Causes of cancer--an alternative view to Doll and Peto (1981).

This paper presents an alternative view to a commonly held misconception. The extent to which the causes of human cancer can be known is less than generally believed on the basis of the formal summation of estimates on the causal role of certain carcinogenic factors that have been tabulated by Doll and Peto (1981). Here it is stressed that the factors with known causal relation in the etiology have not as yet been ascertained for up to 97% of human cancers. Currently only one-third of the cancer deaths registered in the Federal Republic of Germany can be assigned etiologically to known exogenous carcinogenic agents or lifestyle. Accordingly, the causes of less than 50% of all malignant neoplasms are known and amenable to direct causal primary prevention. This, however, does not preclude a concept of "indirect primary prevention" which, based on the probable summation of subcarcinogenic effects of single carcinogens, enables removal from the environment of compounds that show carcinogenicity in animals, even if these compounds lack epidemiological evidence of carcinogenic activity.