Liver disease

Diazo methods.

The most widely used chemical methods for bilirubin measurement are those based on the coupling of bilirubin with a diazo compound. In this reaction ( Fig. 51.9 ), diazotized sulfanilic acid (the diazo reagent) reacts with bilirubin to produce two azodipyrroles (azopigments), which are reddish-purple at neutral pH and blue at low or high pH values. Van den Bergh and Muller applied this reaction to the quantitation of bilirubin in serum. They described the fraction of bilirubin that reacted with the diazo reagent in the absence of alcohol as the direct bilirubin fraction and used the term indirect bilirubin for the difference between total bilirubin (found after the addition of alcohol to the reaction mixture) and the direct bilirubin fraction. Numerous variations of the van den Bergh and Muller method have been developed. All use one of a variety of “accelerators,” which, like alcohol, facilitate the reaction of unconjugated (indirect) bilirubin with the diazo reagent; the most commonly used accelerators are caffeine, dyphylline, and several surface active agents. The diazo method of Malloy and Evelyn, which uses methanol as an accelerator, has substantial matrix effects, negative interference by Hb, turbidity due to protein precipitation by methanol, and a long reaction time. This method, which has been virtually abandoned, is mentioned here for historical reasons only.

The diazo method described by Jendrassik and Grof in 1938 and later modified by Doumas and colleagues gives results for serum total bilirubin that are reproducible and reliable. In this procedure, an aqueous solution of caffeine and sodium benzoate serve as the accelerators. Studies on the mechanism by which the caffeine-benzoate solution facilitates the reaction of unconjugated bilirubin with the diazo reagent have provided strong, albeit indirect, evidence that caffeine, and perhaps benzoate, displaces unconjugated bilirubin from its association sites on albumin. This occurs by (1) formation of hydrogen bonds between bilirubin and caffeine, ^, thus making bilirubin water soluble, or (2) complex formation and disruption of the bilirubin internal hydrogen bonds. With the use of samples prepared by addition of unconjugated bilirubin and authentic human diconjugated (with glucuronic acid) bilirubin to low-bilirubin pooled sera—and a nuclear magnetic resonance technique—Lo and Wu have shown that the modified Jendrassik-Grof total bilirubin assay detects unconjugated and diconjugated bilirubin quantitatively (as unconjugated bilirubin equivalents). This method has acceptable transferability among laboratories ^{, ,} and is currently the method of choice.

Other methods for determining bilirubin include direct spectrophotometric measurement of total bilirubin in serum using analysis of a two-component system by measuring absorbance at two wavelengths and solving a system of two simultaneous equations. This approach is applicable to sera from healthy neonates because only unconjugated bilirubin is present in such sera. Correction for oxyhemoglobin is necessary because it is invariably present in sera from neonates.

Calibrators for bilirubin measurements.

A number of instrument manufacturers use bovine serum, instead of human serum, as the protein base for preparing fluids for calibrating methods for total and direct bilirubin; the protein base is enriched with unconjugated bilirubin or ditaurobilirubin or both. Unconjugated bilirubin in human serum reacts completely with the reference method and with diazo methods available in commonly used clinical analyzers; however, its reaction in bovine serum from commercial sources is incomplete and unpredictable. That makes the assignment of accurate bilirubin values to calibrators virtually impossible, the protein base of which is commercial bovine serum. In human serum, ditaurobilirubin was underestimated by two of seven clinical analyzers tested; the calibrators of these two analyzers were made in bovine serum. Ditaurobilirubin in commercial bovine serum was underestimated by all analyzers and by the reference method; in human serum, it was underestimated by two analyzers only. The practice of using bilirubin calibrators in bovine sera should be abandoned because it compromises the accuracy of bilirubin measurements in jaundiced neonates. However, fresh bovine serum (obtained from a slaughterhouse) has only a small effect on the measurement of unconjugated bilirubin or ditaurobilirubin.

High-performance liquid chromatography.

