Cortisol and free thyroxine determination by time-resolved fluorometry in canine serum

Validation for canine serum of 2 commercially available time-resolved fluoroimmunoassays (TR-FIAs) designed for analysis of cortisol and free thyroxine (fT4) in human serum was carried out. Included was the study of interference by hemolysis, lipemia, and bilirubinemia. With the dissociation enhancement lanthanide fluoroimmunoassay kits, the intra-assay coefficient of variation (CV) ranged from 6.4% to 8.7% for cortisol and from 5.3% to 9.8% for fT4; the interassay CVs ranged from 5.8% to 10.8% and from 3.9% to 14.1%, respectively. Accuracy was evaluated by comparing cortisol and fT4 results obtained with TR-FIA and those obtained with a validated enzyme-linked immunosorbent assay (ELISA) and an equilibrium dialysis (ED) assay, respectively. The regression equations obtained were y = 0.57x + 1.18 (r2 = 0.90) for cortisol and y = 0.87x + 0.82 (r2 = 0.93) for fT4. The limits of detection for cortisol and fT4 were 4.84 nmol/L and 2.68 pmol/L, respectively. The results of adrenocorticotropin-stimulation and dexamethasone-suppression tests were similar to those published previously; likewise, serial dilution of a canine serum sample with a high cortisol content demonstrated that the TR-FIA was immunologically specific. Serial dilution of a serum sample with a high fT4 concentration showed a methodologic bias, a dependence on serum binding capacity, which indicates that the results obtained with this method should be interpreted with caution. Finally, hemolysis and lipemia significantly interfered with cortisol and fT4 measurements, whereas bilirubinemia did not affect the results.