Home › Review — modulators of corneal healing › Emerging therapies in corneal wound healing: cells, genes, exosomes
Emerging therapies in corneal wound healing: cells, genes, exosomes
The therapies gathered here are distinguished not by novelty — some are twenty years old — but by the fact that they target the cause rather than the consequence. Replacing a lost cell population, correcting a gene, restoring innervation: where the previous families accompany healing, these claim to restore its conditions.
Limbal stem cell deficiency is the most accomplished example. It is a disease in which the cornea can no longer repair itself because the stem cell reservoir has been destroyed — chemical burn, aniridia, Stevens-Johnson. No eye drop answers it. HOLOCLAR, the first authorised stem cell therapy in Europe, answers it by culturing the remaining limbal cells of the healthy eye; CALEC is its North American counterpart, under evaluation since 2025.
The other avenues are at very different stages, and this should be said plainly: connexin 43 modulation and the substance P–IGF-1 combination have preliminary clinical results; exosomes and senolytics have preclinical data only.
Overview
| Agent | Mechanism | Level of evidence |
|---|---|---|
| Cultivated limbal stem cells (CLET) | Limbal biopsy (1–2 mm²) expanded ex vivo on fibrin; the graft must contain an adequate fraction of p63-bright stem cells… | Oxford II Moderate–Strong Present (EU) · MA |
| CALEC | 2 mm limbal biopsy from the healthy eye → two-step expansion (plastic culture followed by de-epithelialised amniotic mem… | Oxford II–III Investigational (→ Moderate) Future (USA) |
| Connexin 43 modulation | Injury upregulates Cx43 at the wound edge → ATP release via hemichannels, inflammation, impaired epithelial migration. | Oxford II (Nexagon) · V (corneal aCT1) Moderate (Nexagon) Future · pivotal trial ongoing |
| Mesenchymal stem cells (MSC) & derived exosomes | Effect is essentially paracrine (no engraftment): anti-inflammatory signalling (↓IL-1, TNF-α | Oxford V (preclinical) · IV (rare human) Investigational Future |
| Corneal gene therapy | The avascular, immune-privileged and accessible cornea = an ideal target. | Oxford V (preclinical) Investigational Future · frontier |
| Substance P (FGLM-NH₂) + IGF-1 (SSSR) | Synergy is essential (neither agent alone is effective): substance P (FGLM-NH₂) + IGF-1 (SSSR) cooperatively stimulate e… | Oxford IV Weak–Investigational Future · niche |
| PEDF + DHA | PEDF binds its epithelial receptor (PEDF-R/PNPLA2) → DHA release → conversion to Neuroprotectin D1 (NPD1) and docosanoid… | Oxford V (preclinical) Investigational Future · early-stage |
| Senolytics · Platelet lysate (cord blood) | Oxford V / IV Investigational / Weak Future / emerging |
Cultivated limbal stem cells (CLET) — HOLOCLAR®
Autologous corneal epithelial cells expanded ex vivo · Holostem/Chiesi · first stem-cell-based advanced therapy medicinal product (ATMP) approved in the EU
Limbal biopsy (1–2 mm²) expanded ex vivo on fibrin; the graft must contain an adequate fraction of p63-bright stem cells (release criterion ≥3%), a validated predictor of success. Reconstitutes the limbal niche → self-renewing epithelial phenotype, arrest of conjunctivalisation/neovascularisation.
IndicationsModerate to severe limbal stem cell deficiency (LSCD), uni-/bilateral, secondary to ocular burn (physical/chemical), with superficial neovascularisation ≥2 quadrants and central involvement. Restores the surface before/for subsequent keratoplasty.
Route & dosingSurgical implantation of a single graft (~3.8 cm², ≥79,000 cells/cm²) on a debrided cornea. Single ATMP procedure; re-grafting possible.
Level of evidenceOxford II (prospective/registration cohort; no large RCT — surgical ATMP) · Scale: Moderate–Strong (the strongest evidence base in this section). Success ~66–76%.
Detailed sheet →CALEC — cultivated autologous limbal epithelial cells (Mass Eye and Ear)
Xenobiotic/serum/antibiotic-free GMP process · U. Jurkunas · US analogue of Holoclar
2 mm limbal biopsy from the healthy eye → two-step expansion (plastic culture followed by de-epithelialised amniotic membrane, 13–27 days), graft of 0.4–1.5×10⁶ cells. Restores the limbal stem cell population. Novel feature: an entirely xenobiotic/serum/antibiotic-free process (regulatory advantage).
IndicationsUnilateral LSCD with a healthy contralateral donor eye.
Route & dosingSingle surgical graft on an amniotic membrane scaffold (investigational).
Level of evidenceOxford II–III (single-arm phase I/II, small n) · Scale: Investigational but promising. Nat Commun, 4 Mar 2025 (NCT02592330): complete/partial success 93% at 12 months, 92% at 18 months; no serious AEs.
