Hepatorenal syndrome (HRS) is the most dreaded complication of patients with cirrhosis because of its frequency and high mortality rate.1 The current median survival time of patients with cirrhosis and HRS is only 3 months.2 Until recently, there was no effective therapy for HRS other than liver transplantation. However, waiting-list mortality is higher and posttransplantation survival is lower in patients with HRS, particularly in those with type 1 HRS, than in patients with normal renal function.3 Fortunately, the management of HRS is entering a new era with the use of several treatment options, including vasoconstrictor drugs and transjugular intrahepatic portosystemic shunts, and new methods of albumin dialysis.4 Of all these methods, pharmacological therapy with vasoconstrictors is the one that has been studied the most extensively. A number of recent studies have shown that the administration of vasoconstrictors, particularly vasopressin analogues, is associated with a reversal of HRS in approximately one-half of patients administered this therapy.5, 6 The use of vasoconstrictors in the management of patients with cirrhosis and renal failure is not new. In fact, as early as 1956, Hecker and Sherlock7 reported the effects of treatment with noradrenaline in three patients with cirrhosis, hyponatremia, and renal failure. The rationale for the use of noradrenaline was the observation that in patients with cirrhosis, functional renal failure (later called HRS) could be the result of profound arterial vasodilation causing hyperdynamic circulation and hypotension, with subsequent reduction in renal plasma flow and glomerular filtration rate. Unfortunately, these early attempts to treat HRS with vasoconstrictor drugs were not pursued, and this treatment was rapidly forgotten. A factor that probably contributed to this was the publication of the overflow theory of ascites formation.8 In contrast to the proposal by Hecker and Sherlock, the overflow theory proposed that renal failure was due not to arterial vasodilation but rather to hypovolemia related to the passage of fluid from the overfilled intravascular compartment to the peritoneal cavity or to a hepatorenal reflex causing vasoconstriction of the renal circulation.9 On the basis of this theory, attempts were made to improve renal function by either expansion of plasma volume (by head-out water immersion, administration of plasma expanders, or peritoneovenous shunting) or treatment with vasodilators.10 However, none of these treatments proved to be effective. Fortunately, the pendulum swung back, and the arterial vasodilation theory was reformulated because of overwhelming evidence in both human and experimental cirrhosis in favor of arterial vasodilation as the cause of ascites formation and renal functional abnormalities in cirrhosis.11 In two landmark studies, the group of Robert Schrier in Colorado showed that the combination of head-out water immersion (a maneuver that increases central blood volume by redistributing blood volume from the peripheral vascular bed to the central vascular bed) with administration of noradrenaline to patients with cirrhosis and ascites without renal failure was associated with an increase in arterial pressure, suppression of the renin-aldosterone system, and an increase in sodium and free-water excretion.12, 13 These two studies, in addition to providing support for the arterial vasodilation hypothesis of renal function abnormalities and ascites formation in cirrhosis, paved the way for therapeutic studies in patients with HRS using the approach of combining plasma volume expansion and administration of vasoconstrictor drugs. Thus, subsequent studies performed in the late 1990s demonstrated that HRS could be reversed in a significant proportion of patients with the combination of albumin and ornipressin, a vasopressin analogue, with marked vasoconstrictor activity on the splanchnic circulation.14, 15 More recent studies have used terlipressin, a vasopressin analogue with a better safety profile than ornipressin. The results of randomized controlled studies indicate that the administration of terlipressin (0.5–1 mg/4–6 hours as an intravenous bolus) in combination with albumin is associated with reversal of HRS, as defined by a reduction in serum creatinine below 1.5 mg/dL, in up to 40% of patients and that reversal of HRS is associated with improved survival.16, 17 The sample size of these studies has not been large enough to specifically address whether treatment with terlipressin and albumin is associated with improved survival. Side effects related to the administration of terlipressin and albumin have been reported in some patients, mainly ischemic complications in the heart, extremities, and splanchnic organs, as well as reversible pulmonary edema. Therefore, patients treated for HRS should be subjected to cardiac monitoring and frequent clinical surveillance for early detection of ischemic complications.
