIGINAL RESEARCH
Subarachnoid Hemorrhage in
the Subacute Stage: Elevated
Apparent Diffusion Coefficient in
Normal-appearing Brain Tissue
after Treatment1
Yawu Liu, MD, PhD
Ville Soppi, MD
Timo Mustonen, MD
Mervi Könönen, MD
Timo Koivisto, MD, PhD
Anna Koskela, MD
Jaakko Rinne, MD, PhD
Ritva L. Vanninen, MD, PhD
Purpose:
To prospectively evaluate whether subarachnoid hemorrhage (SAH) is associated with a change in the apparent
diffusion coefficient (ADC) in normal-appearing brain parenchyma.
Materials and
Methods:
Institutional review board approval and informed consent
were obtained for all patient and volunteer studies. One
hundred patients (48 men, 52 women; mean age, 52
years ⫾ 12 [standard deviation]) with aneurysmal SAH
underwent conventional and diffusion-weighted magnetic
resonance (MR) imaging at a mean of 9 days ⫾ 3 after SAH
to evaluate possible lesions caused by SAH, treatment of
SAH, and vasospasm. Aneurysms were treated surgically
(n ⫽ 70) or endovascularly (n ⫽ 30) before MR imaging.
Diffusion-weighted MR imaging was performed at 1-year
follow-up in 30 patients (10 men, 20 women; mean age, 51
years ⫾ 11). Thirty healthy age-matched volunteers (11
men, 19 women; mean age, 54 years ⫾ 16) underwent MR
imaging with an identical protocol. ADC values were measured bilaterally in the gray and white matter (parietal,
frontal, temporal, occipital lobes; cerebellum; caudate nucleus; lentiform nucleus; thalamus; and pons) that appeared normal on T2-weighted and diffusion-weighted MR
images. Linear mixed model was used for comparison of
ADC values of supratentorial gray matter and white matter; general linear regression analysis was used for comparison of ADC values of cerebellum and pons.
Results:
In patients with SAH, the ADC values in normal-appearing
white matter, with a single exception in the frontal lobe
(P ⫽ .091), were significantly higher than they were in
healthy volunteers (P ⱕ .011). The differences disappeared by 1 year, except in parietal white matter (P ⫽
.045). The ADC values of cortical gray matter did not
significantly differ between patients and volunteers (P ⱖ
.121).
Conclusion:
SAH and its treatment may cause global mild vasogenic
edema in white matter and deep gray matter that is undetectable on T2-weighted and diffusion-weighted MR images but is detectable by measuring the ADC value in the
subacute stage of SAH.
1
From the Departments of Clinical Radiology (Y.L., T.M.,
M.K., A.K., R.L.V.), Neurosurgery (V.S., T.K., J.R.), and
Clinical Neurophysiology (M.K.), Kuopio University Hospital, PO Box 1777, FIN-70211 Kuopio, Finland. Received
October 17, 2005; revision requested December 8; revision received January 17, 2006; accepted February 17;
final version accepted May 5. Supported by Kuopio University Hospital (EVO funding 5063515) and the Instrumentarium Science Foundation. Address correspondence to Y.L. (e-mail: yawu.liu@kuh.fi ).
娀 RSNA, 2006
姝 RSNA, 2006
518
Radiology: Volume 242: Number 2—February 2007
NEURORADIOLOGY: Subarachnoid Hemorrhage: ADC in Brain after Treatment
D
iffusion-weighted magnetic resonance (MR) imaging has been
used extensively in the evaluation of ischemic stroke. In ischemic
brain tissue, water mobility is restricted
because of the loss of ionic gradients
and translocation of water from the extracellular to the intracellular compartment (cytotoxic edema). Consequently,
the apparent diffusion coefficient (ADC),
a quantitative value that is calculated
with raw data from diffusion-weighted
imaging and that represents water diffusivity, is decreased. On the other hand,
in vasogenic edema, the ADC value is
increased because of a net increase in
the extracellular space.
Although diffusion-weighted imaging has been used frequently in patients with ischemic stroke, the technique has not been used as much in
patients with subarachnoid hemorrhage (SAH). Some primary studies
(1–4) with a limited number of patients have shown the usefulness of
diffusion-weighted imaging in the detection of acute ischemia in patients
with symptomatic vasospasm after SAH.
Quantitative measurements of ADC are
more sensitive than diffusion-weighted
imaging for detection of even mild
changes in water diffusivity. Thus, the
purpose of our study was to prospectively evaluate whether SAH is associ-
Advances in Knowledge
䡲 In patients with subarachnoid
hemorrhage, the apparent diffusion coefficient values in normalappearing white matter and deep
gray matter, with a single exception in frontal white matter (P ⫽
.091), were significantly increased
(P ⱕ .011) in the subacute stage
compared with values in healthy
volunteers, but they returned to a
normal level at 1-year follow-up
MR imaging.
䡲 This finding indicates the presence of mild diffuse vasogenic
edema in the subacute stage, and
the edema is not detectable on
conventional T2-weighted MR
images.
Radiology: Volume 242: Number 2—February 2007
ated with a change in the ADC in normal-appearing brain parenchyma.
Materials and Methods
Institutional review board approval and
informed consent were obtained for all
patient and volunteer studies.
Study Design
During the study period, all patients
with aneurysmal SAH and eligibility (as
mentioned later) were prospectively
scheduled to undergo conventional MR
imaging and diffusion-weighted imaging
approximately 1 week after SAH to evaluate possible ischemic lesions or other
lesions caused by SAH, treatment of
SAH, and vasospasm. One-year follow-up
conventional MR images and diffusionweighted images were obtained from a
randomly selected subgroup. An agematched group of healthy volunteers
was used as a control group. The number of patients in the follow-up group
and the number of volunteers in the
control group were determined with a
power analysis.
