(Stroke. 2001;32:1588.)
© 2001 American Heart Association, Inc.
Original Contributions |
Prospective Value of Perfusion and X-Ray Attenuation Imaging With Single-Photon Emission and Transmission Computed Tomography in Acute Cerebral Ischemia
Presented in part at the 45th meeting of the Society of Nuclear Medicine, Toronto, Canada, June 1998, and at the 7th World Congress of Nuclear Medicine and Biology, Berlin, Germany, September 1998.
From the Departments of Nuclear Medicine (H.B., S.H.), Radiology (C.D., J.D.), and Neurology (A.R., D.S., J.B.), University of Leipzig, Leipzig, Germany; and Department of Nuclear Medicine (W.H.K.), Hanover Medical School, Hanover, Germany.
Correspondence to Dr Henryk Barthel, PET Oncology Group, MRC Cyclotron Unit, Imperial College School of Medicine, Hammersmith Hospital, Du Cane Road, London W12 0NN, UK. E-mail henryk.barthel{at}ic.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Background and Purpose—The aim of the present study was to test the hypothesis that perfusion single-photon emission computed tomography (SPECT), carried out in addition to transmission computed tomography (TCT), improves the predictive value of brain imaging within the therapeutically relevant time window after acute cerebral ischemia.
Methods—Using TCT and [99mTc]ethyl cysteinate dimer (ECD)-SPECT within 6 hours after symptom onset, we examined 108 patients (44 women, 64 men; mean age 65±13 years) with acute ischemic stroke attributed to the territory of the middle cerebral artery (MCA). In each case, 3 experts prospectively evaluated the early SPECT and TCT images. We correlated these ratings with follow-up TCT findings for the final infarction as well as with clinical outcome (Scandinavian Stroke Scale, Barthel Index, Modified Rankin Scale) after 30 and 90 days.
Results—Severe activity deficits on SPECT, not caused by local atrophy on TCT, were the best predictors (positive predictive value [PPV ]94%, 95% CI 89% to 99%; negative predictive value [NPV] 90%, 95% CI 78% to 100%; P<0.001) for evolving cerebral infarction. Complete MCA infarctions were predicted with significantly higher accuracy with early SPECT (area under receiver operating characteristic curve [AUC] index 0.91) compared with early TCT (AUC index 0.77) and clinical parameters (AUC index 0.73, P<0.05). Logistic regression analysis revealed 1 independent predictor for completed MCA territory infarction: SPECT activity deficits in the corresponding areas (PPV 88%, 95% CI 65% to 100%; NPV 96%, 95% CI 92% to 100%; P<0.001). Furthermore, death after stroke was optimally predicted by [99mTc]ECD-SPECT. Clinical outcome up to 90 days after the stroke event best correlated with the degree of activity deficits in early SPECT (r=0.53, P<0.001).
Conclusions—[99mTc]ECD brain perfusion SPECT that completes TCT definitely improves the predictive value of brain imaging after acute cerebral ischemia. Thus, the combined imaging of brain edema and of cerebral perfusion early after stroke is recommended for clinical use.
Key Words: cerebral ischemia • stroke, acute • tomography, emission computed • tomography, x-ray computed
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
New therapies for acute cerebral infarction have been tested for several years. Among them, intravenous thrombolysis within 3 hours after symptom onset represents the first therapeutical approach that can effectively treat acute ischemic stroke.1 Although adequate treatment should be started as early as possible, most patients still arrive at the hospital too late to receive the maximum benefit from this emerging therapy.2 It is evident, however, that viable brain tissue may persist beyond this time.3 4 Therefore, thrombolysis was also tested up to 6 hours after symptom onset.5 6 7 Despite the recently published PROACT II trial that first demonstrated the efficiency of intra-arterial thrombolysis,7 up until now, thrombolysis has not been approved for intravenous administration >3 hours after stroke onset.
