Comprehensive Right-Sided Assessment for Transcatheter Aortic Valve Replacement Risk Stratification: Time for a Change

Transcatheter aortic valve replacement (TAVR) has been a breakthrough therapeutic advance in the treatment of patients with symptomatic severe aortic stenosis (AS). The continued expansion of this procedure has facilitated access for many patients who previously would have not been considered for further intervention given their high surgical risk. Over the past decade, several clinical and echocardiographic parameters have emerged as important markers in risk stratification, beyond the traditional assessment with Society of Thoracic Surgeons Predictive Risk of Mortality.

In the TAVR literature, most of the focus has been on the left ventricle and its response to pressure overload. Parameters such as left ventricular ejection fraction (LVEF), stroke volume index, valve area, and gradients remain important to understand disease severity and the timing of TAVR intervention.¹ Less attention, however, has been given to the right heart and to pulmonary vascular indices. Specifically, the presence and severity of tricuspid regurgitation (TR), the severity and etiology of pulmonary hypertension (PH), and right ventricular (RV) size and function have not been consistently measured or reported. As we will discuss, all those markers will better delineate the hemodynamic stage of the disease process and improve risk stratification of these patients.

PRESENCE OF CONCOMITANT TR AND OUTCOMES IN AS

At the present time, there is a lack of large, evidence-based, prospective multicenter trials addressing the common clinical dilemma of concomitant valvular disease in patients with severe AS and high surgical risk. Although mitral regurgitation is more commonly seen than TR, both lesions appear to be markers of adverse outcomes in patients undergoing TAVR. Hutter et al.² found in a cohort of 268 consecutive TAVR patients a prevalence of 20% for moderate to severe TR, which although associated with increased 1-year mortality (34% vs 21% for mild or less TR, P = .02) was not an independent predictor after multivariate analysis. Conversely, Barbanti et al.³ showed in a consecutive cohort of 518 TAVR patients that moderate to severe TR, present in 15% of that cohort, was associated with twofold increased 1-year mortality in patients with LVEFs > 40% but not in those with lower LVEFs and/or after multivariate analysis. At 30 days after TAVR, TR response was unchanged in the majority of the patients (68%); 15% experienced TR improvement, while 17% had TR worsening, including 8% without significant TR before TAVR.³ Of note, neither of these two studies integrated quantitative RV systolic function assessment. Expanding on this, a recent analysis of the inoperable cohort of the large multicenter Placement of Aortic Transcatheter Valves II trial showed that moderate or severe TR (particularly when associated with mitral regurgitation) and right atrial and RV dilation were all associated with adverse outcomes after TAVR.⁴ Of note, Lindman et al.⁴ did not find that RV dysfunction was prognostically important after multivariate analysis. However, the investigators acknowledged that the percentage of patients in whom RV function was evaluated by visual estimation as opposed to RV fractional area change was not known, and most of the evaluation was rather qualitative.^4,5

In patients undergoing TAVR, TR is, in most cases, functional and a consequence of right-sided chamber remodeling (dilation, hypertrophy, and dysfunction). PH, a common comorbidity in this group of patients, may also contribute to worsening TR, and differentiating adaptive RV remodeling with poor leaflet coaptation from advanced hemodynamic stress burden secondary to long-standing AS is an important component for future management of the disease. RV remodeling leads to free wall dilation, tricuspid annular enlargement in the anteroposterior direction, flattening of the tricuspid annulus, displacement of papillary muscles, and tethering of the tricuspid valve leaflets.⁶ As such, the vast majority of patients demonstrate concomitant dilation of the right atrium, tricuspid annulus, and right ventricle.

