Deepak V. Gopal, MD, FRCP(C); Hugo R. Rosen, MD
VOL 107 / NO 2 / FEBRUARY 2000 / POSTGRADUATE MEDICINE
CME learning objectives
This is the first of four articles on liver disease
This page is best viewed with a browser that supports tables
Preview: Although markers of liver disease are available, in many
cases, their usefulness is limited by insufficient sensitivity or specificity.
In addition, significant liver damage may already have occurred in patients who
have normal findings on liver function tests. A basic understanding of
abnormalities in liver enzymes is important to assist clinicians in developing a
rational, cost-effective approach in patients with liver disease. In this
article, Drs Gopal and Rosen discuss results of laboratory tests and how to use
them in patient evaluation.
Gopal DV, Rosen HR. Abnormal findings on liver function tests: interpreting
results to narrow the diagnosis and establish a prognosis. Postgrad Med
2000;107(2):100-14
The approach to liver function testing can be challenging, even to an experienced clinician. The large number and wide variety of tasks performed by the liver make evaluation of its function with a single laboratory test difficult. Therefore, a broad array of biochemical tests is often used to provide indirect evidence of hepatobiliary disease. The term "liver function tests" is firmly entrenched in the medical vocabulary, although it has been criticized because the tests most commonly used to evaluate liver disease--measurement of serum aminotransferase and alkaline phosphatase (ALP) levels--assess hepatocyte integrity rather than a known synthetic function.
Biochemical screening of healthy, asymptomatic people has revealed that up to 6% have abnormal liver enzyme levels. However, the prevalence of liver disease in the general population is significantly lower (about 1%) (1). Panels of serum liver chemistry tests, performed serially and interpreted in the context of a person's medical history, physical examination findings, and other laboratory test results, are useful in evaluation of a patient with signs and symptoms of liver disease. Nevertheless, even though the serum biochemical test pattern may suggest a specific diagnosis in such a patient, confirmation usually requires further investigation with serologic or imaging studies and, possibly, liver biopsy (1).
Even mild liver test abnormalities may be an early clue to the presence of potentially significant liver disease (2,3). For instance, patients with chronic hepatitis C virus (HCV) infection are often asymptomatic unless they have advanced liver disease. They usually have mild elevation of the serum alanine aminotransferase (ALT) level, and about one third have persistently normal liver enzyme levels.
Accordingly, lack of sensitivity and specificity limits the usefulness of liver function tests. For example, in some clinical settings (eg, cirrhosis), patients may have serum aminotransferase levels in the normal to near-normal range. In addition, several nonhepatic factors (table 1 [4]) can affect the results of tests that measure specific hepatic function, such as serum albumin, total bilirubin, and prothrombin time (PT).
| Table 1. Nonhepatic causes of abnormal liver function test results | ||
| Test result | Nonhepatic causes | Discriminating tests |
|
|
||
| Decreased serum albumin level | Protein-losing enteropathy | Serum globulins, alpha1-antitrypsin clearance |
| Nephrotic syndrome | Urinalysis, 24-hr urinary collection for protein | |
| Congestive heart failure | Cardiac examination, two-dimensional echocardiogram | |
|
|
||
| Elevated AST level | Myocardial infarction | CK-MB, troponin, ECG |
| Muscle disorders | CK, ESR | |
|
|
||
| Elevated ALP level | Bone disease | GGT, serum leucine aminopeptidase, 59-nucleotidase |
| Pregnancy | GGT, 59-nucleotidase, hCG in serum and urine | |
| Malignant tumor | Alkaline phosphatase electrophoresis | |
|
|
||
| Elevated bilirubin level | Hemolysis | Reticulocyte count, peripheral smear, LDH, haptoglobin |
| Sepsis | Clinical setting, blood cultures | |
| Ineffective erythropoiesis | Peripheral smear, urine bilirubin, hemoglobin electrophoresis, bone marrow aspiration and biopsy | |
| Shunt hyperbilirubinemia | Clinical setting | |
|
|
||
| Elevated PT | Antibiotic use, anticoagulant use, steatorrhea, dietary deficiency | Response to vitamin K |
|
|
||
|
ALP, alkaline phosphatase; AST, aspartate aminotransferase; CK, creatine kinase; ECG, electrocardiogram; ESR, erythrocyte sedimentation rate; GGT, gamma-glutamyltranspeptidase; hCG, human chorionic gonadotropin; LDH, lactate dehydrogenase; PT, prothrombin time. Adapted from Moseley (4). |
||
|
|
||
When analyzing liver enzymes, it is useful to categorize abnormalities on the basis of liver cell necrosis (hepatocellular damage) or cholestatic processes. Serum PT, albumin level, and cholesterol levels assess true synthetic function because they reflect hepatic protein production and cholesterol synthesis and metabolism. Serum bilirubin (further classified as direct and indirect) helps demonstrate abnormalities in hepatic uptake, conjugation, and excretion (5).