Because these entire therapeutic approaches involve the preservation of jeopardized but viable brain tissue with functional relevance, a diagnostic tool is required that is able to predict the presence of salvageable brain tissue.8 9 Transmission computed tomography (TCT) has been the routine brain-imaging method of choice for cases of acute stroke.10 11 TCT allows the exclusion of hemorrhagic stroke with high accuracy.12 In addition, a number of TCT findings provide evidence for an ischemic course of the stroke event: hyperdense middle cerebral artery (MCA) sign,13 focal hypoattenuation,14 and focal brain swelling.15 TCT performed early after stroke onset is able to evaluate brain tissue viability with high specificity16 : the presence of "early signs" in TCT provides strong evidence for an evolving infarction and reliably excludes reversible ischemia. However, the sensitivity of TCT, especially at very early time points after the stroke, is controversial, as discussed in the literature,16 17 and must be further investigated. There also is evidence from the literature that the predictive value of brain imaging in acute ischemia could be improved by the addition of information regarding blood flow to information regarding early ischemic edema, which is obtained with TCT.9 18
In a previous study, we demonstrated that single-photon emission computed tomography (SPECT) of the brain with [99mTc]ethyl cysteinate dimer ([99mTc]ECD) within 6 hours after the onset of stroke symptoms is able to differentiate between reversible cerebral ischemia and evolving cerebral infarction.19 In the case of evolving infarction within the MCA territory, [99mTc]ECD-SPECT, performed at the acute stage, allows a prediction of complete ("malignant") infarction.20 Even clinical outcome can be reliably predicted with early [99mTc]ECD-SPECT.21
Until now, no studies have compared brain TCT with [99mTc]ECD-SPECT within the therapeutically relevant time window of 6 hours, particularly with respect to the identification of (1) patients with spontaneously good prognosis, who do not need specific treatment, (2) patients with evolving infarctions, who would profit from adequate therapy, and (3) patients with the risk of complete (malignant) MCA territory infarctions,22 in whom clinical outcome probably could be improved by alternative treatment strategies such as decompressive hemicraniectomy or hypothermia.23 24 Thus, in the present study, we prospectively tested the hypothesis that [99mTc]ECD-SPECT, performed in combination with TCT, can improve the predictive value of non–contrast-enhanced TCT alone early after stroke.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Subjects
In this study of combined TCT and SPECT, 108 consecutive patients (44 women, 64 men; mean age 65±13 years) were prospectively enrolled. All patients had acute neurological deficits attributed to ischemia within the MCA territory, no stroke history, Scandinavian Stroke Scale (SSS)25 score of <40 points on admission, and TCT as well as SPECT imaging within 6 hours after the time point at which the acute neurological deficit was first observed. Exclusion criteria were presence of coma, rapid improvement in symptoms during the TCT and SPECT investigations, signs of intracerebral hemorrhage on acute TCT, and symptoms of vertebrobasilar stroke. The study was approved by the Ethics Committee of the University of Leipzig. Of the 108 patients, 39 represent the first group of patients, who were included at the University of Leipzig in an ongoing double-blind, placebo-controlled trial (ECASS II [Second European-Australian Acute Stroke Study]26 ) and were treated in a randomized manner with recombinant tissue plasminogen activator (n=20) or placebo (n=19). The SPECT results for the patients who were included in the ECASS II trial at the University of Leipzig were recently published.27
Protocol and Clinical Investigation
Immediately after admission to the hospital, all
patients were clinically examined, including blood pressure
measurement, evaluation of consciousness, eye inspection with respect
to putative gaze palsy or conjugate eye deviation, cardiac examination,
and SSS (46 points maximum, without "gait"). Clinical examination
was followed by TCT and SPECT imaging. To save time, in 43 patients the
radiopharmaceutical agent for SPECT imaging was administered directly
before TCT. TCT was repeated 7 days after the stroke to evaluate extent
and localization of definitive infarctions. SSS was assessed again 30
and 90 days after stroke. Additional functional and disability outcomes
were scored after 90 days using the Barthel Activity of Daily Living
Index (0 to 100 points)28
and the Modified Rankin Disability Scale (0 to 6
points).29
Brain Imaging
For TCT, all patients underwent
non–contrast-enhanced cranial TCT with a
Somatom Plus S scanner (Siemens Inc). Axial
slices of the whole brain were obtained parallel to the orbitomeatal
line. The slice thickness was 5 mm (lower part of brain up to
sellar region) and 10 mm (upper part of brain), respectively.
Window width equaled 128 Hounsfield units, with the center level at 36
Hounsfield units.
For SPECT, all patients were injected intravenously with 400 to 600 MBq [99mTc]ECD.30 SPECT acquisition was begun 15 minutes after the injection. Photons were registered with a brain-dedicated SPECT camera (Ceraspect; DSI) with 3 rotating parallel-hole collimators.31 The SPECT data were acquired, reconstructed, attenuation-corrected, and reoriented according to a standard protocol that has been described elsewhere.32
Image Analysis
Interpretation of the acute TCT and SPECT images was
performed visually.
Separate Analysis of Early TCT
Images
The TCT images were independently evaluated by 3
experienced interpreters. The extent of putative hypodensity and brain
swelling in MCA territory (<33%, 33% to 66%, or >66% of MCA
territory) was rated, as well as the additional involvement of anterior
or posterior cerebral artery territories. On the basis of the TCT
findings, the outcome was rated as "transient ischemia,"
"nontotal MCA infarction," or "total MCA infarction." The
experts were aware of the affected side but blinded to the severity of
symptoms.
Separate Analysis of Early SPECT
Images
In the same manner as for the TCT images, the SPECT
images (transverse, coronal, and sagittal slices) were independently
evaluated by 3 experienced interpreters. These interpreters were also
aware of the side of the symptoms but were blinded to the results of
the TCT scoring and symptom severity. Extent of local activity deficits
within the MCA territory (<33%, 33 to 66%, to >66% subtotal and
total MCA territory) was scored, as well as severity (0=no, 1=mild,
2=moderate, 3=severe decrease in activity). Outcome was predicted as
described for TCT images.
Combined Analysis of Early SPECT and
TCT Images
The SPECT images were reinterpreted as described
together with the corresponding individual TCT
images.
Categorization
The follow-up TCT images were analyzed by 1
of the experts, who was blinded to the ratings of the initial TCT and
SPECT images, concerning the occurrence of hypodensities. According to
these findings, the patients were separated into 3 groups: group A
(n=24) had no hypodensities, group B (n=73) had hypodensities in
subtotal MCA territory, and group C (n=11) had hypodensities in total
MCA territory. This grouping was performed with consideration for
possible subsequent pooling of the study population according to the
different hypotheses to be tested: prediction of evolving subtotal
(groups B plus C versus group A) or total (group C versus groups A plus
B) MCA territory infarctions.
