(Stroke. 2004;35:1657.)
© 2004 American Heart Association, Inc.
Original Contributions |
Editorial Comment—Specificity of Stroke Imaging
Disregarded and Neglected
University of Technology, Uniklinikum-Dresden, Germany
Brain imaging in acute stroke patients can be effective on 6 different levels:1 (1) It will reduce health care costs, if it prevents disability and death of stroke victims. (2) Brain imaging will improve the clinical outcome of stroke patients, if it identifies the patients who benefit from specific treatment. (3) To identify these patients, brain imaging must provide relevant information that is unavailable from other sources for the appropriate choice of treatment. (4) This could be brain images that allow the exclusion of brain hemorrhage and other diseases that mimic ischemic stroke, and allow assessment of ischemic edema and perfusion disturbance, mass effect, arterial wall pathology, and obstruction. (5) The imaging modality should be sensitive and specific for stroke pathology early after symptom onset. (6) This requires the imaging modality to have the technical capacity to reliably detect the relevant stroke pathology.
In this issue, Schramm et al2 compare computed tomography (CT) and MRI (MRI) techniques that assessed brain perfusion, brain water diffusion, and the resulting ischemic infarct in acute stroke patients. In a group of 22 patients, among them 9 patients without a brain infarct on follow-up CT, Schramm et al did not detect a statistical difference regarding the lesion volumes on time-to-peak (TTP) maps and on cerebral blood volume (CBV) maps provided by CT and MRI, observed no difference when comparing the lesion volumes on CT angiography (CTA) source images with the lesions on diffusion-weighted MRI (DWI), and found a significant correlation when comparing the lesion volumes on CT-CBV maps and on follow-up CT.
These valuable observations allow one not only to compare between CT and MRI, but invite one to think about the specificity of imaging information, and raise several questions:
- (1) Does the lack of a statistical difference in lesion volumes and its significant correlation really mean that different imaging modalities assess the same pathology?
The authors subdivided their already small group of patients into those whom they treated with thrombolysis (n=13) and those they did not treat (n=9) after perfusion imaging. The risk to miss small differences between such small groups is rather high. They did not report how the MRI perfusion maps were created and why they obtained CBF maps from CT only. Nevertheless, even if one takes the entire group, the lack of a statistical difference regarding the "lesion" volumes and a high correlation between CT- and MRI-perfusion parameters suggest that CT and MRI detect the same volume of contrast flow disturbance on TTP and CBV maps, which is not a surprise, if similar techniques are used to create such maps. It is interesting that the authors observed a high correlation between the volumes of hypoattenuating brain tissue on CTA source images and of DWI "lesions" with no significant difference of the mean values. Hypoattenuation on CTA source images is caused by ischemic edema and a lack of contrast enhancement reflecting a reduced CBV, whereas the increased signal on DWI is caused by the restricted water diffusion in brain regions with diminished extracellular fluid space and cellular edema.3 A reduction of the extracellular fluid space and a decline of the apparent diffusion coefficient were found at a CBF threshold of 30 mL per 100 g·min.4,5 One may wonder why the "lesions" on DWI in this series of patients were the smallest compared with the other displayed parameters, despite the high sensitivity of DWI for relatively mild degrees of brain ischemia. The authors observed a high correlation between the "lesions" on CT-CBV maps and on CTA source images; on the average the CTA "lesions" were smaller, however. This may suggest that hypoattenuation on CTA source images is mainly caused by the water uptake of ischemic brain tissue. The water content of ischemic brain tissue is indirectly correlated with x-ray attenuation6 and increases only in regions with a CBF below 10 mL per 100 g·min.4 This explains why hypoattenuation of ischemic brain tissue is highly predictive for ischemic damage.7
- (2) Can we consequently conclude that one imaging modality can replace another?
The observations of Schramm et al suggest that it does not matter whether one takes CT or MRI to detect tissue volumes of decreased CBV or prolonged TTP. The CTA source images may serve as a surrogate for DWI although they are likely to display a different pathology. In light of the relatively low feasibility to perform MRI, it is surprising that the authors did not conclude to prefer CT to MRI.
