From the Departments of Radiology and Neurology (L.J.K.), University
Hospital Utrecht (Netherlands).
Methods—We performed DWI, fluid-attenuated inversion recovery,
spin-echo T2-weighted MRI, and spin-echo proton density–weighted MRI
in 42 patients with acute stroke and 15 control subjects. The volume of
ischemic lesions was measured on early (<60 hours after onset)
and follow-up MRI scans. Clinical outcome was measured 4 months after
onset of symptoms with the National Institutes of Health Stroke Scale,
the Barthel Index, and the Rankin Scale.
Results—With DWI, 98% of the ischemic lesions were
detected, and with fluid-attenuated inversion recovery, 91% were
detected, whereas with early T2-weighted or proton density–weighted
scans, only 71% (P=0.002,
Conclusions—DWI is a better imaging method than conventional MRI
in detecting early ischemic lesions in stroke patients. Lesion
size as measured on DWI scans and, to a lesser extent, ADC values are
potential parameters for predicting clinical outcome in
acute stroke patients.
In the ealy stage of ischemia, CT is the technique of
choice to distinguish between hemorrhagic and nonhemorrhagic stroke.
However, with CT 30% to 60% of the ischemic lesions are still
invisible in the acute stage.1 2 3 4 During the
first 24 hours after an ischemic stroke, proton
density–weighted (PD-w) and T2-w MRI have 20% to 30% false-negative
results.5 6 7 This percentage increases to 30% to
50% during the first 3 to 6 hours after
stroke.3 8 As a consequence, CT or conventional
MRI is not generally used to predict the presence and extent of
ischemic damage in the acute stage after stroke.
Cytotoxic edema, which is caused by the accumulation of
intracellular water due to cell membrane damage minutes after onset of
acute cerebral ischemia, causes a restriction of microscopic
proton diffusion. In diffusion-weighted MRI, this decrease in water
diffusion is presumably reflected in a decrease of the apparent
diffusion coefficient (ADC) on ADC trace
maps,9 10 11 12 13 14 15 which is visualized as a
hyperintensity on the diffusion-weighted images
(DWI).12 14 15 Previous animal studies showed
that DWI is able to visualize cerebral ischemic changes within
5 minutes9 to 1 to 3 hours after onset of
symptoms.10 16 17 18 19 In humans, ischemic
changes were detected with DWI as early as 2 to 6 hours after onset of
symptoms.15 20 21 22 Other advantages of DWI are
the low number of false-negative investigations
(5%),23 the clear discrimination between
ischemic lesions and the nonischemic
brain,11 15 24 and the discrimination between
acute and chronic ischemic lesions.11 23
With these features, DWI facilitates the determination of the type,
site, and extent of cerebral ischemia at an early stage. This
might help to predict the clinical outcome of stroke patients.
Recent studies showed that in the acute stage after stroke, DWI is more
sensitive for early ischemic changes than T2-w
MRI.11 14 15 20 21 22 25 However, other studies
showed that both PD-w imaging6 8 and
fluid-attenuated inversion recovery (FLAIR)
imaging26 27 are superior to T2-w imaging in the
detection of acute ischemic lesions. Therefore, DWI should be
compared with PD-w and FLAIR imaging as well.
The purpose of this observational study was to evaluate the use of DWI
in the early phase (<60 hours after onset of symptoms) of
ischemic stroke. We therefore investigated the sensitivity of
DWI in detecting early cerebral ischemic changes in comparison
with T2-w, PD-w, and FLAIR MRI scans. Furthermore, we assessed the
prognostic value of lesion volume on early DWI scans with respect to
final infarct volume and clinical outcome. Finally, we assessed the
prognostic value of ADC measurements in ischemic lesions with
respect to clinical outcome.
The control group consisted of 9 men and 6 women who had no history of
ischemic neurological deficits and had no other intracranial
diseases (mean age, 43 years; range, 22 to 71 years). They underwent
the same DWI protocol as the patients. The study was approved by the
Human Research Committee of our hospital.
MR Protocol
Image Analysis
ADC Measurements
Clinical Assessment
Statistical Analysis
The results for the detection of lesions on early T2-w, PD-w,
FLAIR, and DWI scans are summarized in Table 2
In Table 3
Long-term clinical outcome was measured between 7 and 33 weeks
after onset of symptoms (mean, 17 weeks). During the follow-up period,
4 patients died. One patient with an infarct on follow-up MRI died from
a pneumonia 3 months after onset of symptoms; another patient died from
a retroperitoneal bleeding 2 weeks after the cerebral infarct. The
other 2 patients had transient symptoms of cerebral ischemia
without signs of an ischemic infarct on DWI or follow-up scan.
