Document LKrpbBBZ0Ljz5KbZgwdw232Q5

(^oivmunJ^oUwmn N E W B R U N S W IC K . N. J. 0 8 9 0 3 August 20, 1976 Mr. Armond Welch, Panel Administrator Division of OTC Drug Evaluation (HFD-510) Bureau of Drugs Food and Drug Administration 5600 Fishers Lane Rockville, Maryland 20852 Dear Mr. Welch: RECEIVED 2* W 6 OFDStafrfOSE Qr / 5 5 Bureau of food and DrugAdministraeon/DHEW Mr. C. Burnett, Associate Director Research, Toxicology, of Clairol Research Laboratories, forwarded to ijie toxico logical data on resorcinol which he wanted submitted for review by the Antimicrobial II Panel when appropriate. As such, I am enclosing six published papers relating to toxicological studies conducted with resorcinol. Perhaps you would arrange to have the data distributed to the Panelists. With all best wishes, Sincerely yours, GHS/tm cc: Mr. C. Burnett Gavin Hildick-Smith.M.D., F.R.C.P. Director, Medical Aifairs 0T& DW& Fd Connet. Toxicol. Vol. I t , pp. 641-648. Pergamon Press 1973. Printed in Great Britain August 1973 Study of Long-term Percutaneous Toxicity and Carcinogenicity of Hair Dyes (Oxidizing Dyes) in Rats* H. J. K inkel and S. Holzmann Bioicleticcs Department, Dattelle-lnstilut e. V., 6000 Frankfurt (.Main) 90, Am Rmerhof 35, Bundesrepublik Deutschland (Received 21 December 1972) Abstract--The long-term percutaneous toxicity and carcinogenicity of /Molucncdiamine, resorcinol and /u-diaminoanisolc, which arc all important constituents of hair dyes, were studied in rats. The substances examined, p-tolucncdiaminc alone in an appropriate vehicle ora mixture of all three compounds similarly formulated, were applied to the shaved dorsal skin of the rats twice weekly for 2 yr, in doses considerably greater than those used in practice. One group of control animals remained untreated while a second was treated with the vehicle alone. After the application period of 2 yr, surviving animals were observed for another 6 months. There were no indications that the substances examined had any adverse effects. Behaviour, feed intake, body-weight changes, blood counts, activity of scrum glutamic-pyruvic transamin ase and results of the bromosulphthalcin test for liver function were comparable in the control and experimental groups. No pathological formation of Heinz bodies or mcthacmoglobin was associated with application of the dye components. Similarly, experimental and control animals .showed no difference in respect of their mean lifespan or the type and incidence of tumours. No tumours or other skin reactions occurred at the site of application of the substances and histopathological studies of the liver, kidneys and lungs provided no evidence of degenerative change or functional disturbance. Acute studies involving sc or ip injection of the substances confirmed the lack of effect on crythropoicsis and mcthacmoglobin formation. IN T R O D U C T IO N O f (lie aromatic diamines, aminophcnols and phenols commonly included in preparations used for dyeing human hair, the most important is the oxidizing dye, p-toluenediamine. p-Phenylencdiamine is another major component in some countries, while other para and the corresponding meta compounds are normally used in much lower concentrations as shading dyes. Reports of tumour induction by some dyes administered to experimental animals by various routes (Saruta, Yamaguchi & Matsuoka, 1962) has provoked some suspicion about the carcinogenicity of substances of this type. Ito, Hiasa, Konishi & Marugami (1969) reported the formation of hepatic carcinomas in Wistar rats fed diets containing 01 or 0 06% //i-tolucnediamine. In the work reported here, the percutaneous toxicity and possible carcinogenic effects of /Molucncdiaminc and of a mixture of p-tolucncdiaminc, resorcinol and m-diaminoanisole in concentrations used in practice in hair-dyeing to achieve particularly dark shades were studied in rats over almost the whole of their lifespan. These investigations were conducted on behalf and with the assistance of Elida-Gibbs Gm bH, Hamburg, L'Orcal, Haarkosmetik und Parfmerien Gm bH, Karlsruhe, Hans Schwarzkopf Gm bH, Hamburg,' Thcra-Chcmic Gm bH, Dsseldorf, and Wclla A G , Darmstadt. 641 M2 l i . J . K IN K liL and S. HOLZMANN EXPERIM ENTAL ., Materials. The .substances examined were prepared by incorporating /Moluencdiaminc at a concentration of 4 % (model substance T) or levels of 3 % /Moluencdiaminc, 0-75% resorcinol and 0-75"' m-diaminoanisolc (substance T R A ) in a carboxymcthylccllulose ( C M C ) slime (4 % Tylose I I T supplied by AK.U, Arnhem, The Netherlands), to which 0-5% pure sodium sulphite was added to prevent premature atmospheric oxidation of the amino components. The two dye bases were available commercially as the sulphates from Farbwerke llocchsl A G , Frankfurt a.M., were therefore partly dissolved and con verted with 25% ammonia solution in the absence of iron and, with stirring, dissolved completely in the C M C slime to form slightly brown, transparent jellies with a pH of about 10-1. These were filled into aluminium tubes coated inside with a protective lacquer, in which form they were stable for several years. A colourless control jelly (substance O) was similarly prepared, part of the ammonia being replaced by ammonium sulphate in an amount equivalent to that formed in the experimental preparations by neutralization of the basic dye sulphates (Table 1). A ll three formulations were therefore comparable in free ammonia content and pH, although in each ease control analyses showed that part of the free ammonia was lost during preparation and the filling of the tubes (Table 1). Table 1. Composition o f experimental formulations Componcnt/Analysis Concentration (% w/w) in substance T TRA O Component Tylose I IT Sodium sulphite Ammonia, 25% p-Toluened amine* Resorcinol /-Diaminoanisolc* Ammonium sulphate Deionized water 40 05 no 40 -- -- as formed by neutralization 78-5 T o ta l.. . . 100 0 40 0-5 12-5 30 0-75 0 75 as formed by neutralization 78-5 1000 Analysis Ammonia: total (100%), titrated free (100%) ' PH 2 95 1-96 10t 2-90 1-96 101 Used as sulphate, calculated as free base. 40 05 8-5 _-- ___ -L. 3-7 83-3 1000 303 1-99 10-2 Animals and diet. Spraguc-Dawlcy rats front a closed randomized stock were obtained from Dr. K . !;. Mollegard-1 Jansen, Havdrup-Kopenhagen, Denmark, for the skin-painting study. The rats (about 3 months old and with a mean body weight just under 100 g) were given a 2-wk adaption period befotc treatment was begun and were maintained throughout on a standard diet (Ahromin-R) and water given ad lib. Injection studies were carried out on 6-month-old rats of the Wistar-Elbcrfcld strain bred at Rattcllc. 1*1 K C U I ANLOUS STU D Y OT HAIR D YLS IN HATS Jl.xpcrmental design a n il conduct 643 , S k in application stud y. A preliminary 4-wk experiment established (hat twice-weekly applic.ition of (lie lest substances lo the skin of rals caused no local irritation. In the main expnitnent, three groups, each of 50 male and 50 female rats, were treated with substance T or substance T R A or given no treatment, while a fourth group of 25 males aiid 25 females was treated with the control substance O. T o simulate conventional hair-dyciiig techniques, substances T , T R A and O were mixed with an equal volume of 6% hydrogen peroxide immediately before application. This was done in batches to'ensure that the material was applied to the skin within 10 min of mixing. "I lie solution (0-5g) was applied lo the shaved dorsal skin overan area of 3 X 3 cm and was left on the skin for 30 min, during which the animals were kept apart and fitted with a rubber collar lo prevent any ingestion of the applied material. As under practical conditions, the painted area was then cleaned with a commercial agent and washed with water, and the animal was rubbed with a towel and dried under an infrared lamp. This procedure was repealed twice weekly for 2 yr, after which surviving rats were observed for a further 6 months. The rats were weighed weekly and observed regularly for any change in behaviour. Red and white cell counts and haemoglobin and mcihacmoglobin levels were determined in blood samples from ten males from each group at 3-monthly intervals, and at 4-monthly intervals liver function was checked by means of a bromsulphthalcin test and the determin ation of scrum glutamic-pyruvic transaminase activity (using the d-scalc Biochcmica-Tcst Combination, liochringer A G , Mannheim), again in ten males from each group. Trans aminase activity was also determined in random samples taken from tumour-bearing animals. At month 7 and 16, blood samples from three animals in each group were examined for Hem/, bodies by two methods, the conventional Nile-blue sulphate stajining and a fluorescence microscopic-detection method using morin. All rats that died during the experi ment or were killed when moribund and all survivors, which were killed by decapitation, were autopsied and all tumours were examined histologically. Sections from the application site, liver, kidneys, lungs and any organ that appeared abnormal at autopsy from a total of nine or ten males and ten females from each group killed randomly and singly at intervals of 6, 9, 12 and 24 months during the study were stained with hacmtjtoxylin and cosinTor micro scopic examination. In je ctio n stud y. llein/-body counts and mcihacmoglobin determinations were carried out in groups of 2-4 rats given a single sc injection of 0-75-24-0 mg /Moluencdiaminc sulphatc/kg, or the same dose repealed daily for 3-5 days, a single ip injection of the same compound in doses of 4-32 mg/kg, or a single sc injection of substance T R A in a 1:1 mixture with hydrogen peroxide in a dose equivalent to 1-5 or 3 mg/Moluencdiaminc sulphatc/kg (I mg /Mohicnediamine sulphate being equivalent to 0-555 mg free base). Sodium sulphite was added to the aqueous solution of /Moluencdiaminc sulphate as in the formulation of substance T, used in the skin painting study, and the pH of the preparation was adjusted to 7 by addition of ammonia. RESULTS Growth, fo o d intake anti m o rta lity The behaviour of all the rals was normal and there were no marked differences between the various groups in respect of body-weight changes (Tubfc 2) and food intake^ The mean 6 M 11. J. K IN K L L a n d S. 1I0LZ M A N N lifespan and mortality rate during the long-term experiment were comparable in all experi mental and control groups. 'I alite 2. Mrun hotly hTights o f ynwps o f tonhol rots and o f rats treated with p-tohicncdiaminc formulations by twice-weekly applications to the skin fo r 2 yr 1ic.ilincnt ot None (control) Substance O (control) Substance T Substance T R A None (control) Substance <) (control) Substance T Substance IR A 86 .1 15(50) yi !: 13 (25) ss J: 15 (50) 9U 17 (50) 8.1 :1: 13 (50) SI 11 (25) 82 I-, M (50) 81 -! 16(50) Rody weight* (g) at wk 52 104 Stales 456 49 (36) 452 45 (18) 443 47 (43) 446 53 (41) Females 289 40 (44) 269 33 (19) 282 33 (42) 280 37 (42) 437 71 (26) 421 6 4 (9) 409 64 (27) 411 59 (27) 287 -i- 50 (21) 258 41 (10) 292 47 (14) 277 34 (23) 130 298 35 (9) 361 (3) 343 63 (8) 328 35 (9) IMS 41 (7) 1260 (2) 278 59 (6) (247 31 (7) O Vehicle only T -- /j-Tolucncdiaminc formulation T K A /i-Tolucncdiaminc-resorcinol-/-diammoanisoIc formulation Values given arc means ft) J : standard deviation { \ f lS ( x 1) --.\S(x)H(n--l ) ) for the numbers of animals (/i) indicated in parentheses, tllc i'.inumg of the experiment. /Itte iiu ilo lu x y All the hacmatological data recorded were within the normal ranges. No pathological melhuemoglohin formation was observed even at 3 min or 4 hr after ip injection of a single highly toxic dose of 32 mg /-tolucncdiaminc sulphatc/kg. No Heinz bodies \Vcrc found in annuals treated with the experimental substances for 7 or 16 months or in those given a single sc injection of 1-3 mg p-tolucncdiaminc sulphatc/kg. A very small number of Heinz bodies was found 2 -3 slays after administration of several of these sc injections and also after the single sc injection of a comparable dose of T R A (1-2/1000 erythrocytes), but even in a dose of 12 or 24 mg/kg given ip, p-tolucncdiaminc sulphate caused little or no increase in Heinz bodies. In a positive control experiment, phenylhydrazinc dihydrochloridc given ip as a 1% aqueous solution in a dose of 3 mg/kg body weight produced an incidence of Heinz bodies of 500-700/1000 erythrocytes. Liver function No impairment of liver function was detected in the experimental groups. Bromsulph thalein excretion by the liver was comparable in experimental and control groups and the serum transaminase values were not increased in the treated animals or in any of the random samples from animals with tumours. Autopsy and hislopa/holoyy Autopsies revealed no pathological changes attributable to either of the longt-tcrm treat ments. The main causes of death were pneumonia and lymphadenitis, the latter having the symptoms of a Klebsiella infection and being associated with abscess formation but not l'LRCU I ANI OUS STUDY OT HAIR DYES IN RATS 645 Tabic 3. Tumour incidence i i control rats and rats trcutctl with p-tolucnediuminc formulations by twice-weekly applications to the skin fur 2 yr 1 Treatment Rats with tumours No. of Sex rnts/group No. % of Total group no. Tumours Type (no.) Time of appearance* (months) None (control) M 50 Fibrolipoma Lymphosarcoma Hypcrkcratoiic papilloma Polymorphocellular sarcomat 20 23 29 24 50 12 24 15 Round-cell sarcomaf 11 Round-cell sarcoma 19 Cystadenomas (2) 22,29 Fibroadenomas (8) 16-29 Adenomas (2) 17,24 Fibroma 17 Substance O M 25 14 1 Round-cell sarcoma 18 F 25 4 16 4 Round-cell sarcoma 24 Carcinoma solidum 29 Fibroadenoma * 25 Fibrosarcoma 12 Substance T M 50 Fibromas (2) I.ipoma Round-cell sarcoma Lymphosarcoma 18,31 24 21 21 50 10 20 11 Fibroadenomas (8) 15-26 Carcinoma solidum 17 Adenocarcinoma 23 Adenoma 28 Substance T R A M 50 Lymphosarcoma Adenoma Fibroma Adrenal tumour 29 29 18 25 50 14 28 15 Cystadenoma 18 Fibromas (2) 18, 25 Fibroadenomas (6) 18-29 Fusoccllular sarcoma 15 Lymphosarcomas (2) 28 Hypcrkcratotic papilloma 12 Carcinoma solidumf 21 Squamous cell carcinornaf 27 . O -- Vehicle only T -- p-Tolucncdiaminc formulation T R A -- p-Tolucncdiaminc-resorcinol-i/i-diaminoanisolc formulation * 1ime at which tumour was detected, calculated from beginning of experiment. 1With mtastases in the peritoneal cavity. -L. wyy'fpTwpi/jf/vi . I',,M,*. .i.*.-*--t.ifa. "P 6'Ki JI. ]. KIN K1L and S. IILZMANN with any " non-specific irritation'' (llulin, 1966) of the lymph nodes in the form of lympho*'.cnous tumour cell propagation and formation of mctastascs. Both of these diseases occur frequently in the Sprague - Dawlcy rats used, but were no more common in the experimental groups than in the controls. Histopathological studies similarly revealed no tissue changes that could be ascribed to treatment with the dyes. Tumour incidence. Table 3 shows that tumours only became clinically detectable during the second half of the life of the rats. O f the 350 Spraguc-Dawlcy rats used in the study, 52 (15%) developed tumours and six had more than one tumour, but the incidence, the pro portion or males and females affected and the types of tumours were comparable in the experimental and control groups (Table 3). The 59 tumours found included 16 dilTcrcnt types (Table 4), of which fibroadenomas were by far the most frequent, followed by fibromas, lymphosarcomas and round-cell sarcomas. The two skin tumours (hypcrkcratotic papil lomas) and ten sc tumours (six fibromas, one fibrolipoma, one fibrosarcoma, one lipoma and one round-cell sarcoma) were not found in the area of application of any of the substances. Table Identity and location o f tumours in control rats and rats treated with p-toluenediainine formulations by twicc-wcekly applications to the skin for 2 yr Type of tumour Location of tumour Untreated controls Sex.. ..M F No. found in Animals treated with substance 0 T TRA MF MF MF Adenocarcinoma (m) Adenoma (b) (V cin o m a solidum (in) Fibroadenoma (b) FibroNpoma (b) fibroma (b) f ibrosarcoma (m) llypcikeiaiotic papilloma (b) C'ystadcnoma (b) Lipoma (b) Lymphosarcoma (m) I'liacocliromocyloma (b) Squamous cell carcinoma (m) I'olymoi phoccllular sarcoma (m) Round-cell sarcoma (ivt) Fusocdlular sarcoma (in) E E E E Ms Ms Ms ES E Ms M E M M M 2 8 1 1 1 2 1 1 2 Total number of tum ours.. . . 4 15 ( 1 1 1 l(s) 14 ! 11 11 86 2 12 1 *1 1 ) 12 1 1 1 1 5 1! 