Detailed sheet →Connexin 43 modulation — Nexagon (lufepirsen) & aCT1 peptide (Granexin)
Cx43 antisense oligonucleotide (Amber Ophthalmics) · Cx43 C-terminal mimetic peptide (Xequel Bio) · the most advanced innovative candidate
Injury upregulates Cx43 at the wound edge → ATP release via hemichannels, inflammation, impaired epithelial migration. Nexagon (antisense) transiently reduces Cx43 translation → ↓inflammation, ↑re-epithelialisation. aCT1/Granexin (peptide) binds the C-terminal tail of Cx43 (and ZO-1) → junctional remodelling, ↓inflammation.
IndicationsNon-infectious PED (incl. post severe chemical/thermal burn — Nexagon); impaired diabetic wound healing (peptide, preclinical). ~100,000 PED patients/year in the USA, with no approved treatment.
Route & dosingNexagon: topical ophthalmic gel, low in-office application frequency ("as few as 5" applications, weekly up to week 8 if needed). aCT1: topical (preclinical in the cornea).
Level of evidenceNexagon: Oxford II (positive phase 2) · Scale: Moderate (investigational). Pivotal trial NEXPEDE-1 (phase 2/3, NCT05966493) ongoing. Corneal aCT1: Oxford V / Investigational.
Detailed sheet →Mesenchymal stem cells (MSC) & derived exosomes
MSC (bone marrow, adipose tissue, cord) and extracellular vesicles (MSC-EV) · fast-moving front
Effect is essentially paracrine (no engraftment): anti-inflammatory signalling (↓IL-1, TNF-α ; ↑IL-10), immunomodulatory, anti-apoptotic, anti-fibrotic; exosomal cargo (miRNA/proteins) promotes keratocyte proliferation and epithelial migration and reduces scarring (↓α-SMA). Hydrogel/microneedle carriers for sustained delivery.
IndicationsBurns, PED, prevention of fibrosis/haze, adjunct to LSCD treatment, stromal regeneration.
Route & dosingInvestigational — eye drops, subconjunctival/intrastromal injection, scaffold/hydrogel. No standardised dose.
Level of evidenceOxford V (mostly preclinical) · Scale: Investigational. The FDA has issued warnings about unproven exosome products.
Detailed sheet →Corneal gene therapy
Viral (AAV, lentivirus) / non-viral vectors; gene augmentation, RNAi/antisense, CRISPR editing
The avascular, immune-privileged and accessible cornea = an ideal target. Strategies: anti-fibrotic/anti-angiogenic genes (decorin, sFlt-1, PEDF) ↓scarring/neovascularisation; correction of dystrophies (TGFBI allele-specific silencing, endothelial augmentation); CRISPR editing of dominant mutations. Strong endothelial tropism of AAV6.
IndicationsFibrotic scarring after injury/surgery, neovascularisation, hereditary dystrophies (TGFBI, CHED/Fuchs), graft survival.
Route & dosingInvestigational — intrastromal injection, topical, anterior chamber, or ex vivo graft transduction. No human dosing established.
Level of evidenceOxford V (preclinical / ex vivo human corneas) · Scale: Investigational. No approved corneal product (contrast: retinal Luxturna).
Detailed sheet →Substance P (FGLM-NH₂) + IGF-1 (SSSR) — synergistic combination
Derivative peptides · Nakamura/Nishida (Osaka/Kyoto) · investigational/compounded
Synergy is essential (neither agent alone is effective): substance P (FGLM-NH₂) + IGF-1 (SSSR) cooperatively stimulate epithelial migration and attachment (↑α5 integrins, fibronectin assembly), replacing the neuropeptide support lost in the denervated cornea.
IndicationsNeurotrophic keratopathy with PED (post-herpetic, diabetic, surgical/denervation, leprosy).
Route & dosingCombined eye drops several times/day (investigational/compounded; no standardised commercial formulation).
Level of evidenceOxford IV (open-label, uncontrolled series) · Scale: Weak–Investigational.
Detailed sheet →PEDF + DHA — corneal nerve regeneration
Pigment epithelium-derived factor (± 44-mer peptide) + docosahexaenoic acid · Bazan/Cortina (LSU)
PEDF binds its epithelial receptor (PEDF-R/PNPLA2) → DHA release → conversion to Neuroprotectin D1 (NPD1) and docosanoids → regeneration of corneal sensory nerves, restoration of sensitivity and CGRP⁺ nerve density, ↓inflammation → accelerated epithelial healing (neuro-regenerative mechanism).
IndicationsNeurotrophic states and impaired healing after nerve injury (post-LASIK/PRK, diabetic keratopathy).
Route & dosingTopical (investigational); no human dosing established.
Level of evidenceOxford V (animal + in vitro) · Scale: Investigational.
Detailed sheet →Senolytics · Platelet lysate (cord blood) — complementary approaches
Two distinct approaches, combined here for concision
Detailed sheet →Key points
Only one of these approaches is a treatment today: cultured limbal stem cells, in a precise and rare indication, with organisational requirements — limbal biopsy, cell therapy laboratory, grafting — that restrict it to a few centres.
The others belong to horizon-scanning, and it is worth knowing which to watch. Mesenchymal stem cell exosomes are the most active avenue in the literature: they reproduce the paracrine effect of stem cells without the cells, hence without the tumour risk or the regulatory burden of cell therapy. The next treatment will probably come from there, provided standardisation and scale-up are solved.
The other families in this review
Updated 26 August 2026