Follow-up after tricuspid valve surgery in adult patients with systemic right ventricles

Transcription

1 European Journal of Cardio-Thoracic Surgery Advance Access published March 16, 2016 European Journal of Cardio-Thoracic Surgery (2016) 1 8 doi: /ejcts/ezw059 ORIGINAL ARTICLE Cite this article as: Koolbergen DR, Ahmed Y, Bouma BJ, Scherptong RWC, Bruggemans EF, Vliegen HW et al. Follow-up after tricuspid valve surgery in adult patients with systemic right ventricles. Eur J Cardiothorac Surg 2016; doi: /ejcts/ezw059. Follow-up after tricuspid valve surgery in adult patients with systemic right ventricles David R. Koolbergen a,b, *, Yunus Ahmed a,b, Berto J. Bouma c, Roderick W.C. Scherptong d, Eline F. Bruggemans a, Hubert W. Vliegen d, Eduard R. Holman d, Barbara J.M. Mulder c and Mark G. Hazekamp a,b CONGENITAL a Department of Cardiothoracic Surgery, Leiden University Medical Center, Leiden, Netherlands b Department of Cardiothoracic Surgery, Academic Medical Center, Amsterdam, Netherlands c Department of Cardiology, Academic Medical Center, Amsterdam, Netherlands d Department of Cardiology, Leiden University Medical Center, Leiden, Netherlands * Corresponding author. Academic Medical Center Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam Zuid-Oost, Netherlands. Tel: ; d.r.koolbergen@amc.nl (D.R. Koolbergen). Received 17 September 2015; received in revised form 10 January 2016; accepted 25 January 2016 Abstract OBJECTIVES: In patients with congenitally corrected transposition of the great arteries (cctga) or after atrial (Mustard or Senning) correction for transposition of the great arteries (actga), the right ventricle (RV) supports the systemic circulation. The tricuspid valve (TV) (systemic atrioventricular valve) is prone to regurgitation in these patients and this is associated with impending RV failure and decreased survival. This study evaluates mid-term functional improvements, echocardiographic findings and survival after TV surgery in this patient group. METHODS: From July 1999 to November 2014, 26 patients (mean age 37.1 ± 12.3 years, 14 females) with cctga (n = 15) or actga (n = 11) had TV surgery. All patients had RV dysfunction and more-than-moderate TV regurgitation (TR); 14 underwent TV replacement (TVR) and 12 had valvuloplasty (TVP). Main outcomes were New York Heart Association (NYHA) functional class, TR and RV dysfunction at 1 year postoperatively and at latest follow-up. Complications and freedom from the composite end-point of death or recurrent TR were analysed. RESULTS: The median follow-up time was 5.9 years (range, years). Mean NYHA functional class significantly improved to 1.7 [95% confidence interval (CI): ] at 1 year (P = 0.004) and was 2.1 (95% CI: ) at latest follow-up (P = 0.14). TV competence significantly improved to a mean TR grade of 1.1 (95% CI: ) at latest follow-up (P < 0.001). The mean grade for RV function at latest follow-up was 2.7 (95% CI: ). Most encountered postoperative complications were arrhythmias and temporary haemodynamic instability due to low cardiac output. Early mortality was 11.5% (n = 3); late mortality was 15.4% (n = 4). Estimated freedom from the composite end-point of death or recurrent TR was 76.9% (95% CI: %) at 1 year and 64.8% (95% CI: %) at 5 years. In TVP patients, TV function at 1 year and at latest follow-up was significantly worse than in TVR patients (P < and P = 0.003, respectively). Also, TVP patients had a significantly lower composite end-point survival curve compared with TVR patients (P = 0.018). CONCLUSIONS: In this patient group, TV surgery showed stabilization of RV function and improvement of NYHA functional class for at least several years. In this series, TVR appears superior to TVP with respect to occurrence of recurrent TR. Early and late mortality after TV surgery is substantial, and we believe that patients with significant TR should be referred earlier for surgery for better outcome. Keywords: Transposition of the great arteries Tricuspid valve Surgery Replacement Valvuloplasty Systemic right ventricle INTRODUCTION In patients with congenitally corrected transposition of the great arteries (cctga) or in patients after atrial level correction (Mustard or Senning procedure) of transposition of the great arteries (actga), the morphological right ventricle (RV) sustains the Presented at the 29th Annual Meeting of the European Association for Cardio- Thoracic Surgery, Amsterdam, Netherlands, 3 7 October systemic circulation. With time, this condition is frequently associated with progressive tricuspid (systemic atrioventricular valve) regurgitation (TR), declining RV function and consequently high incidence of adverse outcome due to congestive heart failure and eventually decreased survival [1 4]. Tricuspid valve (TV) surgery is postulated to improve RV function or reduce deterioration of RV function. In an advanced stage of TR and RV deterioration, the results of any surgical intervention are disappointing [5]. In general, a trend towards earlier surgical intervention can be observed and is The Author Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

