Liver Disease

Laboratory Tests And Diagnostic Procedures in Hepatobiliary Disease Patricia Liu, M.D. Liver function tests are best utilized in concert with the clinical situation and in conjunction with serial determinations to ascertain the cause or evolution of the hepatic disorder. In addition, stool and urine tests, radionuclide imaging, contrast cholangiography (transhepatic cholangiography and endoscopic retrograde cholangiopancreatography), and histological assessment (liver biopsy) are often utilized to delineate the nature of the liver disease. Serum Enzymes (Serum Aminotransferase Transaminases) Serum glutamic oxaloacetic transaminase, also referred to as aspartate aminotransferase, and serum glutamic pyruvic transaminase, also called alanine aminotransferase (AST and ALT) are commonly employed to ascertain liver function. Striking elevations in the serum levels of these two enzymes are encountered in acute viral hepatitis, acute drug- or toxin-induced liver damage, and ischemic hepatitis. In addition, levels exceeding 500 IU/L and, on rare occasions, 1,000 IU/L can also be seen in patients with severe chronic active hepatitis, transiently in patients with common bile duct stones, and in patients with Budd-Chiari and veno-occlusive disease. There are a number of important hepatic disorders in which the serum AST and ALT are normal or minimally elevated . These include idiopathic genetic hemochromatosis, methotrexate-induced liver injury, liver dysfunction due to amiodarone, the liver disease associated with jejunal ileal bypass surgery, and patients with chronic hepatitis C virus infection. The ratio of AST and ALT is also sometimes of value in clinical practice. A ratio greater than 2 with both AST and ALT being less than 300 IU/L is characteristic of alcoholic liver disease. On rare occasions, if both the AST and ALT are elevated, patients are subjected to a liver biopsy after a very thorough serologic workup only to find that the liver histology is completely normal. It is very important to exclude a primary muscle disorder in such patients since not only the AST but also the ALT can be elevated in patients with muscle disorders. Serum Alkaline Phosphatase The sources of serum alkaline phosphatase include liver, bone, small intestine, placenta, and, on rare occasions, tumors capable of producing alkaline phosphatase. In general, patients with cholestasis have increased levels. However, the level of serum alkaline phosphatase is not helpful in distinguishing intrahepatic from extrahepatic cholestasis. Rarely, patients with hypernephroma and Hodgkin's disease have elevated levels in the absence of liver involvement. Also, patients with Wilson's disease often have normal values. On rare occasions, patients with a variety of malignant tumors have elevations in the serum alkaline phosphatase level that are not caused by liver or bony metastases. This isoenzyme, referred to as Regan isoenzyme, is biochemically and immunologically indistinguishable from placental alkaline phosphatase, and in addition to being present in serum, can also be present in tumor tissue or in malignant effusion fluids. In normal children with active bone growth, influx of enzyme from osteoid tissue may result in a threefold elevation above normal in alkaline phosphatase values. In the third trimester of pregnancy, the serum level may double as a consequence of the contribution of placental alkaline phosphatase, making interpretation of results of this test difficult during the latter stage of pregnancy. Also, persons over the age of 50 years may normally have one and a half times normal values. Subjects with blood groups O and B, who are ABH secretors and Lewis antigen-positive, may have a significant amount of intestinal alkaline phosphatase in their serum, especially after ingestion of a fatty meal. Thus, unlike other enzymes in the serum, the activity of alkaline phosphatase can be noticeably affected by eating; hence, it is recommended that alkaline phosphatase activity be assayed in the fasting state. Although an elevation of the serum alkaline phosphatase level may be the first clue to hepatobiliary disease, the alkaline phosphatase level is normal on rare occasions despite extensive metastatic hepatic deposits or complete bile duct obstruction. Patients with stage I or II Hodgkin's disease, hypernephroma, congestive heart failure, myeloid metaplasia, peritonitis, diabetes, subacute thyroiditis, or uncomplicated gastric ulcer have been reported to have mild elevations in serum alkaline phosphatase levels that are of probable liver origin in the absence of overt liver involvement. Recently, twofold to fourfold elevations in serum