Serum Enzyme Activities

Liver-specific serum enzyme activities are included routinely in screening serum biochemistry panels and are regarded as markers of hepatocellular and biliary injury and reactivity. Because marked hepatic disease can be present in patients with normal serum enzyme activity, finding normal values should not preclude further investigation, especially if there are clinical signs or other laboratory test results that suggest hepatobiliary disease. Increased serum activity of enzymes normally located in hepatocyte cytosol in high concentration reflects structural or functional cell membrane injury that would allow these enzymes to escape or leak into the blood. The two enzymes found to be of most diagnostic use in cats and dogs are alanine transaminase (ALT; glutamic-pyruvic transaminase [GPT]) and aspartate transaminase (AST; glutamic-oxaloacetic transaminase [GOT]). Because ALT is found principally in hepatocytes and AST (also located within hepatocyte mitochondria) has a wider tissue distribution (e.g., in muscle), ALT is the enzyme selected to most accurately reflect hepatocellular injury. Less is known about the behavior of AST in various hepatobiliary diseases in companion animals, although some studies have indicated that AST is a more reliable indicator of liver injury in cats. Several studies have demonstrated mild to moderately high serum ALT activity (without histologic or biochemical evidence of liver injury), in addition to expected high serum activities of muscle-specific creatine kinase and AST, in dogs with skeletal muscle necrosis.

In general, the magnitude of serum ALT and AST activity elevation approximates the extent, but not the reversibility, of hepatocellular injury. Rather than clinical relevance being assigned to absolute values for ALT or AST activity (e.g., serum ALT activity of 200 IU/L is worse than 100 IU/L), the values should be assessed in terms of number of fold elevations from normal. Twofold to threefold elevations in serum ALT activity are associated with mild hepatocellular lesions, fivefold to tenfold elevations are seen with moderately severe lesions, and greater than tenfold increases suggest marked hepatocellular injury. ALT (and to a lesser extent AST) activity is also often increased by glucocorticoids in dogs, although to a lesser extent than ALP.

Serum enzyme activities that reflect new synthesis and release of enzyme from the biliary tract in response to certain stimuli are alkaline phosphatase (AP) and γ-glutamyltransferase (GGT). Bile retention (i.e., cholestasis) is the strongest stimulus for accelerated production of these enzymes. Unlike ALT and AST, AP and GGT are in low concentration in the cytoplasm of hepatocytes and biliary epithelium and are membrane-associated, so the fact that they simply leak out of damaged cells does not account for increased serum activity. Measurable AP activity is also detectable in nonhepatobiliary tissues of cats and dogs (including osteoblasts, intestinal mucosa, renal cortex, and placenta), but serum activity in healthy adult cats and dogs arises only from the liver, with some contribution by the bone isoenzyme in young, rapidly growing dogs and in kittens less than 15 weeks old. The renal form is mainly measurable in the urine, and the gut form has a very short half-life so is not usually measurable (although the steroid-induced isoenzyme of AP in dogs is believed to be an altered gut isoenzyme with a prolonged half-life). The half-life of feline AP is shorter than that of canine AP; thus serum activity is relatively lower in cats than in dogs with a similar degree of cholestasis, and, conversely, even mild elevations of AP in cats are clinically very significant. Markedly high serum AP activity of bone origin (mean total serum AP values more than fivefold higher than those in nonaffected individuals, with only the bone isoenzyme detected) was identified in certain healthy juvenile (7 months old) members of a family of Siberian Huskies (Lawler et al., 1996). This change is believed to be benign and familial and should be considered when results of serum AP activity are interpreted in this breed. A young, growing dog of any breed can have a mild increase in serum AP. Increased serum AP activity of unknown origin has also been described in adult Scottish Terriers and is believed to be benign and possibly familial (Gallagher et al., 2006).

Certain drugs, the most common of which are anticonvulsants (specifically phenytoin, phenobarbital, and primidone) and corticosteroids, can elicit striking increases (up to hundredfold) in serum AP activity (and to a lesser extent GGT and also ALT activity) in dogs but not in cats. There usually is no other clinicopathologic or microscopic evidence of cholestasis (i.e., hyperbilirubinemia). Anticonvulsant drugs stimulate production of AP identical to the normal liver isoenzyme; GGT activity does not change. Pharmacologic levels of corticosteroids administered orally, by injection, or topically reliably provoke a unique AP isoenzyme that is separable from the others by electrophoretic and immunoassay techniques. This characteristic is useful when interpreting high total serum AP activity in a dog with subtle clinical signs suggestive of iatrogenic or naturally occurring hypercortisolism. The corticosteroid AP isoenzyme is a component of routine canine serum biochemistry profiles at several veterinary colleges and commercial laboratories. However, measurement of AP isoenzymes has been shown to be of limited usefulness either in dogs treated with phenobarbital (Gaskill et al., 2004) or in dogs with hyperadrenocorticism (Jensen et al., 1992). In the latter, it has a high sensitivity but very low specificity, so finding a low steroid-induced isoenzyme rules out hypercortisolism, but a high concentration of steroid-induced isoenzyme may be found in many disease other than hypercortisolism. Serum GGT activity rises similarly in response to corticosteroid influence but less spectacularly. Serum AP and GGT activities tend to be parallel in cholestatic hepatopathies of cats and dogs, although they are much less dramatic in cats. Simultaneous measurement of serum AP and GGT may aid in differentiating seemingly benign drug-induced effects from nonicteric cholestatic hepatic disease in dogs. Assessing serum AP and GGT activities together may also offer clues to the type of hepatic disorder in cats. Both enzymes are in low concentration in feline liver tissue compared with that in the canine liver and have short half-lives, so relatively smaller increases in serum activity, especially of GGT, are important signs of the presence of hepatic disease in cats. In cats a pattern of high serum AP activity with less strikingly abnormal GGT activity is most consistent with hepatic lipidosis (see Chapter 37), although extrahepatic bile duct obstruction (EBDO) must also be considered.

