Refining Severe Tricuspid Regurgitation Definition by Echocardiography With a New Outcomes Based “Massive” Grade

Abstract

Background:

Current echocardiographic guidelines recommend that tricuspid regurgitation (TR) severity be graded into three categories, following assessment of specific parameters. Findings from recent trials have shown that the severity of TR frequently far exceeds the current definition of severe. We postulated that a grading approach that emphasizes outcomes could be useful to identify patients with severe TR at increased risk of mortality.

Methods:

We identified 284 patients with echocardiograms demonstrating severe functional TR, defined as vena contracta (VC) ≥0.7 cm. Demographics and mortality data were obtained from the medical records. Patients were divided into Study (n=122 patients with 3D images) and Validation (n=162) cohorts. VC was measured in both the right ventricular (RV) inflow and apical-4 chamber views and averaged. For the Study cohort, tricuspid annular (TA), RV end-diastolic (basal, mid, long axis) dimensions, tricuspid leaflet tenting height and area, RV free-wall longitudinal strain (Epsilon Imaging), and RV volumes (TomTec) were measured from 2D and 3D datasets, respectively. A K-partition algorithm was used in the Study cohort to derive a mortality-related cutoff VC value, above which TR was termed “massive”. The ability of this VC cutoff to identify patients at greater mortality risk was then tested in the Validation cohort using Kaplan-Meier survival analysis.

Results:

In the Study cohort, VC>0.92 cm (”massive” TR) was optimally associated with worse survival. TA and RV size were larger in the Massive group (p<0.05), while there were no significant differences in demographics between the TR groups. Importantly, in the independent Validation cohort, the above VC cutoff also correlated with increased mortality in the Massive group (log-rank p<0.05).

Conclusion:

Among patients traditionally defined as having severe TR, a subset exists with massive TR, resulting in greater adverse RV remodeling and increased mortality. These patients may derive the greatest benefit from emerging percutaneous therapies.

Introduction

Multiple studies have shown that functional tricuspid regurgitation (TR) is associated with a poor prognosis, independent of left-sided heart failure and pulmonary hypertension^1–3. Accordingly, attention has recently turned toward the development of trans-catheter therapies that correct TR, stimulated in part by the success of percutaneous therapies for aortic stenosis and mitral regurgitation (MR).

Echocardiography remains the primary imaging modality to diagnose the etiology and assess the severity of TR. According to recent American Society of Echocardiography⁴ and the American Heart Association/American College of Cardiology Valvular Heart Disease valvular regurgitation guidelines⁵, TR vena contracta (VC) of >0.7 cm identifies severe TR, as established against the flow convergence method, is independent of flow rate and driving pressure and can be used even with eccentrically direct jets ⁶. Although this index is problematic in the setting of multiple jets, VC is a widely used, rapid technique for the assessment of TR in clinical practice.

While cutoffs of abnormality can be defined by percentiles or standard deviations above and below the mean of normal subjects, partition values describing severity of valvular regurgitation is more challenging, because it cannot be derived from the normal population. Instead it requires disease specific data that determine what is mild, moderate and severe, which is commonly achieved by consensus of experts when robust outcomes data are not available. Current guidelines recommend a multiparametric approach using quantitative and semi-quantitative methods for assessing TR severity⁴. It is important to note, these methods have been more robustly studied in MR⁷ and then applied to the tricuspid valve, which has a more complex shape and is subjected to different loading conditions^{6, 8}. Moreover, TR severity partitions are not based on robust outcomes data. Our study was motivated by the notion that a grading approach that would predict outcomes or prognosis would be preferred.

In clinical practice, patients are commonly encountered with a wide range of TR, all falling within what is defined as severe by VCs of ≥0.7 cm. Up to 5.6% of women and 1.5% of men have clinically significant TR by their eighth decade, but fewer than 8,000 of the estimated 1.6 million patients affected in the United States alone undergo surgical intervention annually^9–11. Given this vast and growing demographic, it is important to develop a method to risk stratify these patients in order to identify those who could derive the most benefit from TR interventions. We therefore hypothesized that within group of patients, there may be a subset with more severe TR resulting in greater adverse remodeling of the right ventricle (RV) and ultimately portending a worse prognosis. Accordingly, we sought to assess the relationship between tricuspid annulus (TA) diameter, and RV size and function with mortality in a large group of patients with a wide range of severe TR, in order to derive and test a new VC cutoff value that would distinguish between severe TR with lesser impact on outcomes and “massive” TR associated with higher risk of mortality.

