INTRODUCTION — Ascites is defined as the pathologic accumulation of fluid in the peritoneal cavity. It is the most common complication of cirrhosis, which is the most common cause of ascites in the United States, accounting for approximately 85 percent of cases. Within 10 years after the diagnosis of compensated cirrhosis, about 50 percent of patients will have developed ascites .
The development of ascites is the final consequence of a series of anatomic, pathophysiologic, and biochemical abnormalities occurring in patients with cirrhosis. The two older theories of ascites formation, the underfill theory  and the overflow theory , appear to be relevant at different stages of the natural history of cirrhosis . However, the most recent theory, the arterial vasodilation hypothesis, appears to match best with the actual hemodynamic data and has become the most widely accepted theory .
The formation of ascites is governed by the same principles as edema formation at other sites: net capillary permeability and the hydraulic and oncotic pressure gradients. (See "Pathophysiology and etiology of edema in adults".)
PORTAL HYPERTENSION — The development of portal hypertension (PHT) is the first step toward fluid retention in the setting of cirrhosis. Patients with cirrhosis but without PHT do not develop ascites or edema . A portal pressure >12 mmHg appears to be required for fluid retention [6,7]; on the other hand, ascites will usually disappear if portal pressure is reduced below 12 mmHg (eg, after a surgical or radiologic portosystemic shunt) . Sinusoidal hypertension appears to be required for fluid retention to occur; presinusoidal portal hypertension, as in portal vein thrombosis, does not result in ascites formation in the absence of another predisposing factor.
PHT leads to profound changes in the splanchnic circulation. Although it was formerly thought that PHT was due solely to a mechanical obstruction to portal flow, data from animal models provide evidence for a component of increased portal venous inflow as a consequence of splanchnic arterial vasodilation [9,10].
Patients with cirrhosis and clinically significant PHT have several circulatory, vascular, functional, and biochemical abnormalities that contribute to the pathogenesis of fluid retention (table 1).
Vasodilation and hyperdynamic circulation — Patients with cirrhosis and ascites usually have a marked reduction in systemic vascular resistance (SVR) and in mean arterial pressure (MAP) plus an increase in cardiac output [11-13]. These abnormalities result in a hyperdynamic circulation which can be found in patients and experimental animals with cirrhosis before the development of ascites [14-16].
The vascular territory where the reduced SVR is most obvious is the arterial splanchnic circulation . The presence of this abnormality in other vascular territories is less obvious and the subject of controversy [15,17].
Mechanisms of vasodilation — Considerable effort has been made to elucidate the exact mechanism(s) of arterial vasodilation and the hyperdynamic circulation of cirrhosis (algorithm 1). Anatomic and functional liver-related causes have been considered in order to explain the presence of vasodilation.
The opening of portasystemic collaterals, a frequent finding in cirrhosis, helps explain the presence of vasodilation . The performance of portocaval shunts in these patients further decreases SVR .
Although portosystemic collaterals may contribute, most of the studies which have examined the vasodilation in patients with cirrhosis have focused on increased levels of circulating vasodilators. Glucagon has been one of the most widely studied [18,19], although its role in the pathogenesis of vasodilation has not been precisely defined [6,12]. Other vasodilators have been considered, such as vasoactive intestinal peptide, substance P, platelet-activating-factor or prostaglandins .
Increased synthesis of systemic prostacyclin has been observed in patients with cirrhosis even before they develop ascites . The synthesis of renal prostacyclin is increased in these patients  and contributes to maintenance of the glomerular filtration rate (as evidenced by reductions in glomerular filtration rate and renal plasma flow following the administration of a nonsteroidal antiinflammatory drug) . Renal prostaglandin synthesis decreases with advanced liver disease and may contribute to the marked renal vasoconstriction in patients with hepatorenal syndrome . (See "Hepatorenal syndrome", section on 'Pathogenesis'.)
Patients with cirrhosis but without portal hypertension (eg, patients after portasystemic shunt) also show increased levels of systemic prostacyclin . It is possible that this activation is related to the presence of endotoxemia, since selective intestinal decontamination with nonabsorbable antibiotics significantly decreases the synthesis of systemic prostacyclin .
