Tricuspid annular plane systolic excursion and central venous pressure in mechanically ventilated critically ill patients

Echocardiography is a noninvasive diagnostic tool and can provide important information regarding certain haemodynamic parameters [1]. Among the measures of the right ventricle (RV) systolic function, the tricuspid annular plane systolic excursion (TAPSE) is easily applied and has low inter-observer variability [2, 3]. The American Society of Echocardiography recommend using the TAPSE routinely as a simple method to estimate the RV systolic function [4]. The TAPSE has also been shown to have prognostic value both in patients with pulmonary hypertension and heart failure and in noncardiac critically ill patients [5‐7].

The Central venous pressure (CVP) is widely recognized as a useful parameter for managing critically ill patients. The CVP is determined by venous return and RV function and plays an important role in the monitoring and management of right ventricular failure patients [8‐11]. Even though the CVP can only provide information about fluid responsiveness in extreme values, it is still useful when it is followed over time [12, 13]. Furthermore, the CVP can be used as a safety limit to avoid extra thoracic organ oedema because the risks for peripheral oedema, renal impairment and liver impairment are related to the absolute CVP value [14, 15]. Additionally, the CVP is of great value in the prognosis of critically ill patients [16, 17].

A previous study found that right ventricular outflow fractional shortening, which is used to assess right ventricular systolic function, can be used to predict the central venous pressure [18]. In a large cohort of healthy subjects, Ferrara F et al. noted that the TAPSE was correlated with echo-Doppler indices reflecting preload [19]. However, the relationship between the CVP and TAPSE has rarely been reported in the critically ill. The present study aimed to investigate the relationship between the TAPSE and CVP in mechanically ventilated critically ill patients.

TAPSE is a clinically feasible parameter of the RV and has been proven to be a valuable prognostic marker in various cardiac diseases [5, 21, 22]. Interestingly, in this study, we observed striking differences on the relationship between the CVP and TASPE in patients with a LVEF below 55% and in those with a LVEF of 55% and greater. We observed that the TAPSE was inversely correlated with the CVP in critically ill patients who had a LVEF below 55%. When adjusting for other possible confounding factors, the TAPSE remained an independent predictor of the CVP, and it was proven to be a good marker for discriminating whether the CVP was greater than 8 mmHg.

Volume is among the first steps of shock therapy [23]. Although dynamic parameters of volume responsiveness are recommended, they have their own innate weaknesses. Pulse pressure variation or stroke volume variation can only be used reliably in a subset of patients, i.e., those who are mechanically ventilated, sedated and without arrhythmias [24]. The passive leg raising test has fewer limitations but is not as simple to perform as it may seem at first glance; it also requires a close monitoring of stroke volume [25]. As a parameter, the CVP is far from perfect but is of great value in fluid therapy of the critically ill [26]. Our results suggest that TAPSE has the potential to predict the CVP in low LVEF patients and provides a noninvasive way to assess the right atrial pressure. The reliability of IVC respiratory variation to assess volume responsiveness has long been debated. No decisive conclusions on the accuracy of IVC respiratory variation can be drawn [27‐29]. In consistency with prior studies, we found that dIVC is directly correlated with the CVP [30, 31]. However, dIVC has its own limitations. In a group of cardiac surgical patients, Lorsomradee S, et al. noted that the correlation between the dIVC and CVP was poor when the CVP was greater than 11 mmHg [32]. Other researchers also noted that the IVC dimension and collapsibility have limited utility in identifying the magnitude of CVP elevation [33, 34]. Thus, the TAPSE could be an alternative parameter in assessing right atrial pressure for low LVEF patients.

RV function, which is often compromised when facing an elevated afterload, could lead to elevation of the CVP. A linear inverse relationship was observed between the TAPSE and pulmonary vascular resistance in a group of PAH patients who had a mean systolic pulmonary pressure of 75 mmHg [5]. However, no such relationship was found among healthy subjects [19]. In this study, patients with severe PAH were excluded. No difference was found between the groups regarding the proportion of ARDS patients and the levels of PEEP and Pplat. We did include the measurement of the R/LVEDA and no difference was found between high and low LVEF patients. These results suggest that the two groups had the same risk for RV afterload elevation. Left ventricle (LV) dysfunction could also precipitate pulmonary arterial pressure through an elevated left atrial pressure [35]. However, in a group of heart failure patients, no relation was discovered between the TAPSE and maximal tricuspid regurgitation pressure gradient [22]. Therefore, the correlation of the TAPSE and CVP cannot be explained by RV afterload elevation in these low LVEF patients.

To date, the accepted definition of septic myocardiopathy is based on a depressed LVEF, but several studies reported that the RV function was also compromised [36, 37]. However, a significant part of the RV systolic function depends on the LV systolic function, and its incidence is difficult to identify [38]. The right ventricle is bounded by its free wall with transverse fibre orientation in the septum; this is essential for ventricular twisting, which is the vital mechanism for RV ejection [39]. An experimental study demonstrated that 30% of the contraction force of the RV comes from the LV [40]. A previous study reported that the TAPSE is reduced with LV dysfunction in heart failure patients, particularly with reduced septal longitudinal motion [41]. We speculated that the TAPSE was determined, to a greater degree, by the right ventricular function in patients with a low LVEF. Therefore, it is likely that TAPSE can reflect the RV function and volume load more precisely when treating patients with low LVEF. Although the mechanism needs further study to confirm it, the result of this study indicated that monitoring of the TAPSE in these patients holds greater value.

There are several limitations to this study. First, TAPSE is a load- and angle-dependent parameter that is reflective of both the right and left ventricle function. Furthermore, this is a single-centre study, and the sample size is insufficient to provide a definite conclusion. Nevertheless, it represents a useful pilot study for further prospective investigations with larger numbers of patients. Second, pulmonary arterial pressure was not measured directly through techniques like Swan-Ganz catheter, and only a few patients were found with measurable tricuspid regurgitation on echocardiography. This may limit the direct interpretation of RV afterload. Third, the subjects in this study were heterogeneous, and the cause of LV dysfunction was not addressed. Previous studies demonstrated that LV dysfunction was very common in ICU patients and that a variety of conditions could result in LV dysfunction, including severe sepsis, prolonged hypoxia, severe metabolic and multiorgan insults, and even tachyarrhythmias [42, 43]. Chockalingam A et al. noted that, to present a unified management approach, acute left ventricular dysfunction could be classified into global LV dysfunction, acute coronary syndrome, stress cardiomyopathy and myocardial injury with minor troponin elevations [42]. The present study excluded patients with severe valvular disease, severe pulmonary hypertension, acute myocardial infarction and Takotsubo syndrome. Therefore, only patients with normal LV function or patients with global LV dysfunction were enrolled. Moreover, the patients in this study reflected the makeup of the population referred for critical care in clinical practice.

TAPSE is inversely correlated with CVP in mechanically ventilated critically ill patients who have a LVEF less than 55%.