Methods

Patient population and study design

This retrospective study was approved by the Institutional Review Board with a waiver of consent. From our database of patients who were referred for a clinically indicated cardiac ultrasound examinations from January 2010 to January 2018, we identified 284 patients diagnosed with severe TR, defined as a VC ≥0.7 cm, in either the RV inflow or apical-4-chamber views. The presence of severe TR was confirmed by remeasuring VC (KYK, RML) in both views. The average VC was used, and in cases in which only one view had an adequate color Doppler jet to measure VC, a single value was used. Patients were excluded if they had primary TR due to an RV catheter or lead, structurally abnormal tricuspid valves (i.e. vegetation, perforation, cleft, repair), or prosthetic tricuspid valves. Patients with prior tricuspid valve surgery or percutaneous intervention were excluded. Patient demographics and all-cause mortality data were obtained from electronic medical records. In patients who had multiple echocardiograms during the study period, only the initial study was used.

The patients were divided into a Study cohort (N=122 patients who had both 2D and 3D images) and a Validation cohort (N=162). The Study cohort was used to derive a cutoff value of VC associated with increased mortality and termed “massive”, and to study the relationship between this cutoff and adverse RV remodeling. Then, the VC cutoff was tested in a Validation cohort to determine whether it could predict mortality in an independent patient population.

Transthoracic 2D and 3D echocardiographic imaging and analysis

In all patients, comprehensive 2D echocardiographic (2DE) imaging with tissue and Color Doppler with optimization of the Nyquist limit was performed by expert sonographers using an iE33 or EPIQ imaging system equipped with an S5 transducer (Philips Healthcare, Andover, MA). In the Study cohort patients, RV-focused 3D echocardiographic (3DE) data sets that included the entire RV in the pyramidal scan volume were also acquired. Simultaneous real-time short-axis multiplanar reconstructions were used at the time of 3DE image acquisition to ensure optimal visualization of the RV free wall. Image settings were optimized to obtain RV full-volume images with high frame rates (>24 Hz) by minimizing the depth and the width of the imaging sector. Digital loops were stored and analyzed offline: 2DE images using Xcelera (Philips) and 3DE data sets using RV Analysis software (TOMTEC, Unterschleissheim, Germany).

In both cohorts, VC was measured in both the RV inflow and apical four-chamber views when the quality of color Doppler images was adequate. As this was a patient population of functional TR, the vast majority of cases had a single, central jet of TR. To assess RV remodeling in the Study cohort, TA dimensions, as well as RV end-diastolic base (RVEDb), RV end-diastolic mid (RVEDm) and RV length (RVEDL) dimensions were measured in the RV-focused apical four-chamber view (Figure 1A). The tricuspid leaflet tenting height and area were measured in the apical four-chamber view during mid-systole (Figure 1B). RV free-wall longitudinal strain was also measured (EchoInsight, Epsilon Imaging) (Figure 1C). In addition, 3DE data sets were used to measure RV end-systolic and end-diastolic volumes and calculate ejection fraction (Figure 1D). Estimated RV systolic pressure was not recorded as it is often inaccurate in the setting of severe TR and rapid equalization of chamber pressures.

Figure 1.

Echocardiographic assessment of the RV including (A) 2DE RV and TA dimensions in the RV focused view at end-diastole, (B) tricuspid valve tenting height and area in the apical four-chamber view during mid-systole, (C) RV longitudinal strain of the free wall, and (D) 3DE endocardial surface reconstruction of the RV in systole (solid) and diastole (mesh).

Statistical analysis

Data are expressed as mean ± SD. Two-tailed, non-paired student’s t-tests were used to in continuous variable measured parameters Chi squared tests were used for categorical parameters to test the significance of the differences between groups. P values <0.05 were used to indicate statistical significance.

