Long-term outcomes of pulmonary valve replacement in patients with repaired tetralogy of Fallot Free

Abstract

Open in new tabDownload slide

OBJECTIVES

The objectives of this study were to evaluate long-term outcomes of pulmonary valve replacement (PVR) in patients with repaired tetralogy of Fallot (TOF) and to identify the factors associated with adverse clinical events (ACEs).

METHODS

A total of 190 patients who underwent PVR between 1998 and 2015 after repair of TOF were retrospectively analysed. ACE was defined as all-cause death, heart transplantation or new-onset sustained arrhythmia. Univariable Cox proportional hazards regression analysis was used to identify the factors associated with ACE after PVR.

RESULTS

The median age at PVR was 19 years. Preoperative magnetic resonance imaging (MRI) was performed in 143 (75%) patients, and the median right ventricular (RV) end-diastolic and end-systolic volume index was 164 and 82 ml/m², respectively. The follow-up completeness was 94%, and the median follow-up duration was 9.8 years. The transplantation-free survival and freedom from ACE at 15 years was 95% and 90%, respectively. The factors associated with ACE were older age at PVR, older age at TOF repair, New York Heart Association functional class III or IV, presence of tachyarrhythmias, longer cardiopulmonary bypass time and concomitant arrhythmia surgery. In a subgroup analysis of 143 patients with preoperative MRI data, larger RV end-systolic volume index, larger left ventricular end-systolic volume index and lower left ventricular ejection fraction were associated with ACE.

CONCLUSIONS

Long-term outcomes of PVR in patients with repaired TOF were satisfactory. Proactive PVR before the onset of advanced symptoms, tachyarrhythmias and ventricular dysfunction may further improve the long-term survival of this patient population.

INTRODUCTION

Repair of tetralogy of Fallot (TOF) often results in pulmonary regurgitation (PR), which can lead to right ventricular (RV) dilatation, RV dysfunction, left ventricular (LV) dysfunction, arrhythmias and sudden death [1–5]. Pulmonary valve replacement (PVR) is being performed with increasing frequency in patients with repaired TOF to prevent or reverse the deleterious consequences of PR. However, the optimal timing of PVR, especially in the asymptomatic patients, is not clearly defined. Several studies using magnetic resonance imaging (MRI) have addressed this issue by reporting cut-off values of preoperative RV volume for successful RV remodelling after PVR [6–10]. According to these studies, the cut-off RV end-diastolic and end-systolic volume for successful RV remodelling after PVR is ∼160 and 80 ml/m², respectively. These cut-off values are now included in the recently updated American Heart Association/American College of Cardiology guideline for the management of adults with congenital heart disease [11].

Although the indications of PVR have been refined, it remains unknown whether PVR will lead to improved long-term survival. We have previously reported satisfactory mid-term outcomes of PVR in patients with repaired TOF [8]. However, data pertaining to the long-term outcomes are limited in the literature and the previously reported long-term survival rates after PVR are far from satisfactory [12–15]. Also, there is little information regarding predictors of adverse clinical outcomes such as death or sustained ventricular tachycardia (VT) after PVR [9, 10, 15–17].

Since we started to perform PVR in patients with repaired TOF, we have maintained a proactive approach towards early PVR in the hope that this will preserve ventricular function and ultimately lead to improved long-term survival [8, 18]. The objectives of this study were to evaluate long-term outcomes of PVR in patients with repaired TOF and to identify the factors associated with adverse clinical events (ACEs) after PVR.

MATERIALS AND METHODS

Ethical statement

The Institutional Review Board of Sejong General Hospital approved this study and waived the need for individual patient consent.

Study population

For this retrospective single-centre study, we identified 190 patients with repaired TOF and chronic PR who had undergone surgical PVR between January 1998 and December 2015. During the study period, transcatheter PVR was not available in our country. Patients with significant (peak pressure gradient ≥40 mmHg) valvar and/or subvalvar pulmonary stenosis, RV to pulmonary artery conduits and other significant cardiac anomalies (pulmonary atresia, absent pulmonary valve, atrioventricular septal defect, etc.) were excluded. Data were obtained by the review of medical records. When a patient was lost to hospital follow-up, we tried telephone contact with the patient or family to determine the survival status. Characteristics of the study population are summarized in Table 1.

