Experiments on the Presence of Carcinogenic Substances in Human Surroundings

The pathogenesis of human cancer constitutes one of the outstanding puzzles not only of medicine but of biology in general. A careful observation of the current literature concerning the pathogenesis of tumours cannot but result in a feeling of disappointment, revealing not only an amazing amount of uncontrolable theories, but also the puzzling etiological divergencies disconnecting experimental, parasitic, occupational, radiational and spontaneous neoplasmas. Histologically artificial tumours, produced by surprisingly various substances, are practically identical with spontaneous neoplasmas, suggesting a single underlying neoplastic principle and the essential unity of the tumor problem. On the contrary, because of the well-known etiologies the production of artificial tumours cannot be considered conclusive for clinical cancers, and spontaneous and artificial tumours are apparently different classes of diseases. Therefore the investigation of tumours has advanced along many different lines more or less independently, and the research worker in the field of spontaneous cancer encounters no little difficulty in keeping in touch with developments of the branches of artificial production of tumours in men and animals. How can these apparently conflicting conceptions be reconciled? The author had the conviction that this discrepancies did not arise from any party being wrong. The possibility of all being correct and helpful to one another, however, has not been much thought about and still less explored. It was in the hope that these diverse researches may be enabled to be considered in interrelation with each other and in respect of the effect of each contribution to the whole that the author has essayed the task of finding carcinogenic substances in ordinary daily human surroundings.

Technique. The technique was always identical and as simple as possible. The substance under test was cut into small pieces, dried to constant weight over calcium chloride, finely divided by grinding in a mill, boiled in benzene under reflux for twelve hours, and filtrated. The hot filtrate was concentrated into a thick oil. This dark, viscous residue was applied twice weekly in the usual way about 50 times to 100 white mice in the interscapular region, 100 mice of the same litters painted only with benzene forming a control series. The diet consisted of wheat, dried yeast, full raw cow milk and water ad libitum. No experiment was continued after the 180th day. The nature of the skin lesions of the mice was checked microscopically by sectioning the tumours, the well-known standard criteria being accepted.

The materials selected for the first experiments were the contents of vacuum cleaners, which proved to give absolutely negative results. A second series was devoted to linen, cotton, wool and leather in an unprepared state; no positive result was obtained. A third series was formed by articles of dress prepared by textile dyeing, with only negative results. A fourth series formed by mixed daily human food yielded the first positive result, a papilloma appearing on the 89th day, which was the reason for the investigation of human food more systematically. In a fifth series raw green vegetables and raw meat were tested; the result was negative. In a sixth series canned vegetables and canned meat were tested; no positive results were obtained. In an attempt to test human articles of food separately, a papilloma was obtained with extracts of commercial bread on the 92th day. The result, however, was unstable; a constant positive series of tumours could not be reproduced regularly. The problem was at this stage as puzzling as ever, notwithstanding the positive result with bread.

Bread is no homogeneous substance. Histologically white bread resembles severe thermal burns of third degree. The outer part often consists of black coal, the inner part is a regularly constructed organic substance, both parts being separated by more or less structured brownish layers. It seemed possible that the inconstancy of the results of the experiments was dependent on the inhomogeneity of the structure of bread. Resolving bread into its 3 factors, the black crust, the white inner part and the brown boundary layer, these more homogeneous compounds were tested separately in three series. The eighth and ninth series, consisting of extracts of the white and the brown parts of bread, were absolutely negative; the tenth series, however, prepared from the black crusts, yielded two papillomata. The idea suggested itself that only black crusts may possess tumour-producing power. To test this possibility, in an eleventh series the black crusts of meat roasted in an ordinary frying-pan were compared with the inner red parts; again only the extract of the black crusts showed positive results. In a twelfth series green cabbage was roasted, and the extract of the black crusts was compared with that of the green parts; only the black crust extract yielded positive results. The result did not change, if no fat was used in the roasting process. The controls were always negative in all series of experiments.

