Association Between Neurotrophic Keratitis and Limbal Stem Cell Deficiency

PURPOSE: To investigate changes in corneal nerves, immune cells, and limbal epithelial stem cells (LESCs) in neurotrophic keratitis (NK) patients; to analyze correlations between these parameters and clinical manifestations; and to monitor long-term corneal changes in NK patients receiving recombinant human nerve growth factor (rh-NGF). DESIGN: Cross-sectional study with an interventional case series. METHODS: A total of 54 NK patients (55 eyes) and 57 healthy controls (59 eyes) were enrolled. In vivo confocal microscopy (IVCM) was performed to evaluate subbasal nerve fiber parameters, density of limbal epithelial stem cells (LESCs) in different directions, and dendritic cell (DC) status. Longitudinal changes in these parameters are recorded at baseline, weeks 4 and 8, and the 2-year follow-up in the rh-NGF-treated subgroup (7 eyes of 7 patients). Correlation analyses are conducted among all parameters in the NK group. RESULTS: In patients with neurotrophic keratopathy (NK), corneal nerve fiber density (CNFD) (P < .01), nerve fiber length (CNFL) (P < .01), and nerve branch density (CNBD) (P < .01) were significantly less than in healthy controls. Similarly, central BCD (P < .001) and peripheral LESCs density (P < .001) were markedly lower in NK group. In contrast, DCs density (P < .001) and activation ratio (P < .001) of patients with NK were higher. Correlation analyses revealed that CNFD was positively correlated with LESCs density (ρ = 0.46; P < .001), and negatively correlated with fluorescein staining scores (ρ = -0.46; P < .001) and DC density (ρ = -0.57; P < .001). LESC density was negatively correlated with both fluorescein staining scores (ρ = -0.62; P < .001) and mean DC area (ρ = -0.44; P < .001), whereas fluorescein staining scores were positively correlated with DC density (ρ = 0.44; P < .001). Compared with baseline, improvements in LESC density, DC density, activation, and fluorescein staining scores became observed at 4 weeks of rh-NGF treatment, whereas nerve regeneration became detectable after 8 weeks and showed significant recovery by year 2. CONCLUSIONS: This study demonstrates that patients with NK exhibit corneal nerve damage, limbal stem cell deficiency, and immune cell activation, with significant correlations observed among these alterations. Short-term treatment with rh-NGF may initiate corneal recovery by first improving the limbal microenvironment and modulating immune responses, thereby promoting a regenerative feedback loop that supports sustained nerve regeneration and epithelial repair.