A K-partition algorithm was used in the Study cohort to derive a mortality-related cutoff VC value. While previous methods either only allowed two-way splits¹² or recursive multi-splits utilizing tree-structured methods, which did not provide always optimal subsequent splits¹³, in our study, the K-Adaptive Partitioning for Survival Data algorithm from the KAPS R package allowed multi-way splits to locate the optimal set of cut-off points. Furthermore, using a resampling technique, the KAPS algorithm also selects an optimal number of subgroups ¹⁴. The K partition algorithm was applied to the Study cohort and resulted in two subgroups divided by a partition value that defined the category of “massive” TR based on survival analysis.

In the Validation cohort, the relationship between this VC cutoff value and mortality was tested using Kaplan-Meier survival analysis. The significance of the difference in mortality was tested using log-rank tests.

Results

In the Study cohort, the median age was 72.5 years (IQR: 59.5–84.75); majority were female (73%), and there was a high prevalence of atrial fibrillation (47%). The K-partition algorithm yielded a partition value of VC≤0.92 cm to define the Severe TR group, (n=79; Figure 2A) and >0.92 cm to define the Massive TR group (n=43; Figure 2B) for survival probability of p=0.043 (Figure 3A). There were no significant differences in baseline demographics between these two groups including LV ejection fraction, LV dilation, moderate or greater mitral regurgitation, and diastolic dysfunction. TA dilation, TV leaflet tenting, and RV enlargement were larger and RV free-wall strain magnitude was more reduced in the Massive compared to the Severe group (Table 1), indicating increased adverse remodeling. Despite the larger TA and RV sizes in the Massive group compared to the Severe group, RV systolic function was similarly moderately reduced in both groups according to the ASE Chamber Quantification Guidelines¹⁵ (Table 1).

Figure 2.

Example of patients in the (A) Severe TR (VC average=0.75 cm) and (B) Massive TR (VC=1.25 cm) groups with color Doppler in the apical four-chamber and RV inflow views.

Figure 3.

Kaplan-Meier survival analysis of the Study cohort (A) and the Validation cohort (B) using the VC of 0.92 cm to separate Severe from Massive TR groups based on the K partition algorithm. Note poorer survival in patients in the Massive TR groups in both cohorts.

Table 1.

Clinical characteristics and right ventricular parameters of the Study cohort including two subgroups defined by the tricuspid regurgitation cutoff of vena contracta of 0.92 cm.

	Study Cohort	Severe TR	Massive TR
Characteristics	Total Cohort (n = 122)	VC ≤ 0.92 (n = 79)	VC > 0.92 (n = 43)	P
Age, years	71 ± 17	71 ± 17	70 ± 16	0.67
Male	33 (27%)	21 (27%)	12 (28%)	0.870
BSA (m2)	1.8 (1.6, 2.0)	1.7 (1.6, 2.0)	1.8 (1.6,2.0)	0.50
History Afib	57 (47%)	33 (42%)	24 (57%)	0.11
LV EF (%)	44 ± 20	48 ± 19	37 ± 19	0.14
LV dilation	32 (29%)	19 (24%)	13 (30%)	0.46
Obstructive CAD	38 (31%)	26 (33%)	12 (28%)	0.57
>Mod mitral regurgitation	48 (39%)	33 (42%)	15 (35%)	0.55
Diastolic dysfunction	69 (57%)	42 (53%)	27 (63%)	0.31
Tenting height (cm)	0.84 ± 0.35	0.76 ± 10.34	0.98 ± 10.34	0.0015
Tenting area (cm2)	1.9 ± 1.0	1.5 ± 0.7	2.6 ± 1.1	<0.0001
RAVi (ml/m2)	70 ± 33	57 ± 25	87 ± 38	<0.0001
TA (cm)	4.39 ± 0.65	4.15 ± 0.58	4.83 ± 0.55	<0.0001
RVEDb (cm)	5.65 ± 0.96	5.37 ± 0.94	6.17 ± 0.77	<0.0001
RVEDm (cm)	4.45 ± 1.03	4.24 ± 1.03	4.85 ± 0.93	0.0008
RVEDI (cm)	8.40 ± 1.08	8.16 ± 1.08	8.84 ± 0.95	0.0005
RV FW Strain (%)	-13.8 ± 5.1	-14.8 ± 5.3	-12.0 ± 4.3	0.007
3DE RVEDVi (ml/m2)	149 ± 68	134 ± 61	177 ± 71	0.0006
3DE RVESVi (ml/m2)	95 ± 47	84 ± 43	114 ± 49	0.0005
3DE RVEF (%)	37 ± 10	38 ± 10	36 ± 10	0.29

