Health assessment of phosgene: approaches for derivation of reference concentration

Abstract

This paper describes the derivation of the chronic reference concentration (RfC) for human inhalation of phosgene that was recently added to the Environmental Protection Agency's (EPA) Integrated Risk Information System (IRIS) data base (U.S. EPA, 2005. Toxicological Review of Phosgene: In Support of Summary Information on the Integrated Risk Information System (IRIS). Available online at: <http://www.epa.gov/IRIS>). The RfC is an estimate of daily phosgene exposure to the human population that is likely to be without appreciable risk of deleterious effects during a lifetime. [For this and other definitions relevant to EPA risk assessments refer to the glossary of terms in the US EPA IRIS website (http://www.epa.gov/IRIS).] Phosgene is a potential environmental pollutant that is primarily used as a catalyst in the polyurethane industry. It is a gas at room temperature, and in aqueous solution it rapidly hydrolyzes to CO2 and HCl. In the absence of chronic human health effects information and lifetime animal cancer bioassays, the RfC is based on two 12-week inhalation studies in F344 rats which measured immune response and pulmonary effects, respectively. The immune response study showed impaired clearance of bacteria that was administered into the lungs of rats immediately after exposure to phosgene at concentrations of 0.1, 0.2 and 0.5 ppm. It also showed that the immune response in uninfected rats was stimulated by phosgene exposure at all concentrations. The pulmonary effects study showed a progressive concentration-related thickening and inflammation in the bronchiolar regions of the lung that was mild at 0.1 ppm and severe at 1.0 ppm. An increase in collagen content, as observed with histological collagen stains, was observed at 0.2 ppm and above. Though there is considerable uncertainty associated with the species and exposure duration employed, this endpoint is considered an indication of chronic lung injury of potential relevance to humans. Three different approaches for RfC derivation were taken in analyzing these studies: (1) the traditional NOAEL/LOAEL method; (2) the benchmark dose (BMD); and (3) the categorical regression for the analysis of severity-graded pulmonary damage data using the recently revised USEPA CatReg software. The BMD approach was selected as the method of choice to determine the RfC for phosgene because it has several advantages compared to the NOAEL/LOAEL: (1) it is not restricted to the set of doses used in the experiments; (2) the result is not dependent on sample size; (3) it incorporates information on statistical uncertainty. The CatReg approach allowed the incorporation of data on the severity of the pathological lesions, and therefore it complemented the other approaches. The BMD approach could not be applied to the immune response data because it was not possible to define an adverse effect level for bacterial resistance. However, NOAEL/LOAEL values for immune responses were consistent with benchmark dose levels derived from lung pathology data and used in the derivation of the RfC. The preferred RfC method and derivation involved dividing the benchmark dose from the collagen staining data (0.03 mg/m3) by a composite uncertainty factor of 100: RfC=0.03/100=3E-4 mg/m3.