871
Measurements of Acute Cerebral Infarction:
Lesion Size by Computed Tomography
Thomas Brott, MD, John R. Marler, MD, Charles P. Olinger, MD,
Harold P. Adams Jr., MD, Thomas Tomsick, MD, William G. Barsan, MD,
Jose Biller, MD, Robert Eberle, Vicki Hertzberg, PhD, and Michael Walker, MD
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
As part of a prospective therapy study of 65 patients with acute, nonhemorrhagic, cerebral
infarction, computed tomographic scans of the head were obtained at admission, 7-10 days, and
3 months. The scans were analyzed for the presence, site, size, and volume measurement of the
infarction. At 7-10 days, the mean infarction volume as measured by computed tomography
was 55 cm3 or about 4 x 4 x 3 . 5 cm (range=0-507 cm3). At 3 months, the mean infarction volume
decreased by 25% to 41 cm3. For the 26 scans showing infarction at the time of admission, the
mean lesion volume was 33 cm3 at admission, 51 cm3 at 7-10 days, and 49 cm3 at 3 months. With
lesion size at 7-10 days expressed as percentage of total brain volume, the mean infarction size
was only 5%. Of the 49 patients with lesions revealed by computed tomography at 7-10 days,
20 had an infarction of 1% or less of total brain volume, while only six had an infarction of 20%
or more of total brain volume. The lesion volumes as measured by the 7-10-day computed
tomography correlated with the neurologic examination scores on admission (Spearman's
rank-order correlation=0.78) and with the scores at 1 week (Spearman's rank-order
correlation=0.79). (Stroke 1989;20:871-875)
C
erebral infarction size as reflected by computed tomography (CT) of the brain provides important clinical information.
Patients with large lesions detected after stroke are
at greater risk of developing life-threatening cerebral edema.1 They are also at greater risk for
subsequent hemorrhagic conversion, with2 or without heparin.3 Lesion size measured by CT is used
by clinicians to assist in prognosis,4 even though
evidence linking lesion size to prognosis is limited.1-5
Lesion volume has not yet been correlated to the
clinical neurologic examination. We describe below a
prospective analysis of CT scanning in the setting of
acute cerebral infarction, carried out as part of a study
of naloxone as therapy for acute ischemic stroke.6
From the Departments of Neurology (T.B., C.P.O., R.E.),
Radiology (T.T.), Emergency Medicine (W.G.B.), and Environmental Health (V.H.), University of Cincinnati, Cincinnati,
Ohio; the Department of Neurology (H.P.A., J.B.), University
of Iowa, Iowa City, Iowa; and the Division of Stroke and
Trauma, National Institute of Neurological Disorders and Stroke
(J.R.M., M.W.), Bethesda, Maryland.
Supported by United States Public Health Service, National
Institutes of Health, and National Institute of Neurological
Disorders and Stroke Contracts N01-NS-2324, TO 1 (University
of Cincinnati) and NO1-NS-2326, TO 1 (University of Iowa).
Address for reprints: Thomas Brott, MD, Department of
Neurology, University of Cincinnati Medical Center, 231 Bethesda
Avenue, Cincinnati, OH 45267-0525.
Received November 7, 1988; accepted January 9, 1989.
Subjects and Methods
Sixty-five patients with the clinical diagnosis of
acute ischemic cerebral infarction were examined
by CT scan without contrast at stroke onset, and
without and with contrast at 7-10 days and at 3
months. The scans were performed on either a
General Electric 9800 scanner or Picker scanner
(600 or 1200) with a 512x512 matrix using 8-10-mm
slices. The lesions' anatomic location and vascular
distribution were determined using the templates of
Damasio.7 All patients had initial CT scans within
48 hours of stroke onset. The mean interval from
time of latest progression of neurologic deficit to
time of naloxone administration was 13 hours, 26
minutes (SD=10 hours, 11 minutes).
Lesion size was measured directly: the area of
abnormal low attenuation was traced on each CT
slice, and the area was summed for the slices
showing the infarct (Figure 1). The volume was
derived from the area and the slice thickness. For
each CT slice, the total brain parenchyma area was
also measured, and then the total brain parenchyma
volume was calculated. Lesion size was also measured qualitatively as normal, lacunar, less than
one-half lobe, up to one lobe, and several lobes
(e.g., a lesion less than one-half lobe in size but
involving two different anatomic lobes was scored
as less than one-half lobe). The measurement sys-
872
Stroke Vol 20, No 7, July 1989
FIGURE 1. Computed tomogram illustrating method
of cerebral infarction volume measurement (see "Subjects and Methods").
