Document 2qQ6YppK7MmGvw6qY13VaNyoR

1 Francis J. C. Roe Cancer inducing agents It has become increasingly difficult to state whether a given agent will or will not induce cancer. It is poss ible that, given the appropriate conditions, all sub stances are carcinogenic ne carcinogens In industry and every day life vniline dyes vnlloxidants such as benzidine and /3>naphthylamine asbestos Certain metals: arsenic, beryllium, cadmium, chromium, obalt, nickel and selenium Certain medicines, particularly some of those used In the reatment of cancer; also some hormones Creosote ipoxides, used widely as adhesives and in the manufacture >f plastics ixhaust gases from petrol and diesel engines Smoked foods and charcoal steaks Boot, coal tar and smoke Tobacco smoke and raw tobacco used by `chewers' onizlng radiation Sunlight Bancer producing viruses ? WIDELY DIFFERING agents have been shown to be :apabla of causing cancer in man. In the past they rave been arbitrarily divided into chemical, physical ind viral. For many reasons this division is not very meaningful. Thus asbestos may cause cancer besausa of its physical rather than chemical nature . Francis John Caldwell Rob is Head of the Departrtent of Experimental Pathology at the Chester Beatty - -.search Institute, and Reader in Experimental Pathov at the University of London. His current research is with the mechanism of carcinogenesis, the dctecj a ofenvironmental cancer hazards and the application | such studies-to the prevention of cancer in man. Cancer can be induced in laboratory animals by deliberately exposing them to a wide variety of physical, chemical or viral agents some of which are known to cause cancer in man. It has become customary to regard the agents concerned as 'carcinogens'. Yet this may not be the best way of looking at the situation, because it suggests too strongly that the many different cancer inducing agents share a common feature which enables them to bring about the induction of cancer by a specific mechanism. Instead it may be better to regard carcinogenesis as an intimate `pas de deux', with the living tissue and the cancer inducing agent as the principal dancers on a stage which is the environment provided by the multi-cellular organism in which cancer will eventually arise. Both the dancers can appear in other roles at other times and the stage can be used for other purposes. Other dancers can act as substitutes, but the pas de deux itself always requires two dancers and a stage on which to appear. The animal body consists of a complex community of cells all derived originally from one cell, the fertilized ovum. Within each cell, coded in the arrangement of bases in its nucleic acids (deoxyribonucleic acid, DNA, and ribonucleic acid, RNA), is a plan of the whole body but, under normal circumstances, only a fragment of the plan is expressed by each cell. It may be presumed that the expression of the majority of the plan is actively suppressed by a pattern of control mechanisms. These include long distance hormonal and neural systems; local systems involving tissue antigen-antibody reactions and cell to cell contacts; and intracellular gene-suppressor systems, built-in to cells during their differen tiation from the undifferentiated state (see "The control of genes", Science ' Journal, March 1966). The normal steady state is maintained in the body by the continual operation and interplay of these control mechanisms, each acting as a negative feed-back system. Repair following injury is controlled by the same mechanisms, the combined function of which is to maintain, or in this case, re-establish the status quo. It is convenient, therefore, to apply the adjective `homeostatic' to these control mechanisms. Each body cell may be regarded as an individual living unit of which the genetic potential is constantly fettered by the chains of homeostatic controls. AP00024443 J not restrict their flying hours. Airlines using Concordes would probably prefer to avoid their Concorde crews becoming marginal radiation workers by apportioning their duties between the North Atlantic route and alternative southern routes. It is simply not feasible to reduce the radiation dose rates in the Concorde by the incor poration of shielding into the aircraft construction since, to be effective, such shielding would have to be equivalent in weight to a thickness of about half a metre of concrete over the fuselage 1 The lack ofreliable ground effects signalling the arrival of solar flare particles means that the experimental investigations of abnormal transient stratospheric radiation situations which must be made before the Concorde comes into service in 1971 will be far mote difficult than the investigation of the normal stratospheric radiation situation. Initially, when it was thought that giant solar flares were the main concern, study was given to the possibility of firing rockets whenever there was an increase in the count rate of a cosmic ray