hMG (Human Menopausal Gonadotrophin)

A mixture, not a molecule

There is no structure to record here, and the registries say so outright. The FDA Global Substance Registration System files menotropins as a mixture under UNII 5Y9QQM372Q, with two components declared must-be-present, follitropin and luteinizing hormone, each held as a pointer to its own substance record rather than as structure. What the record does carry is cross-references: CAS 61489-71-2, MeSH descriptor D008596, WHO ATC code G03GA02, ECHA 232-662-3 and four ChEMBL identifiers. Molecular formula and molecular weight are simply absent; a variably glycosylated mixture of glycoprotein heterodimers has neither. PubChem, queried by name for menotropins and again for human menopausal gonadotropin, returned PUGREST.NotFound both times, so the CID field on this page is blank.

Potency is therefore not expressed in micrograms. The United States prescribing information states that FSH and LH activity assays are standardised using the Fourth International Standard for Urinary FSH and Urinary LH, November 2000, of the World Health Organization Expert Committee on Biological Standardization, and that each vial contains 75 International Units of FSH activity and 75 International Units of LH activity. What is being declared is the response a preparation produces in a reference assay. Two vials agreeing on the declared figure need not agree on protein content, and the label makes no claim that they do.

Three analytical papers describe what else is in the vial. Giudice and colleagues ran SDS-PAGE and Western blotting on six commercial urinary preparations in 1994 and identified tumour necrosis factor binding protein-I, transferrin and immunoglobulin-related proteins in all but one, with urokinase, Tamm-Horsfall glycoprotein and epidermal growth factor in some. Van de Weijer and colleagues reported in 2003 that reversed-phase HPLC on a highly purified preparation put non-gonadotrophin material at a minimum of 30 per cent by peak area, and named leukocyte elastase inhibitor, protein C inhibitor and zinc-alpha-2-glycoprotein among the impurities. Wolfenson and colleagues reported in 2005 that batch-to-batch consistency of urinary preparations matched that of recombinant FSH on isoform profiling, immunoassay, size-exclusion chromatography and an in vitro follicle bioassay. The three groups were based, respectively, at a Serono institute, at Organon and at Instituto Massone; the first two companies marketed competing recombinant or highly purified products, the third manufactures urinary gonadotrophins. Reading the purity literature without reading the affiliations gives a misleading impression of consensus in either direction.

Where the LH activity comes from

Section 2 of the United Kingdom Summary of Product Characteristics for the 75 IU presentation says it plainly: each vial corresponds to 75 IU of FSH bioactivity and 75 IU of LH bioactivity, and hCG obtained from the urine of pregnant women may be added to attain the required total LH bioactivity. The United States label for the same product states only that human chorionic gonadotropin is detected in the product, and does not say that it was added or that it contributes to the declared LH figure. Two labels for one product, differing in how much of the manufacturing they disclose.

Van de Weijer and colleagues measured the ratio behind that sentence. Immunoactivity for hCG in the preparation ran three-fold higher than immunoactivity for LH, and because hCG has the longer circulating half-life, the authors calculated that approximately 95 per cent of in vivo LH-receptor-mediated bioactivity was attributable to hCG. The 95 per cent is a calculation from the measured immunoactivity ratio and the two half-lives, and the paper presents it as such. Their stated inference was that this quantity of hCG could only be explained by addition from an external source. The United Kingdom label, written by the manufacturer rather than by a competitor, says the same thing in one sentence.

Filicori and colleagues had reached a compatible conclusion from the clinical side a year earlier. One hundred and twenty GnRH agonist-suppressed women were randomised across four intramuscular regimens combining hMG with highly purified FSH so that LH activity varied while FSH was held constant. Occurrence of small pre-ovulatory follicles correlated inversely with the administered LH activity dose and with serum hCG concentration, but not with serum LH. The authors concluded that hCG content contributes considerably to hMG activity. Any statement that hMG supplies LH is describing a bioassay figure printed on a label.

