Key takeaways
- Thymosin alpha 1 has been tested against placebo in people: in a 2025 trial of 1,106 adults with sepsis, deaths within 28 days did not differ significantly, and FDA’s 2024 review found a lack of evidence of effectiveness for the uses it examined.
- LL-37 is a peptide the body makes. In a 25-person ulcer trial, two inflammatory markers did not fall with LL-37 cream, and research on psoriasis and rosacea describes LL-37 as driving inflammation.
- BPC-157’s placebo-controlled trial with an inflammation-related measure is known only from a 2005 conference abstract, which FDA staff judged inadequate to support efficacy or safety; an uncontrolled study that measured blood markers of inflammation has posted no results.
- The inflammation-model studies of KPV and GHK-Cu used mice, zebrafish larvae and cells. We found no published human study or registered trial of KPV (21 November 2026). GHK-Cu has randomised trials on skin wounds; the one we found that tracked redness, after laser resurfacing, found no significant difference in how soon it resolved on objective measures.
- None of the five is an FDA-approved medicine. Italy’s medicines agency lists one medicine containing thymosin alpha 1, Zadaxin, as authorised and prescription-only; research-grade material labelled thymosin alpha 1 is not that medicine.
01How far has research on peptides and inflammation gone?
How far the published research has gone for each peptide, dated 21 November 2026. Each bar shows the strongest type of study that gave the peptide in a setting marked by inflammation or infection. A longer bar means a stronger study design, not a better result. Bar lengths show order only; they are not a score.
Thymosin alpha 1: a placebo-controlled trial in 1,106 adults with sepsis. LL-37: randomised trials in people with chronic ulcers, the three described here with 25, 34 and 148 people. GHK-Cu: randomised trials on skin wounds, among them a skin-care trial after laser resurfacing that 13 people completed, plus mouse, zebrafish and cell models. BPC-157: one placebo-controlled trial known only from a conference abstract, plus rat models. KPV: mouse and cell models; we found no human study.
Source: the studies listed in the next section; FDA reviews of 2024 and 2026; searches of 21 Nov 2026.
Is any of them an approved medicine? Not in the US. FDA’s reviews state that BPC-157, KPV and thymosin alpha 1 are not components of any FDA-approved drug, and Drugs@FDA lists no product containing LL-37 or GHK-Cu (searched 21 November 2026). Thymosin alpha 1 is the exception abroad. Italy’s medicines agency lists Zadaxin (timosina alfa 1) as authorised and prescription-only, and FDA’s 2024 review describes that authorisation as “an adjuvant to influenza vaccination in immunocompromised individuals”. The same review records approval in India, for chronic hepatitis B in adults with compensated liver disease, and in Hong Kong. Research-grade material labelled thymosin alpha 1 is not any of those medicines.
02What did each study measure, and in which model?
Each peptide, the strongest study we found in an inflammation-related setting, and what that study cannot tell you. Dated 21 November 2026. Tiers follow our evidence grades; sources are at the end of the page.
| Peptide | What it is | Strongest study in an inflammation-related setting | What was measured | Tier | What we don’t know |
|---|---|---|---|---|---|
| Thymosin alpha 1 (thymalfasin) | A 28-amino-acid peptide that FDA describes as physiologically present in the human body; thymalfasin is the synthetic version | TESTS, 2025: phase 3, double-blind, placebo-controlled; 1,106 adults with sepsis at 22 centres in China | Death from any cause within 28 days | T1 | Any effect in healthy people: the trials described here enrolled people with sepsis or chronic hepatitis B |
| LL-37 | A 37-amino-acid peptide the body makes, cut from a 170-amino-acid precursor protein, cathelicidin | Mahlapuu et al., 2021: phase IIb, randomised, double-blind; 148 adults with hard-to-heal venous leg ulcers. Miranda et al., 2023: randomised, double-blind; 25 adults with mildly infected diabetic foot ulcers | Ulcer closure measures (2021); two inflammatory markers and bacteria in wound fluid, and a wound granulation index (2023) | T1, small | Which way LL-37 moves inflammation: in the 2023 trial the two markers did not fall, and other research describes LL-37 as driving inflammation |
| GHK-Cu | A three-amino-acid peptide, glycyl-histidyl-lysine, bound to copper | Mulder et al., 1994: multicentre, randomised, evaluator-blinded, placebo-controlled trial of a GHK-Cu gel on chronic skin wounds. Miller et al., 2006: randomised; skin-care regimens with or without GHK-Cu after laser resurfacing around the mouth; 13 people completed it | Wound closure and wound infections (1994); redness after the laser, by computer analysis and blinded evaluators (2006) | T1; T3 and T4 for the disease models | Any effect on inflammation in people: the 2006 trial found no significant difference in how soon redness resolved on objective measures, the 1994 abstract reports no inflammation measure, and the disease-model studies used animals and cells |
| BPC-157 | A synthetic peptide of 15 amino acids | Ruenzi et al., 2005, a conference abstract: multicentre, randomised, double-blind, placebo-controlled; 53 patients, 46 completed | A disease-activity score that the abstract did not clearly define | T1 by design; known only from an abstract | Whether it changes inflammation in people: FDA staff found the abstract’s data inadequate to support efficacy or safety |
| KPV | A three-amino-acid peptide, lysine–proline–valine | Dalmasso et al., 2008: mice with two chemically induced forms of colitis, and human cell lines. Kannengiesser et al., 2008: mice with two forms of experimental colitis | Colon tissue under the microscope, cytokine gene activity, body weight and myeloperoxidase activity; inflammatory signalling in cells | T3, T4 | Any effect in people: we found no published human study or registered trial (21 Nov 2026) |
03Which peptides were tested in which models?
