Hepatorenal syndrome and ascites
- questions and answers
zyxwvuts
zyxwv
zyxwvuts
Bernardi M, Blendis L, Burroughs AK, Laffi G, Rodes J, Gentilini P.
Hepatorenal syndrome and ascites - questions and answers.
Liver 1999: 19 (Suppl 1). 0 Munksgaard, 1999
QUESTION 1 QUESTION 2
zy
zyxwvutsrq
1111. Bernardi, 1. Blendis,
A. K. Burroughs, G. Laffi, J. Rodes
and P. Gentiini
For further information see page 74
What is the physiological relationship between the liver Is the congestion of the splanchnic area and hepatic
and kidney? sinusoids during digestion sending signals to the kidney
in order to save sodium and water?
' In 1987 Levy and Wexler (1, 2) suggested that,
probably, there was a physiological relationship be- Classically, portal hypertension was considered to
tween the liver and the kidney because in experi- be due solely to an increased resistance to portal
mental studies it was shown that hepatic dener- venous inflow. However, at present, it is clear that
vation altered urinary sodium excretion in dogs the pathogenesis of portal hypertension is much
with cirrhlosis or with low grade caval constriction. more complex and that increased portal venous in-
These findings are consistent with the hypothesis flow secondary to a generalized arteriolar vasodil-
that probably there is a hepatorenal reflex. Its pres- atation also plays an important role in the in-
ence coulld explain why the activation of in- creased portal pressure (1). This high portal ve-
trahepatic baroreceptors by the increased sinus- nous inflow may explain why, in experimental
zyxwvuts
oidal pressure would produce urinary sodium re- animals, portal pressure remains increased despite
tention in these animal models. However, it has the development of a marked collateral circulation.
been demonstrated that cirrhosis in humans is as- The mechanism by which portal hypertension in-
zyxwvut
sociated with sinusoidal hepatic denervation sug- duces splanchnic arteriolar vasodilatation and in-
gesting that this hypothesis is doubtful (3). It has creased portal venous inflow is not entirely under-
also been speculated that a deficient hepatic syn- stood. There is evidence that some degree of porto-
thesis of iP putative vasodilator hormone may be systemic shunting is almost obligatory for its
the mechanism to explain the development of renal development (2). Portosystemic shunts allow sys-
vasoconstriction in hepatorenal syndrome (4, 5). temic delivery, and perhaps enhance production
However, up to now no data have been presented and release of vasodilating substances such as glu-
to support these mechanisms. Consequently, con- cagon, prostaglandins or nitric oxide (3).
sidering all these data, it seems clear that although There is indirect evidence that splanchnic arteri-
the physiological relationship between the liver and olar vasodilation also occurs in patients with cir-
kidney, is possible, its importance, from the physio- rhosis and portal hypertension. Under normal
logical point of view, is probably minor. (J. Rodes) conditions almost all the flow circulating in the
zyxwvutsrqpo
splanchnic bed reach the liver, and thus, the he-
patic equals the splanchnic blood flow. Since there
References
zyxwvutsrq
1. LEVYM, MXLER M J. Sodium excretion in dogs with low
grade caval constriction: role of hepatic nerve. Am J Physi-
01 1987; 253: F672-8.
2. LEVYM[,WEXLER M J. Hepatic denervation alters first-
phase urinary sodium excretion in dogs with cirrhosis. Am
J Physiol 1987; 253: F66471.
3. LAUGF, TSCHENKO E, SCHULZE E, et al. Hepatorenal reflex
regulating kidney function. Hepatology 1991; 14: 590-4.
is important shunting of blood through the porto-
systemic collateral circulation in cirhrosis and por-
tal hypertension, the hepatic blood flow represents
only a part of the splanchnic blood flow. Between
60 and 80% of the mesenteric and splanchnic
blood flow is shunted through the collateral circu-
lation in cirrhotic patients (3). Since several investi-
gations have shown that the hepatic blood flow is
4. ALVESTFLANI) A, BERCSTROM J. Glomerular hyperfiltration
after protein ingestion during glucagon infusion and insu- normal or even increased in most of such patients
lin-dependent diabetes is induced by a liver hormone and (4), it follows that there must be a marked increase
deficieni production of this hormone in hepatic failure in splanchnic blood flow secondary to arteriolar
causes hepatorenal syndrome. Lancet 1984; 1: 195-7.
