hCG (Human Chorionic Gonadotropin)

One hormone, two manufacturing routes, and a rat bioassay

Two databases hold this hormone at different resolutions. The FDA Global Substance Registration System files chorionic gonadotropin as a single protein substance: UNII 20ED16GHEB, CAS 9002-61-3, two chains of 145 and 92 amino acids, eleven disulfide bonds, four N-linked and four O-linked glycosylation sites. UniProt resolves the chains separately. The alpha subunit is residues 25 to 116 of a 116-residue precursor, and it is the same alpha subunit carried by luteinising hormone, follicle-stimulating hormone and thyroid-stimulating hormone. The beta subunit is residues 21 to 165 of its own precursor. Only that beta chain distinguishes this hormone from the other three, which is why assays are written against it rather than the intact dimer.

Molecular weight is where the registries diverge. GSRS stores a calculated mass of 25,700 Da and the formula C1105H1764N318O337S26 for the recombinant substance, counting the polypeptide and nothing else, while its chorionic gonadotropin record carries a slightly different polypeptide formula, C1105H1769O335N318S26, together with a molecular weight property of 70,000 Da measured by SDS-PAGE. Cole's 2010 review in Reproductive Biology and Endocrinology gives the molecular weight of the hormone as approximately 36,000 and states that 25 to 30 per cent of that mass is sugar side-chains, rising to 35 to 41 per cent in the hyperglycosylated variant, which reaches 40,000 to 41,000. The three figures are not reconcilable as measurements of one object: a calculated polypeptide mass, a glycosylated mass, and an electrophoretic estimate whose basis GSRS does not state. A PubChem query by name returns a small hydrocarbon rather than the heterodimer, a name-matching artefact, and the CID field on this page is blank for that reason.

Two manufacturing routes are in use and they are not the same article. The urinary product is described in its own label as a highly purified pyrogen-free preparation obtained from the urine of pregnant females. The recombinant product, choriogonadotropin alfa, is secreted by Chinese hamster ovary cells. Neither is dosed by mass the way a small molecule is. The recombinant label states that biological activity is determined by the seminal vesicle weight gain test in male rats set out in the European Pharmacopoeia monograph, calibrated against the third international reference preparation IS 75/587. Potency is therefore expressed in international units defined by that bioassay rather than by mass of protein.

What the approval actually covers

Chorionic gonadotropin for injection has held United States approval since 1974 and is one of the few compounds on this site that is an approved medicine. The label lists three indications and no others: prepubertal cryptorchidism not due to anatomic obstruction, selected cases of hypogonadotropic hypogonadism in males, and induction of ovulation and pregnancy in the anovulatory infertile woman already pretreated with menotropins. Administration is intramuscular after reconstitution. Precocious puberty, prostatic carcinoma or other androgen-dependent neoplasm, and prior allergic reaction to the hormone are listed as contraindications. Material sold outside that supply chain is not the approved article, and nothing in the label extends to it.

Pharmacokinetic figures come from the recombinant product, whose label reports them; the urinary label reports none. After a single 250 microgram subcutaneous injection in healthy female volunteers, peak concentration was 121 plus or minus 44 IU/L at a median of 24 hours, area under the curve 7701 plus or minus 2101 hour IU/L, elimination half-life 29 plus or minus 6 hours, and absolute bioavailability 0.4 plus or minus 0.1. Given intravenously the serum profile fitted a two-compartment model with an initial half-life of 4.5 plus or minus 0.5 hours and a terminal half-life of 26.5 plus or minus 2.5 hours. The label does not state how many volunteers produced those means.

