Cirrhotic Cardiomyopathy

Hongqun Liu and Samuel S. Lee

University of Calgary Cumming School of Medicine, Calgary, Canada

Introduction

Scientists and clinicians have gradually come to realize the importance of the abnormalities in the cardiovascular system in patients with cirrhosis or portal hypertension. Although the circulation becomes hyperdynamic when stressed, characterized by increased cardiac output and decreased peripheral vascular resistance and arterial pressure, the ventricular response is blunted, manifested by depressed contractile responsiveness and altered diastolic relaxation. The blunted response plus electrophysiologic abnormalities are termed “cirrhotic cardiomyopathy.” Cirrhotic cardiomyopathy is generally latent and overt heart failure is rare because in cirrhosis the characteristic vasculature is systemic vasodilatation that reduces cardiac afterload thus “autotreating” the ventricle. However, major stresses on the cardiovascular system such as hemorrhage [1], infections [2], insertion of transjugular intrahepatic portosystemic shunt (TIPS) [3], and liver transplantation [4] can further damage the heart and unmask the presence of cirrhotic cardiomyopathy and thereby convert latent to overt heart failure [5]. Taking liver transplantation for example, 25–70% of patients develop cardiovascular complications after orthotopic liver transplantation [6]. Up to 56% of transplant recipients have pulmonary edema following surgery [7]. Moreover, cardiac causes are one of the leading causes of death post-transplantation (23%) [6]. The importance of cirrhotic cardiomyopathy has eventually been recognized. This chapter briefly elucidates the whole picture of cirrhotic cardiomyopathy.

Pathology

The abnormalities of the heart in cirrhosis were first described in autopsies 6 decades ago [8] and later confirmed by other studies [9,10]. The gross change includes myocardial hypertrophy, histologic abnormalities include cardiomyocytes, edema, fibrosis, exudation, nuclear vascuolation, and pigmentation. Lunseth et al. [11] documented a subgroup of 12 of 108 autopsied patients in whom they found idiopathic ventricular hypertrophy and vascular dilatation. Interestingly, cardiac hypertrophy seems to occur in the early and moderate stages of cirrhosis; patients with advanced cirrhosis mitigated the cardiac mass increase [11]. This phenomenon was further confirmed by Pozzi et al. [12] with echocardiography and color echo Doppler. They found that left ventricle wall thickness was increased to peak value and decreased in cirrhosis with tense ascites, both right and left atria and right ventricle diameters were significantly greater than controls.

Pathophysiology

Mechanistic Changes

The mechanistic changes in cirrhotic cardiomyopathy include decreased density of β-adrenergic receptors, changes in membrane lipid content and membrane fluidity, and membrane ion channels such as calcium and potassium channels. Collagen shifts from type III to I (from compliant to a less compliant isoform) also impact diastolic function in the cirrhotic heart [13].

Early investigators noticed that patients with cirrhosis have attenuated cardiovascular responses to endogenous and exogenous catecholamines [14]. Since the β-adrenergic receptor system is the main determinant of ventricular contractility, investigators suspected that the β-adrenergic receptor system of the heart may be abnormal in cirrhosis.

Because of the difficulties in obtaining human heart tissue, Gerbes et al. [15] evaluated β-adrenoceptors on human lymphocytes to reflect the status of cardiac β-adrenoceptors. They found that lymphocyte β-adrenoceptor density was significantly lower in patients with cirrhosis and ascites. Later, using the bile duct ligated (BDL) cirrhotic rat model, it was found that cardiomyocyte sarcolemmal plasma membrane β-adrenoceptor density was significantly decreased [16], while the binding affinity of β-adrenoceptor to isoproterenol was unchanged [16,17]. Moreover, it was found that the β-adrenoceptor signal transduction pathway was impaired at different levels, including membrane content and function of the stimulatory G-protein [18], uncoupling of the receptor–ligand complex from G-protein [17], and impaired activity of the adenyl cyclase enzyme [18,19]. Using microarray, Ceolotto et al. [20] found that the β-adrenergic signaling pathway is impaired in the post-receptor cascades which included a significant overexpression of G-protein alpha-inhibiting subunit 2 (Gαi2), cyclic nucleotide phosphodiesterase (PDE2a), regulator of G-protein signaling 2 (RGS2), and downexpression of adenylate cyclase (Adcy3). These post-receptor cascade alterations favor cardiac contractile inhibition and therefore have an important role in cirrhotic cardiomyopathy.

