Aspartate aminotransferase (AST), also known as serum glutamate-oxaloacetate transaminase (SGOT), and alanine aminotransferase (AST), also known as serum glutamate-pyruvate transaminase (SGPT), are important enzymes commonly measured to identify and follow-up liver damage. Aminotransferases are key enzymes in amino acid metabolism, and they work by transferring amino groups between the vari- ous amino acids and 2-oxoglutarate (α-ketoglutarate), which functions as an acceptor of the amino group to yield glutamate in a reaction involving pyridoxal-50- phosphate (P5P), derived from vitamin B6, as a coen- zyme. The reaction is reversible, and glutamate can function as an amine donor to various ketoacids to form amino acids—for example, alanine from pyruvate with AST and aspartate from oxaloacetate with ALT.
The AST enzyme is coded by two different genes, GOT1 and GOT2. The GOT1 gene codes for c-AST, which is present in the cytosol and most abundant in cardiac and skeletal muscle and red blood cells, whereas the GOT2 gene codes for m-AST, a mitochon- drial isoform of AST most abundant in hepatocytes.
The catalytic activities of the two AST isoforms are similar, but they can be differentiated by immunoas- says or by modified enzymatic assays because m-AST is highly susceptible to proteolysis by proteinases K.
The m-AST isoform has a longer half-life in serum (B87 hr) than does c-AST. Whereas c-AST predomi- nates in the plasma of healthy individuals (B8895%
of the AST activity) and patients with mild liver dis- ease, m-AST is predominant in liver parenchyma (B80% of total AST) and increases more than c-AST in acute hepatitis and more severe liver disease [11]. It is likely that the cytoplasmic location allows c-AST to leak into the systemic circulation with mild, reversible cellular damage, whereas the more abundant m-AST requires injury to the mitochondria (e.g., associated with liver ischemia or severe necrosis). In alcoholic hepatitis, the ratio of m-AST to c-AST is also typically elevated, perhaps indicating a particular susceptibility
132 9. CHALLENGES IN ROUTINE CLINICAL CHEMISTRY ANALYSIS
of mitochondria to ethanol-induced damage, whereas in uncomplicated viral hepatitis the predominant lesion is in the cell membrane, and therefore c-AST predominates.
The GPT gene codes for ALT1, the cytosolic isoform of ALT (c-ALT), whereas the related gene GPT2 codes for the mitochondrial isoform, ALT2 (m-ALT). ALT1 is expressed mainly in liver and kidney, and it is weakly expressed in intestine, heart, skeletal muscle, and sali- vary glands. It is the predominant isoform in serum from healthy individuals, whereas ALT2 is the pre- dominant ALT isoform expressed in skeletal muscle and heart[12,13]. In rodents, ALT2 appears to be pres- ent in liver and seems to undergo a more significant increase than ALT1 after liver damage or steatosis [14,15], but it does not appear to be expressed in human liver [12,13,16]. ALT2 is also expressed in adrenal gland, adipose tissue, neurons, endocrine pancreas, and prostate, where it is induced by andro- gens[13]. ALT1, but not ALT2, is inducible in the liver by peroxisome proliferator-activated receptor (PPAR) αagonists, such as fenofibrate[17]. A similar induction of ALT activity is seen in conditions associated with increased neoglucogenesis, such as fasting, insulin resistance, type 2 diabetes, and the metabolic syn- drome[18,19]. Obesity can increase levels of AST and ALT by approximately 4050%. It is unclear whether the available enzymatic assays show any differential reactivity with ALT1 versus ALT2, although recombi- nant purified ALT1 had 15-fold higher activity than ALT2 in one study[20]. The different cellular localiza- tion, tissue specificity, and inducibility of the two ALT isoforms raise the possibility that isoform-specific immunoassays may have better clinical usability than the currently used enzymatic assay. Despite the pres- ence in other tissues, isolated elevations of ALT are rarely seen in the absence of liver disease.
Both AST and ALT are sensitive markers of liver damage, with ALT being more specific due to its restricted tissue expression and AST being more sensi- tive because it is approximately two or three times more abundant in liver than ALT. The ratio of AST/
ALT (de Ritis ratio) is commonly used to help differen- tiate various causes of liver damage (Table 9.1).
