Validation and application of solid-phase chemiluminescent immunoassays for diagnosis of endocrine diseases in animals

Significant progress has been made in the ability of veterinarians to diagnose endocrine diseases since introduction and development of the radioimmunoassay (RIA) in the 1960s. Alternatives to the radioactive label have been commercialised for human diagnostic applications. Because healthy animals have concentrations of several hormones in serum that are lower than in humans, these non-isotopic assays generally lack the sensitivity for measuring hormones in animals. However, the chemiluminescent enzyme immunoassay may provide this needed sensitivity. We validated the DPC-Cirrus Immulite® automated chemiluminescent immunoassay system for measurement of cortisol in dogs, cats and horses. Serial dilutions of serum inhibited binding of alkaline phosphatase-labelled cortisol in a manner that was parallel with inhibition produced by the stored master calibration curve. Six dogs were injected with adrenocorticotropin (ACTH; 2.2 IU/kg i.m.) and five dogs were injected with dexamethasone (0.015 mg/kg i.v.). Changes in cortisol concentrations measured with the Immulite® were as expected based on published results. Similar tests of biological specificity in other species yielded similar results. Accuracy of the Immulite® cortisol assay was evaluated by comparing cortisol concentrations in serum with concentrations obtained by a validated RIA. The regression equations were: y = 14.6 + 0.8χ for dogs (n = 59, r2 = 0.91), y = 13.5 + 0.9χ for horses (n = 23, r2 = 0.90), and y = 29.8 + 1. Oχ for cats (n = 18, r2 = 0.96), where y = Immulite® result (nmol/l) and χ = RIA result (nmol/l). The regression equation for the canine urinary cortisol: creatinine ratio was y = 0.43 + 1.55χ, where y = ratio calculated from Immulite® cortisol concentrations and χ = ratio using RIA concentrations (n = 35, r2 = 0.93). Initially, the lower limit of the reportable range by the Immulite® was 27.6 nmol/l, a value that was unacceptable for veterinary diagnostics. After consultation with DPC-Cirrus, we adjusted the lower limit to 1.4 nmol/l. Coefficients of variation for two canine, one feline, and two equine quality-control serum samples included in 73 consecutive Immulite® assays ranged from 0.29 (mean = 29.0 nmol/l) to 0.09 (mean = 261.6 nmol/l). We are using the Immulite® routinely for the diagnosis of adrenal disease in animals and have found it to provide diagnostically useful cortisol results for dexamethasone-suppression and ACTH-stimulation tests. Its diagnostic accuracy on single baseline samples is probably less than for the RIA, particularly at concentrations of cortisol less than 27.6 nmol/l.