Clinical biochemistry refers to the analysis of the blood plasma (or serum) for a wide variety of substances—substrates, enzymes, hormones, etc—and their use in diagnosis and monitoring of disease. Analysis of other body fluids (eg, urine, ascitic fluids, CSF) is also included. One test is very seldom specific to one clinical condition, and basic checklists of factors affecting the most commonly requested analytes are given below. Thus, rather than six tests that merely confirm or deny six possibilities, a well-chosen group of six tests can provide information pointing to a wide variety of different conditions by a process of pattern recognition. Biochemistry tests should be accompanied by full hematology, because evaluation of both together is essential for optimal recognition of many of the most characteristic disease patterns (see Clinical Hematology).
Before samples are collected, a list of differential diagnoses should already be established based on the history and clinical examination. Then, additional appropriate tests can be added to the basic panel below.
Making a diagnosis entails establishing a list of differential diagnoses based on the history and clinical examination. Based on this list, tests can be selected to include or exclude as many of the differentials as possible. Yet more tests may be necessary until only one of the original list remains to determine the diagnosis. If all differentials are excluded, then the list must be reevaluated. It is not good practice to order tests without a sensible differential list unless the animal presents without definitive clinical signs.
Most veterinary laboratories offer a basic panel of tests, which represents a minimal investigation applicable to most general situations. For small animals, a typical panel includes total protein, albumin, globulin (calculated as the difference between the first two analytes), urea, creatinine, ALT, and alkaline phosphatase (ALP). In addition, a yellow color seen in the plasma should be considered an indication to measure bilirubin. This panel may be modified as appropriate for other species, eg, glutamate dehydrogenase (GDH) and/or γ glutamyl transferase (γGT) are more appropriate “liver enzymes” for horses and farm animals, or it may be more appropriate to concentrate primarily on muscle enzymes (CK and AST) in athletic animals.
Total protein level increases due to dehydration, chronic inflammation, and paraproteinemia. It decreases due to overhydration, severe congestive heart failure (with edema), protein-losing nephropathy, protein-losing enteropathy, hemorrhage, burns, dietary protein deficiency, malabsorption, and some viral conditions (especially in horses).
Albumin level increases due to dehydration. It decreases due to the same factors as total protein, plus liver failure.
Urea level increases due to excess dietary protein, poor quality dietary protein, carbohydrate deficiency, catabolic states, dehydration, congestive heart failure, renal failure, blocked urethra, and ruptured bladder. It decreases due to low dietary protein, gross sepsis, anabolic hormonal effects, liver failure, portosystemic shunts (congenital or acquired), and inborn errors of urea cycle metabolism. Urea measurement is used especially to indicate renal disease and to a lesser extent liver dysfunction.
Creatinine level increases due to renal dysfunction, blocked urethra, and ruptured bladder. It decreases due to sample deterioration. Animals with a high muscle mass have high-normal creatinine concentrations, whereas animals with a low muscle mass have low-normal creatinine concentrations. Creatinine measurement is used especially for renal disease.
ALT is present in the cytoplasm and mitochondria of liver cells and, therefore, increases due to hepatocellular damage. It has a half life of 2–4 hr and rises higher than AST but recovers quicker. There are minor increases with muscle damage and hyperthyroidism.
ALP level increases due to increased bone deposition, liver damage, hyperthyroidism, biliary tract disease, intestinal damage, hyperadrenocorticism, corticosteroid administration, barbiturate administration, and generalized tissue damage (including neoplasia). The most common causes for an increase is raised levels of circulating steroids and biliary disease. The half-life is 72 hr in dogs but only 6 hr in cats. Levels in the cat are generally much lower than in the dog, and any increase in cats is considered significant. In dogs, ALP levels in the thousands of units are usually associated with increased steroid levels. ALP and ALT levels rarely rise above 1,000 units, even in severe liver disease.
GDH level increases in hepatocellular damage, particularly hepatic necrosis, in horses and ruminants.
γGT increases in longer-term liver damage; it is particularly useful in horses and ruminants.
CK, the classic “muscle enzyme,” increases markedly in rhabdomyolysis and aortic thromboembolism. Slight level increases are reported in hypothyroidism. Only a very small amount of muscle damage such as bruising or IM injections can result in high serum CK levels. In dogs and cats, unless investigating specific muscle disease, increased levels are generally of no clinical significance.
AST level increases in both muscle and liver damage but is of less value than ALT. The half-life is 5 hr in dogs and 77 min in cats. It is also reported to increase in hypothyroidism.
