Long-term results after the réparation à l'étage ventriculaire procedure for transposition of the great arteries and double-outlet right ventricle with pulmonary stenosis

Abstract

Open in new tabDownload slide

OBJECTIVES

The purpose of this study is to describe the long-term results of the ‘réparation à l’étage ventriculaire’ (REV) technique for double-outlet right ventricle and transposition of the great arteries (TGA) with pulmonary stenosis (PS).

METHODS

Between 1980 and 2021, 157 patients underwent a REV procedure (median age and weight: 20.8 months and 7.7 kg). The most frequent anatomical presentation was the association between TGA, ventricular septal defect and PS (n = 116, 73.9%).

RESULTS

Sixty-seven patients (42.7%) underwent a Rashkind procedure, and 67 patients (42.7%) a prior surgical palliation (including 62 systemic-to-pulmonary artery shunts). Resection of the conal septum and/or ventricular septal defect enlargement was performed in 109 patients (69.4%). Thirteen patients (8.3%) died, including 4 during the first postoperative month and 2 after heart transplant. Overall survival at 40 years was 89.3%. Thirty-seven patients (23.6%) required 68 reinterventions on the right ventricular outflow tract (RVOT), including 49 reoperations, with a median delay of 9 years after the REV (8 months to 27 years). Twenty patients (12.7%) underwent RVOT valvulation (16 surgical and 4 interventional). Freedom from RVOT reintervention and reoperation at 40 years were 60.3% and 62.6%, respectively. Four patients (2.5%) required reoperation for left ventricular outflow tract obstruction, with a median delay of 4.8 years.

CONCLUSIONS

The REV procedure is a good alternative for TGA and double-outlet right ventricle with PS patients. Only a quarter of the patients required redo surgery on the RVOT. Reoperations for left ventricular outflow tract obstruction are scarce.

INTRODUCTION

Several surgical techniques can be used to repair anomalies of the ventriculo-arterial connection with pulmonary stenosis (PS). Also named malposition of the great arteries, the common denominator of these anomalies is the need for surgical channelling of the left ventricle (LV) to the aorta and to reconnect the right ventricle (RV) to the pulmonary artery (PA), without obstruction. D-transposition of the great arteries (TGA) with ventricular septal defect (VSD) and left ventricular outflow tract obstruction (LVOTO) and TGA-type double-outlet right ventricle (DORV) with PS can be addressed by the arterial switch operation (ASO) with VSD closure and LVOTO resection, the Rastelli operation, the ‘réparation à l’étage ventriculaire’ (REV) procedure, the Bex–Nikaidoh technique and the half turned conotruncal rotation. Although there is no equivocal evidence that 1 technique is superior to another [1–3], the ASO and Rastelli operations are the preferred options to correct TGA with VSD and LVOTO [4–7]. However, facing the risk of severe aortic stenosis, ASO is contraindicated with severe pulmonary valve hypoplasia. First introduced in 1969 [8, 9], the Rastelli operation involves an extracardiac PA-to-RV prosthetic valved conduit, usually placed leftward of the ascending Aorta. However, long-term results of this technique appear to be far from optimal, due to the extra-anatomic placement of an artificial conduit and the spiral-shaped LV-to-aorta tunnel [10–16]. The (REV) technique, originally described by Pr Yves Lecompte back in 1982 [17], was designed to eliminate the need for an extracardiac conduit, by implanting the PA directly onto the RV [18, 19]. Additionally, the resection of the conal septum, by making the LV-to-aorta connection as direct as possible, is thought to decrease the risk of LVOTO over time [20, 21]. Despite innovative features and satisfactory short- [22–24] and long-term results [25], this technique has not gained large popularity and still struggles to establish itself as a legitimate alternative to the Rastelli operation. By describing our long-term results, we wish to address the misconceptions about the technical aspects and the long-term results of this technique that have been suggested to explain the reluctance towards it.

MATERIALS AND METHODS

Study design

Between January 1980 and December 2022, 157 consecutive patients underwent a REV procedure at Laennec and Necker Sick Children’s Hospitals (Paris, France). Data were collected retrospectively. Both archival paper file and electronic medical records were reviewed while the echocardiographies were reanalysed by a specialized cardiologist anatomist to describe the cardiac phenotype according to the IPCCC-ICD-11 nomenclature.

The protocol was approved by the Clinical Research Ethics Committee of the French Society for Thoracic and Cardiovascular Surgery, approval number CERC-SFCTCV-2022-07-21-21773-IRB00012919. Informed consent was obtained from all patients.