Although the above factors may play a contributory role, much of the recent literature has focused on the possible role of nitric oxide (NO) [24,25]. The following observations suggest that NO is the primary mediator of vasodilation in cirrhosis:
●The activity of endothelial NO synthase (which promotes the synthesis of NO from L-arginine) is increased in the arterial vessels of cirrhotic rats with ascites .
●The serum levels of nitrite and nitrate, an index of in vivo NO synthesis, are significantly higher in patients with cirrhosis than in controls .
●Inhibition of the synthesis of NO in rats with cirrhosis significantly increases the arterial pressure and SVR, decreases the cardiac index  and reverses the impaired response to vasopressors [29,30].
The possible factors responsible for the increased NO synthesis in cirrhosis have been intensively studied. Nitric oxide production may be stimulated by endotoxin or other bacterial products, such as bacterial DNA from the gastrointestinal tract, which are less efficiently cleared due to portal-systemic shunting and decreased reticuloendothelial cell function in cirrhosis. The increased synthesis of NO is mediated by both the endothelial  and inducible forms . Interestingly, it has been shown in animal models of cirrhosis that bacterial translocation may be present prior to the development of ascites, providing a rationale for the increased synthesis of NO even before ascites is present . The following observations support the role of endotoxin or other bacterial products in the stimulation of NO production (algorithm 1) [24,27]:
●A significant correlation has been noted between serum nitrite and nitrate levels and endotoxin .
●Oral administration of the antibiotic colistin to patients with cirrhosis significantly reduces plasma endotoxin levels and the serum concentration of nitrite and nitrate , and norfloxacin increases systemic vascular resistance and decreases plasma renin in a subgroup of patients with ascites .
●Fragments of bacterial DNA in blood and ascitic fluid have been described in one-third of patients with cirrhosis and ascites and up to two-thirds of patients with refractory ascites [37-39]. Bacterial DNA is rarely present in patients with cirrhosis who do not have ascites. Bacterial DNA induces the synthesis of nitric oxide by peritoneal macrophages, suggesting that it may be related to the hemodynamic derangement observed in patients with advanced cirrhosis . In addition, bacterial DNA may further impair the endothelial dysfunction commonly found in this setting . Some data suggest that its presence is correlated with a poor prognosis, but more studies are needed .
Animal models of cirrhosis have shown that the presence of bacterial DNA in biological fluids is invariably associated with its simultaneous presence in mesenteric lymph nodes, either as viable (culture-positive) or nonviable (culture-negative) forms [42-45]. These observations have led to the modification of the concept of bacterial translocation to include the presence of bacterial products in mesenteric lymph nodes, and not only a positive culture.
CONSEQUENCES OF VASODILATION — The progressive vasodilation seen in cirrhosis leads to the activation of endogenous vasoconstrictors, sodium and water retention, and increasing renal vasoconstriction.
Activation of endogenous vasoconstrictor agents — The reduction in pressure (or stretch) at the carotid and renal baroreceptors induced by cirrhotic vasodilation results in activation of the sodium-retaining neurohumoral mechanisms in an attempt to restore perfusion pressure to normal (algorithm 1). These include the renin-angiotensin-aldosterone system, sympathetic nervous system, and antidiuretic hormone (vasopressin). The secretion of these "hypovolemic" hormones is proportional to the severity of the hemodynamic insufficiency (figure 1) [46-48].
The net effect is avid sodium and water retention because the patient is effectively volume depleted even though extracellular sodium stores, the plasma volume, and the cardiac output are increased.
Sodium retention — The retention of sodium and water increases the plasma volume. If this were adequate to refill the intravascular space, the activity of the endogenous vasoconstrictor systems would decrease, with a progressive normalization of the excretion of sodium and water (algorithm 2) . However, impaired sodium excretion after a saline load  and a reduction in central blood volume  have been demonstrated in patients with cirrhosis who have not yet developed ascites. This unstable equilibrium can be affected by infections, drugs, or progressive impairment in liver function. In advanced disease, sodium excretion often falls to less than 10 mEq/day .