alkaline phosphatase levels were reported in several members of a family. No bone or hepatic disorder was present in this family, who demonstrated the enzyme elevation in a pattern suggesting autosomal dominant inheritance. In this context, the elevation of the serum alkaline phosphatase level may be no more specific than that of an elevated sedimentation rate, and the routine use of automated biochemical tests has indeed increased awareness of this phenomenon. The most reliable method of determining the tissue origin of an elevated alkaline phosphatase level is polyacrylamide gel electrophoresis. Since this is not routinely available in most clinical laboratories and since methods using urea denaturation or heat inactivation are quite unreliable, it is common practice to measure serum 5'-nucleotidase, or gamma glutamyl transpeptidase, when dealing with an isolated or disproportionately elevated serum alkaline phosphatase level. In such instances, elevations of the levels of any of these two enzymes generally imply that the source of the elevated alkaline phosphatase level is hapatobiliary and not bony. 5'-nucleotidase levels may increase in normal pregnancy, whereas gamma glutamyl transpeptidase levels do not. However, commonly used drugs such as barbiturates, phenytoin, and alcohol may cause induction of hepatic microsomal gamma glutamyl transpeptidase; therefore, an elevated serum level of this enzyme does not necessarily imply frank liver cell injury. A disproportionately elevated serum gamma glutamyl transpeptidase level for several days often follows moderate alcohol ingestion. This finding is used by many physicians to detect alcohol abuse in patients who underestimate or deny the ingestion of alcohol. Serum Bilirubin Most clinical laboratories use spectrophotometry to measure serum bilirubin. The normal value for total bilirubin is less than 1 mg/dl. Two fractions - a conjugated or direct fraction (normally less than 0.25 mg/dl) and an unconjugated or indirect fraction are obtained fairly routinely. It is useful to classify hyperbilirubinemia into conjugated and unconjugated categories. Patients are considered to have conjugated hyperbilirubinemia if greater than 50 percent of the elevated total bilirubin level is conjugated, and categorized as having unconjugated hyperbilirubinemia if greater than 80 percent of the total bilirubin level is unconjugated or indirect-reacting. Mild unconjugated hyperbilirubinemia (total bilirubin level less than 5 mg/dl) is seen in Gilbert's disease uncomplicated hemolytic disorders, and congestive heart failure. Mild conjugated hyperbilirubinemia is a constant finding in the Dubin-Johnson and Rotor syndromes. Conjugated hyperbilirubinemia of varying intensity is seen in a variety of liver disorders including acute viral, drug-induced, and toxin-induced hepatitis, shock liver, and metastatic disease to the liver. Even in fulminant hepatitis, the liver is capable of conjugating the bilirubin. The height of the serum bilirubin level has no discriminative value in distinguishing intrahepatic cholestasis from extrahepatic obstruction. Indeed, fairly marked elevations in total serum bilirubin levels have been reported in patients with non-biliary tract sepsis. Although patients with fulminant hepatitis may be anicteric, the level of serum bilirubin is of prognostic import in certain conditions such as alcoholic hepatitis, primary biliary cirrhosis, and halothane hepatitis. In primary biliary cirrhosis, persistent elevations of greater than 2.0 mg/dl in the serum total bilirubin level usually occur late in the course of the disease and imply a poor prognosis. Bilirubin levels greater than 10 mg/dl have been associated with a 60 percent mortality in patients with halothane hepatitis. Urine Bilirubin The presence of bilirubin in the urine implies that it is direct bilirubin that is present, as indirect bilirubin is tightly bound to albumin and hence not filtered by the normal kidney. This inexpensive and sensitive test serves several functions: a positive result can rapidly confirm clinically suspected hyperbilirubinemia; it implies hepatobiliary disease and excludes hemolysis; and it may provide an early clue to the presence of hepatobiliary dysfunction, as presence of urine bilirubin often antedates overt icterus. Urine Urobilinogen Urobilinogen results from the breakdown of conjugated bilirubin by intestinal bacteria. Normal excretion in stool averages 100 mg per day. About 10 to 20 percent is absorbed into the portal circulation, from where most of it undergoes entero-hepatic circulation. A small amount, usually less than 2 mg per day, is excreted in the urine. Although urinary