2 2 D.R. Koolbergen et al. / European Journal of Cardio-Thoracic Surgery advocated in the guidelines [6]. In selected patients in the younger age groups, a double switch operation (for cctga patients) or conversion of atrial to arterial switch operation (for actga patients) may sometimes be an option with acceptable long-term results [7]. However, adequate retraining of the left ventricle by pulmonary artery banding remains a challenge and results are not always predictable, especially in adult patients [8, 9]. As cardiac transplantation is the last option and is not always available, treatment of TR is the first and main focus of surgical treatment in adult patients. Our previous report from 2009 showed that adult patients with mild to moderate RV dysfunction and important TR benefit from TV replacement (TVR) or plasty (TVP) [10]. In general, TR and functional class improved after surgery. However, after TVP a high recurrent rate of TR was observed which made our threshold for TVR lower than before. The aim of the present study was to report on the longer-term results of our earlier reported patient group, plus on the outcomes of the patients that have been operated for the same indication since then. MATERIALS AND METHODS Study population Patients included in this study were operated either at the Academic Medical Center Amsterdam or at the Leiden University Medical Center. Timeframe for this study was TV surgery from July 1999 until November Patients were diagnosed with cctga (n = 15) or had previously undergone a Mustard or Senning repair for TGA (n = 11). Four patients with cctga, ventricular septum defect (VSD) and pulmonary stenosis or pulmonary atresia had undergone a Rastelli repair with VSD closure and a valved conduit of the morphological left ventricle to the pulmonary artery. Accordingly, all patients had a morphological right systemic ventricle. All patients had more than moderate TR and underwent surgery on their TV, either a valve plasty or a valve replacement, with or without concomitant procedures. Preoperative work-up included echocardiographic analysis of the RV function and TV function. New York Heart Association (NYHA) functional class was determined. To be eligible for surgery, patients had to meet the following criteria: moderate to severe TR (Grade 2 4), Grade 2 or 3 NYHA class functional impairment and sufficiently preserved RV function, Grade 2 or 3 dysfunction at most. Surgical technique Through a median sternotomy, cardiopulmonary bypass was established. Intermittent cold crystalloid cardioplegia was administered after aortic cross-clamping. In cctga, the right atrium and atrium septum were opened to visualize the TV. In case of actga, a right atriotomy, perpendicular to the atrioventricular groove, was performed to visualize the TV. In earlier years, TVP was preferred above primary TVR and was mostly performed in patients in whom TR was mainly a consequence of annular dilatation. For this type of surgery, a classic Carpentier Edwards or the newer Edwards MC3 annuloplasty ring (Edwards Lifesciences, Inc., Irvine, CA, USA) was used. Since 2009, on the basis of disappointing results of TVP, our threshold for valve replacement was lowered. Transoesophageal echocardiography was done in all cases to assess valve competence and RV function. In case of tricuspid valve replacement a CarboMedics bileaflet mechanical prothesis (Sulzer CarboMedics, Inc., Austin, TX, USA) or a St. Jude Medical bileaflet mechanical prosthesis (St. Jude Medical, Inc.,St.Paul,MN,USA)wasused. A Medtronic Mosaic porcine bioprosthesis (Medtronic, Inc., Minneapolis, MN, USA) was used in 1 case. In all cases of TVR, the sub-valvular apparatus was preserved. Data collection and definitions Complete patient histories with data on previous surgical procedures were retrospectively obtained. For TV surgery, surgical technique, type of surgery (repair or replacement) and additional procedures were documented. Postoperative complications were systematically registered until latest follow-up. In order to determine surgical benefit, NYHA functional class, TV function and RV function were assessed on set times: preoperatively, immediate postoperatively, 1 year postoperatively and at latest follow-up. TV competence and RV function were qualitatively assessed based on echocardiographic examination. TV competence was graded from no regurgitation (0), mild (1), moderate (2), moderate to severe (3), to severe regurgitation (4). RV function was graded using a visual grading system that ranged from normal (1), mildly de_pressed (2), moderately depressed (3), to severely depressed (4). Survival time was defined as the period between TV surgery and the last available follow-up date or the occurrence of the composite survival end-point death or recurrent TR. Early mortality was defined as any death occurring within 30 days of surgery or before hospital discharge. Recurrence of TR was defined as TR grade 3 or 4 on echocardiography. Follow-up The closing date for follow-up was September Follow-up was complete in all survivors but 1 patient. This patient was followed up for 8.7 years postoperatively after which she moved abroad. One patient, who had to be reoperated for recurrent TR 2.6 years after the index TV surgery, was followed up until the reoperation. For 4 patients, no echocardiographic examination was available within an acceptable period around 1 year postoperatively. The median followup time in the total study population was 5.9 years (range, years). For latest echocardiographic follow-up data, the median follow-up time was 6.1 years (range, years). Statistical analysis Results are reported as mean with 95% confidence intervals (CIs), median and interquartile range (IQR) or numbers and percentages. Pre- to postoperative changes in NYHA functional class, TR and RV function were tested using the Wilcoxon matched-pair signed-rank test. To examine differences between sub-groups (TVP vs TVR, cctga vs actga) in postoperative NYHA functional class, TR and RV function, the Mann Whitney U-test was used. Estimated freedom from the composite end-point of death or recurrent TR was calculated using the Kaplan Meier method and expressed as percentages with 95% CI. Survival curves for the above sub-groups were compared using the log-rank test. A P-value of less than 0.05 (twosided) was considered statistically significant. Survival analysis was performed using GraphPad Prism 6 (GraphPad Software, Inc., La Jolla, CA, USA). All other analyses were performed with SPSS Statistics 20 (IBM Corp., Armonk, NY, USA).