Patients

Forty-eight patients (30,6%) were operated before 2000. The median follow-up was 18,1 years (2 months to 41.3 years). The association between TGA, VSD and LVOTO was the most frequent anatomical presentation (n = 116, 73.9%), followed by DORV TGA-like with PS in 23 patients (14.6%) and DORV TOF-like in 9 patients (5.7%), when the distance between the tricuspid valve and the pulmonary annulus precluded LV-to-aorta channelling without the incorporation of the pulmonary annulus within the baffle, leading to the need of pulmonary relocation (Table 1). A double-outlet LV was present in 7 (4.5%) patients and 2 (1.3%) had a congenitally corrected TGA. The median preoperative pulmonary gradient was 64 mmHg (range: 20–100) and 8 patients (5.1%) had a pulmonary atresia. Coronary anomalies were reported in 36 patients (22.9%).

The median age and weight at operation were 10.2 months (range, 2.1 months to 13.7 years) and 8,2 kg (range, 4.2–25 kg). One hundred and eighteen (75.2%) patients were younger than 2 years old. Sixty-seven patients (42.7%) underwent a Rashkind procedure. Sixty-seven patients (42.7%) had a prior surgical palliation, including 62 systemic-to-PA shunts to improve the pulmonary blood flow (Table 2).

Surgical procedure

The operative technique of the REV has already been extensively described in the literature [11, 17, 20, 22–24] (Video 1 & 2). Resection or mobilization of the conal septum, VSD enlargement and/or infundibular resection was performed in 121 patients (77.1%) while 30 patients (19.1%) were diagnosed with absent or undeveloped conal septum (Table 2). The spatial relationship between the great arteries guided the realization of the Lecompte manoeuvre (n = 146, 93.0%). One hundred and five patients (66.9%) had a monocusp valve implantation to reconstruct the right ventricular outflow tract (RVOT) (all before 2008).

Statistical analysis

Continuous data are expressed as median and min–max values. Categorical data are expressed as number and percentage. Univariate analysis was performed using Fisher’s exact test for categorical data. Kaplan–Meier survival curves with time since REV procedure were plotted using years as the time scale. Data were analysed with the Prism, GraphPad software (version 7, GraphPad Software, La Jolla, CA, USA).

RESULTS

Postoperative course

The median mechanical ventilation and inotropic support were 1 (1–30) and 2 (1–34) days respectively, while the median ICU stay was 5 (1–53) days. Nine (5.7%) patients had a delayed chest closure (median of 4 days). Four (2.5%) patients required postoperative extracorporeal membrane oxygenation support, 3 for haemodynamic instability due to postoperative haemoptysis and respiratory disorders and 1 for arrhythmias caused by a left CA stenosis, who underwent early redo surgery for coronary angioplasty. All but 1 patient are still alive.

Survival

Thirteen patients died (8,3%), of whom 4 (2.5%) during the postoperative course. There were 9 late death, all cardiac-related: 4 sudden deaths, 2 cardiac arrests after sustained arrhythmias and 3 postoperative deaths after redo surgery (2 heart transplants and 1 residual VSD closure). Overall survival was 89.3% [95% confidence interval (CI): 81.8–93.8%] at 20 and 40 years (Fig. 1).

Reinterventions

Ninety-four reinterventions were required in 53 patients. Forty-nine patients required 70 reoperations: 6 patients underwent 3 while 9 patients underwent 2 reoperations.

Right ventricular outflow tract

Thirty-eight patients (24.2%) required 68 reinterventions on the RVOT [including 49 reoperations in 34 patients (21.7%)] with a median delay of 9 years after the REV (8 months - 27 years) (Table 3). Indications for RVOT reintervention were RVOT obstruction (RVOTO) (n = 61, associated with endocarditis in 5) and need for pulmonary valvulation (symptomatic moderate-to-severe pulmonary regurgitation with RV dilatation, n = 7). RVOTO was annular in 41, supravalvular in 13, both annular and supravalvular in 5 and subpulmonar in 2. The median RVOT gradient was 65 mmHg (range 40–105) at reintervention, the majority of the patients with RV pressures were iso- or supra-systemic. Freedom from RVOT reintervention and reoperation were 66.7% (95% CI: 56.3–75.2%) and 69.2% (95% CI: 58.8–77.5%) at 20 years, respectively (Fig. 2). Twenty patients (12.7%) required pulmonary valvulation (16 surgical and 4 interventional) for RVOTO in 14 and pulmonary regurgitation with dilated RV in 6. Freedom from pulmonary valvulation was 85.8% (95% CI: 76.5–91.6%) at 20 years (Fig. 3). The implantation of a monocusp was identified as a risk factor for both RVOT reintervention and reoperation (P = 0.0099 and 0.0026, respectively) but not for pulmonary valvulation (P = 0.2132).

