SIR: MRI and Echocardiography Team Up to Find Stroke Cause

SEATTLE, March 7 -- Cardiac MRI used in tandem with echocardiography can enhance the likelihood of finding the cause of cardioembolic stroke, according to a small retrospective study.

Among the 93 patients with suspected cardioembolic stroke, cardiac MRI yielded twice as many findings as echocardiography, said John Sheehan, M.D., of Northwestern University and Northwestern Memorial Hospital in Chicago.

Echocardiography was more effective at detecting shunts and bacterial growths on prosthetic valves whereas MRI found far more clots, he reported at the Society of Interventional Radiology meeting.

Overall, MRI added to the cardiac findings for 31% of patients and to the noncardiac findings for 20% while echocardiography added significant cardiac findings for 26%.

Cardioembolic strokes are usually evaluated with echocardiography. This type of stroke, which carries a high risk of early recurrence, accounts for up to one in six ischemic strokes.

Improved diagnostic capability with the combination of modalities may point which patients should get anticoagulation or other secondary prevention treatment, Constantino S. Pea, of the Baptist Cardiac & Vascular Institute in Miami, in his invited commentary at the session.

"We know we can affect recurrence," he said.

The retrospective study included 93 consecutive patients (52% male, average age 65) with suspected cardioembolic stroke in a cardiac imaging database for a one year period (2005 to 2006). All patients had echocardiography (89% transthoracic, 41% transesophageal, and 32% both) and cardiac MRI.

Among the findings, the researchers reported:

• MRI found thrombus in nine patients whereas echocardiography positively identified only two for a false negative rate of 56%,
• Echocardiography was positive for valvular growths in three patients (two had metal prosthetic valves) whereas MRI found none for a false negative rate of 100%,
• MRI had 60 clinically significant cardiac and noncardiac findings excluding thrombi and valvular growths (one shunt) compared to 30 with echocardiography (15 shunts), and
• MRI showed that 20% of patients had cardiac disease associated with thrombus formation, such as acute infection or scarring, whereas echocardiography found that only 7% did.

Combining the two modalities improved detection and treatment of embolic source, which is certain to improve secondary prevention of subsequent strokes in this high risk group, Dr. Sheehan said.

"We have no doubt that it will have therapeutic impact," he said.

For example, identifying a clot in the heart indicates that part likely fragmented off and traveled to the brain to cause the stroke, so blood thinners may be used to keep the clot from breaking up further, he said.

His research group is now looking at the effect of imaging on management outcomes in this patient group.

However, he cautioned that the study was limited in that it was retrospective, may have been prone to selection bias because all patients had suspected cardioembolic stroke, and there was no external reference standard.

Also, cardiac MRI is an expensive modality and requires expertise in reading the images, he added.

Further study will be needed to define the role for MRI in this setting, Dr. Pea said.

He concluded, "Patients with previous TIA [transient ischemic attack] or stroke, especially those without a pre-existing indication or contraindication to anticoagulation, may benefit."