Origins of cardiac dysfunction in cirrhosis

The clinical course of patients with advanced liver disease is complicated by progressive impairment in circulatory function characterised by low arterial pressure, high cardiac output, and decreased systemic vascular resistance.¹ Clinical and experimental investigations performed during the past two decades have shed light on the multiple mechanisms accounting for these disturbances. These studies have also established the pathogenic role of circulatory dysfunction in organ specific syndromes that commonly develop in cirrhotic patients, such as the hepatorenal and the hepatopulmonary syndromes.^2,3 The heart is another functionally compromised organ in cirrhotic patients. However, whether the hyperdynamic circulation, by overloading the heart, induces cirrhotic cardiomyopathy or whether this is a specific cardiac dysfunction of cirrhotic patients has been subject of extensive discussions.⁴

Cardiac function abnormalities in cirrhosis are clinically not apparent. However, when cardiac function is explored, a reduction in right ventricular volume, probably secondary to reduced venous return, and left ventricular dysfunction, characterised by left ventricular preload and volume, are observed.^5,6 Moreover, cardiac structural abnormalities, including hypertrophy of the myocardium and increased left ventricle thickness and hence diastolic dysfunction, have also been described.⁷ Cirrhotic cardiomyopathy is latent, probably because of the low peripheral vascular resistance presented by these patients, which reduces cardiac afterload. The existence of an abnormal ventricular behaviour can however be unveiled during exercise or following pharmacological stress. It has been demonstrated that left ventricular end diastolic pressure increases and stroke index and left ventricular ejection fraction decrease more in cirrhotic patients than in control subjects.^8–10

Impaired left ventricular performance in cirrhotic patients was initially thought to be due to the so-called alcoholic heart muscle disease, also known as alcoholic cardiomyopathy,¹¹ because almost all earlier studies were performed in alcoholic patients. However, clear dissimilarities between alcoholic and cirrhotic cardiomyopathy exist. Firstly, depressed ventricular responsiveness has been observed in humans and rats with cirrhosis of non-alcoholic aetiology.^12,13 On the other hand, alcoholic heart muscle disease is secondary to impaired contractile protein synthesis and formation of immunogenic cardiac protein acetaldehyde adducts¹⁴ whereas clearly differentiated mechanisms are involved in the pathogenesis of cirrhotic cardiomyopathy.

Several studies have shown sympathetic and parasympathetic autonomic dysfunction in cirrhotic patients.¹⁵ Hypotheses have been raised suggesting that the origin of this abnormality could be located in the nervous system due to damage of the peripheral nerves or because of changes in endogenous neurotransmitters.^8,16 Impaired β adrenergic signal transduction may also be an important element in the pathogenesis of cirrhotic cardiomyopathy. Experimental studies have shown decreased β adrenergic receptor density and receptor desensitisation in cardiocytes of cirrhotic rats.¹⁷ In addition, leucocytes of cirrhotic patients also present decreased abundance of β adrenoreceptor.¹⁸ Heart receptor and post receptor defects are supported by the demonstration of reduced function and expression of cardiac G proteins in cirrhotic animals¹⁷ and impaired cardiac excitation-contraction coupling in portal hypertensive rats.¹⁹ Plasma membrane fluidity and ion channel function are impaired in cirrhosis.²⁰ Recently, Ward and colleagues²¹ described a decrease in K⁺ current in ventricular myocytes of cirrhotic rats, which would result in a tendency to prolong QT intervals. This is in agreement with the results of Bernardi and colleagues²² showing a prolonged QT interval and other electrophysiological abnormalities in cardiac excitation and repolarisation in cirrhotic patients. Exposure of cardiac myocytes for long periods of time to endogenous substances with cardiac function inhibitory properties should also be taken into consideration. There is a wide array of cardiodepressant factors such as nitric oxide, endotoxins, endothelins, bile acids, and cytokines, that have been demonstrated to be increased in cirrhotic patients and experimental models of portal hypertension.^1,23,24

