Longitudinal feature tracking analysis of right ventricular function in patients with pre-tricuspid shunts: a cardiac MRI study

Abstract Background Pre-tricuspid shunts, such as atrial septal defects (ASDs) and partial anomalous pulmonary venous return (PAPVR), result in left-to-right volume overload and may progress to pulmonary arterial hypertension (PAH). Cardiac magnetic resonance (CMR)-based RV feature tracking (RV-FT) has emerged as a non-invasive modality to evaluate RV longitudinal strain and function. Purpose To investigate the utility of longitudinal RV feature tracking in the assessment of RV function in patients with pre-tricuspid shunts. Methods We retrospectively analyzed a cohort of patients with pre-tricuspid shunts and compared them to age- and sex-matched healthy controls. Inclusion criteria comprised eligibility for CMR and absence of claustrophobia. The imaging protocol included cine sequences in the short-axis, long-axis (3CH, 4CH, RV inflow/outflow views), axial stack, velocity-encoded phase contrast imaging of the pulmonary artery, aorta, and atrial septum, as well as late gadolinium enhancement (LGE) sequences. Post-processing was performed using dedicated software (Circle CVI) to derive biventricular volumes and function, Qp:Qs ratio, and longitudinal RV strain from 4CH cine images. Statistical analyses included non-parametric testing (Wilcoxon rank-sum), correlation matrices, and Fisher’s exact test. Results Sixteen participants were included: 8 patients with pre-tricuspid shunts (4 males, 4 females; median age 49 years [IQR: 31–68; range: 18–75]) and 8 controls (5 males; median age 39 years [IQR: 28–56; range: 25–57]). Among the shunt group, three patients had sinus venosus ASDs with PAPVR, three had ostium secundum ASDs, and one had PAPVR involving both the superior and inferior vena cava. Median RV functional parameters were as follows: for the shunt group—EF 54%, RV end-diastolic volume (RVEDV) 121 mL/m², RV stroke volume (RVSV) 100 mL; and for the control group—EF 56%, RVEDV 75 mL/m², RVSV 83 mL. In the shunt group, significant negative correlations were observed between LV stroke volume and Qp:Qs ratio (r = –0.643), right atrial volume (r = –0.731), and main pulmonary artery (MPA) diameter (r = –0.323). A strong positive correlation was found between MPA diameter and age (r = 0.970). While RV-FT values did not significantly differ between patients and controls (p = 0.373, Holm-adjusted), RV longitudinal strain was significantly correlated with RV ejection fraction (r = 0.738) in the shunt group. Conclusions CMR-based longitudinal RV feature tracking may provide valuable insight into RV function in patients with pre-tricuspid shunts. Although RV-FT did not demonstrate significant intergroup differences, its strong correlation with RVEF in the shunt group underscores its potential utility in clinical risk stratification, particularly in patients with incidentally diagnosed congenital heart defects.