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
tem was adapted from the methods of the FullPhase Stroke Data Bank.8 Interpretations and measurements of the CT scans were performed by two
coinvestigators without knowledge of the neurologic assessments.
The mean age of the patients was 64 years; 58%
were white and 42% were black; 52% were men and
48% were women. Of the 65 cerebral infarctions,
57% involved the left hemisphere, 35% involved the
right hemisphere, and 8% 'involved the brainstemcerebellum. On admission, the infarction was thought
to be large-artery atherothrombotic in 52%, smallartery atherothrombotic in 18%, cardiac embolic in
17%, and large-artery atheroembolic in 12%.
Each of the 65 patients was examined neurologically at admission and 24 hours after study entry.
Sixty-three patients survived for examination at 7
days, and 58 patients survived for examination at 3
months. The examinations were performed in a
uniform manner following a format (stroke scale)
shown to be reliable and valid clinimetrically.9
Patient outcome at 3 months was assessed by
determining 1) the patient's performance class, 2)
the patient's placement class, and 3) the patient's
location in the community; these methods were
TABLE 1.
modified from the suggestions of Spence and
Donner.10
Because neither CT measurements nor neurologic examination scores were anticipated to be
normally distributed, Spearman's rank-order method
was used for calculation of correlation coefficients.11
Results
The CT scan was positive for a new cerebral
infarction in 26 of 65 patients (40%) at admission
and positive in 48 of 62 (77%) at 7-10 days, but
sensitivity did not change significantly at 3 months
(78%). The most frequent lesion category was "less
than one-half lobe," accounting for 37% of the
patients scanned at 7-10 days and accounting for
48% of the positive scans at 7-10 days (Table 1).
Lesions involving several lobes were more frequent
than lacunar-sized lesions (19% vs. 10%).
The mean lesion volume at 7-10 days was 55 cm3
(Table 2). Nineteen patients (29%) had a lesion
volume of 10 cm3 or less (Figure 2). With lesion size
expressed as percentage of brain volume, the largest infarction was 47%, but the mean infarction size
was only 5% of brain volume, and one third of
lesions were 1% or less of total brain volume.
Computed Tomography and Cerebral Infarction: Sensitivity and Lesion Size by Qualitative Categonzation
Time of CT
<48 hours from stroke onset
7-10 days from stroke onset
3 months from stroke onset
CT, computed tomography.
Positive CT
26 of 65
48 of 62
45 of 58
Lacunar size
6
8
7
< 1/2 lobe
Up to one lobe
Several lobes
18
23
22
1
6
3
1
11
13
Brott et al Measuring Cerebral Infarction
TABLE 2.
873
Computed Tomography and Cerebral Infarction: Sensitivity and Infarction Size by Volume
Time of CT
<48 hours from stroke onset
7-10 days from stroke onset
3 months from stroke onset
CTs showing
infarction (%)
26 of 65 (40%)
48 of 62 (77%)
45 of 58 (78%)
Mean volume of
infarction by CT (cm3)
14 (range 0-238)
55 (range 0-328)
41 (range 0-291)
Mean size of infarction
(% of total brain volume)
1.0%
4.4%
3.3%
n=65.
CT, computed tomography.
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
After excluding patients with normal CTs at baseline, the mean increase in lesion volume during the
1st week was 18 cm3 (54%) (Table 3). Lesion
volume decreased by 25% between the 1st week
and 3 months when considering all patients, but it
did not decrease significantly for patients with an
initially positive CT (Table 3).
The infarction was shown by CT at 7-10 days to
involve the middle cerebral artery distribution in 40
of the 49 positive scans (82%) (Table 4). The CT
scans did not demonstrate a lesion in four of the
seven patients diagnosed clinically as having brainstem infarctions. At 7-10 days, contrast enhancement was identified in 22 of the 49 positive scans
(45%) (Table 5). Mass effect was seen in 37% of all
scans at 7-10 days and in 47% of the positive CTs
(Table 5).
