Series: Lab medicine in primary care Biochemical liver function tests

W Stuart A Smellie and Stephen D Ryder  consider two common scenarios of "abnormal" liver test results that may be seen in primary care

"Abnormal"liver results (bilirubin or enzymes produced by the liver) are a common finding in the investigation of patients presenting with non-acute illness and often raise questions about the need for, and timing of, further investigation. Although the management of overt liver disease will usually include referral to a secondary care specialist, the thresholds for referral and further investigation in clinically silent situations can be unclear.

CNRI/SPL
Fatty liver disease is a common cause of raised transaminase levels

Case 1

A 27 year old man presented to his general practitioner with a one week history of fever and malaise. On examination he appeared slightly jaundiced and was febrile (38.5°C). He had no evidence of pharyngitis, cervical lymphadenopathy, or organomegaly.

A full blood count was reported as haemoglobin 142 g/l, increased white cell count (10.4¥109/l), and a lymphocytosis (5.1¥109/l). Routine chemistry had been done at the same time and showed urea, creatinine, and electrolytes within laboratory reference ranges, raised bilirubin (42 mmol/l) with normal aspartate and alanine transaminases and alkaline phosphatase. A heterophilic antibody test for Epstein-Barr virus (monospot) was positive, and the patient was diagnosed as having infectious mononucleosis. Further testing showed that the bilirubin was 95% unconjugated, haptoglobins and lactate dehydrogenase were normal, no reticulocytes were present, and a blood film was normal in appearance with atypical lymphocytes present.

His symptoms were treated and when he was seen two weeks later his fever had resolved, he was clinically improved, and a repeat liver profile showed that his bilirubin had fallen but was still raised (28 mmol/l). After discussion with the local laboratory no further investigations were recommended and a presumptive diagnosis of Gilbert's syndrome was made. The patient made an uncomplicated recovery.

Case 2

A 43 year old man with type 2 diabetes was seen for annual review in a diabetes clinic. He had had diabetes for five years and was being managed with diet and metformin. His diabetes control had been good and recently glycated haemoglobin was 7.1%. He had been receiving simvastatin 20 mg lipid lowering therapy for three years, and when he was seen his cholesterol was 4.8 mmol/l, triglycerides 2.1 mmol/l, and high density lipoprotein cholesterol 1.1 mmol/l. His declared long term alcohol consumption, unchanged recently, was about 15 units per week.

Liver enzymes showed alanine transaminase 186 IU/l (normal <40 IU/l), normal alkaline phosphatase of 98 IU/l, bilirubin 21 mmol/l, albumin 42 g/l, total protein 84 g/l. Urea, creatinine, and electrolytes were within the laboratory reference ranges. Two years earlier, alanine transaminase had been 88 IU/l.

A repeat liver profile three months later showed that alanine transaminase was 202 IU/l, alkaline phosphatise 88 IU/l, bilirubin 23 mmol/l, albumin 44 g/l, total protein 84 g/l. Results of further investigations were: autoantibody screen (including antimitochondrial, antinuclear antibodies) negative, hepatitis viral serology negative, a₁ antitrypsin 1.2 g/l (normal range 0.8-2.0 g/l), ferritin 1080 mg/l (normal range 20-250 mg/l).

The patient was referred to secondary care for assessment. He proved to be homozygous for the C282Y mutation of the HFE gene, which is responsible for most genetic iron overload in the United Kingdom.

A liver biopsy confirmed the clinical diagnosis of haemochromatosis with severe hepatic fibrosis.

Discussion

The changes found on liver profiles can occur for reasons other than altered liver function. Incidental rises in bilirubin or in one or more "liver" enzymes are commonly found in everyday practice, and it may be difficult to determine the threshold for further investigation.

Case 1

Epstein-Barr virus infection (infectious mononucleosis) may produce changes in liver profiles in various ways. It may present with appearances of hepatitis, characterised by rises in transaminases, reflecting hepatocellular damage, particularly in postadolescent patients.¹ It may be associated with haemolysis, which is normally autoimmune in origin, and is associated with a rise in unconjugated bilirubin because of haem breakdown, a fall in serum haptoglobins, a rise in lactate dehydrogenase, reticulocytosis, and characteristic blood film appearances.

Viral illness is also a common trigger, producing a rise in unconjugated bilirubin in patients with Gilbert's syndrome, a congenital defect of intrahepatic conjugation of bilirubin with glucuronic acid that seems to be of little clinical or prognostic importance. The defect may be present in up to 6% of white populations (varying in other racial groups) and is typically associated with rises in serum unconjugated bilirubin not exceeding 68-85 mmol/l, depending on the study.²³

Although Gilbert's disease can be diagnosed by subjecting patients to a 48 hour fast, the diagnosis is typically made after excluding other causes of raised unconjugated bilirubin.³⁴ Liver disease is unlikely in the presence of raised unconjugated bilirubin in adults, as most obstructive or hepatitic processes result in a rise in conjugated (indirect) bilirubin or in a mixed picture of abnormal results on liver tests. When patients with alcohol induced changes are excluded, the commonest biopsy findings in patients with transaminases more than twice the upper limit of normal is non-alcoholic fatty liver disease.

