Luteal support in reproduction

Editor, Sexuality, Reproduction & Menopause, Professor of Obstetrics and Gynecology, Alpert Medical School of Brown University, Director, Division of Reproductive Endocrinology and Infertility, Women & Infants Hospital of Rhode Island, Providence, Rhode Island

Medical Director, Stanford Fertility & Reproductive Medicine Center, Stanford, California

Arthur H. Bill Professor and Chair, Departments of Reproductive Biology and Obstetrics & Gynecology, Case Western Reserve Medical School, University Hospitals Case Medical Center, Cleveland, Ohio

DEFINING OUR TERMS

In this panel discussion, reproductive medicine specialists Valerie L. Baker, MD, James H. Liu, MD, and Sandra A. Carson, MD (as moderator), cover practical issues in luteal support in reproduction:

• To prevent recurrent miscarriage

• To treat polycystic ovary syndrome

• To support pregnancy in assisted reproduction

For the most part, the focus is on progestogens, a class of hormones whose main purpose is to support gestation. Progesterone is the major naturally occurring progestogen. And the progestins are synthetically produced progestogens not found in nature.

The progesterone cycle

Progesterone was discovered in 1933 by Dr Willard Allen. It is synthesized in the adrenal glands, ovaries, and placenta, and also by the brain. Progesterone has a triurnal secretion pattern during the day, with levels that go up and down 3 times. The nadir is about 4 ng/mL in the luteal phase. On day 8 after the luteinizing hormone (LH) surge, progesterone peaks are about 35 ng/mL, so that the average serum concentration is between 10 and 15 ng/mL.

Progesterone supports gestation by stimulating secretion of glycogen in the endometrium. Progesterone-primed glands secrete glycogen throughout the luteal phase and—if there is no pregnancy to encourage more progesterone secretion—become exhausted later in the secretory phase. Progesterone starts to rise about 24 hours before ovulation as LH begins to rise, and that rise continues in the presence of a pregnancy.

At 7 weeks’ gestation, progesterone levels drop somewhat, the luteal-placental shift takes place, and the placenta takes care of the secretion thereafter. During pregnancy, progesterone peaks up to about 200 ng/mL around 35 to 36 weeks’ gestation. It falls after separation of the placenta, to drop to almost undetectable levels between 4 days and 2 weeks postpartum.

LUTEAL SUPPORT: IN RECURRENT PREGNANCY LOSS

Miriam is a 32-year-old woman, gravida 5, para 1, who has had regular menstrual periods throughout her life. She conceives easily, taking about 1 to 3 months for each conception, and has a normal physical examination. She has no positive review of systems and no positive medical history, with the exception of 4 prior first-trimester pregnancy losses. Two of those losses had a heartbeat on ultrasound and then proceeded to lose that heartbeat. She had 2 dilation and curettages (D&Cs) from those 4 lost pregnancies.

Miriam’s evaluation at various times from various doctors revealed normal karyotypes for herself and her husband. None of the previous abortuses had been karyotyped. She has a normal hysterosalpingogram, normal thrombophilia evaluation, negative cultures for gonorrhea, chlamydia, and syphilis. Her anticardiolipin, lupus anticoagulant, and anti-beta2-glycoproteins are normal. She has normal thyroid and prolactin concentrations. At day 21, serum progesterone drawn randomly is 5 ng/ml and her husband’s semen analysis parameters were within normal limits.

Serum progesterone levels and luteal phase dysfunction

DR CARSON: Dr Liu, does a serum progesterone level like Miriam’s indicate luteal phase dysfunction?

DR LIU: Not necessarily. Accepting a single progesterone level as a measure of luteal phase adequacy is problematic. In our laboratory, we consider 4 ng confirmatory of ovulation, but we do not use it to measure the adequacy of progesterone production. The problem we have with progesterone secretion is that it is an LH-driven process and is pulsatile in nature. And this pulsatility mirrors the LH pulse frequency driven by the gonadotropin-releasing hormone (GnRH) pulse oscillator, and it occurs at about every 4 hours. This was shown very well by Filicori in 1984 in his classic study and reconfirmed by Wuttke.^1,2 So, in the particular instance of this fertile woman on day 22 shown (FIGURE 1), you can see that the peak level is between 18 and 19 ng/mL, whereas the trough levels are in the 9 to 10 ng/mL range. Miriam’s serum progesterone, which measures 5 ng/mL, is still within normal limits, even though the Filicori study would suggest it does not get that low.

