Background: In photorefractive keratectomy (PRK) procedures, a variable superficial central corneal scar formation ("haze") can be observed following removal of corneal stromal tissue. Today, a near to normal slit lamp finding is observed one year postoperatively in most patients. We employed the slit scanning confocal microscope to study the corneal morphology years after PRK.
Methods: We selected 5 patients, who had been subjected to unilateral photorefractive surgery 1-3 years earlier and who had no corneal haze upon slit lamp examination. As controls we investigated the non-treated corneas of these patients, 5 healthy controls and 5 contact lens wearers. The confocal microscopic investigation was performed with 25x, 40x and 50x water immersion objectives. The video signal was synconized with the slit scan and stored on S-VHS video tape. By reviewing the videos in the single frame mode, all corneal layers could be qualitatively evaluated.
Results: Some minor abnormalities were observed in the epithelium of all PRK-treated eyes. In the epithelial basal cell layer some round structures of about the size of a cell with high reflectivity were observed. These changes were only occasionally found in contact lens wearers, but not in non-treated or normal control eyes. Bowman's layer was absent in the PRK treated eyes, instead, a fine layer of collagen tissue of increased reflectivity was found. The subepithelial corneal nerve plexus was normal in all non-treated eyes, whereas in the PRK-treated corneas nerve shape and branching pattern were changed to quite an extent. In the anterior stroma the keratocyte nucleus patterns indicated an increased cell density and irregular spacing, whereas a normal keratocyte pattern was found in the deeper stromal layers. A significant finding was the observation of rod and needle shaped highly reflective structures, which were limited to the area of the excimer laser keratectomy with a predominance in the anterior stroma. These longitudinal structures themselves consisted of linearly arranged highly reflective granules, which sometimes also were found as isolated dots within keratocyte processes. In long term contact lens wearers a comparable granule type, however with a singular and scattered arrangement, was variably found in all corneal regions and layers. In normal controls none of these findings were present. In contact lens wearers and PRK patients with a contact lens history, the corneal endothelium showed some degree of polymegathisms but no other specific findings.
Discussion: Up to now, refractive surgery with the excimer laser has been reported to elicit no other stromal changes but a mild fibroblast activation with subsequent scar tissue formation. In clinically clear corneas after PRK, we have described a new type of stromal deposit observed 1-3 years after surgery. As acute wound healing responses might have been expected to have passed at this point, this highly reflective stromal deposit can be assumed to consist of linear keratocyte processes filled with some highly reflective (degenerative?) matter as well as a corresponding extracellular stromal deposit arranged parallel to the stromal collagen bundles. Possibly, these stromal deposits represent the result of an inflammatory or degenerative stromal response resulting in the formation of stromal lipofuscein deposits. Visual acuity was not impaired in the patients investigated in this study. As these stromal deposits appear to be persisting years after surgery and possibly are irreversible in nature, a long term effect on the corneal physiology and function should carefully be monitored.