Efficacy in the assisted-reproduction indication rests on a randomised open-label study, numbered 7927 in the label, run at twenty United States centres in women undergoing in vitro fertilisation and embryo transfer. The label describes three arms: 250 micrograms and 500 micrograms of the recombinant product administered subcutaneously against 10,000 USP Units of an approved urinary-derived product administered intramuscularly. Of 297 patients entered, 94 were randomised to the 250 microgram arm. The primary parameter was the number of oocytes retrieved, which the label reports as similar between arms, with both recombinant arms found clinically and statistically equivalent to the comparator and to each other. Safety figures come as counts: alanine aminotransferase elevations up to 1.2 times the upper limit of normal in 10 of 335 patients on 250 micrograms (3 per cent), 9 of 89 on 500 micrograms (10 per cent), and 16 of 328 on urinary-derived hCG (4.8 per cent).

The obesity literature, in the order it happened

Simeons published a two-page communication in the Lancet in November 1954 proposing that chorionic gonadotropin with severe calorie restriction removed fat selectively. PubMed indexes it as a journal article and serves no abstract. The scheme that grew from it divides body fat into reserve, structural and abnormal compartments, and holds that the hormone releases the third while sparing the second. That partition is a claim about physiology.

Asher and Harper reported in 1973 in the American Journal of Clinical Nutrition a benefit for the hormone on weight loss and hunger, against the same 500-calorie diet with placebo. PubMed types the paper as a controlled clinical trial and serves no abstract, so neither the finding nor the group sizes could be read from the retrieved record; the result is described here as Stein and colleagues characterised it in 1976 when setting out to duplicate it, using the same prescribed diet. Fifty-one women aged 18 to 60 were randomised in a 32-day double-blind comparison; 20 of 25 on hormone and 21 of 26 on placebo completed 28 injections. No statistically significant difference appeared in weight loss, percentage weight loss, hip or waist circumference, weight loss per injection, or hunger ratings.

Greenway and Bray ran a double-blind randomised comparison reported in 1977 in the Western Journal of Medicine, and described weight loss as identical between groups, with no differential effect on hunger, mood or localised body measurements. One indexing oddity belongs in the record, because indexing decides what filtered searches return: PubMed types that paper as Case Reports, which does not match the design its own abstract describes. Bosch and colleagues in 1990 gave 40 women with a body mass index above 30 the same 5,000 kJ daily diet with saline or hormone injections for six weeks, and reported no advantage on any variable recorded.

Lijesen and colleagues then scored the field. Their 1995 criteria-based meta-analysis in the British Journal of Clinical Pharmacology traced eight controlled and sixteen uncontrolled trials and scored each on study population, intervention, measurement of effect, and data presentation. Methodological scores ran from 16 to 73 out of 100. Of the twelve studies scoring 50 or more, all of them controlled, one reported the hormone a useful adjunct. The paper's own counts do not reconcile — it reports tracing eight controlled trials, then reports that the twelve studies scoring 50 or more were all controlled. The discrepancy is unexplained in the abstract. The authors concluded there is no scientific evidence of effect on weight loss, fat redistribution, hunger or feeling of well-being.

The approved label reaches the same position and prints it in capitals, in two separate sections. Its indications section states that the hormone has not been demonstrated to be effective adjunctive therapy in the treatment of obesity, and that there is no substantial evidence that it increases weight loss beyond that resulting from caloric restriction, that it causes a more attractive or "normal" distribution of fat, or that it decreases the hunger and discomfort associated with calorie-restricted diets. Its clinical pharmacology section states that the hormone has no known effect on fat mobilisation, appetite or sense of hunger, or body fat distribution. On 6 December 2011 the FDA and the Federal Trade Commission issued seven joint warning letters to marketers of over-the-counter products labelled as homeopathic hCG and sold for weight loss, on the grounds that no approved hCG product exists for that use and the products were unapproved drugs making unsupported claims. A ClinicalTrials.gov intervention search returns zero registered weight-loss studies; the hits filed under obesity and overweight are fertility studies in which body weight is a covariate, plus two androgen-physiology studies in polycystic ovary syndrome in which the hormone is used as a stimulation test.

