Metabolic Clearance Rate of Arginine Vasopressin in Patients with Cirrhosis JOSE

A. SOLIS-HERRUZO, AMELIA G-ONZALEZ-GAMARRA, GREGORIO CASTELLANO AND MARIATERESAM ~ o z - Y A G ~

Gastroenterology Unit, Department of Medicine, Hospital Universitario “Doce de Octubre, ” School of Medicine, Universidad Complutense, 28041 Madrid, Spain

Metabolic clearance rate and half-time of arginine vasopressinwere measured in 43 cirrhoticpatients and 10 control subjects. Syntheticargininevasopressinwas infusedintravenously at a rate of 600 pg/mWg of body weight for 76 min. The metabolic clearance rate was significantly reduced, and the half-time of arginine vasopressin after stopping the infusion was significantly increased in patients with cirrhosis, particularly in those with ascites and in those with moderate or severe liver dysfunction. Changes in metabolic clearance rate and half-time of arginine vasopressin correlated with the score of the liver dysfunction, prothrombin activity and levels of serum albumin and bilirubin but not with parameters of kidney function (serum creatinine levels and clearance of creatinine). We conclude that reduced metabolic clearance rate and proloaged half-time of vasopressin in plasma are frequent findings in cirrhotic patients with poor liver function. This impaired catabolism of antidiuretic hormone may contribute to maintaining elevated plasma levels of this hormone in these patients and may be an additional factor leading to fluid retention and to dilutional hyponatremia. (HEPATOLOGY 1992;16 974-979.)

additional role in the perturbations of water metabolism in patients with liver diseases. Extracts of liver tissue are capable of inactivating ADH in uitro (8, 91, and hepatic clearance of ADH has been demonstrated in isolated perfused livers (10-13). Furthermore, CC1,-induced liver injury resulted in a decreased hepatic clearance of ADH (14). Despite these evidences for a role of the liver in the degradation of ADH, very little attention has been paid to the catabolism ofADH in patients with cirrhosis (15). The purpose of this study was to measure the metabolic clearance rate (MCR) of arginine vasopressin (AW) under steady-state conditions in 43 patients with cirrhosis.

PATIENTS AND METHODS Forty-three hospitalized patients (24 men and 19 women; age range = 32 to 82 yr; mean = 54 yr) with cirrhosis and 10 control subjects were studied. Diagnosis of cirrhosis was established by laparoscopy and liver biopsy in 31 cases. In the remaining 12 cases, liver biopsy was not performed because of severe bleeding diathesis. In these cases diagnosis of cirrhosis was made on the basis of clinical findings (ascites, encephalopathy, esophageal varices, splenomegaly, hepatomegaly and spider angiomas), biochemical findings (prolonged proPatients with advanced cirrhosis and ascites fre- thrombin time, decreased serum albumin levels and hyperquently show a poor diuretic response to water load, gammaglobulinemia) and ultrasound findings. Cirrhosis was dilutional hyponatremia and elevated plasma levels of caused by alcohol intake in 30 patients and by HBsAgantidiuretic hormone (ADH) (1-4). It is accepted that associated disorders in 3, and the cause was idiopathic in 10. The severity of liver dysfunction was estimated accordingto these elevated concentrations of plasma ADH are the result of an increased nonosmotic release of this the Child-Turcotte classification and also by use of a scoring hormone. These patients exhibit hormonal features system, which ranged from 5 to 15 (16). The total score from the assignment of numerical equivalents (A = 1, suggesting a decreased effective blood volume (51, which resulted B = 2, C = 3) to each of the five components of the Childmay stimulate nonosmotic secretion of ADH (6, 7). Turcotte classification (17, 18). Patients were separated into However, plasma concentration of ADH reflects the three groups. Group 1 included patients who had never had balance between production and removal of this ascites, group 2 included patients with ascites and group 3 hormone from the circulation, and it is conceivable that included patients with ascites and hyponatremia (serum an abnormal ADH catabolism might also play some sodium levels below 135 mEq/L). Thirteen patients were included in group 1, 13 in group 2 and 17 in group 3. Ascites was detected in 30 patients, and tense ascites was detected in 16 of them. Table 1 summarizes the clinical and laboratory Received February 5, 1992; accepted June 9, 1992. features of these patients. The control group consisted of 10 This study w88 supported by grant 861729 from Fondo de Investigaciones subjects (eight men and two women; age range = 27 to 78 yr; Sanitarias, Spain. mean = 54.9 yr) without evidence of previous or underlying Address reprint requests to: J. A. Solis-Hemzo, M.D., Ph.D., Servicio de liver, kidney or metabolic disorders. No patient received any Aparata Digestivo, Hospital Universitario “Doce de Octubre,” 28041 Madrid, medication during the week before the study. Daily diet Spain. contained 0.8 to 1.2 gm proteinkg of body weight, 50 to 70 311114oaoa 974