Three groups, by the length of the chain. In the drawings, each bead is one amino acid.
KPV and GHK-Cu
Two peptides of three amino acids each; GHK-Cu also carries a copper ion. Their inflammation-model studies used mice, zebrafish larvae and cells. GHK-Cu also has randomised trials on skin wounds; we found no human study of KPV.
BPC-157
A synthetic peptide of 15 amino acids. Its gut studies used rats; one placebo-controlled trial in people is known only from a conference abstract.
Thymosin alpha 1 and LL-37
Chains of 28 and 37 amino acids. Both have been tested against placebo in people: in sepsis, chronic hepatitis B and chronic ulcers.
More on each: thymosin alpha 1, LL-37, BPC-157, GHK-Cu and KPV.
04What have studies in people measured?
Four of the five have been given to people in a study with an inflammation-related measure or setting. We report what each study measured and whether it met its main goal.
Thymosin alpha 1: deaths in sepsis (T1, n=1,106). The TESTS trial randomised 1,106 adults with sepsis at 22 centres in China between 2016 and 2020 and analysed 1,089. Death from any cause within 28 days occurred in 23.4% of the thymosin alpha 1 group and 24.1% of the placebo group (hazard ratio 0.97, 95% CI 0.76 to 1.24, as corrected in May 2025). The authors found “no clear evidence” that it lowered 28-day mortality, and no secondary or safety outcome differed significantly. The trial’s declared support included SciClone Pharmaceuticals; Italy’s medicines agency lists a SciClone company as the holder of Zadaxin’s authorisation. An older placebo-controlled trial in 97 adults with chronic hepatitis B (1999) did not reach statistical significance on its main measure (14% versus 4%, P=0.084).
What FDA concluded about thymosin alpha 1. FDA’s 2024 review of its proposed uses, sepsis among them, concluded that there is “a lack of evidence to support the effectiveness” for all of them; that review predates the TESTS paper. On safety, the review says that in most clinical studies it “has been found to be well-tolerated”, and also that its use in compounding “may raise safety concerns”, because as a 28-amino-acid peptide given by injection it “may pose a significant risk for immunogenicity”, an immune reaction against the peptide itself.
LL-37: ulcer trials (T1, small). LL-37 has been applied to chronic ulcers in randomised trials, and company scientists are among the authors of the first two. In a 2014 trial in 34 adults with hard-to-heal venous leg ulcers, healing-rate measures were higher than placebo at the two lower strengths tested, significantly so only at the lowest, and no different at the highest. In the largest described here, a phase IIb trial of 148 adults (2021), the analysis of all patients “did not identify any significant improvement in healing” compared with placebo; its authors reported better results in a subgroup with large ulcers, chosen after the fact. A 2023 trial in 25 adults with mildly infected diabetic foot ulcers measured two inflammatory markers, IL-1α and TNF-α, in wound fluid. The markers rose in both groups, and LL-37 cream did not lower them compared with placebo cream; a wound granulation index rose more with LL-37. FDA’s compounding page says the agency “lacks sufficient safety-related information regarding cathelicidin LL-37 to know whether the drug would cause harm when administered to humans”, and adds: “Nonclinical research findings suggest detrimental effects on male reproduction and that this drug can be protumorigenic in some tissues.” Protumorigenic means able to promote tumour growth.
LL-37 on the other side of inflammation (T2, T3, T4). In 2007, researchers working with psoriatic skin and human immune cells reported that LL-37 binds the body’s own DNA into structures that set off interferon release, and called it “the key factor” in that activation. The same year, another group found “abnormally high levels of cathelicidin” in the facial skin of people with rosacea, and reported that putting those cathelicidin fragments into mouse skin increased inflammation. In these studies LL-37 drives inflammation. They are mechanism studies in samples, cells and mice, not tests of giving LL-37 to people.
