Author + information
- Received September 20, 2016
- Revision received January 19, 2017
- Accepted January 27, 2017
- Published online April 3, 2017.
- Daniel Tanase, MDa,∗ (, )
- Peter Ewert, MD, PhDa,
- Stanimir Georgiev, MDa,
- Christian Meierhofer, MDa,
- Jelena Pabst von Ohain, MD, PhDb,
- Doff B. McElhinney, MDc,
- Alfred Hager, MD, PhDa,
- Andreas Kühn, MDa and
- Andreas Eicken, MD, PhDa
- aDepartment of Paediatric Cardiology and Congenital Heart Defects, German Heart Centre of the Technical University Munich, Munich, Germany
- bDepartment of Cardiovascular Surgery, German Heart Centre of the Technical University Munich, Munich, Germany
- cLucille Packard Children’s Hospital Stanford, Palo Alto, California
- ↵∗Address for correspondence:
Dr. Daniel Tanase, German Heart Centre of the Technical University Munich, Department of Paediatric Cardiology and Congenital Heart Defects, Lazarettstrasse 36, D-80636 Munich, Germany.
Objectives This study sought to investigate the impact of tricuspid regurgitation (TR) on right ventricular function after percutaneous pulmonary valve implantation (PPVI).
Background PPVI provides a less invasive alternative to surgery in patients with right ventricular-to-pulmonary artery (RV-PA) conduit dysfunction. Recovery of the right ventricle has been described after PPVI for patients with pulmonary stenosis and for those with pulmonary regurgitation. Additional TR enforces RV dysfunction by supplemental volume overload. Limited data are available on the potential of the right ventricle to recover in such a specific hemodynamic situation.
Methods In a matched cohort study, we compared patients who underwent PPVI with additional TR with those without TR.
Results The degree of TR improved in 83% of the patients. In our patients (n = 36) exercise capacity and right ventricular volume index improved similarly 6 months after PPVI in patients with and without important TR. None of them had significant TR in the long-term follow-up of median 78 months.
Conclusions PPVI improves not only RV-PA-conduit dysfunction, but also concomitant TR. In patients with a dysfunctional RV-PA conduit and TR, the decision whether to fix TR should be postponed after PPVI.
In patients with right ventricular outflow tract (RVOT) conduit dysfunction, percutaneous pulmonary valve implantation (PPVI) is a less invasive therapeutic option than repeated cardiac surgery (1). PPVI proved to be safe and effective in several studies and at present is the preferred treatment option for RVOT conduit dysfunction in many centers (2–5). In prior studies, beneficial effects of PPVI were documented irrespective of the prevailing hemodynamic indication: RVOT stenosis, pulmonary regurgitation (PR), or a combination of stenosis and regurgitation (6,7). Sustained improvements in hemodynamics were shown up to 7 years post-PPVI (8).
In addition to RVOT dysfunction, some patients referred for PPVI present with moderate to severe tricuspid valve regurgitation (TR). This may be caused by various factors, including a congenitally dysplastic tricuspid valve, RV pressure and/or volume overload, or sequela of preceding surgical procedures. Limited data are available on the potential of the RV to recover in such a specific hemodynamic situation. Current guidelines suggest a surgical approach if RVOT dysfunction is combined with moderate to severe TR (9,10). However, if restoring RVOT function by PPVI improves right heart hemodynamics significantly, the clinical impact of TR may be reduced and surgery may thus be avoidable (11). Therefore, the objective of the present matched cohort study was to investigate the impact of TR on right ventricular function after PPVI.
Between December 2006 and December 2014, a total of 173 patients underwent PPVI at the German Heart Center Munich. Twenty-two of these patients had moderate or severe TR according to echocardiographic criteria (12) before PPVI and constitute the study cohort (TR cohort, TR patients). Patients were evaluated with a standardized protocol including the following: history, echocardiography, cardiac magnetic resonance imaging (CMR), cardiopulmonary exercise testing, and assessment of New York Heart Association functional class. For every patient with significant TR, a matched control subject was selected with comparable clinical findings but without additional TR (control cohort). According to study protocols, all cases and control subjects were evaluated before PPVI and 6 months after PPVI.