Evaluating the endometrium

DR CARSON: Dr Liu, would you evaluate the endometrium in this patient?

DR LIU: The timed endometrial biopsy that we have been taught in the past is not easily accomplished. Some women do not have clear LH surges by which to time the endometrial biopsy. The histologic dating of the endometrium is fraught with pitfalls. A study by Murray and colleagues used the same type of criteria that Noyes and colleagues used in describing the characteristic histologic changes in the endometrium in a small number of what they considered infertile women with normal menstrual cycles.^3,4 There are similarities as well as significant disparities in some areas. These data examined mitoses, pseudostratification, basal vacuolation, secretion, stromal edema, pseudodecidual reaction, and leukocyte infiltration.

DR CARSON: Are the timed endometrial biopsies reproducible and predictable for fertility?

DR LIU: The best study probably to date that looks at the utility of timed endometrial biopsies is from the National Cooperative Reproductive Medicine Network.⁵ The authors recruited about 619 women, described as fertile (those individuals with one or more children) or infertile (by the standard definition), and looked at timed biopsies both in the midluteal and late luteal phase. For the midluteal phase biopsies, taken on day 22 to 23, the women in fertile couples were further out of phase—that is, more days’ difference between the postovulatory day and the histologic dating using the Noyes criteria—than the women in infertile couples (TABLE 1). The differences were not statistically significant, but I think the results indicate the tremendous difficulties in our ability to discriminate fertile from infertile women on the basis of midluteal phase data. If you look at the late luteal phase data, the same pattern emerges. Neither midluteal nor late luteal biopsies discriminate accurately between fertile and infertile women, which leads me to believe that timed endometrial biopsy is not a worthwhile test.

DR CARSON: Are there any data showing that endometrial biopsies are useful in patients like Miriam, who conceive yet suffer recurrent pregnancy loss?

DR LIU: I am not aware of any true randomized trials.

Would you evaluate the endometrium?

Using progesterone to prevent miscarriage

DR CARSON: Dr Baker, do you use progesterone in treating patients with recurrent loss?

DR BAKER: In some cases, but its use is not evidence-based. The rationale for doing so is based on classic studies showing that excision of the corpus luteum before gestational age of 7 weeks led to an abrupt decrease in serum progesterone, followed by miscarriage. But if excision occurred later—at 8 or more weeks by menstrual dating—progesterone decreased only slightly, and the pregnancy continued. Exogenous progesterone used after lutectomy at a gestational age of less than 7 weeks prevented miscarriage.⁶ That study demonstrated that early pregnancy is dependent on corpus luteum production of progesterone before the gestational age of 7 weeks. The luteal-placental shift occurs somewhere between 7 and 9 weeks. At a gestational age of 9 weeks or more, progesterone from the trophoblast supports the pregnancy.⁷

Nevertheless, the evidence that progesterone therapy can avert miscarriage is not substantial. A recent Cochrane review found that the odds ratio for miscarriage with progestogen treatment was 0.98, that is, not different from having no treatment. The reviewers concluded that there was no evidence to support routine use of progestogen to prevent miscarriage in early to mid-pregnancy.⁸ In recurrent pregnancy loss, other causes should be sought. On the other hand, the risk of treatment is minimal, and it is certainly within the standard of care to use progesterone if the side effects are not bothersome.

Administering progesterone

DR CARSON: If progesterone is used, how and when would you use it?

DR BAKER: We use vaginal capsules or inserts, 100 mg to 200 mg, 2 to 3 times per day. We have also used compounded suppositories and vaginal gel. Less commonly, we have had patients who prefer to use intramuscular (IM) progesterone. We ask patients to measure the LH surge or temperature rise and then start progesterone 3 to 4 days after the temperature rise or LH surge. We continue to use progesterone through the first trimester. None of these regimens are evidence-based or approved by the FDA for recurrent pregnancy loss.

DR LIU: Our practice is similar. I tell patients that we do not know that the treatment will prevent another miscarriage, because we do not know why as many as half of the cases of recurrent loss occur. But I also emphasize that the treatment is very benign.

Case study follow-up

DR BAKER: Although Miriam had had local side effects from the IM progesterone, she chose it over a proposed vaginal preparation for her next cycle, based on her perception of potentially improved efficacy. However, when her husband began traveling for his job, she switched to a vaginal preparation and did achieve a pregnancy. Her physician told her she could safely discontinue progesterone at a gestational age of 8 weeks, but she chose to continue until the end of the first trimester.