3 D.R. Koolbergen et al. / European Journal of Cardio-Thoracic Surgery 3 RESULTS A total of 26 patients underwent TV surgery, 15 patients with cctga and 11 with actga. In actga patients, eight concomitant Table 1: Patient characteristics in the cctga group and the actga group cctga (n = 15) actga (n = 11) Age (years), mean ± SD 37.6 ± ± 6.3 Female 8 (53) 6 (55) Previous surgery VSD closure 8 (53) 1 (9) ASD closure 1 (7) 0 (0) TVP 3 (20) 0 (0) TVR 1 (7) 0 (0) PA banding 1 (7) 0 (0) Mustard 0 (0) 7 (64) Senning 0 (0) 4 (36) Rastelli repair 4 (27) 0 (0) Year of TV surgery (27) 2 (18) (53) 6 (55) (20) 3 (27) Type of surgery TVR 9 (60) 5 (45) TVP 6 (40) 6 (55) Concomitant procedures Conduit revision 3 (20) 8 (73) Maze 4 (27) 0 (0) VSD closure 2 (13) 0 (0) PA banding 2 (13) 0 (0) Data are expressed as number of patients (%) unless otherwise indicated. cctga: congenitally corrected transposition of the great arteries; actga: atrial ccorrection for transposition of the great arteries; VSD: ventricular septal defect; ASD: atrial septal defect; TVP: tricuspid valve plasty; TVR: tricuspid valve replacement; PA: pulmonary artery; Maze: cox maze procedure for atrial fibrillation; SD: standard deviation. procedures were performed, all of which were atrial baffle revisions. cctga patients additionally underwent maze procedures, VSD closures and pulmonary artery banding. Patient baseline and operative characteristics are displayed in Table 1. As shown in Fig. 1, in 14 patients, annular dilatation with subsequent malcoaptation was identified as the main mechanism of TR. In cases in whom also substantial structural valve abnormalities like leaflet thickening or valvular fibrosis were present, the valve was primarily replaced with a mechanical prosthesis. Three cctga patients had an Ebstein-like valve anomaly. Up until 2009, TVP was attempted when TR was predominantly a result of annular dilatation and no severe tethering of the leaflets was present. In these cases, an annuloplasty ring was used for repair of annular dilatation with good intraoperative results in all except one. All patients operated after 2009 (n = 9), primarily underwent TVR with a mechanical valve. Operation benefit The mean preoperative NYHA functional class in the total study population was 2.6 (95% CI: ). On echocardiographic evaluation, TR was graded moderately to severely impaired (mean TR grade = 3.5; 95% CI: ) and RV function was observed to be mildly to moderately depressed (mean grade = 2.4; 95% CI: ). Postoperatively, mean NYHA functional class for the total study population significantly improved to 1.7 (95% CI: ) at 1 year (10.9 ± 3.0 months) after surgery (P = 0.004). At latest follow-up (6.9 ± 4.7 years), mean NYHA functional class was graded 2.1 (95% CI: ), which did not constitute a significant improvement when compared with preoperative NYHA class (P = 0.14). At each follow-up time point, NYHA functional class was not significantly different between TVP and TVR patients (Fig. 2) or cctga and actga patients. In the total population, TV function significantly improved after surgery. The mean postoperative TR grade was 0.9 (95% CI: ) immediate postoperatively (P < 0.001), 1.2 (95% CI: ) at 1 year postoperatively (P = 0.001) and remained improved, being 1.1 CONGENITAL Figure 1: Surgical intervention flow chart. A total of 26 patients (15 with cctga, 11 with actga) underwent TV surgery. cctga: congenitally corrected transposition of the great arteries; actga: atrial ccorrection for transposition of the great arteries; TV: tricuspid valve; TVP: tricuspid valve plasty; TVR: tricuspid valve replacement.