Indications for pulmonary valve replacement

Indications for PVR were symptoms and signs attributable to RV volume overload, presence of significant residual lesions requiring surgical interventions such as branch pulmonary artery stenosis and tricuspid regurgitation. For asymptomatic patients, we have maintained a proactive approach towards PVR as evidenced by the median age of 19 years at PVR (Table 1). Since 2002, we have performed cardiac MRI as a part of routine preoperative evaluation, and 143 (75%) patients have undergone the examination (Table 2). The median RV end-diastolic and end-systolic volume was 164 and 82 ml/m², respectively. During the recent study period, based on the findings of our previous study [8], we used the RV end-diastolic volume of 163 ml/m² and RV end-systolic volume of 80 ml/m² as a guide to assist decisions regarding PVR in the asymptomatic patients.

Surgery

PVR was performed through median sternotomy and using cardiopulmonary bypass with mild hypothermia. Aortic cross-clamping depended on the surgeon’s preference or on concomitant procedures. Choice of prosthetic pulmonary valves was at the discretion of the surgeon. When deemed necessary, RV reduction was performed according to the surgeon’s preference [8]. Table 3 summarizes the surgical data.

Follow-up

The closing date for this study was 10 September 2017. Follow-up was considered complete if the patient’s status was determined after July 2016. Survival status of the patient who could not be followed by the review of medical records or telephone contact (n = 33) was determined by data obtained from the Korea Statistics Promotion Institute, a government organization providing reliable administrative data regarding vital statistics. The median duration of follow-up was 9.8 years (range 1.0–19.6 years), and 94% of hospital survivors had complete follow-up. Follow-up completeness, as measured by calculating the index devised by Clark et al. [19], was 94%. The most recent electrocardiographic, echocardiographic and MRI data were used for analyses. If necessary, patients underwent Holter examinations to detect arrhythmias and cardiopulmonary exercise testing to determine exercise tolerance.

Definitions of outcomes

Prosthetic pulmonary valve dysfunction was defined as a peak pressure gradient ≥40 mmHg or at least moderate PR on the latest echocardiography. New-onset arrhythmia was defined as sustained (lasting >30 s) atrial fibrillation, atrial flutter, atrial tachycardia or VT. ACE was defined as all-cause death, heart transplantation or new-onset arrhythmia. RV normalization was defined as normal RV volume (end-diastolic volume <108 ml/m² and end-systolic volume <47 ml/m²) and function (ejection fraction ≥49%) [20].

Statistical analysis

Data were presented as frequencies with percentages, means with standard deviations or medians with interquartile ranges as appropriate. Comparisons between paired groups were performed using paired t-test or Wilcoxon signed-rank test as appropriate. Discrete variables were analysed using χ² test. Survival and time-to-event analyses were performed using the Kaplan–Meier method. Univariable Cox proportional hazards regression analysis was used to identify the factors associated with ACE after PVR. The proportional hazards assumption was evaluated by log minus log plots and Schoenfeld residual plots. No corrections were made for multiple testing. All statistical analyses were performed using SPSS version 24 (IBM Corporation, Armonk, NY, USA).