neutron monitor. But when it vas realized that the large, non-giant solar ,arr events were just as capable of producing biologically unacceptable stratospheric radi ation situations it was clear that the necessary procedure would be to fly stratospheric balloons when either large or giant solar flares were thought probable. The full instrument payload would probably consist of the equipment used on the earlier large balloon flights, plus tissue equivalent ionization chambers and an LET spectrometer--an instrument which measures the linear energy transfer, or specific ionization, spectrum: it is this spectrum which determines the biological significance of the radiation and so its quality factor. The experimental in-flight radiation warning meter will accompany these various instruments and experiments so that its response can be tested in strato spheric conditions. The expense of launch ing such a payload would be warranted only when a large or giant solar flare is highly probable. At other times much lighter and more modest balloon payloads, consisting of nuclear emulsions and a tissue equivalent ionization chamber dose rate recorder, would be launched. Obviously flic further development of the warning meter cannot wait on the infre quent and irregular large and giant solar flare events. It is, therefore, proposed to ;mulate the solar flare stratospheric radia- on by using some of the British moderate | and high energy accelerators, with a bath of liquid air to represent the air masses above the Concorde cruising altitude. The response of the warning meter to the resulting secondary radiation situation would be compared with the value of the biologically effective dose rate obtained from separate measurements ofthe neutron, ionization and tissue `star' dose rates, and of the specific ionization (or LET) spectrum. In large and giant solar flares the com position of the energetic particles can vary from protons alone to approximately equal numbers of protons and alpha particles. Although machines have been built to accelerate protons to the energies of solar flare parades and greater, no machines have yet been built to accelerate alpha particles to these energies. However, the secondary radiation caused by energetic alpha particles is generally similar to that caused by protons although there are differences due to the greater mass and charge, but lower average energy, of the alpha parades. In this context flic small balloon flights will provide valuable data on the various solar flare stratospheric radiation situations. Be tween them the data from both the ground level and the stratospheric experiments will indicate the final specification of the in-flight radiation warning meter. Large and giant solar flare events seem most likely before and after, rather than at the time of, a sunspot maximum. Before the next sunspot maximum in 1968 (which, inddentally, is the year that the French and British Concorde prototypes are scheduled to fly) it is expected that balloons launched probably in Britain and/or France will be used for stratospheric radiation measure ments in connection with the further development of the warning meter. During 1968-70 the final development of the warn ing meter will result from detailed com parison of the response of the prototype warning meter with the results of simul taneous, separate and independent radiation measurements made on board all prototype and pre-production Concordes in the course of routine stratospheric test flights. It is important that the warning meter be developed to give an accurate indication of the biologically significant radiation dose rate for the variety of possible solar flare conditions so that, although giving prompt warning of hazardous radiation situations, it does not give false alarms. Any un scheduled reduction in height by the Concorde will cause an increase in the consumption of fuel which, in turn, in creases operating costs and reduces fuel reserves. As a bonus, a warning meter which combines versatility with accuracy will also, be suitable for SSTs at greater altitudes. With the RADIATION warning and inte gration instrument aboard the Concorde, the crew and passengers will fly confident that their radiation safety is assured at all times and that the odd occasion when the radiation situation necessitates a reduction in altitude will be indicated promptly, accurately and without false alarms. The descent will be made gently and it is unlikely that the passengers would be aware of its occurrence. In my opinion something as mundane as bad weather at the airports is more likely to upset the Concorde flying schedules than solar flare stratospheric radiation. FURTHER READING A DIGEST OF DATA ON EXTRA-TERRESTRIAL RADIATION FOR THE EVALUATION OF HAZARDS LIKELY AT HIGH ALTITUDES by D. J. Bland, (in Royal Aircraft Establishment T. N. Mech. Eng., 359, Avgust, 1962) Solar flares by H. J. Smith and E. V. P. Smith (Macmillan, New York, 1963) Some radiation