Contribution of Endothelin-1 to the Circulatory and Renal Abnormalities in Patients With Cirrhosis and Portal Hypertension

Patients who present with cirrhosis and portal hypertension exhibit hyperdynamic circulation characterized by low arterial pressure, high cardiac output, low systemic vascular resistance, and persistent activation of many neurohumoral systems including angiotensin II, norepinephrine, vasopressin, and endothelin-1 (ET-1). These circulatory disturbances worsen with disease progression and are accompanied by a reduced vascular reactivity to these vasopressor agonists. The hemodynamic changes associated with cirrhosis are believed to play a crucial role in the pathogenesis of portal hypertension and in the development of its complications, namely variceal hemorrhage, ascites, and the hepatorenal syndrome. Therefore, a better understanding of the role of such neurohumoral mediators in the pathogenesis of portal hypertension may lead to the development of new therapeutic strategies.

ENDOTHELIN SYSTEM IN HEALTH

Endothelin-1 is one of the most potent vasoconstrictors known, belonging to a 21-amino acid peptide family with a range of biologic effects. Endothelin-1 was originally identified in the culture medium of porcine aortic endothelial cells and is derived from a larger pre-proendothelin-1 (212 amino acids), which is cleaved by endopeptidases to produce big endothelin-1 (38 amino acids), which in turn is converted to ET-1 by specific endothelin-converting enzymes. Molecular studies have identified two endothelin receptor subtypes in mammalian species: endothelin-A (ET_A) and endothelin-B (ET_B). In vascular smooth muscle cells, both receptors are expressed and these mediate vasoconstriction. Only the ET_B receptors are found on endothelial cells and cause vasodilatation through the release of endothelium-derived vasodilators, such as nitric oxide (NO). Endothelin-1-induced vasoconstriction is predominantly mediated by the ET_A receptor, but ET_B receptors may contribute under some circumstances (Fig. 1) (¹). Antagonizing the effects of endogenously produced ET-1 in healthy humans using BQ-123, the ET_A receptor antagonist, decrease mean arterial pressure and systemic vascular resistance. The reverse was shown by systemic infusion of BQ-788, the ET_B receptor antagonist.

Fig. 1

In healthy adult humans, infusion of low dosages of ET-1, which resulted in a twofold increase in plasma levels, decreased sodium excretion by 36%, without a significant effect on systemic and renal hemodynamics. However, infusion of higher doses of ET-1 causes sodium retention and also increased renal vascular resistance by 37%. Pretreatment with the calcium channel blocker nifedipine caused renal vasodilation, compensated for the renal vasoconstriction by ET-1, and prevented sodium retention (²). This observation provides an insight into the mechanism of ET-1-induced vasoconstriction, which in common with other vasopressor agents induces a cascade of second messengers, leading ultimately to an increase in intracellular calcium, which is the signal for smooth muscle contraction.

ENDOTHELIN SYSTEM IN CIRRHOSIS

Since 1991, many studies have shown that patients with advanced cirrhosis and fluid retention have increased plasma ET-1 concentrations, which correlate with disease severity (Fig. 2) and are particularly high in patients with hepatorenal syndrome (^3,4). Studies of liver biopsy specimens have demonstrated synthesis of ET-1 in many cells, including hepatic stellate cells, endothelial cells, and splenic cells. Other studies that assessed regional ET-1 production have shown the splanchnohepatic region as a major source of ET-1 spillover into the circulation, although the kidney and other regions also may contribute. Following liver transplantation, ET-1 levels decrease rapidly and correlate with improvements in both liver and renal function (⁵). The same has been observed after transjugular intrahepatic portosystemic stent shunt placement (⁶), although this seems not to be universal (⁷). The increase in plasma ET-1 concentrations corresponds with up-regulation of pre-proendothelin-1 mRNA and increased plasma concentrations of big ET-1 (⁴), indicating that ET-l levels reflect increased production, rather than decreased clearance. This is expected since the liver plays little role in ET-1 elimination, which occurs mainly through the lungs and kidneys. The mechanisms underlying the elevated ET-1 concentrations are unclear, and do not appear to be related to hypovolemia or endotoxemia (⁸). Moreover, the plasma concentrations of ET-1 correlate positively with the severity of liver disease as measured by Child-Pugh score (^9,10), with the hepatic blood flow as measured by D-sorbitol infusion, and with the hepatic venous pressure gradient (⁷), and correlate inversely with liver cell mass as measured by galactose elimination capacity (¹⁰) and renal function as measured by creatinine clearance (⁹). However, due to its paracrine and autocrine mode of action, the majority of ET-1 is released abluminally to act on adjacent vascular smooth muscle and endothelial cells and, as a consequence, plasma ET-1 concentrations may not truly reflect the underlying activity of the endothelin system, especially if they appear normal (¹¹). It should also be noted that ET-1 may act as an endocrine hormone when plasma concentrations are increased.