What was measured in fat tissue

Two studies measured effects in adipose tissue directly. Tell, Haour and Saez reported in 1977 in Molecular and Cellular Endocrinology that highly purified preparations affected neither adenylate cyclase activity nor lipolysis in rat adipocytes, with or without GTP. The same preparations stimulated cyclic AMP and testosterone production in rat Leydig cells, so the material was biologically active, and rat ovary bound labelled hormone. The authors concluded there are no hCG-LH receptors on the fat cell membrane surface, and suggested that positive results reported elsewhere could be explained by contaminating lipolytic hormones in those laboratories' preparations.

Dos Santos and colleagues revisited the question in human tissue in 2007 in the Journal of Endocrinology and reached a partly different answer. They characterised an hCG/LH receptor in human adipose cells for the first time, and reported that physiological concentrations left both lipolysis and two lipogenesis markers unaltered in mature adipocytes, while stimulating preadipocyte growth through a MAP kinase and c-fos pathway independent of protein kinase A. Preadipocyte differentiation capacity rose moderately, and leptin secretion from adipose tissue rose. Their framing was pro-adipogenic: an explanation for why fat storage increases in the first trimester of pregnancy.

Where the two disagree is on whether the receptor is present, and that disagreement runs across species and across thirty years of assay technique. Where they agree is lipolysis, which is the point the weight-loss claim depends on: neither found the hormone released fat. The human study found an effect in the opposite direction from the one the popular scheme requires.

The Kaposi's sarcoma episode, and what the impurity did

In 1995 Lunardi-Iskandar and colleagues reported in Nature that an immortalised AIDS-Kaposi's sarcoma cell line produced malignant metastatic tumours in nude mice and was killed, apparently by apoptosis, by the beta chain of chorionic gonadotropin, while normal endothelial cells were not killed. That paper carries an erratum, which PubMed records without a PMID of its own. Two commentaries were published alongside it.

Gill and colleagues took it into people. Their 1996 report in the New England Journal of Medicine screened four commercial preparations in vitro, selected the most active, and tested it in 36 patients with AIDS-related Kaposi's sarcoma. Twenty-four received intralesional injections at 250, 500, 1000 or 2000 IU, six patients per dose, with two lesions injected and one given diluent. Twelve more were randomised double-blind to 2000 IU or diluent, six each. Responses ran 1, 5, 5 and 10 of 12 injected lesions across the ascending doses; in the blinded portion no diluent lesion responded against 10 of 12. That paper carries two errata, printed in the same journal in February and April 1997, whose content could not be retrieved this session.

Kachra and colleagues then separated the activity from the hormone. Writing in Endocrinology in 1997, they confirmed that commercial preparations, which they describe as about 25 per cent pure, inhibited Kaposi's sarcoma cell growth dose-dependently, and reported that four different well-characterised preparations of pure dimer, and each of its two isolated subunits, did not alter cellular proliferation at all. Subfractionation by molecular size put the anti-tumour activity in the low molecular weight components and the receptor binding and steroidogenic bioactivity elsewhere. Their conclusion was that hormone activity and anti-Kaposi's activity are separable properties of the same vial.

Malatos and colleagues characterised the contaminants in 2005 in Molecular and Cellular Proteomics. Size-exclusion chromatography isolated a protein complex from urinary preparations, and MALDI time-of-flight mass spectrometry of that complex showed three polypeptides of about 6.2, 11.4 and 15.8 kDa. Peptide mass mapping and N-terminal sequencing identified the polypeptides of about 11.4 and 15.8 kDa as metabolites of placental transforming growth factor beta and of bikunin respectively, alongside the third polypeptide of about 6.2 kDa that was not identified. Of these, only the transforming growth factor beta metabolite was reported in all three pharmaceutical preparations examined, in which no other non-hCG-related protein species were observed. A 1999 letter in Blood had reported the opposite sign of effect, describing paradoxical pro-Kaposi's sarcoma activity of such preparations. No trial in Kaposi's sarcoma is registered on ClinicalTrials.gov. What survived is Tell's 1977 suggestion, reached independently in another tissue: what these preparations do can belong to what else is in them.

