LL-37 Research

LL-37 is the 37-residue C-terminal fragment of human cathelicidin hCAP18, an endogenous human peptide rather than a novel chemical entity ^[1]. LL-37 research is a large field, and most of it does not involve giving LL-37 to anyone. A PubMed search on the term LL-37 across all fields, run on 28 August 2026, returned 2,530 records ^[2]. Only four human studies have administered LL-37 itself, three topically and one intratumorally ^{[1][4][5][6][22]}. A fifth trial administered a bacterium engineered to express LL-37 rather than the peptide ^[21]. Only two of the 20 ClinicalTrials.gov records that name LL-37 in an intervention field administer the peptide ^[22]. The others measure the peptide the body already makes, or administer vitamin D ^[22].

"LL-37 levels are higher in condition X" and "giving LL-37 does X" are therefore different claims. The second does not follow from the first.

What LL-37 is

The sequence is LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES, and PubChem resolves CAS 154947-66-7 to CID 16198951, formula C205H340N60O53 and molecular weight 4493 ^[8]. The molecule also carries the assigned international nonproprietary name ropocamptide, which is a naming action rather than an approval ^[8].

Several related molecules are routinely confused with LL-37. hCAP18 is the parent protein, and most clinical assays measure hCAP18 and LL-37 together rather than free LL-37. CRAMP is the rodent ortholog, and a large share of in vivo cathelicidin work administers CRAMP instead. Truncated forms such as KR-12, GI-20 and GF-17 carry their own separate literatures. PubChem's synonym list for this CAS also wrongly includes a rabbit CAP18 designation ^[8]. FDA's substance-registry record for LL-37 carries the same wrong synonym ^[24]. Rabbit CAP18 is a different molecule.

Mechanism of action

LL-37 is a cationic, amphipathic, membrane-active peptide that permeabilizes bacterial membranes in vitro, which is the origin of the antimicrobial peptide description.

A 2007 study in human cells and human psoriatic skin showed that LL-37 complexes with extracellular self-DNA ^[9]. That complex becomes a plasmacytoid dendritic cell TLR9 stimulus and drives type I interferon production ^[9]. This is a breach of innate tolerance to self. In this setting the mechanism is a liability rather than a benefit.

Human LL-37 research

Three topical trials and one intratumoral trial make up the human administration record.

The largest of the four is a multicenter, double-blind, randomized, placebo-controlled phase IIb trial in hard-to-heal venous leg ulcers, with 148 participants ^[1]. Efficacy analysis in the full study population found no significant improvement in healing over placebo ^[1]. The trial's primary endpoint was confirmed complete wound closure, and the posted registry results quantify the null ^[3]. That endpoint was 26.5% (90% CI 17.1 to 38.7) in the lower-concentration arm, 24.7% (15.8 to 36.4) in the higher-concentration arm, and 25.3% (16.4 to 36.9) on placebo ^[3]. The intervals overlap fully, and the higher-concentration arm sits numerically below placebo ^[3]. A post hoc subgroup with larger wounds did show benefit, and the authors called for a confirmatory study that was never run ^[1]. Company communications and downstream secondary sources describe that trial as having produced positive results in patients with large wounds. That description is the post hoc subgroup, and the trial's primary analysis was null ^[1][3].

The earlier first-in-man dose-ranging trial, with 34 participants, tested three concentrations against placebo ^[4]. Only the lowest-concentration arm reached significance on the primary healing measure ^[4]. The middle arm did not, and the highest showed no difference from placebo ^[4]. That non-monotonic response is a warning sign for a membrane-active peptide. The phase IIb is a failure to replicate the one positive arm, in the same indication and by the same corporate lineage, at larger sample size ^[1][4].