975

CATABOLISM OF ADH IN CIRRHOSIS

Vol. 16, No. 4, 1992

TABLE1. Clinical and laboratory features of cirrhotic patients Characteristics

Age W" Sex ( M F ) Ascites (o/+ / + + P Encephalopathy (01+ )' Nutrition (O/ + I + + )d Serum creatinine level (N = 53 to 113 pmol/L)" Creatinine clearance (N = 90 to 140 ml/min)" Prothrombin activity (N = 70% to 100%)" Serum albumin level (N = 0.58 to 0.77 mmol/L)" Serum bilirubin level (N = 3.4 to 17 ymol/L)" Serum sodium level (N > 135 pmol/L)" Plasma AVP level (N 3.7 pmol/L)" Degree of liver dysfunction (scores)"

Group 1

Group 2

Group 3

5 4 2 c 12 ( 7/61 ( 13/0/0) (0/3) (41712) 75 7 +- 18.5 73.7 i 23.3 67.8 i 16.5 0.51 f 0.07 26.4 f 11.5 141.2 ? 3.1 3.56 t 1.06 7 1 i 1.38

57.2 i 13 3 (7/6) (0110/3) (1013)

53.7 t 10.8 (9/8) (014113) (9/8) (014113) 77.3 -+ 26.1 65.2 -+ 24.1 50.8 t 11.2 0.43 t 0.07 104 2 80.9 131 t 3.17 8.04 -+ 3.0 11.6 t 2.9

(0/8/5)

79.4 f 20.7 61.1 i 18.9 60.9 f 7.8 0.46 2 0.08 40.5 c 24.9 139 ? 1.5 6.57 t 1.8 10.1 f 1.3

N = normal range. "Data are expressed as mean i S.D. bO = none; + = easily controlled; + + = poorly controlled. "0 = none; + = some episode in the past. = excellent; + = good; + + = poor.

mEq sodium and 80 to 100 mEq potassium. Water was not restricted. None of these subjects presented hemodynamic instability, gastrointestinal hemorrhage, hepatocarcinoma or associated cardiac, respiratory or infectious disease at the time of study. In all cases the serum creatinine level was below 132.6 pmol/L. Informed consent for the study was obtained from each patient after a detailed description of the procedure, and the study was approved by the Human Research Committee of the University Hospital "Doce de Octubre." No complications or side effects were observed during the study. Experimental Protocol. After the patients fasted overnight, an indwelling needle was introduced in one antecubital vein for multiple blood collections,and endogenous ADH secretion was suppressed by an infusion of 5% glucose in water (20 ml/kg of body weight) injected for 30 min in an opposite antecubital vein. At the end of this period, synthetic AVP (Pitressin; Parke-Davis, Morris Plains, NJ) diluted in normal saline was infused at a rate of 500 pg/min/kg of body weight for 75 min. After 30,60 and 75 min of infusion, a blood sample was taken from the opposite vein, and the infusion was stopped. Further samples of venous blood were taken 5,15,25,35,45 and 60 min after stopping the AVP infusion. Measurements. All blood samples for measurement of plasma AVP were placed immediately on ice and centrifuged when the procedure had been concluded. Plasma was stored at -40" C for later analysis. Plasma AVP concentration was measured by RIA with a commercial kit (Biihlman Laboratories, Basel, Switzerland). The interassay and intraassay variabilities for this procedure were less than 10%. The sensitivity of this assay was 1 pgiml (0.9 pmoUL), which was determined as the dose at 2 S.D. below the counts at maximum binding. The following vasopressin antiserum cross-reactions have been determined at 50% binding: AVP, 100%; lysine vasopressin, 0.85%;oxytocin, 0.0001%; and vasotocin, 0.015%. The normal range of plasma AVP concentration in our laboratory is less than 3.7 pmoVL. &&ulations. The MCR of AVP was calculated according to the following formula: MCR

