Feasibility of exercise cardiac magnetic resonance imaging for the concurrent assessment of oxygen uptake and its Fick determinants

Background The concurrent assessment of oxygen uptake (VO2) and it’s Fick determinants (cardiac output x arterio-venous oxygen content difference [Qc x a-vO2diff]) provides key insights into the physiologic underpinnings of exercise intolerance across the spectrum of health and disease. However, the accurate assessment of the Fick-determinants is limited by the need for invasive measures of hemodynamics and arterial and venous O2 saturation and content (SaO2 and SvO2, and CaO2 and CvO2, respectively). Purpose To determine the feasibility of an exercise cardiac magnetic resonance imaging (ExCMR) approach that combines real-time free-breathing cardiac volumetric analysis with magnetic susceptometry-based oximetry to non-invasively quantify Qc, a-vO2diff, and VO2peak. Methods Bi-ventricular volumes and Qc were quantified from real-time, free-breathing steady-state free-precession imaging, whilst SvO2 in the inferior vena cava was quantified using magnetic susceptometry-based oximetry. These measures were combined with arterial oxygen saturation (from finger pulse oximetry) and haemoglobin concentration (from venous blood draw) to quantifyCaO2,and CvO2), a-vO2diff and VO2 in accordance with the Fick equation. Imaging was performed in 24 older women free of cardiovascular disease (68±4 yrs; VO2peak: 21.1±5.2 mL/kg/min) at rest, and during light- (Power output: 42±10W; 0-10 rating of perceived exertion [RPE]: 3.6±1.5) and moderate-intensity (Power output: 60±9W; RPE: 5.4±1.8) supine stepping exercise on an MRI compatible ergometer. Results Biventricular volumes and Qc were evaluable in all cases, whilst SVO2 could not be calculated in two participants (due to aliasing artefact or excessive motion during exercise). Consequently, VO2 and Fick determinants were assessed at rest and during exercise in 22 participants (92%). Resting VO2 was 223±77 mL/min (3.3±1.2 mL/kg/min) and increased linearly with increasing power output to 597±191 mL/min (8.9±2.5 mL/kg/min) at light-intensity and 819±225 mL/min (12.2±3.3 mL/kg/min) at moderate-intensity exercise (Fig 1A). The increase in VO2 from rest to exercise was the result of a progressive increase in Qc (Fig 1B), and a widening of a-vO2diff (Rest: 4.2±1.5vs Mod: 9.2±2.0 ml/dL) secondary to a decrease in CvO2 (Fig 1C; Rest: 13.4±1.5 mL/dL vs Mod: 8.3±2.0) and SvO2 (Rest: 73±7% vs Mod: 45±10%). Conclusion A combined non-invasive real-time cardiac and venous susceptometry ExCMR approach was feasible for measuring VO2 and its Fick determinants, and demonstrated normal physiologic responses to light and moderate intensity exercise. Further work is needed to determine the feasibility and utility of this approach during peak exercise in individuals with health and established disease.