Echocardiographic Hemodynamic Heterogeneity of Advanced Heart Failure Patients as Compared to Patients with „Pre-Heart Failure”

Advanced heart failure (HF) represents a clinical entity encompassing severely symptomatic HF with severely dysfunctional left ventricles (LV)¹. The single most important parameter for defining severe LV dysfunction and indicating the prescription of evidence-based therapies is LV ejection fraction (EF)^1,2. While LV EF is a good parameter of cardiac performance as it has the advantage of integrating the interplay between contractility and load (both preload and afterload), it has a poor discriminatory capacity/ability to assess the severity of myocardial disease and hemodynamic dysfunction³. Its limitations in assessing HF hemodynamics have been previously described^4,5,6. The latest therapeutic advances in the most severe HF patients (advanced HF and cardiogenic shock patients) rely on the good understanding of hemodynamics, stepping away from the central mechanism^7,8.

Reliable tools for diagnosis and assessment of cardiac function are crucial for the treatment of HF and the evaluation of efficacy of treatment in clinical trials. Global cardio-vascular performance, including several indexes of myocardial contractility and loading conditions, provides important contributions to the understanding of the pathophysiology, diagnosis, and treatment of various HF phenotypes.

Global cardio-vascular performance is probably best described by analysis of pressure-volume loops (PVL) (Figure 1). Additional information on cardiac function can be found not only in the shape and position of the PV loop in the PV plot but also in the quantification of contractility, compliance of the myocardial tissue and ventricular-arterial coupling. Therefore, PV loops allow for a more comprehensive analysis of a patient’s cardiac function compared with sole volumetric efficiency measurement of EF. Several metrics collected by invasive hemodynamics, although described more than 30 years ago, remain centerpiece for the understanding of HF physiopathology and are now available by echocardiographic imaging and MRI^9,10,11,12. Noninvasive pressure-volume loop analysis enables quantification of stroke work and contractility—clinically important aspects of ventricular function inaccessible by other methods. Echocardiographic LV pressure-strain loop area approximates well invasive pressure-volume loop and allows for regional and global work analysis^10,13. But as invasive methods are more a matter of research and non-invasive methods for PVL computing require complex software and/or good cardiac acoustic imaging, simpler parameters available for daily practice are required.

Figure 1

Effective arterial elastance (Ea), ventricular-arterial coupling (VAC) as the ratio between Ea and ventricular elastance (Ees), LV size, LV ejection force (LVF), cardiac power output (CPO), the ratio between isovolumic contraction time (PEP) and ejection time (LVET) have each been shown to relate to the severity of HF and predict prognosis in both acute and chronic HF settings^{14,15,16,17,18,19,20}.

Nonetheless the most effective method of profiling these patients has remained clinical and most relevant risk scores rely on clinical and biology variables^21,22,23. Beyond the measures of LV EF and congestion, routine echocardiographic evaluation is often focused on the central myocardial function only. Furthermore, when assessed, integrating the multitude of available individual hemodynamic parameters into clinical practice is not standardized and confounding. The latest document of the Joint European, American and Japanese societies for HF still consider the hemodynamic characterization of HF among the gaps in the current definitions and unreliable²⁴. Echocardiographic studies, have not reported so far the concomitant variation of the previously mentioned hemodynamic parameters.

We sought to investigate the hemodynamics by echocardiography in a cohort of advanced HF patients during a hospitalization for HF decompensation and assess the relevant differences when compared to a control cohort of asymptomatic patients with minor structural/functional cardiac abnormalities.

Advanced HF patients were followed up for 6 months after the index hospitalization. We studied the time to a first major event represented by (death, HF hospitalization or ER admission requiring iv medication).

Statistical analysis was performed using SPSS Inc statistical software, version 2020. Continuous variables were expressed as mean +/− standard deviation (SD) and categorical variables as percentages. Variables were tested for normal distribution calculating skewness and using the Shapiro-Wilk test. Patient characteristics were compared using the Fisher’s exact test for categorical variables, the independent t test for normally distributed continuous variables. All significance tests were conducted at the 1% significance level.

Most of the investigated hemodynamic variables, including LV systolic times, LV elastance, VAC, LV ejection force and CPO, differed significantly between the 2 groups (Table 3). Stroke volume and Ea were the only exceptions, with similar values among the two groups (Table 3).

When testing for normal distribution of the variables inside the advanced HF group, we found a fairly symmetrical distribution for LVEF, CPO, VAC, Ees, indexed ESV and LV ejection force.

Significantly, we found a highly skewed distribution for Ea (skewness 2,326, p<0.001). Ea values significantly correlated with Ees (r=0.766, p<0.001).