problems in the super sonic environment by D. L. Dye and W. R. Sheldon (in Society of Automotive Engineers, Paper 633C, April, 1363) Research in geophysics, volume 1 (sun, upper atmosphere and space) edited by H. Odishaw (MIT, Cambridge, Massachusetts, 1964) Solar flares by A. B. Severny (in Annual Review of Astronomy and Astrophysics, 2, 363-397,1964) Cosmic ray physics by A. E. Sandstrom (.North-Holland, Amsterdam, 1965) Radiation dosimetry for protection pur poses NEAR HIGH ENERGY PARTICLE ACCELERA TORS by J. Baarli and A. H. Sullivan of CERN, Geneva (in Health Physics, 11, 5, 353-361, May, 1965) Radiation aspects op the supersonic trans port (a report of the ICRP task group on the biological effects of high energy radiations) by A. C. Upton (Chairman, Committee 1) et. al. (in Health Physics, 12,2, 209, February, 1966) ACKNOWLEDGEMENT UKAEA, AWRE Aldemmton (pages 32 and 33) AP00024444 Whatever else carcinogenesis may involve, it certainly involves an effective breaking out from these chains. Observation indicates that, even under normal circumstances, the two daughter cells resulting from the division of a tissue cell are not completely identical. Their inequality may be trivial, noa-heritable and correctable during subsequent cell divisions. But if inequitable cell division proceeds on a large scale, and if sequential abnormal divisions occur because the mechanism for getting rid of defective cells cannot operate quickly enough, then a mixed popula tion ofcells will arise. Under the pressure ofthe limited availability of nutrients, and the operation of homeostatic mechanisms, a process of natural selection will then operate: the most vigorous cells--those least susceptible to homeostatic suppression--and their progeny will become relatively more numerous than the less vigorous members of the population. Once a cell which carries a heritable defect has gained a measure of autonomy, natural selection operating among its progeny will tend to lead to the dominance of progressively mote vigorous and autonomous cells. In keeping with this view is the fact that it is a character istic of cancer that it is progressive--in the sense that cancerous tissue tends to become more malignant with time. Furthermore it is usual for a time interval, which may be very long, to separate exposure to carcinogenic stimulus and the appearance of cancer. It now seems likely that this interval is occupied by the process of natural selection and the slow emergence of cells sufficiently autono mous to be regarded as cancerous. This is the background against which carcinogenesis should be viewed. In theory, at least, cancer may be induced either directly by an effect on the genetic material of cells or indirectly by interference with extracellular homeostatic control mechanisms. This distinction is far more meaningful than one based on the chemical, physical or viral nature of the agents concerned and it may in feet throw light on the mechanisms by which these agents operate. In 1775 PERCIVALL POTT recorded an association between the occupational exposure of chimney sweepers to soot and cancer of the scrotum: a majority of young men in their teens or twenties with this form of cancer had a history of having been employed as `climbing boys'. One hundred and forty years later two Japanese workers reported the induction of skin cancer in rabbits by the repeated application of coal tar. In 1930, a' team under the late Sir Ernest Kennaway detected and synthesized the first chemical carcinogens of the type present in coal tar. These were aromatic polycyclic hydrocarbons made up of four or five benzene rings. In time many hundreds of such substances were synthesized. Some proved active in the induction of cancer in animals, others-- often closely related chemically to active compounds--proved quite inactive. Theoretical chemists developed hypotheses associating structure and activity. Calculation ofthe electron charge densities ofeach ofthe various active molecules showed that all possessed a region of high density, the `K-region', in their double bond system. In these areas the probability of bond formation occurring with other molecules through charge transfer is greatest. However, not all `K-region' positive compounds proved to be active as carcinogens: high electron density in a second area of the molecule, the `L-region', rendered such com pounds inactive. By the use of this theory, and subsequent modifications of it, it was possible to predict carcinogenic activity within the limited class of the polycydic hydrocarbons. In the meantime, however, a wide variety of quite different chemical agents were found to be carcinogenic. Their extreme diversity led to the abandonment ofthe hope that the theoretical chemists would quickly solve the cancer problem. Another landmark in the history of the subject was the observation of Alexander Haddow in the late 1930s that there is an association between carcino genic activity and