Right-Sided Cardiac Disease: No Longer the ‘Dark Side of the Heart’

This article refers to ‘Pulmonary hypertension and right ventricular dysfunction in heart failure: prognosis and 15-year prospective longitudinal trajectories in survivors’ by E. Santiago-Vacas et al., published in this issue on pages 1214–1225. Since almost two decades, it is known that right ventricular (RV) function is a major determinant of outcome in almost all cardiovascular diseases.1, 2 For instance, the presence of RV dysfunction (RVD) is independently associated with poor prognosis in heart failure (HF) with reduced ejection fraction (HFrEF)1, 3, 4 as well as in HF with preserved ejection fraction (HFpEF).3-6 The most important determinants of RVD in patients with HF are: (i) afterload mismatch due to pulmonary hypertension (PH) secondary to left heart disease, (ii) oxygen–perfusion mismatch in the setting of coronary artery disease and/or low cardiac output, (iii) biventricular involvement in the same intrinsic myocardial process, and (iv) systolic and diastolic ventricular interdependence because both ventricles share common myocardial muscle fibres, the interventricular septum and one surrounding pericardium.2 Despite this long-lasting recognition of RVD in HF, there are still no RV-specific therapies available for patients with chronic HF. More insight is needed in both pathophysiological development of RVD and its possible treatment targets, all this to improve prognosis in patients with HF.2 In addition, the scarce evidence we have so far is based on cross-sectional studies. Longitudinal studies on RVD in HF are therefore clearly welcome. In this issue of the Journal, Santiago-Vacas et al.7 have done just that. They present the results of a 15-year follow-up study on RVD and PH in patients with HF. Between 2001 and 2017, they included 1557 patients with HF within the entire spectrum of left ventricular ejection fraction, of which 72% had HFrEF, 12% had HF with mid-range ejection fraction (HFmrEF) and 16% had HFpEF. The authors performed echocardiography at baseline, at 1 year and thereafter every 2 years up to 15 years. PH was defined as systolic pulmonary artery pressure (SPAP) ≥40 mmHg and RV function was assessed using the tricuspid annular plane systolic excursion (TAPSE) and RVD was determined as TAPSE <17 mm. They also calculated TAPSE/SPAP ratio to estimate RV–pulmonary artery coupling. Clinical endpoints were (i) all-cause mortality, (ii) the composite of all-cause mortality or HF hospitalization, and (iii) the number of recurrent HF hospitalizations. The authors showed that at baseline, 22% of patients had RVD and 62% had PH. As expected, HF patients with RVD and/or PH were older, more symptomatic and had more advanced HF (e.g. more left ventricular diastolic dysfunction, more severe mitral regurgitation and were more often treated with diuretics). Patients with RVD and/or PH also displayed higher prevalence of atrial fibrillation (AF). The presence of RVD and PH at baseline was associated with adverse prognosis for all clinical endpoints. In addition, a reduced TAPSE/SPAP ratio (a measure for impaired RV–pulmonary artery coupling) at baseline was associated with (a small) increased risk of death. Interestingly, in this study patients seemed to show their true colour after 1-year follow-up. While TAPSE and SPAP overall were seen to improve after 1-year follow-up, patients that developed or remained to have RVD had the worst prognosis. In addition, at this time point low TAPSE/SPAP ratio increased the risk of mortality more than threefold compared with patients with a normal TAPSE/SPAP ratio (Figure 1). The association between RVD and PH with adverse prognosis was consistent for each consecutive year thereafter. The authors postulate a possible HF medication ‘responder’ vs. ‘non-responder’ hypothesis that could explain these differences during the first year of follow-up. Unfortunately, RVD cannot be seen as a homogeneous aetiology that responds equally in every patient. While a major strength of this study is the repeated assessment of RVD at pre-specified time points, details about the effect of HF medication on RV function are lacking. Insight into the associations between initiation and up-titration of conventional HF medication on one side, and evolution of RV (dys)function over time on the other side, would allow us to explore: (i) which patients with RVD at baseline would improve with conventional HF medication, (ii) which patients have no improvement in RV function despite optimization of HF medication, and (iii) which patients without RVD at baseline show a deterioration of RV function over time. The first group of patients are likely patients with primarily left-sided HFrEF that improve due to the