Patients
During the study period from January
2002 to July 2004, 245 patients with
aneurysmal SAH were admitted to our
institution. Of these 245 patients with
acute SAH, 92 consecutive patients
were prospectively scheduled to undergo conventional MR imaging and diffusion-weighted imaging 5–19 days after
initial bleeding. Before that, a pilot
study that included eight patients with
SAH was conducted with the same MR
imaging protocol. Thus, a total of 100
patients (48 men, 52 women; mean age,
52 years ⫾ 12 [standard deviation];
range, 19 – 81 years) were enrolled in
our study. Patients were eligible for the
present MR imaging study protocol if
subarachnoid blood was detected at
computed tomography (CT) and a ruptured cerebral aneurysm was verified at
three-dimensional rotational angiography. The mean interval from the bleeding to MR imaging examination was 9
days, with a range of 5–19 days. The
reasons that MR imaging was not performed in the other 153 patients were
Liu et al
the following: lack of informed consent,
ongoing intensive care unit treatment
with a Swan-Ganz catheter (considered
to be a contraindication for MR imaging), the patient died soon after admission, the patient was in extremely poor
condition so that transfer to the Department of Clinical Radiology was considered too risky, temporary lack of resources of the Department Clinical of
Radiology, or treatment delay beyond
the study time range.
Hunt and Hess grade (5) was assessed by four resident neurosurgeons
(with 2– 6 years of experience) and was
confirmed by two senior neurosurgeons
(T.K., with 13 years of experience; J.R.,
with 19 years of experience). In Hunt
and Hess grades I and II, the patient is
asymptomatic or has minimal headache
and slight nuchal rigidity; in grade III,
the patient is confused and drowsy; in
grade IV, the patient is unconscious
with or without a localizing neurologic
deficit; and in grade V, the patient is in a
deep coma with a moribund appearance. The modified Fisher grade (6) was
assessed by the same two senior neurosurgeons who conducted the Hunt and
Hess assessment to evaluate the amount
and distribution of subarachnoid blood.
Fisher grade I indicates that no clot is
seen on the CT scan, grade II indicates
moderate bleeding is observed predominantly in the basal cisterns, grade III
indicates more diffuse severe bleeding,
Published online before print
10.1148/radiol.2422051698
Radiology 2007; 242:518 –525
Abbreviations:
ADC ⫽ apparent diffusion coefficient
ROI ⫽ region of interest
SAH ⫽ subarachnoid hemorrhage
Author contributions:
Guarantors of integrity of entire study, Y.L., R.L.V.; study
concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or
manuscript revision for important intellectual content, all
authors; manuscript final version approval, all authors;
literature research, Y.L., V.S., T.M., T.K.; clinical studies,
Y.L., V.S., T.M., T.K., A.K., J.R., R.L.V.; statistical analysis,
Y.L., M.K., T.K., R.L.V.; and manuscript editing, Y.L., T.K.,
R.L.V.
Authors stated no financial relationship to disclose.
519
NEURORADIOLOGY: Subarachnoid Hemorrhage: ADC in Brain after Treatment
and grade IV indicates an intracerebral
or intraventricular clot with diffuse or
no subarachnoid clot. The patients’ demographics, Hunt and Hess grades,
Fisher grades, locations of aneurysms,
and treatment modalities are shown in
Table 1.
On the basis of data from our previous study (7), a power analysis (␮1 ⫽
[0.80 ⫾ 0.05] ⫻ 10⫺3 mm2/sec, n1 ⫽
100, 5% difference, ␣ ⫽ .05, ␤ ⫽ .05,
where ␮1 is the mean ADC value of sample 1 in our previous study [7] and n1 is
the number in sample 1) showed that
the minimal sample size was 26 for both
patients in the follow-up group and
healthy volunteers in the control group.
MR images and diffusion-weighted
images at 1-year follow-up were obtained from 30 patients (10 men, 20
women; mean age, 51 years ⫾ 11;
range, 28 – 80 years) who were randomly selected from the 100 patients
enrolled in this study. There were no
significant differences in sex (P ⫽ .156,
␹2 test) and age (P ⫽ .290, Mann-Whitney U test) between the 30 patients who
were selected and the 100 patients who
were enrolled.
Volunteers
Thirty healthy age-matched volunteers
(11 men, 19 women; mean age, 54
years ⫾ 16; range, 22–79 years) also
underwent MR imaging with the same
diffusion-weighted imaging sequence
and imaging parameters and with the
same MR imager. The volunteers were
imaged within the same time frame during which the patients with subacute
stage SAH were imaged. Volunteers
were asked about their medical history,
especially any neurologic or cardiovascular disturbances, cranial trauma, or permanent medication they were receiving.
Any individual who was suspected of having any pathologic medical condition or
who was taking permanent medication
was excluded. There were no significant
differences in sex (P ⱖ .274, ␹2 test) and
age (P ⱖ .426, Mann-Whitney U test)
between the group of volunteers and the
group of 100 patients with SAH or the
30 patients who had undergone 1-year
follow-up MR imaging and diffusionweighted imaging.
