Endocrine Abstracts

The duty biochemist at Barts Health noted a testosterone result of 17 nmol/l on a 20-year-old female, measured by Roche E170 electrochemiluminescence immunoassay. Laboratory policy is that all female testosterone results >2.5 nmol/l by immunoassay (upper reference limit 1.92 nmol/l) are checked by LCMS to exclude analytical interferences. The LCMS result on this sample was below the limit of quantification (<0.5 nmol/l).

Clinical details were ‘aplastic anaemia’. There was no previous testosterone result. LH and FSH were 8.6 and 6.3 IU/l respectively. Patient’s notes revealed that she was on danazol to treat anaemia secondary to Dyskeratosis Congenita, at a dose of 200 mg alternating daily with 100 mg.

The sample was diluted in steroid-free serum, and measured neat (confirming previous result) and at three dilutions. The dilution studies gave a linear regression of y=1.01x – 0.25, with a correlation coefficient of r² =0.999.

The sample was checked on an alternative platform; testosterone was 2.3 nmol/l (<4.5) on the Abbott Architect (one-step chemiluminescent microparticle immunoassay).

Danazol (200 mg capsule) obtained from Pharmacy was dissolved in methanol (10 ml), spiked into steroid-free serum to a final concentration of 988 nmol/l, and assayed for testosterone. On the Roche platform the result was 10.2 nmol/l (recovery of 1.04%). By LCMS, testosterone was reported as undetectable.

These data confirm that danazol cross-reacts with the reagent antibodies in the Roche assay. The manufacturer states that cross-reactivity of danazol is ≤0.5% at a concentration of 1000 ng/ml. Abbott state that danazol does not cross-react in their assay. Danazol (17α-ethinyl testosterone) is a derivative of ethisterone, and has significant structural homology with ethisterone and norethisterone. UKNEQAS have demonstrated that norethisterone (at pharmacological concentrations) interferes in the Roche testosterone assay. Users of immunoassay should be aware of interferences due to steroid drugs.