What the approvals cover, and where

The currently marketed United States product is licensed as biologics licence application 021663, sponsored by Ferring and approved in its original form on 29 October 2004; the label header records an initial United States approval for the active ingredient of 1975, and Drugs@FDA carries an original approval dated 22 August 1975 for the earliest menotropin application, BLA 017646. Its single indication is development of multiple follicles and pregnancy in ovulatory women as part of an assisted reproductive technology cycle. Administration is subcutaneous after reconstitution. Nothing in that label addresses men, ovulation induction outside assisted reproduction, or any use in people not undergoing fertility treatment.

United Kingdom labelling for the same product is broader. Section 4.1 of the Summary of Product Characteristics lists anovulation including polycystic ovarian disease in women unresponsive to clomifene citrate, controlled ovarian hyperstimulation in assisted reproduction, and hypogonadotrophic hypogonadism in men, the last in combination with chorionic gonadotrophin for stimulation of spermatogenesis. The same preparation, the same manufacturer, and a male indication that is on-label in one jurisdiction and absent from the other. Patients with primary testicular failure are described in that label as usually unresponsive.

Three earlier United States menotropin products are recorded in Drugs@FDA as discontinued: BLA 017646 (Serono), BLA 020328 (Organon) and BLA 021047 (Ferring). Their submission histories survive in the database; their labels largely predate its electronic document archive, which matters for a specific claim addressed below. ClinicalTrials.gov returns 169 registered studies naming menotropins as an intervention and 178 naming human menopausal gonadotropin. Restricting the menotropins query to male participants returns eight, one of which is a randomised trial discussed below.

The comparison with recombinant FSH, and how it moved

MERIT is the trial the United States label rests on. Andersen, Devroey and Arce randomised 731 women undergoing IVF after a long GnRH agonist protocol to highly purified menotrophin (n = 363) or recombinant FSH (n = 368), assessor-blind and multinational. In that trial, mean oocytes retrieved were 11.8 in the recombinant FSH arm against 10.0 in the menotrophin arm (P < 0.001), and 11.3 per cent against 9.0 per cent of retrieved oocytes developed into top-quality embryos on day three (P = 0.044) (PMID 16873892). Ongoing pregnancy per cycle was 27 per cent against 22 per cent, odds ratio 1.25 with a confidence interval of 0.89 to 1.75. The authors stated that superiority could not be concluded and that non-inferiority was established. Trew published a critical comment in the same journal the following year, with an author reply.

Two later trials repeated the comparison under different transfer rules. MEGASET (NCT00884221) randomised 749 women in a GnRH antagonist cycle with compulsory single-blastocyst transfer across twenty-five centres in seven countries; ongoing pregnancy after a fresh cycle was 30 per cent against 27 per cent per protocol, and cumulative live birth for a single stimulation cycle, counting frozen transfers within one year, was 40 per cent against 38 per cent. MEGASET-HR (NCT02554279) randomised 620 women with anti-Mullerian hormone of 5 ng/mL or above; ongoing pregnancy per cycle start after fresh transfer was 35.5 per cent against 30.7 per cent, and cumulative live birth was 50.6 per cent against 51.5 per cent. All three trials were designed as non-inferiority studies, and all three met that design.

Pooled analyses have not held still. Coomarasamy and colleagues reported in 2008 that seven randomised trials covering 2159 women showed higher live birth with hMG than with recombinant FSH, relative risk 1.18 (1.02 to 1.38), a pooled risk difference of 4 per cent. The Cochrane review three years later pooled recombinant FSH against any of three urinary gonadotrophins together and found no evidence of a difference in live birth, odds ratio 0.97 (0.87 to 1.08) across 28 trials and 7339 couples within a review of 42 trials and 9606 couples, concluding that further research was unlikely to identify substantive differences. The 2026 update reports the hMG stratum on its own: across 15 studies and 4793 participants, live birth is probably lower with recombinant FSH than with hMG or highly purified hMG, odds ratio 0.83 (0.73 to 0.95), at moderate certainty, while the same update finds no difference for recombinant FSH against highly purified urinary FSH, odds ratio 1.02 (0.86 to 1.21). The 2026 review added 25 new studies and applied a trustworthiness checklist that excluded some earlier ones, so the change reflects new data, a changed inclusion rule and a narrower comparator.

