PurposeQuantification of myocardial oxygenation (MO) in heart failure (HF) has been less than satisfactory. This has necessitated the use of invasive techniques to measure MO directly or to determine the oxygen demand during exercise using the cardiopulmonary exercise (CPX) test. We propose a new quantification method for MO using blood-oxygen-level-dependent (BOLD) myocardial T2* magnetic resonance imaging (M-T2* MRI), and investigate its correlation with CPX results. MethodsThirty patients with refractory HF who underwent cardiac MRI and CPX test for heart transplantation, and 24 healthy, age-matched volunteers as controls were enrolled. M-T2* imaging was performed using a 3-Tesla and multi-echo gradient-echo sequence. M-T2* was calculated by fitting the signal intensity data for the mid-left ventricular septum to a decay curve. M-T2* was measured under room-air (T2*-air) and after inhalation of oxygen for 10min at a flow rate of 10L/min (T2*-oxy). MO was defined as the difference between the two values (ΔT2*). Changes in M-T2* at the two conditions and ΔT2* between the two groups were compared. Correlation between ΔT2* and CPX results was analyzed using the Pearson coefficient. ResultsT2*-oxy was significantly greater than T2*-air in patients with HF (29.9±7.3ms vs. 26.7±6.0ms, p<0.001), whereas no such difference was observed in controls (25.5±4.0ms vs. 25.4±4.4ms). ΔT2* was significantly greater for patients with HF than for controls (3.2±4.5ms vs. -0.1±1.3ms, p<0.001). A significant correlation between ΔT2* and CPX results (peak VO2, r=−0.46, p<0.05; O2 pulse, r=−0.54, p<0.005) was observed. ConclusionΔT2* is increased T2*-oxy is greater in patients with HF, and is correlated with oxygen metabolism during exercise as measured by the CPX test. Hence, ΔT2* can be used as a surrogate marker of MO instead of CPX test.