Mild stimulation with clomiphene citrate in combination with recombinant follicle‐stimulating hormone and gonadotropin‐releasing hormone antagonist and its influence on serum estradiol level and pregnancy rate

Abstract

Aim:  The mild ovarian stimulation protocol for in vitro fertilization (IVF) is carried out to minimize adverse side‐effects as well as cost. While performing mild ovarian stimulation with a gonadotropin‐releasing hormone (GnRH) antagonist, the pregnancy rate was examined in cases that exhibited a serum estradiol (E2) drop down.

Methods:  In this study, 174 patients who requested mild ovarian stimulation for IVF began clomiphene citrate on day 3 and recombinant follicle‐stimulating hormone (FSH) on day 5 of their menstrual cycles. A GnRH antagonist was administered when the dominant follicle reached a diameter of 14 mm. Serum luteinizing hormone and estradiol were measured at the time of GnRH antagonist administration and at the time of human chorionic gonadotropin (hCG) injection. Pregnancy rates and implantation rates were compared between 24 cycles in which the E2 level fell at the time of hCG injection and 150 cycles in which it did not fall.

Results:  The pregnancy rate in the cases in which the E2 level fell (25% decrease) at the time of hCG injection was significantly lower than it was in the cases in which it did not fall (16.7 vs 41.0%). The implantation rate for the cases in which the E2 level fell was also lower than that of the control group (7.0 vs 31.0%). There was no significant difference in the number of good‐quality embryos between the two groups.

Conclusion:  When performing the mild ovarian stimulation protocol, serum E2 should be followed. It is prudent to avoid embryo transfer in the same cycle in cases that exhibit E2 drop down. (Reprod Med Biol 2008; 7: 85–89)

INTRODUCTION

IN VITRO FERTILIZATION (IVF) protocols have changed from focusing on clinical results to focusing on decreasing the patient's economical burden and physical discomfort while maintaining an acceptable clinical outcome. In recent years, many infertility centers have accepted the protocol of mild ovarian stimulation, also known as friendly IVF.

Many ovarian stimulation protocols, such as clomiphene citrate (CC) only, as well as CC in combination with human menopausal gonadotropin (HMG) and/or follicle‐stimulating hormone (FSH) have been reported. ¹ , ² , ³ One of the problems of friendly IVF is the possibility of ovulation before oocyte retrieval. Although Branigan and Estes reported the use of CC for the purpose of luteinizing hormone (LH) suppression, ¹ gonadotropin‐releasing hormone (GnRH) antagonists have been developed that have also been applied to IVF for the control of the LH surge.

Several protocols for the use of a GnRH antagonist in IVF have been reported. ⁴ , ⁵ , ⁶ Although the number of retrieved oocytes was lower when compared to the long protocol, which uses a GnRH agonist, there was no significant difference in the clinical pregnancy rate. ³ , ⁷ Furthermore, Williams et al. reported a cancellation rate of 0% and reinforced the validity of a GnRH antagonist for IVF. ³

The literature contains very few reports addressing the use of the mild stimulation protocol and its impact on serum hormone levels using a GnRH antagonist. We have found cases in which the estradiol (E2) level falls when using the mild stimulation protocol with a GnRH antagonist. In this retrospective study, the embryo transfer (ET) outcome with IVF for patients undergoing mild stimulation with CC in combination with recombinant FSH (rFSH) plus GnRH antagonist and its effect on serum hormone levels (especially E2) was reviewed.

MATERIALS AND METHODS

FROM JANUARY 2006 to December 2006, 174 patients aged 27–42 years underwent 174 IVF cycles. Indications of IVF were tubal infertility, male factor infertility, unexplained infertility, and early‐stage endometriosis. Prior IVF cycles numbered less than two. The couples requested the mild stimulation protocol and fully informed consent was obtained. Baseline serum FSH levels (<15 mIU/mL) were taken and CC (100 mg/day) was given during the cycle on days three through seven. Controlled ovarian hyperstimulation was initiated using 150 IU of a rFSH (Follistim, Organon, Tokyo, Japan) or 150 IU of human menopausal gonadotoropin (HMG; Humegon, Japan Organon, Osaka, Japan). Injections were begun on cycle day five and then repeated every other day. Ultrasound was carried out on cycle day nine and the serum levels of LH and estradiol (E2) were determined. Ultrasonographic follicular size, LH and E2 were measured at every visit until the day of oocyte retrieval. An injection of 0.25 mg of a GnRH antagonist (Cetrotide, Nippon Kayaku, Tokyo, Japan) was initiated daily after the lead follicle attained a diameter of >14 mm.

