A compartment model to calculate time-dependent concentration profiles of topically applied chemical compounds in the anterior compartments of the rabbit eye

Hitherto, none of the existing in vitro methods has been convincingly demon strated to be suitable as a replacement for the Draize rabbit eye irritatio n test. We examine the hypothesis that one reason for this is that insuffic ient consideration has been given to the differences in the effective conce ntrations at which chemicals operate in vitro and in vivo. When a chemical is applied topically to the eye, the strength of the observed irritation th at it elicits depends both on its toxic potential toward cells or tissues, and its effective concentration in the tissues of the eye. Most of the exis ting in vitro methods are based on isolated cells or tissues, and thus may be useful in assessing the cytotoxic potentials of chemicals. However, a re liable approach to assessing the effective concentrations of chemicals with in the various tissues of the eye it; lacking. A simplified compartment mod el is presented for calculating the time-dependent, intra-ocular concentrat ion profiles of topically applied chemicals. The model encompasses the oute r surface of the eye, three distinct segments of the cornea (subdivided int o the epithelium, stroma and endothelium) and the conjunctiva. Transport th rough the membranes of these compartments is described as passive diffusion . For the transport coefficients, rate equations are established that conta in, as free parameters, the molecular size and the partition coefficient of the chemical, as well as some intrinsic membrane parameters, such as thick ness, viscosity and pore density. Numerical values for the unknown membrane parameters were estimated by fitting the theoretical rate equations to mea sured permeability coefficients. The compartment model was applied to an in dependent set of 52 test chemicals compiled from the European Commission/UK Home Office validation study The calculated passage times (required to let 95% of the chemical reach the posterior eye tissues) varied between 0.33 m inutes and 50.6 minutes, and are generally much shorter than the typical du ration of observed impairments in the cornea or conjunctiva. This finding s uggests that short-term contacts of the eye tissues: with a chemical are su fficient to elicit long-term eye irritation. An example is given, showing h ow the conventional approach of using in vitro endpoints as predictors of e ye irritation can be improved significantly by incorporating into the predi ction the calculated intra-ocular concentration of a chemical.