The trusted source for
healthcare information and
MRI or CT: Which is Best for Acute Stroke Imaging?
Abstract & Commentary
By Dana Leifer, MD, Associate Professor, Neurology, Weill Medical College, Cornell University. Dr. Leifer reports no financial relationship relevant to this field of study.
This article originally appeared in the June 2007 issue of Neurology Alert, It was edited by Matthew E. Fink, MD, and peer reviewed by M. Flint Beal, MD. Dr. Fink is Vice Chairman, Professor of Critical Care Neurology, NewYork-Presbyterian Hospital, and Dr. Beal is Professor and Chairman, Department of Neurology, Cornell University Medical College. Drs. Fink and Beal report no financial relationships relevant to this field of study.
Synopsis: MRI is more sensitive than CT for detecting acute ischemia and can detect acute hemorrhage with equal sensitivity to CT.
Source: Chalela JA, et al. Magnetic resonance imaging and computed tomography in emergency assessment of patients with suspected acute stroke: A prospective comparison. Lancet. 2007;369:293-298.
Magnetic Resonance Imaging (MRI) can visualize acute strokes with great sensitivity, as well as identify acute and chronic intracranial hemorrhage. Computed tomography (CT), nevertheless, is still the first study obtained on stroke patients at almost all hospitals because of its greater availability and because of concerns about the ability of MRI to diagnose acute hemorrhage. To compare the utility of MRI and CT, Chalela and colleagues performed a blinded, prospective comparison of CT and MRI in 356 acute stroke patients at a single hospital.
All patients referred to the hospital's stroke team were eligible. Patients were excluded if either CT or MRI were not done. Reasons for not doing both studies included contraindication to MRI, clinical syndrome strongly suggestive of subarachnoid hemorrhage, initiation of antithrombotic or thrombolytic therapy before completion of both scans, or inability to complete both scans quickly enough to permit intravenous thrombolysis within 3 hours of symptom onset. Of the 450 patients who were screened, 94 were excluded from the study.
A panel of 4 experts reviewed the images for the study. Positive findings were considered to be present if 3 of the 4 readers agreed. The readers did not have access to clinical information.
The primary result was that MRI detected acute ischemic stroke in 52% of the patients, and CT only in 17% (p < 0.0001). The greater sensitivity of MRI is not surprising, as all patients underwent diffusion-weighted imaging, which can detect acute stroke within one hour of onset.
Of importance, although MRI is more sensitive than CT, there were still 33 false negative MRI results among the 356 studies reviewed by the panel of experts. In other words, the expert panel did not identify an acute stroke on the MRI of 33 patients who had a final diagnosis of acute ischemic stroke. Of note, the treating physicians who had access to clinical information (unlike the panel of experts) identified acute strokes in 23 of these 33 MRI studies. This finding suggests that communication between clinicians and radiologists is important in interpretation of acute stroke imaging.
With regard to false negative MRI studies, it is important to emphasize that imaging confirmation of acute stroke is not necessary to initiate intravenous thrombolytic therapy or other conventional stroke treatments. It is only necessary to rule out acute hemorrhage before intravenous thrombolysis. Nevertheless, imaging confirmation of stroke is important in some settings — for example, if a patient presents with a seizure and an acute deficit that may be ischemic or may be a postictal Todd's paralysis, or if a patient is being considered for intra-arterial treatment.
With regard to acute hemorrhage, MRI and CT had similar sensitivities. MRI found intracranial hemorrhage in 6% and CT in 7%. CT identified 4 patients with hemorrhage that was not seen on MRI, and MRI identified 2 patients with CT scans that were negative. Among the 90 patients imaged within 3 hours, MRI missed 2 acute hemorrhages that were seen on CT, while CT missed one acute hemorrhage seen on MRI. Thus, neither test is 100% accurate for diagnosis of hemorrhage; in the past, one of the reasons for preferring CT over MRI has been the belief that CT is the gold standard for detection of acute hemorrhage. Additional information about the clinical significance of the hemorrhages that were not detected by CT or by MRI might be helpful but was not provided.
MRI was found to detect chronic hemorrhages in 26% of patients, whereas CT, which is poor in detecting chronic blood products, did not identify any chronic bleeds. The extent to which the presence of old hemorrhage is clinically important is not certain, but it may influence choice of thrombolytic or antithrombotic treatment in some cases.
In view of all of these findings, Chalela et al conclude that MRI should be the imaging modality of choice for patients suspected of having an acute stroke. MRI's high sensitivity and inter-rater reliability are in its favor, but the results with expert readers do not necessarily apply to community practice where specialists may not be available to interpret scans acutely. This especially remains a concern with regard to identification of acute hemorrhage in patients being considered for intravenous thrombolysis.
In addition, it should be noted that 20% of patients were excluded from the study primarily because of contraindications to MRI, or lack of time, so the results of the study do not apply to all acute stroke patients. The advantages of MRI should not be taken as reason to perform MRI in unstable or uncooperative patients whose treatment may be excessively delayed by attempts to do so. Indeed, for any patient who is a candidate for intravenous thrombolysis, MRI may introduce significant delay. This paper, nevertheless, demonstrates that there are substantial advantages to MRI. Its use in acute stroke ought to be considered instead of CT in many patients. In addition, when new emergency departments are being designed, rapid access to MRI should be incorporated as a design feature that will facilitate the care of stroke patients.