# Free thyroxine measured by chemiluminescence and equilibrium dialysis is frequently below the reference interval in known euthyroid dogs with nonthyroidal illness syndrome

## TL;DR

This study finds that free thyroxine tests may incorrectly suggest hypothyroidism in euthyroid dogs with nonthyroidal illness syndrome when total thyroxine is low.

## Contribution

The study evaluates agreement and diagnostic accuracy of two free thyroxine assays in euthyroid dogs with nonthyroidal illness syndrome.

## Key findings

- Both fT4ED and fT4CLIA assays frequently report free thyroxine below reference intervals in euthyroid dogs with nonthyroidal illness syndrome.
- Concordant results between fT4ED and fT4CLIA occurred in 77% of dogs with low total thyroxine.
- This misclassification risk highlights limitations in using these assays to distinguish hypothyroidism from nonthyroidal illness syndrome.

## Abstract

Measuring serum free thyroxine (fT4) concentration is recommended to distinguish hypothyroidism from nonthyroidal illness syndrome (NTIS) in dogs having a serum total thyroxine (TT4) concentration below the reference interval (RI). Serum fT4 concentration can be measured by equilibrium dialysis (fT4ED) or chemiluminescent immunoassay (fT4CLIA), neither of which have been extensively evaluated in known euthyroid dogs with NTIS. The objectives of this study were to: (1) determine agreement between fT4ED and fT4CLIA in serum samples from established euthyroid dogs with NTIS, (2) evaluate how frequently each fT4 assay is within the RI when the serum TT4 concentration is within and below the RI, and (3) evaluate concordant and discordant fT4ED and fT4CLIA results in dogs with a serum TT4 concentration below the RI.

This retrospective study used 116 banked serum samples from 38 known euthyroid dogs with NTIS. On each sample, fT4ED and fT4CLIA were measured, and agreement assessed by a Bland–Altman plot. Using the hospital admission serum TT4 concentration, dogs were stratified into two groups: TT4 within the RI (n = 16) and TT4 below the RI (n = 22). The frequency that each fT4 assay was within the RI was calculated in both groups. In dogs having a serum TT4 concentration below the RI, concordant and discordant fT4 results were evaluated.

Assay comparison showed agreement, with no significant systematic or proportional bias. When the serum TT4 concentration was within the RI, fT4ED and fT4CLIA were within the RI in 100% (95% CI: 79.4–100%) and 94% (95% CI: 69.8–99.8%) of dogs, respectively; however, when the serum TT4 concentration was below the RI, fT4ED and fT4CLIA were within the RI in 41% (95% CI: 20.7–63.6%) and 45% (95% CI: 24.4–67.8%) of dogs, respectively. Concordant fT4ED and fT4CLIA were present in 77% (95% CI: 54.6–92.2%) of dogs with a serum TT4 concentration below the RI.

The fT4ED and fT4CLIA were regularly below the RI in dogs with NTIS and a serum TT4 concentration below the RI, potentially resulting in a misdiagnosis of hypothyroidism. This raises concerns about the ability of both fT4 assays to distinguish hypothyroidism from NTIS in this subset of dogs.