RESULTS

Clinical outcomes

There were 2 early deaths (30-day hospital mortality, 1%). A 10-year-old boy, who had undergone TOF repair at the age of 3 years, died of ischaemic brain injury 8 days after the seemingly uneventful operation. He had no intracardiac shunt, and the operation was performed with the aorta cross clamped. A 25-year-old man, who had undergone TOF repair at the age of 5 years, died of pre-existing severe RV dysfunction. At the time of PVR, he was severely symptomatic [New York Heart Association (NYHA) class IV] with severe tricuspid regurgitation and atrial fibrillation. Concomitant tricuspid valve replacement and maze procedure were performed. He could not be weaned from cardiopulmonary bypass because of severe RV dysfunction, and RV assist device support was initiated. He died of uncontrollable bleeding on the day of operation. Two late deaths of unknown cause occurred during follow-up. One patient who had undergone PVR at the age of 20.3 years died 13.1 years postoperatively. The other patient who had undergone PVR at the age of 12.8 years died 8.6 years postoperatively. The overall survival at 10 and 15 years was 98% ± 1% and 96% ± 2%, respectively (Fig. 1A). One patient underwent heart transplantation 9 years after PVR owing to RV failure. This patient, who had undergone TOF repair at the age of 13 months, underwent PVR at the age of 19.3 years. At the time of PVR, the RV end-diastolic and end-systolic volume was 352 and 314 ml/m², respectively. The RV ejection fraction was 11%. The transplantation-free survival at 10 and 15 years was 97% ± 2% and 95% ± 2%, respectively (Fig. 1B). Twenty-four patients underwent repeat PVR surgically. The reasons for repeat PVR were combined pulmonary stenosis and regurgitation (n = 12), pulmonary stenosis (n = 6), endocarditis (n = 3), PR (n = 1) and not available (n = 2). No death occurred after repeat PVR. Freedom from repeat PVR at 10 and 15 years was 88% ± 3% and 76% ± 5%, respectively (Fig. 1C). One hundred eighty-one patients (96% of the hospital survivors) underwent echocardiography at a mean interval of 7.2 ± 4.9 years after PVR. Freedom from repeat PVR and prosthetic pulmonary valve dysfunction at 10 and 15 years was 79% ± 4% and 52% ± 6%, respectively. The patients with tricuspid regurgitation grade ≥ moderate decreased with marginal statistical significance (9% vs 4%, P = 0.060). Seven patients developed new-onset arrhythmias (atrial flutter in 4, atrial fibrillation in 2 and atrial tachycardia in 1). No patient developed VT. QRS duration did not change significantly (n = 112, 147 ± 28 vs 148 ± 23 ms, P = 0.64). Freedom from new-onset arrhythmia at 10 and 15 years was 95% ± 2%. Twelve patients experienced ACE (all-cause death in 4, transplantation in 1 and new-onset arrhythmia in 7), and the freedom from ACE at 10 and 15 years was 92% ± 3% and 90% ± 3%, respectively (Fig. 1D). NYHA functional class improved from 1.6 ± 0.6 to 1.1 ± 0.3 (n = 171, P < 0.001). Cardiopulmonary exercise testing showed no change in peak oxygen consumption (n = 59, 30 ± 7 vs 31 ± 9 ml/kg/min, P = 0.18).

Changes in magnetic resonance imaging parameters

Among the 143 patients who underwent pre-PVR MRI, 90 patients underwent post-PVR MRI after a median interval of 1.6 years (interquartile range 1.0–6.5 years) after PVR. There was significant reduction in RV volumes without significant improvement in RV systolic function (Table 4). RV normalization was achieved in 28 of the 90 (31%) patients.

Factors associated with adverse clinical events after pulmonary valve replacement

Older age at PVR [hazard ratio (HR) 1.01/year, 95% confidence interval (CI) 1.00–1.01; P = 0.030], older age at TOF repair (HR 1.01/year, 95% CI 1.00–1.01; P = 0.020), NYHA functional class III or IV (HR 7.59, 95% CI 2.21–26.03; P = 0.001), presence of atrial or ventricular tachyarrhythmias (HR 6.53, 95% CI 1.76–24.15; P = 0.010), longer cardiopulmonary bypass time (HR 1.01/min, 95% CI 1.00–1.01; P < 0.001) and concomitant arrhythmia surgery (HR 5.53, 95% CI 1.50–20.49; P = 0.010) were associated with ACE after PVR (Table 5). In a subgroup analysis of 143 patients with pre-PVR MRI data, larger RV end-systolic volume (HR 1.01/ml/m², 95% CI 1.00–1.02; P = 0.030), larger LV end-systolic volume (HR 1.06/ml/m², 95% CI 1.01–1.11; P = 0.030) and lower LV ejection fraction (HR 0.91/%, 95% CI 0.84–0.99; P = 0.030) were associated with ACE (Table 5).

DISCUSSION

To our knowledge, this is the largest single-centre study with the longest follow-up duration, which examined long-term outcomes of PVR in patients with repaired TOF. The transplantation-free survival and freedom from ACE at 15 years was 95% and 90%, respectively. Older age at PVR, NYHA functional class III or IV, atrial or ventricular tachyarrhythmias, larger RV and LV end-systolic volumes and LV systolic dysfunction were associated with ACE.

Long-term outcomes of pulmonary valve replacement

Surgical PVR is being performed with increasing frequency in patients with repaired TOF. Cuypers et al. [21], in a longitudinal cohort study consisting of 144 patients with repaired TOF, reported that the cumulative incidence of PVR was 40% at 35 years. Although PVR can lead to the improvement of functional class and successful RV remodelling [6–8], it remains unknown whether such benefits of PVR will translate into improved long-term survival.

