LL-37: Antimicrobial Peptide Explained - Uses, Dosage, Risks, Status

Key takeaways

LL-37 is the only human cathelicidin: a 37-amino-acid antimicrobial peptide that white blood cells and the lining of the skin, gut and airways already make. It kills bacteria in a dish, helps skin wounds close and signals the immune system, in both directions.
The human evidence is topical and small. In 34 patients with chronic venous leg ulcers, the two lower concentrations sped healing over four weeks; a 148-patient follow-up trial found no improvement over placebo (complete closure 26.5 and 24.7 percent versus 25.3 percent). A 25-patient diabetic-foot-ulcer trial improved granulation but not wound area. Nobody has published a trial of LL-37 injected under the skin, the way it is sold.
Seller charts of 100 to 200 mcg injected daily come from no study. What was studied: 0.5 to 3.2 mg/mL applied to open ulcers twice a week, and 250 to 500 mcg injected weekly into melanoma metastases in three cancer patients.
The main risk is that LL-37 is pro-inflammatory as well as antimicrobial. Excess cathelicidin in the skin is implicated in rosacea and psoriasis, injecting LL-37 into mouse skin is the standard way to create rosacea-like lesions, it damages human cells in culture at concentrations close to those that kill bacteria, and it promotes some cancers in laboratory models while suppressing others.
Legal status: never approved; off FDA's Category 2 list since April 15, 2026 after its nomination was withdrawn; due for advisory committee review by February 2027; not on the permitted list, so no US pharmacy may compound it in October 2026, whatever some clinic pages say.

LL-37 at a glance

Also called: LL37, cathelicidin, hCAP18 (the precursor protein), CAMP (the gene), ropocamptide (the developer's name for the topical drug candidate)
Class: 37-amino-acid human antimicrobial peptide, the only human cathelicidin; cut from the precursor protein hCAP18 and made by white blood cells and by the lining of the skin, gut and airways; kills microbes, attracts immune cells and helps wounds close
Route: Sold mainly as research vials for subcutaneous injection (usually 5 or 10 mg); in human trials it was applied to chronic leg and foot ulcers, injected into melanoma skin metastases, and given by mouth as a genetically engineered bacterium
FDA status: Never approved by FDA, and not approved anywhere else as far as we could find. Previously on FDA's Category 2 list and removed on April 15, 2026 after its nomination was withdrawn; FDA's advisory committee is due to review it by February 2027. It is not on the 503A bulks list, so no US pharmacy may compound it as of October 2026.
FDA Category 2 (compounding restricted): No
Available by prescription (US): No. It is not FDA-approved, has no USP monograph, is not part of an approved drug and is not on the 503A bulks list, so no licensed pharmacy may compound it
Best-studied use: Topical treatment of hard-to-heal venous leg ulcers (two randomized trials; the larger one missed its main goal) and diabetic foot ulcers (one 25-patient trial); weekly injections into melanoma skin metastases in three patients; an engineered oral bacterium that makes LL-37, in COVID-19
Typical monthly cost (telehealth, 2026): No legal US price; injectable vials sold online are research-labeled products with no pharmacy, label or recourse behind them

What LL-37 is and how it works

Your cells make a protein called hCAP18, for human cationic antimicrobial protein of 18 kilodaltons, and cut off its last 37 amino acids. The fragment begins with two leucines, hence the name LL-37 (Piktel 2016). Other mammals have several cathelicidins; humans have only this one. It is produced by neutrophils, monocytes and other white cells, and by the epithelial cells that line the skin, gut, airways and testis (Dürr 2006). It belongs to the same broad family of host-defense molecules as the defensins, and like them it is one of the body's first responses to a breach.