The membrane lipid composition is changed in the cirrhotic heart. The increased cholesterol elevated the cholesterol: phospholipid ratio, which determined the membrane lipid biophysical characteristics. The more rigid membrane in cirrhotic heart limited the movement of the protein receptors embedded in the plasma membrane and, thus, hampered the coupling process between G-proteins and the β-adrenergic-receptor–ligand complex which results in abnormal cardiac contractility [17].

Another mechanistic change is in the ion channels in the membrane. Using bile duct-ligated rats, Ward et al. [21] found reduced expression and density of L-type Ca²⁺ channels in cirrhotic cardiomyocytes and the inward cellular calcium current in these cardiomyocytes was decreased. As is well known, intracellular calcium handling is the key factor to determining the contraction of cardiomyocytes, therefore the reduced L-type Ca²⁺ channel and inward cellular calcium current contribute to the depressed cardiac contractility in cirrhotic heart. Interestingly, in a model of portal vein-ligated (PVL) rats, Zavecz et al. [22] found a similar phenomenon; the density of L-type Ca²⁺ channels in heart from PVL was significantly decreased and L-type Ca²⁺ channels together with the sarcoplasmic reticulum Ca²⁺ pool were the cause of changed excitation–contraction coupling. The decreased density of potassium currents in ventricular myocytes mainly contribute to prolonging the QT interval [23].

Contractile Inhibitory Factors

There are many contractile inhibitory factors in cirrhotic cardiomyopathy such as tumor necrosis factor α (TNFα) [24], endocannabinoid [25], nitric oxide (NO) [26], carbon monoxide [27], and oxidative stress. These inhibitors have an important role in the depressed contractility of cirrhotic heart.

TNFα is increased in serum [24] and heart (unpublished data) in cirrhosis. Using gene knockout or antibody to eliminate TNFα, Yang et al. [24] demonstrated the improved cardiomyocyte contractility in cirrhotic heart. Furthermore, when incubated cardiomyocytes isolated from sham operated mice with TNFα, the cell contraction and relaxation velocities were significantly depressed. These results clearly demonstrated the inhibitory effects of TNFα on cardiac contractility. Endocannabinoid was significantly increased in cirrhotic rats [28] not only systematically, but also locally in the heart [29]. Western blot analysis showed that the quantity of cannabinoid receptor-1 (CB-1) has no change in cirrhotic heart [25,29], the local increase of endocannabinoid results in the cardiac contractile inhibition in cirrhotic heart. CB-1 receptor antagonist significantly improved cardiac contractility both in vivo and in vitro [25,29], while endocannabinoid uptake blockers inhibited cardiac contractility [25]. The negative inotropic effect of NO on cardiac contractility in cirrhosis is via guanosine 3,5′-cyclic monophosphate, NO inhibitor nitro-L-arginine methyl ester significantly improved the inhibited cardiac contractility in cirrhotic heart [26]. Another gas molecule, carbon monoxide, also has a significant role in contractile abnormality in cirrhotic cardiomyopathy [27]. Oxidative stress has been confirmed to have an important role in heart failure [30], Fernando et al. [31] verified that oxidative stress was significantly increased in the rat model of portal hypertension. We found not that only the oxidative stress was significantly increased in cirrhotic heart, but also the antioxidant gene was downregulated. Erythropoietin not only significantly decreased oxidative related DNP-derivatized protein but also activated antioxidant gene, nuclear factor (erythroid-derived 2)-like 2(Nrf2). Erythropoietin thus exerts cardioprotective effects [32].

Clinical Features

For clinical features see Table 20.1.

Table 20.1 Clinical feature and prognosis of cirrhotic cardiomyopathy.