Abbreviations: BSA – body surface area, Afib – atrial fibrillation, LV – left ventricular, EF ejection fraction, CAD-coronary artery disease, Mod-moderate, RAVi right atrial volume index, TA – tricuspid annulus, RVED(b/m/l) – right ventricular end-diastolic basal/mid dimensions /length, RV – right ventricular, FW free-wall, 3DE – 3D echocardiographic, EDVi/ESVi - end-diastolic/endsystolic volume index. P-values in the right-hand column indicate significance of the differences between the Severe and Massive groups.

In order to test the validity of the VC cutoff derived from the Study cohort in an independent patient population, we applied the same cutoff to the Validation cohort. In this cohort, of the 162 patients, 95 patients had VC≤0.92 cm, and were defined as having severe TR, while the remaining 67 patients had VC>0.92 cm, and were thus defined as having Massive TR. The ages were similar between Severe and Massive TR groups in the Validation cohort: median of 68.5 years (IRQ: 56.0–78.0) and 67.0 years (IRQ: 53.25–76.25), respectively (p=0.502). Kaplan-Meier survival analysis showed higher mortality in the Massive versus Severe groups (Figure 3B).

In addition, we repeated the Kaplan-Meier survival analysis using a VC cutoff of 1.3 cm, previously suggested by R.T. Hahn and coworkers²⁸, as the upper limit of severe TR, based on visual assessment of the severity of TR. This analysis did not result in significant separation in terms of survival between the two groups defined by this alternative cutoff (p=0.71).

Discussion

The goal of this study was to refine the definition of severe TR based on mortality data, so that this definition could be used in device trials in order to select patients who may derive the most benefit from these emerging therapies and to assess improvement post intervention. The additional VC partition value of 0.92 cm was derived in the Study cohort and subsequently tested in a separate Validation cohort. Our study not only showed worse survival with a higher degree of massive TR, but also confirmed the expected adverse RV remodeling with cavity dilation, annular dilation, and tricuspid leaflet tenting and reduced RV free-wall strain, consistent with prior studies on the mechanisms of functional TR.

It is recognized that patients with functional TR have worse prognosis, irrespective of left ventricular (LV) function, degree of pulmonary hypertension, and TR severity¹. Traditionally, parameters used for assessing TR severity have been initially established for the mitral valve^{7, 16, 17} and then applied to the tricuspid valve⁶, despite the vast differences in three-dimensional geometry and loading conditions between these two valves. Unlike the saddle-shaped mitral annulus, which is attached to two trigones with a thick fibrous structure, the TA is attached only to the posteromedial trigone and has much thinner fibrous structures, more prone to dilation in the setting of pressure or volume overload¹⁸. Each of the mitral papillary muscles supplies chordae to both the anterior and posterior mitral valve leaflets, which helps support the annulus during LV dilation, whereas the chordae of the tricuspid leaflets connect only to the homolateral papillary muscle allowing for more TA dilation. The TA is integral in maintaining normal tricuspid valve mechanics. As it dilates in the anterior direction, as the septal leaflet is fixed and the inferior annulus is adjacent to the diaphragm, it becomes more circular and planar until the area of coaptation is insufficient to maintain competency of the valve^{19, 20}. Dilation of the TA and RV free wall leads to apical displacement of the papillary muscles and tricuspid leaflet tenting. The degree of TV tenting has been shown to correlate with the degree of functional TR²¹. Despite these anatomical and loading condition differences, the current valvular regurgitation guidelines for TR use the same semi-quantitative and quantitative parameters as for the mitral valve, and further, have the same partition values for VC and effective regurgitant orifice area (EROA) of 0.7 cm and 0.40 cm², respectively⁴.