The aminotransferases--aspartate aminotransferase (AST) and ALT--are the most often used and most specific indicators of hepatic injury and represent markers of hepatocellular necrosis. These liver enzymes catalyze transfer of the alpha-amino groups aspartate and alanine to the alpha-keto group of alpha-ketoglutaric acid.
Whereas ALT is primarily localized to the liver, AST is present in a wide variety of tissues, including heart, skeletal muscle, kidney, brain, and liver. AST is present in both the mitochondria and cytosol of hepatocytes, but ALT is found only in the cytosol. In an asymptomatic person with isolated elevation of the AST or ALT level, diagnostic clues can be garnered from the degree of elevation (1,3,5). Although no uniform definition of mild, moderate, or severe elevation exists, a working definition has been outlined (table 2 [6]).
| Table 2. Degrees of elevation of liver enzymes | ||||
| Measurement | Normal (IU/L)* | Mild** | Moderate** | Marked** |
|
|
||||
| AST | 11-32 | <2-3 | 2-3 to 20 | >20 |
|
|
||||
| ALT | 3-30 | <2-3 | 2-3 to 20 | >20 |
|
|
||||
| ALP | 35-105 | <1.5-2 | 1.5-2 to 5 | >5 |
|
|
||||
| GGT | 2-65 | <2-3 | 2-3 to 10 | >10 |
|
|
||||
|
ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; GGT, gamma-glutamyltranspeptidase. *Normal range varies with the assay used and should be obtained from the laboratory performing the test. **Multiples of the upper limit of normal. Adapted, with permission, from Keeffe (6). |
||||
|
|
||||
Serum levels of AST and ALT are elevated to some extent in almost all liver diseases. Mild elevation is typically found in patients with fatty liver, nonalcoholic steatohepatitis, and chronic viral hepatitis. Highest elevation occurs in acute viral hepatitis, hepatic necrosis induced by drugs (eg, acetaminophen) or toxins, and ischemic hepatitis related to circulatory shock (4,5). The degree of elevation of aminotransferase levels does not appear to correlate with the extent of necrosis found in liver biopsy specimens, so it has no prognostic value. In fact, rapidly declining levels may reflect a decrease in viable hepatocytes and indicate a poor prognosis.
Moderately elevated aminotransferase levels (a 3- to 20-fold increase) are typical of acute or chronic hepatitis, including alcoholic hepatitis. A characteristic feature of chronic HCV infection is an episodic, fluctuating pattern of serum ALT levels--periods of elevated enzyme activity alternating with periods of normal or near-normal levels. In patients with extrahepatic biliary tract obstruction, serum levels of AST and ALT may become elevated to greater than 300 IU/L. After peaking (within the first 24 to 48 hours after obstruction), levels usually decline rapidly (1,7).
Although elevated aminotransferase levels may be the first clue to liver disease and screening has proved useful for detecting subclinical disease in asymptomatic persons, patients with normal levels may have significant liver damage. For instance, recent studies have demonstrated that patients with chronic HCV infection may have histologic evidence of chronic hepatitis despite repeatedly normal results on liver tests (7).
The ratio of AST to ALT may be helpful diagnostically. In particular, an AST:ALT ratio of more than 2:1 is characteristic in patients with alcoholic liver disease. Elevation of the AST level out of proportion to the ALT level appears to be the result of a differential reduction in hepatic ALT due to deficiency of the cofactor pyridoxine-5-phosphate (1,6-8). An AST:ALT ratio of more than 2:0 strongly suggests alcoholic liver disease but does not preclude other diagnoses. Elevation of the ALT level to more than 500 IU/L suggests a diagnosis other than alcoholic liver disease, even if the AST:ALT ratio is greater than 2:0 (7,8). Other biochemical clues to the presence of alcoholic liver disease include elevations of serum gamma-glutamyltranspeptidase (GGT) level, erythrocyte mean corpuscular volume, and desialyated transferrin level (8).
In the setting of viral hepatitis, the AST:ALT ratio, which is typically less than 1:0, can rise to greater values as fibrosis and cirrhosis develop. The exact mechanism of AST:ALT ratio alteration in progression of liver disease is unclear, and the correlation with and accuracy in predicting degree of fibrosis and presence of cirrhosis are controversial (7-11).
