[Increased risk of early pregnancy loss and lower live birth rate with GNRH antagonist vs. long GNRH agonist protocol in PCOS women undergoing controlled ovarian hyperstimulation].

to compare standard long GnRH agonist protocol (Triptorelin) and GnRH antagonist regimens (Cetrorelix) in polycystic ovary syndrome (PCOS) patients undergoing controlled ovarian stimulation (COS) for ICSI cycles.Retrospective case-control study. 106 PCOS patients undergoing COS for ICSI with long GnRH agonist protocol (Triptorelin) were matched with age and BMI to 106 PCOS patients undergoing COS for ICSI with GnRH antagonist (Cetrorelix) during the same period. Ovarian stimulation with recombinant follicle stimulating hormone (rFSH) was used in the two groups. Oral contraceptive pill pretreatment was used in all patients undergoing ovarian stimulation using GnRH antagonists. ICSI was performed for male infertility in all cases. The main outcome measures evaluated were: cancellation of the cycles, number of aspirated follicles, oocyte maturity, fertilization rate, Embryo quality, pregnancy and implantation rates, clinical abortion rate, multiple pregnancy rate and the live birth rate rate. Kchi2 test and t Student test were used for differences between normo-ovulatory and PCOS patients and the limit of significance was set at p < 0.05.There was no significant difference in term of cancellation rate (2.8% vs 1.8%; NS). Duration of gonadotrophin stimulation (9.7 +/- 0.7 vs. 11.2 +/- 1.9 days; p < 0.001) and gonadotrophin consumption (2209.0 +/- 548.3 vs. 1411.1 +/- 217.9 UI: p < 0.001) were significantly decreased with GnRH antagonist. The mean oestradiol level on the triggering day was significantly higher in the agonist group (3347.85 +/- 99 vs. 2354.45 +/- 839; p < 0.001 ).A fall in LH level of > or = 50% from stimulation day 8 (S8) to S1 was observed in GnRH antagonist group. Risk of ovarian hyperstimulation syndrome (OHSS) was significantly decreased with GnRH antagonist (1.8% vs 10.7%; p = 0.01). The mean number of retrival oocytes (15.9 +/- 5.9 vs. 17.3 +/- 8.3; ns) and the mean number of mature oocytes (11.43 +/- 4.2 vs. 11.91 6.4; ns) were similar in the two groups, fertilization rate (73.3% vs 75.8%; NS), mean number of grade 1 and 2 embryos (6.3 +/- 2.7 vs. 6.9 +/- 3.9; NS), mean number of transferred embryos (1.9 +/- 0.7 vs. 1.8 +/- 0.7; NS), implantation rate (13.3% vs. 18.45%; ns) and clinical pregnancy rate per transfer (28.6% vs 31.1% ; NS) did not differ statistically in the two groups. Twin and triplet pregnancies rates were also similar in the two groups (7.1% vs. 9.3%; NS) and (3.5% vs. 3.1%; NS) respectively. Live birth rate (12.2% vs. 20.7%; p < 0.001) was significantly lower in GnRH antagonist group and miscarrage rate was significantly higher in this same group (42.8% vs. 18.7%; p < 0.001).GnRH antagonist protocol is a short and simple protocol with a significant reduction in incidence of OHSS and amount of gonadotrophins. However, GnRH antagonist protocol provides a lower live birth rate and an increased risk of early pregnancy loss compared to the GnRH agonist long protocol. Further studies are necessary for more solid conclusions.