Statistical Analysis
Differences between the patient groups in the number
of SPECT and TCT findings were tested for significance using the
Student’s t test for
independent samples after verification of normal distribution with the
statistical software package SPSS. Normal values
were given with 1 SD. Significance levels for differences were set at
P<0.05,
P<0.01, and
P<0.001. Interobserver
variability of the visual SPECT and TCT analyses was assessed
by calculating the 
statistics for the ratings of each of the 3
independent experts. Receiver operating characteristic (ROC)
parameters for results of SPECT and TCT image
analyses, as well as clinical findings, were calculated with
the software CLABROC for continuously distributed variables and
CORROC2 for categorical rating-scale variables. ROC curves were
fitted by application of the software ROCFIT. Accuracy, positive
predictive value (PPV), negative predictive value (NPV), and relative
risk were calculated after cutoff values were defined with the help of
the corresponding ROC data and that of cross-tables, respectively.
Proportions were presented with 95% CI values. After these
univariate analyses, multivariate
analyses were performed to identify independent
parameters for the prediction of infarction, total MCA
territory infarction, or death after stroke. For that, logistic
regression analysis was carried out using stepwise forward
selection of variables with Wald’s testing. Correlations between
clinical, TCT, and SPECT data on admission and clinical follow-up
parameters were characterized by calculation of Kendall’s
rank correlation coefficients and (linear) regression
analyses.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Patient Characteristics
Group characteristics concerning individual and clinical data for the patients are listed in Table 1
. No significant differences were found between
groups A (no hypodensities on final TCT), B (subtotal MCA territory
hypodensities on final TCT), and C (total MCA territory hypodensities
on final TCT) with respect to age and sex distribution or regarding
systolic and diastolic blood pressure and
occurrence of cardiac arrhythmia. SSS score on admission was
significantly higher in group A patients than in group B and C patients
(mean±SD 36±3 versus 29±8 and 25±7, respectively), whereas the
difference between groups B and C was insignificant. Impairment of
consciousness, evaluated on admission, was observed significantly more
often in group C than in group B. Gaze palsy/conjugate eye deviation
occurred in only 4% of patients in group A in comparison to 25% of
patients in group B
(Table 1).
|
Analysis of Early TCT and Perfusion
SPECT Images
In early TCT, hyperdense MCA sign was observed
significantly more often in group B than in group A (10% versus 0%,
P<0.05), as well as focal
hypodensities (64% versus 8%,
P<0.001) and brain swelling
(51% versus 4%, P<0.001). In
a comparison of the findings between groups B and C, a highly
significant group difference was found regarding the occurrence of
focal hypodensities/brain swelling in >66% of the MCA territory (3%
versus 36%, P<0.001).
Interobserver variability concerning the early TCT analysis,
defined as mean variation among the ratings of the 3 experts, was
17.1% (=0.33, P<0.001),
and that of early SPECT analysis was 4.4% (=0.75,
P<0.001). SPECT imaging was
carried out 1.3 to 6.0 hours (mean 4.0±1.6 hours) after the onset of
stroke symptoms, without significant differences within that period,
defined as duration between symptom onset and injection of the
radiopharmaceutical agent, among the 3 patient groups
(Table 1
). With respect to the early SPECT findings, highly
significant differences were found between all patient groups: Activity
deficits with an extent of >33% of the MCA territory were more often
observed in group B than in group A (76% versus 21%,
P<0.001), and those in the
total MCA territory were more often observed in group C than in group B
(64% versus 1%, P<0.001).
Highly significant differences between groups A and B were
further found concerning the frequency of activity deficits with a
degree of >1 in
the visual scoring (17% versus 97%,
P<0.001).
Figures 1 and 2 give examples of early imaging results
of 2 patients from group C: although in the case of the patient in
Figure 1, early TCT revealed hypodensities and brain
swelling in the total MCA territory, in the case of the patient in
Figure 2, hypodensities were rated only in the left
putaminal area. Early SPECT imaging, however, showed activity deficits
in complete MCA territories in both patients.
|
|
Prediction of Evolving Infarction
Patients with final infarctions (groups B and C)
were pooled and compared with the patients without final
infarction (group A). Predictive values, calculated relative risk in
case of the presence of clinical signs on admission, and TCT and SPECT
findings for the prediction of evolving cerebral infarction are listed
in
Table 2. The highest PPVs and NPVs were found for a
parameter of combined interpretation of early TCT and SPECT
images: activity deficits in SPECT with a severity of >1 in visual
scoring, which were not caused by local atrophy (PPV 94%, NPV 90%).
An example of false-positive scoring in the separate interpretation of
the early SPECT images due to local atrophy is illustrated in
Figure 3. Multivariate logistic
regression analysis revealed both SSS scoring on admission and
severity of activity deficits in SPECT to be independent predictors of
evolving infarction. Combination of information from these 2
independent predictors resulted in an accuracy of 91% for the
prediction of evolving infarction
(Table 3). The application of the optimal clinical,
TCT, and SPECT predictors and the combination of their predictive
impact according to the course of events within the clinical setting
(clinical examination, TCT, brain perfusion SPECT) resulted in stepwise
improvement in accuracy from 70% to 83% and 93%. This improvement in
accuracy was attributed to an improvement in NPV
(Figure 4).
|
|
|
|
Prediction of Total MCA Territory
Infarction
The group of patients with a final total MCA territory
infarction (group C) was compared with a pool of patients
without a final total MCA territory infarction (groups A and B). The
highest impact on the prediction of total MCA territory infarction
among all clinical and early TCT and SPECT parameters was
calculated for the extent of activity deficits in SPECT. Deficits with
an extent of 100% of the MCA territory were found in 7 of 11 group C
patients and in only 1 of 97 group A and B patients.