Control subjects were matched to patients with TR according to the following criteria. In detail, we made sure to compare patients: 1) having the same pulmonary valve pathology, either pulmonary stenosis (PS), PR, or a combination of both; 2) with similar indexed right ventricular end-diastolic volume (RVEDVi); and 3) with New York Heart Association functional class state that was the same or differed no more than 1 class. Additionally, the underlying cardiac diagnosis, sex, age at PPVI, and number of previous surgeries were also taken into consideration. PS was defined as a peak invasive right ventricular-to-pulmonary artery (RV-PA) gradient more than 40 mm Hg, and PR was defined as a regurgitant fraction more than 20% by CMR. Patients with mixed valve disease (i.e., meeting both PS and PR criteria) were compared with similar patients with mixed disease. For statistical evaluation, patients with mixed disease were analyzed either with the PS or PR cohorts, depending on which was the dominant pathophysiology.
In 4 patients it was not possible to find a matched pair, in 3 because the RVEDVi differed significantly and in 1 because of a unique congenital heart defect (congenitally corrected transposition of the great arteries). Those 4 patients were excluded from further evaluation. Thus, the study group consisted of 18 patients with pulmonary valve dysfunction and significant TR, who were compared with 18 matched control subjects treated by PPVI who did not have relevant TR.
Cardiopulmonary exercise testing
All patients underwent a symptom-limited cardiopulmonary exercise test on an electronically braked bicycle in a sitting position according to international guidelines (13). Patients cycled according to a ramped protocol with a 3-min warm-up at 0 W followed by a rampwise increase of load with 10, 15, 20, or 30 W/min with the aim of reaching an exercise duration of 8 to 12 min. They were encouraged to exercise until exhaustion. Oxygen uptake was measured breath by breath by a metabolic cart (Vmax 229, SensorMedics, Viasys Healthcare, Yorba Linda, California). Peak oxygen uptake was defined as the mean oxygen uptake in the 30-s period that was highest throughout the exercise test.
Cardiac magnetic resonance imaging
CMR was performed at 1.5 T (Symphony Maestro Series and Avanto, Siemens Medical Solutions, Erlangen, Germany). Retrospective gated cine CMRs of the heart were acquired in the vertical long-axis, 4-chamber, and short-axis views that included the extent of both ventricles, and 2 long-axis planes of the RVOT for through-plane flow quantification. Aortic and PA flow data were acquired with a flow-sensitive gradient echo sequence during free breathing. Regurgitant fractions were calculated from forward and backward flow across the valve. For volume measurements of the RV, endocardial contours were traced, excluding papillary muscles, trabeculae, and the moderator band from the blood volume.
Data were analyzed with a standard statistical package (SPSS version 22.0, SPSS Inc., Chicago, Illinois). All continuous variables were expressed as median and minimum-maximum. Pre- versus post-PPVI data were analyzed with the 2-tailed Mann-Whitney U test. For continuous data, patients with pulmonary valve dysfunction and additional TR were compared with control subjects using a 2-tailed paired Student t test. For statistical analysis of the changes in the degree of TR, the Wilcoxon signed rank test was performed. Statistical significance was inferred when p < 0.05.
The median age of all cases and control subjects included in the study was 21 years (mean, 22.2 ± 7.9 years), the youngest patient being 10 years and the oldest 39 years of age. The group comprised 14 female and 22 male patients. Overall, 18 patients had an underlying cardiovascular diagnosis of tetralogy of Fallot, 9 had pulmonary atresia with ventricular septal defect (n = 9), 8 had truncus arteriosus (n = 8), and 1 had pulmonary atresia with intact ventricular septum (n = 1). Individual patient data are listed in Table 1.
All patients had undergone a median of 2 prior surgical interventions, and all had undergone corrective surgery with RV-PA conduit. Most of the patients were in New York Heart Association functional class II (n = 32), whereas 3 were in functional class I and 1 was class III.