It was now possible to reproduce ad libitum the same univocal experiments, to account for the logical connection of facts, and to predict the result of every new experiment. A series with extracts of human food produces tumours in mice only when black crusts are present. When black crusts of a certain food are tested separately, all the positive results are accumulated in the black crust series. It seemed inevitable to conclude that human food, which originally in a fresh state produces no tumours of any kind in mice, contains compounds with carcinogenic power for mice when these substances are heated to such temperatures that black crusts are formed. In the following tables are reproduced some typical series, all controls being negative:

In the series of Table I relative to black crusts of bread by the 180th day 2 tumours had been obtained. Of these, one was an undoubted cancer in which the epithelium penetrated the panniculus carnosus; the other tumour was a papilloma. The experiment with black crusts of meat of Table II yielded in all 3 tumours. Of these 2 were certainly malignant, for the downgrowing epithelium penetrated the panniculus carnosus, the third being a pailloma with some downgrowth. The tumours showed histologically the well-known characters corresponding to the macroscopic aspects. Both typical series could be reproduced with regular results as to the percentage of tumours, using roasted meat and commercial bread as test materials. This sequence of uniform experiments with uniform results seems to lead to the precise and unique-conclusion that there is a high a posteriori probability that cutaneous application of extracts of human food with black crusts is connected in white mice with carcinogenesis, possibly by the genetic affinity between tar and black crusts.

Table I

Table II

By these experiments is shown clearly not only the well-known importance of heat in the formation of cancer-producing substances from organic compounds having originally no carcinogenic power, but these experiments seem to be the first in which a cancer-producing material has been obtained by ordinary kitchen technique treatment of ordinary human food on an ordinary kitchen fire. The second fact of importance is the apparent potency of the black crusts only, in respect to malignant growth in white mice. Besides accidental burning, black crusts may be produced in the kitchen chiefly in roasted meat and in bread.

It appeared very difficult to determine the temperature of food in a frying-pan. No uniform results could be obtained, the temperature varying in a most irregular way. The boiling point of trimyristin with 14 carbon atoms in the fatty acid chains is known to be 290° C to 300° C at a pressure of 0 mm; that of the triglycerides of the fatty acids with 18 carbon atoms is not to be determined, since decomposition with formation of black substances takes place at about 280°. Stearic acid stills over at 220°-270° at atmospheric pressure; it is blackened and destroyed at 290° with the formation of pitch. It seems very probable that triglycerides heated to their boiling point in the frying-pan give rise to pyrogenous formation of tarry compounds at about 280°, which are related to the pitch formed from stearic acid at 290°.

Heated carbohydrates show caramelization at about 150°. Brown products are formed at this temperature, but no black crusts. The coffee bean is heated for 8 minutes at 150° in the factories; the so-called roasting of the coffee does not transgress the temperature of caramelization. In the production of gun-powder different kinds of wood are heated to 280°; at this temperature complete carbonization takes place. Heating at 270°, however, does not produce black coal, but a reddish-brown substance. In the production of common wood-coal the same limit of temperature is involved: below 270° water and gases are produced, above 270° tarry products still over, and carbonization starts at about 280°.

The temperature of the air in the baker’s oven surrounding the loaves is never allowed to surpass 300°; it is kept at 200°-230° for small rolls and at 230°-270° for loaves. The temperature of the inner white part of bread never surpasses 100°; only the crust reaches a higher temperature in the hot surrounding air. Apparently the inner parts are only boiling during the baking process, the crusts forming a kind of closed pan. A practical determination of the temperature in a baker’s oven, not provided with thermometer or pyrometer, is the observation of the colour of some flour spread out in the oven. If the colour of the flour does not change, the oven is too cold; if it becomes black, the oven is too hot; if the flour becomes brown, however, the oven is all right. From this practical baker’s rule it may be derived that the formation of black crusts is a constant danger during the baking process. This is very comprehensible since the temperature of the air in the oven is allowed to rise to 270°, and carbonization takes place at 280°.

It must be very easy to avoid the formation of black crusts by keeping the temperature of the baker’s oven under 270°, and by avoiding the kitchen technique of roasting at about 280°. A slight modification of the current kitchen habits excludes every human contact with black crusts. Roasted meat must show black crusts inevitably, because the boiling point of fats exceeds the temperature of charring of about 280°.