5. SAFISSTEIN R, LEVITTM F. A hepatorenal depressor reflex:
vasodilation.
a possible clue to the pathogenesis of the hepatorenal syn- Moreover, it has been demonstrated that there
drome. Hepatology 1991; 14: 7 3 4 5 . is a close direct correlation between the degree of
15
,Questions and answers zyxwvutsrq
zyxwvu
portal hypertension and urinary sodium retention
(5). Urinary sodium retention being more irnport-
ant the higher the portal pressure is. In the past
it was considered that this phenomenon might be
10. HAN-CHIEH L, YANG T J, CHENG-CHUN
M, FA-YAUHL, JUNG-JANG
H, HONG-CHIAN
W, et al. Effects of octreotide
on postprandial systemic and hepatic hemodynamics in pa-
tients with postnecrotic cirrhosis. J Hepatol 1994; 21: 424-
9.
explained by the presence of a hepatorenal sym-
pathetic reflex (6, 7). However, nowadays evidence
has shown that the increased urinary sodium reab-
sorption is mediated by increased activity of the
renin-angiotensin-aldosterone and sympathetic QUESTION 3
nervous systems and ADH. The increased activity The physiological relationship between the liver and
of these substances could be a consequence of an kidney. Is it always present and stable in cirrhotic
intense arterial hypotension and reduced periph- patients?
eral resistance induced by the presence of an in-
tense arteriolar vasodilation in the splanchnic area. The fact that there is a hepatorenal connection,
The increased activity of these vasoconstrictors with cirrhotic patients developing renal sodium re-
allows arterial pressure to be maintained within tention, is clear. The question remains as to
the normal or near normal levels (8). whether it is a direct relationship, or indirect, re-
On the other hand, it is well known that food quiring a pertubation of the circulation first? Ani-
intake induces a marked increase in splanchnic mal experiments suggest that both are possible.
blood flow in both normal animals and humans. Animal studies in presinusoidal portal hyperten-
zyxw
zyxwvu
This postprandial increase in portal pressure and sion suggest that portal hypertension is followed
splanchnic blood flow is also observed in patients immediately by vasodilatation, with sodium reten-
with cirrhosis (9, 10). Therefore, by increasing por- tion established by the following day (1). Whereas
tal pressure, food intake may also produce an in- postsinusoidal portal hypertension is followed im-
crease in urinary sodium reabsorption in patients mediately by sodium retention in the absence of
with cirrhosis. However, as far I know, there is no any circulatory pertubation (2). The fact that
data supporting this hypothesis. (J. Rodks) sinusoidal portal hypertension is a trigger for this
phenomenon was indicated by the fact that the
normalisation of portal pressure in sodium retain-
References
ing cirrhotic dogs, via a side to side portocaval
1. VOROBIOFF J, BREDFIELDT J E, GROSLMANN R J. Increased shunt, resulted in a natriuresis (3). In cirrhotic pa-
blood flow through the portal system in cirrhotic rats. Gas- tients with ascites following a TIPS procedure, the
troenterology 1984; 87: 1120-6.
fall in portal pressure to below 10 mmHg eventu-
zyxwvuts
2. SIKULER E, KRAVETZ D, GROSZMANN R J. Evolution of por-
tal hypertension and mechanisms involved in its mainten- ally results in a spontaneous natriuresis, but not
ance in a rat model. Am J Physiol 1985; 28: G618-25. until renin activity and aldosterone levels improve
3. GROSZMANN R J. Mechanisms of portal hypertension. In: (4). However sodium retention occurs if the TIPS
Arroyo V, Bosch J, Rodes J, eds. Treatments in hepatology.
Barcelona: Masson SA, 1995: 3-5. is blocked leading to a rise in portal pressure above
4. BOSCHJ, MASTAI R, KRAVETZ D, BRUIXJ, RIGAUJ, RODES 12 mmHg (5). Therefore in cirrhotic patients, the
J. Measurement of azygos venous blood flow in the evalu- physiological liver-kidney relationship resulting in
ation of portal hypertension in patients with cirrhosis. Clin- renal sodium retention appears to depend on the
ical and hemodynamic correlations in 100 patients. J Hepa- height of the portal pressure. Therefore in answer
to1 1985; 1: 92-9.
