Mortality Prediction by Quantitative PET Perfusion Expressed as Coronary Flow Capacity With and Without Revascularization

Objectives: This study sought to determine the relationship between the severity of reduced quantitative perfusion parameters and mortality with and without revascularization. Background: The physiological mechanisms for differential mortality risk of coronary flow reserve (CFR) and coronary flow capacity (CFC) before and after revascularization are unknown. Methods: Global and regional rest-stress (ml/min/g), CFR, their regional per-pixel combination as CFC, and relative stress in ml/min/g were measured as percent of LV in all serial routine 5,274 diagnostic PET scans with systematic follow-up over 10 years (mean 4.2 ± 2.5 years) for all-cause mortality with and without revascularization. Results: Severely reduced CFR of 1.0 to 1.5 and stress perfusion ≤1.0 cc/min/g incurred increasing size-dependent risks that were additive because regional severely reduced CFC (CFCsevere) was associated with the highest major adverse cardiac event rate of 80% (p < 0.0001 vs. either alone) and a mortality risk of 14% (vs. 2.3% for no CFCsevere; p = 0.001). Small regions of CFCsevere ≤0.5% predicted high risk (p < 0.0001 vs. no CFCsevere) related to a wave front of border zones at risk around the small most severe center. By receiver-operating characteristic analysis, relative stress topogram maps of stress (ml/min/g) as a fraction of LV defined these border zones at risk or for mildly reduced CFC (area under the curve [AUC]: 0.69) with a reduced relative tomographic subendocardial-to-subepicardial ratio. CFCsevere incurred the highest mortality risk that was reduced by revascularization (p = 0.005 vs. no revascularization) for artery-specific stenosis not defined by global CFR or stress perfusion alone. Conclusions: CFC is associated with the size-dependent highest mortality risk resulting from the additive risk of CFR and stress (ml/min/g) that is significantly reduced after revascularization, a finding not seen for global CFR. Small regions of CFCsevere ≤0.5% of LV also carry a high risk because of the surrounding border zones at risk defined by relative stress perfusion and a reduced relative subendocardial-to-subepicardial ratio.