Laboratory evaluation of liver function

Introduction

Laboratory evaluation of the liver requires a minimum database including a complete blood count (CBC), biochemical profile and urinalysis. The biochemical profile will contain most data points utilized to directly interrogate hepatic health and function, but in isolation cannot separate primary from secondary hepatic pathology. This distinction relies on synthesis of all available data and might remain elusive in some instances – for example, dogs with immune-mediated hemolytic anemia (IMHA) will often present with biochemical evidence of hepatocellular injury and cholestasis, due to the effects of hypoxia and inflammatory cytokines on liver function 1,2,3,4. The CBC is essential to facilitate recognition of a primary hematological disease, and to support the conclusion that the hepatic pathology is a secondary process.

The serum biomarkers used clinically to evaluate the liver reflect its main functions. Protein synthesis is gauged by albumin concentration, glucose and cholesterol levels reflect the liver’s role in intermediary metabolism, and bile acids and cholesterol are indicators of biliary tract patency and enterohepatic recirculation. The liver uptakes, conjugates, and excretes bilirubin produced by the breakdown of hemoglobin by phagocytes. Detoxification of toxins and drugs by the liver may be biochemically silent or reflected in changes in specific serum enzymes.

The laboratory data from the minimum database are gathered and synthesized to aid in the detection of four major pathological processes in the liver, namely hepatocellular injury, cholestasis, hepatocellular dysfunction/insufficiency, and alterations in hepatic portal circulation. These processes often occur concurrently, or as a downstream consequence of one another. However, there is also the potential for multiple pathologies to have opposing actions on specific analytes, decreasing their diagnostic sensitivity for hepatic pathology – e.g., cholesterol may be within the reference interval in a dog with a severely cirrhotic liver due to structural cholestasis counteracting the decreased synthesis of cholesterol secondary to a reduced hepatic mass (i.e., insufficient hepatocytes to clear bile acids). A process-centric approach facilitates targeted selection of diagnostic tests, and the generation of a more defined differential diagnosis list.

Detection of liver injury

Injury is a rather nebulous but commonly used term to reflect harm that ultimately results in hepatocellular membrane damage and presumptively some degree of cellular lysis and death. The liver can be injured directly, due to parenchymal inflammation (primary hepatitis), toxins (drugs and plants), neoplasia (primary or metastatic), or as a consequence of a systemic disease impacting blood flow, oxygen delivery, or an endocrinopathy. Regardless of the cause, there are only a few biochemical indicators we can use to identify this type of injury.

Liver “leakage” enzymes

The serum enzymes used to evaluate the liver are divided into two main categories. The first group reside in the cytosol of the hepatocytes and are colloquially termed leakage enzymes. In dogs and cats these enzymes include alanine aminotransferase (ALT) and aspartate aminotransferase (AST). The increase in serum activity of these enzymes is specific for hepatocellular injury, but only if hemolysis and severe muscle injury (as more AST resides in muscle tissue than the liver) are excluded as alternative sources 5,6. Due to the liver’s large cellular reserve and capacity for regeneration, the magnitude of the increases in these enzymes is proportional to the number of hepatocytes damaged, but does not reflect the reversibility of the injury nor provide any indication of hepatic synthetic function 5,7. Serial monitoring is critical; ALT has a half-life of 2-3 days in the dog and only 3-4 hours in the cat, whereas the half-life of AST is less than a day in the dog 5,8. Each biochemical profile is a single snapshot in time, and so enzyme half-life must be considered when weighing the clinical significance of changes over time. It is important to remember that a small fibrotic liver which has been subjected to long-term disease may have few viable hepatocytes remaining to leak, and thus serum activity levels of ALT and AST could be minimally increased or even within the reference intervals despite advanced pathology. Inflammatory or necrotizing disorders are generally associated with the largest increases in leakage enzymes. 