What it does explains both the interest and the risk:

It kills microbes, and not only microbes. LL-37 folds into a helix with one water-loving and one fat-loving face, which lets it disrupt bacterial membranes. In a laboratory, it stopped the growth of E. coli at 5 micromolar; at 13 to 25 micromolar it was toxic to several kinds of human cells; and in the presence of human serum both the antibacterial and the toxic activity were blocked (Johansson 1998). The window between killing bacteria and damaging your own cells is narrow, and serum shuts both down.
It helps wounds close. Human skin produces large amounts of the precursor after injury, peaking at about 48 hours and falling back once the wound closes. In chronic ulcers it is low and absent from the ulcer edge, and antibodies that block LL-37 slowed the closure of wounds in organ-cultured human skin (Heilborn 2003). That observation is the rationale for every human trial of LL-37 as a drug.
It signals the immune system. LL-37 binds and neutralizes bacterial endotoxin, attracts immune cells to wounds and infection sites (Dürr 2006), and binds fragments of the body's own DNA, turning them into a trigger for interferon production by a class of dendritic cells (Lande 2007). That last property is why the same molecule shows up in autoimmune skin disease.

The short version: LL-37 is not an anti-inflammatory peptide. It is a first-responder molecule that kills microbes, calls in immune cells and helps close wounds, and whether that helps or harms depends on the amount, the place and the person.

What the evidence actually shows

There are thousands of laboratory papers on LL-37 and very few human ones. We searched PubMed and ClinicalTrials.gov on October 10, 2026 for LL-37, LL37, ropocamptide and cathelicidin and found five interventional human studies: two venous-leg-ulcer trials, one diabetic-foot-ulcer trial, one melanoma injection trial and one trial of an engineered oral bacterium. None tested LL-37 injected under the skin.

LL-37 claims, graded by the evidence behind them
Claim	Grade	What the evidence shows	Sources
Speeds healing of hard-to-heal venous leg ulcers (topical)	B	First trial, 34 patients, twice-weekly applications for 4 weeks: healing rate about six and three times placebo at 0.5 and 1.6 mg/mL; 3.2 mg/mL no better than placebo. Phase 2b, 148 patients, 13 weeks: no significant benefit overall (complete closure 26.5% and 24.7% versus 25.3% on placebo); a post hoc subgroup with large wounds did better.	Grönberg 2014; Mahlapuu 2021
Helps diabetic foot ulcers heal (topical cream)	B	25 adults, 0.5 mg/g cream twice weekly for 4 weeks: granulation index rose faster than on placebo at every visit, but the reduction in wound area was not significantly different, and inflammatory cytokines and bacterial counts did not fall.	Miranda 2023
Kills bacteria and biofilms in the body; treats Lyme disease, mold illness or SIBO	D	Antibacterial in laboratory dishes at about 5 micromolar, toxic to human cells at 13 to 25 micromolar, and both effects blocked by human serum. No human study has tested administered LL-37 against any infection outside a wound, and we found no human study of it in Lyme disease, mold illness or SIBO.	Johansson 1998; Dürr 2006
Treats cancer when injected into tumors	D	A phase 1/2 melanoma trial injected 250 then 500 mcg into each skin metastasis weekly; 4 were enrolled and 3 treated against a planned 36, with no interpretable efficacy result and no publication. In lab models, LL-37 promotes ovarian, lung and breast cancer and suppresses colon and gastric cancer.	NCT02225366; Piktel 2016
Fights viral infection (the 'oral LL-37' COVID trial)	C	An open-label trial in 238 hospitalized Omicron patients gave a genetically engineered Lactococcus lactis that produces LL-37; started early, it shortened the time to a negative PCR (about 9.8 versus 14.0 days). A bacterium, not the synthetic peptide sold in vials.	Zhao 2023
Clears acne, eczema or aging skin	D	No human trial. The skin evidence runs the other way: people with rosacea overproduce cathelicidin, and LL-37 injected into mouse skin produces rosacea-like lesions, which did not fully recover after 20 days of twice-daily injections.	Yamasaki 2007; Zhang 2023
Calms autoimmune or inflammatory disease	D	In psoriasis, LL-37 converts self-DNA into an interferon trigger, and two-thirds of patients with moderate-to-severe plaque psoriasis carried T cells that recognize LL-37, in step with disease activity.	Lande 2007; Lande 2014
Safe to inject under the skin	D	No human data for that route. Weekly injections into melanoma lesions were followed by a case of widespread wart-like and blistering skin lesions that resolved within 2 months of stopping.	Dolkar 2018; NCT02225366

A = strong human trials or meta-analyses · B = small or short human trials · C = observational, mechanistic or indirect human data · D = animal/in-vitro only, or no evidence. Grades describe the evidence, not whether we recommend the use.

