HMI-115: The Prolactin Antibody Whose Hair Trial Went Unreported
Published by AmpleLab Research
In January 2024, Hope Medicine, a Nanjing-based biotech, announced that its antibody HMI-115 had increased hair counts in a small group of men with androgenetic alopecia, and said it expected full proof-of-concept by the end of the year. The study meant to deliver that proof, 192 men with a placebo arm, is registered as completed. Almost two years after the company's own deadline, to our knowledge, nobody outside the company has seen a number from it.
That leaves HMI-115 in an odd spot. It blocks the prolactin receptor, a target no approved hair loss treatment uses. It has a licence from Bayer, a plausible biological rationale, a small positive signal, and a finished trial with no public results. It is one of two prolactin-receptor antibodies in human trials for hair loss, alongside ABS-201, and this article keeps three things apart: what the biology supports, what the company has claimed, and what is still missing.
The Hypothesis: Prolactin as a Brake on Hair Growth
Prolactin is best known as the hormone that drives lactation, which makes it an odd thing to find in a hair follicle. But follicles make it and respond to it locally, and the case for HMI-115 rests on that finding, largely from one academic group (Foitzik, Paus and colleagues), first in mice and then in human follicles.
In mouse skin, prolactin and its receptor were found in hair follicle epithelium, with expression that shifted across the hair cycle. Adding prolactin (400 ng/mL) to cultured mouse skin accelerated follicle regression (catagen) and reduced the number of proliferating, Ki-67-positive cells in the hair bulb.
Human scalp hair follicles were found to express prolactin and its receptor at both the mRNA and protein level, with prolactin and receptor staining appearing up-regulated as follicles moved from growth toward regression. In organ culture, high-dose prolactin significantly inhibited hair shaft elongation and drove premature catagen. The authors proposed that the follicle's own prolactin acts as a local regulator of regression, rather than depending on prolactin arriving from the bloodstream.
These are mouse and organ-culture findings, and the human work used a high prolactin concentration. They show that prolactin can push follicles toward regression in a dish. They don't show that prolactin signalling drives androgenetic alopecia in people, or that blocking it reverses the condition, and that gap is the whole question. The same group later published a paper titled "Mind the (Gender) Gap", asking directly whether prolactin has gender- and site-specific effects on the human hair follicle¹, a sign the biology is still being worked out.
The Antibody Whose Main Job Is Endometriosis
HMI-115 is a human monoclonal antibody against the prolactin receptor, and Hope Medicine did not invent it. By the company's own account, it licensed the antibody from Bayer in April 2019:
The company describes HMI-115 as a first-in-class monoclonal antibody targeting the prolactin receptor, and states that in April 2019 it "entered a world-wide exclusive license agreement with Bayer AG" for its development and commercialisation. The same release covers US clearance of a planned international Phase 2 in androgenetic alopecia, and notes that a separate endometriosis application had already been cleared in 2021.
That second indication matters for reading the hair programme, because endometriosis-associated pain is now the company's most advanced clinical indication. It has a placebo-controlled Phase 2 behind it and a Phase 3 in China listed as recruiting since early 2026, with 540 women targeted². BioInvent, the Swedish biotech that says it discovered the antibody, announced in March 2026 that the Phase 3 had been initiated³.
A randomised, double-blind, placebo-controlled, 12-week Phase 2 enrolled 142 women with moderate to severe endometriosis across the US, Poland and China. In an interim analysis of 102 patients, the company reported statistically significant reductions versus placebo in dysmenorrhea pain (42% at 240 mg every two weeks) and non-menstrual pelvic pain (50%). It reported no treatment-related serious adverse events, stable estradiol, LH, FSH and progesterone levels, and no significant change in bone mineral density.
This is the closest thing to a safety dataset for the antibody in a larger placebo-controlled group, and the 240 mg every-two-weeks regimen reported there is also the regimen registered for the hair loss Phase 1b below. It is also company-reported interim data, in women, over 12 weeks. What it can't tell us is what happens when men take the antibody for years for a non-life-threatening condition, where the safety bar is usually higher than for a pain indication.
The Phase 1b: One Number, No Placebo
To our knowledge, the entire public human efficacy case for HMI-115 is one sentence in a January 2024 press release, describing a small study in Australia.
An open-label study of 16 people with androgenetic alopecia (12 men, 4 women) over 24 weeks, led by Professor Rodney Sinclair in Australia. In the 12 men, the company reported that mean non-vellus target area hair count (TAHC) rose by 14 hairs/cm² from baseline, described as statistically significant, and that the treatment was safe and well tolerated. The release gives no p-value, baseline count or measure of spread, and results for the four women were not disclosed.
A single-arm, open-label study at one Melbourne site, with participants receiving HMI-115 by subcutaneous injection at 240 mg every two weeks. The primary endpoint is the change in non-vellus TAHC from baseline through week 24, with secondary hair count assessments listed out to week 36. The record shows the study as completed in September 2023 and has no results posted.
Read plainly, this is a 12-man, single-arm, open-label signal with no placebo comparison. Open-label hair counts are more exposed to expectation and measurement effects than blinded ones, and the figure was announced by the sponsor and has not, to our knowledge, been peer reviewed. Fourteen hairs per square centimetre sounds like a result. Without a baseline or a control arm, it is a number looking for context, and the Phase 2 below was designed to supply it. The registry's week-36 timepoint would show what happens after dosing stops, and nothing about it has been reported.
The Phase 2 Nobody Has Seen
The study that was supposed to supply that context is registered as NCT06118866.