One of these 2 patients died from a recurrent stroke in the posterior
fossa 2 weeks after a hemispheric TIA; the other patient died from an
unknown cause 3 months after the TIA. These 4 patients were excluded
from further analysis because the baseline ischemic
event was not the direct cause of their death.
Spearman rank sum correlations of lesion volume, as measured on the
early DWI scans and on the follow-up MRI scans, versus clinical outcome
ratings are shown in Table 4
In Figure 6
Table 5
The volume of the lesions that were imaged within the first 12 hours
after onset of symptoms was significantly increased on the follow-up
MRI. Enlargement of cerebral infarcts has been demonstrated previously
by Baird and coworkers,39 who defined lesions
that were
Our finding that patients with poor outcome had larger lesions
than patients with good outcome was confirmed in a previous study in
which early T2-w MRI was used.41 Compared with
our results, similar correlations were reported between lesion volume
on T2-w images and follow-up BI score42 and
between infarct size on CT scans, performed between 7 and 10 days after
stroke, and the NIHSS score at 3 months after the onset of
symptoms.1 The observed correlations between
lesion volume on early DWI scans and clinical severity ratings have
been described before21 43 and suggest that
lesion volume can be predictive of the clinical outcome of stroke
patients, particularly in patients without previous ischemic
strokes.
For our entire patient population, good outcome could be
predicted with a sensitivity of 79% and a specificity of 89% if
patients had ischemic lesions of
Our measurements of the ADC were reproducible in controls and in
normal-appearing white and gray matter areas of stroke patients. Our
ADC measurements of normal-appearing brain tissue were in the same
range as the ADC measurements of other studies in human
stroke patients.11 12 13 15 20 22 45 46 Marks and
coworkers15 reported that the ADC varies within
ischemic lesions. Our experience confirmed this finding. In our
patient population the ADC varied widely between various locations
within ischemic lesions, and the ADC was usually lower in the
lesion core than in more peripheral parts. A number of
studies11 12 20 22 46 determined the ADC of the
core of the lesion only, whereas in our study the mean ADC was
determined over an entire slice of the ischemic region. This
might explain why in our study the ADCr between ischemic
regions and corresponding contralateral areas were higher than the ADCr
reported elsewhere.11 12 20 22 46 The presence of
cerebrospinal fluid in the investigated regions might have caused
variations in the ADC as well. The variations in ADC within
ischemic lesions might become an important issue for future
research in acute stroke patients. Animal studies indicate that the
variation in ADC corresponds with regional cerebral blood
flow,19 the severity47 and
extent9 47 of neuronal damage, and the degree of
regional metabolic alterations.10
These data suggest that DWI can depict not only severely
ischemic lesions but also moderately compromised regions that
might be salvageable if therapeutic intervention is applied.
The DWI technique used in this study required 1.5 to 2 minutes for each
diffusion direction, which is longer than is required by DWI techniques
reported elsewhere.12 14 20 22 In our study
prolonged scan time was a trade-off with the signal-to-noise ratio and
the number of slices. A disadvantage of longer scan time is the higher
susceptibility to motion artifacts. However, even though we used this
relatively long DWI technique, only 1 patient had to be excluded from
our study because of motion artifacts.
We conclude that DWI is a more reliable imaging method than MRI in
detecting ischemic lesions in acute stroke patients. Lesion
size, as measured on DWI scans, appears to be a potential
parameter for predicting clinical outcome in acute stroke
patients.
Received March 23, 1998;
revision received May 29, 1998;
accepted May 29, 1998.
© 1998 American Heart Association, Inc.
Original Contributions
Diffusion-Weighted Magnetic Resonance Imaging in Acute Stroke
Abstract
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Background and
Purpose—Diffusion-weighted MRI (DWI) is highly sensitive in
detecting early cerebral ischemic changes in acute stroke
patients. In this study we compared the sensitivity of DWI with that of
conventional MRI techniques. Furthermore, we investigated the
prognostic value of the volume of ischemic lesions on DWI scans
and of the apparent diffusion coefficient (ADC). 