Serum Albumin Concentration

The liver is virtually the only source of albumin production in the body; thus hypoalbuminemia could be a manifestation of hepatic inability to synthesize this protein. Causes other than lack of hepatic synthesis (i.e., massive glomerular or gastrointestinal loss or bleeding) must be considered before ascribing hypoalbuminemia to hepatic insufficiency. Renal protein loss can be detected presumptively by routine urinalysis. Consistent identification of positive protein dipstick reactions, especially in dilute urine with inactive sediment, justifies further evaluation by at least measurement of random urine protein : creatinine ratio (normal ratio is <0.5 in cats and dogs). If proteinuria is ruled out, diseases that cause gastrointestinal protein loss should be considered; however, these diseases usually result in equivalent loss of globulins and thus panhypoproteinemia, although this is not invariably the case in inflammatory gastrointestinal disease wherein concurrent increase in gamma-globulins masks the gut loss. Conversely, although panhyproteinemia is reportedly not typical of hypoproteinemia of hepatic origin, globulin concentrations can be low in liver disease, particularly portosystemic shunts, because all plasma globulins except gamma globulins are made in the liver. In fact, globulin concentrations frequently are normal to increased in dogs and cats with chronic inflammatory hepatic disease. Because the plasma half-life of albumin is long in cats and dogs (8 to 10 days) and there must be loss of approximately 80% of functioning hepatocytes before hypoalbuminemia is expressed, the finding of hypoalbuminemia usually indicates severe chronic hepatic insufficiency. The exception to this is the hypoalbuminemia associated with a “negative acute phase” response in acute or acute-on-chronic inflammatory liver disease. Serum albumin can decrease when there is an increase in hepatic production of acute phase proteins in animals without hepatic insufficiency. Serum protein electrophoresis can help differentiate this condition from a true lack of hepatic function: Sevelius et al. (1995) showed that a low albumin concentration combined with a low concentration of acute phase proteins in electrophoresis indicated severe hepatic dysfunction with a poor prognosis, whereas hypoalbuminemia combined with normal or elevated acute phase proteins indicated a good prognosis. Hypoalbuminemia of any cause is unusual in cats, except in those with nephrotic syndrome. When interpreting serum protein concentrations, the clinician should remember that total protein values for young cats and dogs are lower than those for adults and that puppy serum albumin concentration is similar to that in adults, whereas kitten serum albumin concentration is lower than that in adult cats.

Serum Urea Nitrogen Concentration

Formation of urea as a means of detoxifying ammonia derived from intestinal sources takes place only in the liver. Despite this apparent advantage as a specific measure of hepatic function, serum urea concentration is commonly affected by several nonhepatic factors and the capacity of the liver to detoxify urea is so great that it is not noticeably reduced until severe, extensive end-stage liver disease ensues. Prolonged restricted protein intake because of complete anorexia or intentional reduction in protein intake for therapeutic purposes (e.g., chronic kidney disease; urate, cystine, or struvite urolithiasis) is the most common cause of low blood urea nitrogen (BUN) content. Prior fluid therapy and/or marked polydipsia/poluria of nonrenal causes will also result in a decrease in BUN. As always, reference ranges should be considered for each species when interpreting BUN values. For example, a BUN concentration of 12 mg/dl is well within normal limits for dogs but is subnormal for cats. If low BUN values are noted in a cat or dog with normal water intake and a good appetite for a diet with the appropriate protein content for the species (on a dry matter basis: 22% for dogs, 35% to 40% for cats), then the possibility of hepatic inability to convert ammonia to urea should be investigated.

