An evaluation of a cultured human corneal epithelial tissue model for the determination of the ocular irritation potential of pharmaceutical process materials

Irritation and other forms of local toxicity following contact with eyes is a potentially serious problem arising from occupational exposure to chemicals. Traditionally, evaluation of the irritant potential of novel chemicals has relied on the use of in vivo studies with rabbits. Concerns about the predictive potential of in vivo methods for human hazard and demand for economical and rapid screening of chemicals has stimulated a great deal of work to investigate in vitro alternatives for evaluating ocular irritation potential. This publication describes a screening study to assess a reconstituted corneal epithelial culture system, as an alternative for testing for ocular irritation with pharmaceutical process materials, extending the chemical domain with which this system has been tested. A total of 21 test chemicals were applied to commercially supplied reconstituted human corneal epithelial (HCE) cultures and effects on tissue viability (MTT reduction assay), tissue histology and IL-alpha expression were assessed. Positive controls (0.5% and 1% SDS) showed dose- and time-related adverse effects on tissues, consistent with known irritant effects. Negative controls showed no histological changes and retained high viability throughout the time-course of the experiment. Concordance was excellent with accuracy at each sampling time point of over 80% when viability (MTT reduction) was compared with existing EU classification of the test articles for ocular irritation (classification based on results of in vivo evaluation). Tissue viability as estimated by MTT reduction appears most useful as the primary means of assessing the irritation potential of the chemicals. Histopathological examination generally agreed with the results of the MTT assay. However, the use of cytokine analysis will need further consideration as results for this parameter showed no relationship with known irritation potential. These results infer that HCE cultures, alone or as a part of a tiered hazard screening programme, have promise for use in reducing reliance on live subject tests and contribute to generation of an appropriate hazard classification and label advice.