CHAPTER FIVE: ENVIRONMENT AND CANCER
The Importance of Environment
One of the most challenging problems in public health involves the influence of man’s environment on the incidence of cancer. Differences of opinion concerning the extent of this influence are likely to have important practical consequences. If a specialist believes that cancer is caused primarily by genetic factors or by the aging process, his hopes for controlling the disease will focus on advances in surgery, radiology, and chemical therapy. He will tend to regard the occurrence of the disease as inevitable. On the other hand, if he strongly suspects that environmental factors play a major role, he is likely to attach a great deal of importance to cancer prevention. He will regard environmental change as a serious problem requiring careful investigation. Nothing can be done to arrest the aging process or alter the individual’s genetic endowment, but almost any environmental factor that influences the incidence of cancer is subject to some degree of social or individual control. The more effectively control is exercised, the lower the incidence of cancer is likely to be.
To what extent, then, are the two points of view supported by evidence and scientific authority?
Fortunately, it has never been established that cancer results from the aging process. At most it has been shown that the highest incidence of certain types of cancer occurs in the later years of life-a fact that can be explained by the longer exposure of older people to cancer-causing substances. “The senescence theory does no more than paraphrase a fact,” notes Sigismund Peller; “it fails to indicate how and where the changes are initiated which end in cancer.” The theory that cancer is inheritable rests on more impressive evidence and, for a time, occupied a commanding position in cancer research. It had its greatest following in the 1920’S, when Maud Slye managed to produce highly inbred strains of mice in which nearly every specimen acquired a cancer. Even the type of cancer and the age at which it appeared could be predicted with reasonable and often astonishing accuracy. Miss Slye advanced the view that cancer is due to a single genetic factor that determines nearly every aspect of the disease-the type of cancer, the site on which it arises, and the age at which it appears. Whether an individual acquired a cancer or not seemed to depend almost entirely upon his genetic endowment.
But the world is not a genetics laboratory in which individuals are mated to prove a point about hereditary cancer. Fortunately, examples of hereditary cancer in man are rare. After further research, it became evident that seemingly hereditary cancers in certain highly inbred strains of mice depended upon a complex of nongenetic factors. In experiments with one of the most “inheritable” forms of the disease-breast cancer in the mouse-Bittner demonstrated “that the kind of milk given to a litter accounts for more than all hereditary factors together. Whether a high or low probability of spontaneous breast cancer may be expected depends primarily upon the milk (mother versus foster mother taken from a different strain) which the new born mouse of a genetically pure strain sucks in the first twenty-four hours after birth.” Bittner’s “milk factor” is now believed to be a virus. Clarence Little found that the susceptibility of inbred mice to certain tumor grafts could not be explained without taking at least fourteen different genes into account, and even then many problems arose when the tumors were transplanted. Although a predisposition to cancer may be inherited, it is difficult to believe that an individual is foredoomed to acquire the disease by a genetic factor. “On the contrary,” observes Charles Oberling, director of the Institute of Cancer Research in Paris, “cancerous heredity is discovered to be a mosaic of factors that intervene in the most varied domains of the anatomic, physiological, and immunological constitution and cooperate in preparing a soil that is favorable for the emergence of a malignant neoplasm.”
It is tempting to add that if the “soil” is prepared by heredity, the “seed” is planted by the environment. A predisposition to cancer is not an absolute that is either present or absent; it is a matter of degree, varying from person to person. It is likely to manifest itself in malignant tumors to the extent that individuals are exposed to carcinogens. The greater the exposure, the higher the incidence of cancer. In fact, cases are known in which the activity of a particular carcinogen largely determines the rate of death from cancer. For example, Peller found that cancer accounted for 51 per cent of the deaths among the radium miners of Joachimsthal in Czechoslovakia. More than half the miners in this community were confronted with the prospect of dying of cancer, an appallingly high percentage. As cancer “normally” accounts for 20 per cent of the adult mortality rate, Peller reasons, approximately 39 per cent of those who might never have succumbed to the disease died of it. In these circumstances, variations in individual predispositions to cancer are so much overshadowed by the potency of a carcinogen (radon ) that the “seed” can germinate virtually without any “soil” prepared by hereditary or constitutional factors.
Indeed, one may well ask: What is a “normal” cancer death rate? Is the toll predetermined by genetic factors or will rates vary markedly in different environments? The most reliable evidence at hand suggests that the incidence of cancer and the number of lives claimed by the disease vary appreciably from one kind, of environment to another, even within the same general region. An extensive survey of cancer illness in Iowa during I95I, for example, showed that the incidence of the disease was proportionately almost 40 per cent higher in urban areas of the state than in rural areas. Although large differences could be expected in the incidence of lung cancer, it is noteworthy that a higher rate was reported among city dwellers for all forms of cancer. Urban women had a 25 per cent higher incidence of breast cancer and a 50 per cent higher incidence of genital cancer. Surprisingly, urban men had a much higher incidence of skin cancer (45 per cent), despite the fact that people in rural areas are exposed to greater amounts of sunlight, a natural carcinogen. These variations could not be accounted for by any differences in the quality of medical services or in access to hospitals and doctors. “Rural areas in Iowa appear to have access to the same degree and quality of medical care that urban areas have. Also, the cancer incidence rates for rural areas in metropolitan counties were no higher than for rural areas in remote non-metropolitan counties.”
In addition, the incidence of cancer is higher among certain income groups and occupations. The disease occurs most frequently in the poorest strata of the metropolis. Harold F. Dorn and Sidney J. Cutler found that among “white females the incidence of cancer was I4 per cent above average in the lowest income class, but occurred with about equal frequency in the four other income classes…. The most consistent relationship is the observation of relatively high incidence among members of the lower income classes.” Although a few forms of cancer show a mild preference for the rich, most types appear with the greatest frequency among the urban poor. Workers who are exposed to industrial carcinogens have a higher incidence of cancer than clerks, and those in nearly all major urban occupations have relatively more cases of the disease than do farmers.