Athenais/ism/spl
Ultrasound image of a fatty liver

Case 2

Raised transaminases in a diabetic patient are often produced by non-alcoholic fatty liver disease associated with fat deposition secondary to hypertriglyceridaemia and insulin resistance. Transaminases are a poor reflection of underlying liver disease, but raised transaminases (more than twice the upper limit of normal) are often associated with biopsy proved disease.⁵

In the era of widespread prescribing of statins, these drugs are often blamed for increases in transaminases, but randomised trials have not shown this to be a common effect and in the large heart protection study transaminases more than three times the upper limit of normal were not more common in patients treated with statins.⁶

Diabetes mellitus is a presenting feature of haemochromatosis in about 20% of patients, but may also coexist. A raised ferritin concentration is also commonly seen in non-alcoholic fatty liver disease or with chronic alcohol excess, where it is probably the result of hepatic injury rather than true iron overload, although a mild increase in hepatic iron stores may occur.

Haemochromatosis, however, is a rare cause of diabetes.⁷ Recommendations for the investigations of isolated "raised liver function tests" are shown in the boxes.

What are the sources of evidence?

A review of best practice in primary care pathology published in the Journal of Clinical Pathology indicates the varying degrees of evidence that back up these recommendations.⁸ Evidence for further investigation when results of liver tests are out of the normal range is based mostly on observational studies of investigation of patients whose test results have been abnormal and the subsequent findings on liver biopsy. The thresholds that trigger further investigation seem to be based on logistics and availability of resources. Apart from those caused by alcohol, the great majority of transaminase abnormalities are found to be related to non-alcoholic fatty liver disease. No drug treatment for this disorder has been proved, but exercise and weight loss improve results of liver function tests. The principle of investigating these patients seems to be based mainly on detecting potentially treatable causes, as in case 2 above, and on the theoretical assumption that identifying non-alcoholic fatty liver disease will improve the prognosis of the patients concerned and allow treatment in the future.

The principle of interval retesting to distinguish one-off results from stable and progressive increases seems sound. There is guidance, but limited evidence, on retesting interval.⁸ In patients without symptoms, and in view of the slow time course of the potential underlying diseases, retesting with an interval of months rather than weeks seems reasonable when rises are "minor" (up to 3-5 times the upper limit of normal). What is "normal" varies by ethnic origin and age; this confusion could be solved by establishing reference ranges specific for particular groups. The term "normal range" introduces further confusion, as the values reflect a statistical range and are not a definition of what is not pathological. Results outside of these ranges therefore do not necessarily indicate disease.

W Stuart A Smellie, consultant, Clinical Laboratory, General Hospital, Bishop Auckland, DL14 6AD
Email: info@smellie.com
Stephen D Ryder, consultant, Department of Gastroenterology, Queen's Medical Centre, Nottingham NG7 2UH

We thank Susan Richardson for typing this manuscript and IS Young, D Housley (Association of Clinical Biochemists), R Gama (Association of Clinical Pathologists), R Neal, N Campbell (Royal College of General Practitioners), who reviewed the work.

Competing interests: None declared. This article was first published in the BMJ 2006;333:481-3.

studentBMJ 2007;15:1-44 January ISSN 0966-6494

1. Hoagland RJ. Infectious mononucleosis. Prim Care 1975;2:295-307.
2. Tomsen HF, Hardt F, Juhl E. Diagnosis of Gilbert's syndrome. Scand J Gastroenterol 1981;16:699-703.
3. Hultcrantz R, Glaumann H, Lindberg G, Nillson LH. Liver investigation in 140 asymptomatic patients with moderately elevated activities of serum aminotransferases. Scand J Gastroenterol 1986;21:109-13.
4. Green RM, Flamm S. AGA technical review on the evaluation of liver function tests. Gastroenterology 2002;123:1367-84.
5. Skelly MM, James PD, Ryder SD. Findings on liver biopsy to investigate abnormal liver function tests in the absence of diagnostic serology. J Hepatol 2001;35:195-9.
6. Heart Protection Study Collaborative Group. MRC/BHF heart protection study of cholesterol lowering with simvastatin in 20,536 highrisk individuals: a randomised placebo-controlled trial. Lancet 2002;360:7-22.
7. Dubois-Laforgue D, Larger E, Timsit J. Is diabetes mellitus a sufficient condition to suspect hemochromatosis? Diabetes Metab 2000;36: 318-21.
8. Smellie WSA, Forth J, Ryder S, Galloway MJ,Wood AC,Watson ID, et al. Best practice in primary care pathology: review 6. J Clin Pathol 2006 Jul 5. http://jcp.bmjjournals.com/cgi/rapidpdf/jcp.2006.040014v1 (accessed 13 Jul 2006).
9. American Gastroenterological Association medical position statement. Evaluation of liver chemistry tests. Gastroenterology 2002;123:1364-6.