Sensory irritation as a basis for setting occupational exposure limits.

There is a need of guidance on how local irritancy data should be incorporated into risk assessment procedures, particularly with respect to the derivation of occupational exposure limits (OELs). Therefore, a board of experts from German committees in charge of the derivation of OELs discussed the major challenges of this particular end point for regulatory toxicology. As a result, this overview deals with the question of integrating results of local toxicity at the eyes and the upper respiratory tract (URT). Part 1 describes the morphology and physiology of the relevant target sites, i.e., the outer eye, nasal cavity, and larynx/pharynx in humans. Special emphasis is placed on sensory innervation, species differences between humans and rodents, and possible effects of obnoxious odor in humans. Based on this physiological basis, Part 2 describes a conceptual model for the causation of adverse health effects at these targets that is composed of two pathways. The first, "sensory irritation" pathway is initiated by the interaction of local irritants with receptors of the nervous system (e.g., trigeminal nerve endings) and a downstream cascade of reflexes and defense mechanisms (e.g., eyeblinks, coughing). While the first stages of this pathway are thought to be completely reversible, high or prolonged exposure can lead to neurogenic inflammation and subsequently tissue damage. The second, "tissue irritation" pathway starts with the interaction of the local irritant with the epithelial cell layers of the eyes and the URT. Adaptive changes are the first response on that pathway followed by inflammation and irreversible damages. Regardless of these initial steps, at high concentrations and prolonged exposures, the two pathways converge to the adverse effect of morphologically and biochemically ascertainable changes. Experimental exposure studies with human volunteers provide the empirical basis for effects along the sensory irritation pathway and thus, "sensory NOAEChuman" can be derived. In contrast, inhalation studies with rodents investigate the second pathway that yields an "irritative NOAECanimal." Usually the data for both pathways is not available and extrapolation across species is necessary. Part 3 comprises an empirical approach for the derivation of a default factor for interspecies differences. Therefore, from those substances under discussion in German scientific and regulatory bodies, 19 substances were identified known to be human irritants with available human and animal data. The evaluation started with three substances: ethyl acrylate, formaldehyde, and methyl methacrylate. For these substances, appropriate chronic animal and a controlled human exposure studies were available. The comparison of the sensory NOAEChuman with the irritative NOAECanimal (chronic) resulted in an interspecies extrapolation factor (iEF) of 3 for extrapolating animal data concerning local sensory irritating effects. The adequacy of this iEF was confirmed by its application to additional substances with lower data density (acetaldehyde, ammonia, n-butyl acetate, hydrogen sulfide, and 2-ethylhexanol). Thus, extrapolating from animal studies, an iEF of 3 should be applied for local sensory irritants without reliable human data, unless individual data argue for a substance-specific approach.