Age and gender influence left ventricular diastolic filling flow assessed by 4D flow MRI

Assessing left ventricular (LV) diastolic function is critical for diagnosing and managing conditions like heart failure with preserved ejection fraction. Doppler echocardiography is the standard imaging modality but relies on single-directional velocity measurements and is observer-dependent. In contrast, 4D flow MRI offers an advanced alternative by providing comprehensive and reproducible volumetric flow data. Combined with cine-based valve tracking, 4D flow MRI has shown promise in accurately quantifying mitral valve (MV) flow during diastole, which is closely governed by LV diastolic function. However, its potential value for diagnosing diastolic dysfunction remains largely unexplored. This study is a preliminary investigation to determine whether 4D flow MRI can detect age- and gender-related changes in LV diastolic function in healthy adults.

Methods:

The study included 39 healthy volunteers (age 18-82, 20F) without known cardiovascular disease. Cine images were acquired in apical 2- and 4-chamber views and a short-axis stack using a steady-state free precession technique. Whole-heart 4D flow MRI was acquired using a free-breathing, retrospectively ECG-gated GRAPPA (R=2) method. Blood velocities through MV were extracted from 4D flow MRI by tracking the MV hinge points in 2- and 4-chamber cine images and transferring them to the 4D flow domain. The MV plane was reconstructed using multiplanar reformation, and through-plane flow velocities were calculated. MV plane movement velocities along the through-plane axis were calculated to estimate peak myocardial relaxation velocity (e′) and to correct MV through-plane velocities for the moving reference frame. Peak MV flow rates during early LV relaxation (E-wave) and late atrial contraction (A-wave) were quantified, along with the E/A ratio and MV stroke volume (SV). E- and A-wave flow rates were indexed to body surface area to account for individual size differences. Image post-processing and flow quantification were performed using custom MATLAB code, while cine-based MV tracking was performed using Segment v4.0 R12067b (MedViso). Figure 1 shows the post-processing workflow and MV flow waveform examples.

Results:

All subjects were in sinus rhythm during the MRI exam, and the MRI showed no evidence of valve disease in any of the subjects. Table 1 summarizes the demographics, and MV flow metrics and e′ for all subjects and by gender. While age and LV function were similar between genders, females had smaller body surface areas and LV volumes. Mitral SV was in good agreement (ICC = 0.82, 0.67-0.90, p < 0.0001) with cine-derived SV, with negligible bias (0.2 ± 27 ml). Figure 2 shows age-related changes in MV flow metrics and e′ for the combined cohort and by gender. With advancing age, the E/A ratio decreased, and both absolute and indexed A-wave flow rates increased in the combined and gender subgroups. E-wave flow rates, when indexed, decreased with age in the combined cohort and females but not in males. Additionally, e′ declined with age in the combined cohort, though the trend was not significant in gender-specific analyses.

Conclusion:

This study demonstrates that MV flow rates quantified using 4D flow MRI with retrospective valve tracking reveal age-related changes in LV diastolic function, with notable gender differences. These findings suggest potential for this method in assessing diastolic dysfunction, warranting further exploration in patients with LV diastolic dysfunction.