Detection and sizing of myocardial ischemia and infarction by nuclear magnetic resonance imaging in the canine heart.

The usefulness of NMR imaging to size infarcted and hypoperfused, ischemic myocardium was assessed in 16 dogs which underwent coronary artery occlusion and reperfusion. During occlusion, technetium-99 microspheres were injected into the left atrium. Following death, the hearts were excised and underwent NMR imaging with a 0.35 tesla magnet, using multiple spin-echo pulse sequences. The epicardium of the heart was marked to indicate the level of the NMR cross-sectional tomographic image. The heart was subsequently breadloafed into 5 mm sections and the corresponding NMR cross-section was flagged for analysis. Autoradiography was performed to measure the hypoperfused, at-risk zone, and triphenyltetrazolium chloride staining was used to measure infarct size. For the flagged tomographic slice, the size of the NMR abnormality correlated well (r = 0.95), and was comparable to the actual hypoperfused, at-risk zone of the left ventricle. However, NMR estimates of infarct size correlated less well (r = 0.75) with the pathologic measure, and significantly overestimated actual infarct size (p less than 0.005). The T1 and T2 values were consistently increased (p less than 0.0005) in both the hypoperfused and infarct zones, compared to normal myocardium. We conclude that NMR imaging can detect acute myocardial ischemia and infarction, but overestimates infarct size and corresponds better to the area of hypoperfused, ischemic myocardium. In this excised canine heart occlusion-reperfusion model, the NMR abnormality corresponded best to the area including both infarction and the surrounding ischemic region.