2) and 80%
(P=0.02, 2) of lesions, respectively,
were found. Lesion volume on early DWI scans correlated significantly
with clinical outcome ratings (P<0.01). In patients
with a first-ever stroke, a lesion volume of 
22 mL on DWI predicted
good outcome with a 75% sensitivity and a 100% specificity. The mean
ADC of ischemic lesions was 29% lower than the ADC of
normal-appearing parts of the brain (P<0.001). The ADC
ratio correlated significantly with clinical outcome
(P<0.05).
Key Words: cerebral infarction • magnetic resonance imaging • stroke assessment • stroke, acute
Introduction
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
In stroke patients,
imaging of the brain with CT or T2-weighted (T2-w) MRI is an important
tool to distinguish between hemorrhagic and nonhemorrhagic stroke. Once
the presence of an intracranial hemorrhage is excluded, imaging
can also help to classify the stroke subtype by determining the site
and size of the ischemic lesion. Eventually, visualization of
the lesion may help to predict the functional consequences of the
event.
Subjects and Methods
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Patients and Controls
Between October 1996 and November 1997 we investigated 47
patients in the early phase of hemispheric ischemia (<60 hours
after onset), as diagnosed by a neurologist. After a routine CT scan
was performed, MRI was performed after informed consent was obtained
from the patient or a representing family member. Patients
who needed artificial respiration and patients who were otherwise
unable to undergo the MR examinations were excluded. Four patients were
treated with low, intermediate, or high intravenous doses
of heparinoid (Org 10172, Organon) as part of the
EuroTOAST (Trial of Org 10172 in Acute Stroke Treatment) investigation
(Table 1
, patients 8, 10, 12, and 14). Otherwise, patients received
aspirin (n=32) or oral anticoagulants (n=6) together with conventional
supportive therapy for acute stroke. Patients receiving experimental
neuroprotective agents were not included.
View this table:
[in a new window]
Table 1. Demographic Data, Clinical Symptoms, and Infarct
Location on Follow-up
Scan
All MR investigations were performed on a 1.5-T clinical imaging
system (Philips Gyroscan ACS-NT 15 whole-body system, Philips Medical
Systems) in the early phase (<60 hours) after onset of symptoms. A
sagittal spin-echo T1-weighted sequence (repetition time [TR], 427
ms; echo time [TE], 12 ms; 16 slices; slice thickness, 7.0 mm;
0.7-mm interslice gap; 1 average), an axial spin-echo PD-w and T2-w
sequence (dual-echo T2-w MRI: TR, 2014 ms; TE, 30 ms and 100 ms; 19
slices; slice thickness, 7.0 mm; 0.7-mm interslice gap; 1 average;
matrix size, 256x256; scan reduction, 75%) and an axial FLAIR
sequence (Turbo Spin Echo, turbo factor 11 [ie, acquiring 11 echoes
per TR]; TR, 5496 ms; TE, 100 ms; inversion time, 2000 ms; 16
slices; slice thickness, 6.4 mm; 1.3-mm interslice gap; 2
averages, matrix size, 256x256; reduced acquisition, 90% [90% of
data points acquired]) were performed. For DWI, a fat-suppressed
(Spectral Presaturation with Inversion Recovery), multishot,
spin-echo/echo planar imaging sequence was used, with a pair of
diffusion gradients centered around a 180° pulse (TE, 140 ms; TR, 667
or 1200 ms; 2 averages; matrix size, 128x128; reduced
acquisition, 75% [96 of 128 data points acquired]; echo planar
imaging factor, 13 [13 readout gradients per TR]). To reduce motion
artifacts, the head of the patient was supported with soft supportive
wedges and straps. For the DWI scans, navigator echoes were used to
compensate for patient movements during the acquisitions, and cardiac
triggering was used to reduce artifacts from pulsatile brain motion.
Use of this sequence permitted acquisition of 10 to 12 slices
(depending on the heart rate), which allowed a total coverage of 72 to
86.4 mm of the brain (slice thickness, 6.0 mm; 1.2-mm
interslice gap). DWI scans were acquired with diffusion gradients along
each of the three principal axes with 3 different b values (0, 187, and
757 s/mm2). For patient handling and performing
the entire imaging protocol, 
20 minutes were needed, of which 6
minutes were needed for DWI. A follow-up MRI or CT scan was made 
6
days after the onset of symptoms. The follow-up MRI scan consisted of
sagittal T1-w and axial PD-w, T2-w, and FLAIR scans. Total study time
for the follow-up examination was 10 minutes.
