Prognostic validation of partition values for quantitative parameters to grade functional tricuspid regurgitation severity by conventional echocardiography

Abstract

Aims

Quantitative echocardiography parameters are seldom used to grade tricuspid regurgitation (TR) severity due to relative paucity of validation studies and lack of prognostic data. To assess the relationship between TR severity and the composite endpoint of death and hospitalization for congestive heart failure (CHF); and to identify the threshold values of vena contracta width (VCavg), effective regurgitant orifice area (EROA), regurgitant volume (RegVol), and regurgitant fraction (RegFr) to define low, intermediate, and high-risk TR based on patients’ outcome data.

Methods and results

A cohort of 296 patients with at least mild TR underwent 2D, 3D, and Doppler echocardiography. We built statistical models (adjusted for age, NYHA class, left ventricular ejection fraction, and pulmonary artery systolic pressure) for VCavg, EROA, RegVol, and RegFr to study their relationships with the hazard of outcome. The tertiles of the derived hazard values defined the threshold values of the quantitative parameters for TR severity grading. During 47-month follow-up, 32 deaths and 72 CHF occurred. Event-free rate was 14%, 48%, and 93% in patients with severe, moderate, and mild TR, respectively. Severe TR was graded as VCavg > 6 mm, EROA > 0.30 cm², RegVol > 30 mL, and RegF > 45%.

Conclusion

This outcome study demonstrates the prognostic value of quantitative parameters of TR severity and provides prognostically meaningful threshold values to grade TR severity in low, intermediate, and high risk.

Introduction

The presence of haemodynamically significant tricuspid regurgitation (TR), either isolated or in combination with left heart disease, is associated with increased mortality,^1–4 reduced functional capacity,⁵ and end-organ dysfunction.^6,7 The incidence of TR appears to be rising along with age,⁸ prevalence of atrial fibrillation,^9–11 and the use of intra-cardiac devices in patients with heart failure or increased risk for arrhythmias.¹² Lately, severe TR has prompted interest in corrective surgical and transcatheter procedures.¹³ However, the appropriate selection of patients addressed for surgical or transcatheter procedures requires an objective and reproducible quantification of TR severity, using a widespread grading system able to predict patients’ prognosis.

Current American¹⁴ and European¹⁵ guidelines recommend transthoracic echocardiography to assess patients with TR, as it is widely available and it provides information regarding tricuspid valve morphology and function, including the presence of structural abnormalities of the leaflets, measurement of tricuspid annulus diameter and evaluation of TR severity. Current guidelines advocate for a multi-parametric approach, integrating data from both semi-quantitative and quantitative parameters obtained by two-dimensional (2DE) and Doppler echocardiography.^16,17 A vena contracta (VC) width ≥ 0.7 cm, an effective regurgitant orifice area (EROA) ≥ 0.40 cm² and a regurgitant volume (RegVol) ≥ 45 mL currently qualify TR as severe.^16,17 However, the partition values to define mild, moderate, or severe the TR for the diameter of the VC (<0.3, 0.31–0.69, and ≥ 0.7 cm, respectively) and the EROA (<0.2, 0.2–0.39, ≥0.4 cm², respectively) have been derived either in analogy to those used for the mitral valve (e.g. EROA partition values),¹⁸ or cross validated with EROA,¹⁹ without either being validated on different cohorts of patients or assessing their prognostic value. For the RegVol, a single cut-off value of ≥45 mL has been recommended to indicate severe TR, but it is unclear how it has been derived.

The aims of this outcome study were: (i) to assess the relationship between TR severity and the composite endpoint of death and hospitalization for congestive heart failure (CHF) and (ii) to identify the threshold values of the quantitative echocardiographic parameters to define low-, intermediate-, and high-risk TR based on patient outcome data.

Methods

Study population

We retrospectively analysed prospectively acquired echocardiographic studies performed from July 2015 to December 2017. Inclusion criteria were: presence of at least mild TR, and a complete 2DE, three-dimensional (3DE) and Doppler echocardiographic study allowing the assessment of all guideline-recommended semiquantitative and quantitative parameters of TR severity.^16,17 Exclusion criteria were: age <18 years, primary TR, repeated examinations, poor quality echocardiography studies, or lack of follow-up data. We finally selected a cohort of 296 patients (46% men; median age 58 years) with various heart diseases (Table 1). This retrospective analysis of clinically acquired data was approved by the Ethics Committee of the Istituto Auxologico Italiano, IRCCS. The need for patient written informed consent was waived due to the retrospective nature of the study. The data underlying this article will be shared on reasonable request to the corresponding author.