520
Treatment
According to our current practice, ruptured aneurysms should be treated either surgically or endovascularly within
72 hours from initial bleeding. In our
study of 100 patients, this was achieved
in 89 (89%) patients, and 96 (96%) patients were treated within 7 days. The
other four patients were treated surgically after 11–14 days from onset because of delays before admission to
the hospital. All aneurysms, however,
were treated surgically (n ⫽ 70) or
endovascularly (n ⫽ 30) before MR
imaging. All patients with SAH were
routinely treated both pre- and postoperatively in the intensive care unit,
and they all received nimodipine (Nimotop; Bayer Pharma, Paris, France)
and corticosteroid medication (betamethasone, Betapred; Swedish Orphan,
Stockholm, Sweden). Symptomatic hydrocephalus was treated with ventriculostomy in the acute stage and later with
permanent shunt placement if needed.
The Glasgow Coma Scale was recorded
daily by the same resident and senior
neurosurgeons. Patients who developed
symptomatic vasospasm were treated
in the intensive care unit with HHH
(hypertension, hypervolemia, hemodilution) therapy (8–11) used at many
institutions and, if necessary, with angioplasty or intraarterial nimodipine infusion. The clinical criterion for symptomatic vasospasm was a decrease in
the Glasgow Coma Scale of two or more
scores or the appearance of new localizing symptoms (dysphasia, hemiparesis),
with other reasons for deterioration (eg,
hydrocephalus, metabolic disorders, postoperative bleeding, and infections) first
ruled out.
MR Imaging and Evaluation
All MR imaging studies were performed
with a 1.5-T whole-body imager capable
of echo-planar imaging (Magnetom Vision; Siemens Medical Systems, Erlangen, Germany) by using a head coil. The
patient’s head was fixed with standard
restraints used in routine clinical MR
imaging.
Diffusion-weighted imaging was performed with a single-shot echo-planar
spin-echo sequence (repetition time
Liu et al
msec/echo time msec, 6000/100). Other
imaging parameters were as follows:
section thickness, 5 mm; intersection
gap, 1.5 mm; field of view, 240 mm; and
matrix size, 96 ⫻ 200. Nineteen transverse sections parallel to the orbitomeatal line, which included whole supratentorial and infratentorial brain tissue, were imaged. Four images per
same section position were obtained:
one T2-weighted image without diffusion weighting (b value, 0 sec/mm2)
and three diffusion-weighted images
with orthogonally applied diffusion
gradients (b value, 1000 sec/mm2). To
prevent the effects of diffusion anisotropy, isotropic diffusion-weighted imTable 1
Clinical and Imaging Characteristics
of 100 Patients with SAH
Characteristic
Sex
Male
Female
Age (y)*
19 to ⬍60
ⱕ60 to 81
Location of ruptured aneurysm at
digital subtraction angiography
Internal carotid artery
Middle cerebral artery
Anterior cerebral artery
Vertebrobasilar arteries
Hunt and Hess grade
I
II
III
IV
Fisher grade at CT
II
III
IV
Hydrocephalus at CT
Present
Absent
Symptomatic vasospasm before MR
imaging
Present
Absent
Treatment
Endovascular coil
Surgical clip
No. of
Patients
48
52
74
26
20
38
29
13
14
49
26
11
26
48
26
23
77
14
86
30
70
* The mean age was 52 years ⫾ 12.
Radiology: Volume 242: Number 2—February 2007
NEURORADIOLOGY: Subarachnoid Hemorrhage: ADC in Brain after Treatment
ages were reconstructed by calculating the geometric mean of the three
images obtained with a b value of 1000
sec/mm2. The ADC maps were generated off-line. The ADC was calculated
on a pixel-by-pixel basis as the negative slope of the line fitting the two
points for b versus ln(SI), where SI is
the signal intensity.
The ADC values were measured
from bilateral regions in the normal-appearing gray and white matter (parietal,
frontal, temporal, and occipital lobes
and cerebellum), caudate nucleus, putamen, thalamus, pons, and the body of
the lateral ventricle by one radiologist
(Y.L.), who had 6 years of experience in
interpretation of diffusion-weighted images. In all regions of interest (ROIs),
the isotropic diffusion-weighted and conventional MR images were carefully visually reviewed to avoid areas that were
affected by susceptibility artifact or focal
lesions. If any of the locations were affected by metal susceptibility artifact, leukoencephalopathy, intracranial hematoma, edema, or infarction, the corresponding areas were excluded from the
measurements. If possible, identical mirrored regions were used (Figure). The
mean size of the ROIs was 35 mm3 ⫾ 22,
with a range of 2–164 mm3.
Liu et al
Statistical Analysis
All statistical analyses were performed
with software (SPSS, version 11.5.1,
November 16, 2002; SPSS, Chicago,
Ill). After proving no significant difference between the ADC values of the
mirrored ROIs in the right and left
hemispheres with the Wilcoxon signed
rank test, the values of two sides were
averaged and analyzed separately for
each anatomic location. If no ROI
could be analyzed in a certain anatomic location because of an artifact
or a focal lesion, however, only the
ADC value of the ROI on the contralateral unaffected hemisphere was used
in the analysis instead of an averaged
value.
The general linear regression was
used to compare the difference between
the ADC values of patients and those of
healthy volunteers and to analyze the
association between ADC and factors
that were dichotomized with relevant
clinical data (age, presence of hydrocephalus, Hunt and Hess grade, Fisher
grade, location of aneurysm, presence
of symptomatic vasospasm, and treatment) in cerebellar gray matter, cerebellar white matter, the pons, and cerebrospinal fluid. Because the ADC values
in cortical gray or white matter (parietal, frontal, temporal, and occipital
lobes) or deep gray matter (caudate nucleus, lentiform nucleus, and thalamus)
may not be independent of each other,
the linear mixed model was used to
compare the differences in ADC values
between patients and volunteers and to
analyze the association between ADC and
factors that were dichotomized with relevant clinical data. A difference with P ⬍
.05 was considered statistically significant.