estimation is dependent upon several factors including urine pH, rate of urine production, timing of collection (there is a diurnal variation), and method employed, complete absence of urinary urobilinogen implies absence of bilirubin in the intestine and strongly suggests complete bile duct obstruction. The passage of clay-colored stools by the patient is equally meaningful. Increased urinary urobilinogen suggests hemolysis or hepatic dysfunction. Serum Lactate Dehydrogenase Serum lactate dehydrogenase activity may arise from myocardium, liver, skeletal muscle, brain or kidney tissue, and red blood cells. Therefore, an elevated serum 1actate dehydrogenase value is a nonspecific finding. The hepatic origin (1aerate dehydrogenase 5) of serum lactate dehydrogenase can be verified by isoenzyme determination. Increased lactate dehydrogenase levels are seen in patients with a variety of hepatobiliary disorders including acute viral or drug hepatitis, congestive heart failure, cirrhosis, and extrahepatic obstruction. Striking elevations in serum lactate dehydrogenase and alkaline phosphatase levels are highly suggestive of metastatic disease to the liver. Serum Albumin and Globulin Albumin is synthesized exclusively by hepatic parenchymal cells and has a serum half-life of about 20 days. Decreased levels in the serum can occur as a consequence of decreased synthesis or excessive losses. The former occurs in patients with malnutrition or significant liver disease. Hypoalbuminemia secondary to excessive loss of albumin is seen in patients with nephrotic syndrome or protein-losing enteropathy. A fall in the serum albumin level and a rise in levels of globulins, primarily gamma globulins, are frequently seen in patients with chronic hepatitis or cirrhosis. Elevations in IgA levels are common in alcoholic cirrhosis, and elevations in IgG levels are common in chronic active hepatitis. Elevations in IgM levels are frequently seen in primary biliary cirrhosis; of the immunoglobulins, only an elevation of the IgM fraction on immunoelectrophoresis has any significant specificity. Diminished levels of alpha-l-globulins due to deficient alpha-1-antitrypsin activity can be associated with chronic active hepatitis and cirrhosis in children and adults. Coagulation Factors All the clotting factors are synthesized by the liver, with the exception of factor VIII, which is synthesized by reticuloendothelial cells and vascular endothelial cells. The half-life of these coagulation factors is short, ranging from six hours for factor VII to five days for fibrinogen. Therefore, acute liver injury often results in a prolongation of the prothrombin time, which is dependent upon the activity of factors II, V, VII, and X. Factor VIII levels may be normal or increased in severe hepatic injury; hence, assessment of factor VIII levels may help distinguish the coagulopathy of liver disease from that secondary to disseminated intravascular coagulation where factor VIII levels are generally decreased. Vitamin K-dependent factors include factors II, VII, IX, and X. A prolongation of the prothrombin time beyond four seconds that of control despite vitamin K administration (5 to 10 mg parenteral vitamin K; prothrombin time measurement repeated 24 hours later) signifies fairly advanced liver disease. The prognostic significance of a markedly elevated prothrombin time is exemplified by the 100 percent mortality reported in a study of patients with halothane hepatitis who had a prothrombin time greater than 20 seconds. Other clinically relevant aspects of abnormalities in coagulation include the necessity to correct these factor deficiencies in patients with significant bleeding and the fact that certain procedures such as percutaneous liver biopsy are contraindicated in patients with a significant coagulopathy. Alpha Fetoprotein Alpha fetoprotein is a unique alpha-l-globulin normally synthesized in large amounts by embryonic liver cells and in trace amounts by fetal yolk sac cells and the fetal gastrointestinal tract. High serum concentrations (greater than 500 ng/ml by radioimmunoassay) are present in 70 percent of patients with primary hepatocellular carcinoma in the United States. Concentrations of this protein fall dramatically in patients in whom partial curative hepatectomy is carried out. Serial determinations aid in monitoring the response to therapy or detecting early recurrence. Patients with yolk sac tumors with germ cell elements may also have high concentrations of alpha fetoprotein. Modest elevations may be seen occasionally with other tumors, notably gastrointestinal malignancies metastatic to the liver. Levels up to 500 ng/ml are seen during