KGF-1

The native human growth factor FGF-7, secreted by fibroblasts to drive epithelial repair through the FGFR2b receptor; its N-truncated recombinant form palifermin is FDA approved for severe oral mucositis, while KGF-1 itself remains a research reagent.

Also referenced as: Keratinocyte Growth Factor, Keratinocyte Growth Factor 1, KGF, FGF-7, FGF7, Fibroblast Growth Factor 7, rhKGF, Palifermin, Kepivance, sh-Polypeptide-3

Also appears in: Dermal

What is KGF-1?

Keratinocyte growth factor 1 (KGF-1) is the protein encoded by the human FGF7 gene — the two names describe one molecule, so KGF-1 and fibroblast growth factor 7 are interchangeable. It is translated as a 194-residue precursor with a 31-residue signal sequence, secreted as a 163-residue N-glycosylated single chain, and ran at roughly 28 kDa when it was first purified to homogeneity from the conditioned medium of a human embryonic lung fibroblast line (Rubin et al., Proceedings of the National Academy of Sciences, 1989;86(3):802–806). That makes it a folded, glycosylated protein — a biologic — rather than the short synthetic chain this market usually means by “peptide.”

The reason KGF-1 appears in peptide catalogs at all is palifermin, and the two are not the same thing. Palifermin (brand name Kepivance) is a truncated recombinant version: a 140-amino-acid, 16.3 kDa, non-glycosylated protein produced in E. coli, in which the first 23 N-terminal amino acids of the mature chain were deleted to improve protein stability. That engineered molecule — not KGF-1 as the body makes it — is what the FDA approved in December 2004, and only for one narrow transplant setting. Native full-length KGF-1 has never been approved as a drug anywhere; it reaches buyers as a cell-culture reagent sold for laboratory use and, in cosmetics, as a topical ingredient commonly listed under the INCI name sh-Polypeptide-3.

How it works

• A single, epithelium-restricted receptor. KGF-1 signals almost exclusively through FGFR2b (the “KGF receptor”), an alternatively spliced variant of FGFR2 whose IIIb exon choice determines which FGF ligands it will bind — the finding that explained how one gene encodes two receptors with different ligand preferences: the KGF receptor bound both KGF and acidic FGF, while the alternative FGFR-2 transcript bound basic and acidic FGF with no detectable KGF binding at all (Miki et al., Proceedings of the National Academy of Sciences, 1992;89(1):246–250). Per the Kepivance label, that receptor is present on epithelial cells across tongue, buccal mucosa, esophagus, stomach, intestine, salivary gland, lung, liver, pancreas, kidney, bladder, mammary gland, skin including hair follicles and sebaceous glands, and the lens of the eye — and is reported absent from cells of the hematopoietic lineage. That absence is the entire logic of the approved use: the drug can push epithelial recovery during myeloablative therapy without acting on the malignant blood cells being treated.
• A one-way mesenchymal-to-epithelial signal. KGF-1 is made by fibroblasts and other mesenchymal cells but acts only on epithelium. In the original characterization it stimulated DNA synthesis in quiescent keratinocytes more than 500-fold while showing no mitogenic activity on fibroblasts or endothelial cells (Rubin et al., 1989), and the cloned sequence placed it in the FGF family as a paracrine effector of epithelial growth (Finch et al., Science, 1989;245(4919):752–755). Endogenous KGF is upregulated in response to epithelial injury.
• Cytoprotection, not only proliferation. In the mucositis literature the mechanism is described as pleiotropic: inhibition of epithelial apoptosis and DNA damage, upregulation of detoxifying enzymes, downregulation of pro-inflammatory cytokines, and enhanced epithelial migration, proliferation, and differentiation (Blijlevens & Sonis, Annals of Oncology, 2007;18(5):817–826). Animal work behind the label showed increased Ki67 staining and greater tissue thickness in tongue, buccal mucosa, and gastrointestinal tract.
• A genuine but narrow role in hair follicles. Mice engineered to lack KGF develop matted fur, with the defect restricted to the cells that give rise to the hair shaft — while epidermal growth and wound healing remained entirely normal, even in KGF/TGF-α double knockouts (Guo et al., Genes & Development, 1996;10(2):165–175). Recombinant KGF also induced dose-dependent hair growth in athymic nude mice and reduced chemotherapy-induced alopecia in a rat model (Danilenko et al., American Journal of Pathology, 1995;147(1):145–154). These are rodent results; the hair-growth marketing built on them has not been matched by human efficacy data.