Results
Transverse ADC maps show ROIs that were drawn bilaterally in brain parenchyma with a normal appearance
and include cortical gray matter and supratentorial white matter. Top left: ROIs in parietal and frontal lobes.
Bottom left: ROIs in temporal and occipital lobes. Top right: ROIs in caudate nucleus, lentiform nucleus, and
thalamus. Bottom right: ROIs in pons and cerebellar gray and white matter.
Radiology: Volume 242: Number 2—February 2007
MR Imaging Findings
In patients with subacute stage SAH,
findings at MR imaging were normal in
26 patients, a single ischemic lesion was
observed in 48 patients, and multiple
ischemic lesions were observed in 26
patients. In volunteers, MR imaging
findings were normal.
521
NEURORADIOLOGY: Subarachnoid Hemorrhage: ADC in Brain after Treatment
ADC Values
ADC measurements were successfully
performed in every area in each volunteer; thus, a total of 840 ROIs were
drawn. Among the 100 patients, ROIs
could be drawn bilaterally on parietal gray matter in 68 patients, on parietal white matter in 67 patients, on
frontal gray matter in 52 patients,
on frontal white matter in 51 patients, on the caudate nucleus in 31
patients, on the lentiform nucleus in
41 patients, on the thalamus in 57 patients, on temporal gray matter or
white matter in 52 patients, on occipital gray matter in 95 patients, on occipital white matter in 94 patients, on
the pons in 100 patients, and on cerebellar gray matter or white matter in
98 patients. A total of 2460 ROIs were
thus drawn at normal-appearing areas
in patients with the subacute stage of
SAH; in the 30 patients who underwent
follow-up MR imaging, a total of 747
ROIs were drawn 1 year after SAH.
In the patients with SAH, there
was no significant difference in ADC
of the normal-appearing brain parenchyma between the side of the ruptured
aneurysm and the contralateral side in
the measured areas (P ⫽ .056 –.907,
Wilcoxon signed rank test).
In the subacute stage, the ADC values in normal-appearing white matter
and deep gray matter, with a single exception in the frontal lobe of supratentorial white matter, were significantly
higher in patients with SAH than they
were in healthy volunteers (P ⱕ .091)
(Table 2). The values in cortical gray
matter did not significantly differ between patients and volunteers (Table
2). On average, the ADC value was 4%
higher in both normal-appearing white
matter and deep gray matter in patients
with SAH than it was in healthy volunteers.
In the 30 patients who underwent
follow-up diffusion-weighted imaging,
the elevated ADC values had decreased
to a normal level after 1 year from onset, compared with the values in the
volunteers. The single exception to this
finding was the value in the parietal lobe
of supratentorial white matter (P ⫽
.045, comparison of patients at 1-year
522
Liu et al
Table 2
ADC Values of Normal-appearing Brain Parenchyma in 100 Patients with Aneurysmal
SAH in Subacute Stage and 30 Volunteers
Location of Measurements
Supratentorial cortical gray matter
Parietal lobe
Frontal lobe
Temporal lobe
Occipital lobe
Supratentorial white matter
Parietal lobe
Frontal lobe
Temporal lobe
Occipital lobe
Deep gray matter
Caudate nucleus
Lentiform nucleus
Thalamus
Cerebellar gray matter
Cerebellar white matter
Pons
Patients
Volunteers
P Value
0.73 ⫾ 0.007
0.77 ⫾ 0.01
0.79 ⫾ 0.006
0.76 ⫾ 0.005
0.71 ⫾ 0.01
0.76 ⫾ 0.01
0.78 ⫾ 0.01
0.74 ⫾ 0.01
.141
.391
.533
.121
0.77 ⫾ 0.004
0.80 ⫾ 0.006
0.81 ⫾ 0.007
0.78 ⫾ 0.004
0.75 ⫾ 0.01
0.78 ⫾ 0.01
0.77 ⫾ 0.01
0.75 ⫾ 0.01
.004
.091
.006
⬍.001
0.78 ⫾ 0.01
0.78 ⫾ 0.01
0.81 ⫾ 0.004
0.74 ⫾ 0.005
0.69 ⫾ 0.004
0.72 ⫾ 0.01
0.75 ⫾ 0.01
0.74 ⫾ 0.01
0.78 ⫾ 0.01
0.72 ⫾ 0.01
0.66 ⫾ 0.004
0.68 ⫾ 0.01
.011
.008
.012
.120
.004
.008
Note.—The values represent estimated marginal means and standard errors of the averaged ADC values (⫻10⫺3 mm2/sec)
of the left and right hemispheres. Pairwise comparisons of patients and volunteers were based on estimated marginal means
obtained from the linear mixed model for the supratentorial areas and deep gray matter and on data obtained from the general
linear regression for the cerebellum and the pons.
follow-up vs volunteers) (Table 3). The
mean ADC value of cerebrospinal fluid
in the body of the lateral ventricle did
not significantly differ between volunteers and patients with SAH in the subacute stage ([3.46 ⫾ 0.03] ⫻ 10⫺3 mm2/
sec vs [3.51 ⫾ 0.04] ⫻ 10⫺3 mm2/sec,
P ⫽ .232) and between volunteers and
patients 1 year later ([3.46 ⫾ 0.03] ⫻
10⫺3 mm2/sec vs [3.43 ⫾ 0.05] ⫻ 10⫺3
mm2/sec, P ⫽ .284).
