Evaluation of Abnormal Liver Chemistries

by PY Kwo · 2017 · Cited by 467 — Patients with abnormal AST and ALT levels, particularly patients with other autoimmune conditions, should undergo testing for autoimmune liver

157 KB – 18 Pages

PAGE – 1 ============
18The American Journal of GASTROENTEROLOGY VOLUME 112 | JANUARY 2017 www.nature.com/ajg PRACTICE GUIDELINES INTRODUCTION e authors were invited by the Board of Trustees and Practice Guidelines Committee of the American College of Gastroenterol- ogy to develop a practice guideline regarding the evaluation of abnormal liver chemistries. We used the following resources: 1 . A formal review and literature search of the world literature on MEDLINE and EMBASE databases dealing with the evaluation of abnormal liver chemistries, studies that dealt with normal or reference range for alanine aminotransferase (ALT) levels and what thresholds trigger an evaluation for actionable liver disease. Studies detailing the relationship between ALT and nonalcoholic fatty liver disease, as well as studies assessing th cance of elevated liver chemistries on overall mortality and morbidity. 2 . Guideline policies of the American College of Gastroenterology. 3 . e experience of the authors and independent reviewers, as well as communication with senior hepatologists across the United States with regard to the threshold for evaluating abnormal liver chemistries. ese recommendations are intended for use by physicians and health care providers and suggest preferred approaches to the diag- noses and evaluation of those with abnormal liver tests ( Table 1 ). ese guidelines are intended to be exible and should be adjusted as deemed appropriate when applied to individual patients. Recommendations are evidence-based where possible. On subjects lacking rigid scien c data, recommendations are made based on the consensus opinion of the authors. To more fully characterize the available evidence reporting the recommendations, the ACG ACG Clinical Guideline: Evaluation of Abnormal Liver Chemistries Paul Y. Kwo , MD, FACG, FAASLD 1 , Stanley M. Cohen , MD, FACG, FAASLD 2 and Joseph K. Lim , MD, FACG, FAASLD 3 Clinicians are required to assess abnormal liver chemistries on a daily basis. The most common liver chemistries ordered are serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase and bilirubin. These tests should be termed liver chemistries or liver tests. Hepatocellular injury is defi ned as disproportionate elevation of AST and ALT levels compared with alkaline phosphatase levels. Cholestatic injury is defi ned as disproportionate elevation of alkaline phosphatase level as compared with AST and ALT levels. The majority of bilirubin circulates as unconjugated bilirubin and an elevated conjugated bilirubin implies hepatocellular disease or cholestasis. Multiple studies have demonstrated that the presence of an elevated ALT has been associated with increased liver-related mortality. A true healthy normal ALT level ranges from 29 to 33 IU/l for males, 19 to 25 IU/l for females and levels above this should be assessed. The degree of elevation of ALT and or AST in the clinical setting helps guide the evaluation. The evaluation of hepatocellular injury includes testing for viral hepatitis A, B, and C, assessment for nonalcoholic fatty liver disease and alcoholic liver disease, screening for hereditary hemochromatosis, autoimmune hepatitis, Wilson’s disease, and alpha-1 antitrypsin defi ciency. In addition, a history of prescribed and over-the-counter medicines should be sought. For the evaluation of an alkaline phosphatase elevation determined to be of hepatic origin, testing for primary biliary cholangitis and primary sclerosing cholangitis should be undertaken. Total bilirubin elevation can occur in either cholestatic or hepatocellular diseases. Elevated total serum bilirubin levels should be fractionated to direct and indirect bilirubin fractions and an elevated serum conjugated bilirubin implies hepatocellular disease or biliary obstruction in most settings. A liver biopsy may be considered when serologic testing and imaging fails to elucidate a diagnosis, to stage a condition, or when multiple diagnoses are possible. Am J Gastroenterol 2017; 112:18–35; doi: 10.1038/ajg.2016.517; published online 20 December 2016 1 Division of Gastroenterology/Hepatology, Department of Medicine, Stanford University School of Medicine , Palo Alto , California , USA ; 2 Digestive Health Institute, University Hospitals Cleveland Medical Center and Division of Gastroenterology and Liver Disease, Department of Medicine, Case Western Reserve University School of Medicine , Cleveland , Ohio , USA ; 3 Yale Viral Hepatitis Program, Yale University School of Medicine , New Haven , Connecticut , USA . Correspondence: Paul Y. Kwo, MD, FACG, FAASLD, Division of Gastroenterology/Hepatology, Stanford University School of Medicine , 750 Welch Road, Suite 210 , Palo Alto , California 94304 , USA . E-mail: pkwo@stanford.edu Received 11 February 2016 ; accepted 15 September 2016 CME