2D and 3D transthoracic echocardiography

Image acquisition

Comprehensive 2DE, Doppler, and 3DE studies were performed using a Vivid E9 scanner (GE Vingmed, Horten, Norway), equipped with M5S and 4V probes. Acquisitions of the TR jets by colour Doppler were obtained from the right ventricle (RV) inflow, parasternal short-axis, and apical four-chamber views. Three (five in patients with atrial fibrillation) consecutive cardiac cycles were recorded during breath-hold to minimize respiratory changes.

At the end of the 2DE and Doppler study, four- to six-beat (three- to four-beat in patients with atrial fibrillation) electrocardiogram-gated full-volume 3DE data sets of the right atrium and of the RV were obtained from the RV-focused apical 4-chamber view using the 4V matrix-array transducer, taking care to encompass the entire cardiac structure of interest in the data set and to adjust depth and volume size in order to optimize spatial and temporal data set resolution.^20–22 Data sets were stored digitally and exported to a separate workstation for offline analysis.

Image analysis

All 2DE, Doppler, and 3DE data sets were analysed offline, using EchoPAC 203 (GE Vingmed, Horten, Norway). The severity of TR was graded as mild, moderate, or severe using an integrative and semi-quantitative approach, as recommended.^16,17 Quantitative parameters used to assess TR severity were measured (Figure 1): (i) VC width, as an average of measurements obtained from both the apical 4-chamber and parasternal RV inflow views to account for non-circular tricuspid regurgitant orifice,¹⁹ in mid-systole (VC_avg); (ii) EROA, calculated using the formula $EROA=2πPISA rad2×Va/Vmax$⁠, where PISArad is the radius of the proximal isovelocity surface area, Va is the velocity of the aliasing, and Vmax is the maximal velocity of the TR jet measured by continuous wave Doppler; (iii) RegVol, calculated as$RegVol=EROA×VTITR$⁠, where VTI_TR represents the time–velocity integral of the continuouswave Doppler TR jet; and (iv) regurgitant fraction (RegFr), calculated as the ratio between the RegVol and the RV stroke volume obtained by subtracting the RV end-systolic volume from the RV end-diastolic volume. Tricuspid annulus diameter was measured from the RV-focused view, at end-diastole (the frame before TV closure).²³ Pulmonary artery systolic pressure (PASP) was derived from the maximal TR jet velocity and right atrial pressure estimated by the response of the inferior vena cava to a sniff.²⁴ RV end-diastolic volume, end-systolic volume and ejection fraction were quantified using 4D AutoRVQ software package (GE Vingmed Ultrasound, Norway). Right atrial volume was measured using 4D AutoLVQ software package (GE Vingmed Ultrasound, Norway). All the other measurements were analysed according to the current recommendations.²⁴ In patients with atrial fibrillation, we selected the cardiac cycles with preceding and pre- preceding RR intervals within 6 ms of each other, and both exceeding 500 ms in duration.²⁵

Follow-up

Information concerning survival and adverse cardiac events were obtained at regular intervals via: (i) telephone interviews with the patients, or if deceased, with family members; (ii) contact with the patient’s physician(s); and (iii) review of electronic medical records of regular outpatient visits and hospital admission records. Mortality status was verified independently through the Social Security Death Index and death certificates. Cause of death was determined based on a review of death certificates, post-mortem exam reports when available, medical records for patients who died while hospitalized, and contact of patient’s physician(s). CHF requiring hospitalization was diagnosed on the basis of Framingham criteria.²⁶ Briefly, the diagnosis of CHF required two major or one major and two minor criteria to be present concurrently. The major criteria for CHF included elevated jugular venous pressure, right heart distension by imaging, hepato-jugular reflux, and signs of visceral congestion. Minor criteria included: bilateral ankle oedema, dyspnoea on ordinary exertion, hepatomegaly, pleural effusion, or weight loss ≥4.5 kg in 5 days in response to treatment of CHF. Assignment of clinical events was performed by physicians unaware of TR severity. The primary endpoint was the occurrence of either death or CHF.