Most of the above parameters are associated with liver function/dysfunction and are frequently overinterpreted. In small animals, increases in ALT and ALP levels can reach four times normal and still be associated only with fatty change, a nonspecific finding and not, in most cases, a primary liver problem. Some laboratories also frequently receive liver biopsies from dogs that have significant increases in liver enzymes and bile acids >80 but that have normal histologic morphology. The reason for this is unknown.
In general, plasma enzyme levels decrease due to sample deterioration. Uncommonly, atrophy or fibrosis of an organ may result in unusually low plasma activities of the relevant enzymes.
Further tests may be added to the basic panel, according to the principal presenting signs, to create panels for polydipsic animals, collapsing animals, etc. These panels are structured so that the patterns of abnormalities typical of all the likely differential diagnoses applicable to the situation can be discerned. For example, a polydipsia panel may add calcium, glucose, and cholesterol. Calcium allows recognition of hyperparathyroidism and other causes of hypercalcemia (which causes polydipsia and renal insufficiency), glucose may indicate diabetes mellitus and contributes to the pattern characteristic of hyperadrenocorticism, and cholesterol also adds to the appreciation of the “Cushing pattern.” Renal failure is covered by the tests already included in the basic panel. In contrast, in a panel for a "collapsing animal,” calcium and glucose may be added to screen for hypocalcemia or hypoglycemia. Sodium and potassium are included to screen for hypoadrenocorticism or hypokalemia. Analytes that might be considered for incorporation in such expanded profiles are described below.
Sodium level increases due to Conn syndrome (hyperaldosteronism), restricted water intake, vomiting, and most causes of dehydration. It decreases due to hypoadrenocorticism, loss of any high-sodium fluid such as some forms of renal disease, and insufficient sodium provision during IV fluid therapy.
Potassium level increases due to hypoadrenocorticism and severe renal failure (especially terminal cases). It decreases due to Conn syndrome, chronic renal dysfunction, vomiting, diarrhea, and insufficient potassium provision during IV fluid therapy. Congenital hypokalemia occurs in Burmese cats.
Chloride level increases in acidosis, and in parallel with increases in sodium concentration. It decreases in alkalosis, vomiting (especially after eating), and in association with hyponatremia.
Total CO2 (bicarbonate) level increases in metabolic alkalosis and decreases in metabolic acidosis. It is less useful to assess respiratory acid/base disturbances.
Calcium level increases due to dehydration (which is also associated with increased albumin), primary hyperparathyroidism (neoplasia of parathyroid gland), primary pseudohyperparathyroidism (neoplasms producing parathormone-related peptide [PRP], usually perianal adenocarcinoma or some form of lymphosarcoma), bone invasion of malignant neoplasms, thyrotoxicosis (uncommon), and overtreatment of parturient paresis. It decreases due to hypoalbuminemia, parturient paresis, oxalate poisoning, chronic renal failure (secondary renal hyperparathyroidism), acute pancreatitis (occasionally), surgical interference with parathyroid glands, and idiopathic (autoimmune) hypoparathyroidism.
Phosphate level increases due to renal failure (secondary renal hyperparathyroidism). Decreases are seen in some downer cows and as part of the stress pattern in horses and small animals.
Magnesium level increases are rarely seen, including during acute renal failure. It decreases in ruminants due to dietary deficiency, either acute (grass staggers) or chronic, and diarrhea (uncommon).
Glucose level increases due to high-carbohydrate meals, sprint exercise, stress or excitement (including handling and sampling stress), glucocorticoid therapy, hyperadrenocorticism, overinfusion with glucose/dextrose-containing IV fluids, and diabetes mellitus. It decreases due to insulin overdose, insulinoma, islet cell hyperplasia (uncommon), acetonemia/pregnancy toxemia, acute febrile illness, and idiopathically (in certain dog breeds).
β-Hydroxybutyrate level increases in diabetes mellitus. It is a major component of ketoacidosis and as such is also increased in acetonemia/pregnancy toxemia and extreme starvation. It can be measured in both blood and urine.
Bilirubin level increases due to fasting (benign effect in horses and squirrel monkeys, may be caused by hepatic lipidosis in cats), hemolytic disease (usually mild increase), liver dysfunction, and biliary obstruction (intra- or extrahepatic). Theoretically, hemolysis is characterized by an increase in unconjugated (indirect) bilirubin, whereas hepatic and post-hepatic disorders are characterized by an increase in conjugated (direct) bilirubin; however, in practice this differentiation is unsatisfactory. Better appreciation of the source of the jaundice is gained from bile acid measurements.
