Tobacco Documents Online

THE AMERICAN JOURNAL OF CANCER A Continuation of The Journal of Cancer Research VOLUME XXIX FEBRUARY, 1937 I~UMBER 2 CHEMICAL COMPOUNDS AS CARCINOGENIC AGENTS J. W. COOK, G. A. D. HASLEWOOD, C. L..HEWETT, I. HIEGER, E. L. KENNAWAY, Am) W. V. MAYNEORD (From The Royal Cancer Hospital (Free), London) CHEMICAL RELATIONSHIPS Group I: Cholanthrene Derivatives The carcinogenic compounds of the cholanthrene group merit separate consideration on account of their chemical relationship to the cholane (bile acid) class of naturally occurring substances. They are, however, of essen- tially the same .type as the benzanthracene derivatives (see p. 222). Methyicholanthrene was prepared independently by Wieland and Dane (125) and by Cook and Haslewood (35) by dehydrogenation of dehydro- norcholene, a pentacyclic hydrocarbon obtained in 1925 by Wieland and Schlichting (126) from the bile acid, deoxycholic acid, by simple chemical transformations. The molecular structure of methylcholanthrene was proved by its degradation by Cook and Haslewood (36) to 5:6-dimethyl-l:2-benz- anthraquinone, identical with a specimen prepared synthetically by a method which established its constitution. The synthesis of methylcholanthrene by Fieser and Seligrnan (51) is in complete harmony with this structure, and the conversion of cholic acid into methylcholanthrene has been reported by Fieser and Newman (48). Cholic acid and deoxycholic acid are the two principal acids of human bile, so that. in these reactions we have a comparatively simple chemical transformation of important constituents of the human body into a cancer-producing hydrocarbon related in structure to those already known. The interest of these relationships needs no further emphasis. So far it has not been possible to secure evidence that the production in the body of methyl- cholanthrene or a related compound is an Ketiological factor in human cancer. The parent, pentacyclic aromatic hydrocarbon, cholanthrene, has also been synthesised, both in London and at Harvard (38, 53). The biological testing of cholanthrene and methylcholanthrene has shown that these two hydrocar- a Presented before the Second International Cancer Congress, Brussels, September 1936. 219 PLAINTIFF'S EXHIBIT

CHEMICAL COMPOUNDS AS CARCINOGENIC AGENTS 231 and L. R. 10) of the rat, and three tumours (sarcoma M4 and M154 and carcinoma 63) of the mouse---and found in all cases an inhibitory effect. The field of these investigations is certainly not yet exhausted. Photo- chemical changes are known to play a part in some forms of cancer, as xero- derma pigmentosum, and the remarkable work of Roffo (112) on cancer of the skin in whit~ rats exposed to sunlight is of interest in this connection. BostriJm (14) has described experiments suggesting that liver tissue from mice receiving subcutaneous injections of l:2:5:6-dibenzanthracene, sus- pended in agar solution, may show a transient increase in respiration on addition of this hydrocarbon, but E. and M. Boyland have not found such an effect when no protective colloid was used. Reuterwall (111) has described and figured a flattening of an olive oil- serum interface when 1:2:5:6-dibenzanthracene was dissolved in the oil; this change may be observed when as little as 0.02 per cent of the hydrocarbon is present. Anthracene (0.5 per cent) and phenanthrene (1 per cent) had no such effect. Hieger (64) examined the surface tension between drops of serum and solutions in olive oil of pyrene or of 3:4-benzpyrene and could find no difference in effect between the two hydrocarbons, but this form of experiment does not allow any long interval for the attainment of equilibrium. II. Factors Affecting Carcinogenesis 1. Genetic Factors: Andervont ( 1, 2) injected 1:2: 5: 6-dibenzanthracene in lard subcutaneously in mice of 5 pure strains from the Roscoe B. Jackson Memorial Laboratory, having a known incidence of