Author + information
- Received June 19, 2014
- Revision received August 14, 2014
- Accepted September 8, 2014
- Published online December 23, 2014.
- Andreas A. Kammerlander, MD∗,
- Beatrice A. Marzluf, MD, MSc†,
- Alexandra Graf, PhD‡,
- Alina Bachmann, MD∗,
- Alfred Kocher, MD§,
- Diana Bonderman, MD∗ and
- Julia Mascherbauer, MD∗∗ ()
- ∗Department of Cardiology, Medical University of Vienna, Vienna General Hospital, Vienna, Austria
- †Department of Thoracic Surgery, Otto Wagner Hospital, Vienna, Austria
- ‡Department of Medical Statistics, Medical University of Vienna, Vienna General Hospital, Vienna, Austria
- §Department of Cardiothoracic Surgery, Medical University of Vienna, Vienna General Hospital, Vienna, Austria
- ↵∗Reprint requests and correspondence
: Dr. Julia Mascherbauer, Department of Cardiology, Medical University of Vienna, University Hospital, Waehringer Guertel 18–20, A-1090 Vienna, Austria.
Background Significant tricuspid regurgitation (TR) late after left heart valve procedure is frequent and associated with increased morbidity. Surgical correction carries a significant mortality risk, whereas the impact of TR on survival in these patients is unclear.
Objectives This study sought to assess the impact of significant TR late after left heart valve procedure.
Methods A total of 539 consecutive patients with previous left heart valve procedure (time interval from valve procedure to enrollment 50 ± 30 months) were prospectively followed for 53 ± 15 months.
Results Significant TR (defined as moderate or greater severity by echocardiography) was present in 91 (17%) patients (65% female). Patients with TR presented with more symptoms (New York Heart Association functional class ≥II 55% vs. 31%), lower glomerular filtration rates (61 ± 19 ml/min vs. 68 ± 18 ml/min), and a higher likelihood of atrial fibrillation (41% vs. 20%), all statistically significant. Right ventricular (RV) systolic function was worse in patients with significant TR (RV fractional area change 43 ± 11% vs. 47 ± 9%, p < 0.001). A total of 117 (22%) patients died during follow-up. By Kaplan-Meier analysis, overall survival was significantly worse in patients with significant TR (log-rank p < 0.001). However, by multivariable Cox analysis, only RV fractional area change, age, left atrial size, diabetes, and previous coronary artery bypass graft procedure were significantly associated with mortality, but not tricuspid regurgitation.
Conclusions RV dysfunction, but not significant TR, is independently associated with survival late after left heart valve procedure.
Although significant tricuspid regurgitation (TR) is a common finding (1), data about its prognostic relevance remain sparse (2–5). Mostly “functional” in nature, TR is the consequence of geometric alterations caused by right ventricular (RV) dilation, distortion of the subvalvular apparatus, tricuspid annular dilation, or a combination of these factors (6). Significant (moderate and severe) functional TR frequently occurs in combination with left-sided myocardial disease (2,3,5), but also accompanies advanced mitral and aortic valve disease (7–9).
According to current recommendations, patients with primary TR without severe RV dysfunction could benefit from surgical correction at the time of left heart valve procedure if TR is severe or the tricuspid annulus is dilated (10,11). In addition, tricuspid procedure might be considered in patients with previous left heart valve intervention and (recurrent) significant TR (10,11). This setting, however, frequently poses a challenge for the treating physician because the surgical risk in these patients may be substantial, but the prognostic impact of TR is not well defined. Discrepant morbidity and mortality rates have been reported (12–15). Table 1 summarizes the small number of previous studies on the impact of TR late after left heart valve procedure. In addition, no data are available regarding the beneficial effect of isolated redo tricuspid valve (TV) procedure in these patients, whereas the procedure carries a significant operative risk, with 30-day mortality rates between 15% (16) and 19% (17).
To clarify the impact of significant TR late after left heart valve procedure on outcome, we performed the present prospective long-term observational study.
Between January 2007 and December 2008, 571 consecutive patients with previous left heart valve procedure presented to our outpatient heart valve clinic and agreed to participate in the present observational study. The left heart valve procedure had to have occurred at least 6 months before inclusion. Thirty-two patients with significant TR were excluded from further analysis because of a pacemaker or an implantable cardioverter-defibrillator (potentially causing or aggravating TR). Hence, 539 patients were eligible for statistical analysis. From study entry, all data were collected prospectively. Patients who underwent tricuspid repair during follow-up were not excluded. According to the study design (noninterventional, purely observational), written informed consent was not demanded. The ethics committee of the Medical University of Vienna approved the study protocol.
