Assessment of myocardial dysfunction in cirrhotic patients: Should we look at the left atrium rather than at the left ventricle?

Over 50%1 of patients with advanced liver cirrhosis present with cardiac dysfunction due to the dysregulation of the haemodynamic circulation, also known as cirrhotic cardiomyopathy (CCM). In particular, the vasodilatory effects secondary to portal hypertension cause hyperdynamic circulation, resulting in left ventricular (LV) systolic and/or diastolic dysfunction. A timely and proper diagnosis of CCM is of crucial importance, as 15% of postoperative mortality after liver transplantation is a result of cardiac dysfunction.2 The diagnostic criteria for CCM were first determined by the World Congress of Gastroenterology in 2005, defining systolic dysfunction as an LV ejection fraction (LVEF) ≤ 55%. Recently these criteria have been revised by the Cirrhotic Cardiomyopathy Consortium (CCC),3 incorporating LV Global Longitudinal Strain (GLS) measured by speckle tracking echocardiography as a new and more sensitive marker for systolic dysfunction, based on initial evidence that LV GLS was independently associated with mortality and portal hypertension in patients evaluated for liver transplantation.4 However, not many nor large studies tested these new criteria of LV systolic function in the clinical setting. Similarly, the CCC proposed a standardized algorithm for the assessment of LV diastolic function in patients with end-stage liver disease that incorporates multiple standard echocardiographic parameters; furthermore, the integration of left atrial (LA) strain was suggested as an additional tool to potentially improve the diagnosis of CCM, although it was not integrated into the algorithm. In this volume of Liver International, Skouloudi et al. shed light on the clinical utility of strain analysis for the diagnosis of CCM according to the new criteria in patients with severe liver disease. The authors performed a detailed investigation into the alterations of LV GLS, LA and strain and right ventricle (RV) function in relation to the severity of liver disease and to outcome (mortality and liver transplantation) (Figure 1). A total of 135 cirrhotic patients were included with a median follow-up time of 15 months. The presence of systolic dysfunction was defined as LVEF ≤ 50% or an absolute LV GLS < 18% according to the new criteria. Also, LV diastolic dysfunction was defined by the presence of at least three or more standard echocardiographic parameters.5 The authors showed that although LV ejection fraction was similar, an overall higher (i.e. better) LV GLS and RV function (measured as tricuspid annular velocity S′, but not by strain) was observed in patients with more advanced cirrhosis (i.e. MELD > 15). These findings suggest that these measures can well depict and reflect the hyperdynamic circulation characteristic of this patient population and could be used to monitor LV systolic function over time and potentially after treatment or interventions. Of interest, although these patients also had larger LA dimensions, LA strain was similar between the groups. The authors performed an additional exploratory analysis on the association of visceral mass and CCM (Figure 1). They reported that patients with CCM, according to CCC criteria, had a higher mean value of body fat (40% vs 33%, p = 0.006), a higher fat index (12 kg/m2 vs 9 kg/m2, p = 0.005) and a higher volume of visceral adipose tissue (735 cm3 vs 641 cm3, p = 0.04). These findings are novel and hypothesis-generating for the mechanisms underlying cardiac involvement in cirrhotic patients, but further exploration of this association is warranted. Importantly, Skouloudi et al. also explored the prognostic value of CCM and strain analysis in these patients and showed that only the MELD score and LA reservoir strain were significantly associated with the outcome, but neither CCM nor an LV GLS of <18% was associated with mortality and liver transplantation. However, only 11 (8%) patients presented with CCM, according to the new criteria: 5 patients had an LV GLS < 18% (none presented with impaired LV ejection fraction) and 6 had advanced diastolic dysfunction. From our perspective, this study was characterized by a very low prevalence of CCM and therefore underpowered to show possible prognostic value of the associated echocardiographic indices. Because of the low prevalence of LV GLS < 18%, Skouloudi et al. applied a cut-off value for LV GLS of 22.7% to compare patients based on echocardiographic function. However, the majority of patients (42/47, 89%) who were categorized in the ‘low’ or ‘worse’ group had essentially a normal good LV GLS (>18%). This applied cut-off for LV-GLS of 22.7% was therefore possibly too high to be associated with mortality or severity of liver disease. Several other studies have investigated the new CCM criteria in patients with severe liver disease. A study by Razpotnik et al.6 evaluated the prevalence of CCM in severe liver disease patients and the association with disease severity and reported an increased LV GLS in these patients. However, as in the study by Skouloudi et al., a low prevalence of CCM was observed: only 12 patients presented with LV GLS < 18%, 3 with LVEF < 50% and 7 with diastolic dysfunction. Two more studies7, 8 reported on the relationship of CCM with mortality in severe liver disease patients but did not find a significant association. Cesari et al.7 defined CCM as GLS of <18% or >22% and observed a mean LV GLS of 21 ± 3% in the CCM group, which consisted of 21 patients. The mean LV GLS was also within the normal range and therefore possibly too good to find a significant relation with mortality. With a larger study population (n = 278), Singh et al.8 also did not observe a significant association between CCM and mortality, but mortality rate was relatively low compared to other studies2, 9 (resp. 15% vs 32%–50%), which may possibly explain their findings. Conversely, Izzy et al.5 reported a significant relation of CCM with both liver disease severity and the occurrence of cardiovascular outcomes after transplantation. Of note, a higher prevalence of CCM (a total of 49 patients, 35%) was observed compared to the aforementioned studies and the follow-up period was longer. Patients with CCM had a hazard ratio of 2.57 (CI 1.19–5.54, p = 0.016) for new-onset cardiovascular disease (i.e. symptomatic coronary artery disease, heart failure, arrhythmia and stroke) in comparison to patients without CCM. Furthermore, the 7-year event-free survival was significantly worse for patients with CCM than patients without CCM (40% vs 82%). Likewise, Mechelinck et al.4 demonstrated LV GLS as an independent predictor of mortality in liver transplant candidates. In turn, very few studies9 have explored the prognostic value of LA strain in these patients, which was indeed significant in the study of Skouloudi et al. This observation confirmed that evaluation of LA reservoir strain provides incremental prognostic value over the severity of liver disease, assessed by the MELD score, and standard grading of LV diastolic dysfunction, and may help in better risk stratifying these patients. All in all, this newly published study by Skouloudi et al. presents speckle tracking echocardiography as a useful tool to detect and monitor subtle changes in LV systolic and diastolic function in cirrhotic patients. However, the majority of patients in this study were in the earlier phase of hyperdynamic circulation, which was reflected well by high LV GLS values in this population. To validate whether LV GLS < 18% can be used as an effective marker to (early) identify patients with cirrhotic cardiomyopathy and for risk stratification, a larger cohort with a longer follow-up is needed. Interestingly, LV GLS can be applied as a marker for ischaemia detection as well.10, 11 This echocardiographic parameter could therefore be applied to simultaneously detect subtle changes in function and subclinical ischaemia. However, stronger evidence has been produced on the potential prognostic value of LA strain in these patients, which reflects not only LV diastolic function, but also LA stiffness and pulmonary pressure, providing an important tool to the treating physicians to better characterize, monitor and manage these patients (Figure 1).