Objective Patients with surgically repaired tetralogy of Fallot (rTOF) may develop functional tricuspid regurgitation (TR) due to annulus dilation. Guidelines suggest pulmonary valve replacement (PVR) in patients with rTOF with progressive TR, but data on clinical outcomes are lacking. Our objective was to determine whether TR was predictive for adverse events after PVR.
Methods In this retrospective, multicenter cohort study, patients with rTOF who had undergone PVR after preoperative echocardiographic assessment of TR grade were included. Preoperative and postoperative imaging data and a composite of adverse clinical events (death, sustained ventricular tachycardia, heart failure, or supraventricular tachycardia) were collected. Multivariate Cox hazards regression analysis was used to determine which factors were predictive for adverse events after PVR.
Results A total of 129 patients (61% men, age at PVR 32.9±10.4 years) were included. The composite endpoint occurred in 39 patients during 8.4±4.2 years of follow-up. In multivariate analysis, severe preoperative TR (HR 2.49, 95% CI 1.11 to 5.52), right ventricular end-systolic volume (HR 1.02/mL/m2, 95% CI 1.01 to 1.03) and age at PVR (HR 1.07/year, 95% CI 1.04 to 1.09) were predictive for adverse events. Early postoperative TR was not predictive for adverse events (p=0.96). In patients without any risk factor (age >40 years, right ventricular end-systolic volume >90 mL/m2 or severe TR), 5-year event-free survival was 100% as compared with 61% in patients with two or three risk factors.
Conclusions In patients with rTOF, severe preoperative TR was predictive for adverse events after PVR. Close surveillance is warranted in these patients irrespective of postoperative TR.
Statistics from Altmetric.com
In adult patients with repaired tetralogy of Fallot (rTOF), residual pulmonary regurgitation (PR) after relief of pulmonary stenosis is common.1 The resulting right ventricular overload causes right ventricular dilation and eventually dysfunction2 ,3 and is accompanied with annulus dilation, which may lead to functional tricuspid regurgitation (TR).4 In addition, some patients with rTOF have tricuspid leaflet abnormalities or displacement after ventricular septal defect closure or ventricular lead placement, which in itself may cause or aggravate TR.5–7 TR may lead to progressive right ventricular dysfunction and dilation, arrhythmias,4 ,8–10 and symptoms contributable to right-sided heart failure. These effects can cause a vicious circle in which annulus dilation and TR worsen due to progressive right ventricular dilation. Surgical pulmonary valve replacement (PVR) is frequently performed in patients with rTOF with severe PR and leads to a marked decrease in right ventricular volumes and PR fraction,11 ,12 therewith reducing the severity of TR.13 However, it is unclear whether TR reduction after PVR with or without concomitant tricuspid valve repair is sustained during long-term follow-up. Moreover, TR is associated with worse survival in patients with left ventricular dysfunction and/or chronic heart failure.10 ,14 In patients with rTOF, TR is associated with worse functional class and elevated neurohormone levels,15 but data on clinical outcomes are scarce.9 Previous studies have investigated predictors for clinical outcomes in rTOF, but lacked data on TR severity.16–19 Guidelines advocate that PVR should be considered in patients with rTOF with progressive TR in combination with severe asymptomatic PR or pulmonary stenosis.20 Our objective was to determine whether TR severity was prognostic for adverse events in patients with rTOF undergoing PVR.
For this retrospective, multicenter cohort study, we identified adult patients with rTOF who had undergone PVR in one of three participating academic centres (Academic Medical Center, Leiden University Medical Center, Radboud University Medical Center). Patients were eligible for this study if PVR was performed at an age of at least 12 years and preoperative echocardiographic TR assessment was available.
In patients who had undergone subsequent PVRs, the first PVR performed at an age of >12 years with echocardiographic assessment of TR severity was included for the analysis. Medical records for all patients were reviewed. Follow-up was completed from PVR until death or last medical contact. The collected data included patient characteristics, operations and concomitant procedures. Preoperative and postoperative echocardiographic examinations, preoperative cardiovascular MR (CMR) imaging and ECG examinations were collected. Furthermore, adverse clinical events (mortality, sustained ventricular tachycardia, heart failure, supraventricular tachycardia) after PVR were noted. This study complies with the declaration of Helsinki and its conduct in the Netherlands was approved by the medical ethics review committee of the Academic Medical Center.
