Reviewed & approved by Dr. Nova Reyes · Last checked 2026-10-15. ← Back to Evidence Library

LL-37: what the evidence says

Legend: HUMAN data · ANIMAL / LAB data · UNKNOWN · UNVERIFIED: not yet confirmed against a primary source

What "No human data" means here: for immune, infection, gut, healing or anti-aging uses of injected LL-37, there are animal and lab studies only, and no human trials.

---

What is it?

LL-37 kills some bacteria, fungi and viruses in lab dishes, and it also signals to immune cells and helps skin cells move and grow. Vitamin D switches on the gene that makes it, although whether this reliably raises active LL-37 is less clear. That link is one reason many studies measure LL-37 levels rather than give it as a drug. A higher or lower LL-37 level in blood or skin is a lab marker, not a health benefit by itself. Too much LL-37 in skin is linked in research to rosacea and psoriasis, so "more" is not simply "better". ^{R1, R14, R15}

Does it work?

Human Hard-to-heal venous leg ulcers (cream, with compression): - Grönberg 2014: the first trial in people, 34 patients in Sweden. After a placebo run-in, people got one of three LL-37 strengths or placebo, double-blind (strengths and schedule in the study table). Ulcers healed faster at the lowest strength (p = 0.003) and possibly at the middle strength (p = 0.088, which could have been chance). The highest strength did no better than placebo. Run by Pergamum AB, the company developing it; the first author worked there. ^R2 - Mahlapuu 2021 (HEAL LL-37): a phase 2 trial of 148 patients across Poland and Sweden, two strengths vs placebo (schedule in the study table). Across everyone, healing was no better than placebo: about a quarter of ulcers closed in each group. A later look at people with large ulcers (10 cm² or more) found some improvement. That subgroup was chosen after the results were in, so it can only suggest a new trial, not show a benefit. Funded by Promore Pharma; three authors were Promore Pharma employees and one a shareholder. ^R3

Human Diabetic foot ulcers with mild infection (cream): Miranda 2023, a double-blind placebo-controlled trial in Jakarta (2020–2021) in 25 people (13 LL-37, 12 placebo); nobody dropped out. The LL-37 group had a greater rise in healthy healing tissue at every check during the study (schedule in the study table). Inflammation markers and bacteria counts were not better than placebo. Funded by a Universitas Indonesia research grant; the authors declared no conflicts. ^R4

Human Melanoma: MD Anderson Cancer Center ran a phase 1/2 trial of LL-37 injected into melanoma skin tumors, with the US National Cancer Institute as collaborator. It was completed on Nov 24, 2020. Posted results on ClinicalTrials.gov show 4 enrolled and 3 treated in two dose groups, no deaths or serious adverse events, and 1 person stopping for lack of effect. The posted tumor-response table lists all 3 under "antitumor immune response" (complete or partial response), which conflicts with one person stopping for lack of effect; no response details are given, so we don't treat it as evidence of benefit. Other adverse events posted include a squamous cell skin cancer, a precancerous skin patch (actinic keratosis) and anemia. ^R5 One patient from this program is described in a case report (below). ^R6

Animal Animals and lab: infection, wound, gut and immune models. One lab study found LL-37 might help melanoma cells invade nearby tissue. Not reviewed in detail for this page. ^{R1, R12}

Who paid / conflicts: the leg-ulcer trials were run by the developers, Pergamum AB (2014) and Promore Pharma AB (2021, which funded it). The diabetic foot trial was funded by Universitas Indonesia, and the melanoma trial was run by MD Anderson with the National Cancer Institute. ^R2–R5

Is it safe?

• Human In the cream trials: Grönberg 2014 found "no safety concerns" for local or whole-body adverse events (problems reported after use, which may or may not have been caused by the drug). Mahlapuu 2021 called both strengths "well tolerated and safe"; serious adverse events were judged unrelated to the drug. In the diabetic foot trial, one person on LL-37 had mild skin irritation from wound fluid. ^{R2, R3, R4}
• Human Injected into tumors: a 63-year-old woman with advanced melanoma got a course of injections into her tumors (study table). About 45 days in, she developed many wart-like and blistering skin lesions that looked like skin cancer but on biopsy were an inflammatory reaction with abnormal skin cells. All cleared within 2 months of stopping. ^R6
• Caution FDA's concerns: FDA states that compounded LL-37 "may pose risk for immunogenicity" (the body making antibodies against it) and may contain peptide-related impurities. It "lacks sufficient safety-related information" and notes "nonclinical research findings suggest detrimental effects on male reproduction and that this drug can be protumorigenic in some tissues." ^R7
• Dropouts vs placebo: in HEAL LL-37, 2 people on the higher strength stopped treatment because of adverse events (a fractured femur, a deep vein clot) and 3 more on that strength left the study after serious events following their last dose, none judged drug-related. Nobody dropped out of the diabetic foot trial. In the melanoma trial, 1 of 3 stopped for lack of effect. Caution [UNVERIFIED: dropouts by group in Grönberg 2014 (no free full text)] ^R3–R5
• Unknown Nothing is known about injecting it under the skin, about long-term use, or about use in healthy people.

