Predictors of residual tricuspid regurgitation after percutaneous closure of atrial septal defect

Aims: Functional tricuspid regurgitation (TR) associated with atrial septal defects (ASDs) is frequently present due to right-sided volume-overload. Tricuspid valve (TV) repair is often considered in candidates for surgical ASD closure, and percutaneous TV repair is currently under clinical investigation. In this study, we develop a prediction model to identify patients with residual moderate/severe TR after percutaneous ASD closure. Methods and results: In this observational study, 172 adult patients (26% male, age 49 ± 17 years) with successful percutaneous ASD closure had pre- and post-procedural echocardiography. Right heart dimensions/function were measured. TR was assessed semi-quantitatively. A prediction model for 6-month post-procedural moderate/severe TR was derived from uni-and multi-variable logistic regression. Clinical follow-up (FU) was updated and adverse events were defined as cardiovascular death or hospitalization for heart failure. Pre-procedural TR was present in 130 (76%) patients (moderate/severe: n = 64) of which 72 (55%) had ≥1 grade reduction post-closure. Independent predictors of post-procedural moderate/severe TR (n = 36) were age ≥60 years [odds ratio (OR) 2.57; P = 0.095], right atrial end-diastolic area ≥10cm2/m2 (OR 3.36; P = 0.032), right ventricular systolic pressure ≥44 mmHg (OR 6.44; P = 0.001), and tricuspid annular plane systolic excursion ≤2.3 cm (OR 3.29; P = 0.037), producing a model with optimism-corrected C-index = 0.82 (P < 0.001). Sensitivity analysis excluding baseline none/mild TR yielded similar results. Patients with moderate/severe TR at 6-month FU had higher adverse event rates [hazard ratio = 6.2 (95% confidence interval 1.5-26); log-rank P = 0.004] across a median of 45 (30-76) months clinical FU. Conclusion: This study shows that parallel to reduction of volume-overload and reverse remodelling after percutaneous ASD closure, TR improved substantially despite significant TR at baseline. Our proposed risk model helps identify ASD patients in whom TR regression is unlikely after successful percutaneous closure.