Pulmonary valve replacement
Cryopreserved pulmonary homografts were used in almost all patients. All patients were operated on normothermic or moderately hypothermic cardiopulmonary bypass. When no transannular patch was used in the past, the pulmonary artery was opened longitudinally and if necessary this incision was extended across the pulmonary annulus into the right ventricular outflow tract, before insertion of homografts in the pulmonary position. If a transannular patch was used in the primary correction, the patch material was resected as much as possible, after which the right ventricular outflow tract was reefed as far as needed to match the homograft size. Concomitant procedures performed at the time of PVR were noted. Concomitant tricuspid valve suture plasty by any technique and tricuspid valve ring annuloplasty were both considered as tricuspid valve repair.
Echocardiographic examinations were performed on locally available echocardiography systems (Vivid 7 or 9 (General Electric-Vingmed, Milwaukee, Wisconsin) or iE33 (Philips Medical Systems, Andover, Massachusetts, USA)). The last available echocardiographic assessment before PVR, the first assessment after the perioperative period (last before discharge or alternatively first during outpatient follow-up) and late assessment at approximately 5 years after PVR were acquired. Echocardiographic assessments were evaluated by cardiologists at each participating centre. Colour flow Doppler, vena contracta width, pulsed wave Doppler, peak tricuspid systolic inflow, hepatic flow and continuous wave Doppler were used to determine the severity of TR. TR severity was graded into absent/trivial (grade 0), mild (grade 1), moderate (grade 2) or severe (grade 3) according to the European guidelines.21 Vena contract width ≥7 mm and/or systolic flow reversal in hepatic vein were diagnostic for severe TR. In addition, the TR jet was used to estimate right ventricular systolic pressures, using the modified Bernoulli equation. Peak systolic pressure gradient across the right ventricular outflow tract was estimated by continuous wave Doppler echocardiography, using the modified Bernoulli equation. Colour flow mapping of the right ventricular outflow tract and branch pulmonary arteries was used to grade the PR into four categories (0, absent/trivial; 1, mild; 2, moderate; 3, severe).
Cardiovascular magnetic resonance
CMR imaging data were acquired on locally available magnetic resonance systems (1,5-Tesla (NT15 Gyroscan, Philips, Best, The Netherlands or Vision, Siemens, Erlanger, Germany) or 3-Tesla (Intera, Philips, Best, The Netherlands)). CMR studies were performed with the use of previously described imaging protocols.11 ,22 ,23 Briefly, scout images were obtained in three orthogonal planes, and standard 2-chamber and 4-chamber views were acquired. From these views, a stack of 12 to 14 slices was acquired in the short-axis orientation (steady-state free precession). The short-axis orientation was used to calculate both right ventricular and left ventricular volumes from endocardial contours with the use of MASS software (Medis, Leiden, The Netherlands). Trabeculations were included in the blood pool. The EF was calculated as the end-diastolic volume minus the end-systolic volume (which equals the stroke volume) divided by the end-diastolic volume. Velocity mapping was performed with the use of velocity-encoded phase contrast sequence. PR fraction was calculated as PR flow divided by systolic forward flow.
Definition of outcomes
The outcome measures of this study were the course of TR and time to occurrence of adverse events after PVR. Changes in TR grade early after and late after PVR were determined. The composite of adverse events after PVR consisted of mortality, sustained ventricular tachycardia (>30 s or requiring cardioversion), heart failure (increase in New York Heart Association class and requiring increased dose of diuretics) and supraventricular tachycardia (>30 s or requiring cardioversion).
Data were described as number with frequency, median with IQR and mean with SD. Wilcoxon signed-rank test and Mann–Whitney U test were used to determine differences within groups and between groups in TR or change of TR. Multivariate linear regression analysis was used to determine factors independently associated with early postoperative decrease in TR grade. Univariate Cox proportional hazards analysis was used to identify variables prognostic for adverse clinical events after PVR. There was <35% missing data for all variables analysed in univariate analysis. Missing data were imputed five times before performing multivariate analysis. Multivariate analysis was performed with a forward conditional algorithm for which all preoperative variables with a p value <0.10 in univariate analysis were selected. All variables that were independently predictive for adverse events were incorporated into a risk model. Continuous variables predictive for adverse clinical events in the multivariate analysis were dichotomised using relevant cut-off values by plotting receiver operating characteristic curves and assessing the area under the curve. The C-index for the multivariate model was determined and reflects the area under the curve for binary outcome variables with censored survival time data. Kaplan–Meier curves were constructed on the original dataset to determine event-free survival for low-risk, intermediate-risk and high-risk patients. Analyses were performed with SPSS V.20.0 and R statistics with the use of the regression modeling strategies (RMS) package. A p value <0.05 was considered statistically significant.