5. BOSCHJ, ARROYOV, BETRIUA, MAS A, CARRILHOE to the question as to whether this phenomenon is
RIVERAF, et al. Hepatic hemodynamics and the renin- stable the answer must clearly be no! For example
angiotensin-aldosterone system in cirrhosis. Gastroenter- if the cirrhotic patient’s portal pressure rises as a
ology 1980; 78: 92-9. result of excess drinking, sodium retention will in-
6. PREMEN A J. Protein-mediated elevations in renal hemody-
namics: existence of a hepatorenal axis. Med Hypotheses crease and ascites appear. Conversely if cirrhotic
1986; 19: 295-309. patients with alcoholic hepatitis are hospitalised
7. BLANCHART A, RODRIGUEZ-PUYOL D, SANTOSJ L, HER- and the hepatitis resolves, the portal pressure falls,
NANDO L, LOPEZ-NOVOA J M. Effect of chronic and pro- renal sodium handling improves, the ascites will
gressive hepatic outflow blockade in renal function in rats.
J Lab Clin Med 1987; 109: 718-23.
become more responsive to diuretics and a natriur-
8. SCHRIERR W, ARROYO BERNARDI M, EPSTEINM, HEN- esis will follow (6).
RIKSEN J H, RODESJ. Peripheral vasodilation hypothesis: Is this phenomenon always present in cirrhotic
proposal for the initiation of renal sodium and water reten- patients? Even in well compensated patients with
tion in cirrhosis. Hepatology 1988; 8: 1151-7. normal or suppressed renin and aldosterone levels,
9. MCCORMICK P A, DICK R, GRAFFEO M, WAGERAFF D,
MASDDEN A, MCINTYRE N, et al. The effect of non protein who have never had ascites, the renal sodium han-
lipid meals on the hepatic venous pressure gradient in pa- dling abnormality can be demonstrated by sodium
tients with cirrhosis. J Hepatol 1990; l l: 221--5. challenges either in the form of saline infusion (7)
16
,or a high sodium diet (8). Ascitic patients success-
fully treated with peritoneovenous shunting, and
having lost all their ascites, but still portal hyper-
tensive, when challenged with a high sodium diet,
7. WONGF, MASSIE
Questions and answers
decrease in portal pressure with clinical improvement in cir-
rhosis. N Engl J Med 1960; 263: 734-9.
D, HSU P, DUDLEY F. The renal response
to a saline load in well compensated alcoholic cirrhosis.
zy
Hepatology 1994; 20: 873-81.
move into positive sodium balance despite physio- 8. W~RNER L C, CAMPBELL P J, MORALIG A, LOGANA G,
logical suppression of normal renin and aldoster- SKORECKI K L, BLENDIS L M. The response of atrial natriu-
retic factor and sodium excretion to dietary sodium chal-
one levels (9). However as to whether all well com- lenges in patients with chronic liver disease. Hepatology
pensated cirrhotic patients suffer from this pheno- 1990; 12: 480-6.
men on the answer is clearly no. What are the 9. GREICP D, BLENDIS L M, LANGER B, TAYLOR B R, COLAP-
factors determining this? Probably the most im- INTO R E The renal and hemodynamic effects of the per-
portant is that cirrhotic patients with an hepato- itoneovenous shunt. Long term effect. Gastroenterology
1981; 80: 119-25.
portal gradient of less than 8-10 mmHg do not 10. RECTORW G. Portal hypertension; a permissive factor only
show this abnormality (10, 11). Secondly, it is less in the development of ascites and variceal bleeding. Liver
common in cirrhotic patients with presinusoidal 1986; 6: 221-6.
portal hypertension (12), who seem to require ad- 11. MORALIG, SNIDERMAN K W, DEITELK M, TOBES, WITT-
zyxwvutsr
SULLIVAN H, SIMON M, et al. Is sinusoidal portal hyperten-
ditional factors such as systemic vasodilatation sion a necessary factor for the development of hepatic as-
(l).Thirdly, the better the liver function the less cites? J Hepatol 1992; 16: 249-50.
likely it is able to demonstrate sodium retention 12. SOLWAY R D, SUMMERSKILL H W, BAGCENSTROSS A N, GE-
even in paitients with sinusoidal or post sinusoidal ALL M G, GITNICK G L, ELVEBACK L R, et al. Clinical,
portal hypertension (13). Indeed there may be a biochemical and histological remission of severe chronic ac-
tive hepatitis. Gastroenterology 1972; 63: 820-933.
threshold reduction of 50% in functional liver cell 13. WENSING G, LOTTERER E, LINKI, HAHNE G, FLEICW E.
mass below which sodium retention may occur, but Urinary sodium balance in patients with cirrhosis: relation-
above which it is unlikely to occur (13). This may ship to quantitative parameters of liver function. Hep-
be an important explanation why patients with atology 1997; 26: 1149-55.
presinusoidal hypertension rarely develop ascites,
since they can maintain excellent liver function for
many years.
