| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Neuroradiology |
1 From the Department of Radiology, Magnetic Resonance Unit (A.R., P.O., X.A., E.G., A.R.G.) and Department of Neurology, Cerebrovascular Unit (J.A.S., J.F.A., C.M.), Hospital Universitari Vall d’Hebron, Passeig Vall d’Hebron 119–129, 08035 Barcelona, Spain. Received February 19, 2003; revision requested May 7; final revision received November 4; accepted January 5, 2004. Address correspondence to A.R. (e-mail: alex.rovira@idi-cat.org).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Forty-two consecutive patients with stroke involving the MCA territory underwent MR imaging within 6 hours after clinical onset. MR examination included echo-planar T2*-weighted, diffusion-weighted (DW), and perfusion-weighted (PW) imaging and MR angiography. Presence or absence of the susceptibility sign on echo-planar T2*-weighted images, which is indicative of acute thrombotic occlusion involving MCA or ICA, was assessed in consensus by two observers blinded to clinical information and other MR imaging data. Differences in lesion volume on DW and PW images between patients with and those without the susceptibility sign were evaluated with the Mann-Whitney test. P < .05 was considered to indicate a significant difference.
RESULTS: Thirty patients (71%) had a positive susceptibility sign that correlated with MCA or ICA occlusion at MR angiography in all cases (sensitivity, 83%; specificity, 100%). Mean lesion volume on PW images was higher in patients with a positive susceptibility sign (P = .01), but no significant differences were found in mean lesion volume on DW images. Cases in which the susceptibility sign was identified proximal to MCA divisional bifurcation (27 patients) showed a mean perfusion deficit of 83.9% of the total MCA territory (range, 50%–100%).
CONCLUSION: Presence of the susceptibility sign proximal to MCA bifurcation provides fast and accurate detection of acute proximal MCA or ICA thrombotic occlusion.
© RSNA, 2004
Index terms: Arteries, middle cerebral, 174.4311, 174.4312, 174.4352 • Brain, infarction, 17.4352, 17.781, 174.4352, 174.781 • Brain, MR, 174.121411, 174.121412, 174.121413, 174.121415, 174.121416, 174.12142, 174.12143, 174.12144 • Magnetic resonance (MR), vascular studies, 17.12144, 174.12144
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Many authors have suggested that infarct patterns can be differentiated in hyperacute stroke by means of diffusion-weighted (DW) and perfusion-weighted (PW) MR imaging and MR angiography (11–19) and may provide rational selection criteria for therapeutic strategies based on the presence or absence of tissues at risk for irreversible infarction and the presence of an acute clot within an intracranial artery. It has been suggested that the MR imaging protocol for this purpose should include an echo-planar T2*-weighted sequence because of the high sensitivity in detecting acute hemorrhage (9), which would obviate a preliminary CT study. Moreover, this sequence has demonstrated a high sensitivity and specificity in identification of acute middle cerebral artery (MCA) or internal carotid artery (ICA) thrombotic occlusion (20), although the diagnostic value of the sequence for determining therapy decisions has not been well established. The purpose of the study was to evaluate the accuracy of echo-planar T2*-weighted imaging in the detection of acute MCA or ICA thrombotic occlusion.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
Stroke onset was defined as the last time the patient was known to have no neurologic symptoms. Stroke severity was assessed by using the National Institutes of Health Stroke Scale (NIHSS). All stroke severity examinations were performed at admission by a stroke neurologist (J.F.A.) or a senior neurology resident (C.M.). The mean NIHSS score at admission was 15.5 (range, 0–22). Our institutional review board did not require its approval or patient informed consent for this retrospective study.
MR Examination
MR examinations were performed with a 1.5-T whole-body imager (Magnetom Vision Plus; Siemens Medical Systems, Erlangen, Germany) with 24-mT/m gradient strength, 300-µsec rise time, and an echo-planar–capable receiver equipped with a gradient overdrive. The following images were obtained: (a) transverse T2-weighted susceptibility-based echo-planar gradient-echo images (0.8/29 [repetition time msec/echo time msec], one signal acquired, total acquisition time of 2.8 seconds); (b) transverse DW echo-planar spin-echo images (4,000/100, two signals acquired, total acquisition time of 56 seconds); (c) transverse PW echo-planar gradient-echo images (2,000/60, 40 signals acquired, total acquisition time of 80 seconds); and (d) MR angiograms (30/5.4, 15° flip angle, total acquisition time of 156 seconds).