High-performance liquid chromatography (HPLC) methods have been developed for relatively rapid separation and quantification of the four bilirubin fractions. HPLC has been helpful in separating and detecting the various bilirubin photoisomers produced during phototherapy in newborns and thus in elucidating the mechanism by which phototherapy lowers the concentration of bilirubin in newborn blood. ^, Several HPLC methods are available for analysis of bilirubin fractions. In the method of Blanckaert, bilirubin conjugates, but not unconjugated bilirubin, are converted to the corresponding bilirubin methyl esters by base-catalyzed transesterification in methanol followed by extraction with chloroform. With this procedure, the α-, β-, and γ-bilirubin fractions are recoverable, but the δ-fraction (δ-bilirubin) remains in the denatured protein pellet that is produced by the chloroform extraction. In the HPLC method of Lauff and coworkers, all four bilirubin fractions remain in solution after a step that involves salting out globulins with sodium sulfate. Both methods require the use of dim incandescent or yellow light to minimize photodegradation of the various bilirubin species. A simple and fast HPLC method has been published by Adachi and associates ; this method uses a Micronex RP-30 column (Sekisui Chemical Co., Mount Laurel, NJ), which does not require salting out of globulins or chemical transformation of bilirubin conjugates. This method separates serum bilirubin into five fractions; the fifth fraction eluted between the monoglucuronide and the unconjugated bilirubin is the Z,E or the E,Z photoisomer. The elution sequence is the same as in the procedure of Lauff and colleagues. Osawa and associates have successfully developed an isocratic mobile phase. The elution buffer includes 0.8% sodium ascorbate to maintain the stability of the bilirubin species. Isolation was improved with the addition of 1% Brij 35. This method strongly correlates with the HPLC method by Adachi and colleagues. Using the method by Osawa and colleagues, molar absorptivities for unconjugated bilirubin, bilirubin monoglucuronide, bilirubin diglucuronide, and δ-bilirubin were calculated at 450 nm, giving this method the potential for evaluating the accuracy of bilirubin assays.

Additional studies have indicated that the δ-bilirubin fraction consists of one or more bilirubin species that are covalently bound to albumin. Existence of covalent linkage is supported by the fact that the associated bilirubin species are not released from the albumin fraction by treatment with strong acid or base, or a variety of strong denaturing agents, by hydrolysis with proteolytic enzymes, or by boiling in methanol. Delta-bilirubin reacts directly (without a promoter) with diazotized sulfanilic acid. The discovery of δ-bilirubin has solved the mystery of persistent high bilirubin concentrations that mostly direct react in patients with intrahepatic or obstructing jaundice long after hepatitis has subsided or obstruction has been relieved. It is the slowest fraction to clear from serum because it follows the catabolism of albumin, which has a half-life of approximately 17 to 19 days.

HPLC has been helpful in elucidating the nature of the bilirubin species that occur naturally in blood or are formed during phototherapy. Clinically, it offers little, if any, aid to the physician in the differential diagnosis of jaundice, because knowing the percentage of each of the bilirubin fractions in blood is of no diagnostic value. It cannot be considered as a reference method for measuring total bilirubin in blood because its accuracy and precision are inadequate. The method is calibrated with unconjugated bilirubin with the untested assumption that the other three bilirubin fractions have molar absorptivities identical to that of the calibrator, when in fact this is not known. Furthermore, errors in measurement of the four species may be cumulative and may result in a large total error; also, the method is insensitive at total bilirubin concentrations of less than 1 mg/dL (17 μmol/L) and is too laborious for routine clinical analysis. Some of the δ-bilirubin may be lost during pretreatment of samples.

A capillary electrophoresis method for measuring the different types of bilirubin has been developed by Wu and his associates.

Enzymatic methods.

Enzymatic methods for total and direct bilirubin and for bilirubin conjugates with glucuronic acid are based on the oxidation of bilirubin with bilirubin oxidase to biliverdin with molecular oxygen. At a pH near 8, and in the presence of sodium cholate and sodium dodecylsulfate, all four bilirubin fractions are oxidized to biliverdin, which is further oxidized to purple and finally colorless products. The decrease in absorbance at 425 or 460 nm is proportional to the concentration of total bilirubin. Results obtained by the bilirubin oxidase method were in good agreement with those obtained by the Jendrassik-Grof procedure. Direct bilirubin is measured at pH 3.7 to 4.5; at this pH range, the enzyme oxidizes bilirubin conjugates and δ-bilirubin, but not unconjugated bilirubin. ^, At pH 10, the enzyme selectively oxidizes the two glucuronides. ^, Delta-bilirubin is not oxidized at all, and only 5% of unconjugated bilirubin is measured as conjugates.

Transcutaneous measurement of bilirubin.

