The aim was to determine the association of tear fluid cytokine levels and post-PRK corneal haze evaluated by in vivo confocal microscopy. In addition, the possible association between subbasal neural regeneration and haze formation, or epithelial regeneration were investigated. Twenty eyes of 20 patients (16 women and four men, age 30.7 +/- 7.5 years, range 21-48 years) underwent a myopic PRK. The spherical equivalent (SE) of the intended correction was -4.7 +/- 1.5 D (range -2.75 to -9.00 D). ELISA-methods were used to assess tear fluid concentrations of TGF-beta1, PDGF-BB and TNF-alpha pre-operatively, and post-operatively on day 2 and at 3 months. Tear fluid flow in the collection capillary was recorded, and rates of cytokine release (= tear fluid flow-corrected concentrations) were calculated. In vivo confocal microscopy was performed at 3 months to evaluate the corneal morphology and to determine numerical haze estimate. There was wide interindividual variation between pre-operative and post-operative concentrations and rates of release of TGF-beta1, PDGF-BB and TNF-alpha. Subepithelial haze was observed in all corneas and the mean haze estimate was 506 +/- 401 U (100-1410 U). However, no association was found between tear fluid cytokine levels and post-PRK haze. Regenerating subbasal nerve plexus was found in 18 out of 20 corneas; in two corneas it was absent or could not be visualized due to subepithelial haze. The density of the subbasal nerve fiber bundles had a positive correlation with the epithelial thickness (Pearson correlation, r = 0.56, P = 0.011), but not with the haze estimate or the thickness of the haze area. At 3 months post-PRK, haze could be observed in all patients. The results suggest that tear fluid cytokine analysis, as measured, may not be suitable for screening the potential candidates for haze formation. We did not find any correlation between haze and regeneration of subbasal nerve plexus, but we demonstrated that the regeneration of subbasal nerve plexus might have significant influence on regulation of epithelial healing.
Copyright 2001 Academic Press.