Tobacco Documents Online

A,4"c-x F Current cancer risk assessment using linear extrapolation from high doses to low doses is scientifically invalid • The probable causes of cancer in mankind are becoming better known. Major causes are smoking, unbalanced diets, chronic infections and genetic factors. In addition, exposure to the sun and occupational exposures might cause a few percent of human cancer. Pollution seems to be a minor contribution to cancer, accounting for considerably less than 1%; yet public preoccupation with pollution is very high, in good part because of animal cancer tests. • Low levels of carcinogenic agents of natural origin are omnipresent in man's environment, in the air we breathe and the food we eat It is thus impossible to imagine that man could be guaranteed conditions totally free of exposure to such chemicals or to background radiation. Zero risk cannot be achieved. • Major advances in analytical and measurement techniques today enable the detection of extremely low concentrations of all substances, of both natural and man-made origin, often a million times smaller than 20 years ago. • Animal cancer tests, which are done at the maximum tolerated dose (MTD), are beingg misinterpreted to mean that low doses of the chemicals tested and found positive are thereby relevant to human cancer. Animal cancer tests are mainly done on synthetic chemicals and industrial pollutants, yet half of all natural chemicals that have been tested at the MTD are rodent carcinogens. A plausible explanation for the high frequency of positive results in animal cancer tests is due to cell killing and compensatory cell division; this high dose effect does not occur at low doses. In any case 99.99% of the pesticides we eat are naturally present in plants to ward off insects and other predators. More than half of those natural pesticides, tested in high dose animal'tests, are rodent carcinogens. Reducing our exposure to the remaining 0.01%, whether to individual chemicals or mixtures, is both enormously expensive and does not reduce cancer rates. • The reason we can eat the tremendous variety of natural chemical rodent carcinogens in our food is that animals are extremely well defended against all chemicals by many general defense systems, most of which are inducible (more of them are made when they are in use). They are equally effective against natural and synthetic reactive chemicals. Thus, animals are extremely well defended against low doses of chemicals. One does nof expect, nor does one find, a general difference between synthetic and natural chemicals in their carcinogenicity. • A high proportion of rodent carcinogens are likely to have a threshold dose below which no pathological effect is observed. Scientists must determine mechanisms of carcinogenesis, for each substance. Acceptable dose levels should be revised as a function of advances in our understanding of mechanisms. Risk must be estimated using a weight -of- evidence approach, not by a blanket worst-case rule. • Mathematical methods used to extrapolate to low level doses from observed effects in man and in animals at high carcinogenic concentrations, based on linearity have led to grossly exaggerated mortality forecasts. Such methods have never been verified by clinical experimentation. The idea that there is an epidemic of~cancer causedby synthetic industrial chemicals is false. The constant rise in life expectancy in the developed countries provides additional evidence of this error. It is the steady progress of scientific research and technology that will continue to lengthen human life expectancy and improve human walfare. • Those who bearthe responsibitity for formulating regulations should not neglect the risks and costs incurred by their decision. "False positives" -substances hypothesized to cause cancer at low dose exposures, but that do not- harm health by lowering our standard of living and diverting resources from more i'mportant actions to protect health: Risks compete with risks : society must distinguish between significant risks and trivial hypothetical risks_