Sponsorship runs consistently in one direction on each side of this question. The three randomised trials above were run by author groups including Ferring employees, Ferring being the manufacturer of the menotropin product. Of the two 2024 letters published on MEGASET-HR, one is from authors employed by Merck KGaA, which markets recombinant FSH (PMID 37734649); the other is a reply from the trial's own authors, several of them at Ferring (PMID 37995797). The 2026 Cochrane update states that it had no dedicated funding and that its two Cochrane staff authors were excluded from its editorial processing.

Adverse events in the record

Label figures come from a pooled safety population. In two single-cycle, open-label, multinational comparative trials, 434 normal ovulatory infertile women received subcutaneous menotropins as part of an IVF cycle. Adverse reactions at 2 per cent or above were abdominal pain in 29 (6.7 per cent), headache in 27 (6.2 per cent), ovarian hyperstimulation syndrome in 27 (6.2 per cent), injection site pain or reaction in 17 (3.9 per cent), abdominal cramps in 13 (3.0 per cent), enlarged abdomen in 10 (2.3 per cent) and injection site inflammation in 10 (2.3 per cent). Thrombophlebitis was reported in under 1 per cent. The label warns that the preparation contains gonadotropic substances capable of causing ovarian hyperstimulation syndrome with or without pulmonary or vascular complications.

Ovarian hyperstimulation syndrome is where the two preparations separate most clearly. MEGASET-HR recorded it in 9.7 per cent of the menotrophin arm against 21.4 per cent of the recombinant FSH arm, a difference of 11.7 percentage points with a confidence interval of 6.1 to 17.3, in a population selected for high ovarian reserve. Cumulative early pregnancy loss ran 14.5 per cent against 25.5 per cent. The 2026 Cochrane review reaches the same direction on far more data: across 37 studies and 9813 participants, hyperstimulation syndrome is probably higher with recombinant FSH than with hMG, odds ratio 1.42 (1.12 to 1.80). Both comparisons describe relative rates between two stimulation drugs in supervised assisted-reproduction cycles, and neither describes an absolute risk outside one.

The male literature is mostly not randomised

Alexander and colleagues screened 3925 abstracts and identified 103 studies covering 5328 male patients from 21 countries treated with gonadotrophins for pubertal induction in hypogonadotrophic hypogonadism. Human menopausal gonadotrophin appeared in 42 of those studies (40.8 per cent), behind chorionic gonadotrophin at 93 (90.3 per cent) and ahead of FSH at 37 (35.9 per cent). Median reported treatment or follow-up was 18 months. Spermatogenesis rates were 86 per cent (82 to 91) with hCG plus FSH against 40 per cent (25 to 56) with hCG alone. The authors recorded high heterogeneity and treatment variability and called for randomised studies; the field has not published them.

Older controlled work is small. Liu and colleagues at the NICHD followed 16 men with complete isolated hypogonadotrophic hypogonadism from the inception of therapy for two years. Eleven men received a combined gonadotrophin regimen by intramuscular injection and five received pulsatile subcutaneous GnRH; the doses and schedules are in the source. After 24 months, 8 of 11 in the gonadotrophin group and 2 of 5 in the GnRH group had produced sperm, a difference the authors reported as not significant, and the abstract separately states 80 per cent for the gonadotrophin group, a figure that does not follow from 8 of 11. Sperm concentrations in every man in both groups were below 5 million per millilitre, and mean testicular volume reached 4.3-fold and 4.8-fold the pre-treatment value respectively.