Bile acid levels increase when hepatic anion transport is impaired, usually during liver dysfunction (bile acids are more sensitive than bilirubin to hepatic impairment) and in the presence of a portosystemic shunt (congenital or acquired). The latter condition is characterized by a marked increase in bile acid concentration after feeding, from a fasting concentration that may be normal. It also increases in bile duct obstruction; very little increase is seen in feline infectious peritonitis or mild cases of hepatic lipidosis. Very high levels can sometimes be seen without structural histologic changes. The reason for this is not known.
Cholesterol level increases due to fatty meals, hepatic or biliary disease, protein-losing nephropathy (and other protein-losing syndromes to some extent), diabetes mellitus, hyperadrenocorticism, and hypothyroidism. It decreases in some cases of severe liver dysfunction and occasionally in hyperthyroidism.
Lactate dehydrogenase is a ubiquitous enzyme with a number of isoenzymes; electrophoretic separation of isoenzymes is necessary to locate the source of increased activity. It is therefore of very limited value in general clinical practice.
Sorbitol dehydrogenase level increases in acute hepatocellular damage in horses but is a very labile analyte.
α-Amylase level increases in acute pancreatitis but in dogs is also increased in chronic renal dysfunction. It is therefore of limited use in the diagnosis of pancreatitis. Pancreatic lipase immunoreactivity is now the test of choice for diagnosis of pancreatitis in dogs and cats. Amylase is not a useful indicator of pancreatitis in cats.
Lipase level increases in acute pancreatitis in dogs (longer half-life than amylase) and also occasionally in chronic renal dysfunction. Lipase (routine assay) is not a useful indicator of pancreatitis in cats.
Immunoreactive trypsin (trypsin-like immunoreactivity) level decreases in exocrine pancreatic insufficiency in dogs. It will also increase (irregularly) in pancreatitis.
Serum amylase and lipase activities have been used for several decades to diagnose pancreatitis in both people and dogs. Unfortunately, neither of these tests is both sensitive and specific for pancreatitis in dogs. In one study, significant serum amylase and lipase activities remained after total pancreatectomy, indicating there are sources of serum amylase and lipase activity other than the exocrine pancreas. Also, clinical data suggest a specificity for pancreatitis of only ~50% for both of these markers. Many nonpancreatic diseases, such as renal, hepatic, intestinal, and neoplastic diseases, can lead to increases in serum amylase and lipase activities. Steroid administration can also increase serum lipase activity and cause variable responses in serum amylase activity. Thus, in dogs, measurement of serum amylase and lipase activities are of limited usefulness for diagnosis of pancreatitis. Serum amylase and/or lipase activities that are 3–5 times the upper limit of the reference range, in animals with clinical signs consistent with pancreatitis, are suggestive of such a diagnosis. However, it is important to note that ~50% of dogs that fulfill these criteria do not have pancreatitis. In cats, serum amylase and lipase activities are of no clinical value for diagnosis of pancreatitis. Although cats with experimental pancreatitis can show an increase in serum lipase activity and a decrease in serum amylase activity, these changes are not consistent in cats with spontaneous pancreatitis. In one study of 12 cats with severe forms of pancreatitis, not a single cat had serum lipase or amylase activity above the upper limit of the reference range.
Serum trypsin-like immunoreactivity (TLI) concentration measures mainly trypsinogen, the only form of trypsin circulating in the vascular space of healthy individuals. However, trypsin, if present in the serum, is also detected by these assays. Serum TLI concentrations can be measured by species-specific assays that have been developed and validated for both dogs and cats. In healthy animals, serum TLI is low, but during pancreatitis an increased amount of trypsinogen leaks into the vascular space, which can lead to an increase in serum TLI concentration. Trypsin that has been prematurely activated may also contribute to this increase. However, both trypsinogen and trypsin are quickly cleared by the kidneys. In addition, any prematurely activated trypsin is quickly removed by proteinase inhibitors, such as α1-proteinase inhibitor and α2-macroglobulin. In turn, α2-macroglobulin-trypsin complexes are removed by the reticuloendothelial system. Thus, the serum half-life for TLI is short, and a significant degree of active inflammation is required to increase serum TLI concentration. Because of the limited sensitivity of serum cTLI and fTLI concentrations for canine and feline pancreatitis, respectively, and because only a limited number of laboratories measure these assays routinely, serum TLI concentration is of limited usefulness for diagnosis of pancreatitis in dogs and cats.
