Corneal Stromal Remodeling after Photorefractive Keratectomy
MetadataShow full item record
The cornea is an optically clear tissue that contributes 2/3 of the eye's refractive power, making it a target for vision correction procedures. A small percentage of patients who receive corneal surgery experience loss of corneal transparency (haze). Haze occurs when stromal cells in the cornea (keratocytes) become activated and transform into fibroblasts or myofibroblasts, which can generate contractile forces that may disrupt the collagen architecture. We investigated the wound healing process following photorefractive keratectomy (PRK, a clinical vision correction procedure) using 3-D imaging both in vivo and in situ. We hypothesized that following PRK in rabbits, keratocytes located within the injured stroma, where collagen is intact, would align with the collagen lamellae and will not produce fibrosis. In contrast, we expected keratocytes anterior to the wound, where collagen is disrupted, to undergo myofibroblast transformation and produce significant haze. Twelve New Zealand white rabbits were scanned one week before surgery using an in vivo confocal microscope. PRK was performed on the right eyes of the rabbits (9 diopter spherical photocorrection) with subsequent scans at 7 days, 21 days, 3 months, and 6 months. Custom software was used to build 3-D reconstructions and measure stromal thickness and haze. Some rabbits were sacrificed at each time point to obtain in situ confocal images. Thickness measurements at pre-scan, 7 days, 21 days, 3 months, and 6 months were: 335, 208, 265, 293, 313um. The decrease and subsequent increase in thickness is consistent with removal of tissue for PRK, followed by tissue regeneration. Haze measurements at the same time points were: 1797, 4453, 6906, 3212, 3433 intensity units. The increase and subsequent decrease in backscatter suggests two phases of wound healing. In situ confocal imaging showed that cells within the native stroma were aligned with collagen lamellae without prominent stress fibers at 7 and 21 days, while cells anterior to the injured stroma at 21 days aligned randomly and displayed prominent stress fibers. At 3 months, these cells aligned with correlation to the collagen and did not express stress fibers. At 6 months, the cell/collagen arrangement was similar to uninjured tissue. Our results suggest that the collagen lamellae direct fibroblast patterning during repopulation of the native stroma, without inducing fibrosis or significant haze. In contrast, cells accumulating on top of the stroma initially align randomly and produce hazy, fibrotic tissue. Remarkably, over time, cells remodel the fibrotic tissue to produce a lamellar structure that is similar to the native corneal stroma.