WP-001 - Longitudinal evaluation of cardiac function for 16 months of high-intensity exercise using automated CMR scanning and analysis

Background: Exercise is commonly accepted as beneficial to cardiovascular health, and cardiac MR is ideally suited for longitudinal imaging of the heart. The use of automated algorithms for image acquisition and analysis may simplify workflow and improve reproducibility.  In this case study, we used commercially available tools for automated CMR to assess the effects of high-intensity exercise over 16 months.

Methods: A healthy male subject (34 years, BMI 20.8) with no cardiovascular risk factors and a sedentary lifestyle aside from a 2.4 km walk during his commute was enrolled in an exercise program consisting of 150 min/week of high-intensity elliptical or running (average HR >80% of age-adjusted maximum heart rate (MHR)).  Heart rate and exercise activity were monitored using an Apple Watch Series 3 (Apple Inc., Cupertino, United States) and CMR exams performed on a 1.5T MAGNETOM Aera (Siemens Healthcare, Erlangen, Germany).  Long-axis views and a short-axis stack were planned using the AutoAlign Heart feature of Cardiac Dot.  ECG-gated TrueFISP cine images were acquired with typical parameters: field of view (FOV) 340x280 mm, slice thickness 6 mm, TE/TR 1.16/2.81 ms, flip angle 50°, views per segment 13.  Cardiac Dot was used to automatically adjust the FOV and orientation and initiate scans with breath-hold commands.  LV volumes, mass, and function were automatically calculated on the scanner using Inline Ventricular Function (InlineVF).  Linear regression was performed to analyze potential trends.

Results: Average daily “exercise minutes” categorized by the Apple Watch increased from 38±8 min in the 7 months prior to the start of the study to 58±6 min during the study.  High-intensity exercise was well tolerated with an average of 154±5 min/week, during which the mean average HR was 159±6 bpm (85±3% of MHR) and the mean peak HR was 178±6 bpm (95±3% of MHR).  33 CMR exams were performed at a mean interval of 15±8 days.  Slice orientations were prescribed automatically for 32 of 33 exams (97%) – manual identification for 1 of 5 landmarks was required in one exam.  Automated slice orientation was highly consistent (Fig. 1) and InlineVF analyses were successful in all studies (Table 1).  Cardiac output decreased by 20%, coincident with an 18% decrease in HR, and LV mass increased by 8% (p< 0.05 for all, Fig. 2).  No statistically significant changes were found in ejection fraction or volumes.  Inter-study variability of parameters was good, with ≤10% coefficients of variation for all metrics without a significant trend.

Conclusion: In this case study, heart rate, cardiac output, and mass were significantly changed through high-intensity exercise, consistent with previous studies (1,2) although no change was found in end-diastolic volumes.  One hypothesis is that subjects with >30 min/day of walking may respond differently than more sedentary individuals.  Further study is required, and automated scanning and analysis may reduce these as sources of error in longitudinal studies with small effect sizes.