ability to inhibit growth, including tumour growth. Hot only this, but it seemed that carcinogenic compounds often had a selective toxic effect on the process of cell division, and were able to produce mutations in the cells. During the past 20 years this observation has led to the discovery of new rdnogens such as the aminostilbenes and to substances such as the biological xylating agents which both induce cancer and inhibit cell multiplication. The ssential property of alkylating agents is their ability to introduce alkyl CA N C E SCROTI. WHCJAMAZINI hu minis i book 3 RE d< motbis uUfinmi the Celle CXw of PgtAa* it a wtSUhotom A*faper, and f7 body b efqwinsi with the di&rdcr* to which ptialan* plantain, glisten, ud the when to white lud, vc liable j bet there i* ft dlf* \ ttfe t i Cife *a pceelitr to ntais At of pcoptd which hat sot, it keft to top knowledge, beespobliekip aotieedj lamihcchito* iipCincpn' caactr. It Is dlfaft which always malice Its Srft stuck flu, aod Ju lift tpptuucc in the inferior pm of the fcrecum t where it produces a fuperteuJ. painful, rigged. llMaoklnj ftn, with Utdsndrifing edges. The trade call It the het>wirt, t never law It under tha age of ptberry, which it, I fuppefc, osc rctibnf>wfcf It Is gtoenKy taken, both by firieat and forgoes, for veMieal, ud Wing Uesud with tpemriaii If tmby ton, sal otsah raafparatci t to as great length sf tine, k perradsa the Ail, tUrtM, end meohruMcf the (m- PERCiVALL POTT first described the association between exposure of chimney eweept to toot and cancer of the scrotum in 1775. His book, first page of text illustrated, was the first in a long and continuing line of works on occupational and Industrial medicine SIMILAR COMPOUNDS can have dissimilar effects. Thus 1, 2, 5, 6-dibenzanthracene, left, is a potent carcinogen whereas the related substance 1, 2, 3, 4-dibenzanthracene, right, is carcinogenicaliy Inactive HO HiC \c-c H2C' OCH, ^ CH I C:0 PENiCILLIC ACID is the basis of a variety of penicillins. Causes cancer at injection site in rats NH-NO / CO \ NH, NITROSOUREA causes tumours at various sites including the brain in soma laboratory animals ' 0-0 N(CH3)j 4-DIMETHYLAMINOAZOBENZINE (butter yellow) was formerly used to give colour to margarine. However, it has since been found to cause liver cancer in rats ,ch2ch2ci CHjO CH3-S02 0(CH2)40S0j-CH3 ^CHjCHjCI MYLERAN (BUSULFAN) NITROGEN MUSTARD is of value in the treat- ha* been found of value In mentofmyeioidleuksemia the treatment of several --a blood cell cancer different tvoes of cancer AP00024445 A--Adenine C--Cytosine G--Guanine TThymine WO STRANDS OF DNA may become nked together by a bifunctional alkylating gent such as myleran, the molecule of /hich becomes attached at each end to a ueleotide base (A. C. G or T). The effect f such a linkage may lead either to a eduction in the capacity of affected cells o reproduce or to the induction of cancer radicals, such as --CH, or --C4He, into molecular components of cells. The effect of such introduction leads to a dramatic reduction in the capacity of cells to reproduce and, in some cases, to the induction of cancer. The former effect is immediate and the latter delayed. Alkylating agents capable of introducing a single alkyl radical are termed monofunctionaU those able to introduce two or more are termed bifunctional or polyfunctional respectively. When bifunctional or polyfunctional compounds react with two or more molecules the end result may be that the molecules become linked together by the alkylating agent. It is now believed that both the carcinogenic and tumour inhibiting effects are due to reactions of this type with the genetic material of the cell, deoxy ribonucleic arid (DNA), In recent years an ingenious theory has been developed that the most significant effect of the polyfunctional alkylating agents is to cross link the two strands of DNA. If this were true, it would be possible to explain, in terms of a single chemical mechanism, all three types of biological activity of these agents: carcinogenicity, mutagenicity and growth inhibiting activity. As a result of similar thinking a new suggestion was made with regard to the possible mode of action of carcinogenic polycyclic hydrocarbons, and their distinction from inactive analogues. It was pointed out that some of the carcinogenically active compounds, such as 1,2:6,7-dibenzopyrene, were similar in chemical structure to part of the DNA molecule. This suggests that this type of compound might also cross-link the two strands of DNA. These theories arc still actively stimulating new work but, clearly, they do not apply to all forms of carcinogenesis. As more compounds have been examined, the association between carcinogenicity, growth inhibition and mutagenicity has become less dear cut. Monofunctional alkylating agents, capable of reacting with DNA but not of cross-linking its two strands, have been shown to be carcinogenic, though less so than related polyfunctional