Recent evidence suggests that the diagnostic accuracy of myocardial perfusion imaging is improved by quantifying stress myocardial blood flow (MBF) in absolute terms. We evaluated a comprehensive quantitative 13N-ammonia positron emission tomography (13NH3-PET) diagnostic panel, including stress MBF, coronary flow reserve (CFR), and relative flow reserve (RFR) in conjunction with relative perfusion defect (PD) assessments to better detect functionally significant coronary artery stenosis.Methods and Results—
A total of 130 patients (307 vessels) with coronary artery disease underwent both 13NH3-PET and invasive coronary angiography with fractional flow reserve (FFR) measurement. Diagnostic accuracy, optimal cut points, and discrimination indices of respective 13NH3-PET quantitative measures were compared, with FFR as standard reference. The capacity to discern disease with stepwise addition of stress MBF, CFR, and RFR to qualitatively assessed relative PD was also gauged, using the category-free net reclassification index. All quantitative measures showed significant correlation with FFR (PET-derived CFR, r=0.388; stress MBF, r=0.496; and RFR, r=0.780; all P<0.001). Optimal respective cut points for FFR ≤0.8 and ≤0.75 were 1.99 and 1.84 mL/min per g for stress MBF and 2.12 and 2.00 for PET-derived CFR. Discrimination indices of quantitative measures that correlated with FFR ≤0.8 were all significantly higher than that of relative PD (area under the curve: 0.626, 0.730, 0.806, and 0.897 for relative PD, CFR, stress MBF, and RFR, respectively; overall comparison P<0.001). The capacity for functionally significant coronary stenosis was incrementally improved by the successive addition of CFR (net reclassification index=0.629), stress MBF (net reclassification index=0.950), and RFR (net reclassification index=1.253; all P<0.001) to relative PD.Conclusions—
Integrating quantitative 13NH3-PET measures with qualitative myocardial perfusion assessment provides superior diagnostic accuracy and improves the capacity to detect functionally significant coronary artery stenosis.Clinical Trial Registration—
URL: http://www.clinicaltrials.gov. Unique identifiers: NCT01621438 and NCT01366404.