The median peak oxygen uptake (Vo2 peak) was 28.5 ml O2/kg/min (mean, 27.5 ± 7.4 O2/kg/min), and ranged from 15 to 44 ml O2/kg/min. Among patients referred for PPVI because of predominant PR, the median regurgitant fraction calculated by CMR was 33% (21% to 43%). Patients with prevailing PS had a median RV-PA peak systolic pressure gradient at cardiac catheterization of 46 mm Hg (27 to 83 mm Hg). The median central venous pressure (measured in the right atrium) was 8 mm Hg (3 to 17 mm Hg), with no difference between patients with and without TR (p = 0.29). The median RVEDVi of all patients included in the study was 100 ml/m2 (mean, 109.3 ± 34.4 ml/m2) with a range of 61 to 185 ml/m2.
Peak systolic pressure gradients and pulmonary regurgitation
PPVI was performed successfully in all patients, with no significant procedural adverse events. Patients with pulmonary valve pathology and TR received PPVI at a median age of 22 years (10 to 34 years), which was similar to those without TR, whose median age was 22.5 years (13 to 39 years; p = 0.19).
In patients with predominant PR, the regurgitant fraction was significantly reduced after PPVI, from a median of 33% (21% to 43%) to 1% (0% to 3%; p < 0.001). In patients with PS, the peak invasive RV-PA systolic pressure gradient was reduced significantly from 46 ± 12 mm Hg to 14 ± 5 mm Hg (p < 0.001). Comparing patients with TR with control subjects showed no significant difference in the reduction of PR or PS.
Echocardiographic evaluation of TR
In all patients, echocardiographic data were evaluated for at least 8 years before PPVI. In all of them, TR developed gradually over time, with absent or trivial TR initially that increased to mild and later to moderate or severe. None of them had a sudden increase of TR after surgery.
After PPVI, the degree of TR improved in 15 of 18 (83%) study patients (p < 0.001) 6 months after implant. In 3 of 18 (16%) study patients it remained unchanged. Patients in the TR cohort were followed for a median of 6.5 years (8 months to 9.3 years) after PPVI, and at the latest follow-up visit, none had significant TR. In 15 patients, follow-up TR was trivial and in 3 it was mild. In 1 patient whose TR initially improved to trivial, it later become mild in the context of recurrent stenosis of the melody valve. Detailed data are presented in Table 1 and Figure 1.
Right ventricular end-diastolic volume indices
Patients in the TR cohort had a median RVEDVi before PPVI of 100 ml/m2 (71 to 182 ml/m2). After 6 months, RVEDVi decreased significantly to 88 ml/m2 (60 to 152 ml/m2; p < 0.001). Patients without TR also had a median pre-PPVI RVEDVi of 100 ml/m2 (61 to 185 ml/m2), which decreased to 80 ml/m2 (59 to 162 ml/m2; p < 0.001). Complete results are depicted in Table 1. There was no difference in RVEDVi between patients with and without TR, either at baseline or after PPVI (baseline, p = 0.63; after PPVI, p = 0.20). There was also no difference between groups in the magnitude of decrease in RVEDVi (p = 0.57). In TR patients, RVEDVi decreased a median of 16 ml/m2 (3 to 65 ml/m2) and in patients without TR it decreased by a median of 15 ml/m2 (2 to 78 ml/m2). The results are listed in Table 2 and depicted in Figure 2.
Compared with baseline, functional parameters improved in all patients 6 months after PPVI. Pre-PPVI median Vo2 peak in index patients and matched pairs was 28.4 ml O2/kg/min (15.1 to 44 ml O2/kg/min). This improved significantly to 30 ml O2/kg/min (20 to 49 ml O2/kg/min; p < 0.001). The median pre-PPVI mean work load was 2.3 W/kg (1 to 3.4 W/kg) and improved significantly to 2.5 W/kg (1.3 to 3.7 W/kg; p < 0.001). The difference in Vo2 between patients with and without TR did not differ at baseline (p = 0.96) or after PPVI (p = 0.41).
