Conivaptan increases serum sodium in hyponatremic patients with end-stage liver disease

Abstract

Hyponatremia is associated with increased mortality in patients with end-stage liver disease and a greater risk of perioperative mortality with liver transplantation. We performed a retrospective review of our experience with conivaptan as a means of acutely increasing serum sodium in end-stage liver disease patients. The primary group consisted of 15 patients with end-stage liver disease who remained hyponatremic despite discontinuation of diuretics and a 1-L fluid restriction. Twenty milligrams of conivaptan was intravenously administered over 30 minutes, and this was followed by an infusion of 20 mg over 24 hours for 1 to 4 days. A second group of 9 hyponatremic end-stage liver disease patients was treated with 1-L fluid restriction and conivaptan while remaining on diuretics. In the group without diuretics, the mean serum sodium was 124 mmol/L 1 day before and on the day of conivaptan initiation, but the serum sodium rose to a mean of 127.7 mmol/L by day 1 and further increased to 128.6 mmol/L by the second day of the infusion. Despite the continuation of diuretics, the second group of 9 patients also had an increase in serum sodium from the day of conivaptan initiation (125.7 mmol/L) to 2 days after the treatment (130.6 mmol/L). Eleven patients underwent successful liver transplantation, 2 remained on the list for transplantation, and 11 were not candidates for transplantation and either died (7) or were discharged home and lost to follow-up (4). In conclusion, a short course of conivaptan increases serum sodium in patients with end-stage liver disease and may reduce the risk of proceeding to liver transplantation. Further study in a prospective clinical trial is needed to confirm safety and efficacy. Liver Transpl 15:1325–1329, 2009. © 2009 AASLD.

Hyponatremia is an independent predictor of mortality in patients with end-stage liver disease and is associated with more severe liver disease and a greater risk of ascites, hepatorenal syndrome (HRS), and liver-related death.1–4 As a result, it has recently been proposed that it be incorporated into the Model for End-Stage Liver Disease (MELD) score that is currently used to prioritize patients awaiting liver transplantation.2 The MELD score uses serum bilirubin, creatinine, and the international normalized ratio to provide an objective measure of severity of illness that accurately predicts short-term mortality.5 The addition of serum sodium to the MELD score (MELD-Na) would raise the score and priority of approximately 7% of the sicker patients on the waiting list.2 However, it has not yet been shown whether transplanting patients with significant hyponatremia would reduce waiting list mortality. Indeed, hyponatremia has been shown to increase perioperative liver transplant mortality, and prioritization of these patients may negatively affect outcomes.6–8 Furthermore, serum sodium is subject to the influence of diuretic therapy, water restriction, and adrenal insufficiency, which are common in patients with cirrhosis, or spurious reductions due to hyperglycemia.9 As a result, MELD-Na might be subject to manipulation and gaming of the donor organ priority system. Thus, the incorporation of serum sodium into the MELD score remains controversial, and the United Network for Organ Sharing has so far been reluctant to change to MELD-Na.

Medical management to correct hyponatremia remains important in patients with cirrhosis. The standard of care includes fluid restriction and reduction or discontinuation of diuretics.10 However, sodium is corrected slowly, and compliance with fluid restriction is poor outside the hospital. At this time, no drug is Food and Drug Administration (FDA)–approved to treat hyponatremia in patients with end-stage liver disease. However, vasopressin-2 receptor antagonists have been developed, and a recent trial reported the effectiveness of tolvaptan for dilutional hyponatremia in a cohort of patients with congestive heart failure, the syndrome of inappropriate antidiuretic hormone secretion, or cirrhosis.11 Several studies have now found this class of compounds to be safe and effective at increasing serum sodium in patients with cirrhosis, but none of those reported to date are available for use in clinical practice.10–17

Conivaptan is a vasopressin-1 and vasopressin-2 receptor antagonist that is administered intravenously and is FDA-approved for the treatment of euvolemic or hypervolemic hyponatremia in hospitalized patients with normal liver function. We reviewed our experience with conivaptan to treat hyponatremia in patients with end-stage liver disease.

AIH, autoimmune hepatitis; FDA, Food and Drug Administration; HBV, chronic hepatitis B; HCV, chronic hepatitis C; HRS, hepatorenal syndrome; INR, international normalized ratio; MELD, Model for End-Stage Liver Disease; MELD-Na, Model for End-Stage Liver Disease with sodium; NS, not significant; PBC, primary biliary cirrhosis; SBP, spontaneous bacterial peritonitis or systolic blood pressure.