The early T2-w, PD-w, and FLAIR images were examined separately
from the follow-up MRI scans by an experienced neuroradiologist
(L.M.P.R.) without knowledge of the patients' symptoms and without
reference to the DWI results. The ischemic lesions were
measured on a separate workstation: areas of abnormal hyperintensity
were traced on each slice of the PD-w (or FLAIR) images and on each
slice of the DWI scans. These areas of abnormal hyperintensity were
summed and multiplied with the slice thickness (plus interslice gap) to
calculate the volumes of the acute ischemic lesions. To measure
the lesion size on DWI scans we used the images that were acquired with
the highest b value only, and all 3 diffusion directions were
considered. Hyperintensities that were visible on only one or two
diffusion directions were not considered ischemic lesions,
since these hyperintensities are most likely to be caused by diffusion
anisotropy.28 29 30 If patients had 2 fresh
ischemic lesions, the volumes of both lesion were summed to
analyze the prognostic value of infarct volume on clinical
outcome. Preexistent cerebral infarcts were not included in these
measurements.
ADC trace maps were calculated off-line afterward on the basis
of the DWI that were acquired over the 3 principal
axes.13 Measurements of the ADC of
normal-appearing brain were obtained in patients and in controls from
various regions of interest. Regions of interest were between 100 and
300 mm2 in size and were carefully placed in
an attempt to exclude sulcal or intraventricular
cerebrospinal fluid, which has a high ADC value. To calculate the mean
ADC of ischemic lesions (ADClesion), the
region of interest outlining the ischemic lesion on DWI was
transferred to the corresponding ADC trace map. Ratios were calculated
between the ADC of normal-appearing brain in the right and left
hemispheres and between the ADC of the ischemic area and the
corresponding contralateral region.
In the early stage, the patients' clinical status was
determined with the National Institutes of Health Stroke Scale
(NIHSS)31 and the modified Rankin Scale (RS).
Long-term clinical outcome was determined with the NIHSS, RS, and
Barthel Index (BI). The NIHSS is a serial scale of neurological deficit
that measures 11 items of the patients' neurological status. The
reliability of the NIHSS was previously demonstrated in a number of
studies.31 32 33 34 35 36 We used the 31-point scale
proposed by Goldstein et al.31 The modified
RS37 is a simplified overall assessment of
function. A score of 0 indicates no symptoms, and a score of 5
indicates severe disability. In this study a RS score of 0 to 3 was
considered good outcome, whereas a RS score of 4 or 5 was considered
poor outcome. The BI38 is a reliable and valid
measure of the ability to perform activities of daily living such as
eating, dressing, bathing, walking, and using the toilet. A score of
100 indicates no disability, and a score of 0 indicates severe
disability.
For statistical analysis, paired analysis of
differences was performed with the Students' t test (after
the F test was performed) or with the nonparametric
Wilcoxon rank sum test. The paired samples t test
was used for comparison of infarct volumes. Infarct volumes are
expressed as mean and range. Pearson correlations were calculated
between lesion volumes on the early MR images and follow-up scans.
Spearman rank correlations were calculated between lesion volume and
clinical outcome ratings. For the 2 test, the
Yates correction for continuity was used. For the comparison of outcome
ratings between patients with large or small acute ischemic
lesions and between patients with high or low ADC of the acute infarct,
the Mantel-Haenszel 2 test was used. A value
of P<0.05 was considered statistically significant.
Statistical significance was corrected for repeated measures. Receiver
operating characteristics analysis was used to calculate the
lesion volume, which determines clinical outcome of patients with
optimum sensitivity and specificity.