=

I (BSA x C,)

where Z is the infusion rate (pmoVmin), BSA is the body surface area (m2) and C, is the steady-state plasma level (pmow). This level was taken as the mean of the last two points after 60 and 75 min of infusion when they did not deviate from the mean by more than 5% (Table 2). This calculation is valid because endogenous secretion of ADH is minimal in the water-loaded state. In previous studies we have shown that AVP levels after water load are 1.71 5 0.72 pmom in patients with ascites and hyponatremia (4). According to Fabian et al. (191, data of the disappearance curve were expressed as ([CJC,] x 1001, where C, is the plasma concentration of AVP at time t, and they were plotted against time on semilogarithmic paper. These disappearance curves were monoexponential and were resolved with a computational program (Eureka: The Solver, Version 1.0) from the following equation: f(x) = A . e - t K

t'

where x is (ln[C,/C, loo]), K is the rate constant (min-') of the exponential and A is the ordinate intercept. The half-time of this exponential function was calculated by dividing In 2 by the corresponding rate constant. Statistical Anulysis. All data are expressed as mean -+ S.D. Statistical significance of the difference between means was calculated with the Newman-Keuls test. Correlation between data was made by Pearson's correlation coefficient. A p value less than 0.05 was considered statistically significant. 9

RESULTS The MCR of AVP was significantly reduced in patients with cirrhosis. This decrease was particularly marked in patients with hyponatremia (154 & 52 mVmin/m2; p < 0.001) (Fig. 1) and in those of class C of the Child-Turcotte classification (145 L 49 mVmin/m2; p < 0.001) (Table 3). Eight patients (61.5%) without ascites (group 1)but none with ascites (groups 2 and 3) had an MCR in the range of normality (control mean 2 2 S.D.)(Fig. 1). This impairment was also common in patients with episodes of portosystemic

976

SOLIS-HERRUZO ET AL.

HEPATOLOGY

-0.37; p < 0.05), serum albumin levels (r = c 0.01)and serum bilirubin levels (r = 0.636; p < 0.001). Half-time of AVP did not correlate with serum creatinine levels (r = - 0.195;NS) and clearance of creatinine (r = -0.08, NS). (r

=

- 0.42;p

***

Controls

Cirrhosis (I)

Cirrhosis (11) Cirrhosis (111)

Groups of patients

*

FIG. 1. MCR of AVP.Data are expressed as mean S.D. *p < 0.05; ***p < 0.001; (w) = values greater than 500 ml/min/m2.

encephalopathy (100%). In these patients the MCR was significantly lower (177.3f 81 mVmin/m2) than in those without encephalopathy (279 +- 112 ml/min/m2; p c 0.01). The MCR of AVP correlated significantly with the score of liver dysfunction (r = -0.85; p < O.OOl),prothrombin activity (r = 0.49;p < 0.001), serum albumin levels (r = 0.51;p < 0.001)and serum bilirubin levels (r = -0.47;p < 0.01);however, it did not correlate significantly with the serum creatinine levels (r = -0.022; NS) and the creatinine clearance (r = 0.033;NS). The rate constant of the disappearance curve was also markedly reduced in cirrhotic patients, particularly in those of group 2 (0.0107& 0.0032/min; p c 0.001)and group 3 (0.0071f 0.0034/min;p c 0.001)(Fig. 2) and in patients of class C of the Child-Turcotte classification (0.0068? 0.0021/min; p c 0.001)(Table 3). Figure 2 shows the disappearance curve of AVP in control patients and cirrhotic patients. In group 1 38% of patients and in group 2 7.7% of patients had rate constants in the range of normality, but none of the patients in group 3 had rate constants in this range. This constant correlated significantly with the score of liver dysfunction (r = -0.73; p c 0.001), prothrombin activity (r = 0.36;p < 0.051, serum albumin levels (r = 0.33;p < 0.05)and serum bilirubin levels (r = -0.45; p < 0.01);however, no significant correlation was found between the rate constant and serum creatinine levels (r = -0.028; NS) and clearance of creatinine (r = 0.071;NS). The half-time of AVP after stopping the AVP infusion was significantly longer in all three groups of patients (group 1,43.55f 14.1 min, p c 0.01;group 2, 73.7 & 37.1 min, p < 0.001;group 3,101.5f 38.3 min, p < 0.001; and the control group, 30.34 2 4.8 min) (Fig. 3).This abnormality was also significantly marked (109.7 & 33.4 min; p c 0.001)and frequent (100%) in class C patients. A sigdicant correlation existed between the half-time of AVP and the score of liver dysfunction (r = 0.66;p < 0.001),prothrombin activity