spontaneous mammary carcinoma (namely strains A, CsH and D of high incidence, and strains CBA and M "Leaden" of low incidence), and in albino stock mice not of pure strain. Sarcomas appeared most rapidly in C.,H, and less rapidly in the two low-incidence strains than in any of the three high-incidence strains. "At the end of the 18th week, 80 per cent of the C3H mice, 25 per cent of the A mice, and none of the CBA animals had developed tumours . . . tumours began to appear in the CBA strain just as a tumour developed in the last of the C3H mice." These sarcomas produced in.pure strain mice grew on transplantation only in mice of the strain in which the primary tumour arose. Branch (20), in a single experiment upon a small number of mice, found that the production of epithelial tumours by application of 1:2:5:6-dibenz- anthracene to the skin was less in a strain (A) having a high incidence of mammary cancer than in one (C57) having a very low incidence, while the yield of sarcomas from subcutaneous injection of this compound was the same in the two strains. Morton, Branch and Clapp (91) report the production of epithelioma of the skin by tetrapbenylmethane in strain C57 but not in strain A. Dobrovolskaia-Zavadskaia (43) found that mice of six or more strains produced sarcomas as a result of subcutaneous injection of 1:2:5:6-dibenz- anthracene, and that in these various strains there was no exact proportion be- tween the susceptibility to this agent and the number of spontaneous mam- mary carcinomas appearing in the females. 2. Effect o] the Pituitary: Ball and Samuels (4) injected 1:2:5:6-dibenz-

THE AMERICAN -JOURNAL OF CANCER Editor FRANCIS CARTER WOOD ~olumhia ~[nifie~iil~, ~e~v ~o~k Volume XXXIX The Official Organ of The American Association for Cancer Research NINETEEN HUNDRED AND FORTY Col~'lght. 1940, b~ the D¢lmrtm~nt of C~ncer R~search of Columbia University/

urine of a is precipi- floride are on occurs. boiling for ,t~nt prep- s. Calcu- t nitrogen~ ~horus and compound. ~t negative ;howed an a mixture advanced described. also in the han in the CHEMICAL COMPOUNDS AS CARCINOGENIC AGENTS SEco~ro Str~PLE~rENTARY I~PORT: LITERATUR~ OF 1938 ANn 1939 J. W. COOK x~m E. L. KENNAWAY (From the Royal Cancer Hospital (Free), London, and the University o! Glasgow) In two previous articles 1 a review was given of the literature, dealing with the carcinogenic properties of pure chemical compounds, published between the date of the First International Cancer Congress, in 1933, and the end of 1937. The present report summarises the literature of 1938 and 1939. In order that the three reviews shall be as complete as possible, there is appended to the Bibliography in the present communication a list of references to the more important original papers on carcinogenic chemical compounds published prior to the period covered by the first article. Fieser (518) has classified and correlated the data obtained in several laboratories concerned with the biological testing of polycyclic aromatic hydro- carbons and their derivatives. No summary could do justice to his excellent article, and the reader is referred to the original, but an indication may be given of its scope, and comment made on certain features. The paper is mainly concerned with the structural relationships between the various car- cinogenic polycyclic hydrocarbons, with an attempt to assess their relative potencies as accurately as the data warrant? The influence of certain variable factors (e.g., the physical state and dose of compound administered, the age, sex, strain, tissue, and species of the animal selected as the test object) on the production of tumours, is discussed in considerable detail, and we may remark, without wishing to detract from the value of Fieser's exhaustive analysis, that greater prominence might have been given to the uncertainties inherent in any attempt to give generalised quantitative expression