Retrospective cohorts run larger and later. Mao and colleagues compared 182 men treated with combined hCG and hMG against 20 treated with pulsatile GnRH; median time to first appearance of sperm was 18 months (16.4 to 20.0) in the gonadotrophin group against 6 months (1.6 to 10.4) in the GnRH group, P < 0.001, with the gonadotrophin group reaching higher serum testosterone. Mao and colleagues state that prospective randomised studies would be needed to confirm the finding. Liu 1988 is a small prospective non-randomised comparison of two treatment groups run at the Developmental Endocrinology Branch of the NICHD, and its abstract carries no comparable statement. One registered randomised trial addressing the menotropin question in men was returned by the ClinicalTrials.gov query described above: NCT03687606, a phase 4 parallel-group study of chorionic gonadotrophin alone against chorionic gonadotrophin with hMG in men with isolated hypogonadotrophic hypogonadism, 210 participants estimated, overall status UNKNOWN, no results posted. No published head-to-head randomised comparison was retrieved.

Three documents, three half-lives

Elimination half-life for the FSH component is reported differently by three sources describing the same class of preparation. The United States prescribing information gives 11 to 13 hours in the multiple-dose phase, from a study numbered 2000-03, and describes the subcutaneous and intramuscular values as similar to one another. The United Kingdom Summary of Product Characteristics gives 30 plus or minus 11 hours subcutaneous and 27 plus or minus 9 hours intramuscular after repeated dosing in downregulated volunteers. Di Stefano and colleagues, pooling six pharmacokinetic studies in 121 healthy pituitary-suppressed women, reported a range of 39.02 to 53.63 hours across studies for menotrophin and urofollitrophin preparations produced by a different manufacturer.

Peak concentrations across those documents are of similar magnitude — 8.5 mIU/mL in the United States single-dose data, 8.9 plus or minus 3.5 IU/L in the United Kingdom repeated-dose data, 4.98 to 7.50 IU/L across the Di Stefano studies — which makes the half-life spread harder to attribute to assay scale alone. Route does not account for it either: the United States label reports near-identical values for its two routes, and the two United Kingdom figures differ by three hours. The experiments differ in dose, in single against repeated administration, in whether values were baseline-corrected, in the sampling window and in the product analysed, and a terminal half-life is sensitive to all of those. Any secondary page quoting one figure for hMG without naming the document and the dosing regimen behind it has discarded the information needed to judge it.

What is not known

Almost the entire published record describes supervised fertility treatment. The randomised evidence base is a comparison between two stimulation drugs inside assisted-reproduction cycles, and it does not describe what any gonadotrophin preparation does in people not undergoing that treatment. Composition is characterised in only a handful of analytical papers, most written by employees of companies marketing a competing preparation, and none published in the last two decades; whether current manufacturing matches what those papers measured is not established anywhere retrievable. The quantity of hCG added to reach the declared LH bioactivity is disclosed as a practice in the United Kingdom label but is not quantified in any label, so the actual hCG content of a given vial is not a published figure. Half-life is reported at 11 to 13, 27 to 30, and 39 to 54 hours by three different documents, and no study reconciling them was located. In men, the register holds one randomised comparison of chorionic gonadotrophin alone against chorionic gonadotrophin with hMG, NCT03687606, a phase 4 parallel-group trial with 210 participants estimated, whose overall status is UNKNOWN with no results posted; no published head-to-head randomised comparison was retrieved. The male literature is otherwise dominated by retrospective cohorts and single-arm series, and the largest synthesis of it explicitly calls for randomised studies. No published data address use in men with androgen-induced hypogonadism, and none address use outside a fertility indication in either sex. Menotropins are a prescription biological product; material sold outside that supply chain has not been subject to the identity, bioassay-potency or sterility controls that define the labelled article, and the international unit on a label means nothing without the assay behind it.

Questions