Viewed in this perspective, cancer seems to have certain features in common with tuberculosis. Although the two diseases are fundamentally different in many respects, it is noteworthy that the incidence of both cancer and tuberculosis increases with urbanization. Nor can we overlook the fact that cancer, like tuberculosis, claims the largest number of its victims from among the urban poor. The organs in which the diseases take root “are not only determined by the properties of the particular pathogenic [disease-causing] agent involved, but also by the type or route of contact with it,” observes W. C. Hueper. “The conditions of exposure which decide the development of tuberculosis of the lungs and of the intestine-inhalation and ingestion, respectively-are fundamentally no different from those which are responsible for the development of cancers of the skin, bones, lungs, etc., after exposure to radioactive material by skin contact, inhalation, ingestion, or parenteral introduction.”
Tuberculosis became pandemic only after man’s environment underwent a radical change. Although rural workers were familiar with hunger before the Industrial Revolution, they seldom contracted tuberculosis in large numbers, partly because they lived in open surroundings and engaged in healthful work. The disease, we may recall, turned into the “white plague” when men were crowded into large cities and overworked in dismal factories. It spread among them rapidly even though, in certain cases, they were getting more to eat. Tuberculosis promoting factors overshadowed advances that might well have prevented a high incidence of infection. Today we know that control of the disease is achieved primarily through social advances, better sanitation, and improvement of the standard of living.
The same attention to all the aspects of our environment may be required to control cancer. There is good reason to believe that nearly everything that seriously disturbs the proper functioning of the human organism, particularly many of the toxicants produced by the present-day industrial revolution, contributes to the incidence of the disease. We shall find that cancer is promoted by emotional stress, poor diet, and a large variety of environmental pollutants. While the disease remains a “riddle,” to borrow Oberling’s word, it threatens to become the major cause of death in the United States and western Europe. In 1945 one out of every eight Americans was expected to die of the disease; today the figure is one out of six. Twenty-five per cent of the population now living in the United States will eventually acquire cancer, and approximately 450,000 new cases of the disease are discovered annually. Sweeping environmental changes in the nineteenth century produced the “white plague”- pandemic tuberculosis; changes in the present century seem to have spawned a “gray plague”-pandemic cancer.
The complexity of the disease should not be underestimated. It is highly improbable that any cancers arise “spontaneously,” that is, in the absence of a carcinogen. Whether a carcinogen initiates a series of changes in a cell or is continuously involved in them, the biochemical reactions that lead to malignant tumors often occur in stages, many of which are manifested by abnormalities in pre-cancerous tissue. Whether it is in the stomach, the uterus, or the prostate gland, “cancer rarely appears as a bolt from the blue,” Oberling observes. “Almost always it is preceded by nutritional disturbances in the epithelium that masquerade under various names but that are, after all, merely the expression of an abnormal reactivity to certain influences.”
Many cancer specialists believe that once a cell is altered by a carcinogen, the change is irreversible; the cell acquires an “imprint” and transmits it to all its descendants. Whether or not a cancer develops depends upon many factors, such as the potency of the carcinogen, the extent of exposure to other chemical and physical agents, and the individual’s constitution. Malignant cells may become detached from a localized cancer and enter the blood stream without producing new tumors. Apparently, certain resistance mechanisms must be overcome before malignant cells can take root elsewhere in the body and produce invasive, secondary cancers.
A cancer can be induced by a single chemical agent or by several, by small, irregular doses of the carcinogen or by a few large ones. The disease may not manifest itself for many years, even decades. A carcinogen may reach man by way of the air he breathes, the fluids he drinks, and the food he eats. It may harm a specific part of the body or the organism as a whole. Any powerful agent or stimulus that alters the normal functioning of human metabolism over a long period of time is potentially carcinogenic. Let us concede that the word “normal” has yet to be defined and that each individual has different biological needs. But let us also acknowledge that the incidence of cancer can be increased enormously by reckless changes in man’s environment, that millions of lives are placed in jeopardy by thoughtless alterations in man’s manner of life, diet, and physical surroundings.
The General Envionment
What are some of the general, or nonspecific, factors that influence the onset and progress of cancer? If this question is often difficult to answer, the reason is not that these factors are too few; on the contrary, there are too many of them, and the problem is to decide which ones are the most important. There is evidence that both the occurrence of a cancer and its course are influenced by the subject’s emotional make-up, his diet, and his occupation. The adequate treatment of a cancer, in turn, requires knowledge not only of surgery, radiation, and chemical therapeutics but also of the patient’s character and environment. Although the course of many cancers is generally predictable, individual responses to treatment vary so widely that a specific cancer that usually causes death within a few months can, in some cases, be controlled for several years.
Emotional stress is an important element in the progress of certain cancers. At the annual meeting of the American Cancer Society in I 959, several cancer specialists focused attention on the role that anxiety seems to play in the onset and course of malignant tumors. Drawing upon his experiences as a radiologist, Eugene P. Pendergrass, the outgoing president of the
society, observed that even if psychological factors do not determine the eventual outcome of cancer, they seem to influence the development of the disease. “I personally have observed cancer patients who had undergone successful treatment and were living well for years. Then an emotional stress such as a death of a son in World War II, the infidelity of a daughter-in-law, or
the burden of long unemployment seems to have been a precipitating factor in the reactivation of their disease which resulted in death.” A similar viewpoint was expressed by John B. Graham, of the Roswell Park Memorial Institute in Buffalo. “The adversities and despairs that these people [cancer victims] are subject to, and their responses to them, are indeed factors in the development of cancer and the effectiveness of treatment.” At a press conference, Graham added: “It has seemed to us in some cases that despair precedes cancer. I would not say it is a cause, but it may be a factor which decides when cancer will develop.”
