REI Trials
European and Israeli Trials: HP-hMG vs recombinant FSH in IVF/ICSI (EISG and MERIT datasets)
Study design and scope
Type: Multinational, open-label, randomized, parallel-group, phase III trials; noninferiority framework for primary outcome in several analyses.
Settings: 22 centers across Belgium, Germany, Israel, The Netherlands, Switzerland, United Kingdom; later integrated across additional European centers for MERIT analysis.
Trials involved two gonadotropin preparations:
Highly purified hMG (HP-hMG, Menopur) containing FSH and LH activity (via HCG-driven LH activity)
Recombinant FSH (rFSH, follitropin alfa)
Key trials and design notes:
EISG trial (2002): HP-hMG vs recombinant FSH; large, noninferiority design with ongoing pregnancy rate as primary endpoint; randomization occurred before down-regulation; IVF/ICSI cycles; 731 randomized (APT/PP populations described separately).
MERIT trial (2006): Large assessor-blind randomized trial in a similar population; post-randomization exposure to gonadotropins; designed as superiority with possible noninferiority consideration; heavy emphasis on ongoing pregnancy as primary endpoint in original design; integration with EISG data performed subsequently.
Primary endpoint common to these analyses: ongoing pregnancy rate following one IVF/ICSI cycle (clinical pregnancy confirmed by ultrasound at ~10 weeks post-egg retrieval).
Key noninferiority/superiority thresholds:
EISG MERIT analyses used a noninferiority margin of about
ext{Lower CI of difference} > -0.10
(i.e., HP-hMG should be no worse than recombinant FSH by more than 10 percentage points in ongoing pregnancy rate).MERIT integrated analysis later used different framing (live birth and ongoing pregnancy as endpoints) and reported noninferiority margins consistent with study design conventions of that project.
Inclusion/exclusion criteria (highlights)
Age: typically 18–38 years (EISGMERIT cohorts varied slightly in exact ranges across centers).
Infertility duration and cause: diverse etiologies (male factor, tubal factors, unexplained, endometriosis I/II in some cohorts).
BMI: generally within 18–29 kg/m² (with site-specific allowances).
Regular cycles with down-regulation using a long GnRH agonist protocol prior to stimulation.
Baseline reproductive hormones within normal ranges; estradiol <200–>1000 pmol/L at certain milestones for stimulation eligibility.
Exclusions included significant systemic disease, ovarian cysts, contraindications to gonadotropins, poor responder definitions, OHSS history, heavy smoking/alcohol use, and recent participation in other trials.
Stimulation protocol and dosing (common aspects across trials)
GnRH agonist down-regulation: long protocol (e.g., triptorelin 0.1 mg/day or depot formulations) started in midluteal phase; down-regulation confirmed by ultrasound/endocrine parameters before stimulation.
Start of stimulation: fixed dose of either HP-hMG or rFSH at daily for the first 5 days (visit 3).
Dose adjustment: on day 6, dose could be increased up to a maximum of (≈6 vials) or adjusted downward; adjustments allowed every ~4 days.
Criteria for oocyte maturation trigger: three follicles ≥16–17 mm and/or estradiol thresholds; trigger with hCG 5000–10,000 IU (or recombinant hCG in later/alternative protocols).
Oocytes retrieved ~36 hours after hCG; fertilization by standard IVF or ICSI per center practice.
Embryo transfer: typically 1–3 embryos transferred in various trials; single-embryo transfer emphasized in some later antagonist-cycle studies (e.g., MEGASET).
Luteal support: progesterone with local practice (e.g., vaginal progesterone 90 mg/day).
Cryopreservation: available for surplus good-quality embryos; some trials include frozen Embryo Transfer (FET) cycles within a year post-randomization.
Primary and secondary endpoints (highlights)
Primary efficacy endpoint: ongoing pregnancy rate per completed IVF/ICSI cycle (PP for primary noninferiority analyses; ITT used for safety analyses in MERIT-type designs).
Secondary efficacy endpoints: biochemical pregnancy (hCG positive), clinical pregnancy (ultrasound confirmation), number of follicles, number of oocytes retrieved, fertilization rates, number of embryos transferred, estradiol levels (on days 3/6 and day of hCG), endometrial status (thickness and echogenicity), days of stimulation, and number of vials used.
Safety endpoints: adverse events by system, frequency and severity of OHSS (type 2/3), miscarriages, multiple gestations, injection-site reactions; tolerability diaries for injection-site events.