DW imaging was performed with a single-shot spin-echo echo-planar pulse sequence, with three b values of 0, 500 and 1,000 sec/mm2 along all three orthogonal axes over 15 transverse 5-mm-thick sections, 1.5-mm intersection gap, 240-mm field of view, and 96 x 128 matrix. To minimize the effects of diffusion anisotropy, DW data were automatically processed to yield standard isotropic DW imaging. From the DW images, we calculated the apparent diffusion coefficient maps. These apparent diffusion coefficient maps were used to confirm the presence and extent of acute brain infarcts (lower values as compared with those of normal brain tissue) on DW images.
Dynamic first-pass PW imaging was performed with a bolus of 0.1 mmol gadopentetate dimeglumine (Magnevist; Schering, Berlin, Germany) per kilogram of body weight for selected 15-section positions measured sequentially 40 times. The bolus was injected into an antecubital vein with an MR-compatible power injector (Spectris; Medrad, Pittsburgh, Pa) at an injection rate of 5 mL/sec starting 5 seconds after initiation of imaging, followed by a flush with 15 mL of saline. The PW sequence generated a time-to-peak map for each section position that was immediately available for interpretation at the console with all the other images. PW images were obtained by using 5-mm-thick sections, 1.5-mm intersection gap, 240-mm field of view, and 128 x 128 matrix.
For MR angiography, we used a transverse gradient-echo three-dimensional time-of-flight sequence with magnetization transfer suppression and tilted optimized nonsaturating excitation, with 1.5-mm-thick sections (47-mm slab thickness), 200-mm field of view, and 200 x 512 matrix. Maximal intensity projection reconstructions were performed at the time of imaging. The mean MR examination time, from entrance to exit from the MR suite, was 20 minutes (range, 11–30 minutes).
Follow-up MR imaging was performed 3–5 days after the first MR study in 28 patients. In two of these patients, the follow-up MR study confirmed the presence of an acute infarct that was not identified at the initial DW imaging. This examination followed the same imaging protocol as was used initially (with the exception of PW sequence) and included a transverse T2-weighted fast spin-echo (3,000/85, two signals acquired) or fast fluid-attenuated inversion-recovery (9,000/110/2,200 [repetition time msec/echo time msec/inversion time msec], two signals acquired) sequence.
Image Analysis
All images were retrospectively read by two of the authors (A.R. and P.O) who have 11 and 2 years of experience, respectively, in interpreting brain MR images. Hard-copy images of the echo-planar T2*-weighted sequences were separated from the other images, randomly mixed, and evaluated by the two observers who were blinded to the pattern of stroke symptoms and to the findings obtained with use of the other MR sequences. The images were read in three sessions during 1 week, 3 months after inclusion of the last patient in the study. The two observers read the images together and reached the final decision in consensus regarding the presence or absence of the susceptibility sign.
The susceptibility sign on MR images was defined as presence of hypointensity within the ICA or MCA, in which the diameter of the hypointense signal within the vessel exceeded the contralateral vessel diameter (Fig 1). The susceptibility sign was classified according to its location as follows: (a) ICA or proximal MCA trunk, (b) distal MCA trunk, or (c) not present (Fig 2).
|
|
|
|
|
|
Occlusion of the intracranial ICA or MCA on MR angiograms was classified as follows: (a) no occlusion, (b) occlusion proximal to the intracranial ICA bifurcation, (c) MCA occlusion proximal to its divisional bifurcation, and (d) MCA occlusion distal to its divisional bifurcation.
Statistical Analysis
Statistical analysis was performed with a software package (SPSS version 8.0; SPSS, Chicago, Ill). The Mann-Whitney U and 
2 tests were used to assess statistically significant differences in age between men and women and in those with the susceptibility sign. Neither test revealed statistical differences between these variables. In light of this fact, it was not necessary to adjust for age- and sex-related differences in the statistical analysis. Differences in NIHSS scores and lesion volumes on DW and PW images between patients with and those without the susceptibility sign were evaluated with the Mann-Whitney test. P < .05 indicated a significant difference. The diagnostic value of the susceptibility sign with regard to the presence of arterial occlusion was expressed as sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. To quantify uncertainty, 95% CIs for each of the MR measures were determined.