DISCUSSION It is generally accepted that ADH is metabolized by the kidneys and the splanchnic viscera (10,20,21),and it has been assumed that clearance of ADH by these viscera is caused primarily by the liver (20).This view is supported by early experiments that showed that extracts of liver tissue may inactivate vasopressin (8, 9) and by observations that demonstrated that the liver in situ and the isolated perfused liver are effective in removing AVP from the perfusion fluid or blood (10-13, 19). On the other hand, clamping of the splanchnic vessels results in an increased half-time of AVP, which confirms that the splanchnic area is involved in the removal of AVP from circulation (19).Furthermore, destruction of liver cells by CCl, intoxication results in a decreased hepatic clearance of ADH (14). Some workers in the 1940s suggested that an impaired hepatic inactivation of ADH may play an important role in producing fluid retention in patients with cirrhosis (22). However, most investigations during recent years have dealt with the nonosmotic release of ADH in cirrhotic patients. Only Skowsky et al. (23)and Ardaillou et al. (15)reported in preliminary studies that the MCR of exogenous (15,23) and endogenous (15) ADH is decreased in cirrhotic patients with ascites. Our study demonstrates a reduction in the MCR of AVP in hydrated patients with cirrhosis that becomes more pronounced with the degree of liver dysfunction. MCR was particularly reduced in patients of class B and C of the Child-Turcotte classification (Table 31, and consequently catabolism of AVP was significantly decreased in patients with ascites or portosystemic encephalopathy. Nonsuppressed endogenous secretion of ADH might contribute to reducing the MCR in patients with advanced liver disease. Bichet et al. (7),Perez-Ayuso et al. (24)and Castellano et al. (4)have shown that ADH levels may not be fully suppressed after water loading in cirrhotic patients with ascites and hyponatremia. Although we have not measured AVP levels after water loading in this study, our previous investigations have shown that these levels are 1.71 f 0.72 pmol/L in patients with ascites and hyponatremia (4).Because these levels are approximately 60 times lower than those reached during the AVP infusion, it is unlikely that endogenous secretion of ADH could have had any significant impact on the reduced MCR found in our patients. Moreover, assessment of catabolism of ADH from the disappearance curve of AVP led to identical conclusions (Table 3). The rate constant of the disappearance curve was also markedly reduced in cirrhotic patients, particularly in those with poor liver function (Table 3). This rate constant and the MCR of AVP correlate significantly with the score of liver dysfunction. Only a few studies have measured the total MCR of

977

CATABOLISM OF ADH IN CIRRHOSIS

Vol. 16, No. 4, 1992

TABLE 2. Plasma levels of AVP obtained under basal conditions and at 30,60 and 75 min of AVP infusion Subjects

BlWd

30 min

Control group (n = 10) Cirrhotic patients Group 1 (n = 13) Group 2 (n = 13) Group 3 (n = 17)

2.9 t 0.4'

28.9 t- 5.1

4.2 t 1.9 6.5 i- 1.9 8.0 2 3.0

45.7 +- 20 62.4 ? 21 95.0 t 22

60 min

76 min

Ce0

38.9 t- 5.1

38.7 t 4.4

38.8 t 4.6

51.6 2 16 73.7 t 22 113.9 24

50.2 ? 19 73.3 ? 1 7 113.5 ? 22

50.9 t 17 73.5 ? 19 113.7 t 23

"C, = Steady-state plasma level of AVP. 'Data are expressed as mean ? S.D. Units are picomoles per liter

TABLE 3. MCR of AVP plasma Subjects

Control group (n = 10) Cirrhotic patients Class A (n = 12) Class B (n = 16) Class C (n = 15)

K = rate constant of disappearance curve; t, "Data are expressed as mean t S.D. 'p < 0.05. 'Nonsignificant. dp < 0.01. 'p < 0.001.

