IVF outcome in poor responders: Comparison of high-dose gonadotrophin therapy with adjuvant therapy using a GnRH antagonist or a GnRH agonist

Abstract

Recent IVF studies confirmed the efficacy of GnRH antagonists as adjuvants to gonadotropin therapy in good responders. The facilitation of the stimulation of poor responders, however, remains controversial. The objective of this study was to compare IVF results in patients who met the criteria for poor responders and who were subjected to gonadotropin stimulation using adjuvant therapy with either a GnRH antagonist or a GnRH agonist (microflare protocol). This was a retrospective, non-randomized comparison of consecutive patients. Poor responders were defined as follows: age ≥ 40 years, and/or basal cycle day 3 serum FSH ≥ 12 IU/L, and/or basal cycle day 3 serum E2 level of ≥ 90 pg/mL, and/or previous peak serum E2 level of ≤ 500 pg/mL, and/or previous retrieval of ≤ 4 mature eggs, and/or history of previous cycle cancellation. The main outcome measures assessed included parameters of the ovarian response, fertilization, implantation and pregnancy outcomes. In the GnRH antagonist protocol, gonadotropins were started on day 3 and ganirelix (250 mcg/d) was added when the follicular size reached 14 mm. The GnRH agonist microflare protocol consisted of 50 mcg of leuprolide acetate twice-daily beginning on day 2 followed by gonadotropins starting on day 5. In both protocols patients were pretreated for 3 weeks with an oral contraceptive pill and received high-dose gonadotropin stimulation with either recombinant FSH or a combination of recombinant FSH and hMG administered in a step-down fashion. HCG was administered when the mean diameter of at least two follicles reached 18 mm. Uterine transfer of embryos was performed on day 3 under ultrasound guidance. Luteal support was provided with micronized progesterone vaginally. Significance was defined as P< .05. Data are expressed as mean ± SD. Patients’ basal cycle serum day 3 hormone levels, proportion of patients with age ≥40 and FSH ≥12, etiology diagnoses, total units of gonadotropins used per cycle, and proportion of IVF and ICSI cycles were similar among groups. Although the age (38.5 ± 3.8 versus 36.5 ± 4.2), day of hCG administration (12.5 ± 1.4 versus 11.8 ± 1.5), and number of embryos transferred (2.77 + 1.1 versus 2.3 + .95) was significantly higher in the microflare group, the number of mature eggs retrieved (7.1 ± 5.2 versus 7.1 ± 5.9), the fertilization rate (60.4% versus 60.5%), implantation rate (10% versus 9.9%), clinical pregnancy rate (19.6% versus 21.7%), and ongoing pregnancy rate (13.7% versus 15.0%) did not differ between the use of GnRH antagonist or agonist. Although this study showed a small difference in the ovarian response between the two protocols illustrated by a difference in day of hCG administration and the number of embryos transferred, there was no statistically significant difference noted in the fertilization, implantation or pregnancy outcomes of the two groups. We conclude that the use of a high-dose gonadotropin protocol in combination with a GnRH antagonist or a GnRH agonist (microflare) does not result in significantly different ovarian response or pregnancy outcome in strictly defined poor responders for IVF. The optimal therapeutic approach for this group of women remains a considerable challenge.