ADC Values and Clinical Comparisons
Associations between age and the measured ADC values were statistically significant; the elderly patients (ⱖ60 years
old) with SAH had higher ADC values
than did the younger patients (Table 4).
The presence of hydrocephalus, Hunt
and Hess grade, severity of bleeding
(Fisher grade), location of aneurysm,
presence of symptomatic vasospasm before MR imaging, and modality of treatment were not significantly associated
with ADC in the normal-appearing brain
parenchyma. In the control population,
the ADC values did not significantly differ between the elderly individuals (ⱖ60
years old) and the younger individuals
(⬍60 years old) in any area (P ⱖ .074),
with the single exception of the temporal gray matter (P ⫽ .032).
Discussion
We evaluated the ADC values of white
matter and deep and cortical gray matter
that had a normal appearance on diffusion-weighted images and T2-weighted
MR images in patients with SAH. In the
subacute stage of SAH, the ADC values
were increased in the white matter and
deep gray matter, with a single exception in the frontal white matter, in patients with SAH compared with healthy
volunteers. This finding suggested the
existence of mild global vasogenic edema,
rather than cytotoxic edema; cytotoxic
edema indicates ischemia or other irreversible damage of the brain tissue. Cytotoxic edema is characterized by a reduced ADC, whereas vasogenic edema
is characterized by an elevated ADC,
even if both entities may appear similar
on T2-weighted images and diffusion
trace images. To complicate the situation, in ischemic stroke with cytotoxic
edema, the areas show reduced ADC
Radiology: Volume 242: Number 2—February 2007
NEURORADIOLOGY: Subarachnoid Hemorrhage: ADC in Brain after Treatment
in the acute stage but show so-called
pseudonormalization with elevated ADC
approximately 1 week after ictus. Our
patients were imaged 5–19 days after
hemorrhage, but the pseudonormalization effect in an infarcted area can be
ruled out because there was no high T2
signal intensity in the measured areas.
Our results suggest that mild global
edema of the brain tissue is relatively
common after SAH. Previously, the frequency of occurrence of global edema
had been reported to be less than 12%
in patients with SAH at CT (12). In animal models, an increase in the water
content in the brain after rupture of an
aneurysm has been shown by calculating the weight difference between wet
and dried brain tissue and by comparing
this difference with the corresponding
value in control animals (13). To our
knowledge, global vasogenic edema after
aneurysmal SAH in the brain areas with a
normal appearance on T2-weighted MR
images and diffusion-weighted images,
however, has not been previously reported.
A number of mechanisms have been
Liu et al
proposed to explain global edema after
SAH. Those mechanisms include diffuse ischemic injury due to transient
ictal cerebral circulation arrest (14),
diffuse inflammatory or neurotoxic effects of widespread subarachnoid blood
and its degradation products on brain
tissue (15), or abnormal autoregulation (16).
Vasogenic edema with an increased
ADC has been described in hypertensive encephalopathy (17), hyperperfusion after carotid endarterectomy (18),
venous thrombosis (18), reversible posterior leukoencephalopathy (18), and
eclampsia (19). Furthermore, ADC has
been found to be increased in cerebral
edema during acute ketoacidosis (20).
Researchers have speculated that lesions in basal ganglia occur because of
increased permeability of small perforating vessels, which in turn is caused
by endothelial damage and breakdown
in cerebrovascular autoregulation from
hypertension (17). In our population of
patients with SAH, the cause of elevated
ADC values is probably multifactorial
and in part may be explained by hyper-
tensive periods, altered cerebral autoregulation, neurotoxic effects, and treatment in the intensive care unit with hypervolemia and hemodilution.
Vasogenic edema after SAH also can
be considered a direct consequence of
increased permeability of the bloodbrain barrier. In animal models, the
blood-brain barrier permeability changes
after experimental SAH reveal a reversible pattern with time (13). In the clinical setting, an impairment of the bloodbrain barrier has been found in nearly