K (min-')

MCR (ml/min/m') 462

?

65"

0.0238

0.0038

0.0186 ? 0.0084' 0.0115 t 0.0032' 0.0068 t 0.0021'

385 -c 796 244 +- 56' 145 ? 4Y =

?

t, (min)

30.34

?

4.84

42.4 ? 13.1d 68.5 2 34.9 109.7 t 33.4'

half-time of plasma AVP

unlabeled AVP by the steady-state method in human subjects (25,26). Most of the available studies on the outcome of ADH are based on analysis of the disappearance curve of ADH from plasma after a single intravenous injection of vasopressin. These studies have shown that AVP is rapidly cleared from the systemic circulation after a two-exponential curve. Because infused AVP is not bound to plasma proteins (27,28),it diffuses readily into a space approximating the extracellular fluid volume (20,28). Irreversible catabolism of AVP takes place in the second compartment, after AVP is bound to target tissues containing receptor sites and degrading enzymes. Results of earlier studies (29)must be accepted cautiously because of the inaccuracy of bioassays for AVP measurement at that time. Ardaillou et al. (151,using a specific RIA, studied the MCR of synthetic AVP after a single injection of 2 p,g AVP to patients with cirrhosis. They observed that MCR was much lower in cirrhotic patients with ascites than in patients without ascites or in healthy controls (15, 30). Although they also found an abnormal reduction in the MCR in patients with cirrhosis, their results in normal subjects were clearly lower than those we found in our controls. However, differences in results between both studies may be caused by a number of factors including differences in the dose of AVP used, the state of hydration and the conditions in which the clearance was measured. 1261-AVPhas also been used to measure the MCR of AVP in normal subjects and in cirrhotic patients. Using this method, Baumann and Dingman (28)found that the mean MCR of AVP is only 4.1 ml/min/m2. Skowsky et al. (23)found an MCR of 1z61-AVPof 3.98 k 0.42

ml/mm/m2 after infusion of the tracer-labeled AVP for 2 hr. These values of MCR were much lower than those we have found now in our control subjects. However, studies in dogs (31,32)have clearly demonstrated that iodinated-AVP clearance is significantly lower than the clearance of unlabeled AVP. Differences in the volume of distribution between both AVP preparations may account for this discrepancy (32).This volume is considerably larger for the unlabeled hormone because the relative electronegativity of the iodinate molecule impedes its ready diffusion into the extravascular pool. Differences in precision of techniques for measuring both AVP preparations, in affinity of the degrading enzymes or in binding to receptors may also explain the conflicting results between iodinated-AVP and uniodinated-AVPclearance. Using the 1261-AVPmethod under steady-state conditions, Skowsky et al. (23) reported that the MCR of AVP was depressed to about 10% of controls in patients with cirrhosis and that it tended to be lowest in those patients with greater liver impairment. As far as we are aware, only a few authors (25,26,33) have studied MCR of AVP under steady-state conditions. Under these circumstances total MCR is equal to the rate of infusion divided by plasma concentration of AVP. Measurements are comparable only when AVP infusion rates are similar (251,the endogenous secretion of ADH has been properly suppressed and the intravenous infusion of AVP has been maintained during an interval that is long enough to assure stabilization of plasma concentration. Our results in the overhydrated control group were only slightly lower than those reported by Moses and Steciak in normal subjects (25).Under these

978

SOLIS-HERRUZO ET AL.

HEPATOLOGY

100

00 = 0.0107 2 0.0032

GR.- Ill K-0.018 z 0.008

40

-1

\ \

1

GR.- I /

\Y

K- 0.0238

~0.0034

20

I

5

I

I

I

b

I

I

I

5

5

I

I

I

I

b

MINUTES

FIG.2. Disappearance curve of AVP from plasma after stopping the vasopressin infusion. K is the rate constant (min-'). C,is the plasma concentration of vasopressin at time t and C, is the steady-state plasma level of AVP. GR-Z, GR-ZZ and GR-ZZZ represent groups 1, 2 and 3, respectively. Data are expressed as mean f S.D. *p < 0.05; ***p < 0.001.