4 15 ni " Malignant b = Jlcnign E " Epithelial tissue M =* Mesenchymal tissue s *= Subcutaneous tissue S = Skin DISCUSSION Although the amount of solution applied to the animal and the frequency of application both exceeded those likely to be encountered in practice, no evidence of carcinogenic or other toxic effects was detected in the experimental animals. The relatively high tumour incidence was to be expected in these Spraguc-Dawlcy rats, which arc somewhat susceptible I'I.KCU l'ANLOUS STU D Y OI HAIR DYLS IN' RATS 647 to spontaneous tumour formation and were examined or kept under observation for up to 2-5 yr. The tumour incidence and lifespan were comparable in all the experimental and control groups, however, and no tumours developed at the site of application. The other criteria studied (behaviour, feed consumption, body-weight changes, blood parameters, scrum activity of glutamic-pyruvic transaminase and bromosulphthalcin tests for liver function) were also comparable in experimental and control groups. Although there is c\ idence, including the development of allergic reactions in hypersensi tive subjects, that substances of the type examined penetrate through the epidermis and also into the hair shaft, no signs of skin irritation, such as the erythema, ocdcina or scar-tissue formation described by Carson, Weinberg Si Goldhamer (1965), nor any increase in cell proliferation in the skin were observed. Kiesc, Rachor & Rauschcr (1968) showed that phenyicnediamincs were absorbed through the skin of dogs when applied in vehicles used in the dyeing of human hair and appeared unchanged in small amounts in the urine. Absorption was reduced when hydrogen peroxide was added to the mixture before appli cation. Kiesc & Rauschcr ( 1968) demonstrated the absorption of p-toiucncdiaminc through the human scalp during dyeing of the hair with a mixture of p-tolucncdiaminc sulphate with resorcinol and hydrogen peroxide. The absorbed material was slowly excreted in the urine, principally as A'./V'-diacctyl-p-tolucncdiaminc. It was not known, however, whether the p-tolucncdiamlnc was absorbed as such, or was liberated in the body from some compound formed during the oxidation of the p-tolucncdiaminc-resorcinol mixture. REFERENCES Carson, S., Weinberg, M. S. & GoMliamcr. It. (1965). An improved method for testing the safety of hair dye preparations. J . Soc. rosnu'l. Chan. 16, 747. llu lin . T, O. (1966). Morphologie der Tumormctastnsicrung. Deutscher Krebs-Kongress, Mnchen. Ito. N' ., llia s n , Y ., Konishi. Y . & Marugami, M. (1969). The development of carcinoma in liver of rats tie.iled with m-loluylcnedianiinc and the synergistic and antagonistic effects with olbcf chemicals. Cancer fies. 29. I 157. Kiese, M., Rachor, Marga & Rauscher, Llli (1968). The absorption of some phenyicnediamincs through the skill of dogs. 7u\ 1C. (ip/'/. Pharmue. 12, 495. Kiese. M. &. Rauscher. 1Hi (1908). The absorption of p-tolucncdiaminc through human skin in hair dyeing. Toxic, opi'l. Pharmac. 15, 525. - Saruta, N ., Yamaguchi, S. k. Matsuoka, T . (1962). Sarcoma produced by subdermal administration of mcuplicnylcncdiaminc and mctnphcnylcncdiaminc hydrochloride. Kyushu J . med. Sei. 13. 175. Ktude sur le rat de la toxicit percutane long terme et des proprits car cinognes de teintures capillaires (colorants oxydants) Rsume--On a tudie sur des rats la toxicit percutane long terme et le pouvoir carcinogne de la p-tolunediaminc, du rsorcmol et du m-diaminoanisol, produits oui sont d'importants composants de teintures pour les cheveux. On les a appliqus--ta p-loltincdiainint seule dans un porteur approprie ou chaque produit dans des mixtures de compositions similaires--deux fois par sema inc, et teci pendant 2ans. la peau rase du dos des rais. Les doses taient beaucoup plus fortes que celles utilises dans la pratique. Un groupe d'animaux tmoins n'a pas t trait et un second groupe n'a t traite qu'au seul porteur. Aprs les 2 annes de traitement, on a prolong de 6 mois l'observation des animaux survivants. T," '-'Ciy T rcu-.- V'AI TTVrUTHrm<.w s Fil Cnsinct TomcoI. Vnl. !.V pp 1S5 557. Pcrgnrmm Prcw 1975. Printed in Grcm Britain LONG-TERM TOXICITY STUDIES ON OXIDATION HAIR DYES C . Bu rn ett. B. L a n m a n , R. G io v a c c iiin i, G . W olco tt and R. Sca la Cosmetic. Toiletry, Fragrance Association, Washington. D.C. and M . K eplinger industrial Bio-Test Laboratories, Chicago. III., USA (Received 8 Deceipher 1974) Abstract--Three oxidation hair dye formulations, mixed with hydrogen peroxide as in use. were tested for long-term toxicity and carcinogenic activity by topical application to groups of 100 mice weekly or every alternate week for 18 months. Chemical intermediates present in the various formulations were p-phcnylcncdiaminc. 2,5-tolucnodiaminc sulphate, resorcinol, m-phenylencdiaminc. 4-diaminoanisolc sulphate and 2.4-toluencdiaminc. The latter compound has been shown to produce malignant hepatomas when fed to rats in a sub-optimal diet at a level of IM% for 35 wk. In this study none of the formulations produced evidence of systemic toxicity or carcinogenicity. INTRODUCTION 1672 consecutive patients with biopsy-proven cancer The natural vegetable and mineral substances used for centuries to colour the hair have largely been re placed by " permanent" synthetic oxidation dyes. As a result of many technical improvements, the use of oxidation hair dyes has increased dramatically in the last 20 yr. accounting for 75% of the h air colouring products sold today. The dyes are formed inside the hair by the oxi of the skin. This incidence of prim arily sunlightinduced basal-cell carcinomas of the haired scalp area has not changed since the condition was first reported iit 1878 (MacDonald & Bubendorf. 1964). The purpose of this study was to determine whether any toxicity is associated with long-term cutaneous exposure to important ingredients or reaction pro ducts in oxidation hair dyes Included in the study dation of colourless intermediates, principally aro was one compound, m-toluene diamine, which has . matic diamines and aminophenols. to imines which been shown to produce hepatoma in rats when fed at react very rapidly with colour modifiers or couplers a level of 0 1 % in a sub-optimal diet for 35 wk (Ito. to form indo dyes (Corbett & M cnkart, 1973). Although these reactions occur rapidly, it has been shown (Kiese 8c Rauscher. 1968) that in small amounts (less than 0-2% of the applied dose) un reacted intermediates may penetrate the skin during the dyeing of human hair. Heavy industrial exposure to certain aromatic amines has caused bladder cancer in the past (Clayson & Cooper. 1970). Hueper (1963) has shown that although these tumours arose'from prolonged skin absorption, there was no evidence of any effect of Hiasa, Konishi & Marugami. 1969). This compound, which is no longer used in hair dyeing in the United States, was used for a number of years at a con centration of less than )% in oxidation dye formula tions. The extent of its skin penetration and metabo lism has not been determined. Oxidation dyes, which are applied to the hair as a reacting system, can only be meaningfully evaluated from the toxicological viewpoint by the same route of administration as occurs in use, namely skin application. In hair dyeing, the user is exposed to unrcacted these chemicals on the skin itself. The reason for this now seems obvious in view of studies showing that aromatic amines appear to require metabolic conver para components (p-phcnylencdiamine. p-aminophenol) and couplers (m-phcnylenediaminc, 2.4-diaminoanisole, m-aminophenol, resorcinol, hydro- sion to the A-hydroxy derivative to induce cancer (M iller & M iller. 1969). We have found no published data demonstrating that A-hydro.xylation occurs when the aromatic diamines used in permanent hair dyes are oxidized and there is no evidence indicating an increase in malignancies in the likely target organs in women using these dyes regularly for many years (National Cancer Institute, 1971). Evidence that the use of hair dyes is not a signifi cant factor in the production of scalp malignancies has been provided by D r. F. Urbach. Skin and Cancer quinonc). as well as to reactive intermediates, particu larly quinonc imines ami the various indo dyes. The oxidation of the pant components by hydrogen per oxide is relatively slow and has been shown to be incomplete even after 24 hr. However, the coupling reactions are extremely fast and thus there is no sig nificant build-up in the concentration of the interme diate quinonc imines. Nevertheless, it is important that the testing protocol should involve exposure to these extremely reactive species. "this cannot be done by testing with the intermediates themselves since, in Hospital o f the Temple University School of Medi cine (personal communication 1973). Over a period o f 15 yr. Dr. Urbach found only 16 persons (eight men and eight women) with scalp tumours out of aqueous media, they undergo rapid polymerization and or hydrolysis (depending on the pH) to give pro ducts that do not arise in the oxidation dyeing pro cess. Sim ilarly, if aged reaction mixtures were used. F.C.T. f3/3--1> 353 354 G. Burnt tt. I). Lanman. R. G iovacchini, G. Wolcott, R. Scala and M. K lplinclh there would be little or no exposure to the quinone mines since they are no longer formed once the hydrogen peroxide is depleted. The feeding of aged reaction mixtures may, at first sight, appear to be an attractive compromise. How ever, under conditions prevailing in the stomach (a pH around 2 and a temperature of about 38'C). the indo dyes would undergo rapid hydrolysis to give a mixture of para components, hydroxybenzoquinoncs and reaction products thereof. The quinone products would not arise in other metabolic pathways and the ingestion route is thus unrepresentative of the meta bolism of indo dyes. Furthermore, no intermediate quinone mines would be involved in this mode of testing. Thus, from the chemical standpoint, topical application of fresh reaction mixtures is the only pro tocol involving exposure to all the materials that are inherent in.the application of oxidation hair dyes to human subjects. T h is can only be achieved by careful choice o f composites not by the use of individual pre cursors in the presence of peroxide. On the basis of these considerations, the protocols and test composi tions used in this study were selected. EXPERIMENTAL Materials. The hair-dye intermediates used in these studies were obtained from several suppliers and either met standards as purchased or were subjected to further purification. ;>-Phcnylcnediamine (lot 1143) and resorcinol (lot 312), purchased from Nyanz.a, Inc.. Lawrence, Mass., were over 99% pure. />-Tolu,enediaminc (lot K38671) from Naftone, Inc., New York, assayed at 98-5%. The 2,4-toluenediamine and mphenylenediamine from Eastman Kodak, Kingston, Tenn., were recrystallized from ethanol and isopro panol to a final purity of over 98%. On a moisturefree basis, the 2.4-diaminoanisole sulphate (lot A56442) from Lowenstein lnc,, Brooklyn, New York, assayed at 96-97%. Animals and diet. Random-bred Swiss Webster mice of both sexes, 6-8 wk old, from Charles River Labora tories were housed individually and were given R al ston Purina Laboratory Chow and water ad lih. Experimental design and procedure The mice were randomly allotted to nine groups each of 100 mice and one group of 250 mice. There were equal numbers of each sex in each group. The exper imental design is shown in Table L Three oxidation hair-dye formulations were tested (Table 2), each for mulation being mixed with an equal volume of 6% hydrogen peroxide just prior to use. Once weekly or once every other week for 18 months. 0 05 ml of the mixture was applied to the skin of the mid-scapular area, which was kept free of hair by shaving with an electric clipper. The dose was spread just enough to prevent run-olT. After each application all unused material was discarded The vehicle-control group received the base with no added dye intermediate. The positive-control group was housed in a separate room and received weekly applications of 005 ml of a solution of 7,12-dimethy!benz[]anthracene (D M B A ) in acetone. It is not always possible in a topical study to select the single dose of the positive-control carcinogen that will give the best indication of the sensitivity of the Tabic I. Experimental design of IH-lnonih carcinogenicity study pi Itair-dye formulations, innohufig topical application to min' lest Group material* UC None PC DMBA A PP-7588 B PP-7586 C PP-7585 D PP-7588 E PP-7586 F PP-7585 G Vehicle H Vehicle No. of micc/groupt '250 too 100 100 100 100 100 100 100 100 Dose} (ml) 1 005 005 005 005 005 005 005 005 0-05 Dosing frequency __ W W w w F F F W F UC = Untreated control PC = Positive control DMBA = 7.12-dimethvlbenz[ij]anthraccne W = Weekly F = Fortnightly ` For identification o f'P P - samples, sec Table 2. tEach group consisted of equal numbers of males and females. iThc applied dose of the hair-dye was 005 ml of a I : Imix ture of the formulation and 6% hydrogen peroxide. The concentration of the positive control (DMBA in acetone) was adjusted so that 005 ml provided a weekly dosage of 50ng DMBA for 6 months. 10/rg for 4 months and 50 pg for 8 months. test animal. One may simply apply an overwhelming dose producing a high incidence of tumours in a short period of time, or it may be more meaningful in terms of the usual human exposure to apply small doses over longer periods and thus achieve the best estimate of the minimal tumorigenic dose. Poel (1956) has shown that 125 pg benzo[a]pyrene applied topi cally to mice as a single dose is not carcinogenic in the normal life span of the mouse, while I /rg adminis tered three times weekly for 40 wk (total 120 /ig) pro duced tumours in nine of 42 mice by wk 40. Klein (1956) has shown that as little as L2 /tg D M B A applied topically to Swiss mice over a 10-month period caused a 13% incidence of tumours. In the study reported here, a split-dose schedule was used, with mice receiving 50 ng DMBA/\Vk for 6 months, followed by 10 pg/wk for 4 months and finally 50 pg/wk for the remainder of the study. The mice were observed daily for signs of toxicity. Each week they were weighed, the skin graded for irritation and papillomas and other gross lesions were noted. The progress of papillomas and other lesions Tabic 2. Experimental hair-dyeformulations tested in an IHmonth carcinogenicity study in mice Ingredient Oleic Acid Isopropanol Sodium sulphite Ammonia 2.5 Tolucncdiamine sulphatc Resorcinol 14-Toluencdtamine base 2.4-Diaminnanisole sulphate m-Phen\lcncdiamine base Deionized water Level r ,, i in experimental formulation PP-7588 PP-75H6 PP-7585 500 3fX) 0-20 MX) .MX) 1 50 U-40 o:o -- a 70 5-00 MX) <v:o MX) MX) 1 50 040 _ 038 -- XO-70 500 3 00 o:o MX) 3-00 150 0-40 -- -- 0-17 80 70 T ox icily o f oxidation hair dyes 355 was charted monthly. Animals that died or that were killed because of general debility were aulopsied and examined histopabiologically when possible. At the termination o the study all survivors were weighed and killed and a gross autopsy was per formed. A sample of blood taken from a random selection of at least four mice of each sex in each group, except those groups treated only every other week, was analysed for haematocril and hae moglobin, and mean corpuscular volume and mean corpuscular haemoglobin were calculated. A complete blood count, including differential, platelet and reticu locyte counts, was done. The livers of ten animals in each group were weighed and liver/body weight ratios were calulatcd. From each animal were taken sections of skin, thyroid, lung, gastro-intestinal tract (three sec tions), spleen, pancreas, liver, kidneys, adrenals, uri nary bladder, ureter, mesenteric lymph nodes, sternal bone marrow and gonads. These tissues were fixed in 10% buffered formalin. Slides were prepared from haematoxylin/eosin-stained tissues of 13 male and 13 female mice in each test, vehicle-control and positive control group and of 32 mice of each sex in the un treated control group and were examined microscopi cally for evidence of hair dye-induced toxicity. Also, every gross tumour-like lesion, regardless of location, was examined microscopically and classified. RESULTS AND DISCUSSION There were no overt signs of systemic toxicity in any of the dye-treated groups. The survival varied from 58 to 80%, except in the positive Controls in which only 21% of the mice were alive after 18 months. Average body weights were comparable in all groups throughout the study. The absolute and relative liver weights were in the range of normal values for all groups and none of the differences were statistically significant by analysis of variance (Sncdecor, 1966). All of the haematological data recorded were within normal limits. Table 3 shows mean ter minal body weights, liver/body weight ratios and the percentage survival for each group. Disease processes commonly seen in aged mice of this random-bred strain were observed in all groups with essentially the same frequency. These included chronic murine pneumonia, amyloidosis, chronic Tabic 3. Survival, mean terminal hotly weights and liver weight/lwdy weight ratios in mite treatedJar IS months with topically applied liair-dyc formulations Group* Survival o IX months t%i Terminal body weight fg)+ Males Females l.ivcr weight/ body weight ratio* UC 78 37-8 34 6 0-060 PC :i 40-6 U-5 0-05X A (O U>() 313 (HHrf* H 65 >1 X 31-1 1)067 c Oh .'7 1 33 0-06 1) 68 37 7 32 7 0-056 K 69 37-4 35-0 0-053 r HO .36-2 32 5 0-062 G 59 37-4 3'Fk 0-064 H 6: 36-k 33 1 0-062 For identification of groups, see Tabic 1. tValues for body weights arc means for all survivors and those for liver weight expressed relative to body weight are means for ten mice in each group. Tabic 4. Incidence of alvcologenic adenomas and carcinomas in mire treated topically with hair-dye formulations for IS months No of mice with lung tumours Group* No. of micc/group '* Sex A|vcologcnic adenoma Alvcologenic carcinoma UC 125 125 PC 50 50. A 50 t 50 50 50 C SO 50 D 50 50 E 50 50 F 50 50 G 50 so H ' 50 50 M F M F M F M F M F M F M F M F M F M F 23 14 2 s 4 X 6 > X 7 0 .3 9 13 18 9 12 X 11 X For identification of groups, sec Table I 4 2 1 2 1 1 0 2 1 l 1 1 1 0 2 1 1 1 2 4 focal nephritis, focal lymphoid infiltration in lungs, liver and kidneys and testicular atrophy. The inci dence of alvcologenic adenomas and adenocar cinomas (Table 4) was comparable among the various control and test groups and was within the range of control values for this strain of mice (Percy & Jonas. 