MK-677: The Same Molecule Is in a Phase 3 Trial and on the Grey Market

Almost every entry here ends with a compound nobody is developing. This one does not.

MK-677 was never abandoned. Merck's MK-0677 and Lumos Pharma's LUM-201 are the same molecule, and a 2022 paper in Hormone Research in Paediatrics opens by saying so: "LUM-201 (ibutamoren, formerly MK-0677) is an orally administered GH secretagogue receptor agonist under development for treatment of pediatric growth hormone deficiency (PGHD)." (PMID 35354138)

It is in Phase 3 as I write. NCT06948214, sponsored by Lumos Pharma, status recruiting, actual start 20 May 2026. 150 naive to treatment prepubertal children, 12 months, randomised, double blind, placebo controlled, oral LUM-201 at 1.6 mg per kilogram per day. Primary completion December 2027.

At the same moment the same compound is an unapproved drug sold as a research chemical, and FDA's own Pharmacy Compounding Advisory Committee voted on 29 October 2024 against adding ibutamoren mesylate to the 503A bulks list.

The molecule did not change. Its paperwork did.

What it is

A small molecule, not a peptide, orally active, designated L-163,191 by Merck. In rat pituitary cells it released growth hormone with an EC50 of 1.3 nM, and it worked orally in dogs at doses as low as 0.125 mg/kg. (PMID 7624358) It is an agonist at the ghrelin receptor GHSR1a, and cryo-EM structures of the human receptor bound to ibutamoren exist. (PMID 34737341) It is not a SARM and not an anabolic steroid, despite being sold beside both.

The drug arrived before the hormone. Its receptor was identified in 1996. (PMID 8688086, PMID 8838145) Ghrelin was not characterised for another three years. The imitation was in hand before anyone had met the original.

The trial running right now

The pediatric programme is not a footnote. Around the Phase 3 sit four more Lumos Pharma studies and a Phase 3 extension due to start in February 2027. (NCT04614337, NCT04806854, NCT05250063, NCT05796440, NCT07129759) After what turned up for BPC-157, I checked every record here at the registry itself: all sponsors real, nothing mock or fictional.

The pediatric results are also the sharpest argument against how it is marketed. In a six month trial of 68 prepubertal children with growth hormone deficiency, annualised height velocity was 4.5 cm per year on placebo, 6.0 cm per year on LUM-201 at 0.4 mg/kg/day and 6.9 cm per year at 0.8 mg/kg/day. Injected growth hormone produced 11.1 cm per year. (PMID 33982679)

That is "oral growth hormone" tested and answered, in the population it was designed for, by the people developing it. Their own summary: "Interest in GH secretagogues as noninjectable alternatives to daily recombinant human growth hormone (rhGH) was not supported by treatment results in early pediatric trials."

Where 25 milligrams comes from

In 1996, Chapman and colleagues ran a randomised, double blind, placebo controlled dose ranging study in 32 healthy adults aged 64 to 81. Oral MK-677 at 2, 10 or 25 mg once daily, over periods of 14 and 28 days. At 25 mg for two weeks, mean 24 hour growth hormone rose about 97 percent, and serum IGF-I went from 141 µg/L to 265 µg/L by four weeks. Fasting glucose rose significantly. (PMID 8954023)

That is where 25 mg comes from. Not a safety ceiling, not a titration, not anything failing above it. It was the largest of three numbers written on a protocol before the study began, and it produced the biggest response, which is what the largest of three arms usually does. It then became the dose in almost every adult trial since. One programme escalated 30 healthy elderly subjects to 50 mg daily for two weeks. (PMID 10404019) Nothing higher is published.

The community convention is 10 to 25 mg once daily, most often at night. No primary source for it exists beyond vendor pages, retailers and forums, which cite each other rather than data.

1

2

3

4

5

1

10

mg per day · log scale

derived from researchcommunity convention

1. 1Lowest arm of the 1996 dose ranging study · 2 mg per dayChapman tested 2, 10 and 25 mg in 32 adults aged 64 to 81. The bottom of the tested range, and nobody uses it.
2. 2The one documented real world average · 6.7 mg per dayA performance user analysed in a 2026 hair testing paper consumed 60 capsules of 10 mg over 90 days, 600 mg in total. The paper does not record how those doses were spread out, so this is an average and not a daily dose.
3. 3The community convention · 10 to 25 mg per dayWhat vendor pages, retailers and forums converge on. No primary source for it exists. They cite each other.
4. 4Top arm of the 1996 study · 25 mg per dayThe number everything else rests on. It won by being the largest of three arms in one 1996 study, then became the standard dose in almost every adult trial afterwards.
5. 5Highest dose in the published adult literature · 50 mg per dayOne bone turnover programme escalated 30 healthy elderly subjects to 50 mg daily for two weeks. Nothing higher appears in the published adult literature.