PAGE – 2 ============
Evaluation of Abnormal Liver Tests © 2017 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY 19Practice Guideline Committee has adopted the cl cation used by the grading of recommendation assessment, development, and evaluation workup with mo cations. e strength of recommen- dations are cl ed as strong or conditional. e quality of evi- dence supporting strong or weak recommendations are designated by the following level is high, moderate low, or very low quality ( 1 ). is is a practice guideline rather than a review article. Liver chemistries that are commonly ordered in comprehensive metabolic pro les are indirect markers of hepatobiliary disease. ey are not true measures of hepatic function and thus are best referred Table 1 . Recommendations 1. Before initiation of evaluation of abnormal liver chemistries, one should repeat the lab panel and/or perform a clarifying test (e.g., GGT if serum alkaline phosphate is elevated) to confi rm that the liver chemistry is actually abnormal. (Strong recommendation, very low level of evidence). 2. Testing for chronic hepatitis C is conducted with anti-HCV and confi rmation is performed with HCV-RNA by nucleic acid testing. Risk factors for hepatitis C include history of intranasal or intravenous drug use, tattoos, body piercings, blood transfusions, high risk sexual conduct, and those born between 1945 and 1965. Testing for acute hepatitis C is with anti-HCV and HCV RNA by nucleic acid testing. (Strong recommendation, very low level of evidence). 3. Testing for chronic hepatitis B is conducted with HBsAg testing. Testing for acute hepatitis B is with HBsAg and IgM anti-HBc. The following groups are at highest risk: persons born in endemic or hyperendemic areas (HBsAg prevalence >2%), men who have sex with men, persons who have ever used injection drugs, dialysis patients, HIV-infected individuals, pregnant women, and family members, household members, and s exual contacts of HBV-infected persons. (Strong recommendation, very low level of evidence). 4. Testing for acute Hepatitis A (IgM HAV) should occur in patients presenting with acute hepatitis and possible fecal-oral exposu re. Testing for acute hepatitis E (IgM HEV) should also be considered in those returning from endemic areas and whose tests for acute hepatitis A, B, and C are negative. (Strong recom-mendation, very low level of evidence). 5. Patients with elevated BMI and other features of metabolic syndrome including diabetes mellitus, overweight or obesity, hyperli pidemia, or hypertension with mild elevations of ALT should undergo screening for NAFLD with ultrasound. (Strong recommendation, very low level of evide nce). 6. Women consuming more than 140 g per week or men consuming more than 210 g per week who present with AST>ALT should be considere d at risk for alcoholic liver disease and should be counseled for alcohol cessation. (Strong recommendation, very low level of evidence). 7. All patients with abnormal liver chemistries in the absence of acute hepatitis should undergo testing for hereditary hemochromatosis with an iron level, transferrin saturation, and serum ferritin. HFE gene mutation analysis should be performed in patients with transferrin saturatserum ferritin. (Strong recommendation, very low level of evidence). 8. Patients with abnormal AST and ALT levels, particularly patients with other autoimmune conditions, should undergo testing for a utoimmune liver disease including ANA, ASMA, and globulin level. (Strong recommendation, very low level of evidence). 9. Patients with persistently elevated AST and ALT levels, especially patients <55 years of age, should undergo screening for Wils on’s disease with serum ceruloplasmin testing. In the setting of low ceruloplasmin, confi rmatory testing with 24-h urinary copper and slit-lamp eye examination to identify pathogno- monic Kayser–Fleischer rings should occur. (Strong recommendation, very low level of evidence). 