Epithelial plugs in radial keratotomy: the origin of incisional keratitis?

PURPOSE

Incisions in radial keratotomy (RK) patients were examined histopathologically to explore the cause of delayed infections occurring following this procedure.

METHODS

Four corneas from two successful RK patients were obtained postmortem and 1 cornea from an unsuccessful RK was collected following lamellar keratoplasty. The tissue was prepared for light and transmission electron microscopy.

RESULTS

All incisions examined contained an epithelial plug. The average depth of the plugs for left and right corneas in the 2 successful cases were 142.58 microm (range: 36.46-183.04 microm) and 58.44 microm (range: 29.27-115.66 microm), whereas the unsuccessful case these measurements were 121.06 microm (range: 70.03-172.01 microm). In the successful cases, the plugs were on average 4.2 cells deep (range: 2-5) in one case and for the other the plugs were 8.78 cells deep (range: 1.5-11.5) and 2.72 cells wide (range: 2-4). In the unsuccessful case the plugs were on average 6.89 cells deep (range: 5-11) and 8.56 cells wide (range: 4-13). The average epithelial thickness, measured 200 microm on either side away from the plug, was 47.11 microm and 55.09 microm for the successful cases, and 27.52 microm for the unsuccessful case. Degenerate cells were often observed within the plug and along the overlying surface. Lamellae cut during surgery remained severed in all corneas observed. The stroma adjacent to the incision contained an elevated number of stromal cells (8%-78%).

CONCLUSIONS

Healing did not include repair of anterior limiting lamina or stromal lamellae, which all remained severed in all incisions observed. Epithelial plugs were invariably present in all incisions regardless of clinical outcome, thus, increasing the distance a basal cell travels to reach the epithelial surface by 2 to 4 times. Since the epithelial cell has a limited life, it is hypothesized that cells originating in the plug may die before reaching the surface, thus, leading to intermittent loss of surface integrity. The loss of the epithelial barrier allows invasion of microorganisms and the delayed onset of keratitis. The larger the plug the greater the possibility of surface defects and resultant infection.

The effect of age on corneal and lens superoxide dismutase

The enzymatic activity of copper-zinc superoxide dismutase within the cornea and lens of the juvenile (1 week) and adult (26 week) rat has been measured relative to both total soluble protein and immunoreactive protein. Enzymatic activity of corneal superoxide dismutase, localized in the epithelium and endothelium, remains constant with age relative to total soluble protein and the ratio of catalytically active enzyme to immunoreactive enzyme is near unity indicating that the superoxide dismutase present is enzymatically active. Likewise, in the juvenile rat, the majority of lens superoxide dismutase, which is located in the capsule epithelium, is enzymatically active. However, in the adult rat a large portion of the enzyme present is catalytically inactive. The enzymatic activity also decreases relative to total soluble protein with age in both the rat and the rabbit lens. This decrease in protection against toxic radicals and/or the accumulation of enzymatically defective enzyme may be contributing factors to age related lens disorders.