Fig. 2

Whether activation of the endothelin system plays a role in initiation and perpetuation of portal hypertension, or occurs as a secondary event to compensate for the decreased systemic vascular resistance still must be clarified. However, during the past decade, many studies have suggested a role of ET-1 in the pathogenesis of portal hypertension and its complication. Most of these studies are in the rat model of cirrhosis (^12-14), and the direct relevance of this to cirrhosis in humans is unclear. Therefore, much remains to be learned, especially from the use of selective ET-1 antagonists in patients with cirrhosis. This may have therapeutic implications.

ET-1 AND VASCULAR TONE IN CIRRHOSIS

We have demonstrated impaired vasoconstriction to exogenous ET-1 infusion and enhanced vasodilatation in response to the ET_A receptor antagonist BQ-123 in patients with well-compensated cirrhosis who have normal plasma concentrations of ET-1. These patients demonstrated normal responses to the ET_B receptor antagonist BQ-788 (⁴). These findings suggest that modest activation of the endothelin system can occur without a change in plasma ET-1 concentrations and are consistent with: 1) an endothelin system activated before the development of ascites; 2) a greater contribution of endogenous ET-1 to basal peripheral vascular tone acting through the ET_A receptor; and 3) a potential counterregulatory role for ET-1 in the pathogenesis of the circulatory disturbances of cirrhosis.

EFFECTS OF SYSTEMIC ET-1 ANTAGONISTS

As yet, the only available report of systemic infusion of an endothelin receptor antagonist in humans who present with cirrhosis is of infusing the ET_A receptor antagonist BQ-123 in three patients with advanced cirrhosis and hepatorenal syndrome. Interestingly, this study showed improvement in renal blood flow and creatinine clearance (¹⁵). However, experimental studies of cirrhosis and portal hypertension in the rat model have shown reductions in splanchnic blood flow and portal venous pressure by the nonselective endothelin antagonists TAK-044 and bosentan (^12,13) and the selective ET_B receptor antagonist IRL1038, whereas selective ET_A receptor antagonism produced no clear effects (¹⁴). Endothelin B receptor overexpression in the splanchnic vasculature was demonstrated in the same rat model (¹⁵). Taken together, these findings are consistent with a role of ET-1 in the pathogenesis of splanchnic vasodilatation mediated through the ET_B receptor, and in the pathogenesis of hepatorenal syndrome mediated through the ET_A receptor. Further studies assessing the clinical role of systemic ET-1 antagonists in patients with cirrhosis and portal hypertension are awaited eagerly.

As reported in the current issue of the Journal, Bakr et al. (¹⁶) have measured the peripheral venous plasma concentrations of ET-1 in children with cirrhosis (with and without ascites) and in age-and sex-matched healthy controls and have shown a correlation between ET-1 levels and mean arterial pressure, and renal function as measured by creatinine clearance, but not with portal hemodynamic parameters as measured by Doppler ultrasound.

As mentioned earlier, measuring plasma concentrations of ET-1, in isolation, will not accurately assess the endogenous activity of the endothelin system and should be accompanied with simultaneously measuring the plasma levels of its biologic precursor, big ET-1; the in vivo tonic activity of ET-1; or the vascular expression of the mRNA of endothelin and its receptors. In other words, decreased plasma concentrations of ET-1 may accompany enhanced tonic activity of the endothelin system, whereas increased circulating levels always indicate increased activity, provided that the ET-1 clearance mechanisms are not altered.

In patients with cirrhosis and ascites, plasma concentrations of vasoactive mediators such as ET-1, angiotensin II, and vasopressin can be influenced by many factors including the time of obtaining the blood sample, the type of preservative used, whether peptide extraction was performed before the radioimmunoassay, the posture of patients, the time since assuming that posture, the severity and cause of liver disease, any concomitant diuretic therapy, presence and degree of cholestasis, dietary salt intake, and the coexistence of any factor that influences endothelium function such as diabetes mellitus and hypercholesterolemia. Control of these factors is required to decrease the variation seen in the available literature and to allow sound interpretation of the results. In the study of Bakr et al. (¹⁶), diuretic therapy may have influenced the results. Ideally, drugs should be stopped for at least five times their half-life. In addition, there is no indication that the above-mentioned factors were taken into consideration.