Statistical analysis

Patients were grouped according to the occurrence of either CHF and/or death, or to event-free survival during follow-up. Continuous variables were expressed as medians (interquartile range), and categorical variables were described as number of patients (percentages). We used the Kruskal–Wallis test for intergroup baseline characteristic comparison of continuous variables, whereas the Pearson’s χ² test was used for comparison of categorical variables. In the first analysis, we used a primary composite outcome of death and CHF, whereas the second one focused on CHF events only. We will refer to the first analysis as ‘composite outcome’ and to the second one as ‘CHF’. Time-to-event was modelled using multi-variable Cox proportional hazards models. We first implemented a model that included known risk factors for developing the composite outcome, such as age, NYHA class, left ventricular ejection fraction (LVEF), PASP, and severity of TR graded using the multi-parametric approach, as recommended by current guidelines.^16,17 Then, we built one model for each quantitative parameter recommended to assess TR severity (i.e. VC_avg, EROA, RegVol, and RegFr),^16,17 adjusted for NYHA, LVEF, age, and PASP. The relationships between the hazard of the outcome and the various quantitative parameters of TR severity were flexibly modelled using restricted cubic spline with four knots. ANOVA was used to test for both the association and the presence of non-linear relationship between the predictors and the outcome. Proportional hazard assumptions were tested evaluating the correlation coefficients between transformed survival times and scaled Schoenfeld residuals for each covariate in the model. We evaluated the predictive performance of the various models with the quantitative TR parameters and the model with the guidelines-derived multi-parametric grading of TR severity using Somer’s $Dxy$ discrimination index and the slope of the calibration line. The indexes were adjusted with the ‘0.632’ method to corrected for ‘optimism’ in the final model.²⁷ Model’s coefficients were shrunk by a factor equal to the slope of the calibration line to adjust for over-fitting.²⁸ The effects of the various quantitative parameters used to grade TR severity were further depicted on the logarithmic scale of the relative hazard with relative 95% confidence intervals (CIs) for both the composite outcome and the CHF models. The relative hazard values on logarithmic scale were derived from the fitted Cox models given a plausible range of the parameter values, while keeping constant the sum of all the other variables. Since there was a continuous, smooth (i.e. without significant changes in slope) relationship between severity of TR and events, we used the tertiles of the derived logarithmic hazard values to identify the threshold values of the quantitative parameters to grade the severity of TR in mild (parameter value lower than the first tertile of the logarithmic hazard), moderate (parameter value ranging between the first and second tertiles of logarithmic hazard), and severe (parameter higher than the value corresponding to the second tertile of the logarithmic hazard). The presence of differences between the different grades of TR severity in terms of relative logarithmic hazards was jointly tested using χ² test for both CHF and the composite outcome, adjusting for age, NYHA, LVEF, and PASP. P-values were adjusted with Bonferroni correction to avoid multiplicity issues. All data were analysed using STATA 15.1 version (StataCorp. College Station, TX, USA), IBM SPSS Statistics Version 1.0.0-2437 (1.0.0-2437), and R 3.5.0 (R Foundation for Statistical Computing, Vienna, Austria).

Results

Of the 296 patients included in the study, 136 (46%) were men. According to current guidelines, 196 patients (66%) were graded as mild, 56 (19%) as moderate, and 44 (15%) as severe TR. Median follow-up was 47.5 (16.5–80) months. At the end of the follow-up period, we recorded 32 (11%) deaths and 72 (24%) CHF episodes.

Demographics, clinical, and echocardiographic characteristics of study patients

Patients who died and/or were hospitalized for CHF were older (∼76 vs. 50 years old), had higher prevalence of atrial fibrillation, were more symptomatic (i.e. higher NYHA class), and had slightly lower LVEF than patients who remained event-free during follow-up (Table 1). PASP was higher, RV function was lower, right atrial volume was larger, and TR regurgitation was more severe in patients who experienced death or CHF than in patients who remained event-free during follow-up (Table 1). Accordingly, all quantitative parameters of TR severity (VC_avg, EROA, RegVol, and RegFr) were larger in patients who met the composite endpoint than in patients who remained free of events (Table 1). Conversely, prevalence of moderate/severe valvular heart diseases other than TR were similar in patients with and without events (Table 1). Prevalence of RV dysfunction (i.e. RV ejection fraction < 45%) was higher in patients who met the composite endpoint than in patients who remained event-free (53% vs. 22%, respectively; P < 0.001). The prevalence of diabetes (8.3%), systemic hypertension (24.8%), and other conditions such as chronic obstructive pulmonary disease (7.6%), pulmonary artery hypertension (2%), ischaemic heart disease (5.6%), cardiomyopathies (3.2%), and heart failure with both reduced and preserved ejection fraction (18.7%) was similar in patients who met the composite endpoint and in patients who remained event-free.