Our total cohort of PVR patients consisted of 157 patients with rTOF. Of this cohort, all patients (n=129, 61% men) with preoperative (0.40 years before PVR (IQR 0.21–0.71)) echocardiographic assessment of TR were included in our study population (figure 1). Echocardiographic assessment was also performed early (0.02 years (IQR 0.01–0.25)) and late (5.00 years (IQR 4.03–5.77)) after PVR in 125 and 93 patients, respectively. Patient and surgical characteristics are listed in table 1. The mean follow-up duration after PVR was 8.4±4.2 years.
Course of TR after PVR
In the entire study population, TR grade decreased in 53 (42%) patients early after PVR and increased in 26 (28%) patients during follow-up after PVR (figure 2). Preoperative TR grade and early postoperative TR decrease were higher in those who had undergone concomitant tricuspid valve repair (p<0.001 for both). The proportion of patients with an increase of TR grade during follow-up was similar in patients with and without tricuspid valve repair (p=0.30). However, in patients without tricuspid valve repair, TR grade late after PVR was comparable with preoperative TR grade (p=0.37) while TR grade late after PVR remained lower in patients with tricuspid valve repair (p<0.001).
Variables associated with postoperative TR decrease
In univariate linear regression analysis, both tricuspid valve repair (β: 0.96, p<0.001) and preoperative TR grade (β: 0.69 per grade, p<0.001) were associated with a higher early postoperative decrease in TR grade. Both variables were entered in a multivariate analysis; the association with TR decrease was attenuated, but remained significant for both tricuspid valve repair (β: 0.40, 95% CI 0.13 to 0.68, p=0.005) and preoperative TR grade (β: 0.54 per grade, 95% CI 0.38 to 0.71, p<0.001).
Adverse events during follow-up
The composite endpoint occurred in 39 patients during follow-up (figure 3). Five patients died (three after progressive heart failure, two suddenly), seven patients had a sustained ventricular tachycardia, eight patients had heart failure and 19 patients had supraventricular tachycardia. Freedom from adverse events was 83%, 5 years after PVR and 63%, 10 years after PVR (figure 3). Factors predictive for the composite endpoint in univariate and multivariate analysis are listed in table 2.
In multivariate analysis (table 2), preoperative severe TR (HR 2.49, 95% CI 1.11 to 5.52, p=0.028), right ventricular end-systolic volume (HR 1.02/mL/m2, 95% CI 1.01 to 1.03, p=0.002) and age at PVR (HR 1.07/years, 95% CI 1.04 to 1.09, p<0.001) remained independently predictive for adverse events during follow-up. Age at PVR (C-statistic 0.73) was dichotomised at 40 years and right ventricular end-systolic volume (C-statistic 0.71) at 90 mL/m2 before constructing a risk model containing all three dichotomised variables. The C-statistic of the final model containing all three variables was 0.80. In patients without risk factors, 5-year event-free survival was 100%, compared with 81% in patients with one risk factor and 61% in patients with two or three risk factors (figure 4). In addition, we performed an analysis to determine whether the same factors were predictive for adverse events if patients with prior ventricular or supraventricular arrhythmias (n=30) were excluded. In 99 patients without preoperative arrhythmias, 23 events occurred during follow-up. Severe preoperative TR was associated with preoperative arrhythmias (p<0.001) and only four patients with severe preoperative TR did not have preoperative arrhythmias. Therefore, age and preoperative right ventricular end-systolic volume were entered in a multivariate analysis. Both right ventricular end-systolic volume (HR 1.02/mL/m2, p=0.014) and age at PVR (HR 1.12/year, p<0.001) were predictive for adverse events in multivariate analysis.
Our study is the first to reveal that severe preoperative TR is predictive for adverse events after PVR in rTOF. Postoperative TR was not predictive for adverse events. Therefore, patients with severe preoperative TR should remain under stringent follow-up, irrespective of early postoperative TR reduction. Other factors independently predictive of adverse events were right ventricular end-systolic volume (>90 mL/m2) and age at PVR (>40 years).