Key results: ongoing pregnancy and safety (EISG/ MERIT era)
Ongoing pregnancy (PP population, per completed cycle): HP-hMG 25% (n ≈ 85/344) vs rFSH 22% (n ≈ 71/317); difference ≈ +2.62% (P ≈ 0.17). Lower CI limit ≥ -10% confirmed noninferiority.
Ongoing pregnancy per embryo transferred (to assess implantation success): similar noninferiority tendencies; overall, noninferiority of HP-hMG was confirmed in APT population for various end-points.
Biochemical and clinical pregnancies tended to be higher with HP-hMG but differences often not statistically significant.
Endocrine and follicular response differences:
Estradiol on the day of hCG: statistically higher with HP-hMG; estradiol concentration difference between HP-hMG and rFSH was significant (ΔE2 ≈ 1265 pmol/L; 95% CI 680–1864; P < 0.001).
Total follicles and follicle sizes: on day end of stimulation, rFSH produced more total follicles and more follicles ≥10 mm; however, estradiol and progesterone dynamics differed (see below).
Progesterone end-of-stimulation: higher with rFSH (P < 0.001).
Endometrial echogenicity: rFSH group showed a shift toward hyperechogenic endometrium more often than HP-hMG (P = 0.023).
Fertilization and embryo transfer metrics (PP population): similar fertilization rates and number of embryos transferred; HP-hMG often produced a higher proportion of top-quality embryos per retrieval in some analyses.
Safety: OHSS rates were low and similar overall; rare OHSS events occurred with both treatments; injection-site tolerability was similar across groups.
Live birth and total cycle outcomes (integrated MERIT analysis): live birth per started cycle in the fresh cycle tended to favor HP-hMG (e.g., 26% vs 21% in one analysis, OR ≈ 1.36; P ≈ 0.041 for live birth per cycle initiated in MERIT). Overall cumulative live birth rates with fresh and frozen cycles tended to be similar across HP-hMG and rFSH arms in some analyses, depending on the data subset and inclusion criteria.
Embryo quality and implantation: in MERIT-related embryo quality analyses, HP-hMG often yielded a higher proportion of top-quality embryos; top-quality embryos from HP-hMG were more likely to implant and lead to ongoing pregnancy and live birth in some sub-analyses.
Embryo quality-focused findings (MERIT-related studies)
Local vs central embryo assessment (MERIT-associated analyses, Ziebe et al.):
Top-quality embryos per oocyte retrieved (local): HP-hMG 11.3% vs rFSH 9.0%; P = 0.044.
Central review showed a non-significant difference (9.5% vs 8.0%).
Consolidated score: HP-hMG 48% vs rFSH 44% had top-quality embryos across patients.
Embryo quality parameters (DAY 2 vs DAY 3): HP-hMG associated with higher cell number and lower fragmentation on day 2/3 in some analyses; fragmentation >20% more common in HP-hMG but not consistently associated with worse outcomes across all endpoints.
Embryo transfer outcomes with top-quality embryos: when only top-quality embryos were transferred, live birth and ongoing pregnancy rates tended to be higher for HP-hMG (e.g., live birth 48% vs 32%; ongoing pregnancy 48% vs 32%; ongoing implantation 41% vs 27%), though statistical significance varied by analysis and central vs local scoring.
Cryopreservation: similar 1-year cryo-survival metrics; higher proportion of blastomeres surviving thaw in HP-hMG and higher resumption of mitosis after thaw in HP-hMG–derived embryos in some analyses.
MEGASET trial: GnRH antagonist cycle with compulsory single-blastocyst transfer (Fertil Steril 2012)
Design: randomized, open-label, assessor-blind, multicenter; comparison of HP-hMG (Menopur) vs rFSH (Gonal-F) in a GnRH antagonist COS cycle; compulsory single-blastocyst transfer on day 5; some cycles allowed subsequent frozen blastocyst transfers within 1 year.
Primary endpoint: ongoing pregnancy rate per started cycle (fresh cycle).
Key results (fresh cycle): ongoing pregnancy rate per started cycle 30% for HP-hMG vs 27% for rFSH in PP; 29% vs 27% in ITT; noninferiority established (lower limit of 95% CI above -12% margin).
Live birth: fresh cycle live birth rate 52.2% for HP-hMG vs 48.7% for rFSH; frozen cycle live birth rate 63.4% vs 50.8%; both cycles contributed to cumulative live birth rate: 50.6% for HP-hMG vs 51.5% for rFSH (cumulative across fresh+frozen within 1 year).