Multivariate analysis revealed the extent of
SPECT defects to be the only independent predictor of total MCA
territory infarction
(Tables 3 and 4). ROC analysis resulted in a
significant larger area under the ROC curve (AUC) concerning the extent
of the activity deficit scored in the early SPECT images (AUC index
0.91) compared with that concerning SSS scoring on admission (AUC index
0.73) and the extent of hypodensities/brain swelling rated on early TCT
images (AUC index 0.77,
P<0.05)
(Figure 5). The combination of optimal clinical and
early TCT predictors resulted in an improvement in accuracy from 58%
to 92%, with further improvement after the addition of SPECT
information (94%). This stepwise improvement was the result of
substantially higher PPV by adding the parameters of
brain imaging to the clinical score
(Figure 6).
|
|
|
Prediction of Clinical Outcome
Death after stroke occurred in 12 patients (11%).
Among all clinical, TCT, and SPECT parameters, it was
optimally predicted with the information on the extent of activity
deficits in SPECT (PPV 88%, NPV 95%,
P<0.001). Activity deficits on
SPECT in the total MCA territory, as well as coinvolvement of the
territories of the anterior and posterior cerebral arteries to the MCA
territory, were revealed to be independent predictors of death after
stroke (P<0.001 and <0.012
after multivariate logistic regression
analysis). The combination of these 2 independent predictors
resulted in an accuracy of 93%
(Table 3). Correlations between parameters of
clinical outcome (SSS after 30 and 90 days, Barthel Index after 90
days, and Modified Rankin Scale after 90 days) and clinical
parameters on admission as well as early TCT and SPECT
parameters are shown in
Table 5: whereas in the case of SSS scoring after both
30 and 90 days, the highest coefficients were calculated for
correlation with SSS scoring on admission
(r=0.50 and 0.52), with respect
to the Barthel Index after 90 days and the Modified Rankin Scale after
90 days, the highest coefficients were found for the correlation with
the degree of activity deficits on early SPECT
(r=0.46 and
0.53).
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The present study was initiated to test whether brain SPECT with [99mTc]ECD, carried out in combination with TCT, can improve the prospective value of brain imaging within 6 hours after the onset of stroke symptoms. Both imaging modalities were first applied in a time window in which decisions regarding treatment can be derived from the obtained diagnosis. Previous studies that compared brain perfusion SPECT and TCT in early stages (within 48 hours) of stroke were carried out, with the only exception being a study by Giubilei et al33 of 32 patients, >6 hours after stroke symptom onset.34 35 36 37 38 39 Because until now optimal treatment strategy beyond the 6-hour time window has not been found, diagnosis at this time after stroke in most cases lacks therapeutical relevance. Therefore, in the present study, we included only patients (n=108) who underwent both TCT and [99mTc]ECD-SPECT imaging within 6 hours after the onset of symptoms.
Furthermore, we compared TCT with SPECT images obtained using the radiopharmaceutical agent [99mTc]ECD in the acute stage after stroke. The cited studies were exclusively carried out using the local cerebral blood flow (lCBF) tracer [99mTc]hexylmethylpropylene amineoxine (HMPAO) or 133Xe,35 respectively. SPECT studies with 133Xe have lower spatial resolution than 99mTc studies (6 to 7 mm with high-resolution SPECT systems31 ) due to physical characteristics of the isotope. Two brain perfusion radiopharmaceutical agents labeled with 99mTc are available: in comparison to the originally introduced [99mTc]HMPAO,40 the second-generation lCBF tracer [99mTc]ECD41 offers a number of advantages. First, there is a stronger correlation between the uptake of the radiopharmaceutical agent and cerebral blood flow in lower flow rates, which results in higher sensitivity for the detection of smaller infarctions.42 Second, in contrast to the brain uptake of [99mTc]HMPAO, that of [99mTc]ECD is codetermined by metabolic integrity and viability of the brain tissue.43 44 Third, as reported by our group, image quality and radiochemical stability are higher with the use of [99mTc]ECD, even compared with the recently introduced stabilized [99mTc]HMPAO compounds.45
By using [99mTc]ECD for high-resolution SPECT and comparing its prognostic impact with that of parameters of clinical status on admission and early TCT, an improvement in the prediction of evolving infarction was achieved. This is particularly evident in borderline situations concerning therapeutical decisions. The combination of admission symptomatology and information obtained with TCT and SPECT allows the prediction of evolving cerebral infarctions with an accuracy of 93% (95% CI 88% to 98%). Symptom severity and severity of perfusion deficits in SPECT were revealed to be the only independent predictors of evolving infarction. However, in cases of an infarction, involvement of the total MCA territory is predicted with a similar accuracy of 94% (95% CI 89% to 98%). Multivariate regression analysis showed potential to independently predict total MCA territory infarction only for SPECT perfusion defects in total MCA territory, not for any early signs of infarction on TCT or clinical data. This advantage of early [99mTc]ECD-SPECT was clearly confirmed by additional ROC analysis.
