Journal of Clinical Endocrinology & Metabolism, Vol 63, 1284-1291, Copyright © 1986 by Endocrine Society

ARTICLES

Ovulation induction with human menopausal gonadotropin compared to human urinary follicle-stimulating hormone results in a significant shift in follicular fluid androgen levels without discernible differences in granulosa-luteal cell function

ML Polan, A Daniele, JB Russell and AH DeCherney

Follicular fluid estradiol, progesterone, testosterone, and androstenedione levels were compared in 2 groups of spontaneously ovulatory women undergoing ovulation induction with human menopausal gonadotropin (hMG; which contains equal amounts of LH and FSH) or human urinary FSH (huFSH). The results were correlated with the ratios of embryo cleavage and pregnancy. Although significantly more FSH [1268 +/- 38 (+/- SEM) vs. 953 +/- 38 IU; P less than 0.05] was required for equivalent hyperstimulation in hMG compared to huFSH cycles, the number of oocytes retrieved and fertilized and the number of embryos transferred were similar for the 2 ovulation induction protocols. Forty- two follicles from 21 women stimulated with hMG and 38 follicles from 15 women stimulated with huFSH were examined and found to be representative of the total cohort of aspirated follicles. Follicular fluid estradiol and progesterone levels were similar, but hMG- stimulated follicles contained significantly more testosterone [7.83 +/- 0.52 (+/- SEM) vs. 6.30 +/- 0.42 ng/ml; P less than 0.03] and less androstenedione (24.4 +/- 3.6 vs. 37.8 +/- 5.0 ng/ml; P less than 0.03) than did huFSH-stimulated follicles. Embryonic cleavage rates were similar for all fertilized oocytes from both hMG- and huFSH-stimulated cycles, although pregnancy rates were significantly higher in huFSH cycles (40% vs. 9.5%; P less than 0.05). In addition, aromatase activity, progesterone production, and [125I]hCG-binding activity were compared in granulosa-luteal cells isolated from some of these women. Cells from 21 follicles from 9 women stimulated with hMG and 24 follicles from 9 women stimulated with huFSH were studied. There were no significant differences in aromatase activity, progesterone production, or [125I]hCG binding. Thus, the presence or absence of exogenous LH during ovulation induction with FSH has little direct effect on granulosaluteal cell function. However, the presence of LH during ovulation induction with FSH does appear to alter thecal androgen metabolism, resulting in higher testosterone and lower androstenedione levels in follicular fluid. Such a shift in androgen milieu may impair oocyte development and successful implantation.