An injection of 10 000 U of human chorionic gonadotropin (hCG; Profasi, Serono, Tokyo, Japan) was administered 35–36 h before IVF; the oocytes were retrieved via ultrasound‐guided transvaginal needle aspiration. Embryos were cultured in Universal IVF Medium (MediCult a/s, Jyllinge, Denmark) using BlastAssist System 1, 2 (MediCult a/s) in a 5% CO₂, 5% O₂ and 90% N₂ environment. Embryos were transferred 3–5 days after aspiration. Veeck's classification and Gardner's classification ⁸ were used for embryo scoring; good‐quality embryos were: 7 cell G1 or G2 on day 3; and more than 3 AA on day 5. Luteal support was given for 2 weeks (a 222 mg progesterone vaginal suppository was inserted every 12 h and 125 mg of hydroxyprogesterone caproate was injected intramuscularly once per week).

A clinical pregnancy was defined as a hCG level >25 IU/L, and implantation was confirmed by the presence of a fetal heart beat on ultrasound. χ²‐tests and Student's t‐tests were used for statistical analysis and P < 0.05 was defined as statistically significant.

RESULTS

A COMPARISON EXAMINATION of embryo quality, pregnancy rate and implantation rate was conducted on 24 cycles in which the E2 level fell at the time of hCG injection (Group A; day 3 transfer was conducted on six cycles and day 5 transfer was conducted on 18 cycles). As a control, the same comparison examination was conducted on 150 cycles in which the E2 level did not fall (day 3 transfer was conducted on 75 cycles and day 5 transfer was conducted on 75 cycles). The mean age was 36.7 ± 3.7 years for the women who's E2 level had fallen at the time of hCG injection (Group A) and 35.3 ± 3.5 years for the women whose E2 level did not fall (control group). There was no statistically significant difference in the indication for IVF between the two groups. In Group A, 5.5 ± 2.1 oocytes were retrieved; 6.5 ± 3.6 oocytes were retrieved in the control group (not statistically significant). In both groups, the ET cancellation rate was 0%. The mean number of embryos transferred was 1.32 ± 0.62 in Group A and 1.37 ± 0.52 in the control group (not statistically significant). The serum LH level before GnRH antagonist injection was 5.5 ± 2.1 IU/L in Group A and 6.5 ± 3.6 IU/L in the control group. The serum LH level at the time of hCG injection was 3.5 ± 2.2 IU/L in Group A and 5.5 ± 4.9 IU/L in the control group. Serum E2 increased from 1142 ± 569 pg/mL at the time of GnRH antagonist injection to 1782 ± 777 pg/mL at the time of hCG injection in the control group. Conversely, in Group A, serum E2 decreased from 1409 ± 562 pg/mL at the time of GnRH antagonist injection to 1197 ± 479 pg/mL at the time of hCG injection (P < 0.05). The percentage of good‐quality embryos using Veeck's classification and Gardner's classification was 72.7% in Group A and 76.0% in the control group (not statistically significant). The preservation rate of surplus embryos was 37.5% in Group A and 52% in the control group (not statistically significant). The clinical pregnancy rate was 16.7% in Group A and 41.0% in the control group. The implantation rate was 7.0% in Group A and 31.0% in the control group. Both the clinical pregnancy rate and the implantation rate were statistically different between Group A and the control group (P = 0.02; Table 1)

Table 1.

Clinical findings of the group with estradiol (E2) drop down and the control group

	E2 drop down	Control
No. cycles	24	150
Average age (years)	36.7 ± 3.7	35.3 ± 3.5
Past IVF	1.45 ± 0.5	1.21 ± 0.4
Average no. oocyte retrievals	5.5 ± 2.1	6.5 ± 3.6
Embryo transfer	1.32 ± 0.6	1.37 ± 0.52
GnRH antagonist (days)	2 ± 0.6	2.5 ± 0.7
No. MII	3.1 ± 2	4.6 ± 2.7
Cancellation rate (%)	0	0
LH level before GnRH antagonist (IU/L)	5.6 ± 2.9	5.96 ± 3.23
LH level on hCG injection (IU/L)	3.5 ± 2.2	5.5 ± 4.9
E2 level before GnRH antagonist (pg/mL)	1409 ± 562	1142 ± 569
E2 level on hCG injection (pg/mL)	1197 ± 479a	1782 ± 777a
Embryo quality (%)†	72.7	76
Clinical pregnancy (% per ET)	16.7b	41b
Implantation (% per ET)	7c	31c

^† Good quality: 7 cell, G1, G2 on day 3; over 3 AA on day 5.

Values followed by different letters in the same row are statistically different (P < 0.05).

ET, embryo transfer; GnRH, gonadotropin‐releasing hormone; hCG, human chorionic gonadotropin; LH, luteinizing hormone.

