Validation of plasma thyroxine and triiodothyronine methods on the ADVIA Centaur^® XP

Abstract

Standardization programs for thyroid hormones have revealed bias between immunochemical methods and the reference method ED-ID-LC/MS. Lack of standardization between methods, suboptimal reference intervals and replacement of serum with plasma may compromise the capability of the immunochemical thyroid methods to diagnose thyroid disease. To accommodate the demand for faster turn-around times for laboratory replies, we replaced serum with plasma on some serum CE marked thyroid methods. This forced us to do on-board analytical correction for the plasma total T4 (TT4) method on ADVIA Centaur^® XP. We, next, validated the capability of the ADVIA Centaur^® XP thyroid methods on plasma by (1) first carrying out a prospective method comparison with the ED-ID-LC/MS reference method using collected plasma samples, (2) we verified the clinical reference intervals by analyzing collected plasma samples from healthy individuals, and (3) retrospectively compared laboratory results from two different time periods using serum TT4 and serum total triiodothyronine (TT3) versus plasma free thyroxine (FT4) and plasma TT3, respectively, to diagnose thyroid disease. The plasma FT4 method displayed a negative concentration-dependent bias against the reference method. This bias was apparently counteracted by a fitted reference interval for the plasma FT4 method. Indeed, overt hyperthyroid disease was found in 1.0% and 1.1% of the cases using serum and plasma and overt hypothyroid condition were in 1.3% and 0.6% of the cases using serum and plasma, respectively. In conclusion, the ADVIA Centaur^® XP FT4 method displayed a negative bias at high plasma FT4 concentrations against the reference method, but the diagnostic performance was not compromised due to a fitted reference interval.

Keywords:

• Immunochemistry
• mass spectrometry
• thyroid disease
• thyrotropin
• thyroxine
• triiodothyronine

Acknowledgments

The skillful technical assistance of Anette Hansen and Pernille Holm Nielsen is greatly appreciated.

Disclosure statement

No potential conflict of interest was reported by the authors.