View the Disease Management Project
Table of Contents

Revised September 2, 2003

William D.
Carey, MD

Department of
Gastroenterology
and Hepatology

Print Chapter

Laboratory assessment of the patient with suspected or clinically obvious liver disease is context-dependent. For instance, the acutely ill jaundiced patient with a history of prolonged alcohol ingestion requires a different laboratory assessment than the well patient in whom one or more standard liver tests are discovered to be abnormal during routine testing. Additionally, the sequence of liver tests depends heavily on the question asked. If the question is: Does this well person whose brother was recently diagnosed with hemochromatosis also have this genetic disease?, then a series of tests will be initiated to detect iron overload. When the question is: Has this spouse been infected with hepatitis B?, then blood tests related to hepatitis B will be required. For these reasons, algorithms for evaluation of liver disease need to be considered skeptically.

This chapter is designed to discuss a useful way of thinking about liver tests. It emphasizes limitations of and alternative explanations for isolated abnormalities of common liver tests. Information in this chapter should be combined with discussions of specific liver diseases elsewhere in this Disease Management Project. A final caveat relates to terminology. Tests done in clinical laboratories do not measure any functional capacity of the liver. Hence, the commonly used term liver function tests is inaccurate, and the term liver tests will be used in this chapter. A guideline on interpretation and evaluation of abnormal liver tests has recently been published.^1,2 Useful algorithms are presented which parallel the recommendations in this chapter.^1,2

Isolated
Abnormalities
in Liver Tests

Evaluation of Liver Disease Based on Enzymes

Bilirubin
Elevations

Acute Alcoholic Hepatitis

Viral Hepatitis

Viral Hepatitis B

Resolved Hepatitis B and Immunization Status

Viral Hepatitis C

Iron and
Copper Tests

Autoimmune
Liver Diseases

References

A common clinical scenario is the unanticipated discovery of an abnormal liver test, obtained when a bundle of tests have been done for other reasons. Most clinical laboratories offer bundled blood tests, which often contain all or most of the following:

• bilirubin
• aspartate transaminase ([AST], formerly referred to as serum glutamic-oxaloacetic transaminase [SGOT])
• alanine transaminase ([ALT], formerly called serum glutamic-pyruvic transaminase [SGPT])
• gamma-glutamyl-transpeptidase (GGTP)
• alkaline phosphatase
• lactate dehydrogenase (LDH)

An isolated elevation of just one test result should raise suspicion that a source other than the liver is the cause. Table 1 indicates non- hepatic sources of elevated values for certain tests commonly considered as liver tests. When several liver tests are simultaneously out of the normal range, consideration of non- hepatic sources becomes irrelevant.

Special note should be made of the gamma glutamyl transpeptidase (GGTP) and lactase dehydrogenase (LDH) as liver tests. The GGTP is too sensitive (frequently elevated when no liver disease is apparent). The only utility of the GGTP is that it confers liver-specificity to an elevated alkaline phosphatase. An isolated elevation of GGTP does not need to be further evaluated unless there are additional clinical risk factors for liver disease.³ The LDH is insensitive and nonspecific because it is present in tissues throughout the body.

The normal range for aminotransferases in most clinical laboratories is much lower than that for alkaline phosphatase. Accordingly, when considering levels of elevations, it is necessary to consider them relative to the respective upper limit of normal for each test compared. Consider a patient with an AST of 120 IU/ml (normal up to 40) and an alkaline phosphatase of 130 IU/ml (normal up to 120). This represents a hepatocellular pattern of liver injury because the AST is three times the upper limit of normal, whereas the alkaline phosphatase is only marginally higher than its upper limit of normal.

Serum aminotransferases—ALT and AST—are two of the most useful measures of liver cell injury, although the AST is less liver-specific than the ALT. Elevations of AST may also be seen in acute muscle injury (either cardiac or skeletal muscle). Lesser degrees of ALT elevation may occasionally be seen in skeletal muscle injury or even after vigorous exercise. Diseases that primarily affect hepatocytes, such as viral hepatitis, will cause disproportionate elevations of AST and ALT compared to alkaline phosphatase. The ratio of AST/ALT is of little benefit in sorting out the etiology of liver injury except in acute alcoholic hepatitis, where the ratio is usually greater than 2 and the AST is 400 IU/mL or less.