Serum Bilirubin Concentration

Because of the large reserve capacity of the mononuclear-phagocytic system and liver to process bilirubin (e.g., 70% hepatectomy will not cause jaundice), hyperbilirubinemia can occur only from greatly increased production or decreased excretion of bile pigment. Specific inborn errors of bilirubin uptake, conjugation, and excretion have not been documented in cats or dogs. Increased production of bilirubin from red blood cell destruction arises from intravascular or extravascular hemolysis and rarely from resorption of a large hematoma; hyperbilirubinemia also occurs in association with rhabdomyolysis in Greyhounds and other dog breeds. Under these circumstances in dogs, serum bilirubin concentrations are usually lower than 10 mg/dl. Values usually do not increase above 10 mg/dl unless there is a concurrent flaw in bilirubin excretion. This has been borne out clinically in studies of dogs with immune-mediated hemolytic anemia in which high liver enzyme activities are observed, even before treatment with corticosteroids, and moderately delayed bilirubin excretion has been documented. It has been proposed that cholestasis results from liver injury associated with hypoxia and in some cases due to early disseminated intravascular coagulation (DIC). Because increased production and decreased excretion of bilirubin occur in dogs with severe hemolysis, serum bilirubin concentrations therefore can be as high as 35 mg/dl. Icterus in cats with pure hemolytic disease is an inconsistent finding and mild if present; specific bilirubin concentrations associated with experimentally induced or naturally occurring hemolytic diseases in cats are not available.

Because nearly all diseases associated with hyperbilirubinemia in cats and dogs are characterized by a mixture of conjugated and unconjugated bilirubinemia, quantifying the two fractions by use of van den Bergh’s test achieves little in discriminating primary hepatic or biliary disease from nonhepatobiliary disease in a clinical setting. This lack of benefit in using van den Bergh’s test may relate to the time between onset of illness and examination, which is usually at least several days. Under conditions of acute massive hemolysis, the total serum bilirubin concentration may consist primarily of the unconjugated form initially. As hemolysis continues, the liver is able to take up and conjugate bilirubin, accounting for a combination of unconjugated and conjugated bilirubin.

Because red blood cell membrane changes are often a component of many primary hepatobiliary disorders, accelerated red blood cell destruction often contributes to hyperbilirubinemia. In such cases, there is strong clinicopathologic evidence of cholestasis (high serum AP and GGT activities with moderate to high ALT activity), and if there is anemia, it is mild and poorly regenerative. Hyperbilirubinemia is attributed primarily to hemolysis when there is moderate to marked anemia with strong evidence of regeneration (except in the first 1 to 3 days, when the response is less regenerative) and minimal changes in serum markers of cholestasis.

Serum Cholesterol Concentration

Total cholesterol concentration is included in serum chemistry profiles by many commercial laboratories but affords useful information for only a limited number of hepatobiliary diseases. High total cholesterol values are observed in cats and dogs with severe intrahepatic cholestasis involving bile ducts or EBDO because of impaired excretion of free cholesterol into the bile and subsequent regurgitation into the blood. Low total serum cholesterol concentrations have been noted in dogs with chronic severe hepatocellular disease and frequently in cats and dogs with congenital portosystemic shunts (PSS). It has been speculated that hypocholesterolemia is a sign of markedly altered intestinal absorption of (and increased use of) cholesterol for bile acid synthesis when the enterohepatic recirculation of bile acids is disturbed, as occurs with PSS. In other hepatobiliary diseases of cats and dogs, the total cholesterol values vary considerably within the reference range. Normal values in 4-week-old kittens are higher than those for adults; 8-week-old puppy reference ranges are the same as those for adults.

Serum Glucose Concentration

Hypoglycemia is an unusual event associated with hepatobiliary disease in dogs and especially in cats. Lost capacity to maintain normal serum glucose concentrations occurs in animals with acquired chronic progressive hepatobiliary disease when 20% functional hepatic mass or less is remaining. This inability to maintain normal serum glucose concentrations is presumably caused by the loss of hepatocytes with functioning gluconeogenic and glycolytic enzyme systems and impaired hepatic degradation of insulin. Hypoglycemia is often a near-terminal event in dogs with chronic progressive hepatobiliary disease. In striking contrast is the frequent observation of hypoglycemia in dogs with congenital PSS, particularly small-breed dogs. In PSS hypoglycemia may be due to an increase in circulating insulin concentration caused by reduced first pass metabolism in the liver, as observed in humans, but this has never been investigated in dogs. Hypoglycemia is also common as a paraneoplastic syndrome in dogs with large hepatocellular carcinomas and can be associated with production in insulin-like growth factor by the tumour (Zini et al., 2007). In either case, if hypoglycemia is identified and confirmed by repeating the test using sodium fluoride tubes if necessary, and if nonhepatic causes (functional hypoglycemia, sepsis, insulinoma, or other neoplasm producing an insulin-like substance, Addison’s disease; see Chapter 53) are excluded, a primary hepatic tumor (e.g., hepatocellular carcinoma), a PSS, or severe generalized hepatopathy is suspected.

Serum Electrolyte Concentrations

Serum electrolyte determinations facilitate supportive care of cats and dogs with hepatobiliary disease but give no particular hints as to the character of the disorder. The most common abnormality is hypokalemia, which is attributed to a combination of excessive renal and gastrointestinal losses, reduced intake, and secondary hyperaldosteronism in dogs and cats with severe chronic hepatobiliary disease. Metabolic alkalosis, presumptive evidence of which might be abnormally high serum total carbon dioxide content confirmed by blood gas analysis, is usually caused by overzealous diuretic therapy in dogs with chronic hepatic failure and ascites. Hypokalemia and metabolic alkalosis potentiate each other and may also worsen signs of hepatic encephalopathy (HE) by promoting persistence of readily membrane-diffusible ammonia (NH₃).