Results
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Forty-seven patients initially fulfilled our inclusion criteria,
but 3 patients were too uncooperative to undergo the MRI examination,
and in 2 patients the DWI failed because of patient motion (n=1) or
technical problems (n=1). We therefore analyzed the results of
42 patients with signs and symptoms of hemispheric ischemia of
acute onset (25 men, 17 women; mean age, 64 years; range, 36 to 85
years). The demographic data and current clinical symptoms are shown in
Table 1
. Before the current
ischemic event, 9 patients (21%) had suffered from a minor
ischemic stroke and 8 patients (19%) had suffered from a
transient ischemic attack (TIA). The mean time between the
onset of symptoms and the early MR investigations was 23 hours (range,
1.5 to 60 hours). Five patients were imaged in the hyperacute stage
(<6 hours after onset), 5 patients were imaged between 6 and 12
hours after onset, 15 patients were imaged between 12 and 24 hours, and
17 patients were imaged between 24 and 60 hours after onset of
symptoms. The mean time between the onset of symptoms and the follow-up
MRI was 17 days (range, 6 to 49 days). In 3 patients a follow-up CT
scan was performed instead of an MRI (patients 5,12, and 42). . For this
analysis, the follow-up scans were used as the gold standard.
Four of the 42 patients (10%) had no lesion evident on the follow-up
scans. These patients had transient symptoms and also had no lesions on
the early MRI or DWI scans. In 7 patients, 2 infarcts were found on the
follow-up MRI. Thus, a total of 45 infarcts were analyzed.
During the first 60 hours after stroke, more lesions were detected on
the DWI scans (98%) than on the T2-w images (71%; P=0.002,
2) or PD-w images (80%; P=0.02,
2). FLAIR imaging was performed in 32 of the
42 patients (35 lesions). With FLAIR, 91% of the lesions was detected
(FLAIR versus DWI; P=NS, 2). No
false-positive lesions were found with DWI: all hyperintensities
resembling an ischemic lesion on DWI evolved into infarcts on
the follow-up scan. A typical example of DWI, T2-w, and PD-w scans of a
patient in the hyperacute stage after stroke is shown in Figure 1. Figure 2
shows the DWI, T2-w, PD-w, and FLAIR scans of another patient who was
imaged 22 hours after the onset of symptoms.
View this table:
[in a new window]
Table 2. Number (and Percentage) of Ischemic Lesions
Detected on T2-w, PD-w, FLAIR, and DWI Scans in the Acute Stage After
Stroke Compared With the Total Number of Infarcts on Follow-Up
MRI
View larger version (70K):
[in a new window]
Figure 1. Fifty-five–year-old man with an acute left-sided
hemiparesis 6 hours before the first MRI examination (patient 10). On
the early PD-w (A) and T2-w (B) images, no ischemic lesion was
visible. The early DWI scan (C) shows a right-sided hyperintensity in
the frontal lobe (territory of the pericallosal artery), which can be
appreciated as a hypointensity on the ADC trace map (D). The follow-up
MRI (E) was performed 6 days after the onset of symptoms and confirms
the infarct. However, on the follow-up scan the infarct extended into
the corpus callosum.
View larger version (69K):
[in a new window]
Figure 2. Thirty-six–year-old woman with a right-sided
hemiparesis and dysphasia, starting 22 hours before the first MRI
examination (patient 28). The left cortical infarct was initially not
recognized on the early PD-w (A) and T2-w (B) images. In retrospect,
however, the infarct was found on the PD-w images. The early FLAIR (C)
and DWI (D) scans clearly show the infarct. , the sizes of
the lesions as measured on the DWI scans, the early PD-w/FLAIR scans,
and the follow-up MRI scans are summarized. The 3 patients with a
follow-up CT scan only were excluded from this analysis. The
mean lesion volume was significantly larger on the DWI scans (35 mL)
than on the early PD-w/FLAIR images (21 mL) if patients were imaged
between 0 and 12 hours after onset of symptoms (P<0.05). In
patients who were imaged within 6 hours after stroke the difference in
lesion size on the DWI scans (34 mL) and on the PD-w/FLAIR images (15
mL) appears to be even more profound than in patients who were imaged
between 6 and 12 hours (36 mL versus 27 mL, respectively). However,
this difference did not reach statistical significance because of the
limited number of patients. No significant differences in mean lesion
volume between the DWI scans and PD-w/FLAIR images were found if
patients had the first MRI investigation between 12 and 24 hours after
onset of symptoms or 24 hours. Lesions were significantly smaller on
the early DWI scans (mean lesion size, 35 mL) than on the follow-up MRI
scans (mean lesion size, 45 mL) if patients were imaged <12 hours
after onset of symptoms (P<0.05). Again, in patients who
were imaged <6 hours or between 6 and 12 hours after onset of
symptoms, the difference in lesion size did not reach statistical
significance because of the small sample size. The mean volume of
lesions imaged 12 hours after onset of symptoms was not significantly
altered on the follow-up scans.
