Comparison of 64-slice CT with gated SPECT for evaluation of left ventricular function.

UNLABELLED Precise and reliable assessment of left ventricular (LV) function and dimensions is prognostically important in cardiac patients. As the integration of SPECT and multislice CT into hybrid scanners will promote the combined use of both techniques in the same patient, a comparison of the 2 methods is pertinent. We aimed at comparing LV dimensions, muscle mass, and function obtained by electrocardiographically gated 64-slice CT versus gated-SPECT. METHODS Sixty patients (mean age, 64 +/- 8 y) referred for evaluation of coronary artery disease underwent 99mTc-tetrofosmin gated SPECT and 64-slice CT within 4 +/- 2 d. LV ejection fraction (LVEF), end-systolic volume (ESV), and end-diastolic volume (EDV) from CT were compared with SPECT. Additionally, LV muscle mass and quantitative regional wall motion were assessed in 20 patients with both methods. RESULTS CT was in good agreement with SPECT for quantification of LVEF (r = 0.825), EDV (r = 0.898), and ESV (r = 0.956; all P < 0.0001). LVEF was 59% +/- 13% measured by SPECT and slightly higher but not significantly different by CT (60% +/- 12%; mean difference compared with SPECT, 1.1% +/- 1.7%; P = not significant). A systematic overestimation using CT for EDV (147 +/- 60 mL vs. 113 +/- 52 mL; mean difference, 33.5 +/- 23.1 mL) and ESV (63 +/- 55 mL vs. 53 +/- 49 mL; mean difference, 9.3 +/- 15.9 mL; P < 0.0001) was found compared with SPECT. A good correlation for muscle mass was found between the 2 methods (r = 0.868; P < 0.005). However, muscle mass calculated by SPECT was significantly lower compared with CT (127 +/- 24 g vs. 148 +/- 37 g; mean difference, 23.0 +/- 12.2 g; P < 0.001). The correlation for regional wall motion between the 2 methods was moderate (r = 0.648; P < 0.0001). CONCLUSION LVEF and LV functional parameters as determined by 64-slice CT agree over a wide range of clinically relevant values with gated SPECT. However, interchangeable use of the 2 techniques should be avoided for LV volumes, muscle mass, and regional wall motion because of variances inherent to the different techniques.