Cardiac imaging has played an important role for patient selection and pre-procedural planning in transcatheter aortic valve replacement and percutaneous mitral valve interventions, and this situation now expanding to the tricuspid valve arena with the advent of minimally invasive therapies for severe TR.

Similar to other investigators, we have frequently observed patients that appear to have TR that far exceeds the societal guidelines definition of severe. In fact, the European guidelines describe “massive” TR as having a continuous wave Doppler dense/triangular jet with early peaking (peak <2 m/s), but outline no other semi-quantitative or quantitative cutoffs separating “massive” from severe TR²². Tricuspid leaflets that remain tethered open and fail to coapt, a condition referred to as “free” TR²³, with no MR equivalents described in the literature.

Given the progressive nature of functional TR, one could postulate that any reduction in TR would improve prognosis and that further partitions within the severe category would be helpful in patient selection and assessing improvement after TR interventions. To investigate this, the international TriValve Registry was established in 2016 to study high risk patients with severe TR undergoing emerging transcatheter tricuspid valve intervention. Included in this registry are MitraClip (Abbott Vascular, Santa Clara, CA) in the tricuspid position, FORMA spacer (Edwards Lifescience, Irvine, CA), Cardioband (Edwards Lifescience) tricuspid, TriCinch (4TECH, Galway, Ireland), Trialign (Mitraling, Tewksbury, MA), and caval valve implantation (CAVI)²⁴.

In a study of seven patients using FORMA, the inclusion criteria was severe TR with a VC width ≥7 mm, although the mean baseline VC was much larger at 15.5±5.1 mm²⁵. In a series of fifteen patients using the TriAlign device, the pre-procedural mean VC was 1.3±0.3 cm. In the twelve patients with procedural technical success, the 30-day VC was significantly improved at 1.1±0.4 cm (p=0.022), but this still falls within the severe category, where our current grading scheme fails to detect the improvement. The reduction in TR improved echocardiographic parameters of TA dilation and LV stroke volume and importantly clinical outcomes, including 6-minute walk test and NYHA symptom class²⁶. In the TriValve registry of 106 patients undergoing tricuspid intervention, the average VC was 1.1±0.55 cm which is approximately 1.5 times the definition of severe. With the current grading method, over half of the patients had 3+ or greater TR on discharge echocardiography²⁷.

Given the extreme degree of TR encountered, there has been a call for a new grading scheme for TR, which may be better equipped to capture the improvement in TR with these transcatheter interventions. A proposed scheme adds two more defined grades beyond severe TR, namely “massive” and “torrential”, but these involve arbitrary semi-quantitative and quantitative cutoffs, which have not been standardized or validated against outcomes data. The proposed “massive” grade includes patients with VC width 14–20 mm, EROA 60–79 mm², and regurgitant volume 60–74 ml by proximal isovelocity surface area method and the “torrential” grade beyond those partition values^23,28. Our study has established an outcomes-based threshold for VC that defines this at-risk population of patients (Table 2).

Table 2.

Proposed definitions of massive tricuspid regurgitation: Comparison of recent literature sources.

Source	Proposed Definition of Massive Tricuspid Regurgitation		Supporting Evidence
	Qualitative	Semiquantitative/Quantitative
European Association of Echocardiography Guidelines 201022	Continuous wave Doppler: Dense/Triangular profile with early peaking with a peak velocity <2 m/s		No outcomes data
American Society of Echocardiography Guidelines 20174	Valve morphology: Non-coapting tricuspid valve leaflets		No outcomes data
Hahn et al 201828		Biplane VC: 1.4–2.0 cm (Massive), ≥2.1 cm (Torrential)	Extrapolated using this definition, the majority of the 12 AS-Treated SCOUT early feasibility trial patients would have had a 1–2 grade reduction in TR, which was associated with improvement in NYHA functional class (≥1 class, p=0.001), MLHFQ survey (p < 0.001), and 6MWT (p=0.008)26
		EROA by PISA: 60–79 mm2 (Massive), ≥80 mm2 (Torrential)
		3DE VCA or EROA by Doppler: 95–114 mm2 (Massive), ≥115 mm2 (Torrential)
Go et al 201823	Continuous wave Doppler: Dense/Triangular profile with early peaking with a peak velocity <2 m/s	VC (preferably biplane): 1.4–2.0 cm (Massive), ≥2.1 cm (Torrential)
		EROA by PISA: 60–79 mm2 (Massive), ≥80 mm2 (Torrential)
		EROA by 3DE: 95–114 mm2 (Massive), ≥115 mm2 (Torrential)
		Regurgitant volume: 60–74 ml (Massive), ≥75 ml (Torrential)
Our study(Kebedetal)		VC (preferably biplane): >0.92 cm	Massive TR associated with worse survival (p=0.043)