TR severity and outcome

The effects of the multi-variable Cox model including the guideline-derived multi-parametric assessment for the severity of TR as one of the covariates were summarized in Table 2. Ageing, higher NYHA class, and more severe TR were all associated with a higher risk of experiencing both CHF and the composite outcome (Table 2). In this cohort of patients, LVEF had no significant impact on patients’ outcome. Table 3 shows the effects of each quantitative parameter of TR severity on outcomes by comparing the relative hazards at the III with the I tertile of the quantitative parameters of TR severity distributions. All the quantitative parameters of TR severity studied were associated with a significantly higher risk for either the CHF or the composite outcome (Table 3 and Supplementary data online, Table S1). The relationships of the TR parameters with the hazard of experiencing the event on logarithmic scale were significantly non-linear (P-values < 0.05). All the variables used in the Cox models did not violate the proportional hazard assumptions, as suggested by the P-values >0.05 (Supplementary data online, Tables S2 and S3).

Somer’s D_xy and the slope index suggested that models including VC_avg, EROA, RegVol, and RegFr had good discriminating abilities and were well calibrated (Supplementary data online, Table S1). Moreover, their performances were similar to those observed for the model with the guideline-derived severity categories.

To assess the relationship of the echocardiographic parameters with outcome, different logarithmic relative hazard curves were evaluated, showing that, for both CHF (Figure 2) and composite outcome (Figure 3), there was a continuous, direct relationship between the quantitative parameters and the relative hazards of experiencing the event. Derived cut-off values for VC_avg, EROA, RegVol, and RegFr to grade TR severity are shown in Table 4. Figures 4 and 5 show the survival curves corresponding to each cut-off value of TR severity for CHF and the composite outcome, respectively. Survival curves were derived from the multi-variable Cox models at different plausible follow-up times and adjusting for age, NYHA class, LVEF, and PASP. The plots show lower survival probabilities for patients with more severe TR. The relative logarithmic hazards of TR severity levels were significantly different (P-values < 0.05) for both CHF and the composite outcome (Figures 4 and 5), suggesting a clear separation of the hazards at the derived cut-off values for VC_avg, EROA, RegVol, and RegFr.

Discussion

To the best of our knowledge, this is the first study reporting the partition values of quantitative parameters used to grade TR severity based on the associated mortality and occurrence of CHF. The results of our study can be summarized as: (i) TR severity confirms to be independently associated with mortality and morbidity; (ii) for the first time, we calculated the RegFr as the ratio between the RegVol measured with the PISA method and the RV stroke volume measured with 3DE, and provided outcome-based threshold values to grade TR in mild, moderate, and severe using this parameter; (iii) we used models that include single quantitative parameters to grade TR severity are able to stratify the risk of both CHF and the composite outcome; (iv) the predictive power of such models was similar among the various quantitative parameters of TR severity and comparable with the model that includes the multi-parametric assessment of TR severity recommended by current guidelines; (v) the cut-off values of the various quantitative parameters of TR severity that categorize patients into low, intermediate, and high risk to develop either CHF or the composite clinical endpoint are lower than those reported in current guidelines.^16,17

TR severity and outcome

Defining the impact of TR on patient outcome is a difficult task, because TR (particularly secondary TR) is often associated with a variety of clinical conditions like pulmonary hypertension, previous cardiac surgery, heart valve diseases, atrial fibrillation, left ventricular dysfunction and other, that act as confounders when attempting to identify the specific impact of TR on outcome. Accordingly, available evidence was quite controversial, with several studies reporting the independent prognostic role of TR^1,3,29,30 opposed to others.^31,32 However, a recent meta-analysis including >32 500 patients enrolled in 70 studies, showed that both moderate and severe TR were associated with increased cardiac and all-cause mortality, independently on PASP, LVEF, and RV function.³³ Moreover, the study confirmed that the increase in mortality was not associated with severe TR only, but it showed that even moderate TR was associated with a worse outcome compared with mild TR.³³ These findings were confirmed in a large observational study by Topilsky et al.³⁴ showing that, in patients with isolated TR, the risk of death (adjusted for age, sex, LVEF, atrial fibrillation, and Charlson’s comorbidity index) was increased (HR= 1.68; 95% CI 1.04–2.6; P = 0.03). Moreover, they confirmed that, not only severe TR, but also moderate isolated TR is associated with a significant risk of death with a 52±5%, 31±5%, and 29±5% survival at 5, 10, and 15 years, respectively.