A noninvasive approach for measuring bilirubin was introduced in 1980 by Yamanouchi and colleagues. The first bilirubinometer (icterometer) was a reflectance photometer, which used two filters to correct for the color of Hb and required measurements at eight body sites. Efforts to improve the accuracy of such measurements have been successful and led to the development of devices of acceptable performance. Reports indicate that at least one of these devices (Bili Check SpectR _x Inc., Norcross, GA) provides results that are within ±2 mg/dL (34 μmol/L) of those obtained using a serum diazo procedure. ^, Another study found that the Bili Check underestimated serum bilirubin when its concentration was greater than 10 mg/dL (170 μmol/L).

Although transcutaneous bilirubin measurements may not substitute for laboratory quantitative determinations, they provide instantaneous information, reduce the necessity for serum bilirubin determinations, spare infants the trauma of heelsticks, and save money. Furthermore, they are useful in determining whether it is necessary to draw blood in a jaundiced infant before initiating treatment, such as phototherapy or exchange transfusion (currently, this is extremely rare). Another application is predicting those babies who require follow-up according to the “hour-specific” serum bilirubin nomogram developed by Bhutani and coworkers.

Urine bilirubin.

Because only conjugated bilirubin is excreted in urine, its presence indicates conjugated hyperbilirubinemia. The most commonly used method for detecting bilirubin in urine involves the use of a dipstick impregnated with a diazo reagent. Dipstick methods are capable of detecting bilirubin concentrations as low as 0.5 mg/dL (9 μmol/L).

A fresh urine specimen is required because bilirubin is unstable when exposed to light and room temperature, and it may be oxidized to biliverdin (which is diazo negative) at the normally acidic pH of the urine. If the test is delayed, the sample must be protected from light and stored at 2 to 8 °C for no longer than 24 hours. The reagent strip (Chemstrip, Roche Diagnostics, Indianapolis, IL; Multistix, Siemens Healthcare Diagnostics, Deerfield, IL) is immersed in the urine specimen for no longer than 1 second and is read 60 seconds later. During this time, bilirubin reacts with a diazo reagent, yielding a pink to red-violet color, the intensity of which is proportional to the bilirubin concentration. The reaction mechanism for urinary conjugated bilirubin is the same as that described in Fig. 51.9 , except that 2,6-dichlorobenzene-diazonium-tetrafluoroborate is substituted for diazotized sulfanilic acid in the Chemstrip, and 2,4-dichloroaniline diazonium salt in the Multistix. Another commonly used test, more sensitive than the Multistix, is the Ictotest reagent tablet (Siemens Healthcare Diagnostics); in this semiquantitative procedure, the diazo reagent is p- nitrobenzenediazonium- p- toluenesulfonate.

Chemstrip and Multistix strips for bilirubin in urine are highly specific tests and have a low incidence of false-positive results. However, medications that color the urine red or that give a red color in an acid medium, such as phenazopyridine, can produce a false-positive reading. Large quantities of ascorbic acid or of nitrite also worsen the detection limit of the test. In practice, bilirubin is rarely measured in urine.

Chemistry.

Four major bile acids are known. Cholic acid and chenodeoxycholic acid, the primary bile acids, are synthesized in the liver. The sequence of reactions involved in the synthesis of cholic acid from cholesterol is shown in Fig. 51.10 . To date, nine inborn errors of bile acid synthesis have been identified; these can present with neonatal hepatitis, fat malabsorption, or neurologic defects that can progress to chronic liver disease or liver failure and death. The primary bile acids are metabolized (by bacterial 7α-dehydroxylase) in the intestinal lumen to the secondary bile acids—deoxycholic acid and lithocholic acid. Bile acids (through their carboxylate groups) are conjugated in the liver with the amino acid glycine or taurine. This decreases passive absorption in the biliary tree and proximal small intestine but permits conservation through active transport in the terminal ileum. Approximately 0.1 to 0.6 g of bile acids is lost in the feces daily.

Because they possess both polar and nonpolar regions, molecules of bile acids are able to solubilize biliary lipids. Such molecules align at water–lipid interfaces and reduce surface tension, acting as detergents. In an aqueous solution, bile acids aggregate to form small polymolecular aggregates approximately 5 nm in diameter called micelles, which are capable of incorporating cholesterol and phospholipids to form mixed micelles. Micellar solubilization of these water-insoluble constituents maintains cholesterol in solution. In the intestinal lumen, dietary cholesterol and the products of triglyceride digestion (predominantly free fatty acids and monoglycerides) are incorporated into mixed micelles. Micelles deliver lipolytic products to the mucosal surface. To carry out these functions, a critical micellar bile acid concentration of approximately 2 mmol/L is necessary. Bile acids are thus important for ensuring the solubility of cholesterol (a major component of most gallstones) in bile and dietary lipids (including fat-soluble vitamins) in the intestinal lumen.