In view of what we know of Selye’s work on stress and the adrenal corticoids, these conclusions should not be surprising. Cortisone is used effectively to delay death in cases of such hopeless cancers as acute leukemia. The temporary remissions achieved by this hormone are often striking. After being treated with cortisone, a victim of acute leukemia often seems to be on the point of complete recovery, although in time the cancer invariably overcomes the effects achieved by the hormone. There are reasons for suspecting that, just as cortisone inhibits the growth of certain cancers, other adrenal hormones sustain and even propagate cancers of the breast and prostate gland. By removing the adrenal glands of patients suffering from these tumors, physicians often gain control of the disease for long periods of time. New methods are replacing adrenal surgery, but the inhibition of the production of corticoids is a basic technique in managing both types of cancer. According to Selye, the value of adrenal surgery in certain cases of cancer “would suggest that there are adrenal factors which stimulate cancer formation.”
It is tempting to speculate whether emotional stress, by influencing the output of the adrenal glands, contributes significantly to the rising incidence of cancer. The adrenal cortex normally secretes male sex hormones (androgens) and perhaps female sex hormones (estrogens). The corticoids are also directly involved in the secretion of hormones by the male and female sex glands. Many experiments have shown that injections of estrogens induce mammary cancer in cancer-prone mice, and intensive use of estrogen to control cancer of the prostate gland in men has been followed, in some cases, by breast cancer. “There can be little doubt that endocrine imbalance is the basis for chronic cystic mastitis, or maladie de Reclus, which is characterized by fibrous nodules that are sometimes interspersed with cysts, and it has been shown by Warren and by Foote and Stewart that cancer is five times as common in breasts affected by this condition as it is in normal ones.”
Oberling’s opinion, advanced in 1952, received strong support several years later from the results of autopsies performed on 207 women with cancers of the breast. The investigators found evidence of marked endocrine imbalances and particularly of prolonged estrogenic estrogens are steroids, and there is a close resemblance between the chemical structure of the steroid hormones and that of certain carcinogens. Some of these carcinogens, in turn, have a mild estrogenic activity; that is, they produce, on a smaller scale, effects similar to those produced by estrogens and corticoids. There is a possibility either that excessive secretions of the steroids turn pre- cancerous cells into cancerous ones or that the hormones are actually degraded into carcinogens. At any rate, the apparent relationships between stress and cancer and between the steroids and carcinogens can hardly be ignored.
“There is solid evidence that the course of disease, in general, is affected by emotional stress,” Pendergrass observes. “A growing list of authorities have lent support to this field of investigation. Foremost among these is Dr. Hans Selye who devoted much of his life to the subject. In order to learn about such stresses, one must explore beyond the usual limits of a clinical history. This takes time. Busy doctors may be unable to give the time to secure a good history concerned with environmental, constitutional and other factors; a history of emotional stress, or just the opposite; a climate of happiness and security. Could this not be an area for another discipline in medicine, [for] a social scientist for instance; or maybe this is a field where the social worker should expand his or her activities. I believe that research in this area will, almost certainly, lead into investigations of factors that are susceptible to control hormonal systems and other general metabolic influences. Thus, we as doctors may begin to emphasize treatment of the patient as a whole, as well as the disease from which the patient is suffering. We may learn how to influence the general body systems and through them modify the neoplasm which resides within the body.”
Cancer does not always grind inexorably to its logical end. At least 150 spontaneous remissions of cancer in man have been carefully authenticated by reputable scientists. A number of these tumors had already metastasized widely and were regarded as incurable. As Oberling observes: “. . . it is difficult to estimate the frequency of spontaneous cure because cancer is generally not recognized until it has reached an advanced stage. It is entirely possible that many malignant growths disappear in their initial stages, leaving no trace behind. This, of course, is the purest speculation.”
But Oberling’s speculation is not reckless. Research has clearly demonstrated that organisms possess mechanisms for resisting cancer. According to a report by the Sloan-Kettering Institute for Cancer Research, defense mechanisms have been observed in mice receiving transplants of a highly malignant cancer which usually kills 95 per cent of its victims in three weeks. “If these animals were given a small dose of material extracted from yeast cells called zymosan, they defended themselves vigorously against the cancer. Instead of killing the mice, the cancers were gradually dissolved until they disappeared entirely. The animals were then completely cured. No cancer has ever returned, and strong immunity to a second implantation has been demonstrated as long as I I months after the first growth was initiated. Remarkably, these cancer cells grown in tissue culture are not disturbed at all when put in direct contact with zymosan. This indicates that zymosan does not inhibit the cancer directly but acts in some way in mice to enhance their natural defenses, tipping the scales in favor of the endangered host and against the cancer.” The yeast extract, unfortunately, “is too toxic in man for clinical use.”
This line of research is important because high resistance depends not only on hereditary factors but also on good nutrition, a sound state of mind, and balance in work and play. It is fairly well established that the high incidence of liver cancers in depressed areas of the world is due in large part to dietary deficiencies of protein and possibly the B vitamins. Rats nourished on similarly deficient diets have a high incidence of the disease, although it is rare in animals that receive the same food supplemented with wheat, yeast, or powdered liver. Dietary factors may also contribute to the relatively high incidence of chorioepithelioma (a cancer of the uterus) in the Far East. In general, however, experimental data are inconclusive for other forms of cancer because the relationship between animal tumors and nutrition varies with the strain and species used. “It does appear that maintenance of body weight at a minimum compatible with good health may prevent or delay the development of cancer in many individuals,” observes Helen N. Lovell in a review of the available evidence. “A diet adequate in all nutrients seems to be most desirable. Many believe the oral and pharyngeal symptoms of chronic riboflavin and nicotinic acid deficiency are precancerous factors.”
