Luteal Phase Progesterone Support Does Not Improve Ongoing Pregnancies in True Natural Cycle Cryopreserved Blastocyst Stage Embryo Transfers

Background: More than 67% of all embryos transferred in the United States involve frozen-thawed embryos. Progesterone supplementation is necessary in medicated cycles in order to luteinize the endometrium and prepare it for implantation, but little data is available if this is necessary in true natural cycles. We evaluated the use of progesterone luteal phase supplementation for cryopreserved/warmed blastocyst transfers in natural cycles not using an ovulatory trigger. Methods: Retrospective cohort study in a single academic medical center. We studied the use of progesterone supplementation beginning in the luteal phase and continued until 10 weeks gestational age in patients undergoing true natural cycle cryopreserved blastocyst transfer. Our outcome measures were ongoing pregnancy rate, positive serum beta human chorionic gonadotropin (HCG) level, implantation rate, clinical pregnancy rate, miscarriage/abortion rate, ectopic pregnancy rate, and multifetal gestation rate. Categorical data were analyzed utilizing Fisher’s exact test and non-parametric data were analyzed using the Wilcoxon rank sum test. We a priori adjusted for age. Results: 229 patients were included in the analysis with 149 receiving luteal phase progesterone supplementation and 80 receiving no luteal phase support. Patient demographic and cycle characteristics, and embryo quality were similar between the two groups. No difference was seen in ongoing pregnancy rate (49.0% vs. 47.5%, p=0.8738), clinical pregnancy rate (50.3% vs. 47.5%, p= 0.7483), positive HCG rate (62.4% vs. 57.5%, p=0.5965), miscarriage/abortion rate (5.4% vs. 2.5%, p=0.2622), ectopic pregnancy rate (0% vs. 1.3%, p=0.3493), or multifetal gestations (7.4% vs. 3.8%, p=0.3166). Capsule: Progesterone supplementation as luteal phase support in true natural cycle cryopreserved blastocyst transfers does not improve ongoing pregnancies.


Introduction
As of 2017, More than two thirds of all embryo transfers in the United States involve frozen-thawed embryos with an increasing trend towards freeze-thaw cycles(1). This trend has largely been driven by the desire to decrease ovarian hyperstimulation syndrome, the increasing use of elective single embryo transfer, the improvement in cryopreservation techniques from slow-freeze to vitri cation, and the new trend towards freeze-all cycles (2).
There are currently two primary ways in which the endometrium can be prepared for embryo transfer and implantation: natural cycle and medicated prepared cycle. The natural cycle takes advantage of estrogen production by the dominant follicle in the follicular phase and a subsequent shift to progesterone production by the corpus luteum after ovulation to prepare a receptive endometrium for implantation. In this setting, the cycle is referred to as a "true" natural cycle cryopreserved embryo transfer (CET) and the timing of ovulation is usually determined by Luteinizing Hormone (LH) assay. In contrast, a "modi ed" natural cycle CET, involves use of an exogenous ovulatory trigger, which facilitates the timing of ovulation and subsequent embryo transfer. The medicated prepared cycle involves administration of exogenous estrogen until an adequate uterine lining is achieved, and then the addition of progesterone to mimic the intrinsic switch from estrogen dominance to progesterone dominance to prepare the endometrium for implantation.
There are con icting data reporting on the bene ts of luteal phase support in cryopreservation cycles, and no data evaluating utility of progesterone supplementation in true natural cycle CET compared with prepared cycles for blastocyst transfers. The aim of this study was to evaluate the potential bene t of luteal phase progesterone support in true natural cycle CETs compared with no luteal phase progesterone support.

Study population
All patients who underwent day 5 or day 6 true natural cycle cryopreserved transfer between July 30, 2012 and June 30, 2018 were screened for inclusion. Patients were excluded if they received any medication to assist in endometrial preparation, if an exogenous trigger was employed to induce ovulation, or if they received more than two embryos at time of transfer. Evaluation of outcomes following progesterone supplementation was performed and compared to those who did not receive luteal phase progesterone support. No egg donor cycles or gestational carrier cycles were included.
Partners Healthcare Institutional Review Board approved the study (protocol number: 2018P002957).

Management of CET cycles
Patients initiated a true natural cycle CET in discussion with their primary provider. They called the o ce with cycle day 1 of their menstrual period and initiated either home urinary LH testing or serum LH testing per their prescribed protocol. If urinary LH testing, they reported a positive test and a serum progesterone was checked 3 days later to con rm ovulation with a progesterone cutoff value of 3 ng/mL. If measuring serum LH, a serum LH of 24 IU/L was taken as consistent with ovulation and no con rmatory progesterone level was collected. If luteal phase progesterone was ordered, supplementation was initiated 3 days after the LH surge. The route of progesterone administration was given based on the ordering provider's preference and included both vaginal (Progesterone Gel 8% or Progesterone Vaginal Insert 100 mg) and intramuscular routes (Progesterone 50 mg). Embryo transfer was subsequently performed 6 days after either serum or urine LH surge with progesterone supplementation, regardless of whether embryos had been vitri ed on day 5 or day 6. Progesterone was continued so long as pregnancy was viable, or until 10 gestational weeks.