Clinical significance of bile acids.

In view of the multiple processes involved in bile acid synthesis, conjugation, and excretion, and in its hepatic and intestinal uptake, several potential sites for primary or secondary disturbances have been identified ( Box 51.1 ). With hepatocyte dysfunction (which occurs in many liver disorders), decreased bile acid synthesis results in low primary bile acid concentrations and a decreased ratio of primary to secondary bile acids in plasma; in addition, decreased extraction from plasma often leads to increased concentrations of bile acids, particularly in the nonfasting state. With cholestatic disorders, decreased delivery of primary bile acids to the intestine with resulting decreased secondary bile acid production causes an increased ratio of primary to secondary bile acids, as well as increased total bile acid concentrations. With intestinal disease (including bypass operations that may be performed to treat obesity), increased fecal loss of bile acids leads to decreased concentrations of both primary and secondary bile acids and often a decrease in plasma cholesterol concentration caused by an increased need for bile acid synthesis. Although plasma bile acid concentrations are abnormal in many situations, their measurement adds little to standard tests of liver function, and they are rarely used in clinical medicine except in the investigation of unexplained pruritus. This is because of the wide biological variation due to changes in prandial state and due to diurnal rhythm which is independent of food intake.

The value of measuring total bile acid in serum or plasma for diagnosing intrahepatic cholestasis in pregnancy has been recently highlighted and endorsed by a number of scientific societies. In particular, the guidelines of the UK Royal College of Obstetricians and Gynaecologists state that abnormal values of aminotransferases, GGT, and total bile acids are sufficient to support the diagnosis of obstetric cholestasis, and also provide valuable therapeutic information for preventing fetal complications attributable to the toxic effects of bile salts. The reference range of total bile acids in the serum of pregnant women has been recently established at 0.3 to 10 μmol/L, whereas serum or plasma values greater than 40 μmol/L are diagnostic of severe obstetric cholestasis and were found to be strongly associated with impaired fetal outcome.

Analytical methods.

Analytical techniques used to quantify total or individual bile acids in biological fluids include gas-liquid chromatography, HPLC, enzymatic assay, radioimmunoassay, enzyme-linked immunosorbent assay (ELISA), and high-resolution tandem mass spectrometry.

Plasma proteins

Coagulation proteins.

The coagulation proteins that are synthesized in the liver are listed in Table 51.1 . These proteins interact to produce a fibrin clot (see Chapter 79 ). Inhibitors of the coagulation system, including antithrombin, protein C, and protein S, are also synthesized in the liver. Some of the coagulation factors (II, VII, IX, and X) require vitamin K for post-translational carboxylation within the hepatocyte. Proteins C and S are also carboxylated by a vitamin K–dependent enzyme. Activated protein C in plasma inhibits coagulation by inactivating factors V and VIII. Parenchymal liver disease of sufficient severity to impair protein synthesis or obstructive liver disease sufficient to impair intestinal absorption of vitamin K is, therefore a potential cause of bleeding disorders. Because of the great functional reserve of the liver, failure of hemostasis usually does not occur except in severe or long-standing liver disease.

PT depends on the activity of fibrinogen (factor I), prothrombin (factor II), and factors V, VII, and X. Because all of these factors are made in the liver, and several are vitamin K–dependent, a prolonged PT often indicates the presence of significant liver disease. In cholestasis, vitamin K deficiency may cause an increase in PT. In this case, the coagulation abnormality is corrected in 1 to 2 days by parenteral injection of 10 mg of vitamin K. In contrast, if PT is prolonged because of hepatocellular disease, factor synthesis is decreased, and administration of vitamin K does not typically correct the problem. PT is also prolonged in some patients with liver disease because of the presence of dysfibrinogenemia, an abnormal form of fibrinogen that does not clot normally, which may predispose patients to thrombosis.