Hepatic inflammation can be insidious. In a study of Labrador Retrievers that were first-line relatives to dogs with copper-associated hepatopathy, 64% (122/191) of clinically healthy dogs had histopathological evidence of inflammatory infiltrates. Serum enzyme activities were relatively insensitive for the detection of both acute and chronic hepatitis in this population 9. In a separate study, evaluation of 4559 hepatic biopsies from dogs revealed evidence of copper levels > 400 ppm dry weight (reference range 120-400) in 50% of the biopsies, and 19% had Cu > 1000 ppm. In that same study, necroinflammation (i.e., the liver’s immune response to necrosis) was predictive of the copper level 10. Hepatic accumulation of copper in cats is less well characterized, but is emerging as a cause of hepatitis in this species 11. Feline hyperthyroidism is a recognized cause of mild increases in ALT, but the mechanism remains undefined 12. Glutamate dehydrogenase (GLDH) and sorbitol dehydrogenase (SDH) are useful hepatocellular leakage enzymes in other species, there are few studies evaluating their diagnostic utility in cats compared to dogs. 

Inducible enzymes

The second set of enzymes used for liver evaluation are termed inducible enzymes; these include alkaline phosphatase (ALP) and gamma glutamyl transferase (GGT). Hepatocytes increase the protein synthetic pathways that produce these enzymes secondary to what is termed an inductive event, typically exposure to drugs or hormones 13. Bile acids can also induce these enzymes and also solubilize cell membranes, hence the strong association between increases in ALP and GGT with cholestasis 14. The biliary canalicular cells respond to increased pressures by undergoing hyperplasia, increasing the number of cells producing GGT 15. This is often reflected in a proportionately higher increase in GGT when obstructive cholestatic processes are in play. Sustained increases in GGT and ALP can reflect healing and hyperplasia in the biliary tract. Like evaluation of the leakage enzymes, serial monitoring is essential. Bone can be an additional source of ALP in both dogs and cats, and corticosteroids induce this enzyme in the dog, therefore diseases associated with increased bone proliferation (and the age of the animal) and exposure to either endogenous or exogenous corticosteroids (in the dog only) must be considered in laboratory evaluation of the liver 16. ALP increases are also recognized in the context of feline hyperthyroidism 17, whilst phenobarbital treatment can cause increases in ALT, ALP, and GGT in the dog 18. If laboratory results are confusing, levamisole can be used to suppress the isoenzyme that is induced by corticosteroids to help tease out a diagnosis, but is rarely needed. 

Inducible enzymes can be released from intact hepatocytes, but lysis of hepatocytes will release these enzymes as well. In these instances, one typically expects a greater increase in leakage enzymes. Of course, an inductive event prior to a lytic event produces a serum enzyme picture that can be quite obtuse.

Cholestasis

Cholestasis is the stoppage or suppression of bile flow. This can occur due to a decrease in hepatocyte secretion or structural impingement anywhere along the biliary tract, from the small canaliculi to the gallbladder. The excretion of bile from the hepatocyte, across the biliary epithelial cells, and ultimately into the intestinal tract, is highly energy dependent, requires several active transporters, and depends on an osmotic gradient. Therefore, there are several potential points of failure in this process that can result in cholestasis 14. Common pathologies that can cause cholestasis include hepatocellular swelling (e.g., hepatic lipidosis), neoplastic processes, chronic inflammation (fibrosis), choleliths, parasites, pancreatitis, gallbladder mucoceles, and causes of functional cholestasis such as hypoxia and cytokine-mediated suppression of excretion. Compressive lesions can occur diffusely within the liver (e.g., hepatic lipidosis in a cat causing cellular swelling that compresses bile canaliculi) or a mass-like lesion involving the gallbladder or the larger bile ducts. Compressive lesions tend to cause accumulation of other substances that require a patent biliary system for excretion. Therefore, many structural cholestatic lesions are marked by hypercholesterolemia and hyperbilirubinemia. 