10. Patients with persistently elevated AST or ALT should undergo screening for alpha-1 anti-trypsin (A1AT) defi ciency with alpha-1 anti-trypsin phenotype. (Strong recommendation, very low level of evidence). 11. Physicians should ask patients with abnormal liver chemistries about prescribed and over-the-counter medications, non-prescribe d complementary or alternative medicines, and dietary or herbal supplements which may be associated with DILI. (Strong recommendation, very low level of evidence). 12. A liver biopsy may be considered when serologic testing and imaging fails to elucidate a diagnosis, to stage a condition, or when multiple diagnoses are possible. (Strong recommendation, very low level of evidence). 13. An elevation of alkaline phosphatase should be confi rmed with an elevation in GGT. Given its lack of specifi city for liver dise ase, GGT should not be used as a screening test for underlying liver disease in the absence of other abnormal liver chemistries. (Strong recommendation, very low level of evidence). 14. Patients with alkaline phosphatase elevation with or without elevation of bilirubin should undergo testing for PBC (formerly named primary biliary cirrhosis) with testing for anti-mitochondrial antibody. (Strong recommendation, very low level of evidence). 15. Patients with alkaline phosphatase elevation with or without elevation of bilirubin should undergo testing for PSC with MR cholangiography or ERCP in conjunction with IgG4. (Strong recommendation, very low level of evidence). 16. In those with ALT and/or AST levels <5X ULN, the history and laboratory testing should assess for viral hepatitis B and C, alco holic and NAFLD, hemochro-matosis, Wilson’s disease, alpha-1-anti-trypsin defi ciency, autoimmune hepatitis and consider drugs/supplement related injury. (Strong recommendation, very low level of evidence). 17. In those with ALT and/or AST levels 5–15X ULN, evaluation should also assess for acute hepatitis A, B, and C in addition to all etiologies for AST/ALT elevation less than 5x ULN. (strong recommendation, very low level of evidence). 18. In those with ALT and/or AST levels >15X ULN, or massive elevation ALT of >10,000 IU/l, evaluation should also assess for aceta minophen toxicity and ischemic hepatopathy (shock liver). (Strong recommendation, very low level of evidence). 19. A patient presenting with acute hepatitis with an elevated prothrombin time, and/or encephalopathy requires immediate referral to liver specialist. (Strong recommendation, very low level of evidence). ALT, alanine aminotransferase; ANA, anti-nuclear antibody; ASMA, anti-smooth antibody; AST, aspartate aminotransferase; BMI, bo dy mass index; DILI, drug-induced liver injury; GGT, gamma-glutamyl transferase; HAV, hepatitis A virus; HBc, hepatitis B core antigen; HBsAg, hepatitis B surfac e antigen; HCV, hepatitis C virus; HEV, hepatitis E virus; HFE, hereditary hemochromatosis; IgM, immunoglobulin M; MR, magnetic resonance; NAFLD, non-alcoholic fatty liver disease; PBC, primary biliary cholangitis; PSC, primary sclerosing cholangitis; ULN, upper limit of normal.

PAGE – 3 ============
Kwo et al. The American Journal of GASTROENTEROLOGY VOLUME 112 | JANUARY 2017 www.nature.com/ajg 20meal due to increased levels of intestinal alkaline phosphatase ( 8 ). Alkaline phosphatase may be elevated during pregnancy due to placental synthesis of alkaline phosphatase. Typically, alkaline phosphatase elevates with obstruction of the bile ducts, which is due to increased canalicular synth esis of alkaline phosphatase with subsequent translocation to the sinusoid and is also a measure of liver injury ( 9 is occurs even if the obstruction is minor and in cient to increase serum bilirubin levels. To co rm hepatic origin of alkaline phosphatase, the canalicular enzyme GGT may be measured. An elevated GGT suggests that the alkaline phos- phatase elevation is of hepatic origin ( 6 ). Alkaline phosphatase may also be fractionated to better delineate bone, intestinal or hepatic origin of an elevated alkaline phosphatase. In children and the elderly, alkaline phosphatase le vels increase, especially females over 50 years of age, in part due to bone turnover ( 10,11 ). Bilirubin comes from the breakdown of senescent red blood cells and predominantly circulates in its unconjugated form tightly bound to albumin. Unconjugated bilirubin is not excreted in the urine. Conjugation by uridine 5’-diphospho (UDP)-glucuronosyl- transferase makes bilirubin water-soluble (conjugated bilirubin), allowing it to be excreted in bile where it is converted by bacte- ria in the colon to