4 4 D.R. Koolbergen et al. / European Journal of Cardio-Thoracic Surgery (95% CI: ) at latest follow-up (P < 0.001). Patients after TVP, however, showed significantly worse TV function when compared with patients after TVR (Fig. 3). For TVP patients, the mean TR grade was 2.4 (95% CI: )at1yearand2.1(95%CI: ) at latest follow-up. Six TVP patients developed recurrent TR, despite a successful initial repair of annular dilatation with an annuloplasty ring. In these patients, restrictive septal leaflet motion was identified as the main reason for recurrent TR. In 1 patient, the prosthetic annuloplasty ring caused haemolysis resulting in severe recurrent TR. No significant differences in postoperative TV function were observed between cctga and actga patients. RV function in the total population remained unchanged throughout follow-up, the mean grade being 2.7 (95% CI: ) at latest follow-up. Postoperative RV function was not significantly different between TVP and TVR patients (Fig. 4) or cctga and actga patients. Complications Postoperative complications are listed in (Table 2). The median hospital stay after surgery was 12.0 days (IQR, days). In the postoperative period, 16 patients had arrhythmias, all of whom Figure 2: Follow-up of NYHA functional class. Data are presented as mean with 95% confidence interval. P-values denote significance of improvement after TV surgery separately for patients who underwent TVP and TVR at 10.9 months and latest follow-up, respectively. NYHA: New York Heart Association; TV: tricuspid valve; TVP: tricuspid valve plasty; TVR: tricuspid valve replacement. Figure 4: Follow-up of RV dysfunction. Data are presented as mean with 95% confidence interval. P-values denote significance of improvement after TV surgery separately for patients who underwent TVP and TVR at 10.9 months and latest follow-up, respectively. TV: tricuspid valve; TVP: tricuspid valve plasty; TVR: tricuspid valve replacement; RV: right ventricle. Table 2: Complications according to surgical treatment TVR (n = 14) TVP (n = 12) Total (n = 26) Figure 3: Follow-up of tricuspid valve regurgitation. Data are presented as mean with 95% confidence interval. P-values denote significance of improvement after TV surgery separately for patients who underwent TVP and TVR at 10.9 months and latest follow-up, respectively. For difference in TV regurgitation between TVR and TVP patients: *P < 0.001, **P = TV: tricuspid valve; TVP: tricuspid valve plasty; TVR: tricuspid valve replacement. None 1 (7) 2 (17) 3 (12) Arrhythmia 10 (71) 6 (50) 16 (62) Low cardiac output 6 (43) 5 (42) 11 (42) Renal insufficiency 4 (29) 1 (83) 5 (19) Infection 5 (36) 2 (17) 7 (27) Rethoracotomy 1 (7) 3 (25) 4 (15) Death 3 (21) 4 (33) 7 (27) Complications include all events that occurred in the postoperative period up until the latest follow-up. Data are expressed as number of patients (%). TVP: tricuspid valve plasty; TVR: tricuspid valve replacement.

5 D.R. Koolbergen et al. / European Journal of Cardio-Thoracic Surgery 5 were adequately managed by either pacemaker/internal Cardiac Defibrillator implantation or cardioversion combined with medication. Furthermore, 11 patients suffered from postoperative low cardiac output. These patients were predominately treated with medication, and 3 patients required additional support with an intra-aortic balloon pump. All patients with direct postoperative renal insufficiency (n = 5) were managed with temporary haemodialysis or continuous veno-venous haemofiltration. A total of 4 patients underwent a rethoracotomy, 2 patients had persistent thoracic bleeding and 1 patient developed a stomach bleeding. One patient, who underwent valve replacement with a biological prosthesis, developed prosthetic valve stenosis and consequently valve regurgitation during the course of follow-up. This patient has since been hospitalized due to severe heart failure and is currently listed for heart transplantation. Survival In total, 7 patients died. Early mortality occurred in 3 patients (11.5%), late in 4 (15.4%). Characteristics of the early and late deaths are listed in Table 3. For all patients, the estimated composite end-point survival rate was 76.9% (95% CI: %) at 1 year and 64.8% (95% CI: %) at 5 years. The Kaplan Meier composite end-point survival curves for patients who underwent TVP and those who underwent TVR showed a statistically significant difference between groups, in favour of the TVR group (P = 0.018, Fig. 5). The Kaplan Meier curves showed no difference in the composite end-point between cctga and actga patients. results of the present study confirmed that TVR compared with TVP in this patient group leads to a more reliable result with no visible negative impact on RV function. Notwithstanding these beneficial effects, we have to take a close look at the relatively high early and late mortality rates in this series. Early mortality was adversely affected by one very high-risk redo case with endocarditis and a triple valve replacement including a Bentall procedure, who very unfortunately died after an initial recovery. Late mortality was negatively influenced by 2 cases who developed recurrent TR after an initial repair of the TV and ended up in irreversible heart failure. In general, TVP is preferred over TVR for the management of TR in congenital heart disease with lower rates of early and late mortality [11]. However, Fukuda et al. [12] clearly demonstrated that in patients with functional TR who are treated with an annuloplasty ring alone, severe leaflet tethering predicts a high rate of residual TR. Thus, current annuloplasty techniques may reduce leaflet tethering to a certain level but may not be good enough in patients with severe leaflet tethering. Although we excluded patients with structural leaflet malformations or severe leaflet tethering from CONGENITAL DISCUSSION In this update of a previous study [10], we systematically evaluated the benefit and survival of TV surgery for at least moderate TR in symptomatic patients with a dysfunctional RV in the systemic position. The purpose of this surgical intervention is to improve patients clinical condition and to stop the downward spiral of TR and RV dysfunction mutually aggravating each other, thereby aiming for preservation of the RV function for a certain period. Indeed, the present study confirmed that stabilization of RV function seems to hold in the mid-term and clinical condition is improved for at least several years. Furthermore, the mid-term Figure 5: Kaplan Meier curves for estimated freedom from the composite endpoint of death or recurrent TV regurgitation for patients who underwent TVP versus TVR. Dashed lines denote 95% confidence intervals. TV: tricuspid valve; TVP: tricuspid valve plasty; TVR: tricuspid valve replacement. Table 3: Characteristics of patients who died after TV surgery Subject number Year of surgery Type of surgery Days postop Cause of death Recurrence of TR Early mortality TVR 48 Sepsis No TVP 6 Heart failure No TVR + Bentall + Conduit repl. 33 SCD (thrombosis?) No Late mortality TVP 194 VF/cardiogenic shock No TVR 4096 Aspiration pneumonia No TVP 665 Heart failure Yes TVP 1828 Heart failure Yes TR: tricuspid regurgitation; SCD: sudden cardiac death: VF: ventricular fibrillation; TVP: tricuspid valve plasty; TVR: tricuspid valve replacement.

