Parametric quantification of myocardial ischaemia using real-time perfusion adenosine stress echocardiography images, with SPECT as reference method.

Summary Background: Real-time perfusion (RTP) adenosine stress echocardiography (ASE) can be used to visually evaluate myocardial ischaemia. The RTP power modulation technique, provides images for off-line parametric perfusion quantification using Qontrast((R)) software. From replenishment curves, this generates parametric images of peak signal intensity (A), myocardial blood flow velocity (beta) and myocardial blood flow (Axbeta) at rest and stress. This may be a tool for objective myocardial ischaemia evaluation. We assessed myocardial ischaemia by RTP-ASE Qontrast((R))-generated images, using 99mTc-tetrofosmin single-photon emission computed tomography (SPECT) as reference. Methods: Sixty-seven patients admitted to SPECT underwent RTP-ASE (SONOS 5500) during Sonovue((R)) infusion, before and throughout adenosine stress, also used for SPECT. Quantitative off-line analyses of myocardial perfusion by RTP-ASE Qontrast((R))-generated A, beta and Axbeta images, at different time points during rest and stress, were blindly compared to SPECT. Results: We analysed 201 coronary territories [corresponding to the left anterior descendent (LAD), left circumflex (LCx) and right coronary (RCA) arteries] from 67 patients. SPECT showed ischaemia in 18 patients. Receiver operator characteristics and kappa values showed that A, beta and Axbeta image interpretation significantly identified ischaemia in all territories (area under the curve 0.66-0.80, P = 0.001-0.05). Combined A, beta and Axbeta image interpretation gave the best results and the closest agreement was seen in the LAD territory: 89% accuracy; kappa 0.63; P<0.001. Conclusion: Myocardial isachemia can be evaluated in the LAD territory using RTP-ASE Qontrast((R))-generated images, especially by combined A, beta and Axbeta image interpretation. However, the technique needs improvements regarding the LCx and RCA territories. (Less)