DISCUSSION
The main findings of this study are the following: (1) non-invasive 3D RVPAC was correlated with parameters of RV systolic function; (2) RVPAC was significantly lower in patients with severe disease (3) RVPAC was the only independent correlate of severe HF symptoms in our patients with DCM.
While the RV dysfunction has emerged as a powerful predictor in left heart disease, it is well-known that RV assessment with echocardiography is challenging, with no perfect single parameter describing RV function [24]. 3D echocardiography overcomes most of the limitations and geometric assumptions of 2D echocardiography [11] and its use is growing in experienced centres.
The RV adapts to chronic increase in pulmonary vascular resistance by increasing its contractility, which is able to increase 4- to 5-fold [25]. This is done initially by hypertrophy and remodelling, as described by the Frank-Starling law of the heart. If the increase in afterload is uncontrolled and prolonged, in the attempt to maintain an adequate cardiac output, the RV will begin to dilate. This will lead to increased myocyte stress, with progressive decrease of EF, ventriculo-vascular mismatch and ultimately RV failure [25]. Although uncoupling occurs in late stages of pressure overload, it precedes clinically overt RV failure [26]. Consequently, studies aimed to find simplified methods to measure RVPAC, in order to detect patients at risk of developing RV dysfunction. Non-invasive RVPAC – usually estimated as the TAPSE/ PASP ratio – proved to have a prognostic role in patients with PH [27], HF with preserved [10] or reduced EF [28], acutely decompensated HF [29] and secondary TR [30].
In our study, we estimated RVPAC as the 3D SV/ESV ratio, which was significantly more impaired in patients with severe HF symptoms. 3D RVPAC might thus be proposed as a marker of disease severity in DCM patients. The SV/ESV ratio was first validated as a reliable surrogate for RVPAC with CMR, showing good correlation with invasive measurements [7]. Aubert et al. used 3D echocardiography to assess RVPAC in patients with PH, finding that 3D SV/ESV ratio has a good correlation with the reference measurements of ventricular/arterial elastance ratio derived from RHC [13]. A few studies found the SV/ESV ratio to be an independent predictor of adverse outcome in patients with PH [22, 23]. However, this is the first study so far to assess the role of 3D RVPAC in patients with DCM.
The energy transfer from the RV to the arterial bed is maximal when invasive RVPAC is between 1.5 and 2, with significant uncoupling occurring when RVPAC is less than 1 [26]. In our cohort, the mean SV/ESV ratio was 0.77±0.30 and the ratio was less than 1 in 79% of the patients. This apparently high prevalence of uncoupling has two explanations. On one side, it is known that RVPAC is significantly depressed before overt RV failure occurs [31]. On the other side, the volumetric method for RVPAC assessment uses the assumption that RV volume at zero filling pressure is equal to zero, which will lead to an underestimation of coupling [23, 32].
How RV functional parameters reflect the matching of RV contractility to increased pulmonary vascular resistance remains to be clarified. The RVPAC showed a good correlation with the RVEF in our cohort. However, a previous study that assessed both the SV/ESV ratio and the RVEF in PH found only the SV/ESV ratio to be an independent outcome predictor [23]. Since the ESV changes less than the end-diastolic volume at any given change in venous return, the SV/ESV ratio is less load-dependent than the EF and it is thus considered more sensitive to early changes in severe PH [33-35].
DCM is a heterogenous disease in terms of etiology, clinical presentation, regional ventricular function, and outcome. NYHA classification has been long used as a fundamental tool for risk stratification and candidacy for therapeutic strategies [36], since higher NYHA class is a well-known, powerful predictor of adverse outcome [37-38]. Identifying independent correlates of HF in DCM patients is thus of major importance.
We aimed to define the RVPAC value at which significant RV maladaptation begins in patients with DCM. In our study, RVPAC<0.54 accurately predicted severe HF symptoms, independent of age, diuretic use, LV systolic and diastolic function and PASP. It is not surprising that severe symptoms occurred at a significant level of uncoupling, since RVPAC has considerable reserve before the development of overt RV failure [31]. None of the LV functional parameters was an independent correlate of severe HF in our patients. This might be explained by the narrow range of impaired LVEF/GLS-LV and of elevated LV filling pressures in our cohort. PASP was not an independent correlate of severe symptoms either; moreover, it did not modify the prediction power of RVPAC in multivariable regression. In fact, symptom severity in our patients with DCM was not related to the degree of pulmonary hypertension, but to the degree of RV maladaptation to its afterload. This highlights the importance of evaluating the cardiopulmonary unit as a whole.