Despite the influence of all the factors we have been discussing, many well-nourished and serene individuals perish from virulent cancers fairly early in life. Why? What are some of the factors that might produce malignant tumors in persons whose diet, habits, and outlook are sound? These questions introduce a formidable problem. Man is exposed to a large number of cancer-causing agents; some of them are natural in origin and virtually impossible to avoid, but a substantial number of them are man-made. A few of these artificial agents are useful in medicine, but the majority can be either completely dispensed with or at least used in such a way that they do not affect the general public. The presence of these agents in the human environment has aroused a great deal of concern, and their removal requires vigorous social action.
Man is capable of producing at least two types of agents that have cancer-causing properties. One of them, ionizing radiation, will be discussed at some length in the following chapter. Our principal concern here will be with the chemical carcinogens, ranging from confirmed cancer-causing substances to those that are highly suspect.
The suspicion that chemicals in man’s environment can cause cancer first arose nearly two centuries ago, when Sir Percival Pott, a London surgeon, identified “chimney sweep’s disease” as cancer of the scrotum. Pott unerringly pointed to soot, the tarry material formed by the combustion of solid fuels, as the cause of this malignant tumor. In the eighteenth century the job of cleaning chimneys was given to boys who were small enough to negotiate the elaborate flues in the London homes of that period. As the boys worked, their trousers rubbed soot into the sweat glands of the scrotum. In later life, perhaps ten or fifteen years after this kind of work had been abandoned, the scrotum would begin to ulcerate and then develop a cancer which invariably claimed the life of its victim. In 1914 two Japanese investigators, K. Yamagiwa and K. Itchikawa, succeeded in developing skin cancer in 41 of 178 rabbits who had received applications of tar to the internal surface of the ear. Three years later H. Tsutsui was able to similarly induce tar cancer in mice.
The success of these Japanese investigators met with warm professional appreciation-and for good reason. Their work opened a new era in cancer research-the study of chemical carcinogens-and with it there arose a host of disconcerting problems concerning the behavior of these agents in men and animals.
Yamagiwa and Itchikawa would not have been able to produce tar cancer of the skin if they had used guinea pigs instead of rabbits. Although tar induces cancer in man, it does not produce the disease in the guinea pig, an animal commonly used in experimental work. In all probability, the Japanese investigators would also have been unsuccessful if they had used rats. Despite repeated attempts from 1914 on, cancer of the skin could not be produced in this animal through applications of tar until 1935 They might also have failed with dogs, whose response to tar is “sporadic and inconstant,” in Oberling’s words. In short, evidence that tar is carcinogenic in animals was acquired partly through luck and partly through patience. Yamagiwa and Itchikawa had the good fortune to select an animal that is more susceptible to tar cancer than other experimental subjects, but they had to persevere through at least 150 applications of the carcinogen before malignant tumors were induced.
Their experiment was not the first attempt to induce cancer in animals by means of tar. Similar efforts were made in 1888,1890, and 1894, and again as late as 1913, by Haga, another Japanese investigator. All the attempts failed. The experiments in the last two decades of the nineteenth century were performed mostly on dogs and rats. Haga even turned to the rabbit as a subject, but he failed to keep at the tar applications long enough. The medical pundits of the day thereupon concluded that chemical carcinogenesis did not exist. “For decades these few negative [experimental] results weighed more heavily than numerous observations to the contrary in chimney sweeps, tar, paraffin and dye workers, and mule spinners,” Peller observes. “. . . only after positive experimental results on animals had been published, did cancerologists change their opinions. It was formerly considered scientifically sound to deny the relationship; now it became sound to accept it. If Yamagiwa and Itchikawa (I9I4) and Tsutsui (I9I7) had chosen the rat for their experiments, instead of rabbits and mice, then the tar genesis of certain skin epitheliomas in man and the theory of irritation in cancer would have been ‘expertly’ denied for another I7 to 20 years, and the whole development of chemistry of cancerogens would have been retarded.”
The same story can be told about beta-naphthalamine, a compound used in the preparation of a number of coal-tar dyes. Beta-naphthalamine, which induces cancer of the bladder, is one of the most potent carcinogenic substances in man’s environment; it has claimed hundreds of lives among workers in the synthetic-dye industry. Attempts were made as early as I 906 to induce cancer in rabbits by means of synthetic dyes, including betanaphthalamine, but the experimenter, B. Fischer, did not persist long enough in his efforts. After producing cancer-like cells that regressed spontaneously when injections of the carcinogens were stopped, Fischer concluded that the dyes did not induce occupational cancers. This conclusion was generally accepted by the medical profession. Not until the late 1920’S were cancers produced experimentally by beta-naphthalamine, but even then rabbits proved to be poor subjects and rats and mice were found to be resistant to the compound. In 1938, more than thirty years after Fischer’s work, Hueper and his coworkers discovered that the only animal that readily contracted cancer of the bladder from beta-naphthalamine was the dog.
The reason for inflicting these details on the reader should be fairly obvious. Cancers in animals “cannot provide irrefutable proof of the safety or carcinogenicity of a substance for the human species,” observes a report by the Food Protection Committee of the National Academy of Sciences. The committee adds that “it is perhaps reassuring that the known carcinogenic responses of humans to certain chemicals are similar in many ways to those seen in experimental animals”-but the qualifying words “perhaps,” “certain,” and “in many ways” carry a heavy burden. Actually, some carcinogens are handled quite differently in man from the way they are handled in common experimental animals. Beta-naphthalamine, for example, is not carcinogenic as such. In the human body, however, it is metabolized into the carcinogen 2-amino-I-naphthol, whereas resistant animals either fail to produce this metabolic compound or create it in very small amounts.
