BACKGROUND

Direct immunoassay are still the main analytical approach for determining the concentration of testosterone in routine clinical practice. However, the comparability of the results of measurements of testosterone by immunoassay from different manufacturers remains unsatisfactory. Liquid chromatography-mass spectrometry methods (HPLC-MS/MS) overcome these problems, but they are not yet available for widespread use. The aim of our study was to evaluate the possibility of harmonizing testosterone measurements different immunoassay by secondary correction to the HPLC-MS/MS.

MATERIAL AND METHODS

Concentration of total testosterone in serum was measured by various immunoassay (Abbott ARCHITECT i1000SR and i2000SR, Siemens Immulite 1000, Beckman Coulter Access II analyzer) in 4 Ekaterinburg clinical laboratories and by HPLC-MS/MS was carried out in the laboratory of the National Medical Research Center of Endocrinology. Testosterone level was measured in 10 serum pools (from the serum of boys 2-15 years old ) by all methods and in 33 serum pools (from the serum of adult men) only by immunoassay.

RESULTS

Differences in testosterone concentrations determined by various immunoassay ranged from -60 to +29.2%. Agreement testosterone concentration measurements on the three Architect analyzers was good (95% limits of agreement was -8.3% to 8.1%). Mathematical correction of the results of the Architect and Immulite methods to HPLC-MS / MS reduce the bias between methods from 27.9% to 5.6%, but the scatter of results remains rather large.

CONCLUSION

The pronounced dependence of the measured serum testosterone concentrations on the used measurement method was confirmed and fundamental possibility of harmonization of the results by means of mathematical correction to HPLC-MS/MS was shown.