Interconversion of Plasma Free Thyroxine Values from Assay Platforms with Different Reference Intervals Using Linear Transformation Methods

Abstract

Simple Summary Thyroid hormones are extremely potent and exert a broad range of biological actions on many organ systems of all vertebrates including humans. Blood concentration of thyroid hormones mirrors thyroid status quite well. As thyroid hormone excess or deficiency can lead to serious diseases, it is crucial to ensure that measurement techniques of blood thyroid hormones are accurate and precise, especially during the treatment of an overactive or underactive thyroid. Until recently, many laboratories employ different methods of analysis of thyroid hormones, resulting in reports showing values calibrated to dissimilar normal ranges. This becomes a major issue for patients who are tested in different healthcare facilities as it is challenging to interpret their thyroid status and decide if any difference is due to a real change in hormone concentration or whether the variations occurred purely from calibration differences. In this study, we test the reliability of a mathematical system using linear transformation strategies to convert one value in one scale to another value in a separate scale. Via simultaneously analyzed unbound fraction of plasma thyroxine using three different techniques-immunoassay, mass spectrometry and equilibrium dialysis, we show that linear methods are quite successful in achieving accurate inter-scale thyroid hormone conversions. Abstract Clinicians often encounter thyroid function tests (TFT) comprising serum/plasma free thyroxine (FT4) and thyroid stimulating hormone (TSH) measured using different assay platforms during the course of follow-up evaluations which complicates reliable comparison and interpretation of TFT changes. Although interconversion between concentration units is straightforward, the validity of interconversion of FT4/TSH values from one assay platform to another with different reference intervals remains questionable. This study aims to establish an accurate and reliable methodology of interconverting FT4 by any laboratory to an equivalent FT4 value scaled to a reference range of interest via linear transformation methods. As a proof-of-concept, FT4 was simultaneously assayed by direct analog immunoassay, tandem mass spectrometry and equilibrium dialysis. Both linear and piecewise linear transformations proved relatively accurate for FT4 inter-scale conversion. Linear transformation performs better when FT4 are converted from a more accurate to a less accurate assay platform. The converse is true, whereby piecewise linear transformation is superior to linear transformation when converting values from a less accurate method to a more robust assay platform. Such transformations can potentially apply to other biochemical analytes scale conversions, including TSH. This aids interpretation of TFT trends while monitoring the treatment of patients with thyroid disorders.