In conclusion the phenomenon of the cirrhotic
liver sending signals directly to the kidney to retain QUESTION 4
sodium appears to be dependent on at least three Which are the principal factors favouring portal
zy
zyxwvu
factors; the site of the increased resistance, sinus- hypertension in liver cirrhosis?
oidal as opposed to presinusoidal; the height of the
sinusoidal resistance with a critical level around 10 Portal hypertension is by definition a pathological
mmHg hepatoportal gradient; and the permissive, increase in portal pressure. This increase is due to
intrinsic functional liver cell mass. Therefore it is an increased portal blood flow together with an
not alwayis present and is not stable since all three increased intrahepatic resistance. The resistance in
factors are variable. (L.M. Blendis) the liver is increased by 2 main mechanisms - 1)
an obstructive process, nodules, fibrosis, or throm-
bosis whose site may be pre-hepatic, hepatic and/
or post-hepatic (in relation to the sinusoid) (1) and
References 2) an increase in vascular tone, which is reversible
1. ALBILLOS A, COLOMBATA L A, GROZSMAN R J. Vasodilat- (2). Hepatic stellate cells may contribute to the
ation and sodium retention in prehepatic portal hyperten- regulation or modulation of intrahepatic resistance
sion. Gastroenterology 1992; 102: 931-5. (3). The characteristics of activated hepatic stellate
2. CAMPBELL GREIG CRANFORD J, LANCER B, SILVERMAN cells are now considered to be myofibroblasts,
M, BLENDIS L M. A comparison of acute reversible pre-
and postsinusoidal portal hypertension on salt and water
which have been shown to play a role in regulation
retention in the dog. Hepatology 1982; 2: 54-8. of intrahepatic resistance (2). Vascular biology is
3. UNIKOWSKY B, WXLER M J, LEVYM. Dogs with experi- centred on the vascular endothelium, which syn-
mental cirrhosis of the liver but without intrahepatic portal thesises vasodilators (nitric oxide, prostacyclins)
zy
hypertension do not retain sodium or form ascites. J Clin and vasoconstrictors (endothelins, prostanoids),
Invest 1083; 72: 1594-604.
4. WONGF; SNIDERMAN K, LIU P, BLENDISL M. Initial two which act in a paracrine fashion (4). Changes in
stage na triuresis post TIPS. Gastroenterology 1997; 112: this balance lead to abnormal vascular tone. Thus,
899-907. the increased vascular tone in the liver in cirrhosis
5. CASADO M, BOSCHJ, GARCIA-PAGAN J C, BRUC, BANARES could exist by any combination of decreased vaso-
R, BRANDI J C, et al. Clinical events after transjugular in- dilators, increased vasoconstrictors or both. There
trahepatic portosystemic shunt: correlation with hemody-
namic findings. Gastroenterology 1998; 114: 1296-303. is some evidence for decreased nitric oxide ( 5 ) , al-
6. REYNOLDS T B, GELLERH M, KUZMER 0 T. Spontaneous though not all researchers agree. There is an in-
17
, Questions and answers zyxwvutsr
creased responsiveness to endothelins (6). In ad-
dition to the increase in intrahepatic resistance
there is also an increase in portosystemic collateral
lated intracellular calcium transits. J Clin Invest 1992; 90:
642-6.
4. VANEJ R, ANGGARD E E, BOLTINGR M. Regulatory func-
tions of vascular endothelium. N Engl J Med 1990; 323:
resistance. The collateral circulation only develops 21-36.
as a consequence of portal hypertension, but the 5. GIJPTAT K, GROSZMANN R J. Administration of L argi-
vascular resistance of these channels “which nine, the physiological precursor of nitric oxide, reduces
portal perfusion pressure and ameliorates hepatic vascular
should decompress the portal system” is greater hyperreactivity in experimental cirrhosis. Hepatology 1994;
than the resistance of the normal liver. It has been 20: 200A.
shown in animals that preventing the increase in 6. ROCKEY D C. The cellular pathogenesis of portal hyperten-
splanchnic blood flow alone without decreasing sion: stellate cell contractility endothelin and nitric oxide.
