Corneal haze development after PRK is regulated by volume of stromal tissue removal

Purpose. To determine whether excimer laser transepithelial photoablation can reduce the initial keratocyte loss seen after manual epithelial debridement. Second, to establish the relationship between initial depth of keratocyte and stromal loss and the subsequent development of corneal haze. Methods. Five rabbits received a 5-mm diameter monocular epithelial debridement by manual scraping. An additional five rabbits received a 5-mm diameter excimer laser transepithelial photoablation to a preset (intended) depth of 60 μm to ensure complete epithelial removal and to generate a superficial stromal keratectomy in all corneas. At various times during a 3-month period, animals were evaluated by in vivo confocal microscopy through focusing (CMTF), which generates a quantitative image intensity depth profile of the cornea that provides measurements of (i) depth of keratocyte loss, (ii) epithelial and stromal thickness, and (iii) backscattered light from the anterior cornea as an objective estimate of corneal haze. Results. Manual epithelial debridement was associated with an initial loss of anterior stromal keratocytes to a depth of 108 ± 14 μm that was followed by repopulation with migratory keratocytes. These cells showed increased reflectivity producing significant backscattering of light equivalent to clinical haze grade 1-2 (1,442 ± 630 U) at 3 weeks. Furthermore, repopulation occurred without detectable inflammation and was associated with a rapid restoration of normal keratocyte morphology and reflectivity. Transepithelial photoablation induced complete epithelial debridement in all corneas in addition to a superficial stromal keratectomy of 14-44 μm. Photoablation induced 36% less initial keratocyte loss (69 ± 19 μm) in the anterior stroma than manual debridement (p < 0.01) but was associated with intense concomitant inflammation. Photoablated corneas showed significantly more light backscattering (p < 0.01) compared with manually debrided corneas with a threefold increase at 3 weeks (4,397 ± 1,367 U) and a sixfold increase at 3 months (1,483 ± 1,172 compared with 234 ± 91 U). Backscattering of light or haze increased proportionally with increasing stromal keratectomy depth (r = 0.95, p < 0.001) but was unrelated to depth of induced keratocyte death. The increased backscatter in photoablated corneas appeared related to (i) a more pronounced keratocyte repopulation response with a higher density and reflectivity of migratory fibroblasts and (ii) myofibroblast transformation after repopulation. Conclusions. Excimer laser transepithelial photoablation induced significantly less keratocyte loss than manual epithelial debridement; however, photoablation was followed by a more intense inflammatory response and a greater increase in backscattering of light (haze) that was associated with increased keratocyte activation and myofibroblast transformation. Most important, the magnitude of corneal wound repair and the development and duration of corneal haze increased proportionally with increasing stromal photoablation depth (i.e., the volume of stromal tissue removal) but were unrelated to depth of initial keratocyte loss.