Various parameters affect the ratio:
1. Extent of extra-liver disease: Because AST is widely expressed and ALT is predominantly from liver, any conditions involving systemic injury, especially involving skeletal muscles, heart, kidney, lung, pancreas, intestine, or erythrocytes, will increase the ratio. Hemolysis occurringin vitrowill also increase AST with little effect on ALT.
2. Severity of damage: In normal individuals and patients with reversible injury involving the cellular
membranes, thought to occur through cytoplasmic blebbing, c-AST and c-ALT predominate in roughly equal amounts. Severe injury to mitochondria, such as in liver necrosis or degeneration due to drug toxicity or ischemia, leads to a massive release of the much more abundant m-AST, which increases the de Ritis ratio.
3. Elimination of ALT compared to AST: In general, the half-life of ALT is much higher (B2 days) than that of AST (,24 hr); therefore, although AST may predominate during an acute episode of toxic or TABLE 9.1 Conditions Affecting AST and ALT Levels Condition Level of AST AST/ALT
Ratio
Comment
Healthy ,40 U/L 0.71.4
Acetaminophen toxicity
.10,000 U/L .1a Slower increase in 2448 hr Ischemic injury .10,000 U/L .1a Rapid rise in,24
hr Alcoholic
hepatitis
,300 U/L .2
Autoimmune hepatitis
,103ULN ,2
Biliary obstruction
,53ULN ,2 Return to normal
in 72 hr Cholangitis ,103ULN ,2
Cirrhosis ,103ULN .1 Levels may be
normal or decreased Hemochromatosis ,103ULN ,2
Hepatitis, viral, acute
.300 U/L, 101003ULN
,1 AST remains
elevated for 714 days; ALT decays B10%/day, remains elevated 27616 days Hypo- and
hyperthyroidism
,33ULN ,2
Infectious mononucleosis
50800 U/L ,2
Liver metastases ,53ULN .1 Nonalcoholic
steatosis
,103ULN ,2
Primary biliary cirrhosis
,103ULN .2
Reye’s syndrome .2
Wilson’s disease ,103ULN .2
aAST/ALT.1 initially, then reverts to,1.
ULN, upper limit of reference range.
ischemic liver injury, during the recovery phase the ratio may become less than 1. Diseases leading to m-AST release also prolong the elevation of AST activity because m-AST has a longer half-life than c-AST.
4. The extent of liver synthetic function affects ALT more than AST because ALT is more liver-specific.
Situations with reduced liver function, such as in chronic cirrhosis, tend to have elevated de Ritis ratios due to lower ALT production. In some cases of chronic liver disease, as well as in fulminant liver failure, ALT concentrations may decrease below the reference range.
5. Any induction of ALT expression—for example, treatment with PPARαagonists or insulin resistance: Obese patients and those with the metabolic syndrome tend to have higher levels of ALT (4050%) compared to AST[21].
6. The amount of P5P in circulation: Using assays that do not supplement the reaction with exogenous P5P, this becomes a rate-limiting factor, with the ALT enzyme being more susceptible to low levels of P5P seen, for example, with chronic alcoholism, malnutrition, and renal failure, the latter due to P5P binders in the plasma. Also, m-AST is more
susceptible to loss of P5P than is c-AST, and assays using P5P supplementation tend to have higher m-AST/c-AST ratios.
Typically, patients with alcoholic hepatitis or severe alcohol-induced liver disease such as cirrhosis have a de Ritis ratio greater than 2[22], whereas in uncompli- cated acute or chronic liver disease the ratio is usually less than 2. Possible mechanisms involved in elevation of the ratio in severe alcoholic liver disease include mitochondrial oxidative damage, decreased liver ALT production, decreased P5P, and extrahepatic damage such as alcoholic myopathy. The ratio is also helpful in distinguishing alcoholic from non-alcoholic liver stea- tosis, especially when coupled with other determina- tions such as red cell mean corpuscular volume and γ-glutamyl transferase[23], although the ratio tends to be high in advanced cases of alcoholic and non- alcoholic cirrhosis[24]. In contrast, acute viral hepatitis tends to be associated with de Ritis ratios less than 1 and aminotransferase levels greater than 300 IU/L.