The Spearman's rank-order correlation between
the stroke scale score at 7 days and the CT scan
lesion volume at 7-10 days was 0.74 (Table 6). After
converting the lesion volume to a ratio of total brain
volume, the lesion size-scale score correlation was
0.76 (p<0.0001). The admission neurologic deficit
as measured by the stroke scale also correlated
strongly with the 7-10 day CT lesion volume (r=0.74,
301-400
201-300
101-200
UJ
2. Bar graph of cerebral infarction volumes at
7-10 days (as measured by computed tomography), categorized by size. Mean lesion volume was 55 cm3, and
78% of scans showed infarction.
FIGURE
/?<0.0001). Correlations between infarction volume
and neurologic deficit were the same regardless of
the cerebral hemisphere involved (r=0.72 for left
hemisphere, r=0.74 for right hemisphere).
Discussion
The quantitative assessment of CT lesion volume
for acute stroke patients provides an objective
measure of ischemic cerebral infarction. At 7-10
days the mean lesion volume was 55 cm3 or about
4x4x3.5 cm . Fourteen patients had no detectable
lesion at 7-10 days (22%), and 19 patients had a
volume of 10 cm3 or less (29%). Mean infarction size
was only 4.4% of total brain volume (Table 2), and
20 of the 65 patients had a lesion size of 1% or less
of total brain volume.
For future stroke therapy trials, comparison of
the infarction volumes for the treated patients with
those of the control patients could provide objective
means of assessment. If a treatment is found effective, the CT measurements in the study could also
allow comparisons among treated patients to determine any selectivity for that treatment. For example, analysis may reveal the therapy to be primarily
effective for patients with lesions below a particular
volume.
The validity of the volumetric method is suggested
by the relation of the infarction volumes to the other
methods of patient assessment. The neurologic deficit at 7 days correlated highly with the volume of
infarction measured by CT at 7-10 days (Spearman's
rank-order correlation coefficient=0.74). Patient outcome at 3 months also correlated with the CT lesion
volume at 7-10 days (/-=0.54). Of further interest, the
neurologic deficit present on admission also correlated strongly to the 7-10 day CT lesion volume
(r=0.74). This implies that for many stroke patients
the first neurologic examination could be a predictor
of subsequent tissue loss as later shown by CT
measurement. The clinical deficit-infarction volume
correlations also provide incentive for further analysis and improvement of neurologic examination
scales. For example, examination items that have no
correlation to CT lesion size could be deleted.
We recognize that infarction location may be
more important for some patients than infarction
volume in determining the eventual clinical handicap. Certain brain regions have classically been
thought more important or "eloquent" than others.12
We might agree, for example, that a given volume
874
Stroke Vol 20, No 7, July 1989
TABLE 3. Lesion Volume Serially in Patients With Positive Baseline Computed Tomography
Time of CT
<48 hours from stroke onset
7-10 days from stroke onset
3 months from stroke onset
Volume
range
0.3-238cm3
0-327cm3
0-291cm3
Mean
volume
33 cm3 (SD 60.4)
51 cm3 (SD 76.1)
49 cm3 (SD 67.0)
Mean size of infarction
(% of brain volume)
2.7
4.0
3.5
n=26.
CT, computed tomography.
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
of cerebellum is less important than the same volume of left perisylvian cortex. However, comparisons today among most brain regions would be
difficult if not arbitrary. We would rather proceed
from the assumption that all brain regions are nearly
equal in importance regardless of location but that
some regions are more difficult to test than others.
Volume measurements do not require value judgments. More important, perhaps, effective therapies for acute cerebral infarction are not likely to
change lesion location. Effective therapies will be
effective by their ability to decrease infarction size.
The results from the categoric method of infarction size measurement did not accurately describe
our population. The infarction volume distribution
showed a preponderance of very small lesions with
a gradual transition to larger ones. The breakdown
of lesion size by categories resulted in few very
small lesions, with a preponderance of lesions larger
than lacunar size but smaller than one-half lobe in
size (Table 1). We suspect a reluctance to label an
infarction "lacunar"; if the lacunar category had
been designated as "small, deep infarct," then
smaller infarcts may have been more appropriately
categorized.