In our study, we determined the effects of PVR on TR grade in patients with or without concomitant tricuspid valve repair. Kogon et al13 investigated the effects of PVR in patients with rTOF with at least moderate TR. They reported a reduction of TR grade after PVR, irrespective of concomitant tricuspid valve repair. In our study, preoperative TR grade and concomitant tricuspid valve repair were independently associated with a more pronounced TR reduction. We found no difference in increase of TR during follow-up between patients with and without concomitant tricuspid valve repair. This is in contrast with earlier reports describing a higher rate of recurrent TR after mitral valve repair without concomitant tricuspid valve repair.24
We chose to investigate the occurrence of a composite of adverse events after PVR as the mortality rate after PVR in this young population was too low to investigate determinants of mortality. Severe TR may lead to elevated right atrial pressures and/or right atrial dilation, which may contribute to the occurrence of supraventricular arrhythmias.25 ,26 Indeed, earlier studies have revealed a high prevalence of supraventricular arrhythmias in patients with rTOF,26 ,27 which may precede death and sustained ventricular tachycardias.16 Second, we included heart failure in our composite endpoint because volume overload due to TR may lead to irreversible right ventricular dysfunction. Right ventricular dysfunction can be a precursor of heart failure and subsequent sustained ventricular tachycardia or death in rTOF.16 ,19 ,28 Severe TR, whether or not corrected by surgery, has been associated with poor survival in several other patient groups.10 ,29
In our study, patients with severe preoperative TR remained at increased risk for adverse events, irrespective of their postoperative TR grade. There may be several explanations for this unfavourable outcome. First, longstanding volume overload due to PR in patients with rTOF may lead to right ventricular dysfunction2 ,3 and can be aggravated by severe TR. Second, longstanding preoperative right atrial volume and pressure overload due to TR may have contributed to new-onset or recurrent arrhythmias after PVR.25 ,26 Concomitant ablation techniques were performed in only 20% of patients with preoperative arrhythmias. It is conceivable that in some patients recurrent arrhythmias derived from an untreated arrhythmogenic substrate could have been prevented with ablation techniques. The other predictive factors in our multivariate model were older age and larger right ventricular end-systolic volume. Both factors were previously associated with high mortality rates and adverse events in rTOF1 ,4 ,19 but not specifically during follow-up after PVR. Previously, Babu-Narayan et al17 reported aerobic capacity as a predictor for mortality after PVR, while Scherptong et al18 and Oosterhof et al30 reported an association between QRS duration and adverse outcomes after PVR. However, these studies did not investigate the prognostic value of preoperative TR and CMR imaging parameters.
Our study has several limitations, mainly due to the retrospective design. First, no specific criteria were used as an indication for tricuspid valve repair. Therefore, selection bias limits comparison of the outcome of PVR and tricuspid valve repair versus PVR only. Second, preoperative CMR imaging and echocardiographic examinations during follow-up were performed according to the local protocols and were not available for all patients. Third, patients with surgical repair in different surgical periods at a variable age were included in this analysis. This may limit extrapolation of our results to a younger patient group. Finally, preoperative aerobic capacity was not routinely tested in our population and its prognostic value could therefore not be determined.
Patients with rTOF with severe preoperative TR were at high risk for adverse clinical events after PVR, irrespective of early postoperative TR reduction. Severe preoperative TR can be considered a risk marker for adverse events and close follow-up is warranted in these patients.
What is already known on this subject?
Patients with tetralogy of Fallot may develop functional tricuspid regurgitation due to annulus dilation secondary to pulmonary regurgitation. The long-term effects of pulmonary valve replacement on tricuspid regurgitation and clinical outcomes are unknown.
What might this study add?
We found that the degree of tricuspid regurgitation was reduced early after pulmonary valve replacement. However, recurrent tricuspid regurgitation occurred in a subgroup of patients.
Patients with severe preoperative TR were at high risk for adverse events during long-term follow-up.
How might this impact on clinical practice?
Clinicians should incorporate the assessment of tricuspid regurgitation in the risk stratification of patients with tetralogy of Fallot.
Contributors All authors attributed in the conception, design, critical revision and final approval of this manuscript. JPB analysed and interpreted the data and drafted the manuscript under supervision of senior authors BJMM and BJB.
Funding This work was supported by the Parelsnoer Institute and Nuts Ohra Foundation.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.