View this table:
[in a new window]
Table 3. Mean Lesion Size (and Range) on Early PD-w/FLAIR
Images, DWI Scans, and Follow-Up MRI . The corresponding data
are plotted in Figures 3 through 5.
Lesion volume on the early DWI scans correlated significantly with the
early NIHSS and RS scores (n=38, Figure 3) and with the long-term
clinical outcome ratings (n=38, Figure 4). Lesion volume on the follow-up MR
scans correlated significantly with the long-term clinical outcome
ratings as well (n=36, Figure 5). The
correlations between lesion volume on the early DWI scans and long-term
clinical outcome ratings were increased when the 9 patients who had an
ischemic event before the current event were excluded from the
analysis: NIHSS: r=0.71, P<0.001; RS:
r=0.70, P<0.001; BI: r=-0.55,
P<0.001.
View this table:
[in a new window]
Table 4. Spearman Rank Sum Correlations Between Lesion Size
on DWI, Infarct Size on Follow-Up MRI, and ADC Ratios vs Clinical
Outcome
Ratings
View larger version (12K):
[in a new window]
Figure 3. Correlation and 95% CIs for individuals between
lesion volume on early DWI scans and early NIHSS (A)
(r= 0.63, P<0.001) and RS (B)
(r=0.62, P<0.001) scores.
View larger version (10K):
[in a new window]
Figure 4. Correlation and 95% CIs for individuals between
lesion volume on early DWI scans and follow-up NIHSS (A)
(r=0.62, P<0.001), BI (B)
(r=-0.47, P<0.01), and RS (C)
(r=0.62, P<0.001) scores.
View larger version (10K):
[in a new window]
Figure 5. Correlation and 95% CIs for individuals between
lesion volume on follow-up MR scans and follow-up NIHSS (A)
(r=0.66, P<0.001), follow-up BI
(B) (r=-0.49, P<0.01), and RS
(C) (r=0.66, P<0.001) scores. , patients
are grouped by the volume of the ischemic lesion on DWI.
Patients with a relatively good outcome (n=29) had significantly
smaller ischemic lesions on DWI scans (mean lesion volume, 18
mL) than patients with a poor outcome (n=9; mean lesion volume, 92 mL;
P<0.05). With receiver operating characteristics
analysis, a lesion volume of 22 mL on early DWI scans
was found to determine the patients' outcome to be good with a
sensitivity of 79% and a specificity of 89% (95% CI, 0.70 to 0.95).
Clinical outcome ratings were significantly worse in patients with
infarcts >22 mL on DWI than in patients with smaller infarcts (NIHSS:
P=0.001; BI: P=0.01; RS: P=0.001;
Figure 5). When the 9 patients who previously suffered from an
ischemic stroke were excluded from this analysis, the
sensitivity and specificity were 75% and 100%, respectively (95% CI,
0.78 to 0.98).
View larger version (31K):
[in a new window]
Figure 6. Long-term outcome, as measured with the NIHSS
(top), BI (middle), and RS (bottom) scores of acute
stroke patients. Patients are categorized as having large (>22 mL,
n=14) or small ( 22 mL, n=24) ischemic lesions on early DWI
scans. summarizes the ADC values of patients and
control subjects. In both patients and controls, ADC measurements were
performed in various normal-appearing parts of the brain. No
significant difference in mean ADC was found between patients and
controls in corresponding normal-appearing areas in the cerebrum. No
significant difference in ADC was found between the various regions in
the brain parenchyma of patients or controls. The mean ADC of the
ischemic lesions (ADClesion) was 29%
lower than the mean ADC in corresponding areas of the contralateral
hemispheres (ADCcontralateral;
P<0.001). Significant correlations were found between the
ADC ratio
(ADCr=ADClesion/ADCcontralateral)
and the follow-up NIHSS, BI, and RS scores (Table 4). Patients with a
poor outcome had a significantly lower ADCr of the ischemic
lesions (mean ADCr=0.59±0.14) than patients with a good outcome (mean
ADCr=0.77±0.15). An ADCr limit of 0.62 was found to determine the
patients' outcome to be good or poor with a sensitivity of 83% and a
specificity of 75% (95% CI, 0.66 to 0.94). Patients with low ADCr
(<0.62) had significantly worse clinical outcome than patients with
high ADCr (0.62, NIHSS: P=0.024; BI: P=0.001;
RS: P=0.005).
