By Dick Clapp 

Researchers devote a lot of effort to determining what causes cancer, and their findings can help us treat and prevent the disease. Industries that use and manufacture suspected carcinogens have something to fear, though, if research shows their products or processes to be contributing to cancer in workers or nearby communities.  As a result, there has been a three-decade debate about the magnitude of the cancer burden contributed by these sources.

This issue is getting renewed attention because the International Agency for Research on Cancer (IARC) recently released a report on “Attributable causes of cancer in France in year 2000.”  This was a collaborative effort with the French Academy of Sciences, several other cancer-related agencies, and a distinguished international group of reviewers.  It was begun in 2005 with the intention of updating the frequently-cited estimate of the attributable causes of cancer done by Doll and Peto in 1981, specifically as these estimates applied to France.  The report reviews a large literature, goes through a long series of calculations and sensitivity analyses, and comes up with a set of conclusions remarkably similar to those of Doll and Peto’s estimates 26 years earlier.  They attribute even smaller percentages of cancer due to occupation and “pollution” than did Doll and Peto.

The most surprising aspect of the new IARC report comes in the discussion section, though, where the authors suggest the possibility that low-dose environmental and occupational exposures might actually have decreased cancer incidence in France. In doing so, they invoke the notion of “hormesis.”  Say what?

Hormesis is variously defined as “non-monotonic dose-response,” or “J-shaped dose-response at low dose,” but it basically asserts that by stimulating a response at low dose, an organism may actually protect itself from toxicity and the overall effect will be beneficial – a “you could sprinkle a little on your breakfast cereal” kind-of-thing.  This is in contrast to the standard “linear, no threshold” (LNT) approach, which holds that the  response to a substance increases proportionally with dose, and that there is no threshold below which the substance will not provoke a response.

A recent exchange in Environmental Health Perspectives summarizes the opposing viewpoints about hormesis. Ralph Cook and Ed Calabrese argue that there is much animal literature to support hormesis and that it, not the LNT approach, should be the default assumption in regulating things like environmental carcinogens.  In response, Thayer, Melnick, Huff and Burns challenge the scientific basis for the hormesis approach in the public health context and question whether it has become “A New Religion.”

The idea that low-level exposure to toxic chemicals, radiation, medicines, etc. may be beneficial is not new, but it has recently re-appeared in the context of regulating low-level exposure to various chemicals in the environment.  It was raised in the EPA Dioxin Reassessment in 2000 by John Graham and others who cited an animal experiment that showed decreased breast tumors in mice administered low doses of tetrachlorodibenzo-p-dioxin.  There was debate in the Science Advisory Board panel, which included Graham at that point, but the language the panel sent back to the agency seemed to suggest that EPA reconsider regulation of dioxin with a hormesis approach.  The EPA’s leading expert on dioxin, Dr. Linda Birnbaum, called this “irresponsible,” given the voluminous literature about harmful low-dose effects of dioxin on other non-cancer endpoints and in other animal species. 

Radiation hormesis has also been popular in some circles, with T.D. Luckey becoming one of its most vocal proponents over the past few years.  At one point, in a radiation research society newsletter, Dr. Luckey proposed that the U.S. implement a program of radiation supplementation for people who had low background exposure.  This would entail having public health workers drive mobile vans with radiation sources in them around the country and have people come in for supplementation doses (I am not making this up). Although there is clearly some evidence that low-dose ionizing radiation exposure stimulates repair mechanisms at the cellular level, the overall picture from human studies as well as the animal research gives no reason to abandon the linear (LNT) assumption about cancer risk.  This was the conclusion of the recent National Academy of Sciences review (BEIR VII), and it provides no support for Dr. Luckey’s absurd plan.

So, what are the IARC authors doing by invoking hormesis as something to consider when trying to attribute low-dose environmental chemical cancer risks?  They had asserted, without documentation, that environmental pollution might have contributed to the decreased incidence of some cancers in France.  I was not part of the discussion, nor have I talked to Dr. Peter Boyle, the recently chosen President of IARC and the correspondent listed at the beginning of the recent report.  But my suspicion is that the sudden prominence of hormesis is the latest permutation of the “manufacturing uncertainty” strategy pioneered by the tobacco industry several decades ago. The new leadership at IARC deserves our close attention, or this formerly respected international agency will drift further toward becoming an industry mouthpiece.

Stay tuned for further developments on this from colleagues in France and elsewhere.  My own views have been set out at length in a report that is currently being updated and will soon be available from the Collaborative on Health and the Environment.

Dick Clapp is Professor in the Department of Environmental Health at Boston University School of Public Health, and co-Chair of Greater Boston Physicians for Social Responsibility. Dr. Clapp served as Director of the Massachusetts Cancer Registry from 1980-1989 and worked in two environmental health consulting groups in addition to his teaching and research activities. He was a consultant to the U.S. EPA Science Advisory Board in its 1995 and 2000 reviews of the dioxin reassessment.