Serum Bile Acid Concentrations

Recent validation of rapid, technically simple methods for SBA analysis in cats and dogs has provided a sensitive, variably specific test of hepatocellular function and the integrity of the enterohepatic portal circulation. “Primary” bile acids (i.e., cholic, chenodeoxycholic) are synthesized only in the liver, where they are conjugated with various amino acids (primarily taurine) before secretion into the bile. Bile is stored in the gallbladder, where it is concentrated until, under the influence of cholecystokinin, it is released into the duodenum. After facilitating fat absorption in the small intestine, the primary bile acids are efficiently absorbed into the portal vein and returned to the liver for reuptake and resecretion into the bile. A small percentage of primary bile acids that escapes resorption is converted by intestinal bacteria to “secondary” bile acids (i.e., deoxycholic, lithocholic), some of which are also resorbed into the portal circulation. Absorption of bile acids by the intestine is extremely efficient, but hepatic extraction from portal venous blood is not. This accounts for small concentrations of cholic, chenodeoxycholic, and deoxycholic acids that are released into the peripheral blood of healthy cats and dogs in the fasting state (total <5 μmol/L by enzymatic method and 5 to 10 μmol/L by radioimmunoassay [RIA]). During a meal a large load of bile acids is delivered to the intestine and portal circulation for recycling; postprandial values in normal dogs and cats may increase up to threefold to fourfold over fasting values (15 μmol/L with the enzymatic method for cats and dogs; 25 μmol/L with the RIA method for dogs). Normal values for juvenile animals are similar to adult reference ranges. Abnormally high fasting and/or postprandial SBA concentrations reflect disturbance in hepatic secretion into the bile or at any point along the path of portal venous return to the liver and hepatocellular uptake. Low SBA concentrations may be attributable to small intestinal (ileal) malabsorption of bile acids but might be difficult to interpret because both fasting and postprandial SBA concentrations may not be measurable in healthy animals.

The standard way to assess SBA concentrations is outlined in Box 36-1. Collective experience indicates that the likelihood of precipitating an episode of HE during this part of the test is extremely low, even in predisposed animals. After the serum is harvested, the samples may be refrigerated for several days or frozen almost indefinitely before assay. The stability of the blood sample is one of the major advantages over the much more labile serum ammonia test.

Studies of SBAs have confirmed their value in detecting clinically relevant hepatobiliary disease requiring definitive diagnostic testing in cats and dogs, especially in anicteric animals with equivocal clinical signs and unexplained high liver enzyme activity. There continues to be controversy as to whether a fasting or postprandial value alone is sufficient or whether fasting and postprandial measurements are required. If only one sample can be obtained (and the animal will eat or can tolerate being force-fed a small meal), the postprandial value is most useful to determine the presence or absence, but not the type, of clinically relevant hepatobiliary disease in most cats and dogs. Current recommendations state that for animals suspected of having acquired hepatobiliary disease, biopsy is needed when postprandial SBA concentration using the enzymatic method exceeds 20 μmol/L in cats and 25 μmol/L in dogs, although other authors (particularly in the United Kingdom) suggest that SBA between 20 and 40 μmol/L in dogs represents a grey area (Hall et al., 2005). Elevations in this region have been seen with secondary hepatopathies (particularly hyperadrenocorticism) and with small intestinal bacterial overgrowth because of reduced hepatic clearance of deconjugated bile acids. Therefore the authors would recommend a liver biopsy with a higher cut-off for postprandial bile acids of 40 μmol/L. No pattern of preprandial and postprandial values is pathognomonic for any particular hepatic disorder, although it is safe to make certain generalizations. Magnitude of elevation above 20 μmol/L in cats and 25 μmol/L in dogs roughly correlates with the severity, but not the reversibility, of the hepatobiliary disorder, although with PSS, the magnitude of elevation does not correlate with the degree of shunting or severity of clinical signs. The change between the fasting value and the postprandial value likely corresponds to portosystemic shunting, either microscopic (intrahepatic) or macroscopic. There is so much overlap in fasting and postprandial SBA patterns among primary hepatobiliary diseases that no particular statement can be made regarding the specific causative hepatobiliary disease. Occasionally, fasting SBA levels are higher than postprandial levels, which signifies nothing more than occasional, normal, spontaneous gallbladder contraction in fasting. In general, secondary hepatic diseases cause more modest hepatobiliary dysfunction (SBA values <100 μmol/L).

For the diagnosis of congenital PSS, fasting and postprandial SBA determinations are recommended to enhance detection ability because it is relatively common for fasting values to be well within normal limits and for postprandial values to be as high as tenfold to twentyfold higher than normal postprandial values.