pregnancy. High levels indicate multiple pregnancy, significant fetal neural tube defects, fetal distress, or fetal death. Serum Ferritin Serum ferritin levels accurately reflect hepatic and total-body iron stores. However, the serum ferritin can be elevated in a number of conditions, including idiopathic genetic hemochromatosis, hepatocellular necrosis of any etiology (alcoholic liver disease, acute and chronic viral hepatitis, drug-induced liver injury, obesity-related liver dysfunction), Hodgkin's disease, leukemia, hyperthyroidism, uremia, and rheumatoid arthritis. Measurement of serum iron concentration, percent transferrin saturation, and serum ferritin level is the screening regimen currently recommended for idiopathic genetic hemochromatosis. Serum Ceruloplasmin Ceruloplasmin, a blue copper-containing glycoprotein is an acute-phase reactant. 95 percent of patients with Wilson's disease have serum ceruloplasmin concentrations below 20 mg/dl. The serum ceruloplasmin can be normal in 10 percent of patients with Wilson's and chronic active hepatitis. It is also low in 10 percent of heterozygotes with Wilson's and low levels can also be encountered in patients with nephrotic syndrome and protein-losing enteropathy. Antimitochondrial Antibody This autoantibody can be detected in the serum by a variety of methods including immunofluorescence, double immunodiffusion, complement fixation, radioimmunoassay, and enzyme-linked immunoassay. Mitochondrial antibodies are found in 0.8 to 1.6 percent of the general population, 6 percent of patients with chronic active hepatitis, and 85 to 90 percent of patients with primary biliary cirrhosis. Antimitochondrial antibodies are also found in a significant number of asymptomatic relatives of patients with primary biliary cirrhosis and chronic active hepatitis. Although the antibody titer has no discriminative value in distinguishing patients with primary biliary cirrhosis from patients with a variety of other hepatobiliary disorders or autoimmune conditions, recent studies suggest that the mitochondrial antibodies seen in disorders other than primary biliary cirrhosis can be differentiated from those that appear to be specific for primary biliary cirrhosis on the basis of characteristic immunofluorescent patterns that they mediate in rat and human tissues or the effects of trypsin pretreatment of such tissue sections on immunofluorescent staining. Mitochondrial antibodies directed against a purported specific primary biliary cirrhosis antigen are believed to have high diagnostic relevance, and it has been stated that when such antibodies are not detected in the serum, a diagnosis of primary biliary cirrhosis should be made with caution and only after a careful period of clinical follow-up. Mitochondrial antibodies appear to play no part in the pathogenesis of primary biliary cirrhosis and provide no useful prognostic insight in this disorder. Bile Acids Serum levels of bile acids can be measured by sensitive chromatography techniques and radioimmunoassays. Measurements can be made in the fasting state, two hours postprandially, and following intravenous isotopic administration (bile acid tolerance test). The cost of the test and the need for appropriate facilities in the laboratory limit the usefulness of these tests, which currently appear to be not significantly superior to more routine and readily available tests. In addition, an abnormal test result does not discriminate between hepatocellular and cholestatic liver disease. Blood Ammonia Ammonia arises from bacterial degradation of nitrogenous contents (dietary protein, blood) in the intestine. Although the blood and cerebrospinal fluid ammonia levels are elevated in most patients with hepatic coma, there is a poor correlation between blood ammonia level and the degree of hepatic encephalopathy. The correlation is better with arterial ammonia as compared with venous ammonia, which may be increased by muscular exercise, seizure activity, or even clenching the fist. Serial values of arterial ammonia are occasionally helpful in following individual patients. Elevations in arterial blood ammonia levels have also been recognized in patients with portosystemic shunting or inborn disorders of urea metabolism, and in patients following ureteroileostomy. Blood ammonia measurements are of limited utility in the evaluation of most patients with known hepatobiliary disease. Its greatest usefulness may be in evaluating patients with coma or altered mental status of unclear cause. In such patients, an elevated arterial ammonia concentration would suggest hepatocellular dysfunction as an important contributing factor. Serum Lipids and Lipoproteins