The children get a rule. The adults get a number. A per kilogram dose is an instruction that survives meeting somebody it has never met. A flat 25 mg is an artefact of the last group it happened to be measured in, and that group was 32 pensioners in 1996.

Same shape as the BPC-157 dose problem, from the opposite direction. There the number has no derivation and appears to come from vial size. Here it has an impeccable one, in a population with nothing in common with the people now swallowing it.

What the adult trials found

The two year trial. 65 healthy adults aged 60 to 81, double blind, randomised, placebo controlled, 25 mg daily. The longest published human exposure. Fat free mass changed by minus 0.5 kg on placebo and plus 1.1 kg on MK-677, body weight by 0.8 kg against 2.7 kg, limb fat by 0.24 kg against 1.1 kg. Visceral and total fat did not differ. (PMID 18981485)

Then the sentence that matters most in this whole file, from the authors: "Increased FFM did not result in changes in strength or function."

Fat free mass is a subtraction, not a capability. It is everything on the scan that is not fat, and water counts. The same trial reported transient lower extremity oedema. A suitcase that weighs more coming home than going out has definitely got something extra in it. Nothing guarantees the extra thing is anything you can wear.

Hip fracture, twice. Bach 2004, 161 patients aged 65 and over, six months of daily MK-0677 or placebo. IGF-I rose 84 percent against 17 percent, with no significant difference in functional performance or in Sickness Impact Profile score. (PMID 15066065) Adunsky 2011, 123 elderly hip fracture patients, 25 mg daily for a planned 24 weeks. IGF-1 rose 51.4 ng/mL and gait speed improved by a 0.7 score difference, but stair climbing power did not, and neither did several other measures. The trial stopped early for a congestive heart failure signal. (PMID 21067829)

Alzheimer's disease. 563 patients, 416 completing, 25 mg or placebo daily for 12 months. IGF-1 rose 60.1 percent at six weeks and 72.9 percent at 12 months, with no significant differences on any cognitive, functional or dementia measure. (PMID 19015485) The largest trial ever run on this compound, and flatly negative.

Everything else. 24 obese men on 25 mg for eight weeks: IGF-I up about 40 percent, fat free mass up significantly. (PMID 9467542) Eight volunteers held at 18 kcal/kg/day: a week of 25 mg moved mean daily nitrogen balance from minus 1.48 g/day to plus 0.31 g/day, and peak growth hormone fell from 55.9 µg/L after one dose to 22.6 µg/L after a week. (PMID 9467534) 22 haemodialysis patients: IGF-1 up 65 percent over placebo. (PMID 28340044)

Look at the pattern, not any single result. Forty percent, 51 ng/mL, 60 percent, 65 percent, 72.9 percent, 84 percent. IGF-1 is the one endpoint this compound has never missed, across six populations and three decades. Then each trial measured the thing it actually cared about and found nothing. It did not fail to reach its target. It reached it every time, and reaching it turned out not to be the constraint.

The adverse effects

Glucose, and why a fasting panel misses it. The 1996 study found fasting glucose rose significantly. In the two year trial fasting blood glucose rose an average of 0.3 mmol/L, about 5 mg/dL, and insulin sensitivity declined. Cortisol rose 47 nmol/L, about 1.7 µg/dL, against a seven day study in young men where cortisol was unchanged, pointing to an adrenal effect only seen with chronic dosing. (PMID 18981485) Two of that trial's own investigators later called its metabolic data a mild rise in insulin resistance at one year, with a rise in HbA1c they judged unlikely to be clinically significant. (DOI 10.1093/gerona/glad022) They ran the trial.

The obese men trial is the one to read carefully. Fasting glucose and insulin were unchanged after eight weeks. An oral glucose tolerance test showed impaired glucose homeostasis at both two and eight weeks. (PMID 9467542) The fasting labs missed what the challenge test caught. A clean fasting panel here is reassurance from a test already shown to under-report the effect.

Oedema and appetite. In the two year trial the most frequent side effects were increased appetite, which subsided within a few months, and transient mild lower extremity oedema and muscle pain. (PMID 18981485)

Read that list twice. The trial files increased appetite, swollen ankles and muscle pain under side effects. A forum log files the same three under week two going well. Nothing in the body differs between those readings. Only the column heading does.