In the analysis of differences by means of logistic regression, lesion volume on PW images was first transformed into a dichotomous variable. By using the receiver operating characteristic curve, this transformation was based on the determination of a cutoff point for the lesion volume on PW images (>58% of the MCA territory) that maximized both sensitivity and specificity of this variable in relation to the susceptibility sign. The dependence of dichotomous PW imaging on other variables (age, sex, NIHSS, DW imaging, and susceptibility sign) was evaluated by means of logistic regression with the maximum likelihood variable selection method.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
MR angiography demonstrated arterial occlusion related to the clinically affected hemisphere in 36 patients. The occlusion was located in the intracranial segment of the ICA in 10 patients, in the MCA proximal to its divisional bifurcation in 18 patients, and in the MCA distal to its divisional bifurcation in eight patients. In the remaining six patients, no arterial occlusion was detected at MR angiography; although in four of the patients, moderate to severe stenosis of the proximal MCA trunk was readily identified.
In 27 patients with the susceptibility sign proximal to the MCA divisional trunk, MR angiography depicted ICA or proximal MCA trunk occlusion (Fig 3). Moreover, in three patients in whom the susceptibility sign was identified distal to the MCA divisional trunk, MR angiography helped confirm occlusion at this location (Fig 4). Six patients showed a false-negative susceptibility sign; five of them had distal occlusion of the MCA at MR angiography. Only one patient with proximal MCA occlusion at MR angiography and who was receiving anticoagulant therapy did not have a positive susceptibility sign (Fig 5). The six patients with a true-negative susceptibility sign had no arterial occlusions or proximal MCA moderate- to high-grade stenoses at MR angiography.
|
|
|
|
|
|
|
|
|
|
|
|
|
Analysis of the relationship between the susceptibility sign proximal to MCA and perfusion deficit showed that determination of statistically significant differences in lesion volume on PW images between patients with a negative and those with a positive susceptibility sign proximal to MCA divisional bifurcation reached a power of 83.66% for the Mann-Whitney test in detection of these differences, with the significance level set at 
= .05 and with an assumption of logistic distribution.
The mean volume of the PW deficit was 192.2 cm3 (70.3% of the MCA territory) for all patients, 224.6 cm3 (80.5% of the MCA territory) for patients with a positive susceptibility sign, and 109.6 cm3 (44.8% of the MCA territory) for patients with a negative susceptibility sign. Differences between the two groups were significant (P = .01). By considering only those patients who had the susceptibility sign proximal to MCA bifurcation, the mean volume of the lesion on DW images was 68.5 cm3 (21.8% of the MCA territory) and the PW deficit was 237.3 cm3 (83.9% of the MCA territory). Differences between patients with the susceptibility sign proximal to MCA and those without the sign were significant for both DW (P < .05) and PW (P < .001) imaging. As expected, significantly higher NIHSS scores at admission were also observed in these patients (P < .05). A summary of all the results is shown in Table 2.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
The susceptibility-based sequence is the first sequence performed in our routine stroke protocol and is used to rule out hyperacute parenchymal hemorrhage. In our institution, cranial CT is not performed prior to MR imaging to rule out hemorrhage, because therapy decisions are required immediately in these patients. For this reason, we could not compare the sensitivity of the susceptibility sign at MR imaging with the sensitivity of the hyperattenuating MCA sign at CT. Nevertheless, the sensitivity of MR angiography as the reference standard is in keeping with that in a previous study in which this comparison was made (20).
The signal loss that helps identify the susceptibility sign is explained by severe T2 shortening at an acute clot, representing the magnetic susceptibility differences that arise from intracellular deoxyhemoglobin, which is present in high concentrations in acute thromboembolism. This magnetic susceptibility effect produces a nonuniform magnetic field and a rapid dephasing of proton spins, which results in a signal loss best seen on T2*-susceptibility-weighted images (22,23). Other likely contributors to this signal loss are increased hematocrit levels and hemoglobin concentrations due to clot formation and retraction (24) and fibrin polymerization (25).