Another important contributor to TR development and/or potential worsening is implantable RV leads for permanent cardiac devices such as pacemakers or defibrillators. The true magnitude of this problem is currently unknown but is likely underestimated with the use of only two-dimensional (2D) transthoracic imaging.⁷

The redefined TR staging system proposed by Dreyfus et al.⁶ weighs heavily on the correct measurement of the tricuspid annulus, in addition to the assessment of TR severity and leaflet coaptation. Although this is an important step forward toward improving and individualizing the approach to TR intervention, it is crucial to understand that the tricuspid annulus is a complex and dynamic three-dimensional structure that changes in shape throughout the cardiac cycle. This introduces significant measurement variability even in healthy volunteers using standard 2D echocardiography.⁸ Multimodality imaging with the use of three-dimensional echocardiography,^9,10 multidetector computed tomography,¹¹ and cardiac magnetic resonance¹² can, in addition to quantification of the regurgitation severity, provide reliable measurements of these anatomic changes and, importantly, track changes after TAVR intervention.

The impact of aortic valve replacement on TR severity has not been well studied. The largest series, by Jeong et al.,¹³ followed 354 patients after surgical aortic valve replacement for a mean of 4.4 years, with 15% of them (54 patients) having more than mild TR at baseline. Not only did TR frequently persist after aortic valve replacement (49%), but it was also progressive in some and associated with worse outcomes.

The impact of TAVR on TR severity and outcomes remains unclear, and hopefully new prospective studies will be able to address this problem. This is timely given the emergence of the new field of transcatheter therapies for TR treatment.¹⁴

PH IN AS: NOT A BENIGN BYSTANDER

PH is another very common problem in patients undergoing TAVR evaluation. The prevalence of PH in AS varies from 30%¹⁵ to 75%¹⁶; this variation can be attributed to the different threshold criteria used for PH as well as the methods of its assessment (Doppler echocardiography vs gold-standard invasive hemodynamic right-heart catheterization). A recent meta-analysis including 16 studies with 9,204 TAVR patients found that baseline PH has an important prognostic impact on 30-day, 1-year, and 2-year all-cause mortality. Importantly, persistence of PH after TAVR appeared to be associated with higher 2-year mortality; however, there was significant study heterogeneity in the evaluation of its changes.¹⁷ Therefore, the PH response after TAVR is perhaps more important than the baseline PH severity; as it appears also to dictate outcomes.^18–20 If the pulmonary vasculature has remodeled because of the chronic pulmonary venous congestion, then PH is likely irreversible and will persist after TAVR. Indeed, a substantive decrease of pulmonary artery (PA) systolic pressure >15 mm Hg within 1 month of TAVR is seen in only a minority of patients (up to 35%) and more commonly in those without atrial fibrillation, without severely depressed LVEFs, and having functional mitral regurgitation.²⁰

Not only PH severity but also its hemodynamic presentation can provide important insights about classification and prognosis of these patients. O’Sullivan et al.¹⁶ found that TAVR patients with baseline combined or mixed pre-post capillary PH (high pulmonary capillary wedge pressure and pulmonary vascular resistance) had more impaired RV function compared with patients with passive or isolated PH (only high pulmonary capillary wedge pressure). This translated into less improvement in PH after TAVR and a threefold increase in all-cause mortality. It has been suggested that the elevation of transpulmonary pressure gradient, diastolic pressure gradient, and pulmonary vascular resistance are hemodynamic markers of pulmonary vascular remodeling resulting from long-standing, severe AS and chronic pulmonary venous congestion.²¹ It is also possible that undefined genetic and/or environmental “second hits” predispose to persistent alterations in the pulmonary vasculature after years of AS.

THE IMPACT OF RV DYSFUNCTION IN PATIENTS WITH AS

As the medical community moves toward performing TAVR in lower surgical risk categories, the importance of associated comorbidities needs to be heightened. The success of the procedure also depends on the “right-side unit,”²² which cannot be neglected and certainly should be considered in decision making. Importantly, current guidelines for the management of valvular heart disease do not mention the importance of integrating the size and function of the right ventricle in AS. In large part this stems historically from the lack of consistent evaluation of RV size and function.²³