to such influences) Attention is also drawn to a useful section of Fieser's article giving com- mercial sources of some of the carcinogenic hydrocarbons, with references and notes on their preparation and purification. The relation between molecular structure and carcinogenic activity is discussed at length, and the summarising tables provide a convenient source of reference to compounds tested and the results obtained with the groups selected for treatment, which comprise mainly substances related to 1:2-benzanthracene. In a discussion of the results obtained in mice with two different test pro- x Cook, J. W., Haslewood, G. A. D., Hewett, C. L., Hieger, I., Kennaway, E. L., and Mayneord, W. V.: Am. J. Cancer 29: 219, 1937; Cook, J. W., and Kennaway, E. L.: Ibid. 33: 50, 1935. ~ A classification of polycyclic" compounds according to their carcinogenic potency, on the basis of data obtained in tests carried out in this Institute, was independently worked out by Iball (~92), whose paper was submitted for publication before the appearance of Fieser'$ article. s Shear (791) rightly points out that the value for the average latent period in turnout pro- duction is not to be considered as an invariable property inherent in the hydrocarbon, but is merely a convenient summary of particular experiments and is valid ~nly for animals of the strain, age, sex~ etc., of the species employed. 381

• the carcino- , who studied ~lestanol, and tonomolecular :ording to the ere 3:4-benz- s 1'-methyl, ~henanthrene, ne (6-methyl- ). Curiously .fferent group investigation .at the forces may be suffi- influence the ew is further uggest an ex- ,ons and their ~d interaction ~ms was stud- tantity of the :hange in sur- ..as the hydro- ato extremely a parallel is .ctivity of the Lic activity, of -electrons and riew thai car- these valency characterised and whereas ~ two. It is tinguished by that this high an interpreta- 0-5 : 10-acean- 2-methyl-3 : 4- nzanthracene. 2:3:4-dibenz- :ene, and also ,ith 1:2:5:6- oxygen is ex- e conceptions ning electron- CHEMICAL COMPOUNDS AS CARCINOGENIC AGENTS 4O3 densities it is difficult to submit them to experimental test. Moreover, like many other theoretical interpretations they are apt to be more successful in explaining known phenomena than in predicting new ones. If the high elec- tron-density were the chief consideration in determihing carcinogenic activity, it is a little difficult to see why different alkyl groups in the same position (e.g. position 10 of the l:2-benzanthracene molecule) should have such widely differing effects on carcinogenic activity. Also, carcinogenic inertness is not a universal attribute of compounds with functional oxygen groups, as there are now known (see p. 401) compounds of this class with very pronounced carcinogenic activity. In a later paper (785) Schmidt has extended his speculations and has at- tempted a correlation of various theories of cancer with his views on the high B-electron-density of carcinogenic hydrocarbons. In the course of a comprehensive study of the absorption spectra of organic compounds in concentrated sulphuric acid solution Bandow (409, 410) has measured the spectra of 3:4: 5:6-dibenzacridine,. Browning's "styryl 430," 4'-amino-2 :Y-azotoluene, and a number of carcinogenic and non-carcinogenic polycyclic hydrocarbons. It was found, for example, that in the case of 3:4:5:6-dibenzacridine the detailed structure of the spectrum shown in alco- holic solution is lost when the measurements are made in concentrated sul- phuric acid, and there is also a shift of the bands towards the region of longer wavelength. As salt formation would occur in sulphuric acid solution, such changes are to be expected. Sobotka and Bloch (802) failed to obtain either chemical or biological evidence for the presence of carcinogenic hydrocarbons in the llpoid extracts of 2,100 litres of pooled urine from cancer patients. Considerable