Pancreatic lipase immunoreactivity (PLI) concentration measures the concentration of classical pancreatic lipase in the serum. This is in contrast to serum lipase activity, which measures the enzymatic activity of all triglyceridases present in the serum, regardless of their cellular origin. Assays to measure PLI in canine (cPLI) and feline (fPLI) serum have been developed and validated and are commercially available. Serum PLI is highly specific for exocrine pancreatic function. Also, serum PLI is far more sensitive for diagnosis of pancreatitis than any other diagnostic test currently available.
A patient-side semiquantitative assay for diagnosis of canine pancreatitis is also available. A test spot that is lighter in color than the reference spot suggests that pancreatitis can be excluded. A test spot darker in color than the reference spot raises the suspicion of pancreatitis and should prompt the clinician to measure a serum cPLI concentration in the laboratory.
Other tests for diagnosis of pancreatitis in dogs and cats have been evaluated, including plasma trypsinogen activation peptide (TAP) concentration, urine TAP concentration, urine TAP:creatinine ratio, serum α1-proteinase inhibitor trypsin complex concentration, and serum α2-macroglobulin concentration. However, none has been shown to be of clinical usefulness.
In the past, several fecal tests have been used to diagnose exocrine pancreatic insufficiency (EPI). Microscopic fecal examination for fat and/or undigested starch or muscle fibers are at best useful to suggest maldigestion. However, in light of wide availability of tests to diagnose EPI, microscopic fecal examination is no longer justified. Fecal proteolytic activity had been used to diagnose EPI in small animals for several decades. Most of these methods, particularly the radiographic film clearance test, are unreliable. One method, which uses pre-made tablets to pour a gelatin agar, is considered most reliable. However, false-positive as well as false-negative results have been reported. The clinical use of fecal proteolytic activity is limited to species for which more specific assays to estimate pancreatic function are not available and in areas where the more accurate and sophisticated tests are not available.
Serum TLI concentration is the diagnostic test of choice for EPI in both dogs and cats. Assays for TLI measure trypsinogen circulating in the vascular space. In healthy animals, only a small amount of trypsinogen is present in serum. However, in dogs and cats with EPI, the number of pancreatic acinar cells is severely decreased. Serum TLI concentration decreases significantly and may even be undetectable. The reference range for TLI in dogs is 5.7–45.2 mcg/L with a cut-off value of ≤2.5 mcg/L considered diagnostic for EPI. Similarly, the reference range for TLI in cats is 12–82 mcg/L, with a cut-off value of ≤8 mcg/L considered diagnostic. Rarely, animals with serum TLI concentrations below the cut-off value for EPI do not have clinical signs of EPI. This is probably because of the functional redundancy of the GI tract. At the same time, many dogs and cats with chronic diarrhea and weight loss have mild decreases in serum TLI concentration. Most of these animals have chronic small-intestinal disease and should be investigated accordingly. However, a small number of these dogs and cats may have EPI. If there is no evidence of small-intestinal disease in such patients, a trial therapy with pancreatic enzymes and reevaluation of serum TLI concentration after 1 mo is indicated.
PLI is also highly specific for exocrine pancreatic function and could be used to diagnose EPI. However, initial studies showed there is a small degree of overlap in serum PLI concentrations between healthy dogs and dogs with EPI, making the measurement of PLI slightly inferior to that of TLI for accurate diagnosis. In view of these findings, PLI assays for both dogs and cats have been optimized toward higher concentrations, and the current assays are no longer suitable for diagnosis of EPI in dogs or cats.
A fecal canine elastase concentration assay has been developed and validated but is inferior to the widely used TLI measurement.
Most biochemistry tests can be performed on either serum or heparinized plasma. A few (eg, insulin) require serum, whereas potassium is best measured on heparin plasma separated immediately after collection. Glucose measurement requires fluoride/oxalate plasma. Suitable collection tubes with and without anticoagulant are available commercially. Plastic tubes are satisfactory for blood in anticoagulant, but clotted blood must be collected either into glass tubes or plastic tubes specially coated to prevent the clot from adhering to the vessel walls.
Samples for biochemical analysis should be separated as soon as possible after collection to minimize artifacts caused by hemolysis and leakage of intracellular fluid components (eg, potassium) out of the cells. Samples in anticoagulant may be centrifuged immediately, but clotted samples need at least 30 min to allow the clot to form. Fluoride/oxalate samples hemolyze very readily because the cells can no longer respire, so timely separation is especially important. Proprietary gels or plastic beads assist with separation, and these may be incorporated into the collection tube or added before centrifugation.