Testis, testosterone, and the anti-doping entry

Coviello and colleagues reported the only human study in this area to sample intratesticular testosterone directly, in 2005 in the Journal of Clinical Endocrinology and Metabolism. Twenty-nine men with normal reproductive physiology were randomised to weekly testosterone enanthate with either saline placebo or 125, 250 or 500 IU hCG every other day for three weeks, and intratesticular testosterone was sampled by percutaneous fine needle aspiration at baseline and at the end. Baseline serum testosterone at 14.1 nmol/l was 1.2 per cent of intratesticular testosterone at 1174 nmol/l. In the placebo arm intratesticular testosterone fell 94 per cent, from 1234 to 72 nmol/l. It rose linearly with dose: 25 per cent below baseline at the lowest arm, 7 per cent below at the middle, 26 per cent above at the highest.

Downstream of that measurement the retrievable evidence is retrospective. Hsieh and colleagues reviewed 26 hypogonadal men on testosterone with co-administered hormone, in 2013 in the Journal of Urology, and reported no change in semen parameters over more than a year of follow-up and no patient becoming azoospermic. Rainer and colleagues reviewed 28 men with previous exogenous testosterone use switched to monotherapy, in Cureus in 2022, and reported no significant change in mean hormone levels across the cohort. That paper's own numbers do not reconcile: a stated cohort of 28 carries subgroup counts of n equal to 30 and 31, unexplained in the text.

Randomised evidence for testicular volume comes from a different population. Sahib and colleagues randomised 51 men with hypogonadotropic hypogonadism, in Cureus in 2023, to hormone alone, hormone plus human menopausal gonadotropin, or hormone followed by combination after six months, and reported significant increases in mean testicular volume in all three arms with no clinically significant difference between them. A September 2025 correction to that paper amends its ethics statement only. In cryptorchidism, Pyörälä and colleagues pooled 33 articles covering 3282 boys with 4524 undescended testes, 872 boys with 1174 testes of them in randomised trials, and reported a success rate of 19 per cent against 4 per cent with placebo, calling the data scanty.

The anti-doping position is explicit. Section S2.2.1 of the 2026 WADA Prohibited List is headed testosterone-stimulating peptides in males and lists chorionic gonadotrophin first, alongside luteinising hormone, gonadotrophin-releasing hormone and its agonist analogues, and kisspeptin and its agonist analogues. The restriction to males is written into the heading, not appended as a note. On detection, Strahm and colleagues gave three successive injections to ten volunteers in a pilot clinical study and followed blood and urine for two weeks, reporting pharmacokinetics prone to large inter-individual variation and proposing testosterone and the testosterone to luteinising hormone ratio, in both blood and urine, as more relevant markers than direct urinary quantification (PMID 21798680).

What is not known

No trial has established what this hormone does to body composition in people, because no trial has measured it. A PubMed search pairing the compound with lean body mass and weight loss returns nothing, and ClinicalTrials.gov holds no registered weight-loss study, so the controlled trials in the ledger report weight, circumferences and hunger ratings only. The published record also does not establish whether co-administration during testosterone therapy preserves testicular volume, whether it preserves fertility beyond the 26-man retrospective series, or what happens beyond the three weeks Coviello measured; the longest prospective intratesticular measurement remains three weeks. Nothing establishes the composition or content of material sold outside the pharmaceutical supply chain: the only purity characterisations in the literature describe pharmaceutical preparations, and two independent groups working twenty-eight years apart concluded that biological effects attributed to these preparations belonged to contaminants rather than to the hormone. The identity record itself is partly open. The registries disagree on molecular weight because they are describing different objects and using different methods, the two GSRS records carry slightly different polypeptide formulae, glycosylation varies with the source tissue and the manufacturing route, and no glycoform specification exists for anything but the approved products.

Questions