1971). The type and distribution of all other tumours in the groups treated weekly and in the untreated, positive-control and vehicle-control groups are shown in Table 5. ' For the first 5 months of the study, moderate alope cia occurred in about 50% of the animals of each sex in each of the three groups receiving the hair dyes once weekly. This was not seen in the groups treated every other week or in the vehicle or positive controls. The condition diminshed with! time and after i 1 months hair growth appeared normal. Microscopic examination of sections taken fit the final autopsy showed that the skin and its appendages were normal in all mice in these groups. There were no papillomas or carcinomas of the skin except those in the positive control group. The other tumours of the skin, in whatever groups they occurred, were all outside the treated area. The skins of the mice in the positive control group showed no response following the weekly application of 50 ng of the carcinogen for 6 months followed by 10 gg/wk for an additional 4 months. However, when the dose was increased to 50 /rg/wk, there was an almost immediate response arid tumours appeared on six animals. T h is dose proved to be irritating to the skin and mortality increased sharply from 18 to 46% in 3 months. Although the response was somewhat less than expected, it did serve to demonstrate the susceptibility of the animals to the action of the car cinogen. No evidence of carcinogenic activity was seen in the hair dyes tested in this stuejy. The results of a companion study which was carried out by the US Food and Drug Administration and which included groups of mice receiving topical applications of 3% m-to!uenediamine in water/isopropanol, were also 356 C. Burn.tt, B, Lanman, R. G iovacchini, G. Wolcott, R. Scala and M. K eplinger Tabic 5 Distribution of tumours other than ulwologcnic adenomas untl carcinomas in untreated anil positive controls and in mice treated once weekly with hair dyes Type of tumour No. of micc/group... uc 250 No. of mice affected and location of tumours in group* PC AB C loot 100 100 ^ 100 G 100 Lymphosarcoma Fibrosarcoma Fibre mil Adenocarcinoma Grunulosa-ccll lumour Leiomyosarcoma RclieuInr'Cell sarcoma Myxofibroma Squamous carcinoma Papilloma Adcnoacanlhoma Osteogenic ni reoma Serosal fibroma 1 M. liver 1 M. kidney 1 1-, lung 1M. dermis 1 F. dermis I F. dermis 1 F. cervix 1 F. body wall 2. mammary gland 1. ovary 1. uterus 1 F, spleen 2 F. general 1 M, dermis 1 F. body wall l. uterus 1 M. kidney i. prostate. 1. uterus 1 F. skin and lung 2 M. skin 1 F. liver, spleen, lung, lymph node 1 F. liver. lymph node 1 F. dermis 1. proslule 1 F, dermis 2, mammary gland \ mammary gland ' 1 M, liver 1. mammary gland 1 I*. lung 1 F. stomach ` For identification of groups, see Table 1. tTwo positive-control mice with skin papillomas in the treated area were lost as a result of death and autolysis. negative for carcinogenicity and will soon be published (H . Blumenthal, personal communication 1974). Included in these studies was a compound pre viously shown to be a potent hepatocarcinogen when fed to rats. T h is raises the question again of the rele vance of testing topical products by oral administration and of the significance of studies showing the toxic effects of prolonged oral dosing : with high levels of ingredients of products that have been shown to be safe when tested under reasonably exaggerated conditions of use. It is often argued that oral administration is the means of choice to achieve the exaggerated dosage levels necessary to detect, in . small numbers of test animals, effects that may occur in a few exposed individuals in a large population. This oral dosing frequently assumes massive propor tions bearing no reasonable relationship to normal exposure and is often limited only by the amount the animal can tolerate while remaining compensated and apparently normal, T h is kind o f testing may give false and misleading information for a number of rea sons: (I) Blood levels w ill undoubtedly occur that would never be seen in product use. and may result in abnormal metabolism and pathology; (2) due to the effects of pH and gut bacteria, there could be absorption of molecular species that would never occur with topical application; (3) metabolic conver sion to a more (or less) toxic substance may occur, or may occur at a rate quite different from that fol lowing topical application; (4) chemical reaction with essential nutrients in the ingesta can produce toxicity as a result of deficiencies and imbalances that bear no relationship to any inherent toxicity of the com pound tested; (5) the normal gut flora may be drasti cally altered causing additional nutritional problems; (6) the target organs and excretory patterns may not be representative of those occurring during normal exposure. As indicated in the Introduction, these considrations have particular relevance iri the case of oxidative hair dyes. They are. however, valid also for other systems. Conclusions No evidence of toxicity, or carcinogenicity, was seen when several oxidative hair-dye formulations were tested in an 18-month topical study in mice. Included in the formulations were 2.5-toluenediamine sulphate, p-phenylenediamine. resorcinol. 2.4-diaminoanisole sulphate, m-phenylenediamine and 2.4- toluenediaminc. The latter compound has produced liver cancer when fed to rats in a sub-optimal diet. Safety evaluation studies of products must involve actual conditions of use to produce meaningful results. REFERENCES Clayson, D. B. & Cooper. If. H. (1970). Cancer of the uri nary tract. Adv. Cancer Res. 13, 271. Corbett. J. F. & Menkart, J. (1973). Haiq coloring. Cutis 12. 190. ( Hueper, W. C. (1963). Chemically induced skin cancer in man. Conference on Biology of Cutaneous Cancer. Natn. Cancer Inst. Monogr. no. 10, p. 381. Ito. J., Hiasa, Y., Koniski. Y. & Marugami, M. (1969). The development of carcinoma in liver of rats treated with m-tolylenediaminc and the synergistic and antagonistic effects with other chemicals. Cancer Res. 29, 1137. Kiese. M. & Rauscher. E. (1968). The absorption of p- toluene diamine through human skin in hair dyeing. Toxic, appl. Pharmac. 13, 325. Klein, M (1956) Induction of skin tumors in the mouse with minute doses of DM BA alone or with croton oil. Cancer Res. 2, 123. MacDonald. E. & Bubendorf. E. (1964). Tumors of the skin. In Proceedings oj the 7th Annual Conference on the Epidemiologic Aspects o f Skin Cancer, p. 23. Year Book Medical Publishers, Chicago. Toxicity o f oxidation hair dyes 357 Miller. Elizabeth C. & Miller, J. A. (1969). Studies in the mechanism of activation of aromatic amine and amide carcinogens to ultimate carcinogenic electrophilic reac tants. Ann. N.Y Acad. Sci. 163, 731. National Cancer Institute (1971). Preliminary Report, Third National Cancer Survey. DHEW Publication No. 72-128, p. 11. Percy. D. H. & Jonas. A. M. (1971). Incidence of spon taneous tumors in CD(R)-I HaM/ICR mice. J. nalii. Cancer Inst. 46, 1045. Poel, W. (1956). Carcinogens and minimal carcinogenic doses. Science, N.Y. 123, 588. Snedecor, G. W. (1966). Statistical Methods. 5th Ed. p. 291. Iowa State College Press, Ames, Iowa. D ERM A L CA RCIN O G EN ICITY STUDY BY MOUSE-SKIN P A IN T IN G W ITH 2 ,4 - T O L U E N E D I A M I N E A L O N E O R IN R EP R ESEN TA TIV E HAIR D YE FORM ULATIONS Albert L. Giles, )r., Choong W. Chung Division o f To xico lo g y, Food and Drug Adm inistration, Department o f Health, Education, and W elfare, Washington, D .C . Choudari Kommincni Pathobiology Research Branch, Experimental Biology Laboratory, National Environmental Research Center, Environmental Protection Agency, Research Trianglc Park, North Carolina i he chronic to x icity o f 2,4-toluenediamine (2 ,4 -T D A ) alone o r in com bination with a hair dye com plex (2,5-toluenediamine, p-phenylenediamine, and resorcinol) was studied in Swiss-W cbster mice o f both sexes b y a skin-painting technique. The predom inant neoplasms seen in these m ice were prim ary pulm onary adenomas and adenocarcinomas. Shin neoplasms were seen in m ost groups o f mice, ini hiding untreated control mice. Statistical analysis o f the incidences o f skin neoplasms among the various groups o f mice did not show any significant differences. The 2,4-TDA alone o r m ixed with the hair dye com plex did not produce any abnormal proliferation and maturation o f the squamous epithelium o f tlic shin. The 2,4-TD A under o u r experim ental con dition s was fo un d to be non toxic and noncarcinogenic to the skin o f mice. INTRODUCTION Aromatic diamines are used as human hair dyes; one such dye is 2,4-toluenediamine (w-toluylenediamine; 2,4-TDA). Umeda (1955) reported that repeated subcutaneous injections of 2,4-TD A induced rhabdomyo sarcomas in 100% of rats treated. Rats fed diets containing 2 ,4 -T D A developed hepatocellular carcinomas (Ito et al., 1969). T o assess the carcinogenicity and toxicity of 2,4-TD A and other cosmetic ingredients of hair dyes when used topically, a 2-yr mouse-skin painting study was conducted. Both use levels and exaggerated concentra tions of 2,4-TDA were tested. The authors gratefully acknowledge the editorial assistance of Ms. Lettic Stewart and the technical assistance of Mr. Edwin D. Sneed and Mr. lames G. Yates. We are also indebted to Ms, |anct Springer and Ms. Ann Ducca, Division of Mathematics, for the statistical analysis, and to Dr. William S. Monlux of the Division of Pathology, I DA. We express our sincere thanks to f)r. Clilton II. Wilson, Ronald S. Wiight, and Martin Stutsman, |r., of the Division of Cosmetic Technology, I DA, lor their determination of 2,d-lolucncdiamine in tlie hair dye lorinulalions. Requests lor reprints should he sent to Albert L Giles, |r., Division ol Toxicology, f ood and Drug Administration, Bureau of hoods HF F-15d, 200 C Street, S.W., Washington, D.C. 20204. 433 journal of Toxicology and Environmental Health, 1:433-440, 1976 Copyright 1976 by Hemisphere Publishing Corporation 434 A. L. G IL E S , JR . ET A L. M ATERIALS AND METHODS Preparation of Chemicals Form ulations containing 0.2 or 0.6% 2 ,4 -T D A in hair dye complex and hair dye base, hair dye base alone, and 6% peroxide were provided by the Cosmetic, Toiletry and Fragrance Association, Inc. ( C T F A ) , Washington, D .C .; these samples were representative of formulations marketed in the United States. Com positions of the complex and base are given in Table 1. The concentration of 2,4-TD A in the formulations was determined monthly for 12 months by a gas-liquid chromatographic method (Goldstein ct al., 1968) and remained essentially constant. For mouse-skin painting, 2,4 -T D A formulations and base alone were diluted with an equal volume of 6% peroxide and were used within 10 min. Standard 2 , 4 - T D A 1 was prepared as a 6% solution by dissolving 4 0 0 mg in 6.5 ml of water-isopropanol2 ( 7 5 .1 :11.1) in a 20-ml vial with continuous stirring for 2-4 hr over a magnetic stirrer. Aliquots were diluted with an equal volume of either distilled water or 6% peroxide for use in skin painting. Water-isopropanol mixed with an equal volume of 6% peroxide was used to paint the skin of a control group. 9,10-Dimcthylbenzanlhracene (D M B A ),3 the positive control com pound, was prepared as 0.02 and 0.005% solutions in acetone.4 These solutions were stored at 4 0 F and used as needed. Acetone from the same stock bottle was used to paint the skin of a control group. 1Supplied by the Clairol Co., New York, N .Y., which obtained it from the general market. ' Reagent grade, Fislicr Scientific Co., Pittsburgh, Pa. 3ICN KS K Laboratories, Inc., Plainvicw, N.Y. 4 Reagent grade, J. T . Baker Chemical Co., Philadelphia, Pa. T A B LE I. Composition of 2,4-Tolucncdiamine Hair Dye Em mutations Ingredient Weight (%) Dye complex 2,5-Tolucncdiaminc sulfate /;-Phcny lenedi.i mine Resorcinol 2,4-Tolucncdiaminc Base Oleic acid Isopropanol Sodium sulfile 29% Ammonia Deionized w.iier Total 3.00 1.50 0.40 0.20 or 0.60 5.00 3.00 0.20 6.00 80.70 nr 80.30 too M OUSE-SKIN PAIN TIN G WITH A H A IR D YE 435 T A B L E 2. Treatment Protocol Croup no. 1 2 3 4 5 6 7 8A 8B 9 Treatm e n t0 0.2% 2,4-TD A in haii dye com plex in base 4 P 0.6% 2,4-TD A in hair dye com plex in base + P Hair dye base + P 6.0% 2,4-TD A in Wl + distilled water 6.0% 2,4-TD A in Wl + P Wl + P 0.02% 9,10-D M BA in acetone Acetone 0.005% 9,10-D M RA in acetone Unircated controls {no hair clipped) Dose (ue/wk) (2,4-TD A or DM BA) 50 150 0 1,500 1,500 0 0 0 2.5 0 Initial number of mice Males Tmales 28 28 28 28 28 28 28 28 28 28 28 28 28 28 14 14 28 28 76 17 S k in painting was done once a week. Abbreviations: 2,4-T D A = 2,4-tolucncdiam inc; DM BA = dim ethylbcnzanthracenc; P = 6% peroxide; Wl = water-isopropanol (75.1 =11.1). Animals and Treatments Swiss-Webster mice of both sexes, obtained from the National Institutes of Health mouse colony, were 4-7 wk old and weighed 1 5 -2 0 g at the beginning of the experiment. The mice were allotted to nine groups consisting of 28 mice of each sex, except as noted in Table 2. Each mouse was housed individually in a plastic, shoe box type cage kept in a room maintained at 74 2 F and was weighed weekly for the first year. F o o d 5 and water were available ad lib itu m . The hair in the intrascapular region of each mouse was closely clipped with an electric clipper6 each week, one day prior to painting. Each week, 0.05 ml of the appropriate test or control solution (Table 2) was applied to this region by using a tuberculin syringe without a needle. Each mouse was kept under a hair dryer7 without heat to dry the applied solution before being returned to the cage. When signs of marked irritation, ulceration, or tumor formation were evident, the application of the chemical was discontinued until the skin looked "n o rm al." T o minimize variations in treatment, the order of painting was adjusted so that a mouse painted first in one week would be painted last the next week and vice versa. T he hair of untreated controls was not clipped. 5 Purina Laboratory Chow for M ice, R ats, and Hamsters, Ralston Purina C o ., St. Lo uis, Mo. ` Model A 2 , size 40, Ostcr C o rp ., M ilwaukee, Wis. ' Model 202, Oster C orp ., M ilwaukee, Wis. 436 A. L. G ILE S , |R . ET A L. Histopathology Tissue specimens were taken from all major organ systems and tumors of .lice found dead during the study (if not markedly autoly/cd) or sacrificed when moribund or at 2 yr, the termination of the experiment. An overdose of ether was used to sacrifice the animals. These specimens were fixed in buffered 10% formaldehyde. Sections were cut 6 pm thick from the paraffin blocks and stained routinely with H& E before histopathology examination. Complete histopathology examinations were performed on the following organs: skin, spleen, liver, bone marrow, lungs, myocardium, pancreas, kidney, bladder, stomach, large and small intestines, testis and seminal vesicle of males, ovary and uterus of females, pituitary, thyroid, adrenal, and brain. T h e following special stains were used to resolve pathology diagnostic problems: phosphotungstic acid-hematoxylin-(PTAH), Masson, Van Gieson, Alcian blue, periodic acid Schiff (PAS), methyl green pyroninc, toluidine blue, and von Ko ssa's slain. RESULTS Body weight gains of mice in treated groups were not significantly different front those of mice in the untreated control groups. The survival rate of all mice was erratic and did not show sex dependency. Fro m group 8A, which was treated with acetone only, tissues of five female mice were subjected to histopathology examinations. In the untreated control group, although there was a disproportionately large number of male mice (76) at start of the experiment, only 21 males were subjected to histopathology minations. T w o different concentrations of D M BA were used in this study. The 0.02% solution of D M B A produced marked irritation of the skin of mice; ulceration and 15% mortality were observed within 4 wk after the start of the experiment. The utilization of 0.005% D M B A solution obviated these problems, and the compound at this concentration proved to be a potent skin carcinogen. These observations confirm our earlier results (Giles and Byron, 1968). T h e male and female mice in all groups developed both malignant and benign neoplasms (Table 3). There were no statistical differences between the sexes in the total number of neoplasms or in the incidence of neoplasms of a particular organ or type. Primary lung neoplasms predominated in all groups except in the positive control groups 7 and 8B (Table 4) and were cither adenomas or carcinomas. The incidences of skin neoplasms are given in Table 5; these neoplasms included squamous cell carcinomas, fibrosarcomas, myxosarcomas, myoepithelioma, liposarcoma, osteogenic sarcoma (untreated control group 9), and other types. O f the total number of neoplasms in the DMBA-treated T A B L E 3. Incidence of Malignant and Benign Neoplasms in Male and Female Swiss-Webster Mice (2,4-TDA Skin Carcinogenicity Study) Group no. a 1 MF 2 MF 345 MF MF MF 67 MF M 8 F A8B9 MF MF M F No. o f mice examined Total no. of neoplasms No. of mice with neoplasms No. of mice with multiple neoplasms No. of malignant neoplasms No. of benign neoplasms 19 13 19 15 13 17 21 9 11 14 16 14 17 13 14 5 1 1 13 21 12 15 10 11 9 12 14 20 8 4 16 10 4 14 16 7 3 12 15 10 3 11 7 10 6 10 13 15 7 4 12 9 4 13 13 7 3 10 9 9 3 33 12 2 1 51 0 3 1 0 1 3 00 2 5 10 10 7 7 5 8 8 12 4 4 10 5 1 12 15 6 1 10 12 7 3 5 3 4 4 4 6 84 06 532 I 12 2 32 0 a See Table 2 for description of treatment. T A B L E 4. Incidence of Primary Lung Neoplasms in Male and Female Swiss-Webster Mice (2,4-TDA Skin Carcinogenicity Study) Group no. 438 No. of mice examined Total no. of lung neoplasms No. of mice with lung neoplasms No. of malignant lung neoplasms No. of benign lung neoplasms 1 234567 8A8B 9 M F M F M F M F M F M F M F M F M F MF 19 13 19 15 13 17 21 9 11 14 16 14 17 13 14 5 11 13 21 12 10 3 8 5 9 10 17 6 4 8 7 2 2 4 52 436 3 10 3 8 4 9 10 15 6 4 8 7 2 2 4 52 436 3 7 2 6 2 7 4 11 3 4 5 4 0 1 3 40 2 36 3 3 1 2 3 2 6 63 0 3 3 2 1 1 12 20 00 See Table 2 for description of treatment. T A B L E 5. Incidence of Primary Skin Neoplasms in Male and Female Swiss-Webster Mice (2,4-TDA Skin Carcinogenicity Study) Group no 0 1 2 3 4 5 6 7 8A MFMF MFMF MF MF MF MF SB MF 9 MF 439 No. of mice examined Total no. of skin neoplasms No. of mice with skin neoplasms No. of malignant skin neoplasms No. of benign skin neoplasms 19 13 19 15 13 17 21 9 1 1 14 16 14 17 13 14 5 11 13 21 12 2 2_ 1 I 2 00 0 3 0 0 12 10 0 0 8 4 2b 0 2 2_ 1 1 2 00 0 3 0 0 12 10 0 0 8 3 20 2 2_ 1 1 2 00 0 3 0 0 12 10 0 0 S4 10 0 0- 0 0 0 00 0 0 0 0 0 0 00 0 0 10 See Table 2 for description of treatm ent. Dashes indicate that mice were not examined. ^Neoplasm seen at the angle of the lower jaw was considered to be of dermal origin. 