PATIENTS AND METHODS

We performed a retrospective evaluation of the effectiveness of conivaptan treatment of hyponatremic patients with end-stage liver disease. All patients were 18 years old or older, had end-stage liver disease (defined as cirrhosis with decompensation manifested by ascites and synthetic dysfunction), and were admitted to Baylor University Medical Center between November 1, 2006 and September 27, 2008 for moderate to severe hyponatremia (serum sodium < 130 mmol/L) that had failed to be corrected with outpatient management.18 Because all were initially considered to be potential candidates for liver transplantation and our center requires a serum sodium level of at least 128 mmol/L for that procedure, we are aggressive about correcting serum sodium levels. All subjects were placed on 1-L fluid restriction and 2-g dietary salt restriction. Intravascular volume status was clinically assessed as euvolemic or hypervolemic in all patients. Nonetheless, because of elevated serum creatinine, infection, hypoalbuminemia, or low urine output, 83% of patients received 25 g of intravenous albumin 2 to 3 times per day for 3 days, and 67% of patients received an initial fluid challenge of at least 500 mL of intravenous normal saline. All patients had ascites. Paracentesis was performed to rule out spontaneous bacterial peritonitis (SBP) when ascites was clinically apparent and sufficient fluid was found with ultrasound for the procedure to be safely performed.

All patients except 2 were observed for at least 1 day with fluid and sodium restriction alone as a control period. Thereafter, 20 mg of conivaptan was administered intravenously over 30 minutes, and this was followed by an infusion of 20 mg over 24 hours for 1 to 4 days. Diuretics were discontinued in 15 patients and continued in 9 patients. This was a clinical decision at the discretion of the attending physician and was usually based on the degree of ascites and peripheral edema. Patients were analyzed in 2 groups depending on whether diuretics were discontinued or not. Paired t tests compared the day before conivaptan initiation to the day of initiation and the day of initiation to 1 and 2 days after initiation with Bonferroni correction.

This review was approved by the Baylor University Medical Center institutional review board.

RESULTS

A total of 24 patients with end-stage liver disease and hyponatremia (serum sodium ≤ 130 mmol/L) were treated with conivaptan (Table 1). Group 1 consisted of 15 patients whose diuretics were discontinued. Their day 0 median serum sodium was 124 (range, 121–127). All had ascites, 2 had SBP, and 3 had HRS. Twelve of the 15 patients (80%) received intravenous albumin. Their median age was 52 years, 67% were male, and their median MELD score was 20. The cause of cirrhosis was hepatitis C in 7 patients, alcohol in 6 patients, autoimmune hepatitis in 1 patient, and primary biliary cirrhosis in 1 patient.

Table 1. Patient Characteristics According to Whether Diuretics Were Discontinued (Group 1) or Continued (Group 2)

	Group 1	Group 2
Abbreviations: AIH, autoimmune hepatitis; HBV, chronic hepatitis B; HCV, chronic hepatitis C; HRS, hepatorenal syndrome; INR, international normalized ratio; MELD, Model for End-Stage Liver Disease; PBC, primary biliary cirrhosis; SBP, spontaneous bacterial peritonitis.
Diuretics	None	Allowed
Number	15	9
Median age	52	55
Male gender	10 (67%)	7 (78%)
Median MELD (range)	20 (13–25)	23 (16–28)
Median INR (range)	1.7 (1.1–2.6)	1.9 (1.4–2.7)
Median bilirubin, mg/dL(range)	4.1 (1.5–14)	4 (2–19.7)
Median creatinine, mg/dL (range)	1 (0.6–3.5)	1.1 (0.7–1.4)
Median serum albumin, g/dL (range)	2.8 (2.1–3.4)	2.7 (2.1–3.4)
Median serum sodium, mEq/L(range)	124 (121–130)	126 (121–130)
Median urine sodium, mmol/L (range)	2 (0–48)	4 (0–141)
Median admission, mg/day
Spironolactone (range)	100 (0–300)	100 (0–300)
Furosemide (range)	40 (0–120)	40 (0–120)
Intravenous albumin given	12 (80%)	8 (89%)
Complications of cirrhosis
Ascites	15 (100%)	9 (100%)
SBP	2 (13%)	1 (11%)
HRS	3 (20%)	1 (11%)
Etiology of cirrhosis
HCV	7 (47%)	3 (33%)
Alcohol	6 (40%)	3 (33%)
AIH/PBC	2 (13%)	0
HBV	0	2 (22%)
Cryptogenic	0	1 (11%)

Group 2 consisted of 9 patients whose diuretics were continued. Their day 0 median serum sodium was 126 (range, 121–130). All had ascites, and 1 patient had both SBP and HRS. Eight of the 9 patients (89%) received intravenous albumin. Their median age was 55 years, 78% were male, and their median MELD score was 23. The etiology of liver disease was hepatitis C virus in 3 patients, alcohol in 3 patients, hepatitis B in 2 patients, and cryptogenic cirrhosis in 1 patient.