compounds. There are grounds for believing that most potent carcinogens affect the DNA or RNA of cells more or less directly and specifically. The most potent cancer inducing agents of all, namely the tumour viruses, consist simply of DNA or RNA. Carcinogens may be weak either because their reaction with nudeic acids is non-specific, or because they affect nucleic adds indirectly rather than directly. A third possibility is that their primary effect is not on cells at all, but on their environment by interference with the homeostatic control systems. Two facts concerning chemical carcinogenesis deserve special mention. First, several metals are known to be capable of inducing cancer, either in men exposed to them in industry, or in laboratory animals under experiment. A human cancer hazard is now recognized in relation to arsenic, beryllium, chromium and nickel. In the laboratory the latter three, together with cadmium, cobalt and iron, have been shown to induce cancer. Curiously, no one has yet succeeded in inducing cancer by exposing laboratory animals to arsenic. Secondly, in recent years a number of naturally occurring carcinogens have been recognized. These indude cycasin, an alkaloid present in the cycads--types of palm the leaves, stems and roots of which provide starch, for example sago, and form a staple item of diet in some parts of the world; the senedo alkaloids present in some species of ragwort; safrole, a component of several essential oils induding sassafras oil, which is extensively used in North America to flavour `root beer'; and aflatoxin, a lactone formed by some species of the mould Aspergillus flavus when it grows in ground nuts and cereals stored under the hot, humid conditions of the tropics. - SOME CARCINOGENS, such as 1, 2: 6. 7-dibenopyrene, shown coloured, are similar in chemical tructure to parts of the DNA molecule, for example adenine and thymine, black. This suggests that this type of compound may be able to cross link the DNA It has long been puzzling that hormones, produced naturally by the body, act as carcinogens under certain circumstances. For instance, cancers of many kinds may be induced in animals by giving them excessive amounts of oestrogen, one of the hormones normally produced by the ovary. Many attempts have been made to show structural resemblances between particular hormones and carcinogens ofthe polycyclic hydrocarbon type. Indeed, some of the latter have been shown to exhibit weak oestrogenic activity. Neverthdess, the links between carcinogenesis by hormones and by other agents are very tenuous. It is much more likely that the primary action of hormones is not on cells but on one of the homeostatic mechanisms which control their growth and self-expression. The continued presence of excessive amounts of a hormone may, by blocking a normal suppressing mechanism, permit the proliferation of a tissue. This may favour the emergence of autonomous cells in one of two ways. First, the con- AP00024446 ditions existing in a proliferating tissue may be more favourable to the survival of cells which deviate from normal. Secondly, if because of previous exposure to a carcinogen the cancerous process has been started and there is already a mixed population of cells, the process of selection of more vigorous and autonomous cell lines will operate more quickly under conditions of active proliferation. It has been shown in the laboratory that cancer may be induced by exposing an animal to two agents, A and B, in. that order, whilst exposure to A alone, B alone, or B followed by A does not lead to cancer. Despite several attempts to do so, no one has yet found an agent which has only A-type activity (tumour initiating activity) or only B-type activity (tumour promoting activity). According to the theory presented here, tumour initiators act primarily on cells and tumour promoters primarily on homeostatic control mechanisms. Doses of carcinogens too low to complete the process of carcinogenesis may, nevertheless, be sufficient to initiate it. It is not surprising that pure promoters have not been demon strated since it is not possible to escape background exposure to initiating agents in the form of low doses of environmental chemical carcinogens and ionizing radiation. It is reasonable to suspect that homeostatic control mechanisms are weakened as a result of the ageing process. The tumour promoters may do no more than bring forward in time a process which would eventually occur spontaneously as a result of ageing. There is some experimental evidence that this is the case. The term `co-cardnogen' has been applied to an agent which enhances the activity of a carcinogen. There are many possible mechanisms by which such enhancement may be brought about; for example, the co-cardnogen may aid the absorption ofcarcinogen or block its breakdown to non-cardnogenic metabolites. All tumour promoters are co-carcinogens, but not all co-carcinogens are tumour promoters. H2C------- CHZ 