Many subtle physiological differences between animal species-indeed, between individuals in the same species-are likely to affect the potency of a chemical carcinogen. “The only universal carcinogen is apparently ionizing radiation,” Hueper notes. Whether or not a cancer is produced by an environmental carcinogen depends on the organ with which it comes in contact, on the solubility of the carcinogen in the body’s fluids and therefore “on the speed of removal from the site of primary contact of a carcinogen, upon the metabolism of the carcinogen, on its route of excretion and on the tissue or organ of retention or deposition of a carcinogen. The target organ may vary in different species.” Hueper concludes that the ability of a carcinogen to affect different species and organs in different ways seems to depend on the distinctive metabolic processes of the animals involved.
These factors apparently account for the fact that several types of cancer that are common in man are difficult to produce in animals. For example, until very recently the most common form of stomach cancer in humans (adenocarcinoma) could not be produced in rabbits, mice, rats, and dogs by feeding them various carcinogens. On the other hand, adenocarcinomas of the small intestine, while rare in man, are fairly easy to produce in animals. The lung tumors produced in mice do not resemble the lung tumors in humans that have aroused so much professional and public concern. Many cancer specialists even disagree on whether these tumors in mice are malignant or benign. Spontaneous regressions of cancer in mice, animals widely used in cancer research, are not uncommon, but the regression of diagnosed cancers is very rare in man.
When the researcher finds that a particular compound does not induce cancer in rats, mice, or dogs, he can only hope that it will not cause the disease in man. He can never be certain of this unless carefully controlled experiments on a large number of human volunteers are continued for decades without causing a cancer that is clearly attributable to the particular chemical. Experiments with animals are likely to be more painstaking if the compound being tested is chemically related to a number of known carcinogens. But there are no clear-cut groups of carcinogenic substances. Some known carcinogens are highly complex hydrocarbons; others, such as chromium, nickel, and arsenic, usually reach man as simple inorganic compounds. Often the most trivial chemical change can transform an apparently non-carcinogenic substance into a potent carcinogen. A mere shift of the NH2 group in alpha-naphthalamine, a non-carcinogenic compound, from one position to another in the molecular ring of the chemical produces the carcinogen beta-naphthalamine.
Many carcinogens have been able to enter man’s environment because of the relative crudeness of experimental techniques. Even today, despite advances in testing methods, carcinogenic properties are being discovered every few years in established food additives. At least ten chemicals listed as recently as I 956 by the National Academy of Science in The Use of Chemical Additives in Food Processing are currently suspected of being cancer-causing agents. These include a cheese preservative (8-Hydroxyquinilone), a flavoring agent for root beer (safrole), a cheese stabilizer (carboxymethyl cellulose), and several coal-tar dyes. Perhaps the worst offenders are the dyes, notably Yellow AB and OB, which were used for decades to color butter, cheese, cake, candy, cookies, drugs, and cosmetics. When ten commercial samples of these food colors were analyzed by Walter D. Conway and Elizabeth J. Lethco, of the National Cancer Institute, they were found to contain very large residues of beta-naphthalamine, ranging from a low of 76 to a high of 908 parts per million.
Although recent legislation stipulates that no carcinogen may be used as a food additive, there are several noteworthy exceptions. Lead arsenate, for example, is still employed as an insecticide in orchards and vineyards. Residues of the compound (usually 7 parts per million) are allowed to remain on such common fruits as apples, pears, plums, peaches, cherries, and grapes. ‘`It has been known for many years that exposure to arsenicals produces a certain type of cancer of the skin,” Francis E. Ray, director of the University of Florida’s Cancer Research Laboratory, told the Delaney Committee. “More recent evidence is that inhalation of arsenic dust-and sprays-may cause cancer of the lungs. It is possible that other types of internal cancer may be caused by the long-continued ingestion of so-called non-toxic doses of arsenicals.” Ray declared that the use of arsenicals on crops “should be prohibited.” The demand, although reiterated by many outstanding cancer specialists, has largely been ignored.
“While not within the scope of this investigation,” Ray added, “it may be mentioned that female sex hormones [estrogens] have been incorporated into cosmetics. These hormones have been shown to induce cancer in female animals when administered for a prolonged period. Their use in cosmetics should be prohibited.” (The most commonly used estrogens in facial creams are estradiol and estrone. Hueper warns that although estrogenic creams are applied to the skin, they occasionally produce systemic effects.) Thus far, however, no action has been taken to forbid the use of estrogens in facial creams. In fact, the use of synthetic hormones has been extended, not reduced. For example, synthetic hormones, notably stilbestrol, are now added to 85 per cent of the cattle feed produced in the United States. A woman who purchases a cosmetic that contains hormones exercises a choice, but a consumer who acquires residues of stilbestrol in meat belongs to a “captive” population whose wishes have not been consulted by food producers or government agencies. Whether the stilbestrol fed to livestock makes its way into meat has been the subject of heated controversy. Regulations prepared by the F.D.A. stipulate that daily feedings of the drug should not exceed IO milligrams for steers and 2 milligrams for sheep. According to the F.D.A., residues of stilbestrol are avoided when hormone-containing feed is withdrawn forty-eight hours before slaughtering. Actually the evidence for this claim is equivocal. Although F. O. Gossett and his co-workers found that residues did not appear in meat when the F.D.A.’s regulations were followed, C. W. Turner detected traces of stilbestrol in the lungs and kidneys of animals slaughtered forty-four hours after the prescribed 10-miligram dosages had been withdrawn.
Experimental work on mice by a group of researchers at the National Cancer Institute indicates that stilbestrol is a highly potent drug. Practically all the mice that had received a subcutaneous implantation of I milligram of the synthetic hormone acquired tumors of the testicles within seven months. When the pellet was removed after eight weeks, no gross tumors developed, but when it was reimplanted twenty-four weeks later, the incidence of tumors rose sharply, as though the pellet had never been removed. The effect of a brief stimulus is never lost; exposure to stilbestrol seems to be cumulative.