6 6 D.R. Koolbergen et al. / European Journal of Cardio-Thoracic Surgery repair, a high recurrence rate of TR was observed in the cohort that we have reported on in Probably, for the TV in the systemic RV, more than mild tethering should be a reason to replace the valve after an attempt to repair or maybe no tethering at all should be accepted. Our frequent observation of restrictive septal leaflet motion during redo surgery at least suggests ongoing septal papillary muscle displacement. On the other hand, it cannot be excluded that a better, more durable result could have been obtained with a more restrictive annuloplasty ring. The difficulty to obtain an exact measurement of the tricuspid annular size in general makes it difficult to make strong statements on this topic. However, the unpredictable results made us cautious and the threshold for valve replacement much lower, which have resulted in replacement of all valves since For younger age groups with cctga and Ebsteinoid dysplasia of the TV, an anatomical repair results in significant improvement of TV function, even without additional valvuloplasty [13]. Concisely, all considerations above and the wide variety on functional and structural valve abnormalities and differences in underlying pathologies plead for a strongly individualized surgical approach. In the Mayo Clinics, Mongeon et al. [14] updated the largest series of TV surgery in cctga patients earlier described by Van Son et al. [15] and retrospectively analysed preoperative RV ejection fraction (RVEF) as a predictor of postoperative outcome. They clearly demonstrated that for best long-term results TV surgery should be considered at an early stage, i.e. before the RVEF drops below 40% and the sub-pulmonary ventricle systolic pressure rises above 50 mmhg. It should be noted that they replaced all valves with a prosthesis. In our series, preoperative RV function was graded mildly to moderately depressed and, probably, some patients had an RVEF around or below 40%. Together with the 2 patients with adverse outcome after repair of the TV, this might explain the lower survival rates in our series. Beauchesne et al. [16] have emphasized the problem of late recognition and referral of un-operated adult cctga patients despite symptomatic TR and significant RV dysfunction. They followed a cohort of 44 patients in 66% of whom the first diagnosis was made in adulthood. They concluded that poor preoperative systemic RV function is predictive for the eventual need of heart transplantation and that early surgical intervention (TVR) leads to favourable results. Patient factors may also play a role, as congenital patients tend to underestimate their own functional status. In a cohort of 42 actga patients, Szymanśki et al. [17] found a significant percentage (23.8%) of patients with asymptomatic RV dysfunction. Otherwise, in the same population they also found patients with heart failure symptoms and preserved RV dysfunction, where in most cases the symptoms could be ascribed to the presence of significant TR. In a series of cctga patients, Prieto et al. [18] found strong evidence that long-standing significant TR almost always precedes RV dysfunction. In only 2 patients out of 40, RV dysfunction was clearly not associated with TV dysfunction. They concluded that TR represents the major risk factor for cctga patients and suggested that decisions related to surgical interventions should be strongly influenced by the status of the TV. With respect to the mode of TR, RV failure and eventually heart failure, the clinical course of cctga and actga patients clearly shows similarities although most studies in this field have analysed only one of these groups, making an accurate comparison difficult. In a recent case study, speckle tracking and 3D echocardiographic analysis showed that in a cctga patient, the RV adapted to systemic load by a shift in contraction pattern from normal right to normal left ventricle [19]. In addition to a predominant increase in circumferential over longitudinal strain, a twist was observed although of less magnitude than the normal left ventricle. This ventricular torsion was not observed in the RV of 14 Senning operated patients [20]. In our study, cctga and actga patients showed no difference in the composite end-point of survival or recurrence of TR, but it is important to note that underlying pathological mechanisms are not necessarily the same and may be determined by the ability of the RV to adapt to systemic conditions. Authors from a large follow-up study on 468 Mustard and Senning operations in Sweden and Denmark [4] recently concluded that long-term survival in these patients is primarily determined by TV and RV factors and not the timing or type of surgery in childhood. As no other risk factors could be determined, outcome seemed to depend on how well the TV and RV tolerate or rather adapt to the systemic afterload and blood pressure. However, their survival curves showed that the occurrence of RV failure and need for heart transplantation was equally distributed along the different decades of life indicating strong individual differences. Similarly, the age spectrum of clinical presentation in cctga patients varies widely. Patients in the sixth decade of life have been reported with no signs of RV dysfunction and good clinical performance [15]. Thus, strong individual differences within patient groups hinder the ability to predict long-term outcome and timely identification of patients at risk for heart failure. This emphasizes the need for frequent and close follow-up of these patient groups. Modern magnetic resonance imaging (MRI) techniques may help with the early detection and risk stratification of