ALP and GGT are the markers normally used to identify cholestasis. The term "serum ALP" is applied to a group of enzymes that catalyze hydrolysis of phosphate esters at an alkaline pH. The enzymes are widely distributed and may originate from bone, liver, intestine, kidney, or placenta. In children and adolescents, in whom bone growth is active, serum ALP may increase by up to threefold. In patients with hepatobiliary disorders, increased ALP results from increased hepatic production with leakage into the serum rather than from failure to clear or excrete circulating ALP.
Markedly elevated levels of ALP suggest the possibility of such disorders as extrahepatic biliary obstruction, primary biliary cirrhosis, drug-induced cholestasis, primary sclerosing cholangitis, and infiltrative processes (eg, amyloid, granulomatous hepatitis, neoplasm). The degree of elevation does not differentiate intrahepatic from extrahepatic cholestasis. In evaluating patients with mild ALP elevation, use of the algorithm in figure 1 may be helpful (6). Depressed serum levels of ALP have been associated with congenital hypophosphatasia, hypothyroidism, pernicious anemia, and zinc deficiency (1,2,5,12).
GGT is a sensitive indicator of hepatobiliary disease, but it is not specific. Elevated GGT levels can occur in renal failure, myocardial infarction, pancreatic disease, and diabetes mellitus. In addition, increased GGT levels are inducible and may be seen with ingestion of phenytoin (Dilantin) or alcohol in the absence of other clinical features of liver disease. The major clinical utility of the GGT level is to exclude a bone source of ALP elevation. Because of its long half-life (26 days), GGT is limited as a marker of surreptitious alcohol ingestion. Many patients with isolated serum GGT elevation have no other clinical features of liver disease and therefore do not warrant extensive evaluation.
Found in the liver in association with canalicular and sinusoidal plasma membranes, 5'-nucleotidase, when present in serum, reflects hepatobiliary release by the detergent action of bile salts on plasma membranes. It is distributed in other organs as well. An elevated 5'-nucleotidase level along with an elevated ALP level is specific for hepatobiliary disease, and the finding is more useful than results of a GGT measurement alone. This marker is also helpful in differentiating physiologic elevation of ALP from hepatobiliary disease in children.
Bilirubin, an endogenous organic anion, binds reversibly to albumin and is transported to the liver, where it is conjugated to glucuronic acid and excreted in the bile. It is derived primarily from catabolism of red blood cell heme and to a lesser extent from degradation of myoglobin, cyochromes, catalase, and peroxidase. Healthy people have a small amount of unconjugated (indirect) bilirubin but no conjugated (direct) bilirubin in their blood. However, commonly used laboratory tests often incorrectly identify some serum bilirubin as being conjugated in healthy people, so most testing centers report a range of normal for conjugated bilirubin (7).
Hepatobiliary disease is indicated when the conjugated fraction of total bilirubin exceeds the upper limit of normal, even if the total serum bilirubin concentration is normal or near normal. The presence of conjugated, water-soluble bilirubin in the urine (bilirubinuria) always indicates hepatobiliary disease. Hemolysis and Gilbert syndrome are common conditions that cause benign elevation of unconjugated bilirubin (1,7). Diagnostic factors in hyperbilirubinemia and jaundice are summarized in table 3 (12).
Most circulating proteins in plasma are synthesized in the liver, and levels indicate synthetic capability of the liver. For instance, albumin accounts for 65% of serum protein and has a half-life of about 3 weeks. The blood concentration depends on albumin's synthesis rate (normal, 12 g/day) and plasma volume. Therefore expanded plasma volume or decreased albumin synthesis can result in hypoalbuminemia. Hypoalbuminemia is often associated with ascites and expansion of the extravascular albumin pool at the expense of intravascular albumin levels. In general, albumin is a good marker of severity of chronic liver disease, but levels may be affected by chronic renal insufficiency, urinary protein losses, or gastrointestinal losses (5,13).
Increases in serum globulin levels are common in chronic liver disease but are nonspecific. However, the pattern of elevation may give a clue to the underlying cause. For example, in autoimmune hepatitis, the serum IgG level is elevated, whereas in primary biliary cirrhosis, the serum IgM level is elevated (5).
Coagulation factors also reflect synthetic liver function. Most of them--including fibrinogen, the vitamin K-dependent factors (prothrombin and factors VII, IX, and X), and factor V--have a much shorter half-life than albumin and are synthesized by the liver. Factor VII has the shortest half-life and therefore decreases first, followed by factors X and IX. Factor V is not vitamin K-dependent, so its measurement can help distinguish vitamin K deficiency from hepatocellular dysfunction in patients with prolonged PT. In patients with fulminant hepatic failure, factor V values that are less than 20% of normal portend a poor outcome in the absence of liver transplantation (4,5).