Apparently it may be regarded as a fact that black crusts of human food possess carcinogenic power for the skin of mice, and it is tempting to imagine that the formation of spontaneous human tumours may also be connected with the presence and ingestion of these cancer-producing compounds in black crusts of human food. Deductions from the white mouse, however, cannot be carried over to the human sphere, even with due reservations. Therefore these experiments cannot give a solution of the problem of human cancer, but they only create new lines of investigation in human pathology parallel to the results in mice, namely the attempt to test the possibility of an interrelation between the formation of human tumours and the superheating of human food above 280°, and of a parallelism between the frequency of cancer in man and the ingestion of black crusts of food. Only a gigantic field test during about a century would be conclusive for this problem: of a series of hundreds of monochorionic, monozygotic, identical human twins one member of each couple ought to avoid the ingestion of black crusts from birth till death, and the other member of each couple ought to be treated the other way round. It does not seem probable that this convincing mass experiment can be executed, and we are therefore reduced to accidental information about the occurrence of tumours and about habits concerning the preparation of food in different populations and in different layers of society. Especially investigations in geographical pathology may be reviewed from this new point of view of the possibility of a parallelism between differences in kitchen technique and corresponding differences in tumour frequency. The low frequency of cancer in tropical countries compared with Europe and U.S.A. has no significance for this question; only populations are comparable living their whole life in identical surroundings under identical conditions, except their kitchen habits. Unfortunately almost all peoples of the world, even the most primitive, use the techniques of roasting, frying and baking, resulting in the production of black crusts in their food. These techniques seem to be missing only in a few areas of the globe where constant lack of fuel exists, as in polar regions and in deserts under primitive conditions.

It is well-known that the Eskimoes living in their original way did not suffer from tumours, neither in Greenland nor in America; their neighbours, however, the North American Indians, showed tumours. It is known that the boundary between these two populations coincides with the limit of trees. The old-fashioned Eskimoes of Alaska and Canada transgressed the limit of trees only once a year to obtain wood for their sledges, and returned as soon as possible to their barren grounds, since they believed that the spirits of the trees would kill them if they stayed more than five nights in the forests. Inversely the Indians went out of the forests into the tundra only to follow the wild reindeer migrating northward during the few summer months; mainly for lack of fuel they returned to their forests as soon as possible. The Indians, disposing of an abundance of fuel, roasted their meat on big wood fires. The Eskimoes, however, were confined to the use of seal blubber burning in their soapstone lamps. It took about three hours to bring an Eskimo family soup to the boiling point. Therefore only on rare occasions would the Eskimoes slightly boil their meat; but there was no doubt that they considered this a waste of fuel and that they preferred the meat raw. Even after eating some cooked meat they would finish up with some raw as a dessert. The technique of roasting, frying and baking was not known to the old-style Eskimoes. Here were two peoples living under the same conditions, in the same climate, hunting and eating the same wild reindeer, but whose kitchen techniques differed widely from each other, the production of black crusts being apparently parallel to the frequency of tumours. It would be of interest to review this problem in inner Alaska and in the barren grounds of Central Canada, since precise figures for an exact comparison are not available; and it seems likely that very soon there will be no more chance of their being available by the general acceptance of the white man’s techniques. Perhaps it is already too late now!

The Hadendowa, a Nubian tribe living between Port Sudan on the Red Sea and Berber on the Nile have their special kitchen technique, probably forced on them by lack of fuel. They do not use the techniques of roasting, frying or baking; they take their food raw or boiled. This fact is mentioned by a medical officer interested in tumour investigations, Dr. Pridie of the Sudan Medical Service, recording in the British Med. J., 1925, II, p. 649:….”among Sudanese natives generally, in the Red Sea Province, the commonest form of carcinoma is cancer of the breast”. These facts agree with those in other tropical and non-tropical populations, but differ widely from that in the Hadendowa, living in a tree-less desert,country under constant want of fuel: “The Hadendowa tribes (Fuzzies) live an isolated life scattered over a wide area in dry desert country rising to a maximum of 6ooo feet in the hills among the Red Sea coast.” “Malignant disease is very rare. During the last twelve months I have seen in the whole province two cases–one of an epithelioma of the vulva and the other of a malignant parotid tumour; both were verified by the Wellcome Laboratories of Khartoum.”—-“It is curious that both cancer and deficiency diseases should be so rare, as the Hadendowa are liable to most ordinary complaints, particularly arthritis, rheumatism and syphilis.” “They subsist principally on the milk of the camel, goat, sheep or cow, supplemented by a small amount of millet and meat. The milk is drunk cold, and the millet and meat cooked”….Deficiency diseases may be almost absent in this population because raw milk contains enough of all vitamins essential for man. As for the rarity of tumours in the Hadendowa contrasting with their presence in the surrounding Nubian tribes, their kitchen techniques may possibly be responsible. The Nubians roast their meat and bake their millet, but the Hadendowa only boil their meat and millet. Apparently the Hadendowa food is not superheated till black crusts are formed, and as a matter of fact they show hardly any malignant tumours. Here again there seems to exist a parallelism.