Patients with cholestatic disorders (both intrahepatic and extrahepatic) frequently have an elevation in the concentration of serum phospholipids and unesterified cholesterol, skin xanthomas commonly appear if the total serum cholesterol level exceeds 450 mg/dl for longer than three months. The major component of unesterified cholesterol is an abnormal low-density lipoprotein designated lipoprotein X. Contrary to earlier claims, quantitation of serum lipoprotein X levels is not helpful in distinguishing intrahepatic cholestasis from extrahepatic obstruction. Patients with cirrhosis or malignant biliary obstruction who are malnourished may have low total serum cholesterol levels. Breath Tests Several radiolabeled compounds have been used to assess hepatic function in humans. Most studies have employed 14C aminopyrine. The test is inexpensive and entails minimal radioactive exposure to patients (generally, less than 0.5 µCi per test) but should not be performed in children or pregnant women. Studies in humans suggest that the aminopyrine test (14CO2 - a metabolic product of the methyl group of aminopyrine - is measured) is an indirect measure of mixedfunction oxidase mass. Although not yet widely available, breath tests may emerge as excellent non-invasive screening tests for detecting liver disease or prove useful in the periodic follow-up of patients in whom liver biopsy is contraindicated. Hepatic Scintiscanning Technetium-99m sulfur colloid is taken up selectively by the reticuloendothelial cells and is used occasionally in clinical practice. It often provides useful information regarding the size and shape of the liver, although on occasion virtually no hepatic uptake of colloid is present in patients with severe or far-advanced liver disease. Heterogeneous uptake of colloid by the liver with "shift" of colloid to the spleen and vertebral bone marrow is characteristic but not specific for cirrhosis of the liver. Focal defects of the liver are seen in many conditions, including hepatic cysts, hemangiomas, abscesses, tumors, amyloidosis and "regenerative" cirrhotic nodules, and peliosis hepatis. Lesions smaller than 2 cm in size and present deep within the hepatic parenchyma may escape detection on liver scanning. Hepatic scintiscanning may aid in the distinction between focal nodular hyperplasia and hepatic adenoma, as scanning usually shows normal findings in the former condition, whereas a cold defect is usually present in patients with a hepatic adenoma. Avid uptake by an enlarged caudate lobe is characteristic of Budd-Chiari syndrome. In this condition, there is occlusion of the major right and left hepatic veins. The caudate lobe has a separate arterial supply and venous drainage and hence characteristically undergoes compensatory hypertrophy. Hepatic scintiscanning is also useful in the evaluation of hepatic or splenic trauma (currently CT is more often employed) and in the detection of remnant or accessory splenic tissue in patients who have undergone splenectomy. Hepatobiliary Scanning (Hida Scan) Agents being used with increasing frequency for biliary scanning are technetium-99m-labeled acetanilide iminodiacetic acid derivatives (dimethyl, paraisopropyl, and diisopropyl iminodiacetic acids). These agents are rapidly taken up by functioning hepatocytes and excreted into the bile. These radiopharmaceuticals are superior to Rose-Bengal as biliary scanners. In a broad sense, they evaluate hepatic excretory function and the patency of the biliary tree. Adequate visualization is often obtained in patients with serum bilirubin levels between 5 and 10 mg/dl. Nonvisualization of the gallbladder within two hours with good visualization of the common bile duct is highly suggestive of cystic duct obstruction or calculous cholecystitis. Caution in interpretation must be exercised in those patients with concomitant hepatocellular dysfunction who' exhibit poor hepatic uptake and concentration of the isotope. Dimethyl iminodiacetic acid scanning is thus most useful in excluding cystic duct obstruction and acute cholecystitis if the gallbladder is clearly visualized. Dimethyl iminodiacetic acid scanning has been used to demonstrate the gallbladder as the site of the defect observed on the inferior border of the liver on technetium-99m sulfur colloid scanning. Dimethyl iminodiacetic acid can also be used to diagnose choledochal cysts and demonstrate biliary leaks or assess the patency of biliary-enteric shunts. Angiography Selective celiac, superior mesenteric, and hepatic angiography is employed in the evaluation of selected patients with liver disease. Enlargement of the hepatic arteries, neovascularity, arteriovenous