ADMSC-Seeded Fibrin Membrane Improved Histologic Repair After Rabbit Corneal Alkali Burn
Severe corneal alkali injury can trigger a cascade of epithelial loss, stromal inflammation, destructive neovascularization, and collagen disorganization that leaves the cornea opaque and structurally unstable. In a rabbit model, investigators tested whether a fibrin membrane loaded with adipose-derived mesenchymal stem cells could blunt that cascade and favor repair rather than scarring.
The experiment used 21 New Zealand White rabbits divided into three arms: a treatment group receiving fibrin membranes embedded with ADMSCs, a comparator group receiving acellular fibrin membranes, and an untreated control group. All animals sustained a standardized alkaline burn in the left eye on day 0. Allocation to the treatment groups occurred on day 14, after corneal neovascularization had already developed, which makes the study especially relevant to the difficult phase in which injury has moved beyond the initial chemical insult and into active tissue remodeling.
Epithelial Integrity Diverged Sharply From the Comparator Arms
The most striking terminal finding involved epithelial preservation. Epithelial degeneration was absent in all 7 rabbits treated with the ADMSC-seeded membrane, while every rabbit in the acellular membrane group and every rabbit in the untreated group showed epithelial degeneration. The overall three-group test was highly significant, and pairwise comparison confirmed a meaningful separation between the cell-seeded construct and the acellular scaffold.
Corneal surface restoration is not a cosmetic endpoint. Once epithelial continuity fails, the cornea remains exposed to persistent inflammatory signaling and proteolytic injury, conditions that favor stromal breakdown and melt. In this model, the presence of viable mesenchymal cells on a fibrin scaffold appears to have altered that surface biology in a way a bare membrane could not.
Corneal melting followed the same pattern. Only 1 of 7 rabbits in the ADMSC group developed corneal melting, compared with all rabbits in both comparator groups. For a chemically injured cornea, that difference points to a clinically meaningful reduction in tissue destruction, even though the study was designed as a terminal histologic experiment rather than a functional vision trial.
Neovascularization, Collagen Organization, and Stromal Inflammation Shifted in Favor of Cell Therapy
The treatment effect extended beyond the epithelium. Neovascularization differed significantly across the three groups, as did stromal inflammation, collagen organization, and vimentin-positive stromal cell staining. These findings fit the expected biology of mesenchymal stromal support: dampening inflammatory recruitment, modulating fibroblastic activation, and promoting a matrix environment that is less chaotic than the fibrotic response seen after severe alkali burn.
Neovascularization in the cornea is particularly damaging because the healthy cornea is normally avascular. Once new vessels invade, transparency declines and immune traffic increases, both of which can worsen chronic opacity. A therapy that lowers neovascular burden in this setting may be doing more than preserving appearance; it may be interrupting the structural remodeling that drives long-term visual loss.
Collagen organization also improved in the ADMSC arm, suggesting more orderly stromal architecture. In the cornea, alignment of collagen lamellae is central to optical clarity, and disorganization usually tracks with scarring and haze. Vimentin staining, a marker tied to stromal cellular activity and mesenchymal phenotype, also differed significantly across groups, reinforcing the impression that the seeded membrane altered the repair milieu rather than serving as a passive bandage.
The Biomaterial Scaffold Likely Mattered as Much as the Cells
Fibrin is not a neutral carrier. As a biologic scaffold, it can support cell adhesion, deliver local paracrine signals, and create a temporary matrix that helps injured tissue retain structure during repair. Seeding that scaffold with ADMSCs likely added another layer of activity, since mesenchymal stem cells are known in regenerative settings for immunomodulatory and trophic effects rather than direct tissue replacement alone.
That mechanism is plausible in a corneal burn model where the target problem is not just cell loss but dysregulated wound healing. A bare scaffold can cover the defect, yet it does not necessarily alter the inflammatory and angiogenic signals that continue to reshape the cornea. In this experiment, the cell-seeded membrane outperformed the acellular membrane on the most relevant histologic endpoints, which supports the idea that the living component contributed real biologic activity.
Preclinical Signal Is Strong, but Translation Needs More Than a 14-Day Snapshot
Several design features keep the findings in the preclinical lane. The observation window ended 14 days after treatment, and the outcomes were terminal ordinal histologic grades rather than serial molecular or functional measures. No uninjured normal-cornea comparator was included, so the study cannot define how close repair moved toward true physiologic normalization. Sample size was also small, with only 7 rabbits per group.
Those limitations matter because corneal repair is a long game. A therapy that looks effective at the level of histology in a short rabbit experiment still has to prove durability, safety, and reproducibility before any human translation can be considered. Longer follow-up would be needed to see whether the early gains in epithelial integrity and matrix organization persist or fade as remodeling continues. Molecular readouts could also clarify whether the observed benefits reflect reduced pro-inflammatory signaling, altered fibroblast behavior, or both.
Even with those caveats, the direction of effect was consistent across the major tissue readouts. The ADMSC-seeded fibrin membrane produced more favorable histologic repair than either the acellular scaffold or no treatment at all. For now, that positions the construct as a candidate regenerative strategy for chemically injured corneas, not a ready-made clinical intervention, and the next step is a larger study with longer observation and functional endpoints that matter at the bedside.