Metabolic Cost And Reproducibility Of Exercise Cardiac Magnetic Resonance Imaging

PURPOSE: Cardiac magnetic resonance imaging (cMRI) is ideal for non-invasive assessment of cardiovascular structure and function. Advancements in imaging technology and hardware development has made exercise cMRI more robust and increasingly more feasible. The purpose of this study was therefore three-fold: First, to compare the oxygen cost of upright cycling with supine exercise using a commercially available MRI compatible exercise ergometer. Second, to evaluate the bi-ventricular response to graded sub-maximal exercise using cMRI. Third, to evaluate the reproducibility of exercise cMRI across a broad range of exercise intensity. METHODS: To accomplish these goals, we first performed incremental exercise tests, on two separate days, using a conventional upright cycle ergometer and an MRI compatible stepping ergometer; beginning with three-minute stages (4 x 0.3 W/kg/stage), followed by one-minute stages to exhaustion (0.2 W/kg/stage). Then, on a separate visit, participants exercised inside the bore of a 3 T MRI at 20-, 40-, and 60% of their upright maximum. Left ventricular stroke volume was assessed by the method-of-disks, using high-resolution short-axis cine imaging, with brief (~2-3 sec) end-expiratory pauses. To assess the reproducibility of exercise cMRI, imaging was repeated on a separate visit. RESULTS: Peak oxygen consumption, peak heart rate, and peak work rate were lower during supine stepping exercise with the MRI-compatible ergometer than during upright cycle exercise (supine: upright = 86 ± 7%, 93 ± 8%, and 91 ± 9%). The metabolic cost of performing submaximal exercise, however, was very similar between the two ergometers (0.21 ± 0.06 ml/kg/min per watt versus 0.20 ± 0.05 ml/kg/min per watt, upright versus supine, respectively; p = 0.34). During exercise cMRI, left ventricular stroke volume increased from rest with 20% and 40% exercise, returning towards baseline at 60% (100.7 ± 21.5 mL, 114.6 ± 29.0 mL, 116.6 ± 33.8 mL, 104.2 ± 27.7 mL, rest, 20%, 40%, and 60%, respectively), with excellent agreement between trials (ICC =0.96). CONCLUSION: Taken together, these results support the use of exercise cMRI for evaluation of cardiovascular structure and function, establishing both feasibility and reproducibility.Funded by NIH (P01HL137630) and the Potratz Family Endowment.