Despite the limited comparability to the above studies, our results of higher predictive values of early SPECT compared with those of early TCT for the prediction of infarction correspond with the results of Giubilei et al,33 who published the only available study that systematically compared brain perfusion SPECT and TCT within 6 hours after acute stroke. They used [99mTc]HMPAO as the lCBF tracer in 32 patients and found that 7 of 8 patients (88%) with large infarctions on follow-up TCT, but without early signs of infarction on acute TCT, had severe perfusion defects in SPECT (asymmetry >40% compared with contralateral hemisphere). Furthermore, 10 of 11 patients (91%) with smaller final infarctions had accordingly milder perfusion disturbances.33 These results are paralleled by studies that compare brain perfusion SPECT and TCT for the diagnosis of infarction within time windows of 36 to 48 hours.34 37 39 However, we were able to reproduce these findings for the newer radiopharmaceutical agent [99mTc]ECD and, of more importance, for the therapeutically relevant time window of 6 hours.
In a comparison of the 2 imaging modalities during the acute stage of cerebral ischemia, it should not be disregarded that the information obtained with [99mTc]ECD-SPECT substantially differs from that obtained with TCT: Although in the case of cerebral infarction [99mTc]ECD-SPECT directly detects local perfusion abnormalities, TCT reveals changes in brain tissue density due to ischemic edema. However, this edema develops as a consequence of the initial ischemia, which means there is a varying time delay, and it occurs only when brain perfusion decreases under the critical threshold of 15 mL · 100 g-1 · min-1 (Pulsinelli46 ) (normal 50 mL · 100 g-1 · min-1). However, a comparison of the predictive values of brain perfusion SPECT and TCT is appropriate, because both procedures are applied within an identical clinical context after acute ischemic stroke.
Our results of a higher predictive value with brain perfusion SPECT compared with TCT are not hampered by a lack of quality in early TCT evaluation: Fiorelli et al47 evaluated the occurrence of early signs of cerebral infarction in TCT (within 5 hours after symptom onset) in a study of 158 patients and reported a PPV of 97% (95% CI 93% to 100%; present results, 95% to 100%) and an NPV of 40% (95% CI 29% to 51%; present results, 25% to 49%) for focal parenchymal changes. In addition, von Kummer et al14 reported a PPV of 70% to 85% and an NPV of 83% to 88% of hypodensities and focal brain swelling covering >50% of the MCA territory for the prediction of fatal clinical outcome (Oxford Handicap Scale >3). In the present study, focal brain swelling/hypodensities in >66% of MCA territory predicted total MCA territory infarction with a PPV of 67% (95% CI 29% to 100%) and an NPV of 93% (95% CI 88% to 98%). However, the PPV and NPV of our early TCT evaluation are in accordance with those of Fiorelli et al47 and von Kummer et al.14
An interesting finding resulted from the analyses of predictive values of early TCT and [99mTc]ECD-SPECT for the prediction of evolving cerebral infarctions. Optimal PPV and NPV were calculated for perfusion defects that were not caused by local atrophy, whereby this information was obtained from the combined TCT and SPECT image interpretation. In the future, such combined measurements of different pathophysiological parameters could play an important role, particularly in the determination of ischemic penumbra,48 which represents the target of most stroke therapies49 and could be diagnosed by combining evaluations of brain tissue viability and blood flow/oxygen concentration.18 9 50
Weir et al51 investigated a subgroup of 28 stroke patients within 16 hours of symptom onset by using [99mTc]HMPAO-SPECT and concluded that the accuracy of clinical follow-up prediction decreases the longer the SPECT imaging is delayed, so we focused on the prediction of clinical outcome of the stroke patients within the early time window of 6 hours. In the comparison of early brain perfusion SPECT with TCT and clinical data on admission, only 2 SPECT parameters regarding the extent of lCBF defects, and no early TCT parameters or clinical data on admission, were found to be independent predictors of death after the stroke (overall accuracy 93%, 95% CI 88% to 98%). Among all clinical and early TCT and SPECT parameters, the functional and disability outcome after 3 months was best correlated with the degree of the initial perfusion defect (r=0.06 and 0.53, P<0.001). Our results are paralleled by studies of Lees et al52 and Laloux et al,37 who reported the volume of SPECT, but not of TCT defects or Canadian Neurological Score on admission, to be a significant predictor of clinical outcome. In contrast to our study, these authors exclusively compared SPECT and TCT results, which were acquired later than the therapeutical relevant time window of 6 hours.52 37
In present study study, the interobserver variability of [99mTc]ECD-SPECT interpretation was revealed to be substantially lower than that of TCT interpretation. Only moderate interobserver agreement for TCT analyses was confirmed by von Kummer et al.53 This problem with the quality of TCT analysis resulted, for example, in the inclusion of 52 patients in the ECASS I trial (8.4% of the ECASS I population) despite extended hypoattenuation in early TCT, which was detected by additional retrospective evaluation.54 It can be speculated that in the future, not only diagnosis of acute cerebral ischemia but also quality of studies for the testing of new therapeutical strategies of acute stroke can be improved by including such additional diagnostic information, which can be obtained with lower interobserver variability.
A possible limitation of the present study must be
considered. Discrimination of the patient groups was performed
depending on the results of a follow-up TCT examination after 7 days.