Patients with a de Ritis ratio less than 0.6 in acute viral hepatitis had better prognoses than those with ratios greater than 1.2[25].
Many drugs affect AST and ALT levels, and monitor- ing of transaminase elevation is routine when evaluat- ing or following therapy with potentially hepatotoxic drugs. Some drugs also increase ALT and AST levels by inducing its expression. Commonly used drugs that may affect ALT and AST levels include acetaminophen,
aminoglycosides, angiotensin-converting enzyme (ACE) inhibitors, anticonvulsants, cephalosporins, clofibrate, fenofibrate, fluconazole, ganciclovir, griseo- fulvin, isoniazid, isoniazid, macrolides, nicotinic acid, nonsteroidal anti-inflammatory drugs, omeprazole, opi- ates, statins, sulfonamides, and sulfonylureas.
The following are additional patient-related factors to take into consideration when evaluating ALT or AST measurements:
1. ALT and, more important, AST levels in normal individuals tend to parallel muscle mass and therefore are 1020% higher in males, African Americans, and athletes.
2. Intense exercise can increase AST and ALT levels, usually less than threefold, subsiding within 17 days[26].
3. Both hypothyroidism and hyperthyroidism tend to increase AST levels.
4. Levels of AST and ALT increase with age until approximately 40 years and then decline
approximately 1020% to the age of 60 years. Both AST and ALT tend to be higher in elderly
individuals—approximately 10% after age 70 years[27].
5. Circadian variation affects ALT by approximately 45%, peaking in the afternoon and a nadir at night, whereas AST varies by less than 10%, although some studies show no significant variation[27].
6. Both AST and ALT are elevated in obese patients (4050% higher) and those with
hypertriglyceridemia or high-sugar or high-fat diets [27,28].
7. Hospitalization by itself and even placebo treatment may be associated with mild elevations in
transaminases, particularly ALT[29,30].
CASE REPORT A 32-year-old female was referred to a hepatologist for possible liver biopsy[31]. She had a history of AST levels between 120 and 500 U/L per- sisting for more than 7 years without any symptoms, family history, or significant past medical history of liver disease, except for a limited acute EpsteinBarr virus infection at age 14 years, which caused a revers- ible 10-fold elevation of ALT and AST. Medical exami- nation was normal, an abdominal ultrasound was unremarkable, and serum levels of other liver enzymes including ALT, hepatitis and autoimmune serology, ferritin, ceruloplasmin, and thyrotropin were normal.
The patient denied taking any medications or herbal supplements, except for oral contraceptives.
Discontinuation of the contraceptive resulted in no change in AST levels. Before performing the liver biopsy, the patient’s serum sample was treated with protein A Sepharose, which binds immune
134 9. CHALLENGES IN ROUTINE CLINICAL CHEMISTRY ANALYSIS
complexes, resulting in a change in AST activity from 459 to 57 U/L, whereas another healthy control sample showed no change. The elevation in AST was attrib- uted to macro-AST, and the patient was subject to no further studies.
Although rarely seen, macro-AST is more common in patients with liver disease, particularly those with autoimmune, viral, or neoplastic liver disease, but it can be found in healthy individuals and persist for several years[31,32]. Isolated increases in AST without ALT or creatine kinase elevations suggest macro-AST, which can be eliminated by polyethylene glycol pre- cipitation, ultrafiltration, dialysis, and exclusion or affinity chromatography.
Specimen Processing
Serum or plasma should be separated from cells to avoid leakage of enzymes from red blood cells, which increases markedly after 24 hr at room temperature.
The ALT and AST activities tend to decay when speci- mens are stored at room temperature, particularly in specimens with low P5P levels, because apoenzymes are more susceptible to degradation than are holoen- zymes [33]. Because the enzymes are irreversibly degraded, lower activities may be seen even with methods using P5P supplementation. However, several studies show only minimal decay of the enzymes at room temperature, when refrigerated, or when frozen for at least 3 days [27,3436]. Repeated freezing and thawing may decrease ALT activity.