Chronic leg and foot ulcers: the one real test

The wound program is the best human evidence on LL-37, and it is honest about its limits. The first trial, run by the Swedish company then developing the peptide, enrolled 34 people with venous leg ulcers that had not healed on standard care. After a three-week run-in on placebo, they were randomized to LL-37 at 0.5, 1.6 or 3.2 mg/mL or placebo, applied twice a week for four weeks. The healing-rate constants for the two lower concentrations were about six and three times those on placebo, mean ulcer area fell 68 and 50 percent, the highest concentration did no better than placebo, and there were no local or systemic safety concerns (Grönberg 2014).

The phase 2b trial was designed to confirm it. It randomized 148 patients in Poland and Sweden (mean age 67.6, median ulcer duration 20.3 months) to 0.5 or 1.6 mg/mL or placebo on top of compression, the standard care, applied as 25 microliters per square centimeter of ulcer twice a week for 13 weeks. Across the whole population there was no significant difference in any healing measure: confirmed complete closure was 26.5 percent at 0.5 mg/mL, 24.7 percent at 1.6 mg/mL and 25.3 percent on placebo. In a post hoc analysis of patients with wounds of at least 10 square centimeters, closure was 28.1 percent on the lower concentration against 8.1 percent on placebo; in smaller wounds the full text shows placebo ahead, at 44.2 percent against 26.0 and 31.4 percent, a difference that was not statistically significant either. Three authors worked for the developer, which calls the drug ropocamptide (Mahlapuu 2021). A subgroup found after the fact is a reason for another trial, not a result.

An independent Indonesian group tested an LL-37 cream at 0.5 mg per gram, applied twice a week for four weeks, in 25 adults with diabetic foot ulcers, 13 on LL-37 and 12 on placebo. Granulation tissue filled the wound faster on LL-37 at every visit, but wound area shrank by a similar amount in both groups, and the cream did not lower inflammatory cytokines or bacterial counts (Miranda 2023).

Notice the dose pattern across those trials: the lowest concentration did best, and more was not better.

Cancer: three patients and one skin reaction

At MD Anderson Cancer Center, a phase 1/2 trial injected LL-37 into two to four skin metastases of melanoma once a week for up to eight weeks, starting at 250 mcg per tumor and moving to 500 mcg. It planned for up to 36 participants; the results posted to the registry show 4 enrolled and 3 treated. No serious adverse events were recorded; other events included one squamous cell carcinoma, one actinic keratosis, skin hypopigmentation and anemia. The posted efficacy counts cannot be read as tumor responses, and we found no publication of the trial (NCT02225366). A case report from the same center described a 63-year-old woman with stage IIIC melanoma whose injected lesions shrank over eight weekly injections, but who developed multiple wart-like papules and a blistering lesion on her trunk and limbs about 45 days in; 11 of 12 biopsies showed lichenoid inflammation over atypical, keratoacanthoma-like squamous growth, and all of it resolved within two months of stopping (Dolkar 2018).

The laboratory picture is mixed in a way that matters: overexpressed LL-37 promoted ovarian, lung and breast cancer in models and suppressed colon and gastric cancer (Piktel 2016). A peptide that can feed some tumors is not a general anticancer agent.