interest attaches to a statement made by Winterstein (846). in a review article, to the effect that he has isolated from 500,000 litres of pregnant mares' urine $ grams of a phenolic substance, C1,HloO~, which has the same absorption and fluorescence spectra as carcinogenic compounds. M. R. Lewis (640) reported that sulphanilamide is not carcinogenic to mice. This was confirmed by Zamecnik and Koletsky (855), who found that repeated subcutaneous injections Of sulphanilamide and Prontosil Soluble into mice during a year failed to produce turnouts. In recent years much attention has been devoted to the possible presence in food of carcinogenic substances as an aetiological factor in human cancer, but .although some observations of considerable interest have been made there is as yet no convincing evidence that any specific dietary constituent is impli- cated in this connection. Nevertheless, in an extensive series of publications A. H. Roffo has described the production of tumours by food materials sub- jected to various treatments, and claims that he has identified the substances responsible for these tumours. In the main these identifications are unsup- ported by the evidence put forward, but in view of the importance of the subject, and of the plausibility of many of Roffo's speculations it may be well to summarise the present position. The extensive experimental work of Roffo on the production of skin cancer in rats by prolonged exposure to ultraviolet light has been confirmed inde- pendently, as has his observation that the cholesterol content of the skin is

404 j.w. COOK AND E. L. KENNAWAY inc~reased locally on exposure to ultraviolet light. But it is unjustifiable to conclude from this that the increase in cholesterol concentration is associated with tumour production, as has been pointed out by Bergmann, Staveley, Strong and Smith (428), who have been unable to obtain skin tumours in mice with irradiated cholesterol. Roffo (762) obtained tumours of the stomach (adenocarcinoma) and of the liver (sarcoma) in many rats which received irradiated cholesterol with their food, and attributed this to the carcinogenic action of a photochemical oxidation product of cholesterol to which is attributed on quite insufficient grounds the formula C~IH,eO~ (compare Roffo and Correa, 772). It is sug- gested that similar oxidation is brought about by heating cholesterol. Roffo and Corre.a (773) claim to have demonstrated the presence of anthracene and phenanthrene in the products of distillation of irradiated cholesterol, thereby establishing a link with the carcinogenic hydrocarbons. The meagre spectro- scopic evidence advanced in support of this claim is insufficient. The tar obtained from cholesterol by irradiation and heat, under vaguely defined conditions, was found by Roffo (766, 771) to give tumours in rats when added to the food. According to Roffo the same type of oxidation which takes place when cholesterol is irradiated is aiso brought about in the cholesterol of fats when these are heated as in cooking. In a recent paper (768; compare 764) he describes the production of numerous malignant and non-malignant tumours of the alimentary tract in rats which received with their food a quantity of fat (lard, mutton fat, veai fat, olive oil) which had been heated at about 350° C. for thirty minutes. These results are interpreted as due to oxidation by heat of the cholesterol in the fats to a carcinogenic "oxycholesterol." Again, the evidence in support of this interpretation is wholly inadequate. It consists in showing that the oxygen content of the fat is slightly increased by heating for an hour at 350°, that the cholesterol (estimated by digitonin precipitation) is destroyed in the process, and that the ultraviolet absorption increases in the region of the phenanthrene absorption bands. It is not sur- prising that the combined cholesterol in the fat is destroyed