patients with a dysfunctional systemic RV who are at risk for adverse outcome. In actga patients, who in general have standard clinical follow-up, Rydman et al. [21] were able to demonstrate that RV fibrosis detected by cardiovascular magnetic resonance (CMR) with late gadolinium enhancement (LGE) is strongly associated with adverse clinical outcome. They suggested that the routine use of LGE CMR should be incorporated in follow-up protocols and risk stratification. Van der Bom et al. [22] clearly demonstrated that regular cardiovascular MRI and exercise testing are important tools in the risk assessment of patients with a systemic RV. Using this protocol, they identified an RV end-diastolic volume index above 150 ml/m 2 and peak exercise systolic blood pressure below 180 mmhg as important risk factors that predict a 20-fold higher annual event rate in this population. Study limitations This study is retrospective in nature and due to low incidence of the condition under study has small patient numbers. Therefore, results should be interpreted with caution. Only echocardiographic qualitative analysis was used to access TV and RV function. We are aware that cardiac MRI is far more superior to echocardiography with respect to estimations of RV function. However, an important number of patients had pacemakers and in earlier years, no systematic preoperative and postoperative MRI was done. For completeness, consistency and comparison of follow-up data, we had to fall back on echocardiographic data. CONCLUSIONS In symptomatic patients with a dysfunctional systemic RV, the surgical treatment of TR can result in the stabilization of RV function and improvement of NYHA functional class for at least several years. In

7 D.R. Koolbergen et al. / European Journal of Cardio-Thoracic Surgery 7 this series, TVR appears to be superior to TVP with respect to early postoperative TR as well as to freedom of the composite end-point of death or recurrent TR which may be a reason to lower the threshold for valve replacement. A wide variety in clinical presentation and pathological mechanisms emphasizes the need for frequent and close follow-up of these patients and a strongly individualized treatment strategy. Timely referral, accurate patient selection with the use of modern diagnostic tools and well-timed surgery are probably the key to a better long-term outcome. Conflict of interest: none declared. REFERENCES [1] Graham TP, Bernard YD, Mellen BG, Celermajer D, Baumgartner H, Cetta F et al. Long-term outcome in congenitally corrected transposition of the great arteries: a multi-institutional study. Jac 2000;36: [2] Voskuil M, Hazekamp MG, Kroft LJM, Lubbers WJ, Ottenkamp J, van der Wall EE et al. Postsurgical course of patients with congenitally corrected transposition of the great arteries. Am J Cardiol 2015;83: [3] Bogers AJJC, Head SJ, de Jong PL, Witsenburg M, Kappetein AP. Long term follow up after surgery in congenitally corrected transposition of the great arteries with a right ventricle in the systemic circulation. J Cardiothorac Surg 2010;5:74. [4] Vejlstrup N, Sørensen K, Mattsson E, Thilén U, Kvidal P, Johansson B et al. Long-term outcome of Mustard/Senning correction for transposition of the great arteries in Sweden and DenmarkCLINICAL PERSPECTIVE. Circulation 2015;132: [5] Said SM, Burkhart HM, Schaff HV, Dearani JA. Congenitally corrected transposition of great arteries: surgical options for the failing right ventricle and/or severe tricuspid regurgitation. World J Pediatr Congenit Hear Surg 2011;2: [6] Baumgartner H, Bonhoeffer P, De Groot NMS, De Haan F, Deanfield JE, Galie N et al. ESC Guidelines for the management of grown-up congenital heart disease (new version 2010). Eur Heart J 2010;31: [7] Hiramatsu T, Matsumura G, Konuma T, Yamazaki K, Kurosawa H, Imai Y. Long-term prognosis of double-switch operation for congenitally corrected transposition of the great arteries. Eur J Cardiothorac Surg 2012;42: [8] Devaney EJ, Charpie JR, Ohye RG, Bove EL, Mee RBB, De Leval MR. Combined arterial switch and Senning operation for congenitally corrected transposition of the great arteries: patient selection and intermediate results. J Thorac Cardiovasc Surg 2003;125: [9] Poirier NC, Yu JH, Brizard CP, Mee RBB, Mavroudis C, Fraser CD. Long-term results of left ventricular reconditioning and anatomic correction for systemic right ventricular dysfunction after atrial switch procedures. J Thorac Cardiovasc Surg 2004;127: [10] Scherptong RWC, Vliegen HW, Winter MM, Holman ER, Mulder BJM, Van Der Wall EE et al. Tricuspid valve surgery in adults with a dysfunctional systemic right ventricle repair or replace? Circulation 2009;119: [11] Said SM, Dearani JA, Burkhart HM, Connolly HM, Eidem B, Stensrud PE et al. Management of tricuspid regurgitation in congenital heart disease: is survival better with valve repair? J Thorac Cardiovasc Surg 2014;147: [12] Fukuda S, Song JM, Gillinov AM, McCarthy PM, Daimon M, Kongsaerepong V et al. Tricuspid valve tethering predicts residual tricuspid regurgitation after tricuspid annuloplasty. Circulation 2005;111: [13] Myers PO, Bautista-hernandez V, Baird CW, Emani SM, Marx GR, Nido PJ. Tricuspid regurgitation or Ebsteinoid dysplasia of the tricuspid valve in congenitally corrected transposition: is valvuloplasty necessary at anatomic repair? J Thorac Cardiovasc Surg 2014;147: [14] Mongeon F-P, Connolly HM, Dearani JA, Li Z, Warnes CA. Congenitally corrected transposition of the great arteries ventricular function at the time of systemic atrioventricular valve replacement predicts long-term ventricular function. J Am Coll Cardiol 2011;57: [15] Van Son J, Danielson G, Huhta J, Warnes C, Edwards W, Schaff H et al.late results of