73.7****

(mmmm)

37.1

(3

l20r

30.94 24.04

6ol :t

-

101.5

***

238.3

43.55*** 14.15

:

40

1

CONTROLS

CIRRHOSIS (I)

CIRRHOSIS (11)

CIRRHOSIS (111)

Groups of patients

FIG.3. Half-time of AVP in control subjects and cirrhotic patients. Data are expressed as mean f S.D. **p < 0.01; ***p < 0.001; (m) = values over 120-min period.

conditions cirrhotic patients showed significantly reduced total MCR. Although the liver is involved in inactivation of ADH, other organs also take part in this process. The role of the kidney in the breakdown of ADH has been clearly demonstrated in experiments with the perfusion of isolated kidneys (10, 27) and by in v i m studies on nephrectomized dogs (31). Experiments in animals have revealed that the kidney avidly removes ADH from the circulation through glomerular filtration and peritu-

bular secretion (27) and that most of the filtered ADH is reabsorbed or inactivated in the proximal nephron (34). The relative contribution of the liver and the kidney in the catabolism of ADH has been a matter of debate through the years; however, it is generally accepted that both organs inactivate approximately equal amounts of ADH (21). Cirrhotic patients with ascites frequently have kidney failure that may account for the decreased MCR of AVP found in cirrhotic patients. In human beings, Ardaillou et al. (15) and Benmansour et al. (30) demonstrated that the MCR of AVP is much lower in uremic patients than in normal subjects. It is unlikely that reduced kidney function accounts for the abnormal MCR of AVP in our patients. None of them had serum creatinine levels above 132.6 pmoVL, and we could not find any significant correlation between the MCR of AVP and serum creatinine levels. Although creatinine clearance was reduced in many of our patients, no correlation was found between this clearance and the MCR of AVP. In fact, Rabkin et al. (27) showed that the rate at which the kidney clears ADH remains constant, even though the glomerular filtration rate falls. Under these conditions, decreased amounts of ADH cleared by filtration are compensated by an increased peritubular secretion of the hormone. Delayed inactivation of ADH by the liver may contribute to maintaining increased serum ADH concentrations in cirrhotic patients, particularly when nonosmotic release of this hormone is increased. This study shows that AVP half-time is significantly prolonged in cirrhotic patients with poor liver function (Table 3). Although half-time of plasma AVP ranged between 20 and 40 min in control subjects, in cirrhotic patients of

Vol. 16, No. 4, 1992

CATABOLISM OF ADH IN CIRRHOSIS

class C half-time averaged 109.7 f 33.4 min. Skowsky et al. (23)came to a similar conclusion. They found that serum disappearance half-time of 1251-AVPwas prolonged in patients with cirrhosis. This delayed catabolism of the ADH may be an additional factor leading to fluid retention and to dilutional hyponatremia. Our investigation demonstrates that the MCR and the rate constant of the disappearance curve were markedly decreased in patients with ascites and hyponatremia. We conclude that the MCR of AVP is reduced and half-time of AVP is prolonged in patients with cirrhosis, particularly in those with poor liver function. These changes, in addition to the increased ADH release by the neurohypophysis, may contribute to maintaining elevated plasma levels of ADH in patients with cirrhosis and may be an additional factor for the fluid retention and hyponatremia in these patients. REFERENCES 1. Arroyo V, Rodes J. A rational approach to the treatment of ascites. Postgrad Med J 1975;51:558-562. 2. Birchard WH, R o u t TE, Williams TF, Rosenbaum JD.Diuretic responses to oral and intravenous water loads in patients with hepatic cirrhosis. J Lab Clin Med 1956;48:26-35. 3. Klinger EL Jr, Vaamonde CA, Vaamonde LS, et ai. Renal function changes in cirrhosis of the liver. Arch Intern Med 1970;125:10101015. 4. Castellano G, Solis-Herruzo JA, Morillas JD, Larrodera L, Coca C, Gonzalez-Gamarra A, MuAoz-Yagtie MT. Antidiuretic hormone and renal function after water loading in patients with cirrhosis of the liver. Scand J Gastroenterol 1991;26:49-57. 5. Schrier RW. Pathogenesis of sodium and water retention in hlgh-output and low-output cardiac failure, nephrotic syndrome, cirrhosis and pregnancy (1). N Engl J Med 1988;319:1065-1072. 6. Schrier RW. Pathogenesis of sodium and water retention in high-output and low-output cardiac failure, nephrotic syndrome, cirrhosis and pregnancy (2). N Engl J Med 1988;319:1127-1134. 7. Bichet D, Szatalowicz V, Chaimovitz C, Schrier RW. Role of vasopressin in abnormal water excretion in cirrhotic patients. Ann Intern Med 1982;96:413-417. 8. Dicker SE, Greenbaum AL. Inactivation of the antidiuretic activity of vasopressin by tissue homogenates. J Physiol (Lond) 1956;132: 199-212. 9. Miller GE, Townsend CE. The in vitro inactivation of pitressin by normal and cirrhotic human liver. J Clin Invest 1954;33:549-554. 10. Rabkin RS, Ghazeleh S, Share L, Crofton J , Unterhalter SA. Removal of immunoreactive arginine vasopressin by the perfused rat liver. Endocrinology 1980;106:930-934. 11. Lauson HD, Bocanegra M, Beuzeville CF. Hepatic and renal clearance of vasopressin from plasma of dogs. Am J Physiol 1965;209:199-214. 12. Little JB,Klevay LM, Radford EP Jr, McGandy RB. Antidiuretic hormone inactivation by isolated perfused rat liver. Am J Physiol 1966;211:786-792. 13. Ginsburg M, HelIer H. The clearance of injected vasopressin from the circulation and its fate in the body. J Endocrinol 1953;9: 282-291. 14. Schnieden H. Comparison between the effects of intravenous and

