Association of early beta-human chorionic gonadotrophin values with pregnancy wastage and multiple implantation in a donor oocyte programme.

An early marker predictive of a viable pregnancy would ease the anxiety associated with positive pregnancy tests after the use of donor oocytes. We examined the predictive value of an early serum quantitative human chorionic gonadotrophin (Q-HCG) concentration on pregnancy outcome following oocyte donation. Embryo transfers after oocyte donation resulting in a positive serum beta-HCG were examined beginning 9 days after embryo transfer from those samples assayed in our laboratory (n = 77). Q-HCG concentrations were measured in our laboratory by an immunoradiometric assay utilizing the first International Reference Preparation. Implantations were defined as the number of gestational sacs visualized by transvaginal ultrasound 21 days after embryo transfer. Biochemical pregnancies were those with transient elevations in beta-HCG concentration but without implantation sites. Spontaneous abortions were characterized by an implantation site with the eventual arrest of development. Ongoing/delivered pregnancies developed appropriately and proceeded beyond the first trimester. Day 9 Q-HCG concentrations did not differentiate between biochemical pregnancies/spontaneous abortions and ongoing/delivered pregnancies, although mean +/- SD concentrations for biochemical pregnancies were significantly lower than those for the other groups (P < 0.0001): biochemical pregnancies, n = 18, 5.8 +/- 8.9 mIU/ml, range 0-35; spontaneous abortions, n = 2, 46.0 +/- 10.0 mIU/ml, range 39-53; ongoing/delivered pregnancies, n = 57, 41.5 +/- 35.4 mIU/ml, range 0-214. In addition, day 9 Q-HCG concentrations did not differentiate between multiple implantations, although the implantation of four sacs had a significantly higher mean Q-HCG concentration compared with the implantation of fewer sacs (P < 0.0001): one sac, n = 22, 32.2 +/- 21.5 mIU/ml, range 3-78; two sacs, n = 25, 35.8 +/- 21.3, range 0-81; three sacs, n = 7, 47.1 +/- 37.1 mIU/ml, range 22-126; four sacs, n = 4, 122.3 +/- 62.4 mIU/ml, range 76-214. The positive predictive value of a Q-HCG > 10 mIU/ml was 0.91 (sensitivity 91%, specificity 75%). These initial data suggest that early day 9 serum Q-HCG determinations do not accurately identify viable pregnancies or multiple implantations. Even an early negative pregnancy test should be repeated because it can be associated with a normal pregnancy.