Workshop summary: phosgene-induced pulmonary toxicity revisited: appraisal of early and late markers of pulmonary injury from animal models with emphasis on human significance.

A workshop was held February 14, 2007, in Arlington, VA, under the auspices of the Phosgene Panel of the American Chemistry Council. The objective of this workshop was to convene inhalation toxicologists and medical experts from academia, industry and regulatory authorities to critically discuss past and recent inhalation studies of phosgene in controlled animal models. This included presentations addressing the benefits and limitations of rodent (mice, rats) and nonrodent (dogs) species to study concentration x time (C x t) relationships of acute and chronic types of pulmonary changes. Toxicological endpoints focused on the primary pulmonary effects associated with the acute inhalation exposure to phosgene gas and responses secondary to injury. A consensus was reached that the phosgene-induced increased pulmonary extravasation of fluid and protein can suitably be probed by bronchoalveolar lavage (BAL) techniques. BAL fluid analyses rank among the most sensitive methods to detect phosgene-induced noncardiogenic, pulmonary high-permeability edema following acute inhalation exposure. Maximum protein concentrations in BAL fluid occurred within 1 day after exposure, typically followed by a latency period up to about 15 h, which is reciprocal to the C x t exposure relationship. The C x t relationship was constant over a wide range of concentrations and single exposure durations. Following intermittent, repeated exposures of fixed duration, increased tolerance to recurrent exposures occurred. For such exposure regimens, chronic effects appear to be clearly dependent on the concentration rather than the cumulative concentration x time relationship. The threshold C x t product based on an increased BAL fluid protein following single exposure was essentially identical to the respective C x t product following subchronic exposure of rats based on increased pulmonary collagen and influx of inflammatory cells. Thus, the chronic outcome appears to be contingent upon the acute pulmonary threshold dose. Exposure concentrations high enough to elicit an increased acute extravasation of plasma constituents into the alveolus may also be associated with surfactant dysfunction, intra-alveolar accumulation of fibrin and collagen, and increased recruitment and activation of inflammatory cells. Although the exact mechanisms of toxicity have not yet been completely elucidated, consensus was reached that the acute pulmonary toxicity of phosgene gas is consistent with a simple, irritant mode of action at the site of its initial deposition/retention. The acute concentration x time mortality relationship of phosgene gas in rats is extremely steep, which is typical for a local, directly acting pulmonary irritant gas. Due to the high lipophilicity of phosgene gas, it efficiently penetrates the lower respiratory tract. Indeed, more recent published evidence from animals or humans has not revealed appreciable irritant responses in central and upper airways, unless exposure was to almost lethal concentrations. The comparison of acute inhalation studies in rats and dogs with focus on changes in BAL fluid constituents demonstrates that dogs are approximately three to four times less susceptible to phosgene than rats under methodologically similar conditions. There are data to suggest that the dog may be useful particularly for the study of mechanisms associated with the acute extravasation of plasma constituents because of its size and general morphology and physiology of the lung as well as its oronasal breathing patterns. However, the study of the long-term sequelae of acute effects is experimentally markedly more demanding in dogs as compared to rats, precluding the dog model to be applied on a routine base. The striking similarity of threshold concentrations from single exposure (increased protein in BAL fluid) and repeated-exposure 3-mo inhalation studies (increased pulmonary collagen deposition) in rats supports the notion that chronic changes depend on acute threshold mechanisms.