It will be of interest to investigate more accurately and if possible statistically exactly the frequency of tumours in populations having little fuel at their disposal, as for instance the Eskimoes and the Hadendowas, compared with their neighbours using the current kitchen techniques. Perhaps more such populations may be found in the three circular belts of the earth, where fuel is a luxury: the region to the north of the limit of trees, identical with an August-isotherm of 10° C, and the two desert zones of the northern and southern hemispheres, each covering three continents around the globe.

The same contrast as between Eskimoes and Indians exists between the Chukchee and Koryak peoples on the Bering Sea on one side, and on the other side their Tungus and Yakut neighbours. Chukchees and Koryaks, like the Eskimoes in their old style of life, use seal fat as well for lighting as for heating; on the contrary their neighbours of the wood belt live like the Indians. It is the contrast between tundra and taiga life.

The Samoyeds of the European and Siberian tundras eat most of their food uncooked, even whole reindeers; on the contrary, their related Ostjak neighbours on the southern side of the limit of trees use wood as fuel.

The universal food of the Tibetans of the mountain deserts of Central Asia is tsamba, barley caramelized like coffee beans and eventually ground to a brownish flour, and eaten without further cooking with salt in buttered tea. It may be of interest to compare in the mission hospitals of Western China the tumour frequency of the Tibetans with that of the Chinese with their well-known high cancer frequency. The Chinese possess perhaps the most elaborate kitchen technique of the world; they do not only fry their meat, but even most of their vegetables, and for this purpose special vegetable oils are generally used.

The principal article of food of the genuine Bedouins of the Arabian desert is camel’s milk, with a few dates for a dessert and meat and rice only as an addition for festivals. In the wooded oases, in the two southern corners of Arabia, along the western edge of the peninsula where mountain ranges take the place of the desert plateau, and in the Arabian cities, however, the date palm furnishes the main fuel supply and permits of frying and roasting of food.

The staple diet of the nomadic Tuaregs of the Central Sahara is milk and cheese, and only the more sedentary eat grain. Meat is not much eaten, for it is a luxury. They neither roast nor fry, but they cook or stew their food in a pot, or when on the march they broil it in the hot sand under the hot embers of a fire until it becomes shredded. Their neighbours to the South, the negroid Kamuri and Eulani, and the sedentary Arabs and Berbers of Algeria, Tunisia and Tripolitania nearer to the north coast fry and bake their food in the usual way.

The Bushmen of the Kalahari desert, hunters and collectors with an Aurignacian culture prefer their food broiled, but often eat it half raw, or sometimes only slightly heated or singed. They assert that boiled food makes them ill, and much of the food is eaten raw, without their even. attempting to cook it. The sedentary Bechuana surrounding them are agriculturists and stock raisers, and make use of the techniques of frying, baking and roasting.

The name “Eskimo”, which is the Algonquian word “Usquemow” for the Inuit, as well as the old Russian name “Syrojedets” for the “Samojedets” both mean “raw-eaters” or “omophagi”; the present Russian name Samojedets means “self-eaters” or “homophagi”, and is probably a mutilation of the older name Syrojedets; a third mutilation is the term “Samoyeds” used by non-Russians. Similar terms exist in East-Siberia.

Cancer is infrequent in lunatic asylums. Mathematically correct figures derived from the available statistics seem to indicate that the populations of lunatic asylums show a cancer frequency which is 3 to 4 times lower than the frequency in the civil populations of the same areas. This is comprehensible from the point of view that a parallelism may exist between frequency of cancer and kitchen technique, since in such great communities food and especially meat is cooked in great boilers, but is never roasted. Only the bread baked in the ordinary way may contain superheated black crusts. Assuming this possibility, it is even seductive to deduce from these figures the ratio of cancer frequency related to meat, compared with the cancer frequency related to bread, as causal factors. Comparing in this way the possible shares of each of these two possible culprits, both being parent-substances of black crusts, 70 to 75 per cent would be due to meat and only 25 to 30 per cent to bread.

