TESTING FOR I,ONG-TERM TOXICITY 831 Tannenbaum (23). Mice were fed either a low or high calorie diet during cutaneous application of benzo(a)pyrene or during the subse- quent holtting period when the tumors appeared. The animals kept on a restricted low calorie intake during the entire experiment had a tumor incidence less than half that of mice fed ad libitum on the high calorie diet (Table VIII). Another illustration of the effect that diet may have on carcinoge- nicity comes from the work of Kawachi et al. (24). Addition of l r/o tryptophan, an essential amino acid, to the diet of rats receiving a relatively low level of the hepatocarcinogen N-nitrosodiethylamine in- creased the carcinogenic action on the liver almost fourfold (Table IX). Heated (thermally oxidized) oil, a common dietary component, may also have a synergistic effect on a carcinogen and enhance its action. Thus, Sugai et al. (25) found that rats fed a low level of the carcinogen 2-acetylaminofluorene developed virtually no tumors. However, simul- taneous feeding of this low level of 2-acetylaminofluorene with fractions of oxidized corn oil led to an 80-100c•o tumor incidence (Table X). The influence of ultraviolet light on the response to topical applica- tion of a test substance should also be kept in mind. Previously, the Table VIII Induced Skin Tumors and Diet in Mice• Diet during Benzopyrene Subsequent Diet Tumor Incidence Application (10 Weeks) • (52 Weeks) (%) High calorie High calorie 69 High calorie Low calorie 34 Low calorie High calorie 55 Low calorie l,ow calorie 24 Data from Tannenbaum (23). A 0.05-rag dose of benzo(a)pyrene was applied to the skin twice weekly during this period. Table IX Effect of L-Tryptophan on N-Nitrosodiethylamine Carcinogenesis • L-Tryptophan N-Nitrosodiethylamine Tumor Incidence (%) (rng/day) (%) 1 None 0 None 0.79 17 1 0.77 62 • Data from Kawachi et al. (24). All the animals were killed and examined for tumors after 192 days on experiment.

839 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS Table X Tumor Incidence from 2-Acetylaminofluorene (AAF) and Oxidized Corn Oil Fractions in Rats• Tumor Incidence at 30 Months (%) Oxidized Mammary AAF (%) Oil (%) Sex Liver Ear Duct Gland o. 005 ... M 0 0 0 0.005 ... F 0 0 12.5 0.005 2.5 M 96 21 70 0. 005 2.5 F 89 21 94 ... M 0 0 0 Controls ''' ... F 0 0 0 Data from Sugai et al. (25). greatest concern was that photosensitization might occur (26). However, fairly recent results point toward the need for reassessing whether cer- tain compounds may not enhance the weak carcinogenic action of ultra- violet radiation sufficiently so that tumors are produced. Thus, Bing- ham and Falk showed that three separate optical brighteners had such capability (27) (Fig. 2). Other factors to be noted are that percutaneous absorption of various materials, including carcinogens, does occur (28). In certain such in- stances tumors did not develop on the skin but in various internal organs (29, 30). Thus, despite efforts of Hoffmann and Graffi (30) to elicit tumors at the site of cutaneous application of the strong' hepato- carcinogen, N-nitrosodiethylamine, none were found. One liver tumor, and many tumors of the nasal cavity were produced, however. OPTICAL BRIGHTENERS + UV LIGHT E• Brighteners • Brighteners + UV Vehicle • # I ,/,///• x// //// // h /,A #3 / / • i 50 I00 SKIN TUMOR INCIDENCE (%) Figure 2. Tumor incidence in mice from repeated topical application of three optical brighteners and exposure to ultraviolet light. Data from Bingham and Falk (27)