DISCUSSION

SEVERAL OVARIAN STIMULATION protocols for IVF are currently available because of development and improvement in ovulation inducers; thus, an appropriate stimulating mechanism can be selected for each patient. In particular, for couples with male and/or tubal factors, there is no need for hyper‐stimulation. ⁵ , ⁶ The mild ovarian stimulation protocol is currently used in many institutions. It has the advantages of low cost and low risk for IVF procedures. ¹ , ³

Pelinck et al. obtained a good pregnancy rate using a GnRH antagonist with rFSH (without CC) and noted that their pregnancy rates after IVF with minimal, late follicular phase stimulation were encouraging. ⁹ Weghofer et al. found it to be an effective method of mild ovarian stimulation for IVF in women over 40 years of age. ¹⁰ For women over 40 years undergoing IVF via our method, there was no significant difference in the number of oocytes retrieved, pregnancy rate or implantation rate between the long protocol and the mild ovarian stimulation protocol (data not shown).

According to recent reports, there is no difference in clinical results with the mild ovarian stimulation protocol when compared with the conventional long protocol. ³ , ¹⁰ Furthermore, studies have shown that there is no cancellation as a result of ovulation. ³ This finding occurred because ovulation accompanying the LH surge is controlled by the GnRH antagonist; thus, oocyte retrieval can be assured.

In our mild stimulation IVF protocol, cases in which the serum E2 was lower on the day of hCG injection than it was at the initial GnRH antagonist injection were examined; in these cases, the pregnancy rate and implantation rate were markedly decreased. Hwang et al. reported in depth variations in the serum hormone level using the mild stimulation protocol with a GnRH antagonist; however, no data was presented with regard to E2 drop down. ⁶

Once follicular growth begins either FSH or LH is capable of sustaining follicular E2 production. ¹¹ Therefore, Pelinck et al. apparently prevented E2 drop down by simultaneously injecting rFSH. In our cases, rFSH was also injected at the time that a GnRH antagonist was added; however, cases existed in which the serum E2 was falling at the time of hCG injection. No morphological difference was found between the two groups when the quality of the transferred embryos was compared.

Eldar‐Geva et al. reported that a lower successful IVF embryo transfer rate using a GnRH‐antagonist/GnRH‐agonist protocol does not appear to be related to decreased oocyte quality. ¹² No difference was found in terms of implantation rates or pregnancy rates when a GnRH agonist or GnRH antagonist was used in the previous oocyte retrieval. Lee et al. suggested that GnRH antagonists do not have a detrimental effect on oocyte quality or embryo development. ¹³ Our data supports their results. It is possible that the low pregnancy rate in the E2 drop down cases results from the GnRH antagonist adversely affecting the endometrium (implantation environment).

No significant change was observed in endometrial development during the early and mid‐luteal phases in women undergoing controlled ovarian stimulation for oocyte donation following daily treatment with a standard or high dose GnRH antagonist. ¹⁴ Saadat et al. found no significant difference in luteal phase endometrial histology between cycles using a GnRH antagonist or a GnRH agonist. ¹⁵ Conversely, ovarian stimulation with a GnRH agonist rather than a GnRH antagonist may partially restore the endometrial physiological secretion and improve uterine receptivity. ¹⁶ Currently, the influence of a GnRH antagonist on the endometrium is under debate; however, CC may exert some influence on the implantation environment. ¹⁷

Although there was no correlation between the histology of the endometrial biopsy and the duration of the luteal phase, CC treatment significantly increases the prevalence of out‐of‐phase endometrium. Homburg et al. could not explain the significant difference between ovulation and pregnancy rates with CC. ¹⁸ The administration of CC produces aberrant endometrial beta3 integrin expression in conjunction with a failure of the downregulation of progesterone receptor during the implantation window. ¹⁹

It can be postulated that CC alone is not adequate to explain the pregnancy rate in the E2 drop down group; obviously, the rapid fall of the E2 level before oocyte aspiration impacts on the endometrial environment. E2 and progesterone are largely responsible for endometrial changes as implantation approaches. ²⁰ It is possible that implantation failure results from changes in various factors necessary for implantation, which are induced by E2. ²¹ , ²² However, when mild stimulation is conducted, the E2 level rises above physiological levels; thus, serum E2 drop down alone cannot explain the decreased implantation rate. Gonadotropin‐releasing hormone receptors are localized on the granulosa cells ²³ and GnRH may participate in the production of E2 directly or indirectly; furthermore, a GnRH antagonist may produce an E2 fall in the process.

There was no difference in patient background between the E2 drop down cases and the control group. It is currently unknown why E2 drop down occurred in the former group. Moreover, E2 drop down cannot be predicted before stimulation. If there is no difference in embryo quality, increasing FSH, adding HMG at the day of GnRH antagonist injection or frozen embryo transfer may improve implantation. Further research, including prospective studies, is required.