440 A. L. G ILE S , ) K . E l A L. 'Hips 7 and 8B, 77 and :)4%, respectively, were skin neoplasms; statistical alysis of the incidence of skin neoplasms among all other groups did not show any significant differences. The adrenal glands, the reticuloendothelial system, and the female reproductive organs developed a few neoplasms. DISCUSSION Swiss-Wcbslcr mice manifest a high incidence of spontaneous lung neoplasms (Stoats, 1972; C. Kommincni, unpublished data, 1972). Statistical analysis of the incidence of lung neoplasms among the groups in this study did not show any significant differences; thus the 2 ,4 -T D A and other hair dvc ingredients did not augment the development of primary lung neoplasms. The incidence of skin neoplasms did not show statistically significant differences in any of the groups under test except for the positive control groups 7 and 8B (Table 5) Moreover, no deleterious effects, including signs ol skin irritation, abnormal proliferation and maturation of squamous epithelium of the skin, or changes in the roots of the hair shafts, were seen in the skin sections examined from the painted mice exhibiting no skin neoplasms. The absence in these mice of the type of neoplasms that occur in rats treated with 2 ,4 -T D A (Umeda, 1955; llo cl al., 1969) is of particular interest. The difference is probably due to (1) mode of treatment, (2) quality of diet, and (3) the species of animals used; the route of administration may be most important. 2,5-Toluenediaminc and 2,4diaminoanisole, which arc similar hair dyes, did not cause any skin toxicity oi carcinogenicity when given to rats by the percutaneous route (K in kcl and ' il/mann, 1973). A chronic skin-painting study in mice with 2,4-TD A , -diaminoanisole, or 2,4-phenylenediamine in hair dye formulations failed ,o show evidence of toxicity or carcinogenicity in mice ( C T F A , personal com munication). Our present results in mice are in accord with these reports. R EFER EN CES G iles, A. I... ]r. .m il Byron. VV. R . 1968. A mouse-skin painting study of two colors and several cosm clieal ingredients compared with D M BA . Inlcrbureou B y tin e s 5:105-1 1 I. G oldstein, S ., K o p l, A . A . and I cinlam l, R . 1968. Analysis of oxidation dyes in hair colorants by thin-layer and gas chromatography. In Proceedings o f the jo in t conference on cosm etic sciences, pp. 19-38. Washington, Q .C..: I he to ilet Goods Assoc. Ito , N ., Hiasa, Y ., Konishi, Y . and Marugami, M. 1969. The development of carcinoma in liver of rats treated with toluylenediam inc and the antagonistic effects with other chemicals. Cancer Res. 29:1 137-1 145, K in k c l, H . |. and MoUmann, S. I9 7 3 . Study of long-term percutaneous to xicity and carcinogenicity of hair dyes (oxidizing dyes) in rats, f o o d Cosm et. T o xico l. 11:64 I-6 4 8 . Stoats, |, I9 7 2 . Standardized nomenclature for inbred strains of mice: Fifth listing. Cancer Res. 3 2 :1 6 0 9 - 1 6 4 6 . Umeda, M. 1955. Production of rat sarcoma by injections of propylene glycol solution of oM oluylcncdiam inc. Carm 46:597-601). R eceived M arch 20, 19 75 A ccep ted M ay 7, 1975 TOXICOLOGY AND APPLIF.D PHARMACOLOGY .TO, 7-13 (1974) Lack of Toxicity and Carcinogenicity of Som e C om m only U s e d C u t a n e o u s A g e n t s 1-2 F ri'j Stf.nhack and PniLippr: S iiuoik F.pptcy Institute fo r 'Research in Cancer, University o f Nebraska M etlical Center, Ornalia, Nebraska 68105 R eceived Septem ber 21, 1973\ accepted M a rch 22, 1974 Lack of To xicity and Carcinogenicity of Some Commonly Used C ut aneous Agents, S iin h a Ck , F . a n d S iiu u ik , P. (1974). basical. Appl. Pharmacol. 30, 7-13, The potential carcinogenicity and toxicity of several comm only used cutaneous agents (ben/ophenone, propylene glycol, isopropyl myristate, resorcinol, 2-cthyl-l,3-hc.xanediol. p-aminobcnzoic acid and pyrogallol) were studied in female Swiss mice by administering repeated applications of the chemicals on the skin for the life-span o f the anim als. Tum o rs seen in both control and treated anim als were mainly lym phom as, hemangiomas o f the liver and lung adenomas, as well as tumors of other organs. A .statistically significant increase in Iumor incidence caused by the chemical treatment was not seen. Skin lesions, slight inflamm a tion and ulceration were seen, but no persistent cutaneous abnormalities occurred. A few skin tumors were seen in treated areas as well as in untreated areas and in control.anim als. Thus a carcinogenic or toxic potential which would affect the use o f these agents in man was not detected. Skin cancers arc one of the largest groups of environmentally induced tumors; both radiation (sunlight) and certain chemicals arc responsible (llucper, 1942). These tumors may be reproduced readily in animal models as demonstrated with coal tar and petroleum oil constituents ( I lartwell, 1951; Sltubik and Hartwell. 1969). Because regulations governing the testing of externally used chemicals arc less extensive than those governing drugs, little is known about their toxicity and possible carcinogenicity. In this study, 7 compounds that may come into contact with human skin were selected because of their use as solvents, insect repellents, fixatives, sun protecting agents, dispersing agents, ns well as chemicals used in photographic develop ing ami dye manufacture. These agents were applied to mouse skin, and local changes, systemic response and tumor occurrence were noted. M ETHODS Chemical.';. Benzophenone, propylene glycol, resorcinol. 2-cthyl-1,3-hcxanediol,3 /7-aminobcnzoic acid, isopropyl myristate and pyrogallol4 and 7.12-dimeilnIbenz- 1 This w o rk was supported by Pu blic H ealth Service contract PH-43-6S-959 from the N ation al Cancer Insiilule. ` ,>;in f ,hi<; w o rk "-as reported at the T w e lfth A n n u a l M eeting o f the Society or T o x ic o lo e v N ew Y o r k C ity , M arch 18-22. 1973. ' 4 Purchased from A ld rich Chem ical Com pany, M ilw aukee, W isconsin. 4 Obtained from N utritional Hioeltcmical Com pany, C leveland, O hio. Khl T l')74 hy Academic l*rc*\. Inc. ol reproduction in any form reserved. -d in Great llritam 7 TABLIZ I N u m ber of D if fe r e n t T yp es o r H is t o l o g ic a l l y V e r if ie d T umors in S w iss M ic e T r ea t ed w it h D if f e r e n t C h em ic a ls Treatment (50 mice con- centra tion) Toial No. '0 of tumorConcentration bearing (in acetone) animals % of tumorhearing animals Total No. Lung Licer Skin o f tumors L> mphomas adenomas hemangiomas Thym om as tumors Other n t ln iia c k an d s iiu iiik Pyrocallol 50 22 44 2S 11 8 3 _ 6" 25 IS 36 23 12 6 3 -- -- 5 25 50 30 20 4 1 -- -- 2 - E ih d - 1,3-he\ancdiol 100 32 64 40 14 11 9 -- 2 50 28 56 34 17 7 2 -- 2 6' 10 23 46 2S 11 5 i -- 1 1' Prop' lene d> col 100 20 40 23 9 7 4 50 26 52 31 13 13 2 10 26 52 30 15 7_ i 1 __ 19 2-- P i -- 6' Isopri'psl im rRtate 100 21 42 27 14 7 3 50 25 50 30 13 6 4 10 21 42 25 13 5 1 i -- 2` 2 1 41, i 1 4' /-A mi no hen zoic acid 10 20 40 24 S 8 5 5 22 44 26 16 4 1 i IS 36 25 8 8 1 2 __ r 3 1V 5 -- lo Resorcinol Henrophenone Acetone 50 27 54 37 25 1 44 24 5 20 40 29 IS 13 16 7 4 -- __ S " 6-- -- -- 1- - A* 5 1 -- 1 50 14 28 IS 6 6 1 -- __ 4' 25 16 32 22 11 3 5 26 52 38 . 15' .3 1 1 1 1 5' 1 1 16" 100 2~> 44 31 12 9 "> -- 1 DM B \ 0.5 59 7S 86 64 1 -- 75 -- U n t r e a t e d conliol 1150 mice i -- 64 42 72 26 17 4 6 3 16- * 2 mammary fibroadenomas, 2 m ammary adenocarcinom as. 1 suhcutancoi. ..angiom a, I osaran i ' 2 mammary adenocarcinomas. f I ovarian hem angiom a. 1 o saran cystadenom a. ! lung ad enocarcinom a. 1 hem angioma ot the c a r, I m.eve o.td cn o m a. ` 2 forestomach papillom as, 1 subcutaneous fibrom a. 2 subcutaneous fibrosarcomas. * 2 m am m ary adenocarcinom as, 1 m am m ary fibroadenom a, 1 adnc.sal carcinom a ol the eye. ! subcutaneous : m o m a. 1 spleen hem angioma. 1 I forcstomaeh papillom a. * 2 mammary adenocarcinomas. * 1 forestomach papilloma. 1 I osaran cystadcnocarcinom a, 1 ovarian fibrom a. 1 ovarian hemangioma, 2 subcutaneous fibrom as, I forcstomaeh papillom a. ` I subcutaneous fibrom a. I udnesal adenocarcinom a ol the eye. CARCIN O G EN ICITY OT CUTANEOUS AGENTS 1 I forcstomaeh papillom a. I mammary fibroadenoma. 2 ovarian hemangiomas. 12 ovarian hemangiomas. 1 m am m aiy fibroadenoma, I forcstomaeh papilloma. " 1 forcstomaeh papilloma. " I subcutaneous libren-,y som a. I cecal adenocarcinom a. I ovarian hemangioma. I subcutaneous fibroma. r 2 mamm ary fibroadenomas, 2 mammary' adenocarcinom as. I uterine leiom yosarcom a, 1 ovarian fihiom a, ! uterine hem angioma, I subcutaneous fibroma. " I forcstomaeh papillom a. I subcutaneous fibrosarcom a, I m am m ary adenocarcinom a. I ovarian hemangioma. ' I ovarian hem angioma. I spleen hemangioma. I subcutaneous hemangioma. I ovarian (brom a. I m am m ary fibroadenoma. 1 2 spleen hem angiom as. I retroperitoneal lib ro m yso m a. 1 ovarian eystadenoma. 1 2 ovarian fibrom as, I spleen hemangioma. I ovarian hem angioma, I subcutaneous fibrosarcom a. " t spleen hem angiom as. 3 ovarian hem angiom as. 2 m am m ary adenocarcinom as, I m am m ai y fibroadenom a. 1 subcutaneous hem angiom a. 2 subcutaneous fibrosarcomas, 2 uterine adenomas. 1 metine adenocarcinom a, i retroperitoneal librom yvom a, I vaginal hemangioma. r I mammary fibroadenoma. 1 m am m aiy adenocarcinom a, 2 ovarian hemangiomas, I subcutaneous fibrosarcom a, I lorcstom ach papillom a. **t m am m ary fibroadenom a. 2 mammary' ad enocarcin om as, i .subcutaneous tib io m a. 3 subcutaneous fibrosarcom as, 1 subcutaneous hem angiom a, I retroperitoneal tibromy.soma. 2 renal adenomas, 2 ovarian hemangiomas. 2 ovarian lih io in as. 2 forcstomaeh papillomas. 10 SII.NMACK AM) SIIUIIIK anthracene' (D M B A ) were dissolved in acetone.5 The concentrations tire shown in Table I . I he chemicals were user! as obtained from the manufacturer. Animals. Seven-week-old female Swiss mice from the kpplcy colony were used. The animals were randomized prior to the start of the experiments, and the httermales were separated. There were 50 animals per concentration group. Animals were housed in plastic cages with commercial heckling/' 10.per cage, and fed a commercial diet7 and water ad libitum. One hundred thirty-live untreated animals served as controls, as well as 50 animals trented with the solvent alone and 50 animals treated with 7 ,12-dime thy Ibcn/anlhraccne as a positive control. 'Trcainicni. The chemicals (0.02 ml) were dropped on the dorsal skin between the Hanks twice a week on a I-inch square area which was shaved regularly. The animals were checked weekly; and all lesions, as well as tumors, were recorded. Animals were allowed to die spontaneously or were killed vvjien moribund. Complete autopsies were performed on a11 animals. The skin from all animals, all grossly observed tumors and other lesions in the lungs, livers, kidneys, etc., from treated as well as control groups were, studied histologically. Formalin-fixed paraffin-embedded specimens were cut and stained with hematoxylin-cosin and other stains when needed. The statistical significance of the results was evaluated using the methods presented by Armitnge (1971). R ESU LTS Anim als treated with pyrogallol showed no significant changes, in the skin. Hyper plasia, ulceration or inflammation did not occur. The numhet of tumors produced was not significantly dilie rent from that in untreated controls (Tabic 1). Skin tumors were not found The number and distribution of the tumors were similar to that in other groups. Lymphomas, lung adenomas and liver hcmagiuiomas were the most common tumors. The.survival rate was slightly lower than in the untreated control group, but comparable to other groups (Table 2). Anim als treated with 2-ethyl-l ,3-hcxanedio! had a 46-64 % tumor incidence which was higher than for untreated controls (42% ). However, this incidence was not speci fically related to tumor type ( fable I). Isopropyl myristatc led to 2 skin tumors, I on the abdomen and I on the eyelid Table 5). Anim als treated with resorcinol showed skin lesions with ulceration, inllammation and hyperplasia. Two skin tumors were seen, I on the hack and I on the cur as well as I subcutaneous fibrosarcoma. Anim als treated with benzophenone had a small number of skin tumors. One animal receiv ing the 25% concentration had a tumor on the lip and 2 animals treated with the 5 % concent ration had tumors on l lie dorsal skin. In animals receiving repeated applica tions of DM I$A a large number of skin tumors were seen- -squamous cell papillomas, kcraloacnnlhomas and squamous cell carcinomas (Table 2). The number of tumors of other organs was small owing to the short life-span of the animals (7 able 3) as a result of the skin tumors. ' fis h e r Scientific C o m p an y, F a ir b w n . Nov/ Jersey. h S;m -i-ccl. I'dMon Processing Com pany. In c., P axio n , Illino is. 7 Wa>uc. Allied M ills. Chicago, Illinois. CARCINO GIJN ICri Y O r CUTANEOUS AGENTS TABLE 2 S urvival Rath of S wiss M ice T reated with D ifferent C hemicals O O O O O O IO O O O O O O 'O --r-l OJ -- ' O ro 0 1 D n o O O O O O O Oo " --o I vr> vO O n VO O co vo i~" co O' n (TV fN vo W o cj -- ro o - T - f -- co 0 1 0 1 nO O co n o n on -- i o v o c o r ^ r ^ O c o c o O' ------------------------ r i - - n - ------------------- c i -- -- r t -- __ r-i o -- r - ON D- n ^ W O ' r l _^ _ N n n ri ri o -- r i ro OI r ir in CO O ' ri ri so o O O' -- O io> n r j r i r 1 -- - I __________r , IT. c>d r 1 ^ 00 ^ n r 1! CM C I fO ro co ro O O r i C \ r-" o O' O C' w co O' o rO ro C l ro r I ro 'O ro ro ro C l -- -- rI ro _ TZoo IT; *0 It It IT i r C J r^ o ^ r* ri oo oncmo odsoc- o- C I ro TO ~T 'T O' ' r f Tj- CO ro FO CO T f r}- ro ro co -"1 r i C l rO r f r- C HUH 22 ^ *0 -- ^ LO LO co ID IO| CO O f ' ' co VO or^r-ct m m- *j o- o- o ^ ^ rt o- o- o- m t *r coroor o- CO ro -t -i-'T T-r o- K-r Oo'' O-r' CoO- C-Or rOt' -Tr" rox- V-Ot roo- voO- CoO- O*o CrnO --o- V^Cr iCrO CD C*rl O icvoovo o O' O' OOO' a> c- r- CO ' O' 'rT V | T *T ID >D -T TT T f c|- C O O C s a s O o v c o r - CO CO O O' O' O O VO r n O' O' O' CO O' P i ST9 Ct 9 9 9 O Os o O CO ooas O O O O On r - Os >o CO r 1 ' t ~T 'O </D *o c, - r 'DI ID O' ID ID T f on >o 'O t TT TT O' O' O' 9 P 22 9 -9 O O O O o O' O O O O O O O' r- o O o I V* /". O^ 'O /) */~k Dj V i IDI 'D i.iy , O' */D 'O o V t !/> >/! O' O' O' o V | </. ' O vo vo O O O O O O O O O 0 * 0 1 'O C 1 0 * 0 -- 0 * 0 -- O VO -- -- Ovc. *coi 'O *oocoio oo O no U Ipcfi oC oo Oo -3 O -C u oc. Co. OCL 12 S 1I NUACK AND S IIU B Ik TABU-. 3 N u m iilr an d L o c a tio n o r D iit e r k n t T um o rs in -Sw iss M ic e in R i c a tio n to T r e a t m e n t Concent ra- Trealmcnt lion" (50 micc/eonccntraiion) (% ) Location 11istopalhology 2-Cthyl-1 ,3-licxancdiol Isopropyl myrislate /;-Aminobcnzoic acid Resorcinol Benzoplicnonc Acetone Dimethyl benzanthracene Unticalcd control (150 animals) 100 50 10 50 10 25 5 25 5 100 0.5 0.5 0.5 I on dorsal skin. I On tail I on dorsal skin, 1 on car I on eyelid I on eyelid I on abdominal skin I on car 1 on car 1 on dorsal skin 1 on 1i[> 2 on dorsal skin I on tail. I on ear 3I on dorsal skin 20 on dorsal skin II on dorsal skin I on car I on dorsal skin. I on tail Sqiiamous cell papillomas Squamous cell carcinomas Squamous cell papilloma Kci aloacanihoma Squamous cell papilloma Squamous cell carcinoma Squamous cell carcinoma Squamous cell papilloma Squamous cell carcinoma Squamous cell papillomas Squamous eel] papillomas Squlumuis cell papillomas Squamous cell carcinomas Kcrtuoacaiiilionias Squamous cell carcinoma Squamous cell papillomas " Dissolved in acetone. DISCUSSION None of (lie chemicals studied produced a statistically significant increase in skin tumor incidence, when compared to untreated controls, and to those seen alter repealed applications of DM BA. The skin tumor incidence seen in the treated animals (2-4 was similar to that of the acetone control and untreated eoiithol animals. Also, a significant percentage of the tumors were on the tail. ear. eyelid, lip and abdomen, which did not have direct contact with the chemicals, this speaks against a causal relationship between lumnr incidence and treatment. The possibility exists that the mice rub against each other, which might spread the chemical. However, in the test groups 72% were outside the treated area, in the positive control Only S%. Thus from our_results it cannot be concluded that any of the chemicals studied are carcinogenic for skin. Swiss mice are very sensitive to skin tumors as seen in tIre positive control group, which shows a 78% skin tumor incidence after painting with a known chemical carcinogen such as DM BA 10 //g tw-icc weekly. The chemicals administered did not produce lesions in other.organs directly related to the treatment. Resorcinol has been reported to he a goitrogenic agent in rats when injected subcutaneously ( Doniach and Logoilictopoulos, 1453), however, the reported goitrogenic cficct related only to rcsorcinol-diaeelale and resorcinol in peanut oil. The systemic or local cllccts of the other chemicals have not been studied extensiscly. Propylene glycol, a solvent used for pharmacologic studies, was originally suggested ;is a safe solvent by Seidenfeld and Hanzlik (1932). Davis and Jenner (1959) reported --------------u . *j - , .