View this table:
[in a new window]
Table 5. Mean Absolute ADC of Both Hemispheres and
Ipsilateral/Contralateral Ratios in Various Normal-Appearing Cerebral
Regions in Patients and Controls and Mean Absolute ADC (and
Ipsilateral/Contralateral Ratio) of Acute Ischemic Lesions and
Corresponding Contralateral Regions
Discussion
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
Our results confirm that DWI is a useful imaging method to detect
ischemic lesions in the early stage after stroke. All but one
of the infarcts on follow-up MRI were depicted as a hyperintense lesion
on DWI, whereas with T2-w or PD-w imaging, 29% and 20% of lesions,
respectively, were not found in the early stage. Although less accurate
than DWI, the FLAIR imaging provided better results in detecting
ischemic lesions than the T2-w or PD-w scans. 20% larger on follow-up MRI scans than on early DWI scans
as significantly enlarged. When we analyzed our data
accordingly, 41% of the lesions were significantly enlarged, which is
comparable to the 43% reported by Baird et al. A probable explanation
for the enlargement of the ischemic lesions is that areas
surrounding the core of the lesion became infarcted after we performed
the MRI investigation.40 Another explanation for
lesion enlargement is the presence of vasogenic edema. Since we
performed the follow-up examination between 6 and 49 days (mean, 17
days) after onset of symptoms, vasogenic edema surrounding the
infarcted area may have mimicked lesion enlargement. 22 mL on early DWI scans.
One of our patients had very poor clinical outcome ratings, although
she only had a small infarct. This patient had suffered from an
ischemic stroke in the contralateral hemisphere 2 years
earlier, from which she was left with a mild left-sided hemiparesis.
The presence of infarcts in both hemispheres explained her low clinical
outcome ratings. In patients who had never suffered from an
ischemic event before or patients who previously had a TIA
only, good outcome could be predicted with a specificity of 100%. This
means that all patients without a previous stroke had good outcome if
the acute ischemic lesion was 22 mL. These results
show that lesion size measurements on early DWI scans can provide
prognostic information to the clinician. Early information about the
prognosis of acute stroke patients is useful if acute treatment with
tissue plasminogen activator is
considered.44 One might speculate that patients
with ischemic lesions 22 mL on DWI scans should not be
treated with tissue plasminogen activator
because the potential benefit is not outweighed by the risk of
complications. However, the presented cutoff value of 22 mL
should be interpreted with care because ischemic lesions tend
to increase in size during the first 12 hours after onset of symptoms.
To establish a critical lesion size in the hyperacute stage after
stroke (0 to 6 hours), more patients should be investigated.
Acknowledgments
Dr van der Grond is clinical investigator of the Netherlands
Heart Foundation. We thank Dr A. Algra of the Julius Center for Patient
Oriented Research for help in statistical analysis, N.W.A.M.
Swinkels-Noppen of the Department of Radiology for assistance in image
analysis, and Dr Ir P.J.M. Folkers from Philips Medical
Systems for technical assistance in optimizing the MRI
protocol.
Footnotes
Reprint requests to K.J. van Everdingen, MD, Department of Radiology, E01.132, University Hospital Utrecht, Heidelberglaan 100, 3584 CX Utrecht, Netherlands.