Some statistical reports about cancer frequency seem to indicate a higher frequency in the higher social layers. Here again the same cause may be present as in asylums. The preparation of food of poor people is simpler; roasting and baking are less frequent than boiling and cooking, since water is cheaper than fat; and generally the use of meat as well as of fuel is more limited than in higher social layers for the same reason.

The frequency of tumours in domestic animals may be explainable in the same way. The high frequency of spontaneous tumours in dogs and especially in cats may be due to the human food on which they principally subsist; the almost exclusively carnivorous cat may be still more in danger than the other domesticated animals. Many other animals which subsist on food originally prepared for human use by the application of heat show spontaneous tumours, as e.g. wild animals living in a free state in human surroundings, or in cages in laboratories.

The animals in zoological gardens, well protected and living long enough to develop tumours, seem to get cancer only when subsisting on human food. Beasts of prey eating raw meat do not frequently contract genuine tumours; they do show, however, the well-known neoplasmata due to the interference of parasites. Malignant growths are induced by viral and helminthic endo-parasites, as by neoplastic viruses in frogs (Lucké), in fowl (Rous), in rabbits (Shope), in condylomata acuminata of men (Serra); in rats by Spiroptera and Cysticercus, and in men by Schistosoma and Opistorchis. The few tumours observed in the Hadendowa are perhaps connected with an analogous chain of causes of parasitical character, the possibility of which being demonstrated in the bilharziasis of their Egyptian neighbours with the well-known tumours of the bladder.

It is evident that at the present state of the investigation a definite opinion must be deferred until more facts are available, but all facts known at present seem to point in the same direction. If we accept the theory that the presence of carcinogens in food may be an indispensable link in the chain of events causing the “spontaneous” transformation of a normal human cell into a cancer cell, the general situation is clarified, the antithesis between artificial animal cancer and spontaneous human cancer disappears, and it seems that the pieces of the cancer puzzle begin to fall into their places. This problem going beyond direct experiment, a direct proof of the correctness of this theory is impossible. It is really the whole system of guesses which is to be either proved or disproved by experience, since not one of the assumptions can be isolated for separate testing. Once the correct arrangement of the experimental and observational material into an adequate theory being found, however, such a widespread agreement seems to result, that the doubt as to its validity gradually diminishes. There is at least a strong a posteriori probability that this first tentative attempt at a complete theory of carcinogenesis is an approximation to a correct description of the chain of causal processes necessary for the evolution of malignant growth.

The active substances in the black crusts of human food being carcinogenic for mice may be 3.4-benzpyrene and 1.2.5.6-dibenzanthracene, the products with cancer-producing power for white mice present in natural tar. Superheated organic substances seem to contain only these two carcinogenic compounds. It will perhaps be possible to determine the presence of carcinogens in the black crusts of superheated food by the demonstration of their specific bands in the fluorescence spectrum, after isolation by the technique of the formation of a chromatographic spectrum in an absorbing column. The possible presence of murine carcinogens in human food is not a proof of their causal function in human malignant growth, but only an indication in favour of this theory.

If a carcinogenic substance in superheated human food is assumed by this “black crust theory” to be an exogenous factor in cases of the kind that we agree to call malignant growth, we have by no means exhausted the study of all the correlations that may be called “causal”. Seeking to explore why and how, of so many persons exposed to the ingestion of black crusts of superheated human food, only some become diseased, not only the peristatic influences, the susceptibility, and the hereditary disposition of the reacting organism may be of importance, but also the mitoses in parent-cells of future cancer cells. Neoplastic conversions of somatic vertebrate cells in vivo and in vitro may take place during mitosis, presumably by mutant change, for this transformation is discontinuous, permanent, irreversible, inheritable, non-specific, and disproportionate to the presence of carcinogens. If a certain cell nucleus is in an active phase of division, but the whole organism on the contrary is in a passive stage of atrophy as in old age, a mitotic agens or toxin may be able to alter the normal sequence of karyokinetic processes. Cell-divisions are seen regularly in so-called “irritated” tissues, in ulcera ventriculi, in lesions caused by stones in the vesica fellea and urinaria, in chronical inflammations of the epithelium of the portio vaginalis uteri, of the sulcus coronarius in phimosis, of the lips and mouth of smokers and of chewers of tobacco and betel, of the skin in kangri-cancer and mule spinners’ cancer, and in neoplasmas elicited by radium, X-rays, ultraviolet rays of 2537 A, and by chemical products. Since cell-divisions are inevitable in operation wounds, a mitotic poison may have a special chance in the operation field. On the contrary the genuine nerve cell, showing no cell-divisions after the first years of life, never produces malignant tumours; of the nerve tissue only the less differentiated glial cells and the epithelium cells of the canalis centralis of the central nervous system show both as well persistence of cell-division capacity as tendency to malignant growth.