Bile acids have vasodilator properties and circulate in increased levels in patients with primary biliary cirrhosis. Indeed, plasma concentrations of bile acids correlate with disease severity and with the blood levels of liver function tests. Pronounced cholestasis may increase the production of and alter the vascular responses to any vasopressor agent, including ET-1. Indeed, plasma ET levels have been shown to be high in patients with biliary atresia, especially those with severe biliary cirrhosis, and may contribute to the development of portal hypertension as well as being a useful marker in the postoperative follow-up of these patients (¹⁷). Interestingly, in the study of Bakr et al. (¹⁶), there was no significant difference in plasma ET-1 levels between the cirrhotic patients with cholestasis and those without cholestasis. This is consistent with our previous finding of a normal response to norepinephrine in patients with mild cholestasis (¹⁸) and confirms our view that a mild degree of cholestasis (bilirubin levels < 50 μmol/L) has no significant effect on vascular reactivity or vasopressor production.

We (⁶) and others (^19,20) have shown normal plasma ET-1 concentrations in adult patients with preascitic cirrhosis. In addition, the two studies cited by Bakr et al. did not show any significant difference in plasma ET-1 concentrations between pediatric cirrhotic patients without ascites and healthy controls. Therefore, high ET-1 levels in the preascitic children included in the study of Bakr et al. is unexplained.

Duplex Doppler ultrasonography allows identification of a number of parameters related to the presence of portal hypertension, such as the increase in portal vein diameter, the lack of respiratory change in portal, the splenic and mesenteric vein diameters, the presence of a portal-collateral circulation, a decreased or reversed portal vein velocity, an increased congestion index of the portal vein, and the presence of ascites and splenomegaly. However, the association between splanchnic Doppler parameters and portal pressure has been reported as generally weak (^21,22). Indeed, recent practice guidelines suggest the use of Doppler ultrasonography in the first-line assessment of patients with suspected portal hypertension, but do not support its use in graduating the severity of portal hypertension (²³). Therefore, the lack of correlation between the plasma concentrations of ET-1 with portal flow volume and portal flow reported by Bakr et al. is not surprising and does not exclude an association between plasma ET-1 concentrations and the severity of portal hypertension. When interpreting the results of this article, it should be noted that: 1) the authors measured ET-1 concentrations in the peripheral blood but not in the splanchnic circulation, and that regional differences in ET-1 production is well documented; 2) plasma ET-1 concentrations may not reflect the actual activity of the endothelin system due to its abluminal release; 3) the examiners were only masked to the laboratory data but not to whether the child was a patient or a healthy control; 4) plasma ET-1 concentration might have been affected by the use of diuretics in some patients and the presence of cholestasis in others; and 5) Doppler parameters are operator-dependent and poor predictors of portal pressure.

ENDOTHELIN-1 AND KIDNEY FUNCTION

Patients who present with liver cirrhosis and ascites frequently experience impaired renal function, which is characterized by increased serum creatinine concentration with sodium and water retention in the absence of any identifiable specific causes of renal dysfunction. The range of this dysfunction varies from subtle sodium retention in the preascitic stage of cirrhosis to the complete picture of hepatorenal syndrome. The exact basis of the impaired renal function in these patients is not yet understood fully. However, altered renal hemodynamic responses appear to be involved. This is evidenced by the successful transplantation of kidneys from patients with hepatorenal syndrome.

Increased levels of ET-1 have been demonstrated not only in cirrhosis but also in various pathologic conditions that are characterized by sodium retention or renal vasoconstriction, such as heart failure, hepatorenal syndrome, and chronic renal failure and have been demonstrated during administration of cyclosporine and radiocontrast. Bakr et al. reported a negative correlation between creatinine clearance and plasma ET-1 concentrations. This supports the available literature indicating that ET-1 plays a role in the pathogenesis of renal dysfunction in adults and children with cirrhosis (^3,9,24). However, no correlation was observed between renal vasoconstriction, as detected by Doppler ultrasound technique, and plasma concentrations of ET-1 in these patients (²⁵). Again, the unmeasured intrarenal production of ET-1, and not the circulating peptide, may contribute to the observed renal vasoconstriction and consequently may explain this lack of correlation. The potential beneficial therapeutic effect of the use of ET_A antagonists suggested by Soper et al. (¹⁵) must be verified in a bigger cohort of patients with hepatorenal syndrome.

In summary, the endothelin system is activated in adults and children with cirrhosis starting from the preascitic stage, and its activity correlates with disease severity. Endothelin-1 plays a role in the maintenance of basal vascular tone and also probably contributes to the pathogenesis of portal hypertension and renal dysfunction in these patients. The effects of systemic doses of endothelin-receptor antagonists on systemic, renal, and hepatic hemodynamics in patients with cirrhosis may lead to novel drug therapies for the complications of portal hypertension.