Evaluation of an Automated Fluorescence Enzyme Immunoassay for Quantification of Equine Insulin and Comparison to Five Other Immunoassays.
Authors: Nolen-Walston Rose D, Kulp Jeaneen C, Stefanovski Darko, van Eps Andrew W
Journal: Journal of veterinary internal medicine
Summary
# Editorial Summary: Equine Insulin Assay Comparison Accurate insulin measurement is critical for identifying hyperinsulinemia and assessing laminitis risk in horses, yet the reliability of different laboratory assays has remained unclear until now. Nolen-Walston and colleagues evaluated the Tosoh AIA-360 automated fluorescence enzyme immunoassay against a reference radioactive immunoassay (RIA) and five alternative methods across 100 serum samples, comparing their accuracy, precision, and agreement across different insulin concentration ranges. The Tosoh assay demonstrated excellent correlation with RIA (r² = 0.94) and clinically acceptable precision (1.8% intra-assay, 5.7% inter-assay variation), though it showed a constant negative bias of approximately 24 µIU/mL; notably, both ELISA and the Cobas e chemiluminescent assay performed acceptably at insulin concentrations below 100 µIU/mL, whilst both Immulite analysers produced unacceptably high results in roughly 10% of samples. These findings suggest that practitioners should be cautious about which assay their laboratory uses—results from Immulite systems warrant verification with alternative methods—and that the choice of assay may influence interpretation of borderline hyperinsulinaemia cases, particularly when insulin levels cluster around diagnostic thresholds for metabolic disease screening.
Read the full abstract on PubMed
Practical Takeaways
- •The Tosoh AIA-360 is a reliable automated platform for measuring insulin in horses and can be confidently used to identify and monitor hyperinsulinemia, a treatable laminitis risk factor
- •Do not use Immulite 2000/2000XPi analyzers for equine insulin measurement as they produce unreliable results; ELISA and Cobas e are acceptable alternatives if Tosoh is unavailable
- •When monitoring insulin-sensitive horses or those at laminitis risk, consistent use of the same assay is important given inter-assay variation; document which method was used for longitudinal comparisons
Key Findings
- •TOS-FEIA showed excellent correlation with RIA (r² = 0.94) with acceptable agreement (constant bias -23.8 μIU/mL) and good precision (1.8% intra-assay, 5.7% inter-assay)
- •ELISA and Cobas e CLIA demonstrated good agreement with RIA in samples <100 μIU/mL, though Cobas e showed marked bias
- •Immulite 2000 and 2000XPi CLIA analyzers produced spuriously high results in ~10% of samples (2 to >10-fold higher than RIA), rendering them unsuitable for clinical use
- •TOS-FEIA had acceptable accuracy and precision for clinical insulin measurement including at concentrations <100 μIU/mL