- i. CAKCIN O G I N ir n Y Ol CUTANEO US AGI N IS |3 propylene glvcol lo be less toxic tluin A'.A'-dimelliylformamide ;h k I A',A'-diacel> Iformamidc. I lowcvcr, Zaroxlinski c l al. ( 1971) concluded that propylene glvcol cannot be recommended as ;m inert cosolvenl in pharmacologic studies. Gaunt c l al. (1972) did not report any pathological findings or carcinogenic potential after feeding propylene glycol in concentrations up to 50,000 ppm to rats. Our findings arc similar. Fit/gerald cl al. (1968) reported that isopropyl myristatc had a local cITect when applied on skin, causing erythema, followed by licheni/ation and fissure formation. Histologically, acanthosis, para- and hyperkeratosis, erosions and hemorrhages were seen, which regressed. Our results were similar In that initially cutaneous lesions were seen, but these regressed. Moderate dermatitis has also been reported in rabbits following 2-cthyl-1,3-hcxancdiol treatment (Drai/.e cl al.. 1948). In our studies no persistent changes occurred except for a slight inflammation and atrophy of the skin. ACKN OW LEDGM ENTS The authors thank M rs. D . Janccek for editorial assistance and D r. J . Pattil for the statistical analysis. REFEREN CES A k sii t a g e , P. (1971). S tatistical M ethods in M edical Research, pp. 129-131. Blackw ell, Oxford, England. D a v is, K . an d J en n eh , P. (1959). T o xicity o f three drug solvents. Toxicol. Appl. Pharmacol I, 576-578. D o n ia c h . I. a n d LoGOTiir.TOPOt'Los, T . (1953). The goitrogenic action o f resorcinol in rats. Brit.-J. E\/t. Pathol. 34, 146-151. D R A iz r, J. I I . , A t v i.h i./, E ., W in i r.sn t,, M . F ., W o o d a r d ,G ., H a g a n , E . C . a n d N llso n. A . A . (1948). Toxicological investigations o f compounds proposed for use as insect repellents. A . Local and systemic clTeets following topical skin application. It. Acute Oral I oxicitv. C . Pathological Exam ination. J . Phot m m ol. Pxp. 'liter. 93,26-39. Fti/.G i RAi.D, J. E , K titn /.. S. M ., S o t A iuii in , J. E. a n d K a u m p, D . H. (1968). Cutaneous and parenteral studies with vehicles containing isopropyl myristatc and peanut oil. T o xic o l. Appl. Pharmacol. 13,448-453. G a u n t , I F ,, C a k p a n in i. F . M. B ., C-hasso, P. a n d L ansdovvn, A . B. G . (1972). Long-term toxicity o f propylene glycol in rats. Pood Co.smet. Toxicol. 10, 151-162. H A n r w u .L , .1. L . (1951). S urrey o j Compounds Which Hove Been Tested fo r Carcinogenic /U til ity. PI IS publication 149, United Stales Government Printing Office, Washington, D .C . I U iErrit. W. C. (1942). Occupational Tumors and A llied Diseases. Thom as, Springfield. Illinois. SriDt.M tt.n , M. A . a n d H a n / i i k . P. .1. (1932). T he general properties, actions and toxicity o f propylene glycol. J . Plnnmacol. /-2.x/*. liter. 44. 109-121. , S u im iK , P. a n d 11a r tw!_!.!., J. I.. (1969). S u rrey o] Compounds Which H are Been Tested lo r Carcinoycnic A ctirity. PU S Pub! 149, Suppl. 2. United Slates Government Printing Office, Washington, D .C . Z arosunski. J. 12, Brown, R. K . and Posst.rv, L . H. (1971). Propylene glycol as a drug solvent in pharmacologic studies. Toxicol. Appl. Pharmacol. 19, 573-578. /J d o pC| njOtoo'C C o vUv IO a s ^ clV o rtvC L,/ ..z L TER A TO LO G Y AND PERCUTANEOUS TO XICITY ST U D IES ON HAIR D Y ES 1133 15th Street,N hashington, D.C. 20005 Ilattawan, Wienigan 49071 C. Burnett W. Cosmetic, Toiletry and Fragrance Assoc. , E. I. Goldcnthal, S. B. Harris, F. X. Wazeter International Research and Development Corp. 9200 Leesburg Tnpk Vienna, VA 22180 J. Strausburg, R. Kapp, R. Voelker Hazleton Laboratories America Inc. f Twelve hair dye form ulations were tested fo r system ic to xicity by topical application tw ice w eekly fo r 13 w k to groups o f 12 N ew Zealand w hite rabbits and fur teratologic effects following applications to groups o f 20 pregnant Charles fin e r CD rats on days I, 4, 7, 10, 13, 16, and 19 o f gestation. The three sem iperm anent fo rm ula tions were applied as is, and the nine o xid a tio n d yes were m ix e d 1 : t. w ith 6% hydrogen peroxide ju st p rior to application, as in norm al use. The forlm ilutions included a broad spectrum o f dyes and dye interm ediates used o r considered useful in oxidative and semipermanent hair color products. In the teratology stu d y no biologically sign ificant so ft tissue o r skeleta l changes were noted. Siniilurly, the mean numbers o f corpora lutea, im plantation sites, live fetuses, and resorp tion s p er pregnancy, as w ell as num bers of[htcrs^\vith reso rp tio n s, were not significantly affected b y the dye treatment. \ |V I+ G ^ In the percutaneous to xicity study there was no evidence o f com pound-induced --{sy stem a t i\ e f f e c t s . M icro sco p ic exam ination o f 25 tissues from each anim al gave no indication o f liistom orphologic evidence o f to xicity. No dye discoloration o f urine was seen at any time during the test o r at n ecro psy. So m e o f the d ye groups show ed epiderm al hyperplasia, which was p rob a b ly a re fle ctio n o f slight irrita tio n due to the frequency o f application o f the oxidation form ulations. IN T R O D U C T IO N Q,,- -:;ons of the safety of hair color products have been raised as a result of studies (Am es et al., 1975) showing mutagenic activity in Requests fo r reprints should be sent to C B u rn e tt, C lairol In c ., 2 B lachley Road, ^tam ford, Connecticut 06902. 001 Journal of Toxicology and Environmental Health, 000-000, 1976 C opyright O 1976 by Hemisphere Publishing Corporation I( f1 7* o' 002 C. B U R N E T T ET A L. J Salm onella bacteria. On the basis of these results, Am es and cp-workers speculate that hair dyes may be mutagenic to humans and probably carcinogenic. However, this extrapolation of the results o f tests in' bacteria to human safety has been questioned by the proponents o f the A m es test and other screening procedures (dc Serres, 1975). The regulatory agencies arc taking a deliberate position in this matter and are considering |hc more relevant and weighty evidence emerging from animal studies in evaluating the safety of hair dyes (Anon., 1975). T he hair color industry has performed a number of chronic studies on major dye ingredients over the past 10 yr. T hree such studies (Burnett et al., 1 9 75 ; Kink-el and Holzm an, 1973; Wernick el a I., 1975) showed no evidence of carcinogenicity when a variety of dyes and dye intermediates were applied topically or were fed to rodents and dogs for up to 2 yr. Unpublished data (C. Burnett and co-workers) show that 11 hair dye chemicals found to be mutagenic (Am es et al., 1975) and to cause cytogenic abnormalities in some cases (Venitt et al., 1975) in In vitro tests had no effect on the germ cells of male rats treated with high doses of the dyes in a dominant lethal mutagenicity study. The studies reported here are part of an ongoing hair dye ndustry safety testing program that includes a three-generation reproduction/ carcinogenicity study. . T E R A T O L O G Y S T U D Y IN R A T S Methods Twelve hair dye composite formulations were tested, along with one positive and three negative contro' &ivwtfjcs, The dves were obtained from commercial supplieis and :rc eithei curren iC<^ or arc considered potentially useful in hair color products t o d j L ^ 1he formulations (Tables 1-3) were applied topically at a dose o f 2 ml/kg on days 1, 4, 7, 10, 13, 16, and 19 of gestation to groups of 20 mated Charles River C D female rats (presence of sperm in the vagina coo*--Acred day 0 of gestation). Pilot studies had shown that the potential skin irritancy of most formulations would not permit more frequent application. The 12 formulations tested represented the two major classes of hair color produ c t s f u s a D Three formulations (P-22, P-23, and P-24) 'jvere of die semipermanent type that consist of preformed pigments tjtat are deposited on the hair shaft from solvent-detergent bases. Th e nine remaining formulations were of the type that result in the oxidajtion of colorless intermediates, which arc deposited as colored polym ers in the hair shaft. These oxidation formulas have an am m oniacal soap b is c and i kr U - J 'O T A B L E 1. Com position of Form ulations l^oi -7 4 0 6 , P-21, P-22, P-25, and P-26 L_ Form ulation0 Compounds tested 7401 7402 7403 7404 7405 7406 P-21 P-22 P-25 P-26 p -P hcnylcne d iam inc 3.0 2.0 4.0 1.0 p-TolucneJiam ine sulfate 3.0 6.0 0.25 0.4 0.6 Resorcinol 1.7 1.7 1.0 2.0 0.25 0.2 2.4- l)iam inoani$olc sulfate 2.0 4.0 0.02 2.5- Diaminoanisole sullate 6.0 rzr-Aniinuphenol 0.7 0.7 0.09 0.02 o-A minuphenol 0.3 p -A m in o p h cn o l 1.0 0.2 0.04 / Naphthul 0.5 rzt-Phcnylenediamine 1.5 o-Phenylenc ilia mine 1.0 2-N it r o-p-pi ic ny le ne diamine 1.1 4-N itr o-o-pheny letted iam inc 0.25 2-A m in o -1 -n itro p h en u l 0.4 2 -A m in o -5 -n itro p h en o l 0.5 4-Chlororesorcinol 2.0 p-Aminodiphenylamii^c hydro- chloride 2.0 N-mcthy I-p-ami nophenol stilt ate 1.0 0.05 2 .4 - D iam inoph enol 0.2 2-Mct hylrcsorcinol 1.0 4-A m in o -2 -n itro p h en o l Flydroquinone 0.3 0.2 Pyrogallol 0.4 f-Butyl hydroquinone 0.3 Sodium picramaic 0.1 Toluencti iol (2 ,4 ,5 ) 0.5 Benzcnetriol (1 ,2 ,-I) 0.5 N ,/V-bis(2-hydroxy e th y l)-p ph enyl encdiaminc sulfate 1.0 0.5 2.5- Diam ino-4-m ethyljnisole di-HCI 0.6 6-fly d ro xy ben zo m orpholine 1.1 2-M c[hyl-5-(carbamylinethyl)am ino- phcnol 2.4 3-Car bam ylm eih ylam i nophenol 0.7 Oleic acid 5.0 5.0 15.0 5.0 5.0 5.0 15.0 8.0 8.0 Ispropanol 3.0 3.0 10.0 3.0 3.0 15.0 10.0 3.0 3.0 Sodium sulfite 0.2 0.2 0.2 0.2 0.2 0.2 0.1 1.6 29% Am monia . 6.0 6.0 9.0 6.0 6.0 6.0 9.0 6.0 6.0 G lyce rin e 4.5 Propylene glycol 9.0 5.0 40% D ie thy lene triam ine pent acetic acid (Na salt) 2.4 2.4 35% Sodium bisulfite 1.3 1.3 Ascorbic acid 0.2 Laurie diethanolamide 2.0 Sodium lauryl sulfate 2.95 SulfonateJ castor oil 4.0 Carbitol 5.0 Ethoxylatcd octyl phenol 6.0 Ethoxylatcd nonyl phenol 3.0 Ethoxylatcd cholesterol 1.0 Perfume 0.4 V inyl pyrrolidone copolymer quaternary 2.0 Polyoxylated ethylated fatty alcohol 0.533 Polyoxyethylene (23) lauryl ether 0.04 Perfume 0.0127 Sf- ryi aim cth y l benzyl am monium chloride 0.0667 Silicone 0.533 SDA 40 alcohol 25.0 Eth yl acetate 1.992 Trlethnolam inc 0 .1 0 - 0 .1 5 J AII form ulations Water Q .S . 100. 003 1 1 ' : 'I t1 004 C. B U R N ETT ET A L. J arc mixed with an equal volume of 6% hydrogen peroxide just prior to use. T he hair at the site o f application on the dorso-scapular area was shaved closely the day before the solutions were applied. The negative control animals were untreated but were shaved. Three separate and discrete negative control groups were maintained in order to determine the degree of variability among small groups of control animals and to utilize this information in assessing treatment effects. T A B L E 2, Com position of Form ulation P-23 Chem ical or trade name Basic violet 3 UC black I Solvent black 5 (nig. SS) Acid black 2 (nig. WS blue) Acid black 2 (nig. WS jet) Direct red 23 Solvent black 5 (nig. WS jet) Solvent blue 16 B rill, red safranine M ethylene blue HC brown No. 1 A C red 259 . Solvent black 5 (nig. S S J) Solvent yellow 63 Acid orange.89 Acid black 107 Acid blue 168 Solvent brown 43 Acid red 213 Acid blue 188 HC blue No. 3 Solvcni yellow 90 Acid yellow 12 7 Acid orange 8S 4 -Amino -l-dihydroxy -5-nilrojnihraquinonc 4-Amino-4 -nilroa/oben/ene 2.4- Diniuo-6-chloro-2'-acetamido-4-bis- (2-hytlroxyeihyl)am ino-5'-methoxyatrobcnienc 2.5- Dijnrino 6-niirobcimhi.uole 4-|(5 -m eih ylsu llo n ylb en 2 th iaro le)-2-aro l-/V -(2- cyanocthyI)-/V-2-chloro-2-hydro.xy propyl (aniline Carbitol Triethano lam ine H y d ro x ye th y lce llu lo sc C itric acid Partial sodium salt of /V-lauryl(3-iminodipropionate Nonionic surfactant (dextrol X L IS ) 1,3-Dim e'hyiol-5,5-dim ethyl hydantoin M ethyl paraben `^aier Percent 0.002 0.003 0.200 0.200 0.200 0.200 0.150 0. t 50 0.005 0.100 0.200 0.150 0.200 0.050 0.070 0.200 0.100 0.080 0.030 0.060 0.200 0.100 0.200 0.200 0.200 0.200 0.200 0.200 0.200 5.000 1.500 2.400 0.300 1.500 0.480 0.100 0.050 8 4 .6 2 0 (H ^ {/ f v- y 006 C. B U R N E T T E T A L . skeletal anomalies. Fetal examinations were performed in a random order' with treatment group identity unknown to the examiner. All statistical analyses compared the treatment groups with the control groups. T he number of females exhibiting resorption sites, numbei of females exhibiting two or more resorptions, number of dead or resorbed fetuses, and tlie number of fetuses with soft-tissue or skeletal anomalies and accessory ribs was compared using chi-square test criterion with Y a t e s ' correction on 2 X 2 contingency tables as described by Steel a r d T o n i c ( 1960) or Fisher's exact probability test (Siegel, 1956) as appropriate to judge the significance of difference. T h e mean number o f corpora lutea, implantation sites, live fetuses, and resorption sites was compared by analysis of variance (one-way classifica tion) as described by Steel and Torrie (1960) using Dunnctt's (1964) multiple comparison tables to judge the significance of differences. T he live fetal weights were compared by analysis of variance (hierarchal classification) as described by Steel and Torr ie (1960) using D u n n c t t 's ( 1964) multiple comparison tables to judge the significance of differences. Statistically significant differences between groups were judged valid only when there were significant differences between any one of the dye treated groups and each of the three untreated control groups. Results and Discussion No signs of toxicity were seen throughout the study. Ex ce pt for the changes in the color of the skin and hair at the site of dye application, no irritation or other changes in appearance were seen. Changes in female body weights were similar for rats in the untreated controls and all dye-treated groups. A marked reduction in maternal weight gain through gestation was observed in the rats receiving acetyl- salicylic acid as compared with either the untreated control rats or dye-treated rats. /T) Mean food consumpti on for all groups throughmTf gestation was similar except for rats in the ac et y Isa Iicy I ic acid g r ou p / l h e s e rats showed a moderate decrease in food consumption from days 7 to 13 of gestation. This decrease was not seen from days 13 to 20 of gestation. T he dye formulations produced no significant differences in the mean number of corpora lutea, implantation sites and live fetuses, and the sex ratio when compared with the untreated control groups. No differences between groups were seen regarding the number of females exhibiting resorption sites or mean resorptions per pregnancy. No significant changes were observed regarding soft-tissue anomalies between the dye-treated groups and the untreated control groups. Normal l y occurring skeletal variations were present in all groups; the most frequent variation noted was accessory ribs. '7 . io H A IR D Y E T E R A T O LO G IC AND T O X IC IT Y STU D IES j 007 Ik b ^ A statistically significanl increase in skeletal anomalies appeared in the group receiving fcaiinulation 7402. Nine fetuses of the 169 (described in I I ahles 7 and o;| examined revealed anomalies that arc regarded as minor skeletal changes. Seven of the nine Ictuses displayed notched ribs and two displayed short ribs. These minor skeletal changes were seen in o n l y . 