Comparison of low-mechanical index pulse sequence schemes for detecting myocardial perfusion abnormalities during vasodilator stress echocardiography

Abstract

Myocardial contrast echocardiography has the potential to accurately detect functionally significant coronary artery disease during pharmacologic stress testing. Different low-mechanical index modalities, including triggering replenishment imaging (TRI) and real-time imaging (RTI), are currently used to identify myocardial perfusion defects. We compared the ability of TRI with that of RTI for detecting and localizing perfusion abnormalities. Thirty-six patients (62 +/- 14 years old, 15 men) underwent single-photon emission computed tomography (SPECT) with technetium-99m sestamibi and myocardial contrast echocardiography at baseline and after infusion of 0.56 mg/kg of dipyridamole. Sixteen of these patients also underwent quantitative angiography. Contrast-enhanced images were obtained in 4-, 3-, and 2-chamber views after intravenous bolus injections of lipid-encapsulated microbubbles (0.1 ml of Definity). A myocardial perfusion defect was defined by myocardial contrast echocardiography as a delay of >2 seconds in contrast replenishment after high-mechanical index flash impulse. The myocardial segments were divided into 3 major coronary territories. There was agreement in detecting perfusion defects between SPECT and TRI in 26 patients (72%, kappa = 0.46) and between SPECT and RTI in 27 patients (75%, kappa = 0.50). Agreements between myocardial contrast echocardiography and SPECT for localizing coronary territories with perfusion defects were 81% for TRI (kappa = 0.43) and 85% for RTI (kappa = 0.61). Accuracy of RTI for detecting >50% diameter stenoses by quantitative angiography was 79%, that of TRI was 71%, and that of SPECT was 65%. These data indicate that the different low-mechanical index imaging schemes are equivalent to radionuclide SPECT in accurately detecting diseased coronary artery territories during vasodilator stress.