Four-Dimensional Magnetic Resonance Pulmonary Flow Imaging for Assessing Pulmonary Vasculopathy in Patients with Postcapillary Pulmonary Hypertension.

Background

Noninvasive techniques for diagnosing combined postcapillary pulmonary hypertension (CpcPH) are unavailable.

Objective

To assess the diagnostic performance of cardiac magnetic resonance (CMR)-based four-dimensional (4D)-flow analysis in identifying CpcPH.

Methods

Prospective observational study of heart failure (HF) patients with suspected pulmonary hypertension (PH) who underwent simultaneous CMR and right heart catheterization. The 4D-flow biomarkers were calculated using an automatic pipeline. A predictive model including 4D-flow biomarkers associated with CpcPH with a p-value < 0.20 was built to determine the diagnostic performance of 4D-flow analysis to identify CpcPH.

Results

A total of 46 HF patients (55.4 ± 14 years, 63% male) with confirmed PH (19 [41%] isolated postcapillary PH [IpcPH], 27 [59%] CpcPH) were included. No differences were found in baseline characteristics, echocardiography, or CMR anatomical and functional parameters, except for a higher Doppler-estimated systolic pulmonary pressure and larger pulmonary artery in CpcPH patients. The 4D-flow CMR analysis was performed in 31 patients (67%). The maximal peak velocity (67.1 [62.2-77.5] cm/s-IpcPH vs. 58.2 [45.8-66.0] cm/s-CpcPH; p = 0.021) and maximal helicity (339.9 [290.0-391.8]) cm/s2-IpcPH vs. 226.0 (173.5-343.7) cm/s2-CpcPH; p = 0.026) were significantly lower in patients with CpcPH. A maximal multivariable model including sex, maximal average, and peak velocities, Reynolds number, flow rate, and helicity showed fair diagnostic performance (area under the curve: 0.768 [95%-CI: 0.572-0.963]; sensitivity: 100%; specificity: 55%).

Conclusions

In HF patients with PH, 4D-flow-derived maximal peak velocity and maximal helicity were significantly lower in CpcPH patients. A multiparametric model including maximal 4D-flow-derived biomarkers showed good diagnostic performance for identifying CpcPH.