3 of 20 litters examined and arc not regarded as biologically significant. The data arc shown in Tables 4-6. T h e results observed in the positive control group, namely an increase in e mbr yo t o x i c i t y (e.g., increase in teratogenicity, increase in embr yo death, and decrease in fetal weight), arc in accord with the literature concerning the effects of aspirin upon fetal rat development ( Ki mm el and Wilson, 1973; Ki mmel et al., 1971; McCol l ct al., 1965). Conclusion From this investigation, the administration of formulations 7401, 7402, 7403, 7404, 7405, 7406, P-21, P-22, P-23, P-24, P-25, and P-26 every third day of the gestation period produces no embryotoxic or teratogenic effects. The dyes and dye intermediates tested in these formulations include most of those used in oxidative and semipermanent hair color products today in the United States. P E R C U T A N E O U S T O X I C I T Y IN R A B B I T S Methods T h e 12 hair dye formulations shown in Tables 1-3 were applied topically twice weekly for 13 wk to groups of 12 adult New Zealand white rabbits (six of each sex). E x ce pt for P-22, P-23, and P-24, all formulations were mixed with an equal volume of 6% hydrogen peroxide prior to application. The sites of application on the dorsolateral aspects of the thoracic-lumbar area (one on each side of the midline) were alternated to minimize skin irritation. The dose was 1 ml/kg of the 1:1 oxidation mixtures and of those formulations used as is. T h i s was the max im u m dose that could be applied on the side of the rabbits without m n of f. T h e hair at the site of application on the back and sides of each rabbit was clipped short throughout the study. T he application sites on three animals of each sex in each group were abraded on the first treatment day of each week. The rabbits were restrained in holding stocks for 1 hr following each application and were then shampooed, rinsed, dried, and returned to their cages. T h e rabbits in the three independent control groups of 12 rabb ts each were treated identically except that no dyes were applied. T he animals were weighed weekly during the study. Hematologic and clinical chemistry determinations and examination of urine was performed on all animals at 0, 3, 7, and 13 wk. These studies included determination of complete - TABLE 4. Sum m ary of Teratology Study In Rats Receiving 7401, 74 02 , 74 03 , 7 4 0 4 , 74 05 , and 7406 Observations Maternal parameters Total no. females Mean no. corpora luiea Mean no. im plantation sites No. females exhibiting resorp tion sites No. females exhibiting 2 or more resorptions No. females aborting Control 1 untreated Control II untreated Control III untreated A c e tyls a licy lic acid (250 mg/kg-day) 7401 (2 ml/kg) 7402 (2 ml/kg) 20 15.35 12.40 13 9 0 20 13.55 12 .1 0 c 14 7 0 20 15.25 13.90 12 6 0 20 16.15 13.25 15 15* 0 20 14.05 13.40 10 6 0 20 14.05 13.30 13 6 0 7403 (2 ml/kg) 7404 (2 ml/kg) 20 12 .85 *' 12.60 14 7 0 20 15.15 1 3 .6 0 11 6 0 7405 (2 ml/kg) 20 15.55 1 4 .3 0 '* 11 7 0 7406 (2 ml/kg) 20 14.55 13.00 t. 14 8 0 Feta/ porometers Mean no. live fetuses/group Mean live fetal weight (g) No. dead or resorbed fetuses (%) Mean no. resorptions/pregnancy Sex ratio, M : F No. fetuses with soft-tissue anomalies (%) No. fetuses with skeletal anomalies (%) No. fetuses with accessory ribs only (%) 10.65 3.38 35 (1 4 .1 1 ) 1775 1 0 6 :1 0 7 10.85 3.58 25 (1 0 .3 3 ) 1.25 1 0 2 :1 1 5 12.50 3.62 28 (10.07) 1.40 1 1 9 :1 3 1 8.70e 2.89* 91 ( 3 4 .3 4 / 4.5 S>' 1 0 0 :7 5 4 (6.35) 4 (6.06) 6 (7 .7 9 ) 21 ( 3 6 .8 4 / 0 (0.00) 1 (0.67) 2 (1.16) 40 (3 4 .1 9 / 75 (5 0 .0 0 ) 56 (3 7 .0 9 ) 72 (4 1 .6 2 ) 32 (2 7 .3 5 )'* 12.40 3.60 20 (7 .4 6 / 1.00 1 1 6 :1 3 2 12.25 3.70 21 ( 7 .8 9 / 1.10 1 1 8 :1 2 7 11.15 3.829 29 (1 1.51) 1.45 121 : 102 12.40 3.63 24 (8.82) 1.20 1 3 7 :1 1 1 12.95 3.799 27 (9.44) 1.35 1 1 7 :1 4 2 11.15 3.68 37 (1 4 .2 3 ) 1.80 1 0 0 :1 2 4 4 (5.19) 3 (3.95) 4 (6.06) 8 (10.81 ) 4 (5.00) 3 (4.54) 2 (1.17) 9 (5.32)*'9 0 (0.00) 1 (0.57) 1 (0.56) 1 (0.63) 86 (5 0 .2 9 / 57 (3 3.73)9 64 (4 0 .7 6 ) 35 (20.1 l / 72 (4 0 .2 2 ) 78 (4 9 .3 7 / S ig n ifican tlyd iffe re n tfro m co n tro lllatp < 0 .0 1 . ^ Significantly different from control I at p < 0 .0 5 . c Sig nificantly different from control II I at p < 0 .0 5 . ^ Significantly different from controls II , III at p < 0 .0 5 . e Sig nificantly different from control II I at p < 0 .0 1 . i^ Significantly d ifferent from controls I , I I , I I I at p < 0 .0 1 . ^ Significantly d ifferent from control I at p 0 .0 5 . ^Significantly different from control I at p < 0.05 . 'S ig n ifica n tly different from controls I, II at p < 0 .0 5 . ^Significantly different from control II at p < 0 .0 5 . - i~ O -O f T A B L E 5. Su m m ary o f Te raio lo g y S lu d y in Rats Receiving P-21, P-22, P-23, an Observations Control 1 untreated C ontrol II untreated Maternal parameters To tal no. fem ales gravid Mean no. corpora lutea Mean no. im plantation sites No. females exh ibiting resorp tion sites No. females exhibiting 2 or more resorption sites No. females aborting 20 15.35 12.40 13 9 0 20 13.55 12 .1 0C 14 7 0 Fetal parameters Mean no. live fetuses/group Mean live fetal weight (g) No. dead or resorbed fetuses (%) Mean no. resorptions/prcgnancy Sex ratio, M : F No. fetuses with soft-tissue anomalies [%) No. fetuses with skeletal anomalies (%) No. fetuses with accessory ribs only [%) 10.65 3.38 35 (14.1 1) 1.75 1 0 6 :1 0 7 10.85 3.58 25 (1 0 .3 3 ) 1.25 1 0 2 :1 1 5 4 (6.35) 4 (6.06) 0 (0.00) 1 (0.67) 75 (5 0 .0 0 ) 56 (3 7 .0 9 ) " Significantly different from control II at p < 0.05. ^ Significantly different from control III at p < 0 .0 5 . c Srg-mficantly d ifferent ffe trtc o n tro l HI at p < 0 .0 1 . ^ Significantly d ifferent from controls I I , III at p < 0 .0 5 . f Sig nifican tly different from controls I , I I , III at p < 0 .0 1 . Control III untreated 20 15.25 13.90 12 6 0 12.50 3.62 28 (10.07) 1.40 1 1 9 :1 3 1 6 (7.79) 2 (1.16) 72 (41.62) d P-24 A c e ty ls a lic y lic acid (250 mg/kg-day) P-21 (2 ml/kg) L_ P-22 (2 ml/kg) P-23 (2 ml/kg) P-24 (2 ml/kg) 20 10.15 1 3.25 15 \Sd 0 20 1 3.45 1 1.65* 8 4 0 20 14.30 1 3 .8 0 11 5 0 20 14.80 13.50 I2 6 0 20 14.35 13.50 13 7 0 8.70c 2.89e 91 (3 4 .3 4 )e 4.5 5h>' 1 0 0 :7 5 1 1.00 3.79'' 13 (5 .5 8 / 0 .6 5 9 1 1 0 :1 0 4 1 2 .5 0 3.75^ 26 (9 .4 2 ) 1.25 1 3 3 :1 1 8 1 2.15 3.96^ '' 27 (10.00) 1.35 1 2 3 :1 2 0 1 2.35 3.88'' 23 (8.52 1.10 1 2 9 :1 1 8 21 (3 6 .8 4 )* 7 (10.29) 5 (6.33) 9 (12.16) 3 (4.05) 40 (34.1 9 / 1 (0.65) 2 (1.16) 3 (1.78) 1 (0.06) 32 (27.35) 83 (5 4 .6 0 / '^ 57 (3 3 .1 4 / 66 (39.05) ''sig n ifica n tly different from control I at p < 0 .0 1 . ^Significantly different from control I at p < 0.05. ^Signffrearrtly d iffe re n t from co ntrol H at V < 0 .0 1 . 'S ig n ifica n tly d ifferent from controls I, II I at p < 0.05 . 60 (38.15) r 010 T A B L E 6. Sum m ary o f Terato lo g y Stu d y in Rats Receiving P-25 and P-26 Observations Maternal parameters Control 1 untreated C ontrol II untreated Control III unircated A c e tyls a licy lic acid (250 mg/kg-day) P-25 (2 ml/kg) Total no. females Mean no. corpora lutea Mean no. im plantation sites No. females exhibiting resorp tion sites No. females exhibiting 2 or more resorptions No. females aborting Fe ta 1param eters 20 15.35 12.40 13 9 0 20 13.55 12.10 14 7 0 20 15.25 13.90* 12 6 0 20 16.15 13.25 15 15 0 20 14.20 12.70 16 10 0 . Mean no. live fetuses/group Mean live fetal weight (g) No. dead or resorbed fetuses (%) Mean no. resorptions/pregnancy Sex ratio, M : F No, fetuses with soft-tissue anomalies (%) No. fetuses with skeletal anomalies (%) No. fetuses with accessory ribs only (%) 10.65 3.38 35 (14.1 1) 1.75 1 0 6 :1 0 7 4 (6.35) 0 (0.00) 75 (5 0 .0 0 ) 10.85 3.58 25 (1 0 .3 3 ) 1.25 1 0 2 :1 1 5 4 (6.06) 1 (0.67) 56 (37.09) 12.50 3.62 28 (10.07) 1.40 1 1 9 :1 3 1 6 (7.79) 2 (1.16) 72 (4 1 .6 2 ) Z.lQ d 2.89 91 (3 4 .3 4 ) 4.5 5'? 1 0 0 :7 5 21 (3 6 .8 4 ) 40 (34.19) 32 (2 7 .3 5 / '* 10.85 3.91*'^ 37 (1 4.57) 1.85 1 0 6 :1 1 1 2 (3.08) 1 (0.66) 63 (41.44) ^ Significantly different from control 11at p < 0.01 . ^Significantly different from control II at p < 0.05 . c STg-mffcanrly different from controls I I , III a tp < 0705. ^ Significantly different from control II I at p < 0 .0 1 . ^ Significantly different from controls I, II , III at p < 0.01 ^Significantly different from control 1at p < 0.01. ^ Sign ificantly different from controls 1, III at p < 0 .0 5 . ^ Sig n ifican tly different from control 111 at p < 0 .0 5 . , P-26 (2 ml/kg) 20 14.75 1 4 .0 5 * 12 8 0 12.50 3.86^ 31 (1 1 .0 3 ) 1.55 1 2 7 :1 2 3 3 (3.90) 0 (0.00) 76 (43.93) L r HAIR DYE T ER A T O LO G IC AND T O X IC IT Y STUDIES OH blood count, mclhcmoglobin, fasting blood sugar, blood urea/nitrogen, alkaline phosphatase, and serum was examined for color, pH, , u _, , microscopic elements. The clinical chemistry determinations were performed oh blood obtained by cardiac puncture by SMA-1 2 analysis using the T echni con auto analyzer. Hematology studies utilized the automated electronic Coulter F counter and reference methods (Schalm, 1975). Ames Labsux were used for the urinalysis. All survivors were sacrificed after 13 wk and examined for gross abnormalities. Organ-body weight ratios were determined for liver, kidneys, adrenals, heart, thyroid, spleen, and brain. Twenty-five tissues [skin from treated and untreated areas, lymph nodes (mcsanteric, axillary, and cervical), spleen, stomach, du o de n u m, colon, liver, gal! bladder, adrenals, nerve with adjacent muscle, eyes, pancreas, kidneys, urinary bladder, ovaries, testes, bone, bone marrow, heart, lung, thyroid, brain, and skeletal muscle under the site of application and elsewhere (thigh) | from each animal were stained with hematoxylin and'eosin and examined microscopically. Statistical analysis of the data on body weight gains, hematology, clinical chemistries, and absolute and relative organ weights was performed using the analysis of variance F test and S t u d e n t 's / test. When variances differed significantly, S t u d e n t's t test was modified (/') and C o c h r a n 's approximati on was utilized (Snedecor and Coc hr an j 1967). Results and Discussion No evidence of compound-induced toxicity was seen. Body weight gain of all test groups was at least equal to that of the c o n t r o s . Five control and five test animals died during the study due to complications resulting from cardiac puncture while collecting blood. There were scattered statistically significant differences in lhe clinical chemistry and hematologic values between test and control groups at the various sampling intervals (Tables 7-10). However, these differences were not considered to be of toxicologic significance because of cither the direction or continuity of the differences or the fact that they fell within the range of historical control values. There were a few instances when there were statistically significant differences in relative organ weights between a test group and the combined controls where the differences were not significant when the group was compared with each control group separately. In no instance were any o f these differences accompanied by histomorphologic evidence of toxicity. T h e results of the urinalyses were unremarkable. No dye discoloration of the urine was seen at any time during the test. T h e treated skin showed slight thickening in some groups, particularly groups 7403 and P-21. This was not unexpected, due to the frequency pf dye application. T A B L E 7. Mean Blood C hem istry Values fo r Female New Zealand Rabbits Subjected to 13-wk Repeated Derm al A p plications o f Hair Dyes w ith Standard Deviations 012 Group no. C - 1 ,2 ,3 47 1 P-22 e P-23 P-24 1 P-21 i P-25 t P-26 7401 7402 1 7403 t 7404 t 7405 7406 1 Glucose (mg%) 124.69 22.31 134.20 9.01 142.67 S+b 5.92 143.17 25.22 122.67 8.98 127.00 6.87 122.20 6.91 121.33 4.89 125.33 6.92 127.83 8.86 130.17 10.91 141.00 15.05 132.67 14.88 BUN (mg%) 22.19 3.94 24.20 4.38 24.83 4.12 26.33 S + 3.72 23.50 3.45 25.50 5.28 28.80 S + 5.40 20.00 1.41 21.83 2.71 20.83 4.31 20.17 1.60 23.83 3.49 21.83 2.48 Alkaline phosphatase (K - A units) 9.26 3.-19 9.62 4.16 8.85 1.46 7.98 2.07 7.07 S - c 1.12 1 1.42 5.09 8.62 3.15 8.83 3.65 6.52 1.75 9.28 4.57 9.70 2.68 8.68 2.79 7.47 2.42 SGOT ( K units) 27.94 1 1.85 24.60 1 1.99 26.67 9.16 23.17 10.78 22.67 8.26 18.00 10.68 19.60 1 1.13 27.83 1 2.21 20.50 7.87 2 2 .3 3 7.71 26.67 12.31 26.00 16.53 28.50 4 5 .3 2 Combined controls. b S + , statistically sig nificantly higher than control (p < 0 .0 5 ). CS--, statistically sig n ifica n tly low er than control (p < 0 .0 5 ). !__ T A B L E 8. Mean Blood C hem istry Values o f 1 3-wk Repeated Derm al A p p li cations of Hair Dyes on Male New Zealand Rabbits w ith Standard Deviation Group no. Glucose. (mg%) 0 1 ,2 ,3 P-22 i P-23 P-24 + P-21 i P-25 P-26 + 7401 + 7402 7403 + 7404 7405 7406 + 130.29 21.44 144.17 1 1.64 138.33 19.74 128.67 8.36 130.50 7.94 135.50 14.31' 121.50 3.45 131.50 12.36 128.00 16.73 132.83 3.4 3 1 3 4 .2 0 13.88 137.33 5.65 140.67 _________ 9.16 BUN (mg%) 1 7 .0 0 3.27 20.50 S+* 1.97 21.67 S + 3.33 23.33 S + 3.67 20.33 6.22 1 7.1 7 0.98 16.83 1.83 17.83 3.71 18.00 1.79 18.33 4.41 18.40 1,52 16.17 2.56 16.00 1.67 A lk a lin e phosphatase ( K -A units) 11.13 4.25 S. 70 1.54 9.48 5.91 10.73 2.92 7.32 S - c 1.75 9.25 3.34 7.75 S 1.77 7.97 2.57 14.53 15.41 8.98 2.90 7.28 2.51 1 1.43 5.35 8.02 1.86 SGOT (K units) 30.06 1 3 .9 3 23.83 7.31 21.50 12.18 26.67 20.50 33.67 12.37 43.33 29. SS 22.33 9.61 26.1 7 14.99 34.50 13.72 28.67 10.29 28.40 13.72 18.50 1 1.64 23.00 2.10 ` 'Combined controls. , * S + , statistically sig n ifica n tly higher than control (p < 0 .0 5 ). CS--, statistically sig n ifica n tly low er than control (p < 0 .0 5 ). r T A B L E 9. Mean Hem ato logical Values of 13-wk Repeated Dermal A p plications o f Hair Dyes on fem ale New Zealand Rabbits with Standard Deviations . Group no. HCT (%) lib ifciv) RBC (10` /m3) WBC (1 0 3/rn 3) M ct-llb (%) C -1 ,2 ,3 0 37.31 12.54 5.7^5 6.51 1.46 t 1.67 0.68 0.29 1.52 0.85 P-22 37.80 12.66 5.79 7.86 0.54 S - c i 1.30 0.52 0.30 0.97 0.39 P-23 36.25 12.57 6.30 6.80 0.88 1 2.84 0.68 1.14 1.18 0.41 P-2-1 38.00 12.77 6.17 7.33 0.77 i 0.89 0.37 0.61 0.96 0.43 P-21 i 40.00 2.97 13.30 S-t-* 0.91 6 05v f ) 6.48 1.12 0.83 0:46 P-25 35.67 11.87 S - 5.69 7.37 0.95 i 1.63 0.59 0.35 2.54 0.86 P-26 34.60 S - 1 1.56 S-- 5.37 5.78 1.34 1 1.14 0.45 0.5 5 1.42 1.10 7401 35.50 S - 12.67 5.77 6.42 0.90 1 1.64 0.84 0.46 1.33 0.75 7402 37.33 2.07 12.52 0.82 5.77 0.46 6.30 1.27 1.47 0.59 7403 38.42 1 1.4 7 S - 6.10 S+ 7.5 3 0.83 1 2.42 0.72 0.35 3.46 0.94 7404 36.83 11.95 5.55 5.40 1.42 1 1.83 0.88 0.29 1.23 1.04 7405 39.50 S + 13.00 6.03 7.10 1.33 t 2.59 0.69 0.26 1.57 0.88 7406 39.92 13.12 6.05 S + 7.33 0.17 S - 1 3.69 1.11 0.29 3.17 0.45 ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:-- T 0 Combined controls. * S + , statistically significantly higher than control (p < 0 .0 5 ). . CS--, statistically significantly lower than control (p < 0 .0 5 ). T A B L E T 0. - Mean-flea T A B L E 10. Mean Hematological Values o f 13-wk Repeated Dermal Applications of Hair Dyes on Male New Zealand Rabbits with Standard Deviations Group no. C -1 ,2 ,3 0 i P-22 t P-23 i P-24 i P-21 P-25 i P-26 1 7401 t 7402 i 7403 t 7404 t 7405 t 7406 I HCT (%) 38.74 2.29 38.00 2.10 38.42 1.56 40.00 2.10 38.50 1.22 39.17 2.40 38.00 2.83 39.08 3.61 40.25 1.33 38.17 1.17 38.80 3.35 40.83 2.93 39.83 3 13 (lb (s%) 13.35 0.79 12.53 S - c 0.50 13.17 0.56 13.27 0.73 13.1 7 0.56 13.23 0.87 12.93 0.88 13.35 2.12 13.60 0.75 1 1.98 S-- 0.77 12.66 0.81 13.80 0.91 13.60 1.12 RBC ( l0 6/m 1) 6.33 0.53 5,90 0.29 5.98 0.39 6.19 0.12 6.09 0.32 6.19 0.28 5.98 0.19 6.30 0.55 6.28 0.20 5.99 0.31 5.82 0.63 6.19 0.28 6.26 0.47 WBC (1 03/m3) 6.54 1.37 8.02 S +6 0.70 8.18 S + 2.17 8.57 S+ 1.22 7.32 , 1.34 6.52 2.17 6.72 1.36 6.53 0.85 5.30 1.20 6.73 1.95 7.30 0.58 6.95 1.30 6.45 1.49 Mct-Hb (%) 1.42 l- 11 0.50 - ( S - ) 0.45 V-- ^ 0.87 0.45 0.23 S~ 0.26 1.00 0.43 1.40 1.04 0.87 0.77 0.52 0.57 0.4 8 S - 0.51 0.78 0.65 2.86 S+ 0.94 1.63 1.19 1 0.27 S - 0.45 XV 0 014 C. B U R N ETT ET A L. J No gross abnormalities were seen at necropsy, and no microscopic lesions were seen that were judged to be due to the administration of the hair dye formulations. The incidence and severity of disease processes c ommon to laboratory rabbits was not affected by the experimental treatments. Conclusion Hair dye formulations, representing oxidative and semipermanent hair color products widely used in the United States today, were applied to rabbits twice weekly for 13 wk. This represents a considerable exaggera tion of once-monthly hair dyeing in normal product usage. The re was no evidence of systemic toxicity. REFERENCES Am es, B . N ., Kam m en, H. O . and Yam asaki, E . 1975. Hair dyes are m uiagenic: Id em ification o f a variety of muiagenic ingredients. Proc. Null. A cud. Sci. U .S .A . 72:2423-24,27. A n on. 1975. Mutagenesis b-y hair dye constituents. Br. M ed . /. October 25 , p. 188. - * Bu rn e tt, C. M ., Lo ehr, R . and Corbett, ). F . 1976. Dom inant lethal m utagenicity study on hair d y e s ./ . T o xicol. Tin-iron. 7/:u/i/tTlh~puss!\ T e<J. 4c) 4 o r |-'ub B urnett, C. M ., Lanm an, B ., G iovacchini, R ., W olcott, G ., Scala, R . and Kep lingcr, M. |9 7 5 . Long term to x icity studies on oxidation hair dyes, f o o d C osm ct. T o x ic o l. 1 3 :3 5 3 -3 5 7 . de Serres, F . 1975. Meeting on evaluation o f systems to detect mutagenic activity o f chem icals, National Institutes of H ealth, Bethesda, M aryland, Ju ly 16 and 17. D un nctt, C . W. 1964. New tables for m ultiple comparisons w ith a co n tro l. B io m etrics pp. 4 8 2 - 4 9 1.% K im m el, C. A . and W ilson, J. G. 1973. Skeletal deviations in rats: M alform ations or variations? Teratology 8:309-316. K im m e l, C. A ., W ilson, J. G . and Schum acher, H. I. 1971. Studies on m etabolism and identification o f the causative agent in aspirin teratogenesis in rats. Teratology 4 :1 5 - 2 4 . K in k e l, H. J. and H oU m an, S. 1973. Study of long term percutaneous to x ic ity and carcinogenicity o f hairdyes (o xid irin g dyes) in rats, f o o d C osniel. T o x ic o l. 11 :6 4 1 -6 4 8 . M cC oll, J. D ., Globus, M. and Robinson, S. 1965. Effect of some therapeutic agents on the developing rat fetus. T o xic o l. A p p l. Pharm acol. 7 :4 0 9 - 4 1 7 . Schalm , O. W. 1975. Veterinary hem a to log y, 3d ed. Philadelphia: Lea and Febiger. Siegel, S. 1956. Nonpuram ctric statistics fo r the behavioral sciences, pp. 9 6 -10 4. New Y o rk : M cGraw-Hill. Snedecor, G . W. and Cochran, W. G . 19 67 . S ta tistica l m eth od s, 6th ed. A m es: Iowa State University Press. . Staples, R . E . and Schn ell, V . L . 1964. Refinem ents in the rapid clearing technique in the K O H -alitatin red S method for fetal bone. Stain Technol. 3 9 :6 1 -6 3 . Steel, R . G . D. and T o rrie , ) . H. 1960. Principles and procedures o f statistics. New Y o r k : M cGraw-Hill. V e n itt, D ., etal. 1975. Carcinogenicity and m utagenicity test on some hair colourants and constituents. Nature 225:506-507. W ernick, T ., Lanm an, B. and F ra u x , J. L . 1975. Chronic to x ic it y , teratologic, and reproduction studies with hair dyes. T o xico l. A p p l. Pharmacol. 3 2 :4 5 0 -4 6 0 . W ilson, J. G. and W arkany, ). (ed .). 