Pulmonary Vein Flow Indices by MRI for Diastolic Function Evaluation

Transthoracic echocardiography (TTE) is the first-line and most useful imaging modality for evaluating diastolic dysfunction, but its ability to measure pulmonary vein (PV) flow is limited. We sought to establish MRI-based PV flow for diastolic dysfunction by comparing it to TTE and evaluating its performance in patients. First, near same-day (2 ± 3 days) dedicated TTE and MRI were performed in a young cohort (N = 12, 36 ± 20 years old) for comparison of PV flow measurements at 3 T. Secondly, an older healthy cohort (N = 11) was also evaluated for PV flow. Finally, 21 patients with a recent TTE (within 4 months) were prospectively enrolled from 2019 to 2022 and imaged at 1.5 T to measure PV flow. In these patients, a dedicated phase-contrast scan was performed to assess PV flow as an addition to their clinical scan. With near same-day dedicated TTE and MRI to measure PV velocity and flow, S/D (measured by total flow) agreed moderately between TTE and MRI (r = 0.72, p = 0.03, N = 12), and the average S/D was 1.2 ± 0.4. The Bland-Altman limits of agreement revealed bias ±2 SD = 0.13 ± 0.64 mL/s. MRI-based S/D using peak velocity was highly correlated to S/D by total flow (R = 0.97, p < 0.001) in this cohort. Six of 21 patients had LV diastolic dysfunction (four indeterminate, two with Grade II dysfunction). S/D (by flow) was not different in patients with DD vs. without (1.5 ± 0.6 vs. 1.4 ± 0.5) or in healthy age-matched controls (1.5 ± 0.6). The D wave DT (by flow) was prolonged in patients with diastolic dysfunction vs. patients without (253 ± 42 ms vs. 207 ± 41 ms, p < 0.05) and was 225 ± 61 in healthy age-matched controls. In ROC analysis, D wave DT > 241 ms provided an AUC of 0.78 to predict patients with LV diastolic dysfunction. PV flow by MRI has potential to be a useful method for diastolic function evaluation.