The 'oral LL-37' COVID trial is a different product

The trial most often cited for oral LL-37 gave 238 hospitalized adults with the Omicron BA.5.1.3 variant either placebo or a genetically engineered Lactococcus lactis bacterium that produces LL-37, in an open-label, single-center design. In patients who started treatment within six days of diagnosis, the time to a negative PCR test was about 9.8 days versus 14.0 on placebo, with no severe adverse events (Zhao 2023). It measured viral clearance, not symptoms; it was unblinded; and it says nothing about swallowing or injecting the synthetic peptide.

The skin problem: rosacea and psoriasis

For anyone considering LL-37 'for skin', this is the most important section. People with rosacea express abnormally high levels of cathelicidin in their facial skin, processed into unusual fragments by an overactive skin enzyme; injecting those fragments into mouse skin increased inflammation, and mice lacking the cathelicidin gene did not develop the enzyme-driven inflammation (Yamasaki 2007). Injecting LL-37 into mouse skin has since become the standard laboratory model of rosacea: two days of injections every 12 hours produce lesions that recover, and 20 days produce lesions with collagen deposition that did not fully recover (Zhang 2023).

Psoriasis points the same way. LL-37 is overexpressed in psoriatic skin and binds the body's own DNA, turning it into a signal that makes plasmacytoid dendritic cells pour out interferon, a pathway proposed as a driver of the disease (Lande 2007). Two-thirds of patients with moderate-to-severe plaque psoriasis carried T cells that recognize LL-37 itself, and their presence tracked disease activity (Lande 2014).

None of that proves that an injection would cause rosacea or psoriasis in a given person; it does mean that someone with either disease, or a family history of it, is the person most likely to be harmed by adding more of this molecule to their skin.

What you are promised versus what is documented

LL-37 sales pages and forum protocols promise a 'natural antibiotic' that dissolves biofilms, treats Lyme disease, mold illness and SIBO, heals the gut, clears acne and eczema, speeds wound healing, boosts immunity and fights cancer. The evidence sorts those claims into three bins.

Documented in humans: tolerability of topical LL-37 on chronic leg ulcers; a healing signal in a 34-patient trial that a 148-patient trial did not confirm; faster granulation in diabetic foot ulcers; faster viral clearance with an engineered bacterium in one unblinded COVID trial.
Plausible but untested in people: help for large or infected chronic wounds, which is where the developer's subgroup pointed; any effect of an injected dose anywhere.
Not supported, or contradicted: systemic antibiotic or anti-biofilm action, which serum blocks in the laboratory; Lyme disease, mold illness and SIBO, which have no data at all; 'anti-inflammatory for skin', contradicted by its role in rosacea and psoriasis; 'anticancer', contradicted by its tumor-promoting effects in several cancer models.

LL-37 versus the realistic alternatives

	LL-37	KPV	Thymosin alpha-1	What a clinician can prescribe
What it is	37-amino-acid human antimicrobial peptide	Three-amino-acid fragment of alpha-MSH	28-amino-acid thymus peptide	Compression and wound care for venous ulcers; debridement and infection control for foot ulcers; culture-guided antibiotics; dermatology treatment for rosacea or psoriasis
Human evidence	Topical ulcer trials (the larger one negative overall); three melanoma patients; one engineered-bacterium trial	None	Randomized trials in hepatitis B, vaccine response and sepsis; the 1,106-patient sepsis trial showed no mortality benefit	Randomized trials and guidelines
Direction on inflammation	Both: antimicrobial and wound-healing, but pro-inflammatory in rosacea and psoriasis	Dampens NF-kappa-B signaling in cells and mice	Immune modulator	Targeted to the diagnosis
Can a US pharmacy dispense it in October 2026?	No	No	No	Yes
FDA committee status	Review due by February 2027	Recommended 8-6-1 in July 2026; no rule	Voted 4 yes, 17 no in December 2024	Not applicable

The pairing people ask about most is LL-37 with KPV for gut and skin; our LL-37 versus KPV comparison takes it apart. The KPV profile and the thymosin alpha-1 profile cover the other two, and the gut and inflammation guide ranks all of them, LL-37 last. For skin, the best peptides for skin guide covers what has human data.