shunting, and portal vein thrombosis are frequently encountered in patients with primary hepatocellular carcinoma. However, differentiation from other primary hepatic lesions such as focal nodular hyperplasia or angiosarcoma and from vascular metastatic tumors such as hypernephroma, carcinoid, and choriocarcinoma is not always possible. Angiography may be useful in defining the extent of tumor and delineating the anatomic vascular supply in preparation for hepatic lobectomy. Angiography has been routinely performed prior to elective shunt surgery for patients with variceal bleeding in order to define anatomy. Magnetic resonance angiography may prove to be an adequate option in the future. Ultrasonography Ultrasound examination of the abdomen is a useful test in the evaluation of liver disease. It is available in most hospitals, is inexpensive, and entails no radiation exposure to patients. Thus, it is particularly applicable in the evaluation of liver and biliary tract disease in children and pregnant women. Marked obesity and excessive intestinal gas can be limiting factors in obtaining good resolution of the images. Ultrasound examination of the liver will often identify mass lesions 1 to 2 cm in size in the hepatic parenchyma and do this independent of hepatic function. The nature of defects seen on technetium-99m sulfur colloid scanning - solid or cystic - can readily be ascertained and it can facilitate guided aspiration of cysts or biopsy specimens of lesions. Ultrasonography is an appropriate procedure for detecting gallstones and confirming the presence of ascites; study in the fasting state is important. It is often used as the first test in the evaluation of patients with cholestatic jaundice. Dilated bile ducts can be readily seen on ultrasound examination in patients with mechanical extrahepatic biliary tract obstruction. Dilation of the bile ducts may not be evident if the obstruction is incomplete or intermittent or if it has been present for a short duration. Serial ultrasound examinations may provide valuable information in these circumstances. The common bile duct is frequently dilated following cholecystectomy. Therefore, an enlarged duct in this situation does not necessarily signify ongoing biliary tract obstruction. Computed Tomographic Scanning This modality is useful in detecting mass lesions (cysts, tumors, abscesses) within the liver, and computed tomographic scanning with contrast enhancement can often accurately predict the nature of the lesion identified. It has an advantage over ultrasound in obese patients and those with excessive gas, but its widespread use is limited by the high cost; an additional consideration is radiation exposure to patients. In patients with cholestatic jaundice, computed tomographic scanning can distinguish intrahepatic cholestasis from mechanical extrahepatic obstruction with an accuracy rate of 90 per cent. Computed tomographic scanning after intravenous injection of iodinated contrast material has been reported to demonstrate hypertrophy of the caudate lobe, stagnation of contrast material at the periphery of the liver, narrowing of the inferior vena cava in its intrahepatic section, and failure of opacification of the major hepatic veins in patients with Budd-Chiari syndrome. This constellation of findings appears to be specific for Budd-Chiari syndrome, although further confirmatory experience is necessary before this diagnosis can be confidently made in the absence of liver biopsy or venography. Hepatic computed tomography has also been used to evaluate iron overload in the liver. It appears to be moderately sensitive and highly specific for this purpose. However, liver biopsy with quantitation of liver iron content remains the "gold test" in the diagnosis of hemochromatosis. Computed tomographic scanning can also be useful in assessing liver fat content. Monoenergetic computed tomographic scanning has been shown to accurately (good agreement with histologic and chemical liver fat determination) predict liver fat content in patients with alcoholic liver disease. Magnetic Resonance Imaging (MRI) This modality is useful in evaluating the extent of hepatic metastases. It is also of value in assessing the patency of hepatic and portal veins and, hence, is a useful test in patients with suspected Budd-Chiari syndrome and in patients with portal vein thrombosis. Hemangiomas of the liver also have a characteristic MRI appearance. In patients with iron overload, a characteristic black or hypointense liver is often encountered. Percutaneous Transhepatic Cholangiography Percutaneous transhepatic cholangiography using a "skinny" Chiba needle is performed in most