Mild elevations of AST (<2x) the upper limit of normal are common. In part, this is explained by the way normal values are calculated, where normal is defined as the mean of the distribution ± 2 standard deviations. By this definition, 2.5% of normals will have values above the normal range.² Common causes of mild increases in AST and ALT include reduction effect (eg, status) and fatty liver disease seen most often in those with obesity, diabetes or elevated blood lipids. Fatty liver is also seen on those who drink alcohol.

Serum alkaline phosphatase is comprised of a heterogeneous group of enzymes. Hepatic alkaline phosphatase is most densely represented near the canalicular membrane of the hepatocyte. Accordingly, diseases that predominately affect hepatocyte secretion (eg, obstructive diseases) will be accompanied by elevations of alkaline phosphatase. Bile duct obstruction, primary sclerosing cholangitis, and primary biliary cirrhosis are some examples of diseases in which elevated alkaline phosphatase levels are often predominant over transaminase elevations (Table 2).

It is apparent that infiltrative liver diseases most often result in a pattern of liver test abnormality similar to that of cholestatic liver disease. Differentiation often requires imaging studies of the liver. Liver imaging by ultrasound, computerized tomography, or magnetic resonance imaging most often identifies infiltration of the liver by mass lesions such as tumors. Imaging by cholangiography (by endoscopic retrograde cholangiography, transhepatic cholangiography, or magnetic resonance cholangiography) identifies many bile duct lesions causing cholestatic liver disease. Liver biopsy is often needed to confirm certain infiltrative disorders (eg, amyloidosis) and microscopic biliary disorders such as primary biliary cirrhosis.

Most laboratories report only total bilirubin levels (the sum of the conjugated, and unconjugated portions). It is sometimes useful to determine the fraction of total serum bilirubin that is unconjugated versus conjugated. This is usually referred to as fractionation of bilirubin. The main clinical situation in which this is useful is when all of the standard liver tests are normal except the total bilirubin. Laboratories report the total bilirubin as well as the conjugated and unconjugated portions. To make matters more confusing, the conjugated bilirubin is sometimes referred to as the direct- reacting bilirubin, and the unconjugated as the indirect- reacting bilirubin (Table 3).

Normally, 90% or more of measured serum bilirubin is unconjugated (indirect-reacting). When the total bilirubin level is elevated and fractionation shows that the major portion (90% or more) unconjugated, liver disease is never the explanation. Instead, the clinical suspicion should turn to one of two explanations. If the patient is young and healthy, an inherited decrease in inability to conjugate bilirubin is liklely. This is referred to as Gilbert's syndrome. It causes no symptoms, and is associated with no liver disease. Interestingly, fasting and intercurrent illnesses such as influenza, often makes the level of unconjugated bilirubin even higher in those with Gilbert's syndrome. This syndrome is easily diagnosed when all the standard liver tests are normal, and 90% or more of the total bilirubin is unconjugated. There is no need for an imaging study or for liver biopsy in cases of suspected Gilbert's syndrome.

Elevations of unconjugated bilirubin, when the conjugated bilirubin remains normal may also indicate an increased load of bilirubin due to hemolysis. Anemia and an elevated reticulocyte count is usually present in such cases (Table 4).

Many mistakenly interpret elevations of direct-reacting bilirubin to indicate that cholestatic (obstructive) liver disease is present. It is apparent from Table 2 that the serum bilirubin level plays no useful role in categorizing a case as hepatocellular, cholestatic, or infiltrative. The bilirubin level may be normal or elevated in each type of disorder. Viral hepatitis A (a prototypical hepatocellular disease) may frequently be associated with bilirubin levels that are quite high, whereas primary biliary cirrhosis (a prototypical cholestatic disorder) is associated with a normal serum bilirubin level except in later-stage disease. Serum bilirubin levels should be disregarded when trying to decide if the liver test pattern is more suggestive of hepatocellular or cholestatic disease.