Is it legal?

• FDA: not approved. Drugs@FDA (searched via openFDA Oct 15, 2026 for "LL-37," "cathelicidin" and "ropocamptide") has no matching product. ^R9
• EU: no authorised medicine and no orphan designation for LL-37, cathelicidin or ropocamptide in EMA's published medicines and orphan-designation data (checked Oct 15, 2026). ^R10
• Compounding (503A/503B): "Cathelicidin LL-37" is on FDA's "significant safety risks" page under the heading for substances that were in Category 2 but whose nominations were withdrawn (content current Apr 22, 2026). It is in no category of the 503A nominations list (updated May 14, 2026) and not on the 503B nominations list (updated Mar 21, 2025). With no list entry and no approved drug, pharmacies have no approved basis to compound it. ^{R7, R8, R11}
• Athletes: WADA's 2026 list does not name LL-37. Unknown The Open Peptide Project's reading (not a WADA ruling): because it has no approval from any health authority, it falls under S0, "non-approved substances," banned at all times. Athletes should ask their anti-doping organization. ^R13

What's in the vial?

Gray-market "LL-37" vials have no quality control and no guarantee of identity. It could contain the wrong compound, the wrong amount, impurities or bacterial toxins, or be fake (counterfeit). The trials above tested a cream on wounds or injections into tumors, so their results don't transfer to it. We test retail vials blind under the Open Peptide Standard. See results at /tests/. A passing test covers one batch, tested on one date. It does not mean the compound itself is safe.

What do users report?

Self-report, not evidence.

We haven't summarized user reports for this peptide yet.

Study details

^R2–R6 Human Human studies

Study	People	Route	Dose used (study arm, not a recommendation)	Follow-up	Reported	Side effects / dropouts	Limitations	Citation
Grönberg 2014 (first in people)	34 with venous leg ulcers, Sweden	Topical, 4 wk after 3-wk placebo run-in	LL-37 0.5, 1.6 or 3.2 mg/mL vs placebo, applied twice weekly for 4 weeks	4 wk after treatment	Healing rate ~6× placebo at 0.5 mg/mL (p=0.003), ~3× at 1.6 mg/mL (p=0.088); 3.2 mg/mL = placebo	No local or systemic safety concerns; dropouts by group UNVERIFIED	Tiny groups; company-run	PMID 25041740
Mahlapuu 2021 (HEAL LL-37, phase 2b)	148 with venous leg ulcers, Poland and Sweden	Topical + compression	LL-37 0.5 or 1.6 mg/mL vs placebo, applied twice weekly for 13 weeks after a 3-week placebo run-in	4 months after treatment	Closed ulcers ~26.5% / 24.7% vs 25.3% placebo (no difference); post hoc benefit in large ulcers	"Well tolerated and safe"; serious adverse events not drug-related; on 1.6 mg/mL, 2 stopped treatment for adverse events and 3 left after serious events	Subgroup chosen after the fact; company-run	PMID 34687253
Miranda 2023	25 with diabetic foot ulcers and mild infection (13 vs 12), Jakarta	Topical cream	LL-37 0.5 mg/mL cream (per registry) vs placebo cream, applied twice a week for 4 weeks	28 d	Granulation index better on days 7–28 (p=0.006–0.037); IL-1α, TNF-α, bacteria not better	No drug-caused side effects; 1 mild irritant dermatitis on LL-37; no dropouts	Small; single country	PMID 37480520; NCT04098562
MD Anderson phase 1/2	Melanoma skin metastases; 3 treated per posted results (registry enrollment 4)	Into tumors	250 µg per tumor (cohort 1, 2 people) or 500 µg per tumor (cohort 2, 1 person), weekly for up to 8 weeks	Up to 8 wk	Posted "antitumor immune response" count lists all 3, conflicting with 1 stopping for lack of effect; no response details	No deaths or serious adverse events; all 3 had other adverse events (incl. 1 squamous cell skin cancer, 1 actinic keratosis, anemia in 2); 1 of 2 in cohort 1 stopped for lack of effect	3 people; no control	NCT02225366 (posted results)
Dolkar 2018 (case report)	1 woman, 63, stage IIIC melanoma	Into tumors	8 weekly injections; amount per injection UNVERIFIED	~2 mo after stopping	Injected lesions shrank	Wart-like, keratoacanthoma-like skin lesions from ~day 45; cleared within 2 months of stopping	One person	PMID 29665030

