Congenital Heart Disease
Tobias Rutz, MD
Consultant/MD
Lausanne University Hospital and University of Lausanne
Lausanne, Vaud, Switzerland
Estelle Tenisch
Consultant/MD
Lausanne University Hospital and University of Lausanne, Switzerland
Mariana Falcao, MSc
PhD Candidate
Lausanne University Hospital and University of Lausanne
Lausanne, Vaud, Switzerland
Christopher W. Roy, PhD
Post Doc
Lausanne University Hospital and University of Lausanne (UNIL)
Prilly, Vaud, Switzerland
Sara Faessler
MS
Lausanne University Hospital and University of Lausanne, Vaud, Switzerland
David Rodrigues, MSc
Deputy chief radiology technician
Lausanne University Hospital and University of Lausanne, Switzerland
.jpg "Liliana Ma, MD, PhD photo")
Liliana Ma, MD, PhD
Resident
Ascension RMC and Northwestern University
Mark Markl, PhD
Professor
Northwestern University
Matthias Stuber, PhD
Professor
University Hospital (CHUV) and University of Lausanne (UNIL)
Lausanne, Switzerland
Davide Piccini, MD, PhD
PhD
Siemens Healthcare
Lausanne, Vaud, Switzerland
Juerg Schwitter, MD
Professor
Lausanne University Hospital and University of Lausanne
Lausanne, Vaud, Switzerland
Milan Prša, MD
Attending physician
Lausanne University Hospital and University of Lausanne, Switzerland
Flow quantification by cardiac magnetic resonance (CMR) is particularly important in right-sided congenital heart disease (CHD). 2D phase contrast (2DPC) flow sequences are limited by the need for precise prescription of image planes, long scan time and impreciseness in the presence of valvulopathies.1 Recently introduced accelerated free-breathing 3D whole heart flow sequences promise to simplify CMR exam allowing faster image acquisition and retrospective flow measurements. This study compares therefore standard 2DPC to an accelerated 4D flow sequence with prospective respiratory navigation2 and a free-running radial fully self-gated respiratory and cardiac motion-resolved 5D flow.3,4
Methods:
Patients with right-sided CHD were scanned on a 1.5T MAGNETOM Sola system (Siemens Healthcare, Erlangen, Germany). 2DPC was performed in the ascending, descending aorta (AA, DA), main, right and left pulmonary artery (MPA, RPA, LPA) and superior vena cava (SVC). A prototype whole-heart free-running 3D radial PC CMR sequence was also acquired.3 Using a compressed-sensing-based image reconstruction framework3,4, 5D flow images were obtained, and the end-expiratory phase was used for analysis. Finally, 4D flow data were collected.2 3D vessel segmentation was based on retrospectively computed phase-contrast angiography images for each 5D and 4D flow datasets. Flow rate was measured in the AA, DA, MPA, RPA, LPA and SVC on images obtained with all three sequences.
Results:
Fifteen patients (age 31±3 years, 7 women, tetralogy of Fallot 10, Ross operation 4, pulmonary valvuloplasty 1) were included. Mean pulmonary regurgitation fraction (PR) was 16 ± 4 % (range 1; 51) and peak velocity of MPA/ right ventricle to pulmonary artery conduit was 2.0 ± 0.2 m/s (range 0.9; 3.5).
Compared to 2DPC, both 4D and 5D flow generally tended to underestimate absolute flow rate with significant reduction in the RPA and LPA for 4D (table 1).
Analysis of RPA flow revealed a significantly smaller bias for the comparison of 2DPC vs. 5D flow rate compared to 2DPC vs. 4D flow rate (table 2).
The bias i.e. the difference of RPA flow rate between 2DPC vs. 4D as well as 2DPC vs. 5D flow, was directly related to the RPA peak velocity (RPA PV, figure): RPA PV to RPA 2DPC vs. 4D: r2=0.413, p=0.013, RPA PV to 2DPC vs. 5D: r2=0.317, p=0.045. The bias of RPA flow rate between 2DPC vs. 4D flow correlated to the MPA PR: r2=0.308, p=0.039. There were no similar observations for other vessels.
Conclusion:
4D and 5D flow provide results comparable to 2DPC, however, both sequences showed a trend to lower absolute flow values which is particularly true for the pulmonary arteries. Flow turbulences due to the pulmonary valvulopathies could explain these observations, as their severity appears to influence the agreement between 2DPC and both 4D and 5D flow sequences.
Larger studies are planned to better elucidate the reasons for our results and to understand which sequences determine most accurately the true flow rate.