Now that simplified methods for SBA measurement have been developed (i.e., enzymatic, RIA) and are accessible, determination of total SBA has become a convenient, practical test of hepatobiliary function in cats and dogs. Some reference laboratories use an adapted enzymatic method, a commercial enzymatic kit (Enzabile; Nyegaard and Co., Olso, Norway), or a commercial RIA (Conjugated Bile Acids Solid Phase Radioimmunoassay Kit ¹²⁵I; Becton Dickinson, Orangeburg, N.Y.). Each yields comparable diagnostic results, although the sample size needed for the RIA assay is quite small (50 μl) compared with the enzymatic method (400 to 500 μl). Because the measurement of fasting and postprandial SBA concentrations assesses the same functions as the ammonium chloride (NH₄Cl) tolerance test without potentially dangerous consequences, it is the preferred test. As with any specially requested test, the laboratory chosen should use methods verified for clinical use in the target species and be able to provide reference ranges.

A benchtop “snap” test for bile acids has recently become available from IDEXX Laboratories (see www.idexx.com/animalhealth/analyzers/vetlabnotes/2005snapreader.jsp). The disadvantage of the snap test is that it has a low cut-off, which means that it does not differentiate secondary from primary hepatobiliary disease.

Several factors may affect SBA values and therefore their interpretation. One aspect of the SBA challenge test that has not been standardized is the feeding step. The ideal quantity and composition of the test meal have not been determined. Size of the test meal and therefore consumption of the entire meal or only part of the meal may affect gastric emptying. Delayed gastric emptying could cause the peak SBA concentration to occur more than 2 hours later. Hastened or delayed intestinal transit time or the presence of intestinal disease (especially of the ileum) may also impede and blunt peak absorption of the test meal. It is likely that fat content of the test meal is important because fat is the primary stimulus for the small intestinal mucosa to secrete cholecystokinin, which causes gallbladder contraction. Expulsion of bile during periodic physiologic gallbladder contraction between meals may complicate interpretation of the fasted sample result. Lipemia of the sample can seriously affect the validity of the test, particularly on heparinized blood. For this reason, it is far preferable to use serum, both for the external samples and for the snap test.

Several questions remain to be answered regarding the clinical use of SBA measurement in cats and dogs. Investigation of individual SBA profiles in cats and dogs with various hepatobiliary diseases has provided interesting information but no clear and specific profile for any one disease. Can sequential SBA values be used to more precisely monitor a cat’s or dog’s progress? Until this and other questions are answered, use of SBA analysis is limited to measuring total serum values as a sensitive and relatively specific screening test for the presence or absence of clinically significant hepatobiliary disease. Additional diagnostic testing must always follow to identify the specific cause.

Urinary Bile Acid Concentrations

Determination of bile acids accumulating in urine over time can be used to assess hepatobiliary function. Urine bile acids are believed to reflect average serum bile acid concentrations during the interval of urine formation. Expression of urine bile acid concentrations as a ratio with the urine creatinine concentration eliminates the influence of urine concentration and flow. Random urine sampling for bile acid determination does not require attending to the timing of an enterohepatic challenge or obtaining a sample after withholding food. In recent studies urinary bile acid concentrations were increased in dogs and cats with hepatobiliary disease and portosystemic vascular anomalies, compared with dogs and cats with nonhepatic disorders (except for hepatic neoplasia in dogs; Balkman et al., 2003; Trainor et al., 2003). The urine nonsulfated bile acid : creatinine ratio and the urine sulfated plus nonsulfated bile acid : creatinine ratio positively correlated with serum bile acid test results and had similar overall diagnostic performance and substantially higher (dogs) or similar (cats) specificity, compared with the serum bile acid test, and are recommended. The urine sulfated bile acid : creatinine ratio had lower sensitivity in dogs and cats, compared with the serum bile acid test.

Plasma Ammonia Concentration

One test that is not included in a standard screening battery of tests but is available through reference or human hospital laboratories is plasma ammonia concentration. Fasting plasma ammonia can be measured in any cat or dog with historic or physical examination findings suggestive of HE. Signs of HE (see Box 35-2), whether they have a congenital or acquired basis, appear the same. Quantifying plasma ammonia concentration not only can confirm HE, although normal fasting values in animals with hepatobiliary disease are relatively common, but can also provide baseline data and help in evaluating response to treatment. However, SBA values (particularly postprandial) provide very similar information. Some investigators have suggested that arterial ammonia concentrations may more accurately represent blood ammonia status in dogs with hepatobiliary disease than venous measurements because skeletal muscle can metabolize ammonia. High plasma ammonia concentration usually indicates reduced hepatic mass available to process ammonia and/or the presence of portosystemic shunting, which disrupts presentation of ammonia to the liver for detoxification. However, ammonia is very labile in the blood sample and, for example, can be falsely elevated if the blood sample is taken in an environment contaminated with urine. Sample handling has to be undertaken with caution, and some benchtop analyzers are inaccurate, particularly in the moderately elevated range. For these reasons, SBAs are often a preferred test. The exception to this would be an animal with suspected hepatic encephalopathy and concurrent cholestasis. As outlined in the preceding paragraphs, bile acid concentrations will be high in cholestasis (because they are excreted in the bile) independent of any reduction in liver function or shunting. Measuring blood ammonia in these circumstances will give useful additional information about potential shunting and HE.

