The ability to predict pregnancy loss in cattle with ELISAs that detect pregnancy associated glycoproteins is antibody dependent

The concentration of circulating pregnancy associated glycoproteins (PAGs) early in pregnancy may serve as markers to predict late embryonic mortality or fetal mortality in cattle. In this study, pregnancies were established in dairy cows, by either fixed-time AI (FTAI) or fixed-time embryo transfer (FTET) with in vitro produced embryos. Circulating PAGs were measured with different combinations of antibodies in either a laboratory-based ELISA or a commercial ELISA. For the in-house ELISA, three monoclonal 'trapping' antibodies (A6, J2, and L4) and two polyclonal 'detection' antisera (antibodies F2 or 45) were used to quantify PAGs in serum from the same cows. The different assays were identified as follows: 'Mix-45' (A6, J2, and L4 with 45), 'Mix-F2' (A6,12, and L4 with F2), and 'L4-F2': (L4 with F2); the commercial assay was from IDEXX. Ovulation was synchronized and FTAI or FTET was performed on day 0 or 7, respectively. Ultrasound-based diagnosis of pregnancy and serum collections occurred on day 30. The proportion of cows that subsequently experienced pregnancy loss between days 30 and 60 was 23% (43 of 183) and 16% (21 of 131) for the FTAI or FTET groups, respectively. In the FTAI group, mean serum concentration of PAGs detected with Mix-45 was higher in cows that maintained pregnancy (9.2 +/- 0.4 ng/ml; mean +/- SEM) compared with cows that experienced pregnancy failure (3.9 +/- 0.6 ng/ml) between day 30-60 (P < .001). However, there was no difference (P > .69) in circulating concentrations of PAGs between cows that experienced loss or survival between days 30 and 60 when Mix-F2 or L4-F2 were used in an in-house ELISA. Likewise, a commercial assay also did not result in measurable differences in PAG concentrations between those animals that experienced loss or survival. Following HET, circulating concentrations of PAGs on day 30 were lower (P <.001) in cows that experienced pregnancy failure compared to cows that maintained pregnancy when the Mix-45 and the commercial assay were used, but not with the other antibody combinations. A receiver operating characteristic curve showed that only the Mix-45 antibody combination was predictive (95% accuracy) of pregnancy loss but not the other antibody combinations following FTAI. However, both Mix-45 and the commercial assay were predictive of losses following FTET. In summary, although multiple PAG assay formats have been shown to accurately detect pregnancy, the ability to predict embryo survival during early gestation appears to be antibody dependent. (C) 2017 Elsevier Inc. All rights reserved.