Abstract
1728
Objectives To evaluate non-gated CT scans obtained with SPECT/CT MPI ability to depict coronary artery calcium (CAC) by comparison with Agatston scores (AS) from gated CT, and to compare AS with the prevalence of MPI abnormalities.
Methods 6-slice, non-gated SPECT/CT of 136 pts were analyzed by 3 blinded, NM Physicians to generate a clinically relevant, visual coronary calcium classification (vCCC). Class 0 is defined as no evidence of CAC (AS < 0-20), Class 1 as mild CAC in 1 or 2 vessels (AS 20-100), Class 2 as moderate CAC (AS > 101-399) and Class 3 as severe CAC (AS >400). Star artifact along the CA were also considered as CAC. The sensitivity, specificity, PPV and NPV and accuracy for predicting vCCC were calculated for each reviewer. Patients with stents, CABG and pacemakers were excluded.
Results Of 136 pts, 69 had 0-20 AS, 19 had 21-100 AS, 30 had 101-399 AS and 18 had AS >400. Table 1 summarizes reviewers ability to detect CAC. The average reviewer accuracy per class was 91% for Class 0, 63% for Class 1, 60% for Class 2 and 83% for Class 3. In the AS 0-20 group, 13/69 (19%) had abnormal MPI, 12 ischemia and 1 scar. In the AS 21-100 group, 4/19 (21%) had MPI abnormalities, 2 ischemia and 2 scar. In the AS 101-399 group, 5/30 (17%) had MPI abnormalities, 4 ischemic and 1 scar. In the AS >400 group, 6/18 (33%) had perfusion abnormalities, 4 scar and 2 ischemia.
Conclusions Class 0 vCCC obtained from a non-gated CT, excluded CAC with 91% accuracy. vCCC less reliably estimates AS. vCCC directly correlates with the extent of coronary calcium and the prevalence of MPI abnormalities. By estimating CAC, vCCC may improve the specificity of MPI interpretation and help identify pts requiring aggressive risk factor management