Hyperbilirubinemia and icterus

An overt clinical sign of potential hepatic disease is icterus, the visible yellowing of the skin, mucus membranes and sclera due to the accumulation of bilirubin (Figure 1). This is due to hyperbilirubinemia (Figure 2), generally above 2-3 mg/dL (34-51 μmol/L). It is important to recognize that significant hepatic pathology can exist in an animal that is not icteric. Bilirubin is produced as the components of senescent red blood cells (RBC) are broken down and recycled, and occurs in two major steps. Firstly, the RBC is engulfed by a macrophage and digested, and the hemoglobin molecule is enzymatically converted to bilirubin. This step generally occurs in splenic macrophages, and to a lesser degree in Kupffer cells (the macrophages residing in the liver). Secondly, the unconjugated bilirubin produced by the macrophages is non-covalently bound to albumin and transported to the liver where it is taken up by hepatocytes, and enzymatically conjugated to facilitate excretion in the bile. Hyperbilirubinemia can occur when there is a hemolytic process, a “pre-hepatic” icterus. Hemolysis accelerates the turnover of RBC and breakdown of hemoglobin. In these instances, the liver is simply overwhelmed and unconjugated bilirubin initially accumulates in the blood 19. Hepatic hypoxia secondary to acute anemia reduces cellular energy in the hepatocytes and slows bilirubin excretion. Inflammatory cytokines also suppress excretion of bilirubin. Thus, in an animal with a hemolytic anemia (pre-hepatic icterus) there is a secondary hepatic component that is due to biochemical lesions. 

The “hepatic” or “post hepatic” cases of icterus are either intracellular biochemical lesions or structural lesions that physically inhibit the excretion of bilirubin into the bile (e.g., cholestasis). The causes of hepatic and post-hepatic icterus often overlap. In terms of generating a differential diagnosis, once pre-hepatic icterus is excluded, the quest really becomes about determining if there is a biochemical lesion in the hepatocytes and/or biliary epithelial cells, or a structural impingement (which typically relies on imaging rather than laboratory data).

Liver synthetic function and portal blood supply

The liver produces a plethora of proteins, the most abundant being albumin, but also acute phase proteins and several coagulation factors. In addition to this, the liver is the main site of gluconeogenesis, cholesterol synthesis, and houses the enzymes of the urea cycle. Thus albumin, glucose, cholesterol and urea are used in tandem to indicate the liver’s ability to participate effectively in intermediary metabolism. If a dog or cat lacks enough functioning hepatocytes due to either injury or atrophy, these four analytes can all be decreased. However, these analytes are also impacted by other processes. Albumin is a negative acute phase protein and can be lost from the body via several pathways, including the urinary or gastrointestinal routes or from hemorrhage or exudation. A CBC and urinalysis can help rule in/out these alternative etiologies. Cholesterol synthesis is decreased in malabsorptive diseases, hypoadrenocorticism, and some cancers, so these processes must be excluded if synthetic liver failure is to be considered the leading differential 20,21,22. Imaging can be helpful in evaluating overall hepatic size and looking for evidence of potential fibrosis. 

If altered blood flow and subsequent decreased clearance functioning of the liver is suspected, bile acids can be measured to detect this. Increased serum bile acids can indicate a decreased hepatocellular mass and/or portosystemic shunting (with bile acids remaining in the systemic circulation due to blood by-passing the liver). Bile acids will likely be increased if there is biochemical evidence of cholestasis, and this test then becomes difficult to utilize to gauge hepatic mass and blood flow. If there is concurrent ileal disease, the absorption of bile acids secreted into the gastrointestinal tract is compromised. Therefore, in animals with significant hepatic disease and ileal malabsorption, bile acids may be within normal limits. Dynamic (pre- and post-prandial measurement) testing increases the sensitivity of bile acids, although 15-20% of dogs will have a higher level on the fasting sample, presumptively due to delayed gastric emptying or spontaneous gallbladder contraction. In these cases, the higher of the two values should be used for interpretation 23. 

