PHOENIX—Noncontrast CT is the imaging method of choice for initial stroke evaluation, but MRI may be a better tool for the job, according to the results of two prospective multicenter investigations presented at the 28th International Stroke Conference.

In one of the investigations, Chelsea S. Kidwell, MD, and colleagues compared CT to multimodal MRI in 169 patients who presented with focal stroke symptoms within six hours of symptom onset. “MRI … appears to be as accurate as CT in detecting hyperacute intracerebral hemorrhage,” related Dr. Kidwell, who is an Assistant Professor of Neurology at the University of California, Los Angeles (UCLA), as well as a Co-Director of the UCLA Stroke Center. She added that, “MRI is superior to CT in detecting evidence of chronic hemorrhage, including microbleeds.”

The ability to uncover intracerebral hemorrhage is crucial because such hemorrhage is a contraindication to thrombolytic therapy in stroke patients, pointed out Peter D. Schellinger, MD, PhD, Assistant Professor at the Neurology Clinic at the University of Heidelberg in Germany. Dr. Schellinger was the lead investigator in an MRI trial involving 62 ischemic stroke patients and 62 patients with intracerebral hemorrhage.

“Hyperacute intracerebral hemorrhage has a characteristic stroke MRI pattern and is detectable with excellent accuracy even from readers with limited experience,” Dr. Schellinger and his colleagues reported. They asserted that CT is no longer necessary to rule out intracerebral hemorrhage before ischemic stroke therapy and that MRI should be considered the imaging method of choice in patients with acute stroke.

RESULTS OF THE HEME STUDY

There has been growing interest in using multimodal MRI for acute stroke evaluation because it provides detailed information about ischemia, tissue injury, and perfusion and vessel status. The main impediment to this MRI application has been uncertainty about its ability to detect acute hemorrhage. That is why acute stroke evaluation currently calls for CT first and MRI second, Dr. Kidwell explained.

In the HEME (Hemorrhage Early MRI Evaluation) study, she and her colleagues reversed the usual order of the tests and tried to limit the interval between tests to 30 minutes to ensure that MRI did not receive an advantage over CT in detecting early signs of evolving hemorrhage. MRI testing included echo-planar diffusion and susceptibility-weighted imaging, as well as standard gradient echo imaging.

One of two neurologists blindly analyzed each patient’s imaging tests, specifically looking for acute hemorrhage, acute ischemic stroke, and chronic hemorrhage (including microbleeds). The patients in the study averaged age 74 and had a median score of 7 on the National Institutes of Health Stroke Scale (NIHSS). The median times to MRI and CT from symptom onset were 2.7 and 3.5 hours, respectively. Overall, imaging tests led to the diagnosis of acute or chronic hemorrhage in 60 patients, acute hemorrhage in 27 patients, and ischemic stroke in 110 patients.

The sensitivity and overall accuracy of MRI for acute hemorrhage were 96% and 99%, respectively, versus 89% and 98% for CT. For detecting acute or chronic hemorrhage, MRI was just as effective, while CT’s sensitivity and accuracy fell to 40% and 78%, respectively. As expected, MRI was far better than CT for diagnosing ischemic stroke; the latter had a sensitivity of only 20% and an overall accuracy of 60% in that setting. “Of the 26 patients with MRI-identified [acute] hemorrhage, 23 were seen on CT and three were not detected at all [by CT],” remarked Dr. Kidwell. One CT that initially showed an acute hemorrhage was actually false-positive due to meningioma, she noted.

The HEME study provides sufficient evidence of MRI’s accuracy and reliability in detecting acute hemorrhage, Dr. Kidwell stated. “Therefore, we suggest that MRI may be employed as the first and sole imaging study for the evaluation of acute stroke patients,” she proposed.

A STANDARDIZED MRI PROTOCOL