Larger bucket-type centrifuges will accept almost any type or size of tube, but the rotors require careful balancing. They should be spun at 3,000 rpm for 10 min. Dual-purpose, high-speed microhematocrit centrifuges are favored for in-practice use, because they separate samples more quickly and the same machine can be used to measure PCV. However, they can handle only a limited range of small-volume tubes.
Some “gel-tube” products may provide a permanent separation of serum or plasma; otherwise, this must be separated into a fresh tube. The new tube must be adequately labeled. Samples may then be sent to a professional laboratory or analyzed in the practice.
A number of biochemical analytes may be estimated in the practice without the need for large analytical instruments.
Total protein level is measured by refractometry, using the same instrument as is used to measure urine specific gravity, provided the instrument has a total protein scale. It is also valid for protein measurement of ascitic and pleural fluids. The readout may be in g/dL, in which case multiplying the result by 10 will yield the SI unit of g/L.
Urea level may be estimated by chromatographic reaction strips, which correlate well with standard laboratory methods. A rapid whole-blood color comparison strip is also available, but these read only up to ~20 mmol/L and are thus of limited use. A dedicated reflectance meter for urea estimation is not available.
Glucose meters for use on whole blood are widely available for home use by human diabetic patients. These yield acceptably accurate results on animal blood, although an unexpected hypoglycemia should be confirmed by a professional laboratory. Fresh whole blood may be used, but fluoride blood or plasma is the preferred sample if analysis is not immediate.
Ketone levels may be estimated on either urine (preferred sample) or plasma/serum. This can be achieved by using the ketone patch of a urine dipstick, giving a qualitative result. However, there are a number of point-of-care instruments for measurement of blood glucose and ketone levels, including specifically β-hydroxybutyrate.
Triglyceride levels may be visualized in a plasma or serum sample as lipemia. If the milkiness rises to the top of the tube on storage, chylomicrons are present. Otherwise, the milkiness is caused by triglycerides. This is a qualitative judgment but is nevertheless useful, especially in equine patients.
Bilirubin level may also be appreciated by eye in most species. Equine and bovine plasma is normally yellow, which makes determination problematic, but in other species, any yellow color is abnormal and indicates an increased bilirubin level. Visual assessment of the depth and shade of color may provide additional information.
Other point-of-care tests include C-reactive protein as a marker for inflammation and cardiac troponin as a marker for cardiac muscle damage.
For emergency in-clinic use, the most important analytes beyond these simple basics are sodium and potassium. A dedicated ion-specific electrode meter is the best way to measure these. Instruments are available that can analyze whole blood, although great care must be taken to avoid artifacts due to unappreciated hemolysis. Critical care meters are also available that can estimate a variety of analytes, including glucose, urea, and electrolytes; however, these have not been extensively validated on nonhuman blood, and results should be interpreted with caution.
Extending in-practice analysis beyond these emergency basics requires a dedicated instrument capable of measuring multiple analytes. Two types are available—those based on transmission/absorbance photometry (wet chemistry) and those based on reflectance photometry (dry-reagent chemistry). Transmission/absorbance photometry is the reference method on which all reference values and interpretive guidelines are based. Reflectance photometry methods do not always compare well with the reference method and are best confined to simpler tests such as glucose and urea. For wider applications, such as enzyme analysis, wet chemistry instruments are preferred.
In-clinic analysis is inevitably more expensive than the same investigations done by a professional laboratory, and the range of analytes available is more restricted. Additionally, the level of accuracy or reliability is likely to be lower. Therefore, it is still best practice to regard in-practice analysis as an interim emergency investigation, with the results to be confirmed as appropriate by a professional laboratory. Detailed case laboratory evaluation of nonemergency patients is best referred to a professional laboratory from the outset, for reasons of cost, accuracy, range of analytes available, and the assistance of the clinical pathologist in interpretation of the results.
If in-clinic analysis is to be relied on for general case laboratory evaluation, meticulous attention must be paid to quality assurance. Samples of known composition must be run at least daily for each analyte, in both normal and pathologic ranges; unless these are within the tolerance limits, no patient samples should be tested. Participation in an external quality assessment program is also strongly recommended. Employing a trained technician will address some of these issues but has implications for availability of results during off hours and holidays. The veterinarian in charge of the laboratory is responsible for all the results issued and incurs a legal liability to prove accuracy and reliability. If these cannot be guaranteed to the same standard as a referral laboratory, then results should not be relied on without external confirmation.