The method for reporting PT in liver disease remains controversial. PT measures the time for plasma to clot after exposure to tissue factor. Reagents differ in the amount of tissue factor present; in patients on warfarin, clotting times are more greatly prolonged when lower amounts of tissue factor and other reagents that stimulate clotting are in the reagents. This makes a reagent more sensitive to clotting factor abnormalities but makes standardization of results among laboratories difficult. The international normalized ratio (INR) was developed by the World Health Organization (WHO) and the International Committee on Thrombosis and Hemostasis for reporting the results of blood coagulation (clotting) tests. All results are standardized using the international sensitivity index (ISI) for the particular thromboplastin reagent and instrument combination used to perform the test. In practice, it requires determination of the ISI based on the slope of the relationship between PT using the reagent and that using a reference method in patients on warfarin. The INR is then calculated as follows:

INR has been found to standardize interpretation of PT measurements among laboratories for those taking warfarin. Unfortunately, INR does not have the same relationship with impairment of clotting in individuals with liver disease. The liver is responsible for the synthesis of nearly all clotting factors and their inhibitors. As a result, patients with chronic liver disease and cirrhosis experience a rebalancing of their hemostatic variables. Patients with acute liver failure (ALF) likely experience minimal effects on their in vivo coagulation profiles as assessed with thromboelastography (TEG) despite mean INR values greater than 3. Furthermore, these patients have significant rates of hypercoagulable (35%) and hypocoagulable (20%) states. To further complicate matters, the presence of a hypercoagulable state does not exclude the presence of a tendency toward increased bleeding risk, and conversely, increased bleeding risk does not rule out the development of a new thrombus. The apparent explanation lies in the mechanism of clotting factor deficiency in liver disease and warfarin administration. Although liver disease inhibits synthesis of clotting factors, warfarin impairs vitamin K–dependent carboxylation, which impairs the ability of the factors to bind calcium. These noncarboxylated clotting factors (termed proteins induced by vitamin K absence [PIVKAs]) appear to act as inhibitors of coagulation; thus when lower amounts of tissue factor are present, clotting times are more prolonged. In contrast, in liver disease, factor deficiency is due to impaired factor synthesis, and no PIVKAs are present (except in HCC, as discussed later in this chapter). This leads to lesser increases in PT in individuals with liver disease and an underestimation of the degree of clotting impairment when reagents with a low ISI are used. Studies have shown that calculation of a different ISI using plasma samples from patients with liver disease can standardize PT results with different reagents, ^, but to date, such liver ISI information is not readily available to laboratories.

Lipid and lipoprotein synthesis.

The liver plays a key role in the metabolism of lipids and lipoproteins (see Chapter 36 ). On a daily basis, approximately 33% of the fatty acids originating from adipose tissue enter the liver, where they undergo esterification into triglycerides or are oxidized. Oxidation is favored in the fasting state and esterification is favored in the nonfasting state. Excessive esterification results in fatty liver, a disorder in which excess triglycerides are deposited in large vacuoles that displace other cellular components. Most cholesterol is synthesized endogenously, in the liver. The endogenously synthesized cholesterol and that of dietary origin enter the hepatic pool, where they are converted to bile acids, incorporated into lipoproteins, or used in the synthesis of liver cell membranes. The relative rates of secretion of bile acids, cholesterol, and lecithin are important factors in the pathogenesis of cholesterol gallstones.

Urea synthesis.

Patients with end-stage liver disease may have low concentrations of urea in plasma (see Chapters 34 and 49 ). The rate of urea excretion in urine is lower in these patients than that in healthy individuals. In addition, plasma concentrations of urea precursors—ammonia and amino acids—are increased. In nonalcoholic steatohepatitis, the function of urea cycle enzymes may be affected, resulting in hyperammonemia and the risk of disease progression. In nonalcoholic steatohepatitis (NASH) animals, gene and protein expression of ornithine transcarbamylase (OTC) and carbamoylphosphate synthetase were reversibly reduced. Hypermethylation of urea cycle enzymes is a potential underlying mechanism. The functional changes of urea synthesis in NASH are associated with hyperammonemia. Hyperammonemia can cause progression of liver injury and fibrosis.

Hypermethylation of Otc gene promoters has been observed. Additionally, in animal models of NASH and in patients with NAFLD, OTC concentration and activity were reduced and ammonia concentrations were increased, which was further exacerbated in those with NASH. In primary hepatocytes, induction of steatosis was associated with Otc promoter hypermethylation, a reduction in the gene expression of Otc and Cps1 , and an increase in ammonia concentration. These findings suggest that patients with liver disease have an impaired ability to metabolize protein nitrogen and to synthesize urea. The rate of hepatic urea synthesis also depends on exogenous intake of nitrogen and on endogenous protein catabolism.

Biochemistry and physiology.