To assess the relationship between baseline TR and functional outcome, the improvement of peak Vo2 was compared between patients with significant TR and control subjects. In patients with TR, the Vo2 peak improved from a pre-PPVI median of 25.5 ml O2/kg/min (16 to 44 ml O2/kg/min) to 27.5 ml O2/kg/min (20.4 to 49 ml O2/kg/min; p = 0.009). In control subjects, only the median peak Vo2 improved from 29.4 (15.1 to 36) to 31 (20 to 43.6) ml O2/kg (p = 0.001). Despite the fact that in both groups the Vo2 peak improved 6 months after PPVI, the improvements were similar and not significantly different between groups (p = 0.32). These results are depicted in Figure 2.
Similar results were observed regarding improvement in workload. Before PPVI, the median workload in patients with TR was 2.0 (1.2 to 3.5) W/kg, which increased to 2.2 (1.5 to 3.6) W/kg after PPVI (p = 0.001). In control subjects without TR, median workload increased from 2.5 (1 to 3.4) W/kg to 2.6 (1.3 to 3.7) W/kg, (p = 0.001). As shown in Table 2, there were no significant differences in workload between patients and control subjects either at baseline (p = 0.56) or after PPVI (p = 0.52), and no difference in the amount improvement (p = 0.85).
This matched cohort study shows that in most patients with RVOT conduit dysfunction and moderate to severe TR, PPVI leads to a sustained reduction in TR. The increase in exercise tolerance and reduction of RVEDVi after PPVI did not differ according to the presence or absence of significant baseline TR.
In the setting of increased afterload from PS or volume overload from PR, the RV undergoes a series of adaptations to maintain stroke volume. The potential of the RV to recover when abnormal loading conditions are relieved has been reported for patients with PS and/or PR (6,14). In addition to dysfunction of the RV-PA conduit, some patients present with important TR, which may be secondary/functional and the result of adaptive mechanisms of the RV (15), a sequela of previous surgical procedures, or related to a primary congenital abnormality of the tricuspid valve. Additional RV volume overload caused by TR can lead to a further rightward shift on the Frank-Starling curve. The potential for recovery of the RV in that specific hemodynamic condition has not been reported so far.
In our study population, exercise capacity and RVEDVi improved equally in patients with and without important TR. Following the current guidelines, surgery is recommended for treatment of RVOT dysfunction with concomitant moderate to severe TR (9,10). Surgery enables addressing both the pulmonary and tricuspid valves directly. The hypothesis of the current study was that treatment of RVOT dysfunction with PPVI, relieving PS, and providing a competent pulmonary valve would improve right heart hemodynamics and mechanics, with consequent reduction in the severity and clinical impact of TR. If this hypothesis were true, surgery could be postponed or potentially avoided altogether.
Do we have to address secondary TR?
Symptomatic patients with severe primary tricuspid valve dysfunction need surgery. However, in cases of a secondary TR, there is evidence that treatment of the underlying cause improves tricuspid valve function. For example, patients with severe pulmonary arterial hypertension caused by mitral regurgitation or other left heart disease often show regression of TR and almost one-third of them experience complete resolution of TR following mitral valve surgery without direct intervention on the tricuspid valve (16). Because of its functional physiology, secondary TR may diminish or disappear with improvement of right ventricular function and loading. In the context of right atrial or RV dilation caused by an atrial septal defect, amelioration of functional TR after interventional or surgical repair was described (17). In patients with at least moderate TR, significant improvement in tricuspid valve function occurs after surgical pulmonary valve replacement (PVR), irrespective of concomitant tricuspid valve annuloplasty (18).