Research status

Every meaningful human result attached to this molecule comes from palifermin, given intravenously in a hospital. There are no human efficacy trials of native, full-length KGF-1 for recovery, hair, or skin in healthy people.

The approved indication and its pivotal trial. In a double-blind study of 212 patients with hematologic cancers receiving fractionated total-body irradiation plus high-dose chemotherapy and autologous stem-cell transplantation, 106 received palifermin at 60 µg/kg/day for three days before conditioning and three days after transplant, and 106 received placebo. WHO grade 3 or 4 oral mucositis occurred in 63% on palifermin versus 98% on placebo; grade 4 in 20% versus 62%; median opioid use fell from 535 to 212 mg of morphine equivalents, and total parenteral nutrition from 55% to 31% (Spielberger et al., New England Journal of Medicine, 2004;351(25):2590–2598). The FDA approved palifermin on that basis in December 2004, worded narrowly: to decrease the incidence and duration of severe oral mucositis in patients with hematologic malignancies receiving myelotoxic therapy in the setting of autologous hematopoietic stem cell support.

Two failures are printed in the drug’s own label. In a post-approval multiple myeloma study of 281 patients using melphalan 200 mg/m² as the conditioning regimen, severe mucositis occurred in 38% of the palifermin pre-post arm versus 37% on placebo — no difference — and treatment-emergent infections were significantly more common with palifermin (50% and 47% across the two dosing arms versus 25% on placebo). In an allogeneic transplant study, palifermin (n=77) did not reduce severe acute graft-versus-host disease compared with placebo (n=78, 16% versus 17%), and the incidence of WHO grade 3–4 mucositis was nominally higher on palifermin (81% versus 73%). Both results appear on the label as explicit “Lack of Efficacy” sections and as Limitations of Use.

Head and neck cancer: mucositis fell, patient experience did not. Two randomized placebo-controlled trials reported in the same 2011 issue both hit their mucositis endpoint and neither changed how patients actually felt. In definitive chemoradiotherapy for locally advanced disease, severe oral mucositis fell from 69% to 54% (P = .041) and its median duration shortened from 26 to 5 days, but opioid analgesic use, mouth and throat soreness scores, and chemoradiotherapy compliance did not differ, and no secondary endpoint held up after multiplicity adjustment; overall and progression-free survival were similar after a median 25.8 months (Le et al., Journal of Clinical Oncology, 2011;29(20):2808–2814). In postoperative radiochemotherapy, severe mucositis fell from 67% to 51% (P = .027) with duration shortened from a median 22.0 to 4.5 days, yet patient-reported soreness scores and treatment breaks again did not differ (Henke et al., Journal of Clinical Oncology, 2011;29(20):2815–2820). Palifermin was never approved for head and neck cancer.

Systematic review. A Cochrane review of 35 randomized trials in 3,102 participants made KGF its principal finding, and the strength of that finding depends entirely on the population. In adults undergoing transplant for hematological cancers — the approved setting — there might be a reduction in moderate-to-severe oral mucositis (RR 0.89, 95% CI 0.80 to 0.99; 6 studies; 852 participants), graded low-quality evidence, with a number needed to treat of 11; the reduction in severe mucositis in that same population had confidence intervals crossing no effect (RR 0.85, 95% CI 0.65 to 1.11). The evidence is stronger in the settings where palifermin is not approved: in adults receiving head and neck radiotherapy with cisplatin or fluorouracil, the reduction in severe mucositis was graded high-quality (RR 0.79, 95% CI 0.69 to 0.90; 3 studies; 471 participants; NNT 7), and in adults receiving chemotherapy alone for mixed solid and hematological cancers, moderate-to-severe mucositis fell on moderate-quality evidence (RR 0.56, 95% CI 0.45 to 0.70; 4 studies; 344 participants). The reviewers state they are confident KGF is beneficial in those latter two populations and explicitly less confident in the transplant population (Riley et al., Cochrane Database of Systematic Reviews, 2017;11:CD011990). The paradox is worth naming: the one indication that carries an FDA approval rests on the weakest tier of evidence in the review.