Viral hepatitis most often produces a hepatocellular pattern of injury (AST and ALT elevations predominate). Individuals with no symptoms and normal aminotransferase levels may be infected. In addition, a great deal of confusion is caused by abnormal viral markers, many of which do not indicate active infection but rather immunity. These concepts are more fully developed in the chapters on viral hepatitis. A clinical practice guideline on viral hepatitis is also available.⁴

Hepatitis A:

Hepatitis A is an acute self-limited disease in most, although it may rarely be fatal. Diagnosis is made through the use of antibody tests (anti HAV). The standard screening tests contain reagents, which will test positive for the presence of either IgM anti HAV or IgG anti-HAV. This test will, therefore, be positive with any exposure (acute, remote, or via immunization). Thus, it is not useful to determine of a patient with acute hepatitis has hepatitis A.

Because IgM anti-HAV is present for only a few months after acute infection, the key to diagnosis of acute hepatitis A is to measure IgM anti-HAV. Selective testing of IgM anti-HAV is required to establish the serologic diagnosis of acute hepatitis A (Table 5).

Acute Hepatitis B:

Within 2 weeks of exposure (but often delayed for weeks or months), hepatitis B surface antigen (HBsAg) emerges. This antigen is present in the blood for a variable period, usually encompassing the time during which the patient is clinically ill and most likely to seek medical attention. For patients with mild symptoms whose testing may be delayed, the HBsAg levels may have already declined. In this case, a second chance to make the diagnosis comes from detection of IgM antibody directed against the hepatitis B core antigen (anti-HBc-IgM (Table 6).

Chronic Hepatitis B:

Chronic hepatitis B is characterized by a long-lasting persistence of HBsAg and anti-HBc (IgG). Anti-HBs is absent. An additional antigen-antibody system requires mention, the hepatitis B e antigen (HBeAg) and its antibody (anti-HBe). These tests are only relevant in the individual in whom the HBsAg is chronically positive. In chronic B infection in which HBeAg is also positive, there is usually active viral replication and significant liver injury. In time, HBeAg may be lost, replaced by its antibody, anti-HBe. This transformation is often associated with lower-level infection (less viral replication) or HBV DNA, lower AST and ALT values, and less (or no) hepatic inflammation. These concepts and exceptions are discussed more fully in the chapter on viral hepatitis. A recently developed clinical practice guideline on viral hepatitis B provides additional information on laboratory testing in various contexts of hepatitis B infection.⁵

In acute hepatitis B where delay in testing has occurred, the HBsAg may be absent. In such cases the anti-HBc will be positive (but not anti-HBs). In remote (resolved) infection, only IgG anti HBc is present. In acute infection, IgM anti-HBc will be detected.

To test for chronic HCV infection the most commonly used anti-HCV antibody test is an enzyme immunoassay (EIA) or a variant. False-positive results may occur. The radioimmunoblot assay (RIBA) adds specificity to a positive anti-HCV EIA, but is probably of lesser use now that direct measurement of viral products in serum (HCV RNA) is widely available. HCV RNA is usually determined by polymerase chain reaction (PCR) although a simpler test, HCV RNA by bDNA, is still done in some laboratories. This latter test is not quite as sensitive as the PCR assay. HCV RNA in serum establishes with certainty the presence of hepatitis C infection. Some have wondered if the initial screening test for HCV should, therefore, be the HCV RNA, rather than an antibody test. Currently however, because of cost considerations, the initial test for test for HCV should be an anti HCV antibody test.

Once the presence of HCV is established, the genotype should be determined. There are six major HCV genotypes (1-6). Genotype has increasing importance for treatment determinations. This is discussed more fully in the chapter on hepatitis C. (Table 7).

Iron Tests:

Two paths may be considered to establish a diagnosis of hemochromatosis: genotypic and phenotypic. Diagnosis is most often made by phenotypic expression of disease, ie, by demonstration of excess circulating iron; ferritin; or iron accumulation within organs, especially the liver. Despite the excitement surrounding genotypic diagnosis of hereditary hemochromatosis, it has recently been shown that phenotypic demonstration of iron overload is the most cost-efficient strategy.⁶ A practice guideline has recently been published that confirms this approach.⁷

The most useful tests for iron overload in serum are iron, iron-binding capacity, and percent transferrin saturation. Serum ferritin levels are also useful and easily obtained. Hemochromatosis should be suspected in:

• any adult with liver disease, especially men;
• a transferrin saturation of greater than 55%; or
• ferritin elevations of greater than 200 µg/L in premenopausal women or greater than 300 µg/L in men and postmenopausal women.⁸

These thresholds are low; most patients who exceed them will not prove to have iron overload. Many inflammatory conditions, and especially other liver disease, will result in elevations of ferritin and/or iron levels in the absence of total body iron excess.