For the detection of the susceptibility sign, we used 5-mm section thickness and fast two-dimensional echo-planar T2*-weighted MR imaging, which covers the entire brain in less than 3 seconds. This technique demonstrated high sensitivity in depiction of hyperacute thrombotic occlusion within the proximal MCA and intracranial ICA and a low sensitivity in depiction of occlusion within the distal MCA branches. Explanations for the single false-negative finding observed in a patient with proximal MCA occlusion might include limited spatial resolution of the sequence or nonacute arterial thrombotic occlusion. However, these reasons seem unlikely, since a repeat susceptibility-based sequence with thinner sections (3 mm) and different orientations that was performed in this patient also failed to show the susceptibility sign. Also, the arterial occlusion was associated with a large acute infarct in the corresponding territory, a fact that almost excludes the possibility of chronic arterial occlusion. Delayed formation of deoxyhemoglobin within the clot is a more likely explanation in this patient, who is undergoing anticoagulant therapy.
No significant differences were observed in the extent of the infarct on DW images between patients with the sign and those without it, a fact that is explained by the wide range of lesion sizes in both groups. Therefore, the presence of this sign should not be used to exclude patients from receiving thrombolytic therapy. These findings are in accordance with the therapeutic implications given to the hyperattenuating MCA sign as detected on early CT scans (26–30), which is not considered a contraindication for reperfusion treatment. However, patients with the susceptibility sign had worse clinical status at admission, as measured with the NIHSS, than did patients without the sign.
MR criteria to identify candidate patients for thrombolysis have not been definitively established. The most generally accepted criteria include extension of the infarct to less than 33% of the MCA territory on DW images, PW imaging–DW imaging mismatch greater than 20%, and demonstration of an occluded intracranial artery at MR angiography that is anatomically relevant to the lesion identified on DW and PW images (31).
Presence of the susceptibility sign proximal to the MCA bifurcation in our patients was always associated with a large perfusion deficit of more than half of the MCA territory. In fact, the 95% CI of the perfusion deficit ranged from 77% to 91% of the MCA territory. Therefore, in patients with the susceptibility sign proximal to MCA divisional bifurcation (about two-thirds of our series), the information obtained with PW images would be redundant, since the extent of the perfusion deficit can be accurately anticipated.
Since no false-positive cases of acute thromboembolism were registered with the echo-planar T2*-weighted sequence in our study, the need for confirmation of arterial occlusion with MR angiography seems unnecessary.
The data from this study indicate that in almost two-thirds of patients with acute MCA stroke due to ICA or MCA occlusion, treatment decisions could be made on the basis of information obtained with use of a short MR protocol, including echo-planar T2*-weighted and DW sequences aimed toward selection of patients who might benefit from thrombolytic therapy. Our results contrast partially with those from a previous study (20) in which patients with the susceptibility sign were reported to have limited perfusion deficits, indicating good collateral blood supply. However, proximally and distally located thromboembolism was grouped together in that study, whereas in our study, we considered these groups separately.
The standard set of MR imaging sequences commonly used in the initial assessment of hyperacute stroke (echo-planar T2*-weighted, DW and PW imaging, and MR angiography) can be performed in a total examination time of 20 minutes; in noncooperative patients, the time required may be longer. This relatively long examination time, together with the limited availability of MR units, are the main drawbacks to the use of MR imaging in the triage of patients for reperfusion therapy. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Trial demonstrated that a delay in therapy of 20–30 minutes may diminish the likelihood of favorable outcome by 10%–20% (32,33). Although sequence performance and image reconstruction time only partially account for the total examination time (transport, access, and positioning may also contribute substantially), the use of a faster MR protocol with a total acquisition and interpretation time of less than 2 minutes (echo-planar T2*-weighted and DW imaging) should reduce the total examination time and produce a benefit to the outcome of these patients.
On the basis of the findings obtained in this study with the combination of echo-planar T2*-weighted and DW imaging, we propose the following four categories: category A, a positive susceptibility sign located proximal to the MCA divisional bifurcation; category B, a positive susceptibility sign located distal to the MCA divisional bifurcation and a lesion of less than 33% of the MCA territory on DW images; category C, a negative susceptibility sign and a lesion of less than 33% of the MCA territory on DW images; and category D, a negative susceptibility sign and a lesion greater than 33% of the MCA territory on DW images. For two of the categories (A and D), which represented 66% of our patients, the decision on whether to treat patients with thrombolytic agents can be made simply by combining the findings of echo-planar T2*-weighted and DW images. Patients in category B (7%) would require PW imaging, and patients in category C (26%) would require both PW imaging and MR angiography.