Because Baron and Marchal55
noted that TCT within this period after stroke onset is not optimum for
assessment of the extent of the final infarction due to vasogenic edema
and mass effects, it cannot be guaranteed that all patients were
correctly grouped. In addition, it cannot be completely excluded that
the "fogging" effect, in which initially hypodense ischemic
areas become isodense compared with normal brain
tissue,56 could have
compromised the quality of the follow-up TCT evaluation. However,
literature shows that many stroke investigators schedule the follow-up
TCT to evaluate the final infarction within the time range of 
5 to 7
days,36 39 57
even in large multicenter
trials.6 26
In conclusion, the hypothesis that the prospective value of brain imaging within 6 hours after the onset of cerebral ischemia is improved by the application of [99mTc]ECD brain perfusion SPECT in addition to TCT can be confirmed by our results. In the case of irreversible cerebral ischemia, total MCA infarction is predicted with high reliability. Furthermore, the prediction of clinical outcome after acute stroke is improved. Thus, early brain perfusion SPECT in combination with TCT has the potential to become a valuable diagnostic tool in the acute phase after stroke.
|
Acknowledgments |
|---|
This work was supported by DuPont Pharmaceuticals Co. The authors are also grateful to Jochen Halpick for the acquisition of a number of the acute brain perfusion SPECT studies, to Johannes Köster for his support in preparation of the analyses of the early TCT images, and to Hisham Taha for his careful correction of the manuscript. Furthermore, the important suggestions of Rüdiger von Kummer during the preparation of the manuscript are acknowledged.
Received October 19, 2000; revision received February 2, 2001; accepted March 8, 2001.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
1. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med. 1995;333:1581–1587.
2.
Smith MA, Doliszny
KM, Shahar E, McGovern PG, Arnett DK, Luepker RV. Delayed hospital
arrival for acute stroke: the Minnesota Stroke Survey.
Ann Intern Med. 1998;129:190–196.
3.
Marchal G,
Beaudouin V, Rioux P, de la Sayette V, Le Doze F, Viader F, Derlon JM,
Baron JC. Prolonged persistence of substantial volumes of potentially
viable brain tissue after stroke: a correlative PET-CT study with
voxel-based data analysis.
Stroke. 1996;27:599–606.
4.
Barber PA,
Darby DG, Desmond PM, Yang Q, Gerraty RP, Jolley D, Donnan GA, Tress
BM, Davis SM. Prediction of stroke outcome with echoplanar perfusion-
and diffusion-weighted MRI.
Neurology. 1998;51:418–426.
5.
Hacke W, Kaste M,
Fieschi C, von Kummer R, Davalos A, Meier D, Larrue V, Bluhmki E, Davis
S, Donnan G, Schneider D, Diez-Tejedor E, Trouillas P.
Intravenous thrombolysis with recombinant
tissue plasminogen activator for acute
hemispheric stroke: the European Cooperative Acute Stroke Study
(ECASS). JAMA. 1995;274:1017–1025.
6.
The Multicenter
Acute Stroke Trial-Europe Study Group. Thrombolytic therapy
with streptokinase in acute ischemic stroke.
N Engl J Med. 1996;335:145–150.
7.
Furlan A, Higashida
R, Wechsler L, Gent M, Rowley H, Kase C, Pessin M, Ahuja A, Callahan F,
Clark WM, Silver F, Rivera F. Intra-arterial prourokinase
for acute ischemic stroke.
JAMA. 1999;282:2003–2011.
8. Brass LM. Brain SPECT in clinical neurology: stroke. Neurology. 1997;48(suppl 1).
9. Hurst RW. Imaging of pathophysiology of infarction in the clinical setting. AJNR Am J Neuroradiol. 1998;19:1947–1948.[Medline] [Order article via Infotrieve]
10.
Culebras A,
Kase CS, Masdeu JC. Practice guidelines for the use of imaging in
transient ischemic attacks and acute stroke: a report of the
Stroke Council, American Heart Association.
Stroke. 1997;28:1480–1497.
11. Vuadens P, Bogousslavsky J. Diagnosis as a guide to stroke therapy. Lancet. 1998;352:5–9.[Medline] [Order article via Infotrieve]
12.
Gilman S. Imaging
the brain. N Engl J
Med. 1998;338:812–820.
13. Yock DH Jr. CT demonstration of cerebral emboli. J Comput Assist Tomogr. 1981;5:190–196.[Medline] [Order article via Infotrieve]
14. von Kummer R, Meyding-Lamade U, Forsting M, Rosin L, Rieke K, Hacke W, Sartor K. Sensitivity and prognostic value of early computed tomography in middle cerebral artery trunk occlusion. AJNR Am J Neuroradiol. 1994;15:9–15.[Abstract]
15.
Tomura N, Uemura
K, Inugami A, Fujita H, Higano S, Shishido F. Early CT findings in
cerebral infarction. Radiology. 1988;168:463–467.
16.
Wardlaw JM,
Dorman PJ, Lewis SC, Sandercock P. Can stroke physicians and
neuroradiologists identify signs of early cerebral infarction on CT?
J Neurol Neurosurg
Psychiatry. 1999;67:651–653.
17.
von Kummer
R. CT of acute cerebral ischemia.
Radiology. 2000;216:611. Reply.
