We aimed to determine the efficacy of ovarian hyperstimulation protocols employing a GnRH antagonist to prevent a premature LH rise allowing final oocyte maturation and ovulation to be induced by a single bolus of either a GnRH agonist or hCG.A total of 122 normogonadotrophic patients following a flexible antagonist protocol was stimulated with recombinant human FSH and prospectively randomized (sealed envelopes) to ovulation induction with a single bolus of either 0.5 mg buserelin s.c. (n = 55) or 10,000 IU of hCG (n = 67). A maximum of two embryos was transferred. Luteal support consisted of micronized progesterone vaginally, 90 mg a day, and estradiol, 4 mg a day per os.Ovulation was induced with GnRH agonist in 55 patients and hCG in 67 patients. Significantly more metaphase II (MII) oocytes were retrieved in the GnRH agonist group (P < 0.02). Significantly higher levels of LH and FSH (P < 0.001) and significantly lower levels of progesterone and estradiol (P < 0.001) were seen in the GnRH agonist group during the luteal phase. The implantation rate, 33/97 versus 3/89 (P < 0.001), clinical pregnancy rate, 36 versus 6% (P = 0.002), and rate of early pregnancy loss, 4% versus 79% (P = 0.005), were significantly in favour of hCG.Ovulation induction with a GnRH agonist resulted in significantly more MII oocytes. However, a significantly lower implantation rate and clinical pregnancy rate in addition to a significantly higher rate of early pregnancy loss was seen in the GnRH agonist group, most probably due to a luteal phase deficiency.

GnRH agonist (buserelin) or hCG for ovulation induction in GnRH antagonist IVF/ICSI cycles: a prospective randomized study