institutions by skilled radiologists. The success rate with multiple passes approaches 90 to 95 percent in patients with extrahepatic biliary tract obstruction and 70 percent in patients with intrahepatic cholestasis and normal-sized bile ducts. Percutaneous transhepatic cholangiography may be preferred over endoscopic retrograde cholangiopancreatography in patients with surgically distorted gastroduodenal anatomy such as following hemigastrectomy and Billroth II anastomosis. Marked ascites and irreversible coagulopathy are contraindications to this procedure. Serious complications occur in about 5 percent of patients and include cholangitis, hemorrhage, bile peritonitis, pneumothorax, and drug reaction, and are responsible for a 0.5 percent attendant mortality. Endoscopic Retrograde Cholangiopancreatography (ERCP) This procedure is usually performed by a gastroenterologist and has a success rate of 80 to 90 percent in skilled hands. Serious complications occur in about 5 percent of patients and include pancreatitis, instrumental injury, cholangitis, sepsis, and aspiration pneumonia. Endoscopic retrograde cholangiopancreatography would appear to be the procedure of choice if periampullary carcinoma, common duct stone, or postcholecystectomy biliary tract disease is strongly suspected on clinical grounds. In patients with common duct stones, endoscopic papillotomy can provide definitive therapeutic benefit. Needle Biopsy of the Liver Percutaneous needle biopsy of the liver is a commonly utilized and safe procedure that can be performed at the bedside. It often provides precise and accurate tissue diagnosis without resorting to general anesthesia and laparotomy, and most agree that it can be performed as an outpatient procedure provided facilities are available for short-term observation and hospitalization should the need arise. Accepted indications for liver biopsy include hepatocellular disease of uncertain cause, unexplained hepatomegaly and/or splenomegaly, hepatic filling defects demonstrated by radionuclide scanning, ultrasound examination, or computed tomographic scanning, chronic hepatitis, fever of unknown origin, and staging of malignant lymphoma. Liver biopsy is often helpful in the evaluation of patients with clinically suspected alcoholic liver disease to confirm the diagnosis and ascertain the severity of damage to the liver and in the evaluation of patients with portal hypertension. Liver biopsy is the single best "liver function test" (the gold standard) in documenting hemochromatosis, Wilson's disease, certain varieties of glycogen storage disease and type I Crigler-Najjar syndrome. Liver biopsy is not particularly helpful in distinguishing intrahepatic from extrahepatic obstruction. Although liver biopsy is not apparently associated with an excessive risk of complications in patients with extrahepatic obstruction, the pathognomonic features of extrahepatic obstruction - a bile infarct with feathery degeneration of surrounding hepatocytes - are not invariably present. Additionally, it provides no information regarding the site or nature of the obstructing lesion. With the advent of imaging modalities to visualize the extrahepatic biliary tree, needle biopsy of the liver is rarely performed today when extrahepatic obstruction is suspected. Contraindications for percutaneous needle biopsy of the liver include uncooperative or comatose patients, hydatid cyst disease, hemangioma or angiosarcoma of the liver, right pleural disease or local infection at the proposed biopsy site, and significant coagulopathy. Arbitrarily employed contraindications include a prolongation of the prothrombin time greater than four seconds over control, a partial thromboplastin time greater than 15 seconds over control, and a platelet count less than 75,000/mm3. A bleeding time should be obtained in patients with borderline clotting abnormalities or uremia, and in those taking antiplatelet drugs. In patients with a severe coagulopathy, the transvenous (transjugular) approach may be used. In this technique, a catheter is placed in the jugular vein and manipulated into a hepatic vein. If bleeding occurs, it occurs within the vascular system itself. Percutaneous needle biopsy of the liver is not contraindicated in amyloidosis unless the liver is massively enlarged or there is an associated bleeding tendency. Tense ascites and severe anemia are relative contraindications. Complications are numerous but rare and the overall mortality rate is less than 0.02 percent. The incidence of complications appears to increase significantly if multiple passes (greater than four using the percutaneous route) are employed. The most common untoward