two-fifths of patients within 5 days of
SAH. The majority of these patients developed vasospasm and ischemic complications in the later stage of SAH and
had a poor prognosis (21). In our study,
the patients underwent initial MR imaging at approximately 9 days after ictus,
and there were no significant differences
in ADC in cortical areas between patients and volunteers. One explanation
might be that the possible blood-brain
barrier dysfunction in cortical gray matter had already returned to normal by
the time of initial MR imaging. Increased
ADC values, however, were still ob-
Table 3
ADC Values of Volunteers and Patients with Aneurysmal SAH Who Participated in 1-year Follow-up
Location of Measurements
Supratentorial cortical gray matter
Parietal lobe
Frontal lobe
Temporal lobe
Occipital lobe
Supratentorial white matter
Parietal lobe
Frontal lobe
Temporal lobe
Occipital lobe
Deep gray matter
Caudate nucleus
Lentiform nucleus
Thalamus
Cerebellar gray matter
Cerebellar white matter
Pons
Volunteers
(n ⫽ 30)
Patients (n ⫽ 30)
Subacute
SAH
Follow-up
0.71 ⫾ 0.01
0.76 ⫾ 0.01
0.78 ⫾ 0.01
0.74 ⫾ 0.01
0.75 ⫾ 0.01
0.78 ⫾ 0.01
0.79 ⫾ 0.01
0.75 ⫾ 0.01
0.71 ⫾ 0.01
0.76 ⫾ 0.01
0.79 ⫾ 0.01
0.75 ⫾ 0.01
0.75 ⫾ 0.01
0.78 ⫾ 0.01
0.77 ⫾ 0.01
0.75 ⫾ 0.01
0.78 ⫾ 0.01
0.81 ⫾ 0.01
0.84 ⫾ 0.01
0.79 ⫾ 0.01
0.74 ⫾ 0.01
0.75 ⫾ 0.01
0.78 ⫾ 0.01
0.72 ⫾ 0.01
0.66 ⫾ 0.01
0.68 ⫾ 0.01
0.78 ⫾ 0.01
0.78 ⫾ 0.01
0.81 ⫾ 0.01
0.74 ⫾ 0.01
0.69 ⫾ 0.01
0.72 ⫾ 0.01
P Value
Follow-up vs
Volunteer
Subacute SAH vs
Follow-up
.025
.161
.227
.440
.912
.860
.371
.553
.025
.142
.906
.968
0.78 ⫾ 0.01
0.79 ⫾ 0.01
0.79 ⫾ 0.01
0.77 ⫾ 0.01
.013
.062
⬍.001
.001
.045
.757
.172
.140
.859
.230
.051
.447
0.74 ⫾ 0.01
0.77 ⫾ 0.01
0.78 ⫾ 0.01
0.72 ⫾ 0.01
0.68 ⫾ 0.01
0.71 ⫾ 0.01
.002
.014
.005
.107
.004
.009
.939
.075
.993
.591
.079
.151
.029
.643
.017
.065
.228
.441
Subacute SAH vs
Volunteer
Note.—The values represent estimated marginal means and standard errors of the averaged ADC values (⫻10⫺3 mm2/sec) of the left and right hemispheres. Pairwise comparisons were based
on estimated marginal means obtained from the linear mixed model for the supratentorial areas and deep gray matter and on data obtained from the general linear regression for the cerebellum
and the pons.
Radiology: Volume 242: Number 2—February 2007
523
NEURORADIOLOGY: Subarachnoid Hemorrhage: ADC in Brain after Treatment
Liu et al
Table 4
ADC versus Clinical Data in 100 Patients
Clinical Data
Age (y)
⬍60
ⱖ60
Hydrocephalus
Absent
Present
Hunt and Hess grade
II
III and IV
Fisher grade
I and II
III and IV
Aneurysm location
Anterior circulation
Vertebrobasilar artery
Symptomatic vasospasm
before MR imaging
Absent
Present
Treatment
Surgical clip
Endovascular coil
Cerebellar White
Supratentorial
Supratentorial White
Cerebellar Gray
Matter†
Cortical Gray Matter*
Matter*
Deep Gray Matter*
Matter†
No. of
P
P
P
P
P
Patients ADC
Value ADC
Value ADC
Value ADC
Value ADC
Value ADC
Pons†
P
Value
74
26
0.76 ⫾ 0.01 .198
0.77 ⫾ 0.01
0.77 ⫾ 0.01 .001
0.80 ⫾ 0.01
0.80 ⫾ 0.01 .01
0.82 ⫾ 0.01
0.74 ⫾ 0.01 .719
0.74 ⫾ 0.01
0.69 ⫾ 0.01 .208
0.71 ⫾ 0.01
0.70 ⫾ 0.01 .001
0.75 ⫾ 0.02
77
23
0.77 ⫾ 0.01 .516
0.76 ⫾ 0.01
0.79 ⫾ 0.01 .512
0.78 ⫾ 0.01
0.80 ⫾ 0.01 .448
0.81 ⫾ 0.01
0.74 ⫾ 0.01 .609
0.75 ⫾ 0.01
0.69 ⫾ 0.01 .102
0.71 ⫾ 0.01
0.71 ⫾ 0.01 .058
0.74 ⫾ 0.02
63
37
0.77 ⫾ 0.01 .702
0.76 ⫾ 0.01
0.79 ⫾ 0.01 .106
0.78 ⫾ 0.01
0.81 ⫾ 0.01 .67
0.81 ⫾ 0.01
0.74 ⫾ 0.01 .664
0.74 ⫾ 0.01
0.70 ⫾ 0.01 .87
0.70 ⫾ 0.01
0.73 ⫾ 0.01 .743
0.72 ⫾ 0.01
26
74
0.76 ⫾ 0.01 .33
0.77 ⫾ 0.01
0.78 ⫾ 0.09 .456
0.79 ⫾ 0.07
0.81 ⫾ 0.01 .89
0.81 ⫾ 0.01
0.74 ⫾ 0.01 .262
0.75 ⫾ 0.01
0.70 ⫾ 0.01 .782
0.70 ⫾ 0.01
0.72 ⫾ 0.02 .245
0.73 ⫾ 0.01
87
13
0.76 ⫾ 0.01 .962
0.76 ⫾ 0.01
0.79 ⫾ 0.01 .585
0.78 ⫾ 0.01
0.81 ⫾ 0.01 .967
0.81 ⫾ 0.01
0.74 ⫾ 0.01 .893
0.74 ⫾ 0.02
0.70 ⫾ 0.01 .799
0.70 ⫾ 0.01
0.73 ⫾ 0.01 .916
0.72 ⫾ 0.02
86
14
0.76 ⫾ 0.01 .125
0.77 ⫾ 0.01
0.79 ⫾ 0.01 .777
0.78 ⫾ 0.01
0.80 ⫾ 0.01 .298
0.82 ⫾ 0.01
0.74 ⫾ 0.01 .597
0.75 ⫾ 0.02
0.70 ⫾ 0.01 .406
0.71 ⫾ 0.01
0.73 ⫾ 0.01 .615
0.72 ⫾ 0.02
70
30
0.77 ⫾ 0.01 .588
0.76 ⫾ 0.01
0.78 ⫾ 0.01 .279
0.79 ⫾ 0.01
0.81 ⫾ 0.01 .533
0.81 ⫾ 0.01
0.74 ⫾ 0.01 .524
0.75 ⫾ 0.01
0.70 ⫾ 0.01 .749
0.70 ⫾ 0.01
0.73 ⫾ 0.02 .363
0.72 ⫾ 0.01
Note.—The values represent estimated marginal means and standard errors of the averaged ADC values of left and right hemispheres.