urobilinogen, which is subsequently excreted in the urine and stool. e absence of urobilinogen gives stool its classic clay-colored appearance in those with impaired bi ow. Unconjugated bilirubin is reported as indirect bilirubin as deter- mined by the van den Bergh reaction and accounts for ~70% of the total serum bilirubin ( 12 e total serum bilirubin is usually <1.1 mg/dl and an elevated direct bilirubin (conjugated bilirubin) indicates hepatocellular dysfunctio n or cholestasis. Fractionation of the bilirubin level to conjugated and unconjugated forms is not done routinely as many laboratories only report total serum bili- rubin, which is the sum of conjugated and unconjugated portions. Fractionation of total bilirubin is most helpful when the ALT, AST, and alkaline phosphatase levels are normal or near normal. If the total bilirubin is elevated and fractionation shows the majority of the elevation is unconjugated bilirubin, hepatocellular disease is unlikely to be the explanation. Conjugated bilirubin elevations are present in hepatocellular disorders as well as cholestatic disorders with impairment in bi ow. e delta bilirubin is derived from the reaction of conjugated bilirubin and albumin and has a half- life similar to albumin ( 5 e delta bilirubin accounts for the prolonged jaundice noted in patients recovering from hepatitis or cant obstruction as its decay is directly related to the half-life of albumin which is 3 weeks. Two markers of hepatocellular function are albumin and pro- thrombin time. Albumin is a plasma protein exclusively synthe- sized by the liver with a circulating half-life of 3 weeks ( 6 ). A reduction in albumin (norg/dl) usually indicates liver disease of more than 3 weeks duration, although an cant illness can decrease albumin levels due to cyto ects. Pro- thrombin time is a far more sensitive measure of liver function than albumin because prothrombin time may be prolonged in patients with severe liver disease of <24 h duration ( 6 ). Prothrom- bin time measures the extrinsic pathway of coagulatio e pro- thrombin time is a measurement of the clotting tendency of the to as liver chemistries or liver tests, and should not be referred to as liver function tests. True tests of liver function are not com- monly performed but include measurement of hepatic substrates that are cleared by hepatic uptake, metabolism, or both processes ( 2 ). Because of the widespread use of the comprehensive metabolic le testing that is done in routine practice to screen those who present for routine evaluation as well as those who are symptomatic and/or referred for elevation of abnormal liver chemistries, such abnormalities require a rational approach to interpretation. To date, there are no controlled trials that have been performed to deter- mine the optimal approach to evaluate abnormal liver chemistries. is guideline has been developed to assist gastroenterologists and primary care providers in the interpretation of normal and abnor- mal liver chemistries as well as an approach to prioritize and evalu- ate those who present with abnormal liver chemistries. Summary statements: 1 . Liver chemistries including ALT, aspartate aminotransferase (AST), alkaline phosphatase and bilirubin are markers of liver injury, not liver function, and should be referred to as liver chemistries, or liver tests. 2 . Albumin, bilirubin, and prothrombin time are markers of hepa- tocellular function that can be uenced by extrahepatic factors. 3 . e laboratory measurements of ALT, AST, and alkaline phosphatase are highly reproducible. 4 . Elevations of AST and/or ALT, alkaline phosphatase, and bilirubin suggest hepatocellular injury and are the abnormal liver chemistries that require assessment and potential evalu- ation. 5 . ALT is a more sp c marker of hepatic injury than AST. 6 . An elevated alkaline phosphatase level of hepatic origin may be co rmed by elevation of gamma-glutamyl transferase (GGT) or fractionation of alkaline phosphatase. e standard comprehensive metabolic pro le panel includes AST, ALT, alkaline phosphatase, bilirubin, and albumin. In addi- tion, a prothrombin time may be ordered. Aminotransferases including AST and ALT are enzymes involved in the transfer of amino groups of aspartate and alanine to ketoglutaric acid and are markers of hepatocellular injury and are also referred to as transaminases ( 3 ). AST is present in the liver and other organs including cardiac muscle, skeletal muscle, kidney, and brain. ALT is present primarily in the liver, and thus is a more sp c marker of