systemic atrioventricular valve replacement in corrected transposition. J Thorac Cardiovasc Surg 1995;109: [16] Beauchesne LM, Warnes CA, Connolly HM, Ammash NM, Tajik AJ, Danielson GK. Outcome of the unoperated adult who presents with congenitally corrected transposition of the great arteries. J Am Coll Cardiol 2002;40: [17] Szymanśki P, Klisiewicz A, Lubiszewska B, Lipczynśka M, Michałek P, Janas J et al. Application of classic heart failure definitions of asymptomatic and symptomatic ventricular dysfunction and heart failure symptoms with preserved ejection fraction to patients with systemic right ventricles. Am J Cardiol 2009;104: [18] Prieto LR, Hordof AJ, Secic M, Rosenbaum MS, Gersony WM. Progressive tricuspid valve disease in patients with congenitally corrected transposition of the great arteries. Circulation 1998;98: [19] Sim M-M. Adaptation of the systemic right ventricle in a congenitally corrected transposition of the great arteries. Circulation 2013;127:e [20] Pettersen E, Helle-Valle T, Edvardsen T, Lindberg H, Smith HJ, Smevik B et al. Contraction pattern of the systemic right ventricle. Shift from longitudinal to circumferential shortening and absent global ventricular torsion. J Am Coll Cardiol 2007;49: [21] Rydman R, Gatzoulis MA, Ho SY, Ernst S, Swan L, Li W et al. Systemic right ventricular fibrosis detected by cardiovascular magnetic resonance is associated with clinical outcome, mainly new-onset atrial arrhythmia, in patients after atrial redirection surgery for transposition of the great arteries. Circ Cardiovasc Imaging 2015;8:e [22] Van Der Bom T, Winter MM, Groenink M, Vliegen HW, Pieper PG, Van Dijk APJ et al. Right ventricular end-diastolic volume combined with peak systolic blood pressure during exercise identifies patients at risk for complications in adults with a systemic right ventricle. J Am Coll Cardiol 2013; 62: APPENDIX. CONFERENCE DISCUSSION Scan to your mobile or go to to search for the presentation on the EACTS library Dr T. Tlaskal (Prague, Czech Republic): You have analyzed a very challenging group of patients with systemic right ventricular dysfunction and tricuspid valve insufficiency in congenitally corrected transposition of the great arteries, and after atrial correction of transposition of the great arteries in whom repair or replacement of the tricuspid valve was done. The long-term outcome of these patients is very unpredictable and is not clear so far. However, in your study of 26 patients, only adult patients are involved. In this group of patients, after tricuspid valve surgery, a rather high mortality was found. There was a global close to 27% mortality rate. The long-term follow-up after the surgery in your study has shown improvement of tricuspid valve function and the clinical condition. However, the right ventricle function did not improve. It was more or less stable during follow-up, but it did not improve. So I would like to ask you, as you advocate the possibility of earlier surgery, do you have any experience with earlier treatment of such patients with systemic tricuspid valve regurgitation in the settings of systemic right ventricle, in children or younger adults? Dr Koolbergen: Of course, we know from the studies, especially from the Mayo Clinics, that earlier referral is better. They have found the cutoff point of an ejection fraction of the right ventricle of 40%, and when it drops below that, results will be worse. We only had our echocardiographic data to rely on in this series, and it is very well possible that some of the patients had a right ventricular function, which was too bad, actually. So hopefully, with better, modern assessment techniques like MRI in the future, we will be able to identify those patients who are beyond surgery for that part. For children, it is another story. This is different than what we presented here in the adult patients. In children with serious tricuspid regurgitation and congenitally corrected transposition, we would start with pulmonary artery banding to treat the transient ischemic attack, which usually works perfect by the shift of the septum; while at the same time, the left ventricle stays in good condition, or it is trained; and then we will perform a double switch operation at the age of one year. Dr Tlaskal: My second question: do you have any experience with double switch repair of congenitally corrected transposition or retraining of the left ventricle with subsequent anatomic correction in patients with failing right ventricle and tricuspid regurgitation after atrial correction of transposition? Dr Koolbergen: Yes, we do have some experience, and it is our experience that regarding the possibilities of left ventricular retraining and conversion, that we think that the left ventricle cannot be trained after the age of 5 to 8 years. Nobody knows exactly the cutoff point here, but it is probably something like this. The only exception is, when there is a long-standing pulmonary venous pathway obstruction present in a patient with a Senning or Mustard. Then a CONGENITAL