1965. T era to lo g y: Principles and tech n iq u es, pp. 2 6 3 -2 7 7 . Chicago: University o f Chicago Press. . '0rl Sear le, C.E., Itarnden, D.G. and G yde, G. H. . R e ce iv e d D e ce m b e i 19, 1 9 75 A c c e p te d M arch 1, 1976 -b/Jca o v \)c>\. p o . I Cocarcinogenic and Tumor-Promoting Agenf in Tobacco Carcinogenesis 112 B. L. Van Duuren and B. M. Goldschmidt3-4 ' 1MARY-- A se rie s of 21 tobacco smoke components and ted com pounds were tested for cocarcinogenic activity on m o u se skin. T h e co m po u nd s w ere applied to m o u se skin (5 0 fem ale IC R /H a Sw iss m ice/group) three tim es weekly with a low dose (5 fig/application) of benzo[a]pyrene (B [a ]P ). The test com pounds were of five classes: aliphatic hydrocarbons, arom atic hydrocarbons, phenols, and long-chain acid s and alcohols. The following com pounds enhanced rem arkably the carcinogenicity of B [a ]P : catechol, pyrogallol, decane, unde cane, pyrene, benzo[e]pyrene, and fluoranthene. Itie following com pounds inhibited B [a]P carcinogenicity com pletely: esculin, quercetin, squatene, and oleic acid. Phenol, eugenol, resorctt nol, hydroquinone, hexadecane, and lim onene partially inhibited B [a]P carcinogenicity. Six of the 21 com pounds were also tested a s tum or prom oters in two-stage carcinogenesis. No direct correlation existed between tumor-promoting activity and cocarcinogenic activity. The cocarcinogens pyrogallol and cate chol did not show tumor-promoting activity. Decane, tetrade- cane, anthralin, and phorbol m yristate acetate showed both types of activity. Structure-activity relationships and possible m odes of action were described.-- J Natl Cancer Inst 56: 1237-1242, 1976. From carcinogenicity tests of cigarette-smoke conden sate (CSC) and its known content of aromatic hydrocar bon carcinogens on mouse skin, it was concluded in 1958 that these carcinogens do not account for the observed carcinogenicity of CSC (1-3). Subsequent studies by oth" S (4) led to the same conclusion, and several investiga- (5-iS) undertook studies on the nature of tumor-proting agents in CSC. However, none of the compounds or tobacco tar, fractions tested as tumor promoters in two-stage carcinogenesis showed convincing tumor-pro moting activity. Some reports described cocarcinogenic activity on mouse skin by aromatic hydrocarbons (9) and long-chain aliphatic hydrocarbons (10-12). In (9-12) and our study, cocarcinogenesis, in the restricted sense, is a regimen in which the cocarcinogen, applied repeatedly on mouse skin with low doses of a carcinogen such as ben/o[a]pyrene(B[a]P), enhances considerably the carci nogenic activity of the carcinogen. This type of experiment, i.e., cocarcinogenesis, was only recently applied to the problem of tobacco carcino genesis. In preliminary reports, the cocarcinogenic ac tivity of several known CSC components such as benzofe]pyrene (B[e]P), pyrene, and catechol was described (13, 14); B[a]P was the carcinogen. Since cocarcinogenesis testing is more relevant to animal or human exposure to CSC or cigarette smoke, respectively, compared with tumor promotion, we made a more extensive study of the cocarcinogenic activity of known and suspected com ponents of CSC. Included in these cocarcinogenesis experiments were eight phenolic compounds, four aromatic hydrocarbons, six nonaromatic hydrocarbons (aliphatic, olefinic, and alicyclic), oleic acid, stearic acid, and lauryl alcohol. We also tested six of these compounds for promoting activity in two-stage carcinogenesis tests on mouse skin with ']P as the initiating agent. The results of these new eriments, as well as the final results of several in- complete experiments published in (14), are described here. M ATERIALS AND METHODS Chemicals.--The 21 test compounds as well as B[a]P, l,8-dihydroxy-9-anthrone (anthralin), and phorbol myris tate acetate (PMA) were prepared and/or purified as summarized in table I. The criteria of purity are also given in table 1. Preparation and storage of test solutions.--B[a]P and the compounds to be tested were made up biweekly in amber containers and stored at 4 C. The B[a]P and each test or control compound were made up in one solution so that they were applied simultaneously on the same area of mouse skin. Animals.--Female ICR/Ha Swiss mice (ARS/SpragueDawley Inc., Madison, Wis.) were vaccinated against ectromelia and tested beginning at age 6-8 weeks. They were housed on sterile hardwood chips (Iso-Dri; Fisher 8c Son, Bound Brook, N.J.) in stainless-steel cages (10/cage), fed Purina Laboratory Chow and water ad libitum, and weighed regularly. The animal rooms were maintained at 22-24 C. Bioassay procedure.--The backs of the mice (50 mice/ experimental group) were clipped 2 days before the ini tial treatment and thereafter as needed for the duration of the experiment. The solutions were applied by micro pipette in acetone or in dimethyl sulfoxide (DMSO) by calibrated paintbrush, three times weekly.. Animals were observed regularly; tumors were re corded, and those greater than 1 mm in diameter were counted and charted regularly. Only tumors persisting for 30 days or more were counted in the cumulative totals; the data presented are based on these chartings. Animals bearing tumors that appeared grossly to be car cinomas were killed approximately 2 months after the tumors were clinically classified as malignant or when animals were moribund. All animals were autopsied. Specimens from tumors and any abnormal tissues were excised, fixed in unbuffered 10% formalin, blocked in paraffin, stained with hematoxylin and eosin, and con firmed histologically. Included in the experimental pro tocol were control groups receiving carcinogen or test compound only, solvent only, and a group given no treatment. 1 Received September 18, 1975; accepted Decem ber 19, 1975. -Su p p o rte d by P u b lic H ealth Service (P H S ) contract N01 CB33241 from the D ivisio n of Cancer Cause and Prevention, National C a n cer In stitu te ; and by P H S grants C A 13343 from the N ational Cancer Institute and ES00260 from the N ational Institute of E n vironm ental H ealth Sciences. E a rlie r phases of this work were sup ported by a grant from the Am erican M edical Association, E d u ca tion anti Research Foundation. a Laboratory of Organic Chem istry and Carcinogenesis, Institute of Environm ental M edicine, New York U niversity Medical Center, New Y o rk , N .Y . 1001. t T h is work was carried out in collaboration w ith M rs. C . Katz. Mrs. S. M clchionne, and Miss A Sm ith . Histopathologic diagnoses were done by l)r. I. Seitlinan, Departm ent of Pathology, and D r. 1). R o th , D epartm ent of E n viro n m en ta l M edicine, New Y o rk U niveisily Medical Center. JO U R N A L OF T H E N A T IO N A L CANCER IN S T IT U T E , VOL. 56, NO. 6, JU N E 1976 1237 1238 VAN DDURF.N AND GOLDSCHMIDT T a b l e 1.-- Purification of lest compounds Compound Source'1 Method df purification and properties 6 Phenols Phenol. ____________________________ ________ Esoulin (6-glycosy1-7-hydroxycoum arin)___ Eugenol (4-aly1-2-mcthoxy plinn ol)_____ ___ Quercetin (3,3,.4',5,7-pcnl aliyxhoxyfiavone) Catechol ( 1,2-dihvdroxybenzene)___ _______ Peso reimil ( 1,3-dihydroxybenzene)_________ Ilydroquinone ( 1,4-dihvdro.xybenzene)____ Pyrogailol (1,2,3-trihyclroxybeiizcne)_______ Anthralin ( 1,8-dihydroxy--anthrone)______ A Distillation; bp: 182" C/7C0 mm A RecrystaUization; water; mp: 100-201 C A Distillation; bp: 78-81" C/0,08 mm B RecrystaUization; ethanol/waler; mp: 313-315 C A RecrystaUization; toluene; mp: 102-105 C A RecrystaUization; toluene; mp: 100-109 C A RecrystaUization; water; mp. 109-171 C B RecrystaUization ; ethyl acetate/toluene; mp: 129-132 C K RecrystaUization; hexane/benzene; mp: 178-179 C(16) Hydrocarbons (nonaromalic) c Stiualene___ Limonene. Derane_____ Uiidecftiie--Telradeeane Hexadecane. E nu281.4892 E Distillation; bp: 42-45 C/2.5 mm A no281.4050 A no281.4110 B nn2"1.4221 A no281.4281 Aromatic hydrocarbons d Pyrone____________ Ben zoljAfipery lene B M P ___________ Fluoranthene_____ B[a]P A RecrystaUization; ethanol; mp: 149-150 C A RecrystaUization; xylene; mp: 274-275 C C RecrystaUization; benzene; mp: 179-180 C A Zone refined; mp: 107-109 C A RecrystaUization; benzene/methanol (1 :3 ); mp: 179 C Long-chain alcohols and acids Oleic acid---------------------------------------------------------------------- S, B Fractional1distillation; bp: 164 C/O.lTnm Stearic a c id ._____ ___________ _______ ______ ______ ______ ______ ______ ______ E Vacuum dried; mp: 07-68 C Lauryl alcohol_______________ T_____________________________________________ E Distillation; bp: 99-100 C/2.5 mm P M A .. _______ _____________________________________________________________ Croton oil Solvent fractionation and column chromatography (16) a A =A ldrich Chem ical Co., M ilwaukee, W is.; B = J . 1'. B aker Laboratory, New York, N.Y.,; K =I< & K Laboratories, Plainview, N .Y .; C =C onsolidatcd M idland Corp., Brewster, N .Y .; II * ICaatwmn Organic Chem icals, Rochester, N .Y .; S =Signm Chem ical Co., St. Louis, Mo. 6 N uclear mugnetic resonance, IR an<l U V spectra, and thin-layer chrom atography, with the use of a t least two solvent systems, were obtained, where applicable, for the purified samples; these analyses confirmed the assigned stru ctu res and purity of the test compounds. Jap =boiling point; m p = m cltin g point. e n n 16 = refractive index, each 2ti C (sodium (D) line). 4 In addition to the other analyses (footnote 6). fluorescence and m ass spectra also were obtained for these compounds. Twenty-one compounds were tested for cocarcinogenic activity with B[a]P. Each compound, at the dosages indi cated under "Results," was tested by simultaneous appli cation with 5 /g B[]P/().1 ml solvent. Anthralin and PMA were used as positive controls in the cocarcino genesis experiments, since they are potent cocarcinogens (3, 14). . Six of the 21 compounds were also tested as tumor- promoting agents, i.e., by repeated application of the test compound on mouse skin (50 ntice/group) after an ini tiating dose of 150 ;ig B[a]P. Fourteen days after the primary treatment, animals were given applications of the compound three times weekly for the duration o f,the tests. Duration of tests, dosages, and vehicles are shown in the tables under "Results." PMA and anthralin were positive controls for these experiments also. The dosages of test compounds used were based on preliminary short-term (2-4 wk) evaluations and were the highest dosages at which skin-damaging effects were minimal. RESULTS Findings from the two series of cocarcinogenesis experi ments are in tabic 2. The first series was terminated at 368 days; the second, at 440 days. The rates of tumor appearance for a few of the test compounds are illus trated in text-figure 1. Pyrene, benzo[g/tt]perylene, and B[<?]P were each tested at two or three dosages. For all three compounds the increased dosages resulted in nota bly higher tumor incidences. Except the aromatic hydro carbons, the test compounds were all applied at high dosages (2-25 mg/application). The dosages of aromatic hydrocarbons ranged from 4-40 ,pg per application. The importance of dosage in the determination of tumor induction is discussed later. It is instructive first to com pare the effects of the whole range of chemicals and dosages used on B[a]P carcinogenesis. The following compounds more than doubled the numbers of mice with papillomas and carcinomas (our indexes of cocarcinogenic activity) compared with the numbers of such mice among the B[aJP-acetone controls: catechol, pyrogallol, decane, undecane, pyrene (40-p,g dose), B[c]P (15-^g dose), and fluoranthene. In the B[a]P control group (440-day expt), the total number of papil lomas among 50 mice v^as 26. The seven compounds listed above, when tested as cocarcinogens, increased this number by a low of 66 for pyrene to a high of 126 for fluoranthene; i.e., these cocarcinogens also largely in creased tumor multiplicity compared with the B[a]P con trol. Except fluoranthene, these cocarcinogens did not notably decrease the day$ to the appearance of the first tumor compared with findings for the B[a]P controls. Benzo[g/it]perylene, tetradecane, and lauryl alcohol gave results that suggested weak to moderate cocarcino- COCRCIN OGEN IC AND TUMOR-PROMOTING AGENT 1239 T a b l e 2.-- Cocarcinogenesis: Bioassay in mouse skin " Carcinogen4 Cocarcinogen (dose) Days on test Days to first papilloma Mice with papillomas/ total papillomas e Phenols B[ajP None B[a|P None B[a|P None IMalP None B[a|P None Hla|P Nonti B|a|P None B[a|P None Phenol (3 mg) Phenol (3 mg) Maculili (25 mg)11 Esculin (25 mg)<< Eugenol (10 mg) Eugenol ( 10 mg) Quercetin ('25 mg)J Quercetin (25 mg)*' Catechol (2 mg) Catechol (2 mg) Ite-sorcinol ( 10 mu) Resorcinol (10 mg) Ilydroquinone (5 mg) Hydroquinone (5 mg) Pyrogaflol (5 mg) Pyrogallol (5 mg) 308 308 308 308 308 308 308 308 308 308 308 308 308 308 440 440 207 7/9 (3) 355 1/1 (0) --0 --0 209 4/4 (2) --0 --0 --0 299 36/90 (31) 292 1/1 (1) 249 5/0 (2) --0 254 7/11 (3) --0 253 40/95 (33) "0 Hydrocarbons ( nonaromatic) B|a]P None B[a|P None B[a|P None B|a|P None B[a|P None B[u|P None Squalene (25 mg) Squalene (25 mg) Limonene ( 10 mg) Limonene ( 10 mg) Decane (25 mg) Decane (25 mg) Undecime (25 mg) Undecane (25 mg) Tetradeeane (25 mg) Tetradecune (25 mg) Hexadecane (25 mg) Hexadecane (25 mg) 440 410 440 440 440 440 440 440 440 440 440 440 400 1/1 (0) --0 295 13/13 (4) --0 230 38/73 (34) 400 1/1 (0) 195 44/105 (41) --0 120 32/58 (19) --0 314 5/8 (2) 0----- Aromatic hydrocarbons B|a|P B|a'JP Bl|P Pyrene (4 ag) Pyrene ( 12 ag) Pyrene (40 Mg) 308 250 12/14 (0) 308 180 20/42(20) 440 229 35/00 (20) a Wo used 50 female IC R /H a Swiss mice per g r o u p . Two sots of experim ents arc presented bore . T he lirst experim ent lasted for 308 days; the second, for 440. M edian survival times were greater than the days of testi nu, except for those ex perim ents in wind.i lite carcinom a incidences were liitfli. T he hih incidences of curcinem a occurred with the p o ten t cocarcinoKCr.s: pyrogailol, undccuuc, tetradccane, pyrene (at 12 and 40 g/application), fluoranthene, lauryl alcohol, and 1JM A. when applied simultaneously with D(alP. Carcinogen11 Cocarcinogen (dose) Days pn test Days to first papilloma Mice with papillomas/ tptal papillomas ` None None ' IMalP B[a|P None' B|a.|P B[a|P None B[a|P None. Aromatic hydrocarbons--Continued Pyrene (12 ag) Pyrene (40 ag) Benzoly/rilperylene (7 u&) Benzoli/fulperylene (31 ag) Benzo|fti|perylene (21 ag) B|e|P (5 ag) B|e|P (15 ag) B|e|P (15 ag) Fluoranthene (40 ag) Fluoranthene (40 ag) -1108 4140 208 2IG8 308 308 308 308 440 <140 -- ` -- 238 222 -- 249 210 -- 99 -- 0 0 19/31 (10) 20/39 ( 18) 0 24/33 (9) 33/79 (27) 0 39/120(37) 0 Long-chain alcohols and acids B|a)P None B|a|P None B||P None Oleic acid (25 mg) Oleic acid (25 mg) Stearic acid (4 mg)` Stearic acid (4 mg)' Lauryl alcohol (10 mg) Lauryl alcohol (10 mg) -140 `140 -140 `140 140 140 --0 --0 195 25/38 (7) --0 226 21/27 (13) --0 Control groups BlalP B|a|P B|a|P lMa|P None IMalP None None None None Acetone Acetone DMSO PM A (2.5 ag) PM A (2.5 ag) Anthralin (80 ag) Anthralin (80 ag) Acetone None7 None7 1308 440 :308 308 308 308 308 440 308 440 251 14/16 (10) 210 10/20 (12) 252 13/10 (0) GO 45/200 (37) 174 5/5 (0) 159 25/58 (15) 307 2 /2 (0 ) . --0 --0 --0 6 B|a]P was applied in the same solution as the cocarcinogen (5 mk/ 0.1 ml acetone or DMSO) three tim es weekly to the dorsal skin. Acetone and DMSO solutions were applied by m icropipelte and a calibrated brush, respectively. c Xumbers in parentheses are num bers of mice with squam ous carcinoma. __ d DMSO was the solvent; solvent for all others was acetone. * Applied in 0.2 ml acetone due to limited solubility. / 100 animals. genic activity. The following four compounds completely inhibited BfaJP carcinogenesis on mouse skin at the dos ages used: esculin, quercetin, squalene, and oleic acid. The compounds phenol, eugenol, resorcinol, and hexadecane reduced the lumorigenicity of B[a]P by 50% or more in ali three tumor indexes given in table 2. Hydro- qninone and limonenc showed weak inhibitory activity. Stearic acid was the only compound that gave incon clusive results. 'I'he number of animals with tumors and total papillomas was higher than that in the control groups, whereas the number of animals with carcinomas was not. Of the 21 compounds tested for carcinogenic activity, only catechol induced a skin carcinoma when applied alone, i.c., without li[]P. Phenol and decane, when ap plied without li[a]P, each induced a papilloma.in 1 of 50 ice. PMA, a known potent tumor promoter and cocartogen (13, II), was as active as, or slightly more active dian, the seven most poient cocarcinogens in this study. However, PMA was used at a dose several orders of magnitude lower than that used for the other compounds and caused a markedly higher tumor multiplicity and shorter latent period compared with the other active test compounds. Anthralin, also a known tumor pro mote] and cocarcinogen, showed the expected cocarcino genicity (//) but was considerably less active than the seven most potent cocarcinogens of the 21 test com pounds described here. No tumors existed in the solventtreated and untreated control groups. Concerning the dosages used, as dictated by short-term toxicity evaluations, the following observations are note worthy: 1) The four compounds that completely inhib ited B[a]P carcinogenesis were all given at 25 mg per application. However, 2) the long-chain aliphatic hydro carbons were all used at the same high dose; decane, undccane, and iciradecane showed cocarcinogenic activity, whereas hexadecane caused marked inhibition. 3) Of the five simple phenols tested, all in the range of 2-10 1240 VAN DUUREN AND GOI.DSCIIM IDT on mouse skin, with B[rt]P as the initiating agent (table 3). Anthralin and PMA were positive controls and showed the expected tumor response (15, 16). The results of control experiments for these test compounds, i.e., promoter without B[a]P pretreatinent, are in table 2. Of the six compounds, only fluoranthene and pyrene each caused a skin carcinoma in I of 50 mice. Pyrogallol caused a papilloma in 1 of 50 mice. The other three phenols did not show any tumor-promoting activity. Days on Test T e x t -fig u rf. 1.