LVEF did not correlate to any other relevant hemodynamic parameter (Table 4).

All variables appeared normally distributed in the control group.

Intra- and interobserver variability has been performed for all patients from group D and 5 patients from group B, with less than 5% coefficient of variation.

The interaction between a dysfunctional LV and the arterial system is complex and governed by many parameters, but echocardiography can investigate macro-hemodynamics. Simple measurements of LV function (i.e.LVEF, PEP/LVET) and LV size (iESV and iEDV) can appropriately identify severe HF patients and the progressive deterioration and differentiate them from those with pre-HF at risk of developing HF. This approach continued without major changes even as newer medications that impact reverse-remodeling and acute and chronic hemodynamics have been consistently introduced over the time. Nonetheless, more thorough evaluation is important to understand the differences among these groups of patients. In the group of the most severe patients, LVEF doesn’t have the incremental value or accuracy to further identify the extent of cardiac dysfunction, since it is highly dependent of loading conditions.

Successful chronic HF therapies such as angiotensin converting enzyme inhibitors (ACEI) and angiotensin receptor neprilysin inhibitor (ARNI) act as vasodilators and acutely influence hemodynamics by decreasing Ea and with direct result of increasing cardiac index and thus, CPO^35,36. Nonetheless, with the exception of VAC and Ea, and arguably of left ventricular filling pressures (PCWP), none of the previously mentioned parameters have been proven to be of consistent value as therapeutic targets³⁶.

As mean values for E/A and E/E’ ratios did not differ between the two groups, our data give further proof that basic diastolic function evaluation cannot be used to guide decongestive therapy. Inferior vena cava evaluation is probably more informative and better reflects elevated filling pressures.

The most significant finding is the identification of a highly asymmetrical distribution of Ea among advanced HF patients, who are otherwise a homogenous selected group (relatively to ‘classic’ echocardiographic and clinical parameters), suggesting different patterns of hemodynamic adaptation to a severely diseased LV. Despite of a relatively small sample size, patients with advanced HF included in this study display minimal dispersion of the standard echocardiographic measurements, but a large dispersion of elastances values, suggesting a wide range of ventricular-arterial coupling and possible distinct responses to inotropic therapies, both findings being clinically relevant.

Figure 4 depicts the important variation of Ea as compared to an otherwise linear distribution of LVEF. Ea phenotypes are very diverse for the same LVEF intervaland Ea variation should not be considered erratic. Its strong correlation with Ees, suggests that ‘adaptation’ of peripheral hemodynamics can have different patterns. Full assessment of changes in contractility requires as well, accounting for changes in Ees at markedly reduced levels of contractility and in hearts that have undergone extreme degrees of remodeling - as LVEF is no longer sufficiently informative.

Figure 4

This observation reinforces the idea that the most efficient mean to influence HF macro-hemodynamics is through Ea modulation.

Our data show conclusively, that inside the most severe HF group, ‘classic’ hemodynamic parameters fail to further sub-classify patients and to identify those at highest risk for adverse events. The cut-off values of classical parameters for selecting severe cardiac dysfunction appear too broad, as newer medications acting on peripheral resistance and preload, may enable some patients to better adapt their hemodynamics than others.

As rest echocardiographic examination appears unsensitive, we suggest that the ability to modify Ea during low dose dobutamine stress testing could be more informative for prognosis and better guide therapies. To note, such granular hemodynamic evaluation by echocardiography requires good acoustic window. Further research, including less complex/sophisticated parameters, such as PEP/LVET ratio or LV ejection force, may contribute to the understanding of the contractility abnormalities seen in advanced HF patients.

The small sample size of the present study is one of the main limitations of this type of research, and is found in all such studies reported in literature. Although the study included a small sample size, the patients were highly selected to define a homogenous group of advanced HF patients. We found similar cut-off values to the previously published data for CPO and PEP/LVET that were able to identify severe patients. Dynamic tests were not performed in the present study. Any type of sensitivity analysis was beyond of the scope of this research, and limited due to the low sample size. We cannot exclude that many other variables, not collected in the present study, might be influenced the results. Our main scope was to validate a research protocol to demonstrate a significant dispersion of the elastance values regardless of baseline LVEF.

Echocardiographic measures of LV function as LVEF, LV force, LV size, CPO, VAC maintain their ability to differentiate advanced HF patients from pre-HF patients, but are load dependent and miss discriminatory capacity to further characterize advanced HF in clinical relevant and hemodynamic phenotypes. Among the advanced HF patients, the hemodynamic interaction between the dysfunctional LV and the compensatory response of the peripheral system is heterogenous and cannot predict outcome by single parameters. In these patients, assessment of cardiac performance should no longer rely on LEVF alone.