Comparison of immunofluorescence and chemiluminescence assays for measuring ACTH in equine plasma.

BACKGROUND: The chemiluminescence (CL) and immunofluorescence (IF) assays yield different results for basal adrenocorticotropin hormone concentrations [ACTH] in pony plasma. It is unclear whether this difference also occurs in basal samples from horses or samples from ponies following thyrotropin-releasing hormone (TRH) stimulation.

OBJECTIVES: To compare the results of [ACTH] analysis by CL and IF methods in basal samples from horses and pony samples following TRH stimulation.

STUDY DESIGN: Method comparison.

METHODS: Plasma [ACTH] was measured concurrently using CL and IF methods in 12 ponies (basal and post-TRH stimulation) in November and basal samples from horses (n = 45; November and May).

RESULTS: CL and IF methods yielded different results (P < .01). The median difference (CL-IF) (95% CI) for ponies was 5.9 (0.1-7.5) pg/mL at baseline and 227.9 (61-1001) pg/mL post TRH; and horses 1.9 (1.1-5.4) pg/mL in November and 9.4 (8.2-11.5) pg/mL in May, at baseline. Correlation was good in ponies at baseline (R = 0.80, P = .003) but not post-TRH, and good in horses in November and May (R = 0.68 and 0.71, P < .001). Bland-Altman analysis demonstrated moderate bias and wide 95% limits of agreement (95% LOA) in ponies at baseline (bias 5.5 pg/mL; 95% LOA -9.9 to 20.9 pg/mL) and horses in May (bias 10.6 pg/mL; 95% LOA -9 to 30.3 pg/mL) and very large bias and wide 95% LOA in ponies post-TRH (bias 477 pg/mL; 95% LOA -633 to 1587 pg/mL). Using CL cut-offs of >29 and >110 pg/mL, agreement was moderate (ƙ = 0.67) and very good (ƙ = 0.82) for binary classification of PPID in ponies at baseline and post-TRH; and good (ƙ = 0.73) for horses in November, but poor (ƙ = 0.40) in May.

MAIN LIMITATIONS: Limited numbers of horses with [ACTH] above threshold values.

CONCLUSIONS: The assays yielded different absolute values, particularly in post-TRH samples from ponies, suggesting TRH stimulates secretion of cross-reacting peptides other than ACTH. Agreement for binary classification for PPID was moderate to good, except in basal samples from horses in May.