Ammonia is a by-product of nitrogen metabolism, and its formation in the body is predominantly a consequence of the action of the enzyme glutaminase, located within enterocytes of the small intestine and colon, as well as the action of the vast number of urease-producing bacteria located in the gut. Plasma ammonia concentration in the hepatic portal vein is typically fivefold to tenfold higher than that in the systemic circulation. Under normal circumstances, most of the portal vein ammonia load is metabolized to urea in hepatocytes through the Krebs-Henseleit (urea) cycle during the first pass through the liver; this process includes intramitochondrial and cytosolic enzyme-catalyzed steps ( Fig. 51.11 ).

Ammonia enters the tissue of the central nervous system by passive diffusion. The rate of entry increases in proportion to the plasma concentration and is dependent on pH. Ammonia crosses the blood–brain barrier more readily than the ammonium ion. As pH increases, the rate of entry of ammonia into the central nervous system tissue increases as the result of an increase in ammonia relative to ammonium. Because the acid dissociation constant (pK _a ) of ammonia is 9.1 at 37 °C, approximately 3% of total blood ammonia is ammonia at the normal physiological pH of 7.4. An increase in pH to 7.6 produces an increase in ammonia to approximately 5% of total blood ammonia, which is a 67% increase in concentration.

Clinical significance.

Animal and human studies have shown that an increased concentration of ammonia (hyperammonemia) exerts toxic effects on the central nervous system. Several causes, both inherited and acquired, of hyperammonemia are known. Inherited deficiencies of urea cycle enzymes are the major cause of hyperammonemia in infants. The two major inherited disorders are those that involve the metabolism of the dibasic amino acids lysine and ornithine and those that involve the metabolism of organic acids, such as propionic acid, methylmalonic acid, isovaleric acid, and others (see Chapter 60 ). Insult to the liver, whether acute or chronic in nature, reduce its capacity to metabolize ammonia and this creates an ammonia burden on extrahepatic tissues which can result in hyperammonemia up to five times that of normal blood ammonia concentrations. The occurrence of hyperammonemia is not specific to liver dysfunction and can also be observed in various other disease states including, but not limited to, inborn errors of the urea cycle, Reye syndrome, and valproate poisoning.

The main acquired causes of hyperammonemia are advanced liver disease and renal failure. Severe or chronic liver failure (which occurs in fulminant hepatitis or cirrhosis, respectively) leads to significant impairment of normal ammonia metabolism. Reye syndrome, which is primarily a central nervous system disorder with minor hepatic dysfunction, is also associated with hyperammonemia. Hepatic encephalopathy in the cirrhotic patient is often precipitated by GI bleeding, which enhances ammonia production through bacterial metabolism of protein found in blood. Other precipitating causes of encephalopathy include excess dietary protein, constipation, infection, drugs (particularly central nervous system depressants and those that alter blood biochemistry such as diuretics), and electrolyte and acid-base imbalance (alkalosis). Because cirrhosis is accompanied by portosystemic shunting, ammonia clearance is impaired, leading to increased concentrations of blood ammonia. Impaired renal function also causes hyperammonemia. As blood urea concentration increases, more diffuses into the GI tract, where it is converted to ammonia.

The fasting venous plasma ammonia concentration is useful in the differential diagnosis of encephalopathy, when it is unclear whether encephalopathy is of hepatic origin. It is especially helpful in diagnosing Reye syndrome and the inherited disorders of urea metabolism, as well as increased ammonia concentrations due to drugs such as salicylates or valproate. In acute liver injury, ammonia concentrations more than 200 μmol/L (340 μg/dL) are associated with cerebral edema and a poor prognosis, and it has been suggested that ammonia concentrations should be used as part of the evaluation of prognosis in ALF. However, plasma ammonia is not useful in patients with known chronic liver disease. Although ammonia concentrations are higher as the degree of encephalopathy worsens, significant overlap between concentrations is seen in different stages of encephalopathy, and approximately 70% of those with cirrhosis without encephalopathy have increased ammonia concentrations. Ammonia concentrations may actually better reflect the presence of shunting blood around the portal veins than the degree of liver dysfunction. There is growing recognition of the complex and synergistic relationship between ammonia, inflammation (sterile and nonsterile), and oxidative stress in the pathogenesis of hepatic encephalopathy which develops in patients with liver dysfunction.

Preanalytic issues.

The concentrations of analytes may be affected by numerous factors that should be explored in order to establish ideal sampling conditions, processes, handling, and storage. These include diurnal variation, effects of feeding and fasting, donor position, sample containers, preservatives, sample handling, and sample processing and storage, including the influence of storage time, temperature, and freeze-thawing (see Chapter 5 ).

Analytical methodology.