Baseline assessment at study entry (as detailed in the preceding text) included medical history, assessment of current medication, physical examination, electrocardiogram, blood tests, and a transthoracic echocardiogram.
The following data were collected:
• Age, sex, body mass index, and body surface area using the Mosteller formula (18)
• Type and number of previous cardiac procedures
• Additive EuroSCORE (19)
• New York Heart Association functional class
• Presence of comorbidities: coronary artery disease (coronary artery stenosis of >50% lumen loss or fractional flow reserve <0.8), previous coronary artery bypass grafting (CABG), chronic obstructive pulmonary disease (long-term use of bronchodilators or use of steroids for lung disease), arterial hypertension (blood pressure ≥140/90 mm Hg at repeated measurements; or use of antihypertensive agents), atrial fibrillation (AF) (present on the electrocardiogram at index examination or verified episode of AF within the last 6 months), hypercholesterolemia (total serum cholesterol ≥240 mg/dl or cholesterol-lowering medication), diabetes (fasting blood glucose level >126 mg/dl or use of antidiabetic medication), estimated glomerular filtration rate (GFR) using the simplified Modification of Diet in Renal Disease formula (20)
• Pacemaker/implantable cardioverter-defibrillator
All transthoracic echocardiography studies were performed by board-certified physicians, using high-end scanners such as the GE Vivid 5 and Vivid 7 (GE Healthcare, Wauwatosa, Wisconsin). The evaluation included M-mode echocardiography, 2-dimensional echocardiography, and conventional and color Doppler ultrasonography according to current recommendations (21–23). Left ventricular (LV) ejection fraction was assessed with the biplane Simpson’s method. RV function was assessed by the percent RV fractional area change, defined as , according to recent recommendations (24). In addition, tricuspid annular plane systolic excursion was measured.
TR was quantified by an integrated approach (Table 2). Echocardiographic parameters used for grading included TV morphology; RV, right atrial, and inferior vena cava size; vena contracta width; proximal flow convergence radius; and hepatic venous flow pattern (21,23,25–27). Moderate and severe TR were considered “significant” TR and were compared with no and mild TR. The graduation into nonsignificant TR and significant TR was chosen to account for inaccuracies as a result of the semiquantitative assessment of TR by echocardiography and has previously been deemed reasonable (5,28).
Continuous data are expressed as mean ± SD or as median with corresponding interquartile range. Categorical variables are presented in percent and/or total numbers. Differences between 2 groups were analyzed using the Wilcoxon rank sum test. Chi-square tests and Fisher exact tests were used for categorical variables. Kaplan-Meier estimates were used to calculate 1-, 2-, 3-, 4-, and 5-year survival rates. Differences between survival curves were analyzed using a log-rank test. Differences in these and all other tests were considered significant at p ≤ 0.05.
A multivariable Cox regression was performed to identify parameters associated with overall mortality. All baseline variables (Table 3) and echocardiographic measurements (Table 4) were assessed. Variables with a significant univariate influence (p < 0.05 in the univariate Cox regression model) were included in the multiple regression analysis with backward selection.
In a previous work (5), a strong interaction of TR and LV systolic function was found. Therefore, data were also tested for significant interactions with TR, particularly with respect to LV and RV systolic function, kidney function, and systolic pulmonary artery pressure. Statistical analyses were performed using SPSS Statistics version 18 (IBM, Armonk, New York) and SAS release 9.2 (SAS Institute, Cary, North Carolina).
Clinical and surgical data
Figure 1 depicts patient baseline characteristics according to type of valve procedure. The mean period from left heart valve procedure to enrollment was 50 ± 30 months. The majority of patients underwent previous aortic valve replacement (65%). The remaining patients had previous isolated mitral valve procedure (19%), combined aortic and mitral valve procedure (10%), or previous TV surgery in addition to any other valve (6%). Significant TR most frequently occurred in patients with previous surgery of the TV (34%), followed by previous combined aortic and mitral valve procedure (25%) and mitral valve procedure alone (17%). In patients with previous aortic valve replacement, significant TR appeared in 14%.
Table 3 shows baseline patient characteristics according to TR severity. Patients with significant TR were more often female (p = 0.004), had lower GFR (p = 0.002), more often presented with AF (p < 0.001), and were more symptomatic (p < 0.001). They also had more previous valve (redo) procedures (p = 0.019).
Table 4 displays echocardiographic data at baseline. TR patients presented with larger left (p = 0.008) and right (p < 0.001) ventricles, larger left (p < 0.001) and right (p < 0.001) atria, and worse RV systolic function (p < 0.001).
Patients were followed for 53 ± 15 months, and all completed follow-up. A total of 117 (22%) patients died during follow-up. Causes of death are given in Table 5. Cardiovascular death was the leading cause of mortality and was more frequent among patients with significant TR (p = 0.045). Only 2 patients had TV repair procedure during follow-up: 1 died perioperatively, but the other survived and remained in the analysis.
