Luteal Phase Estradiol Support for In Vitro Fertilization/Intracytoplasmic Sperm Injection Cycles

Background:

It is well established that the luteal phase is compromised in IVF cycles using GnRH analogues. Use of GnRH agonists or antagonists results in decreased production of estrogen and progesterone, a decrease in luteal phase length, and impaired endogenous gonadotropin secretion caused by persistent pituitary suppression. While the benefits of progesterone support during the luteal phase of an IVF cycle are well established, the role of estrogen support during the luteal phase is less clear.

The corpus luteum produces both progesterone and estradiol in support of the endometrium for implantation. Estrogen produced during the luteal phase modulates the concentration of progesterone receptors within the secretory endometrium in an effort to maintain sufficient receptor concentrations for progesterone stimulation. Stewart et al. highlighted the importance of luteal phase serum estradiol concentration after observing a significant difference in midluteal serum estradiol concentration between conception and non-conception cycles. This observation appears to hold true in patients undergoing IVF as well. Shahara et al. demonstrated that, not only the absolute estradiol level, but also the magnitude of estradiol decline (as measured by the ratio of peak estradiol at the time of hCG administration to midluteal estradiol) was predictive of IVF success. These studies support the idea that luteal phase estradiol supplementation may be important to implantation and IVF success.

Studies investigating the effect of luteal phase estradiol supplementation have produced conflicting results. In a small randomized, controlled trial (RCT) by Farhi et al., IVF patients receiving both oral estradiol and vaginal progesterone luteal phase support achieved higher clinical pregnancy and implantation rates than patients received vaginal progesterone alone. Lukaszuk et al. confirmed these findings and found that the most significant benefit occurred at the highest dose of estradiol administered (6mg). Elgindy et al. observed a correlation between mid-luteal serum estradiol concentration and clinical pregnancy rates in patients receiving 6mg of oral estradiol15. Similar benefits of luteal phase estradiol support have been demonstrated in both vaginal and transdermal formulations.

Conversely, multiple older studies have failed to demonstrate a benefit of luteal phase estradiol support in IVF cycles. Lin et al., in a RCT of 402 patients undergoing IVF, found no benefit of luteal phase oral estradiol. This finding was supported by studies investigating both vaginal and transdermal formulations of luteal phase estradiol support. A recent meta-analysis also did not demonstrate a statistically significant benefit of luteal phase estradiol administration to IVF outcomes, though the common odds ratio was 1.18 (95% CI: 0.98, 1.41) with a p value of 0.07. Significant heterogeneity was observed between the included studies and the authors admit that more large-scale RCTs are needed to appropriately address this question.

Since these trials were done, there has been a major shift away from long GnRH agonist stimulation to short protocol GnRH antagonist use. All but three of the studies of luteal phase estradiol support have focused on IVF cycles using the old GnRH agonist protocols given the frequency of their application. The few studies using GnRH antagonist protocols have not shown a benefit of luteal phase estradiol support but these studies have been small in scale (total n=426) or failed to report on important outcomes such as clinical pregnancy rate. Both GnRH agonists and antagonists suppress pituitary gonadotropin production and lead to a luteal phase deficiency. GnRH antagonists appear to promote premature luteolysis causing a reduction in luteal phase length and ultimately decreased pregnancy rates. Given the frequency with which these protocols are now used in IVF, a large-scale study is greatly needed to address the role of estradiol luteal phase supplementation.

It is well established that the luteal phase in IVF cycles using GnRH analogues is compromised and some form of luteal phase support is required. Compelling evidence exists to support the use of luteal phase progesterone but the role of estradiol remains controversial. Most of the existing studies are small in size and lack sufficient power. Few studies exist to determine the role of estrogen supplementation in GnRH antagonist IVF cycles. This well-powered RCT will address this important clinical question.

Methods:

Design: This is a single center RCT, conducted at ONE Fertility Burlington.

Inclusion criteria: Indications for IVF/ICSI include male factor, diminished ovarian reserve, tubal factor, ovulatory dysfunction and unexplained infertility. Female age will be 42 years or less.

IVF protocol: A short, GnRH antagonist protocol will be used for all patients. Starting on cycle day 3, ovarian stimulation will be performed using a combination of recombinant follicle-stimulating hormone (FSH) (Puregon, Ferring, or; Gonal-F, EMD Serono), medications mimicking luteinizing hormone (LH) activity, i.e recombinant human choriogonadotropin (hCG, Ferring) and/or medications containing both FSH and LH (Menopur, Ferring). Medications will be dosed according to the patient's diagnosis and ovarian reserve measurements. Follicular development will be monitored with serial transvaginal ultrasounds, and serum estradiol, progesterone and LH concentrations. A GnRH antagonist (Orgalutran, EMD Serono) will be commenced for pituitary suppression between cycle day 6 to 9. Oocyte retrieval will be performed 36 hours after administration of choriogonadotropin alpha (Ovidrel, EMD Serono), or a GnRH agonist (Decapeptyl, Ferring), to complete oocyte maturation once a sufficient cohort of mature follicles has been identified. A maximum of 2 embryos will be transferred on either day 3 or day 5 following oocyte retrieval. Serum beta hCG measurement will be performed 17 days after oocyte retrieval and, if positive, clinical pregnancy will be confirmed every 2 weeks commencing at 6 weeks gestation until 12 weeks gestation.

Recruitment and Randomization: Recruitment will occur and informed consent will be obtained at the time of treatment consent. Participants will be randomized at the time of recruitment by way of numbered, sealed envelopes to receive either 17-beta estradiol 3 mg PO/PV BID plus micronized progesterone 200 mg PV TID (treatment group) or micronized progesterone alone (control group) for luteal support commencing the day after oocyte retrieval. 17-beta estradiol will be continued until the time of pregnancy testing and, if β-hCG is positive, until 6 weeks gestation (4 weeks total). Progesterone will be continued until the time of pregnancy testing and, if β-hCG is positive, until 10 weeks gestation (8 weeks total). There will be no blinding to group allocation and no placebo. Subjects will be assigned a unique subject number prior to data analysis in order to avoid the use of any identifying information. Data will be collected using a standardized patient data form in a secure computerized database.

Outcomes: The primary outcome is clinical pregnancy, defined as the presence of fetal heart activity on ultrasound at or beyond 6 weeks gestation. Secondary outcomes include ongoing pregnancy rate (number of clinical pregnancies on ultrasound continuing beyond 12 weeks gestation/number of clinical pregnancies), implantation rate (number of clinical pregnancies/number of embryos transferred), luteal phase serum estradiol and progesterone concentration (performed at oocyte retrieval, 10 days after retrieval, and at the time of serum pregnancy testing, i.e. 17 days after retrieval), miscarriage rate (number of pregnancy losses before 20 weeks/number of clinical pregnancies) and ectopic pregnancy rate (number of ectopic pregnancies on ultrasound/number of clinical pregnancies).

Data Analysis: Descriptive analyses will be performed using SPSS software (IBM Corp., Version 22). Associations between categorical variables will be analyzed using a combination of Chi-square and Fisher's Exact tests. Associations between continuous variables will be analyzed using a combination of independent sample t-tests and logistic regression.

Sample Size Calculation: Based on a sample size calculation with assumptions of ß = 80%, α = 0.05 and an effect size of 25% in clinical pregnancy rate, a total of 506 subjects (253 per arm) will need to be recruited. This is feasible within this center over a two-year period or less, based on the current number of stimulated cycles per year of 360 and a high rate of compliance and patient interest in the study.