Background: We sought to determine the diagnostic performance of a recently developed combined supine-prone quantification algorithm for myocardial perfusion single photon emission computed tomography (MPS) for the detection of coronary artery disease (CAD) in women.
Methods and results: Consecutive MPS scans of women without known CAD and coronary angiography within 3 months of MPS (n = 168) and with a low likelihood of CAD (n = 291) were considered. Total perfusion deficit (TPD) was automatically derived for supine (S-TPD), prone (P-TPD), and combined prone-supine (C-TPD) data sets. The low-likelihood patients were grouped by bra cup size (A/B, n = 102; C, n = 101; and D, n = 88). The areas under the receiver operator characteristic curves for S-TPD, P-TPD, and C-TPD were 0.84 +/- 0.03, 0.88 +/- 0.03, and 0.90 +/- 0.03, respectively. C-TPD had a higher specificity than S-TPD and P-TPD for identification of CAD (stenosis > or =70%) without compromising sensitivity (61%, 76%, and 94% for S-, P-, and C-TPD, respectively; P < .0005 vs S-TPD and P < .05 vs P-TPD). Normalcy rates were higher for C-TPD than for S-TPD or P-TPD.
Conclusions: Combined supine-prone quantitative MPS in women yields significantly increased specificity and normalcy rates without compromising sensitivity for the detection of CAD compared with standard analysis.