In a recent study the 12-hour fasting plasma ammonia concentration had higher sensitivity and specificity than the 12-hour fasting bile acid concentration for detecting portosystemic shunting in a general population of dogs and in dogs with liver disease (Gerritzen-Bruning et al., 2006). However, a bile acid stimulation test (fasting and 2-hour postprandial bile acid) has a much higher sensitivity for detecting PSS than a single fasting bile acid, and a single postprandial bile acid concentration is likely as sensitive as a fasting ammonia concentration, although the authors did not test this.

Although reference ranges vary among laboratories, fasting plasma ammonia values for normal dogs are typically 100 mg/dl or less and 90 mg/dl or less for normal cats. At least 6 hours of fasting should precede sample collection. Samples must be collected into iced ammonia-free heparinized tubes and spun immediately in a refrigerated centrifuge. Plasma must be removed within 30 minutes so that values will not be spuriously elevated by hemolysis because red blood cells contain two to three times the ammonia concentration of plasma. To obtain accurate values, feline plasma can be frozen at −20° C and assayed within 48 hours; canine plasma must be assayed within 30 minutes.

If signs are compatible with HE at the time of sample collection, a single fasting sample will suffice. If there are no signs of HE and results of other tests are equivocal, a postprandial challenge test may be performed (see Box 36-1). The older ammonium chloride challenge tests (either oral or rectal) are contraindicated because of the significant potential for either test to trigger a severe encephalopathic crisis in the patient. The postprandial ammonia test is a safer test and has a 91% sensitivity for portosystemic shunting but only a 31% sensitivity for diffuse hepatocellular disease (Walker et al., 2001).

Plasma Protein C Activity

Plasma protein C activity was recently evaluated as a marker of hepatobiliary disease in dogs. Protein C is an anticoagulant protein that is synthesized in the liver and circulates as a plasma zymogen. Low protein C activity has been associated with thrombotic disorders in humans and animals. Low protein C activity has also been documented in dogs with acquired and congenital hepatobiliary disorders, and dogs with PSS appear to develop the lowest protein C activity. In a recent study by Toulza et al. (2006), protein C acivity was significantly lower in dogs with congenital or acquired portal systemic shunts, compared with dogs without PSS. Plasma protein C activity improved or normalized after surgery for the shunt. These findings suggest that plasma protein C activity reflects the adequacy of hepatoportal perfusion in dogs and that protein C activity may prove useful as a means to monitor improvement of hepatic-portal perfusion after ligation of portosystemic vascular anomalies. Plasma protein C activity may also help differentiate dogs with intrahepatic portal vein hypoplasia from those with portal systemic vascular anomaly (plasma protein C activity ≥70% versus <70%, respectively).

General Considerations

For many primary hepatobiliary diseases of cats and dogs, a hepatic biopsy is needed to establish a final diagnosis and prognosis. In some cases bile culture is also imperative. Biopsy is indicated to (1) explain abnormal results of hepatic status and/or function tests, especially if they persist for longer than 1 month; (2) explain hepatomegaly of unknown cause; (3) determine hepatic involvement in systemic illness (although biopsy is not always necessary for this); (4) stage neoplastic disease; (5) objectively assess response to therapy; or (6) evaluate progress of previously diagnosed, not specifically treatable disease. Percutaneous hepatic biopsy is not performed if there is a good chance that the disease can be corrected surgically, such as in some cases of EBDO or congenital PSS; instead, a specimen is obtained at the time of surgery to complete the diagnostic evaluation. Several approaches are available; choice is dictated by patient and operator considerations (Box 36-2). In addition, in most cases of hepatic disease the accuracy of histological diagno sis is better with larger (i.e., surgical or laparoscopic) rather than smaller (i.e., needle) biopsies.

All cats and dogs undergoing hepatic biopsy are fasted for at least 12 hours, regardless of the approach selected. In general, percutaneous needle core biopsy or aspiration (for cytologic analysis) of a single cavitary or solid lesion highly likely to be nonlymphoid cancer should be avoided unless the owner is unwilling to permit surgery for complete resection. Fine-needle aspiration of the liver for cytologic analysis is rarely advisable because of low diagnostic yield and often misleading results. The exceptions to this are for quick diagnosis of hepatic lipidosis in cats and possibly for suspected hepatic lymphoma, although even then the diagnosis may need to be confirmed histologically (Fig. 36-13). However, an overall correlation of only 30% in dog and 51% in cats was found in one study comparing the cytologic diagnosis with the histopathologic diagnosis of a variety of liver diseases (Wang et al., 2004).

FIG 36-13 A 4-year-old spayed female domestic short-haired cat with suspected hepatic lipidosis positioned in right lateral recumbency for blind fine-needle aspirate for cytology. With care taken to avoid the spleen, the needle is directed craniomedially into the liver.

In an especially small and/or firm fibrotic liver, it is difficult to obtain a biopsy specimen by percutaneous needle methods; small, fragmented specimens that are challenging to interpret are often the result (Fig. 36-14). There is less than a 40% correlation between 18-gauge needle biopsy and wedge biopsy for certain hepatobiliary diseases (i.e., chronic hepatitis/cirrhosis, cholangitis, portovascular anomaly, fibrosis). If a needle technique is selected, the largest available instrument is used (preferably 14 gauge; minimum 16 gauge) and multiple samples are taken to ensure samples adequate for examination.

