913-120 New Evidence for the Ability of Gated Blood Pool Scanning to Calculate Volumes in Normal and Myopathic Left Ventricles

Abstract

Gated blood pool scanning (GBPS) with Tc-99m labeled red blood cells has been proposed as a method for determining left ventricular volumes (LVV) after correction for soft tissue attenuation; this has previously been the subject of controversy. Accurate determination of the attenuation correction factor depends upon estimating the distance (D) from the center of the LV to the chest wall. The purpose of this study was to compare LVVs measured by GBPS with an independent noninvasive measure not based on counts or the geometric assumptions used in angiography, comparing three different methods for determining D and then LVV. We studied 10 consecutive patients referred for GBPS (7/10 male, LVEF 20–72%) The standard nuclear method divided the distance from the center of LV activity on GBPS to a chest wall marker on an anterior view by the sine of the LAO image angle. The second method averaged the distance from the chest wall marker to the LV apex and the LV outflow tract on an anterior view, and divided that average by the sine of the LAO image angle. 2D echo determined D by bisecting a line between apex and mid-mitral valve and measuring D to the chest wall. The attenuation factor for Tc-99m was calculated by e (-0.15 * D) Volumes were calculated after adjusting for background counts and peripheral blood activity (counts/ccl. The reference volumes were determined by ECG gated 3D LV echo (3DE) reconstructions from rotated apical 2D echo views. All 3 methods agreed well with 3DE volumes: Method for D GBPS vs 3DE r SEE, ml Mean Error Standard y = 0.85x + 6.7 0.89 22.0 8.1 ± 21.6 (NSvsO) Average y = 0.92x + 10.7 094 15.6 -2.4 ± 15.7 (NSvsO) 2DEcho y = 1.0x + 1.6 0.93 19.9 -2.0 ± 19.3 (NSvsO) The mean errors of the averaging and 2DEcho methods were not significantly different (p > 0.85), and the methods correlated well (slope = 0.99, r = 0.97). The standard D method had a larger mean error (p < 0.01) but no different variance (F = 2, p > 0.15). These results provide further support for the ability of GBPS to measure LVV with confidence. Two geometric methods for calculating soft tisue attenuation D are roughly equivalent to a direct 2D echo method method. The radionuclide averaging method to obtain D compared favorably, and relies upon easily demarcated landmarks, yielding smaller mean errors.