On the other end, Montalto et al.³⁵ documented the ominous effect of more than moderate TR on venous congestion of the visceral organs and showed that renal and hepatic venous congestion contribute to excess morbidity and mortality associated with significant TR.

However, the interpretation and clinical applicability of the above reported studies is hindered by the lack of quantitative assessment of TR severity.^36,37 Among the 70 studies included in the meta-analysis by Wang et al.³³, only one had used a quantitative assessment of TR severity.

Our data add to current literature by confirming the independent prognostic power of TR about morbidity and mortality, and showing a significant, direct relationships among the various quantitative parameters used to quantify TR severity and the risk of both death and hospitalization for CHF.

Quantitative assessment of TR severity

Quantification of TR severity is objectively challenging due to its dynamicity throughout the respiratory cycle, dependency on loading conditions, peculiar haemodynamic environment, and unpredictable regurgitant orifice geometry.³⁸ Despite these limitations, quantitative parameters of TR severity obtained with Doppler echocardiography (e.g. EROA by PISA method) have been reported to be powerful independent predictors of outcome, superior to the recommended multi-parametric qualitative assessment.^3,29 However, quantitative variables have not been implemented neither in the clinical routine nor in research due to paucity of validation studies and lack of prognostic data.³⁷

In our study, using either CHF or the composite endpoint of death and CHF as outcomes, we computed separate logarithmic hazards for VC_avg, EROA, RegVol, and RegFr. The relationships of the quantitative TR parameters with the hazard of experiencing the event on logarithmic scale were significantly non-linear. The models using quantitative parameters to grade TR severity showed good calibration and discriminating ability, i.e. they assign higher risks to patients that experienced the event and lower risks to those that did not experience the event. Furthermore, the performances of the models using quantitative echocardiography parameters were similar to the model with TR guidelines-derived severity categories. In addition, we found significant differences by tertiles, and we were able to define threshold values of the quantitative parameters to grade TR into three categories, mild, moderate, and severe, based on clinical outcome.

By using patient outcome data as reference, we found that the threshold values to define severe TR were > 6 mm, > 0.30 cm², > 30 mL, and > 45% for VCavg, EROA, RegVol, and RegFr, respectively. Interestingly, these cut-off values were significantly smaller than those recommended by current guidelines.^16,17 Our results are consistent with recent data by Dahou et al.³⁹ who showed that in 115 patients who underwent echocardiographic evaluation of TR severity in the context of transcatheter treatment of TR, the best cut-off for severe TR by EROA calculated with proximal isovelocity surface area method was ≥ 0.34 cm². Moreover, Bartko et al.⁴⁰ reported that TR EROA ≥ 0.2 cm² and RegVol ≥ 20 mL were both associated with decreased survival in 382 patients with heart failure and reduced LVEF.

Our study adds to the literature on TR grading by showing the association of the partition values of the quantitative parameters used to define moderate TR with patient outcome, and by providing the partition values for RegFr to be used to grade TR severity. The former is becoming clinically relevant as data are accumulating about the increase in mortality^33,34,40 and morbidity,^29,35 associated with moderate TR and not only to severe TR. The latter parameter (i.e. RegFr) may be particularly useful in patients with secondary TR since absolute metrics (e.g. VC_avg, EROA, RegVol) are heavily dependent on loading conditions and they increase by intra-vascular volume loading, salt intake, or intra-venous fluids, or may decrease after aggressive diuretic treatment. However, RegFr was not introduced in current guidelines on the assessment of TR severity, because the calculation of RegFr by conventional 2DE and Doppler is quite challenging.^16,17 In our study, we have shown an alternative way to obtain the RegFr as the ratio between the RegVol measured with the PISA method and the RV stroke volume measured with 3DE.²¹

