HORMONAL PROFILES IN SUCCESSFUL AND UNSUCCESSFUL IMPLANTATION IN IVF-ET AFTER COMBINED GNRH AGONIST GONADOTROPIN TREATMENT FOR SUPEROVULATION AND HCG LUTEAL SUPPORT

The luteal phase of in vitro fertilization-embryo transfer (IVF-ET) cycles has come under great scrutiny as the cause for the discrepancy between fertilization rates (> 70%) and pregnancy rates (around 20%) in most IVF programs is sought. The effects of the various stimulation protocols on the subsequent hormonal events of the luteal phase are both important and controversial but information regarding cycles where ovarian stimulation has been carried out with gonadotropins under pituitary suppression is scanty. The effect of high levels of estrogen in the late follicular phase and around the time of implantation is a matter of concern. As combined gonadotropin-releasing hormone analog (GnRH-a) gonadotropin treatment for superovulation is associated with supraphysiological ovarian steroid levels, both in the follicular and luteal phase of IVF cycles, we compared preovulatory (estradiol), midluteal (estradiol, progesterone and prolactin) and late luteal (estradiol and progesterone) hormone levels in on-going pregnancies, abortions and non-conception cycles in 222 patients accomplishing their first IVF-ET attempt who received such ovarian stimulation therapy. For both successful and unsuccessful implantation cycles, estradiol on the day of human chorionic gonadotropin (hCG) administration correlated positively with progesterone and estradiol levels and negatively with the progesterone/estradiol ratio in the midluteal phase. Mean peak follicular estradiol, midluteal estradiol and progesterone levels, mean mid- and late luteal progesterone/estradiol ratio, and mean midluteal prolactin concentration, were similar in the three groups studied. The mean late luteal estradiol and progesterone in the on-going pregnancy group were significantly higher than in non-conception cycles (p < 0.005 and p < 0.001, respectively) as a reflection of trophoblastic hCG production. Forty-eight patients (21.6%) had hyperprolactinemia. Estradiol and progesterone at around the time of implantation were significantly higher (p < 0.001) in the hyperprolactinemic group but the progesterone/estradiol ratio, conception rate and pregnancy outcome were similar in patients with normal prolactin levels and those having hyperprolactinemia. It is concluded that combined GnRH-a/gonadotropin treatment for superovulation in IVF adversely affects neither the hormonal milieu of the luteal phase nor implantation, provided that hCG luteal phase support is given.