Initiation of ovarian stimulation independent of the menstrual cycle: a case–control study

In the GnRH-antagonist protocol, ovarian stimulation with gonadotropins typically commences on cycle day 2 or 3. Initiation of ovarian stimulation with a spontaneously occurring menstruation, however, poses significant organizational challenges for treatment centres and patients alike. It has previously been demonstrated in the context of fertility preservation that initiation of stimulation in the luteal phase is feasible in terms of retrieval of mature oocytes for cryopreservation. Herein, we report the extension of this concept to a routine IVF setting with the aim of establishing an ovarian stimulation protocol, which can be utilized independent of menstruation. Because of asynchrony of endometrium and embryo in such a setting, all fertilized oocytes have to be cryopreserved for a later transfer.

This was a prospective, case–control study (trial registration: NCT00795041) on the feasibility of starting ovarian stimulation in a GnRH-antagonist protocol in the luteal phase. Inclusion criteria were: IVF or ICSI; 18–36 years; ≤3 previous IVF/ICSI attempts; BMI 20–30 kg/m²; regular cycle (28–35 days); luteal phase progesterone >7 ng/ml at initiation of stimulation. Exclusion criteria were: PCOS, endometriosis ≥AFS III°, unilateral ovary, expected poor response. Stimulation was performed with highly purified uFSH (Bravelle^®) 300 IU/day and 0.25 mg/day GnRH-antagonist starting on cycle day 19–21 of a spontaneous menstrual cycle and commencing until hCG administration when three follicles ≥17 mm were present. All 2PN stage oocytes were vitrified for later transfers in programmed cycles. Feasibility was defined as the achievement of ongoing pregnancies progressing beyond the 12th gestational week in at least 2/10 study subjects (primary outcome). Secondary outcomes were gonadotropin consumption per oocyte obtained, stimulation duration, and fertilization rates. Study subjects were matched in a 1:3 ratio with concomitantly treated control cases of similar age, BMI, and duration of infertility who were treated in a conventional GnRH-antagonist protocol with 150–225 rFSH or HP-HMG/day.

The study group consisted of ten subjects, mean age 31.4 years, BMI 25.4 kg/m², of which one had fertilization failure. Mean stimulation duration was 11.7 (SD 1.6) vs. 9.1 (SD 1.3) days, mean cumulative FSH dose was 3,495.0 (SD 447.5) vs. 2,040.5 (SD 576.2) IU, and mean number of oocytes was 8.8 (SD 5.0) vs. 10.0 (SD 5.4) in study vs. control group, respectively. Per follicle ≥10 mm, the cumulative FSH dose was 274.5 (SD 130.8) IU vs. 245.2 (SD 232.3) IU in study and control groups, respectively. Cumulative ongoing pregnancy rates were 1/10 (10 %) and 6/30 (20.0 %) in study and control group, respectively (difference: 10 %, 95 % confidence interval of the difference: −29.2–22.2 %, p = 0.47). Fertilization rate was similar between groups, with 63.5 % (SD 32.9) in the study and 61.3 % (SD 26.7) in the control group, respectively. Serum estradiol levels were significantly lower on the day of triggering final oocyte maturation with 1,005.3 (SD 336.2) vs. 1,977.4 pg/ml (SD 1,106.5) in study and control group, respectively. Similarly, peak estradiol biosynthesis per growing follicle ≥10 mm was lower in the study group (134 pg/ml, SD 158.4 vs. 186.7 pg/ml, SD 84.7).

Per retrieved oocyte, a nearly threefold higher dose of FSH had to be administered when ovarian stimulation had been initiated in the luteal phase. Furthermore, the present study casts doubt on the efficacy of initiating ovarian stimulation in the luteal phase in terms of pregnancy achievement. Thus, this concept is currently not feasible for routine use, and it should also be explored further before using it at larger scale in the context of emergency stimulation for fertility preservation.