LL-37 is the human body's own cathelicidin, a 37-residue antimicrobial peptide made by skin, gut and white blood cells. As a research compound it is sold in 5 mg and 10 mg vials and described on protocol pages as an injectable immune peptide. The human record says something different: every completed efficacy trial put LL-37 on a wound as a solution or cream, and the only injections were into melanoma tumours in three patients. This page sets out those doses as the trials state them (abstracts and registry records read 2026-10-12), converts the circulating figures into syringe units, and compares the vial concentrations with the concentrations the trials used. It reports; it suggests no dose.
The arithmetic for any vial size is on the LL-37 calculator. The other immune-signalling peptide on this site with a human record is covered on thymosin alpha-1 dosage, and the anti-inflammatory tripeptide often grouped with it on KPV dosage.
What the human trials gave
| Trial (registry, PMID) | Who | Route | Dose | Result |
|---|---|---|---|---|
| Grönberg 2014 (PMID 25041740) | 34 with venous leg ulcers | Topical, twice weekly, 4 weeks | 0.5, 1.6, 3.2 mg/mL vs placebo | Healing rate ~6× placebo at 0.5, ~3× at 1.6, none at 3.2 |
| HEAL LL-37, Mahlapuu 2021 (PMID 34687253) | 148 with venous leg ulcers | Topical with compression | 0.5, 1.6 mg/mL vs placebo | No significant overall benefit; post hoc benefit in wounds ≥10 cm² |
| Miranda 2023 (NCT04098562, PMID 37480520) | Diabetic foot ulcers, mild infection | Cream, twice weekly, 4 weeks | 0.5 mg/mL vs placebo | Greater granulation index days 7–28; cytokines and bacteria not different |
| NCT02225366, MD Anderson | 3 with cutaneous melanoma metastases | Into the tumour, weekly, up to 8 weeks | 250 or 500 mcg per tumour | Results posted; no serious adverse events, one squamous cell carcinoma |
Two features of this table matter for anyone reading a dose figure. The trials report concentrations, not amounts: a wound received a solution at so many milligrams per millilitre, and the amount reaching the body depended on the wound's size. And the dose-response ran backwards — the weakest solution did best and the strongest did nothing measurable — which is a reason the follow-up trial dropped 3.2 mg/mL.
The melanoma injections
The one injected human study started at 250 mcg per tumour, injected into skin or subcutaneous metastases at least 1 cm across, every 7 days for up to 8 weeks, with a second cohort at 500 mcg. Its posted results list two patients at 250 mcg (one stopped for lack of efficacy) and one at 500 mcg; no deaths and no serious adverse events were recorded, but one patient at 250 mcg had a squamous cell carcinoma and one had actinic keratosis among the other adverse events. A pathology report from the same trial describes a 63-year-old woman who, about 45 days after starting eight weekly injections, developed multiple verrucous papules on her trunk and limbs, 11 of 12 biopsies showing an atypical squamous proliferation (Dolkar 2018).
These are doses into a tumour, chosen to provoke a local immune reaction. They are not a body-wide dose and do not convert into one.
The figures on protocol pages
| Source (read 2026-10-12) | What it states |
|---|---|
| Peptide Dosing Protocols, LL-37 (last reviewed August 2026) | 100–250 mcg a day under the skin, some write-ups up to 500 mcg; once daily, often 5 days on and 2 off, for 2–4 weeks; "This has not been tested in human trials" |
| Peptide Protocol Wiki, LL-37 dosing (updated 2026-02-01) | 100–200 mcg daily, 5 days a week, 2–4 weeks on and 2 off; "no established human SC dosing from clinical trials" |
Both pages say plainly that the subcutaneous figures are community conventions. By molecule count, 100 mcg of LL-37 is about 22 nmol and 250 mcg about 56 nmol — LL-37 is a large peptide, roughly three and a half times the weight of a ten-residue peptide such as kisspeptin-10, so the same microgram figure carries far fewer molecules.