by heat. Berg- mann and Hirshberg (426) showed that irradiated cholesteryl acetate is con- verted by heat into A3:~-cholestadiene and this compound was also obtained by Veldstra 1~ (827) by heating cholesteryl oleate at 300-340°. Cholesterol is by no means the only constituent of fats likely to absorb atmospheric oxygen at high temperatures, and a mere increase in selective absorption of light.is of little diagnostic value for the presence of a particular type of compound. A topic of related interest has also been investigated by Roffo, who found (765, 770) that coffee-tar produces epitheliomas when applied to the ears of rabbits. The tar is apparently prepared by roasting the coffee at 200-250°, although the nature of the subsequent treatment is not quite clear, and it is stated that only a minimal quantity of the carcinogenic material is present in coffee as ordinarily prepared as a beverage. The carcinogenic action of the coffee tar is attributed, again without experimental justification, to the oxida- tion of cafesterol, the "sterol " of coffee oil, to carcinogenic hydrocarbons, ~x Veldstra states that Waterman obtained 4 papillomas Of the stomach, one of which ~howed infiltrating properties, in a group of mice fed with A~:~-cholestadiene. and hyd~ I duct toluf " 275~ lived of ~ brow one r of' ra paigz propl pyre~ relat~ fluort U cinog III. IV. V. VI.. VII. VIII. ' ~. ' X. I. M~ subje( ascor~ sues; yeast. solubl~ was el The t

~tifiable to associated Staveley, umours in a) and of terol with :ochemical nsufficient It is sug- ol. Roffo acene and .I, thereby e spectro- The tar y defined aen added ace when fats when .~ 764) he : turnouts tantity of at about oxidation ~lesterol." adequate. increased digitonin bsorption . not sur- f. Berg- te is con- obtained ~olesterol c oxygen f light is ,mpound. ho found ~ ears of 30-250°, and it is resent in n of the te oxida- ~carbons, ich showed CHEMICAL COMPOUNDS AS CARCINOGENIC AGENTS 40S and it is claimed that the absorption spectra show bands of carcinogenic hydrocarbons, e.g. benzpyrene. It was found by Baumann, Rnsch, Kline, and Jacchi (418) that the pro- duction of tumours by ultraviolet light, by 5:4-benzpyrene, by o-aminoazo- toluene, or by the Shope papilloma virus, was not accelerated by cholesterol. Widmark (842) has recorded experiments in which mice were painted with alcoholic and petroleum ether extracts of (a) horse muscle treated at 275° C., (b) browned butter, and (c) roasted coffee. Controls were treated with the unheated materials or were untreated. Of 25 female mice which lived more than eleven months, 9 developed malignant mammary turnouts of the adenocarcinoma type (2 in mice treated with roasted Coffee; one with browned butter; 6 with heated muscle extracts). In 25 control animals only one mammary cancer was observed during the observation period of 578 days. The production of epitheliomas by application of tobacco tar to the ears of rabbits has also been described by Roffo (763,769), who advocates a cam- paign for the reduction of smoking, especially by women, as a measure of prophylaxis against cancer, goffo's claim (767) to have isolated 3:4-benz- pyrene from tobacco tar is unsupported by his experimental results, which relate merely to a fraction which is stated to have the spectroscopic and fluorescence characteristics of benzpyrene. BIOLOGICAL CONSIDERATIONS Under this heading the following aspects of tumor production by the car- cinogenic chemicals are considered in the order given: IL III. IV. V. VI. VII. VIII. IX. X. Mechanism of Cancer Production by Chemical Compounds, and Factors Affecting Carcinogenesis and the Growth of Transplanted Turnouts Action of Carcinogenic Compounds in Different Spedes and Tissues Oestrogens in Relation to Tumours Systemic Effects of Carcinogenic Compounds and of Turnouts Minimum Amount of a Compound, and Mode of Administration, for Carcinogenesis Action of Carcinogenic