effect is pain at the biopsy site or right shoulder. Moderate to severe pain with or without hypotension usually manifests within the first three hours of the procedure. Serious bleeding occurs in less than 0.3 percent of patients, although asymptomatic subcapsular or intrahepatic hematomas are probably considerably more common. Extremely rare serious sequelae include pneumothorax, hemothorax, hemobilia, arteriovenous fistula formation, and perforation of colon or gallbladder. If suspicion for highly vascular tumors such as carcinoid is high, angiography or nuclear medicine flow scanning should be performed before percutaneous biopsy. Suspected primary hepatocellular carcinoma is not a contraindication to biopsy. Liver biopsy in unsuspected echinococcal cyst disease can be followed by dissemination of daughter scoleces or fatal anaphylaxis. Peritoneoscopy and Laparotomy Peritoneoscopy may be performed using local anesthesia and mild sedation. It permits adequate visualization of the anterior surface of the liver, gallbladder, and stomach as well as parts of the small intestine and colon. It has proved useful in the diagnosis by direct visual inspection and directed needle biopsy of a variety of hepatic disorders including cirrhosis, hepatomegaly, hepatic malignancy, and portal hypertension. Peritoneoscopy has helped in determining the cause of ascites and in the staging of Hodgkin's disease. Contraindications include tense ascites, an uncooperative patient, marked obesity, surgical abdominal scars, and severe coagulopathy. Serious complications occur in 1 to 2 percent of patients and include air embolism, subcutaneous emphysema, ascites leak, bowel perforation, and hemorrhage. The overall mortality rate is approximately 0.03 percent. Laparotomy is now seldom performed for the express purpose of diagnosing hepatobiliary disease. On occasion, however, it is necessary to undertake this procedure to obtain biopsy and culture material and perform operative cholangiography. The increasing availability of and familiarity with the vast array of both noninvasive (eg, ultrasonography, computed tomographic scanning, dimethyl iminodiacetic acid scanning, etc.) and invasive (eg, skinny needle transhepatic cholangiography, endoscopic retrograde cholangiopancreatography, angiography, liver biopsy, etc.) hepatic tests have resulted in the infrequent use of peritoneoscopy and laparotomy for the evaluation of hepatobiliary disease. References 1. Chopra S, Griffin PH: Laboratory tests and diagnostic procedures in evaluation of liver disease. Am J Med 1985;79:221-230. 2. 3. Kaplan MM: Alkaline phosphatase. N Engl J Med 1972; 286:200-202. Schapiro JM, Smith H, Schaffner F: Serum bilirubin, a prognostic factor in primary biliary cirrhosis. Gut 1979;20:137-140. Blanckaert N, Kabra PM, Farina FA, Stafford BE, Marton LJ, Schmid R: Measurement of bilirubin and its mono- and di-conjugates in human serum by alkaline methanolysis and high performance liquid chromatography. J Lab Clin Med 1980;96:198-212. Klion FM, Schaffner F, Popper H: Hepatitis after exposure to halothane. Ann Intern Med 1969; 71:467-477. Howard JM, Ghent CN, Carey LS, Flanagan PR, Valberg LS: Diagnostic efficacy of hepatic computed tomography in the detection of body iron overload. Gastroenterology 1983;84:209-215. Elias E, Hamlyn AW, Jain S, et al: A randomized trial of percutaneous transhepatic cholangiography with the Chiba needle versus endoscopic retrograde cholangiography for bile duct visualization in jaundice. Gastroenterology 1976;71:439-443. Mueller PR, Van Sonnenberg E, Simeone JF: Fine-needle transhepatic cholangiography: indications and usefulness. Ann Intern Med 1982; 97:567-572. Knauer MC: Percutaneous biopsy of the liver as a procedure for outpatients. Gastroenterology 1978; 74:101-102. Rösch J, Lakin PC, Antonovic R, Dotter CT: Transjugular approach to liver biopsy and transhepatic cholangiography. N Engl J Med 1973; 289:227-231. Perrault J, McGill DB, Ott BJ, Taylor WF: Liver biopsy: complications in 1000 inpatients and outpatients. Gastroenterology 1978; 74:103-106. Losowsky MS: Needle biopsy of the liver. A review. J R Soc Med 1982; 75:736-741. Kaplan MM: Evaluation of hepatobiliary diseases. In: Stein JH, ed. and chief. Internal Medicine,4th ed. Mosby-Yearbook, 1994; 544-552. Bernardino ME, Galambos JT: Computed tomography and magnetic resonance imaging of the liver. Sem Liv Dis 1989; 9:32-49. Johnson CD: Magnetic resonance imaging of the liver: current clinical applications. Mayo Clin Proc 1993; 68:147-156. Sheth SG, Gordon FD, Chopra S. Nonalcoholic steatohepatitis. Ann Intern Med 1997;126:137-145. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.