Limitations of Serum-based Tests of
Iron Overload:

Because both iron and ferritin are stored in liver cells, any condition that results in hepatocyte injury and release of intracellular contents into the blood will falsely raise iron, transferrin saturation, and ferritin levels. Therefore, in acute hepatic injury of any source, these tests will falsely suggest iron overload. Acute inflammation outside the liver may also falsely elevate serum-based iron tests. Tests of serum ferritin, iron, iron-binding capacity, and percent saturation determined in the setting of markedly elevated aminotransferase (AST and ALT), such as that seen in acute viral hepatitis or massive hepatic necrosis, will be identical to those seen in hemochromatosis. Iron studies are uninterpretable in the face of major elevations of transaminases.

Another limitation of iron studies relates to the time it takes for an individual with genetic hemochromatosis to accumulate excessive iron. In a young patient with this condition who has not yet had enough time to accumulate iron (especially the premenopausal woman), screening tests for iron overload may be normal even though the individual is at risk for subsequent development of iron overload.

Confirmation of suspected iron overload from serum- based tests requires demonstration of increased hepatic iron, usually by liver biopsy. The value of the biopsy is two-fold: it provides information about the degree of fibrosis/cirrhosis present, which is vital in predicting the risk of subsequent development of hepatoma, and it provides assessment of iron stores. Because there is an age- dependent increase in hepatic iron in normals, it is necessary to create an index that takes this into account. The hepatic iron index is calculated as follows:

Hepatic iron index = hepatic iron concentration (µmole/g dry weight) ÷ patient age in years.

A hepatic iron index of less than 2.0 is normal; values greater than 2.0 are seen in hemochromatosis.⁹ Recent rekindling of interest in assessment of hepatic iron by evaluation of iron stains of liver biopsy material indicates that this is a satisfactory alternative to quantitative iron determination.¹⁰ Bone marrow iron stores are not adequate to assess total body iron stores; indeed, cases of hemochromatosis with absent stainable bone marrow iron have been reported.

Genotypic Diagnosis of
Genetic Hemochromatosis:
It has been known for years that many cases of hemochromatosis are inherited as an autosomal recessive trait. In many cases, a defective gene called the HFE gene is implicated. The presence of this inherited gene results in the production of a protein in which a tyrosine amino acid rather than a cysteine amino acid is present at position 282 of the HFE protein. A second missense gene that results in an aspartic acid (instead of histidine) at position 63 of the same protein may increase iron absorption in some patients. The abnormalities are called C282Y and H63D mutations, respectively.

The individual with hereditary hemochromatosis usually must have two abnormal genes (homozygosity). Most often, two C282Y genes are present, but occasionally a compound heterozygote (C282Y-H63D) will also have excess iron. Homozygosity for H63D, interestingly, does not usually result in excess iron absorption.

The value of genotypic diagnosis is primarily limited to identification of at- risk family members after an index case has been discovered. HFE determination can identify young individuals at risk for iron overload before iron has become excessive. In Australia, where virtually all individuals with genetic hemochromatosis demonstrate such genetic abnormalities, gene diagnosis can replace iron testing. However, in most parts of the world, only 60% to 80% of those with hemochromatosis have an HFE abnormality. Therefore, HFE testing is not a suitable screening test for iron overload generally. Reliance on phenotypic expression (iron overload) is still required.

Copper Tests:

Although copper may accumulate to moderate excess in the liver in any chronic cholestatic liver condition, it does not appear to be injurious in these conditions. Wilson's disease is the main disease in which pathologic copper deposition results in serious liver injury, cirrhosis, and death. In Wilson's disease, copper also accumulates in the basal ganglia of the brain where it produces a wide gamut of neurologic abnormalities. Patients may present with either liver disease, brain disease, or both. This disorder is discussed in more detail in the Inherited Liver Disease chapter.