Despite the relatively large number of studies reported on the management of type 1 HRS with terlipressin (reviewed by Moreau and Lebrec5 and Arroyo et al.6), a number of issues remain either unanswered or poorly understood. First, the dose of terlipressin with the best efficacy/safety ratio is not known. Second, serum creatinine levels have been shown to be a predictor of response in several studies, so the higher the levels are at the start of therapy, the lower the probability of response is; this suggests that treatment of HRS should be started as early as possible in the course of renal failure. Whether or not very high creatinine levels (that is, greater than 4–5 mg/dL) are indicative of superimposed acute tubular necrosis is not known and deserves specific investigation; patients with these creatinine levels should not be excluded from future HRS studies but should be randomized separately from those with lower creatinine levels. Other potentially important predictors of response have not been investigated, including the severity of circulatory failure, the presence of associated cardiomyopathy, the existence of systemic inflammatory response, and the possible presence of adrenal insufficiency. A further issue also unknown is whether the response rate of patients with HRS with concomitant bacterial infections is similar to that of patients with HRS without bacterial infections. Finally, it is not known whether vasoconstrictor drugs other than terlipressin, such as noradrenaline, have an efficacy/safety ratio similar to that of terlipressin; the various small studies that have been carried out suggest that noradrenaline is also effective, but large studies are needed to definitively answer this question.18, 19
Most of the existing information on the use of terlipressin or other vasoconstrictor drugs in the management of patients with HRS refers to type 1 HRS, the more aggressive form of HRS with a high mortality rate.2 By contrast, little information exists on the use of terlipressin or other vasoconstrictor drugs in the management of type 2 HRS. Compared to patients with type 1 HRS, patients with type 2 HRS have a less severe impairment of renal function (average glomerular filtration rate and serum creatinine, 30 mL/minute and 1.8 mg/dL, respectively), which is usually stable for weeks or months (although it may eventually progress to type 1 HRS) and is associated with less severe impairment of liver function, lower Model for End-Stage Liver Disease scores, absence of multiorgan failure, and a better prognosis (3-month probability of survival of 70% in type 2 HRS compared to 10% in type 1 HRS).2 HRS type 2 represents a potential indication for therapy with vasoconstrictor drugs, but only a few studies have analyzed the effect of treatment in these patients. Although the existing information is very limited, it appears that treatment with terlipressin is effective in improving renal function in patients with type 2 HRS.20, 21 Recurrence of HRS appears to be high, although this has not been completely established. It would be of major importance to investigate whether the improvement of renal function obtained with vasoconstrictor therapy in patients with type 2 HRS results in a lower incidence of development of type 1 HRS and improved survival.
Another clinical condition in which vasoconstrictor therapy needs to be evaluated is refractory ascites, a condition that represents a major challenge for the physician caring for patients with liver diseases. Refractory ascites is characterized by ascites with poor or no response to diuretics, need for large-volume paracentesis or transjugular intrahepatic portosystemic shunt insertion, and poor long-term survival.4 In approximately one-third of patients, refractory ascites is associated with type 2 HRS, whereas in the remaining patients, the glomerular filtration rate is relatively preserved.22 The pathogenesis of refractory ascites is similar to that of HRS, with circulatory dysfunction due to splanchnic arterial vasodilation resulting in activation of antinatriuretic and vasoconstrictor systems, but to a lesser extent in comparison with type 1 HRS. Therefore, it is possible that the administration of vasoconstrictors may reverse some of the pathogenic events that result in increased renal sodium retention and refractoriness to diuretic therapy. In this regard, the study reported by Krag et al.23 in the current issue of HEPATOLOGY supports this hypothesis. This study included 23 patients with cirrhosis and ascites, 8 with refractory ascites and 15 with nonrefractory ascites. All patients except 4 of the nonrefractory ascites group who received placebo were treated with a single dose of 2 mg of terlipressin. Renal function was evaluated comprehensively before and after the administration of terlipressin or placebo. The administration of terlipressin had marked beneficial effects on renal function in patients with and without refractory ascites, whereas no significant effects were observed in patients treated with placebo. Specifically, terlipressin administration was associated with a significant increase in renal plasma flow, glomerular filtration rate, sodium clearance, lithium clearance, osmolar clearance, and sodium excretion. Moreover, there was suppression of activity of the renin-angiotensin system and sympathetic nervous system, which was indicative of an improvement in circulatory function. These effects are reminiscent of the observations in patients with HRS type 1 treated with terlipressin21 and indicate that terlipressin also improves renal and circulatory function in patients with cirrhosis without HRS. Therefore, if further studies using multiple doses of terlipressin for prolonged periods of time confirm these observations, terlipressin could theoretically be used in the management of refractory ascites. However, the drawback of terlipressin is that it has to be administered by an intravenous route, which precludes applicability in clinical practice where most patients with refractory ascites are outpatients. An alternative to terlipressin that could be used in outpatients is the orally active α-adrenergic agonist midodrine.24 A recent study showed that the administration of midodrine for 7 days to patients with cirrhosis and ascites was associated with an improvement in circulatory and renal function and sodium excretion.25 Further studies are clearly needed to evaluate the possible beneficial effects of long-term administration of midodrine in patients with cirrhosis and ascites.
In summary, randomized studies have demonstrated that vasoconstrictor drugs are effective in the management of HRS. Moreover, the study by Krag et al.23 and other studies suggest that therapy with vasoconstrictor drugs may also be useful for treatment of refractory ascites, a condition for which no effective pharmacological therapy exists. Finally, the possible role of orally active vasoconstrictors in the prevention of HRS needs to be investigated. Ideally, these new potential indications for vasoconstrictor drugs should be evaluated in large, randomized, double-blind studies with comparison to placebo. Because most of these drugs have been on the market for many years for indications different than those currently proposed, these studies are unlikely to be undertaken by the pharmaceutical industry. Therefore, a large collaborative effort should be initiated by clinical investigators to study these drugs.