Ascites and Spontaneous Bacterial Peritonitis

Sodium Retention and Extracellular Fluid Volume Expansion

Sodium retention is the most frequent and earliest renal abnormality in patients with cirrhosis and is the key factor in the expansion of the extracellular fluid volume and the development of ascites and edema. Sodium is retained isosmotically together with water, and, therefore, sodium retention is associated with extracellular fluid volume expansion. The amount of sodium retained depends on the balance between sodium in the diet and sodium excreted in the urine. If the sodium excreted in the urine is lower than that ingested, ascites and edema will develop. The central role of sodium retention in the pathogenesis of ascites is supported by the observation that ascites can resolve as a result of either a reduction in dietary sodium intake or enhancement of sodium excretion by diuretics. In fact, the achievement of a negative sodium balance by increasing urinary sodium excretion is the goal of pharmacologic therapy for ascites in patients with cirrhosis (see later).

The degree of sodium retention in patients with cirrhosis and ascites is highly variable from patient to patient. Some patients have relatively high urinary sodium excretion, whereas others have a low urine sodium concentration ( Fig. 93.2 ). Most patients who require hospitalization because of severe or difficult-to-control ascites have marked sodium retention (urine sodium excretion <10 mEq/day), and sodium retention is particularly intense in patients with refractory ascites. By contrast, in patients with cirrhosis and mild-to-moderate ascites, the proportion of patients with marked sodium retention is low, and most such patients excrete more than 10 mEq/day (without diuretic therapy). In addition, response to diuretic treatment is usually better in patients with moderate sodium retention than in those with marked sodium retention.

In healthy subjects, approximately 95% of filtered sodium is reabsorbed in the renal tubules (approximately 60% to 70% in the proximal tubules, 30% to 40% in the thick ascending limb, and 5% to 10% in the collecting ducts). In most cases, sodium retention in cirrhosis is due to increased tubular reabsorption of sodium because it occurs in the presence of a normal or only moderately reduced GFR. The contribution of the different segments of the nephron to increased sodium reabsorption in patients with cirrhosis is not completely known, as experimental and clinical studies have yielded discrepant findings. Studies using lithium clearance, which estimates sodium reabsorption in the proximal tubule, suggest that cirrhotic patients with ascites have a marked increase in proximal sodium reabsorption. On the other hand, clinical studies using spironolactone to antagonize the mineralocorticoid receptor indicate that this agent induces natriuresis in a large proportion of patients with cirrhosis and ascites in the absence of renal failure, thereby suggesting a major role for increased sodium reabsorption in distal sites of the nephron.

The overall data suggest that in patients with cirrhosis without renal failure, sodium retention is caused by enhanced reabsorption of sodium in both the proximal and distal tubules. As described earlier, the increased activity of the RAAS and SNS plays a major role in increased renal sodium reabsorption. ^{, ,} Sodium retention is usually more marked in patients with renal failure than in those without renal failure as a result of both a reduction in filtered sodium and a marked activation of sodium-retaining systems.

Portal Hypertension

Portal hypertension represents the triggering factor for the development of circulatory dysfunction in patients with advanced cirrhosis. ^, Cirrhosis is the result of a prolonged process, usually more than 20 years, of progressive liver inflammation and fibrosis in response to chronic injury (e.g., alcohol consumption, chronic viral hepatitis, NAFLD) (see Chapter 74 ). The development of cirrhosis causes marked structural abnormalities in the liver, thereby resulting in a marked disturbance in the intrahepatic circulation, which in turn causes increased resistance to portal flow and subsequent hypertension in the portal venous system (see Chapter 92 ). Progressive collagen deposition and formation of nodules in the hepatic parenchyma lead to architectural distortion of sinusoidal blood flow, thereby resulting in increased intrahepatic resistance. ^{, ,} In addition to passive resistance to portal flow due to architectural changes, a significant component of the increased resistance results from dynamic interaction between injured hepatocytes, contraction of hepatic stellate cells (HSCs) and hepatic endothelial cells, together with an imbalance in the levels of intrahepatic vasodilator and vasoconstrictor substances. Nitric oxide (NO) has been shown to be a key regulator of intrahepatic vascular tone. Abundant evidence suggests that, despite the overproduction of vasodilator factors such as NO in the splanchnic and systemic circulation in cirrhosis, the production of NO from endothelial nitric oxide synthase is reduced in the intrahepatic circulation of cirrhotic livers and contributes to increased intrahepatic resistance. Moreover, intrahepatic vascular tone is also regulated by HSCs that show a myofibroblastic phenotype after activation. Activated HSCs have increased contractility, leading to increased vascular tone and increased intrahepatic resistance. ^, Finally, intrahepatic inflammation has also been described to play a role in the increased vascular resistance leading to portal hypertension. In advanced cirrhosis, Kupffer cells have been involved in the development of hepatic inflammation and oxidative stress, leading to increased intrahepatic vascular resistance. In response to pathogen-associated molecular patterns and Toll-like receptor signaling, Kupffer cells induce the production of proinflammatory cytokines, reactive oxygen species, and vasoactive mediators, leading to hepatic and systemic inflammation and, thus, to increased intrahepatic vascular tone.