hepatocellular cell injury ( 4–6 ). AST increase without elevation in ALT is suggestive of cardiac or muscle disease. Alkaline phosphatase is part of a family of zinc metalloprotein- ases enzymes that catalyze the hydrolysis of phosphate esters at an alkaline pH ( 7 is enzyme is found in hepatocytes on the canalicular membrane, not the bile duct cell. In addition to being present on the canalicular membrane of the hepatocyte, alkaline phosphatase is also found in bone, placenta, intestine, and kidney with the most common extrahepatic location originating from bone. Although rarely used in practice, in those with blood type O and B, serum alkaline phosphatase may increase a er a fatty PAGE - 4 ============ Evaluation of Abnormal Liver Tests © 2017 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY 21blood and measures factors 1, 2, 5, 7, 9, and 10. Because factors 2, 7, 9, and 10 are vitamin K dependent, the presence of cholestasis, where vitamin K is not absorbed, will prolong the prothrombin time. Also cant hepatocellular dysfunction can result in prolongation of the prothrombin time. is does not typically occur until concentrations of clotting factors fall below 10% of normal. As a general rule, in the absence of liver disease, a pro- thrombin time that is prolonged is due to vita ciency and/or steatorrhea. It should be noted that prothrombin time can also be elevated with warfarin, heparin bolus, disseminated intra- vascular coagulation (DIC), and hypothermia. is practice guideline will discuss the interpretation and evalu- ation of those with elevation of the major chemistries including ALT, AST, alkaline phosphatase, and bilirubin. Other liver tests (including GGT, albumin and prothrombin time) will be incorpo- rated into the evaluation of these major liver chemistries but will not be discussed separately. WHAT ARE TRULY NORMAL LIVER CHEMISTRY TESTS? Summary statements: 1 . A true healthy normal ALT level in prospectively studied pop- ulations without iden able risk factors for liver disease ranges from 29 to 33 IU/l for males and 19 to 25 IU/l for females, and levels above this should be assessed by physicians. 2 . Elevated ALT or AST above the upper limit of normal (ULN) in a population without iden able risk factors is associated with increased liver-related mortality. 3 . ere is a linear relationship between ALT level and body mass index (BMI) that should be assessed by physicians. 4 . A normal ALT level may not exclude cant liver disease. 5 . ALT levels are higher in males than females. 6 . AST and ALT ULN ranges can vary betw erent labs. 7 . Clinicians may rely on local lab ULN ranges for alkaline phosphatase and bilirubin. Normal lab values are generall ned as the mean value of a healthy population±2 s.d.’ is incorporates 95% of subjects. By nition, 2.5% of the population will be greater than the ULN of the reference population. For alkaline phosphatase and bilirubin levels, establishing normal liver enzyme levels that can be replicated acr erent reference labs has not been reported as problematic, whic ers from the wide variations in ranges reported as normal for ALT levels. However, establishing normal ranges for ALT and AST levels have been problematic due to dif- ferences in th nition of healthy control populations that are used to establish the normal reference ranges. One report exam- ined the local reference laboratory ranges for ALT used by the non-alcoholic steatohepatitis (NASH) Clinical Research Network and demonstrate cant erences in th ned ALT ULN (range 35–79 IU/l for men and 31–55 IU/l for women) ( 13 ese wide ranges appeared to be due to the use of erent reference populations utilized by th erent laboratories with local popu- lations used to establish the normal range of ALT, apparently with- out consideration of factors such as BMI. Another study found that 67 reference laboratories within one state us erent ALT ULN levels ranging from 31 to 72 U/l ( ref. 14 ). In this report, the majority of the labs used equipment from one of four manufactur- ers, but us erent methods to ne the ULN, with 40% uti- lizing manufacturer’s recommendations and local healthy control testing, 38.5% only using the manufacturer’s recommendations, 17% only using local healthy controls, and 8% using published normal levels from textbooks. However, inter-laboratory er- ences for ALT levels have not been reported to er cantly ( 13,15 ). When ning a normal population to be used for the establishment of a reference range, the possible presence of under- lying liver disease must