optimization of HF care. This is also the vast majority of patients in the current study. In this setting, an increase in left ventricular ejection fraction, which is typically seen after initiation and up-titration of HF medication in new-onset HFrEF, is directly coupled with improvement in RV contractility because both ventricles are strongly interdependent of each other.2 In HFpEF patients, the mechanism of RV function after HF medication initiation is less clear, because for these patients there are currently no specific drugs recommended. In addition, the mechanisms of ventricular interaction between both ventricles are different and in HFpEF it seems that diastolic ventricular interaction is much more important than systolic interaction, which is predominantly the case in HFrEF.8 Any clues from observational studies about potential drugs that may improve RV function in HFpEF would support a trial in patients with HFpEF and RVD. The other two groups are more difficult to disentangle. HF patients in whom RVD maintains or develops despite HF medication may for instance have not reached their target dose (yet), or there are other ongoing processes involved that allow the RV to further deteriorate despite neuro-hormonal blockade. For instance, post-capillary PH in the setting of left-sided heart disease can progress into combined pre- and post-capillary PH with more profound RVD.9 The present study by Santiago-Vacas et al. suggests that the latter mechanism may play a role, especially because TAPSE/SPAP ratio <0.36 at 1-year follow-up was strongly predictive of poor prognosis.7 Indeed, TAPSE/SPAP ratio <0.36 also identifies HF patients with a pre-capillary component of PH in HFpEF.10 While conventional HF medication may improve left ventricular systolic function, concomitant pulmonary vascular disease may progress leading to progressive RV–afterload mismatch and subsequent high risk of death. However, attempts to use PH-specific drugs in patients with post-capillary PH due to left-sided HF were rather unsuccessful and sometimes even harmful, and thus these drugs are not recommended for patients with PH due to left heart disease.11 Continuous monitoring of pulmonary pressures using an implantable pulmonary artery pressure monitoring device (CardioMEMS™) has been more promising.12 Such a device may be helpful to better up-titrate HF medication, vasodilators and diuretics in order to reduce the risk of recurrent decompensation and to prevent the onset of a vicious cycle of progressive RV remodelling, dysfunction and failure.13 Besides a potential RV–afterload effect described above, the study by Santiago-Vacas et al. also clearly demonstrates the importance of AF in the development and/or maintenance of RVD in HF.7 This finding is consistent with prior studies focusing on RVD in patients with HF.3-6, 10 AF in HF has an incremental effect on pulmonary capillary wedge pressure,14 increases pulsatile load to the RV and worsens pulmonary vascular disease. Moreover, development of new persistent AF in patients with HFpEF has previously been associated with incident RVD, and this association was coupled with biatrial dilatation and worsening of tricuspid regurgitation.6 The association between AF and RVD can be seen as partly load-independent, since cardioversion of AF into sinus rhythm has been associated with improvement of RV longitudinal contraction.15 All these findings together support further investigation whether restoring and maintaining sinus rhythm could improve outcomes in HF patients with RVD. Some comments can be made regarding the study by Santiago-Vacas et al. First, the primary endpoint was all-cause mortality and in such an elderly population, many patients will die of non-cardiac causes during 15-year follow-up. In addition, the vast majority of patients were diagnosed with HFrEF, which could possibly hamper direct extrapolation to all types of HF (i.e. HFmrEF and HFpEF). Despite these potential limitations, the present study by Santiago-Vacas et al. on RVD in patients with HF is timely and highly relevant. It reveals clinical correlates associated with RVD and specifically demonstrates that despite optimization of HF care, a significant number of patients still maintained or developed RVD during follow-up and these patients have the worst prognosis (Figure 1). These observations can possibly form the basis for future studies focussing on possible different forms of RVD within the HF spectrum, and on the effects of HF medication up-titration on RV function in HF patients. Such knowledge would be especially relevant for patients with HFpEF, for which we still have not figured out the right path for a more tailored therapy for those HF patients with RVD. Conflict of interest: none declared.