In this series, TR gradually developed over years in all patients and was not related to surgery. This suggests that the mechanism might be associated with hemodynamic changes in the pressure or volume loaded RV. After PPVI, TR improved in most patients. The RVEDVi decreased significantly, which may have improved remodeling of the RV. The hemodynamic improvements and RV remodeling seemed to positively influence tricuspid valve function. Indeed, the larger RVEDVi after PPVI in the TR group suggests that the RV remained volume loaded in patients who had concomitant TR at baseline. This seems to be supported by the exercise testing findings, with higher workload and peak Vo2 in patients without TR. However, even if differences were minimal and in 3 patients (16%) TR did not improve immediately after PPVI, none of the patients had significant TR during long-term follow-up. These results are in concordance with a recent study that documented sustained hemodynamic improvement of tricuspid valve function after PPVI at midterm follow-up (11).
Do patients with RVOT dysfunction benefit from concomitant surgery on the tricuspid valve?
The hypothetical advantage of surgery over PPVI in patients with RVOT dysfunction and significant TR is that there is the possibility to repair both dysfunctional valves at the same time. But in some patients with severe TR who undergo tricuspid annuloplasty, significant TR remains (19). Tricuspid valve repair usually is associated with a low perioperative risk (20). However, when reconstruction fails or is not feasible, valve replacement becomes inevitable. Compared with tricuspid valve repair, valve replacement is associated with reduced late survival (21). Additionally, in patients with repaired tetralogy of Fallot and both RVOT dysfunction and at least moderate TR, significant improvement in RV size and function occurs after PVR with or without tricuspid valve repair. This suggests that tricuspid valve repair at the time of PVR may not be superior to PVR alone (22).
Bokma et al. (23) compared the results of PVR alone with those with concomitant tricuspid annuloplasty in patients with repaired tetralogy of Fallot and severe TR. Irrespective of early post-operative TR reduction, patients with severe preoperative TR were at high risk for adverse clinical events after PVR in long-term follow-up. This may reflect the fact that the presence of TR is a marker of more advanced disease and an indication for PPVI or PVR to prevent irreversible RV damage. However, meticulous evaluation is needed to rule out patients with dysplastic tricuspid valves, because addressing just the pulmonary valve is a potentially deficient strategy in such patients.
When should we treat RVOT dysfunction?
The optimal timing for PVR is still under debate because no controlled, randomized studies are available. In 1 study, RV size did not return to normal after PVR in patients with a RVEDVi more than 170 ml/m2 (24). Oosterhof et al. (25), reported that RV volume returned to normal if PVR was performed before the RVEDVi reached 160 ml/m2. Another large study reported that RV remodeling was possible in patients <17.5 years of age with a RVEDVi <150 ml/m2 (26). In contrast to these data, it was shown that in young patients with a mean age of 14 years, the RVEDVi normalized after PVR even if RVEDVi exceeded 200 ml/m2 before surgery (27). This suggests that the potential of the RV to remodel decreases with age and that there is not completely fixed “point of no return” for RV dilation. In our patients, PPVI was performed at a mean RVEDVi of 111 ml/m2 at a mean age of 21 years. Compared with published surgical series, the patients included in our study were older and had less severe RV dilation. This may be explained in part by the method of assessing the RVEDV in CMR. Exclusion of papillary muscles, trabeculae, and the moderator band from the RV volume automatically results in smaller volume indices. However, PPVI has improved functional outcome even in patients with preserved RV function and underlines the earlier timing of PPVI (28).
This is a retrospective matched cohort study with a relatively small number of patients, and thus the findings cannot be generalized. Also, the study might be underpowered and accordingly subject to type II error. In addition, the patients in this series with significant TR generally had modest functional impairment and RV enlargement at baseline, and do not represent the more severe end of the clinical spectrum of patients with RVOT dysfunction and RV volume overload related to PR and TR. Accordingly, the findings of this study may not reflect expected outcomes in patients with more severe RV dilation and/or symptoms of right heart failure.
This study shows that in patients with RVOT conduit dysfunction and moderate to severe TR, PPVI leads to improvement in TR in most cases. The beneficial effects of PPVI, including improved clinical symptoms and exercise tolerance, along with RV remodeling, were similar in patients with and without significant baseline TR. This study supports the idea of primary catheter intervention in cases of RVOT dysfunction and secondary TR. However, if significant TR persists, close clinical surveillance is indicated to prevent irreversible damage of the RV.