Thousands of man-made chemicals with undetermined properties enter our food supply, water, and air. Some may be co-carcinogens; these lack significant cancer-causing properties of their own but increase the potency of known carcinogenic agents. Others may be weak carcinogens whose cancer-causing effects are additive and synergistic. Any such carcinogens and cocarcinogens would escape the attention of the researcher unless they were brought into combination with one another. But this is a rare experimental procedure. “Potential carcinogenic contaminants,” Hueper warns, “also may be introduced into foodstuffs if vegetables, fruits, fish, oysters, and livestock are grown on soil or in water polluted with known carcinogens, such as radioactive matter, arsenicals, selenium and polycyclic hydrocarbons contained in ship fuel oils, since these chemicals may be taken up and stored by the vegetable and animal matter growing in such contaminated media…. There exists also the possibility that originally noncarcinogenic additives and contaminants may interreact with each other or with food constituents and form new compounds possessing carcinogenic properties in the foodstuffs. They may be produced under the influence of processing procedures or during the preparation of food in the kitchen. Plastics used as wrapping material, sausage skins and coating material of fruits, cheese, meat, butter, and can linings may carry similar hazards.”
These are not irresponsible vagaries. A number of chlorinated hydrocarbons, such as DDT, seem to have mild tumor-inducing effects.. Nearly all insecticides are dissolved in petroleum distillates, many of which are also suspected of being carcinogens. Crude and semirefined paraffin waxes, from which cosmetics and food containers are made, have carcinogenic properties and, in some cases, contain residues of highly potent cancer-causing substances. Hans L. Falk, Paul Kotin, and Adele Miller found that the protein in ordinary milk could eluate, or extract, hydrocarbons from the wax coating of their containers. The investigators had deliberately impregnated the wax with these hydrocarbons so that the eluant properties of milk could be tested. “However,” they write, “an analysis of several dairy waxes revealed the presence of a small amount of 1,2,5,6-dibenzanthracene of the order of one part per million.” Dibenzanthracene, a highly potent carcinogen, appears as an impurity in many semi-refined parafflns.
It is difficult to specify all the potential hazards to public health that are created by saturating man’s environment with poorly tested, exotic chemicals. Many cancers are stealthy; they develop slowly and imperceptibly. Decades may pass before mistakes committed today finally bear their malignant fruit; we may pay a heavy price in human life, not so much as a result of brief exposures to a few potent carcinogens as because of the exposure of millions of individuals to small quantities of “harmless” compounds over long periods. The potent carcinogens “by their very strength are almost sure to be discovered clinically,” Kotin writes. “In essence, a strong carcinogen identifies itself by the very circumstances which aroused suspicion in the first place. It is assuredly the less potent carcinogens that seem to be more important in human cancer, and it is these that provide the real problem for evaluation.”
Environment and Lung Cancer
The incidence of one kind of malignant tumor, lung cancer, has begun to increase at an appalling rate. “A century ago, Rokitansky of Vienna, the most experienced pathologist of his time, considered lung cancer a rarity” Peller observes. “In other and older books with detailed descriptions of cancer, the lungs are not mentioned at all, whereas cancer of the larynx and of the nose are discussed. Fifty years ago, in hospitals, larynx cancer in the United States was 1.5 times as frequent as lung cancer. In 1947-48, 9 per cent of all fatal cancers, namely 18,144,, developed in the respiratory organs and only 1,870 of them originated in the larynx. The number of recorded lung cancers is rising more rapidly than that of any other subgroup.”
Although it could be claimed that before X rays were discovered many lung cancers were mistaken for other respiratory disorders, the opinion of an experienced, conscientious medical man such as Rokitansky cannot be ignored. If lung cancer had been a common disease a century ago, it would not have been regarded as a medical curiosity by many of the leading pathologists of the day. During the last five decades the incidence of lung cancer increased at a startling rate, and today the disease is very common. In 1914 seven white American males per million died of lung cancer. By 1956, this figure had reached 284 a fortyfold increase in less than half a century.
It is difficult to attribute these higher figures to improvements in diagnostic technique. By the 1930’S, X rays were being used extensively in the United States and biopsies involving the microscopic analysis of cancerous tissue were frequently performed. Nevertheless, the number of deaths from lung cancer among American males rose steadily after 1930 until it had increased 500 per cent by 1951. Even more conclusive evidence against the “diagnostic improvement” argument is supplied by data from the Copenhagen Tuberculosis Station in Denmark. “In a tuberculosis referral service, used extensively by local physicians, where diagnostic standards and procedures including systematic bronchoscopy remained virtually unchanged between 1941 and 1950, the lung-cancer prevalence rate among male examinees increased at a rate comparable to that recorded by the Danish cancer registry for the total male population,” Jerome Cornfield and his colleagues observe in an excellent review of the lung cancer problem. “This can be regarded as evidence that the reported increase in Danish incidence is not due to diagnostic changes.”
It is equally difficult to attribute the rise in the number of cases of lung cancer to the increased longevity of the population. Whatever may be the factors that account for the high incidence of cancer in older people, careful study has shown, Cornfield observes, that “only one sixth of the over-all increase in lung-cancer mortality among males in the United States (from 4 to 24 deaths per 100,000 males between 1930 and 1951) could be attributed to an aging population. Similar findings have been presented for England and Wales…. Allowance for increased average age of the population could account for only half this rise in lung-cancer mortality, with a 24-fold difference between 1900 and 1953.”
Nor is it a convincing argument that the rise in deaths from lung cancer is due to cancers that begin in one part of the body and later migrate to the lungs. Autopsies have proved this to be a fallacy. Cornfield and his associates reviewed thirteen series of such autopsies, the first begun early in the present century and the last completed in the 1950’S. These autopsies support the claim that mortality from primary lung cancer has increased sharply during the past fifty years.