Methodology
Transaminase levels are measured by supplying 2- oxoglutarate and the specific amino acid (L-aspartate for AST and L-alanine for ALT), which generate glu- tamate and the respective ketoacids. In some methods (e.g., Vitros), P5P is supplied to enhance the activity and remove circulating P5P levels as an analytical vari- able. A signal is generated by dehydrogenases specific to the ketoacid generated in the transaminase reaction.
For AST, the oxaloacetate is converted to malate by malate dehydrogenase, with generation of NAD1, whereas the pyruvate resulting from the ALT reaction is converted to lactate by lactate dehydrogenase (LDH). Because the method uses serum blanking and all the components are in vast excess to usual plasma levels of the corresponding endogenous substances, interference by pyruvate, lactate, and LDH is uncom- mon. In rare cases with very high levels of LDH (.12,000 U/L), the preincubation step during serum blanking may consume all the NADH substrate and cause a very low result or an analyzer flag. A 20-fold
dilution of the sample will mitigate this problem. In the Beckman SYNCHRON assay, pyruvate levels of less than 6 mg/dL (normal levels are,1 mg/dL) show minimal interference. However, pyruvate can be generated from lactate and cause interference with ALT measurements in samples in which plasma was in contact with cells for more than 48 hr, especially in samples with high LDH levels, which converts lactate to pyruvate [36]. Other interferences include heparin greater than 50 U/L with the Vitros method [37] and also significant lipemia, hyperbilirubinemia, and hemolysis (particularly AST, which is 1030 times more abundant in red cells than in plasma)[27].
γ -GLUTAMYL TRANSFERASE AND ALKALINE PHOSPHATASE ANALYSIS
γ-Glutamyl transferase (GGT) and alkaline phospha- tase (ALP) are often measured to determine if a patient has intra- or extrahepatic cholestasis because these enzymes are abundant in the plasma membrane of hepatocytes, particularly in the biliary canaliculi and the basolateral surface of biliary epithelia. ALP is the product of one of three genes—ALPI, expressed in the intestine; ALPP, expressed in placenta; and ALPL, expressed in most tissues, particularly in liver, kidney, and bone, in which it can be differentially glycosy- lated, resulting in changes in protein conformation, stability, and clearance. For example, ALPP is heat stable at 65C, the bone form of ALPL, derived from osteoblasts, is extensively glycosylated and heat labile, and liver ALP is heat stable at 56C. Heterogeneity of glycosylation affects clearance of the various forms of ALP by the liver galactose receptor[38].
Most plasma ALP activity derives from liver, with up to 50% from bone and a small amount of intestinal form. The release of ALP into the circulation during cholestatic disease may be mediated by increased phospholipase activity induced by bile acids because the enzyme is anchored to the membrane by a glycosyl phosphatidyl inositol (GPI) moiety [39]. The various isoforms of ALP can be separated by electrophoresis, isoelectric focusing, wheat germ lectin precipitation, high-performance liquid chromatography, or specific immunoassays; however, it is simpler to determine GGT levels because this enzyme is increased in liver but not bone disease. The liver form of ALPL is ele- vated in cholestatic disease (up to 12 times normal), particularly in extrahepatic obstruction, whereas parenchymal liver disease, such as viral, alcoholic, or autoimmune hepatitis, usually increases ALP less than 3-fold. Liver metastases can also result in high levels of ALP, depending on the extent of the infiltration.
The bone ALPL is elevated in growing children and
in disorders with osteoblastic activity and bone remo- deling, such as Paget’s disease (up to 25-fold), rickets and osteomalacia (2- to 4-fold), and in osteogenic bone cancer (very high levels). Elevated ALPP is seen in pregnancy (2- or 3-fold) and some tumors (Regan iso- enzymes), and ALPI is increased in liver cirrhosis and blood group O or B secretors.
