Evaluation of potassium-43 scintillation images during early myocardial ischemia in an animal model.

To assess the validity of myocardial imaging with potassium-43 (43K) early after the onset of ischemia, the left anterior descending artery was occluded with a baloon tip catheter in 32 intact anesthetized dogs. 99mTechnetium ventriculograms localized the left ventricle. 43K was administered intravenously and serial images were obtained in four views using an Anger camera with a pinhole collimator. The heart was arrested after 60 minutes and removed for imaging and tissue counts to ascertain extracardiac and geometric factors. In normals (group 1) left ventricular images were relatively homogeneous, except for the thin walled apex, both in vivo and in the isolated heart. Equilibration with 43K prior to ischemia (group 2) gave similar images to group 1, associated with a small reduction in tissue count after one hour of ischemia. Group 3 was infused with 43K after initiation of ischemia. Despite a reduction of 43K counts in the ischemic area to less than one-fourth of the nonischemic site (P less than 0.001), demonstration of a "cold area" in vivo was inconstant, occurring in only 34% of studies. Lead shielding did not improve accuracy. In the isolated heart the ability to detect the cold area was improved to 73%. However, when the left ventricle was incised and spread flat, so that low and high activity areas were contiguous rather than superimposed, a widespread area of ischemia was present without exception in the anterior wall. Use of a rectilinear scanner in seven animals failed to improve diagnostic yield; areas of reduced radioactivity were seen at the apex in normals by both techniques. Thus, while detection of low flow areas in the isolated heart is feasible by isotopic imaging early after the onset of ischemia, both extracardiac and geometric factors can contribute to qualitative and quantitative errors in vivo.