Superfund Research Program

Improving Health Risk Assessment Through Mechanistic Research

Release Date: 05/19/1999

The hazardous substances remaining in the soil or groundwater on Superfund sites present potential health risks to humans. Predicting how much risk is associated with these sites is the purpose of risk assessment, which is the process used by the U.S. Environmental Protection Agency to determine the probability of adverse health effects that may result from exposures to environmental contaminants. A risk assessment involves the collection of relevant biological, dose-response, and exposure data, which are subsequently entered into risk-prediction models to generate estimates of health risks. These "risk estimates" are used to establish the cleanup standards for hazardous waste sites.

While a practical tool for guiding environmental remediation, risk estimates are inherently uncertain due in part to the inadequacy of the data and the methods used to predict health risks. Oftentimes, the best available biological and dose-response data for a health risk assessment are the results obtained from experiments in which rodents have been exposed to high doses of test compound. In these experiments, the doses far exceed what humans typically encounter, and the route of exposure often does not match actual human exposure conditions. To generate a risk estimate from these high-dose bioassays, extrapolations must be made from high to low dose, from species to species, and sometimes from route to route of exposure. Extrapolations are purposefully conservative to ensure adequate protection of public health; however, there is concern that this cautious approach may result in substantial overestimates of risk.

Recognizing the limitations of risk assessment, there has been increasing interest in strengthening its scientific foundations to include improving our understanding of the fundamental processes underlying chemical carcinogenicity.

Researchers at the University of North Carolina are improving the scientific basis of cancer risk assessment by investigating "DNA adducts," a type of DNA damage that many environmental carcinogens can produce in cells. DNA adducts are sites on DNA to which reactive chemicals have covalently bonded. If not repaired by cellular defense mechanisms, they can lead to genetic mutations, an early step in the multistage process of cancer.

Initially, the UNC scientists focused on the development of new analytical methods for measuring DNA adducts and DNA repair in cells. Their long-term plan is to use these methods to study the biological mechanisms of cancer induced by vinyl chloride and other environmental carcinogens.

Three highly sensitive and specific methods were developed. The first method takes advantage of the powerful analytical capabilities of gas chromatography/high resolution mass spectroscopy (GC/HRMS). GC/HRMS has been optimized to detect and measure a DNA adduct known as the "etheno" adduct, which is the major promutagenic DNA adduct of vinyl chloride. The second method was designed to measure the expression of the enzyme methylpurine glycosylase, which is the main DNA repair enzyme for etheno adducts in cells. The third method can measure abasic sites, which are the sites on DNA where etheno adducts have been removed by DNA repair enzymes.

Using these newly developed methods, the UNC researchers discovered that hepatic endothelial cells, the target cells for vinyl chloride-induced liver cancer in rodents, develop 2.5-fold greater numbers of etheno adducts than the cell type (hepatocytes) in a neighboring tissue layer. Endothelial cells also appear to be deficient in methylpurine glycosylase, the major DNA repair enzyme for etheno adducts. In comparison to hepatocytes, the endothelial cells were found to have 500% more abasic sites.

The GC/HRMS technique is so sensitive that it gave the UNC researchers the opportunity to discover that cells have relatively high endogenous amounts of etheno adduct. The source of these endogenous DNA adducts is unknown, but they are not likely associated with exposure to environmental contaminants, such as vinyl chloride. This discovery suggests that normal metabolic processes may be responsible for the formation of endogenous etheno adducts and that these adducts may be important causes of human cancer in the absence of chemical exposure.

These new findings are providing crucial mechanistic data that is important for the accurate prediction of human risk to vinyl chloride exposures. By examining the dose-response relationships for etheno adduct formation and repair, risk assessors can then determine if non-linearities exist in the vinyl chloride dose-response curve. This knowledge, along with data on endogenous etheno adducts, will provide important comparisons for high versus low exposures. The quality of risk assessment is being improved by this research, which will allow society to better prioritize hazardous site cleanups and make the best use of available resources.

For More Information Contact:

James A Swenberg
University of North Carolina-Chapel Hill
Department of Environmental Sciences and Engineering
253C Rosenau Hall, CB# 7431
Chapel Hill, North Carolina 27599-7431
Phone: 919-966-6139

To learn more about this research, please refer to the following sources:

  • Nakamura J, Walker V, Upton PB, Chiang S, Kow YW, Swenberg JA. 1998. Highly sensitive apurinic/apyrimidinic site assay can detect spontaneous and chemically induced depurination under physiological conditions. Cancer Res 58(2):222-225. PMID:9443396
  • Swenberg JA. 1997. Generating data for scientifically-based risk assessment. Fundam Appl Toxicol 36:286.
  • Wu K, Ranasinghe A, Scheller NA, Upton PB, Swenberg JA. 1997. Mechanism-based risk assessment: utility of DNA adducts in model selection. Fundam Appl Toxicol 36:97.

To receive monthly mailings of the Research Briefs, send your email address to