The Department of Health and Human Services' National Toxicology Program recently published its 12th Report on Carcinogens (RoC). RoCs have historically classified substances as either “known” or “reasonably anticipated” human carcinogens. Yet, some of NTP's recent classifications are based on incomplete and unbalanced scientific reviews, which have resulted in incorrect conclusions.
For each substance NTP reviewed, there are human, animal toxicity and mechanistic studies available, although the quality and quantity of studies vary greatly. For some substances, results within and among studies are consistent, but for others, this is not the case.
To determine whether any substance is carcinogenic, therefore, scientists must use a clear framework that includes methods for evaluating the strengths and weaknesses of studies, and apply this methodology consistently to each study. Scientists must then assess the whole body of evidence — that is, conduct a weight-of-evidence analysis — using consistent criteria to judge the concordance of results across different types of studies. NTP's 12th RoC did not pursue this approach.
For example, NTP concluded that formaldehyde is a “known” human carcinogen. NTP's formaldehyde evaluation was similar to a recent Environmental Protection Agency evaluation that was reviewed by a National Academy of Sciences panel. The NAS panel concluded that the formaldehyde data were not reviewed in a balanced manner, and that this led to an overestimation of risks.
Using a similar analysis methodology, NTP concluded styrene is “reasonably anticipated” to be a human carcinogen. As was the case for formaldehyde, NTP did not conduct a proper weight-of-evidence analysis. Were styrene truly a human carcinogen, cancer should have been observed in patterns within and across human studies and such observations should have been coherent with animal toxicity and mechanistic studies. This is not the case.
An exhaustive risk assessment prepared by European Union scientists concluded that cancer risk was not elevated among workers in styrene-using and production industries and that animal toxicity studies did not indicate styrene would be a human carcinogen.
Some people are under the mistaken notion that misclassifying substances as carcinogens will not have any consequences. Quite the opposite is true. Misclassification could lead to unnecessary worry, a loss of jobs and consumer products, and to resources being diverted from measures that could truly benefit public health.
Goodman is a principal at environmental and risk science consulting firm Gradient of Cambridge, Mass., and an adjunct faculty member at the Harvard School of Public Health.