Comparison of left ventriculography and coronary arteriography with positron emission tomography in assessment of myocardial viability

Background: Assessment of viability of myocardium after an ischemic insult is an important clinical question that affects decisions pertaining to potential revasculatization. The results of contrast left ventriculograms and coronary angiography were compared to positron emission tomography (PET) in 64 patients with coronary artery disease and reduced left ventricular function. Hypothesis: The study was undertaken to determine the relative utility of the invasive studies in the assessment of viability. Methods: Right anterior oblique ventriculograms were assessed for hypokinesis, akinesis, or dyskinesis in six segments. The PET scans were assessed for viability by visual estimation of flourodeoxyglucose (FDG) uptake in six segments that corresponded to the segments analyzed on the ventriculograms. Results: Of a total of 373 segments successfully analyzed by PET, 272 were judged to be viable (normal or hypokinetic) by contrast ventriculography. Of these, 253 (93%) were considered viable by PET. Of 177 segments deemed either normal or mild-to-moderately hypokinetic by ventriculography, 170 (94%) were viable by PET. Of 95 severely hypokinetic segments, 83 (84%) were viable by PET. Of 79 akinetic segments, 44 (56%) were considered viable by PET. For segments that were dyskinetic and thought to be nonviable by ventriculography, 19 of 22 (86%) were also considered nonviable by PET. For 294 segments for which a determination on viability was made based on the presence of wall motion on the ventriculogram (normal, hypokinetic, or dyskinetic; not akinetic), there was excellent agreement with PET (93%; p < 0.001). In 49 patients there was akinesis in no more than one segment in either the anterior or inferior territories, indicating the potential for assessment of viability by ventriculography in at least two of three segments in each territory. Coronary anatomy was analyzed to assess whether coronary patency could help in assessing viability. Segments supplied by patent arteries were more likely to be viable by PET than segments supplied by occluded arteries (p < 0.001). Akinetic segments were more likely to be supplied by occluded arteries (56 vs. 23, 72%). Dyskinetic segments were predominantly nonviable by PET (86%) and were usually supplied by occluded arteries (77%). Conclusion: In patients in whom the assessment of viability is clinically relevant, the presence of systolic inward motion on the contrast left ventriculogram correlates well with segment viability by PET, while outward or dyskinetic movement correlates well with nonviability. Thus, the use of PET to assess viability in many patients may be unnecessary.