Compounds on Tumours Carcinogenic Compounds and Viruses Tumours of the Lung Tissue Culture and Cytology Characters and Composition of Tumours I. Mechanism o] Cancer Production by Chemical Compounds, and Factors Affecting Carcinogenesis and the Growth o/ Transplanted Turnouts The abstracts in this section are arranged in relation to the following subjects: haemolysis; light and other radiations; dietary factors; vitamin A; ascorbic acid and glutathione; metabolism and changes undergone by com- pounds in the tissues; the pituitary; chemotherapy; local changes in the tis- sues; genetic and extrachromosomal factors; age; effects on planariam and yeast. Haemolysis: The haemolytic action on mouse red corpuscles of the water- soluble sodium salts of the endosuccinic acids of 17 polycyclic hydrocarbons was examined by Warren (831) in association with the work of L. D. Parsons. The derivatives of all the five strongly carcinogenic compounds tested--

reported. Tumor .atively large doses vas given multiple ignant growths de- n. ections of 1,2,5,6- 14 transplant gen- a of the same agent 'umor No. 1 was a seventh transplant type became domi- bowed a regression nary tumor No. 28 ;hour the 22 trans- 1 cells which divide :livisions show ab- .nd of the common :hemically induced ~nown spontaneous sue are interpreted 0ciated with tumor l developed tumors of the tumors was [~ENNAWAY, E. L., AND 1055, 1938. 1936. 1936. 1937, pp. 141. Abst. 1938. CHEMICAL COMPOUNDS AS CARCINOGENIC AGENTS SECOND SUPPLEMENTARY REPORT: LITERATURE OF 1938 AND 1939 J. W. COOK Am~ E. L. KENNAWAY (From the RoyM Cancer Hospital (Free), London, and the University of Glasgow) BIOLOGICAL CONSIDERATIONS (Continued from page 428, July 1940) II. Action o/ Carcinogenic Compounds in Different Species and Tissues A number of reports of the action of carcinogenic compounds on human tis- sues have appeared. Klar (601) developed a nodule on the forearm after completion of a series of experiments with 3:4-benzpyrene. He applied a solution of the hydrocarbon (0,25 per cent in benzol) to the skin of mice with a paint brush and for at least part of the period wore rubber gloves. He also conducted experiments, of which no description is given, with the powdered hydrocarbon contained in a glass vessel. Three months after the completion of the e.xperiments a small nodule appeared on the dorsum of the left forearm. This was excised in May 1938 and described by Professor Htickel as a "so- called benign calcifying epithelioma." The growth extended into the subcu- taneous fatty tissue; the connection with the superficial epithelium is not described nor is it evident in the two photomicrographs which illustrate the report. The author does not state his age. Gordonoff and Walthard (562) record the occurrence of a tumor in a labo- ratory assistant, aged forty-two, engaged in applying methylcholanthrene (0.3 per cent in benzol) to the skin of mice. The site was in the nasolabial fold, at a spot often touched by the patient when smoking. The microscopic ap- pearance was that of a "still well delimited stage of an incipient squamous- cell sarcoma." Cottini and Mazzone (479) deliberately applied 3:4-benzpyrene (1 per cent in benzene) to the skin, generally of the arm or thigh, of 26 patients with various cutaneous diseases, usually daily for periods up to 120 days. The changes observed were (1) pigmentation (" an increase in melanin in the basal layer of the epidermis "), which was more distinct in older people and in parts exposed to light, and (2) verrucae. The latter are described as " an accentua- tion of the relief pattern of the skin secondary to a deepening of the sulci," but no dimensions are given. The changes retrogressed completely within two months after cessation of the applications. In a child aged four, suffering from xeroderma pigmentosum with multiple squamous-cell cancers of the face, the skin reacted to the hydrocarbon in the ordinary way, though the reaction to ultraviolet radiation was intense. This latter effect was less in areas treated with the hydrocarbon. Application of an ointment containing 1 per cent 3:4- benzpyrene to one of the ulcerated cancers in this patient had at first a favour- 521

