Elective and Onco-fertility Preservation: Factors Related to IVF Outcomes

A. Cobo; J. García-Velasco; J. Domingo; A. Pellicer; J. Remohí


Hum Reprod. 2018;33(12):2222-2231. 

In This Article

Materials and Methods

Study Design and Population

A multicenter retrospective study with 6362 women who had their oocytes vitrified for the purpose of FP during the period between January 2007 and May 2018 at Instituto Valenciano de Infertilidad clinics in Spain. Of them, 5289 women (7044 vitrification cycles) opted for Elective Fertility Preservation (EFP group) due to age-related fertility decline, and 1073 women (1172 vitrification cycles) chose this option for oncological reasons (Onco-FP group). Institutional Review Board approval was obtained (1505-VLC-033-AC). Data were collected from computerized clinical charts and remained anonymous according to Spanish law on assisted reproduction (Law 14/2007 on Biomedical Research). Data about baseline characteristics were compared between both populations (EFP vs. Onco-FP), and data on survival and clinical outcomes were analyzed in the group of women who came to use their oocytes (N = 641 and N = 80 patients, respectively).

Controlled-ovarian Stimulation Protocols

Antagonist and agonist protocols were used in both the EFP and Onco-FP groups for non hormonal-dependent diseases (Table I). Letrozole was added to the antagonist protocol in the Onco-FP group exclusively for breast cancer patients with hormone-responsive tumors. The controlled-ovarian stimulation (COS) protocols have been described elsewhere (Oktay et al., 2006; Garcia-Velasco et al., 2013; Cobo et al., 2016; Domingo and Garcia-Velasco, 2016) and were applied as indicated below.

The starting dose and adjustment of gonadotropins were based on our standard of care in conventional follicular phase and random-start COS protocols. The initial dose of gonadotropins was selected according to the patient's age, her BMI, and the assessment of ovarian reserve as estimated by antral follicle count (AFC). In detail, protocols were applied as follows.

Antagonist protocol. COS was initiated on Day 2 or 3 of a spontaneous cycle. An initial dose of 225–300 IU recombinant FSH (rFSH) (Gonal-F, Merck-Serono, Spain; Puregon, MSD, Spain) and 75–150 IU highly purified hMG (Menopur, Ferring Pharmaceuticals, Spain) was administered until triggering. When a leading follicle reached ≥14 mm, a GnRH antagonist (Cetrotide, Merck-Serono; Orgalutran, MSD) was administered at 0.25 mg/day. Usually, final oocyte maturation was triggered with a single dose of a GnRH agonist (0.1 mg of Decapeptyl, Ipsen Pharma) when the mean diameter of two follicles was ≥18 mm. In a few cases, especially when during the FP program was still in its early years, triggering was performed with 250 μg of recombinant hCG (rhCG) (Ovitrelle, Merck-Serono, Madrid, Spain). Oocyte retrieval was scheduled 36 h later. In some cases, Clomiphene (Omifin, Sanofi Aventis, Madrid, Spain) was added to the antagonist protocol at a daily dose of 50–100 mg for 5 days, initiated on Days 2–3 of the cycle, and combined with 150–300 IU of hMG from Day 3 and every other day until triggering. The antagonist (0.25 mg/day) was added as explained above and ovulation was triggered as previously described.

Agonist protocol. The GnRH agonist (0.1 mg/day) was initiated on Day 21 of the previous cycle (long protocol). On Day 3 of the menstrual cycle, rFSH and hMG were administered as explained above. Once the criteria for triggering were met, ovulation was induced by administering rhCG (250 μg). In some cases, COS was initiated on Day 1 or 2 of a spontaneous cycle by administering the GnRH agonist (0.2 mg/24 h). From Day 3 onward, the dose was adjusted to 0.1 mg/24 h and gonadotropins started being administered.

Antagonist protocol with letrozole. Letrozole (5 mg/day; Femara; Novartis, Switzerland) was initiated on Day 2–3 of a cycle and was maintained until oocyte retrieval. After 2 days of letrozole administration, rFSH was added (150–225 IU). When the leading follicle reached ≥14 mm, 0.25 mg/day of the GnRH antagonist was added. Final oocyte maturation was triggered with a single bolus of a GnRH when the leading follicle reached >20 mm 1 day later than when Letrozole was not used. Oocyte retrieval was scheduled 36 h after triggering and Letrozol was reinitiated until menstruation appeared.

If the Onco-FP group, patients were not close to the beginning of a cycle, their menstrual cycle phase was evaluated by the onset of the last menstrual period by either ultrasound or progesterone concentrations, and stimulation was started randomly in either the late follicular phase or the luteal phase (Domingo and Garcia-Velasco, 2016). The late follicular phase was defined as being after menstrual cycle Day 7 with the emergence of a dominant follicle (>13 mm) and a progesterone level of <2 ng/ml. The luteal phase was determined by a progesterone level of >5 ng/ml or by the presence of a corpus luteum. Ovarian stimulation was started immediately after the initial consultation when the patient was in the luteal phase or after triggering ovulation if she was in the late follicular phase. Serum E2 levels were measured during every visit. Similarly to the conventional form, the GnRH antagonist was added once the leading follicle reached 14 mm, with triggering for the final oocyte maturation with 18–20 mm follicles.