* Linear mixed model was used.
†
General linear regression was used.
served in cerebral and cerebellar white
matter, the pons, and deep gray matter
in the subacute stage.
The brain seems to be more vulnerable to damage in elderly patients with
SAH. In our study, the ADC values were
significantly higher in white matter and
deep gray matter in elderly patients
(ⱖ60 years old) than in younger patients (⬍60 years old). This finding indicates that the blood-brain barrier was
more severely impaired in elderly patients. This result seems to agree with
findings in previous studies (12,22) in
which older age was related to higher
morbidity and poorer neurologic outcome after SAH.
Our study results suggest that diffuse vasogenic edema is associated with
neither the patients⬘ clinical neurologic
condition nor the location of the ruptured aneurysm. Severe global edema
detectable at CT is associated with loss
524
of consciousness at ictus (12,23,24), but
no corresponding association was found
in our study. The ADC measurement is
much more sensitive than is CT in the
detection of even mild edema. In our
study, there was no significant difference in ADC values between patients
with a good clinical condition versus
those with a poorer clinical condition
(Hunt and Hess grades I and II vs grades
III and IV) nor between patients with
symptomatic vasospasm versus patients
without symptomatic vasospasm before
MR imaging. The most likely explanation for that finding might be the fact
that only normal-appearing brain tissue
was examined, and the areas with more
severe edema that showed hyperintensity on T2-weighted MR images were
excluded. Moreover, patients who currently were receiving hypertension, hypervolemia, hemodilution (HHH) treatment because of clinical vasospasm did
not undergo MR imaging because of the
use of the Swan-Ganz catheter. Thus,
we were unable to investigate the possible changes in ADC values related to
coexistent vasospasm. Patients who had
minor symptoms indicative of vasospasm at the time of MR imaging did not
receive active HHH therapy in the intensive care unit. Rupture of an aneurysm
in the posterior circulation was a predictor of vasospasm in a recent study
(25). Our results indicate that the location of the ruptured aneurysm does not
have an important role in the existence
of mild vasogenic edema.
The ADC in cortical gray matter did
not significantly differ between the volunteers and the patients on ADC maps
at both the subacute stage and 1-year
follow-up. Performance of the ROI analysis only in the cortical gray matter,
however, is demanding. The layer of the
gray matter is relatively thin, and beRadiology: Volume 242: Number 2—February 2007
NEURORADIOLOGY: Subarachnoid Hemorrhage: ADC in Brain after Treatment
cause of the partial volume effect, the
voxel in the 5-mm-thick section often
also contains subcortical white matter.
It is also important to avoid the cerebrospinal fluid in the sulci to prevent falsely
high ADC values. The ADC values in the
white matter and deep gray matter
measured in our study were comparable with those reported in the literature. There exists variation in the published values in cortical gray matter,
however, and our results tended to be
low (26,27). This finding may be related
to the different selection and size of
ROIs, and, probably, our results reflect
values that are for a combination of cortical gray and subcortical white matter.
There were limitations of our study.
First, although the ADC values in patients with SAH in the subacute stage
were significantly higher in normal-appearing white matter and deep gray
matter than in healthy volunteers, with
a single exception in frontal white matter, it is difficult to apply a universal
cutoff value in the daily clinical practice
to detect global edema. Second, all the
patients in our study received nimodipine and corticosteroids. The medication
might affect the ADC values. The effects
of these medications could not be evaluated in this study, and it was not possible to define the mechanism of vasogenic edema, nor its potential effect on
neurologic and psychological outcome,
merely with the findings of our study.
Third, the metal artifact from clips might
potentially affect the ADC; however,
the areas with a metal artifact were excluded from the ADC measurements.
Furthermore, results of the 1-year follow-up study suggested that the ADC
values had returned to normal levels
while the surgical clips and endovascular
coils remained in place. In addition, only
13 of 100 patients had aneurysms in the
posterior circulation. In the majority of
the patients, the clips were far from the
cerebellum, but the ADC values of cerebellar white matter were still higher than
they were in the control population.
To conclude, in the subacute stage
of SAH after treatment, ADC values in
the normal-appearing white matter and
deep gray matter, with a single exception in the frontal white matter, are
Radiology: Volume 242: Number 2—February 2007
higher in patients than they are in a
healthy control population. The possible explanations for this observation
include hypertensive periods, altered
autoregulation, neurotoxic effects, and
prophylactic treatment with hypervolemia and hemodilution. Clinical and neuropsychological outcome after SAH depends on many different factors that often
are interrelated. The possible predictive
value for outcome of elevated ADC values
in the subacute stage needs further study.
Acknowledgment: We express our gratitude to
Vesa Kiviniemi, PhLic, for statistical consultation.
Liu et al
bral vasospasm after aneurysmal subarachnoid
hemorrhage. Stroke 1995;26:1620 –1626.
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:
1225–1232.