The course of serum sodium levels in group 1 is shown in Fig. 1A. The mean serum sodium was 124 mmol/L on the day of entry (day −1) and just prior to the administration of conivaptan (day 0; P = not significant). Serum sodium rose during the infusions in all subjects. One may define a significant response as an increase in serum sodium of ≥5 mmol/L. This occurred by day 2 in 7 of the 15 patients (47%) and by the end of treatment (up to 4 days) in 9 of the 15 (60%). It rose to 127.7 mmol/L by day 1 (P < 0.001) and to 128.6 mmol/L by day 2 (P < 0.001) of conivaptan administration and was 131.3 mmol/L on the first day after infusion. One of the 15 patients had a fall in serum sodium of ≥5 mmol/L after a median follow-up of 4 days (range, 1–88 with at least weekly testing) after discontinuation of therapy. Of the 15 patients, 10 were listed for liver transplantation. Nine underwent successful transplantation a median of 30 days (range, 7–241 days) after conivaptan was initiated. Their mean and median serum sodium at the time of transplantation was 132 mmol/L (range, 126–139 mmol/L). None had perioperative neurological complications or renal failure, and 8 were well a median of 202 days (range, 9–516 days) after transplant. The other patient died of acute renal insufficiency that presented 10 months after transplant. Two patients remained on the list for transplantation with persistently normal serum sodium. Of the remaining 5 patients, 2 were lost to follow-up after discharge, and 3 were discharged to hospice and died of liver failure.

Figure 1. Serum sodium levels (A) in individuals in whom diuretics were discontinued before the administration of intravenous conivaptan (levels rose in all 15 treated patients) and (B) in individuals in whom diuretics were continued throughout the administration of intravenous conivaptan (levels rose in 8 of the 9 treated patients). The asterisks represent the mean serum sodium levels, the gray boxes represent the 25% to 75% intervals, the black lines in the boxes represent the medians, and the whiskers represent the ranges of serum sodium on particular days. The tables show the daily serum sodium levels for each patient during the study. Abbreviation: NS, not significant.

Figure 1B shows the changes in serum sodium in the group that remained on diuretics during the admission. The mean serum sodium on days −1 and 0 were 125.9 and 125.7 mmol/L, respectively (P = not significant). It rose during the infusions in 8 of the 9 subjects and improved by at least 5 mmol/L by the end of treatment in 6 of the 9 patients (67%). The sodium level was 129.9 mmol/L on day 1 (P = not significant versus day 0) and 130.6 mmol/L by day 2 (P = 0.04 versus day 0). None of the patients experienced a fall in serum sodium of ≥5 mmol/L after a median follow-up of 4 days (range, 0–10 with at least weekly testing). Of the 9 patients, 3 were listed for liver transplantation. Two underwent successful transplantation 10 and 38 days after conivaptan was initiated. Their serum sodium levels at the time of transplantation were 128 and 129 mmol/L. Neither had perioperative neurological complications or renal failure, one was well 9 months post-transplant, and the second died of recurrent hepatitis C virus 2 years post-transplant. One patient remained on the list for transplantation and maintained a normal serum sodium level. Of the remaining 6 patients, 2 were discharged and lost to follow-up, whereas 4 died of liver failure.

Conivaptan was well tolerated. There was no significant change in systolic blood pressure or serum creatinine during the infusion over time (Table 2). Serum potassium also remained normal in all subjects during the infusions. There were no cases of hypernatremia. Although the aquaretic effect of conivaptan is mediated via inhibition of the V2 receptor, the drug is also a V1 receptor antagonist and might therefore increase portal hypertension. No episodes of variceal hemorrhage or worsening of ascites occurred during the infusion period. One patient had an upper gastrointestinal bleed from portal gastropathy 6 weeks after receiving conivaptan.

