KY-19382
guinea pig
The Dose Guy’s verdictOct 27, 20263 sources
One of the more exciting ideas in hair loss, and nowhere near proven. KY-19382 flips a growth switch in the follicle that nothing on the proven shelf touches, and it regrew hair in mice and lengthened human follicles in a dish. That’s the whole record: no human trial, no human safety data, no tested scalp dose. If you run it, run it on top of finasteride and minoxidil, never instead of them.
What it is
A small synthetic molecule out of a Korean university lab, built from indirubin, a compound from an old herbal remedy. Your follicles grow when a signal called Wnt/beta-catenin is on, and a protein called CXXC5 acts as the brake on it. KY-19382 jams the brake by blocking CXXC5 from grabbing its partner, Dishevelled, so the growth signal stays on. That’s a different lever from the one finasteride pulls by cutting DHT and the one minoxidil pulls through blood flow and growth-phase timing, which is why it makes sense as an add-on rather than a swap.
The same lab built PTD-DBM, a peptide that hits the same CXXC5 target. KY-19382 is the small-molecule version: easier to dissolve into a scalp solution, cheaper to make, more stable on the shelf. Among the other Wnt-pathway bets, SM04554 went further, into phase 2 trials in actual people. KY-19382 never left the lab, and anything you buy is a research chemical.
What the research shows
Every result here comes from cells, cultured follicles, and mice. The core paper is good preclinical work. It’s still preclinical work.
Realistic protocols
There’s no human dose to hand you, so I won’t fake one. What people actually do is buy the powder, dissolve it into an alcohol and propylene glycol base, and apply a milliliter or so to the thinning area once a day, at concentrations that vary from one forum thread to the next. Nobody has translated the mouse dose to a human scalp, so any percentage you see is a guess dressed up as a recipe. If you run it anyway: keep finasteride and minoxidil going as the base, apply KY-19382 at a separate time so the solutions are not fighting each other on your scalp, start low, and patch-test a small area for a week first. Keep it off broken skin and off any mole or spot that is changing. Take photos in the same light every month and give it four to six months before you judge it. If you see nothing past what your proven stack was already doing, stop, because a growth switch with zero human safety data is not something to run on hope.
Side effects
Zero human safety data, so everything below is user reports or open questions:
