Determination of ventricular volumes in coronary artery disease. Comparison of two gated SPECT analysis tools with MRI

Aim: Comparison of two gated SPECT analysis tools - gated SPECT quantification (GSQ) and emory cardiac tool-box (ECT) - in patients with coronary artery disease (CAD) and severely impaired left ventricular function (preoperative: EF <35% by cardiac catheter). Patients, methods: A total of 56 gated SPECT examinations (one-day hybrid-protocol with 201Tl-chloride for rest and 99mTc-sestamibi for stress applied during low-dose dobutamine stress MR-examination; temporal resolution; 8 phases per cardiac cycle) were performed in 36 patients (31 preoperatively, 25 postoperatively) and compared with MRI in 48 cases. Left ventricular end-diastolic (LV-EDV) and end-systolic (LV-ESV) volumes as well as the left ventricular ejection fraction (LV-EF) were calculated. Results: The total volumetric assessment by both analysis algorithms (n = 56) showed good introclass correlation coefficients preoperatively (n 31), but even better postoperatively (n = 25). The mean reconstruction time was approximately 3 minutes (±2 SD) for GSQ and 15 minutes (±5 SD) for ECT. In comparison to MRI the results of both analysis tools also correlated well, but the agreement decreased in the presence of scared tissue. The mean LV-EF (MRI) preoperatively was 30.4%, in 6/36 patients above the values calculated from cardiac catheter, postoperatively 34.6%. Conclusion: Both gated SPECT analysis tools showed reliable volumetric assessments in high-risk patients with CAD and severely reduced LV-EF in comparison to MRI, with advantages for GSQ in terms of postprocessing time. However, for the calculation of LV-EF a markedly lower concordance with MR-results was observed for both methods depending on the presence of myocardial scars. (orig.)