Corneal wound healing after photorefractive keratectomy: a 3-year confocal microscopy study.

PURPOSE: To perform a sequential quantitative analysis of corneal wound healing after photorefractive keratectomy (PRK) by using confocal microscopy in vivo. METHODS: In a prospective, nonrandomized, comparative trial performed in an institutional setting, 24 eyes of 14 patients received PRK to correct refractive errors between -1.25 and -5.75 D. Central corneas were examined preoperatively and at 1 day, 5 days, and 1, 3, 6, 12, 24, and 36 months after PRK by using confocal microscopy. A masked observer randomly examined 3 to 6 confocal scans per eye per visit to determine epithelial and stromal thickness, keratocyte density in 5 anterior-posterior stromal layers, corneal nerve density in the subbasal region and the stroma, and corneal light backscattering (corneal haze). RESULTS: Epithelial thickness increased 21% (P < .001) by 12 months after PRK and thereafter remained unchanged to 36 months after PRK. There was no change in stromal thickness between 1 and 36 months after PRK (P = .35). The dense keratocyte population in the preoperative anterior 10% of the stroma (32,380 +/- 5,848 cells/mm3) that was partially or completely removed during photoablation was not reconstituted at 36 months in the anterior 10% of the post-PRK stroma (17,720 +/- 4,308 cells/mm3, P < .001). Subbasal nerve fiber bundle density was decreased 60% at 12 months after PRK (P < .001) before returning to densities at 24 and 36 months after PRK that were not significantly different from preoperative values (P = 1.0). Activated keratocytes and corneal haze peaked at 3 months after PRK. CONCLUSIONS: Wounding of the cornea by PRK alters the normal structure, cellularity, and innervation of the cornea for up to 36 months.