Ammonia, produced as a byproduct of protein metabolism, is also cleared from circulation by the liver. Unlike bile acids, there is no enterohepatic component to this clearance, therefore plasma ammonia concentration is not impacted by cholestasis. This test can be a sensitive and specific test for hepatocellular mass and blood delivery to the liver, but it requires special sample collection and handling (Box 1) and is not performed by all laboratories, so the clinician should call the laboratory prior to submission. In cats, ammonia is more specific than bile acids in detecting shunting or decreased hepatic mass, but less sensitive 24. Urea cycle defects can also cause a hyperammonemia, but other biochemical evidence of hepatic pathology will typically be lacking in these animals.

Box 1. Notes on ammonia testing.

Pulling in the urinalysis

The canine kidney has a low renal threshold for conjugated bilirubin, thus in cholestatic animals a bilirubinuria may often precede hyperbilirubinemia. In long-standing cholestatic diseases, conjugated bilirubin may covalently bind to albumin, forming delta-bilirubin. This retains the bilirubin in the blood, thus a lack of bilirubinuria in an animal with other evidence of cholestasis should not be used to discount a cholestatic process. Male dogs can conjugate some bilirubin in their renal tubules, so a small amount of bilirubin in the urine of a male dog may be clinically insignificant. If hepatic synthesis of urea is decreased, this may have a secondary impact on the kidney, because urea is essential to the kidney’s ability to concentrate the urine. Many animals with hepatic disease will have poorly concentrated urine and could be polyuric (Figure 3). Bilirubinuria in a cat always warrants further investigation. The spill-over of ammonia into the urine can result in the formation of ammonium biurate crystals.

Additional findings on test profiles

RBC morphology

RBC morphology can be impacted by serum lipoprotein changes that occur with hepatic disease. This is reflected in the poikilocytosis often seen in cats with hepatic lipidosis (Figure 4). Animals with a portosystemic shunt can present with microcytosis with or without anemia. Acanthocytes are loosely associated with hepatic pathologies. 

Hepatic coagulopathy

The liver not only produces coagulation factors, but also several proteins that inhibit coagulation, including protein C and protein S. Animals with a decrease in synthetic liver function can present with either a bleeding tendency, or a thrombotic one. Gastrointestinal diseases that decrease vitamin K prevent hepatocytes from carboxylating factors and this can cause ineffective coagulation. Factor VII has the shortest half-life, thus the prothrombin time may be prolonged prior to the activated partial thromboplastin time in these instances.

Protein C and CRP

Protein C and C-reactive protein (CRP) are both made in the liver, and both can be accurately measured in companion animals. However, the unfortunate similarity in nomenclature can precipitate an occasional error in test selection. Protein C is a proven biomarker of hepatic synthetic function and hepatoportal perfusion; as a component of a major anti-coagulative pathway, serum concentrations have been shown to be useful in identifying portosystemic shunts versus microvascular dysplasia (as it is consistently decreased in dogs with shunts). Increasing concentrations of protein C can also be used to monitor portosystemic shunt ligation. Protein C is measured utilizing an activity assay, using a citrated plasma sample; the acute phase protein CRP can be measured in serum or plasma, depending on the assay used 25. 

Peritoneal effusions

Hypoalbuminemia can contribute to a decrease in oncotic pressure, whilst fibrotic changes in the liver can increase pressure within the hepatic vasculature and sinusoids. Thus, an animal with significant hepatic pathology can present with a peritoneal effusion (Figure 5), more often than not a transudative effusion. 

Conclusion

When considering possible liver disease in companion animals it is important to remember that systemic processes can cause secondary hepatic pathology, so the clinician should separately consider the evidence for injury or cholestasis and identify the predominant pattern. It is helpful to look for evidence of altered synthetic function, using glucose, urea, cholesterol, and albumin levels, and the CBC and urinalysis should always be part of the investigative work-up. Most importantly, since a biochemical profile is a snapshot of the body’s functioning at the point of sampling, temporal monitoring is critical for most cases.