0:C------- 0 CHj-CHj-OSOj-CHj 0-PROPIOLACTONE is an alkylating agent with virus killing properties. It In duces cancer on injection Into rats or when it is applied to the skin of mice ALKYLATING AGENT (ethyl msthanesulphonate) has been shown to induce formation of tum ours of the lung when injected into infant mice / ch3no-n-ch2-o-ch chch2oh \ choh \------- c HH 0H QH CYCASIN--an alkaloid found in various speciss of cycad palm which are an important source of starch in many tropical countries--can induce cancers in rats ch2-ch-ch2 Until the publications of F. C. Turner in 1941 and of B. S. and E. T. Oppenheimer and A. P. Stout from 1948 onwards the term physical carcino genesis was applied, for the most part only, to the induction of cancer by agents such as ultraviolet light, ionizing radiation, physical trauma and chronic irritation of various kinds. It was fairly easy to interpret the effects of both ultraviolet and ionizing radiation in chemical terms, and the interpretation was supported by the introduction of the idea of `radiomimetic agents' in connection with chemical agents whose effects, in general, resembled those of ionizing radiation. The fart that physical trauma seems to be able to initiate or precipitate cancer has never been very palatable to those who seek a neat chemical explanation of all forms ofcarcinogenesis. To them the discovery by Peyton Rous that wounding the ears of rabbits enhanced carcinogenesis by coal tar, though it did not induce cancer on its own, was a windfall. Physical trauma could then be dismissed as a co-carrinogenic factor and need not be explained directly in chemical terms. On the other hand, the demonstration by first Tumer and later the Oppenheimers that cancer may be induced by the implantation of chemically inert materials into the tissues made it necessary once more to consider physical carcinogenesis seriously in its own right. At the Chester Beatty Research Institute we have induced cancer in the bladders of mice by the implantation of simple glass beads. Some have argued that nothing is so chemically inert that chemical carcinogenesis is ruled out. Nevertheless, the induction ofcancer by the implantation ofobjects of various shapes, and the simultaneous failure to induce cancer by implanting the powdered chemicals from which the objects are made, have compelled the acceptance of the view that the implanted objects induce cancer because of their physical and not their chemical characteristics. In our researches at the Institute we have found that a high proportion of a group of rats developed cancer at the site of implantation in the subcutaneous tissues of pieces of polyvinyl sponge (as used in plastic surgery) measuring 20x20x5 mm, whereas only 1 out of 24 rats did so in response to implants measuring 33x33x2 mm. The amount of sponge was the same in both cases but the shape was different. What possible relation can this type of cancer induction have to that idue to the various chemical agents discussed above? In the past the distinction between chemical and physical carcinogenesis has , tended to be artificial. A better distinction is between agents which primarily 1 xrtect cells tad those which primarily affect homeostatic control mechanisms. 4-ALLYL--1, 2-METHYLENEDIOXYBENZENE (safrole) it a natural ingredient of sassafras tea and has been used to flavour 'root beer" in North America. It has been shown to cause liver tumours when fed to rats AFLATOXIN B is one of a group of toxins produced by strains of the mould AspargiUus flavus which grows in ground nuts and cereals stored under hot, humid conditions. Besides acting aa a poison--it killed more than 100,000 turkeys and game birds In 1960--it can induce cancers of the liver and kidney in rodents CANCER, arrowed, in the bladder of a mouse following the implantation of e glass bead. 40 weeks earlier.. AP00024447 i M PLANT ED OBJECTS can induce cancer because if their physical characteristics. The cancer, irrowed, arose in the region of a piece of plastic .ponge implanted In the subcutaneous tissue of the at several months earlier. A similar piece of sponge if the same weight but of different shape vas found to induce cancer far less frequently ASBESTOS induces cancer possibly because of its jhysical rather than chemical properties. This lodule of cancerous tissue from the abdominal tavity of a rat arose several months after the last of t series of four subcutaneous injections of asbestos FURTHER READING Chemical carcinogenesis by David B. Clayson J. and A. Churchill, London, 1962) Mechanisms of carcinogenesis: chemical, physical and viral (in British Medical Bulletin, 20, No. 2,1964) Chemical carcinogenesis and cancers by W. C. Hueper and W. D. Conway (Charles C. Thomas, Springfield, Illinois, 196i) JKNOWLEDGEMENT E. Boyiand (page 40, bottom) It is now suggested that, as in the case of the induction of cancer by the admini stration or withdrawal of hormones, carcinogenesis by inert objects