-- A : N um ber of mice w ith p ap illo m as; B : total num ber of papillom as. Curves 1: B [n jP + u iu lc ca n c ; curves 2: B[]P'4-tlecanc; curves 3: B[]P + tetrauccane; curves 4: B[a]P alone; curves 5: B[(iJP + hcxaclccanc. Dosages anti other data arc given in "M aterials and Methods'' and table 2. T a b l e 3.-- Two-slage carcinogenesis: 1Vumor-prnmnlini7 activity of roenrrinogens and inactive analogues " Secondary 6 treat ment. (dose) Days of testing Median Mice with survival Days to papillomas/ lime, first total days papilloma papillomas c Pyrene (10 mk) ____ Fluoranthene (10 og)_ Caleehol (2 m g)______ Resoreinol (10 mg)___ Hydroquinone(5 mg)_ Pyrogallol (5 mg)____ PM A (2.5 n f ) ________ Anthralin (80 mS?)____ Acetone______ . No treatment''________ 448 448 448 440 409 449 449 434 450 443 >448 >448 >448 >449 >409 >119 357 >431 >4.r)0 >427 414 401 -- -- -- 328 54 85 -- -- 1/1 (1) 1/1 ( I ) 0 0 0 1/1 (0) 43/155 (18) 9/14 (2) 0 0 a Wo used .r)0. frmalo I C H /lla Swiss mice i>or croup, oxcopt for the nnt.lirhlin experiment- in which 20 m ire wore used. 6 1f0 B(rtJIV0.1 ini acetone applied to dorsal skin once !>y miero))ipctte. For the an th ralin experiment', the iinliatinc dose was 100 ic B |a ]l\ For the duration of the test, the prom oters were applied to the dorsal skin 3 times weekly in 0.1 ml ace tone beginning M days after initiator. For data on the application of promoting compounds, see table 2. ` Numbers in parentheses are num bers of mice with squam ous carcinom a. d 100 mice. mg per application, catechol and pyrogallol were active and phenol, resorcinol, and hydroquinone inhibited B[a]P carcinogenesis. Six of the 21 compounds tested as cocarcinogens were also tested as tumor promoters in two-stage carcinogenesis D ISC U SSIO N Cocarcinogenic Activity The results obtained in this work are noteworthy, not only with respect to tobacco carcinogenesis, but also with respect to possible modes of biologic action of a variety of chemical structural types. Using incidences of skin papillomas and carcinomas and papilloma multiplicity as indexes of biologic activity, one can differentiate four categories within 20 of the 21 test compounds (table 4). Because inconclusive results were obtained with stearic acid it is not included in this table. A few of the compounds listed in table 4 have been reported in earlier studies. Phenol was tested earlier with a different protocol (17): B[tr]P was the carcinogen, ap plied three times weekly, with twice weekly applications of a 5 or 10% solution of phenol on alternate days. A weak cocarcinogenic response was observed. Our earlier work (13), in which we used the same protocol as that used in the present studies, showed partial inhibition of B[a]P carcinogenesis by phenol, which agrees with the present findings. Fluoranthene and pyrene were tested earlier (9) as cocarciiiogens by application with B[a]P. Although de tails of these experiments were not reported, the data presented indicated significant cocarcinogenic activity for both compounds, which agrees with our present findings. Hoffmann and Wynder (9) observed tumor inhibitory effects when phenanthrcnc and benz[a]anthracene were applied with B[a]P. Horton and associates (10-12) tested a variety of long-chain hydrocarbons, including some of those listed in table 4 as "accelerators" in carcinogenesis on mouse skin, and reported on the cocarcinogenic activ ity of some of these materials. In another study (18), tumor inhibitory elfccts were observed for several noncarcinogcnic aromatic hydrocarbons administered simul taneously witli 3-methylchoIanthrene to mice by sc injec tion. To our knowledge the other compounds listed in table 4 have not been tested for cocarcinogenic activity. Structure-Activity Relationships Although all four aromatic hydrocarbons tested showed weak to potent cocarcinogenic. activity, the same was not T a b l e 1.-- Cocarcinogcnesis experiments: Summary Cocarcinogenic activity Potent Weak or moderate Inhibitory activity Partial Complete Catechol Pyrogallol Ducane Undecane Pyrene ll|c|P Fhtoranthene Ilcnzo((7/ii)pery lene I.auryl alcohol Telradccanc Phenol Uugcnol Resorcinol Ilexadecanc Hydroquinone Limonelle F.sculin Quercetin iSqttalcnc Oleic acid COCARCIN OGEN IC AND TUMOR-PROMOTING AGENT 1241 true for the long-chain hydrocarbons and phenols. There T adle 5.-- Comparison of cocarcinogenic and fore, though decane, undecane, and tctradecane showed tumor-producing activities cocarcinogenic activity, hexadecane, tested by the same procedures and at the same dosage, partially inhibited B[rt]P carcinogenesis. Horton and co-workers (10-12) com Compound Cocarcinogenic activity Tumor-promoting activity " mented on the effect of chain length on cocarcinogenic Phenols ctivity. In the case of the phenols, both catechol and pyrogallol contain vicinal phenolic functions and are Phenol __ ____ potent cocarcinogens, whereas all the other phenols tested partially or completely inhibited B[a]P carcino Esculin..................... . . . . Eugenol _ . . . . . . Q uercetin.................- - - - genesis. None of these compounds except quercetin con Catechol _ - - - - tains a vicinal phenolic function in its structure. This suggests that hydrogen bonding and/or metal chelation . Resorcinol____ ilydroquinonc Pyrogallol. . . _,. . . . . .... . .... have some role in cocarcinogenic activity. PMA and Antlirnlin .--- 1 + -- 3 -- (TR ) -(T R ) -(T R ) + (T U ) -(T R ) -(T R ) -t-(TR) + (T lt) + (5, 17, 19) -m (20) -(5 ) -(19), -(T R ) (19), - (T R ) -(19), - (T R ) -(19), -(T R ) +(16), + (T R ) anthraliu, which are potent carcinogens, also contain functional groups that will facilitate hydrogen bonding Hydrocarbons (nonaromatic) and metal chelation. Squalene................... . . . -(T R ) -(17) Effect of Dosage Limonene._ . . . . . Decane. _____ . . . -(T R ) + (TR) NT +(21) . The dose effect has been alluded to under "Results." Undecane____ . -- Tetradecane____ -- In addition, the relative dosages of B[a]P and test com Hexadecane. _ __ --- + (TR) + (TR) -(T R ) NT +(21) (21) pounds used did not necessarily correspond to their re spective levels in CSC. Except those of the aromatic Aromatic hydrocarbons hydocarbons, the concentrations of test compounds are known only in terms of order of magnitude. Additional bioassays will be needed to clarify the matter of relative concentrations of B[a]P and cocarcinogens. Pyrene................ ....... . . . Benzo|(//'lperylene . . . B M P --------------- . . . Fluoranthene.. . . . . +(), + (T R ) + (TR) + (TR) +(), + ( T R ) (TR) NT NT (TR) Comparison of Tumor-Promoting and Cocarcinogenic Activity There is a clear-cut difference in the experimental pro tocols used in cocarcinogenesis and tumor-promotion (two-stage carcinogenesis) experiments on mouse skin. Since both procedures were used in this study, it is rele vant to compare the biologic activities, in the two types of experiments, of the 21 compounds that were exami ned. The comparison (table 5) includes results from the present wprk and from earlier reports. The rcsuPts of some of these experiments were incon clusive; nevertheless, some observations can be made. Catechol and pyrogallol, both potent cocarcinogens, are devoid of promoting activity. Phenol, a weak promoter (5, 19, 23) is not a cocarcinogen and, in fact, partially inhibits B[z;]P carcinogenicity. Decane, tetradccane, anthralin, and PMA show both types of activity; anthralin and PMA are the most potent tumor-promoting agents known. The pattern that emerges from this short list of compounds indicates that the terms "cocarcinogen" and "tumor promoter" as defined by us, cannot be used inter changeably. Mode of Action of Cocarcinogens and Inhibitors Because of the wide differences in chemical structures, reactivity and physical properties of the cocarcinogens and tumor inhibitors described in this report, it is ex pected that they exert their biologic activities in dillerent ways. There are several possible cllccts. The test com pounds may simply alter the rate of absorption, and hence disappearance, of the carcinogen B[rt]P from mouse skin. This type of action will depend largely on the lipo philic and/or hydrophilic nature of the test compound. A second possible effect of the cocarcinogen or inhibi tor is to alter the metabolic pathway of B[a]P by activa tion or deactivation of various enzymes, e.g., epoxidizing or hydroxylating enzymes. In this respect, all four aromatic hydrocarbons tested showed cocarcinogenic activity in our present work. Since aromatic hydrocarbons can in- Long-chairi alcohols and acids Oleic acid............. .... --Stearic acid_______ . . . Lauryl alcohol . . . -( TR) ( TR) (TR) +(22) -(22) NT P M A ______________ . . . +(13), + (T R ) +(5), + (T R ) a + --positive; -- = negative; = inconclusive. T R - this report; N T = n o t tested duce epoxidizing enzymes, this possible mode of action was examined in preliminary studies (24). With standard procedures (25), all four aromatic hydrocarbons induced aryl hydroxylating enzymes in mouse skin. However, the level of activation was low and did not correlate directly with cocarcinogenic activity. The phenolic cocarcinogens and inhibitors decreased the aryl hydroxylating enzyme activities in mouse skin. It can be concluded that, even though alteration of metabolic pathways of B[a]P involv ing aryl hydroxylating enzymes is a possible mode of action for the aromatic hydrocarbons, the same probably does not apply for phenolic compounds and for the longchain aliphatic compounds. This conclusion is strength ened by our observation that chain length is important in the determination of cocarcinogenic or inhibitory activity for the aliphatic hydrocarbons and that the posi tions of hydroxyl functions are important in the determi nation of the same biologic activities for the phenolic compounds tested. It is also noticeable that the unsatu rated aliphatic compounds squalcne and oleic acid com pletely inhibit B[a]P carcinogenicity. The preferential oxidation of these materials possibly interferes with the conversion of B[]P to its activated carcinogenic inter mediate. Precursors Of the 21 compounds tested for cocarcinogenic activity, 18 reportedly are components of CSC (26, 27). Pyrogallol, quercetin, and esculin have not been found, though it 1242 VAN DUURFN AND GOLDSCHMIDT lias been reported that pyrogallol dimethyl ether occurs in CSC (2,9). ' Catechol and pyrogallol are two of the most potent cocarcinogens uncovered in this study. Catechol does not occur in tobacco leaf, yet it is the most abundant phenol in CSC. The yield of catechol-is 0.4-0.5 mg per cigarette from 85-mm nonfiltcr cigarettes and 0.2-0.3 mg from filler cigarettes (29). It is lielievcd that catechol is formed in the process of tobacco combustion from tobacco flava- noids such as rutin, quercetin, and chlorogcnic acid (29). Other precursors arc probably also responsible for cate chol in smoke. Another study (30) showed an inverse rela tionship between the catechol content of CSC and the nitrate content of the tobacco leaf used for the prepara tion of the CSC. Pyrogallol may be in fresh CSC but rapidly reacts with other smoke components; in addition, pyrogallol is rapidly oxidized by air. Most fractionation procedures used in the past to prepare subfractions of tobacco tar are expected to completely degrade pyrogal lol, since alkali has been used extensively in the fractiona tion of tobacco tars; pyrogallol is rapidly degraded in air in the presence of alkali. Gallic acid (3,4,5-trihydroxy- bcnzoic acid), a likely precursor of pyrogallol, has not been reported in tobacco leaf, (26, 27) but it occurs widely in the plant kingdom (31). The other compounds, i.e., aliphatic and olefinic hydrocarbons and long-chain acids and alcohols, occur in both tobacco leaf and smoke (26, 27). Conclusion The results in this report demonstrate the important function of cocarcinogcnic agents in tobacco carcino genesis. Not only are the active agents potent but also phenols, such as catechol, and the long-chain aliphatic hy drocarbons constitute a part of whole CSC. Moreover, the results suggest that other compounds in these categories contribute to cocarcinogcnic activity. The nature of other cocarcinogcnic compounds, the mode of action of cocarcinogcns, and further studies on tumor inhibitory agents are all areas that deserve further study in the development of a less hazardous cigarette. REFERENCES (/) V an D u u rfn B L : Identification of some polynuclear arom atic hydrocarbons in cigarette-smoke condensate. J N atl Cancer In st 21:1-16. 1958 (2) O rris I ,, V an D u u rfn B L , K o sak A I, ct a l: T h e carcino genicity for mouse skin and the arom atic hydrocarbon content of cigarette-smoke condensates. J N atl Cancer Inst 21:557-561, 1958 (7) G e i .l iio r n A : T h e cocarcinogcnic a ctiv ity of cigarette tobacco tar. C ancer R es 18:510--517, 1958 (4) H o f f m a n n D , W y n d f r E L : A study o f tobacco carcinogenesis. X I. T u m o r initiators, tum or accelerators, and tumor pro m oting activity of condensate fractions. Cancer 27:848-86/l, 1971 (5) V a n D u u r in B L . S iv a k A , L a n o si.t i i L . e l a l: In it ia t o r s and promoters in tobacco carcinogenesis. N all Cancer Inst Monogr 28:175--180, 1968 (6) W yn d fr E L , H o ff m a n n D : A stu d y of tobacco carcinogenesis. X . T u m o r prom oting a c tivity. C ancer 24:298-301, 1970 (7) V an D u u r en B L , S iv a k A . K a tz C, el a l: Cigarette smoke carcinogenesis: Im portance of tum or promoters. J Natl C ancer In st 47:235-240, 1971 (8) B ock E C , S w a i n A P , Stfd m a n R L : C om position studies on tobacro. X I.1 V . Tum or-prom oting activity of suhfraclions of the weak acid fraction of cigarette smoke condensate. J N a ll C ancer Inst 47:429^136, 1971 (9) H o f f m a n n D, W v n d e r E L : Studies on gasoline engine e x haust. J A ir P o llu l C o ntro l Assoc 13:322-327, 1963 (10) H o r to n A W , D e n m a n D T , T ro ssft R P : C a rc in o g e n e sis o f th e skin. I I . T h e accelerating properties of aliphatic and related hydrocarbons. C ancer Res 17:758-766. 1957 (//) B in g h am E , H orto n A W : Environm ental carcinogenesis: E n vironm ental observations related to occupational cancer. In Advances in Biology of the Skin. V II. Carcinogenesis (M ontagna W , Dobson R I... eds.). O xford and New Yo rk, Pcrgam on, 1966. pp 183-19.1 (12) H o r to n A W , V a n D rf.a i. P A , B in g h a m E L : Physicochem ical mechanisms of acceleration of skin carcinogenesis. In A d vances in Biology of the Skin. V II. Carcinogenesis (M on tagna W . Dobson R I., eds.). O xfortl and New Yo rk. Pcrgam on. 1966, pp 165-181 (11) V an D u u r e n B L . B l a z e j T , C o i.d s g h m id t B M , et a l: C o ca rci nogenesis studies on mouse skin and in h ib itio n of tum or in d uctio n. J N atl C ancer lu st 46:1039-1044, 197! ( I I ) V an D uurfjs B L , K a t z C, G o ld sch m id t BM : Cocarcinogcnic agents in tobacco carcinogenesis. J N atl C ancer In st 51:70.3 705, 1973 (15) S f c a i. A , K a t z C , V a n D u u r e n B L : S tru c tu re an d tu m o rpromoting activity of anthralin (l,8-dihydroxy-9-anthrone) and related com pounds. J Med C hcm 14:1152-1154, 1971 (16) V an D u u r f n B L , S iv a k A , Se g a l A , et a l: Dose-response studies w ith a pure tumor prom oting agent, phorbol myris- la tc acetate. C ancer Res 3.3:2166-2172, 1973 (17) W y n d f r E L , H o f f m a n n D: A stu d y o f tobacco carcin o g e n e sis. V I1 1. T h e role of the acidic fractions as promoters. Cancer 14:1.306-1315, 1961 (18) F a l k H L , K o t in P, T h o m pso n S: In h ib itio n of carcinogenesis. A rch E n v iro n H e alth 9:169-179, 1964 ". (19) B o u t w e l l R K , B o sch D K : T h e tum or-prom oting action of phenol and related compounds for mouse skin. Cancer Res 19:413^124, 1959 (20) V an D u u r fn B L , Siv a k A , S e g a l A , et al: T h e tumorprom oting agents of tobacco leaf and tobacco smoke con densate. | N all C ancer In st: 37:519-526, 1966 (21) S ic t j : T u m o r-p ro m o tin g a c tiv ity o f n -alkan cs and l-a lkan o !s. T o x ic o l A p p l Ph arm aco l 9:70-74, 1966 (22) H o is t i P: T u m o r prom oting effects of some long-chain fatty acids in experim ental skin carcinogenesis in the mouse. Acta Pathol M icrobiol Scand 46:51-58, 1959 (27) Sa i a m a n M i l , G l e n d f n n in g O M : T u m o u r prom otion in mouse skin by sclerosing agents. B r J Cancer 11:434-444, 1957 (24) V an D u u r e n B L , S iv a k A , G o l d s c h m id t B M , et a l: To b acco carcinogenesis: Cocarcinogens, promoting and inhib itory agents. Proceedings of the 62(1 A n n u al M eeting o f the A m erican Association for C ancer Research, Chicago, 111., 1971, p 2 (25) A rcos |C , C o n n ey A H , Buu -H ot N P : Ind uctio n of m icro somal enzyme synthesis by polycyclic arom atic hydrocarbons of different m olecular sizes. I B io l Chcm 236:1291-1296, 1961 (26) W yn d fr E L , H o ffm a n n D: Tobacco and Tobacco Smoke. New Y o rk and Lond on, Academ ic Press, 1967 (27) St e d m a n R L : T h e chem ical composition of tobacco and to bacco smoke. C hcm Rev 68:153-206, 1968 (28) K a to K , S h iiia y a m a Y . N a k a iia t a T : Som e phenols in c ig a r ette smoke. Sci Papers C ent R es Inst 105:209-212, 1963 (29) W a l t z P, I I au ser m a n n M. K r u l e A : Mcthodcn tier qu an tita tive Bestim m ung des Brcnzcalcchin s im C ig arctte n rau ch . B c itr T a b a k fo rsch 3:263-277, 1965 (70) K a e u a n o s A G , M e a n s R I$, M o ld J D : Effect of nitrates in tobacco on the catechol yield in cigarette smoke. Tobacco Sci 12:125-129, 1968 (91) R odd E H , cd.: Chem istry of Carbon Compounds, vol 3B. Am sterdam , E lse vie r, 1956, p 780