A randomized double-blind trial in mildly infected diabetic foot ulcers reported a greater increase in granulation index with topical LL-37 than placebo ^[5]. It was mixed to negative on everything else that was reported. IL-1alpha and TNF-alpha rose in both arms with no between-group difference, and the reduction in aerobic bacterial colonization did not differ significantly between arms ^[5]. That reduction was in fact greater in the placebo arm by day 28 ^[5]. The registry record lists ten co-primary outcomes with no stated multiplicity correction, and it is still marked unknown status ^[5].

The only trial that has injected LL-37 into people is a phase I/II intratumoral study in melanoma cutaneous metastases ^[6]. It enrolled 4 people across five years and treated 3, one of whom discontinued for lack of efficacy ^[6]. No efficacy conclusion is supportable from a sample that size ^[6].

Safety findings in the human data

The phase IIb posted results show a serious adverse event imbalance against LL-37 ^[3]. Serious adverse events occurred in 8.33% of the lower-concentration arm and 12.24% of the higher-concentration arm, against 1.96% on placebo ^[3]. Non-serious adverse events were 31.25% and 28.57% against 19.61% ^[3]. The counts are small, no between-group comparison is posted, there were zero deaths in every arm, and investigators attributed no serious event to treatment ^[3]. This is not proof of harm.

Erysipelas, a streptococcal skin infection, appeared only in the LL-37 arms ^[3]. It was a non-serious event in 2 of 48 and 3 of 49 LL-37 participants against 0 of 51 on placebo, and a serious event in one participant per LL-37 arm against none on placebo ^[3]. Wound infection was flat across all three arms ^[3]. Numbers this small carry no statistical weight on their own. Alongside the null on bacterial colonization in the only human trial that measured it ^[5], the antimicrobial framing has no support from human bacterial-burden data.

The intratumoral trial posted adverse events for its three treated participants ^[6]. They include squamous cell carcinoma, actinic keratosis, skin hypopigmentation, hypothyroidism, anemia and decreased lymphocyte and white cell counts ^[6]. Three treated participants cannot establish a rate for any of these. The squamous cell carcinoma entry does line up with the squamoproliferative findings in the case report from the same trial, and with FDA's protumorigenic language ^[6][7][20].

A case report from inside the intratumoral trial documents dermatologic toxicity from administered LL-37 ^[7]. A woman with stage IIIC melanoma developed verrucous papules and a vesiculobullous eruption after repeated LL-37 injections ^[7]. Biopsies showed lichenoid infiltrate with eosinophils plus atypical squamous proliferation with keratoacanthoma-like features ^[7]. All lesions resolved within two months of stopping LL-37 ^[7].

LL-37 permeabilizes human sperm membranes in vitro within a physiological concentration range, causing loss of motility and premature acrosome reaction ^[10]. In mice, treated sperm lost fertilizing ability, and all 26 inseminated females failed to conceive ^[10]. A second group replicated the human sperm membrane effect in vitro ^[11]. LL-37 was under development as a candidate vaginal contraceptive, so both groups' incentive ran toward finding spermicidal activity ^[10][11].

What correlational human research shows

Most human LL-37 data are observational, and the correlation often runs opposite to a supplement narrative.

A meta-analysis of observational studies found significantly higher circulating cathelicidin in adults with active pulmonary tuberculosis than in non-TB controls ^[12]. Higher circulating levels marked the diseased group. The same pooled analysis reported low local LL-37 expression in those patients, so the two measures point in opposite directions inside one dataset ^[12]. A systematic review of host defense peptides in sepsis found cathelicidin expression associated with septic condition ^[13]. That review criticized its own source literature as largely cross-sectional, so no temporal relationship can be drawn from it ^[13].

In psoriasis, LL-37 is a T-cell autoantigen in human patients, and the circulating frequency of LL-37-specific T cells correlates with disease activity ^[14]. In rosacea, work in human tissue plus mouse intradermal challenge showed abnormally processed cathelicidin peptides driving the inflammation ^[15]. The rosacea treatment literature that followed is about blocking or suppressing cathelicidin rather than supplying it. In these two skin conditions, more LL-37 is the disease rather than the therapy.