Wilson's disease is rare. Untreated, it usually produces death before age 40. Therefore, it is most appropriate to consider this potential cause in a child or young adult with otherwise unexplained liver disease. Laboratory diagnosis is most often based on the finding of a low ceruloplasmin level. Because most acute and chronic liver diseases cause the ceruloplasmin to be elevated, the finding of a low-normal or depressed ceruloplasmin level in a young patient with liver disease is suggestive of Wilson's disease. There are a few exceptions to this. A patient with acute fulminant liver failure of any sort may no longer have a liver capable of ceruloplasmin synthesis, so that individual may have a low serum level. Similarly, the patient with terminal end-stage liver disease may have a falling ceruloplasmin level. Finally, a few individuals have congential hypoceruloplasminemia without copper accumulation and are well.

The two most common forms of autoimmune liver disease are autoimmune chronic hepatitis and primary biliary cirrhosis. Ninety percent of those with each disorder are women. Autoimmune chronic hepatitis is characterized by very high serum aminotransferase (ALT and AST) levels, whereas primary biliary cirrhosis is associated with predominant elevations of alkaline phosphatase (a cholestatic disorder). Each is associated with autoantibodies in the serum. The treatment for each is quite different, so accurate diagnosis is essential. Table 8 contrasts the laboratory findings in these two autoimmune liver disorders.

Interpretation of autoimmune markers in a patient with liver disease is highly context-dependent. Autoantibodies are common in low titer in a number of acute and chronic liver conditions such as viral hepatitis. Therefore, the finding of autoantibodies in low titer is not sufficient evidence upon which to make a diagnosis of either autoimmune chronic hepatitis or primary biliary cirrhosis

Autoimmune hepatitis should be rapidly recognized by its propensity to occur in women (90%) and to be associated with very high transaminase levels (200 IU/mL, or higher). In this disease elevations of the gamma globulins (especially IgG) are pronounced. A myriad of autoimmune markers may be positive in autoimmune chronic hepatitis, but only a few are measured regularly: smooth muscle antibody, antinuclear factor, liver kidney microsomal (LK-M) antibody. A liver biopsy is often done to establish the diagnosis of autoimmune chronic hepatitis

Primary biliary cirrhosis is discussed in detail in the Intrahepatic Cholestatic Liver Disease chapter. In this condition, serum based liver tests reveal a predominant elevation of the alkaline phosphatase. Auto- antibodies most likely to be present in high titer in primary biliary cirrhosis is the antimitochondrial antibody but may not be needed in clear-cut cases.¹¹ An occasional patient may have features of both autoimmune chronic hepatitis, and primary biliary cirrhosis.

Return to Medicine Index

1. American Gastroenterological Association. Medical position statement: evaluation of liver chemistry tests. Gastroenterology. 2002;123:1364-66.
   
   
2. AGA Technical Review on the Evaluation of Liver Chemistry Tests. Gastroenterology. 2002;123:1367-84.
   
   
3. Carey WD. How should a patient with an isolated GGT be evaluated? Cleve Clin J Med. 2000;67:315-316.
   
   
4. Younossi ZM. Viral hepatitis guide for practicing physicians. Cleve Clin J Med. 2000;67(Suppl 1):S16-45.
   
   
5. Lok AS, McMahon BJ. Chronic hepatitis B. Hepatology. 2001;34:1225-1241.
   
   
6. Adams PC, Valberg LS. Screening blood donors for hemochromatosis: decision analysis model comparing genotyping to phenotyping. Am J Gastroenterol. 1999;94:1593-1600.
   
   
7. Tavill AS. Diagnosis and management of hemochromatosis:AASLD Practice Guidelines (Cosponsored by AGA and ACG) Hepatology. 2001;33:1321-1328.
   
   
8. Powell LW, George DK, McDonnell SM, Kowdley KV. Diagnosis of hemochromatosis. Ann Intern Med. 1998;129:925-931.
   
   
9. Bassett ML, Halliday JW, Powell LW. Value of hepatic iron measurements in early hemochromatosis and determination of the critical iron level associated with fibrosis. Hepatology. 1986;6:24-29.
   
   
10. Deugnier YM, Turlin B, Powell LW, et al. Differentiation between heterozygotes and homozygotes in genetic hemochromatosis by means of a histological hepatic iron index: a study of 192 cases. Hepatology. 1993;17:30-34.
    
    
11. Heathcote EJ. Management of primary biliary cirrhosis. The American Association for the Study of Liver Diseases practice guidelines. Hepatology. 2000;31:1005-1013.

Disclaimer