Both enzymatic and chemical methods are used to measure ammonia in body fluids. An enzymatic assay with glutamate dehydrogenase is the most frequently used method. Plasma ammonia measurement is particularly susceptible to contamination, leading to falsely increased concentrations. Common preanalytical problems are discussed in Chapter 4 on Specimen Collection and Processing.

Reference intervals.

For the enzymatic method, the reference interval is 15 to 45 μg/dL (11 to 32 μmol/L). For more details on age-dependent values, see the Appendix on Reference Intervals. Laboratories should verify that these ranges are appropriate for use in their own settings.

Dye excretion tests.

Dye excretion tests, such as BSP and indocyanine green clearance, were formerly used as indicators of liver disease. With the development of more sensitive and specific indicators of liver disease, dye excretion tests have become obsolete, although until the 1970s, BSP was the most frequently used dye excretion test. Because of reports of fatalities resulting from hypersensitivity and other adverse effects, BSP use has been discontinued. Indocyanine green clearance is still occasionally used for investigating hepatic blood flow and for predicting clearance rates of drugs that undergo first-pass clearance by the liver, such as lignocaine. Typical indocyanine green clearance values in healthy subjects range from 6.5 to 14 mL/min/kg body weight.

Drug clearance tests.

A variety of drugs that are metabolized by the liver have been used to study the action of various P ₄₅₀ (mixed-function oxidase) enzymes. Aminopyrine is demethylated to form carbon dioxide and aminoantipyrine. With the use of ¹³ C- or ¹⁴ C-labeled aminopyrine, the resulting isotopically labeled carbon dioxide is measured in breath as a reflection of functioning liver mass. Decreases in metabolism are common in persons with cirrhosis, but metabolism is also affected by other factors such as cigarette smoking and use of drugs such as oral contraceptives; significant intraindividual variation in results has been noted. Overall diagnostic sensitivity is similar to that of other more routine laboratory tests.

Caffeine clearance is altered during hepatic injury; it is prolonged in both chronic hepatitis and cirrhosis. Caffeine is rapidly and nearly completely absorbed from the GI tract and then undergoes N-demethylation by the hepatic mixed-function oxidase system. A single dose of caffeine (3.5 mg/kg to a maximum dose of 200 mg, dissolved in water, fruit juice, or milk for oral administration) is administered. This caffeine dose is equivalent to that found in one cup of brewed coffee or in one can of a commercial soft drink.

Blood (or salivary samples) obtained before and at timed intervals after caffeine ingestion can be analyzed by reversed-phase HPLC or immunoassay. A close correlation is found between plasma and salivary caffeine concentrations. Caffeine half-life is approximately 5.5 hours in healthy adults and 3 hours in healthy children, with clearance of approximately 2 mL/min/kg in healthy adults and 10 mL/min/kg in healthy children. Caffeine clearance correlates with the aminopyrine breath test and has similar limitations, although it is less subject to effects of variables, such as smoking and oral contraceptive use. Lidocaine undergoes N-deethylation in the liver by cytochrome P ₄₅₀ to form monoethylglycinexylidide (MEGX); the rate of appearance of MEGX in plasma reflects hepatic lidocaine clearance. Because lidocaine is highly extracted, its clearance is flow dependent. Thus alterations in hepatic blood flow also influence lidocaine elimination. Lidocaine (1 mg/kg) is given by intravenous bolus; plasma is obtained at baseline and at 15 minutes for MEGX concentration (time of plateau concentration in healthy individuals). MEGX is most commonly measured using an immunoassay. Lidocaine clearance has been used to assess liver transplantation function, but its use is limited by the effect of hypoperfusion (which occurs in sepsis or volume depletion).

Hepatic stellate cells.

HSCs reside in the space of Disse, interposed between the endothelium and hepatocytes, where they encircle the liver sinusoids. After liver injury, HSCs become activated by the products of apoptotic mesenchymal cells, which leads to the conversion of a resting vitamin A–rich cell (a quiescent HSC) to one that has lost vitamin A droplets by autophagy, which leads to increased proliferation and contraction, and the release of proinflammatory, profibrogenic, and promitogenic cytokines. The activated HSCs become contractile myofibroblasts that generate a scar that forms around the injury site.

HSC activation can be divided into two phases: initiation and perpetuation. Initiation, which is also known as the preinflammatory stage, refers to early changes in gene expression and phenotype. It is the result of primarily paracrine stimulation from damaged parenchymal cells. Maintenance of these stimuli leads to a perpetuation phase that is regulated by autocrine and paracrine stimuli. Perpetuation involves at least six distinct changes in HSC behavior, including proliferation, chemotaxis, fibrogenesis, contractility, matrix degradation, and retinoid loss.

