Early detection of cirrhotic cardiomyopathy using myocardial strain imaging and its correlation to liver stiffness and severity of liver disease – a pilot study

Background and Aim:

The prevalence of cardiomyopathy in cirrhotics remains unknown because of its latent nature, characterized by blunted contractile responsiveness to stress, altered diastolic relaxation, and electrophysiological abnormalities. Our aim was to detect early myocardial dysfunction using new echocardiography technologies in cirrhotic patients and correlating them to liver stiffness (LS) and severity of liver disease.

Methods:

Consecutive patients with liver cirrhosis without structural heart disease and portal vein thrombosis were included. Conventional and speckle-tracking echocardiography (Vendor GE, EchoPAC PC software) were performed by a single investigator (EACVI TTE certified). Subclinical myocardial dysfunction of left ventricle was defined as average global longitudinal strain (GLS) < -18% (Lang RM, et al. J Am Soc Echocardiogr 2015;28:1 – 39). LS was assessed by transient elastography (TE, Fibroscan®, Echosens) and share wave elastography (SWE) from Hitachi (Arietta V70). Reliable results were defined as median value of 10 valid measurements with an IQR/Med < 30% and expressed in kPa. The presence of ascites, esophageal varices, splenomegaly and/or thrombocytopenia were considered as sign of portal hypertension.

Results:

We evaluated 48 patients, but 8 did not fulfilled the inclusion criteria (1 portal vein thrombosis, 2 coronary artery disease, 1 cor pulmonale and 4 patients with valvular dysfunction). The final analysis included 40 patients, with mean age of 58.7 ± 11.1 years (67.5% males). Compensated cirrhosis (Child-PughA) was present in 60% of patients, 22.5% were classified as Child-PughB and 17.5% as Child-PughC. LS could be evaluated in 77.5% of cases by TE and in all patients by HitachiSWE. Slightly reduced left ventricular ejection fraction (EF) was observed in 10% and diastolic dysfunction in 40% of cases. Subclinical systolic dysfunction as assessed by GLS was present in 22.5% of cases. The presence of systolic dysfunction tends to correlate with LS (Table).

Tab. 1

Parameter	GLS < -18% or reduced EF (n = 10)	GLS ≥ -18% and normal EF (n = 30)	p value
LS by TE (kPa)	41.3 ± 25.1	29.1 ± 17.6	0.09
LS by SWE (Hitachi)(kPa)	10.8 ± 3.8	14.6 ± 6.3	0.006
MELD	9.9 ± 3.8	10.8 ± 4.7	0.48
Child-Pugh: – compensated (A) – decompensated (B+C)	7 (70%) 3 (30%)	18 (60%) 12 (40%)	0.85 0.85
Portal vein velocity (cm/s)	17.1 ± 3.9	15.9 ± 2.9	0.23
Signs of portal hypertension	6 (60%)	23 (57.5%)	0.81
Etiology: – alcoholic – other etiologies	5 (50%) 5 (50%)	19 (63.3%) 11 (36.7%)	0.71 0.71
Age (years)	61.5 ± 12.2	58.8 ± 10.6	0.54
Gender: – male – female	5 (50%) 5 (50%)	22 (73.3%) 8 (26.7%)	0.33 0.33

Conclusion:

Clinical or subclinical left ventricular dysfunction was identified in a quarter of cirrhotic patients and this seems to correlate with LS, but not with the severity of liver cirrhosis.