Purpose: To establish the main indicator of progressive corneal exposure to dry eye conditions in an animal model.
Methods: Cell surface covered by microvilli (SCM), intercellular junctions (IJs), cell area, cell shape, and cell shade were measured in 988 epithelial corneal cells from rabbit eyes that were kept open for different drying times (DT) between 0 and 3 h. Two multivariate logistic regressions were used to model the relationship between these morphometric cell characteristics and DT. To corroborate the results, terminal differentiation was assessed by cluster analysis.
Results: Scanning electron microscopy images of the epithelial cells showed typical desiccation changes that increased in frequency and intensity as DT were prolonged. Binomial logistic regression, to distinguish cells exposed to desiccation from the control cells, displayed SCM, IJ, and cell shade as significant indicators (p < 0.01) of desiccation exposure. Multinomial logistic regression analysis showed SCM, IJ, cell shade, and cell area as significant indicators (p < 0.01) differentiating the four levels of desiccation exposure, and SCM was the variable that showed elevated odds ratio in all the analysis. In addition, the cells were grouped in three groups (G1, G2, and G3) by cluster analysis. G2 cells increased in percentage as DT grew (p < 0.05, χ2 test) in detriment of G1 cells, while the percentage of G3 cells remained stable. These changes were consistent with a terminal differentiation process from G1 cells to G2 cells, which was the group with the lowest values of SCM (p < 0.001, Mann-Whitney test).
Conclusions: The variable SCM could be the best determining indicator of progressive corneal exposure to dry eye conditions. The development of clinical methods to analyze microvilli density could allow a more accurate assessment of the level of degradation of the corneal epithelium under evaporative conditions.