Cardiopulmonary exercise testing is gaining increased recognition as a useful tool for assessing pulmonary hypertension (PH)¹. The increased pulmonary artery pressure associated with the disease can shift the ventricular septum, resulting in decreased left ventricular (LV) filling and aortic flow. This response becomes more pronounced when stressed through exercise. Due to its increased accuracy in measuring cardiopulmonary function, MRI is preferred to echocardiography for assessing such physiologic responses. Traditionally, however, MRI relies on pharmacologically induced stress due to inherent constraints within the MRI bore. Previously we demonstrated the use of MRI-compatible exercise equipment² to assess aortic flow and LV function in healthy subjects^3,4. In this study, we aim to determine if changes to aortic flow caused by PH are detectable using MRI following exercise stress.Three subject cohorts were imaged: 20 young, healthy controls (26±4 years,8M,12F), 8 older, age-matched controls (58±10 years,1M,7F), and 6 PH patients (60±9 years,1M,5F). Imaging was performed on a clinical 1.5 T system (HDx/Discovery 450w, GE Healthcare, Waukesha, WI). A custom-built MRI-compatible stepping device was used which allowed subjects to exercise in a supine position in the scanner bore (Figure 1). Figure 2 shows the exercise/scanning protocol for each cohort. A low power of 30W was targeted for the PH patients whose exercise capacity can be greatly compromised. The control groups exercised at slightly higher powers to induce measurable responses. Each exercise stage was followed immediately by an ECG-gated 2D cine PC acquisition (TR/TE=6.1/3.7ms; FA=30^o; ASSET=2; VENC=200cm/s; resolution=256x256) with 20 cardiac phases in the ascending aorta across a 15s breath hold. Changes in heart rate, stroke volume, cardiac output, peak systolic velocity and flow, relative aorta area change, and pulse wave velocity (using the QA method⁵) were analyzed with CV Flow (Version 3.3, Medis, Leiden, Netherlands) and MATLAB (Mathworks,Natick,NJ). A paired t-test and 1-way ANOVA with Tukey’s HSD test were used to determine the statistical significance (p<0.05) of intra-group and inter-group changes respectively between exercise and baseline. To reduce the effect of varying exercise power between subjects, the exercise stage measuring closest to 30 W was analyzed for each subject.The average exercise powers were 36±8 W, 33±6 W, and 30±9 W with an average increase in heart rate of 35%, 39%, and 65% (Figure 3) for the young controls, older controls, and PH subjects respectively, all of which were statistically significant. Figure 4 displays boxplots comparing the results for each group for the remaining parameters investigated. A bracket between two boxplots indicates the two groups have a statistically significant difference.The increase in heart rate was much higher for the PH subjects than the control groups, indicating a higher stress level for this modest exercise. The results for stroke volume and cardiac output indicated that younger, healthy subjects showed a greater ability to increase blood supply in response to exercise stress. For many of the parameters measured at stress the older controls and PH groups showed a wider range of values than the younger controls. We hypothesize this is due to a greater perceived difficulty for each exercise challenge to these groups, resulting in higher respiratory and cardiac motion compromising image quality, making ROI measurements more difficult. Differing degrees of PH and age-related loss of CV function could also be responsible for the spread found in these subjects. Many of the parameters investigated did not show strong intra-group statistical significance between rest and stress. This may be a result of low statistical power from the small sample size for the PH subjects, and work-rates that were too low to elicit a strong cardiopulmonary response, as evidenced by the lower heart rates, in the healthy controls. Future studies may benefit from standardizing comparisons based on heart rate instead of exercise power. The implementation of improved motion-correction and real-time imaging would also allow for more accurate measurements, possibly allowing us to detect more subtle hemodynamic changes brought on by exercise.This study demonstrates the use of MRI to measure the effect of modest exercise stress on aortic flow in both healthy controls and patients with pulmonary hypertension. The PH patients showed a good response to the exercise challenge (an average 65% increase in heart rate), yet their response was heterogeneous for measures of cardiac output and stroke volume, likely reflecting different stages of loss of CV function. The response to the exercise challenge was less pronounced for the control groups due to the low exercise power. Future studies comparing larger cohorts of subjects based off of heart-rate instead of work-rate may provide comparisons with more statistical power.