- (3) Are the lesions that were compared relevant for treatment of ischemic stroke?
What is the "diagnostic value" of perfusion maps (not "perfusion-weighted imaging") the authors were looking for? Time-to-peak maps appeared most sensitive in detecting brain perfusion disturbance; however, a reference standard is not available. All 16 patients with TTP prolongation developed infarcts on the follow-up CT with the exception of 3 patients who were treated with thrombolysis. The follow-up CT remained normal in all 6 patients with normal TTP at baseline. This supports the view that reperfusion therapy is unnecessary in patients without a deficit on perfusion CT. The authors may have missed the therapeutic impact of this specific information. They treated 2 patients with IV thrombolysis who had no arterial occlusion and no perfusion deficit, but did not treat 5 patients with perfusion deficits or arterial obstructions, among them 1 patient with basilar artery occlusion and 1 patient with a proximal MCA occlusion. It appears as if the diagnostic and therapeutic impact of a normal CT or MRI in acute stroke patients is widely underestimated. Schramm et al regard the "diagnostic yield" of unenhanced CT as low by neglecting the impact of normal CT in patients with acute stroke. A normal CT excludes major ischemic damage with high specificity7 and thus allows reperfusion therapy – in my personal experience – even beyond accepted time windows. From the data of this paper, I conclude that reperfusion therapy makes no sense if CT does not show a perfusion deficit of the affected brain region.
- (2) Can we consequently conclude that one imaging modality can replace another?
 |
References |
|---|
 Top References |
|---|
- Fryback D, Thornbury J. The efficacy of diagnostic imaging. Med Decis Making. 1991: 88–94.
- Schramm P, Schellinger P, Klotz E, Kallenberg K, Fiebach J, Külkens S, Heiland S, Knauth M, Sartor K. Comparison of perfusion computed tomography and computed tomography angiography source images with perfusion-weighted imaging and diffusion-weighted imaging in patients with acute stroke less than 6 hours. Stroke. 2004; 35: 1652–1658.
[Abstract/Free Full Text] - Mintorovitch J, Yang G, Shimizu H, Kucharczyk J, Chan P, Weinstein P. Diffusion-weighted magnetic resonance imaging of acute focal cerebral ischemia: comparison of signal intensity with changes in brain water and Na+,K(+)-ATPase activity. J Cereb Blood Flow Metab. 1994; 14: 332–336.[Medline] [Order article via Infotrieve]
- Hossmann KA, Schuier FJ. Experimental brain infarcts in cats. I. Pathophysiological observations. Stroke. 1980; 11: 583–592.
[Abstract/Free Full Text] - Wang Y, Hu W, Perez-Trepichio A, Ng T, Furlan A, Majors A, Jones S. Brain tissue sodium is a ticking clock telling time after arterial occlusion in rat focal cerebral ischemia. Stroke. 2000; 31: 1386–1392.
[Abstract/Free Full Text] - Dzialowski I, Weber J, Doerfler A, Forsting M, von Kummer R. Brain tissue water uptake after middle cerebral artery occlusion assessed with CT. J Neuroimaging. 2004; 14: 42–48.[CrossRef][Medline] [Order article via Infotrieve]
- von Kummer R, Bourquain H, Bastianello S, Bozzao L, Manelfe C, Meier D, Hacke W. Early prediction of irreversible brain damage after ischemic stroke by computed tomography. Radiology. 2001; 219: 95–100.
[Abstract/Free Full Text]
Related Article:
- Comparison of Perfusion Computed Tomography and Computed Tomography Angiography Source Images With Perfusion-Weighted Imaging and Diffusion-Weighted Imaging in Patients With Acute Stroke of Less Than 6 Hours’ Duration
- Peter Schramm, Peter D. Schellinger, Ernst Klotz, Kai Kallenberg, Jochen B. Fiebach, Sonja Külkens, Sabine Heiland, Michael Knauth, and Klaus Sartor
Stroke 2004 35: 1652-1658.[Abstract] [Full Text] [PDF]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||