Effect of metformin on mouse embryo development

Polycystic ovary syndrome (PCOS) is a common disorder affecting women of reproductive age. Its cardinal features are hyperandrogenism, chronic anovulation, and infertility. Some of the most important pathogenic associations of PCOS are insulin resistance and compensatory hyperinsulinemia (1, 2, 3). The hyperinsulinemia in turn stimulates ovarian androgen production. Several clinical studies have demonstrated that a reduction in the insulin level results in decreased ovarian androgen production (4, 5, 6). Metformin is a biguanide used extensively in type 2 diabetes. It inhibits hepatic gluconeogenesis and increases peripheral insulin sensitivity and glucose uptake but does not cause hypoglycemia (6, 7, 8). The correction of the hyperinsulinemia by metformin in PCOS patients is accompanied by normalization of the elevated serum LH and ovarian androgen levels, leading to resumption of spontaneous ovulatory cycles, or responsiveness to ovulation induction by clomiphene citrate (CC) (1, 4, 5, 6, 8, 9). Clinical use of this agent for ovulation induction in CC-resistant PCOS patients has increased with little knowledge of the possible effect on early embryo development. We sought to assess whether metformin was embryotoxic using a mouse embryo model. During controlled clinical trials, plasma metformin levels do not exceed 5 μg/mL, even at maximum doses (10). Metformin hydrochloride (N, N-dimethylimidodicarbonimidic diamide hydrochloride; Sigma, St. Louis, MO) was dissolved in human tubal fluid (HTF) media (Irvine Scientific, Santa Ana, CA) to give working concentrations of 5, 25, and 100 μg. The culture dishes containing 1 mL of media at each concentration were incubated for overnight equilibration at 37°C and 5% CO2. Thawed two-cell mouse embryos (Embryotech, Wilmington, MA) were pooled and randomly distributed into the culture dishes containing the various concentrations of metformin. The control consisted of HTF media alone (group 1). Groups 2, 3, and 4 were supplemented with metformin (5, 25, and 100 μg/mL). There were 40 embryos in each group. The embryos were incubated in an atmosphere of 5% CO2 at 37°C for 72 hours. The blastocyst development rate (BDR) was then calculated by dividing the number of embryos reaching the blastocyst stage by the total number of embryos cultured. The relationship between BDR and concentration was assessed with repeated measures logistic regression using generalized estimated equation (GEE) methodology with a compound symmetry correlation structure. The sample size was sufficient to detect whether a specific concentration reduced the odds of development by a factor of 2.5. Statistical significance was assessed using two-tailed P<.05. Statistical computations were performed with SAS version 8.1 (SAS Institute Inc, Cary, NC). . .