Validation of an embryotoxicity assay.

Abstract

Culture of mouse blastocysts has served as a tool for identifying various embryotoxic factors in human serum. While inactivated, sera from recurrently aborting women inhibit mouse blastocyst development in vitro. Variation in results from individual serum samples has limited the usefulness of this assay in establishing a new classification of idiopathic recurrent spontaneous abortion (RSA). Two-cell embryos were collected from superovulated mated CB6F1/J mice and cultured in Ham's F-10 media supplemented with 10% fetal bovine serum (FBS) or tested human serum at 37 degrees C with 5% CO2 and high humidity. Each sample was assayed in triplicate using three mice with at least five embryos from the same mouse per dish. Development was evaluated at 72 h and the frequency of atretic embryos was recorded. Intrasample (interassay) variation yielded a coefficient of variation of 9%. When repeated, samples from a given individual were evaluated and the coefficient of variation was 8.7%. Interoperator variability was 4% interassay and 2% intrassay. Atresia of embryos was 23% when incubated with FBS (N = 122), 21% in FC (N = 122), and in the sera of patients with RSA 34.6% (N = 95). Results of percentage of atresia from the fertile control group had a nonparametric distribution. Using 2.2 multiples of the median to determine the 95% confidence interval, a threshold at 44.0% of atresia was established. The critical step in maintaining low variability in this bioassay is to control mouse variability by averaging the percentage atresia from different mice as embryo donors for each tested serum. A subgroup of 24% (23/95) RSA patients who displayed embryotoxic activity was identified with a specificity of 95% and positive predictive value of 83%, P = 0.001.