be considered. Conditions such as non- alcoholic fatty liver disease (NAFLD), viral hepatitis, alcoholic liver disease and the use of medications and herbal agents or sup- plements need to be factored into the development of these nor- mal ranges. Most importantly, multiple studies have demonstrated that ALT levels correlate with increasing BMI ( 16–18 ). Determining an ALT level that is normal is clinically relevant to practicing clinicians and patients as there is substantial clinical sig- cance to thes erent reference ranges for ALT levels between labs. Both practice guidelines and diagnostic and therapeutic stud- ies base clinical decisions including evaluation of abnormal liver tests, therapy for hepatitis B, and evaluation of potential drug- induced liver injury on multiples of the ULN of ALT ( 19–23 ). Several studies and guidelines have proposed a standardized ULN for ALT based on prospectively acquired data using various meth- odologies ( Table 2 ). ese st ned normal reference popu- lations by excluding subjects with viral hepatitis, high-risk behavior and NAFLD risk factors (elevated BMI, triglycerides, glucose, and cholestero ese proposed normal ALT values are lower than commonly reported reference ranges an er by gender. Of the major liver chemistries, ther cient data on the measurement of ALT leve e true ULN for ALT was proposed in a large study of 6,835 blood donors with normal viral serologies, and BMI under 24.9 kg/m 2 to be 30 IU/l for men, and 19 IU/l for women ( 24 ). In a Korean study of 1,105 potential liver donors with normal liver biop- sies, they reported that age, BMI, and metabolic factors cantly ected ALT levels ( 25 ey proposed ULN for ALT to be 33 IU/l for men and 25 IU/l for women. In an examination of the National Health and Nutrition Examination Survey (NHANES) 1999–2002 and 2005–2008 databases, a er eliminating subjects with viral hep- atit cant alcohol use, diabetes, BMI>25, or enlarged waist circumference, and using statistical analysis, the calculated “maxi- mum correct cl cation” for ULN of ALT was found to be 29 IU/l for men and 22 IU/l for women ( 26 ). ELEVATED AMINOTRANSFERASE LEVELS AND THE EFFECT ON MORBIDITY AND MORTALITY ere is an accumulating set of data demonstrating that AST and ALT elevations correlate with morbidity and mortality ( Table 3 ). An initial report from Germany noted that those with AST>18 U/l had a 3X increased risk of all-cause mortality ( 27 ). A Korean

PAGE – 5 ============
Kwo et al. The American Journal of GASTROENTEROLOGY VOLUME 112 | JANUARY 2017 www.nature.com/ajg 22For example, applying the calculated “maximum correct cl -cation” for ULN of ALT (29 IU/l for men and 22 IU/l for women) to the NHANES databases would result in 36% of men and 28% of women bein ned as having elevated ALT ( 26 ). However, mul- tiple studies from the NHANES database have demonstrated that indeed the prevalence of individuals with elevated ALT levels has increase cantly in the US, likely from the obesity epidemic and resultant NAFLD ( 32 ). Some have argued against lowering of the ALT ULN due to major clinical an nancial implications including increased health care costs and unnecessary evaluations, increased mental anguish and anxiety, and reducing the blood donation pool ( 33 ). Supporting these concerns was a study of 235 asymptomatic work- ers, in which 27% had abnormal liver tests (using lower ULN criteria), yet only six were found to have any liver disease on further evaluation although long-term mortality was not assessed in this study ( 34 ). Our belief, given the increased liver-related mor- tality demonstrated across multiple populations for ALT>33 IU/l for men and >25 IU/l for women, is that clinicians should be edu- cated about the adverse long-term outcomes of these historically non-elevated levels and that a nation ort should be undertaken to standardize ALT levels across all populations. Further evidence for lowering the ULN of ALT and AST comes from reports demonstrating that cant liver disease may occur in the presence of “normal” liver chemistry levels. One report study found that, compared with men with AST or ALT<20 IU/l, the 30–39 IU/l group had an 8X (AST) or 9.5X (ALT) relative risk (RR) for liver-related death ( 28 ). Similar results were dem- onstrated from a study comparing the standardized mortality ratios in subjects from Olmsted County where higher ALT levels correlated with higher mortality with the standardized mortality ratio being 0.95 for normal AL ned as ULN 45 IU/l for men, 29 IU/l for females), 