The inescapable conclusion seems to be that the rising incidence of lung cancer is due to recent changes in man’s environment and manner of life. What are the changes? The majority of American biostatisticians and cancer specialists believe that cases of lung cancer are multiplying because of the increased consumption of cigarettes, and they have collected an enormous amount of statistical evidence to support this belief. “Twenty-two. . studies of smoking in relation to lung cancer have now been carried out independently in eight different countries,” notes E. C. Hammond, of the American Cancer Society. “In every one of these, a far larger proportion of smokers (particularly heavy cigarette smokers) was found among lung cancer patients than among people without this disease. This has been shown for both men and women.” Hammond adds that “similar studies have shown an association between smoking and cancer of the buccal cavity, cancer of the larynx, cancer of the bladder, coronary artery disease (the most common form of heart disease), and Buerger’s disease (a circulatory disease of the extremities ).”
Thus far the most important attempt to correlate the rising incidence of lung cancer with increased smoking has been the survey made under the direction of E. C. Hammond and Daniel Horn for the American Cancer Society. The society mobilized 22,000 volunteers as researchers, and asked each of them to have about ten men between fifty and sixty-nine years of age complete a questionnaire on their smoking habits. Precautions were taken to query only individuals who were not seriously ill, who apparently were not suffering from lung cancer, and with whom the research volunteers enjoyed some measure of personal contact, so that thorough follow-ups could be made. Nearly I 90,000 acceptable questionnaires were obtained during the first half of 1952 Almost 188,000 were followed up by the volunteers through October 1955. A careful study of the death certificates of about 12000 men who died during the three-year period yielded remarkable results. Comparing smokers with non-smokers in the same age groups, Hammond and Horn found that the general death rate was 9 per cent higher for occasional smokers, 12 per cent higher for pipe smokers, and 68 per cent higher for cigarette smokers. The death rate for cigarette smokers increased progressively with the quantity of cigarettes smoked, nearly doubling with a rise in consumption from a half pack to two or more packs daily. It was found that the incidence of lung cancer among cigarette smokers was almost twice as high as it was among non-smokers. Comparing moderate smokers (a half pack daily) and heavy smokers (two or more packs daily ) with non-smokers, Hammond and Horn found that the ratio of death from lung cancer soared to 17 to I and 60 to I, respectively.
Those who maintain that there is no relationship between lung cancer and smoking have been placed in an untenable position. At first it was held that there was no direct evidence to implicate tobacco tars in cancer in animals. This contention was shaken when at least six different laboratories reported that such tars induced skin cancer in mice. It was then claimed that although tobacco tars may cause skin cancer in mice, it had not been shown that they could cause lung cancer in mice, men, or any other species of animal. This argument is difficult to sustain, at least as far as men are concerned. In a co-operative project initiated in I 954 under the direction of two leading American pathologists, Oscar Auerbach and Arthur Sout, the entire breathing apparatus was systematically removed from all individuals autopsied at the Veterans Hospital in East Orange, New Jersey, and at eleven hospitals in New York State. The investigators were able to obtain reliable histories of smoking on 402 of the subjects, all of them white males, 63 of whom had died of lung cancer. After preparing nearly 20,000 usable slides of lung tissue from the 402 subjects and carefully examining each slide under the microscope, the pathologists found that tissue abnormalities (changes in the epithelium) appeared in only 16.8 per cent of the slides from non-smokers, in 92.3 per cent of those from light smokers, in 97.4 per cent of those from moderate smokers (a half pack to a pack daily), and in nearly IOO per cent of the slides from those who smoked more than a pack daily. The changes in the lung tissue from atypical, potentially cancerous cells to invasive, cancerous cells increased with the average intake of cigarettes. Focusing their attention on all the subjects who had died from causes other than lung cancer, Auerbach and Sout found that 22.6 per cent of the slides from heavy smokers had atypical cells and 11.4 per cent had cancerous cells. On the other hand, only 0.2 per cent of the slides from nonsmokers had atypical cells and none had cancerous cells.
However, disinterested critics of the theory that smoking and lung cancer are related have performed a very useful function. By looking for other factors besides smoking, they have drawn a great deal of attention to the role that air pollution plays in producing the disease. The relationship between air pollution and lung cancer has largely been ignored despite evidence that lung-cancer rates are higher in cities than they are in rural areas. But as more research is done, zealots on both sides in the lung-cancer controversy are moving toward a more balanced position. An impressive amount of statistical evidence implicates air pollution as well as smoking in the rising incidence of lung cancer. White South African males, for example, are probably the heaviest smokers of packaged cigarettes in the world, and the number of deaths from lung cancer among them has increased from 11.7 per 100,000 in 1947 to 24.6 in 1956. But South Africa presents an interesting opportunity for statistical research. The country has acquired a large number of British immigrants, most of whom came from English cities with heavily polluted air. Is the incidence of lung cancer higher among British immigrants than among native-born South Africans? After a careful statistical study, Geoffrey Dean, of the Provincial Hospital in Port Elizabeth, South Africa, established that in the forty-five-to-sixty-four age group, deaths from lung cancer were 44 per cent higher among a group of British male immigrants than among native-born males. The incidence of the disease rose with the age at which immigration occurred, indicating that longer residence in England increased the possibility of dying from lung cancer. The difference in the death rate between native-born and British-born males could not be explained by differences in cigarette brands; the two groups smoked very much the same brands. But Dean’s data in no way indicate that prolonged smoking is not hazardous. On the contrary, the steadily rising rate of death from lung cancer observed in both groups of males supports the contention that cigarette consumption plays a major role in causing the disease.
