Progestogens for prevention of luteinising hormone (LH) surge in women undergoing controlled ovarian hyperstimulation as part of an assisted reproductive technology (ART) cycle

Background: Currently, gonadotrophin releasing hormone (GnRH) analogues are used to prevent premature ovulation in ART cycles. However, their costs remain high, the route of administration is invasive and has some adverse effects. Oral progestogens could be cheaper and effective to prevent a premature LH surge.

Objectives: To evaluate the effectiveness and safety of using progestogens to avoid spontaneous ovulation in women undergoing controlled ovarian hyperstimulation (COH).

Search methods: We searched the Cochrane Gynaecology and Fertility Group trials register, CENTRAL, MEDLINE, Embase and PsycINFO in Dec 2021. We contacted study authors and experts to identify additional studies.

Selection criteria: We included randomised controlled trials (RCTs) that included progestogens for ovulation inhibition in women undergoing controlled ovarian hyperstimulation (COH).

Data collection and analysis: We used standard methodological procedures recommended by Cochrane, including the risk of bias (RoB) assessment. The primary review outcomes were live birth rate (LBR) and oocyte pick-up cancellation rate (OPCR). Secondary outcomes were clinical pregnancy rate (CPR), cumulative pregnancy, miscarriage rate (MR), multiple pregnancies, LH surge, total and MII oocytes, days of stimulation, dose of gonadotropins, and moderate/severe ovarian hyperstimulation syndrome (OHSS) rate. The primary analyses were restricted to studies at overall low and some concerns RoB, and sensitivity analysis included all studies. We used the GRADE approach to assess the certainty of evidence.

Main results: We included 14 RCTs (2643 subfertile women undergoing ART, 47 women used oocyte freezing for fertility preservation and 534 oocyte donors). Progestogens versus GnRH antagonists We are very uncertain of the effect of medroxyprogesterone acetate (MPA) 10 mg compared with cetrorelix on the LBR in poor responders (odds ratio (OR) 1.25, 95% confidence interval (CI) 0.73 to 2.13, one RCT, N = 340, very-low-certainty evidence), suggesting that if the chance of live birth following GnRH antagonists is assumed to be 18%, the chance following MPA would be 14% to 32%. There may be little or no difference in OPCR between progestogens and GnRH antagonists, but due to wide Cs (CIs), we are uncertain (OR 0.92, 95%CI 0.42 to 2.01, 3 RCTs, N = 648, I² = 0%, low-certainty evidence), changing the chance of OPCR from 4% with progestogens to 2% to 8%. Given the imprecision found, no conclusions can be retrieved on CPR and MR. Low-quality evidence suggested that using micronised progesterone in normo-responders may increase by 2 to 6 the MII oocytes in comparison to GnRH antagonists. There may be little or no differences in gonadotropin doses. Progestogens versus GnRH agonists Results were uncertain for all outcomes comparing progestogens with GnRH agonists. One progestogen versus another progestogen The analyses comparing one progestogen versus another progestogen for LBR did not meet our criteria for primary analyses. The OPCR was probably lower in the MPA 10 mg in comparison to MPA 4 mg (OR 2.27, 95%CI 0.90 to 5.74, one RCT, N = 300, moderate-certainty evidence), and MPA 4 mg may be lower than micronised progesterone 100 mg, but due to wide CI, we are uncertain of the effect (OR 0.81, 95%CI 0.43 to 1.53, one RCT, N = 300, low-certainty evidence), changing the chance of OPCR from 5% with MPA 4 mg to 5% to22%, and from 17% with micronised progesterone 100 mg to 8% to 24%. When comparing dydrogesterone 20 mg to MPA, the OPCR is probably lower in the dydrogesterone group in comparison to MPA 10 mg (OR 1.49, 95%CI 0.80 to 2.80, one RCT, N = 520, moderate-certainty evidence), and it may be lower in dydrogesterone group in comparison to MPA 4 mg but due to wide confidence interval, we are uncertain of the effect (OR 1.19, 95%CI 0.61 to 2.34, one RCT, N = 300, low-certainty evidence), changing the chance of OPCR from 7% with dydrogesterone 20 to 6-17%, and in MPA 4 mg from 12% to 8% to 24%. When comparing dydrogesterone 20 mg to micronised progesterone 100 mg, the OPCR is probably lower in the dydrogesterone group (OR 1.54, 95%CI 0.94 to 2.52, two RCTs, N=550, I² = 0%, moderate-certainty evidence), changing OPCR from 11% with dydrogesterone to 10% to 24%. We are very uncertain of the effect in normo-responders of micronised progesterone 100 mg compared with micronised progesterone 200 mg on the OPCR (OR 0.35, 95%CI 0.09 to 1.37, one RCT, N = 150, very-low-certainty evidence). There is probably little or no difference in CPR and MR between MPA 10 mg and dydrogesterone 20 mg. There may be little or no differences in MII oocytes and gonadotropins doses. No cases of moderate/severe OHSS were reported in most of the groups in any of the comparisons.

Authors' conclusions: Little or no differences in LBR may exist when comparing MPA 4 mg with GnRH agonists in normo-responders. OPCR may be slightly increased in the MPA 4 mg group, but MPA 4 mg reduces the doses of gonadotropins in comparison to GnRH agonists. Little or no differences in OPCR may exist between progestogens and GnRH antagonists in normo-responders and donors. However, micronised progesterone could improve by 2 to 6 MII oocytes. When comparing one progestogen to another, dydrogesterone suggested slightly lower OPCR than MPA and micronised progesterone, and MPA suggested slightly lower OPCR than the micronised progesterone 100 mg. Finally, MPA 10 mg suggests a lower OPCR than MPA 4 mg. There is uncertainty regarding the rest of the outcomes due to imprecision and no solid conclusions can be drawn.