Left ventricular outflow tract

Four patients (2.5%) required reoperation for LVOTO, with a median delay of 4.8 years (range 3.2–13.5 years). Two of the 4 patients required concomitant RVOT relief. Freedom from left ventricular outflow tract (LVOT) reoperation was 96.3% (95% CI: 90.3–98.6%) at 20 years (Fig. 4).

Atrioventricular block

Eight patients (5%) required a pacemaker (PM) implantation for postoperative complete atrioventricular block. Four patients (2.5%) underwent early PM implantation after the REV procedure at a median delay of 2 weeks (range: 13–20 days). Two patients underwent a PM implantation within the month after their first reoperation (1 for residual VSD, 1 for LVOTO relief which required refection of the LV-to-aorta channel). Two patients required late PM implantation 11 and 25 years after the REV procedure.

Clinical status and late cardiologic assessment

Among the 144 survivors, only 6 (4.2%) patients were reported to be mildly symptomatic (class II of the New York Heart Association). All but 4 patients regularly attended school at their age level or had permanent employment. Regular physical activity was reported by 107 patients and 6 patients reported 11 successful pregnancies.

An effort test was carried out in 35 patients, all with normal results and a median VO2max of 75% (range 50–130).

Arrhythmias were reported in 7 patients (flutter/supraventricular tachycardia in 4 and ventricular tachycardia occurred in 3). Two patients had successful ablation of atrial arrhythmias, and 1 patient required a cardioverter implantation to treat ventricular tachycardia.

Long-term echocardiographic data were available for 115 patients (79.9%). Regarding the RV assessment (ejection fraction and Tricuspid Annular Plane Systolic Excursion (TAPSE)), only 4 patients were reported with decreased RV function. Despite a free pulmonary regurgitation is described in all patients who did not require valvulation in a subsequent reintervention, most patients had either no (n = 77/115, 67.0%) or mild-to-moderate (n = 25/115, 21.7%) dilatation, and only 13 patients (of whom 6 had underwent RVOT reintervention) had severe dilatation of the RV. The median RVOT gradients at last follow-up were 34 mmHg (range: 9–100) for the patients who required RVOT reinterventions and 27 mmHg (range: 4–80) in the rest of the cohort. None of the patients have an impaired LV function and only 3 patients have a maximal LVOT gradient over 20 mmHg.

DISCUSSION

The REV procedure

The REV procedure has been introduced as an alternative to the Rastelli procedure when an ASO is not feasible, with expected reduced risks of LV and RV outflow tract obstructions, and has the advantages of avoiding aortic root translocation and coronary transfer as requested in the Bex–Nikaidoh procedure or the double root rotation. To achieve such outcome, it encompasses important surgical stages [11, 17, 20, 22–24]: (i) extensive resection of the conal septum usually through the infundibulotomy, which can be associated with VSD enlargement and/or infundibular resection, in order to establish an adequate and large LV-to-aorta channel; (ii) a tailored LV-to-aorta tunnelization with a patch whose diameter and shape is determined by the distance between the posterior edge of the VSD and the aortic annulus; (iii) extensive pulmonary branch dissection associated with a Lecompte manoeuvre and a cylindrical resection of the ascending Aorta in order to create a tension-free RV-to-PA posterior autologous connection; and (iv) a conduit-free reconstruction of the RVOT using an anterior infundibular patch.

Even if the REV procedure seems to have theoretical advantages over the Rastelli and Bex–Nikaidoh, it lacks popularity worldwide and is underrepresented in the literature, where the larger series comparing surgical outcomes of the different surgical techniques include <10 REV patients representing a small percentage of the patients’ cohort (3–12%) [3, 7, 16]. Seese et al. [7] reported a higher operative mortality after the REV procedure in comparison with the other techniques (11% vs 4%) whereas in Hazekamp et al. [21] series, the late mortality appears to be lower in the REV subgroup. They however all describe significant higher rates of reinterventions on both the RVOT and LVOT in patients who underwent the Rastelli procedure [2, 3, 7, 15, 16, 21]. Although it initially seemed that the REV procedure was limited to France and extended to French-speaking countries, it has gained popularity in Asia with several teams reporting their results on increasingly larger cohorts [1, 13, 26–29]. They all report a low early mortality (<5%) with an overall survival ranging from 87% to 100% at mid and long terms, what is consistent with the results reported in our cohort.

