76 Can serum pregnancy-specific protein B concentration be used to accurately diagnose early pregnancy in dairy cattle with a single sample from days 18 to 24 post-AI?

Our aim was to determine if a single-day sample of serum pregnancy-specific protein B (PSPB) concentration from 18 to 24 days post-AI can be used to diagnose pregnancy with a high accuracy in dairy cattle. Ovulatory response between AI (d = 0) and 1 day later was evaluated in 614 heifers and cows from the same herd that were submitted to timed-AI programs. Animals that ovulated (n = 514) had blood samples collected daily from Days 18 to 24 (D18 to D24) and at D27 post-AI. On D32 post-AI, pregnancy was diagnosed by ultrasound. Serum PSPB concentrations were measured by ELISA (BioPRYN, BioTracking LLC) on a preliminary subsample of animals (n = 395). Animals with a 20% daily increase in PSPB on 2 consecutive days with at least one sample of PSPB ≥ 1.00 ng mL⁻¹ were considered with a conceptus presence (CP) from D18 to D24, as an early pregnancy diagnosis. Single-day serum PSPB concentrations were tested as an early and D32 pregnancy predictor using ROC curve analyses in RStudio. Early P/AI and D32 P/AI were 62.5% (247/395) and 55.7% (220/395), respectively. Pregnancy loss from early to D32 was different between parities (P < 0.01): heifer, 0%, n = 38; primiparous, 6.6%, n = 61; and multiparous, 15.5%, n = 148. A requirement of an early pregnancy diagnosis is high specificity combined with a sensitivity of 100% (0% false negative rate) because a re-insemination strategy using PGF_2α may lead to abortion. Day for achieving 100% sensitivity was different according to parity group independent of whether we used CP or D32 pregnancy as a reference (Table 1). Because heifers did not have pregnancy loss between early conceptus presence detection and D32 post-AI, PSPB achieved 100% sensitivity and 100% specificity (AUC = 1) on D22 post-AI in heifers as a predictor for CP or D32 pregnancy. In primiparous cows, PSPB achieved AUC = 1 on D23 post-AI as a predictor for CP, but maximum specificity for predicting D32 pregnancy with 100% sensitivity was 92.7% on D23 post-AI. In multiparous cows, the maximum specificity achieved with 100% sensitivity was on D24 for predicting CP (95.1%) or D32 pregnancy (79.8%). To summarize, our results demonstrate that a single-day measurement of serum PSPB concentration can identify nonpregnant heifers and cows around conceptus attachment with a high accuracy. Yet, parity affected accuracy of the diagnosis and day post-AI with better accuracy, and it should be considered when using serum PSPB concentrations as an early pregnancy diagnosis.