13. Germano A, d’Avella D, Imperatore C, Caruso G,
Tomasello F. Time-course of blood-brain barrier
permeability changes after experimental subarachnoid haemorrhage. Acta Neurochir (Wien)
2000;142:575–580.
14. Shigeno T, Fritschka E, Brock M, Schramm J,
Shigeno S, Cervos-Navarro J. Cerebral edema
following experimental subarachnoid hemorrhage. Stroke 1982;13:368 –379.
15. Germano AF, Dixon CE, d’Avella D, Hayes RL,
Tomasello F. Behavioral deficits following experimental subarachnoid hemorrhage in the rat.
J Neurotrauma 1994;11:345–353.
1. Rordorf G, Koroshetz WJ, Copen WA, et al.
Diffusion- and perfusion-weighted imaging in vasospasm after subarachnoid hemorrhage. Stroke
1999;30:599 – 605.
16. Handa Y, Hayashi M, Takeuchi H, Kubota T,
Kobayashi H, Kawano H. Time course of the
impairment of cerebral autoregulation during
chronic cerebral vasospasm after subarachnoid
hemorrhage in primates. J Neurosurg 1992;76:
493–501.
2. Condette-Auliac S, Bracard S, Anxionnat R, et al.
Vasospasm after subarachnoid hemorrhage: interest in diffusion-weighted MR imaging. Stroke
2001;32:1818 –1824.
17. Schwartz RB, Mulkern RV, Gudbjartsson H,
Jolesz F. Diffusion-weighted MR imaging in hypertensive encephalopathy: clues to pathogenesis. AJNR Am J Neuroradiol 1998;19:859 – 862.
3. Leclerc X, Fichten A, Gauvrit JY, et al. Symptomatic vasospasm after subarachnoid haemorrhage:
assessment of brain damage by diffusion and perfusion-weighted MRI and single-photon emission
computed tomography. Neuroradiology 2002;44:
610 – 616.
18. Schaefer PW, Gonzalez RG, Hunter G, Wang B,
Koroshetz WJ, Schwamm LH. Diagnostic value
of apparent diffusion coefficient hyperintensity in
selected patients with acute neurologic deficits.
J Neuroimaging 2001;11:369 –380.
References
4. Phan TG, Huston J 3rd, Campeau NG, Wijdicks
EF, Atkinson JL, Fulgham JR. Value of diffusionweighted imaging in patients with a nonlocalizing
examination and vasospasm from subarachnoid
hemorrhage. Cerebrovasc Dis 2003;15:177–181.
5. Hunt WE, Hess RM. Surgical risk as related to
time of intervention in the repair of intracranial
aneurysms. J Neurosurg 1968;28:14 –20.
6. Fisher CM, Kistler JP, Davis JM. Relation of
cerebral vasospasm to subarachnoid hemorrhage
visualized by computerized tomographic scanning. Neurosurgery 1980;6:1–9.
7. Liu Y, Karonen JO, Vanninen RL, et al. Detecting
the subregion proceeding to infarction in hypoperfused cerebral tissue: a study with diffusion
and perfusion weighted MRI. Neuroradiology
2003;45:345–351.
8. American Heart Association. International cardiovascular disease statistics. Dallas, Tex:
American Heart Association, 2003.
9. Bederson JB, Awad IA, Wiebers DO, et al. Recommendations for the management of patients
with unruptured intracranial aneurysms: a statement for healthcare professionals from the
Stroke Council of the American Heart Association. Circulation 2000;102:2300 –2308.
10. Awad IA, Carter LP, Spetzler RF, Medina M,
Williams FC Jr. Clinical vasospasm after subarachnoid hemorrhage: response to hypervolemic hemodilution and arterial hypertension.
Stroke 1987;18:365–372.
11. Mori K, Arai H, Nakajima K, Tajima A, Maeda
M. Hemorheological and hemodynamic analysis
of hypervolemic hemodilution therapy for cere-
19. Engelter ST, Provenzale JM, Petrella JR. Assessment of vasogenic edema in eclampsia using diffusion imaging. Neuroradiology 2000;42:818 –
820.
20. Glaser NS, Wootton-Gorges SL, Marcin JP, et al.
Mechanism of cerebral edema in children with
diabetic ketoacidosis. J Pediatr 2004;145:164 –
171.
21. Doczi T. The pathogenetic and prognostic significance of blood-brain barrier damage at the acute
stage of aneurysmal subarachnoid haemorrhage:
clinical and experimental studies. Acta Neurochir
(Wien) 1985;77:110 –132.
22. Ogilvy CS, Carter BS. A proposed comprehensive grading system to predict outcome for surgical management of intracranial aneurysms. Neurosurgery 1998;42:959 –968.
23. Lagares A, Gomez PA, Alen JF, Lobato RD, Campollo J. Global cerebral edema after subarachnoid hemorrhage [letter]. Stroke 2002;33:2153–
2154.
24. Claasen J, Mayer SA. Reply. Stroke 2002;33:
2154.
25. McGirt MJ, Mavropoulos JC, McGirt LY, et al.
Leukocytosis as an independent risk factor for
cerebral vasospasm following aneurysmal subarachnoid hemorrhage. J Neurosurg 2003;98:
1222–1226.
26. Ulug AM, Beauchamp N Jr, Bryan RN, van Zijl
PC. Absolute quantitation of diffusion constants
in human stroke. Stroke 1997;28:483– 490.
27. Helenius J, Soinne L, Perkiö J, et al. Diffusionweighted MR imaging in normal human brains in
various age groups. AJNR Am J Neuroradiol
2002;23:194 –199.
525