979

intraportal vasopressin in normal rats, malnourished rats, and rats treated with carbon tetrachloride. J Endocrinol 1962;24: 397-402. 15. Ardaillou R, Benmansour M, Rondeau E, Caillens H. Metabolism and secretion of antidiuretic hormone in patients with renal failure, cardiac insufficiency and liver insufficiency. Adv Nephrol Necker Hosp 1984;13:35-49. 16. Conn HO. A peek at the Child-Turcotte classification. HEPATOLOGY 1981;1:673-676. 17. Child CG 111. Turcotte JG. Surperv and mrtal hvoertension. In: Child CG II1,'ed. The liver and Po& h y p k e n s i i i . Philadelphia: WB Saundera Co, 1964:50. 18. Campbell DP, Parker DE, Anagnostopoulos CE. S u M v d prediction in portocaval shunt: a computerized statistical analysis. Am J Surg 1973;126:748-751. 19. Fabian M, Forsling ML, Jones JJ, Lee J . The release, clearance and plasma protein binding of oxytocin in the anaesthetized rat. J Endocrinol 1969;43:175-189. 20. Lauson HD. Metabolism of antidiuretic hormones. Am J Med 1967;42:713-744. 21. Lauson HD. Metabolism of the neurohypophyseal hormones. In: Knobil E, Sawyer WH, eds. Handbook of physiology. Washington DC: American Physiological Society, 1974;4:287-293. 22. Ralli EP, Robson JS, Clarke DH, Hoagland CL. Factors influencing ascites in patients with cirrhosis of the liver. J Clin Invest 1945;24:316-325. 23. Skowsky R, Riestra J , Martinez I, Swan L, Kikuchi T. Arginine vasopressin kinetics in hepatic cirrhosis. Clin Res 1976;24:101A. 24. Perez-Ayuso RM, Arroyo V, Camps J , Rimola A, Gaya J , Costa J , Rivera F, et al. Evidence that renal prostaglandins are involved in renal water metabolism in cirrhosis. Kidney Int 1984;26:72-80. 25. Moses AM, Steciak E. Urinary and metabolic clearances of arginine vasopressin in normal subjects. Am J Physiol 1986;251: R365-R370. 26. Moses AM, Jones C, Yucha CB. Effects of sodium intake, furosemide, and infusion of atrial natriuretic peptide on the urinary and metabolic clearances of arginine vasopressin in normal subjects. J Clin Endocrinol Metab 1990;70:222-229. 27. Rabkin R, Share L, Payne PA, Young J , Crofton J . The handling of immunoreactive vasopressin by isolated perfused rat kidney. J Clin Invest 1979;63:6-13. 28. Baumann G, Dingman JF. Distribution, blood transport and degradation of antidiuretic hormone in man. J Clin Invest 19?6;57:1109-1116. 29. Czaczkes JW.Kleeman CR. K o e n i ~M. Phvsioloeic studies of antidiuretic hormone by its direct mekuremeit in h