The figures in units
Vials are 5 mg and 10 mg. At the calculator's water volumes:
| Vial and water | Concentration | mcg per unit | 100 mcg | 200 mcg | 250 mcg | 500 mcg |
|---|---|---|---|---|---|---|
| 5 mg in 1 mL | 5,000 mcg/mL | 50 | 2 units | 4 units | 5 units | 10 units |
| 5 mg in 2 mL | 2,500 mcg/mL | 25 | 4 units | 8 units | 10 units | 20 units |
| 5 mg in 3 mL | 1,667 mcg/mL | 16.7 | 6 units | 12 units | 15 units | 30 units |
| 10 mg in 2 mL | 5,000 mcg/mL | 50 | 2 units | 4 units | 5 units | 10 units |
| 10 mg in 3 mL | 3,333 mcg/mL | 33.3 | 3 units | 6 units | 7.5 units | 15 units |
A 5 mg vial holds fifty 100 mcg amounts or twenty-five 200 mcg amounts whatever the water volume. On the protocol pages' five-days-a-week pattern, that is ten weeks or five weeks of the stated figures — longer than the 2–4 week blocks the same pages describe, so most of a vial would sit reconstituted between blocks. How small unit marks read on each barrel is on insulin syringe units.
Vial concentration against trial concentration
The ulcer trials' most effective solution was 0.5 mg/mL (500 mcg/mL). Every standard reconstitution of a research vial is stronger: 2,500 mcg/mL for a 5 mg vial in 2 mL is five times stronger, and 5,000 mcg/mL ten times. Reaching 0.5 mg/mL would need 10 mL of water in a 5 mg vial and 20 mL in a 10 mg vial. This is arithmetic about solutions, not a statement that a trial concentration for a wound applies to an injection; the two routes are not comparable.
Regulator status
| LL-37 | |
|---|---|
| Approved product anywhere | None (openFDA, 2026-10-12: no label) |
| FDA safety-risk list (content current 2026-04-22) | Listed: immunogenicity; "detrimental effects on male reproduction"; "can be protumorigenic in some tissues" |
| ClinicalTrials.gov | Intervention trials are the topical and intratumoral studies above; most other records measure the body's own LL-37 |
| WADA 2026 | Not named |
PubMed held 3,020 records with LL-37 or LL37 in the title or abstract on 2026-10-12. Most are laboratory work on the body's own peptide in infection, skin disease and cancer — including papers describing it as both anti-tumour and tumour-promoting depending on tissue — which is the literature FDA's entry summarises.
What is not established
No trial has given LL-37 under the skin or into a vein for a body-wide effect, so no subcutaneous dose has a human basis. The topical trials disagree with each other at their own doses, and the highest concentration did nothing measurable. The only injected doses were into tumours in three people. Product and dose decisions belong with a clinician. Medibact's LL-37 guide covers the compound.
Sources and dates
Opened 2026-10-12: PubMed abstracts PMIDs 25041740 (Grönberg 2014), 34687253 (Mahlapuu 2021), 37480520 (Miranda 2023) and 29665030 (Dolkar 2018); PubMed record count for LL-37/LL37; ClinicalTrials.gov v2 API records NCT02225366 (arms, participant flow, adverse events) and NCT04098562 (arms), and the intervention search for LL-37; PubChem CID 16198951; openFDA drug-label searches for LL-37, cathelicidin and ropocamptide (no records); FDA, "Certain bulk drug substances for use in compounding that may present significant safety risks" (content current as of 2026-04-22); the 2026 WADA Prohibited List as published in Tractatenblad 2026 no. 25; Peptide Dosing Protocols' LL-37 page (last reviewed August 2026) and Peptide Protocol Wiki's LL-37 dosing page (updated 2026-02-01). Molar, unit and per-vial arithmetic by this site. Corrections go to the contact page.