FIG 36-14 A, Liver specimen obtained percutaneously (with ultrasound guidance) from a dog with hepatic fibrosis and nodular regeneration (B). The specimen was difficult to obtain because the liver was firm and rubbery in texture. C, The resultant sample was difficult to interpret histologically.

The animal’s coagulation status is determined before a liver biopsy is performed, regardless of the approach. Ideally, a complete coagulation profile (one-stage prothrombin time [OSPT], APTT, fibrin degradation products, fibrinogen content, platelet count) is obtained; a platelet count and an activated clotting time or whole blood clotting time in a glass tube, as a screening test of the intrinsic coagulation cascade, are also acceptable. Bleeding after ultrasound-guided biopsy is more likely if the platelet count is less than 80,000 cells/μl or if the OSPT (dogs) or APTT (cats) is prolonged (Bigge et al., 2001). If possible, von Willebrand’s factor is measured in susceptible breeds in advance of biopsy because results of standard coagulation tests are usually normal in affected dogs. A buccal mucosa bleeding time test provides indirect assessment of platelet function (see Chapter 87). In dogs with von Willebrand’s disease, desmopressin acetate (DDAVP) is given (1 μg/kg intranasal preparation subcutaneously) before surgery to enhance shift of von Willebrand’s factor activity from endothelial cells to the plasma.

Mild abnormalities in coagulation test results do not preclude liver biopsy. In fact, results of routine coagulation tests may not correlate with liver bleeding times, as was found in one study of human patients. Liver biopsy should be delayed if there is clinical evidence of bleeding or marked abnormalities in results of coagulation tests. Because animals with complete EBDO may be vitamin K deficient (manifested by prolongation of both OSPT and APTT), treatment with vitamin K₁ (0.5 to 1.0 mg/kg [maximum of 10 mg] subcutaneously q12h for 3 treatments) is indicated for 1 or 2 days before surgery. Vitamin K supplementation can also improve coagulation times in animals with other liver disease, particularly cats. Repeating the OSPT and APTT within 24 hours after administration of vitamin K₁ should demonstrate normal or near-normal values. If not, the dose can be adjusted and the procedure delayed. Although it may not seem rational to give vitamin K₁ to animals with severe parenchymal hepatic disease before surgery, it has been of benefit to some animals and, if given properly, can do no harm. These animals may have high serum activity of proteins induced by vitamin K antagonism (PIVKA) that could impart bleeding tendencies. If there has been minimal improvement in coagulation test results after vitamin K₁ has been administered, fresh frozen plasma is administered before biopsy. If bleeding is excessive during or after biopsy and cannot be controlled locally with direct pressure or application of clot-promoting substances, fresh whole blood or plasma is given (see Chapter 83 for transfusion guidelines).

Techniques

Information gathered before biopsy must support the fact that the likelihood of acquiring a diagnostic sample without complications is high. A specimen procured from any area of the liver is considered representative of the disease. Because only a small stab incision large enough to accommodate the biopsy needle is needed (a No. 11 blade is the perfect choice for this purpose), healing in hypoalbuminemic animals is not compromised. If the operator is confident with the biopsy procedure, there is little time involved and only heavy sedation is required. If the results are nondiagnostic, a larger specimen is obtained for histopathologic examination, usually by laparoscopy or laparotomy.

Biopsy can be performed blindly if the cat or dog has generalized hepatomegaly and the operator is confident of the path of the needle. The most common needle biopsy instruments are the Tru-Cut (Cardinal Health) and Jamshidi Menghini (Cardinal Health; Kormed Inc.) needles. The Jamshidi Menghini biopsy needles can be operated with one hand, and aspiration is used to sever and contain the specimen within the barrel of a 6- or 12-ml syringe. The Tru-Cut needle requires two hands to operate and relies on the tissue falling into the specimen trough and then being severed by the sharp outer cannula (Fig. 36-15). One-handed operatable semi-automatic (e.g., Tenmo Evolution biopsy needle, Cardinal Health; VET-core biopsy needle, Global Veterinary Products Inc) and automatic (e.g., Pro-Mag Ultra Automatic biopsy instrument, Manan Medical Products; Bard Biopty biopsy instrument and Bard Biopty-Cut biopsy needle, Bard Inc) versions of this instrument are also available. These biopsy needles are intended for single use. Either the automatic biopsy instrument or the semi-automatic biopsy needle device can be used to obtain liver biopsies in dogs, but only the semi-automatic biopsy needle device should be used in cats. A recent study identified a high risk of fatal complications (i.e., unexpected fatal shock reaction) when an automatic biopsy instrument was used to obtain liver biopsies in cats (Proot and Rothuizen, 2006).

FIG 36-15 A, Tru-Cut biopsy needle with the specimen trough exposed (left) and then covered by the sharp outer cannula (right). B, Liver tissue filling the specimen trough (between arrows).

