Derivation of thresholds for inhaled chemically reactive irritants: Searching for substance-specific common denominators for read-across prediction

Abstract

Emergency response planning guideline values are used to protect the public when there has been a short-term chemical release. These values serve the purpose of identifying areas where a hazard exists if the concentration of hazardous chemicals is exceeded for the specified exposure duration. This paper focuses on carbonyl chlorides, a class of highly irritant/corrosive chemical intermediates characterized by the reactive moiety R–COCl. Despite their unifying property of reacting with nucleophilic biopolymers/peptides lining the airways of the respiratory tract, their adverse outcome pathway (AOP), in addition to surface area dose, appears to be dominated by their site(s) of major deposition (liquid) or retention (gas) within the respiratory tract. Thus, the physicochemical properties “phase” and “lipophilicity” become more decisive for the AOP than the chemical structure. This complicates the grouping of portal-of-entry irritant chemicals for the read-across prediction of chemicals, especially those with semivolatile properties. Phosgene (COCl2) served as a template to predict emergency response planning levels 2 (non-incapacitating, reversible injury) and 3 (nonlethal) for related chemicals such as SOCl2, formates, and acid chlorides. A rationale and guide to the systematic characterization of uncertainties associated with the lung region, water solubility of the vapor phase, and chemical specificity is given. The approach described in this paper highlights the regional differences and outcomes that are phenotypically described as irritation of the respiratory tract. Especially for such a data-lean group of chemicals, reliable read-across predictions could reduce the uncertainty associated with the derivation of values used for emergency-related risk assessment and management. Likewise, the approach suggested could improve the grouping and categorization of such chemicals, providing a means to reduce animal testing with potentially corrosive chemicals. Overall, the course taken for read-across predictions provided valid estimates as long as emphasis was directed to the physicochemical properties determining the most critical regional injury within the respiratory tract.