Topical LL-37 aided ulcer healing in small trials; the largest missed

What is LL-37?

LL-37 is the active C-terminal fragment of hCAP18, the product of the CAMP gene, released when proteases in skin, neutrophils, and mucosal surfaces cleave it. Its 37 amino acids form a cationic, amphipathic helix that inserts into and disrupts microbial membranes, the basis of its "natural antibiotic" framing.

Membrane disruption is only half of what LL-37 does. Reviews of the immunomodulatory literature describe LL-37 as bidirectional: in infectious contexts it promotes antimicrobial defense and cytokine production, while in sterile inflammation it suppresses pro-inflammatory cascades such as TNF-alpha and IL-1beta ^[5].

The same review documents LL-37's role across wound healing, autoimmunity, and cancer biology. It describes LL-37 as a context-dependent signal, not a single clean "healing" or "killing" agent. Its effect depends on where it is released, how much is present, and what tissue state it meets.

What did the small LL-37 ulcer trials show?

Two small randomized trials of topical LL-37 at 0.5 mg/mL produced positive healing signals, one in venous leg ulcers and one in diabetic foot ulcers ^[1]⁠^[3].

The first was a randomized, placebo-controlled trial in hard-to-heal venous leg ulcers (n=34). Topical LL-37 at 0.5 mg/mL produced healing rate constants roughly six times higher than placebo (p=0.003) and a 68% mean reduction in ulcer area, with no local or systemic safety concerns during the trial ^[1]. It is the study most often cited as "proof," and an encouraging early signal.

The second was a small RCT in diabetic foot ulcers (n=40) that used the same 0.5 mg/mL concentration as a topical cream, twice weekly for four weeks. The granulation index rose consistently at all four weekly checkpoints (p ranging from 0.006 to 0.037). Inflammatory cytokines and bacterial colonization did not change significantly compared with placebo ^[3].

What did the 148-patient Phase IIb trial of LL-37 find?

The largest LL-37 trial, a multicentric Phase IIb study of 148 patients with hard-to-heal venous leg ulcers, failed its primary endpoint of improved complete wound closure ^[2].

It was the largest and best-powered human LL-37 study to date. Benefit appeared only in a post-hoc subgroup of patients with large ulcers (at least 10 cm²). The authors describe that result as hypothesis-generating and as warranting further, properly powered study, not as confirmation ^[2].

Post-hoc subgroup findings help plan the next trial. They are not evidence that a treatment works. The study with the most patients and the most statistical power did not replicate the earlier small positive results, so it carries the most weight, and it points toward "unproven," not "proven." LL-37's evidence grade is D, limited. The pexiganan Phase 3 diabetic foot ulcer trials show another peptide's path through wound trials.

Does LL-37 kill bacteria in human wounds?

LL-37 did not reduce bacterial colonization compared with placebo in the diabetic foot ulcer trial ^[3]. Its antimicrobial activity in vitro is real but concentration-dependent.

In that trial, the tissue-repair signal showed up without a measurable antimicrobial effect ^[3]. That pattern fits the immunomodulatory framing of LL-37 better than the antibiotic one.

Is there evidence for injecting LL-37?

No human trial has tested systemic LL-37 administration. The closest evidence is a mouse model of MRSA wound infection, where combined topical and intraperitoneal LL-37 produced better re-epithelialization, granulation tissue, collagen organization, and angiogenesis than the comparator antibiotic teicoplanin ^[4].

Much of the confidence behind injectable LL-37 protocols traces to preclinical work like this study. The finding is real, and it is the closest the evidence comes to showing that systemic administration did something.

Three gaps separate the mouse result from human injection. It is a single animal model, and mouse immunology and skin biology differ substantially from human. Intraperitoneal dosing does not translate directly to subcutaneous self-injection, and route effects differ by molecule. No human trial has replicated systemic administration at all.

Forums that cite the mouse study for injection protocols extrapolate across species and route of administration at the same time. That double gap turns "mechanistically plausible" into "unproven" the moment the protocol is applied to a person. The same topical-versus-injectable question comes up for GHK-Cu.

What LL-37 dose has been tested in people?

The positive human trials used topical LL-37 at 0.5 mg/mL, applied to open chronic wounds a few times weekly under clinical monitoring, for short defined courses ^[1]⁠^[3]. The diabetic foot ulcer course lasted four weeks ^[3].

That is the entire evidence-backed dosing picture. No controlled human trial has tested subcutaneous injection at any dose. No trial has evaluated acne, general skin quality, gut dysbiosis, or "immune support," and no trial has evaluated nasal or systemic delivery.

Any mg amount or cycle length circulating outside that narrow topical-wound context is an extrapolation, often an untraceable one, not a figure derived from controlled evidence.

Is LL-37 safe because the body already makes it?

Natural production does not make self-dosed LL-37 safe: the same molecule is mechanistically implicated in psoriasis and lupus pathology when dysregulated ^[5].

The body produces LL-37 locally, in small amounts, for short windows, largely at the site of injury or infection. Exogenous LL-37, applied topically at a fixed concentration or injected systemically, is a fundamentally different exposure. It has no local feedback, no natural clearance timing, and a dose chosen by a vendor protocol instead of physiology.

What side effects are reported with LL-37?

Reported effects with exogenous LL-37 include injection-site redness and discomfort, plus transient fatigue, headache, or mild nausea, more often at higher doses. In vitro, LL-37 is cytotoxic to host cells at concentrations not far above its antimicrobial range.

Research-grade LL-37 sold outside pharmaceutical channels also carries its own purity and identity uncertainty.

What is still unknown about LL-37?

The largest LL-37 evidence gaps concern safety, not efficacy:

• Systemic dosing safety. No controlled trial has measured the safety of systemic LL-37 dosing.
• Injection pharmacokinetics. No human data describe what happens to LL-37 after injection.
• Autoimmune flare. LL-37 is implicated in driving inflammatory skin and autoimmune conditions like psoriasis and lupus when dysregulated ^[5]. Whether raising LL-37 outside its normal local, transient window pushes susceptible people toward inflammatory flare instead of repair is a plausible question, and nobody has run that trial.
• Other uses. No trial shows that the topical wound signal carries over to the skin or immune uses forums claim.

The mechanism is real. The topical wound-healing signal is real but inconsistent at scale. The systemic, self-dosed use that forums describe sits almost entirely outside anything the evidence has tested.