Biopsy can be done of any palpably enlarged lobe as long as care is taken to angle the needle to avoid puncturing the gallbladder. Most often, the animal is placed in right lateral recumbency for this purpose and biopsy of the left lateral lobe is done. Elevating the head and thorax slightly may assist in “presenting” the liver to the operator. Two or three complete core specimens are obtained; if indicated, one core specimen is placed in a sterile container for culture and sensitivity testing. Gently rolling a specimen on a slide for cytologic assessment is a good way to attempt to identify the disease process quickly and inexpensively. Each of the remaining core specimens is placed on a piece of stiff paper (e.g., filter paper) in correct orientation (Fig. 36-16) before immersion in fixative for histologic examination and/or special testing.

FIG 36-16 Needle biopsy specimen affixed to a stiff piece of paper to preserve orientation of the sample during formalin fixation for histopathologic examination.

After biopsy, a small bandage is applied to keep the site clean during recovery, and the animal is placed in a position to allow body weight to compress the region of the biopsy sites in the liver (e.g., left lateral recumbency). Consideration should be given to postoperative analgesia; puncture of the liver capsule can be painful. The animal should be monitored carefully for any evidence of hemorrhage for several hours after the procedure. As long as the biopsy procedure proceeded smoothly and without unpleasant surprises (animal awake and struggling), only basic monitoring of mucous membrane color and the skin puncture site is needed. Naturally, if excessive hemorrhage or damage to other organs occurs with this blind technique, detection and treatment may be delayed.

Visualized percutaneous needle biopsy, with the aid of either US (Fig. 36-17) or modified laparoscopic equipment (Fig. 36-18), allows selection of the site(s) and direct or indirect inspection after the biopsy. When the procedure is properly performed, serious complications are few. In an animal in which diffuse or multifocal hepatobiliary disease is suspected, multiple biopsy specimens are obtained safely. General anesthesia is usually required for use of a modified laparoscope. Aspiration of the gallbladder for cytologic analysis and culture can be accomplished with US guidance or by laparoscopy. Bile leakage may occur, even if a small-gauge needle is used, so attempts are made to completely evacuate the gallbladder, and the needle should be placed in the gallbladder through the liver parenchyma to help prevent leakage. Some surgeons prefer to obtain bile during laparotomy when a purse-string suture can be applied to the aspiration site to prevent seepage. Large-volume abdominal effusion hinders direct inspection of the liver and associated structures and must be removed before laparoscopic biopsy is attempted.

FIG 36-17 Biopsy gun instrument with accompanying biopsy needle used for obtaining liver specimens with ultrasonographic guidance.

FIG 36-18 Modified laparoscopic approach for liver biopsy. A, Readily available materials needed for the procedure. B, A Tru-Cut biopsy needle is used for obtaining liver specimens. C, The liver is first inspected, and then the needle is passed through a sterile otoscope cone into the liver for tissue sampling. See Bunch et al. (1985) in Suggested Readings for further details on this procedure.

An operative approach (laparoscopy [Fig. 36-19], laparotomy) is preferred for liver biopsy if the liver is small, US equipment is not available, or the operator is not experienced with the aforementioned percutaneous needle methods. Laparotomy is perfectly acceptable for dogs and cats that are good anesthetic risks and allows thorough examination of the liver, biliary tract, and portal vein as well as other abdominal structures, such as lymph nodes. Bile can be acquired easily and safely. The procedure is more prolonged, and healing complications may arise in severely hypoalbuminemic animals, notably those with intractable ascites, but larger samples for histopathologic examination and special staining techniques are obtainable (Fig. 36-20) and hemorrhage can be arrested directly. Use of nonabsorbable suture material and small cranial or flank incisions may lessen incision complications. Obviously, this is the approach of choice for surgically correctable diseases; a liver biopsy specimen is obtained concurrently.

FIG 36-19 A, Laparoscopic liver biopsy. B, Tip of the biopsy instrument that is passed through one of the preplaced cannulas. C, Intraabdominal view of a dog with chronic hepatic disease and portal hypertension. Note the prominent, tortuous omental veins.

FIG 36-20 Comparison of liver specimens obtained by different methods and mounted on glass slides. These samples are adequate for histopathologic examination: percutaneous needle sample (left); samples obtained intraoperatively by use of an 8-mm skin biopsy punch (middle) or removal of a wedge specimen (right).

As for the percutaneous biopsy techniques, liver and/or bile specimens for microbiologic culture are aseptically processed first. Impression smears for cytologic analysis are then made by gently touching the specimen to a slide before placing it in fixative. Excess blood is removed by blotting the sample on gauze before slides are made. Abnormal populations of cells (e.g., mast cells, lymphoblasts) are readily detectable using rapid stain systems such as Diff Quik (Harleco, Gibbstown, N.J.). For routine processing and histopathologic examination, liver tissue specimens are submerged in buffered 10% formalin at a ratio of at least 10 parts formalin to 1 part tissue. Samples for copper histochemical staining or tissue quantification are harvested and fixed or preserved according to the specifications of the pathology laboratory selected to do the assays. Other special stains for infectious agents or fibrous tissue, amyloid, glycogen, and other metabolic products are available, and their use is discussed with the attending pathologist before the tissue specimen is obtained. A portion of the specimen can be frozen for molecular studies (e.g., PCR for organisms or tumor clonality).