In the setting of acute stroke, analysis of lesion
size as measured by CT has been limited. Correlation of aphasia severity and lesion volume13-15 as
well as correlation between nonverbal performance
and lesion volume13 have been demonstrated. Correlation between upper limb weakness and lesion
volume (r=0.59) and between presence of a lesion
in the corticospinal system and lesion volume
(r=0.75) have also been reported.16 Yamaguchi et al
estimated infarction size by comparing the area of
CT hypodensity with the area measured by CT of
the affected cerebral hemisphere, deriving what
they called an infarct index.17 Infarctions thought to
be embolic were more than twice as large as infarcTABLE 4.' Cerebral Infarction Arterial Distribution Measured by
Computed Tomography at 7. Days
Arterial distribution
Anterior cerebral artery
Middle cerebral artery
Posterior cerebral artery
Internal carotid artery
Vertebrobasilar
Total positive
Positive scans (%)
i
2
82
2
8
6
CT, computed tomography.
40
1
4
3
tions thought to be thrombotic («=209). The infarct
index was not compared to a detailed neurologic
examination.
As has been reported in previous studies,5-18 we
found CT sensitivity at 7-10 days to be considerably
higher for infarction (77%) than CT sensitivity at
admission (40%). Not previously reported, this CT
sensitivity did not decline at 3 months (78%) as
might have been expected (Table 2). The increase in
infarction size (mean 54%) over the 1st week in
those patients with initially positive scans was not a
surprise, but the modest change in size thereafter
(25% decrease at 3 months for all patients) was
encouraging. We conclude that the timing for CTs
performed in future therapy studies to detect and
measure cerebral infarction may involve a generous
window from 1 week or earlier to 1-3 months.
Prospective therapy studies may require only two
CT scans, one at admission to rule out hemorrhage
and one at a later date (e.g., 1-2 weeks) for optimal
infarction detection and measurement.
Our study has several limitations. Interrater reliability of the CT scan volume measurements was
not assessed, and the 65 stroke patients may have
had somewhat smaller infarctions and less severe
neurologic deficits than the general acute stroke
population (e.g., our patient mortality at 3 months
was 8%). All patients received a 24-hour infusion of
naloxone, which could have affected lesion imaging
(we were unable to detect any positive or negative
clinical effects that may have had a bearing on
lesion size). Finally, the correlations reported should
be interpreted cautiously, given the wide range of
the CT infarction volumes and the clustering of
patients with normal CT scans.
Measurements of cerebral infarction will become
increasingly important as effective early treatment
options are developed. Clinical assessment techniques will continue to miss silent cerebral infarctions.19"21 Likewise, high-resolution CT scanners
will continue to miss clinically significant cerebral
infarctions.5 With magnetic resonance imaging,
greater sensitivity is possible, but imaging may be
less specific.22 We suggest that clinical and anatomic measures of cerebral infarction are fundamentally complementary and that neither measurement approach can be adequately evaluated
independent of the other.
Acknowledgments
The authors wish to thank Drs. Carlos Kase,
Oscar Reinmuth, and Percy Karanjia for their assis-
Brott et al Measuring Cerebral Infarction
875
TABLE 5. Edema, Mass Effect, and Contrast Enhancement Measured by Computed Tomography
Time of CT
<48 hours from stroke onset
7-10 days from stroke onset
3 months from stroke onset
Edema
No. (%)
Mass effect
Contrast enhancement
No. with infarct
No. (%)
26
49
45
16 (62)
29 (59)
1 (2)
10 (38)
23(47)
0
No. (%)
CTs noncontrast
22 (45)
2 (4)
n=65.
CT, computed tomography.
TABLE 6. Spearman's Rank-Order Correlation of Infarction Volume (cm3) to Neurologic Deficit
Neurologic exam on
admission
Time of CT
<48 hours from stroke onset
7-10 days from stroke onset
3 months from stroke onset
w
0.35
0.78*
0.76*
Neurologic exam at
1 week
M
0.28*
0.79*
0.70*
Neurologic exam at
3 months
(r)
0.38*
0.68*
0.62*
CT, computed tomography.
*p<0.0001.
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
tance in design of the CT assessment methods and
the stroke scale (neurologic examination and scale).
References
1. Ropper AH: Lateral displacement of the brain and level of
consciousness in patients with an acute hemispheral mass. N
EnglJMed 1986;314:953-958
2. Cerebral Embolism Study Group: Cardioembolic stroke,
early anticoagulation, and brain hemorrhage. Arch Intern
Med 1987;147:636-640
3. Lodder J: CT-detected hemorrhagic infarction; relation with
the size of the infarct and the presence of midline shift. Ada
Neurol Scand 1984;70:329-335
4. Kinkel WR, Jacobs L: Computerized axial transverse tomography in cerebrovascular disease. Neurology 1976;26:924-930
5. Tatemichi TK, Mohr JP, Rubinstein LV, Kase CS, Nichols
FT, Price TR, Wolf PA: CT findings and clinical course in
acute stroke: The NINCDS Pilot Stroke Data Bank (abstract).