Data pertaining to the long-term outcomes of PVR are limited in the literature, and most of the reported outcomes are far from satisfactory (Table 6). Reported survival rates at 15 years after PVR range between 70% and 80% [13–15]. On the contrary, survival of our patients was excellent with 96% at 15 years. This difference in survival may be interpreted in terms of optimal timing of PVR. For instance, Sabate Rotes et al. [15] reported that the mean age at PVR was 31 years and 42% of the patients had advanced symptoms. In their study, survival was 80% at 15 years and NYHA class III/IV was identified as one of the independent risk factors for death. In our study, the median age at PVR was 19 years and only 6% of the patients had NYHA class III/IV. Harrild et al. [13] also reported unsatisfactory outcomes of PVR with freedom from death or VT of only 41% at 10 years. Notably, the mean RV end-diastolic volume before PVR was 196 ml/m², which was much larger than that of our study (164 ml/m²). These differences imply that a proactive approach towards PVR before it is too late may be beneficial in improving long-term survival of the patients with repaired TOF and PR. Although current guideline recommends PVR in asymptomatic patients with severe RV dilatation (RV end-diastolic volume ≥160 ml/m² or RV end-systolic volume ≥80 ml/m²) [11], we believe that proactive PVR before these cut-off values are exceeded may translate into better clinical outcomes in these patients. To this end, judicious follow-up of the patients may be important. Wald et al. [25] reported that the optimal time interval between MRI examinations for the detection of rapidly progressive RV dilatation and dysfunction was a 3-year interval.

One of the concerned aspects of a proactive approach is an issue of repeat PVR due to structural valve deterioration of the bioprosthetic valves [26, 27]. In our study, the freedom from repeat PVR at 15 years was 76%, which was similar to those of previous studies (70–75% at 15 years) [14, 15, 23]. Other valve substitutes such as polytetrafluoroethylene valves used in some of our patients may be promising [28]. Importantly, the role of transcatheter PVR for dysfunctional surgical bioprostheses is expanding and changing the treatment paradigm for patients with PR [29]. We have to emphasize that the operative mortality (1%) was not negligible, and this should be considered when assessing the risk/benefit balance of PVR.

Predictors of adverse clinical outcomes after pulmonary valve replacement

Predictors of adverse clinical outcomes after PVR such as death, VT, heart failure or supraventricular tachycardia have been reported in previous studies [9, 10, 15–17, 30]. These include older age at PVR, older age at TOF repair, NYHA class III/IV, severe tricuspid regurgitation, larger RV end-systolic volume, RV or LV systolic dysfunction and so on. These are in line with those identified in our study. Bokma et al. [9] reported that patients with preoperative RV end-systolic volume >95 ml/m² were at increased risk for ACE, and this was probably the first study showing an association between larger preoperative RV volume and adverse clinical outcomes after PVR. Geva et al. [17] found that age at PVR ≥28 years was one of the predictors of death and sustained VT after PVR. Of note, we found that atrial/ventricular tachyarrhythmias, concomitant arrhythmia surgery and longer cardiopulmonary bypass time were associated with ACE. These seem to be closely related to variables, which can act as confounders for outcomes. Not being able to perform multivariable analysis to account for confounding because of an insufficient number of ACE is a significant limitation of our study. According to the study by Khairy et al. [2], the prevalence of atrial/ventricular tachyarrhythmias increases as the age of the patients increases. Again, all these findings support the potentially beneficial role of proactive PVR in improving long-term survival of the patients with repaired TOF and PR.

Limitations

The present study was limited by its retrospective nature, absence of a control group and incomplete follow-up (94%). The causes of 2 late deaths, which were ascertained by data obtained from the Korea Statistics Promotion Institute, were unknown. Also, information regarding other outcomes (heart transplantation, new-onset arrhythmia, etc.) was not available in 33 patients whose vital status had been ascertained through the Korea Statistics Promotion Institute. Multivariable analysis to identify predictors of ACE was not performed due to an insufficient number of ACE in our patients.

CONCLUSIONS

Long-term outcomes of surgical PVR in patients with repaired TOF were satisfactory. Proactive PVR before the onset of advanced symptoms, tachyarrhythmias and ventricular dysfunction may further improve the long-term survival of this patient population.

Conflict of interest: none declared.

Author contributions

Cheul Lee: Conceptualization; Data curation; Formal analysis; Methodology; Writing—original draft. Eun Seok Choi: Data curation; Writing—review & editing. Chang-Ha Lee: Writing—review & editing.

REFERENCES

Abbreviations

 
• ACE

  Adverse clinical event

 
• CI

  Confidence interval

 
• HR

  Hazard ratio

 
• MRI

  Magnetic resonance imaging

 
• NYHA

  New York Heart Association

 
• PR

  Pulmonary regurgitation

 
• PVR

  Pulmonary valve replacement

 
• RV

  Right ventricular

 
• TOF

  Tetralogy of Fallot