I0: II0, O: 421, . 1938. cancer • 1939. .e biol. 1937. e biol. CHEMICAL COMPOUNDS AS CARCINOGENIC AGENTS 579 731. OWEN, S. E., WEISS, H. A., A~D P~NCE, L. H.: Science 87: 261, 1938. 732. OWEN, S. E., WEISS, H. A., ~ PSJ.NCE, L. H.: Am. J. Cancer 35: 424, 1939. 733. PAoE, R. C.: Arch. Path. 26: 800, 1938. 734. P~RSONS, L. D.: Nature 142: 480, 1938. 735. PArsoNs, L. D.: J. Path. & Bact. 47: 501, 1938. 736. PASSEy, R. D. : J. Roy. Inst. Pub. Health & Hyg. 2: 16, 1939. 737. PF.~COCK, P. R., AND BECK, S.: Brit. J. Exper. Path. 19: 315, 1938. 738. P~cocx, P. R., AND BECK, S.: Brit. J. Exper. Path. 19: 434, 1938. 739. P~zRs, J. H.: Am. J. Path. 15: 261, 1939. 740. PEISA~HOVITCH, I., EINHORN, G., AND VORONIANSmI, G.: T~sissT docladov na pervome siezd~ oncologov, Oukrainy, Kiev, May 1938, p. II. (Read in abstract only. Index analyt: cancerol. 13:210 (Abst. 556), 1939.) 741. P~RRY, I. H.: Proc. Soc. Exper. Biol. & Meal. 39: 346, 1938. 742. P~RRY, I. H., AND GX~Z~rON, L. I.: Am. J. Cancer 29: 680, 1937. 743. P~.uRy, I. H., AND LOCKHY~D, M. S.: Am. J. Cancer 35: 422, 1939. 744. PIERSON, H.: Ztschr. f. Krebsforsch. 47: 1, 336, 1938. 745. PmRSON, H.: Ztschr. f. Krebsforsch. 48: 177, 1938. 746. POL~.IA, J. A.: Am. J. Cancer 32: 545, 1938. 747. POLLL% J. A.: J. Indust. Hyg. & Toxicol. 21 : 219, 1939. 748. PONDER, E.: J. Exper. Biol. 16: 38, 1939. 749. PovRs~x, Y.: Act*, Unio internat, contra cancrum 3: 31, 1938. 750. PovltS~x, Y.: Acta, Unio internat, contra cancrum 4: 719, 1939. 751. POV-RsA~, Y.: Compt. rend. Soc. de biol. 127: 364, 1938. 752. Povmt~v¢, Y.: Compt. rend. Soc. de biol. 127: 1475, 1938. 753. PROHASKA, J. vAN, BRUNSCHWIO, A., AND W~LSON, H., Arch. Surg. 38: 328, 1939. 754. RA~AC.E, G. R.: J. Chem. Soc., 1938, p. 397. 755. RA~I, B., AND Gv~a~z~., H.: Ztschr. f. Krebsforsch. 48: 355, 1939. 756. REOOIANI, M., DA~Sr, A., AND MORELLL E.: Tumori 13: 635, 1939. 757. RE,MANN, S. P., AND CHATALBASH, N.: Growth 1: 247, 1937. 758. P, EITANO, R.: Riv. real. 18: 185, 1939. (Read in abstract only. Chem. Zentralbl. 1: 903, 1940.) 759. REvoI.~EI.~, G.: Boll. $oc. ital. di biol. sper. 13: I106, 1938. (Read in abstract only. Chem. Abstr. 33: 8764, 1939.) 760. RmEAL, E. K., AND SCHULMAN, J. H.: Nature 144: 100, 1939. Cf. Rm~L, E. K.: Science 90: 217, 1939. 761. ROBSON, J. M., AND BONSER, G. M.: Nature 142: 836, 1938. 762. Ro~Fo, A. H.: Bol. Inst. reed. exper. 14: 673, 1937. 763. RoF~o, A. H.: Bol. Inst: reed. exper. 15: 5, 349, 1938. 764. Ro~o, A.'H.: Bol. Inst. reed. exper. 15: 407, 1938. ' 765. ROll, O, A. H.: BoL Inst. reed. exper. 15: 741, 1938. 766. ROrl~O, A. H.: BoL Inst. reed. exper. 15: 837, 1938. 767. Roll, o, A. H.: Bol. Inst. reed. exper. 16: 1, 1939• 768. RoF~o, A. H.: Bull. Assoc. franq, p. l'~tude du cancer 28: 556, 1939. 769. Ro~o, A. H.: Deutsch. reed. Wchnschr. 65: 963, 1939. 770. Ro~o, A. H.: Deutsch. reed. Wchnschr. 65: 1382, 1939. 771. Ro~l~O, A. H.: Ztschr. f. Krebsforsch. 49: 341, 1939. 772. Ro~o, A. H., ^Nv C6~, L. M.: Bol. Inst. reed. exper. 14: 681, 1937. 773. Ro~o, A. H., ANy Co~, L. M.: Bol. Inst. reed. exper. 15: 847, 1938. 774. RoNvoNI, P.: R. Ist. San. Pubb. 2: 345, 1939. (Read in abstract only.) 775. RONDONI, P., ~ B.v~I, W.: Tumori 13: 585, 1939. 776. RozHm~m~, S., ANn I-I~m~AN, J. R.: Arch. Path. 28: 212, 1939. 777. ROWNI~-E, L. G., $1w__.iNlm.~, A., ~ BROWN, W. R.: Kongressbericht des XVI. Int. Physiologenkongress, Ziirich. !I, 1938, p. 65. 778. RvscH, H. P., B~v~, C. A., ~,NV KLn~E, B. E.: Proc. Soc. Exper. BioL & Meal. 42: $08, 1939. 779. Sas~tz/~s, J., Bmm~, J., ~ GEVF.~, A.: Compt. rend. Soc. de biol. 127: 4~3, 1938. 780. SazA~trr~ H.: Fukuoka acta reed. 31: 164, 1938.