^R7–R13 Regulatory sources (as of Oct 15, 2026)

• ^R7 FDA, Certain Bulk Drug Substances for Use in Compounding that May Present Significant Safety Risks (content current Apr 22, 2026; "nominated but withdrawn" section): https://www.fda.gov/drugs/human-drug-compounding/certain-bulk-drug-substances-use-compounding-may-present-significant-safety-risks
• ^R8 FDA 503A nominations list (May 14, 2026): https://www.fda.gov/media/94155/download
• ^R9 Drugs@FDA: https://www.accessdata.fda.gov/scripts/cder/daf/
• ^R10 EMA medicines and orphan designations data (JSON reports, downloaded Oct 15, 2026): https://www.ema.europa.eu/en/documents/report/medicines-output-medicines_json-report_en.json ; https://www.ema.europa.eu/en/documents/report/medicines-output-orphan_designations-json-report_en.json
• ^R11 FDA 503B nominations list (Mar 21, 2025): https://www.fda.gov/media/94164/download ; list page: https://www.fda.gov/drugs/human-drug-compounding/503b-bulk-drug-substances-list
• ^R13 WADA 2026 Prohibited List (S0): https://www.wada-ama.org/en/resources/world-anti-doping-code-and-international-standards/prohibited-list
• ClinicalTrials.gov: https://clinicaltrials.gov/study/NCT02225366 ; https://clinicaltrials.gov/study/NCT04098562

Full reference list (PMIDs verified on PubMed, Oct 15, 2026)

• ^R1 Background reviews on LL-37 biology: Dürr UH, Sudheendra US, Ramamoorthy A. Biochim Biophys Acta. 2006;1758(9):1408-25. doi:10.1016/j.bbamem.2006.03.030. PMID 16716248 (37-residue peptide; made in skin, gut and respiratory lining and white blood cells; antimicrobial, immune-signaling and wound-closure roles); Vandamme D, et al. Cell Immunol. 2012;280(1):22-35. doi:10.1016/j.cellimm.2012.11.009. PMID 23246832 (antibacterial, antifungal and antiviral activity; immune signaling; wound healing; possible role in cancer); Svensson D, et al. Inflamm Res. 2016;65(1):25-32. doi:10.1007/s00011-015-0884-z. PMID 26433491 (vitamin D switches on the hCAP-18/LL-37 gene).
• ^R2 Grönberg A, et al. Wound Repair Regen. 2014;22(5):613-21. doi:10.1111/wrr.12211. PMID 25041740. Authors from Pergamum AB.
• ^R3 Mahlapuu M, et al. Wound Repair Regen. 2021;29(6):938-950. doi:10.1111/wrr.12977. PMID 34687253. Funded by Promore Pharma AB. Mahlapuu, Björk and Ekblom were Promore Pharma employees; Ekblom a minority shareholder.
• ^R4 Miranda E, et al. Arch Dermatol Res. 2023;315(9):2623-2633. doi:10.1007/s00403-023-02657-8. PMID 37480520. Funded by Universitas Indonesia (PUTI Q1 grant). No conflicts declared.
• ^R5 ClinicalTrials.gov NCT02225366, "Intratumoral Injections of LL37 for Melanoma," M.D. Anderson Cancer Center; collaborator National Cancer Institute; results posted (checked Oct 15, 2026).
• ^R6 Dolkar T, et al. J Cutan Pathol. 2018;45(7):539-544. doi:10.1111/cup.13262. PMID 29665030.
• ^R12 Ohuchi K, et al. Cancers (Basel). 2023;15(6):1678. doi:10.3390/cancers15061678. PMID 36980564 (lab study: patient tissue staining, mouse and cell experiments).
• ^R14 Yamasaki K, et al. Nat Med. 2007;13(8):975-80. PMID 17676051 (cathelicidin and rosacea).
• ^R15 Lande R, et al. Nature. 2007;449(7162):564-9. PMID 17873860 (LL-37 and psoriasis).

---

Disclaimer

This page is for education and harm reduction. It is not medical advice and not a recommendation to use LL-37. Please talk to a clinician you trust, especially if you have a chronic wound, cancer, a skin condition such as rosacea or psoriasis, or are pregnant, or are a tested athlete. Seek care right away for a serious reaction, and report it to FDA MedWatch (https://www.fda.gov/safety/medwatch) or your national regulator.