Gastrin is an important hormone involved in gastric acid secretion. Despite its importance, validated methods other than radioimmunoassay (RIA) to assess serum gastrin concentrations in horses are lacking. This study aims to determine the agreement between ELISA and RIA in quantifying equine serum gastrin concentrations. Serum gastrin concentrations were quantified using two ELISA kits and RIA. Samples (196) from 14 horses at different time points were analyzed using one ELISA kit and RIA, selected samples (7) were analyzed using a second ELISA kit, and the correlation between methods was calculated. The level of agreement was analyzed by Bland-Altman analysis and differences between ELISA and RIA were plotted against averages for each sample. The Pearson correlation between gastrin concentrations measured by ELISA and the RIA was 0.27 and -0.32 for ELISA kit 1 and kit 2, respectively. Mean bias (ELISA-RIA) was 198.40 pg/mL (95% CI: -142.95-539.76) and -17.90 pg/mL (95% CI: -89.98-54.19) for ELISA kit 1 and kit 2, respectively. Measurements of horse gastrin by both ELISA methods were highly variable, with an unacceptable correlation to the reference method, RIA. Using non-validated ELISA methods to quantify horse gastrin cannot be recommended.
Read full abstract