Hepatology 1997; 25: 2-5.
portal pressure leads to an improvement in the col- 7. LEEF Y, COLOMBATO A, ALBILLOS A, et al. Administration
lateral circulation (7). However, there is no direct of N Omega nitro L arginine ameliorates portal systemic
relationship between the portal venous inflow and shunting in portal hypertensive rats. Gastroenterology
the size of the collateral circulation. Studies with 1993; 105: 1464-70.
propranolol have shown that collateral circulation 8. HALVORSEN J F, MYKINGA 0. The portosystemic collateral
pattern in the rat. Eur Surg Res 1974; 6: 183-95.
can be prevented in part, thus its action may also 9. MOSCAP, LEEF Y, KAUMANN A J, et al. Pharmacology of
be via reduction in portal pressure (8) as well as portal systemic collaterals in portal hypertensive rats: role
reducing portal inflow. It is likely that nitric oxide of endothelium. Am J Physiol 1992; 263: G540-4.
may be the principal factor initiating collateraliz- 10. GARCIA-TSAO G, GROSZMANN R J, FISHER R L, et al. Portal
pressure, presence of gastroesophageal varices and variceal
ation of the portal system, probably resulting from bleeding. Hepatology 1985; 5: 419-24.
the increase in splanchnic flow. This opens up pre-
formed channels as it is known that initially there
is no de novo formation of new channels (9).
The increased portal blood flow is considered to
be related to the peripheral vasodilatation in cir- QUESTION 5
rhosis, leading to decreased systemic resistance, an Is the precise identification of the pathogenetic
expanded plasma volume, an increased cardiac mechanism of sodium retention necessary for correct
index and a rise in splanchnic arterial blood flow
treatment?
resulting in increased portal blood flow.
The portal pressure increase once above a It is now becoming clearer and more accepted that
threshold of 10 mmHg leads to the formation of there are at least two important pathogenetic
oesophageal varices (lo), but not all patients de- mechanisms of sodium retention in cirrhosis. First
velop these collaterals. there is a primary sodium retention which develops
Drug therapy for portal hypertension is aimed in patients with predominantly sinusoidal (for ex-
at reducing the increased splanchnic blood flow ample, alcoholic) or post sinusoidal (for example,
and intrahepatic resistance. Current interest is in chronic Budd Chiari Syndrome) portal hyperten-
patients who as yet, have not reached a sufficient sion (1). The initiation of this process at the very
portal pressure rise to develop varices. Varices least involves portal hypertension, which is associ-
might be prevented in human cirrhosis by earlier ated with localised splanchnic dilatation. The
administration of portal hypotensive agents. This message from the liver to the kidney to retain so-
has yet to be proven, but an NIH funded trial is dium is still unclear but may well involve a neuro-
zyxwvu
currently in progress. As yet “normalisation” of genic pathway (2). In the pre-ascitic patients, kid-
the decreased systemic resistance has not been ney activation of tubular factors, of the renin-
tried in order to reduce the increased splanchnic angiotensin-aldosterone system (RAAS), affecting
zyxwvutsrqpo
blood flow, but reduction in cardiac output does initially both proximal and distal segments, causes
occur with beta-blockers, as it is a mode of action increased salt and water reabsorption (3). This re-
zyxwv
for lowering portal pressure. Nitrates act by reduc- sults in expansion of the extracellular volume, in-
ing intrahepatic resistance. (A.K. Burroughs) cluding the total and central blood volumes (4).
These pre-ascitic patients return to sodium balance
by compensatory suppression of RAAS and in-
References creased levels of plasma atrial natriuretic peptide
H, ZAKF G. Pathological aspects of cirrhosis. Am
1. POPPER (ANP) (5), but retain extracellular volume, ex-
J Med 1958; 24: 593-625. panded with an increased total body sodium.
2. BHATHAL P S, GROSSMAN H J. Reduction of the increased Both animal experiments (6) and clinical investi-
portal vascular resistance of the isolated perfused cirrhotic
rat liver by vasodilators. J Hepatol 1985; 1: 325-37. gations (7) have shown that sodium restriction par-
3. PINZANI M, FAILLI P, Ruocco C, et al. Fat storing cells as tially reverses these phenomena, including blunting
liver specific pericytes spatial dynamics of agonist stimu-
- the blood volume expansion and reducing portal
18