References
Top
Abstract
Introduction
Subjects and Methods
Results
Discussion
References
This article has been cited by other articles:
 |
|
 |
|
  E. S. Roach, M. R. Golomb, R. Adams, J. Biller, S. Daniels, G. deVeber, D. Ferriero, B. V. Jones, F. J. Kirkham, R. M. Scott, et al. Management of Stroke in Infants and Children: A Scientific Statement From a Special Writing Group of the American Heart Association Stroke Council and the Council on Cardiovascular Disease in the Young Stroke, September 1, 2008; 39(9): 2644 - 2691. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. P. Adams Jr, G. del Zoppo, M. J. Alberts, D. L. Bhatt, L. Brass, A. Furlan, R. L. Grubb, R. T. Higashida, E. C. Jauch, C. Kidwell, et al. Guidelines for the Early Management of Adults With Ischemic Stroke: A Guideline From the American Heart Association/American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups: The American Academy of Neurology affirms the value of this guideline as an educational tool for neurologists. Circulation, May 22, 2007; 115(20): e478 - e534. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. N.E. Redgrave, S. B. Coutts, U. G. Schulz, D. Briley, and P. M. Rothwell Systematic Review of Associations Between the Presence of Acute Ischemic Lesions on Diffusion-Weighted Imaging and Clinical Predictors of Early Stroke Risk After Transient Ischemic Attack Stroke, May 1, 2007; 38(5): 1482 - 1488. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. P. Adams Jr, G. del Zoppo, M. J. Alberts, D. L. Bhatt, L. Brass, A. Furlan, R. L. Grubb, R. T. Higashida, E. C. Jauch, C. Kidwell, et al. Guidelines for the Early Management of Adults With Ischemic Stroke: A Guideline From the American Heart Association/ American Stroke Association Stroke Council, Clinical Cardiology Council, Cardiovascular Radiology and Intervention Council, and the Atherosclerotic Peripheral Vascular Disease and Quality of Care Outcomes in Research Interdisciplinary Working Groups: The American Academy of Neurology affirms the value of this guideline as an educational tool for neurologists Stroke, May 1, 2007; 38(5): 1655 - 1711. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Cook, J. Huston III, M. R. Trenerry, R. D. Brown Jr, K. J. Zehr, and T. M. Sundt III Postcardiac Surgical Cognitive Impairment in the Aged Using Diffusion-Weighted Magnetic Resonance Imaging Ann. Thorac. Surg., April 1, 2007; 83(4): 1389 - 1395. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Banks and C. A. Marotta Outcomes Validity and Reliability of the Modified Rankin Scale: Implications for Stroke Clinical Trials: A Literature Review and Synthesis Stroke, March 1, 2007; 38(3): 1091 - 1096. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Schiemanck, G. Kwakkel, M. W. M. Post, and A. J. H. Prevo Predictive Value of Ischemic Lesion Volume Assessed With Magnetic Resonance Imaging for Neurological Deficits and Functional Outcome Poststroke: A Critical Review of the Literature. Neurorehabil Neural Repair, December 1, 2006; 20(4): 492 - 502. [Abstract] [PDF]
|
|
|
|
|
|
M. Luby, J. L. Bykowski, P. D. Schellinger, J. G. Merino, and S. Warach Intra- and Interrater Reliability of Ischemic Lesion Volume Measurements on Diffusion-Weighted, Mean Transit Time and Fluid-Attenuated Inversion Recovery MRI Stroke, December 1, 2006; 37(12): 2951 - 2956. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Kastrup, T. Nagele, K. Groschel, F. Schmidt, E. Vogler, J. Schulz, and U. Ernemann Incidence of New Brain Lesions After Carotid Stenting With and Without Cerebral Protection Stroke, September 1, 2006; 37(9): 2312 - 2316. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. A. Barber and W. Powers MR DWI does not substitute for stroke severity scores in predicting stroke outcome Neurology, April 25, 2006; 66(8): 1138 - 1139. [Full Text] [PDF]
|
|
|
|
|
|
P. J. Hand, J. M. Wardlaw, C. S. Rivers, P. A. Armitage, M. E. Bastin, R. I. Lindley, and M. S. Dennis MR diffusion-weighted imaging and outcome prediction after ischemic stroke Neurology, April 25, 2006; 66(8): 1159 - 1163. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Pierallini, F. Caramia, C. Falcone, E. Tinelli, A. Paonessa, A. B. Ciddio, M. Fiorelli, F. Bianco, S. Natalizi, L. Ferrante, et al. Pituitary Macroadenomas: Preoperative Evaluation of Consistency with Diffusion-weighted MR Imaging--Initial Experience Radiology, April 1, 2006; 239(1): 223 - 231. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. K. Schiemanck, G. Kwakkel, M. W.M. Post, L. J. Kappelle, and A. J.H. Prevo Predicting Long-Term Independency in Activities of Daily Living After Middle Cerebral Artery Stroke: Does Information From MRI Have Added Predictive Value Compared With Clinical Information? Stroke, April 1, 2006; 37(4): 1050 - 1054. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P L Tan, D King, C J Durkin, T M Meagher, and D Briley Diffusion weighted magnetic resonance imaging for acute stroke: practical and popular. Postgrad. Med. J., April 1, 2006; 82(966): 289 - 292. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. E. Chen, J. E. Simon, M. D. Hill, C.-H. Sohn, P. Dickhoff |