The Carcinogenicity of Certain Azo Dyes

Information on the relation of molecular structure to carcinogenicity is not nearly so extensive for the azo dyes as for the polycyclic hydrocarbons (10). Most of the pertinent publications appeared in Japan shortly after the carcinogenicity of azo dyes was first recognized, but before experimental conditions were established for an adequate survival of animals ingesting these compounds. One of the most active investigators in this field has published in English only his conclusions, without sufficient data for an evaluation of the results (14-16). Virtually all the earlier assays were made with a basal diet consisting of rice and carrot. Although this particular dietary combination was useful in the original discovery of the carcinogens o-aminoazotoluene (35) and p-dimethylaminoazobenzene (14), and also in the discovery that the development of hepatic tumors varies with diet (27), it may be questioned whether its continued use has not actually hindered further progress. All too frequently the feeding of an azo dye with rice and carrot has resulted in so high a mortality before tumors developed that any conclusion based on the surviving animals was without quantitative significance. The deficiencies in this diet have been discussed elsewhere (23), and the progress made with other diets is a matter of record (4, 8, 21-24, 34). It is now possible to examine the carcinogenicity of a series of closely related derivatives of p-dimethylaminoazobenzene under uniform dietary conditions that will permit an adequate survival of the experimental animals. An extension of the older data appeared particularly desirable, because such information is necessary to evaluate an attempt made by Kensler, Dexter, and Rhoads (11) to explain how azo dyes produce tumors. The essence of their idea is that it is not the azo dye that initiates the carcinogenic process, but an