Evaluation of a new immunoassay for measuring equine serum insulin and plasma adrenocorticotrophic hormone: A method comparison study

Insulin dysregulation (ID) and pituitary pars intermedia dysfunction (PPID) are common equine endocrinopathies detected by measurement of circulating insulin and adrenocorticotrophic hormone (ACTH) concentrations. Different assays may yield variable results, limiting comparability.

To compare immunoassays developed specifically to measure equine insulin (VIDAS® Equine Insulin; enzyme-linked fluorescent assay [ELFA]) and equine ACTH (VIDAS® Equine ACTH; ELFA) with results obtained from a chemiluminescence immunoassay (CLIA) measured using IMMULITE 2000XPi, a chemiluminescence enzyme immunoassay (CLEIA) measured using TOSOH Automated Immunoassay Analyser-360, and an equine insulin enzyme-linked immunosorbent assay (ELISA; Mercodia).

Residual serum (n = 242) and plasma (n = 337) from samples submitted to a commercial laboratory were simultaneously assayed using the CLIA and ELFA as appropriate. Concentrations were also measured in a subset using CLEIA (n = 126 insulin; n = 76 ACTH) and ELISA (n = 76 insulin) assays. Additionally, plasma [ACTH] was measured using the ELFA in samples from 139 healthy horses.

The median (95% confidence interval) difference for insulin assays was 1.6 (0.1-3.3) μIU/mL for CLIA-ELFA, 11.3 (8.7-13.4) μIU/mL for CLEIA-ELFA, 3.7 (1.0-6.3) μIU/mL for ELISA-ELFA, 7.3 (5.1-9.5) μIU/mL for CLIA-CLEIA; and for ACTH assays was 19.7 (16.8-25.1) pg/mL for CLIA-ELFA, 10.7 (6.9-15.5) pg/mL for CLEIA-ELFA, 23.7 (9.0-42.5) pg/mL for CLIA-CLEIA. Correlation was strong (R > 0.8) or very strong (R > 0.9) between pairs of insulin assays and between ELFA and CLEIA ACTH assays, and moderate (R > 0.6) between other ACTH assay pairs. Bland-Altman analysis revealed moderate bias and wide 95% limits of agreement between all assay pairs. Three-way classification using adapted diagnostic cut-offs for [insulin] and [ACTH] measured using ELFA indicated substantial agreement. A reference range for plasma [ACTH] in healthy horses was determined.

Residual samples were collected between July and November and from healthy horses but not ponies.

The assays yielded different absolute values, but agreement between most assay pairs and for three-way classification for PPID diagnosis and laminitis risk was good.