The effect of the presence of a carcinogenic agent will not only be connected with its specific potency, with its concentration and with the duration of application, but also with conditions of solubility. It is known that 3.4-benzpyrene is readily soluble in aqueous solutions of sodium desoxycholate. Endogenous sterols, gonadal and adrenal hormones (dehydrogenated analogies of which do not seem to exist in the organism), and exogenous, habitual inciting agents like sterols, fats, nicotine and alcohol, with their selective distribution, localization, excretion, affinity and toxicity, may act as solvents, vehicles and carriers of carcinogens. In ordinary life occupational and experimental carcinogens may not be involved, such as benzene, aniline, dimethylaniline, tolidine, benzidine, dianisidine, naphthylamine, anthraquinoline, crude and processed mineral and gas tar oils, tar, pitch, soot, arsenic, chromates, nickel carbonyl, asbestos, active radiations, urethane, azo compounds, CCl4, and dehydrogenated polycyclic hydrocarbons.

The transformation of a normal human cell into a cancer cell may be due to a chain of causal factors, all indispensably necessary for the effect of the interaction of the carcinogenic substance and the protein of a susceptible cell: action of a sufficiently potent agent in solution in sufficient concentration during a sufficient time as a mitotic poison on the karyokinetic processes of a nucleus in division, in an organism with an adequate hereditary susceptibility, or with endocrinal or other degenerations, or with resident, indigenous viruses. Only if the whole chain of causal factors is intact, the transition or mutation of a normal cell into a cancer cell may be involved at the moment of mitosis.

The present day attack against malignant growth, consisting in an early diagnosis followed by the prompt and expert application of well established methods of treatment, being the only way of evading a fatal progress of existent malignant disease, may be completed but by no means neutralized by preventive measures. The chain of causal events, of which therapeutic attack is the last link, may also be broken at the very first link by avoiding the ingestion of black crusts of human food. The prevention of cancer seems to be surprisingly easy, even easier than that of infectious diseases. As experiments in mice seem to indicate, the colour of the dangerous crusts is a kind of “signatura rerum”: the black colour due to partial carbonization indicates the possibility of the presence of carcinogenic substances, the brown colour due to caramelization on the contrary indicating their absence. In human food caramelized brown compounds may prove to be inoffensive, but there seems to exist some possibility that coal-black crusts can cause cancer. In all human surroundings, except under some very special conditions, carcinogenous compounds may be absent, but the possibility might exist that in human food pitch-black pyrogenous products possess cancer-producing power. A slight correction of our basic kitchen techniques seems to be advisable: boiling and stewing, even with the formation of brown crusts being apparently inoffensive, the avoidance of the formation of black crusts in the processes of roasting, frying and baking might be desirable.

Summary

1. These experiments seem to be the first in which materials which induce cancer of the skin in white mice by current laboratory methods have been obtained by ordinary kitchen treatment of ordinary human food; other sources of such materials seem to be absent in normal human surroundings under the circumstances of ordinary daily life.
2. Human food which produces no tumours of any kind in white mice shows cancer-producing properties for mice after exposure to a temperature sufficient for the formation of black crusts; brown crusts, however, being inoffensive for mice according to these experiments.
3. The probability of a significance of these results obtained in white mice for human pathology is open to discussion, in as far as the theory presents itself that the ingestion of a carcinogenic substance in black crusts of human food might indeed be an indispensable link in the chain of causal factors leading to “spontaneous” malignant growth.
4. Planned research in geographic neoplasmology is urgently needed.