Oocyte vitrification/warming. Oocytes were denuded 2 h after oocyte retrieval. Only MII oocytes were selected for vitrification. All the vitrification and warming solutions were obtained from Kitazato® (Kitazato, Shizuoka, Japan). The vitrification device was the Cryotop® obtained from the same suppliers. The protocols and tools for vitrification and warming were applied following the manufacturer's instructions, as explained elsewhere (Cobo et al., 2015). Briefly, oocytes were equilibrated gradually in a mixture of 15% (v/v) ethylene glycol (EG) + dimethylsulfoxide (DMSO) for 12 min, before being washed in vitrification solution containing 30% (v/v) of the same cryoprotectants for 50–60 s. Synthetic serum substitute (SSS) or hydroxy-propyl cellulose (HPC) was used as a substitute for the protein source, and sucrose or trehalose was used as the osmotic agent (Kuwayama, 2007; Coello et al., 2016). Oocytes were loaded on the Cryotop® surface contained in a minimum volume drop. The loading of samples took no more than 10 s. Vitrification was induced immediately by plunging the Cryotop® into liquid nitrogen. Four oocytes (maximum) were loaded per device. Oocytes were stored in vapor tanks (V1500-AB Isothermal Freezer; CBS®; USA) for a variable storage time (Cobo et al., 2010). Warming consisted in an initial phase of dilution of Cryoprotectant. Rehydration of oocytes took place in two subsequent steps by exposing oocytes to two hyperosmolar solutions with decreasing concentrations of trehalose for 1 min and 3 min, respectively. The warming procedure was completed with two washes lasting 5 min and 1 min in buffer solution at room temperature. Oocytes were kept in culture for 2 h until the ICSI procedure (Cobo et al., 2015). Whenever necessary, the vitrification and warming procedures of the surplus embryos were performed as explained in detail elsewhere (Cobo et al., 2012). The procedures were similar to those described above for oocytes, except that equilibration was performed in a single step that lasted ~10 min. The vitrification, loading storage and warming/dilution steps were carried out as described above for oocytes.

Endometrial preparation for embryo transfer. The endometrial preparation protocol has been described elsewhere (Soares et al., 2005; Cerrillo et al., 2017).

Briefly, after menses, the subjects received oral estradiol valerate (EV) (Progynova®, 6 mg/day Schering, Madrid, Spain). Approximately 10 days after initiating EV, serum E2 levels and endometrial thickness were measured. Administration of micronised progesterone (P) (800 mg/day, vaginally; Progeffik, Effik Laboratories, Madrid, Spain) was initiated 3 days prior to embryo transfer (ET) for Day 3 ETs. For the ET of blastocysts, P was initiated 5 days prior to transfer. If pregnancy was achieved, administration of EV and P was maintained until gestation week 12. ET was usually scheduled 4–8 h after warming for either Day 3 or blastocyst transfers.

Definition of Outcomes and Statistics

The main outcome measurements were oocyte survival and cumulative live birth according to the indication for FP (EFP and Onco-FP) and patient's age at vitrification. Survival was determined by oocytes' morphology immediately upon warming, and was reconfirmed at 2 h immediately before ICSI. The live births included the live births obtained from fresh transfers, plus those from frozen embryo transfers. Secondary outcomes were the number of retrieved and vitrified oocytes, patients' return rate, age at return, implantation, and clinical and ongoing pregnancy rates.

Chi-square tests were used to compare the categorical data, and T tests or analyzes of variance with post hoc comparisons (Bonferroni, Tuckey, Sheffé) were employed to compare the means within groups whenever appropriate. A difference was considered statistically significant when the P value was <0.05. The SPSS software (version 19.0) was used for the statistical analyzes. A binary logistic regression (BLR) analysis was performed to determine if the indication for FP, age at vitrification and the COS protocols were associated with oocyte survival and live births. The binary response parameter for survival was the occurrence of a survival rate of ≥90% categorized as '1', and a survival rate of <90% categorized as '0'. For the CLBR, the binary response parameter was the achievement of at least one live birth categorized as '1', or no live birth achieved and categorized as '0'. The odds ratio (OR) of the effect that the variables indication, age and COS protocols had on survival and the CLBR was expressed in 95% confidence interval (95%CI) and significance terms.

To determine the required number of oocytes to achieve at least one child, the cumulative probability of having at least one baby according to the total number of oocytes utilized in consecutive procedures was estimated by the Kaplan–Meier method, as previously described (Cobo et al., 2016). The analysis comprised all the procedures performed per patient, including those that resulted in embryo transfer or not, and those from both fresh or frozen embryo transfers, if this were the case, until at least one live birth was achieved. To compute all the oocytes, all the women who used their oocytes were counted in this analysis, which also comprised those who utilized oocytes and did not get pregnant. We considered 'utilized oocyte' to be all the oocytes that survived the vitrification process, or did not; those that fertilized and those that did not; those that evolved into usable embryos (transferred or frozen) and those that, on the contrary, stopped while developing. Those embryos still cryopreserved at the time of the analysis were excluded. The analysis was performed separately for each indication, and data were further stratified and analyzed according to age with the log-rank, Breslow, and Tarone–Ware tests to compare the Kaplan–Meier survival curves according to each case.