Embryo transfer and blastocyst data: mean embryos transferred per cycle ~1.7; most transfers were single; 82% of randomized patients underwent embryo transfer in the study; higher proportion of single transfers when comparing across arms due to policy and clinical practice.
OHSS safety: OHSS incidence lower with HP-hMG (approximately 9.7%) than with rFSH (≈21.4%), a notable safety difference; early OHSS rates also lower with HP-hMG (≈6.1% vs ≈17.5%).
Cumulative outcomes: despite more transfers in the rFSH arm, cumulative live birth rates were very similar between arms; the study supports noninferiority and favorable safety with HP-hMG.
Neonatal outcomes: overall neonatal health similar across arms; monozygotic twinning observed but in line with expected rates for blastocyst transfers.
Interpretive conclusions: HP-hMG yields comparable efficacy with a favorable safety profile (lower OHSS, reduced pregnancy losses), suggesting a favorable risk/benefit balance for HP-hMG in GnRH antagonist cycles with single-embryo transfers; differences in ovarian response and endometrial parameters were observed and discussed.
MEGASET-HR trial: high responders (Fertil Steril 2020)
Design: randomized, open-label, assessor-blind, multicenter trial in the U.S. (MEGASET-HR); included women predicted as high responders (AMH-based criteria: AMH ≥5 ng/mL) undergoing COS; HP-hMG vs rFSH + rLH combination (fixed-dose 2:1 ratio; HP-hMG equivalent vs rFSH with LH supplementation).
Primary endpoint: ongoing pregnancy rate per cycle start after fresh transfer.
Baseline and eligibility: AMH-based high-responder identification; AMH measured centrally; Bologna/POR-like criteria applied for risk stratification; inclusion criteria were designed to reflect high responder populations.
Treatment regimens: HP-hMG vs rFSH+rLH with GnRH antagonist coadministration; baseline dose was 150 IU/day for initial days (adjustments allowed); GnRH antagonist used to prevent premature LH surge.
Outcomes (fresh cycle): ongoing pregnancy rate per cycle start 35.5% for HP-hMG vs 30.7% for rFSH; noninferiority demonstrated (1-sided p-value, margin pre-specified); written as noninferiority with a lower bound exceeding the pre-specified margin.
Cumulative live birth outcomes: cumulative live birth rate per patient over a year (including frozen cycles) 40% for HP-hMG and 38% for rFSH; results are reported with 95% CIs around the differences; no significant superiority for HP-hMG over rFSH; the study supports a comparable performance with potential safety advantages for HP-hMG.
Pharmacodynamics and safety: notable differences in pharmacodynamics observed (e.g., LH exposure with LH-containing regimens); OHSS rates were lower with HP-hMG; endometrial and embryonic development metrics suggested differential influences on endometrial receptivity and embryo development, though primary outcomes were noninferior.
Safety and tolerability: TEAEs occurred but did not reveal major safety concerns; OHSS incidence was lower with HP-hMG; injection-site tolerability and other adverse events were similar across arms.
Mechanistic and practical implications
LH activity appears to modulate folliculogenesis, oocyte maturation, and endometrial receptivity; HP-hMG provides FSH activity plus LH activity via HCG-derived LH activity.
Higher estradiol levels with HP-hMG at the time of hCG trigger may reflect LH-driven steroidogenesis leading to endometrial enhancement; however, elevated progesterone late in stimulation (end-of-stimulation) was more associated with rFSH, potentially affecting endometrial timing and implantation in some cycles.
Endometrial echogenicity shifts toward hyperechogenic patterns with rFSH, which has been associated with poorer implantation in some contexts; HP-hMG tended toward more favorable endometrial patterns in certain analyses.
Embryo quality: HP-hMG tended to produce a higher proportion of top-quality embryos in some embryo assessments, and top-quality embryos from HP-hMG transfers showed higher implantation and live birth rates in selected analyses.
Oocyte yield vs embryo yield: rFSH often yielded more oocytes, but quality and implantation potential were not necessarily superior; HP-hMG’s LH activity may enhance oocyte/embryo quality and endometrial receptivity, contributing to higher ongoing pregnancy and live birth in some circumstances.