Recently, brain natriuretic peptide (BNP), a cardiac hormone belonging to the natriuretic isopeptide family, has attracted increasing attention as an accurate marker of left ventricular dysfunction. In fact, BNP is an independent predictor of high left ventricular pressure,²⁵ estimates left ventricular systolic dysfunction, and closely correlates with the New York Heart Association (NYHA) classification.²⁶ The accuracy of BNP for the detection of left ventricular systolic dysfunction is similar to that of prostate specific antigen for the detection of prostate cancer, and is superior to that of mammography for breast carcinoma and Papanicolau smears for cervical cancer.²⁷ BNP is released from cardiac ventricles in response to ventricular volume expansion and pressure overload, suggesting that BNP levels are a more sensitive and specific indicator of ventricular disorders than other natriuretic peptides. Data from heart failure investigations suggest that the increased release of BNP is a compensatory response elicited by ventricular remodelling aimed at reducing systemic pressure load hypertrophy through sodium and water diuresis. Thus BNP has become a specific marker of ventricular damage rather than just an indicator of volume overload.²⁸

Cardiac natriuretic peptides, namely atrial natriuretic peptide and BNP, have long been known to be elevated in cirrhotic patients as a consequence of increased cardiac release and not because of impaired hepatic extraction.^29,30 However, they have been generally regarded as markers of volume overload rather than markers of cardiac dysfunction. Recently, Wong and colleagues³¹ proposed that BNP could be an indicator of cirrhotic cardiomyopathy. These authors measured cardiac natriuretic peptide levels and cardiac structural parameters in a group of 36 cirrhotic patients with and without ascites. Increased circulating levels of BNP were related to septal thickness and left ventricular diameter at the end of diastole.³¹

In fact, although some authors considered cirrhotic cardiomyopathy, at best, a complication of alcoholic liver disease and, at worst, a non-existent medical invention, numerous evidence supports the concept of a specific cardiac disorder peculiar to cirrhosis. What is still an unanswered question is whether this abnormality results from the hyperdynamic circulation also present in these patients. In the current of Gut issue, Herikssen and colleagues³² use an elegant experimental approach to solve this dilemma [see pages 1511–7]. These investigators have simultaneously assessed plasma levels of BNP and total proBNP, and indicative parameters of liver and cardiac dysfunction and hyperdynamic circulation in a large group of cirrhotic patients. ProBNP is the high molecular precursor of functionally active BNP. Cleavage of proBNP is mainly located in the myocyte and results in secretion to the systemic circulation of equimolar amounts of the N terminal fragment of proBNP (NT-proBNP) and BNP. NT-proBNP circulates at considerable concentrations in human plasma, is stable in human blood, and is less dependent on pulsatile fluctuations, produced by postural changes or other physiological responses, than BNP. Total proBNP measurement is performed after in vitro plasma trypsinisation and it has been suggested that this is a more reliable method to assess BNP secretion as it does not depend on precursor processing.³³ Confirming previous investigations, cirrhotic patients showed increased circulating levels of BNP, which paralleled the results obtained on analysing total proBNP. The most interesting finding of this study is that ventricular natriuretic peptide secretion closely correlates with indicative parameters of abnormal liver (Child score, hepatic venous pressure gradient, and serum albumin) and cardiac function (plasma volume, heart rate, and QT interval) but not with those characteristic of the hyperdynamic circulation (cardiac output and systemic vascular resistance). Therefore, these results seriously jeopardise the concept that increased BNP levels in cirrhotic patients are due to the hyperdynamic circulation. Rather, they support the fact that increased secretion of this natriuretic peptide is a consequence of ventricular dysfunction, which seems to progress in parallel with the severity of the liver disease. These findings should also stimulate further research to clearly delineate the molecular and cellular mechanisms responsible for the structural and functional abnormalities distinctive of cirrhotic cardiomyopathy. Identification of well defined therapeutic targets will certainly improve life quality and expectations of patients with advanced liver disease.