The pathology of surfactant-induced ocular irritation, especially in the context of accidental human exposures and animal tests used to assess a surfactant's potential ocular irritation, is not well understood. The purpose of this study was to characterize the microscopic changes in rats at 3 hr and on days 1, 2, 3, 4, 7, 14, and 35 following treatment with anionic, cationic, and nonionic surfactants of differing irritancy. The right eye of each rat was treated by placing 10 mu l of a surfactant directly on the cornea. Untreated left eyes served as the controls. At each time point, eyes and eyelids were macroscopically examined and collected for microscopic examination. Macroscopically, the differing levels of irritation were characterized by differences in incidence and magnitude of scores, reflecting involvement of the cornea, conjunctiva, and iris, as well as by the incidence of neovascularization and time to recovery. Microscopically, differences in the area and depth of injury paralleled the differences seen grossly and the relative irritancies of the various surfactants. All surfactants affected the corneal and conjunctival epithelium. All surfactants, except the slightly irritating anionic surfactant, caused corneal stromal changes, with this involvement being proportional to their overall level of irritation. Corneal endothelial cell effects principally occurred with only the severely irritating cationic surfactant. Over time, responses to surfactants of differing irritancy were qualitatively and quantitatively different, and these differences correlated with the extent of initial injury. Qualitative differences in response included presence of keratocyte regeneration, corneal neovascularization, and conjunctivalization of the corneal epithelium with all of the surfactants except the slight irritant. Quantitative differences in response occurred in the extent of epithelial regeneration, edema, and inflammation for surfactants of slight to severe irritancy, and with neovascularization, keratocyte regeneration, and conjunctivalization for surfactants of mild to severe irritancy. These results suggest that by defining initial area and depth of injury associated with an ocular irritant, it may be possible to predict the subsequent response and final outcome. Such an approach would be applicable to the development of mechanistically based in vitro assays.
