Myocardial perfusion imaging by 15O-H2O positron emission tomography predicts clinical revascularization procedures in symptomatic patients with previous coronary artery bypass graft

Abstract

Abstract Aims We wanted to assess if 15O-H2O myocardial perfusion imaging (MPI) in a clinical setting can predict referral to coronary artery catheterization [coronary angiography (CAG)], execution of percutaneous coronary intervention (PCI), and post-PCI angina relief for patients with angina and previous coronary artery bypass graft (CABG). Methods and results We analysed 172 symptomatic CABG patients referred for 15O-H2O positron emission tomography (PET) MPI at Aarhus University Hospital Department of Nuclear Medicine & PET Centre, of which five did not complete the scan. In total, 145 (87%) enrolled patients had an abnormal MPI. Of these, 86/145 (59%) underwent CAG within 3 months; however, no PET parameters predicted referral to CAG. During the CAG, 25/86 (29%) patients were revascularized by PCI. Relative flow reserve (RFR) (0.49 vs. 0.54 P = 0.03), vessel-specific myocardial blood flow (MBF) (1.53 vs. 1.88 mL/g/min, P < 0.01), and vessel-specific myocardial flow reserve (MFR) (1.73 vs. 2.13, P < 0.01) were significantly lower in patients revascularized by PCI. Receiver operating characteristic analysis of the vessel-specific parameters yielded optimal cutoffs of 1.36 mL/g/min (MBF) and 1.28 (MFR) to predict PCI. Angina relief was experienced by 18/24 (75%) of the patients who underwent PCI. Myocardial blood flow was an excellent predictor of angina relief on both a global [area under the curve (AUC) = 0.85, P < 0.01] and vessel-specific (AUC = 0.90, P < 0.01) level with optimal cutoff levels of 1.99 mL/g/min and 1.85 mL/g/min, respectively. Conclusion For CABG patients, RFR, vessel-specific MBF, and vessel-specific MFR measured by 15O-H2O PET MPI predict whether subsequent CAG will result in PCI. Additionally, global and vessel-specific MBF values predict post-PCI angina relief.