Stroke 1985;16:138
6. Adams HP Jr, Olinger CP, Barsan WG, Butler MJ, GraffRadford NR, Brott TG, Biller J, Damasio H, Tomsick T,
Goldberg M, Spilker JA, Berlinger E, Dambrosia J, Biros M,
Holleran R: A dose escalation study of large doses of
naloxone for treatment of patients with acute cerebral ischemia. Stroke 1986;17:404-409
7. Damasio H: Vascular territories denned by computed tomography, in Wood JH (ed): Cerebral Blood Flow. New York,
McGraw-Hill Book Co, 1987
8. Shinar S, Gross CR, Hier DB, Caplan LR, Mohr JP, Price TR,
Wolf P, Kase CS, Fishman IG, Barwick JA, Kunitz SC:
Interobserver reliability in the interpretation of computed tomographic scans of stroke patients. Arch Neurol 1987;44:149-155
9. Brott T, Adams HP, Olinger CP, Marler JR, Barsan WG,
Biller J, Spilker J, Holleran R, Eberle R, Hertzberg V,
Rorick M, Moomaw CJ, Walker M: Developing measurements of acute cerebral infarction: A clinical examination
scale. Stroke 1989;20:864-870
10. Spence JD, Donner A: Problems in design of stroke treatment trials. Stroke 1982;13:94-99
11. Hollander M, Wolfe DA: Nonparametric Statistical Methods. New York, John Wiley & Sons, Inc, 1973, pp 191-192
12. Spetzler RF, Martin NA: A proposed grading system for
arteriovenous malformations. JNeurosurg 1986;65:476-483
13. Kertesz A, Harlock W, Coates R: Computer tomographic
localization, lesion size, and prognosis in aphasia and nonverbal impairment. Brain Lang 1979;8:34-50
14. Naeser MA, Hayward RW, Laughlin SA, Zatz LM: Quantitative CT scan studies in aphasia. I. Infarct size and CT
numbers. Brain Lang 1981;12:140-164
15. Knopman DS, Seines OA, Niccum N, Rubens AB, Yock D,
Larson D: A longitudinal study of speech fluency in aphasia:
CT correlates of recovery and persistent nonfluency. Neurology 1983;33:1170-1178
16. Knopman DS, Rubens AB: The validity of computed tomographic scan findings for the localization of cerebral functions. Arch Neurol 1986;43:328-332
17. Yamaguchi T, Minematsu K, Choki J, Ikeda M: Clinical and
neuroradiological analysis of thrombotic and embolic cerebral infarction. Jpn Circ J 1984;48:50-58
18. Inoue Y, Kazamasa T, Miyamoto T, Yoshikawa N, Taniguchi S, Saiwai S, Nishimura Y, Komatsu T: Sequential
computed tomography scans in acute cerebral infarction.
Radiology 1980;135:655-662
19. Turnbull IW, Bannister CM: CT observations on the natural
history of asymptomatic cerebral infarction following transient ischemic attacks. Neurol Res 1985;7:190-193
20. Awad I, Modic M, Little JR, Furlan AJ, Weinstein M: Focal
parenchymal lesions in transient ischemic attacks: Correlation of computed tomography and magnetic resonance imaging. Stroke 1986;17:399-403
21. Berguer R, Sieggreen MY, Lazo A, Hodakowski GT: The
silent brain infarct in carotid surgery. / Vase Surg 1986;
3:442-447
22. DeWitt LD, Kistier JP, Miiier DC, Richardson EP, Buonanno FS: NMR-neuropathologic correlation in stroke. Stroke
1987;18:342-351
KEY WORDS • cerebral infarction • neurologic examination •
tomography, emission computed
Measurements of acute cerebral infarction: lesion size by computed tomography.
T Brott, J R Marler, C P Olinger, H P Adams, Jr, T Tomsick, W G Barsan, J Biller, R Eberle, V
Hertzberg and M Walker
Stroke. 1989;20:871-875
doi: 10.1161/01.STR.20.7.871
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1989 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/20/7/871
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click Request
Permissions in the middle column of the Web page under Services. Further information about this process is
available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Stroke is online at:
http://stroke.ahajournals.org//subscriptions/