GGT is abundantly expressed in kidney, liver, pan- creas, and intestine, but most of the circulating activity in healthy individuals is of hepatic origin. Like ALP, GGT is cleared by the liver, and various glycosylation isoforms are cleared at different rates by the liver asia- loglycoprotein receptor [40]. Interestingly, GGT is inducible in the liver by a variety of xenobiotic agents, most notably microsomal inducers such as the anti- convulsants phenytoin and phenobarbital, oral contra- ceptives, cimetidine, furosemide, and methotrexate, and in patients with chronic ethanol exposure, who can have hepatic GGT levels approximately 3-fold higher than normal [41]. Chronic alcoholism may also be associated with increased release of GGT into the circulation because serum levels can be as high as 530 times the upper limit of normal. In general, it is thought that GGT levels reflect oxidative stress and glutathione consumption in the liver, which makes sense given the role of GGT in the metabolism of glu- tathione, a key antioxidative agent in the liver. The same liver and biliary diseases that cause increased ALP will also cause GGT elevations to a similar extent, although GGT tends to be more sensitive than ALP to cholestasis, with elevations averaging 12-fold in greater than 93% of cases compared to 3-fold for ALP [42]. Elevations of GGT originating from other organs can be seen in acute and chronic pancreatitis or pan- creatic cancer and mildly in chronic kidney disease, diabetes, acute myocardial infarction, hyperthyroid- ism, rheumatoid arthritis, and lung disease[42,43].
The following additional pre-analytical factors should be taken into consideration when interpreting ALP and GGT measurements[27,34,42]:
• Normal levels of ALP increase in growing children with a peak approximately fivefold the adult levels at approximately age 13 years, mostly due to the bone isoenzyme. The levels then decay until age 25 years when they reach relatively stable levels, similar in men and women. GGT levels increase linearly with age until approximately 4050 years and are approximately 4050% higher in men than in women.
• In African Americans, GGT activity is
approximately twice that of Caucasians, whereas ALP is 1015% higher.
• Obesity increases ALP and GGT by approximately 2030%, and in individuals with a body mass index
greater than 30 m2, levels of GGT may be 50%
higher.
• Diurnal variation of less than 10% in ALP and GGT levels has been reported with activities highest in the morning and lowest in the afternoon.
• Ingestion of fatty meals can increase intestinal ALP as much as 30 U/L for approximately 12 hr,
especially in Lewis-positive secretors of blood types O or B.
• Starvation can decrease GGT, whereas meals cause an initial decrease followed by a small increase.
• Smoking increases GGT by 1050% and ALP by approximately 10%.
• Oral contraceptives increase GGT and decrease ALP by approximately 20%, whereas pregnancy
increases ALP 200300% and decreases GGT by approximately 2030%.
• ALP and GGT activities are generally stable in separated serum or plasma at room temperature, 4C or frozen for at least 7 days. Frozen serum or plasma should be thawed for at least 18 hours at room temperature to achieve full ALP enzyme reactivation.
Analytical Issues
Most commercially available methods for ALP anal- ysis use the chromogenic substrate 4-nitrophenyl phos- phate, which upon loss of the phosphate group catalyzed by ALP under alkaline conditions forms a yellow product that absorbs at 405410 nm. Massive blood transfusion, plasmapheresis, or citrate anticoag- ulant inhibits ALP activity because citrate chelates the required zinc and magnesium ion co-factors. Citrate, oxalate, and fluoride also inhibit GGT activity up to 15% and should be avoided. The GGT assay also uses a chromogenic substrate such asγ-glutamyl-p-nitroani- line, which is cleaved by GGT into p-nitroaniline, mea- sured at 405120 nm, and a glutamyl group that is transferred to an acceptor such as glycylglycine.
Hemolysis and heparin may interfere with some ALP and GGT methods, whereas others show no such interferences. In patients with hepatobiliary dis- eases, macro-GGT may occur as complexes with IgA autoantibodies and apolipoproteins A and B [44].
Binding to lipoproteins may increase GGT half-life from approximately 9 to 20 hr [40]. IgM paraproteins in heparinized plasma may also interfere with some GGT assays [45]. Macro-ALP has also been rarely observed, without clear clinical significance, although increased prevalence in patients with inflammatory bowel disease and chronic peritoneal dialysis has been suggested [46,47].
136 9. CHALLENGES IN ROUTINE CLINICAL CHEMISTRY ANALYSIS