Progesterone determination in equine plasma using different immunoassays

Several assay systems (3H radioimmunoassay (RIA) with and without extraction; microplate enzyme-linked immunoassay (ELISA); qualitative ELISA (tube test)] were used to measure plasma progesterone concentration in mare plasma. The direct RIA showed a close correlation (R = 0.94) with the extraction RIA. The direct RIA and the microplate ELISA were compared in two different studies. In the first study 1155 samples of postpartum mares were used for progesterone determination with both assays. The ELISA resulted in more elevated values both in oestrus and dioestrus (0.19+/-0.3 and 2.44+/-3.62 nmol/l for oestrus, n = 436, and 8.94+/-4.29 and 27.88+/-18.34 nmol/l for dioestrus, N = 719, for the RIA and ELISA, respectively, R = 0.71). The evaluation of individual progesterone profiles has revealed that the microplate ELISA detects the time of ovulation at the same time as it is determined by the RIA and clinical examination. The sensitivity and specificity were calculated for different progesterone threshold values. In the second study including 7 non-pregnant, cycling mares the progesterone concentration of 240 samples was determined by both assays. Basal values (Day 0) obtained with the ELISA were higher (1.57 nmol/l) than those of the RIA (0.2 nmol/l). Both curves reached the same maximum concentration (12.11 and 12.45 nmol/l) 5 days after ovulation. The correlation between the RIA and ELISA values was high (R = 0.90). The tube test was compared to the microplate ELISA as reference using 576 plasma samples of 34 non-pregnant, non-cycling mares included in an ovulation induction study. Of these samples 118 had higher and 458 had lower values than 3.18 nmol/l. In most cases the tube test was in complete agreement with the microplate ELISA. The sensitivity, specificity, + predictive and - predictive values for the tube test were 79.7%, 95.4%, 81.7% and 94.8%, respectively.