Computed tomography myocardial perfusion vs 15O-water positron emission tomography and fractional flow reserve

Abstract

Objectives: Computed tomography (CT) can perform comprehensive cardiac imaging. We compared CT coronary angiography (CTCA) and CT myocardial perfusion (CTP) with 15O-water positron emission tomography (PET) and invasive coronary angiography (ICA) with fractional flow reserve (FFR). Methods: 51 patients (63 (61–65) years, 80 % male) with known/suspected coronary artery disease (CAD) underwent 320-multidetector CTCA followed by “snapshot” adenosine stress CTP. Of these 22 underwent PET and 47 ICA/FFR. Obstructive CAD was defined as CTCA stenosis >50 % and CTP hypoperfusion, ICA stenosis >70 % or FFR <0.80. Results: PET hyperaemic myocardial blood flow (MBF) was lower in obstructive than non-obstructive territories defined by ICA/FFR (1.76 (1.32–2.20) vs 3.11 (2.44–3.79) mL/(g/min), P < 0.001) and CTCA/CTP (1.76 (1.32–2.20) vs 3.12 (2.44–3.79) mL/(g/min), P < 0.001). Baseline and hyperaemic CT attenuation density was lower in obstructive than non-obstructive territories (73 (71–76) vs 86 (84–88) HU, P < 0.001 and 101 (96–106) vs 111 (107–114) HU, P 0.001). PET hyperaemic MBF corrected for rate pressure product correlated with CT attenuation density (r = 0.579, P < 0.001). There was excellent per-patient sensitivity (96 %), specificity (85 %), negative predictive value (90 %) and positive predictive value (94 %) for CTCA/CTP vs ICA/FFR. Conclusion: CT myocardial attenuation density correlates with 15O-water PET MBF. CTCA and CTP can accurately identify obstructive CAD. Key Points: •CT myocardial perfusion can aid the assessment of suspected coronary artery disease. • CT attenuation density from “snapshot” imaging is a marker of myocardial perfusion. • CT myocardial attenuation density correlates with15O-water PET myocardial blood flow. • CT attenuation density is lower in obstructive territories defined by invasive angiography. • Diagnostic accuracy of CTCA+CTP is comparable to invasive angiography + fractional flow reserve.