Predictors of clinical outcome in adults with repaired tetralogy of Fallot and the prognostic value of surrogate markers of right ventricular dyssynchrony

Abstract

Abstract Introduction Sequelae or residual lesions after surgical repair of tetralogy of Fallot (rTOF) are inevitable and have an impact on long-term outcome. So far, most of our attention was directed to negative consequences of residual pulmonary regurgitation. Less attention has been paid to the impact of right bundle branch block and intraventricular dyssynchrony on biventricular function, exercise capacity and clinical outcome. Purpose Using the Swiss Adult Congenital HEart disease Registry we aimed to investigate the 5-year clinical course of adults with rTOF late after surgical repair. We used echocardiography strain analysis as surrogate to quantify right and left ventricular dyssynchrony and correlated our findings with routine clinical measures from cardiopulmonary exercise testing, neurohumoral blood work and cardiac magnetic resonance imaging with the aim to identify patients at increased risk for an adverse clinical event. Methods 2D speckle-tracking strain analysis of the right and left ventricle was performed with dedicated software (Tomtec, Germany). RV dyssynchrony index was measured as the standard deviation of time to peak shortening (TTP) in six RV segments. Interventricular shortening delay (IVSD) was defined as the maximal delay of TTP between ventricular segments. We constructed a multivariate model with NT-proBNP, ECG, clinical, imaging and exercise testing variables to predict the composite outcome of all-cause mortality, relevant arrhythmias and hospital admission for decompensated heart failure. Results A total of 285 patients were included. During a mean follow-up of 48±21 months, 33 patient (12 %) suffered an adverse event with a mean time to event of 30 ± 21 months after the baseline visit. RV dyssynchrony index and IVSD did not differ statistically between the groups with and without events (49±31ms vs 41±18ms, p-value 0.21, 130±65ms vs. 103±50ms, p-value 0.9). In the multivariate Cox-regression model, NT-proBNP, right atrial area and peak heart rate at exercise testing were independent predictors of outcome. An integrated risk score with the median values of these 3 variables as cut-off was highly accurate in separating patients at low risk (score 0; n=39) and high risk (score 3; n=38) for an adverse event. After 6 years of follow-up, no patient at low risk (score 0) had an adverse event, while in 62% of patients at high risk (score 3) an adverse event occurred (p<0.001). Conclusion In our cohort of rTOF patients, echocardiographic surrogates of RV dyssynchrony did not correlate with clinical events. An integrated multimodality approach proved to be effective in predicting outcomes.