8 8 D.R. Koolbergen et al. / European Journal of Cardio-Thoracic Surgery conversion to arterial switch can be considered, which we have done in a few cases. Dr Freund (Oldenburg, Germany): Tricuspid valve regurgitation in congenital corrected transposition is not a disease of the valve; it is a disease of the right ventricle in the sub-aortic position. So when you do a repair of the tricuspid valve, perhaps you have to do improve the remodeling of the right ventricle. For example, in old patients, as you do in the young patients, you should perform a pulmonary banding; to have a shift of the interventricular septum, even when you are older, from the left to the right ventricle. Do you consider doing something like this, because some people suggested this? Dr Koolbergen: Yes, although there is no consensus that the tricuspid regurgitation is caused by right ventricular dysfunction. There are also papers that support very strictly that the primary focus is the tricuspid regurgitation. But I hear what you say. Of course, there are methods to address right ventricular dilatation next to valvuloplasty, but we did not perform these in our series. Dr Kaarne (Helsinki, Finland): You showed nicely that the results are a bit unpredictable in this patient group. You had 4 patients with severe dysfunction and severe tricuspid regurgitation, and you operated on them. Did you have any patients which you did not operate on? If you did not offer an operation with residual volume right ventricle dysfunction, a severe dysfunction and a tricuspid regurgitation, and thought that the transplantation is a better option than a valve repair. Also if you had those patients, how do you decide which go to the transplant route and which patients go to valve repair or replacement? Dr Koolbergen: Yes, well, in the Netherlands, the transplant program is very low, so this is an option we cannot always consider. So these are really patients sometimes on waiting lists actually for transplantation, and then with this operation, we try to win time. Eventually, some of those patients will come for heart transplantation eventually. But we hope of course with modern MRI techniques that we can identify the patients who are not suitable for this procedure. Dr Kaarne: You mean that your practice is to offer repair or replacement for all the patients with poor ventricles and severe tricuspid regurgitation? Dr Koolbergen: Yes. Dr D. Barron (Birmingham, UK): I think you are not alone, and I think it is a really important message that the first operation should be valve replacement; and repair does not have a place in these patients. Carpentier showed it many years ago, the very early failure rate. I guess the problem, as you said, is we should refer these patients sooner and this is another important message. But did you have the opportunity to offer these patients earlier surgery, or are they coming to you with impaired function already? Do you wait for symptoms? Dr Koolbergen: I think also our cardiologists actually recently published a paper where they presented that with special MRI techniques and functional testing, you can predict the patients who are at risk for poor outcome. I think awareness now is good, and I think in the future, patients will be referred earlier, and will be operated earlier. Dr Barron: Also any role for cardiac resynchronization therapy, in the impaired ventricles? Dr Koolbergen: Maybe in some selected cases, this was done, but I am not sure. Dr A. Corno (Leicester, UK): As a consequence of your study, what are you doing now? Are you considering valve replacement in all the patients, or are you still considering valve repair first? Second question. Our colleague s cardiologists, are getting quite good with percutaneous tricuspid valve implantation. Would you consider this option in case of failed repair? In other words, do you go now directly to the valve replacement or do you still consider a repair? Dr Koolbergen: Yes, in principle, we go for replacement now, because it has been clearly shown, now that it is not similar as in the left ventricle where you can accept leaflet tethering up to 1 cm and still have good results in the long term. So in these patients, I think you cannot accept any leaflet tethering; but if in patients with only a dilated annulus, we would be very cautious to do a repair, I think. So replacement will be the first treatment options for us. Dr Corno: The other question: in the case of failed repair, would you call your cardiologist for a percutaneous implantation? Dr Koolbergen: We have not done this yet. There is no experience, but this may be an option in the future, yes. Dr T. Karl (St. Petersburg, FL, USA): These results are in opposition to what you would expect in a patient with a leaking mitral valve in a systemic left ventricle. Could you comment on the repair techniques that you have used for the systemic tricuspid valves, and do you think, that would have any bearing on what the outcome might be? There are many new techniques available for tricuspid valve repair, some of them which address leaflet tethering. In the era when you were doing the repairs, what sorts of repairs were they specifically? Dr Koolbergen: Actually, we only treated the annular dilation in these patients with a ring. This was most important and of course, now, you can do things on the subvalvular apparatus to treat the tethering more, and we did this in a lot of Ebstein patients. But in this group, we did not use these techniques. Dr Karl: The other question is, do you think these results would be relevant for single systemic right ventricles, in terms of timing of surgery and replacement, rather than repair. Dr Koolbergen: I think there will be similarities between those groups and also, in these right ventricles, we have to be very cautious with repair. Dr A. Aazami (Mashhad, Iran): I think there are very different pathophysiological mechanisms of tethering for tricuspid valve and mitral valve. The tethering valve in a mitral valve is mainly due to ventricular dilatation. But if you look at a variation in the tricuspid valve, I would say that tethering is partly the result of posterior leaflet dilation. So we should seek a solution to fix posterior segment and posterior leaflets, by its remodeling to reduce the tethering. Also I do not think that tethering is alone a reliable argument to replace the valve.