In human melanoma immunohistochemistry, LL-37 expression correlates positively with depth of vertical invasion ^[16]. The in vivo arm of that paper administered CRAMP rather than LL-37 ^[16]. The wider cancer literature on LL-37 runs in both directions. Inhibitory effects are reported in osteosarcoma and in gastric and colon models, and promoting effects in melanoma, ovarian, lung and hepatocellular models. Those reports come from cell and animal work rather than from human administration.

Being an endogenous human peptide is not a safety argument for this molecule. FDA flags a risk of immunogenicity for compounded LL-37 by certain routes of administration, despite the sequence being one the body already makes ^[20].

Literature integrity and conflicts of interest

A 2012 paper identifying LL-37 as an agonist for the type I insulin-like growth factor receptor was retracted in 2023 ^[17]. PubMed indexes the original as a retracted publication and the notice as a retraction notice ^[17]. That retracted paper is the origin of the claim that LL-37 activates IGF-1R.

One of that paper's five authors, A. Grönberg, is also first author of the 2014 first-in-man venous leg ulcer trial ^[4][17]. That authorship overlap is not an allegation about the trial, which carries no correction or expression of concern of its own ^[4].

Both venous leg ulcer trials were run by the company developing the compound ^[1][4]. The phase IIb states that three authors were sponsor employees and one a minority shareholder ^[1]. PubMed carries no conflict-of-interest statement at all for the 2014 first-in-man paper, despite the first author's affiliation being the sponsor ^[4]. That sponsor-employed team did publish its own negative primary result rather than burying it ^[1].

Company disclosures report that the developer, Promore Pharma AB, entered voluntary liquidation in October 2023 ^[23]. The phase IIb was registered in the EU Clinical Trials Register rather than on ClinicalTrials.gov ^[3]. No phase 3 trial of LL-37 is registered on ClinicalTrials.gov ^[22].

Regulatory status

LL-37 is not an approved medicine in the United States. Field-qualified searches of FDA's drug label, approvals and product databases return no approved product and no marketing application for LL-37, ropocamptide or cathelicidin ^[24]. The only FDA record of the substance is a registry identifier, which is not an approval ^[24].

Category 2 of FDA's 503A bulk drug substances list is its category for substances that raise significant safety risks ^[18]. FDA removed cathelicidin LL-37 from that category on 22 April 2026, because the nomination was withdrawn by the nominator ^[18]. FDA states it intends to consult the Pharmacy Compounding Advisory Committee on potential inclusion before the end of February 2027 ^[18]. The current version of that list, updated 14 May 2026, places LL-37 in no category at all ^[19]. FDA's own stated position is that removal does not establish that a substance meets the 503A criteria ^[18]. Removal is not a safety clearance, and any claim that FDA cleared LL-37, or moved it to Category 1, is false.

FDA's safety-risks page still carries the LL-37 entry, under its heading for bulk drug substances nominated but withdrawn ^[20]. It states that compounded LL-37 may pose a risk for immunogenicity for certain routes of administration, and may carry complexities regarding peptide-related impurities and characterization of the active ingredient ^[20]. It states that FDA lacks sufficient safety-related information regarding cathelicidin LL-37 to know whether the drug would cause harm when administered to humans ^[20]. It adds that nonclinical findings suggest detrimental effects on male reproduction, and that the drug can be protumorigenic in some tissues ^[20]. That is a regulator's finding about the evidence base itself, and it postdates all three topical trials.

No EMA record for ropocamptide, cathelicidin or LL-37 could be found. EMA's orphan and committee records are not fully searchable from outside the agency. This is an absence of evidence rather than proof of absence.

Limitations

No human pharmacokinetic, bioavailability or half-life study of LL-37 has been published. Every human efficacy datum for LL-37 itself comes from topical application to a wound bed or intratumoral injection, both under medical supervision. There is no human evidence of any kind from systemic administration of the peptide.