In clinical practice, portal hypertension is assessed by the measurement of the hepatic venous pressure gradient (HVPG), defined as the difference between wedged and free hepatic venous pressure, measured by hepatic vein catheterization (see Chapter 92 ). The normal portal pressure by this method is up to 5 mm Hg, whereas clinically significant portal hypertension is defined as an HVPG above 10 to 12 mm Hg, because this is the threshold for clinical manifestations of portal hypertension (e.g., ascites) to develop. Moreover, the severity of portal hypertension has also been associated with prognosis; an HVPG above 16 mm Hg identifies patients at high risk for mortality, and an HVPG above 20 mm Hg is associated with treatment failure and mortality in patients with cirrhosis and acute variceal bleeding (see Chapter 92 ).

Systemic Circulatory Dysfunction

A large body of evidence indicates that impairment in circulatory function is the main cause of renal dysfunction in patients with cirrhosis and leads to the development of the major complications of the disease, such as ascites, hyponatremia, and HRS. ^{, ,} The arterial vasodilatation theory, proposed in 1988, describes the hemodynamic disturbances that occur in patients with decompensated cirrhosis as characterized by systemic arterial vasodilatation, particularly in the splanchnic circulation. As mentioned earlier, portal hypertension is the initial event, resulting in splanchnic arterial vasodilatation due to the release of several vasodilating factors, such as NO, carbon monoxide, and endogenous endocannabinoids. Splanchnic arterial vasodilatation leads to decreased vascular resistance and, consequently, to reduction in effective arterial blood volume and arterial pressure.

In the early stages of cirrhosis, when patients are still asymptomatic, the increase in hepatic vascular resistance and, therefore, in portal pressure are moderate. In this setting, systemic vascular resistance is slightly reduced due to moderate splanchnic arterial vasodilatation, which can be counterbalanced by an increase in cardiac output, permitting the maintenance of arterial volume and arterial pressure within normal limits. In advanced stages of cirrhosis, when patients have already developed complications of the disease, splanchnic arterial vasodilatation is intense, leading to a marked reduction in systemic vascular resistance that cannot be compensated for by a further increase in cardiac output. At this stage, effective arterial hypovolemia develops due to the disparity between intravascular blood volume and the enlarged intravascular arterial circulation due to vasodilatation. Moreover, at this stage an associated decrease in cardiac output also contributes to the arterial underfilling. Vasoconstrictor systems such as the RAAS, SNS, and, at later stages of the disease, vasopressin are activated, as a homeostatic response to maintain arterial pressure within normal limits. The activation of these systems helps maintain effective arterial blood volume and arterial pressure within normal limits but has important detrimental effects on kidney function, with sodium and solute free-water retention leading to ascites and edema and to dilutional hypernatremia, respectively. At late stages of the disease, if the activation of these systems is extreme, patients develop marked renal vasoconstriction, leading to a reduction in the GFR and the development of HRS (see Chapter 94 and Fig. 93.1 ). ^,