Table 2. SBP and Creatinine Levels over Time in the Two Groups

		Day
		−2	−1	0	1	2
NOTE: An analysis of variance between days 0, 1, and 2 for each group did not reveal a significant change in SBP or creatinine. Data are presented as means and standard deviations. Abbreviation: SBP, systolic blood pressure.
No Diuretics	SBP (mm Hg)	101 ± 15	104 ± 18	103 ± 15	101 ± 14	97 ± 8
	Creatinine (mg/dL)	1.44 ± 0.99	1.37 ± 0.97	1.42 ± 0.85	1.43 ± 0.91	1.41 ± 0.86
Diuretics Continued	SBP (mm Hg)	101 ± 22	107 ± 14	105 ± 11	100 ± 13	101 ± 13
	Creatinine (mg/dL)	1.26 ± 0.55	1.25 ± 0.44	1.13 ± 0.44	1.34 ± 0.82	1.44 ± 0.95

DISCUSSION

Hyponatremia is a frequent complication in patients with decompensated cirrhosis, is commonly associated with SBP and/or HRS, and is an independent predictor of mortality.1–4, 16, 19 Medical management of hyponatremia is difficult because it involves fluid restriction that is not well accepted by patients and interruption of diuretic therapy that usually leads to worsening fluid retention. Many of these patients are awaiting liver transplantation, and the presence of significant hyponatremia may delay transplantation because of the risk of developing perioperative neurological disorders, renal failure, and infectious complications. On the other hand, some have suggested that the higher mortality related to hyponatremia justifies additional priority for liver transplantation. Although the addition of serum sodium to the MELD score that is currently used to allocate organs for liver transplantation would certainly increase the priority for an organ in a subset of patients, the societal benefit in terms of reduced waiting list mortality and improved outcomes is less clear, particularly if perioperative problems were more common in these patients. Thus, more effective medical management of hyponatremia is what is most needed. Our study demonstrates that the serum sodium level can be raised in the majority of patients with intravenous conivaptan: when “response” was defined as a ≥5 mmol/L increase in serum sodium, 60% to 67% of patients responded. Patients who subsequently underwent liver transplantation did well, and none had neurological sequelae of hyponatremia. Although this study was small and did not include an untreated control group, the impact of conivaptan is clear.

Several other vasopressin-2 receptor antagonists have been used to treat hyponatremia in patients with cirrhosis.11–17 These agents previously have been shown to be safe and effective at increasing serum sodium in hyponatremic patients with end-stage liver disease and helping achieve better control of ascites. Despite the data available, no oral or intravenous agent is available for clinical use in patients with cirrhosis. Conivaptan is the only vasopressin-2 receptor antagonist that is FDA-approved in the United States for the treatment of hyponatremia in any patient population. Although the medication is available for the treatment of hospitalized patients with hyponatremia, its side effect profile and efficacy in patients with advanced liver disease and/or renal impairment are not known. The cases that we report in this retrospective experience suggest that short-term use of conivaptan to treat hyponatremia is safe in patients with advanced decompensated cirrhosis. Indeed, our study group had quite advanced decompensation with high MELD scores. Unfortunately, the severity of the liver disease also limited our ability to confidently assess the durability of the increase in serum sodium because most patients underwent liver transplantation after just a median of 30 days or died from liver failure. However, we would certainly expect the effect of conivaptan to be short-lived. Nonetheless, future studies with this and other V2 receptor antagonists should prospectively follow serum sodium after treatment to determine how long the effects persist.

In this retrospective review, we report that intravenous conivaptan can be used to safely and rapidly improve hyponatremia in patients with advanced liver disease. Although our data suggest that the effect of conivaptan on serum sodium occurs with or without the continuation of diuretics, our experience was limited to a small number of patients and was not controlled. Thus, these findings will need to be confirmed in a randomized study. This pilot study does indicate that conivaptan can be used in patients with end-stage liver disease to treat hyponatremia. Although we assume that the treatment of hyponatremia in patients with decompensated cirrhosis is beneficial and might decrease transplant morbidity and mortality, our study was not designed to answer this question. As a result, it is essential that prospective trials be undertaken to evaluate the longer term safety and efficacy of conivaptan and other vasopressin-2 receptor antagonists on outcome with and without transplantation. In addition, it will be important to determine if vasopressin-2 receptor antagonists can, by raising the serum sodium, eliminate hyponatremia as an independent predictor of mortality in patients with end-stage liver disease, thereby allowing MELD, and not MELD-Na, to continue to be used to list patients for liver transplantation.