Clinical outcomes
The primary outcome was ongoing pregnancy rate de ned as evidence of an intrauterine pregnancy (IUP) at discharge of care to the patient's primary obstetrician typically around 8-12 weeks gestational age. Secondary outcomes were positive serum human chorionic gonadotropin (HCG) level, implantation rate, clinical pregnancy rate, miscarriage/abortion rate, ectopic pregnancy rate, and multifetal gestations.
Human chorionic gonadotropin (HCG) was checked 11 days after embryo transfer and trended for 3 values at 48-hour intervals. A viability ultrasound was performed at 7-9 weeks gestational age or sooner if felt to be warranted by clinical indications. Implantation rate was de ned as the number of gestational sacs divided by the number of embryos transferred. Clinical pregnancy was de ned as one or more gestational sacs con rmed on ultrasound. Miscarriage/abortion rate was de ned as the loss of pregnancy once there was con rmation of a clinical pregnancy. Ectopic pregnancy was de ned as abnormally rising serum HCG quantitative values without evidence of a pregnancy on pathology after intrauterine aspiration or ultrasound con rmation of an extrauterine location. Multifetal gestation was de ned as greater than one fetus ever detected on ultrasound.
Embryo quality was classi ed based on a modi cation of Gardner's system (3), which assessed for stage based on degree of expansion (Stages 5 through 8) and quality of the inner cell mass (ICM; A through D) and trophectoderm (TE; a through d), as previously described (4). Embryos were then classi ed as good, fair, or poor quality as shown in Table 1. If two embryos were transferred, the embryos were classi ed as the higher of the two embryo qualities.

Statistical analyses
Values were expressed as total number of patients included (percentage) and as mean ± standard deviation. Categorical data were analyzed utilizing Fisher's exact test and non-parametric data were analyzed using the Wilcoxon rank sum test. We a priori adjusted for age. All statistical analyses were performed with SAS® version 9.4 software (Cary, NC, USA). Statistical signi cance was assumed for p < 0.05 with risk ratios and 95% con dence intervals included.

Results
A total of 229 cycles were analyzed with 149 cycles receiving progesterone supplementation and 80 cycles receiving no progesterone supplementation. In the group receiving supplementation, no apparent differences were noted in the progesterone supplementation group compared with the control group for patient demographics as related to age, BMI, gravity, parity, infertility diagnosis, AMH, race/ethnicity, number of prior transfers at our institution, or lack of prior transfers ( Table 2)  Table 3). There was no difference in embryo quality between the progesterone supplementation group and the control group (Table 2.).  Values are total number (%) or p-value and risk ratio with 95% CIs

Discussion
The major nding of this study was that no difference was observed in the incidence of ongoing pregnancies with progesterone supplementation in true natural cycle cryopreserved blastocyst transfers compared with transfer not supplemented with progesterone. Also, no difference between the two groups was found regarding the incidence of cycles resulting in positive HCG, implantation rate, clinical pregnancy, miscarriage/spontaneous abortion, ectopic pregnancy, or multifetal gestation.
Our nding of no difference between the two groups for ongoing pregnancy contrast that of a signi cantly higher live birth rate reported by Bjuresten et al in true natural cycle CET cycles supplemented with vaginal progesterone in a randomized controlled trial (30% vs. 20%, P = 0.0272). Similar to our ndings for the other clinical outcomes, however, Bjuresten et al did not nd a difference in the percentage of cycles with a positive HCG, clinical pregnancy, miscarriage, or spontaneous abortion (5).
The primary difference between our study and the Bjuresten et al study is that we included only day 5 transfers whereas Bjuresten et al included only transfers on day 3. Another difference between our two study populations was that we included all types of progesterone supplementation and not just vaginal, which could be a possible reason for variation in our study ndings.
A similar large retrospective study evaluated vaginal progesterone luteal phase support for 2 weeks after day 3 and day 5 transfers and found a decrease in live birth rate for those true natural CET cycles without progesterone luteal phase support, compared with those receiving progesterone luteal phase support with an odds ratio of 0.58 (p = 0.003) (6).
Lee et al performed a RCT looking at true natural cycle CET of cleavage stage embryos with HCG luteal phase support and found no difference in regard to ongoing pregnancy rate, implantation rate, or miscarriage rate (7). This is consistent with our ndings that luteal phase support in true natural cycle CET does not offer clinical improvement.
Several subsequent studies evaluated luteal phase support in modi ed natural cycle CET with an HCG trigger. Eftekhar et. al performed a RCT of luteal phase support with IM progesterone on modi ed natural cycle with HCG trigger and found no difference in clinical pregnancy rate, implantation rate, or spontaneous abortion rate (8). Two other retrospective studies evaluated modi ed natural cycles with vaginal progesterone and found no difference in ongoing pregnancy rate (9), clinical pregnancy rate, implantation rate, and multiple pregnancy rate, but Kim et. al. did nd a decrease in miscarriage rate and improvement in LBR with progesterone supplementation(10). Lee et. al. performed a retrospective study which did not nd an improvement in luteal phase support with HCG injections on implantation rate, clinical pregnancy rate, or miscarriage rate (11).
Collectively, our study adds to the literature reporting no difference in ongoing pregnancy with luteal phase support in true natural cycle blastocyst cryopreserved transfers. However, there is discrepancy in the literature in which patients might bene t from luteal phase support which needs further study. It would seem the presence of a corpus luteum and the absence of a true luteal phase defect would indicate the lack of need for progesterone supplementation, particularly in blastocyst stage transfers. At this time, we can neither support nor strongly refute giving luteal phase support.
The primary strength of this paper is that this is the rst study to evaluate luteal phase progesterone supplementation in true natural cycle blastocyst cryopreserved transfers. The limitations of the study are its retrospective design, modest study size, no single regimen by which the luteal phase support was given although we did control for the regimen used, and the primary outcome was ongoing pregnancy not live birth rate. Future studies addressing these issues would certainly be warranted.

Conclusion
In true natural cycle blastocyst cryopreserved transfers there does not appear to be a bene t to giving progesterone luteal phase support regarding ongoing pregnancy, or the incidence of cycles resulting in a positive HCG, clinical pregnancy or an increase in the implantation rate.