RV functional performance is closely coupled with the pulmonary circulation. Normally, the pulmonary vascular bed has increased compliance and lower resistance compared with the systemic arterial circulation, so that RV afterload is lower than LV afterload. The right ventricle is well adjusted to lower afterload with its relatively thinner and more compliant wall. Consequently, RV-PA coupling has a significant impact on RV performance and is linked with outcomes in patients with PH.^22,24

Although the pathway is certainly nonlinear, and not so well understood, the progressive pressure overload caused by worsening AS creates left-sided chamber remodeling, worsening diastolic function, and eventually increased left atrial pressure. Backward pressure transmission continues through the pulmonary venous circulation and capillaries and eventually affects the pulmonary arterial circuit. Several possible pathogenic triggers (i.e., genetic susceptibility, hypoxia, inflammation, viral infection, damage to deoxyribonucleic acid, metabolic dysregulation, and hemodynamic and shear stress, among others) may promote vasoproliferative and antiapoptotic processes and consequent adverse pulmonary vascular remodeling.²⁵ This leads to worsening compliance in the pulmonary circulation, with increased PA stiffness, and ultimately RV-PA uncoupling. Depending on the extent, severity, and duration of the insult, compensatory RV remodeling takes place with hypertrophy, dilation, and eventually impairment of RV systolic function (Figure 1). Work to decipher these enigmatic pathogenic events at the molecular level has yet to be initiated.

Figure 1.

Pathophysiology of right-heart involvement in AS backward pressure transmission. LA, Left atrial; PCWP, pulmonary capillary wedge pressure; PHTN, pulmonary hypertension; PVR, pulmonary vascular resistance.

The prevalence of RV dysfunction in AS will vary according to the disease stage and patient comorbidities. Galli et al.²⁶ found a prevalence of 24% in a population of patients with normal or high-gradient AS with preserved LVEFs. In our own single-center cohort of patients with low-flow, low-gradient AS and reduced LVEFs,²⁷ we found that the presence of RV dysfunction (defined as tricuspid annular plane systolic excursion [TAPSE] < 16 mm) was very common (57% of patients). Although guidelines generally recommend proceeding with dobutamine stress echocardiography for these patients to determine the true AS severity and presence of flow reserve,²³ little was known about the prognostic impact of RV function in these patients.

We found that RV dysfunction in these patients was by far the most important prognostic marker of all-cause mortality. More so than baseline LV global longitudinal strain, the presence of flow/contractile reserve, severity of TR, and PH.²⁷ Our findings were confirmed by the Truly or Pseudo-Severe Aortic Stenosis group, which further demonstrated the prognostic value of RV longitudinal strain at both rest and stress²⁸ in these patients with reduced LV systolic function.

HOW TO QUANTIFY RV SYSTOLIC FUNCTION?

Given the peculiar anatomy and shape of the right ventricle, its evaluation using 2D echocardiography is complex. The American Society of Echocardiography guidelines recommend the use of several 2D views that will need to be integrated for the complete assessment of RV function.²⁹ Although the common qualitative evaluation using the “eyeball method” can be used to assess RV systolic function subjectively, quantitative evaluation is necessary and should be implemented in the routine echocardiographic evaluation of these patients. Parameters such as TAPSE describe longitudinal displacement, which although angle dependent, is a simple and reproducible quantitative method to assess RV systolic function and prognostically important marker in patients with PH.²⁹ This measurement is based on the observation that RV stroke volume is largely related to shortening in the longitudinal axis rather than reduction in the cavity diameter in the radial direction; the latter is more important in the case of the left ventricle. Yet there are many other markers of RV function by echocardiography, including fractional area change, tissue Doppler (of both the tricuspid valve annulus and myocardium), strain imaging by speckle-tracking, and measures of function such as the myocardial performance index. No one measure is perfect for assessing systolic function of this complex chamber.

