Endocrine Abstracts

Introduction: Alpha-fetoprotein (AFP), produced by the yolk sac and liver, is the most abundant serum protein in the human foetus. In adults, elevated AFP maybe associated with germ-cell or non-germ cell tumours, gastrointestinal tumours, active liver disease and hepatocellular carcinoma. Our case illustrates diagnostic uncertainties due to persistently elevated AFP values in the absence of any obvious pathology.

Case Presentation: A 22-year-old male was referred for assessment of gynaecomastia. He had no breast discomfort or galactorrhoea. On examination there was bilateral symmetrical gynaecomastia with no palpable breast lumps. Investigations showed normal liver function, gonadotropins, prolactin, testosterone, oestradiol and HCG. AFP was mildly elevated at 8.9 kU/l (normal range 0.1 – 7 kU/l). Ultrasound testes showed a normal solitary left testis with no intra testicular mass. Hepatitis serology and autoimmune screen was normal, but AFP remained elevated at 9.5 kU/l. Whole body cross sectional imaging to look for an occult germ cell tumour showed no significant abnormality within the liver and no pathological retroperitoneal or mesenteric nodes. Assay interference was suspected and paired samples were analysed in different laboratories. AFP analysed using the Siemens assay was elevated at 8.2 kU/l, whereas AFP analysed using the Roche assay in a different laboratory was normal at 4 kU/l (normal range 0 – 6 kU/l), indicating clear biochemical evidence of assay interference with spuriously elevated AFP values in our laboratory.

Conclusion: AFP assays are susceptible to interference by heterophile antibodies which can cause nonspecific false-positive results through usual mechanisms involving the capture and detection antibodies in immunoassays. Assay interference should be considered as a potential cause of persistently raised AFP values if other investigations are normal. This avoids the need for unnecessary surveillance imaging which has both cost and radiation safety implications.