Salivary cortisol measurement in horses: immunoassay or LC-MS/MS?

This research aims to compare two methods for measuring cortisol levels in horse saliva: liquid chromatography-tandem-mass spectrometry (LC-MS/MS) and cELISA immunoassay. The scientists used these methods to measure the cortisol levels before and after stimulation of the hypothalamic-pituitary-adrenal (HPA) axis. The study found that immunoassay provided higher but correlated results for elevated cortisol levels, while LC-MS/MS delivered more accurate results, especially for baseline cortisol levels.

Study Design and Execution

• The study involved nine clinically healthy horses undergoing ACTH stimulation tests. In these tests, cortisol levels were measured before (T0) and 60 minutes after (T60) the horses were intravenously injected with synthetic ACTH1-24. This synthetic substance stimulates the HPA axis, causing an increase in cortisol levels.
• For the determination of equine salivary cortisol, LC-MS/MS was used, a proven analytical technique known for its precision and accuracy. The researchers compared these results with data obtained using a cELISA — an immunoassay previously validated for use in horses.

Results Interpretation

• The comparison of the two methods showed no correlation at T0 (r = -0.2452; P = 0.5249), indicating that immunoassay and LC-MS/MS did not align in their measurements of baseline cortisol levels. However, significant correlation was observed at T60 (r = 0.8334; P = 0.0053), meaning both methods showed similar trends in increased cortisol levels after HPA axis stimulation.
• A detailed analysis (Bland-Altman-Plots of T60) revealed that measurements obtained through immunoassays often resulted in higher values compared to LC-MS/MS. In fact, on average, immunoassay outcome values were 2.3 times higher.

Underlying Factors and Conclusions

• The lack of agreement between the methods at T0 is attributed to potential inaccuracies of the immunoassay in the low measurement range, structurally similar cross-reacting agents, and matrix effects, which can all bias the results.
• The overestimation of results by the immunoassay at T60 is possibly due to differences in how the methods are standardized, unavoidable matrix effects on the antigen-antibody interaction in the ELISA, and potential cross-reactions of other steroids.
• The research concludes that while immunoassay results of equine salivary cortisol yielded higher but reasonably correlated results after stimulation of the HPA axis, LC-MS/MS provided better accuracy, particularly for baseline cortisol concentrations.