Diagnostic accuracy of contrast echocardiography during adenosine stress for detection of abnormal myocardial perfusion: a prospective comparison with technetium-99 m sestamibi single-photon emission computed tomography.

Myocardial contrast echocardiography (MCE) utilizes compressible microbubbles behaving similarly to red blood cells. Destruction of microbubbles and observation of the gradual refill into the myocardium are key to evaluating perfusion using real-time MCE. We aimed to assess the feasibility and diagnostic accuracy of qualitative MCE utilizing a 17-segment model for localization of myocardial perfusion abnormalities compared with simultaneous technetium-99 m sestamibi single-photon emission computed tomography (SPECT). From July 2005 through August 2007, 97 patients with known or suspected coronary artery disease underwent simultaneous SPECT and realtime MCE during adenosine stress. Qualitative MCE and tracer uptake were analyzed visually using a 17-segment model in a blinded manner. Diagnostic accuracy and 95% confidence interval (CI) were determined. Myocardial contrast echocardiography was completed in 91 patients (age, mean [SD], 69.3 [10.9] years; body mass index, 30.0 [6.3]; 59 males [65%]). Myocardial contrast echocardiography analysis was feasible in 88 (97%) patients (261 of 264 [99%] territories; 1299 of 1497 [87%] segments). At patient level, MCE sensitivity was 88% (95% CI, 79%-94%); specificity was 85% (77%-90%). For disease detection in individual coronary territories, sensitivity and specificity were 84% (71%-92%) and 79% (72%-84%) for the left anterior descending artery; 62% (38%-80%) and 88% (83%-91%) for the left circumflex artery; and 73% (57%-82%) and 94% (89%-97%) for the right coronary artery. For MCE combined with wall-motion analysis, concordance with SPECT improved from 80% to 86%. Myocardial contrast echocardiography interobserver concordance was 81% (kappa [SE], 0.611 [0.78]). Myocardial contrast echocardiography accuracy was comparable in patients classified in accordance with presence of diabetes mellitus, myocardial infarction, hypertension, or percutaneous coronary intervention. Improved MCE specificity in detecting perfusion defects was seen in patients with no history of coronary bypass graft surgery (P = 0.005). Real-time MCE with a 17-segment model for analysis has good feasibility and accuracy in evaluation of myocardial perfusion during adenosine stress.