Our study had some limitations. MR angiography was used as the reference standard for arterial occlusion, instead of conventional digital subtraction angiography. Although MR angiography has proved to be highly accurate in depicting high-grade stenosis and occlusions (34), in a small number of cases it may cause misdiagnosis of subtotal occlusions as total occlusions. From the viewpoint of decision making, however, this fact is of limited importance, since thrombolytic therapy is targeted toward thrombotic emboli, regardless of whether they produce total or subtotal arterial occlusion.
Another point to consider is that most of our patients were examined beyond the 3-hour window and always no later than 6 hours after the onset of symptoms. Also, no serial MR examinations were performed in the first days. Therefore, there are not enough data to establish the minimum time needed to study an acute thrombus with the susceptibility-based T2* sequence, and the signal changes that may occur over time are unknown. For this reason, we cannot ensure that the high level of diagnostic accuracy obtained in this study will be reproduced in patients in whom the MR examination is performed within the first 3 hours or beyond the 6 hours after symptoms onset.
Fast echo-planar T2*-weighted MR imaging performed as the first sequence in the MR protocol seems to be useful in the triage of patients with hyperacute stroke. Presence of the susceptibility sign proximal to the MCA bifurcation provides fast and accurate detection of acute proximal MCA or ICA thrombotic occlusion and helps predict the extent of perfusion deficit. In patients with a positive susceptibility sign proximal to MCA bifurcation, which corresponded to almost two-thirds of our patients with acute stroke, the extent of the infarct on DW images might be the only additional imaging information required to establish therapy decisions, a fact that implies a considerable savings in examination time and evaluation analysis.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, A.R., A.R.G.; study concepts, A.R.; study design, A.R., P.O.; literature research, A.R., P.O., A.R.G.; clinical studies, J.F.A., C.M., J.A.S.; data acquisition, E.G., A.R., A.R.G., J.F.A., C.M.; data analysis/interpretation, A.R., P.O., X.A.; statistical analysis, X.A.; manuscript preparation, definition of intellectual content, editing, and final version approval, A.R.; manuscript revision/review, A.R.G., J.A.S., E.G.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Siemonsen, T. Fitting, G. Thomalla, P. Horn, J. Finsterbusch, P. Summers, C. Saager, T. Kucinski, and J. Fiehler T2' Imaging Predicts Infarct Growth beyond the Acute Diffusion-weighted Imaging Lesion in Acute Stroke Radiology, September 1, 2008; 248(3): 979 - 986. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. S. Kim, D. H. Lee, C. G. Choi, S. J. Kim, and D. C. Suh Progression of middle cerebral artery susceptibility sign on t2*-weighted images: its effect on recanalization and clinical outcome after thrombolysis. Am. J. Roentgenol., December 1, 2006; 187(6): W650 - W657. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Srinivasan, M. Goyal, F. A. Azri, and C. Lum State-of-the-Art Imaging of Acute Stroke RadioGraphics, October 1, 2006; 26(suppl_1): S75 - S95. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Selim Vertebral artery susceptibility sign as a marker of vertebral thromboembolism on magnetic resonance imaging. Arch Neurol, September 1, 2006; 63(9): 1330 - 1330. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Rubiera, M. Ribo, R. Delgado-Mederos, E. Santamarina, P. Delgado, J. Montaner, J. Alvarez-Sabin, and C. A. Molina Tandem Internal Carotid Artery/Middle Cerebral Artery Occlusion: An Independent Predictor of Poor Outcome After Systemic Thrombolysis Stroke, September 1, 2006; 37(9): 2301 - 2305. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Molina Imaging the Clot: Does Clot Appearance Predict the Efficacy of Thrombolysis? Stroke, November 1, 2005; 36(11): 2333 - 2334. [Full Text] [PDF]
|
|
|
|
|
|
M. Ribo, C. A. Molina, A. Rovira, M. Quintana, P. Delgado, J. Montaner, E. Grive, J. F. Arenillas, and J. Alvarez-Sabin Safety and Efficacy of Intravenous Tissue Plasminogen Activator Stroke Treatment in the 3- to 6-Hour Window Using Multimodal Transcranial Doppler/MRI Selection Protocol Stroke, March 1, 2005; 36(3): 602 - 606. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||