Dosage: what was studied versus what is sold

Setting	Dose and route	Population	Result
First leg-ulcer trial, 2014	0.5, 1.6 or 3.2 mg/mL applied to the ulcer twice weekly for 4 weeks	34 adults with hard-to-heal venous leg ulcers	Faster healing at the two lower concentrations; 3.2 mg/mL no better than placebo
Phase 2b leg-ulcer trial, 2021	0.5 or 1.6 mg/mL at 25 microliters per square centimeter (12.5 or 40 mcg per square centimeter) twice weekly for 13 weeks	148 adults, mean age 67.6	No overall benefit; post hoc benefit in large wounds
Diabetic foot-ulcer trial, 2023	Cream at 0.5 mg/g, 0.025 mL per square centimeter, twice weekly for 4 weeks	25 adults	Faster granulation; wound area not significantly different
Melanoma trial, 2015 to 2020	250 mcg, then 500 mcg, injected into each of 2 to 4 skin metastases weekly for up to 8 weeks	3 adults with metastatic melanoma	No interpretable efficacy result
Oral COVID trial, 2023	Engineered Lactococcus lactis producing LL-37; amount not stated in the abstract	238 hospitalized adults	Faster viral clearance when started early
Seller charts, 2026	100 to 200 mcg subcutaneously once daily, often 5 days on and 2 off, in 2- to 4-week blocks; some charts go to 500 mcg; 5 or 10 mg vials	Anyone with a credit card	No data of any kind

The arithmetic is worth doing once, labeled as our arithmetic. At the start of the phase 2b trial, the average ulcer of 11.6 square centimeters received about 145 mcg of LL-37 per application at 0.5 mg/mL, or about 465 mcg at 1.6 mg/mL, twice a week: roughly 290 to 930 mcg a week, applied to an open wound. A seller chart of 100 to 200 mcg injected five days a week is 500 to 1,000 mcg a week, pushed into intact tissue by a route no trial has tested. The charts are not a cautious fraction of anything studied.

We give no reconstitution example, because there is no legal LL-37 product to reconstitute and no pharmacy label to check the arithmetic against; FDA's 2026 warning letters cite exactly that kind of dosing help as evidence that research vials are drugs for human use. Our reconstitution calculator is for medications a pharmacy has dispensed. The LL-37 dosage article goes through each studied dose in more detail.

Side effects and who should not use it

What is known, from topical use on wounds. In the 148-patient trial, 6 adverse events in 4 patients were judged possibly related to treatment: wound infection and dermatitis in one patient each on the lower concentration, peripheral swelling in one on the higher, and deep phlebitis in one on placebo; none of the 12 serious events was judged related (Mahlapuu 2021). In the foot-ulcer trial, one of 13 patients on LL-37 developed mild irritant contact dermatitis, attributed to wound exudate (Miranda 2023).

What is known, from injection into tumors. The case of widespread wart-like and blistering skin lesions after eight weekly injections, resolving within two months of stopping (Dolkar 2018), and a registry adverse-event list that includes a squamous cell carcinoma, an actinic keratosis and skin hypopigmentation in three treated patients (NCT02225366).

What is not known. Anything about injecting LL-37 under the skin: local reactions, how far it travels, how long it lasts, whether antibodies form, and what weeks of daily injections do. The theoretical risks follow from its biology: inflammation at and around the injection site, since it is cytotoxic to human cells in culture at concentrations close to those that kill bacteria (Johansson 1998); rosacea-like or psoriasis-like reactions; immune activation against the body's own DNA; possible support for some tumors; and the product itself, with no verified identity, purity, endotoxin testing or sterility.

