“My paper reevaluated the data from the original ED01 study from a totally new approach. My evaluation is based on the fundamental laws of chemistry and comes to a different conclusion.”

This paper has been highly cited primarily because it is so controversial. The monumental ED01 study was done on more than 24,000 mice about 30 years ago in an effort to finally answer the important question of extrapolating carcinogenicity data from animal experiments to humans. The doses used in animal experiments on carcinogenicity are very large and human exposures to those same chemicals are usually very low. The question, therefore, is, and has been, how should those results be extrapolated to human exposure? There has been considerable discussion about the interpretation of those experiments; however, all of the interpretations have been based on a statistical approach. The prevailing opinion has been that only a zero dose could result in zero cancer from a particular chemical; that is, a single molecule could cause cancer in a sufficiently large number of people. This opinion has dominated public opinion and the policy of governmental agencies ever since.

My paper reevaluated the data from the original ED01 study from a totally new approach. My evaluation is based on the fundamental laws of chemistry and comes to a different conclusion. Instead of finding that the extrapolation should be linear to zero dose and zero tumors, it concludes that the animal data show that there is a threshold for carcinogenicity only slightly below the large doses given to the animals. Consequently, human exposures at even slightly lower doses should be safe and the usual, much lower, doses to which humans are exposed are of no consequence whatsoever. Obviously this has profound implications and many are interested in deciding which approach is correct. Many toxicologists are excited about and support my approach, but many who have supported the prevailing opinion are vehemently opposed. Unfortunately the paper seems to have created a conflict between toxicologists, who in general support the concept, and statisticians, who in general oppose it. The surprise to me has been the intense emotional reaction by some well-known scientists who are opposed; the new analysis needs careful, deliberate consideration on its scientific merits alone.

In layman’s terms, my analysis has enormous implications. We, as a society, have assumed that if we are exposed to any amount of a chemical that has been shown to be carcinogenic in animal experiments, then we have some risk of developing cancer as a result of that exposure. My analysis would change that notion completely. Governmental agencies such as the EPA, etc. would have to change their positions on what is and is not carcinogenic. The evaluation would be based on the dose received. Any exposure below a certain dose would not be labeled carcinogenic, even if the animal experiments demonstrated that it was "strongly carcinogenic."

My involvement in this analysis was entirely fortuitous. I was a member of a committee evaluating the animal experiments on the carcinogenicity of methyl eugenol, which is naturally present in many foods that are consumed by humans. I decided to evaluate those animal data in the way that I thought was most appropriate from the laws of chemistry. The agreement with that analysis was so good (correlation coefficient = 0.999983) that I decided to examine other substances and studies by the same approach. That led me to the major ED01 study, and to others, which also have been published. My decision to base the analysis on fundamental principles of chemistry was the result of a much earlier interest in intracellular pH. The concept and definition of pH was a big issue in biology in the 1960s and I spent a lot of time then reconciling my concept with the laws of chemistry. Those concepts from chemical thermodynamics have been with me ever since.