A 2023 randomized trial tested an intervention named "Oral LL-37" against the Omicron BA.5.1.3 variant in 238 participants ^[21]. The intervention was a recombinant Lactococcus lactis engineered to express LL-37, meaning a live bacterial product rather than the peptide ^[21]. It was open-label and single-center, and its primary endpoint was a virological surrogate ^[21].

Material sold as LL-37 is supplied for laboratory research use only, and nothing on this page is guidance for use in people.

Frequently Asked Questions

Has LL-37 been administered to people in clinical trials? Yes, in a small number of studies. Three topical trials and one intratumoral trial make up the human administration record ^[1][4][5][6]. A PubMed search on the term LL-37 returned 2,530 records ^[2]. Only those four administered LL-37 itself ^{[1][4][5][6][22]}.

Was the largest placebo-controlled trial of LL-37 positive? Not on its primary analysis. Efficacy analysis in the full study population found no significant improvement in healing over placebo, and the confidence intervals overlap fully ^[1][3]. A post hoc subgroup with larger wounds did show benefit, and the confirmatory study the authors called for was never run ^[1].

Is "cathelicidin levels" the same measurement as LL-37? Usually not. hCAP18 is the parent protein, and most clinical assays measure hCAP18 and LL-37 together rather than free LL-37. CRAMP, the rodent ortholog, is what a large share of in vivo cathelicidin work actually administers.

Did FDA clear LL-37 when it left the 503A Category 2 list? No, it was removed on 22 April 2026 because the nomination was withdrawn by the nominator ^[18]. FDA intends to consult its advisory committee on potential inclusion before the end of February 2027 ^[18]. FDA's safety-risks page still states that it lacks sufficient safety-related information to know whether the drug would cause harm in humans ^[20].

References

1. 1

   Mahlapuu M, et al. Evaluation of LL-37 in healing of hard-to-heal venous leg ulcers: A multicentric prospective randomized placebo-controlled clinical trial. Wound Repair Regen. 2021;29(6):938-950. PMID 34687253.
2. 2

   NCBI E-utilities esearch, db=pubmed, term=LL-37 (query translation "LL-37"[All Fields]), run 2026-08-28, count 2,530. Narrower queries on the same date: "LL-37"[Title/Abstract] returned 2,521, and "cathelicidin LL-37" returned 755.
3. 3

   EU Clinical Trials Register, EudraCT 2018-000536-10, posted results section. Sponsor Promore Pharma AB. Fetched and text-extracted 2026-08-28.
4. 4

   Grönberg A, et al. Treatment with LL-37 is safe and effective in enhancing healing of hard-to-heal venous leg ulcers: a randomized, placebo-controlled clinical trial. Wound Repair Regen. 2014;22(5):613-21. PMID 25041740.
5. 5

   Miranda E, et al. Efficacy of LL-37 cream in enhancing healing of diabetic foot ulcer: a randomized double-blind controlled trial. Arch Dermatol Res. 2023;315(9):2623-2633. PMID 37480520. Registry record ClinicalTrials.gov NCT04098562.
6. 6

   ClinicalTrials.gov NCT02225366, Intratumoral Injections of LL37 for Melanoma, M.D. Anderson Cancer Center. Verified against the ClinicalTrials.gov v2 API 2026-08-28: completed, actual enrollment 4, 3 treated, 1 discontinued for lack of efficacy, results posted.
7. 7

   Dolkar T, et al. Dermatologic toxicity from novel therapy using antimicrobial peptide LL-37 in melanoma: A detailed examination of the clinicopathologic features. J Cutan Pathol. 2018;45(7):539-544. PMID 29665030.
8. 8

   PubChem Compound record CID 16198951. CAS 154947-66-7 round-tripped via PUG-REST 2026-08-28. Formula C205H340N60O53, molecular weight 4493, INN ropocamptide.
9. 9

   Lande R, et al. Plasmacytoid dendritic cells sense self-DNA coupled with antimicrobial peptide. Nature. 2007;449(7162):564-9. PMID 17873860.
10. 10

