Study shows MRI superiority over CT for diagnosis of ischemic stroke

January 27, 2007

A prospective trial published in the Jan. 27 issue of The Lancet has confirmed suspicions among neuroradiologists about the power of MRI to diagnose acute stroke, while finding noncontrast CT surprisingly insensitive to the potentially lethal disease.

A prospective trial published in the Jan. 27 issue of The Lancet has confirmed suspicions among neuroradiologists about the power of MRI to diagnose acute stroke, while finding noncontrast CT surprisingly insensitive to the potentially lethal disease.

The blinded study sponsored by the National Institute of Neurological Disorders and Stroke examined the diagnostic capabilities of the two modalities on 356 consecutive patients with stroke symptoms at Suburban Hospital, a community acute care facility in Bethesda, MD, from September 2000 to February 2002. Dr. Julio A. Chalela, then associated with NINDS, was principal investigator. He now practices at the Medical University of South Carolina. Dr. Steven Warach, section chief of stroke diagnostics and therapy, supervised.

The findings represented a vote of no confidence in CT, the consensus choice for initial diagnosis of acute stroke since the early 1990s.

"Our study shows that use of CT is no longer justifiable on the basis of diagnosis accuracy alone," Chalela wrote.

The report also concluded that MRI can be used as the sole modality for the emergency imaging of patients with suspected ischemic or hemorrhagic acute stroke.

The overall sensitivities of MRI and CT for detecting acute stroke were 83% and 26%, respectively.

MRI detected stroke in 164 of 356 patients, compared with 35 patients for CT. For cases evaluated within three hours of symptom onset, the period when tPA thrombolysis may be performed, MR identified the presence of ischemic stroke in 90 cases, compared with six for CT. The relative sensitivities for ischemic stroke during the three-hour window were 76% for MR and 27% for CT. The relative specificities were 96% for MR and 100% for CT.

The two modalities were equivalent for identifying acute intracranial hemorrhage, though neither modality detected all clinically confirmed cases. Four cases were misdiagnosed with MRI, and three false-negative cases were reported with CT.

The authors linked MRI's superiority to the ability of diffusion-weighted MR to detect ischemic injury earlier than CT. Acute ischemic stroke was diagnosed with MRI in 46% of patients but only 10% of patients with CT. Diffusion-weighted MRI was responsible for a 17% false-negative rate for ischemic stroke, however. Most of these cases were associated with small lesions located in the brain stem.

Chalela and colleagues recommended another head-to-head comparison to measure the effect of MRI on patient outcomes and healthcare expenditures.

For more information from the Diagnostic Imaging archives:

Perfusion CT in acute stroke patients

Diffusion-weighted MR findings new niche in stroke therapy

MRI spots patients prone to poststroke bleeds

Report from ASNR: CE-MRI identifies susceptibility to hemorrhage after stroke

The window expands for more effective stroke treatment