By Kaplan-Meier analysis, overall survival was significantly worse in patients with significant TR, with 1-, 3-, and 5-year survival rates of 92%, 81%, and 64%, respectively, compared with 96%, 89%, and 81% in patients without significant TR (log-rank test, p < 0.001) (Figure 2).
Per the results of the univariable and multivariable Cox regression analyses (Table 6), significant TR (p < 0.001), age (p < 0.001), GFR (p < 0.001), coronary artery disease (p = 0.001), previous CABG (p < 0.001), New York Heart Association functional class ≥2 (p = 0.001), chronic obstructive pulmonary disease (p = 0.030), diabetes (p = 0.002), left and right atrial and RV size indexed to body surface area (all p < 0.001), peak TR velocity (p < 0.001), and RV systolic function (p < 0.001) were significantly associated with survival in the univariable analysis.
By multivariable analysis, however, only age, left atrial size, RV systolic function, diabetes, and previous CABG, but not TR, remained significantly associated with outcome.
We also tested our data for potential interactions, particularly with respect to LV and RV systolic function, kidney function, and peak TR velocity. No significant interactions were detected.
The present prospective, long-term, observational study of 571 patients examined the impact of significant TR late after left heart valve procedure. Our data show that RV dysfunction, but not TR, is independently associated with outcome in this setting.
Reoccurrence or new development of significant TR late after left heart valve procedure is found frequently and varies between 9% and 49% (13–15,29,30). These patients represent a challenging population because they often experience severe symptoms, conservative treatment options are limited, and redo surgery carries a high risk (10,16,17).
However, for patients presenting with TR late after left heart valve procedure, the evidence with respect to the impact of TR on outcome remains limited. Song et al. (12) retrospectively investigated more than 600 patients after left heart valve procedure. Late significant TR was associated with worse outcome, defined as a combination of cardiovascular death, need for redo surgery, and hospital admission due to congestive HF. Conversely, Garcia Fuster et al. (13) reported higher frequencies of chronic HF in 801 patients with late significant TR after left heart valve procedure, but no significant increase in mortality. Kwak et al. (14) retrospectively evaluated 335 patients after left heart valve procedure. The endpoint, defined as cardiovascular death and redo surgery, was more frequently reached by patients with significant TR, but only after 10 years of follow-up.
It must be emphasized that all these studies were retrospectively designed and lack assessment of RV function as well as multivariable regression analyses.
Therefore, the present large prospective study adds important information. Our data show that although significant TR was strongly associated with mortality by univariable Cox and Kaplan-Meier analysis (Figure 2, Table 6), only RV function, but not TR, remained associated with outcome in the multivariable model (Table 6).
Recently, Gulati and coworkers (34) reported on the important prognostic value of RV function by cardiac magnetic resonance imaging (CMR) in patients with dilated cardiomyopathy. RV ejection fraction of ≤45% was shown to be significantly associated with adverse outcome. The present data underline the similar importance of RV function for prognosis in patients presenting late after left heart valve procedure. Although assessment of RV function by echocardiography is challenging, an integrated approach including assessment of RV size, comprehensive visual assessment of contractility, and tricuspid annular plane systolic excursion allows the distinction between normal and abnormal RV function with good accuracy (24).
The development or progression of “functional” TR late after left heart valve procedure follows as a consequence of geometric alterations of the RV. It has recently been shown in aortic stenosis patients that after aortic valve replacement, LV hypertrophy decreases, but the degree of diffuse interstitial myocardial fibrosis remains unchanged (35). Diffuse myocardial fibrosis has been linked with diastolic dysfunction and elevated LV end-diastolic pressures, promoting post-capillary pulmonary hypertension (36). The elevation of pulmonary pressure induces RV pressure overload, RV dilation, distortion of the tricuspid valvular apparatus, and finally, significant TR. The assumption that TR late after left heart valve procedure develops secondary to post-capillary pulmonary hypertension is supported by more prominent left atrial dilation in patients with severe TR, and also higher estimated pulmonary artery pressures by echocardiography (Table 4). The Central Illustration displays the pathogenetic processes underlying TR late after left heart valve procedure.
Pulmonary hypertension as a result of left heart disease (Dana Point 2) can cause significant TR, but correcting TR may not be sufficient to modify the outcome of these patients. Because of this, RV function seems to be a more important predictor of outcome. However, from the pathophysiological point of view, a double load on the RV (pressure plus volume related to the presence of TR) will produce consequences. The presence of significant TR leads to more dilation and more RV systolic dysfunction, and through this, to worse outcome. However, when we excluded patients with dilated right heart chambers from the multivariable Cox regression analysis, similar results were obtained.
