Progesterone prepares the endometrium for pregnancy by stimulating proliferation in response to
human chorionic gonadotropin (hCG), which is produced by the corpus luteum. This occurs in the luteal phase of the menstrual cycle. In assisted reproduction techniques (ART) the
progesterone or hCG levels, or both, are low and the natural process is insufficient, so the luteal phase is supported with either
progesterone, hCG or
gonadotropin releasing hormone (
GnRH) agonists. Luteal phase support improves implantation rate and thus pregnancy rates but the ideal method is still unclear. This is an update of a Cochrane Review published in 2004 (Daya 2004).
OBJECTIVES: To determine the relative effectiveness and safety of methods of luteal phase support in subfertile women undergoing assisted reproductive technology.
SEARCH STRATEGY: We searched the Cochrane Menstrual Disorders and
Subfertility Group (MDSG) Specialised Register, Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE, PsycINFO, CINAHL, Database of Abstracts of Reviews of Effects (DARE), LILACS, conference abstracts on the ISI Web of Knowledge, OpenSigle for grey literature from Europe, and ongoing clinical trials registered online. The final search was in February 2011.
SELECTION CRITERIA: We extracted data per women and three review authors independently assessed risk of bias. We contacted the original authors when data were missing or the risk of bias was unclear. We entered all data in six different comparisons. We calculated the Peto odds ratio (Peto OR) for each comparison.
MAIN RESULTS: Sixty-nine studies with a total of 16,327 women were included. We assessed most of the studies as having an unclear risk of bias, which we interpreted as a high risk of bias. Because of the great number of different comparisons, the average number of included studies in a single comparison was only 1.5 for live birth and 6.1 for clinical pregnancy.Five studies (746 women) compared hCG versus placebo or no treatment. There was no evidence of a difference between hCG and placebo or no treatment except for ongoing pregnancy: Peto OR 1.75 (95% CI 1.09 to 2.81), suggesting a benefit from hCG. There was a significantly higher risk of
ovarian hyperstimulation syndrome (OHSS) when hCG was used (Peto OR 3.62, 95% CI 1.85 to 7.06).There were eight studies (875 women) in the second comparison,
progesterone versus placebo or no treatment. The results suggested a significant effect in favour of
progesterone for the live birth rate (Peto OR 2.95, 95% CI 1.02 to 8.56) based on one study. For clinical pregnancy (
CPR) the results also suggested a significant result in favour of
progesterone (Peto OR 1.83, 95% CI 1.29 to 2.61) based on seven studies. For the other outcomes the results indicated no difference in effect.The third comparison (15 studies, 2117 women) investigated
progesterone versus hCG regimens. The hCG regimens were subgrouped into comparisons of
progesterone versus hCG and
progesterone versus
progesterone + hCG. The results did not indicate a difference of effect between the interventions, except for OHSS. Subgroup analysis of
progesterone versus
progesterone + hCG showed a significant benefit from
progesterone (Peto OR 0.45, 95% CI 0.26 to 0.79).The fourth comparison (nine studies, 1571 women) compared
progesterone versus
progesterone + oestrogen. Outcomes were subgrouped by route of administration. The results for clinical pregnancy rate in the subgroup
progesterone versus
progesterone + transdermal oestrogen suggested a significant benefit from
progesterone + oestrogen. There was no evidence of a difference in effect for other outcomes.Six studies (1646 women) investigated
progesterone versus
progesterone +
GnRH agonist. We subgrouped the studies for single-dose
GnRH agonist and multiple-dose
GnRH agonist. For the live birth, clinical pregnancy and ongoing pregnancy rate the results suggested a significant effect in favour of
progesterone +
GnRH agonist. The Peto OR for the live birth rate was 2.44 (95% CI 1.62 to 3.67), for the clinical pregnancy rate was 1.36 (95% CI 1.11 to 1.66) and for the ongoing pregnancy rate was 1.31 (95% CI 1.03 to 1.67). The results for
miscarriage and multiple pregnancy did not indicate a difference of effect.The last comparison (32 studies, 9839 women) investigated different
progesterone regimens:intramuscular (IM) versus
oral administration, IM versus vaginal or
rectal administration, vaginal or rectal versus
oral administration, low-dose vaginal versus high-dose vaginal
progesterone administration, short protocol versus long protocol and micronized
progesterone versus synthetic
progesterone. The main results of this comparison did not indicate a difference of effect except in some subgroup analyses. For the outcome clinical pregnancy, subgroup analysis of micronized
progesterone versus synthetic
progesterone showed a significant benefit from synthetic
progesterone (Peto OR 0.79, 95% CI 0.65 to 0.96). For the outcome multiple pregnancy, the subgroup analysis of IM
progesterone versus oral
progesterone suggested a significant benefit from oral
progesterone (Peto OR 4.39, 95% CI 1.28 to 15.01).
AUTHORS' CONCLUSIONS: This review showed a significant effect in favour of
progesterone for luteal phase support, favouring synthetic
progesterone over micronized
progesterone. Overall, the addition of other substances such as
estrogen or hCG did not seem to improve outcomes. We also found no evidence favouring a specific route or duration of administration of
progesterone. We found that hCG, or hCG plus
progesterone, was associated with a higher risk of OHSS. The use of hCG should therefore be avoided. There were significant results showing a benefit from addition of
GnRH agonist to
progesterone for the outcomes of live birth, clinical pregnancy and ongoing pregnancy. For now,
progesterone seems to be the best option as luteal phase support, with better pregnancy results when synthetic
progesterone is used.