THE AMERICAN .,JOURNAL OF CANCER Editor FRANCIS CARTER WOOD THE INSTITUTE OF CANCER RESF_2kRCH OF Volume XXXIII The Official Organ ot The American Association for Cancer Research and The American Association for the Study of Neoplastic Diseases NINETEEN HUNDRED AND THIRTY-EIGHT Cop/tight, 1938. bg the Institute of Cancer Research of Columb;- Univ~,~ity

"CHEMICAL COMPOUNDS AS CARCINOGENIC AGENTS Frost SUPPLEMENTARY REPORT: LITERATURE OF 1937 J. W. COOK A~m E. L. KENNAWAY (From the Roya~ Cancer Hospital (Free), London) A review of the progress made in the study of the carcinogenic properties of chemical compounds, incorporating literature published during the triennial period between the First and Second International Cancer Congresses, was presented to the Second International Congress in Brussels, in September 1936, and was subsequently reprinted with an addendum covering most of the literature published up to the end of 1936 (231). The year 1937 has again witnessed a large output of new work in this field and this is summarised in the present report. For convenience of reference the general classification adopted in the original report has been retained and the references are num- bered continuously. CHEMICAL RELATIONSHIPS Group I. Cholanthrene Derivatives An interesting review of the carcinogenic properties of a number of com- pounds structurally related to cholanthrene was published early in the year by Fieser, Fieser, Hershberg, Newman, Seligrnan, and Shear (250). In the case of many of the compounds discussed the results of biological testing were necessarily of a preliminary character, and the points of view expressed may eventually require slight modification. Reference is made to some of these compounds under Group III (below). The synthesis of 4:10-dimethylene- 1:2-benzanthracene, an isomeride of cholanthrene, has now been described by Fieser and Seligman (256) and the same compound was prepared by Dansi (239, 240) by a different method. This compound, when injected into rats, produces tumours of connective tissue (308); so the interesting fact emerges that of the four known isomeric " ace-l:2-benzanthracenes,, the three in which substitution occurs at a meso position of the benzanthracene molecule are all carcinogenic (cholanthrene, 8: 9 -dimethylene- 1 : 2 -benzanthracene, 4:10- dimethylene-1:2-benzanthracene), whereas the fourth, 3:4'-dimethylene-1:2- benzanthracene (228), in which this condition is not fulfilled, is inactive. Cholanthrene is related to 5:10-dimethylene-1:2-cyclopentenoanthracene, which has been synthesised by Fieser and He~shberg (251), in the same way that 1:2:5:6-dibenzanthracene is related to the slightly more carcinogenic 5:6-cyclopenteno-l:2-benzanthracene. Fieser and Hershberg (253) have re- cently reported that Shear obtained tumours at the site of injection in 3 of 20 mice within ten months after injection of a lard solution of 5 : 10-dimethylene- 1: 2-cyclopentenoanthracene. The loss of carcinogenic activity following the introduction of a methyl 5O group int, our previ~ cholanthr, tion in th~ Bruce ant A hey Bergmam method i~ the final : eight hou conversio pound wt (333), w] arises fro derivative with phe compoun, the chemJ group fo~ carbon g derivatiw suggests anthrene centres t¢ ring of tt Of im obtained solution