18. Rowley HA. Noninvasive imaging of the ischemic penumbra. Ann Neurol. 1997;42:539–541.[Medline] [Order article via Infotrieve]
19. Berrouschot J, Barthel H, Hesse S, Köster J, Knapp WH, Schneider D. Early differentiation between TIA and ischemic stroke within the first six hours after onset of symptoms by using Tc-99m-ECD-SPECT. J Cereb Blood Flow Metab. 1998;18:921–929.[Medline] [Order article via Infotrieve]
20.
Berrouschot J,
Barthel H, von Kummer R, Knapp WH, Hesse S, Schneider D.
99mTechnetium-ethyl-cysteinate-dimer
single-photon emission CT can predict fatal ischemic brain
edema. Stroke. 1998;29:2556–2562.
21. Barthel H, Berrouschot J, Hesse S, Dannenberg C, Schneider D, Knapp WH, Burchert W. Prognostic potential of Tc-99m-ECD-SPECT within 6 hours after onset of stroke symptoms. In: Höfer R, Bergmann H, eds. Radioactive Isotopes in Clinical Medicine and Research XXIII. Stuttgart/New York: Schatthauer; 1999:37–42.
22.
Hacke W, Schwab
S, Horn M, Spranger M, De Georgia M, von Kummer R. "Malignant"
middle cerebral artery territory infarction: clinical course and
prognostic signs. Arch Neurol. 1996;53:309–315.
23.
Schwab S, Steiner
T, Aschoff A, Schwarz S, Steiner HH, Jansen O, Hacke W. Early
hemicraniectomy in patients with complete middle cerebral artery
infraction. Stroke. 1998;29:1888–1893.
24. Ginsberg MD, Sternau LL, Globus MY, Dietrich WD, Busto R. Therapeutic modulation of brain temperature: relevance to ischemic brain injury. Cerebrovasc Brain Metab Rev. 1992;4:189–225.[Medline] [Order article via Infotrieve]
25.
Scandinavian
Stroke Study Group. Multicenter trial of hemodilution in
ischemic stroke: background and study protocol.
Stroke. 1985;16:885–890.
26. Hacke W, Kaste M, Fieschi C, von Kummer R, Davalos A, Meier D, Larrue V, Bluhmki E, Davis S, Donnan G, Schneider D, Diez-Tejedor E, Trouillas P. Randomised double-blind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischemic stroke (ECASS II). Second European-Australian Acute Stroke Study Investigators. Lancet. 1998;352:1245–1251.[Medline] [Order article via Infotrieve]
27.
Berrouschot J,
Barthel H, Hesse S, Knapp WH, Schneider D, von Kummer R. Reperfusion
and metabolic recovery of brain tissue and clinical outcome
after ischemic stroke and thrombolytic therapy.
Stroke. 2000;31:1545–1551.
28. Mahoni FI, Barthel DW. Functional evaluation: the Barthel Index. Md State Med J. 1965;14:61–65.[Medline] [Order article via Infotrieve]
29. Rankin J. Cerebral vascular accidents in patients over the age of 60, II: prognosis. Scott Med J. 1957;2:200–215.[Medline] [Order article via Infotrieve]
30.
Walovitch RC,
Hill TC, Garrity ST, Cheesman EH, Burgess BA, O’Leary DH, Watson AD,
Ganey MV, Morgan RA, Williams SJ. Characterization of
99mTc-L,L-ECD
for brain perfusion imaging, part 1: pharmacology of
99mTc-ECD in non-human primates.
J Nucl Med. 1989;30:1892–1901.
31.
Holman BL,
Carvalho PA, Zimmerman RE, Johnson KA, Tumeh SS, Smith AP, Genna S.
Brain perfusion SPECT using annular single cristal camera: initial
clinical experience. J Nucl
Med. 1990;31:1456–1561.
32. Barthel H, Wiener M, Dannenberg C, Bettin S, Sattler B, Knapp WH. Age-specific cerebral perfusion in 4 to 15 years old children: a high resolution brain SPECT study using Tc-99m-ECD. Eur J Nucl Med. 1997;24:1245–1252.[Medline] [Order article via Infotrieve]
33.
Giubilei F, Lenzi
GL, Di Piero V, Pozzilli C, Pantano P, Bastianello S, Argentino C,
Fieschi C. Predictive value of brain perfusion single-photon emission
computed tomography in acute ischemic stroke.
Stroke. 1990;21:895–900.
34. Yeh SH, Liu RS, Hu HH, Wong WJ, Lo YK, Lai ZY, Huang JC, Chang SL, Wang SJ, Chu FL. Brain SPECT imaging with Tc-99m-hexamethylpropyleneamine oxime in the early detection of cerebral infarction: comparison with transmission computed tomography. Nucl Med Commun. 1986;7:873–878.[Medline] [Order article via Infotrieve]
35. Kurokawa H, Iino K, Kojima H, Saito H, Suzuki M, Watanabe K, Kato T. Regional cerebral blood flow in the acute stage with ischemic cerebrovascular disease studies by xenon-133 inhalation and single photon emission computerized tomography. No To Shinkei. 1987;39:437–446.[Medline] [Order article via Infotrieve]
36. Feldmann M, Voth E, Dressler D, Henze T, Felgenhauer K. Tc-99m-Hexamethylpropylene amine oxime SPECT and x-ray CT in acute cerebral infarction. J Neurol. 1990;237:475–479.[Medline] [Order article via Infotrieve]
37.
