Author + information
- Published online October 9, 2017.
- Olivier Villemain, MD∗ (, )
- Fidelio Sitefane, MD,
- Mathieu Pernot, PhD,
- Sophie Malekzadeh-Milani, MD,
- Mickael Tanter, PhD,
- Damien Bonnet, MD, PhD and
- Younes Boudjemline, MD, PhD
- ↵∗M3C-Necker Enfants malades, AP-HP, Université Paris Descartes, Sorbonne Paris Cité, Cardio-Vascular Department, 149 rue de Sèvres, Paris 75015, France
Heart failure is the main cause of morbidity and mortality in cardiac patients, and may cause the dysfunctions of multiple organs as a result of various interactions. A right heart dysfunction can induce signs of congestive right heart due to the increase of filling pressures, with direct repercussions on the liver (hepatic congestion). Currently, the indirect estimation of filling pressures of the right heart is achieved through clinical examination and ultrasound parameters. Right heart catheterization (RHC) is the gold standard to measure central venous pressure (CVP) (1) but it is invasive, costly, and cannot be repeated for close follow-up, particularly on children.
Liver elastography was initially developed to analyze the tissue properties of this organ, especially for the grading of cirrhosis. It is important to note that congestion alters the intrinsic rigidity of the liver (2). For this reason, recent studies have demonstrated that liver stiffness (LS) and CVP are correlated (3,4). Until now, however, these studies have had technical limitations and the potential usefulness of LS measurement using shear wave elastography (SWE) as a reliable and quantitative surrogate of CVP in clinical practice needs further developments. To our knowledge, no clinical studies have been conducted on humans to link the CVP and the LS estimated by SWE in real time.
Here we sought to determine whether LS estimated by SWE could reliably estimate the measurement of CVP during RHC. In addition, we investigated whether acute changes in CVP paralleled changes in LS.
A total of 103 children (6.8 ± 5.5 years of age) referred to our institution for diagnostic or interventional RHC were prospectively enrolled. CVP and LS were measured simultaneously at baseline and after 15 ml/kg of volume loading. Inferior vena cava diameter and pulsed-Doppler profile of hepatic veins were also evaluated. Plasma level of N-terminal pro–B-type natriuretic peptide was assayed during the RHC.
SWE was used to image LS with the Aixplorer ultrasound imaging system (Aixplorer, Supersonic Imagine, Aix-en-Provence, France) with an abdominal curved probe (SC6-1), during the RHC procedure. These acquisitions were performed at 2 moments during the procedure, before and after the rapid saline loading (at the same time of the pre-CVP and post-CVP measurements). The physicians performing the LS measurement were unaware of the CVP results.
At baseline RHC, the mean CVP was 7.4 ± 2.9 mm Hg (range 3 to 16 mm Hg) and the mean LS was 9.0 ± 5.8 kPa (range 4 to 46.1 kPa). After volume loading, the mean CVP increased significantly to 10.0 ± 3.3 mm Hg (range 3 to 18 mm Hg) (p < 10–4) and the mean LS increased significantly to 14.4 ± 9.1 kPa (range 4.3 to 72 kPa) (p < 10–4). LS strongly correlated with CVP, pre-loading (r = 0.86; p < 10–4) and post-loading (r = 0.87; p < 10–4) (Figure 1).
All patients who had an increasing CVP also had an increasing LS (93 of 103 patients). In addition, all patients who had a decreasing CVP also had a decreasing LS (10 of 103 patients).
Optimal cutoff value of LS for detection of CVP >10 mm Hg was 10.8 kPa (sensibility = 89.3%, specificity = 86.0%), with an area under the curve of 0.946 (95% confidence interval: 0.920 to 0.971; p = 0.01). Beyond this correlation, LS is sufficient to provide an indirect and reliable measurement of quantitative CVP variations (r = 0.86, ρ = 0.756, p < 10-4; multivariate model). Inferior vena cava diameter (r = 0.40, ρ = −0.408, p = 0.01), pulsed-Doppler profile of hepatic veins (r = 0.19, ρ = 0.078, p = 0.483), and N-terminal pro-B-type natriuretic peptide (r = 0.10, ρ = 0.038, p = 0.736) were less robust than LS to estimate CVP.
The intraoperator and interoperator reproducibility of the LS measurement technique were studied in 20 patients, and we found no significant statistical difference.
For children with congenital heart disease and no argument for tissue liver anomalies, noninvasive and quantitative estimation of filling pressures can be very difficult to perform. The paradox is that the filling pressures (and specifically CVP) have been recognized as a major predictive parameter of adverse events. The management of the acute heart failure at the child’s bedside depends on the evaluation of volume or pressure overload (5). That is why a quantitative and real-time parameter as LS by SWE could be clinically necessary.
Here, we show that LS measurement using SWE is a reliable surrogate of quantitative estimation of the CVP. It can also be used to measure CVP changes in real time. LS could potentially be a useful noninvasive tool for evaluation and follow-up of acute and chronic right heart failure.
Please note: Dr. Tanter is cofounder of SuperSonic Imagine. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2017 American College of Cardiology Foundation
- Nagueh S.F.,
- Kopelen H.A.,
- Zoghbi W.A.
- Jalal Z.,
- Iriart X.,
- De Lédinghen V.,
- et al.
- Hsu D.T.,
- Pearson G.D.