Myofibroblasts.

The profibrotic myofibroblasts are the master regulators of the fibrotic response because of their scar-producing, proliferative, migratory, contractile, immunomodulatory, and phagocytic properties. Myofibroblasts are the prototypical mesenchymal cell type that regulates repair after an injury in a range of tissues, including liver, kidneys, skin, lungs, and bone marrow, as well as the central nervous system. Myofibroblasts, once activated, are capable of enhanced migration and deposition of ECM components. Although HSCs are the primary source of this fibrogenic population in the liver, other cells such as bone marrow–derived cells, portal fibroblasts, and epithelial-to-mesenchymal transition from hepatocytes and cholangiocytes also contribute to fibrogenesis, although their exact role in disease is not completely understood.

Role of the extracellular matrix.

In the normal liver, the ECM provides structural and biochemical support to the surrounding cells and is composed mainly of a number of structural proteins (including collagens IV and VI), as well as a range of growth factors and matrix metalloproteinases (MMPs) that are specifically bound and preserved in latent forms. The ECM can modulate the activation and proliferation of HSCs, angiogenesis, and the availability and activity of growth factors and MMPs. The ECM also provides cells with signals for polarization, adhesion, migration, proliferation, survival, and differentiation. ECM–cell interactions are determined largely by specific membrane adhesion receptors. The ECM may prevent apoptosis in the damaged liver and also prevent growth factor proteolysis. Interactions between ECM and its surrounding cells are bidirectional. After injury, the fibrillary collagens I and III predominate together with fibronectin. Liver fibrosis as a consequence of liver injury entails both qualitative and quantitative changes in ECM composition as a result of an imbalance between the rates of matrix synthesis and degradation. The ECM becomes progressively insoluble and resistant to protease digestion because of the thickening of fibrotic septae and increased cross-linking. ^,

Matrix metalloproteinases.

MMPs, also known as matrixins, are the major family of calcium-dependent enzymes that degrade collagenous and noncollagenous ECM substrates. There are 25 members of this tightly regulated family, which are classified on the basis of their substrate specificity: interstitial collagenases, gelatinases, stromelysins, membrane types, and metalloelastases. MMPs are secreted as inactive proenzymes, have complex transcriptional control, and their action is inhibited by a family of endogenous proteinase inhibitors known as tissue inhibitors of metalloproteinases (TIMPs). Four TIMP members bind reversibly to the active site of all MMPs and have different affinities for specific MMPs. Thus TIMPs play an important role in preventing degradation of the accumulating matrix during liver injury by antagonizing the activity of MMPs. TIMP-1 also has an antiapoptotic effect on HSCs; it prevents clearance of activated HSCs during injury and promotes their survival through induction of B-cell lymphoma. HSCs are a key source of MMPs, especially MMP-2, -3, -9, and -13. In chronic human liver disease and animal models of fibrosis, concentrations of MMP-1 and/or -13 do not change, but there is a progressive increase in TIMP-1 and -2 as fibrosis advances. TIMP expression can be detected soon (6 hours) after liver injury and may precede the induction of procollagen I.

Cytokines.

Fibrosis usually follows an inflammatory insult; therefore certain cytokines secreted by a range of cells, including Kupffer cells, HSCs, hepatocytes, natural killer cells, lymphocytes, and dendritic cells play a key role in the response. These include the chemokines (monocyte chemotactic protein-1, RANTES [Regulated on Activation, Normal T Expressed and Secreted; also known as CCL5 or C-C motif chemokine ligand 5], IL-8), interferons (IFN-α, IFN-γ), ILs (IL-1, IL-6, IL-10), growth factors, adipokines, and soluble neurohumoral ligands (endocannabinoids). Adipokines (adipose tissue cytokines) are polypeptides secreted mainly by adipocytes, and to a lesser extent, by stromal cells, including macrophages, fibroblasts, and infiltrating monocytes. Leptin and adiponectin are the main adipokines implicated in liver injury. ^,

Hepatocytes.

The hepatocytes are the major cell type in the liver and are also involved in the process of fibrosis and cirrhosis. Normally, hepatocytes can regenerate removed liver tissue rapidly, but in chronic disease states they appear to become senescent with respect to this function; however, the HSCs become activated and are sufficient to regenerate the biliary and hepatocellular epithelium.