On March 29, 2005, the U.S. Environmental Protection Agency (EPA) issued a revised set of Guidelines for Carcinogenic Risk Assessment to replace those adopted in 1986. The revisions reflect a gradual evolution of the process by which EPA performs risk assessment for possible cancer-causing agents. The major changes in this new document include recommendations to: (1) increase the use of scientific data, particularly those related to understanding the way that agents act to cause cancer; (2) perform separate evaluations of cancer risks for possible sensitive subpopulations, particularly children; and (3) use safety factors rather than mathematical formulae in estimating cancer potency for certain types of chemicals.
Science and EPA's Real Goal
To understand these new guidelines and how they differ from the original ones, it is important to examine the goal of the assessment and the way that EPA is trying to achieve this. Despite its title, the ultimate goal of the new, as well as the original, guidelines is not to assess risk but rather to protect health. The method for achieving this aim, as described in the document, is to first consider the science, and then if this is inadequate, to apply policy judgments, called defaults, to develop risk values that are used in risk management.
As stated in the document, the EPA position on the use of science is that, if the scientific data are of "sufficient" quality, the results of scientific studies alone can be used as the bases for judgments about carcinogenic risk. However, if the data are "limited" or of "insufficient quality," or if there is "too much uncertainty in the data," then policy choices (also known as defaults) must be applied. These defaults are chosen so as to increase the likelihood that a given chemical will be identified as a cause of cancer and also that the potency of a given cancer-causing chemical will be overestimated. This approach is considered health protective since it leads to an overstatement of the risk, and it is expected that higher risk estimates will lead to more stringent regulation. More stringent regulation is then equated with greater "health protection."
Policy First, Science Second?
However, close examination of the guidelines indicates that the actual process is not consistent with the EPA description of the approach. Rather than policy being considered only after scientific studies have been performed and evaluated, policy is the first consideration and is of primary importance. This is evident in the numerous statements that explicitly and implicitly indicate that defaults, based on policy choices, are assumed correct unless there are "sufficient" data to abandon these defaults; i.e., policy, not science, comes first. The central role of policy in the guidance is further emphasized by the use of policy considerations to judge the "sufficiency" of the scientific data.
Further, the assertion that applying the defaults described in the guidelines will lead to improved health is not borne out after careful analysis. Even if overstating the risk leads to more stringent regulation, this is not necessarily equivalent to greater health protection. For example, more stringent regulation can have a negative impact on health if limiting the use of a chemical leads to loss of its health benefits without significantly decreasing any actual cancer risk. For another, the costs associated with over-regulating a chemical based on overstatement of risk may reduce the resources available to address more significant health concerns. Thus, the idea that overestimation of risk is equivalent with greater health protection is flawed and can lead instead to reduced health protection.
In summary, the fundamental approach to cancer risk assessment in the new document is the same as that used in the original carcinogen guidelines of 1986 -- including the primacy of policy over science and the belief that overstatement of risk is equivalent to greater health protection. It is unlikely that EPA's newly-stated effort to include more science in the assessment process and to show more flexibility in the use of scientific data will result in any essential change in the way that cancer risks are assessed. In light of these flaws, the new guidelines do not represent a significant advance for science or for achieving increased health protection.
Michael Kamrin, Ph.D., is a Michigan State University professor emeritus of environmental toxicology and an ACSH Advisor (ACSH.org, HealthFactsAndFears.com).
