Background
Positron emission tomography (PET) enables robust and reproducible measurements of myocardial blood flow (MBF). However, the relatively limited resolution of PET till recently prohibited distinction between the subendocardial and the subepicardial layers in non-hypertrophied myocardium. Recent developments in hard- and software, however, have enabled to identify a transmural gradient difference in animal experiments. The aim of this study is to determine the feasibility of subendocardial and subepicardial MBF in normal human hearts assessed with ^15^O-labeled water PET.
Methods
Twenty-seven healthy subjects (mean age 41 ± 13 years; 11 men) were studied with ^15^O-labeled water PET to quantify resting and hyperaemic (adenosine) MBF at a subendocardial and subepicardial level. In addition, cardiac magnetic resonance imaging was performed to determine left ventricular (LV) volumes and function.
Results
Mean rest MBF was 1.46 ± 0.49 in the subendocardium, and 1.14 ± 0.342 mL · min^−1^ · g^−1^ in the subepicardium (P < .001). MBF during vasodilation was augmented to a greater extent at the subepicardial level (subendocardium vs subepicardium: 3.88 ± 0.86 vs 4.14 ± 0.88 mL · min^−1^ · g^−1^, P = .013). The endocardial-to-epicardial MBF ratio decreased significantly during hyperaemia (1.35 ± 0.23 to 1.12 ± 0.20, P < .001). Hyperaemic transmural MBF was inversely correlated with left ventricular end-diastolic volume index (LVEDVI) (r ^2^ = 0.41, P = .0003), with greater impact however at the subendocardial level.
Conclusions
^15^O-labeled water PET enables MBF measurements with distinction of the subendocardial and subepicardial layers in the normal human heart and correlates with LVEDVI. This PET technique may prove useful in evaluating patients with signs of ischaemia due to coronary artery disease or microvascular dysfunction.