Thus, sodium retention is a sensitive marker of the overall status of the patient with cirrhosis. A significant nonlinear relationship has been identified between urinary sodium excretion and the aminopyrine breath test, a measurement of true liver function ; the presence of sodium retention was indicative of at least a 50 percent reduction in liver function . True liver function usually drops below 60 percent before ascites forms according to a newer measure of liver function, the perfused hepatic mass or quantitative liver-spleen scan . This nuclear medicine-based test received approval from the US Food and Drug Administration in January of 2015.
Fluid overload is an important landmark in the natural history of these patients, and the degree of sodium retention is inversely related to survival. In one series, for example, patients with ascites and urinary sodium excretion below 10 mEq/day had a mean survival as low as five to six months, in comparison to over two years in those with ascites and a higher rate of sodium excretion .
Water retention — Water excretion is usually normal in patients with cirrhosis before the development of ascites and then becomes increasingly impaired as the liver disease progresses. This abnormality is largely related to the increased release of antidiuretic hormone (ADH) described above [46-48], since suppression of ADH release is required to excrete a water load. The pathogenetic importance of ADH in water retention in cirrhosis has been demonstrated in rats with cirrhosis in which the administration of an ADH receptor antagonist restores near normal water excretory ability . The inability to excrete water regularly leads to the development of hyponatremia and hypoosmolality . (See "Hyponatremia in patients with cirrhosis".)
Thus, patients with cirrhosis and ascites usually demonstrate urinary sodium retention, increased total body sodium, and dilutional hyponatremia. It can be challenging to explain to patients and their families (and even to some clinicians) that the hyponatremia does not indicate a deficiency of sodium.
Because the increase in ADH secretion (and therefore the degree of water retention) is roughly proportional to the severity of the cirrhosis (figure 1), the degree of hyponatremia parallels the severity of the liver disease and is, along with the degree of sodium retention, of prognostic value. The severity of hyponatremia correlates with worsening survival [55-58].
Renal vasoconstriction — The activation of vasoconstrictor systems tends to reduce renal blood flow . Renal perfusion may initially be maintained due to vasodilators such as prostaglandins and perhaps nitric oxide (algorithm 2). However, the natural progression of liver disease overcomes these protective mechanisms, leading to progressive renal hypoperfusion, a gradual decline in the glomerular filtration rate, and, in some patients, the hepatorenal syndrome [15,60]. (See "Hepatorenal syndrome", section on 'Pathogenesis'.)
The importance of splanchnic vasodilation in the genesis of renal ischemia has been indirectly illustrated by the response to ornipressin and terlipressin, analogs of antidiuretic hormone (arginine vasopressin) that are preferential splanchnic vasoconstrictors [10,61,62]. In patients with advanced cirrhosis, the administration of ornipressin or terlipressin raised the mean arterial pressure, increased renal blood flow, glomerular filtration rate, and urinary sodium excretion and volume. (See "Hepatorenal syndrome", section on 'Terlipressin plus albumin where available'.)
The reduction in glomerular filtration rate in patients with liver disease is often masked clinically. Both urea and creatinine production may be substantially reduced in this setting, due to the liver disease and to decreased muscle mass. The net effect is that a serum creatinine concentration that appears to be within the normal range (1 to 1.3 mg/dL or 88 to 115 micromol/L) may, depending primarily upon muscle mass, be associated with a glomerular filtration rate that ranges from as low as 20 to 60 mL/min to a clearly normal value above 100 mL/min [63,64].
SUMMARY AND RECOMMENDATIONS
●Ascites is defined as the pathologic accumulation of fluid in the peritoneal cavity. It is most commonly caused by cirrhosis. (See 'Introduction' above.)
●The development of portal hypertension (PHT) is the first step toward fluid retention in the setting of cirrhosis. Patients with cirrhosis but without PHT do not develop ascites or edema. A portal pressure >12 mmHg appears to be required for fluid retention. (See 'Portal hypertension' above.)
●Patients with cirrhosis and clinically significant PHT have several circulatory, vascular, functional, and biochemical abnormalities that contribute to the pathogenesis of fluid retention (table 1). (See 'Portal hypertension' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff thank Dr. José Such, MD, PhD for his contributions as author to prior versions of this topic review.
آیا می خواهید مدیلیب را به صفحه اصلی خود اضافه کنید؟