Severity and prognosis of liver disease can be assessed using PT, which reflects deficiency of one or more of the liver-synthesized factors. In cholestatic liver disease, PT prolongation can result from vitamin K deficiency, but differential diagnosis includes disseminated intravascular coagulation, inherited deficiencies of coagulation factors, and medications that antagonize the prothrombin complex. To correct PT, vitamin K is often given subcutaneously or intravenously at a dosage of 10 mg/day for 3 days. Correction by at least 30% within 24 hours suggests that hepatic function is intact and that vitamin K deficiency may be the source of the problem (4,5,14).
In general, the international normalized ratio (INR) is believed to be a better indicator of degree of coagulopathy than PT, because it is a standardized value and is not subject to laboratory variability as is PT. However, PT and the INR are not sensitive indexes of liver function in chronic disease, and they may show normal values in patients with cirrhosis. Nevertheless, PT (along with serum albumin concentration) is a criterion used in the Child-Turcotte-Pugh prognostic classification (table 4), which grades the degree of hepatic dysfunction in patients with cirrhosis.
| Table 4. Modified Child-Turcotte-Pugh prognostic classification for grading degree of hepatic dysfunction in patients with cirrhosis | |||
| Factor | Points* | ||
|
|
|||
| 1 | 2 | 3 | |
|
|
|||
| Encephalopathy (grade) | 0 | 1-2 | 3-4 |
|
|
|||
| Ascites | None | Slight | Moderate |
|
|
|||
| Bilirubin (mg/dL) | 1-2 | 2-3 | >3 |
|
|
|||
| Albumin (g/dL) | >3.5 | 2.8-3.5 | <2.8 |
|
|
|||
| Prothrombin time (sec) | 1-4 | 5-6 | >6 |
|
|
|||
|
*Total points = 5 or 6, grade A; 7 to 9, grade B; 10 to 15, grade C. |
|||
|
|
|||
In most forms of parenchymal liver disease, liver biopsy is the definitive test for making a diagnosis and establishing a prognosis. The typical use of liver biopsy is assessment of abnormal liver enzyme levels when serologic, autoimmune, metabolic, and other diagnostic measures yield negative results (figure 2 [1]: not shown).
For instance, biopsy is useful in diagnosis of nonalcoholic steatohepatitis and in confirmation and differentiation of primary hemochromatosis from secondary iron overload states (5,15,16). In addition, it is essential in assessing degree of chronicity of viral hepatitis and cirrhosis. Other indications for biopsy are detection of systemic disorders of the liver (including fever of unknown origin), assessment of severity of drug-induced liver injury, confirmation of suspected hepatocellular carcinoma (primary or metastatic) and multisystem infiltrative disorders (eg, granulomatous hepatitis), and screening of relatives of patients with familial disease.
Liver biopsy is also useful in acquiring hepatic tissue samples to send for culture, in evaluating response to therapies for viral and metabolic liver disease (eg, Wilson's disease, hemochromatosis, autoimmune hepatitis, primary biliary cirrhosis), and excluding allograft failure, reinfection, or ischemia after orthotopic liver transplantation.
In general, liver biopsy is safe and is usually performed as an outpatient procedure. However, important contraindications include severe coagulopathy, presumed hemangioma or echinococcal disease, ascites, pleural effusion, and lack of cooperation by the patient. Transjugular liver biopsy may be feasible when the standard percutaneous approach is contraindicated. Samples of focal defects are best taken using radiologic (ultrasonography or computed axial tomography) guidance (3,5,15,16).
Evaluating abnormal liver test results requires careful attention to the corresponding clinical data obtained during history taking and physical examination. Generally, it is helpful to separate liver tests into three categories: tests that assess synthetic function, tests that assess hepatocellular necrosis (hepatocellular enzymes), and tests that assess cholestasis (3,15,17). The clinical setting together with the specific pattern of liver function abnormalities can narrow differential diagnosis and provide a cost-effective approach to assessing patients and identifying those who need liver biopsy.
Dr Gopal is a fellow in gastroenterology and Dr Rosen is assistant professor of medicine, division of gastroenterology and hepatology, Oregon Health Sciences University School of Medicine, and both are on staff at the Veterans Affairs Medical Center, Portland. Correspondence: Hugo R. Rosen, MD, Veterans Affairs Medical Center, 3710 SW US Veterans Hospital Rd, PO Box 1034, P3-GI, Portland, OR 97201. E-mail: hugo.rosen@med.va.gov.