Humans are exposed to a myriad of environmental pollutants through inhalation, ingestion, and dermal absorption. The pollutant's dose to internal target organs (the absorbed dose) is partially determined by the route of exposure. The structure, function, and metabolism of the portal of entry modulate the dose of pollutant and its metabolites that enter the circulation system for distribution to the rest of the body. In addition, the pollutant may express its toxic effect in the portal of entry itself.
Under a variety of legislation, the U.S. Environmental Protection Agency (the U.S. EPA) and other federal and state regulatory agencies are charged with protecting human health from harmful effects of environmental chemicals. For pollutants, the U.S. EPA conducts a quantitative risk assessment to determine the daily exposure dose from a particular route of exposure that is not anticipated to cause significant risk over a lifetime of exposure. Frequently, these risk assessments are conducted based on limited data. In some cases, data are not available for the route of exposure being considered, but are available for another route. For these substances, the U.S. EPA must determine whether the data can be extrapolated to the route being assessed.
At present, the U.S. EPA has no formal guidelines for route-to-route extrapolation for noncarcinogens. However, route-to-route extrapolation is done routinely by different EPA program offices, often using empirically derived factors that are not necessarily applicable to the case at hand. For example, the U.S. EPA has derived approximately 15 to 20 oral reference doses from inhalation data [1]. For these risk assessments, it is assumed (in the absence of other information) that absorption via the oral route is complete (i.e., 100%), whereas absorption via inhalation is 50% that of oral.
Currently, the U.S. EPA 1 method for assessing the risk of carcinogens (the linearized multistage model) assumes that the same total daily body burden will give the same tumor incidence regardless of the route of exposure. Thus, the approach generally does not consider that some tumors at the site of contact (e.g., following topical application) may be site-specific, or that the dose to a target organ may be modulated by the route of exposure.
Route-to-route extrapolations are also performed by other federal and state regulatory agencies. The state of California, for example, considers it prudent risk assessment policy to assume, in the absence of data, that a substance that causes cancer when ingested will also cause cancer when inhaled, and vice versa [2]. The Department of Health and Human Services assumes 100% absorption across species, regardless of the route of exposure, in the absence of valid evidence to the contrary [3].
For most route-to-route extrapolations, the lack of data, lack of ability to interpret data, and underutilization of existing data due to insufficient models and statistics reduce the validity of these extrapolations. At present, little is known about absorption characteristics, the potential for portal-of-entry effects, and the potential for first-pass metabolic effects for most compounds by most routes. Few studies are currently designed to test toxicity across routes. Despite this lack of data, immediate regulatory needs often necessitate route-to-route extrapolation.
Development of scientifically based principles, procedures, and data for route-to-route extrapolation would improve the validity of risk assessments, which in turn would better ensure protection of human health while avoiding overregulation and underregulation. It would also help to make the risk assessment process more efficient, an important factor considering that federal and state agencies must regulate hundreds of chemicals in the environment. Identification of research needs will help these agencies plan future research to address the problems of route-to-route extrapolation.
The workshops documented in this publication were intended to be the first step in an iterative process. The principles, conclusions, and recommendations derived from the workshops and presented in this summary report will help provide a scientific foundation for route-to-route extrapolation. Additional research on assumptions and scientific aspects of route-to-route extrapolation will require periodic revision of this foundation and any methodologies developed from it to incorporate new information.