Longitudinal changes in systemic right ventricular remodeling in adult patients with transposition of the great vessels as assessed by cardiovascular magnetic resonance imaging.

Background: Systemic right ventricular (sRV) physiology occurs in patients with congenitally corrected transposition of the great arteries (ccTGA) and D-TGA post atrial switch repair, and the natural history is of progressive sRV dysfunction. No study has assessed longitudinal changes in sRV remodeling by serial CMR.

Methods: Patients evaluated at two adult congenital heart disease centers and who underwent ≥2 CMR exams were studied. Indexed sRV end-diastolic volume (sRVEDVi), end-systolic volume (sRVESVi), and ejection fraction (sRVEF) were determined by a core laboratory. Concurrent echocardiograms were assessed for degree of systemic TR (sTR). Tricuspid valve events were defined as ≥moderate sTR, or interval tricuspid replacement (TVR). Generally, the earliest and most recent studies were compared. A subset of patients were followed with ≥moderate sTR, and then subsequently underwent interval TVR. For these patients, two study time-intervals were defined to analyze the impact of each event independently.

Results: 67 patients were studied (33±11 years, 47% male, 33% ccTGA), with 72 total time intervals studied (median interval 9.0 years [IQR 4.6-13.3]). There was a small increase in sRVEDVi over time (ΔsRVEDVi 5.5±15.8ml/m², p<0.001), but mean change in sRVEF was not significant (ΔsRVEF 0.1±6.9%, p=0.86); notably, confidence intervals were wide for both. ccTGA patients had a trend towards greater decrement in sRVEF (ΔsRVEF -1.7±6.8 vs 1.3±6.7%, p=0.06). For each 25ml/m2 increase in baseline sRVEDVi, there was a 1.8% decrease in sRVEF (95% CI -3.2% to -0.5%, p=0.01). Patients without significant sTR had lesser deterioration in sRVEF compared to those with ≥moderate sTR or with interval TR intervention (ΔsRVEF 1.8±6.9% vs -2.1±6.6% and -2.6±4.5, p<0.05). Interval sRV conduction delay was associated with a trend towards greater decrements in sRVEF (ΔsRVEF -3.9±6.3 vs. 0.9±6.8%, p=0.07). Overall, underlying congenital anatomy, baseline sRVEDVi, advanced sTR or interval TVR, and sRV conduction delay explained only 16% of the variability in ΔsRVEF over time.

Conclusions: Longitudinal changes in sRV remodeling were small, with great heterogeneity. Apparent risk factors in our study, namely underlying congenital anatomy, baseline sRVEDVi, TR events, and sRV conduction disease accounted for only 16% of the variability seen in the longitudinal change of sRVEF.