The guidelines on the management of valvular heart disease currently recommend surgical correction of severe primary TR and of significant TR at the time of left heart valve procedure (10). In patients who present with TR after previous left heart valve procedure, the guidelines propose a Class IIa, Level of Evidence: C recommendation for surgical correction. The present data challenge this recommendation because they imply that TR in these patients is a result of elevated pulmonary pressures, which cannot be modified by TV procedure.
We collected the presented data in a single-center setting; therefore, a center-specific bias cannot be excluded. However, the major advantages of limiting data collection to a single center are: 1) inclusion of a homogenous patient population; 2) adherence to a constant clinical routine; 3) constant quality of echocardiographic work-up; and 4) constant follow-up.
Renal impairment is a common finding in patients with severe TR and RV dysfunction. The majority of our patients presented with renal dysfunction (GFR = 66.3 ± 18.1 ml/min) (Table 3). However, in only 1.9% of patients was GFR <30 ml/min. Thus, our data are not necessarily transposable to patients with severe renal failure. The present study is also limited by the lack of data on rehospitalization for HF. LV volumes and conduction abnormalities were not evaluated.
Accurate determination of RV dysfunction in the presence of significant TR is difficult because RV unloading as a result of TR may be misleading, on top of technical limitations of 2-dimensional echocardiography used to assess the RV. CMR is currently the gold standard for the assessment of RV function. Unfortunately, systematic CMR data are not available for the present population.
As impairment of RV function in our patients seems to be linked with elevated pulmonary artery pressures, caused by high LV filling pressures, invasive hemodynamic data would be of great interest. Unfortunately, systematic right heart catheterization was not performed; thus, hemodynamic data are only available in a small proportion of our patients.
Significant TR late after left heart valve procedure is frequent but, in contrast to RV dysfunction, age, left atrial size, diabetes, and previous CABG, not independently associated with survival.
In light of these results, isolated surgical correction of significant TR late after left heart valve procedure may not be beneficial. Further studies are needed to define specific groups of patients who would clearly benefit from such a procedure.
COMPETENCY IN MEDICAL KNOWLEDGE: Tricuspid regurgitation frequently develops late after left heart valve surgery. However, in contrast to right ventricular dysfunction, age, left atrial size, diabetes, and previous coronary artery bypass procedure, it is not independently associated with reduced survival.
COMPETENCY IN PATIENT CARE: Isolated surgical correction of significant tricuspid regurgitation may not be beneficial late after left heart valve procedure.
TRANSLATIONAL OUTLOOK: Further studies are needed to define specific groups of patients who would benefit from isolated repair of tricuspid regurgitation late after left heart valve procedure.
This study received support from the Austrian Society of Cardiology (to Dr. Mascherbauer), the Österreichischer Herzfonds (to Dr. Mascherbauer), and the Austrian fellowship grant KLI 245 (to Dr. Mascherbauer). The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- atrial fibrillation
- coronary artery bypass graft
- cardiac magnetic resonance imaging
- glomerular filtration rate
- heart failure
- left ventricle/ventricular
- right ventricle/ventricular
- tricuspid regurgitation
- tricuspid valve
- Received June 19, 2014.
- Revision received August 14, 2014.
- Accepted September 8, 2014.
- 2014 American College of Cardiology Foundation
- Hung J.,
- Koelling T.,
- Semigran M.J.,
- Dec G.W.,
- Levine R.A.,
- Di Salvo T.G.
- Nath J.,
- Foster E.,
- Heidenreich P.A.
- Nishimura R.A.,
- Otto C.M.,
- Bonow R.O.,
- et al.
- Song H.,
- Kim M.J.,
- Chung C.H.,
- et al.
- Levey A.S.,
- Coresh J.,
- Greene T.,
- et al.
- Pierard L.A.,
- Moonen M.,
- Lancellotti P.
- Rudski L.G.,
- Lai W.W.,
- Afilalo J.,
- et al.
- Topilsky Y.,
- Khanna A.,
- Le Tourneau T.,
- et al.
- Tribouilloy C.M.,
- Enriquez-Sarano M.,
- Bailey K.R.,
- Tajik A.J.,
- Seward J.B.
- Taramasso M.,
- Vanermen H.,
- Maisano F.,
- Guidotti A.,
- La Canna G.,
- Alfieri O.
- Sagie A.,
- Schwammenthal E.,
- Newell J.B.,
- et al.
- Gulati A.,
- Ismail T.F.,
- Jabbour A.,
- et al.
- Mascherbauer J.,
- Marzluf B.A.,
- Tufaro C.,
- et al.