Heart failure. The Adunsky trial was terminated early for a congestive heart failure signal, and its authors concluded the safety profile was unfavourable. That much is published. The breakdown that circulates everywhere, four cases on drug against one on placebo, is not. The abstract says only "a limited number of patients", the full text is paywalled, and FDA's briefing document that would carry the counts returns a 404. Those counts are reported from industry summaries rather than the primary record, so I am not publishing them as fact. And the other hip fracture trial excluded patients with diabetes, cancer, uncontrolled hypertension or existing congestive heart failure, so the reassuring language here comes from screened people.

Case reports. Transaminitis in a man in his early thirties after two months on MK-677, resolving after he stopped, no dose recorded. (PMID 40675653) Bilateral gynaecomastia and hypogonadotropic hypogonadism in a 40 year old after six months on products containing RAD-140, MK-677 and cardarine, where analysis found undisclosed testosterone, estradiol and growth hormone in the products themselves. (PMID 39145153) That is about adulteration as much as MK-677, which is why product testing matters: of 44 products sold online as SARMs, only 18 held the labelled amount of the labelled compound, and 17 held a different unapproved drug, ibutamoren among the contaminants named. (PMID 29183075) A 2026 splenic rupture report exists, but its own authors call the link speculative. It is not an established risk. (PMID 42064485)

What nobody knows

The pharmacokinetics in people. Searching for these returns an assay paper and animal work. The roughly 24 hour half life and the 60 to 70 percent bioavailability figure on every vendor page have no primary human source I could locate, and the pages contradict each other. Treat any half life number you are quoted as unsourced.

Why Merck stopped. No published statement gives a reason. Anyone telling you it was killed by the heart failure signal is inferring a chain no primary source states. Separately, a completed 24 week Phase 2 in 64 women with fibromyalgia has never reported. (NCT00116129)

Whether it does anything for sleep. The whole human polysomnography record is one 1997 study in 14 subjects, whose older cohort of six had no placebo arm. (PMID 9349662) Morning against night dosing has never been compared.

Regulatory status

FDA. Not approved for human use, which makes it an unapproved drug, and not legal as an ingredient in dietary supplements or any consumer product. It is on the Department of Defense Prohibited Dietary Supplement Ingredients List. Products carry "Supplement Facts", "Research Facts" or "For Research Use Only" labels.

On 29 October 2024 the Pharmacy Compounding Advisory Committee reviewed ibutamoren mesylate for the 503A bulks list alongside L-theanine, ipamorelin and kisspeptin-10, and voted against inclusion for all four. No tally is retrievable: FDA's briefing document and meeting minutes both return 404, and the outcome is reported from an industry association summary rather than FDA's own meeting materials. The written recommendation is on record. The agency concluded that "the physicochemical characterization, limited information on historical use, lack of evidence of effectiveness, and the specific safety concerns identified for ibutamoren mesylate weigh against inclusion of this substance on the 503A Bulks List", and committee members raised fluid retention, congestive heart failure and hyperglycemia.

WADA. Prohibited at all times, in and out of competition, under section S2, added by name for the 2024 List alongside capromorelin. WADA's own site would not render for me, so this comes from USADA's athlete advisory rather than WADA's document, and I give the section as S2 only. It shows up in hair: one 10 mg oral dose produced 1.3 pg/mg in the segment from 0 to 1 cm four weeks later. (PMID 40882886)

One naming trap. Anamorelin, capromorelin, ipamorelin, macimorelin and hexarelin are distinct compounds in the same class. Their data is not MK-677 data, and CJC-1295 acts on a different receptor.

My take

The interesting thing about MK-677 is not that the evidence is thin. It is that the evidence is unusually thick, and it says no.

Hundreds of people in randomised placebo controlled trials, a solved receptor structure, a real sponsor, an active Phase 3. On published evidence it outranks most of what is sold beside it. And in adults, every trial that measured something a person would notice came back blank. Strength and function did not move in two years. Cognition did not move in Alzheimer's. Recovery did not move after hip fracture, and that trial was stopped.

The dose sits on top of that. The children in the Phase 3 get 1.6 mg per kilogram per day, a rule that produces a different answer for every child in the trial. Adults get 25 mg, inherited from the top arm of a 1996 study and carried forward because it was already there. Thirty years of research, and the number in circulation still describes a room of pensioners.

If you have run MK-677 and had bloods done, I want to know one thing. Did anyone give you a glucose tolerance test, or only a fasting panel? The obese men trial is why that distinction matters, and I have never once seen it mentioned in a log.

---

This content is for educational and informational purposes only and is not medical advice. Peptides discussed are research compounds and may not be approved for human use. Nothing here should be used to diagnose, treat, cure, or prevent any disease. Always consult a qualified healthcare professional before starting any peptide, supplement, or protocol. Individual responses vary. Do not self-administer compounds without proper medical supervision.