Diverse Right Ventricular Remodeling Evaluated by MRI and Prognosis in Eisenmenger Syndrome With Different Shunt Locations

Gong C, He S, Chen X, Wang L, Guo J, He J, Yin L, Chen C, Han Y, Chen Y.

J Magn Reson Imaging. 2022 May;55(5):1478-1488. doi: 10.1002/jmri.27791. Epub 2021 Jun 21.

PMID: 34152058


Take Home Points

Eisenmenger syndrome patients with pre-tricuspid shunts show worse right ventricular decompensation remodelling features on MRI and have worse prognosis than those with post-tricuspid lesions.


Commentary from Dr Simon MacDonald (London, UK):

Eisenmenger syndrome (ES) is a type of pulmonary hypertension in congenital heart disease where paradoxical right-left shunting occurs with associated systemic cyanosis with multiorgan complications. How the right ventricle remodels in this process and what this entails for prognosis has been unclear.


Chao Gong and colleagues investigated right ventricular (RV) remodelling with MRI in patients with different Eisenmenger syndrome (ES) subgroups classified by location of shunt. It was a single centre prospective study of 54 ES patients, 16 with pre-tricuspid and 38 with post tricuspid shunts. Patients with atrial septal defects (ASDs) and/or anomalous pulmonary venous drainage were classified as pre-tricuspid with those with ventricular septal defect (VSD) and/or patent ductus arteriosus (PDA) as post-tricuspid. Fourteen pre-tricuspid patients had ASDs only, with the largest post-tricuspid group being VSD only (19 of 38) with 11 having PDAs only.


MRI investigations detailed RV morphology, systolic function, RV-pulmonary artery (RV-PA) coupling and fibrosis, measured via gadolinium enhancement. They were also able to compare this to right heart catheterization data, and examined all-cause mortality or readmission for heart failure as clinical endpoints.


The pre-tricuspid patients were significantly older than post-tricuspid (43 vs 34 years, p=0.02), as a possible confounder, and 66% of patients were female. Blinding to shunt position and patient group during MRI analysis will not have been possible. Demographic data were as shown in Table 1 below:


Table Description automatically generated


Patients with pre-tricuspid shunts had lower mean PA pressures and pulmonary vascular resistance at catheterisation but greater pulmonary than systemic blood flows than those with post-tricuspid lesions.


MRI added additional information, not just on shunt location, but ventricular function and size, scar (assumed from gadolinium enhancement) and RV-PA decoupling, RV-PA decoupling being defined as the ratio of RV stroke volume to end-diastolic volume. The authors found that those with pre-tricuspid lesions had larger RVs, RV/LV volume ratios, worse RV function and RV-PA decoupling and this also correlated with worse outcomes. Myocardial tissue characterisation was possible, the authors finding that pre-tricuspid lesion patients had higher extracellular volumes, T1 values, and fibrosis as detailed in tables 3 and 4 below:



Sharing of volume load between LV and RV may explain why these patients do better than those with idiopathic PAH. LV volumes or shunts may not have been large enough to impact RV function in the post-tricuspid group. Numbers were not large enough for lesion specific subanalysis.


This study adds to previous studies that describe adverse RV remodelling using echo, now using MRI techniques, and also compares MRI findings to invasive haemodynamic data. The mismatch in PAH severity measured invasively and RV function may mean these additional measures that can be assessed on MRI will be increasingly useful, adding to the utility of cardiac MRI in this group of patients.