The issue of grading the severity of TR

There may be several explanations for the apparent discrepancies between the data reported in guidelines^16,17 and the results of recent studies,^39,40 including the present one. Threshold values of quantitative parameters of TR severity included in current guidelines have been adapted from those developed for the mitral valve,¹⁸ validated in only few comparative, small-size studies,^18,19 and their prognostic value has never been tested. However, the peculiar anatomical and haemodynamic characteristics of TR suggest that the same metrics cannot be used interchangeably for the left and right atrio-ventricular valves.³⁸

The non-circular and unpredictable geometry of TR orifice may explain the underestimation of TR EROA using the PISA method. Chen et al.⁴¹ showed that PISA EROA systematically underestimated VC area planimetered using 3DE and that the underestimation was related to the ellipticity of the TR orifice. Song et al.⁴² reported that PISA EROA underestimated VC area planimetered using 3DE by two-fold. Finally, Dahou et al.³⁹ showed that, in patients with TR, PISA EROA significantly underestimated EROA calculated using quantitative Doppler and VC area planimetered using 3DE, and the underestimation was larger in more elliptical regurgitant orifices.

Moreover, the lower atrio-ventricular systolic gradient of the right compared with the left heart explains the observation that for the same EROA, RegVol will be smaller for TR than for mitral regurgitation⁴³ due to the lower TR velocity, suggesting that in clinical practice different thresholds of RegVol should be used to grade TR and mitral regurgitation.

Recently, there has been a trend towards expanding TR severity grading system and include ‘massive’ and ‘torrential’ as ascending severity beyond severe TR.⁴⁴ Availability of prognostically validated continuous metrics of TR severity will allow to measure whether or not the extent of the reduction in TR severity has been enough to change the risk category of the patient.

Study limitations

This study is a retrospective analysis performed on patients studied in a single tertiary-centre. Since we cannot exclude referral and selection biases, our results should be confirmed by properly designed prospective studies. Moreover, the composite endpoint was mainly driven by heart failure re-hospitalizations rather than the hard endpoint of mortality. However, our results about the prognostic importance of moderate TR and the need of lower thresholds of the quantitative echocardiographic parameters to identify patients with severe TR are consistent with recent reports.^33,39,40

We did not compare our new echocardiography threshold values used to grade TR in low, intermediate, and high risk with any reference technique. However, there is no actual gold standard to grade TR severity. Accordingly, we used tertiles of the proportional hazard models to categorize patients into low, intermediate, and high risk to develop either CHF or the composite clinical endpoint and to obtain a clinically meaningful grading of TR severity.

Since our study aims were to assess the relationship between TR severity and the composite endpoint of death and hospitalization for CHF, and to identify the threshold values of the echocardiography quantitative parameters used to define low, intermediate, and high-risk TR based on patient outcome data, and not to identify the variables associated with poor prognosis in patients with functional TR, we did not adjusted our multi-variable model for concomitant clinical conditions, BNP, and other laboratory test of either renal or liver function, and RV function. Moreover, the number of events in our cohort allowed us to include only a limited number of variables in the multi-variable model. Finally, the introduction of the RegFr among the quantitative parameters of TR severity was a way to normalize the severity of functional TR for the size and function of the RV.

Finally, conventional Doppler echocardiography parameters used to quantify the severity of TR are affected by many technical, geometrical, and fluid dynamic limitations.³⁸ However, these parameters have shown strong prognostic value in various studies from different Institutions.^29,34,40

Conclusions

This is the first study using patient outcome data to identify the threshold values of quantitative echocardiography parameters used to grade TR severity by conventional 2D and Doppler echocardiography. We showed that the tertile of patients with the highest incidence of deaths and CHF events can be identified with cut-off values of VC_avg, EROA, RegVol, and RegFr that are significantly lower than those reported in current guidelines. Implementing these new partition values to the multi-parametric assessment of TR may prevent the underestimation of TR severity and may allow patients to be treated earlier. Moreover, since quantitative parameters of TR severity predict prognosis and the multi-parametric system grading, and since the new cut-off values are able to separate patients in low, intermediate, and high risk of death and/or CHF, these continuous metrics of TR severity could be applied to assess the effectiveness of transcatheter procedures used to treat TR.

Supplementary data

Supplementary data are available at European Heart Journal - Cardiovascular Imaging online.

Funding

A.C.G. received a research grant from the Romanian Society of Cardiology.

Conflict of interest: none declared.

References