WHAT IS KNOWN? PPVI is a less invasive therapeutic option in patients with RVOT conduit dysfunction than repeated cardiac surgery. Current guidelines suggest a surgical approach if RVOT dysfunction is combined with moderate to severe TR.
WHAT IS NEW? In patients who underwent PPVI with additional TR, degree of TR improved in 83% 6 months after the intervention. None of them had significant TR in the long-term follow-up.
WHAT IS NEXT? In patients with a dysfunctional RV-PA conduit and additional TR, the decision whether to fix TR should be postponed after PPVI if TR is secondary in nature.
Dr. Ewert has received personal fees (proctor) for the Medtronic Melody valve; and is a proctor for Edwards pulmonic transcatheter valve. Dr. McElhinney is proctor and consultant for Medtronic. Dr. Hager has received speakers honoraria from AbbottVirology Encysive GmbH, Pfizer, Actelion, Abbott, Medtronic, Schiller Medizintechnik GmbH, GlaxoSmithKline, AOP, Orphan Pharmaceuticals AG, and OMT; honoraria for writing informational material from Actelion; travel compensations from Braun, Guidant, Arrows, Medtronic, Actelion, GlaxoSmithKline, Pfizer, Lilly, AOP, and Orphan Pharmaceuticals AG; is a shareholder of Johnson & Johnson, Gilead, Merck Sharp & Dohme Inc., Pfizer, Medtronic, and Roche; his institution contributed to company-driven clinical trials from Actelion, Medtronic, Edwards, Occlutec, Novartis, and Lilly; and his institution received unrestricted scientific grants for investigator initiated trials from Pfizer, GlaxoSmithKline, Abbott, Actelion, and Medtronic. Dr. Eicken is a proctor for the Medtronic Melody valve. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- cardiac magnetic resonance imaging
- percutaneous pulmonary valve implantation
- pulmonary regurgitation
- pulmonary stenosis
- pulmonary valve replacement
- right ventricle to pulmonary artery
- right ventricular end-diastolic volume index
- right ventricular outflow tract
- tricuspid regurgitation
- Vo2 peak
- peak oxygen uptake
- Received September 20, 2016.
- Revision received January 19, 2017.
- Accepted January 27, 2017.
- 2017 American College of Cardiology Foundation
- Lurz P.,
- Coats L.,
- Khambadkone S.,
- et al.
- McElhinney D.B.,
- Hellenbrand W.E.,
- Zahn E.M.,
- et al.
- Eicken A.,
- Ewert P.,
- Hager A.,
- et al.
- Coats L.,
- Khambadkone S.,
- Derrick G.,
- et al.
- Cheatham J.P.,
- Hellenbrand W.E.,
- Zahn E.M.,
- et al.
- Baumgartner H.,
- Bonhoeffer P.,
- De Groot N.M.,
- et al.
- Nishimura R.A.,
- Otto C.M.,
- Bonow R.O.,
- et al.
- Jones T.K.,
- Rome J.J.,
- Armstrong A.K.,
- et al.
- Lancellotti P.,
- Tribouilloy C.,
- Hagendorff A.,
- et al.,
- for the Scientific document committee of the European Association of Cardiovascular Imaging
- Balady G.J.,
- Arena R.,
- Sietsema K.,
- et al.
- Coats L.,
- Khambadkone S.,
- Derrick G.,
- et al.
- Kogon B.,
- Mori M.,
- Alsoufi B.,
- et al.
- Guenther T.,
- Mazzitellli D.,
- Noebauer C.,
- et al.
- Bokma J.P.,
- Winter M.M.,
- Oosterhof T.,
- et al.
- Oosterhof T.,
- van Straten,
- Vliegen H.W.,
- et al.
- Frigiola A.,
- Tsang V.,
- Bull C.,
- et al.
- Buechel E.R.,
- Dave H.H.,
- Kellenberger C.J.,
- et al.
- Borik S.,
- Crean A.,
- Horlick E.,
- et al.