Clinical feature	Prognosis	Evidence	Comments	Reference
Diastolic dysfunction	High mortality after TIPS insertion	E/A <1 at day 28 post-TIPS predicted 1-year survival (60% mortality)	E/A ratio might be a useful predictor in patients who receive TIPS	Cazzaniga et al. (2007) [33]
	Mortality after TIPS insertion	E/A <1 at baseline predicted 5-year survival (60% mortality)	Large two-center study; day-28 post-TIPS E/A not available	Rabie et al. (2009) [34]
	Poor diuresis after TIPS insertion	E/A ≤1 at baseline associated with slow/poor diuresis post-TIPS	Diastolic dysfunction impedes the increase of CO after TIPS, “underfilling” vascular slows clearance of the ascites	Rabie et al. (2009) [34]
Low cardiac output or systolic dysfunction	Patients with low CI have high mortality	The one year mortality was 63% in patients with CI <1.5 l/min/m²	Of 24 patients, the group with the lowest CO had highest mortality at all time points examined	Krag et al. (2010) [35]
	Lower CO precipitates or aggravates HRS post-SBP	Six out of eight patients (75%) with low CO and HRS died after infection resolution	Of 23 patients, 8 developed HRS; this group had lower CO which declined further after infection resolution	Ruiz del Arbol et al. (2003) [36]
	Lower CO Precipitates hepatorenal syndrome	Of 66 patients, 27 who went on to develop HRS during follow-up had lower CO	Decreased CO and MAP were more pronounced in the HRS group	Ruiz del Arbol et al. (2005) [37]
Overt left ventricular failure	Poor outcomes post-liver transplantation?	Reversible LVF occurred in 7/754 patients post-transplant	True prevalence of LVF may be higher than 1% reported in this retrospective series	Sampathkumar et al. (1998) [38]
QTc prolongation	Unclear; risk of torsade de pointes arrhythmia?	Found in 30–60% of patients with advanced cirrhosis	Probably subclinical but no large prospective study to date	(Many studies)
CI, cardiac index; CO, cardiac output; E/A, ratio of early- to late-diastolic filling wave velocity; HRS, hepatorenal syndrome; LVF, left ventricular failure; MAP, mean arterial pressure; QTc, corrected QT interval; SBP, spontaneous bacterial peritonitis; TIPS, transjugular intrahepatic portosystemic shunt.

Systolic Dysfunction

The circulation of cirrhotic patients is hyperdynamic which includes lower mean arterial pressure (MAP), lower systemic vascular resistance (SVR), and higher cardiac output. It is well accepted that cardiac function is normal in most patients at rest due to peripheral vasodilatation [39,40], However, when challenged, the systolic dysfunction is revealed. Wong et al. [41] found that under the challenge of peak physical exercise, the heart rate, ejection fraction, and cardiac index were significantly decreased in cirrhotic patients compared with control subjects; these decreases are more pronounced in decompensated patients and were not related to the etiology of the cirrhosis. Limas et al. [42] infused angiotensin to correct the decreased peripheral vascular resistance back to within a normal range, and found that the pulmonary wedge pressure was increased tremendously but the cardiac output did not change. Using ouabain, a short-acting cardiac glycoside, the same group demonstrated that this significantly shortened the total electromechanical systole, pre-ejection period, and left ventricular ejection time in normal volunteers. Ouabain has no effect on systolic time intervals and cardiac output in cirrhotic patients [42]. Mikulic et al. [43] showed that the infusion of dobutamine, a β₁-adrenergic receptor agonist, did not change cardiac stroke volume significantly in cirrhotic patients. Isoproterenol, a β-adrenergic receptor agonist, required a higher dose to increase 25 beats/minute in cirrhotic patients compared with control volunteers [44]. Lee et al. [16] infused isoprenaline to test the cardiac response and found that compared with sham-operated controls, cirrhotic rats needed a significantly higher dose (four times) of isoprenaline to raise basal heart rate by 50 beats/min and that the maximal chronotropic response in cirrhotic rats was 30% lower than that in controls. Krag et al. [45] investigated the effects of terlipressin on the cardiovascular system in cirrhotic patients. They found that terlipressin increased end diastolic and end systolic volumes, MAP and SVR, but there were no increases in stroke volume or ejection fraction. The total effects on heart were to decrease cardiac output and cardiac index. The reduction in wall motion and wall thickening is related to the severity of the liver disease. Krag et al. concluded that the suppression of cardiac systolic function is related to the degree of hepatic decompensation [45]. The TIPS procedure leads to a sudden increase in cardiac preload. The exact risk of precipitating acute congestive heart failure after TIPS insertion remains inconclusive. Huonker et al. [3] found that TIPS increased right atrial pressure, mean pulmonary artery pressure, pulmonary capillary wedge pressure, left atrial diameter, and left ventricular end diastolic volume. Braverman et al. [46] reported that TIPS caused high-output congestive heart failure; Gines et al. [47] showed that 12% of the patients with TIPS developed heart failure. A recent study by Pimenta et al. [48] showed that systolic function was impaired in those whose brain natriuretic peptide (BNP) was elevated even at rest.

As techniques improve, the conception of “normal cardiac function” in cirrhotic patients at rest is challenged. Unlike conventional two-dimensional echocardiography, which detects fluid movement, tissue Doppler perceives tissue dynamics. Using tissue Doppler, a newer echocardiographic modality, Kazankov et al. [49] demonstrated that myocardial dysfunction is present even at rest. They found that both systolic and diastolic myocardial functions were compromised in cirrhotic patients at rest. Left ventricular ejection fraction (LVEF), mean peak systolic tissue velocity, and mean systolic strain rate were all reduced in cirrhotic patients.