The air over the metropolitan districts of the United States and Europe contains a large variety of known carcinogens. Perhaps the most potent of these agents are the so-called polycyclic hydrocarbons, such as benzpyrene and anthracene. By painting the skin of mice with benzpyrene three times weekly for six months, E. L. Wynder and his co-workers were able to produce cancers in nearly half the animals. An individual who smokes a pack of cigarettes a day acquires about 60 micrograms (millionths of a gram) of benzpyrene a year. If he lives in Helena, Montana, he inhales 0.8 micrograms of benzpyrene from the air annually; in Los Angeles, he inhales 20; in Detroit, IIO; and in London, 320. That is, he acquires the equivalent of the benzpyrene in a third of a pack of cigarettes daily merely by breathing the air in Los Angeles, the benzpyrene in nearly two packs from the air in Detroit, and the benzpyrene in five packs from the air in London. It should not be surprising to learn that death rates for lung cancer among males in England are more than twice those in the United States.
Kotin found that when an efficient automobile engine revolves 500 times per minute for one minute, it produces a gasoline exhaust containing 235 micrograms of benzpyrene. When the engine is accelerated to 3500 revolutions per minute, the amount of benzpyrene in the exhaust decreases to IO micrograms. A congested thoroughfare cluttered with idling or slow-moving motor vehicles is a hazard to human lungs. Individuals who spend a large part of their time in such areas, irrespective of the particular metropolis in which they live, are likely to acquire a lifetime dosage of benzpyrene that is thousands of times greater than the amount required to produce cancer in mice.
In addition to smoking and air pollution, socioeconomic factors have been found to play a role in the incidence of lung cancer. It is a tempting hypothesis that susceptibility to the disease is increased by poor nutrition, economic insecurity, and low physical resistance. “The existence of other important lung-cancer effects associated with such characteristics as socioeconomic class cannot be questioned,” Cornfield and his co-workers state. “Cohart found that the poorest economic class had a 40 per cent higher lung-cancer incidence than the remaining population of New Haven, Connecticut. Results of the 10-city morbidity survey [by Dorn and Cutler] have revealed a sharp gradient in lung-cancer incidence, by income class, for white males, which is consistent with Cohart’s findings. Since cigarette smoking is not inversely related to socioeconomic status, we can agree with Cohart ‘. . . that important environmental factors other than cigarette smoking exist that contribute to the causation of lung cancer.’ these and other findings are convincing evidence for multiple causes of lung cancer. It is obviously untenable to regard smoking of tobacco as the sole cause of lung cancer.” (This conclusion is apparently shared by Hammond and Wynder, who not only were co-authors of Cornfield’s report but also played an outstanding role in focusing public attention on the hazards of cigarette smoking.)
What can we learn from these studies? There no longer seems to be any doubt that the rising incidence of lung cancer is due to recent changes in our synthetic environment. So compelling are the data now at hand that virtually the entire scientific community has been obliged to accept this conclusion. One can only wonder why it was not accepted earlier. The growing incidence of lung cancer was noted as early as 1929 in the German Journal for Cancer Research by Fritz Lickint, who thought that the disease was related to smoking. In 1932 the same relationship was explored by W. D. McNally in the American Journal of Cancer Research. Lickint and McNally were followed by other investigators, including Alton Ochsner, who wrote on the subject in I94I, but the scientific community remained largely unmoved. More than a decade passed before Hammond and Horn finally presented the problem of the relationship between lung cancer and smoking in a form that was too compelling to be ignored. Until then, the rise in the number of cases of lung cancer was generally attributed to genetic factors or to increased longevity.
It should be noted that it has taken many years for an obvious link between environmental changes and the rising incidence of a major disease to gain wide acceptance among scientists. Because of its highly charged social, economic, and political implications, man’s environment is an arena that scientific research tends to enter with reluctance. There is a strong disposition to probe heredity, a more neutral field, or to dismiss grave problems of disease as the consequence of greater longevity. But now the evidence of the direct connection between the rising incidence of lung cancer and changes in man’s environment has affected this traditional approach profoundly. We now have good reasons for questioning whether any type of cancer need burden mankind as heavily as it does today. We have begun to take a closer look at the irrationalities in our synthetic environment. Perhaps we shall begin to eliminate them. But we cannot wait indefinitely before making man’s environment and manner of life more rational. Mortality from cancer is rising steadily. In the United States, known deaths from the disease increased 22 per cent between 1940 and 1959, an increase that can scarcely be accounted for by any monumental improvements in diagnostic technique. Recently a survey of 610 women by the New York City Cancer Committee showed that more than a third had symptoms suggestive of malignant tumors or a pre-cancerous condition. None of the women was more than forty-two years old. Robert A. Loberfeld, vice president of the committee, is quoted in a news item as describing the situation as “shocking.” Cancer has even reached epidemic proportions. At the Fourth National Cancer Conference, in September 1960, Lester Breslow reported that “tens of thousands” of trout in western hatcheries suddenly acquired liver cancers. Apparently the disease was “associated with a new type of diet,” not with a virus.
“Technological changes affecting man’s environment are being introduced at such a rapid rate and with so little control that it is a wonder man has thus far escaped the type of cancer epidemic occurring this year among the trout,” Breslow told his colleagues. “Or have we? Lung cancer is not too dissimilar except that the time period is decades, not weeks. If community air pollution of the types now prevailing or being introduced by technological changes is proved to be a causative factor in lung cancer, we would be dealing with exposure of millions of people. New pollutants are frequently added nowadays to community air through changes in industrial process, fuel composition and the like. They are added for reasons of economy or efficiency in industry or transportation-essentially the same reasons given for changing the diet of the fish. It is not inconceivable that although cigarette smoking has been the major factor in the lung cancer epidemic during the first half of the 20th century, some form of community air pollution could be responsible for an even larger epidemic of the disease during the second half of the 20th century.”