LUTEAL SUPPORT: FOR POLYCYSTIC OVARY SYNDROME

Myra is 39 years old, gravida 1, para 1. She conceived on clomiphene citrate 5 years ago. She has known polycystic ovary syndrome (PCOs) and has 2 to 3 menses per year. Her body mass index is 29 kg/m². She has a normal physical examination, with no acanthosis nigricans. Her blood pressure and fasting glucose screen are normal, and an oral glucose tolerance test also revealed no diabetes. What would progesterone accomplish for myra as a contraceptive, if she does not want to get pregnant, or to support a pregnancy, if that is what she wants?

Endometrial evaluation

DR CARSON: Dr Baker, would you evaluate the endometrium in this patient?

DR BAKER: Myra should undergo an endometrial biopsy because she is at risk for endometrial cancer and PCOS. In practice, some patients may not be at as high a risk for endometrial hyperplasia and cancer as Myra because they have been on oral contraceptives throughout their lives to regulate their cycles. But, in general, it is important to err on the side of performing the endometrial biopsy if a patient has been anovulatory and has not had regular progestin withdrawal.

Patients such as Myra are at a risk of developing endometrial polyps. I frequently perform saline infusion sonography to rule out polyps. In our practice, we also have the ability to perform office hysteroscopy to detect and remove endometrial polyps and to sample the endometrium to rule out hyperplasia.

DR CARSON: Dr Liu, how would you evaluate the endometrium?

DR LIU: I probably would do an endometrial biopsy. I think the patient in this case—with only 2 to 3 cycles per year—is at increased risk for hyperplasia, with or without atypia. We do not know if those cycles actually reflect sporadic ovulation or an anovulatory type of bleeding. Unfortunately, ultrasound and endometrial thickness are not good criteria in reproductive-age women.

When pregnancy is not desired

DR CARSON: If Myra does not want to become pregnant, Dr Liu, what would you recommend for long-term management?

DR LIU: I would recommend any type of contraceptive with a progestational agent: Combination oral contraceptives, long-cycle regimens that reduce the number of periods to 4 each year, such as Seasonale or Seasonique, or generic formulations used in a long-cycle fashion for 3 to 4 months at a time. A progestin-releasing intrauterine device (IUD) like Mirena would be a reasonable option, or a long-acting progestin preparation, such as Depo-Provera or depomedroxyprogesterone acetate, which is available in a subcutaneous form. The long-lasting Implanon etonogestrel implant would also be an option.

When pregnancy is desired

DR CARSON: Dr Baker, if this patient does want to become pregnant, would you use progesterone with clomiphene or the aromatase inhibitors?

DR BAKER: Not routinely. There really are no good data to suggest that progesterone supplementation would be necessary or helpful with either of those treatments. There is some evidence that progesterone might be helpful with gonadotropin therapy, and we tend to use it in those cases.

Studies from the 1980s demonstrated a shorter luteal phase much more commonly in women who were on gonadotropin therapy. In a recent randomized trial from Turkey of patients with unexplained infertility being treated with recombinant follicle stimulating hormone (FSH), the live birth rate was approximately double in the group that used progesterone supplementation in the luteal phase compared with those who did not.⁹

LUTEAL SUPPORT: TO SUSTAIN IVF PREGNANCY

ART with stimulation regimens

DR CARSON: Our final topic is the role of progesterone supplementation in assisted reproduction with various stimulation regimens. Dr Baker, do you use luteal support for in vitro fertilization (IVF)? And if so, what kind and at what dose?

DR BAKER: Yes. We base that practice on data from the 2004 Cochrane review, which found a statistically significant odds ratio of 2.38 for maintaining IVF pregnancy with luteal phase support. The review included cycles using a GnRH agonist for downregulation. The review was withdrawn in July 2008, but this is a routine process Cochrane uses whenever a review has not been updated within several years and does not imply that any errors have been detected.¹⁰

A number of explanations for the beneficial effect of progesterone supplementation on IVF pregnancy rates have been suggested. Luteal support may be helpful because the GnRH agonists and antagonists suppress pituitary production of LH, which is required for progesterone production. Progesterone secretion may also be inhibited by follicle disruption involved in oocyte retrieval. In addition, estradiol levels in IVF cycles are very high; progesterone supplementation may restore a more physiologic estrogen to progesterone ratio.¹¹ We could achieve similar pregnancy rates using human chorionic gonadotropin (hCG) rather than progesterone, but the risk of ovarian hyperstimulation syndrome (OHSS) is higher with hCG. The case for using progesterone to support IVF pregnancy is made in the ASRM Practice Committee report published this year in Fertility and Sterility.^7,10

Route of administration

DR CARSON: Dr Baker, how is progesterone administered?