Key numerical highlights and formulas ( LaTeX-formatted)
Primary endpoints and noninferiority concepts:
Ongoing pregnancy rate per cycle: HP-hMG − rFSH =
with a 95% CI lower bound above
, establishing noninferiority.Endocrine and follicular metrics (examples):
Estradiol on day of hCG: ext{E}2^{ ext{hCG, HP-hMG}} - ext{E}2^{ ext{hCG, rFSH}} = 1265 ext{ pmol/L} ext{ (95% CI: 680–1864)}
(P < 0.001).Progesterone end of stimulation: P{ ext{end}}^{ ext{HP-hMG}} e P{ ext{end}}^{ ext{rFSH}} ext{ with } P < 0.001.
OHSS incidence (safety):
HP-hMG OHSS (types 2/3) ≈ 7/344 ext{ (1.9%)} vs recombinant FSH 4/317 ext{ (1.2%)} in the EISG PP dataset.
Live birth and ongoing pregnancy (MERIT integrated data):
Live birth rate per cycle initiated (fresh cycles): HP-hMG vs rFSH ; OR ≈ 1.36 ext{ (95% CI: 1.01–1.83)}; P = 0.041.
Ongoing pregnancy rate per started cycle (fresh): HP-hMG ; OR ≈ 1.38 (95% CI around 1.03–1.86).
Cumulative live birth rate per patient (fresh+frozen within 1 year): HP-hMG ≈ vs rFSH ≈ ; difference ≈ ; 95% CI around difference broad, includes 0.
Embryo-quality-related metrics (MERIT):
Top-quality embryos per oocytes retrieved (local): HP-hMG 11.3 ext{%} vs rFSH 9.0 ext{%}; P = 0.044.
Central scoring showed a similar directional trend but with non-significant difference.
Conclusions and practical implications
Across major RCTs and integrated analyses, HP-hMG is at least as effective as recombinant FSH for COS in IVF/ICSI under long GnRH agonist protocols, with certain studies showing a trend toward higher ongoing pregnancy and live birth rates when HP-hMG is used, particularly in IVF cycles with LH activity from the start of stimulation.
HP-hMG generally demonstrates a more favorable or comparable safety profile, notably with lower OHSS incidence in several trials.
In antagonist-cycle trials with single-blastocyst transfer (MEGASET), HP-hMG maintained noninferiority to rFSH for ongoing pregnancy and showed lower OHSS and pregnancy loss, with cumulative live birth rates that were similar between arms.
In high-responder populations (MEGASET-HR), HP-hMG demonstrated noninferiority for ongoing pregnancy, with safety advantages (lower OHSS) and potential implications for personalized stimulation strategies based on biomarkers like AMH.
Overall, HP-hMG provides an attractive alternative to rFSH, offering a biologically balanced LH/FSH activity profile, potential endometrial and embryonic quality advantages, and favorable safety considerations, particularly in IVF cycles employing LH-active gonadotropins from cycle start.
Connections to broader ART principles and ethical/practical considerations
Personalization of COS: AMH-based identification of high responders was explored (MEGASET-HR) to tailor stimulation intensity and reduce OHSS risk, illustrating a move toward individualized ART regimens.
Embryo transfer strategy: several trials emphasize single-embryo transfer or blastocyst-stage transfers to optimize live birth rates while minimizing multiple gestations; results interact with direct gonadotropin choice by influencing embryo quality and endometrial receptivity.
Economic and health implications: noninferiority plus safety advantages can translate into cost-savings and reduced maternal-fetal risk; meta-analyses and subsequent economic evaluations have supported HP-hMG as a cost-effective alternative in many practice settings.
Ethical considerations: the balance between maximizing live birth rates and minimizing OHSS and pregnancy-related complications remains central; trials increasingly probe not just pregnancy rates but cumulative outcomes and neonatal safety across fresh and frozen cycles.
Takeaway for exam-ready understanding
HP-hMG (FSH + LH activity) often matches or modestly exceeds rFSH in ongoing pregnancy and live birth outcomes, especially in IVF cycles with long down-regulation or antagonist protocols where LH activity may influence follicular development, oocyte/embryo quality, and endometrial receptivity.
Across multiple large RCTs and integrated analyses, HP-hMG tends to reduce OHSS risk and may improve certain embryo-quality metrics, though findings can vary by study design and endpoint definitions.
The landscape supports a role for LH-active gonadotropins in specific IVF contexts, with HP-hMG offering a practical, cost-effective option that can align with patient-specific risk profiles and transfer policies.