TAVR does not seem to immediately affect RV systolic function in those with normal baseline function.^30,31 However, changes in RV systolic function do not seem to be uniform, as worsening of RV systolic function after TAVR, assessed by cardiac magnetic resonance, also has been reported.³² Advanced imaging methods such as speckle-tracking strain imaging,^28,33 three-dimensional echocardiography,³⁴ computed tomography, and cardiac magnetic resonance³⁵ complement 2D echocardiography and invasive hemodynamic testing by providing useful surrogate end points for disease progression and may help track response to interventions for future clinical trials in PH.

Another approach has been the integration of RV-PA coupling (RV stroke volume/RV end-systolic volume), which takes into account RV systolic performance at a given degree of afterload. It has greater sensitivity to subtle changes and potentially may be a better metric than the traditional RV ejection fraction.³⁶ In fact, RV-PA coupling, but not RV ejection fraction, was an independent predictor of outcomes in patients with pulmonary arterial hypertension.³⁷ Noninvasive surrogates for RV-PA coupling are emerging, such as tricuspid annular plane systolic excursion to PA systolic pressure ratio, but have yet to be evaluated in the TAVR population.³⁸

PUTTING EVERYTHING TOGETHER: THE INTERPLAY AMONG RV FUNCTION, PH, AND TR AND THE CHANGES PRODUCED BY TAVR TREATMENT

In this issue of JASE, Schwartz et al.³⁹ retrospectively evaluated 519 patients treated with TAVR (access route, valve type, and procedural complications are not mentioned). They quantified RV systolic function, TR, and PH using 2D echocardiography at two time points (baseline and 6 months after TAVR). For RV function assessment, they used both subjective qualitative evaluation and several quantitative 2D methods, of which TAPSE emerged as the best parameter. The investigators observed that TAVR patients with moderate or greater TR (11.3%) had a more advanced stage of the disease with a higher burden of comorbidities (i.e., more RV remodeling with larger tricuspid annular diameter, worse PH, atrial fibrillation, etc.). They conclude that RV systolic function, but not TR severity, is an independent predictor of adverse outcomes, which replicates our findings²⁷ and those of other groups.⁴⁰ There was also a suggestion that persistence of moderate or greater TR at 6 months after TAVR was associated with lower survival; however, their follow-up had a limited number of patients with severe TR and a high attrition rate.

The work by Schwartz et al.³⁹ highlights the importance of considering TR severity in the context of other relevant comorbidities commonly seen in TAVR patients, in particular the presence of RV dysfunction, which might be the main driver of outcomes. It is important to highlight, however, that only 15 patients (3%) had severe TR, and the majority of patients had preserved RV systolic function at baseline. In addition, gold-standard invasive hemodynamic assessment of PH was not available, which could have added further information about the hemodynamic presentation and outcomes. Last, this study is not suited to answer the question of whether baseline moderate or severe TR should be a clinical target for ancillary treatment beyond TAVR; larger prospective studies will be needed to address this question with a proper study design.

FUTURE DIRECTIONS

Schwartz et al.³⁹ should be commended for evaluating and quantitating these parameters before and after TAVR intervention. Their work certainly builds on the current evidence that RV function, rather than TR severity, dictates outcomes in TAVR patients. This work produces several other questions. For example, what happens to RV function after TAVR treatment in patients with baseline RV dysfunction? What drives the improvement or lack thereof? Should indices of RV function and pulmonary vascular indices be included in preoperative risk assessment scoring models? What should be the thresholds?

The importance of the right heart in patients with AS being considered for TAVR is a topic about which much remains to be learned. The multifactorial pathogenesis of PH and RV dysfunction in these patients will require our continued research efforts to better understand the complex left-right interplay and pulmonary vascular biology. Schwartz et al. should be congratulated for their work emphasizing the need to perform and integrate the comprehensive assessment on both ventricles and hemodynamics in the decision making process of AS patients.

Identification of specific hemodynamic, molecular, and genetic pathways are critically need in patients with group 2 PH, in particular as it relates to valvular heart disease. A better understanding of the importance of right heart function will hopefully guide further strategies beyond TAVR to improve the risk stratification and ultimately long-term outcomes of these patients.