Fate of the right ventricular outflow tract

Right ventricular outflow tract reinterventions and reoperations

One of the main concerns regarding long-term outcome of patients with anomalies of the ventriculo-arterial connection with PS who underwent biventricular repair is the occurrence of further RVOTO. When comparing the several surgical techniques, outcomes are the worst after the Rastelli procedure, up to half of the patients requiring RVOTO reoperations within the 5 years after surgical repair, with reported freedom from RVOTO reoperations reduced to 25% and 20% at 10 and 15 years in most of the series [2, 10, 14, 21]. A much lower prevalence of RVOTO reoperations has been described after the REV procedure [2, 13, 21, 25–29]. In our cohort, RVOT reintervention rate was 25% at 10 years, obstruction being the main indication in 90% of the patients. It encompasses both interventional and surgical reinterventions, regardless of the cause or level of obstruction. It seemed important to analyse the fate of the RVOT as a whole and to take into consideration the long-term implications of every aspect of the REV procedure (shape of the LV-to-aorta channel that can bulge into the RVOT, RV muscular resection beneath the right ventriculotomy, partial ascending aorta resection and direct reconstruction of the RVOT without conduit with or without monocusp and finally, the Lecompte manoeuvre itself with its impact on the pulmonary branches). Obstruction can be seen at any levels if these specific key points are not closely followed at surgery. Obstruction can be managed surgically, but transcatheter-based reinterventions are also an option, even the risk of coronary compression is not negligible. Stenting of the RVOT in REV patients might also lead to subsequent future surgical difficulties, with a stent stack into the sternum. Long-term outcome of REV patients is not only a surgical concern and require management in a congenital/GUCH Heart-Team. In the end, long-term freedom from RVOT reinterventions and reoperations did not really differ from one another, exceeding 60% at 40 years. Long-term RVOT gradients appeared to be really low in our series, such as in An et al. [28, 29] series where the RVOT peak gradient was less than 40 mmHg in 75% of the patients.

Monocusp or no monocusp

The implantation of a monocusp for the RVOT reconstruction, as initially described in the REV procedure [17], was performed in our centre until 2008 but has since been replaced by the insertion of a simple pericardial patch to reconstruct the anterior wall of the RV-to-PA connection. As the RVOT was reconstructed without a valved conduit, the expected aim of the monocusp was to avoid or minimize pulmonary regurgitation. However, it is prone to thickening and degeneration therefore leading to RVOTO and the need for RVOT reintervention [26, 27]. Kim et al. [26] reported that the main reason for PS after the REV procedure was a calcified monocusp in 71% of the reoperated patients. We found a statistically significant risk for both RVOT reintervention and reoperation after implantation of a monocusp, whereas there was no statistically significant difference regarding the need for pulmonary valvulation due to pulmonary regurgitation. Yet, these results should be interpreted carefully as it concerns 2 groups of patients with different timelines and follow-up. Like Weyand et al. [16] stated, we now prefer to reconstruct the RVOT without the insertion of a monocusp valve to decrease the risk of calcification and reintervention, as pulmonary regurgitation seems to be well tolerated in these patients. We believe that the acute angle of take off of the pulmonary trunk in such patients provides a sufficient degree of restriction to counter the adverse effects of the pulmonary regurgitation due to the unvalved RVOT reconstruction, with well-preserved RV function and no to mild–moderate RV dilatation in the vast majority of patients. In a forthcoming study, we aim to compare the fate of the RV and the RVOT (clinical, echocardiographic and Magnetic resonance imaging (MRI)) in REV patients versus Fallot patients who underwent repair with a transannular patch.