A harm signal outside oncology. The KARE trial randomized 60 patients with acute respiratory distress syndrome to palifermin 60 µg/kg daily for up to six days or placebo. The primary endpoint, oxygenation index at day 7, showed no benefit (P = 0.13). In outcomes not specified a priori, the KGF group had a median of 1 ventilator-free day versus 20 on placebo (P = 0.0002), a longer median duration of ventilation among survivors (16 versus 11 days, P = 0.002), a 28-day mortality of 31% versus 10% (nine versus three deaths; RR 3.2, 95% CI 1.0 to 10.7) that did not reach statistical significance (P = 0.054), and significantly more adverse events (OR 4.9, 95% CI 1.3 to 20.3, P = 0.008). The authors concluded KGF did not improve outcomes and might be harmful (McAuley et al., The Lancet Respiratory Medicine, 2017;5(6):484–491).

Other completed programs, all negative or inconclusive. A randomized ACTG trial of palifermin at 20, 40, or 60 µg/kg in 99 people with HIV and poor CD4 recovery on antiretroviral therapy found no significant change in CD4 count, thymus size, naive T cells, or recent thymic emigrants (Jacobson et al., Journal of Acquired Immune Deficiency Syndromes, 2014;66(4):399–406; NCT00376935). Long-term follow-up of a randomized trial in 100 allogeneic transplant recipients found no differences in CMV or invasive fungal infection, chronic GVHD, or long-term survival, concluding the benefit is limited to mucotoxicity (Levine et al., Biology of Blood and Marrow Transplantation, 2008;14(9):1017–1021). A more recent phase 1/2 study escalated palifermin to 720 µg/kg on day −7 in 31 unrelated-donor transplant recipients: no patient at that dose developed grade 2–4 acute GVHD (0/19), but the primary endpoint — severe chronic GVHD — was unchanged versus historical controls (Schulz et al., Blood, 2025;146(8):944–950; NCT02356159). Registered studies also exist in asthmatic airway permeability (NCT01386151, completed 2011) and toxic epidermal necrolysis (NCT02037347, terminated after enrolling one patient).

Hair and skin: almost nothing registered. The single registered human study directly relevant to the cosmetic pitch is a 28-participant early-phase study of a topical KGF hair serum for preventing chemotherapy-induced alopecia (NCT04554732, University of Arizona, completed June 2022), with no results posted at the time of writing. A separate laser-plus-growth-factor alopecia study (NCT04882969) was withdrawn without enrolling anyone.

Regulatory footprint. Kepivance remains marketed in the United States under BLA 125103, now held by Swedish Orphan Biovitrum. The European Commission withdrew its EU marketing authorization on 1 April 2016 at the request of the marketing authorization holder, which had decided to permanently discontinue marketing there for commercial reasons — an availability decision, not a safety action.

Common dosage forms

• Prescription biologic (palifermin). Kepivance is supplied as a preservative-free lyophilized powder in single-dose vials containing 5.16 mg, reconstituted with 1.2 mL of Sterile Water for Injection to a 5 mg/mL solution and given as an intravenous bolus in a transplant unit. This is a hospital-administered product with no consumer format.
• Life-science reagent. Recombinant human KGF/FGF-7 is sold lyophilized in microgram fills — commonly 10 µg to 1 mg — most often expressed in E. coli, labeled for laboratory research use only, and supplied with reconstitution instructions and carrier-protein storage guidance for cell culture. These constructs are usually full-length (roughly 163–164 residues), not the 140-residue palifermin molecule.
• Topical cosmetics. Serums, creams, and scalp or hair formulas listing sh-Polypeptide-3 at parts-per-million levels, frequently blended with other recombinant growth factors so the KGF content is not stated separately.
• Multi-ingredient “growth factor” blends, in which KGF appears alongside EGF, IGF-1, or bFGF analogs under their own INCI names.