is brought about by interference with an extracellular homeostatic mechanism. It is possible, for instance, that the presence of the object interrupts tissue communication and negative feed-back control mechanisms over an area. From studies on cells grown in tissue culture it is apparent that an environment in which cells may divide and move freely leads sooner or later to the appearance ofcancer-like cells. It is possible that this is also the explanation of some examples of what has up to now been regarded as chemical carcinogenesis. Certain iron-carbohydrate complexes, such as iron-dextran, iron-dextrin and saccharated iron oxide, induce local cancer when injected subcutaneously or intramuscularly into animals. The carbohydrate moieties of the complexes do not themselves induce cancer, nor does iron in other forms do so. The complexing of iron with the carbohydrates leads to the formation ofvery large molecules. These are ingested by scavenging white blood cells which may remain for long periods at the site ofinjection, particularly in animals which are not generally deficient ofiron. The question is do these large masses of iron complex loaded cells act in the same way as implanted inert objects and do they interfere with trans-tissue communication in the same way as such objects ? Certainly the whole sequence of events which precedes the development of cancer is very similar in the two instances. At first SIGHT it may seem that this is a merely academic problem but, in fact, it has important practical implications. In most countries, governments now require that constituents of cosmetics and pharmaceutical preparations, sub stances added to food, and substances such as pesticides or herbicides which may contaminate food should be tested for carcinogenicity in laboratory animals. A positive result in any carcinogenicity test renders a substance, if not completely unacceptable, then at least suspected of being dangerous for man. Should a potentially useful food additive or drug, intended for oral administration, be banned on the grounds that it induces cancer at the site of its subcutaneous or intramuscular injection in animals ? Might not the latter be an implantationisduced effect rather than an example of true chemical carcinogenesis ? As more AND more different types of agent--chemical, physical and viral--have been shown to be capable of inducing cancer, the avoidance of exposure to known carcinogens which was once easy became at first difficult and then impossible. Substances capable of inducing cancer in laboratory animals are present in vehicle exhausts, tobacco smoke and the general atmosphere---even where there is no marked pollution. They have also been isolated from certain cooked foods and even from so called `health' foods. Dangerous materials are used in many branches of industry. They are present in freshly mined minerals, such as asbestos, in the materials used for household or garden maintenance, such as creosote, and in a variety ofpharmaceutical preparations. No man, be he primitive or civilized, can completely avoid exposing himself to potentially carcinogenic factors. Why then bother to try ? Why stop smoking ? Why take precautions ? The answer is that it is a matter ofdose and probability: the larger the quantity of carcinogen taken into the body the greater the risk of cancer. In the consideration of an environmental factor the question "Is this carcino genic ?" is the wrong one--since it only permits oftwo answers, "Yes" or "No". The better question is, "What is the risk that exposure to a measured dose ofthis particular agent by a specified route of exposure will lead to cancer during a stated interval of time iter exposure ?" The question framed in this way takes into account the vital part played by the exposed tissue in the carcinogenesis process. It also stresses the distinction between major and minor carcinogenic hazards. Such a distinction can be made tentatively on incomplete evidence and given as an expression of informed opinion, whereas the inflexibility of the `all' or `none' distinction between `carcinogenic* and `non-carcinogenic' stultifies dear thinking and paralyses legislative action. For those who insist on an `all' or `none' distinction perhaps the most reasonable one is that there are two types of substances, materials or agents, those which have been shown to induce cancer and those which have not yet been shown to do so. In other words there may be no agents in the `none' category. AP00024448 THE CANCER PROBLEM Almost all cancers appear to be caused by exposure to factors in the environment. The most promising approach to the control of the disease is to identify those factors and eliminate them by John Cairns uring the past 150 years the West thought of as an unpredictable disease D ern world has virtually elimi nated Infectious diseases as a sig that strikes indiscriminately at rich and poor, fat and thin, old and middle-aged, nificant cause