A randomised, multicentre, double-blind, placebo-controlled, 24-week study in men aged 18 to 65 with Norwood-Hamilton III to V androgenetic alopecia, with Peking University People's Hospital in Beijing as the listed site. The April 2024 version of the record planned 180 participants across four arms: placebo, and HMI-115 at 120 mg every four weeks, 240 mg every four weeks, or 240 mg every two weeks, all by subcutaneous injection. The primary endpoint is the mean change in non-vellus TAHC from baseline at week 24. Later versions of the record, as mirrored by third-party registry sites, list 192 participants and a completed status, last updated in June 2025. No results are posted.
Then, silence. In January 2024 the company said it aimed to reach full proof-of-concept by the end of 2024. As of early October 2026, we have found no Phase 2 efficacy result in the registry, in a company release, or in a conference abstract. There are ordinary explanations for a long gap: data held back for partnering or regulatory discussions, resources going to the endometriosis programme, or results that did not warrant an announcement. Public information cannot distinguish between them, so the silence should not be read as failure, and it should not be read as success either.
What it does mean is that every public human hair-efficacy claim about HMI-115 in circulation today, however confidently it is repeated, traces back to the 12-man open-label study above.
Why Is There a New Trial Already?
In mid-2026, a new efficacy trial of HMI-115 in male androgenetic alopecia appeared in China's trial registry. Opening a second efficacy trial before the first has reported publicly is, at minimum, worth noticing.
A 24-week, randomised, multicentre, double-blind, placebo-controlled study in male androgenetic alopecia, registered as Phase IIb/III (the aggregator listing states that only the IIb portion is currently being conducted). The listing shows a target of 160 participants in China, first disclosure on 26 May 2026, first participant enrolled on 12 June 2026, Peking University People's Hospital as principal investigator site, and Hope Medicine's Nanjing entity as sponsor.
A sponsor opening another efficacy trial tells us the hair programme is still alive. It does not tell us what the 192-man Phase 2 found. A combined IIb/III design could follow a positive Phase 2, a need to refine the dose, or a route chosen for Chinese regulators, and the registry entry doesn't say which. Until a readout appears, treat the new trial as a sign of life, not as indirect evidence that the drug works.
HMI-115 vs ABS-201: A Race Run Mostly on Paper
Both HMI-115 and ABS-201 block the prolactin receptor, so the two programmes will tend to be read together. At the most recent update we found (August 2026), Absci had reported safety and pharmacokinetic data from healthy volunteers and no efficacy data in people with androgenetic alopecia, with interim proof-of-concept findings expected in the second half of 2026.
That leaves HMI-115 as the only antibody in the class with any human hair count figure, and that figure is the 12-man open-label result above. Neither programme has placebo-controlled hair data in the public domain. The one head-to-head comparison in circulation comes from Absci itself, and it concerns pharmacokinetics, not hair:
In a single-dose, 56-day non-human primate pharmacokinetic study, Absci reports a more than threefold longer half-life for ABS-201 than for HMI-115, and uses it to project dosing every 8 to 12 weeks in humans, against an assumed every-2-to-4-week interval for HMI-115. This is a company-run animal comparison and a modelled projection, not a human head-to-head, and it has not been independently confirmed.
A longer half-life, if it holds in people, would mean fewer injections. It says nothing about which antibody grows more hair, and no public data currently allows that comparison.
What We Know, What We Don't, and What We're Waiting For
The biology supports a hypothesis, and a reasonable one. Hair follicles express prolactin and its receptor, and high-dose prolactin pushes mouse and human follicles in culture toward regression. HMI-115 has been given to humans in several studies, and the company's interim endometriosis data report no treatment-related serious adverse events and stable sex hormones in women over 12 weeks.
What has not been shown in public is whether HMI-115 beats placebo for hair count in androgenetic alopecia, whether the effect depends on dose, whether any gain lasts after dosing stops, how women respond, and whether men taking it long term stay safe. A 2010 paper from the original research group explicitly asked whether prolactin's effects differ by sex or scalp site. The Phase 2 readout, or the results of the new IIb/III trial, is the event that would turn this from a mechanism with a small signal into something that can be judged. Until then, HMI-115 is a candidate worth tracking, not a result to rely on.
Frequently Asked Questions
What is HMI-115?
HMI-115 is an investigational monoclonal antibody that blocks the prolactin receptor. It is developed by Hope Medicine, which holds a worldwide licence from Bayer, and it is being studied for androgenetic alopecia and for endometriosis-associated pain.
Has HMI-115 been shown to regrow hair?
Not in a result that has been made public. The company reported a mean increase of 14 hairs/cm² in 12 men in an open-label Phase 1b, without a placebo group or a published baseline. A placebo-controlled Phase 2 in 192 men is registered as completed, and to our knowledge its results have not been released.
How is it different from finasteride or minoxidil?
Finasteride and dutasteride lower DHT, and minoxidil acts through a separate, non-hormonal mechanism. HMI-115 targets prolactin signalling at the receptor instead, and in the trials registered so far it is given by subcutaneous injection.
Is HMI-115 approved or available?
No. It is investigational and, to our knowledge, not approved by any regulator. It has only been given in clinical trials. Any product sold online under its name would be unregulated, and its identity and purity could not be assumed.
When might it become available for hair loss?
No approval filing for hair loss has been announced that we know of, and a Phase IIb/III began enrolling in China in June 2026. Compounds at this stage typically need further trials and regulatory review, which generally takes years and often ends in failure.
Selected Research
This article is provided for educational purposes and does not constitute medical or investment advice. HMI-115 is an investigational compound not approved for any use by the FDA, MHRA, NMPA, or any other regulatory body, and its safety and efficacy have not been established. AmpleLab products are unrelated to HMI-115 and are not intended to diagnose, treat, cure, or prevent any condition.
AmpleLab.