DR BAKER: Intramuscular progesterone is still quite commonly used, 50 mg qd, in sesame oil, peanut oil, or ethyl oleate. Other possibilities include the vaginal inserts either 2 or 3 times a day at a 100 mg dose; vaginal gel, 90 mg daily; vaginal capsules, 200 mg, 3 times daily; and compounded vaginal suppositories. Oral progesterone—except for dydrogesterone, which is not available in the United States—is not typically recommended. And of these, the only 2 that are FDA-approved currently are the inserts and the gel (TABLE 2).

DR CARSON: Dr Liu, why use painful IM progesterone when vaginal and oral preparations are available?

DR LIU: One reason is cost. Insurance companies may not cover branded progesterone preparations that are new and expensive. If I had a choice of what preparation to use, I would definitely treat using a vaginal preparation but probably not the oral.

Options for progesterone luteal phase support in IVF cycles

Impact on live birth rate

DR CARSON: Dr Baker, is there a difference in live birth rate with IM versus vaginal progesterone?

DR BAKER: This is a very controversial question. In the 2004 Cochrane review that was withdrawn, there was a suggestion of higher ongoing live birth rates with IM progesterone.¹⁰ I think that is the reason many physicians are committed to using IM progesterone. However, these analyses often used much lower doses of vaginal progesterone than are commonly recommended today, and they used formulations that were locally compounded and not necessarily consistent. So I think we need to be open-minded as more data are collected.

A study done in 2001 at Brigham and Women’s Hospital showed a higher rate of live birth with IM progesterone than with vaginal gel (TABLE 3).¹² The vaginal gel was voluntarily withdrawn from the market that year and reformulated. A 2008 study from the same group found no difference between IM and vaginal gel.¹³ This second study began the vaginal progesterone 2 days after the retrieval, rather than 1 day afterward, as done in the 2001 study. We do not know whether this difference in timing accounts for the difference in outcome, but the authors of the study emphasized that point.

More studies are being published comparing available vaginal preparations with IM progesterone, breaking out subanalyses from larger trials. One such study looked at matched pairs comparing vaginal with IM administration and found no difference in live birth rate between the 2 routes of administration.¹⁴

The ASRM Practice Committee Opinion this year says that the optimal route of progesterone administration has not been established. We should remember that although providers are comfortable with IM progesterone, many patients are not. They find the shots painful and cumbersome. So I think continuing to look for new formulations and collecting more data is important.

Progesterone gel versus IM

Luteal support with differing ovulation-induction regimens

DR CARSON: Dr Liu, are there any particular ovulation-induction regimens that alter your choice of luteal phase support?

DR LIU: Any stimulation program that targets the pituitary GnRH receptor will need to consider luteal phase support. In a study by Beckers that provides evidence of LH insufficiency after GnRH stimulated cycles, patients who received hCG support during the luteal phase maintained both estradiol and progesterone at very robust amounts, whereas groups that did not have luteal phase support show a significant lower peak level of progesterone. What is even more important is that patients who did not have luteal phase support had a truncation of progesterone secretion before day 10, which may not be sufficient to allow hCG rescue.¹⁵

When and for how long?

DR CARSON: Dr Baker, when do you start the luteal progesterone and how long do you continue it?

DR BAKER: Starting the progesterone some time after retrieval but before embryo transfer is fairly standard. The recent large study initiated vaginal progesterone the day after the retrieval, and I think that is fairly common.¹⁶

DR CARSON: How long would luteal support be continued?

DR BAKER: The available evidence says that support is justified until about the time of diagnosis of pregnancy, typically 2 weeks after the retrieval. It is common practice to continue luteal support up to the detection of a fetal heart beat or at a gestational age of 10 to 12 weeks. The recent large trial that was done on vaginal inserts versus gel did continue into that time frame.¹⁶ Part of the reason for continuing so long is that if patients miscarry after they stop using progesterone, they think stopping the progesterone caused the miscarriage. In fact, it is more likely that the miscarriage results from an abnormal embryo because the lutectomy studies show that exogenous progesterone or even progesterone from the corpus luteum should not be necessary after 9 weeks’ gestation.

Sexuality, Reproduction & Menopause ©2009 Dowden Health Media