1.32 for 1–2X ULN, and 1.78 for >2X ULN with a similar relationship for AST levels ( 29 ). Studies have used the data from the NHANES databases to assess risk of morbidity and mortality in relationship to abnormal liver tests with one study demonstrating that elevated ALT (ULN ned as 30 U/l for men and 19 U/l for women) was associated with cant increases in liver-related mortality (11.2X) and diabetes-related mortality (3.3X) ( 30 ). Another analysis demon- strated that ALT>43 IU/l for men and >30 IU/l for women was related to the presence of coronary heart disease, even when patients with obesity, chronic viral hepatitis, and excessive alcohol use were excluded ( 31 ). IMPACT OF USING A LOWER ALT ULN ON CLINICAL PRACTICE Utilizing a cantly lower ULN for ALT will have implications by ning many more patients as having abnormal ALT levels. Table 3 . ALT and AST levels and liver related mortality Author /year Proposed ALT/AST cutoff level ALT/AST level for increased mortality Comments Arndt et al. ( 27 ) AST 18 AST>18 3X increase in all cause mortality Kim et al. ( 28 ) ALT<20 ALT 30–39 RR of liver mortality 2.9 (2.4–3.5) and 9.5 (7.9–11.5) in men, 3.8 (1.9–7.7) and 6.6 (1.5–25.6) in women Lee et al. ( 29 ) ALT (ULN 45 IU/l for M, 29 for F ALT 45–90 M 29–58 for F SMR risk 1.32 for 1–2X ULN, and 1.78 for >2X ULN Ruhl and Everhart ( 30 ) ALT 30 IU/l M, 19 IUL for F ALT>30 for M ALT >19 for F Increased liver related mortality ALT, alanine aminotransferase; AST, aspartate aminotransferase; F, female; RR, relative risk; M, male; ULN, upper limit of norm al. Table 2 . Summary of studies proposing ULN for ALT and or AST levels Author /Year Proposed ALT ULN (male) Proposed ALT ULN (female) Comments Neuschwander-Tetri et al. ( 13 ) 40 IU/l 40 IU/l Gender not specifi ed, not derived from data Piton et al. ( 18 ) 42 66 IU/l if BMI>23 31 44 IU/l if BMI>23 Derived from reference population with HCV Prati et al. ( 24 ) 30 IU/l 19 IU/l Derived from reference population Lee et al. ( 25 ) 33 IU/l 25 IU/l All had normal liver biopsies Ruhl and Everhart ( 26 ) 29 IU/l 22 IU/l Derived from NHANES Wright et al. ( 34 ) 33 IU/l – Gender not specifi ed ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; ULN, upper limit of normal.

PAGE – 6 ============
Evaluation of Abnormal Liver Tests © 2017 by the American College of Gastroenterology The American Journal of GASTROENTEROLOGY 23found that 9% of hepatitis C patients with normal AL ned as <50 IU/l) or near-normal AL ridg- in brosis and 11% had cirrhosis ( 35 ). Another study treated hepatitis C patients with ALT levels <30 IU/l with peg interferon alfa-2a/ribavirin and noted further reduction in ALT levels by up to 10 IU/l (( ref. 36 )). In a large cohort of patients with hemochro- matosis, including 32% of cirrhotics, 40% had normal AST and ALT although the precis nition of normal ALT and AST levels was not reported ( 37 ). Another study demonstrated that an ALT cut o of 40 IU/ml was associated with a high prevalence of stea- tosis in a cohort where proton magnetic resonance spectroscopy was used to determine hepatic triglycerides as a marker of hepatic steatosis with 79% of those with elevated hepatic triglyceride levels having ALT<40 U/l in men and 31 U/l in women ( 38 ). Previous guidelines have recommended basing clinical deci- sions to evaluate abnormal ALT levels based on multiples of the ULN of ALT (i.e., 2X ULN, 3X ULN, etc.) without sp cally ning the ALT level ( 19,20 is has led to variability in clinical practice due to the wide variation in the ULN of ALT across labo- ratories. For the purposes of this guideline, we have opted to ne a “normal” ALT based on the available literature correlating ALT levels and liver-related mortality. However, clinical judgment still remains of paramount importance. If a patient has signs and/or symptoms of clinical liver disease, even in the absence of abnormal liver chemistries, an evaluation should be initiated. In addition, the linear relationship of ALT to BMI should also be considered when assessing for the presence of cant liver disease as studies have demonstrated that those with higher BMIs may have higher ALT levels and some have suggested that an ALT correction for being overweight should be considered ( 18,39,40 ). SPECIFIC DISEASES OF THE LIVER INCLUDING DIAGNOSTIC TESTING Viral hepatitis Chronic hepatitis B and C infections are common in the United States. Approximately 4.1 million Americans are positive for anti- bodies against hepatitis C, and an estimated 3.0 million harbor chronic infection