    Srakaew N, et al. Antimicrobial host defence peptide, LL-37, as a potential vaginal contraceptive. Hum Reprod. 2014;29(4):683-96. PMID 24549217.
11. 11

    Kiattiburut W, et al. Antimicrobial peptide LL-37 and its truncated forms, GI-20 and GF-17, exert spermicidal effects and microbicidal activity against Neisseria gonorrhoeae. Hum Reprod. 2018;33(12):2175-2183. PMID 30357408.
12. 12

    Acen EL, et al. Impact of vitamin D status and cathelicidin antimicrobial peptide on adults with active pulmonary TB globally: A systematic review and meta-analysis. PLoS One. 2021;16(6):e0252762. PMID 34115790.
13. 13

    Ho J, et al. Pathological Role and Diagnostic Value of Endogenous Host Defense Peptides in Adult and Neonatal Sepsis: A Systematic Review. Shock. 2017;47(6):673-679. PMID 27941592.
14. 14

    Lande R, et al. The antimicrobial peptide LL37 is a T-cell autoantigen in psoriasis. Nat Commun. 2014;5:5621. PMID 25470744. Carries a corrigendum, PMID 25759123.; the paper stands.
15. 15

    Yamasaki K, et al. Increased serine protease activity and cathelicidin promotes skin inflammation in rosacea. Nat Med. 2007;13(8):975-80. PMID 17676051.
16. 16

    Ohuchi K, et al. LL-37 Might Promote Local Invasion of Melanoma by Activating Melanoma Cells and Tumor-Associated Macrophages. Cancers (Basel). 2023;15(6). PMID 36980564.
17. 17

    Girnita A, et al. Identification of the cathelicidin peptide LL-37 as agonist for the type I insulin-like growth factor receptor. Oncogene. 2012;31(3):352-65. PMID 21685939. RETRACTED; PubMed publication type "Retracted Publication". Retraction notice: Oncogene. 2023;42(8):626. PMID 36494433.
18. 18

    US Food and Drug Administration. Bulk Drug Substances Nominated for Use in Compounding Under Section 503A, version dated 22 April 2026 (archived copy of fda.gov/media/94155/download). Text-extracted 2026-08-28; carries the Cathelicidin LL-37 removal note verbatim.
19. 19

    US Food and Drug Administration. Bulk Drug Substances Nominated for Use in Compounding Under Section 503A, updated 14 May 2026 (fda.gov/media/94155/download). Downloaded and text-extracted 2026-08-28: 7 pages, zero occurrences of "LL-37" or "cathelicidin".
20. 20

    US Food and Drug Administration. Certain Bulk Drug Substances for Use in Compounding that May Present Significant Safety Risks. Fetched and text-extracted 2026-08-28; LL-37 entry live under the heading "Bulk drug substances nominated but withdrawn".
21. 21

    Zhao Y, et al. Efficacy and safety of Oral LL-37 against the Omicron BA.5.1.3 variant of SARS-COV-2: A randomized trial. J Med Virol. 2023;95(8):e29035. PMID 37605995.
22. 22

    ClinicalTrials.gov v2 API, 2026-08-28. query.intr=LL-37 returns 20 records, enumerated in full; the broader query.term=LL-37 returns 61. Only NCT02225366 and NCT04098562 administer LL-37; the remainder measure endogenous levels or administer vitamin D. No phase 3 LL-37 record appears among them. This registry never held the two Promore venous leg ulcer trials, which are EudraCT records, so it is evidence about ClinicalTrials.gov only.
23. 23

    Promore Pharma AB company disclosures, October 2023 (news.cision.com; Nasdaq First North).
24. 24

    openFDA field-qualified searches of drug/label, drug/drugsfda, drug/ndc and other/substance for LL-37, ropocamptide and cathelicidin, all six returning NOT_FOUND. The only FDA record is substance-registry UNII 3DD771JO2H, substance class protein, whose synonym list also carries the rabbit CAP18 designation.