Do not use LL-37, or treat it strictly as an unknown, if you: have rosacea, psoriasis or another autoimmune disease, or a strong family history of one; have a current or past cancer, particularly ovarian, breast or lung cancer, where it promoted tumors in models; have an infection that needs treatment, since delaying antibiotics for a research peptide can turn a treatable infection into a serious one; have a chronic wound, which needs a cause found (venous disease, arterial disease, diabetes, pressure); are pregnant, trying to conceive or breastfeeding; or are a tested athlete, because an unapproved substance falls under the S0 catch-all of the anti-doping rules. If you develop spreading redness, blistering, new skin growths, fever or a flare of a skin condition, stop and see a clinician.

Legal status in the United States

The full story is in the legal tracker; the short version:

Never approved. LL-37 has never been approved by FDA, has never been a component of an approved drug and has no USP monograph. The topical drug candidate, ropocamptide, completed a phase 2b trial that missed its main goal (Mahlapuu 2021).
Category 2, then off it. LL-37 was on FDA's Category 2 list of bulk substances that may present significant safety risks in compounding. On April 15, 2026 FDA announced its removal, with eleven other peptides, because the nomination had been withdrawn (Frier Levitt 2026). That ended the safety flag; it did not make LL-37 compoundable.
Review pending. FDA's Pharmacy Compounding Advisory Committee is due to review LL-37, with dihexa, injectable GHK-Cu, PEG-MGF and melanotan II, by February 2027. A favorable vote would only be advice; FDA would still have to finalize a rule, which it has not done for any of the six peptides recommended in July 2026.
Not compoundable today, whatever you read. Some clinic pages and AI-generated summaries list LL-37 among peptides a pharmacy may compound. As of October 2026 that is false: it passes none of the three 503A tests, and a prescription cannot make a non-compoundable substance compoundable.
Research labels do not change the analysis. FDA's 2026 warning letters to online peptide sellers rejected 'research use only' disclaimers on products sold with dosing guides, reconstitution calculators or bacteriostatic water. The letters go to sellers, not buyers; possessing LL-37 is not a crime.

Cost and the prescription question

There is no legal LL-37 product in the United States in October 2026, so there is no legitimate price to quote, and this page offers no purchase guidance. What exists are research vials and clinics that attach a consultation to them; our guide to what telehealth can legally prescribe explains why that prescription is cosmetic, and our cost guide covers what the peptides with a legal path actually cost. What you can be prescribed depends on the problem: for a leg ulcer that will not heal, a wound clinic, compression and a search for the cause; for recurrent or chronic infections, a diagnosis and an antibiotic chosen from a culture; for rosacea, acne or psoriasis, a dermatologist, with treatments that have randomized trials behind them. If you are new to how peptide prescriptions work, start with our complete guide to peptide therapy and what peptides are.

Women, older adults and athletes

There are no human data on injected LL-37 by sex or age. The leg-ulcer trials enrolled older adults (mean age 67.6 in the larger one), and topical LL-37 was well tolerated on their wounds, which says nothing about injections. Nothing is known about pregnancy or breastfeeding, so avoid it then. For older adults, more medications and a higher cancer risk make an unmonitored immune-active peptide a worse bet. For athletes, LL-37 is not named on the WADA list but falls under its S0 catch-all for unapproved substances.

Bottom line

LL-37 is a real and important part of human immunity: an antimicrobial, wound-closing, immune-signaling peptide that the body makes where it needs it. As a drug, it has been tested properly in one setting, topical treatment of chronic leg ulcers, and the larger trial did not confirm the smaller one. Nobody has studied injecting it under the skin, and what is known about its biology argues for caution: it drives rosacea-like inflammation in mice, it is part of the machinery of psoriasis, it damages human cells near the concentrations that kill bacteria, and it feeds some tumors in laboratory models.

In October 2026 there is no legal LL-37 product in the United States, and the committee review due by February 2027 will not change that on its own. If you have a wound, an infection or a skin disease, treatments with evidence are available on prescription today; a research vial of LL-37 is a dose a seller chose, for uses no human study supports.

Frequently asked questions

+What is LL-37 peptide used for?