Laloux P,
Richelle F, Jamart J, De Coster P, Laterre C. Comparative correlations
of HMPAO SPECT indices, neurological score, and stroke subtypes with
clinical outcome in acute carotid infarcts.
Stroke. 1995;26:816–821.
38.
Laloux P, Jamart
J, Meurisse H, De Coster P, Laterre C. Persisting perfusion defects in
transient ischemic attacks: a new clinically useful subgroup?
Stroke. 1996;27:425–430.
39.
Baird AE, Austin
MC, McKay WJ, Donnan GA. Sensitivity and specificity of
99mTc-HMPAO SPECT cerebral perfusion
measurements during the first 48 hours for the localization of cerebral
infarction. Stroke. 1997;28:976–980.
40. Ell PJ, Jarritt PH, Cullum I, Hocknell JM, Costa DC, Lui D, Jewkes RF, Steiner TJ, Nowotnik DP, Pickett RD, et al. A new regional cerebral blood flow mapping with Tc-99m-labelled compound. Lancet. 1985;2:50–51.
41.
Leveille J,
Demonceau G, De Roo M, Rigo P, Taillefer R, Morgan RA, Kupranick D,
Walovitch RC. Characterization of
technetium-99m-L,L-ECD
for brain perfusion imaging, part 2: biodistribution and brain imaging
in humans. J Nucl Med. 1989;30:1902–1910.
42. Matsuda H, Li YM, Higashi S, Sumiya H, Tsuji S, Kinuya K, Hisada K, Yamashita J. Comparative SPECT study of stroke using Tc-99m ECD, I-123 IMP, and Tc-99m HMPAO. Clin Nucl Med. 1993;18(9):754–758.
43.
Ogasawara K,
Mizoi K, Fujiwara S, Yoshimoto T.
99mTc-Bicisate and
99mTc-HMPAO SPECT imaging in early
spontaneous reperfusion of cerebral embolism.
Am J Neuroradiol. 1999;20:626–628.
44.
Jacquier-Sarlin
M, Polla B, Siosman D. Cellular basis of ECD brain retention.
J Nucl Med. 1996;37:1694–1697.
45. Barthel H, Kämpfer I, Seese A, Dannenberg C, Kluge R, Burchert W, Knapp WH. Improvement of brain SPECT by stabilization of Tc-99m-HMPAO with methylene blue or cobalt chloride: comparison with Tc-99m-ECD. Nuklearmedizin. 1999;38:80–84.[Medline] [Order article via Infotrieve]
46. Pulsinelli W. Pathophysiology of acute cerebral ischemia. Lancet. 1992;359:533–536.
47. Fiorelli M, Toni D, Bastianello S, Sacchetti ML, Sette G, Falcou A, Argentino C, Lorenzano S, Di Angelantonio E, Bozzao L. Computed tomography findings in the first few hours of ischemic stroke: implications for the clinician. J Neurol Sci. 2000;173:10–17.[Medline] [Order article via Infotrieve]
48.
Astrup J,
Siesjö BK, Symon L. Threshold in cerebral ischemia: the
ischemic penumbra.
Stroke. 1981;12:723–725.
49.
Fisher M.
Characterizing the target of acute stroke therapy.
Stroke. 1997;28:866–872.
50. Lythgoe MF, Williams SR, Busza AL, Wiebe L, McEwan AJ, Gadian DG, Gordon I. The relationship between magnetic resonance diffusion imaging and autoradiographic markers of cerebral blood flow and hypoxia in an animal stroke model. Magn Reson Med. 1999;41:706–714.[Medline] [Order article via Infotrieve]
51. Weir CJ, Bolster AA, Tytler S, Murray GD, Corrigall RS, Adams FG, Lees KR. Prognostic value of single-photon emission tomography in acute ischaemic stroke. Eur J Nucl Med. 1997;24:21–26.[Medline] [Order article via Infotrieve]
52. Lees KR, Weir CJ, Gillen GJ, Taylor AK, Ritchie C. Comparison of mean cerebral transit time and single-photon emission tomography for estimation of stroke outcome. Eur J Nucl Med. 1995;22:1261–1267.[Medline] [Order article via Infotrieve]
53. von Kummer R, Holle R, Gizyska U, Hofmann E, Jansen O, Petersen D, Schumacher M, Sartor K. Interobserver agreement in assessing early CT signs of middle cerebral artery infarction. AJNR Am J Neuroradiol. 1996;17:1743–1748.[Abstract]
54. von Kummer R. Effect of training in reading CT scans on patient selection for ECASS II. Neurology. 1998;51(suppl 3):50–52.
55.
Baron JC, Marchal
G. What is the predictive value of increased
technetium-99m-HMPAO uptake for brain survival/necrosis in
the acute stage of ischemic stroke?
J Nucl Med. 1995;36:2392.
Letter to the Editor.
56. Becker H, Desch H, Hacker H, Pencz A. CT fogging effect with ischemic cerebral infarcts. Neuroradiology. 1979;18:185–192.[Medline] [Order article via Infotrieve]
57.
Alexandrov
AV, Black SE, Ehrlich LE, Bladin CF, Smurawska LT, Pirisi A, Caldwell
CB. Simple visual analysis of brain perfusion on HMPAO SPECT
predicts early outcome in acute stroke.
Stroke. 1996;27:1537–1542.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||