Pregnancy associated glycoprotein predicts pregnancy losses in dairy cattle

The objective of this study was to determine the association between concentrations of pregnancy associated glycoprotein (PAGs) and pregnancy loss in lactating dairy cattle following fixed-timed artificial insemination (TAI) or timed embryo transfer (TET). Three experiments were performed, on experiment 01 (TAI - artificial insemination on d0) and 02 (TET - embryo transfer on d7) the pregnancy diagnosis was performed by ultrassonography at d31. In pregnant cows blood samples were collected and serum concentrations of PAG were determined by ELISA. On d59, a second pregnancy check was performed and the pregnancy loss was determined when a cow had a viable embryo on d31 of gestation but not on d59. In study 03 (TET - TE d7) cows were considered pregnant based on the presence of a vascularized corpus luteum on d24 (evaluated with a color doppler US) ipsilateral to the CL at TET, on days 31, 38, 45, 52 and 59 the cows were scanned to evaluate a viable embryo/fetus and blood samples were collected to evaluate the progesterone concentration (P4) and PAGs. One-way ANOVA (SAS 9.4) was used to evaluate differences of PAGs among d31, the LOGISTIC procedure was used to determine the probability of pregnancy maintenance. To test of the effectiveness of a single circulating PAG sample on d31 to predict pregnancy loss, a receiver-operating characteristic (ROC) curve was generated on MedCalc software package to determine PAG concentrations on d31 that should predict pregnancy loss with ≥95 accuracy in both TAI and TET. A total of 77 cows experienced pregnancy loss on TAI (exp 01/n=413) and 47 cows experienced pregnancy loss on TET (exp 02/n=238). Cows that were pregnant at d31 and maintained the pregnancy until d59 had higher (P 0.05) of PAGs and P4 concentration on d24 in pregnancy maintenance until d31. The P4 concentration at d31 not interfere (P>0.05) in pregnancy maintenance; however, there was effect (P<0.05) of PAGs at d31 in pregnancy loss, as observed in experiments 01 and 02. The majority of pregnancy losses occurred between 24 and 31 (n=20; 20%), followed by days 31 and 38 (n=8, 10%), 38 and 45 (n=6, 7.5 %) and 45 to 59 and (n=1, 1%). In conclusion, PAGs concentration on d31 of pregnancy can be successfully used to detect pregnancy losses occurring after d31.