Echocardiography of the Right Ventricle for Outcome Prediction in Heart Failure

In the present issue of Echocardiography, Asarcikli et al. report an interesting study on the prediction of death or hospitalization for clinical worsening by echocardiographic measurements of the right ventricle (RV) in 191 heart failure patients with either reduced or preserved left ventricular (LV) ejection fraction (HFrEF or HFpEF) [1]. Only longitudinal diameter and wall thickness of the RV were independently associated with short-term 6-month outcomes after adjusting for other clinical and echocardiographic variables. The results are in keeping with the now established prognostic relevance of the RV in any cause of HF [2]. However, the study by Asarcikli et al. offers incremental knowledge of the evolution of structural versus function changes of the RV and their assessment by echocardiography. Heart failure is associated with an increase in pulmonary vascular pressures as a cause not only of lung congestion but also of excessive afterload-induced RV failure [3]. Accordingly, symptomatology and survival in HF have been shown to be predicted by isotopic or thermodilution-derived RV ejection fraction (EF) [4, 5] echocardiography of tricuspid annular plane systolic excursion (TAPSE) or longitudinal strain referred to systolic pulmonary artery pressure (sPAP) [6, 7], and semi-quantitative visual assessment integrating these variables [8]. Of note, outcome prediction by imaging of the RV in these studies was not different in HFpEF versus HFrEF [6-8] and, somewhat surprisingly only loosely related to the level of PAP [4-8]. Pulmonary hypertension, either isolated post-capillary or combined pre- and post-capillary is in itself a potent predictor of survival in HF [3, 9]. Thus factors other than just increased afterload affect the RV in HF [2]. Why does the RV fail in diseases which predominantly affect the LV? Both ventricles are encircled by common fibers, share the septum, and are constrained within a nonacutely distensible pericardium resulting in diastolic and systolic interdependence [10]. It has been estimated that 40% of RV contraction is dependent on LV contraction [10]. This explains earlier RV failure in experimental models of HFrEF [11] or occasional clinical observations of preserved coupling of the RV to the pulmonary circulation in HFpEF [12]. However, depressed RV-PA coupling in the presence of marginally increased PAP also occurs in experimental models of HFpEF [13]. This is related to the variable impact of comorbidities and pro-inflammatory mediators [14]. In the study by Asarcickli et al., 72% of the patients suffered from HFrEF, which compares with 81% and 76% reported in other HF cohorts [6, 8]. The authors argued that a too small subgroup of HFpEF patients precluded a meaningful separate analysis. They have a point, but this would probably not have altered their results, indeed also shown in studies with matched numbers of HFrEF and HFpEF [7]. What is the natural history of RV failure on increased loading conditions? There is basically a sequence of “homeometric” increase in contractility with myocardial hypertrophy followed eventually by a “heterometric” increase in dimensions to preserve flow output adapted to peripheral demand [3, 15, 16]. In cardiac diseases with depressed contractility, the homeometric response is diminished and the RV undergoes an earlier heterometric increase in size with associated systemic congestion and negative ventricular interactions defining a right heart failure phenotype [3]. In the study by Asarcikli et al., only structural changes emerged as independent predictors of outcome while low-normal TAPSE did not [1], as expected in advanced HF [7] when systolic function adaptability is exhausted [3]. In the analysis of their results, Asarcikli et al., a lot of echocardiographic measurements were correlated to each other so they implemented a Bayesian logistic regression model to define independent predictors of outcome. Tricuspid annular systolic velocity S' almost made it, and in the end, only one dimension measurement emerged along with the expected hypertrophic response. This relatively disappointing result may be related to a high number of measurements in a relatively small patient population with a necessarily limited number of events. Other larger scale studies showed a good prediction capability of measurements of systolic function by TAPSE, S' or fractional area change, FAC [8], or better: TAPSE corrected for systolic PAP (sPAP) to estimate the coupling of systolic function to afterload [3, 6, 7]. It may also be that composite echocardiography scores integrating function and structure measurements may prove more feasible and clinically useful in HF, like recently reported in patients with pulmonary arterial hypertension [17]. The RV is inherently coupled to the pulmonary circulation and is therefore best evaluated in relation to its hydraulic load [15, 16]. Direct or indirect measurements of PAP are indispensable. However, sufficient quality tricuspid regurgitation velocity (TRV) signals to recalculate PAP are not always recovered. Failure rates of up to 20% have been reported in HF studies [7]. Asarcickli et al. mention that the recovery rate of TRV signals was too inconsistent to incorporate PAP estimations into outcome prediction models [1]. In the multicenter Right Heart International Network (RIGHT-NET) consortium, which relied on standardized training and centralized quality control, recovery of good TRV signals was reported in 894 of 954 patients with cardiorespiratory conditions (93.7%) at rest and in 816 of them (85.5%) at exercise [18]. Lower recovery rates of 83% at rest and 73% at exercise were reported in healthy controls who present with generally milder tricuspid regurgitation [18]. Thus while measurements of TRV are part of standard echocardiographic examination and generally feasible, exceptions in a small percentage of patients occur even in highly expert centers. The assessment of the pulmonary circulation may then be completed by the analysis of RV outflow tract flow-velocity waves and/or pulmonary regurgitation [19]. However, the clinical relevance of the addition of these alternative measurements needs to be validated. As the expertise and use of echocardiography with a focus on the RV in HF patients is expanding, evidence is being gathered about the interest of measurements not only at rest but also in exercise stress conditions [20]. Protocols combining echocardiography with lung ultrasound not only at rest but also during exercise or after a fluid challenge are being developed for a better characterization of the RV in HF [21, 22]. In the meantime, Asarcikli adds an important piece to the puzzle of the clinical relevance and pathophysiology of echocardiographic assessments of the RV in HF. This commendable effort calls for more. The data that support the findings of this study are available from the corresponding author upon reasonable request.