Sympathetic Nervous System

The plasma concentration of norepinephrine (NE) in the systemic circulation, a marker of the activation of the SNS, is increased in most cirrhotic patients with ascites and normal or only slightly elevated in patients without ascites. The increase in plasma NE levels is due to an increase in the activity of the SNS rather than to impaired elimination of NE, because the release of NE into plasma is markedly increased in cirrhotic patients with ascites but clearance of NE from plasma is normal. ^, The activity of the SNS is increased in many vascular territories, including the kidneys, splanchnic organs, heart, muscle, and skin, thus supporting the concept of a generalized activation of the SNS. ^,

Ample evidence suggests that the SNS is involved in sodium and water retention in cirrhosis. The activity of the SNS correlates inversely with sodium and water retention. ^, In addition, a study in a small number of patients with ascites showed that administration of diuretics together with clonidine to inhibit SNS activity is more effective than diuretics alone in inducing diuresis. The cause of the increased activity of the SNS in patients with cirrhosis and ascites is not completely understood; the most likely explanation is a baroreceptor-mediated response to decreased effective arterial blood volume due to arterial vasodilatation, ^, as supported by data showing that the activity of the SNS can be suppressed by increasing effective arterial blood volume with, for example, the administration of vasopressin analogs and albumin or the insertion of a TIPS or peritoneovenous shunt (see later and Chapter 92 ).

Systemic Inflammation

Growing evidence indicates that decompensated cirrhosis is associated with chronic and progressive systemic inflammation that may also play an important role in the progression of disease and development of complications (see Chapter 2 ). ^, Decompensated cirrhosis is associated with increased serum levels of inflammatory markers, such as C-reactive protein and the leukocyte count, which increase in parallel with disease severity and independently of the presence of bacterial infections. ^, Moreover, patients with advanced cirrhosis exhibit increased serum levels of proinflammatory cytokines such as interleukin (IL)-6, IL-8, and TNF-α. ^,

Patients with cirrhosis and ascites experience bacterial translocation, the passage of bacteria or bacterial products from the gut to mesenteric lymph nodes, mainly due to an increase in gut permeability (see later). The hypothesis has been raised that these bacterial products, known as pathogen-associated molecular patterns (PAMPs), may activate pattern recognition receptors present in circulating innate immune cells, in turn leading to the activation of immune cells, the release of proinflammatory mediators and reactive oxygen species, and the consequent development of an inflammatory response. In addition, damage-associated molecular patterns (DAMPs) derived from the injured liver due to local inflammation and cell death may also contribute to the activation of pattern recognition receptors. The systemic release of these inflammatory mediators contributes to further impairment of circulatory dysfunction (see Fig. 93.1 ). ^,

Laboratory Tests

Liver function should be evaluated by standard liver biochemical and coagulation tests (see Chapter 73, Chapter 94 ). Assessment of renal function should include the serum creatinine level and serum and urine electrolyte concentrations, as well as a 24-hour urine collection for sodium and protein. These laboratory tests should be performed before diuretic treatment is initiated. ^{, ,}

Assessment of Renal Sodium Excretion

Assessment of the urinary excretion of sodium is useful for the management of patients with cirrhosis and ascites because it allows the quantification of sodium retention. Ideally, urine should be collected under conditions of controlled sodium intake (a low-sodium diet of approximately 90 mEq/day during the previous 5 to 7 days), because sodium intake may influence sodium excretion. Although the measurement of sodium concentration on a “spot” analysis of urine provides an estimate of sodium excretion, the assessment of sodium excretion in a 24-hour period is preferable because it is more representative of sodium excretion throughout the day. Sodium excretion should be measured without diuretic therapy in patients with a first episode of ascites or with worsening of pre-existing ascites (e.g., a marked increase in ascites despite treatment). ^{, ,} The measurement of sodium excretion in patients on diuretic therapy may be useful for monitoring the response to treatment (see later).