BPC-157: one abstract and one unpublished study. Reviewing BPC-157 in 2026, FDA staff found a single meeting abstract of a multicentre, randomised, double-blind, placebo-controlled study in 53 people with mild to moderate ulcerative colitis, of whom 46 completed it. Its primary measure was a disease-activity score that the abstract did not clearly define. Because key details were missing, FDA staff concluded that the data “are therefore inadequate to support the efficacy and safety” of BPC-157 in that setting. A second study is registered. NCT07752381 gave a commercial product containing BPC-157, made by the study’s sponsor, to 40 physically active adults with no control group. Its primary measures are two blood markers of inflammation, hs-CRP and IL-6, and self-reported swelling. It was completed in November 2025, has posted no results, and we found no publication (ClinicalTrials.gov and PubMed, 21 November 2026). FDA staff also found “insufficient clinical safety information to characterize the safety profile” of BPC-157, and wrote that as a peptide “administered through a parenteral or nasal” route it “may pose a significant risk for immunogenicity”. More: BPC-157 human studies.
GHK-Cu: trials on skin wounds (T1). A 1994 multicentre, randomised, evaluator-blinded, placebo-controlled trial applied a GHK-Cu gel to foot ulcers in people with diabetes, alongside standard wound care, and measured wound closure and ulcer infections; its abstract reports no inflammation measure. The randomised trial we found that tracked redness, a sign of inflammation (21 November 2026), randomised people having laser resurfacing around the mouth to skin-care regimens with or without GHK-Cu; 13 completed it. It tracked redness by computer analysis and blinded evaluators, and assessed wrinkles, skin appearance and a patient questionnaire at 12 weeks. On objective measures it found no statistically significant difference between groups: not in how soon redness resolved, and not in evaluators’ ratings of wrinkles or skin quality.
05What have animal and cell studies measured?
KPV, BPC-157 and GHK-Cu have inflammation-model studies in rodents, zebrafish and cells. We describe the models and what was measured, not the authors’ findings. Animal and cell results are not evidence of an effect in people.
KPV (T3, T4). In 2008, researchers at Emory University exposed two human intestinal cell lines and a human T-cell line to inflammatory signals, with and without KPV, and gave KPV to mice with two chemically induced forms of colitis. They measured inflammatory signalling in the cells, the cytokines the cells released, and inflammation in the mouse colon under the microscope and by cytokine gene activity. The same year, a group at the University of Münster gave KPV to mice with chemically induced colitis and to mice whose colitis was set off by transferred immune cells, and tracked body weight, colon tissue under the microscope and myeloperoxidase activity in colon tissue. FDA staff wrote in 2026 that the molecular targets of KPV “remain unknown”. They found no clinical studies or human exposure data by any route, and concluded that “potential safety risks associated with the use of KPV in humans are unknown”. HHS read the record differently: its 2026 policy appendix says KPV’s “historical use is now documented and quantified”, and that the safety conclusion of FDA’s drug centre “reflects an absence of information rather than an observed signal”. Use is not a study.
BPC-157 (T3). The gut studies are in rats, and the two cited here come from one research group at the University of Zagreb. A 2013 study used rats given cysteamine to induce colitis and rats with a surgical join in the colon, examined between days 3 and 14. A 2017 study tied off the blood supply to a segment of rat colon and recorded how the blood vessels filled, along with tissue levels of a marker of oxidative stress and of nitric oxide. FDA’s 2026 review describes further rat models, including surgically made fistulas whose size was assessed by eye. FDA staff cautioned that these studies had not established how effects change with the amount given, that BPC-157’s molecular targets “have not been identified” and that its mechanisms of action “remain poorly understood”.
GHK-Cu (T3, T4). The disease-model studies we found gave GHK-Cu to mice whose lungs had been inflamed by a bacterial toxin (2016), cigarette smoke (2022), an allergen (2023) or silica dust (2024), and to mice with chemically induced colitis (2025). A 2026 study used zebrafish larvae. They measured cytokines such as TNF-α and IL-6, inflammatory cells in tissue and tissue changes under the microscope. A 2012 cell study measured IL-6 released by human skin fibroblasts. Gut research in context: KPV and BPC-157 gut studies.
06What has not been shown about peptides and inflammation?
An effect on inflammation in healthy people
The published human studies above enrolled people with sepsis, chronic hepatitis B, chronic ulcers or colitis, or people having a laser procedure. The one study in physically active adults had no control group and has posted no results.
That animal and cell findings carry over to people
We found no human study of KPV (PubMed and ClinicalTrials.gov, 21 Nov 2026). The disease-model studies of GHK-Cu and BPC-157 used rodents, zebrafish and cells.
Which way LL-37 moves inflammation
In a 25-person ulcer trial, two inflammatory markers did not fall with LL-37. In psoriasis and rosacea research, LL-37 is described as driving inflammation.
Whether they cause harm in people
FDA’s compounding page (current as of 22 April 2026) says the agency lacks the information to know whether BPC-157, KPV or LL-37 would cause harm when given to humans.
How these tiers are assigned: evidence grades. Why we publish no amounts for any compound: our policy.
This is a plain-English summary of public documents and studies, checked on 21 Nov 2026. It is not medical or legal advice.