Dynamic 3D analysis of myocardial sympathetic innervation: an experimental study using 123I-MIBG and a CZT camera.

Abstract

Data on the in vivo myocardial kinetics of (123)I-metaiodobenzylguanidine ((123)I-MIBG) are scarce and have always been obtained using planar acquisitions. To clarify the normal kinetics of (123)I-MIBG in vivo over time, we designed an experimental protocol using a 3-dimensional (3D) dynamic approach with a cadmium zinc telluride (CZT) camera. We studied 6 anesthetized pigs (mean body weight, 37 ± 4 kg). Left ventricular myocardial perfusion and sympathetic innervation were assessed using (99m)Tc-tetrofosmin (26 ± 6 MBq), (123)I-MIBG (54 ± 14 MBq), and a CZT camera. A normal perfusion/function match on gated SPECT was the inclusion criterion. A dynamic acquisition in list mode started simultaneously with the bolus injection of (123)I-MIBG, and data were collected every 5 min for the first 20 min and then at acquisition steps of 30, 60, 90, and 120 min. Each step was reconstructed using dedicate software and reframed (60 s/frame). On the reconstructed transaxial slice that best showed the left ventricular cavity, regions of interest were drawn to obtain myocardial and blood pool activities. Myocardial time-activity curves were generated by interpolating data between contiguous acquisition steps, corrected for radiotracer decay and injected dose, and fitted to a bicompartmental model. Time to myocardial maximum signal intensity (MSI), MSI value, radiotracer retention index (RI, myocardial activity/blood pool integral), and washout rate were calculated. The mediastinal signal was measured and fitted to a linear model. The myocardial MSI of (123)I-MIBG was reached within 5.57 ± 4.23 min (range, 2-12 min). The mean MSI was 0.426% ± 0.092%. Myocardial RI decreased over time and reached point zero at 176 ± 31 min (range, 140-229 min). The ratio between myocardial and mediastinal signal at 15 and 125 min and extrapolated at 176 and 4 h was 5.45% ± 0.61%, 4.33% ± 1.23% (not statistically significant vs. 15 min), 3.95% ± 1.46% (P < 0.03 vs. 125 min), and 3.63% ± 1.64% (P < 0.03 vs. 176 min), respectively. Mean global washout rate at 125 min was 15% ± 14% (range, 0%-34%), and extrapolated data at 176 min and 4 h were 18% ± 18% (range, 0.49%-45%) and 25% ± 23% (range, 1.7%-56.2%; not statistically significant vs. 176 min), respectively. 3D dynamic analysis of (123)I-MIBG suggests that myocardial peak uptake is reached more quickly than previously described. Myocardial RI decreases over time and, on average, is null about 3 h after injection. The combination of an early peak and variations in delayed myocardial uptake could result in a wide physiologic range of washout rates. Mediastinal activity appears to be constant over time and significantly lower than previously described in planar studies, resulting in a higher heart-to-mediastinum ratio.