Measurement of baseline sodium excretion is also useful because it helps predict the response to diuretic treatment and has been associated with prognosis. Patients with moderate sodium retention (urine sodium ≥10 mEq/day) are more likely to respond to lower doses of diuretic treatment than those with marked sodium retention. Finally, the degree of sodium retention also provides prognostic information in patients with cirrhotic ascites. Patients with a baseline urine sodium excretion lower than 10 mEq/day have a median survival time of only 1.5 years, compared with 4.5 years in patients with urine sodium greater than or equal to 10 mEq/day ( Fig. 93.3 ). ^,

Abdominal US

In all patients with a first episode of ascites, abdominal imaging should be performed in order to support the diagnosis of cirrhosis by evaluating the liver parenchyma and to assess the patency of the portal vein and suprahepatic veins and rule out a liver tumor. Abdominal US is the technique of choice because it is simple and cost effective. In addition to all patients with the first presentation of ascites, US should be performed in patients with known ascites who experience unexplained loss of response to treatment. ^,

Ascitic Fluid Analysis

The analysis of ascitic fluid is essential for detecting ascitic fluid infection and excluding causes of ascites other than cirrhosis, in cases in which the diagnosis is not clear. A diagnostic paracentesis with a standard 1.5-inch (longer in obese persons), 22-gauge steel needle should be performed in all patients who present with a first episode of grade 2 or 3 ascites, as well as in those patients with ascites admitted to the hospital for any intercurrent complication. The ascitic absolute polymorphonuclear leukocyte (neutrophil) count and total protein and albumin concentrations should always be assessed, along with an ascitic fluid culture. ^{, ,}

An ascitic neutrophil count higher than 250/mm ³ is diagnostic of SBP (see later). The ascitic fluid protein concentration has been shown to be related to prognosis. Moreover, an ascitic fluid protein less than 1.5 g/dL is also associated with an increased risk of developing SBP (see later).

Ascitic fluid culture should be performed by inoculating at least 10 mL of ascitic fluid into blood culture bottles immediately after paracentesis. ^{, ,} Culture of the fluid will be highly helpful for guiding treatment if ascitic fluid infection is confirmed. The most common cause of ascitic fluid infection is SBP, and, in that case, culture is expected to be monomicrobial. In cases of polymicrobial culture results, secondary bacterial peritonitis should be suspected.

The serum-ascites albumin gradient (SAAG) is a sensitive and specific measurement to determine whether ascites is related to portal hypertension ( Box 93.2 ). ^{, ,} Calculating the SAAG involves measuring the albumin concentration in serum and ascitic fluid and simply subtracting the ascitic fluid value from the serum value; the gradient is calculated by subtraction and is not a ratio. If the SAAG is 1.1 g/dL (11 g/L) or greater, the patient can be considered to have portal hypertension with an accuracy of approximately 97%. By contrast, if the SAAG is less than 1.1 g/dL (11 g/L), the patient is unlikely to have portal hypertension. The SAAG does not confirm the diagnosis of the cause of ascites but is an indirect and accurate index of portal hypertension. The SAAG may be useful when the cause of ascites is not clear after the initial medical history, physical examination, standard blood tests, and abdominal US.

The assessment of other tests should be performed based on clinical presentation or the need to exclude causes of ascites other than cirrhosis. ^{, ,} If ascitic fluid is infected and secondary bacterial peritonitis rather than SBP is suspected, the measurement of glucose, amylase, lipase, and LDH in ascitic fluid may be useful. Glucose is a small molecule that can diffuse easily into extravascular fluid. Therefore, ascitic fluid glucose levels are usually similar to those in plasma, unless glucose is being consumed by leukocytes or bacteria. In the setting of secondary bacterial peritonitis, glucose levels are markedly low and close to 0 mg/dL due to significantly increased numbers of leukocytes and bacteria in ascites. In addition, ascitic LDH levels increase markedly due to their release from neutrophils and are typically several-fold higher than serum levels. ^, Finally, assessment of a Gram stain of ascitic fluid may be useful in the setting of secondary peritonitis by showing polymicrobial ascites, which is typical of this condition.

Levels of pancreatic enzymes or mycobacterial culture should only be performed when pancreatic disease or tuberculosis, respectively, is suspected. Finally, ascitic fluid cytology should be performed if malignancy is suspected as the cause of ascites; cytology is likely to be positive in patients with peritoneal tumor but not in those with tumor limited to the liver.