Dipyridamole radionuclide ventriculography: a test with high specificity for severe coronary artery disease.

Ventricular dysfunction induced by dipyridamole would be evidence of myocardial ischemia in patients with limited ability to undergo standard exercise testing. Radionuclide ventriculography before and after intravenous dipyridamole infusion was compared with the results of exercise radionuclide ventriculography in a prospective study of 31 patients undergoing coronary angiography. Among these patients, 21 (68%) had significant coronary artery disease (greater than or equal to 50% stenosis), 19 (61%) had severe coronary disease (greater than or equal to 70% stenosis) and 10 (32%) were "normal" (less than 50% stenosis). The left ventricular ejection fraction was calculated, and regional wall motion was scored on a 6 unit scale. In the normal patients, the ejection fraction (+/- SEM) increased 5.6 +/- 2% (units) during exercise and 7.9 +/- 1 units after dipyridamole (both p less than or equal to 0.004 compared with that during rest). However, in patients with coronary artery disease, the ejection fraction failed to increase during exercise or after dipyridamole. In the patients with coronary artery disease, regional wall motion decreased by 4.1 +/- 0.5 units during exercise (p less than 0.003) and by 1.8 units after dipyridamole (p less than 0.02). Receiver operating characteristic analysis demonstrated general comparability between the sensitivity and specificity of exercise and dipyridamole ventriculography, with "optimal" operating points that favored choosing high sensitivity for the former and high specificity for the latter. Specific subsets of patients with severe coronary atherosclerosis were analyzed with use of these criteria. In patients with severe stenosis (greater than or equal to 70%), the sensitivity of dipyridamole ventriculography was 67% compared with 89% for exercise ventriculography. However, at these levels of sensitivity, the specificity of dipyridamole ventriculography was 92% compared with 67% for exercise ventriculography. In this and other subsets of patients, the specificity of dipyridamole ventriculography exceeded that of exercise ventriculography. Thus, it is concluded that dipyridamole radionuclide ventriculography is moderately sensitive and highly specific for detecting severe coronary atherosclerosis. This technique provides a widely applicable, useful alternative to exercise ventriculography in the diagnosis of coronary atherosclerosis in patients who have limited exercise tolerance.