VC-vena contracta, EROA-effective regurgitant orifice area, VCA-vena contracta area, 3DE-three dimensional echocardiography, PISA-proximal isovelocity surface area, TR-tricuspid regurgitation, NYHA-New York Heart Association, MLHFQ-Minnesota Living with Heart Failure Questionnaire, 6MWT-six minute walk test

When the RV is exposed to pressure overload, the adaptive response is myocardial hypertrophy followed by contractile dysfunction. RV chamber dilation ensues as a compensatory mechanism to maintain right-sided cardiac output. As the base of the RV expands, the TA also dilates leading to malcoaptation and functional TR. Accordingly, the RV is capable of adapting to a large degree of TR for extended periods of time. This adaptive RV dilation and systolic dysfunction have been shown to be markers of poor prognosis. In a study of patients with idiopathic LV dilated cardiomyopathy, concomitant RV enlargement is a strong predictor of poor prognosis with nearly triple the mortality over four years²⁹. In a study of patients with New York Heart Association Class II symptoms and heart failure with reduced LV systolic function, RV ejection fraction but not LV ejection fraction was an independent predictor of survival³⁰. Given the crescent shape of the RV, accurate assessment of systolic function is harder than the bullet-shaped LV. Another technique to assess RV function is free-wall longitudinal strain and strain rate. Advantages of strain over traditional surrogate measures of RV function include angle independence, less load dependence, and ability to detect regional variation. To date, there is a paucity of data on RV strain and outcomes, and it is not currently one of the parameters recommended for the routine assessment of RV function³¹. Although the recommended methods, such as tricuspid annular plane systolic excursion, S’-velocity, and RV fractional area change only moderately correlate with cardiac magnetic resonance derived RV ejection fraction, whereas RV strain showed a good correlation ³². As more data becomes available, normal reference ranges and correlations to clinical outcomes are likely to make RV longitudinal strain an integral part of the standard assessment of the RV.

Limitations

This was a single-center study of patients with functional TR due to varying disease states. This is a retrospective study with the intrinsic limitations of no fully standardized protocol of image acquisition, resulting in variable image quality that may have affected our measurements. Because the patients were identified from our clinical database, we had to use the semi-quantitative 2D VC measurements performed routinely in our laboratory, rather than regurgitant volume or EROA. 3DE studies have shown that VC can be non-circular, therefore we used the average of two views to get the most accurate assessment. The RV parameters were only obtained on the Study cohort, as the Validation cohort was used only to test the ability of the new VC cutoff to identify patients at higher risk of mortality.

One might view as a limitation the fact that RV parameters were not measured in the Validation cohort, and therefore greater adverse RV remodeling was not independently confirmed in patients with massive TR. However, we studied RV remodeling only to provide supportive evidence to the presence of more severe TR in patients with VC above the newly derived mortality-based cutoff. Furthermore, the cutoff was not derived to differentiate degrees of adverse RV remodeling, and therefore did not need to be validated to confirm its ability to detect such differences.

Conclusion

Patients with severe TR defined by current guidelines are not a homogenous population in terms of TR severity or potential outcomes. Within the group of patients with VC ≥0.7 cm, currently defined as having severe TR, there are varying degrees of severity and coexisting RV remodeling. When these patients are divided by a VC of 0.92 cm, two distinct categories can be appreciated which have divergent outcomes. The ability to distinguish patients with severe TR at particularly high risk may be helpful in selecting the most appropriate patients who can derive the greatest benefit from interventional procedures.