In the body, LL-37 is part of the first-line immune defense: it kills bacteria, calls in immune cells and helps skin wounds close. As a drug, it has been tested mainly as a topical solution or cream on chronic leg and foot ulcers, where a small first trial looked promising and a 148-patient trial found no overall benefit over placebo. Sellers market injections for infections, biofilms, Lyme disease, gut healing and skin, none of which has been tested in people.

+Is LL-37 FDA approved or legal in 2026?

LL-37 is not FDA-approved, and no licensed US pharmacy may compound it as of October 2026. It was on FDA's Category 2 safety-risk list until April 15, 2026, when it came off because its nomination was withdrawn; that lifted the flag but did not put it on the list of substances pharmacies may use. FDA's advisory committee is due to review it by February 2027. Vials sold online as research chemicals are unapproved drugs when marketed for human use; possessing one is not a crime.

+Can a doctor prescribe LL-37?

A clinician can write anything on a prescription, but no licensed pharmacy may legally compound LL-37 to fill it in October 2026, because it passes none of the three tests in section 503A: it has no USP monograph, it is not part of an approved drug, and it is not on FDA's permitted bulks list. Some clinic pages and online summaries say otherwise; they are wrong. A clinic that supplies LL-37 is supplying a research product.

+What are the side effects of LL-37?

On open leg ulcers, topical LL-37 was well tolerated: in the 148-patient trial, 6 adverse events in 4 patients were judged possibly related, including wound infection and dermatitis, and one of those patients was on placebo. Injected into melanoma metastases it produced a case of widespread wart-like and blistering skin lesions that cleared after it was stopped. No study has recorded the effects of injecting it under the skin. The theoretical risks are inflammation rather than infection: LL-37 drives rosacea-like lesions in mice, is implicated in psoriasis, and damages human cells in culture at concentrations not far above those that kill bacteria.

+Is LL-37 good for skin, acne or rosacea?

The evidence points the other way for rosacea and psoriasis. People with rosacea have abnormally high cathelicidin in their facial skin, injecting LL-37 into mouse skin is the standard way researchers create rosacea-like lesions, and in psoriasis LL-37 turns the body's own DNA into an immune trigger and is recognized by patients' T cells. There is no human trial of LL-37 for acne, eczema or skin aging. The only skin data in people are topical applications to chronic ulcers.

+Does LL-37 work for Lyme disease, biofilms or chronic infections?

No human study has tested it for any of them. LL-37 kills bacteria in laboratory dishes, but human serum blocks its antibacterial activity, and nobody has measured what an injected dose does to an infection anywhere in the body. If you have a chronic or recurrent infection, a diagnosis and an antibiotic chosen from a culture have evidence behind them; a research vial does not.

+What is the LL-37 dosage?

There is no studied dose for injecting LL-37 under the skin. The human doses that exist are topical concentrations of 0.5, 1.6 and 3.2 mg/mL applied to venous leg ulcers twice a week, a 0.5 mg/g cream on diabetic foot ulcers twice a week, and 250 to 500 mcg injected weekly into each melanoma metastasis in three cancer patients. The 100 to 200 mcg daily injections on seller charts come from no study. Our LL-37 dosage article lays them side by side.

+Can you take LL-37 orally or as a nasal spray?

There is no human study of oral or nasal synthetic LL-37. The trial reported as 'oral LL-37' used a genetically engineered Lactococcus lactis bacterium that produces the peptide in the gut, given to hospitalized COVID-19 patients in an open-label study; it is a different product from a capsule or spray of the peptide. Whether any swallowed or sprayed synthetic peptide survives and is absorbed has never been measured in a person.

+LL-37 vs KPV: which is better for gut and skin?

Neither can be compounded in the US, and neither has a human trial for a gut condition. LL-37 has topical human data in chronic ulcers but is also part of how rosacea and psoriasis develop; KPV has no human data at all but dampens inflammatory signaling in mouse colitis and in cell studies, and FDA's advisory committee recommended it 8-6-1 in July 2026 without a rule following. Our comparison article goes through both.