Usefulness of ~(99m)Tc-MIBI Stress Myocardial SPECT Bull's-Eye Quantification in Coronary Artery Disease

摘要

~(99m)Tc-methoxy-isobutyl isonitrile (MIBI) myocardial SPECT quantification performed using a Bull's-eye polar map, was evaluated and compared with visual analysis in 120 patients with proven or suspected CAD. The study series comprised 106 men and 14 women, age 37-75 years (mean 51±6), 68 of whom had had a prior myocardial infarction. Coronary angiography was taken as the gold standard: one-vessel disease was present in 24 patients, two-vessel disease in 39, and three-vessel disease in 44, whereas no significant stenosis was documented in 13 cases. Forty age-matched subjects (26 men, 14 women), with less than a 5% chance of having CAD, were enrolled to establish the normal database for males and females. ROC analysis was used to calculate the optimal thresholds for the definition of the disease extension in each vascular territory of the Bull's-eye polar map: 10% for LAD, 8% for LCX, and 20% for RCA territory. The sensitivity/specificity ratio of the scintigraphy was: 75/82% with the visual and 78/74% with the quantitative analysis for LAD; 60/90% with visual and 72/64% with quantitative analysis for LCX; and 78/66% with visual and 70/62% with quantitative analysis for RCA territory. The sensitivity/specificity ratios for the CAD diagnosis were similar with the visual and the Bull's-eye analysis in 92/61% and 93/61% respectively. Bull's-eye analysis agreed with visual analysis in 296/360 vessels. Two and three-vessel disease were most frequently observed using the Bull's-eye approach. However, a greater number of false positive (FP) cases were found with Bull's-eye than with visual analysis (28 versus 3 cases): FP cases were detected principally (21/28) in patients with previous myocardial infarction, in whom the corresponding hypoperfused area involved an adjacent vascular territory. It is interesting to note that this phenomenon, commonly observed in the LCX or RCA territory, was almost always correctly interpreted as a FP case by visual analysis. In conclusion, we found the sensitivity and specificity for CAD diagnosis with the Bull's-eye approach to be similar to that of visual analysis, but the former method seems to be somewhat more sensitive for the localization of LAD and LCX disease. However, particular attention should be paid to patients with previous myocardial infarction, as FP cases are not infrequently observed with quantitative analysis, especially in the LCX and RCA territories.