Fate of the left ventricular outflow tract

For these malpositions of the great arteries with PS, the Bex–Nikaidoh procedure has been reported to be a very valuable technique, as it places the aortic root directly over the LV. Most authors report no late reoperation for LVOTO [1–3, 21, 30] after this surgical procedure. In the literature, the rate of LVOTO reoperation after the Rastelli procedure varies from 8.7% to 30% [10, 15, 16, 21] whereas it is relatively low after the REV procedure (<5%) [1, 25–29]. Lee et al. even reported a statistically significant difference in LVOTO reoperation rates between the Rastelli and REV procedures (40% vs 4%, P = 0.02). On the contrary, in the European Congenital Heart Surgeons Association multicentre study, higher LVOTO reoperation were found after the REV procedure in comparison to the Rastelli group (16 vs 5%) [21]. It is important to highlight that a systematic resection of the conal septum is advocated and probably realized for the Rastelli procedure in the new era, which should lead to a comparable result. Only 3 (2.5%) of our patients required reoperation for LVOTO and estimated survival without redo surgery on the LVOT was over 96% at 40 years, which is consistent with our centre’s previous publications [25]. We assume that the association of conal septum resection and tailoring of the LV-to-aorta tunnelization patch allows LVOT realignment reducing the risk for further obstruction. Some authors report a freedom from LVOT reoperation of 100% after the REV procedure but in cohorts with shorter follow-ups (5–7 years) [26, 28, 29] whereas Lim et al. [27], in their 25-year experience of the modified REV procedure with a 14-year follow-up, report a 5% risk of LVOT reoperation at 10 years, increasing to 12% at 15, 20 and 25 years. Hu et al. [1] analysed the haemodynamic performance of the reconstructed LVOT after REV, Rastelli and Bex–Nikaidoh procedures. Although none of their patients required LVOTO reoperations, they reported a more physiologically natural flow pattern after in the Bex–Nikaidoh group whereas patients with REV and Rastelli procedures had a bending-shaped LVOT with blood flow turbulences in the LV-to-aorta channel. All but 3 of the survivors in our cohort have no to under 20 mmHg LVOT gradients but assessment of the LVOT channel turbulences should be considered during follow-up and be the subject of another study to try and predict the occurrence of LVOTO after the REV procedure.

Limitations and perspectives

This was a retrospective and single-centre study. It includes 157 patients over a 4-decade period, which results in a long inclusion time for a rather small yearly sample size (<4 patients per year in average). Although it relates the outcome of a large series of patients, it only focuses on a specific surgical technique rather than the overall management of patients with malposition of the great arteries, VSD and PS. In addition to monitoring the occurrence of LV and RV obstructions, assessment of the unvalved RVOT with free pulmonary regurgitation and its consequences on the RV seems necessary. Further evaluations of the RV’s morphology and function via MRI and of the occurrence of arrhythmias or rhythm disturbances are required to assess the long-term outcome of the REV procedure in adulthood.

CONCLUSION

The REV procedure allows relatively low postoperative and acceptable overall mortalities, low RVOT reinterventions and the need for recurrent LVOTO reoperation is exceptional. This surgical technique appears as a reliable surgical correction for TGA or DORV with PS patients and can be considered as the alternate procedure, by specifically trained surgeons, when an ASO with LVOTO relief is not feasible. Optimal technical realization by a trained surgeon mastering the procedure is the key to achieve such results.

Funding

None to declare.

Conflict of interest: none declared.

DATA AVAILABILITY

The data underlying this article will be shared on reasonable request to the corresponding author.

Authors contributions

Margaux Pontailler: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Resources; Software; Writing—original draft; Writing—review & editing. Alexander Moiroux-Sahraoui: Conceptualization; Data curation; Investigation; Writing—original draft. Ségolène Bernheim: Conceptualization; Investigation; Resources. Régis Gaudin: Conceptualization; Investigation; Resources. Lucile Houyel: Resources; Writing—review & editing. Damien Bonnet: Conceptualization; Investigation; Methodology; Resources. Pascal Vouhe: Conceptualization; Investigation; Methodology; Resources. Olivier Raisky: Conceptualization; Investigation; Methodology; Resources; Supervision; Validation; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Prem Sundar Venugopal and the other anonymous reviewer(s) for their contribution to the peer review process of this article.

Presented at the 36th EACTS/ESTS Conference 5-8 October 2022, Milan, Italy.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• ASO

  Arterial switch operation

 
• CI

  Confidence interval

 
• DORV

  Double-outlet right ventricle

 
• LV

  Left ventricle

 
• LVOT

  Left ventricular outflow tract

 
• LVOTO

  Left ventricular outflow tract obstruction

 
• PA

  Pulmonary artery

 
• PM

  Pacemaker

 
• PS

  Pulmonary stenosis

 
• REV

  Réparation à l’étage ventriculaire

 
• RV

  Right ventricle

 
• RVOT

  Right ventricular outflow tract

 
• RVOTO

  RVOT obstruction

 
• TGA

  Transposition of the great arteries

 
• VSD

  Ventricular septal defect