Carcinogenicity bioassays with arsenic in chemical mixtures

Abstract

The carcinogenicity of arsenic has been indisputably demonstrated through diverse epidemiological data sets gathered from around the world. Evidence from environmental and occupational exposures has unequivocally established a link between exposure to arsenic and a variety of cancers, most notably in the skin, lung, liver, and urinary bladder. Yet, despite its known carcinogenic potential in humans, laboratory animal evidence for the carcinogenicity of arsenic is equivocal at best. Consequently, much remains unknown about its carcinogenic mode of action. In recent years, the scientific community invested significant resources to study the carcinogenic effects and mode of action of arsenic. This intensive research effort was driven partially by the recognition of large-scale, high-dose exposures to arsenic in drinking water in certain parts of Asia, as well as concerns in the United States regarding the protectiveness against cancer endpoints of the Environmental Protection Agency's Maximum Contaminant Level (MCL) of 50 parts per billion (ppb) in drinking water (revised in February, 2002 to 10 ppb). The need for a more complete understanding of arsenic's carcinogenic mode action, to facilitate the development of more accurate and reliable risk assessments, had become strikingly apparent. The preponderance of experimental evidence indicates that arsenic does not act as an initiator in the process of carcinogenesis, but rather as a promoter, progressor, or cocarcinogen. A logical experimental application of this observation, therefore, is the evaluation of arsenic in chemical mixtures composed of compounds the carcinogenic effects of which arsenic is likely to enhance. Given that humans generally encounter hazardous chemicals in the environment as mixtures rather than single compounds, the consideration of environmentally realistic chemical mixtures is equally important. Both of these issues are critical considerations in the development of a valid experimental animal model of arsenic carcinogenesis. The overall purpose of this research is to evaluate the carcinogenic potential of a chemical mixture consisting of 1,2-dichloroethane (DCE), vinyl chloride (VC), and trichloroethylene (TCE) in addition to arsenic (As). Composition of the experimental chemical mixture was determined based on considerations of environmental relevance and the existence of a common carcinogenic endpoint. All four chemicals are frequently identified as groundwater contaminants near hazardous waste disposal sites, and three (As, DCE, and VC) have been associated with the development of hepatic angiosarcoma in humans and/or animals. The significant results of this research were: (1) observation of hepatic sinusoidal dilatation, an early change reported in association with development of hepatic angiosarcoma, in the livers of rats treated with DCE; (2) discovery, in the framework of a multiple organ carcinogenicity bioassay, of a dose-dependent antagonistic effect of chemical mixtures containing As, DCE, VC, and TCE on the development of a) preneoplastic lesions in the lung and liver and b) neoplastic lesions in the lung; (3) validation, in a medium-term liver foci bioassay, of the suppressive effect of the quaternary mixture on the development of preneoplastic glutathione-S-transferase π(GSTP) positive hepatic foci; (4) attribution of the suppressive effect on GSTP positive focus development to As alone; (5) development of an HPLC method for measuring S-adenosylmethionine (SAM) and S-adenosylhomocysteine (SAH) in liver tissue; (6) evaluation of the potential involvement of altered levels of hepatic SAM and SAH in the suppressive effect of As, alone and in chemical mixtures, on the development of hepatic GSTP positive foci; and (7) in vitro observations of modulation of the toxicity of As and its metabolites in endothelial cells associated with altered microenvironmental pH. In contrast to data from a variety of recent studies which support the hypotheses that As may act as a promoter, progressor, or cocarcinogen, evidence from the research presented herein indicates that under certain circumstances, As may exert an antagonistic or suppressive effect on the development of neoplasia. Indeed, the efficacy of arsenic trioxide in the treatment of a variety of hematologic malignancies is well-documented, and it is currently undergoing clinical evaluation as a chemotherapeutic agent for a variety of other cancers as well. These apparently conflicting effects of As highlight the need for a greater understanding of its paradoxical nature as both a carcinogen and a therapeutic agent. Furthermore, reconciliation of the disparity in carcinogenic effects between animals and humans represents another challenge facing researchers in the arena of As carcinogenesis. With continued investigations, deeper knowledge of these issues will enable the development of more accurate and reliable risk assessment strategies for arsenic.