This section describes formats only and is not dosing guidance.

Key considerations

• The approval belongs to palifermin, in one setting, and the label says where it failed. Kepivance is approved only to decrease severe oral mucositis in hematologic malignancy patients receiving myelotoxic therapy with autologous stem-cell support. Its Limitations of Use state that safety and efficacy are not established in non-hematologic malignancies, that it was not effective in the allogeneic transplant setting, and that it is not recommended with melphalan 200 mg/m². Marketing that reads “FDA approved” onto native KGF-1 is transferring a different molecule’s authorization to a different product, route, population, and purpose.
• The tumor-growth question is the central safety issue. FGFR2b is expressed on epithelial cells across most organs, and the label carries “Potential for Stimulation of Tumor Growth” as its only Warning and Precaution. Nonclinical data behind it: palifermin enhanced growth of human epithelial tumor cell lines in vitro at ≥10 µg/mL, and in nude mouse xenografts at 25- and 67-fold the human dose it produced a dose-dependent increase in growth rate in 1 of 7 KGF-receptor-expressing tumor lines. The effect in patients with receptor-expressing non-hematopoietic tumors is stated as not known.
• Timing and context change the result — sometimes for the worse. The label prohibits administration within 24 hours before, during, or within 24 hours after myelotoxic chemotherapy because a clinical trial found doing so increased the severity and duration of oral mucositis. Heparin raises systemic exposure, requiring line flushes. Infections were more frequent in the melphalan study, and the ARDS trial’s mortality and ventilator-free-day signals point the same direction: a proliferation signal given in the wrong context is not neutral.
• Documented adverse reactions are common, not rare. In controlled trials versus placebo: rash 62% versus 50%, elevated serum amylase 62% versus 54% (mainly salivary in origin), fever 39% versus 34%, pruritus 35% versus 24%, erythema 32% versus 22%, edema 28% versus 21%, mouth or tongue thickness or discoloration 17% versus 8%, altered taste 16% versus 8%, and dysesthesia 12% versus 7%. A post-marketing safety study found numerically more cataracts on palifermin. Animal reproduction studies showed embryo-fetal mortality and growth alterations, and fertility studies showed decreased epididymal sperm counts, so the label warns of possible fetal harm and impaired fertility. Anti-drug antibodies appeared in 12 of 645 patients (2%), none neutralizing, with unknown clinical significance including possible cross-reactivity with other FGF family members.
• The naming is genuinely confusing, in three directions. KGF-1, KGF, and FGF-7 are one molecule; KGF-2 is a different protein entirely — FGF-10, developed as repifermin, whose program ended without approval: its ulcerative colitis trial was negative (clinical remission 0–19% across five dose groups versus 11% on placebo) and its mucositis work never advanced past a small phase 1/2. KGF-2 carries its own separate cosmetic INCI designation, distinct from the sh-Polypeptide-3 used for KGF-1. Palifermin is a 140-residue N-truncated, non-glycosylated E. coli construct, not the 163-residue glycosylated human protein, so literature on one is not automatically evidence about the other. Searches on the bare abbreviation “KGF” also collide with unrelated keratin and keratinocyte literature.
• Identity is hard to verify for a protein of this size. A 163-residue glycoprotein cannot be checked the way a short synthetic peptide can: purity by mass alone says nothing about correct folding, glycosylation, or biological activity, which requires a cell-based assay. Reagent-grade material is also sold with carrier proteins and buffers intended for culture work, and both full-length and N-truncated constructs circulate under the same “KGF” label.