of death. A child bomaisnif it usually owed nothing to external the U.S. today can look forward to a life causes. If that were true, our only hope untroubled by fear of diseases such as of overcoming cancer would be to im scarlet fever, diphtheria, tuberculosis, prove the treatment of the disease. One typhoid fever and dysentery, which were object of this article is to show that most major causes of death three or four gen of the common kinds of cancer seem to erations ago. Life expectancy has been be caused in large part by environmen increasing steadily since the middle of tal factors; because we can act to alter the 19th century. The longer average life the environment, those cancers are po span is a result mainly of improvements tentially avoidable. in public health; the more spectacular fruits of scientific research, such as the introduction of vaccines and antibiotics, Incidence merely completed the process. Groups of abnormally proliferating Death has now been confined mostly cells can arise in any part of the body. to old age and can therefore be attrib Those that cannot invade the surround uted to diseases that are either peculiar ing tissues and so remain striedy local to old age or lethal only in old people. growths are called benign tumors. Those Although innumerable changes in the that spread from their site of origin and body that accompany advancing age can therefore reach the bloodstream and could be classified as diseases, two par the lymphatic system are called malig ticular conditions commonly arising in nant tumors, or cancers. old age are often a direct cause of death: The cancers are divided into three arterial disease (atheroma and arterio broad groups. The carcinomas arise in sclerosis) and cancer. Arterial disease is the epithelia, the sheets of cells covering lethal when it affects the arteries supply the surface of the body and lining the ing the heart or the brain; it now ac various glands. The much rarer sarcomas counts for about 50 percent of all deaths arise in supporting structures such as in the U.S. Cancers are lethal when they fibrous tissue and blood vessels. The leu spread from their site of origin; they now kemias and lymphomas arise in - the account for almost 20 peroent of all U.S. blood-forming cells of the bone marrow deaths. and lymph nodes. These three words-- People have very different attitudes carcinoma, sarcoma and leukemia--are toward these two diseases. It is not just so entrenched in everyday usage that that death from arterial disease is often they must be mentioned, but I do not rapid, whereas death from cancer can be mean to imply by their use that there are painfully drawn out. For some reason three basically different forms of carcino heart attacks and strokes tend to be genesis or that the three kinds of cancer thought of as natural hazards of age, and have different prospects for prevention either a normal end to a satisfactorily and cine. That kind of information can long life or, when they occur in middle- be obtained only through a finer system aged men, the wages of overeating and of classification. lade of exercise. In contrast, cancer is Cancers are classified mainly by the organ in which they originate and by the kind of cell involved. When they are considered in this way, there are 100 or so distinct varieties of the disease. Such an elaborate classification would be of no general interest were it not that the different varieties plainly have different causes, since the incidence of each one changes independently when the en vironment is altered. Most of the 100 varieties are rare, and so we can account for most cancer mortality by considering a fairly short list of diseases. Roughly half of all cancer deaths are caused by cancers of three organs: the lung, the large intestine and the breast [see illustration on page 66], There can therefore be no major inroad on overall cancer mortality until some means are found for curing or preventing these three kinds of cancer. Each of them can be considered a discrete entity because the frequency of each varies indepen dently when factors in the environment are changed. It could reasonably be argued that we are not interested in total numbers of deaths as much as in loss of life span. The death of a 90-year-old man from cancer of the prostate is less of a tragedy than the death of a young man from leukemia. In determining our priorities we should therefore take into considera tion the age distribution of the victims of each cause of death. There are vari ous ways of doing this. For example, it is possible to calculate how much each ma jor cause of death or each kind of cancer diminishes the average life expectancy or, in particular, how much each reduces our working life up to the age of 65. The main effect of such a weighting proce dure is to increase the relative impor tance of accidents among the general causes of death, and of the leukemias and lymphomas among the cancers. Can- AP00024449