on the basis of positive hepatitis C RNA in the serum ( 41 ). e risk for acquiring hepatitis C is highest among individuals with parenteral exposure, such as from intravenous or intranasal drug use, blood transfusions before 1992, needle stick exposures, tattoos or body piercings, as well as high-risk sexual contac e screening test for chronic hepatitis C is the hepatitis C antibody, which has a sensitivity of 92–97% ( 42 e positive predictive value of the antibody test is highest among individu- als with risk factors for exposure; false positives may be observed in 5% of cases, and as high as 30% among individuals without reported risk factors. Co rmation of chronic infection is estab- lished by a highly sensitive HCV RNA PCR assay which has high sensitivity and sp city (AASLD/IDSA/IAS-USA. HCV testing and linkage to care. Recommenda tions for testing, managing, and treating hepatitis C ( http://www.hcvguidelines.org/ full-report/hcv-testing-and-lin kage-care )). Individuals co rmed to have positive RNA should be referred to a specialist for further characterization of the infection and assessment of liv brosis to guide the decision to pursue antiviral therapy. As recommended by the Centers for Disease Control (CDC) and United States Pre- ventive Services Task Force (USPSTF), individuals born within the 1945–1965 birth cohort should be considered for universal HCV antibody testing independent of abnormalities in AST/ALT levels, which may be normal in the presence of chronic infection including advanced liver disease ( 41,43 ). Acute infection with hepatitis C is a rare presentation compared with chronic hepatitis C and is associated with higher levels of aminotransferase levels, but is typically anicteric and without clinical symptoms of hepati- tis ( 44 ). e diagnosis of acute hepatitis C can be made in the set- ting of recent risk factors including drug use, and more recently men having sex with men has been iden ed as an increasingly prevalent risk factor. Testing includes anti-HCV, which is typically positive 6–8 weeks a er exposure, with co rmation by measur- ing HCV RNA by PCR testing which must be done to co rm a case of acute hepatitis or chronic hepatitis C. Testing for hepatitis B should be performed in all patients with persistently elevated AST/ALT levels. Chronic hepatitis B infects ~1.5 million Americans and over 280 million individuals world- wide ( 23 ). e route of transmission is predominantly vertical or horizontal in Asia and Africa where hepatitis B is endemic, and pre- dominantly through parenteral or sexual routes in the U.S. and other western nations. Characterization of an individual’s hepatitis B status can be achieved with three serologic tests, including the hepatitis B surface antigen (HBsAg) which is indicative of hepatitis B infection, the hepatitis B core antibody total which signals prior exposure, and the hepatitis B surface antibody, which signal immunity to the infec- tion, either natural or vaccine-mediated ( 23 ). Chronic infection is co rmed by presence of the HBsAg and/or positive viremia on a highly sensitive HBV DNA assay. Individuals co rmed to have chronic hepatitis B should be referred to physicians with exper- tise in its management, and undergo further characterization of their infection with tests such as hepatitis B e antigen, hepatitis B e antibody, hepatitis B genotype, hepatitis B viral load (quantitative DNA), an brosis assessment, which may guide the decision for antiviral therapy e diagnosis of acute hepatitis B is made by a pos- itive immunoglobulin M (IgM) hepatiti s B core antibody (hepatitis B core antibody IgM) and HBsAg in the setting of an acute hepatitis ( 45 ). Unlike acute hepatitis C in adults, acute hepatitis B infection is more commonly associated with signs and symptoms of hepatitis. Non-alcoholic fatty liver disease NAFLD is a highly prevalent condition associated with the meta- bolic syndrome, and is observed most commonly in patients with co-existing disorders including obesity, diabetes mellitus, dys- lipidemia, and hypertension and should be strongly considered in individuals with mild elevations of AST/ALT levels ( 46,47 ). Unlike alcoholic liver disease, there is no unique pattern of eleva- tion for ALT and AST levels, although in general, ALT is higher than AST levels, and levels are rarely above 300 IU/l. Although one-third of Americans may meet a br nition of NAFLD based on the presence of hepatic steatosis, a much smaller sub- set have NASH which is characterized by ammation, brosis, 157 KB – 18 Pages